Vous êtes sur la page 1sur 559

Debmalya Barh


Omics Approaches
in Breast Cancer

Towards Next-Generation
Diagnosis, Prognosis, and

Omics Approaches in Breast Cancer
Debmalya Barh

Omics Approaches in
Breast Cancer
Towards Next-Generation Diagnosis,
Prognosis and Therapy
Debmalya Barh, MSc, MTech, MPhil, PhD, PGDM
Institute of Integrative Omics
West Bengal

ISBN 978-81-322-0842-6 ISBN 978-81-322-0843-3 (eBook)

DOI 10.1007/978-81-322-0843-3
Springer New Delhi Heidelberg New York Dordrecht London

Library of Congress Control Number: 2014951826

© Springer India 2014

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or
part of the material is concerned, specifically the rights of translation, reprinting, reuse of
illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way,
and transmission or information storage and retrieval, electronic adaptation, computer software,
or by similar or dissimilar methodology now known or hereafter developed. Exempted from this
legal reservation are brief excerpts in connection with reviews or scholarly analysis or material
supplied specifically for the purpose of being entered and executed on a computer system, for
exclusive use by the purchaser of the work. Duplication of this publication or parts thereof is
permitted only under the provisions of the Copyright Law of the Publisher’s location, in its
current version, and permission for use must always be obtained from Springer. Permissions for
use may be obtained through RightsLink at the Copyright Clearance Center. Violations are liable
to prosecution under the respective Copyright Law.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are
exempt from the relevant protective laws and regulations and therefore free for general use.
While the advice and information in this book are believed to be true and accurate at the date of
publication, neither the authors nor the editors nor the publisher can accept any legal responsibility
for any errors or omissions that may be made. The publisher makes no warranty, express or
implied, with respect to the material contained herein.

Printed on acid-free paper

Springer is part of Springer Science+Business Media (www.springer.com)

My loving Gul: Shaurya Shree

The modern trends of disease diagnosis, prognosis, and therapy are changing
from conventional to rationally designed personalized approaches. “Omics”
based next-generation technologies have boosted this revolution. The fact is
true in the case of breast cancer also, which is one of the most common cause
of morbidity and mortality in women worldwide.
This book Omics Approaches in Breast Cancer: Towards Next-Generation
Diagnosis, Prognosis and Therapy, edited by Debmalya Barh, is one of the
endeavour efforts that presents omics based research outcomes and their real-
time applications in breast cancer diagnosis, prognosis, and therapy. The
book is designed in such a way that blends the basic topics such as biology,
conventional diagnosis and treatment approaches along with the latest
advances in tackling breast cancer using next-generation omics based tech-
nology outcomes. The book has highlighted several latest technology and
fields in the subject area such as metabolomics, nutrigenomics, RNAomics,
stem cell and cellomics, pharmacogenomics, animal and in silico models,
exhaled volatile biomarkers, minimal/non-invasive molecular biomarkers,
targeted therapy, personalized medicine, and gene therapy of breast cancer
among others. Male breast cancer, although it is very rare, is also included in
the book. There are 27 chapters in this book written by 65 breast cancer spe-
cialists across the world. Therefore, the book provides state-of-the-art con-
tents, real-life personal experiences, and future prospects of breast cancer
For the benefit of readers, topics are grouped under two sections where the
first section describes various omics technologies and their outcomes in
breast cancer research and the second section is having chapters that contain
the applications of the latest omics technologies in next-generation breast
cancer diagnosis, prognosis, and therapy. In the first part, the book contains
chapters on: Overview of omics technologies applied in breast cancer
research; Omics of hereditary breast cancer; Oncogenes and tumor suppres-
sor genes as biomarkers in breast cancer; Breast cancer genomics;
Epigenomics approaches in breast cancer; Breast cancer neutrigenomics;
Implications of long non-coding RNAs in breast cancer pathogenesis, diag-
nosis and therapy; micro-RNA as clinical biomarkers for diagnosis and treat-
ment of breast cancer; Breast cancer proteomics; Metabolomics in breast
cancer; Lipidomics in breast cancer; Breast cancer stem cells and cellomics;
Omics of male breast cancer; Omics approaches in chemoresistant and metastatic

viii Foreword

breast cancer; Animal and in silico breast cancer disease models; and Systems
biology and integrative omics approaches in breast cancer. In the application
part, the book includes chapters on: Gynaecological aspects and consider-
ations for women suffering from breast cancer; Currently available imaging
technologies and their applications in early diagnosis and prognosis of breast
cancer; Non-invasive or minimal invasive molecular biomarkers for risk
assessment, screening, detection, diagnosis, and prognosis of breast cancer;
Circulating tumor cells for diagnosis and prognosis of breast cancer;
Molecular diagnosis of metastasizing breast cancer using liquid biopsy;
Volatile biomarkers in breast cancer; Classical and targeted therapy in breast
cancer; Personalized medicine in breast cancer; Gene therapy in breast can-
cer; and Clinical trials endpoints in breast cancer.
It is a great effort by Debmalya Barh to cover almost all aspects of breast
cancer in this omics era in this book. The broad-coverage, latest information,
and rich contents of the book in the field of breast cancer is the uniqueness of
the book that will definitely help in next-generation diagnosis, prognosis, and
therapy of breast cancer. I highly recommend the book for readers who are
working in the field of breast cancer research, diagnosis, and treatment.

Sanjeev Misra, MBBS, MS, MCh, FRCS (Eng.),

Director and CEO, All India Institute of Medical Sciences (AIIMS)
Jodhpur, India

Breast cancer is the leading cause of cancer specific deaths in women world-
wide. Screening and early diagnosis provides better patient management,
improves treatment efficacy, and reduce mortality. With the advent of next-
generation “omics” technologies; several early markers, novel targets, and
personalized targeted therapeutics are now available that are either under
development or in use. However, the new cases are increasing rapidly and
control of the incidences and the mortality rate is not coming down signifi-
cantly due to several factors associated with the disease. Therefore, identifi-
cation of the disease biology at a deeper level and search for the gold standard
molecular markers for screening, early diagnosis, prognosis, and novel thera-
peutics irrespective to the type of breast cancer will continue till we identify
them. In this “omics” era, a large amount of data have been generated and
analyzed in various aspects of breast cancer to achieve these ultimate goals.
However, the outcomes of these data in an organized form is not readily avail-
able so far, so that, the summery and the advancements in these fields can be
glanced in a single resource.
This book entitled Omics Approaches in Breast Cancer: Towards Next-
Generation Diagnosis, Prognosis and Therapy is introduced to fill these gaps
by providing all basic and latest developments in various “omics”-based
breast cancer research outcomes and applications in a single volume. The
book also contains basic topics such as types of breast cancers, conventional
treatment strategies, currently used diagnostic tools, etc. so that readers can
get the entire spectrum of breast cancer.
The book is a successful effort of more than 65 experts (scientists, clini-
cians, pharmacists, etc.) from nearly 20 countries who are either working on
various “omics” aspects of breast cancer biology or developing breast cancer
biomarkers and therapeutics or treating breast cancer for last several decades.
Therefore, the book reflects richest and up-to-date contents, personal and
real-life experiences, and most importantly, provides the future directions of
breast cancer research.
Omics Approaches in Breast Cancer: Towards Next-Generation Diagnosis,
Prognosis and Therapy contains 27 chapters covering most of the aspects of
female breast cancer “omics” and is divided into 2 parts. While Part I con-
tains applied technologies and outcomes of various “omics” approaches in
breast cancer, Part II provides real-life applications of “omics”-based research
outcomes in breast cancer diagnosis, prognosis and therapy along with basic
clinical and therapeutical aspects of the disease. A special chapter is also

x Preface

included on male breast cancer to describe up-to-date “omics”-based out-

comes on this rare type of breast cancer.
Part I starts with an introductory chapter (Chap. 1) by Dr. Cusati and col-
leagues to provide an overview of “Omics technologies applied in breast can-
cer research”. In Chap. 2, the “Omics of hereditary breast cancer” has been
described by Dr. Catherine’s group. Dr. Uctepe et al. provide the insights on
how the oncogenes and tumor suppressor genes can act as biomarkers in
breast cancer in the Chap. 3. Chapter 4 by Dr. Kumar and Dr. Mandal gives a
detail account on various aspects of “Breast cancer genomics”. In the next
chapter (Chap. 5), Dr. Yaykasli and colleagues have provided the epigenom-
ics approaches in breast cancer. How the nutrients and genes interplay in
breast cancer have been covered by Dr. Dwivedi and colleagues in Chap. 6.
In Chap. 7, the implications of long non-coding RNAs in breast cancer patho-
genesis, diagnosis and therapy have been discussed by Dr. Juracek and col-
leagues. The subsequent chapter (Chap. 8) by Dr. Shafi’s team gives a detail
account on how microRNA can be utilized as clinical biomarkers for diagno-
sis and treatment strategies in breast cancer. In Chap. 9, Prof. Minafra has
demonstrated a detail account on “Breast cancer proteomics”. The next chap-
ter (Chap. 10) by Dr. Calomarde et al. covers the “Metabolomics in breast
cancer”. The “Lipidomics in breast cancer” written by Dr. Kamili and Dr.
Byrne is included in Chap. 11. “Breast cancer stem cells and cellomics” by
Dr. Demir and colleagues in Chap. 12 has provided the emerging field of
potential stem cell therapeutic aspects in breast cancer. In this book, breast
cancer generally means female breast cancer. However, there are cases where
males are also diagnosed with breast cancer. The book has included a chapter
(Chap. 13) on “Omics of male breast cancer” by Dr. Nur Unal et al. to give
up-to-date “omics”-based strategies, outcomes, and other aspects of this very
rare type of breast cancer. In Chap. 14, omics approaches in chemoresistant
and metastatic breast cancer have been discussed by Dr. Aguilera. In the next
two chapters (Chaps. 15 and 16), animal and in silico breast cancer disease
models and their various aspects have been described by Dr. El-Abd and Dr.
Munshi’s groups, respectively. The last chapter (Chap. 17, by Dr. Hernández-
Lemus) under Part-I provides systems biology and integrative omics
approaches in breast cancer to give the latest developments in this area.
Part II of this book provides information on how the “omics”-based
research outcomes are used in real-life diagnosis, prognosis and therapy of
breast cancer. The section starts with a chapter (Chap. 18) by Dr. Robinson
and Dr. Ali that provides the basic clinical or gynaecological aspects and
considerations for women suffering from breast cancer. The next chapter
(Chap. 19) developed by Dr. Mar Gil and colleagues gives the detail account
of currently available imaging technologies and their applications in early
diagnosis and prognosis of breast cancer. In Chap. 20, Dr. Verma and I have
demonstrated various molecular biomarkers that are either under develop-
ment or in practice for risk assessment, screening, detection, diagnosis, and
prognosis of breast cancer. Next three chapters deal with very important
aspects in breast cancer diagnosis and prognosis through minimal- or non-
invasive strategy. Dr. Van Pham in Chap. 21 has provided how the circulating
tumor cells can be used for diagnosis and prognosis of breast cancer. In the
Preface xi

same direction, Dr. Dwivedi and colleagues in Chap. 22 have described the
molecular diagnosis of metastasizing breast cancer using liquid biopsy, and
in Chap. 23, Dr. Barash and Dr. Haick have demonstrated the emerging
potential of exhaled volatile biomarkers in breast cancer. Chapters 24, 25,
and 26 cover therapeutic aspects of breast cancer. Dr. Ch Yiannakopoulou in
Chap. 24 has given a detail account of classical therapy and drug targets along
with targeted therapy in breast cancer. In the next chapter (Chap. 25),
Dr. Ch Yiannakopoulou and me have discussed the pharmacogenomics or
personalized medicine and their therapeutic implications in breast cancer.
The current status and future prospects of gene therapy in breast cancer has
been discussed in Chap. 26 by Dr. Büyükköroğlu and her colleagues. The last
chapter (Chap. 27) is on clinical trial, and in this chapter Dr. George and
Dr. Selvarajan have described the essentiality of optimum end points in breast
cancer clinical trials to select the right drug having the best efficacy and mini-
mal toxicity.
I believe that this book and its up-to-date contents and broad coverage will
be worthwhile to cutting-edge breast cancer research, diagnosis, and clinician
communities. I highly appreciate your comments and suggestions to improve
the next edition of the book.

Purba Medinipur, Debmalya Barh

West Bengal, India (Editor)
About the Editor

Debmalya Barh (MSc, MTech, MPhil, PhD, PGDM) is the founder of the
Institute of Integrative Omics and Applied Biotechnology (IIOAB), India, a
virtual global platform for multidisciplinary research and advocacy. He is a
biotechnologist and an active researcher in integrative omics-based
biomarkers, targeted drug discovery, and personalized medicine. He works
with over 400 well-regarded researchers from more than 40 countries and has
to his credit over 125 high-impact international publications, several books,
and book chapters. He is a globally branded editor for editing omics-related
research reference books and an editorial and review board member for a
number of highly reputed international journals. Due to his significant
contributions to the field and in promoting biomedical sciences using unique
research strategies, in the year 2010 he was recognized by Who’s Who in the
World and in 2014 he entered the Limca Book of Records, the Indian equivalent
of the Guinness Book of World Records.


2DE Two-dimensional electrophoresis

2D-PAGE Two-dimensional polyacrylamide gel electrophoresis
ABC ATP binding cassette transporter family
aCGH Array-comparative genome hybridization
ACOG American Congress of Obstetricians and Gynecologists
ACP American College of Pathology
ACR American College of Radiology
ACS American Cancer Society
ADP Adenosine diphosphate
AE Adverse event
AI Aromatase inhibitor
AIDS Acquired immunodeficiency syndrome
ALA Alphalinolenic acid
ALDH Aldehyde dehydrogenase activity
AMBP α-1-microglobulin/Inter-α-trypsin inhibitor light chain precursor
ANN Artificial neural network
A-T Ataxia-telangiectasia
ATAC Arimidex, Tamoxifen Alone or in Combination (trial or study)
ATM Ataxia-telangiectasia mutated
BBN Bayesian belief network
BC Breast cancer
BCC Breast cancer cells
BCI Breast cancer index
BCRAT Breast cancer risk assessment tool
BCSC Breast cancer stem cells
BIG Breast International Group (1–98 study/trial)
BIP Immunoglobulin heavy chain binding protein
BI-RAD Breast Imaging Reporting and Database
BLBC Basal-like breast cancer
BM Bone marrow
BMAC Breath methylated alkanes contour
BSE Breast self-exam
CAFs Cancer-associated fibroblasts
CAGE Cap analysis of gene expression
CAN Copy number aberrations
CAV1 Caveolin-1
CD Cluster of differentiation (molecules or markers)

xvi Abbreviations

cDNA Complementary DNA

CGH Comparative genomic hybridization
CMTC ClinicoMolecular Triad Classification
CNV Copy number variations
COMT Catechol-O-methyltransferase
CT Computerized tomography
CTC(s) Circulating tumor cell(s)
CTOP Cancer therapy outcome predictor
DCIS Ductal carcinoma in situ
DCR Disease control rate
DEP Dielectropheresis
DFS Disease-free survival
DIC Ductal infiltrating carcinoma
DIGE Difference in-gel electrophoresis
DNMTs DNA methyltransferases
DTCs Disseminated tumor cells
ECD Ectodomain
ECGC Epigallocatechin-3-gallate
ECM Extracellular matrix
EGF Epidermal growth factor
EGFR Epidermal growth factor receptor
EMEA European Agency for the Evaluation of Medical Products
EMT Epithelial/mesenchymal (or epithelial to mesenchymal) transition
EORTC European Organization for Research and Treatment of Cancer
EpCAM Epithelial cell adhesion molecule
ER Estrogen receptor(s)
ERRα Estrogen-related receptor α
ERα Estrogen receptor α
Erβ Estrogen receptor β
ESI Electron spray ionization
ESR1 Estrogen-receptor 1
FA Fanconi anemia
FASN Fatty acid synthase
FAST Fiber-optic array scanning technology
FC Flow cytometry
FDG 2-deoxy-2-[18F]fluoro-d-glucose
FFA Free fatty acids
FFPE Including formalin-fixed paraffin-embedded (samples)
FGFR Fibroblast growth factor receptor
FISH Fluorescence in situ hybridization
FK-228 Romidepsin
FM Fluorescence microscopy
FNA Fine-needle aspiration
FTDT Finite difference (modelings)
GASP-1 G-protein coupled receptor associated sorting protein 1
GC Gas chromatography
GES Gene expression signatures
GLA Gamma-linolenic acid
Abbreviations xvii

GNP Gold nanoparticles

GnRH Gonadotropin-releasing hormone
GR Glucocorticoid receptor
GRNs Gene regulatory networks
GS Glutamine synthetase
GSEA Gene Set Enrichment Analysis
GWAS Genome-wide association studies
HATs Histone acetyltransferases
HBOC Hereditary breast and ovarian cancer
HDAC Histone deacetylase
HDMs Histone demethylases
HER2 Human epidermal growth factor 2
HIC1 Hypermethylated in cancer 1
HIV Human immunodeficiency virus
HME Human mammary epithelial (cells)
HMTs Histone methyltransferases
hNMSC Human normal mammary stem cell (signature)
HNPCC Hereditary non-polyposis colon cancer
HPLC High-performance liquid chromatography
HPV Human papillomavirus
HR Homologous recombination
HR-MAS High-resolution magic angle spinning
HSV Herpes simplex virus
HT Hormonal therapy
HTOTs High-throughput omic technologies
IDC Infiltrating ductal carcinoma
IDFS Invasive disease-free survival
IEL Intraepithelial lesion
IGF Insulin-like growth factor
IHC Immunohistochemical/immunohistochemistry
ILC Infiltrating lobular carcinoma
ITC Isolated tumor cells
IVD In vitro diagnostic
IVF In vitro fertilization
LC-MS/MS High-performance liquid chromatography and tandem mass
LD Linkage disequilibrium
LN Lymph node
lncRNA Long non-coding RNA
LSINCT5 Long stress-induced non-coding transcript 5
LV(s) Latent variable(s)
m/z Mass to charge ratio
MALDI Matrix-assisted laser desorption/ionization
MAPs Microtubule-associated protein
MATES Multidrug toxin extrusion proteins
MBC Male breast cancer
MBP Methyl CpG binding protein
MEK Mitogen-activated protein kinase
xviii Abbreviations

miRNA MicroRNA
MPA Metabolic phenotypic analysis
MRGs Master regulator genes
MRI Magnetic resonance imaging
mRNA Messenger RNA
MS Mass spectrometry
MSP Methylation specific PCR
mTOR Mammalian target of rapamycin
MudPit Multidimensional protein identification technology
MUFAs Monounsaturated fatty acids
NCBI GEO National Center for Biotechnology Information Gene
Expression Omnibus
NCI National Cancer Institute
NCNN National Comprehensive Cancer Network
ncRNA Non-coding RNA
NGS Next-generation sequencing
NIH National Institutes of Health
NLST National Lung Screening Trial
NMR Nuclear magnetic resonance
NPV Negative predictive value
OA Oleic acid
ODE Ordinary differential equations
ORR Objective response rate
OS Overall survival
PAM50 Prediction Analysis of Micro-array 50
PARPs Poly ADP-ribose polymerase inhibitors
PB Peripheral blood
PC Principle component
PCA Principle component analysis
pCR Pathologic complete response
PCR Polymerase chain reaction
PET Positron emission tomography
PFS Progression-free survival
PHGDH Phosphoglycerate dehydrogenase
PLS-DA Partial least squares discriminant analysis
PNN Probabilistic neural network
ppb Parts per billion
ppt Parts per trillion
PR Progesterone receptor
PTMs Post-translational modifications
PtNP Platinum nanoparticles
PTR Proton transfer
PUFAs Polyunsaturated fatty acids
QM-MSP Quantitative multiplex-methylation specific PCR
QOL Quality of life
qRT-PCR Quantitative real-time polymerase chain reaction
RASSF1A Ras association domain family protein 1
RECIST Response Evaluation Criteria in Solid Tumors
Abbreviations xix

RIP RNA immunoprecipitation

RNAi RNA interface
ROLL Radionuclide lesion localization
ROS Reactive oxygen species
RT Reverse transcription
SAGE Serial analysis of gene expression
SAHA Vorinostat
SELDI-MS Surface-enhanced laser desorption/ionization mass
SERM Selective estrogen receptor modulator
SLN Sentinel lymph node
SLNB Sentinel lymph node biopsy
SNPs Single nucleotide polymorphisms
SNRI Serotonin noradrenalin reuptake inhibitor
SNS Single-nucleus sequencing
SP Side population (technique/cells)
SPME Solid phase micro-extraction
SRA Steroid receptor RNA activator
SSRI Serotonin reuptake inhibitor
TCA Tricarboxylic acid
TCGA The Cancer Genome Atlas
TERRA (or Tel RNA) Telomeric repeat-containing RNA
TGF-β Transforming growth factor-β
TKI Tyrosine kinase inhibitor
TLC Thin-layer chromatography
TOF Time of flight
TRAM Transverse rectus abdominus myocutaneous
TTF Time to treatment failure
T-UCRs Transcribed-ultraconserved regions
UGTs Glucurosyltransferases
UPSIO Ultrasmall paramagnetic iron oxide
VEGF Vascular endothelial growth factor
VOCs Volatile organic compounds
VOMs Volatile organic metabolites
XIST X-inactive specific transcript
Zfas1 Zinc finger antisense

Part I Omics Approaches in Breast Cancer

1 Omics Technologies Applied in Breast Cancer Research. . . . . 3

Mariana Panal Cusati, Maria Herrera de la Muela,
and Ignacio Zapardiel
2 Omics of Hereditary Breast Cancer . . . . . . . . . . . . . . . . . . . . . . 17
Catherine A. Moroski-Erkul, Burak Yilmaz, Esra Gunduz,
and Mehmet Gunduz
3 Oncogenes and Tumor Suppressor Genes as a Biomarker
in Breast Cancer. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
Eyyup Uctepe, Muradiye Acar, Esra Gunduz,
and Mehmet Gunduz
4 Breast Cancer Genomics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
Birendra Kumar
5 Epigenomics of Breast Cancer . . . . . . . . . . . . . . . . . . . . . . . . . . 105
Kursat Oguz Yaykasli, Ertugrul Kaya, and Emine Yaykasli
6 Nutrigenomics in Breast Cancer. . . . . . . . . . . . . . . . . . . . . . . . . 127
Shailendra Dwivedi, Shailja Shukla, Apul Goel,
Praveen Sharma, Sanjay Khattri, and Kamlesh Kumar Pant
7 Long Noncoding RNAs in Breast Cancer: Implications
for Pathogenesis, Diagnosis, and Therapy . . . . . . . . . . . . . . . . . 153
Jaroslav Juráček, Robert Iliev, Marek Svoboda,
and Ondrej Slaby
8 Breast Cancer MicroRNAs: Clinical Biomarkers for
the Diagnosis and Treatment Strategies. . . . . . . . . . . . . . . . . . . 171
Gowhar Shafi, Tarique N. Hasan, Naveed Ahmed Syed,
Ananta Paine, Jesper Tegner, and Anjana Munshi
9 Breast Cancer Proteomics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
Ida Pucci-Minafra
10 Metabolomics in Breast Cancer . . . . . . . . . . . . . . . . . . . . . . . . . 211
Maria C. Calomarde, Javier De Santiago, and Ignacio Zapardiel

xxii Contents

11 Lipidomics in Breast Cancer. . . . . . . . . . . . . . . . . . . . . . . . . . . . 225

Alvin Kamili and Jennifer A. Byrne
12 Breast Cancer Stem Cells and Cellomics. . . . . . . . . . . . . . . . . . 245
Esin Demir, Bilge Atar, Dipali Dhawan, Debmalya Barh,
Mehmet Gunduz, and Esra Gunduz
13 Omics of Male Breast Cancer . . . . . . . . . . . . . . . . . . . . . . . . . . . 265
Zahide Nur Unal, Gülhan Kaya, Debmalya Barh, Esra Gunduz,
and Mehmet Gunduz
14 Omics of Chemoresistant and Metastatic Breast Cancer. . . . . 277
Margarita Aguilera and Juan Antonio Marchal
15 Animal Models of Breast Cancer . . . . . . . . . . . . . . . . . . . . . . . . 297
E.A. El-Abd, A.S. Sultan, E.A. Shalaby, and F. Matalkah
16 In Silico Disease Models of Breast Cancer. . . . . . . . . . . . . . . . . 315
Anjana Munshi and Vandana Sharma
17 Systems Biology and Integrative Omics in Breast Cancer. . . . 333
Enrique Hernández-Lemus

Part II Breast Cancer: Next-Generation Diagnosis, Prognosis,

and Therapy

18 Gynecologic Considerations for Women

with Breast Cancer. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 355
William R. Robinson III and Kaneez Fatima Ali
19 Imaging Technologies and Applications in Early
Diagnosis and Prognosis for Breast Cancer. . . . . . . . . . . . . . . . 371
M. Mar Gil, Marcos Cuerva, Sara Iacoponi,
Jose I. Sanchez-Mendez, and Ignacio Zapardiel
20 Breast Cancer Biomarkers for Risk Assessment, Screening,
Detection, Diagnosis, and Prognosis. . . . . . . . . . . . . . . . . . . . . . 393
Mukesh Verma and Debmalya Barh
21 Breast Circulating Tumor Cells: Potential Biomarkers
for Breast Cancer Diagnosis and Prognosis Evaluation. . . . . . 409
Phuc Van Pham
22 Molecular Diagnosis of Metastasizing Breast Cancer
Based Upon Liquid Biopsy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 425
Shailendra Dwivedi, Apul Goel, Sadashiv, Arti Verma,
Shailja Shukla, Praveen Sharma, Sanjay Khattri,
and Kamlesh Kumar Pant
23 Exhaled Volatile Organic Compounds as Noninvasive
Markers in Breast Cancer. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 461
Orna Barash and Hossam Haick
Contents xxiii

24 Breast Cancer Therapy–Classical Therapy, Drug Targets,

and Targeted Therapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 483
Eugenia Ch Yiannakopoulou
25 Pharmacogenomics–Pharmacoepigenomics of Breast
Cancer Therapy: Clinical Implications . . . . . . . . . . . . . . . . . . . 499
Eugenia Ch Yiannakopoulou and Debmalya Barh
26 Breast Cancer Gene Therapy . . . . . . . . . . . . . . . . . . . . . . . . . . . 519
Gülay Büyükköroğlu, Duygu Abbasoğlu, and Candan Hızel
27 Clinical Trial Endpoints in Breast Cancer . . . . . . . . . . . . . . . . 535
Melvin George and Sandhiya Selvarajan

Duygu Abbasoğlu, MSc Faculty of Pharmacy, Yunusemre Campus,

Anadolu University, Eskişehir, Turkey
Muradiye Acar, PhD Department of Medical Genetics, Turgut Özal
University Medical School, Ankara, Turkey
Margarita Aguilera, PhD Centre for Biomedical Research (CIBM),
Insituto Biosanitario de Granada (ibs.GRANADA), Hospitales
Universitarios de Granada-Universidad de Granada, Granada, Spain
Kaneez Fatima Ali, MD Tulane Cancer Center, Tulane University School
of Medicine, New Orleans, LA, USA
Bilge Atar, BS, MSc Department of Medical Genetics, Turgut Özal
University Medical School, Ankara, Turkey
Orna Barash, PhD Department of Chemical Engineering, Russel Berrie
Nanotechnology Institute, Technion-Israel Institute of Technology,
Haifa, Israel
Debmalya Barh, MSc, MTech, MPhil, PhD, PGDM Centre for Genomics
and Applied Gene Technology, Institute of Integrative Omics and Applied
Biotechnology (IIOAB), Nonakuri, Purba Medinipur, West Bengal, India
Gülay Büyükköroğlu, PhD Pharmaceutical Biotechnology,
Faculty of Pharmacy, Anadolu University, Tepebaşı, Eskişehir, Turkey
Jennifer A. Byrne, PhD Molecular Oncology Laboratory, Children’s
Cancer Research Unit, Kids Research Institute, The Children’s Hospital at
Westmead, Westmead, NSW, Australia
Maria C. Calomarde, MD Department of Gynecologic Oncology, La Paz
University Hospital, Madrid, Spain
Marcos Cuerva, PhD Gynecologic Oncology Unit, La Paz University
Hospital, Madrid, Spain
Mariana Panal Cusati, MD Department of Gynecologic Oncology, La Paz
University Hospital, Madrid, Spain
Maria Herrera de la Muela, MD, PhD Department of Gynecologic
Oncology, La Paz University Hospital, Madrid, Spain

xxvi Contributors

Esin Demir, BS, MSc Department of Medical Genetics, Turgut Özal

University Medical School, Ankara, Turkey
Javier De Santiago, MD, PhD Department of Gynecologic Oncology, La
Paz University Hospital, Madrid, Spain
Dipali Dhawan, PhD Institute of Life Sciences, Ahmedabad University,
Ahmedabad, Gujarat, India
Shailendra Dwivedi, MSc, PhD Department of Biochemistry, All India
Institute of Medical Sciences, Jodhpur, India
E.A. El-Abd, PhD Department of Radiation Sciences, Medical Research
Institute, Alexandria University, Alexandria, Egypt
Faculty of Science, Biology Department, University of Hail, Saudi Arabia
Melvin George, MBBS, MD, DM Clinical Trials and Research Unit,
Department of Cardiology, SRM Medical College Hospital and Research
Centre, Chennai, Tamil Nadu, India
M. Mar Gil, MD Gynecologic Oncology Unit, La Paz University Hospital,
Madrid, Spain
Apul Goel, MS, Mch Department of Urology, King George Medical
University, Lucknow, India
Esra Gunduz, PhD Department of Medical Genetics, Turgut Özal
University Medical School, Ankara, Turkey
Mehmet Gunduz, MD, PhD Department of Medical Genetics, Turgut
Özal University Medical School, Ankara, Turkey
Hossam Haick, BSc, PhD Department of Chemical Engineering, Russel
Berrie Nanotechnology Institute, Technion-Israel Institute of Technology,
Haifa, Israel
Tarique N. Hasan, MSc Department of Biotechnology, R & D Center,
Bharathiar University, Coimbatore, Tamil Nadu, India
Enrique Hernández-Lemus, PhD Computational Genomics Department,
National Institute of Genomic Medicine, , Mexico City, Mexico
Candan Hizel, PhD Faculty of Pharmacy, Anadolu University, lTepebaşı,
Eskişehir, Turkey
Sara Iacoponi, PhD Gynecologic Oncology Unit, La Paz University
Hospital, Madrid, Spain
Robert Iliev, MS Department of Molecular Oncology II – Solid Cancer,
CEITEC: Central European Institute of Technology, Brno, Czech Republic
Jaroslav Juráček, BS Department of Molecular Oncology II – Solid
Cancer, CEITEC: Central European Institute of Technology, Brno, Czech
Alvin Kamili, PhD Molecular Oncology Laboratory, Children’s Cancer
Research Unit, Kids Research Institute, The Children’s Hospital at
Westmead, Westmead, NSW, Australia
Contributors xxvii

Ertugrul Kaya, MD Department of Medical Pharmacology, School of

Medicine, Duzce University, Duzce, Turkey
Gülhan Kaya, BS, MSc Department of Medical Genetics, Turgut Özal
University Medical School, Ankara, Turkey
Sanjay Khattri, MBBS, MD Department of Pharmacology and
Therapeutics, King George Medical University, Lucknow, India
Birendra Kumar, PhD Department of Pathology, Vardman Mahavir
Medical College and Safdarjung Hospital, New Delhi, India
F. Matalkah, MSC Department of Biomedical Sciences, School of
Medicine, West Virginia University, Morgantown, WV, USA
Catherine A. Moroski-Erkul, BA Department of Medical Genetics,
Turgut Özal Üniversity Medical School, Ankara, Turkey
Anjana Munshi, PhD Department of Molecular Biology, Centre for
Human Genetics, School of Health Sciences, Central University of Punjab,
Bathinda, Punjab, India
Ananta Paine, PhD Deparment of Medicine, Karolinska University
Hospital, Karolinska Institute, Stockholm, Sweden
Kamlesh Kumar Pant, MBBS, MD Department of Pharmacology and
Therapeutics, King George Medical University, Lucknow, India
Ida Pucci-Minafra, PhD Centro di Oncobiologia Sperimentale, Palermo
University, La Maddalena Oncology Clinic, Palermo, Italy
William R. Robinson III, MD Maxwell E Lapham Professor of
Gynecologic Oncology, Department of Obstetrics and Gynecology, Tulane
University School of Medicine, New Orleans, LA, USA
Sadashiv, PhD Department of Physiology, King George Medical
University, Lucknow, India
Jose I. Sanchez-Mendez, PhD Gynecologic Oncology Unit, La Paz
University Hospital, Madrid, Spain
Sandhiya Selvarajan, MBBS, MD, DNB, DM Department of Clinical
Pharmacology, Jawaharlal Institute of Postgraduate Medical Education and
Research (JIPMER), Puducherry, India
Gowhar Shafi, PhD Department of Medicine, Karolinska University
Hospital, Karolinska Institute, Stockholm, Sweden
E.A. Shalaby, PhD Biochemistry Department, Cairo University, Giza,
Praveen Sharma, MSc, PhD Department of Biochemistry, All India
Institute of Medical Sciences, Jodhpur, India
Vandana Sharma, MSc Centre for Human Genetics, School of Health
Sciences, Central University of Punjab, Bathinda, Punjab, India
Shailja Shukla, BDS, PhD Department of Pharmacology and
Therapeutics, King George Medical University, Lucknow, India
xxviii Contributors

Ondrej Slaby, PhD Research Group Molecular Oncology II, Central

European Institute of Technology, Masaryk University, Brno, Czech
Ahmed Samir Sultan, MSc, PhD, MPH Biochemistry Department,
Faculty of Science, Alexandria University, Alexandria, Egypt
Marek Svoboda, PhD, MD Department of Molecular Onocology II –
Solid Cancer, CEITEC: Central European Institute of Technology, Brno,
Czech Republic
Naveed Ahmed Syed, MSc Department of Biotechnology, R & D Center,
Bharathiar University, Coimbatore, Tamil Nadu, India
Jesper Tegner, PhD Department of Medicine, Karolinska University
Hospital, Karolinska Institute, Stockholm, Sweden
Eyyup Uctepe, PhD Department of Medical Genetics, Turgut Özal
Üniversity Medical School, Ankara, Turkey
Zahide Nur Unal, BS, MSc Department of Medical Genetics, Turgut Özal
University Medical School, Ankara, Turkey
Phuc Van Pham, PhD Laboratory of Stem Cell Research and Application,
University of Science, Vietnam National University, Ho Chi Minh City,
Arti Verma, MSc, PhD Department of Physiology, King George Medical
University, Lucknow, India
Mukesh Verma, PhD Epidemiology and Genomics Research Program,
National Cancer Institute, National Institutes of Health (NIH), Rockville,
Emine Yaykasli, MSc Department of Medical Genetics and Biology,
Graduate School of Health Sciences, Duzce University, Duzce, Turkey
Kursat Oguz Yaykasli, PhD Department of Medical Genetics, School of
Medicine, Duzce University, Duzce, Turkey
Eugenia Ch Yiannakopoulou, MD, MSc, PhD Faculty of Health and
Caring Professions, Technological Educational Institute of Athens, Athens,
Burak Yilmaz, BA Department of Medical Genetics, Turgut Özal
Üniversity Medical School, Ankara, Turkey
Ignacio Zapardiel, MD, PhD Department of Gynecologic Oncology,
La Paz University Hospital, Madrid, Spain
Juan Antonio Marchal, MD, PhD Department of Human Anatomy and
Embryology, Faculty of Medicine, Centre for Biomedical Research (CIBM),
Insituto Biosanitario de Granada (ibs.GRANADA), Hospitales
Universitarios de Granada-Universidad de Granada, Granada, Spain
Part I
Omics Approaches in Breast Cancer
Omics Technologies Applied
in Breast Cancer Research 1
Mariana Panal Cusati, Maria Herrera de la Muela,
and Ignacio Zapardiel

Omics are technologies used to quantify cellular components on a large
scale. Studying the genome with high-throughput research tool gene
expression profiling has produced a wide knowledge of the molecular
level process of different pathologies, particularly about cancer. Omics
technologies have been linked from the beginning to breast cancer
research, achieving the exposure of cancer heterogeneity, its genomic
complexity, and the molecular events that drive cancer biology. This whole
set of genetic information brought the understanding of breast cancer as a
heterogeneous disease with diverse morphologies, molecular characteris-
tics, and clinical behavior; therefore omics technologies are currently
being used to identify gene signatures in a need for more accurate diagno-
sis, prognosis, and treatment. A valuable quantity of genetic information
from breast cancer and multiple research groups is being publicly stored,
and analyzing and integrating these will achieve a complete and deep
understating of the pathogenesis of this disease and help to drastically
improve its clinical outcome.

Breast cancer • Omics technologies • High throughput • Gene expression
profiling • Molecular profiling technologies • Genomics • Epigenomics •
Proteinomics • Transcriptomics


Nowadays breast cancer is being regarded as

several diseases with the same name, due to its
M.P. Cusati, MD • M.H. de la Muela, MD, PhD numerous subtypes and its different histological,
I. Zapardiel, MD, PhD (*)
biological, and molecular characteristics. This
Department of Gynecologic Oncology,
La Paz University Hospital, Madrid, Spain variability brings different responses to the ther-
e-mail: ignaciozapardiel@hotmail.com apy applied, and therefore a change in its previous

D. Barh (ed.), Omics Approaches in Breast Cancer: Towards Next-Generation Diagnosis, 3

Prognosis and Therapy, DOI 10.1007/978-81-322-0843-3_1, © Springer India 2014
4 M.P. Cusati et al.

prognosis [1]. Tumor size, histological grade, cancer heterogeneity, and the molecular events
lymph node status, and hormone receptor status that drive cancer biology, hence recognizing
expression have classically been used in order to them as the pathways to creating improved thera-
set a treatment strategy and prognosis. pies specific for cancer phenotypes [6].
Nevertheless, these prognostic factors lack in Summarizing, we can use omics technologies
accuracy because there are many genetic and to refer to the study and application of the infor-
molecular factors of breast cancer disease that are mation obtained from the genes involved in
not fully understood yet [2]. It seems like some tumors, using genetic resources such as the num-
patients are over-treated, out of fear for the still ber of copies of DNA or RNA, the transcription,
unexpected and uncontrolled relapses. It is very and methylation, with the possibility to scrutinize
significant to find new markers in breast cancer cancer cells from gene to metabolite [7]. We will
that may improve detecting and classifying this describe the main technologies that study
disease, as well as predicting and tracking patient each step of the central dogma of biology, gene–
response [3, 4]. DNA–RNA–protein, which in the same order are
teinomics, showing how they are applied in breast
The Birth of the Omics Era cancer research, although there are many more
omics technologies. Figure 1.1 explains the rela-
In 1920 the botanist Hans Wrinkler used a new tionship between gene, DNA, RNA, and protein
word to describe the entire genetic material of an (the central dogma of molecular biology) with all
animal or plant: genome, a combination of the the information obtained about them through
words gene and chromosome. Some scientists omics technologies.
interpreted the suffix “ome” as a collective from
a unit, with gene being the unit, and genome col-
lectively, so then this suffix was applied to create Genomics
more words. The first example is the use of pro-
teome as a word to describe the whole set of pro- Genomics may be described as the comprehen-
tein derivate from genome. At the end of 1990, sive analysis of genes and their DNA structure
the word genomics was beginning to be used to and function or as the scientific discipline look-
describe the study and application of the infor- ing for information about the entire genome.
mation obtained from the genome, which resulted Before the existence of omics technologies, DNA
in the suffix “omics.” This new genetic language was sequenced gene by gene, but improving
has started the omics age, in which the word informatics technology and integrating them to
omics now refers to the comprehensive analysis the biology have allowed the sequencing of the
of biological systems [5]. entire genomes of many types of organisms and
Omics are high-throughput technologies used the classification and storage of DNA sequences
to quantify cellular components on a large or in genomic databases. Variations of the genes or
wide scale as a genome or proteome. This tech- DNA sequences can be created through gene
nological breakthrough has brought cancer amplification, gene deletion, or gene rearrange-
genomics, or the study of tumor genome, via ment [8].
various profiling data such as DNA copy number, Copy number aberrations (CNA) or copy
DNA methylation, transcription, and genome number variations (CNV) are the names that
sequencing of cancer cells. This allows identifi- encompass the alterations of the organization and
cation of genetic pathways of cancer, bringing a amount of DNA within a cell. Genetic alterations
greater understanding of the biology of cancer are generated by deletions, duplications, inver-
and leading to the discovery of novel diagnostic, sions, or translocations of chromosomes; more-
prognostic, and therapeutic options. Omics tech- over these modifications are capable of being
nologies have exposed the genomic complexity, inherited. CNA or CNV can disturb the normal
1 Omics Technologies Applied in Breast Cancer Research 5

Fig. 1.1 Relationship

between gene, DNA, RNA,
and protein (the central
dogma of molecular biology)
with omics technologies

transcriptional activity of a cell, increasing or surface fixed with either a tissue or a panel of dif-
decreasing it, changing the normal expression of ferent DNA or RNA molecules, proteins, or anti-
genes [9]. bodies that have the possibility of linking up with
Gene amplification is one of the mechanisms a corresponding DNA/RNA/protein/antibody
in which a gene can be overexpressed and cause from a sample. Genomic microarrays, also called
the activation of oncogenes; therefore it is related array comparative genomic hybridization (array-
to disease progression and poor prognosis. It is CGH), are used to study and quantify chromo-
created by an increase in the copy number of a somal abnormalities, microdeletions and
loci or restricted area of a chromosome arm. microduplications, and copy number aberrations
Amplification occurs recurrently on some chro- (CAN), at a wide level in an entire genome.
mosomal locations, indicating the common acti- Array-CGH has a refined process and resolution,
vation of some oncogenes during tumor which enables evaluation of the genome of any
development [10]. The HER2 is the most studied cell, while chromosomal-CGH allows the identi-
breast cancer oncogene; it is located on chromo- fication of gene regions where there is DNA gain
some 17q21.1 and encodes a transmembrane pro- or DNA loss [1, 12].
tein that is similar to the epidermal growth factor Chromosomal-CGH studies have identified
the HER1. A normal cell presents two copies of that the loss of 16q is one of the most consis-
the HER2 gene and about 50,000 copies of the tent DNA aberrations found in infiltrating lobular
protein, whereas by gene amplification in a breast carcinoma (ILC). Etzell et al. studied genomic
cancer cell, there can be more than one copy of alterations in lobular carcinoma in situ (LCIS).
the HER2 gene and more than one million copies Using comparative genomic hybridization, they
of the protein [11]. found loss of chromosome 16q was in 88 % of
High-throughput technologies are needed to cases [13]. Mastracci et al. studied the genetic
discover, monitor, and quantify these DNA modi- profile of atypical lobular hyperplasia (ALH)
fications. The most widely used technique and lobular carcinoma in situ (LCIS), using
involves the microarrays that simultaneously microarray comparative genomic hybridization
monitor the expression levels of thousands of (CGH). They found a common alteration in both
genes inside different samples. This technique ALH and LCIS—the loss of 16q21-q23.1 [14].
has advanced rapidly in recent years. It involves a This common genetic alteration highlights the
6 M.P. Cusati et al.

Table 1.1 Confirmation of the intrinsic subtypes with several studies

Study Main technique Other technique Novel finding
Sørlie et al. 2001 [18] cDNA microarrays Hierarchical clustering Division of luminal A
and B subtypes
Sørlie et al. 2003 [19] cDNA microarrays Hierarchical clustering Association of BRACA1
with basal subtype
Sørlie et al. 2003 [19] cDNA microarrays Hierarchical clustering Correlation of ER status
with intrinsic subtypes
Abd El-Rehim et al. 2004 [21] Tissue microarray Immunohistochemistry Correlation with
cytokines expression
Carey et al. 2006 [22] Microarrays Immunohistochemistry Population-based
distribution of subtypes
Hu et al. 2006 [65] Microarrays Hierarchical clustering Prediction of survival

possibility of a relationship between ILC and being studied and recognized as distinct diseases
LCIS as they may be different forms of the same with different treatment options [16, 17].
disease, and it seems that lobular carcinoma in Table 1.1 summarizes the main findings of sev-
situ (LCIS) is a precursor to infiltrating lobular eral studies that confirm the existence of the
carcinoma and therefore a marker of risk for intrinsic subtypes, the technology used, and the
breast cancer. This serves as an example of the novel finding that each study provided to this new
application of genomic studies in order to dis- breast cancer classification.
cover breast cancer initiation, progression, and
metastasis [13, 14].
Breast cancer gene expression profiling per- Novel Classification of Breast
mits the collection of all the information about a Cancer Based on Genome Profile:
whole set of genes in a tumor cell, including vari- The Intrinsic Subtypes
ations, gene expression, and the way those genes
interact with each other and with the environ- Luminal Type
ment. The associations of specific genes with a
common characteristic of expression or pheno- This type responds to endocrine therapy and has
types are called gene profiles or gene signatures. the best prognosis of all subtypes although it
Gene expression profiling has been used to create shows only limited chemosensitivity. Sorlie et al.
a complete new classification of breast carcino- studied a subdivision of luminal-like carcinoma
mas. The different kinds of clinical breast cancer Type A and Type B with characteristic molecular
have been correlated with diversity in gene profiles and different prognoses (better for the A
expression profiles, which usually are studied type than the B) [18]. In luminal B type, the HER2
using DNA microarray techniques [1]. Perou expression is greater and responds only slightly
et al. using DNA microarrays in samples of breast better to chemotherapy than luminal A, which
cancer and normal breast tissue studied the pres- seems to have lower risk for relapse [18, 19].
ence of sets of genes that they called intrinsic
genes, because these genes were found repeat-
edly inside the same patient biopsies but were not Basal-Like
found repeatedly in all samples from other
patients. A new group of genes associated with Defined by the expression of cytokeratins with-
different breast cancer types were discovered, out ER and HER2 expression, it has shown poor
showing four distinct molecular subgroups: ER+/ survival independent of nodal status and size and
luminal-like, basal-like, HER2 enriched, and is the subtype most common when the patient has
normal breast-like [15]. This new classification is BRCA1 [16, 19, 20].
1 Omics Technologies Applied in Breast Cancer Research 7

HER2 Enriched intrinsic subtypes, they found that the rate and
location of relapses are correlated to different
This type is characterized by the overexpres- molecular subtypes and metastases conserve
sion of the HER2 gene, and it has no expression genetic similarities with their primaries [24].
of genes characteristic of the luminal subtypes. Following the intrinsic gene research of today,
Not all breast cancers with HER2 positive by a new classification of breast cancer exists using
immunohistochemistry are classified as HER2- omics technologies. New trials show the possibil-
enriched type by molecular profiling. The basal- ity of more subtypes, but this pathway is barely
like and the Her2 enriched seem the ones with started, and currently it is not possible to exactly
worst prognoses, the luminal A the best, and the compare the new subtypes with the classical
luminal B in the intermediate [18, 19]. ones, and it is necessary for more studies to set up
Breast cancer tumors had been separated by many new targeted therapies for these new
expression of hormone receptor status defining molecular subtypes [11, 15].
biologically distinct phenotypes that are used on Applying this new genetic information on the
clinical basis for prognosis and treatment. cluster of genes related to breast cancer, and
Genomics links phenotypes to genotypes of using the new subtypes or intrinsic types of
breast cancer by means of the study of the intrin- breast cancer, there is now a connection between
sic subtypes and their clinical characteristics. The the lab and the patient with the creation of sev-
relationship between breast cancer molecular eral molecular assay tests that use genomics to
subtypes and prognosis has been studied exten- calculate risk assessment and prognosis, creat-
sively by gene expression profiling in order to ing a personalized diagnosis, prognosis, and
find the differences that are used in clinical basis therapy. This genomic test calculates certain
using hormone receptor status, HER2 status, and clinical outcomes, such as patient risk of relapse
other classical characteristics. Among the intrin- if avoiding chemotherapy or if treated only with
sic subtypes, luminal A and B have been identi- hormonal therapy and patients having more
fied as the ones with high expression of hormone global risk of relapse. Although these tests
receptor, while the HER2-enriched and basal-like searched different genes, when they are applied
subtypes have low expression of hormone recep- within the same patient data, they show similar
tors. In addition, low claudin, interferon rich, results on prognosis of breast cancer. The
androgen receptor, and normal-like are the other MammaPrint and Oncotype DX are the tests
subtypes studied in new trials [19]. There are most used and studied [1, 3].
many studies that find a relationship between tra- The MammaPrint analyzes 70 genes and
ditional prognostic factors with gene expression shows which patients have a low-risk molecular
signatures [21–23]. profile and therefore can avoid chemotherapy. It
also identifies those women with good clinical
factors but who carry poor prognosis genes [1, 3,
Genomic Testing 4]. Van de Vijver used inkjet-synthesized oligo-
nucleotide microarrays to study the 70-gene
The molecular subtypes in breast cancer have expression profile that is associated with progno-
shown different tumor characteristics, tumor sis in patients with breast cancer and validate the
aggressiveness, and different response to certain classification system to predict the likelihood of
types of chemotherapy. They all have the ability distant metastases within 5 years [25].
to progress to metastases, but with differences Oncotype DX uses the intrinsic subtype model
regarding their leaning toward relapse into differ- with 21 genes and is used in estrogen and proges-
ent organs. Smid et al. used microarrays and ana- terone receptor-positive and lymph node-negative
lyzed them with statistical significance analysis breast cancer, to identify patients who can avoid
of microarrays (SAM). Studying the preference adjuvant chemotherapy [1, 3]. Paik et al., using
for organ-specific metastases of each of the reverse transcription polymerase chain reaction
8 M.P. Cusati et al.

(RT-PCR), studied 21 selected genes and found One example of storage of new public oncoge-
the correlation with the likelihood of distant nome data is the Cancer Genome Atlas (TCGA),
recurrence in patients with node-negative disease which is “a comprehensive and coordinated effort
and treated with tamoxifen, validating then a to accelerate the understanding of the molecular
recurrence score [26, 27]. basis of cancer through the application of genome
Although these tests are beginning to be used analysis technologies, including large-scale
in daily medical practice, there is still no consen- genome sequencing.” This atlas is created with
sus about their reliability or guidelines for use, the efforts of the National Cancer Institute (NCI)
because there is little information about the and the National Human Genome Research
impact that these tests may have on clinical out- Institute (NHGRI) that are applying omics tech-
comes in the long term. Where there seems to be nologies to profile DNA, RNA, protein, and oth-
agreement is that the information provided by ers that can be related to this disease. This data is
these tests must be taken into account as addi- being stored within several samples of cancer,
tional information to that obtained by conven- creating a new list of genes, mutations, and
tional methods [16, 17]. molecules.
Regarding the possibility of properly selecting Now there are tons of genetic data related to
the patient who should use chemotherapy and the breast cancer, but there is lack in the final connec-
one who should avoid it, as personalized medi- tion with the actual patient [6, 11]. Curtis et al.
cine should be able to do, it seems necessary to used almost 2,000 samples of breast cancer to
relate a set of genes and its regulation systems analyze their genome and transcriptome within
with a clinical outcome [4, 11]. In order to copy number variants, single nucleotide poly-
accomplish more scientific evidence, prospective morphism, and acquired somatic copy number
clinical trials—MINDACT for MammaPrint and aberrations, and they found new subgroups to
TAILORx for Oncotype Dx— are trying to create a novel molecular stratification of the
improve their clinical application [28–30]. breast cancer population [33]. The Cancer
One of the main drawbacks of applying past Genome Atlas network 2012 used the analysis of
genomics research on a daily clinical basis is the large amounts of data through the TGGA plat-
lack of validation because these trials were based form studying the DNA copy number arrays,
on small samples. That is why the latest research DNA methylation, exome sequencing, messen-
is focused on using more biopsy samples and ger RNA arrays, microRNA sequencing, and
more genetic analysis than the first trials, creating reverse-phase protein arrays from breast cancer,
a genomics landscape of breast cancer [6]. Using showing the existence of four main breast cancer
larger samples in order to integrate breast cancer classes but associating them with several gene
genome and transcriptome, and search for molec- mutations and subgroups of proteins expression.
ular drivers and gene expression, seems to be They also found many molecular commonalities
more reliable than obtaining information through between basal-like breast tumors and high-grade
the analysis of smaller samples [6, 31]. The cre- serous ovarian tumors [34].
ation of biobanks allows the storage of multiple
samples of different types of cancer, from differ-
ent patients, and treatments using different tech- Epigenomics
niques of conservation. The existence of a useful
biobank offers multiple and extensive databases Epigenetic modifications are the alterations
using many categories, such as family samples, found in DNA not affecting its primary structure,
pre- and posttreatment samples, and population such as DNA methylation, histone modifications,
characteristics, to classify and identify the sam- and chromatin or nucleosomal remodeling; these
ples. To improve the biobank utilization, ethical alterations are heritable and reversible. Therefore
issues must be addressed regarding the use of the epigenomics is the study of the complete set of
samples for different studies [32]. those epigenetic modifications [7, 35].
1 Omics Technologies Applied in Breast Cancer Research 9

DNA methylation, histone modifications, and gene-specific methylation analysis. The global
chromatin or nucleosomal remodeling mutually methylation analysis measures the overall level
interact with each other to regulate gene expres- of methyl cytosines. The gene-specific meth-
sion. The modifications obtained through them ylation analysis uses enzymes to digest DNA
sometimes end up with aberrant transcription dis- to process it and quantify and identify meth-
rupting the normal regulation of a cell. The epi- ylated genes. One of the processes used is the
genetics processes modulation of chromatin polymerase chain reaction (PCR), a main tech-
structure to form a loosely packed DNA that is nique for omics that amplifies a DNA sequence
transcriptionally active, called euchromatin, or a using an enzymatic reaction to obtain thousands
tightly packed DNA that is transcriptionally inac- to millions of copies of it. Methylation-specific
tive, the heterochromatin. Therefore, these pro- PCR (MSP) is able to identify methylated
cesses can either turn on or turn off the gene sequences or hypermethylated zones inside of
expression. the genome [39, 40].
Cancer seems to be driven by these epigenetic Dejeux et al. studied DNA methylation of cer-
alterations, which cause aberrant cell regulation tain genes from samples of patients with locally
that result in a change in expression patterns of advanced breast cancer before the neoadjuvant
genes implicated in cellular proliferation, sur- treatment with doxorubicin. To quantify the
vival, and differentiation. Therefore, in breast modification, they used DNA methylation anal-
cancer cells, DNA methylation, histone modifi- ysis by pyrosequencing. They found aberrant
cations, and chromatin or nucleosomal remod- methylation in 9 of 14 genes studied, and three
eling can lead to initiation, promotion, and of the variations found are possible biomarkers
maintenance of carcinogenesis and treatment fail- for prognosis for patients with this disease and
ure. Understanding epigenetic alterations that ini- the prediction of response of this neoadjuvant
tiate and maintain gene silencing and oncogene drug [41].
activation in cancer cells is necessary in order to Hsu et al. studied the methylation of BRCA1
find clinical applications of epigenomics [36–38]. promoter by methylation-specific PCR. They
found a significant correlation within triple-
negative cancer and with poor overall survival
DNA Methylation and disease-free survival in patients with breast
cancer with methylated BRCA1 promoter [42].
DNA methylation is the addition of a methyl
group into a region of DNA with participation of
the DNA methyltransferase enzymes. They cre- Histones
ate genomic instability and rearrangement, acti-
vation of oncogenes, and inactivation of tumor Histones are protein components of chromatin
suppressor genes, resulting in a cancerous cell that package and order the DNA into structural
[36]. Inside a cell there may be a loss of DNA units called nucleosomes. The nucleosome is the
methylation (hypomethylation) or there can exist fundamental unit of chromatin and it is composed
an excess of DNA methylation (hypermethyl- of a structure of four histones around which the
ation). Currently genes that are aberrantly meth- DNA is wrapped. Histone modifications and
ylated in breast cancer can be identified; with chromatin or nucleosomal remodeling are medi-
DNA methylation mapping technologies, it ated by various enzymes, and it seems that a
would be possible to identify DNA methylation deregulation of normal DNA arrangement pat-
patterns within a tumor subgroup that are corre- terns turns on modifications that generate a new
lated to breast cancer subtypes, clinical stages, histone code, resulting in chromatin regions for
and prognosis. transcription activation or repression, leading to
Methods for DNA methylation analysis can changed gene expression as tumor suppressor
be divided into global methylation analysis and genes and oncogenes [37].
10 M.P. Cusati et al.

The study of histone modifications is more includes different kinds of transcript, such as a
difficult than finding DNA methylation and needs messenger RNA (mRNA), noncoding RNA, and
high-throughput techniques such as proteomic small RNA; the structure of genes where the
techniques that are described in more detail in the transcriptions are located; and the expression of
section “Proteomics.” One such technique is sta- each transcript under different conditions [47].
ble isotope labeling with amino acids in cell cul- Transcriptomic techniques profile all types of
ture (SILAC), based on mass spectrometry [43]. transcript, find the start and end sites of genes to
Cuomo et al. used SILAC technology to quantify locate where transcriptions occur, and under-
and study histone modification changes in breast stand the posttranscriptional modifications,
cancer cells and compared them to normal breast enabling to decode all the RNA sequences in
tissue histone modifications. They found signifi- order to map and quantify it [46].
cant changes for histone modifications in cancer
cells that could represent biomarkers or what
they called “breast cancer-specific epigenetic sig- Gene Expression
nature” [39].
Gene expression is the study of the activity of a
gene that can be quantified by the measurement
Immunohistochemistry (IHC) of DNA transcription into mRNA and mRNA
into proteins [12]. Knowing oncogene expression
Another combined omics technique, IHC uses provides additional information about the genetic
the linkage between antigen and antibody to processes that drive cancer. Various high-
localize specific antigens in microarray tissues throughput technologies enable the measurement
by labeling antibodies with fluorescent material and classification of the transcriptome as differ-
[44]. Elsheikh et al. used this technology with ent types of microarrays (cDNA and oligonucle-
breast cancer samples in order to detect specific otide microarrays) and next-generation sequence
histone marks. They found significance between techniques [46].
low and high levels of individual histone marks
with a range of clinical and pathologic variables Complementary DNA (cDNA)
of the disease; also they could relate histone cDNA is the name used for a fragment of DNA
modifications with tumor type beside a rela- that is complementary to another mRNA frag-
tionship among histone modifications with bio- ment. cDNA can be formed in the laboratory by
logical markers and phenotypic groups of breast extracting mRNA and then using mRNA as a
cancer such as estrogen receptor, progesterone matrix for formation of cDNA sequence. cDNA
receptor, expression of luminal cytokeratins, and microarrays study gene transcription or expres-
others [45]. sion through a solid surface with several frag-
ments of cDNA attached in fixed locations that
would match and link up with their complimen-
Transcriptomics tary mRNA in a process called hybridization.
If the mRNA is chemically or fluorescently
Transcription is the process by which the DNA is labeled, it is possible to measure the fluorescence
copied into RNA. The RNA can be translated intensities from each location on the array.
into peptides or proteins and used as a guide for An oligonucleotide is a short sequence of
the synthesis of proteins by ribosome. Also RNA nucleotides. An oligonucleotide microarray is a
can work as gene regulation and enzymatic activ- microarray that uses synthetically created DNA
ity [46]. Transcriptome is the collection of all oligonucleotide sequences that are chosen to
cellular RNA, and transcriptomics is the study of obtain better hybridization characteristics [1, 46].
the whole transcriptome, measuring and classify- Yao et al., trying to define genetic changes
ing all the RNA within a cell. Transcriptome involved in the progression of breast cancer and
1 Omics Technologies Applied in Breast Cancer Research 11

using cDNA array, identified two overexpressed variability of the proteome exceed that of the
genes (H2AFJ and EPS8) from samples of ductal genome [7, 50].
carcinoma in situ, invasive breast carcinomas, Proteins can have multiple types of
and lymph node metastases. They also observed modifications, such as phosphorylation, acetyla-
that the overall frequency of copy number altera- tion, and ubiquitylation, which are called post-
tions was higher in invasive tumors than in ductal translational modifications (PTMs). Multiple
carcinoma in situ. This study demonstrates that proteomic techniques are used to separate, iden-
cDNA arrays are useful tools to identify gene tify, quantify, and classify proteins existing in
expression aberrations that can in the future be low levels inside breast cancer tissues. They
used as markers for targeted therapy for breast include electrophoresis, mass spectrometry, and
cancer [12]. novel and combined high-throughput techniques
Among the novel or next-generation sequenc- such as surface-enhanced laser desorption/ion-
ing technologies is the RNA-seq. It is achieved by ization mass spectrometry (SELDI-MS), two-
taking the RNA that is studied and converting it dimensional difference in-gel electrophoresis
into a library of cDNA fragments with known (DIGE), and multidimensional protein identifica-
sequences that can be studied and classified. tion technology.
Those fragments can be matched to a reference Proteomics uses microarrays as well. The pro-
genome or reference transcripts or reassembled in teomic microarray technique uses a surface fixed
order to identify the whole set of RNA molecules with either a tissue or a panel of different proteins
detected and to measure them [47]. Huber-Keener or antibodies, and these are matched with a cor-
et al. studied gene expression in tamoxifen-resis- responding protein/antibody sample. The identi-
tant breast cancer cells using RNA-seq technol- fication and characterization of breast cancer
ogy. They studied the genes involved in the expressed proteins provide future drug targets for
development of tamoxifen resistance, indentify- more specific therapies and could find new bio-
ing the transcriptomes within tamoxifen-sensitive markers to detect breast cancer [44, 40, 51].
breast cancer cells, and compared them with the Two-dimensional electrophoresis (2DE) is a
transcriptomes inside tamoxifen-resistant breast protein profiling technique that uses a base made
cancer cells. Through this comparison they found of gel where proteins can be separated according
gene expression alterations of several mRNA and to certain characteristics, such as the isoelectric
small RNA transcripts that are related to tamoxi- point and molecular weight. Further methods are
fen resistance mechanisms as estrogen receptor needed to identify the proteins previously sepa-
function, cell regulation, transcription regulation, rated as mass spectrometry (MS). MS is a basic
and mitochondrial dysfunction. Therefore, the proteomic technology that is capable of showing
gene expression alterations found can be used for the different peptides inside a sample, ionizing
the diagnosis and treatment of tamoxifen resis- their chemical compounds to generate charged
tance [48]. molecules and measuring their mass and charge
ratios. MS can be used to directly identify pro-
teins previously separated on gels [44]. Using the
Proteomics combination of these two techniques, Rowell
et al. studied the primary isoflavone component
Proteomics involves the systematic study of pro- of soy (genistein) and its effect on breast tissue
teins in order to provide a comprehensive view from rats. They found changes on the proteomics
of their structure and function and their regula- of the cells exposed to this isoflavone, suggesting
tion of biological systems [49, 50]. Proteins are that this exposure enhances cell proliferation, cell
the working force of a cell—they catalyze enzy- differentiation, and gland maturation therefore
matic reactions, provide cell structure, serve as increasing the possibility of breast cancer [52].
cellular signals, and provide many other cellular This is an example of how proteomic studies can
and intercellular functions. The complexity and allow deep knowledge of how breast cancer
12 M.P. Cusati et al.

starts, and it could lead to find novel molecular cellular processes [7]. The new lab technologies
markers and targets for breast cancer. enable profiling more kinds of cellular metabo-
Surface-enhanced laser desorption/ionization lites. The applications of a global metabolome
mass spectrometry (SELDI-MS) is a high- analysis, or the detection of the whole set of mol-
throughput technique for proteomic analysis that ecules in organic fluids and tissues produced by
permits the selection of proteins and peptides cellular processes, include disease diagnosis and
with specific properties using different chromato- the identification of drugs or chemical exposures
graphic surfaces [44]. Hu et al. used proteomics inside a cell. Using metabolomics technologies,
with the help of SELDI-MS on breast cancer it is possible to know the metabolic profile of an
patients, patients with benign breast diseases, and individual and predict toxicity of certain drugs,
healthy women and found four new biomarkers depending on the metabolizing capacities of the
that are able to create a diagnostic model with patient. The increasing availability of these novel
significant sensitivity and specificity [53]. methods with the capacity to detect, with high
Two-dimensional difference in-gel electropho- accuracy, the levels of the metabolome will lead
resis (DIGE) is a technology that separates pro- to discovering the molecular response to a bio-
teins from different samples inside the same gel, logical treatment or to the different metabolic
making it possible to detect and quantify the dif- characteristics of a tumoral cell [58].
ferences between two or three samples studied on
the same 2D gel, enabling simultaneous visual-
ization of relatively large portions of the pro- Pharmacogenomics
teome [54, 55]. Davalieva et al. made a proteomic
analysis of breast cancer tissues and normal tis- Pharmacogenomics is an omics technology that
sue samples from patients with infiltrating ductal studies how the response of a patient to different
carcinoma using DIGE. They found overex- drugs is related to the patient’s genetics. A com-
pressed proteins inside the tumor samples not bination between pharmacology and genomics,
previously associated with breast cancer and that pharmacogenomics pursues the optimization of
are involved in cancer pathways [56]. therapies using the genotype of the patient and
Multidimensional protein identification tech- the genetic characteristics of disease in order to
nology (MudPit) uses two-dimensional liquid obtain the best drug efficacy with the lowest
chromatography combined with mass spectrome- adverse effects. Pharmacogenomics could clas-
try to identify several peptide sequences and their sify and determine the relations between patient
related proteins. It is a technology that improves genetics and drug response. This is the basis of
the classification of proteins inside different sam- individualized medicine. The understanding of
ples. Sandhu et al. used this technology to profile pharmacogenomics includes drug targets, drug
protein expression of breast cancer cells. They metabolism, drug transport, disease susceptibil-
found alterations associated with the malignant ity, and drug safety [59].
breast cancer phenotype, including differences in
the apparent levels of key regulators of the cell
cycle, signal transduction, apoptosis, transcrip- Interactomics
tional regulation, and cell metabolism [57].
The network of all interactions between mole-
cules and proteins from a cell is called interac-
Other Omics Techniques tome; therefore interactomics is the study of both
the interactions and the consequences of those
Metabolomics interactions [60]. This complete network of pro-
tein–protein interactions or molecule–protein
Metabolomics is the study of the whole set of interactions conform the signaling pathways,
small molecules that can be detected in organic metabolic pathways, and cellular processes that
fluids. These fluids represent the end products of are required for a cell survival. Therefore the
1 Omics Technologies Applied in Breast Cancer Research 13

study of these interactions allows for a deep and the implementation of better cancer screen-
understanding of the pathogenesis of many dis- ing and prevention [7, 63, 64].
eases, including cancer [61, 62]. These high-throughput molecular profiling
studies have provided great advances in the
understanding of cancer biology. Researchers are
Conclusion and Future Perspective hopeful to use the large data now stored and
transform it into a new era of cancer manage-
Personalized cancer medicine is defined as “a ment. Omics technologies are increasingly used
medical model using molecular profiling tech- and today are being regarded almost as standard
nologies for tailoring the right therapeutic strat- research tools, notwithstanding that relatively
egy for the right person at the right time, to only a small part of the information obtained via
determine the predisposition to disease at the these novel techniques are currently being used
population level, and to deliver timely and strati- on a clinical basis. The innovation of omics tech-
fied prevention” [63]. nologies applied in breast cancer research is
The appearance of cancerous cells requires focused to completely decode its pathogenesis, to
some alterations of their genetic material. Better endure the new classification using molecular
treatment of cancer could be achieved if it is subtypes with more specific characteristics, to
possible to decipher every single step that fol- accurately define its prognosis, and to find the
lows the creation of a tumor cell. There are best use of the actual therapy and novel drugs
many genetic alterations in a single tumor; with the right targets.
maybe only a few of these alterations are
responsible for the instauration of the disease,
and we have to select the alterations that should References
be targeted in order to eradicate cancer [7, 64].
Every patient has different combinations of 1. Gevaert O, De Moor B. Prediction of cancer outcome
using DNA microarray technology: past, present and
altered genes and should be treated according to
future. Expert Opin Med Diagn. 2009;3(2):157–65.
their specific oncogenic characteristics, and 2. Dowsett M, Dunbier AK. Emerging biomarkers and
perhaps each patient should receive more than new understanding of traditional markers in personal-
one treatment with different oncogenic pathway ized therapy for breast cancer. Clin Cancer Res.
targets [64].
3. Daidone MG, Zaffaroni N, Cappelletti V. Strategies
Trastuzumab is a monoclonal antibody that to translate preclinical information to breast cancer
interferes with the HER2 receptor. The HER2 patient benefit. J Natl Cancer Inst Monogr. 2011;
receptors are proteins that turn genes on and off; 2011(43):55–9.
4. Van’t Veer LJ, Dai H, van de Vijver MJ, He YD, Hart
therefore the HER2 proteins stimulate cell pro-
AA, Mao M, et al. Gene expression profiling predicts
liferation. As an example of personalized can- clinical outcome of breast cancer. Nature.
cer medicine, if we were testing trastuzumab 2002;415(6871):530–6.
(Herceptin) in a population of breast cancer 5. Lederberg J, McCray AT. ‘Ome sweet’ omics – a gene-
alogical treasury of words. Scientist. 2001;15(7):8.
patients without knowing their HER2 receptor
6. Ellis MJ, Perou CM. The genomic landscape of breast
status, the results would be different from the cancer as a therapeutic roadmap. Cancer Discov.
reality of the trastuzumab (Herceptin) effective- 2013;3(1):27–34.
ness, because as almost everyone knows, trastu- 7. Damia G, Broggini M, Marsoni S, Venturini S,
Generali D. New omics information for clinical trial
zumab shows good results but only in patients
utility in the primary setting. J Natl Cancer Inst
with HER2-positive tumors [7, 64]. Monogr. 2011;2011(43):128–33.
The use of a treatment previously studied and 8. National Human Genome Research Institute. A brief
targeted for the right patient would improve guide to genomics. 2010. Genome.gov. Retrieved 3
Mar 2011.
healthcare services and may reduce cost, increase
9. Redon R, Ishikawa S, Fitch KR, Feuk L, Perry GH,
drug efficacy, and reduce toxicity. The research Andrews TD, et al. Global variation in copy number
goal of omics technologies is to identify groups in the human genome. Nature. 2006;444(7118):
of at-risk patients with less-invasive diagnoses 444–54.
14 M.P. Cusati et al.

10. Savelyeva L, Schwab M. Amplification of oncogenes 24. Smid M, Wang Y, Zhang Y, Sieuwerts AM, Yu J, Klijn
revisited: from expression profiling to clinical appli- JG, et al. Subtypes of breast cancer show preferential
cation. Cancer Lett. 2001;167(2):115–23. site of relapse. Cancer Res. 2008;68(9):3108–14.
11. Vucic EA, Thu KL, Robison K, Rybaczyk LA, 25. Van de Vijver MJ, He YD, van’t Veer LJ, Dai H, Hart
Chari R, Alvarez CE, et al. Translating cancer AA, Voskuil DW, Schreiber GJ, Peterse JL, Roberts
‘omics’ to improved outcomes. Genome Res. C, Marton MJ, et al. A gene-expression signature as a
2012;22(2):188–95. predictor of survival in breast cancer. N Engl J Med.
12. Yao J, Weremowicz S, Feng B, Gentleman RC, Marks 2002;347:1999–2009.
JR, Gelman R, Brennan C, Polyak K. Combined 26. Paik S, Shak S, Tang G, Kim C, Baker J, Cronin M,
cDNA array comparative genomic hybridization and et al. A multigene assay to predict recurrence of
serial analysis of gene expression analysis of breast tamoxifen-treated, node-negative breast cancer.
tumor progression. Cancer Res. 2006;66(8):4065–78. N Engl J Med. 2004;351(27):2817–26.
13. Etzell JE, Devries S, Chew K, Florendo C, Molinaro 27. Paik S, Tang G, Shak S, Kim C, Baker J, Kim W,
A, Ljung BM, Waldman FM. Loss of chromosome Cronin M, Baehner FL, Watson D, Bryant J, et al.
16q in lobular carcinoma in situ. Hum Pathol. Gene expression and benefit of chemotherapy in
2001;32(3):292–6. women with nodenegative, estrogen receptor-positive
14. Mastracci TL, Shadeo A, Colby SM, Tuck AB, breast cancer. J Clin Oncol. 2006;24:3726–34.
O’Malley FP, Bull SB, et al. Genomic alterations in 28. Perou CM, Børresen-Dale AL. Systems biology and
lobular neoplasia: a microarray comparative genomic genomics of breast cancer. Cold Spring Harb Perspect
hybridization signature for early neoplastic prolifera- Biol. 2011;1:3(2).
tion in the breast. Genes Chromosomes Cancer. 29. Lyng MB, Lænkholm AV, Tan Q, Vach W, Gravgaard
2006;45(11):1007–17. KH, Knoop A, Ditzel HJ. Gene expression signatures
15. Perou CM, Sortie T, Eisen MB, van de Rijn M, Jeffrey that predict outcome of tamoxifen-treated estrogen
SS, Rees CA, et al. Molecular portraits of human receptor-positive, high-risk, primary breast cancer
breast tumours. Nature. 2000;406:747–52. patients: a DBCG study. PLoS One. 2013;8(1):e54078.
16. Strehl JD, Wachter DL, Fasching PA, Beckmann 30. Cardoso F, Van't Veer L, Rutgers E, Loi S, Mook S,
MW, Hartmann A. Invasive breast cancer: recogni- Piccart-Gebhart MJ. Clinical application of the
tion of molecular subtypes. Breast Care (Basel). 70-gene profile: the MINDACT trial. J Clin Oncol.
2011;6(4):258–64. 2008;26(5):729–35.
17. Kao J, Salari K, Bocanegra M, Choi YL, Girard L, 31. Stephens PJ, Tarpey PS, Davies H, Van Loo P,
Gandhi J, et al. Molecular profiling of breast cancer Greenman C, Wedge DC, et al. The landscape of can-
cell lines defines relevant tumor models and provides cer genes and mutational processes in breast cancer.
a resource for cancer gene discovery. PLoS One. Nature. 2012;486(7403):400–4.
2009;4(7):e6146. 32. Cambon-Thomsen A, Ducournau P, Gourraud PA,
18. Sørlie T, Perou CM, Tibshirani R, Aas T, Geisler S, Pontille D. Biobanks for genomics and genomics
Johnsen H, Hastie T, et al. Gene expression patterns of for biobanks. Comp Funct Genomics. 2003;4(6):
breast carcinomas distinguish tumor subclasses with 628–34.
clinical implications. Proc Natl Acad Sei U S A. 33. Curtis C, Shah SP, Chin SF, Turashvili G, Rueda OM,
2001;98:10869–74. Dunning MJ, et al. The genomic and transcriptomic
19. Sørlie T, Tibshirani R, Parker J, et al. Repeated obser- architecture of 2,000 breast tumours reveals novel
vation of breast tumor subtypes in independent gene subgroups. Nature. 2012;486(7403):346–52.
expression data sets. Proc Natl Acad Sci U S A. 34. Koboldt DC, Fulton RS, McLellan MD, Schmidt
2003;100:8418–23. H, Kalicki-Veizer J, McMichael JF, et al, Cancer
20. Nielsen TO, Hsu FD, Jensen K, Cheang M, Karaca G, Genome Atlas Network. Comprehensive molecu-
Hu Z, et al. Immunohistochemical and clinical charac- lar portraits of human breast tumours. Nature.
terization of the basal-like subtype of invasive breast 2012;490(7418):61–70.
carcinoma. Clin Cancer Res. 2004;10(16):5367–74. 35. Bernstein BE, Meissner A, Lander ES. The mamma-
21. Abd El-Rehim DM, Pinder SE, Paish CE, Bell J, lian epigenome. Cell. 2007;128(4):669–81.
Blamey RW, Robertson JF, et al. Expression of lumi- 36. Widschwendter M, Jones PA. DNA methylation and
nal and basal cytokeratins in human breast carcinoma. breast carcinogenesis. Oncogene. 2002;21:5462–82.
J Pathol. 2004;203(2):661–71. 37. Esteller M. Cancer epigenomics: DNA methylomes
22. Carey LA, Perou CM, Livasy CA, Dressler LG, and histone-modification maps. Nat Rev Genet.
Cowan D, Conway K, et al. Race, breast cancer sub- 2007;8:286–98.
types, and survival in the Carolina Breast Cancer 38. Baylin SB, Ohm JE. Epigenetic gene silencing in can-
Study. JAMA. 2006;295(21):2492–502. cer – a mechanism for early oncogenic pathway
23. Wirapati P, Sotiriou C, Kunkel S, Farmer P, Pradervand addiction? Nat Rev Cancer. 2006;6:107–16.
S, Haibe-Kains B, et al. Meta-analysis of gene expres- 39. Cuomo A, Moretti S, Minucci S, Bonaldi T. SILAC-
sion profiles in breast cancer: toward a unified under- based proteomic analysis to dissect the “histone modi-
standing of breast cancer subtyping and prognosis fication signature” of human breast cancer cells.
signatures. Breast Cancer Res. 2008;10(4):R65. Amino Acids. 2011;41(2):387–99.
1 Omics Technologies Applied in Breast Cancer Research 15

40. Herman JG, Graff JR, Myöhänen S, Nelkin BD, discovery of genistein action in the rat mammary
Baylin SB. Methylation-specific PCR: a novel PCR gland. J Nutr. 2005;135(12 Suppl):2953S–9.
assay for methylation status of CpG islands. Proc Natl 53. Hu Y, Zhang S, Yu J, Liu J, Zheng S. SELDI-TOF-MS:
Acad Sci U S A. 1996;93(18):9821–6. the proteomics and bioinformatics approaches in the
41. Dejeux E, Rønneberg JA, Solvang H, Bukholm I, diagnosis of breast cancer. Breast. 2005;14(4):250–5.
Geisler S, Aas T, et al. DNA methylation profiling in 54. Gharbi S, Gaffney P, Yang A, Zvelebil MJ, Cramer R,
doxorubicin treated primary locally advanced breast Waterfield MD, et al. Evaluation of two-dimensional
tumours identifies novel genes associated with sur- differential gel electrophoresis for proteomic expres-
vival and treatment response. Mol Cancer. 2010;9:68. sion analysis of a model breast cancer cell system.
42. Hsu NC, Huang YF, Yokoyama KK, Chu PY, Chen Mol Cell Proteomics. 2002;1(2):91–8.
FM, Hou MF. Methylation of BRCA1 promoter 55. Marouga R, David S, Hawkins E. The develop-
region is associated with unfavorable prognosis in ment of the DIGE system: 2D fluorescence differ-
women with early-stage breast cancer. PLoS One. ence gel analysis technology. Anal Bioanal Chem.
2013;8(2):e5625–6. 2005;382(3):669–78.
43. Lo PK, Sukumar S. Epigenomics and breast cancer. 56. Davalieva K, Kiprijanovska S, Broussard C,
Pharmacogenomics. 2008;9(12):1879–902. Petrusevska G, Efremov GD. Proteomic analysis
44. Lau TY, O’Connor DP, Brennan DJ, Duffy MJ, of infiltrating ductal carcinoma tissues by coupled
Pennington SR, Gallagher WM. Breast cancer pro- 2-D DIGE/MS/MS analysis. Mol Biol (Mosk).
teomics: clinical perspectives. Expert Opin Biol Ther. 2012;46(3):469–80.
2007;7(2):209–19. 57. Sandhu C, Connor M, Kislinger T, Slingerland J,
45. Elsheikh SE, Green AR, Rakha EA, Powe DG, Emili A. Global protein shotgun expression profiling
Ahmed RA, Collins HM, et al. Global histone modifi- of proliferating mcf-7 breast cancer cells. J Proteome
cations in breast cancer correlate with tumor pheno- Res. 2005;4(3):674–89.
types, prognostic factors, and patient outcome. Cancer 58. Yang C, Richardson AD, Smith JW, Osterman A.
Res. 2009;69(9):3802–9. Comparative metabolomics of breast cancer. Pac
46. Culhane AC, Howlin J. Molecular profiling of breast Symp Biocomput. 2007;12:181–92.
cancer: transcriptomic studies and beyond. Cell Mol 59. Ma Q, Lu AY. Pharmacogenetics, pharmacogenom-
Life Sci. 2007;64(24):3185–200. ics, and individualized medicine. Pharmacol Rev.
47. Wang Z, Gerstein M, Snyder M. RNA-Seq: a revolu- 2011;63(2):437–59.
tionary tool for transcriptomics. Nat Rev Genet. 60. Kiemer L, Cesareni G. Comparative interactomics:
2009;10(1):57–63. comparing apples and pears? Trends Biotechnol.
48. Huber-Keener KJ, Liu X, Wang Z, Wang Y, Freeman 2007;25(10):448–54.
W, Wu S, et al. Differential gene expression in 61. Rak J. Extracellular vesicles – biomarkers and effec-
tamoxifen-resistant breast cancer cells revealed by a tors of the cellular interactome in cancer. Front
new analytical model of RNA-Seq data. PLoS One. Pharmacol. 2013;4:21.
2012;7(7):e41333. 62. Lo SH. Reverse interactomics: from peptides
49. Anderson NL, Anderson NG. Proteome and pro- to proteins and to functions. ACS Chem Biol.
teomics: new technologies, new concepts, and new 2007;2(2):93–5.
words. Electrophoresis. 1998;19(11):1853–61. 63. Gonzalez-Angulo AM, Hennessy BT, Mills
50. Goncalves A, Bertucci F. Clinical application of pro- GB. Future of personalized medicine in oncology:
teomics in breast cancer: state of the art and perspec- a systems biology approach. J Clin Oncol. 2010;
tives. Med Princ Pract. 2011;20(1):4–18. 28(16):2777–83.
51. Stein RC, Zvelebil MJ. The application of 2D 64. Ocaña A, Pandiella A. Personalized therapies in the
gel-based proteomics methods to the study of cancer “omics” era. Mol Cancer. 2010;9:202.
breast cancer. J Mammary Gland Biol Neoplasia. 65. Hu Z, Fan C, Oh DS, Marron JS, He X, Qaqish BF,
2002;7(4):385–93. et al. The molecular portraits of breast tumors are
52. Rowell C, Carpenter DM, Lamartiniere CA. conserved across microarray platforms. BMC
Chemoprevention of breast cancer, proteomic Genomics. 2006;7:96.
Omics of Hereditary Breast Cancer
Catherine A. Moroski-Erkul, Burak Yilmaz,
Esra Gunduz, and Mehmet Gunduz

Breast cancer is the leading cause of cancer-related deaths among women
worldwide. Although advances in our understanding of this disease have
been made in the last decade, the available treatments remain inadequate,
particularly for the more intractable forms of breast cancer. Hereditary or
familial breast cancer poses a particularly difficult challenge as only a few
susceptibility genes with high penetrance have been identified, namely,
BRCA1 and BRCA2. It is now suspected that the majority of hereditary
and familial breast cancers are caused by various combinations of several
moderate- and/or low-penetrance genes. Recent developments in research
methodologies and conceptual frameworks within biology have revolution-
ized the study of cancer. This systems approach, which emphasizes a holis-
tic understanding of biological systems, is referred to generally as “omics.”
A decade of omics research has led to the identification of many new thera-
peutic targets and biomarkers, allowing for more accurate and earlier diag-
nosis and treatment of the wide spectrum of diseases that are collectively
referred to as breast cancer. Here we review the contributions of several
omics fields to our understanding of hereditary and familial breast cancer,
namely, genomics, transcriptomics, proteomics, and metabolomics.

Hereditary breast cancer • Genomics • Transcriptomics • Proteomics •


C.A. Moroski-Erkul, BA • B. Yilmaz, BA With the sequencing of the human genome,

E. Gunduz, PhD • M. Gunduz, MD, PhD (*) the study of biological systems underwent a
Department of Medical Genetics,
Turgut Özal Üniversity Medical School,
major transformation. Many researchers began
Ankara, Turkey to approach their work with a more global
e-mail: mehmet.gunduz@gmail.com perspective. New fields of study have developed,

D. Barh (ed.), Omics Approaches in Breast Cancer: Towards Next-Generation Diagnosis, 17

Prognosis and Therapy, DOI 10.1007/978-81-322-0843-3_2, © Springer India 2014
18 C.A. Moroski-Erkul et al.

Table 2.1 Publicly available Web-based databases con- ments in disease prevention (genome sequencing)
taining microarray datasets
detection (biomarkers), better and more individu-
Database Curator Publication alized treatments for patients (pharmacogenomic
Gene Expression National Center for [111] profiling), as well as a more thorough and accu-
Omnibus (GEO) Biotechnology
rate picture of disease prognosis (biomolecular
Information (NCBI)
Riken Riken [112]
Expression Although we saw in the first decade of the
Array Database twenty-first century an explosion in new methods
(READ) that allow for more detailed and comprehensive
ArrayTrack U.S. Food and Drug [113] exploration of biological systems, improvements
Administration (FDA)
in both detection and analysis of omics data
ArrayExpress European Molecular [114]
Biology Laboratory – are needed [1, 3, 4]. Ultimately, in the context
European of medicine, the goal of the omics revolution is
Bioinformatics Institute for a better understanding of pathophysiological
(EMBL-EBI) processes and better prevention, detection, and
BioGPS Genomics Institute of [115]
treatment/management of disease. This chapter
the Novartis Research
Institute aims to describe some of the main omics meth-
Microarray Leiden University [116] ods currently utilized in cancer research and how
Retriever Medical Center they have contributed to our current understand-
(MaRe) (LUMC) ing of hereditary breast cancer.

consisting of high-throughput data-rich meth- A Growing Problem

ods aimed at understanding the intricacies of the
biological systems around and within us. The Cancer places a heavy economic burden on
development of automated high-throughput tech- health-care systems, making the need for early
nologies able to carry out complicated experi- detection and more effective treatments not only
mental procedures and acquire detailed imaging a medical imperative but also an economic one as
and other data in a fraction of the time compared well [5–7]. Current treatment regimens, while
to previous labor-intensive methods has led to the improved and often more targeted, are still harm-
generation of an almost unmanageable amount of ful to healthy cells and tissues. This harm to
data (see Table 2.1 for a list of publically avail- healthy cells is responsible for unpleasant side
able microarray dataset databases). Along with effects and carries the possibility of causing sec-
the development of these new technologies in ondary cancers [8–10].
the lab have come advancements in computer Cancer is the leading cause of death in high-
science that allow researchers with little or no income nations and the second leading cause of
background in computer software engineering to death in nations of low to moderate income. For
sift through these mountains of data in the hopes women, breast cancer is the leading cause of
of mining relevant patterns [1]. These new and cancer-related death. In 2008, breast cancer
exciting fields of study have come to be collec- accounted for 23 % of newly diagnosed cancers
tively referred to as “omics” [2]. and 14 % of cancer-related deaths in women.
The power of the various omics fields lies in Fifty percent of breast cancers are diagnosed in
the vast and detailed information that can now be economically developing countries, and 60 % of
extracted relatively quickly and easily from a bio- breast cancer-related deaths worldwide occur in
logical sample. The integration of data from sev- these nations, suggesting an even more urgent
eral areas of omics (e.g., genomics, proteomics, need for better early detection and targeted, cost-
and metabolomics) can offer a more informative, effective treatments [11, 12]. The impact of
holistic view of the system under investigation, as breast cancer on a patient and their family is both
shown in Fig. 2.1. This should lead to advance- physically and emotionally devastating, and in
2 Omics of Hereditary Breast Cancer 19

Fig. 2.1 Visual representation of the

synthesis of high-throughput data for
the identification of more robust pre-
dictive biomarkers

Combine and filter multiplatform microarray data from tumor cell population

Strong candidate biomarkers

some nations, like the United States, which lack a deal of confusion and frustration among laypeo-
comprehensive social health care system, it can ple who may not be aware of the immense hetero-
also be economically crippling for a family. geneity both between and within different cancer
While great strides forward have been made in types. Owing to breakthroughs in understanding
early detection and identification of new treat- from the omics world, we have come to even bet-
ments, there still remains much work to be done. ter appreciate this aspect of cancer. Breast cancer
A concentrated effort is required to improve our is a particularly good example of why this term,
understanding of the genetic, biochemical, and while useful, is at the same time woefully inade-
environmental factors that contribute to the quate [13]. Breast cancer is often broadly catego-
development of breast cancer. rized as either hereditary or familial and sporadic.
Tumors are also classified into subtypes based on
various histological, genetic, and biomolecular
Heterogeneity characteristics. What has become increasingly
clear in the past decade is that each tumor, while
One of the major challenges in understanding any similar to others in many characteristics, is also
form of cancer is the heterogeneity inherent in unique. So while the search for new targets
the disease [12]. In fact, the word “cancer,” while focuses on the similarities within subtypes, we
useful as a general descriptor, has led to a great must also remain aware of the unique nature of
20 C.A. Moroski-Erkul et al.

each tumor, which may make it resistant to any Cancer Genomics

number of available therapies. The promise of the
omics revolution is that routine, inexpensive The discipline of genomics, as it is known today,
molecular profiling of individual tumors will lead started with the invention of DNA cloning in the
to truly personalized treatment modalities. 1970s and then the sequencing of the human
genome [16]. “Classical” genomics is primarily
concerned with the sequencing of genomes, the
Oncogenic Transformation identification of all genes contained within a par-
ticular genome, and understanding gene structure
Oncogenic transformation is a complex, multistep and the complex interplay between genes and
process that differs widely between and even environment. There are now many subdisciplines
within cancer types. However different each can- within this field, such as structural and functional
cer case may be, common to all are the character- genomics, epigenomics, and pharmaco- and toxi-
istics of oncogene activation and mutations in cogenomics. All aim to better understand the
tumor suppressors and other genes involved in a relationship between genetic sequences and bio-
multitude of different signaling pathways that logical processes or outcomes.
cumulatively produce the phenotype of a cancer We are now living in the so-called “post-
cell [14]. Monitoring biological samples (e.g., genomic” age. Gene mutations that increase a
blood, serum, or urine) taken from high-risk candi- person’s risk of developing various types of can-
dates over time using global transcription, metabo- cer have been identified. In high-risk breast can-
lomic, and proteomic methods may help us to cer families, genetic screening can be carried out
understand the early changes that occur during this so that preventive measures can be taken, such as
transformation. Some recent studies have utilized lifestyle changes, beginning mammograms at an
breast cancer cell lines and/or patient-derived sam- earlier age, or prophylactic mastectomy [17–20].
ples in order to examine the global changes that
occur during the transformation to metastasis with
the aim of identifying more specific and sensitive Genomics of Hereditary
biomarkers. The hope is that early identification Breast Cancer
and treatment can prevent a cancer’s advancement
to metastasis [15]. Perhaps one day our under- Many attempts have been made to classify breast
standing of the disease along with advancements cancers into meaningful subgroups to aid in diag-
in detection will even allow us to detect oncogenic nosis, optimal treatment determination, and prog-
transformation at a stage where its progression to nosis. Breast cancer tumor classification systems
cancer can even be blocked. have evolved over time as our understanding of
Four omics disciplines and their contributions the heterogeneity of this disease has increased.
to our understanding of hereditary breast cancer Breast cancer tumors may be separated into four
will be described in this chapter: genomics, tran- main types based on clinical and therapeutic char-
scriptomics, proteomics, and metabolomics. The acteristics. The luminal group is the most numer-
order in which they are presented is meant to ous and diverse subtype and is often subclassified
represent the flow of cellular information from into luminal A and luminal B, and several genomic
genomics, the relatively fixed, molecular code of tests are available to predict outcomes to endo-
life; to transcriptomics, the first step in translat- crine therapy. The second group is the human epi-
ing this code into “usable” parts; to proteomics, dermal growth receptor 2 (HER2 or ERBB2)
representing the workhorses of cellular activity; amplified or HER-2 enriched group, which has
and finally ending with metabolomics, the down- responded very well to targeting of HER2 with
stream “end products” of the myriad cellular monoclonal antibodies. The third group is referred
processes carried out by the aforementioned to as normal breastlike. The fourth group is
molecules. referred to as triple negative (or basal like) and is
2 Omics of Hereditary Breast Cancer 21

so called because they lack estrogen receptor although primarily associated with breast cancer,
(ER), progesterone receptor (PR), and HER2 the BRCA genes are more highly associated with
expression [21]. They have higher incidence in ovarian cancer, with an overall mutation rate of
patients with germline BRCA1 mutations or who about 12 % in women diagnosed with ovarian
are of African ancestry and account for about cancer [31]. After their identification, there was a
15 % of all breast cancer [22]. great deal of excitement, with many hoping that
In 2009, Parker et al. reported subtype predic- more high-penetrance genes would be discov-
tion by 50 genes using qRT-PCR and microarray ered. However, this has not been the case and this
technology, which came to be known as the is one reason why many have great hope for
Prediction Analysis of Microarray 50 (PAM50) advancements in understanding breast cancer via
and is commonly used to predict the best treat- omics methodologies.
ment modalities for individual cases [23]. In Although the two BRCA genes function in the
2011, Ebbert et al. reported that the PAM50 sys- same DNA repair pathway, homology-directed
tem is generally accurate and that the assay is recombination repair (Fig. 2.2), the tumors that
resistant to errors in the multivariate analyses result from BRCA1 and BRCA2 mutation are
(MVAs) used for classification. However, in the remarkably different. BRCA2 tumors have char-
case of tumors that do not fit existing parameters acteristics similar to sporadic cases. BRCA1
very well, the system can lead to inaccurate con- tumors, on the other hand, are uniformly aggres-
clusions [24]. In 2012, the IMPAKT task force sive, difficult to treat, and are typically ER nega-
compared the effectiveness of the PAM50 assay tive [30, 32]. In a recent review, Roy et al. propose
with a three-gene immunohistochemical (IHC) several theories to explain why tumors arising
approach using antibodies against ER, HER2, from two genes involved in the same DNA repair
and Ki67 and found that the former was “insuffi- pathway may vary so significantly, both geneti-
ciently robust” to make systemic treatment deci- cally and clinically. It is possible that other gene
sions. They recommend instead the combined mutations or polymorphisms are co-inherited
use of ER and HER2 IHC. with BRCA1; although, they note, there is no evi-
In addition to the PAM50, there are germline dence currently available to support this. Another
genetic tests for BRCA1, BRCA2, and CYP2D6 possibility they propose is that the role of
and the Breast Cancer Index (BCI). OncotypeDX BRCA1 in transcriptional co-activation or co-
and MammaPrint assays are used in the United repression, which is not shared by BRCA2, may
States and Europe for clinical decision-making be able to modify expression of the ER bio-
[25]. Recently, more extensive and meaningful sub- marker. For this to be proven correct, the expres-
grouping has been made possible by genomic (as sion profiles of ER-negative BRCA1 and
well as other omic) profiling of large sample groups ER-negative sporadic tumors would need to be
[12, 26]. Such subtyping is essential for identifying compared to identify a common mechanism.
and applying rational treatment combinations. Finally, they suggest the possibility that BRCA1
The first genes to be associated with heredi- and BRCA2 heterozygosity induce different
tary breast cancer are also probably the best mutational spectrums, perhaps resulting from
known. These breast cancer susceptibility genes, their different roles in homologous recombina-
BRCA1 and BRCA2, are inherited in an autoso- tion (HR) repair. Analyses using array-
mal dominant fashion and have high penetrance comparative genome hybridization (aCGH) show
[27, 28]. Together, they account for about 30 % some similarities between BRCA1 and BRCA2
of familial cases of breast cancer [29]. Germline cancers, including large deletions and amplifica-
mutations in these genes result in what is called tions, but some differences are also seen [30].
hereditary breast and ovarian cancer (HBOC) In a response to this review, Simon A. Joosse
syndrome, which is associated with a lifetime puts forward an intriguing alternative to these
risk of developing breast cancer of 50–80 % and explanations. He starts by noting that all mammary
of 30–50 % for ovarian cancer [30]. Interestingly, stem cells, from which all epithelial breast cells
22 C.A. Moroski-Erkul et al.

Double strand break (DSB)


DSB recognition
RAD50 MRN complex
DNA end resection RAD51


Strand invasion & synthesis


Fig. 2.2 Relationship between various genes involved in/ BRCA2, BRCA1 breast cancer 1, early onset, BRCA2 breast
implicated in the development and/or progression of breast cancer 2, early onset, XRCC2 X-ray repair complementing
cancer. Role of several proteins implicated in breast cancer defective in Chinese hamster cells 2, RAD51 RAD51 homo-
susceptibility in homologous recombination directed repair log (S. cerevisiae), RAD52 RAD52 homolog (S. cerevisiae),
of double strand breaks (DSBs) (left); Signaling between ATM ataxia telangiectasia mutated, CHK2 checkpoint
various proteins implicated in breast cancer susceptibility kinase 2, Topoisomerase (DNA) II binding protein 1, P53
(right). Gene/protein abbreviations: RAD50 RAD50 homo- tumor protein p53, PTEN phosphatase and tensin homolog,
log (S. cerevisiae), NBS1 Nijmegen breakage syndrome pro- AKT v-akt murine thymoma viral oncogene homolog 1,
tein 1, MRE11 MRE11 meiotic recombination 11 homolog MDM2 MDM2 oncogene, E3 ubiquitin protein kinase,
(S. cerevisiae), RPA replication protein A, BARD1 BRCA1 PI3K phosphatidylinositol-4,5 bisphosphate 3-kinase, cata-
associated RING domain 1, BRIP1 BRCA1 interacting pro- lytic subunit alpha, BACH1 BTB and CNC homology 1,
tein C-terminal helicase 1, PALB2 partner and localizer of basic leucine zipper transcription factor 1

originate, begin as ER negative and that BRCA1, but genes are the tumor suppressors P53, PTEN,
not BRCA2, is required for their maturation to and STK11 [33–35]. Also implicated are genes
ER-positive cells. Thus, BRCA1 deficiency may of moderate penetrance, which include ATM,
result in an accumulation of undifferentiated CHK2, RAD51D, and RAD51B [36–39]. Other
ER-negative stem cells with oncogenic potential. moderate-penetrance genes that are part of the
Following this line of thought, BRCA1 tumors Fanconi anemia (FA) pathway are PALB2,
would originate from a common cell lineage, BRIP1, RAD51C, and XRCC2 [38, 40–43]. A
explaining why they are so homogeneous as com- further 21 low-risk alleles have been identified
pared to tumors in BRCA2 mutation carriers. It [44–47]. Together these genes, including BRCA1
could also explain why BRCA1 mutation carriers and BRCA2, account for approximately 35 % of
have a higher overall risk for developing breast can- all familial breast cancer worldwide. This leaves
cer and why they are typically diagnosed at an ear- a gaping hole in our knowledge of what causes
lier age than BRCA2 mutation carriers [32]. Further the majority of familial breast cancer; a hole that
genomic and proteomic analyses will hopefully pro- is slowly being filled with valuable omics data.
vide answers to these questions in the near future. (See Table 2.2 for a list of genes mentioned in
Apart from the BRCA genes, several oth- this chapter, including brief descriptions.)
ers have been associated with familial cases Gracia-Aznarez et al. analyzed seven BRCA1/
of breast cancer. Among the high-penetrance BRCA2-negative families, each with six to ten
2 Omics of Hereditary Breast Cancer 23

Table 2.2 Brief description of genes reviewed in this chapter: their basic function and role in various cancer types
Gene name Function Role in various cancer types
PPP2R2A (Protein Negative control of cell growth Associated with complexes that directly
phosphatase 2, and cell division dephosphorylate ATM at DSBs; loss inhibits
regulatory subunit B, homologous recombination directed DNA repair [143];
alpha) somatic deletion predicts prostate cancer [144]
MTAP Polyamine metabolism, adenine Loss is common in human cancer, addition of MTA to
(methylthioadenosine and methionine scavenger MTAP negative tumor cells increased sensitivity to 6TG
phosphorylase) and 5FU without affecting MTAP positive cells [145]
MAP2K4 (MKK4) Serine threonine protein kinase, Metastasis repressor in prostate and ovarian cancers
(mitogen-activated phosphorylates and activates [146]
protein kinase kinase 4) JNK
STK11 (LKB1) (serine/ Serine threonine protein kinase, Gastrointestinal polyposis-associated mutations [147];
threonine kinase 11) regulates cell polarity, tumor inactivating mutations present in 20 % NSCLC [148];
suppressor pancreatic cysts resembling precancerous lesions in
Lkb1 mutant allele knock-in mouse model [149]
DUSP4 (MKP2) Negatively regulates ERK, p38 Low expression activates RAS-ERK signaling in
(dual-specificity and JNK residual breast cancer cells after neoadjuvant treatment
phosphatase 4) [76]; frequently overexpressed in MSI-H colorectal
cancer [150]
RUNX1 (AML1 or Transcription factor regulating Inactivation common in many hematopoietic and solid
CBFA2) (runt-related differentiation of HSCs, ERα tumors, as expression decreases with increasing breast
transcription factor 1) antagonist cancer aggression [151]; central to miRNA circuits
involved in normal and malignant hematopoiesis [152]
AMBP (alpha-1- Found in complexes with Differentially expressed in bladder cancer [153]
microglobulin/bikunin prothrombin, albumin, and
precursor) immunoglobulin A (CD79a) in
ABAT (4-aminobutyrate Catabolizes neurotransmitter Differentially expressed in ARMS
aminotransferase) GABA into succinic
CDH1 (cadherin 1, type Suppresses re-accumulation of Germline alterations associated with various gastric
1, E-cadherin mitotic cyclins by recruiting cancer syndromes [154]; normally considered tumor
(epithelial)) them to APC for ubiquitination suppressor, but expression may cause progression of
and subsequent proteolysis some ovarian and brain cancers [155]
RB1 (retinoblastoma 1) Tumor suppressor, regulates cell Demonstrated role in various cancers [156]; possible
growth, interacts with proteins role in EMT in TNBC [157]
involved in apoptosis and
HSP90 (Hsp90 Role in folding and activating HSP90 inhibitors used in treatment refractory HER2
chaperone) proteins in various signal BC and various other cancers [158, 159]
transduction pathways
PIK3CA Gene with highest frequency of gain-of-function
(phosphatidylinositol- mutations in breast cancer [160]
3-kinase, catalytic
subunit alpha)
MLL3 (myeloid/ Histone methyltransferase, Germline mutations, i.e., by exome sequencing in colon
lymphoid or mixed- involved in circadian cancer and AML [161]; loss of expression common in
lineage leukemia 3) transcription MSI-H gastric cancers [162]
GATA3 (GATA binding Transcription factor regulating Suppresses breast cancer metastasis by inducing
protein 3) luminal epithelial cell miR-29b [163]; loss of expression in PTEN-deficient
differentiation in the mammary prostates accelerates tumor invasion [164]; may impact
gland, involved in regulation of ESR1 enhancer accessibility [165]; mutant status
T-cell development correlated with suppression of proliferation upon
aromatase inhibitor treatment [166]
24 C.A. Moroski-Erkul et al.

Table 2.2 (continued)

Gene name Function Role in various cancer types
MAP3K1 (mitogen- Serine threonine kinase Correlated with breast cancer susceptibility in BRCA2
activated protein kinase carriers [167]
kinase kinase 1, E3
ubiquitin protein ligase)
CDKN1B (P27KIP1) Controls cell cycle progression HSP90 inhibitors alter its expression in melanoma cells
(cyclin-dependent kinase at G1 by binding cyclin E-CDK2 [168]; its inhibition by Id3 may result in more aggressive
inhibitor 1B) or cyclin D-CDK4 complexes prostate cancers [169]; induction by vitamin E
δ-tocotrienol inhibits proliferation in PDCA cells [170]
TBX3 (T-box 3) Transcription repressor Overexpressed in HNSCC, represses PTEN [171];
BRAF/Tbx3/E-cadherin pathway may promote
metastasis of BRAF-mutant melanomas [172];
methylation status of promoter identified glioblastoma
patients with MGMT-methylated tumors [172]
CBFB (core-binding Beta subunit of a core-binding CBFB-MYH11 fusion protein associated with AML
factor, beta subunit) transcription factor complex [173]; decreased expression may be involved in
involved in development and malignant phenotype of some prostate and ovarian
stem-cell homeostasis cancers [174]
NF1 (neurofibromin 1) Negative regulator of several Mutations strongly associated with myeloid
signal transduction pathway malignancies [175]; loss-of-function mutations cause
involved in proliferation, neurofibromatosis 1, a tumor predisposition syndrome
including Ras pathway [176]; somatic mutations occur in OSCs [177]; allelic
loss detected in CRC [178]
SF3B1 (splicing factor Subunit of U2 snRNP complex Mutations common in B-CLL [179]; mutations at codon
3b, subunit 1, 155 kDa) and minor U12-type spliceosome 625 in uveal melanoma [180]; mutated in PDCA [181]
CCND3 (cyclin D3) Forms regulatory subunit of Molecule targeting kinase function of cyclin
CDK4 and CDK6 which are D3:CDK4/6 inhibits cell cycle entry in human T-ALL
required for G1/S cell cycle [182]; expression upregulated in AC [183]; may be
transition regulated by miR-138 which is often downregulated in
HCC [184]; overexpressed in laryngeal squamous cell
carcinoma [185]
PDZK1 (PDZ domain Scaffolding protein mediating Overexpression likely associated with drug resistance in
containing 1) localization of cell surface multiple myeloma [186]; upregulated by 17beta-
proteins and involved in estradiol in some ovarian cancer cell lines [187]
cholesterol metabolism
PTX3 (pentraxin 3, Involved in innate immunity and May be used as FGF2 antagonist in tumor cells resistant
long) extracellular matrix formation to anti-VEFG therapy [188]
5FU fluorouracil, 6TG thioguanine, AC lung adenocarcinoma, AML acute myeloid leukemia, APC anaphase-promoting
complex, ARMS alveolar pediatric rhabdomyosarcoma, B-CLL B-cell chronic lymphocytic leukemia, CRC colorectal
cancer, EMT epithelial-to-mesenchymal transition, GABA gamma-aminobutyric acid, HCC hepatocellular carcinoma,
HNSCC head and neck squamous cell carcinoma, MSI-H microsatellite instability high, MTA methylthioadenosine,
NSCLC non-small cell lung cancer, OSCs ovarian serous carcinomas, PDCA pancreatic ductal cancer, snRNP small
nuclear ribonucleoproteins, T-ALL T-cell acute lymphocytic leukemia

affected family members across generations who association with breast cancer was not clear
were diagnosed under the age of 60. A known due to insufficient statistical power. Targeted re-
moderate susceptibility indel variant, CHEK2 sequencing of these gene candidates would need
1100delC, was identified. CHEK2 (or CHK2) is to be carried out in a larger cohort to determine
a gene integral to cell cycle checkpoint regula- whether or not an association actually exists [48].
tion and is found within the same signaling path- In 2011, Rebbeck et al. published the results
way as ATM and p53 (see Fig. 2.2). Additionally, of a study in which they analyze a set of genes
11 rare variants were identified, although their known to code for BRCA1 interacting proteins in
2 Omics of Hereditary Breast Cancer 25

2,825 BRCA1 mutation carriers to try to identify DNA methylation is an epigenetic mecha-
breast cancer risk-modifying genes. The follow- nism that is thought to contribute to the control
ing genes were identified as potential modifiers: of gene expression [56]. In a recent review the
ATM, BRCC45, BRIP1, CTIP, MERIT40, NBS1, possibility of targeting epigenetic enzymes to
RAD50, and TOPBP1 [49]. ATM, BRIP1, NBS1, specific DNA sequences to attain a more thor-
and RAD50 had previously been associated with ough understanding of epigenetic effects on gene
hereditary breast cancer cases [50]. Mutation expression was discussed [57]. With recent
screenings of the MERIT40 and TOPBP1 gene major advances in genome editing, it seems that
had been previously carried out by the Winqvist epigenomic editing may be a reality in the near
group. In their work, MERIT40 was not found to future [58, 59]. Recent research linking DNA
be associated with disease in familial breast can- methylation of particular genes with breast and
cer cases. However, the sample size of the study other types of cancers suggests that the ability to
was relatively small (125 families) and geograph- edit epigenetic marks could be immensely useful
ically limited (families originating in northern both to basic and translational cancer research
Finland); thus, their results may not be relevant [60–63].
to other populations [51]. The same group had Swift-Scanlan et al. reported the use of quan-
performed a similar study in 2006 in which they titative multiplex-methylation specific PCR
examined TOPBP1 and identified several variants (QM-MSP) to examine the precise levels of
in familial breast cancer cases [52]. Two other methylation of genes known to be hypermethyl-
studies examined the possible role of TOPBP1 in ated in breast cancer [64]. In a set of 99 formalin-
modifying breast cancer risk. The first found fixed archival breast cancer tissue samples from
aberrant subcellular localization of the protein in patients with germline mutations in BRCA1 or
breast carcinoma from an unselected consecutive BRCA2 and/or a family history of breast cancer,
cohort of 61 patients [53] and the other, specifi- the authors were able to identify associations
cally examining familial breast cancer cases in between levels of DNA methylation in several
Poland, found that decreased mRNA levels and genes (APC, RASSF1A, TWIST, ERα, CDH1,
increased protein levels of TOPBP1 were associ- and cyclin D2) and tumor stage, hormone recep-
ated with disease progression [54]. tor status, growth receptor status, and history of
The power of genomic analysis is well illus- recurrent or metastatic disease. While not as high
trated in a study published by Banerji et al. in throughput as other methods, QM-MSP is very
2012 in which whole-exome sequences of 103 sensitive, allowing analysis of samples that are
breast cancers from patients in Mexico and very limited in size (50–1,000 cells) [64]. Other
Vietnam were compared to matched normal studies investigating DNA methylation in breast
DNA. They also performed whole-genome cancer have found that GSTP1 and FOXC1 pro-
sequencing for 22 breast cancer/normal pairs. moter methylation status could be used as a prog-
Results confirmed a number of previously identi- nostic marker [65].
fied somatic mutations as well as discovering Another interesting and promising method
some new mutations, including a recurrent that is increasingly being utilized is single-
MAGI3-AKT3 fusion enriched in triple-negative nucleus sequencing (SNS) from flow-sorted
breast cancers. The fusion causes constitutive nuclei. This has clearly illustrated that tumors
activation of AKT kinase. They found that treat- are composed of a number of distinct subpopu-
ment with an AKT small-molecule inhibitor was lations, each with unique genetic characteristics
able to abolish AKT activation [55]. Although but also shared genomic mutations. These vari-
this work does not specifically assess cases of ous subpopulations may then travel to different
hereditary breast cancer, it illustrates the power- parts of the body, forming genetically distinct
fully informative nature of high-throughput metastases [15].
sequencing technologies that are now at many According to a number of different studies,
researchers’ disposal. the majority of mutations present in metastases
26 C.A. Moroski-Erkul et al.

are also present in the primary tumor [66]. This is throughput, expensive, and imprecise. Microarray
potentially good news in that the transformation and gene chip technologies are currently the main
to metastasis may be more easily inhibited than tools of choice in this omics field. RNA-Seq (RNA
previously thought. Then again, cancer is a very sequencing), a relatively recent development, is a
“smart” disease with quickly evolving genetic method that uses deep sequencing technology and
characteristics allowing tumor cells, even if only has vastly improved precision in transcript mea-
a small subpopulation, to escape our attempts at surement, does not rely on known genomic
its eradication. sequences, and can identify single nucleotide poly-
morphisms (SNPs) present in transcripts [74].

Cancer Transcriptomics
Transcriptomics of Hereditary
With the development of microarray technologies Breast Cancer
and advanced bioinformatics analysis software,
the focus of many researchers turned to such The heterogeneity inherent in triple-negative
efforts as identifying patterns of differential gene breast cancer (TNBC), frequently associated
expression in cells under various conditions. In with BRCA1 germline mutations, makes it an
1999, Golub et al. showed that identification of especially intractable disease. A recent study
tumor subtypes could be carried out using global published by Cascione et al. examined miRNA
gene expression data rather than histological and and mRNA expression profiles of samples (for-
clinical observations [67, 68]. However, some malin fixed and paraffin embedded) that had been
recent reports suggest that a combination of vari- obtained from women with TNBC between 1995
ous methods is currently the most effective way to and 2005 (see Table 2.3 for a complete list of
make accurate diagnoses and prognoses [69]. the miRNAs identified). Samples were obtained
Transcriptomics is the study of all the tran- from tumor, adjacent non-tumor, and lymph node
scripts of a particular organism, including metastatic lesions from 173 patients. Due to low
mRNAs, small RNAs (microRNA and siRNA), RNA yield, a somewhat limited array analysis
and noncoding RNAs. It also includes the charac- was necessary. A human cancer-specific mRNA
terization of transcriptional structure, splicing array and the human miRNA expression profil-
patterns, and other posttranscriptional modifica- ing v1 panel were used. Two miRNA signatures
tions of genes. The characterization of differen- were linked to patient survival (miR-16, 155,
tial gene expression between different types of 125b, 374a and miR-16, 125b, 374a, 374b, 421,
cells or in the same cell type under variable con- 655 497) and miRNA/mRNA anticorrelations
ditions is an invaluable tool in cancer research. were used to identify four distinct molecular
The gene expression profile of a cancer cell is subclasses. One (subclass) group included seven
strikingly different from surrounding noncancer- mRNAs overexpressed in tumors compared to
ous cells. It can also be used to define tumor sub- normal tissue (SPP1, MMP9, MYBL2, BIRC5,
types for a variety of different cancers, including TOP2A, CDC2, and CDKN2A). The second
breast cancer [21, 70–73]. Examination of the group had 43 mRNAs downregulated in tumors
changes in gene expression between cells of pri- with the top gene ontologies being enriched in
mary and metastatic tumors adds to our under- NF-κB, PPAR, and PTEN signaling pathways.
standing of the mechanisms underlying metastatic The third group had ten deregulated mRNAs and
transformation. was enriched with gene ontologies associated
While traditional sequencing techniques have with growth factors. Finally, the fourth group is
been modified for transcriptional profiling (serial composed of 64 mRNAs with NF-κB signaling
analysis of gene expression (SAGE), cap analysis pathway as the most enriched gene ontology [22].
of gene expression (CAGE), massively parallel sig- Adjuvant and neoadjuvant treatments are
nature sequencing (MPSS)), these methods are low often coupled with primary therapeutic modalities
2 Omics of Hereditary Breast Cancer 27

Table 2.3 Deregulation of miRNAs identified by and molecularly, with the highest positive stain-
Cascione et al. (in TNBC expression signatures)
ing observed in BLBC cases. Expression profil-
miRNA Dereguation of identified miRNAs in other ing data from BLBC samples with high Ki67
types of cancer cells staining, when compared with the Molecular
miR-16 Prostate [117]
Signatures Database, indicated activation of the
Myeloma [118]
Ras-ERK pathway.
Breast [119]
To rule out KRAS mutation, which is
Colon [120]
infrequent in breast cancer, DNA sequencing
Oral [121]
Lung [122]
was performed and no mutations were found.
Liver [123] However, expression of DUSP4, a negative regu-
Brain [124] lator of the Ras-ERK pathway, was significantly
miR-155 Colon, cervix, pancreas, lung, thyroid, downregulated in these samples. Low DUSP4
lymphoma, leukemia [125] expression had previously been correlated with
Pancreas [126] shorter DFS in a cohort of 286 patients who had
miR-374a Lung [127] not received adjuvant therapy. To further verify
Breast [128] the significance of DUSP4, the authors measured
Colon [129] its expression in another cohort composed of
miR-421 Head and neck [130] samples obtained from 89 TNBC patients after
Stomach [131] neoadjuvant treatment and found a similar pattern
Liver [132]
of high Ki67 staining together with low DUSP4
Pancreas [133]
expression. Experiments conducted in BLBC cell
Prostate [134]
lines with siRNA knockdown of DUSP4 resulted
Breast [135]
miR-497 Cervix [136]
in decreased apoptosis, increased mitogen-acti-
Breast [137] vated protein kinase (MEK)-dependent prolifera-
Skin [138] tion, and an increased half-maximal inhibitory
Colon [139] concentration (IC50) of docetaxel, an antimitotic
Brain [140] drug used in the clinic.
Head and neck [141] After restoring DUSP4 expression in three
Stomach, lung [142] BLBC cell lines, phosphorylation of ERK was
inhibited and viability in two of the three cell
lines was reduced. The addition of MEK inhibi-
with the aim of improving the effectiveness of the tors was found to increase sensitivity to docetaxel
primary therapy. For tumors that do not respond in 17 BLBC cell lines. In cell lines with loss of
well to treatment, there is a good chance of dis- PTEN expression, the PI3K pathway is activated
ease recurrence and/or progression. Some theo- resulting in what is likely MEK-independent pro-
rize that this is due to the difficulty of eradicating liferation and evasion of apoptosis. Thus, the
tumor cells especially resistant to cancer therapy, authors conclude DUSP4 expression coupled
what are commonly referred to as cancer stem with PTEN status may effectively predict effi-
cells [75]. To identify genes associated with drug cacy of MEK inhibitors in patients with BLBC
resistance in TNBC (also referred to in the paper tumors [76].
as basal-like breast cancer, BLBC), transcrip- Taking advantage of the vast number of tumor
tional profiling was performed on 49 archival samples available in tissue banks, Curtis et al. car-
samples that had been surgically resected follow- ried out an analysis of copy number variation and
ing neoadjuvant treatment. To estimate long-term its effects on the transcriptome using a discov-
clinical outcome, IHC staining with Ki67 (a com- ery set of 997 fresh-frozen primary breast tumor
monly used marker of proliferation) was per- samples with accompanying clinical information.
formed on all samples. Ki67 staining highly Another set of 995 tumors was then used as a test
correlated with tumor subtype, both clinically set to verify the predictive ability of data gleaned
28 C.A. Moroski-Erkul et al.

from the discovery set. The aim was to identify of available methods for MS data analysis and
the underlying genetic mechanisms that trans- protein identification via database search, we
late into observed variance between and among refer the reader to Brusniak et al. and Eng et al.,
breast cancer subgroups. Patients were clinically respectively [80, 81]. For a basic look at how to
homogenous with most ER-positive/lymph node analyze protein-protein interaction networks and
(LN)-negative patients having not received treat- regulatory networks, we recommend Koh et al.
ment while ER-negative/LN-positive patients had and Poultney et al. [82, 83]. Finally, for integrated
received treatment. A number of putative can- analysis of omics data from multiple platforms,
cer genes were identified including PPP2R2A, the reader is referred to Chavan et al. [84].
MTAP, and MAP2K4. The patients also stratified
into a high-risk, ER-positive 11q13/14 cis-acting
subgroup and a subgroup without any copy num- Proteomics of Hereditary
ber aberrations, which corresponded to favorable Breast Cancer
prognosis, providing a new method for identify-
ing breast cancer subgroups [77]. Cohen et al. examined plasma from 76 breast can-
cer patients and were able to identify a signature
consisting of four proteins previously found to be
Cancer Proteomics associated with breast cancer tissue: fibronectin,
clusterin, gelsolin, and α-1-microglobulin/inter-
Proteomics is the study of the entire comple- α-trypsin inhibitor light chain precursor (AMBP).
ment of proteins expressed by a particular bio- The plasma levels of these proteins differed
logical system. As with genomics, subspecialties between the two tumor types such that they were
of proteomics have developed. The four major able to distinguish between infiltrating ductal and
subfields are expression proteomics, functional invasive mammary breast carcinomas [85].
proteomics, structural proteomics, and the pro- A powerful methodology for identifying puta-
teomics of posttranslational modifications [78]. tive breast cancer biomarkers was used in a recent
The aim of proteomics is not only to identify all study by Pavlou et al. First, proteomic data was
proteins in a particular system, but also to collected from the secretome (the full complement
understand the regulation of their expression, of proteins secreted by a cell) of eight different
the interactions that occur between them, and breast cancer cell lines, representing the three
their effects on cellular function. An example of major breast cancer tumor subtypes. Out of 5,200
one of the many programs available for visual- nonredundant proteins identified by MS, 23 were
izing protein-protein interactions is presented in unique to basal breast cancer cells, 4 were unique
Fig. 2.3. to HER2-neu-amplified, and another 4 were
The primary method in the proteomics tool- unique to luminal breast cancer cells. These results
box is mass spectrometry (MS), a technology that were then compared with four publicly available
measures the mass-to-charge ratio of ions in the breast cancer mRNA microarray data sets queried
gas phase. Throughout the twentieth century, MS from the National Center for Biotechnology
technologies developed, but it was not until the Information Gene Expression Omnibus (NCBI
late 1980s that its widespread use in biological GEO). In total, 24 out of the 30 candidate proteins
research became feasible. This was made possi- had microarray expression patterns similar to
ble by the development of electron spray ioniza- those identified in the proteomic approach.
tion (ESI) and matrix-assisted laser desorption/ They next tested the clinical applicability of
ionization (MALDI) [78, 79]. this data by performing MS on cytosol collected
A recent review outlines the main challenges from eight ER-positive and eight ER-negative
facing the field of proteomics. According to the breast cancer tissue samples. Eighteen out of the
authors, the primary bottleneck in proteomics 30 subtype-specific proteins were identified and
development is in data analysis [3]. For a review three proteins in particular (ABAT, PDZK1, and
2 Omics of Hereditary Breast Cancer 29






Fig. 2.3 Protein interaction networks for BRCA1 (left) The thickness of lines represents strength of association,
and BRCA2 (right) generated using Search Tool for the with thicker lines indicating stronger associations
Retrieval of Interacting Proteins/Genes (STRING) 9.05.

PTX3) had significantly different expression in physical examination, biopsy, and imaging)
the different subtypes. Finally, they examined the alongside adjacent normal tissue collected
2-year and 5-year disease-free survival (DFS) from patients with infiltrating ductal carcinoma
data that accompanied the four gene expression (IDC). Using two-dimensional polyacrylamide
array data sets. ABAT was the most robust candi- gel electrophoresis (2D-PAGE) and high-per-
date of the three potential biomarkers. Expression formance liquid chromatography and tandem
levels of ABAT were, on average, 2.3 times mass spectrometry (LC-MS/MS), they found
higher for patients with DFS of more than 2 years. a number of proteins upregulated specifically
ABAT expression remained significantly differ- in metastatic tissue and also identified possible
ent in all four datasets at the 5-year DFS mark. markers to distinguish between the various
They also queried the Gene Expression-Based metastatic stages. Calreticulin was significantly
Outcome for Breast Cancer Online Database and upregulated in metastases of all three stages with
found that patients with higher ABAT expression a rate of 77 % in stage N0, 92 % in stage N1, and
had slightly longer disease-free survival than 83 % in stage N2. Tropomyosin alpha-3 chain
those with low expression. Patients with was also upregulated in all three stages, albeit
ER-positive disease and high ABAT expression as at lower overall incidence (N0 and N1 69 %, N2
well as tamoxifen-treated patients with high 75 %). They suggest that HSP70 is a possible
ABAT expression had better prognosis than those marker for stage N0 metastases, 80 k protein H
with low expression [86]. This work demonstrates precursor and PDI may serve as biomarkers for
that in vitro proteomic analysis of breast cancer N1 stage metastases, and immunoglobulin heavy
cell lines combined with publicly available tran- chain binding protein (BIP) is a potential identi-
scriptomic data from patients can be used to suc- fier of stage N2 metastases [87].
cessfully identify new candidate biomarkers that Another study aimed at stratifying tumors into
are breast cancer subtype specific. subgroups based on protein expression profiles
Lee et al. carried out protein expression found increased expression of STAT1 and CD74
profiling of 38 sample pairs from lymph node to be associated with metastatic potential in
metastases of varying grades (classified accord- TNBC, both in patient samples and in
ing to the TNM staging system, which includes MDA-MB-231 cells. The authors suggest that the
30 C.A. Moroski-Erkul et al.

mechanism by which this increased capability major types of macromolecules (carbohydrates,

occurs is likely the CD74/CD44/ERK, MIF lipids, proteins, and nucleic acids) [14, 91, 92].
receptor pathway with a positive feedback loop This phenomenon was first formally described
between CD74 and STAT1 [88]. by Otto Warburg in the 1920s and refers specifi-
cally to a cancer cell’s “preference” for perform-
ing glycolysis even in the presence of oxygen
Cancer Metabolomics (aerobic glycolysis) [14, 93, 94]. Known as
the Warburg effect, this characteristic of “glu-
Metabolomics is another piece of the omics puz- cose addiction” is exploited in the clinic for
zle that will improve our understanding of cancer the identification of cancerous lesions. Positron
cells and their transformation to the metastatic emission tomography (PET) is used to detect
state. Metabolomics may be defined as “the com- radioactively labeled glucose (2-deoxy-2-[18F]
prehensive analysis of the low-molecular-weight fluoro-D-glucose, FDG), which accumulates
molecules, or metabolites, that are the intermedi- more in tumor cells relative to other cells due
ates and products of metabolism” [89]. The to their heavy reliance on glycolysis [93]. This
Human Metabolome Database (www.hmdb.ca) is has helped thrust cancer metabolism back into
a publicly available collection of detailed infor- the spotlight in recent years. Evidence that acti-
mation about the 40,250 small-molecule metabo- vated oncogenes and mutant tumor suppressors
lites that have been thus far identified in human can impact metabolism has also helped feed
cells. “The large number of different metabolites, this interest. Nature and Nature Reviews Cancer
differences in their relative concentrations and published a “Web focus” on cancer metabolism
variability in their physicochemical properties where they highlight some recent developments
(polarity, hydrophobicity, molecular mass or in the field [27]. One review notes that the oxy-
chemical stability) require the application of dif- gen and nutrient-rich environment in which
ferent technologies and a huge range of experi- cancer cell lines are typically maintained and
mental conditions” [90]. The most common studied is markedly different from the in vivo
techniques used in metabolomic profiling are tumor microenvironment [91]. Cocultures of
nuclear magnetic resonance (NMR) and mass breast cancer cells with fibroblasts may more
spectroscopy (MS) [89]. accurately reflect the conditions in which tumors
While metabolomics, like the other omics grow and can add to our understanding of how
disciplines, offers great hope, it also comes cancer cells evade death during treatment.
with many challenges. The complete human An example of how metabolomics, integrated
metabolome is very large, almost twice that of with the other omics fields, can enrich our under-
the human proteome, and they exist in a con- standing of biological systems comes from a
stant dynamic flux. While collection of samples study published in 2011 in which a panel of 59
for metabolomic analysis is relatively easy and cell lines from different cancer types is used to
noninvasive (typically serum, plasma, or urine), identify links between genetic and metabolic
because the molecules of interest are so easily profiles. They examined these cell lines before
modified during the process of sample transport and after treatment with a variety of chemother-
and preparation, this presents potential variability apy agents, including a variety of platinum-
that may be very difficult to control for. For trans- containing drugs. Their approach required an
lation to use in medicine, standard protocols and integrated analysis of two very different sets of
conditions for collection, storage, and processing data. Using the common method of overrepre-
must be designed and strictly adhered to. sentation (OR) analysis to first analyze the two
One of the distinguishing characteristics of sets of data individually and then develop a
cancer cells is their unique “reprogramming” method to integrate analysis of both data sets
of metabolic pathways, in which they acquire together, they illustrate the potential power of
changes that affect the metabolism of the four inter-omics analysis [95].
2 Omics of Hereditary Breast Cancer 31

Metabolomics of Hereditary and subsequently studied in vitro in metastatic

Breast Cancer and nonmetastatic breast cancer cell lines.
Consistent with results from previous studies
A flurry of papers specifically investigating breast examining other types of cancers, they found that
cancer metabolism were published by the Lisanti metastatic breast cancer cells had a proliferative
group in 2009 and 2010. As a result of their find- capacity similar to that of cells in the primary
ings, they proposed that it is not only the cancer tumor; that is, metastatic cells did not have higher
cells that have altered metabolic pathways but proliferative activity than primary tumor cells.
that cancer-associated fibroblasts (CAFs) present There was also an apparent trade-off between
in the tumor microenvironment also have major proliferation and detoxification of reactive
alterations in metabolism [25, 96, 97]. Known as oxygen species. Production of lipids is necessary
the “reverse Warburg effect” or “stromal-epithelial for cell proliferation but hinders a cell’s ability to
metabolic coupling,” this theory posits that epithe- detoxify oxidative molecules. The authors identi-
lial cancer cells actually induce the Warburg effect fied metabolic differences between ER-positive
in stromal fibroblasts located in the tumor micro- and ER-negative tumor cells, which suggested
environment so that they produce and secrete that the latter have a lower capacity for producing
additional pyruvate and lactate that can then be lactate from glucose. The authors suggest that
utilized by cancer cells as inputs to the mitochon- their findings may apply broadly to many differ-
drial TCA cycle, oxidative phosphorylation, and ent types of cancer, but also note that their work
ATP production [98]. They proposed that CAFs, was in cell lines and thus in vivo experiments are
which are thought to be derived from mesenchy- necessary to explore the applicability of their
mal stem cells of the bone marrow, are essentially results to disease in the organismal context [103].
like TGFβ-activated fibroblasts (myofibroblasts) In 2011, Possemato et al. presented a new
that cannot be “turned off” [99]. They, and others, tool for target identification with an in vivo
have shown that CAFs exhibit a loss of caveolin-1 RNA interference-based loss-of-function screen
(Cav1), an inhibitor of TGFβ signaling. of metabolic genes associated with aggressive
In another paper, the Lisanti group shows that breast cancer and stemness in a human breast
Cav1 was dramatically downregulated in breast cancer xenograft model. Increased expression
cancer CAFs as compared with normal fibro- of phosphoglycerate dehydrogenase (PHGDH)
blasts taken from the same patients. Other studies was found to be required for increased serine
have shown that Cav1 is an effective predictor of pathway flux in some breast cancers. Inhibition
breast cancer tumor recurrence, lymph node of PHGDH in cell lines with elevated PHGDH
metastasis, tamoxifen resistance, and poor clini- resulted in decreased proliferation and a reduc-
cal outcome even independent of ER, PR, and tion in serine synthesis. While overall cellular
HER2 tumor status [100–102]. serine levels were not affected, a reduction in
Jerby et al. designed and validated an in silico α-ketogluterate, another output of the serine bio-
metabolic phenotypic analysis (MPA) to measure synthesis pathway, was observed [104].
whole metabolomic flux. These phenotypes were Another study published in 2011 found that
inferred by integration of transcriptomic and pro- basal, but not luminal, breast cancer cells are
teomic data and this technique was applied to highly glutamine dependent and that the gluta-
conduct the first genome-scale study of breast mine independence of luminal breast cancers is
cancer metabolism. This method differs from associated with cell lineage-specific expression
previous models in that it does not require an of glutamine synthetase (GS). Glutamine synthe-
optimal fit to the data. The model used in MPA tase is induced by GATA3 and glutaminase
allows the data to deviate somewhat from the expression is repressed by GATA3 [105]. This is
optimal fit so that one can estimate the cell or sys- yet another aspect of tumor biology that must be
tem’s adaptive potential. Predictions were made considered when formulating treatment plans for
based on data from nearly 400 clinical samples individual patients.
32 C.A. Moroski-Erkul et al.

For a review of breast cancer metabolism, we were also identified and include TBX3, RUNX1,
refer the reader to a piece by Davison and Schafer, CBFB, AFF2, PIK3R1, PTPRD, NF1, SF3B1,
published in 2010 [106]. A review from 2011 that and CCND3 [26].
looks specifically at the role of the PI3 kinase
pathway in breast cancer is also recommended
[107]. For a more recent review, we refer the Conclusion and Future Perspective
reader to a piece by Deblois and Giguere about
estrogen-related receptors and their role in breast The Oxford English Dictionary defines the suffix
cancer cell growth and metabolism [108]. –ome in cellular and molecular biology applica-
tions as “forming nouns with the sense ‘all of the
specified constituents of a cell, considered col-
Pharmacogenomics lectively or in total’” [110]. This definition very
neatly captures the broad vision shared by the
Perhaps equally important to an understanding of many omics fields. Each field has its own unique
the genetic and metabolic profiles of cancer cells methods and instruments designed by engineers,
is knowledge of the status of the genes involved biologists, chemists, and physicists working
in the body’s response to cancer therapeutics. In together. And each has its own set of molecules
terms of chemotherapy, the status of genes of interest. However, combined they all share the
involved in drug metabolism and transporters vision of a more complete and nuanced under-
may be used to determine the suitability of a par- standing of biological systems.
ticular treatment for a patient. Likewise, the sta- The task of sifting through the data produced
tus of genes involved in the body’s response to by high-throughput omics methodologies is a
radiation therapy (e.g., DNA repair and radiation- daunting task. However, after about a decade of
induced fibrosis) can also aid in determining the work in the various omics fields, the value of
best treatment options for each individual [109]. doing so has become evident. Here we reviewed
This will certainly be a key component in the some of the recent literature relevant to our
“personalized medicine” revolution. understanding of breast cancer as seen through
the prism of various omics fields. It is expected
that these powerful methods will continue to
The Power of Omics Integration provide a more holistic understanding of cancer
and other diseases and contribute to the quest for
The Cancer Genome Atlas Research Network truly personalized medical treatment.
published a paper in 2012 presenting results from
a comprehensive molecular analysis of tumors
and germline DNA samples from 825 breast can-
cer patients. Samples were analyzed by genomic References
DNA copy number arrays, DNA methylation,
1. Ahn NG, Wang AH. Proteomics and genomics: per-
exome sequencing, messenger RNA arrays,
spectives on drug and target discovery. Curr Opin
microRNA sequencing, and reverse-phase pro- Chem Biol. 2008;12(1):1–3. PubMed PMID:
tein arrays. Data from the six different platforms 18302945. Pubmed Central PMCID: 2386992.
were analyzed both individually and in an inte- 2. Pina AS, Hussain A, Roque AC. An historical over-
view of drug discovery. Methods Mol Biol.
grated manner, which resulted in identification of
2009;572:3–12. PubMed PMID: 20694682.
four basic breast cancer classes, each with a dis- 3. Cappadona S, Baker PR, Cutillas PR, Heck AJ, van
tinct molecular signature. Almost all genes that Breukelen B. Current challenges in software solutions
had been previously identified in breast cancer for mass spectrometry-based quantitative proteomics.
Amino Acids. 2012;43(3):1087–108. PubMed PMID:
were confirmed in this study, including PIK3CA,
22821268. Pubmed Central PMCID: 3418498.
PTEN, AKT1, TP53, GATA3, CDH1, RB1, 4. Ferrer-Alcon M, Arteta D, Guerrero MJ, Fernandez-
MLL3, MAP3K1, and CDKN1B. Novel genes Orth D, Simon L, Martinez A. The use of gene
2 Omics of Hereditary Breast Cancer 33

array technology and proteomics in the search of 19. Euhus D. Managing the breast in patients who test
new targets of diseases for therapeutics. Toxicol positive for hereditary breast cancer. Ann Surg Oncol.
Lett. 2009;186(1):45–51. PubMed PMID: 2012;19(6):1738–44. PubMed PMID: 22395981.
19022361. Epub 2012/03/08.
5. Ershler WB. Cancer: a disease of the elderly. J Support 20. Smith KL, Isaacs C. BRCA mutation testing in deter-
Oncol. 2003;1(4 Suppl 2):5–10. PubMed PMID: mining breast cancer therapy. Cancer J.
15346994. Epub 2004/09/07. 2011;17(6):492–9. PubMed PMID: 22157293.
6. Lyman GH. Economics of cancer care. J Oncol Pract. Pubmed Central PMCID: PMC3240813, Epub
2007;3(3):113–14. PubMed PMID: 20859394. 2011/12/14.
Pubmed Central PMCID: PMC2793796. Epub 21. Perou CM, Sorlie T, Eisen MB, van de Rijn M, Jeffrey
2007/05/01. SS, Rees CA, et al. Molecular portraits of human
7. Russell HV, Panchal J, Vonville H, Franzini L, Swint breast tumours. Nature. 2000;406(6797):747–52.
JM. Economic evaluation of pediatric cancer treat- PubMed PMID: 10963602. Epub 2000/08/30.
ment: a systematic literature review. Pediatrics. 22. Cascione L, Gasparini P, Lovat F, Carasi S, Pulvirenti
2013;131(1):e273–87. PubMed PMID: 23266919. A, Ferro A, et al. Integrated microRNA and mRNA
Epub 2012/12/26. signatures associated with survival in triple negative
8. Zhu G. Risk of second primary cancer after treatment breast cancer. PLoS One. 2013;8(2):e55910. PubMed
for esophageal cancer: a pooled analysis of nine can- PMID: 23405235. Pubmed Central PMCID:
cer registries. Dis Esophagus. 2012;25(6):505–11. PMC3566108. Epub 2013/02/14.
9. Duman BB, Paydas S, Disel U, Besen A, Gurkan 23. Parker JS, Mullins M, Cheang MC, Leung S, Voduc D,
E. Secondary malignancy after imatinib therapy: eight Vickery T, et al. Supervised risk predictor of breast
cases and review of the literature. Leuk Lymphoma. cancer based on intrinsic subtypes. J Clin Oncol.
2012;53(9):1706–8. PubMed PMID: 22329351. Epub 2009;27(8):1160–7. PubMed PMID: 19204204. Pubmed
2012/02/15. Central PMCID: PMC2667820. Epub 2009/02/11.
10. Karran P, Offman J, Bignami M. Human mismatch 24. Ebbert MT, Bastien RR, Boucher KM, Martin M,
repair, drug-induced DNA damage, and secondary Carrasco E, Caballero R, et al. Characterization of
cancer. Biochimie. 2003;85(11):1149–60. PubMed uncertainty in the classification of multivariate assays:
PMID: 14726020. Epub 2004/01/17. application to PAM50 centroid-based genomic predic-
11. Jemal A, Bray F, Center MM, Ferlay J, Ward E, tors for breast cancer treatment plans. J Clin
Forman D. Global cancer statistics. CA Cancer J Clin. Bioinforma. 2011;1:37. PubMed PMID: 22196354.
2011;61(2):69–90. PubMed PMID: 21296855. Pubmed Central PMCID: PMC3275466. Epub
12. Gatza ML, Lucas JE, Barry WT, Kim JW, Wang Q, 2011/12/27.
Crawford MD, et al. A pathway-based classification 25. Martinez-Outschoorn UE, Balliet R, Lin Z, Whitaker-
of human breast cancer. Proc Natl Acad Sci U S A. Menezes D, Birbe RC, Bombonati A, et al. BRCA1
2010;107(15):6994–9. PubMed PMID: 20335537. mutations drive oxidative stress and glycolysis in the
Pubmed Central PMCID: 2872436. tumor microenvironment: implications for breast can-
13. Bertos NR, Park M. Breast cancer – one term, many cer prevention with antioxidant therapies. Cell Cycle.
entities? J Clin Invest. 2011;121(10):3789–96. 2012;11(23):4402–13. PubMed PMID: 23172369.
PubMed PMID: 21965335. Pubmed Central PMCID: Epub 2012/11/23.
PMC3195465. Epub 2011/10/04. 26. Cancer Genome Atlas N. Comprehensive molecular
14. Hanahan D, Weinberg RA. Hallmarks of cancer: the portraits of human breast tumours. Nature.
next generation. Cell. 2011;144(5):646–74. PubMed 2012;490(7418):61–70. PubMed PMID: 23000897.
PMID: 21376230. Pubmed Central PMCID: 3465532.
15. Trape AP, Gonzalez-Angulo AM. Breast cancer and 27. Miki Y, Swensen J, Shattuck-Eidens D, Futreal PA,
metastasis: on the way toward individualized therapy. Harshman K, Tavtigian S, et al. A strong candidate for
Cancer Genomics Proteomics. 2012;9(5):297–310. the breast and ovarian cancer susceptibility gene
PubMed PMID: 22990109. Epub 2012/09/20. BRCA1. Science. 1994;266(5182):66–71. PubMed
16. Galas DJ, McCormack SJ. An historical perspective PMID: 7545954. Epub 1994/10/07.
on genomic technologies. Curr Issues Mol Biol. 28. Wooster R, Bignell G, Lancaster J, Swift S, Seal
2003;5(4):123–7. PubMed PMID: 12921227. Epub S, Mangion J, et al. Identification of the breast
2003/08/19. cancer susceptibility gene BRCA2. Nature.
17. Riscuta G, Dumitrescu RG. Nutrigenomics: implica- 1995;378(6559):789–92. PubMed PMID: 8524414.
tions for breast and colon cancer prevention. Methods Epub 1995/12/21.
Mol Biol. 2012;863:343–58. PubMed PMID: 29. Vuillaume ML, Uhrhammer N, Vidal V, Vidal VS,
22359305. Chabaud V, Jesson B, et al. Use of gene expression
18. Singh K, Lester J, Karlan B, Bresee C, Geva T, profiles of peripheral blood lymphocytes to distin-
Gordon O. Impact of family history on choosing risk- guish BRCA1 mutation carriers in high risk breast
reducing surgery among BRCA mutation carriers. cancer families. Cancer Inform. 2009;7:41–56.
Am J Obstet Gynecol. 2013;208(4):329.e1–6. PubMed PMID: 19352458. Pubmed Central PMCID:
PubMed PMID: 23333547. Epub 2013/01/22. PMC2664702. Epub 2009/04/09.
34 C.A. Moroski-Erkul et al.

30. Roy R, Chun J, Powell SN. BRCA1 and BRCA2: dif- anemia J gene BRIP1 are low-penetrance breast can-
ferent roles in a common pathway of genome protec- cer susceptibility alleles. Nat Genet. 2006;38(11):
tion. Nat Rev Cancer. 2012;12(1):68–78. PubMed 1239–41. PubMed PMID: 17033622. Epub
PMID: 22193408. 2006/10/13.
31. Walsh T, Casadei S, Coats KH, Swisher E, Stray SM, 42. Meindl A, Hellebrand H, Wiek C, Erven V,
Higgins J, et al. Spectrum of mutations in BRCA1, Wappenschmidt B, Niederacher D, et al. Germline
BRCA2, CHEK2, and TP53 in families at high risk of mutations in breast and ovarian cancer pedigrees
breast cancer. JAMA. 2006;295(12):1379–88. establish RAD51C as a human cancer susceptibility
PubMed PMID: 16551709. Epub 2006/03/23. gene. Nat Genet. 2010;42(5):410–14. PubMed PMID:
32. Joosse SA. BRCA1 and BRCA2: a common pathway 20400964. Epub 2010/04/20.
of genome protection but different breast cancer sub- 43. Shamseldin HE, Elfaki M, Alkuraya FS. Exome
types. Nat Rev Cancer. 2012;12(5):372; author reply. sequencing reveals a novel Fanconi group defined by
PubMed PMID: 22525577. XRCC2 mutation. J Med Genet. 2012;49(3):184–6.
33. Borresen AL, Andersen TI, Garber J, Barbier-Piraux PubMed PMID: 22232082. Epub 2012/01/11.
N, Thorlacius S, Eyfjord J, et al. Screening for germ 44. Turnbull C, Ahmed S, Morrison J, Pernet D, Renwick
line TP53 mutations in breast cancer patients. Cancer A, Maranian M, et al. Genome-wide association study
Res. 1992;52(11):3234–6. PubMed PMID: 1591732. identifies five new breast cancer susceptibility loci.
Epub 1992/06/01. Nat Genet. 2010;42(6):504–7. PubMed PMID:
34. Lynch ED, Ostermeyer EA, Lee MK, Arena JF, Ji H, 20453838. Epub 2010/05/11.
Dann J, et al. Inherited mutations in PTEN that are 45. Easton DF, Pooley KA, Dunning AM, Pharoah PD,
associated with breast cancer, cowden disease, and Thompson D, Ballinger DG, et al. Genome-wide
juvenile polyposis. Am J Hum Genet. 1997;61(6):1254– association study identifies novel breast cancer sus-
60. PubMed PMID: 9399897. Pubmed Central ceptibility loci. Nature. 2007;447(7148):1087–93.
PMCID: PMC1716102. Epub 1997/12/18. PubMed PMID: 17529967. Pubmed Central PMCID:
35. Giardiello FM, Brensinger JD, Tersmette AC, PMC2714974. Epub 2007/05/29.
Goodman SN, Petersen GM, Booker SV, et al. Very 46. Ahmed S, Thomas G, Ghoussaini M, Healey CS,
high risk of cancer in familial Peutz-Jeghers syn- Humphreys MK, Platte R, et al. Newly discovered
drome. Gastroenterology. 2000;119(6):1447–53. breast cancer susceptibility loci on 3p24 and 17q23.2.
PubMed PMID: 11113065. Epub 2000/12/13. Nat Genet. 2009;41(5):585–90.
36. Renwick A, Thompson D, Seal S, Kelly P, Chagtai T, 47. Thomas G, Jacobs KB, Kraft P, Yeager M, Wacholder
Ahmed M, et al. ATM mutations that cause S, Cox DG, et al. A multistage genome-wide associa-
ataxia-telangiectasia are breast cancer susceptibility tion study in breast cancer identifies two new risk
alleles. Nat Genet. 2006;38(8):873–5. PubMed alleles at 1p11.2 and 14q24.1 (RAD51L1). Nat Genet.
PMID: 16832357. Epub 2006/07/13. 2009;41(5):579–84.
37. Meijers-Heijboer H, Wijnen J, Vasen H, Wasielewski 48. Gracia-Aznarez FJ, Fernandez V, Pita G, Peterlongo
M, Wagner A, Hollestelle A, et al. The CHEK2 P, Dominguez O, de la Hoya M, et al. Whole exome
1100delC mutation identifies families with a heredi- sequencing suggests much of non-BRCA1/BRCA2
tary breast and colorectal cancer phenotype. Am J familial breast cancer is due to moderate and low pen-
Hum Genet. 2003;72(5):1308–14. PubMed PMID: etrance susceptibility alleles. PLoS One. 2013;
12690581. Pubmed Central PMCID: PMC1180284. 8(2):e55681. PubMed PMID: 23409019. Pubmed
Epub 2003/04/12. Central PMCID: 3568132.
38. Loveday C, Turnbull C, Ramsay E, Hughes D, Ruark 49. Rebbeck TR, Mitra N, Domchek SM, Wan F, Friebel
E, Frankum JR, et al. Germline mutations in RAD51D TM, Tran TV, et al. Modification of BRCA1-
confer susceptibility to ovarian cancer. Nat Genet. associated breast and ovarian cancer risk by BRCA1-
2011;43(9):879–82. PubMed PMID: 21822267. Epub interacting genes. Cancer Res. 2011;71(17):5792–805.
2011/08/09. PubMed PMID: 21799032. Pubmed Central PMCID:
39. Orr N, Lemnrau A, Cooke R, Fletcher O, Tomczyk 3170727.
K, Jones M, et al. Genome-wide association study 50. Walsh T, King MC. Ten genes for inherited breast
identifies a common variant in RAD51B associated cancer. Cancer Cell. 2007;11(2):103–5. PubMed
with male breast cancer risk. Nat Genet. PMID: 17292821.
2012;44(11):1182–4. PubMed PMID: 23001122. 51. Solyom S, Patterson-Fortin J, Pylkas K, Greenberg
Epub 2012/09/25. RA, Winqvist R. Mutation screening of the MERIT40
40. Rahman N, Seal S, Thompson D, Kelly P, Renwick A, gene encoding a novel BRCA1 and RAP80 interacting
Elliott A, et al. PALB2, which encodes a BRCA2- protein in breast cancer families. Breast Cancer Res
interacting protein, is a breast cancer susceptibility Treat. 2010;120(1):165–8. PubMed PMID: 19572197.
gene. Nat Genet. 2007;39(2):165–7. PubMed PMID: Pubmed Central PMCID: 2863093.
17200668. Pubmed Central PMCID: PMC2871593. 52. Karppinen SM, Erkko H, Reini K, Pospiech H,
Epub 2007/01/04. Heikkinen K, Rapakko K, et al. Identification of a
41. Seal S, Thompson D, Renwick A, Elliott A, Kelly P, common polymorphism in the TopBP1 gene associ-
Barfoot R, et al. Truncating mutations in the Fanconi ated with hereditary susceptibility to breast and
2 Omics of Hereditary Breast Cancer 35

ovarian cancer. Eur J Cancer. 2006;42(15):2647–52. tumours identifies novel genes associated with sur-
PubMed PMID: 16930991. vival and treatment response. Mol Cancer. 2010;9:68.
53. Going JJ, Nixon C, Dornan ES, Boner W, Donaldson PubMed PMID: 20338046. Pubmed Central PMCID:
MM, Morgan IM. Aberrant expression of TopBP1 in PMC2861056. Epub 2010/03/27.
breast cancer. Histopathology. 2007;50(4):418–24. 66. Conklin MW, Keely PJ. Why the stroma matters in
PubMed PMID: 17448016. breast cancer: insights into breast cancer patient out-
54. Forma E, Krzeslak A, Bernaciak M, Romanowicz- comes through the examination of stromal biomark-
Makowska H, Brys M. Expression of TopBP1 in ers. Cell Adh Migr. 2012;6(3):249–60. PubMed
hereditary breast cancer. Mol Biol Rep. PMID: 22568982. Pubmed Central PMCID:
2012;39(7):7795–804. PubMed PMID: 22544570. 3427239.
Pubmed Central PMCID: 3358587. 67. Golub TR, Slonim DK, Tamayo P, Huard C,
55. Banerji S, Cibulskis K, Rangel-Escareno C, Brown Gaasenbeek M, Mesirov JP, Coller H, Loh ML,
KK, Carter SL, Frederick AM, et al. Sequence analy- Downing JR, Caligiuri MA, Bloomfield CD, Lander
sis of mutations and translocations across breast can- ES. Molecular classification of cancer: class discov-
cer subtypes. Nature. 2012;486(7403):405–9. ery and class prediction by gene expression monitor-
PubMed PMID: 22722202. ing. Science. 1999;286:531–7.
56. Varley KE, Gertz J, Bowling KM, Parker SL, Reddy 68. van’t Veer LJ, Dai H, van de Vijver MJ, He YD, Hart
TE, Pauli-Behn F, et al. Dynamic DNA methylation AA, Mao M, et al. Gene expression profiling predicts
across diverse human cell lines and tissues. Genome clinical outcome of breast cancer. Nature.
Res. 2013;23(3):555–67. PubMed PMID: 23325432. 2002;415(6871):530–6. PubMed PMID: 11823860.
Epub 2013/01/18. Epub 2002/02/02.
57. de Groote ML, Verschure PJ, Rots MG. Epigenetic 69. Rakha EA, Ellis IO. Modern classification of breast
Editing: targeted rewriting of epigenetic marks to cancer: should we stick with morphology or convert
modulate expression of selected target genes. Nucleic to molecular profile characteristics. Adv Anat Pathol.
Acids Res. 2012;40(21):10596–613. PubMed PMID: 2011;18(4):255–67. PubMed PMID: 21654357. Epub
23002135. Pubmed Central PMCID: PMC3510492. 2011/06/10.
Epub 2012/09/25. 70. Cancer Genome Atlas Network. Comprehensive
58. Cong L, Ran FA, Cox D, Lin S, Barretto R, Habib N, molecular portraits of human breast tumours. Nature.
et al. Multiplex genome engineering using CRISPR/ 2012;490(7418):61–70. PubMed PMID: 23000897.
Cas systems. Science. 2013;339(6121):819–23. Pubmed Central PMCID: PMC3465532. Epub
PubMed PMID: 23287718. 2012/09/25.
59. Mali P, Yang L, Esvelt KM, Aach J, Guell M, 71. Chen HY, Yu SL, Li KC, Yang PC. Biomarkers and
DiCarlo JE, et al. RNA-guided human genome transcriptome profiling of lung cancer. Respirology.
engineering via Cas9. Science. 2013;339(6121): 2012;17(4):620–6. PubMed PMID: 22372638. Epub
823–6. PubMed PMID: 23287722. 2012/03/01.
60. Hammoud SS, Cairns BR, Jones DA. Epigenetic reg- 72. Langer CJ. Individualized therapy for patients with
ulation of colon cancer and intestinal stem cells. Curr non-small cell lung cancer: emerging trends and chal-
Opin Cell Biol. 2013;25(2):177–83. PubMed PMID: lenges. Crit Rev Oncol Hematol. 2012;83(1):130–44.
23402869. Epub 2013/02/14. PubMed PMID: 22280915. Epub 2012/01/28. Eng.
61. Asuthkar S, Velpula KK, Chetty C, Gorantla B, Rao 73. Vivekanandan P, Singh OV. High-dimensional biol-
JS. Epigenetic regulation of miRNA-211 by MMP-9 ogy to comprehend hepatocellular carcinoma. Expert
governs glioma cell apoptosis, chemosensitivity and Rev Proteomics. 2008;5(1):45–60. PubMed PMID:
radiosensitivity. Oncotarget. 2012;3(11):1439–54. 18282123. Epub 2008/02/20.
PubMed PMID: 23183822. Epub 2012/11/28. 74. Wang Z, Gerstein M, Snyder M. RNA-Seq: a revolu-
62. Lu F, Zhang HT. DNA methylation and nonsmall cell tionary tool for transcriptomics. Nat Rev Genet.
lung cancer. Anat Rec (Hoboken). 2011;294(11):1787– 2009;10(1):57–63. PubMed PMID: 19015660. Pubmed
95. PubMed PMID: 21956844. Epub 2011/10/01. Central PMCID: PMC2949280. Epub 2008/11/19.
63. Baylin SB. The cancer epigenome: its origins, contri- 75. Reya T, Morrison SJ, Clarke MF, Weissman IL. Stem
butions to tumorigenesis, and translational implica- cells, cancer, and cancer stem cells. Nature.
tions. Proc Am Thorac Soc. 2012;9(2):64–5. PubMed 2001;414(6859):105–11. PubMed PMID: 11689955.
PMID: 22550245. Pubmed Central PMCID: Epub 2001/11/02.
PMC3359110. Epub 2012/05/03. 76. Balko JM, Cook RS, Vaught DB, Kuba MG, Miller
64. Swift-Scanlan T, Vang R, Blackford A, Fackler MJ, TW, Bhola NE, et al. Profiling of residual breast can-
Sukumar S. Methylated genes in breast cancer: asso- cers after neoadjuvant chemotherapy identifies
ciations with clinical and histopathological features in DUSP4 deficiency as a mechanism of drug resistance.
a familial breast cancer cohort. Cancer Biol Ther. Nat Med. 2012;18(7):1052–9. PubMed PMID:
2011;11(10):853–65. 22683778. Epub 2012/06/12.
65. Dejeux E, Ronneberg JA, Solvang H, Bukholm I, 77. Curtis C, Shah SP, Chin SF, Turashvili G, Rueda OM,
Geisler S, Aas T, et al. DNA methylation profiling in Dunning MJ, et al. The genomic and transcriptomic
doxorubicin treated primary locally advanced breast architecture of 2,000 breast tumours reveals novel
36 C.A. Moroski-Erkul et al.

subgroups. Nature. 2012;486(7403):346–52. PubMed 89. Van QN, Veenstra TD. How close is the bench to the
PMID: 22522925. Pubmed Central PMCID: bedside? Metabolic profiling in cancer research.
PMC3440846. Epub 2012/04/24. Genome Med. 2009;1(1):5. PubMed PMID:
78. Kislinger T. Cancer informatics in the post genomic 19348692. Pubmed Central PMCID: 2651582.
era. In: Rosen ST, series editor. Toward information- 90. Cascante M, Benito A, Zanuy M, Vizan P, Marin S, de
based medicine. Springer, New York; 2007. Atauri P. Metabolic network adaptations in cancer as
79. Hommerson P, Khan AM, de Jong GJ, Somsen targets for novel therapies. Biochem Soc Trans.
GW. Ionization techniques in capillary electrophore- 2010;38(5):1302–6. PubMed PMID: 20863303.
sis-mass spectrometry: principles, design, and appli- 91. Cairns RA, Harris IS, Mak TW. Regulation of cancer
cation. Mass Spectrom Rev. 2011;30(6):1096–120. cell metabolism. Nat Rev Cancer. 2011;11(2):85–95.
PubMed PMID: 21462232. Epub 2011/04/05. PubMed PMID: 21258394.
80. Brusniak MY, Chu CS, Kusebauch U, Sartain MJ, 92. Frezza C, Gottlieb E. Mitochondria in cancer: not just
Watts JD, Moritz RL. An assessment of current bioin- innocent bystanders. Semin Cancer Biol.
formatic solutions for analyzing LC-MS data acquired 2009;19(1):4–11. PubMed PMID: 19101633.
by selected reaction monitoring technology. 93. Bensinger SJ, Christofk HR. New aspects of the
Proteomics. 2012;12(8):1176–84. PubMed PMID: Warburg effect in cancer cell biology. Semin Cell
22577019. Epub 2012/05/12. Dev Biol. 2012;23(4):352–61. PubMed PMID:
81. Eng JK, Searle BC, Clauser KR, Tabb DL. A face in 22406683.
the crowd: recognizing peptides through database 94. Koppenol WH, Bounds PL, Dang CV. Otto Warburg’s
search. Mol Cell Proteomics. 2011;10(11): contributions to current concepts of cancer metabo-
R111.009522. PubMed PMID: 21876205. Pubmed lism. Nat Rev Cancer. 2011;11(5):325–37. PubMed
Central PMCID: PMC3226415. Epub 2011/08/31. PMID: 21508971. Epub 2011/04/22.
82. Koh GC, Porras P, Aranda B, Hermjakob H, Orchard 95. Cavill R, Kamburov A, Ellis JK, Athersuch TJ,
SE. Analyzing protein-protein interaction networks. J Blagrove MS, Herwig R, et al. Consensus-phenotype
Proteome Res. 2012;11(4):2014–31. PubMed PMID: integration of transcriptomic and metabolomic data
22385417. Epub 2012/03/06. implies a role for metabolism in the chemosensitivity
83. Poultney CS, Greenfield A, Bonneau R. Integrated of tumour cells. PLoS Comput Biol. 2011;7(3):
inference and analysis of regulatory networks from e1001113. PubMed PMID: 21483477. Pubmed
multi-level measurements. Methods Cell Biol. Central PMCID: 3068923.
2012;110:19–56. PubMed PMID: 22482944. Epub 96. Brauer HA, Makowski L, Hoadley KA, Casbas-
2012/04/10. Hernandez P, Lang LJ, Roman-Perez E, et al. Impact
84. Chavan SS, Shaughnessy Jr JD, Edmondson of tumor microenvironment and epithelial phenotypes
RD. Overview of biological database mapping on metabolism in breast cancer. Clin Cancer Res.
services for interoperation between different ‘omics’ 2013;19(3):571–85. PubMed PMID: 23236214. Epub
datasets. Hum Genomics. 2011;5(6):703–8. PubMed 2012/12/14.
PMID: 22155608. Pubmed Central PMCID: 97. Chaudhri VK, Salzler GG, Dick SA, Buckman MS,
PMC3525252. Epub 2011/12/14. Sordella R, Karoly ED, et al. Metabolic alterations
85. Cohen A, Wang E, Chisholm KA, Kostyleva R, in lung cancer-associated fibroblasts correlated with
O’Connor-McCourt M, Pinto DM. A mass increased glycolytic metabolism of the tumor. Mol
spectrometry-based plasma protein panel targeting Cancer Res. 2013;11(6):579–92. PubMed PMID:
the tumor microenvironment in patients with breast 23475953. Epub 2013/03/12. Eng.
cancer. J Proteomics. 2012;81:135–47. PubMed 98. Pavlides S, Tsirigos A, Vera I, Flomenberg N, Frank
PMID: 23174118. PG, Casimiro MC, et al. Loss of stromal caveolin-1
86. Pavlou MP, Dimitromanolakis A, Diamandis leads to oxidative stress, mimics hypoxia and drives
EP. Coupling proteomics and transcriptomics in the inflammation in the tumor microenvironment, con-
quest of subtype-specific proteins in breast cancer. ferring the “reverse Warburg effect”: a transcriptional
Proteomics. 2013;13(7):1083–95. PubMed PMID: informatics analysis with validation. Cell Cycle.
23386393. Epub 2013/02/07. 2010;9(11):2201–19. PubMed PMID: 20519932.
87. Lee HH, Lim CA, Cheong YT, Singh M, Gam Epub 2010/06/04.
LH. Comparison of protein expression profiles of dif- 99. Pavlides S, Whitaker-Menezes D, Castello-Cros R,
ferent stages of lymph nodes metastasis in breast can- Flomenberg N, Witkiewicz AK, Frank PG, et al. The
cer. Int J Biol Sci. 2012;8(3):353–62. PubMed PMID: reverse Warburg effect: aerobic glycolysis in cancer
22393307. Pubmed Central PMCID: PMC3291852. associated fibroblasts and the tumor stroma. Cell
Epub 2012/03/07. Cycle. 2009;8(23):3984–4001. PubMed PMID:
88. Greenwood C, Metodieva G, Al-Janabi K, Lausen B, 19923890. Epub 2009/11/20.
Alldridge L, Leng L, et al. Stat1 and CD74 overexpres- 100. Witkiewicz AK, Dasgupta A, Sotgia F, Mercier I,
sion is co-dependent and linked to increased invasion Pestell RG, Sabel M, et al. An absence of stromal
and lymph node metastasis in triple-negative breast caveolin-1 expression predicts early tumor recur-
cancer. J Proteomics. 2012;75(10):3031–40. PubMed rence and poor clinical outcome in human breast
PMID: 22178447. Epub 2011/12/20. cancers. Am J Pathol. 2009;174(6):2023–34.
2 Omics of Hereditary Breast Cancer 37

PubMed PMID: 19411448. Pubmed Central 113. Tong W, Harris S, Cao X, Fang H, Shi L, Sun H,
PMCID: PMC2684168. Epub 2009/05/05. et al. Development of public toxicogenomics soft-
101. Sloan EK, Ciocca DR, Pouliot N, Natoli A, Restall ware for microarray data management and analysis.
C, Henderson MA, et al. Stromal cell expression of Mutat Res. 2004;549(1–2):241–53. PubMed PMID:
caveolin-1 predicts outcome in breast cancer. Am J 15120974. Epub 2004/05/04.
Pathol. 2009;174(6):2035–43. PubMed PMID: 114. Parkinson H, Sarkans U, Shojatalab M,
19411449. Pubmed Central PMCID: PMC2684169. Abeygunawardena N, Contrino S, Coulson R, et al.
Epub 2009/05/05. ArrayExpress – a public repository for microarray
102. Witkiewicz AK, Dasgupta A, Nguyen KH, Liu C, gene expression data at the EBI. Nucleic Acids Res.
Kovatich AJ, Schwartz GF, et al. Stromal caveo- 2005;33(Database issue):D553–5. PubMed PMID:
lin-1 levels predict early DCIS progression to 15608260. Pubmed Central PMCID: PMC540010.
invasive breast cancer. Cancer Biol Ther. Epub 2004/12/21.
2009;8(11):1071–9. PubMed PMID: 19502809. 115. Wu C, Orozco C, Boyer J, Leglise M, Goodale J,
Epub 2009/06/09. Batalov S, et al. BioGPS: an extensible and custom-
103. Jerby L, Wolf L, Denkert C, Stein GY, Hilvo M, izable portal for querying and organizing gene anno-
Oresic M, et al. Metabolic associations of reduced tation resources. Genome Biol. 2009;10(11):R130.
proliferation and oxidative stress in advanced breast PubMed PMID: 19919682. Pubmed Central
cancer. Cancer Res. 2012;72(22):5712–20. PubMed PMCID: PMC3091323. Epub 2009/11/19.
PMID: 22986741. Epub 2012/09/19. 116. Ivliev AE, t Hoen PA PA, Villerius MP, den Dunnen
104. Possemato R, Marks KM, Shaul YD, Pacold ME, JT, Brandt BW. Microarray retriever: a web-based
Kim D, Birsoy K, et al. Functional genomics reveal tool for searching and large scale retrieval of public
that the serine synthesis pathway is essential in microarray data. Nucleic Acids Res. 2008;36(Web
breast cancer. Nature. 2011;476(7360):346–50. Server issue):W327–31. PubMed PMID: 18463138.
PubMed PMID: 21760589. Pubmed Central Pubmed Central PMCID: PMC2447788. Epub
PMCID: PMC3353325. Epub 2011/07/16. 2008/05/09.
105. Kung HN, Marks JR, Chi JT. Glutamine synthetase 117. Hellwinkel OJ, Sellier C, Sylvester YM, Brase JC,
is a genetic determinant of cell type-specific gluta- Isbarn H, Erbersdobler A, et al. A cancer-indicative
mine independence in breast epithelia. PLoS Genet. microRNA pattern in normal prostate tissue. Int J
2011;7(8):e1002229. PubMed PMID: 21852960. Mol Sci. 2013;14(3):5239–49. PubMed PMID:
Pubmed Central PMCID: PMC3154963. Epub 23459235. Epub 2013/03/06.
2011/08/20. 118. Yyusnita N, Zakiah I, Chang KM, Purushotaman
106. Davison CA, Schafer ZT. Keeping a breast of recent VS, Zubaidah Z Z, et al. MicroRNA (miRNA)
developments in cancer metabolism. Curr Drug expression profiling of peripheral blood samples in
Targets. 2010;11(9):1112–20. PubMed PMID: multiple myeloma patients using microarray. Malays
20545609. Epub 2010/06/16. J Pathol. 2012;34(2):133–43.
107. Baselga J. Targeting the phosphoinositide-3 (PI3) 119. Zhang N, Wang X, Huo Q, Li X, Wang H, Schneider
kinase pathway in breast cancer. Oncologist. 2011;16 P, et al. The oncogene metadherin modulates the
Suppl 1:12–9. PubMed PMID: 21278436. Epub apoptotic pathway based on the tumor necrosis
2011/02/10. factor superfamily member TRAIL in breast cancer.
108. Deblois G, Giguere V. Oestrogen-related receptors J Biol Chem. 2013;288(13):9396–407. PubMed
in breast cancer: control of cellular metabolism and PMID: 23408429. Epub 2013/02/15.
beyond. Nat Rev Cancer. 2013;13(1):27–36. 120. Ma Q, Wang X, Li Z, Li B, Ma F, Peng L, et al.
PubMed PMID: 23192231. Epub 2012/11/30. MicroRNA-16 represses colorectal cancer cell
109. Ayoub N, Lucas C, Kaddoumi A. Genomics and growth in vitro by regulating the p53/survivin signal-
pharmacogenomics of breast cancer: current knowl- ing pathway. Oncol Rep. 2013;29(4):1652–8.
edge and trends. Asian Pac J Cancer Prev. 2011; PubMed PMID: 23380758. Epub 2013/02/06.
12(5):1127–40. PubMed PMID: 21875255. Epub 121. Santhi WS, Prathibha R, Charles S, Anurup KG,
2011/08/31. Reshmi G, Ramachandran S, et al. Oncogenic
110. Dictionary OE. “ome, comb. form”: Oxford microRNAs as biomarkers of oral tumorigenesis and
University press, 2013. OED online. minimal residual disease. Oral Oncol. 2013;49(6):567–
111. Edgar R, Domrachev M, Lash AE. Gene expression 75. PubMed PMID: 23380617. Epub 2013/02/06.
omnibus: NCBI gene expression and hybridization 122. Diaz-Garcia CV, Agudo-Lopez A, Perez C, Lopez-
array data repository. Nucleic Acids Res. 2002; Martin JA, Rodriguez-Peralto JL, de Castro J, et al.
30(1):207–10. PubMed PMID: 11752295. Pubmed DICER1, DROSHA and miRNAs in patients with
Central PMCID: PMC99122. Epub 2001/12/26. non-small cell lung cancer: implications for out-
112. Bono H, Kasukawa T, Hayashizaki Y, Okazaki comes and histologic classification. Carcinogenesis.
Y. READ: RIKEN expression array database. 2013;34(5):1031–8. PubMed PMID: 23349018.
Nucleic Acids Res. 2002;30(1):211–13. PubMed Epub 2013/01/26.
PMID: 11752296. Pubmed Central PMCID: 123. Zeng L, Yu J, Huang T, Jia H, Dong Q, He F, et al.
PMC99103. Epub 2001/12/26. Differential combinatorial regulatory network
38 C.A. Moroski-Erkul et al.

analysis related to venous metastasis of hepatocel- tate cancer cells. Cancer Res. 2011;71(5):1956–67.
lular carcinoma. BMC Genomics. 2012;13 Suppl PubMed PMID: 21343391. Epub 2011/02/24.
8:14. PubMed PMID: 23282077. Pubmed Central 135. Cui XY, Guo YJ, Yao HR. Analysis of microRNA in
PMCID: PMC3535701. Epub 2013/01/11. drug-resistant breast cancer cell line MCF-7/
124. Li X, Ling N, Bai Y, Dong W, Hui GZ, Liu D, et al. ADR. Nan Fang Yi Ke Da Xue Xue Bao.
MiR-16-1 plays a role in reducing migration and 2008;28(10):1813–15. PubMed PMID: 18971180.
invasion of glioma cells. Anat Rec (Hoboken). Epub 2008/10/31. Chinese.
2013;296(3):427–32. PubMed PMID: 23175429. 136. Luo M, Shen D, Zhou X, Chen X, Wang
Epub 2012/11/24. W. MicroRNA-497 is a potential prognostic marker
125. Faraoni I, Antonetti FR, Cardone J, Bonmassar E. in human cervical cancer and functions as a tumor
miR-155 gene: a typical multifunctional microRNA. suppressor by targeting the insulin-like growth fac-
Biochim Biophys Acta. 2009;1792(6):497–505. tor 1 receptor. Surgery. 2013;153(6):836–47.
PubMed PMID: 19268705. Epub 2009/03/10. PubMed PMID: 23453369. Epub 2013/03/05.
126. Teng G, Papavasiliou FN. Shhh! Silencing by 137. Shen L, Li J, Xu L, Ma J, Li H, Xiao X, et al. miR-
microRNA-155. Philos Trans R Soc Lond B Biol 497 induces apoptosis of breast cancer cells by tar-
Sci. 2009;364(1517):631–7. PubMed PMID: geting Bcl-w. Exp Ther Med. 2012;3(3):475–80.
19008191. Pubmed Central PMCID: PMC2660923. PubMed PMID: 22969914. Pubmed Central
Epub 2008/11/15. PMCID: PMC3438749. Epub 2012/09/13.
127. Vosa U, Vooder T, Kolde R, Fischer K, Valk K, 138. Poell JB, van Haastert RJ, de Gunst T, Schultz IJ,
Tonisson N, et al. Identification of miR-374a as a Gommans WM, Verheul M, et al. A functional
prognostic marker for survival in patients with early- screen identifies specific microRNAs capable of
stage nonsmall cell lung cancer. Genes Chromosomes inhibiting human melanoma cell viability. PLoS
Cancer. 2011;50(10):812–22. PubMed PMID: One. 2012;7(8):e43569. PubMed PMID: 22927992.
21748820. Epub 2011/07/13. Pubmed Central PMCID: PMC3425484. Epub
128. Cai J, Guan H, Fang L, Yang Y, Zhu X, Yuan J, et al. 2012/08/29.
MicroRNA-374a activates Wnt/beta-catenin signal- 139. Guo ST, Jiang CC, Wang GP, Li YP, Wang CY, Guo
ing to promote breast cancer metastasis. J Clin XY, et al. MicroRNA-497 targets insulin-like
Invest. 2013;123(2):566–79. PubMed PMID: growth factor 1 receptor and has a tumour suppres-
23321667. Pubmed Central PMCID: PMC3561816. sive role in human colorectal cancer. Oncogene.
Epub 2013/01/17. 2012;32(15):1910–20. PubMed PMID: 22710713.
129. Wang YX, Zhang XY, Zhang BF, Yang CQ, Chen Epub 2012/06/20.
XM, Gao HJ. Initial study of microRNA expression 140. Yang C, Wang C, Chen X, Chen S, Zhang Y, Zhi F,
profiles of colonic cancer without lymph node et al. Identification of seven serum microRNAs from
metastasis. J Dig Dis. 2010;11(1):50–4. PubMed a genome-wide serum microRNA expression profile
PMID: 20132431. Epub 2010/02/06. as potential noninvasive biomarkers for malignant
130. Mansour WY, Bogdanova NV, Kasten-Pisula U, astrocytomas. Int J Cancer. 2013;132(1):116–27.
Rieckmann T, Kocher S, Borgmann K, et al. Aberrant PubMed PMID: 22674182. Epub 2012/06/08.
overexpression of miR-421 downregulates ATM and 141. Lajer CB, Garnaes E, Friis-Hansen L, Norrild B,
leads to a pronounced DSB repair defect and clinical Therkildsen MH, Glud M, et al. The role of miR-
hypersensitivity in SKX squamous cell carcinoma. NAs in human papilloma virus (HPV)-associated
Radiother Oncol. 2013;106(1):147–54. PubMed cancers: bridging between HPV-related head and
PMID: 23199656. Epub 2012/12/04. neck cancer and cervical cancer. Br J Cancer.
131. Zhang X, Cui L, Ye G, Zheng T, Song H, Xia T, et al. 2012;106(9):1526–34. PubMed PMID: 22472886.
Gastric juice microRNA-421 is a new biomarker Pubmed Central PMCID: PMC3341860. Epub
for screening gastric cancer. Tumour Biol. 2012; 2012/04/05.
33(6):2349–55. PubMed PMID: 22926798. Epub 142. Zhu W, Zhu D, Lu S, Wang T, Wang J, Jiang B, et al.
2012/08/29. miR-497 modulates multidrug resistance of human
132. Zhang Y, Gong W, Dai S, Huang G, Shen X, Gao M, cancer cell lines by targeting BCL2. Med Oncol.
et al. Downregulation of human farnesoid X receptor 2012;29(1):384–91. PubMed PMID: 21258880.
by miR-421 promotes proliferation and migration of Epub 2011/01/25.
hepatocellular carcinoma cells. Mol Cancer Res. 143. Kalev P, Simicek M, Vazquez I, Munck S, Chen L,
2012;10(4):516–22. PubMed PMID: 22446874. Soin T, et al. Loss of PPP2R2A inhibits homologous
Epub 2012/03/27. recombination DNA repair and predicts tumor sensi-
133. Hao J, Zhang S, Zhou Y, Liu C, Hu X, Shao tivity to PARP inhibition. Cancer Res. 2012;
C. MicroRNA 421 suppresses DPC4/Smad4 in pan- 72(24):6414–24. PubMed PMID: 23087057. Epub
creatic cancer. Biochem Biophys Res Commun. 2012/10/23.
2011;406(4):552–7. PubMed PMID: 21352803. 144. Cheng Y, Liu W, Kim ST, Sun J, Lu L, Sun J, et al.
Epub 2011/03/01. Evaluation of PPP2R2A as a prostate cancer suscep-
134. Ostling P, Leivonen SK, Aakula A, Kohonen P, tibility gene: a comprehensive germline and somatic
Makela R, Hagman Z, et al. Systematic analysis of study. Cancer Genet. 2011;204(7):375–81. PubMed
microRNAs targeting the androgen receptor in pros- PMID: 21872824. Epub 2011/08/30.
2 Omics of Hereditary Breast Cancer 39

145. Tang B, Testa JR, Kruger WD. Increasing the thera- 156. Manning AL, Dyson NJ. pRB, a tumor suppressor
peutic index of 5-fluorouracil and 6-thioguanine by with a stabilizing presence. Trends Cell Biol.
targeting loss of MTAP in tumor cells. Cancer Biol 2011;21(8):433–41. PubMed PMID: 21664133.
Ther. 2012;13(11):1082–90. PubMed PMID: Pubmed Central PMCID: PMC3149724. Epub
22825330. Pubmed Central PMCID: PMC3461815. 2011/06/15.
Epub 2012/07/25. 157. Jiang Z, Jones R, Liu JC, Deng T, Robinson T,
146. Taylor JL, Szmulewitz RZ, Lotan T, Hickson J, Chung PE, et al. RB1 and p53 at the crossroad of
Griend DV, Yamada SD, et al. New paradigms for EMT and triple-negative breast cancer. Cell Cycle.
the function of JNKK1/MKK4 in controlling growth 2011;10(10):1563–70. PubMed PMID: 21502814.
of disseminated cancer cells. Cancer Lett. Epub 2011/04/20.
2008;272(1):12–22. PubMed PMID: 18572308. 158. Jhaveri K, Modi S. HSP90 inhibitors for cancer ther-
Epub 2008/06/24. apy and overcoming drug resistance. Adv Pharmacol.
147. Ngeow J, Heald B, Rybicki LA, Orloff MS, Chen JL, 2012;65:471–517. PubMed PMID: 22959035. Epub
Liu X, et al. Prevalence of germline PTEN, 2012/09/11.
BMPR1A, SMAD4, STK11, and ENG mutations in 159. Chiosis G, Dickey CA, Johnson JL. A global view of
patients with moderate-load colorectal polyps. Hsp90 functions. Nat Struct Mol Biol. 2013;20(1):1–
Gastroenterology. 2013;144(7):1402–9. PubMed 4. PubMed PMID: 23288357. Epub 2013/01/05.
PMID: 23399955. Epub 2013/02/13. Eng. 160. Cizkova M, Susini A, Vacher S, Cizeron-Clairac G,
148. Shackelford DB, Abt E, Gerken L, Vasquez DS, Seki A, Andrieu C, Driouch K, et al. PIK3CA mutation
Leblanc M, et al. LKB1 inactivation dictates therapeutic impact on survival in breast cancer patients and in
response of non-small cell lung cancer to the metabo- ERalpha, PR and ERBB2-based subgroups. Breast
lism drug phenformin. Cancer Cell. 2013;23(2):143–58. Cancer Res. 2012;14(1):R28. PubMed PMID:
PubMed PMID: 23352126. Pubmed Central 22330809. Pubmed Central PMCID: PMC3496146.
PMCID: PMC3579627. Epub 2013/01/29. Epub 2012/02/15.
149. Lo B, Strasser G, Sagolla M, Austin CD, Junttila M, 161. Li WD, Li QR, Xu SN, Wei FJ, Ye ZJ, Cheng JK, et al.
Mellman I. Lkb1 regulates organogenesis and early Exome sequencing identifies an MLL3 gene germ line
oncogenesis along AMPK-dependent and -indepen- mutation in a pedigree of colorectal cancer and acute
dent pathways. J Cell Biol. 2012;199(7):1117–30. myeloid leukemia. Blood. 2013;121(8):1478–9.
PubMed PMID: 23266956. Pubmed Central PubMed PMID: 23429989. Epub 2013/02/23.
PMCID: PMC3529533. Epub 2012/12/26. 162. Je EM, Lee SH, Yoo NJ, Lee SH. Mutational and
150. Groschl B, Bettstetter M, Giedl C, Woenckhaus M, expressional analysis of MLL genes in gastric and
Edmonston T, Hofstadter F, et al. Expression of the colorectal cancers with microsatellite instability.
MAP kinase phosphatase DUSP4 is associated with Neoplasma. 2013;60(2):188–95. PubMed PMID:
microsatellite instability in colorectal cancer (CRC) 23259788. Epub 2012/12/25.
and causes increased cell proliferation. Int J Cancer. 163. Chou J, Lin JH, Brenot A, Kim JW, Provot S, Werb
2013;132(7):1537–46. PubMed PMID: 22965873. Z. GATA3 suppresses metastasis and modulates the
Epub 2012/09/12. tumour microenvironment by regulating microRNA-
151. Chimge NO, Frenkel B. The RUNX family in breast 29b expression. Nat Cell Biol. 2013;15(2):201–13.
cancer: relationships with estrogen signaling. PubMed PMID: 23354167. Epub 2013/01/29.
Oncogene. 2012;32(17):2121–30. PubMed PMID: 164. Nguyen AH, Tremblay M, Haigh K, Koumakpayi
23045283. Epub 2012/10/10. IH, Paquet M, Pandolfi PP, et al. Gata3 antagonizes
152. Rossetti S, Sacchi N. RUNX1: a microRNA hub in cancer progression in Pten-deficient prostates. Hum
normal and malignant hematopoiesis. Int J Mol Sci. Mol Genet. 2013;22(12):2400–10. PubMed PMID:
2013;14(1):1566–88. PubMed PMID: 23344057. 23428429. Epub 2013/02/23.
Pubmed Central PMCID: PMC3565335. Epub 165. Theodorou V, Stark R, Menon S, Carroll JS. GATA3
2013/01/25. acts upstream of FOXA1 in mediating ESR1 binding
153. Lei T, Zhao X, Jin S, Meng Q, Zhou H, Zhang by shaping enhancer accessibility. Genome Res.
M. Discovery of potential bladder cancer biomarkers 2013;23(1):12–22. PubMed PMID: 23172872.
by comparative urine proteomics and analysis. Clin Pubmed Central PMCID: PMC3530671. Epub
Genitourin Cancer. 2013;11(1):56–62. PubMed 2012/11/23.
PMID: 22982111. Epub 2012/09/18. 166. Ellis MJ, Ding L, Shen D, Luo J, Suman VJ, Wallis
154. Oliveira C, Pinheiro H, Figueiredo J, Seruca R, JW, et al. Whole-genome analysis informs breast
Carneiro F. E-cadherin alterations in hereditary dis- cancer response to aromatase inhibition. Nature.
orders with emphasis on hereditary diffuse gastric 2012;486(7403):353–60. PubMed PMID: 22722193.
cancer. Prog Mol Biol Transl Sci. 2013;116:337–59. Pubmed Central PMCID: PMC3383766. Epub
PubMed PMID: 23481202. Epub 2013/03/14. 2012/06/23.
155. Rodriguez FJ, Lewis-Tuffin LJ, Anastasiadis 167. Fanale D, Amodeo V, Corsini LR, Rizzo S, Bazan V,
PZ. E-cadherin’s dark side: possible role in tumor pro- Russo A. Breast cancer genome-wide association
gression. Biochim Biophys Acta. 2012;1826(1):23– studies: there is strength in numbers. Oncogene.
31. PubMed PMID: 22440943. Pubmed Central 2012;31(17):2121–8. PubMed PMID: 21996731.
PMCID: PMC3362679. Epub 2012/03/24. Epub 2011/10/15.
40 C.A. Moroski-Erkul et al.

168. Wu X, Marmarelis ME, Hodi FS. Activity of the heat Cancer. 1994;70(5):813–18. PubMed PMID:
shock protein 90 inhibitor ganetespib in melanoma. 7947085. Pubmed Central PMCID: PMC2033544.
PLoS One. 2013;8(2):e56134. PubMed PMID: Epub 1994/11/01.
23418523. Pubmed Central PMCID: PMC3572008. 179. Filip AA. New boys in town: prognostic role of
Epub 2013/02/19. SF3B1, NOTCH1 and other cryptic alterations in
169. Sharma P, Patel D, Chaudhary J. Id1 and Id3 expres- chronic lymphocytic leukemia and how it works.
sion is associated with increasing grade of prostate Leuk Lymphoma. 2013;54(9):1876–81. PubMed
cancer: Id3 preferentially regulates CDKN1B. Cancer PMID: 23343182. Epub 2013/01/25.
Med. 2012;1(2):187–97. PubMed PMID: 23342268. 180. Harbour JW, Roberson ED, Anbunathan H, Onken
Pubmed Central PMCID: PMC3544440. Epub MD, Worley LA, Bowcock AM. Recurrent muta-
2013/01/24. tions at codon 625 of the splicing factor SF3B1 in
170. Hodul PJ, Dong Y, Husain K, Pimiento JM, Chen J, uveal melanoma. Nat Genet. 2013;45(2):133–5.
Zhang A, et al. Vitamin E delta-tocotrienol induces PubMed PMID: 23313955. Epub 2013/01/15.
p27(Kip1)-dependent cell-cycle arrest in pancreatic 181. Biankin AV, Waddell N, Kassahn KS, Gingras MC,
cancer cells via an E2F-1-dependent mechanism. Muthuswamy LB, Johns AL, et al. Pancreatic cancer
PLoS One. 2013;8(2):e52526. PubMed PMID: genomes reveal aberrations in axon guidance path-
23393547. Pubmed Central PMCID: PMC3564846. way genes. Nature. 2012;491(7424):399–405.
Epub 2013/02/09. PubMed PMID: 23103869. Pubmed Central
171. Burgucu D, Guney K, Sahinturk D, Ozbudak IH, PMCID: PMC3530898. Epub 2012/10/30.
Ozel D, Ozbilim G, et al. Tbx3 represses PTEN and 182. Sawai CM, Freund J, Oh P, Ndiaye-Lobry D, Bretz
is over-expressed in head and neck squamous cell JC, Strikoudis A, et al. Therapeutic targeting of the
carcinoma. BMC Cancer. 2012;12:481. PubMed cyclin D3:CDK4/6 complex in T cell leukemia.
PMID: 23082988. Pubmed Central PMCID: Cancer Cell. 2012;22(4):452–65. PubMed PMID:
PMC3517435. Epub 2012/10/23. 23079656. Pubmed Central PMCID: PMC3493168.
172. Boyd SC, Mijatov B, Pupo GM, Tran SL, Epub 2012/10/20.
Gowrishankar K, Shaw HM, et al. Oncogenic 183. Meng X, Lu P, Bai H, Xiao P, Fan Q. Transcriptional
B-RAF(V600E) signaling induces the T-Box3 tran- regulatory networks in human lung adenocarcinoma.
scriptional repressor to repress E-cadherin and Mol Med Rep. 2012;6(5):961–6. PubMed PMID:
enhance melanoma cell invasion. J Invest Dermatol. 22895549, Epub 2012/08/17.
2012;133(5):1269–77. PubMed PMID: 23190890. 184. Wang W, Zhao LJ, Tan YX, Ren H, Qi ZT. MiR-138
Epub 2012/11/30. induces cell cycle arrest by targeting cyclin D3 in
173. Kundu M, Liu PP. Function of the inv(16) fusion gene hepatocellular carcinoma. Carcinogenesis.
CBFB-MYH11. Curr Opin Hematol. 2001;8(4):201– 2012;33(5):1113–20. PubMed PMID: 22362728.
5. PubMed PMID: 11561156. Epub 2001/09/19. Pubmed Central PMCID: PMC3334515. Epub
174. Davis JN, Rogers D, Adams L, Yong T, Jung JS, 2012/03/01.
Cheng B, et al. Association of core-binding factor 185. Pignataro L, Sambataro G, Pagani D, Pruneri
beta with the malignant phenotype of prostate and G. Clinico-prognostic value of D-type cyclins and
ovarian cancer cells. J Cell Physiol. 2010;225(3):875– p27 in laryngeal cancer patients: a review. Acta
87. PubMed PMID: 20607802. Epub 2010/07/08. Otorhinolaryngol Ital. 2005;25(2):75–85. PubMed
175. Ward AF, Braun BS, Shannon KM. Targeting onco- PMID: 16116829. Pubmed Central PMCID:
genic Ras signaling in hematologic malignancies. PMC2639874. Epub 2005/08/25.
Blood. 2012;120(17):3397–406. PubMed PMID: 186. Inoue J, Otsuki T, Hirasawa A, Imoto I, Matsuo Y,
22898602. Pubmed Central PMCID: PMC3482854. Shimizu S, et al. Overexpression of PDZK1 within
Epub 2012/08/18. the 1q12-q22 amplicon is likely to be associated
176. Patil S, Chamberlain RS. Neoplasms associated with with drug-resistance phenotype in multiple
germline and somatic NF1 gene mutations. myeloma. Am J Pathol. 2004;165(1):71–81. PubMed
Oncologist. 2012;17(1):101–16. PubMed PMID: PMID: 15215163. Pubmed Central PMCID:
22240541. Pubmed Central PMCID: PMC3267808. PMC1618545. Epub 2004/06/25.
Epub 2012/01/14. 187. Walker G, MacLeod K, Williams AR, Cameron DA,
177. Sangha N, Wu R, Kuick R, Powers S, Mu D, Fiander Smyth JF, Langdon SP. Estrogen-regulated gene
D et al. Neurofibromin 1 (NF1) defects are common expression predicts response to endocrine therapy in
in human ovarian serous carcinomas and co-occur patients with ovarian cancer. Gynecol Oncol.
with TP53 mutations. Neoplasia. 2008;10(12): 2007;106(3):461–8. PubMed PMID: 17624412.
1362––72, following 72. PubMed PMID: 19048115. Epub 2007/07/13.
Pubmed Central PMCID: PMC2586687. Epub 188. Alessi P, Leali D, Camozzi M, Cantelmo A, Albini
2008/12/03. A, Presta M. Anti-FGF2 approaches as a strategy to
178. Cawkwell L, Lewis FA, Quirke P. Frequency of compensate resistance to anti-VEGF therapy: long-
allele loss of DCC, p53, RBI, WT1, NF1, NM23 and pentraxin 3 as a novel antiangiogenic FGF2-
APC/MCC in colorectal cancer assayed by fluores- antagonist. Eur Cytokine Netw. 2009;20(4):225–34.
cent multiplex polymerase chain reaction. Br J PubMed PMID: 20167562. Epub 2010/02/20.
Oncogenes and Tumor Suppressor
Genes as a Biomarker in Breast 3

Eyyup Uctepe, Muradiye Acar, Esra Gunduz,

and Mehmet Gunduz

Breast cancer is the most common cause of cancer in women in the United
States and the Western world. The important question is what can be done
to limit the human suffering associated with cancer and to reduce the bur-
den on society? One solution is early detection. Early diagnosis of breast
cancer before symptoms emerge is the most effective prevention of breast
Currently, mammography is the gold standard for breast cancer screen-
ing. The procedure is suggested and often reimbursed for women between
the ages of 50 and 75. Yet it is presumed that between 15 and 25 % of
women with early-stage breast cancers are presently missed by commonly
used diagnostic procedures such as mammography. Since breast cancer is
also diagnosed in an increasing number of younger women, the screening
strategy should be modified. Hence, oncogenes and tumor suppressor
genes could be used as biomarkers for early detection of breast cancer.
Eventually, researchers aim to use the molecular data collected from an
individual tumor for prognostication and personalized therapy for each
patient. Genetic profiles of tumors are now providing information about
clinical outcome, and some prognostic and predictive indicators have
appeared based on this research. In the near future, prospective tissue col-
lection for molecular analysis may become routine in order to classify
patients for alternative treatment options and to optimize treatment strate-
gies based on molecular structure of the cancer.

Breast Cancer • Early Diagnosis • Oncogenes • Tumor Suppressor Genes
• Biomarker

E. Uctepe, PhD • M. Acar, PhD

E. Gunduz, PhD • M. Gunduz, MD, PhD (*)
Department of Medical Genetics, Turgut Özal
Üniversity Medical School, Ankara, Turkey
e-mail: mehmet.gunduz@gmail.com

D. Barh (ed.), Omics Approaches in Breast Cancer: Towards Next-Generation Diagnosis, 41

Prognosis and Therapy, DOI 10.1007/978-81-322-0843-3_3, © Springer India 2014
42 E. Uctepe et al.

Introduction Molecular Predictors of Response

to Therapy for Breast Cancer
Breast cancer is the second most common cause
of cancer death in women in the United States Breast cancer is a clinically heterogeneous dis-
and the Western world. Recently, an increase in ease; patients with the same stage of disease and
incidence has occurred in many of the countries similar pathological diagnoses can present very
that have screening programs. Significantly, such different clinical outcomes [6]. This clinical het-
an increase is especially observed in younger erogeneity is due to the genetic variability of
women in Europe and Asia [1]. patients and tumors. Eventually, researchers aim
Breast cancer causes disability, psychological to use the molecular data collected from an indi-
trauma, and economic burden. The burden of vidual tumor for prognostication and personal-
cancer affects not only the patients and their ized therapy for each patient.
loved ones, who suffer physically, emotionally, The US NIH Consensus Conference explained
and financially, but society in general. The loss that a clinically useful prognostic biomarker must
of human potential due to cancer-related mor- be a proven independent, significant factor that is
bidity and mortality is very hard to estimate, par- easy to determine and interpret and that has ther-
ticularly when the effect of the disease on the apeutic consequences.
patient’s family and friends is also counted in. As yet, measurements from tumor tissue
As the prevalence of cancer is highest among biomarkers have been used to diagnose breast
older people [2] and the population is getting cancer patients who may profit from specific
older, the overall costs to society will most likely therapies. Estrogen receptor testing has been
increase in the next years. The important ques- routinely executed since the 1980s on breast car-
tion is that what can be done to limit the human cinoma samples in order to decide whether hor-
suffering associated with cancer and to reduce monal therapy is indicated. Recently, estrogen
the burden on society. One solution is early receptor, progesterone receptor, and human epi-
detection. dermal growth factor receptor type 2 testing to
Breast cancer morbidity increases meaning- direct treatment decisions are standard of care.
fully when it is not diagnosed early in its progres- Nowadays, multigene assays have been intro-
sion. Early diagnosis of breast cancer before duced to estimate breast tumor behavior.
symptoms emerge is the most effective preven- Specifically, the Oncotype Dx and MammaPrint
tion of breast cancer. When the cancer is diag- assays have been used in North America and
nosed early, patients live longer and need less Europe to direct clinical assessments. Others,
extensive treatment [3]. Early detection of breast including the Breast Cancer Index (BCI; bio-
cancer decreases the suffering and cost to society Theranostics) and PAM50 (Expression Analysis,
associated with the disease. Hence, research Inc.), are gaining acceptance as validated assays
leading to an understanding of the mechanisms with related clinical results. Also, some germ line
that lead to breast cancer and ways to avoid it genetic tests are now reported to estimate
should be increased. response to specific treatments (e.g., BRCA1,
Currently, mammography is the gold standard BRCA2, CYP2D6) [7].
for breast cancer screening. It is presumed that
between 15 and 25 % of women with early-stage
breast cancers are presently missed by commonly Important Oncogenes and Tumor
used diagnostic procedures such as mammogra- Suppressor Genes for Breast Cancer
phy [4]. Yet this screening is suggested and often on Chromosome 17
reimbursed for women between the ages of 50
and 75 [5]. Since breast cancer is also diagnosed Abnormalities of chromosome 17 are significant
in a progressing number of younger women, the molecular genetic changes in human breast can-
screening strategy should be modified. cers. Some very well-known oncogenes (HER2,
3 Oncogenes and Tumor Suppressor Genes as a Biomarker in Breast Cancer 43

Table 3.1 Breast cancer predisposition genes on chromosome 17 and their basic functions
Gene ID Symbol Function of gene
2064 ERBB2/HER2 Epidermal growth factor (EGF) receptor family of receptor tyrosine kinases.
Amplification and/or overexpression has been reported in numerous cancers
7153 TOP2A DNA topoisomerase controls and alters the topologic states of DNA during
transcription. It is associated with the development of drug resistance
7157 P53 P53 responds to diverse cellular stresses to regulate target genes that induce cell cycle
arrest, apoptosis, senescence, and DNA repair. It is accumulated in a variety of
transformed cells
672 BRCA1 BRCA1 plays a role in maintaining genomic stability. It acts as a tumor suppressor.
BRCA1 combines with other tumor suppressors, to form a BRCA1-associated genome
surveillance complex (BASC). Mutations in this gene are responsible for
approximately 40 % of inherited breast cancers and more than 80 % of inherited breast
and ovarian cancers
3090 HIC-1 Hypermethylated in cancer 1, a candidate tumor suppressor gene which undergoes allelic
loss in breast and other human cancers. The human HIC-1 gene is a target gene of p53
4137 TAU Microtubule-associated protein TAU (MAPT) functions to keep cell shape,
microvesicle transportation, and spindle formation. Interfering spindle microtubule
dynamics will cause cell cycle arrest and apoptosis. TAU detection helps to identify
those patients who are most likely to benefit from taxane treatment and resistant to
paclitaxel treatment
This table has been modified by utilizing [60]

TOP2A, and TAU) and tumor suppressor genes than that in ductal carcinoma, the HER2 amplifi-
(p53, BRCA1, and HIC-1) are located on chro- cation is still an important adverse prognostic fac-
mosome 17, as shown in Table 3.1. tor in lobular breast cancer [11]. It is suggested
that tumor-initiating cells (side population frac-
tion by cytometry, SP) of breast cancer present
HER2 important HER2 expression—the SP fraction was
reduced by HER2 inhibitors [12]. A meta-analysis
HER2 amplification is widely utilized as molecu- for cohort randomized trials on women with
lar markers for trastuzumab target treatment. HER2-positive early breast cancer explored that
Hence, analysis of HER2/neu expression on trastuzumab-based adjuvant chemotherapy
breast tissue is used as the standard of care now derived profit in disease-free survival, overall sur-
[8]. The HER2 gene encodes a transmembrane vival, and recurrence to adjuvant chemotherapy
tyrosine kinase receptor (the human epidermal [13]. The candidate gene is regarded as amplifica-
growth factor receptor). It is located on 17q21.1, tion when demonstrating more than six HER2
which encodes two oncogenes. Most recently, a copies per nucleus or with a ratio of HER2 to cen-
series of HER2-targeting agents were developed, tromere 17 greater than 2.2. The HER2 amplifica-
including trastuzumab, pertuzumab, ertumax- tion and overexpression have been associated with
omab, and lapatinib. HER2 gene amplification or negative responses to conventional chemotherapy
protein overexpression is demonstrated in and poor prognosis, but better overall survival rate
approximately 25 % of newly diagnosed breast for trastuzumab in breast cancer [14, 15].
cancers. HER2/neu gene amplification is associ-
ated with a worse prognosis in patients with
node-positive breast cancer due to increased pro- TOP2A
liferation and angiogenesis and inhibition of
apoptosis [9, 10]. DNA topoisomerase II alpha (TOP2A) is a new
Even though the incidence and importance of marker of cell cycle turnover. TOP2A gene is
HER2 amplification in lobular carcinoma are less located on chromosome 17q21-q22. TOP2A is a
44 E. Uctepe et al.

major target of anthracycline activity [16]. In endocrine-sensitive but chemotherapy-resistant.

breast cancer, TOP2A expression has been asso- On the contrary, low TAU expression in
ciated with HER2 protein overexpression and ER-positive cancers presented a poor prognosis
cell proliferation. with tamoxifen alone, but may profit from tax-
The TOP2A gene showed an increase in ane-containing chemotherapy [22]. Patients with
responsiveness to anthracycline-containing che- TAU-negative expression showed better response
motherapy modalities relative to non-anthracycline to paclitaxel administration compared to patients
regimens [17]. Because the gene locus of TOP2A with TAU-positive expression (respectively,
is fairly close to the HER2 gene, the amplification 60 % and 15 %) [23]. Amplified TOP2A and
of TOP2A is often cooperated with HER2 gene TAU genes are correlated to an important
amplification. Twenty-three patients with T2–T4 response to anthracycline-based chemotherapy,
ER-negative and HER2-overexpressed breast can- taxane, or cisplatin, respectively.
cers treated with anthracycline-based chemother-
apy were evaluated by Orlando et al. TOP2A was
amplified in five (22 %) of the tumors. In all BRCA1
patients with TOP2A amplification, HER2 gene
amplification was determined as well. It was The breast cancer predisposition gene BRCA1 is
reported that the pathological complete remission located on chromosome 17q12-21. The breast
ratio in TOP2A amplified tumors is higher than in cancer genes BRCA1 and BRCA2 are involved
tumors without TOP2A amplification (respec- in tumor suppressor pathways, such as DNA
tively, 60 % and 15 %). It is postulated that in repair and cell cycle control, but are not directly
endocrine-unresponsive/HER2 overexpression interacting [24], and the respective protein prod-
cases, TOP2A amplification or the polysomy of ucts do not show any homology [25]. However, it
chromosome 17 is associated with meaningfully is interesting that breast cancers related to muta-
high remission after anthracycline-based chemo- tions of either BRCA1 or BRCA2 result in tumors
therapy [18]. TOP2A is used as a molecular target with a similar phenotype of high genomic insta-
of anthracycline drug and is pretty helpful as a bility [26]. The BRCA1 protein presents its func-
predictive marker of response to anthracycline tion via its ubiquitin ligase activity, which
therapy [19]. implicate in DNA repair pathways in response to
DNA damage, cell cycle checkpoints, and mito-
sis by targeting proteins for degradation [27, 28].
TAU BRCA2 acts with BRCA1 and RAD51 in
genotoxic stress response [29]. BRCA2 also
TAU is one of the microtubule-associated pro- functions in exit of mitosis and is essential for
teins (MAPs) and is located on chromosome formation of the contractile ring and appropriate
17q21.1. Tubulin-targeting agents alter the abscission [30]. This function is in accordance
microtubule function to disrupt cell shape, spin- with the observed genetic instability in cells lack-
dle formation, and microvesicle transportation. ing BRCA2. Loss of heterozygosity for BRCA1
Detection of TAU expression may direct doctors and BRCA2 is determined in nearly all BRCA1-
to choose the patients who are likely to benefit and BRCA2-associated carcinomas, correspond-
more from taxane treatment. The decreased ingly [31]. Hereditary breast cancer, appearing in
expression of TAU is accompanied with high carriers of mutations in the BRCA1 and BRCA2
responsiveness to paclitaxel in vitro. TAU genes, differs from sporadic breast cancer and
enhances microtubule assembly and stabilizes from non-BRCA1/2 familial breast carcino-
microtubules and it is likely that TAU competes mas, which imply defects in specific pathways
with taxanes for microtubule binding [20, 21]. It [32]. Particularly, BRCA1 carcinomas have the
was shown that high TAU mRNA expression in basal-like phenotype and are high-grade, highly
ER-positive breast cancer presented as proliferating, estrogen receptor-negative, and
3 Oncogenes and Tumor Suppressor Genes as a Biomarker in Breast Cancer 45

HER2-negative breast carcinomas, characterized targets p53 for proteasomal degradation through
by the expression of basal markers such as basal E3 ubiquitin ligase activity. Regulating p53 activ-
keratins, P-cadherin, and epidermal growth fac- ity with MDM2 inhibitor is a striking approach
tor receptor. for treatment of cancer [41]. A small-molecule
Inherited mutations in breast cancer predispo- MDM2 antagonist, nutlin-3, has been developed
sition genes, mainly BRCA1 and BRCA2, at last. Cancer cells with MDM2 gene amplifica-
account for approximately 5–10 % of breast can- tion are most susceptible to nutlin-3 in vitro and
cer cases [33]. Most mutations in BRCA1 and in vivo. It is suggested that patients with wild-
BRCA2 result in unstable protein products; how- type p53 tumors may profit from antagonists of
ever, how this led to cancer predisposition is not the p53-MDM2 interaction [42].
obvious, and as yet, no treatment methods have
been developed to target BRCA1 functions and
mutations [34]. However, it is possible that muta- HIC-1
tions that are related with increased cancer risk
might produce gene products that interfere with Hypermethylated in cancer 1 (HIC1, also named
tumor suppressor pathways or support oncogenic ZBTB29 or ZNF901) is located at 17p13.3. It is a
pathways. candidate tumor suppressor gene, which gener-
Germ line mutation of BRCA1 increases the ally undergoes allelic loss in breast and other
risk of having breast cancer. Genetic factors com- human cancers. High HIC1 protein expression
prise about 5 % of all breast cancer cases. has been accompanied with better outputs in
However, somatic BRCA1 mutations are seldom breast cancers. Lastly, it is demonstrated that
determined in sporadic breast tumors. BRCA1 HIC1 can hold down the ephrin-A1 transcription
methylation has been described to take place in and it involves in the pathogenesis of epithelial
sporadic breast tumors and to be accompanied cancers. Restoration of HIC1 in breast cancer
with decreased gene expression. Hence, epigen- cells causes a growth arrest in vivo [43]. It is
etic modification and deletion of the BRCA1 shown that HIC1 controls breast cancer cell
gene might act as Knudson’s two “hits” in spo- responses to endocrine therapies. A demethylat-
radic breast tumorigenesis [35, 36]. ing drug 5-aza-2′-deoxycytidine can restore the
Retrospective studies explored a high level of HIC1 expression in MDAMB231 cells. Hence,
responsiveness to platin derivatives in BRCA- restoration of HIC1 function by demethylation
associated tumors [37] and first clinical trials dem- may suggest a therapeutic opportunity in breast
onstrate good efficacy and tolerability for PARPs, cancer [44].
or poly ADP (adenosine diphosphate)-ribose
polymerase inhibitors, in mutation carriers with
advanced breast and ovarian cancers [38]. But, the Hereditary Breast Cancer
advantages of risk-reducing prophylactic surgery
in mutation carriers have been verified [39]. Monogenic Inheritance of BRCA1
and BRCA2 Mutations

P53 Hereditary breast and ovarian cancers are brought

about by an autosomal dominant inheritance with
The tumor suppressor gene p53 is located on incomplete penetrance. Population-based studies
17p13.1. P53 somatic change is determined in put its reentrance for breast cancer at 45–65 %
approximately 50 % of all human cancers [40]. [37, 38]. This emphasizes the effect of modifying
MDM2 can inactivate the p53 through binding to factors and lifestyle. Lastly, approximately 50 %
the transactivation domain of p53. High expres- of the monogenically determined breast and
sion level of this gene can cause excessive inacti- ovary cancers are through a mutation in one or
vation of p53 tumor suppressor protein. MDM2 the other of the highly penetrate BRCA genes.
46 E. Uctepe et al.

Table 3.2 Breast cancer-related genes and their effect on risk

Risk genes Increase in risk Genes/syndromes
Highly penetrant genes 5- to 20-fold BRCA1/BRCA2/RAD51C: hereditary breast and ovarian cancer
TP53: Li-Fraumeni syndrome
STK11/LKB1: Peutz-Jeghers syndrome
PTEN: Cowden syndrome
Moderately penetrant genes 1.5- to 5-fold CHEK2, PALB2, BRIP1, ATM
Mildly penetrant genes 0.7- to 1.5-fold FGFR2, TOX3, MAP3K1, CAMK1D, SNRPB, FAM84B/c-MYC,
COX11, LSP1, CASP8, ESR1, ANKLE1, MERIT40, etc.
This table adapted from [61]

Women carrying a mutated gene have an 80–90 % For instance, ATM, CHEK2, BRIP1, and PALB2
risk of breast cancer and a 20–50 % risk of ovar- have been classified in moderate-risk gene groups
ian cancer [45]. with low heterozygote frequency [47]. Some
low-risk variants located within the intron or reg-
ulatory areas were identified in the following
Monogenic Inheritance in Mutations genes: FGFR2, MAP3K1, TNRC9, and LSP1
of the Gene RAD51C and as Yet (2q35, 6q22.33, 8q24) [48, 49]. The risks inher-
Unidentified, Highly Penetrant Genes ent in these variants are very low, with relative
risks (RRs) of just about 1.1–1.3; however, their
The third highly penetrant gene for breast and heterozygote frequencies are high.
ovarian cancer, RAD51C, was determined in
the summer of 2010 [46]. It is mutated in almost
1.5–4 % of all families predisposed toward breast Other Predisposition Genes
and ovarian cancer with high or moderate pen- Associated with Breast Cancer
etrance. Like BRCA1 and BRCA2, it has an
important role in DNA repair as a tumor suppres- G-Protein-Coupled Receptor-
sor gene. First studies in other populations verify Associated Sorting Protein 1
that mutation in RAD51C is a predisposing fac- (GASP-1)
tor for development of breast cancer. But, as it is
rarely mutated and the data available on its pen- Zheng et al. identified a specific fragment of
etrance are as yet inadequate, it is most recently G-protein-coupled receptor-associated sorting
not offered as part of routine diagnostics. protein 1 (GASP-1) which was found in the sera
of patients with early-stage disease but absent in
sera of normal patients. They immunohisto-
Moderately and Mildly Penetrant chemically determined overexpression of
Gene Variants GASP-1 in all 107 cases of archived ductal
breast carcinoma tumor samples, although nor-
A significant proportion of BRCA1/2- negative mal adjacent breast tissue from 12 cases of duc-
high-risk families possibly may have mutations tal carcinoma presented little or no staining.
in highly penetrant genes, which have not been Moreover, all 10 cases of metastatic breast carci-
identified yet. Therefore, it is postulated that the noma present in lymph nodes were positive.
total effect of moderately and mildly penetrant These studies point to GASP-1 as a potential
gene variants is probably the cause for the major- new serum and tumor biomarker for breast can-
ity of carcinomas [47]. This may be correct for cer and suggest that GASP-1 may be a novel tar-
50 % of cases of hereditary breast cancer and get for the development of new breast cancer
20 % of all cases of breast cancer (see Table 3.2). therapeutics [4].
3 Oncogenes and Tumor Suppressor Genes as a Biomarker in Breast Cancer 47

Estrogen-Related Receptor-α (ERR-α) suggested as a prognostic factor in node-negative

breast cancer patients and STAT5 has a role in
Jarzabek et al. showed that breast cancer tis- estimating response to endocrine therapy [54,
sues showed a slightly higher expression level 55]. In a retrospective study of 346 node-negative
of estrogen-related receptor-α (ERR-α) mRNA tumors, nuclear expression of STAT3 was deter-
compared to the normal breast tissue (mean, mined in 23.1 % of cases and phospho-STAT3
57.7 ± SD 58.7, 46.2 ± SD 42.0, respectively). (Tyr705) in 43.5 %; both were accompanied with
But, ERR-α mRNA levels in breast cancer tis- a reduced risk of recurrence and longer survival.
sues demonstrated greater diversity than in nor- It is demonstrated by multivariate analysis that
mal tissues. It is probable that ERR-α could play phosphorylation of STAT3 is an independent
a significant role in the alternative pathway to prognostic factor for survival in this group, with
classical estrogen receptor-dependent pathway in no relation to hormone-receptor expression, pro-
cell signaling. The development and use of ERR liferation index, or HER-2 amplification [56].
modulators in the near future could help to design
new well-tolerated and individualized therapeu-
tic agents [50]. Visfatin

Mean serum visfatin was significantly greater in

Survivin cases than in controls and patients with benign
breast lesions (BBL) BBL (p < 0.001). In cases,
Petrarca et al. showed that survivin overex- visfatin was significantly correlated with CA
pression in the primary tumor may be used as 15-3 (p = 0.03), hormone-receptor status
a promising predictive biomarker of complete (p < 0.001), and lymph node invasion (p = 0.06)
pathological response (pCR) to neoadjuvant che- but not with metabolic and anthropometric vari-
motherapy in patients with stage II and stage III ables (p > 0.05). Multivariable regression analysis
breast cancer [51]. explored that absence of estrogen and progester-
one receptors (ER−PR−) was the strongest deter-
minant of serum visfatin level (p < 0.001) in cases
BCL2 adjusting for demographic, metabolic, and clini-
copathological features [57].
Callagi et al. showed that the meta-analysis
strongly supports the prognostic role of BCL2 in
breast cancer and this effect is independent of TGF-β-Signaling Pathway
lymph node status, tumor size, and tumor grade in Breast Cancer
as well as a range of other biological variables on
multi-variety analysis. Further large prospective The transforming growth factor-β (TGF-β) path-
studies are now essential to establish the clini- way has dual effects on tumor growth. Discordant
cal utility of BCL2 as an independent prognostic results have been apparently demonstrated on the
marker [52]. association between TGF-β-signaling markers
and prognosis in breast cancer. It is shown that
tumors with high expression of TβRII (TGF-β
STAT3 and STAT5 receptors II), TβRI (TGF-β receptors I), and
TβRII and p-Smad2 (p = 0.018, 0.005, and 0.022,
Also in recent years, recognition of the intrinsic respectively) and low expression of Smad4
subtypes of breast cancer has enabled the disease (p = 0.005) had a poor prognosis concerning
to be categorized into different types, with the progression-free survival. Low Smad4 expres-
use of adjuvant therapies targeted to the biologi- sion combined with high p-Smad2 expression or
cal profile of each type [3, 53]. STAT3 has been low expression of Smad4 combined with high
48 E. Uctepe et al.

expression of both TGF-β receptors showed an appeared based on this research. While progress
increased hazard ratio of 3.04 (95 % confidence of therapeutics in this field is rapid and laudable,
interval (CI) 1.390–6.658) and 2.20 (95 % CI many hindrances must be overcome for these
1.464–3.307), respectively, for disease relapse. molecule-based therapies to become a reality for
As a result, combining TGF-β biomarkers use in common cancers.
give us prognostic information for patients with Pharmacological difficulties include develop-
stages I through III breast cancer. This can deter- ing safe, effective, and site-specific delivery
mine patients at increased risk for disease recur- mechanisms for these molecule-directed thera-
rence, so they might be candidates for additional pies. Despite these challenges, the remarkable
treatment [58]. potential of miRNAs as cancer biomarkers and
therapeutics cannot be undervalued. If the current
studies in these molecular targets can be sus-
PTEN or PIK3CA tained, it will bring a new dimension to the field
of diagnostics and therapeutics for breast cancer.
MK-2206 was shown to repress Akt signaling Hence, studies with the molecules mentioned in
and cell cycle progression and increased apopto- this chapter have the potential to transform cur-
sis in a dose-dependent manner in breast cancer rent practice to the ideal of individualized care
cell lines. Cell lines with PTEN or PIK3CA for breast cancer patients. Additionally, prospec-
mutations were meaningfully more sensitive to tive tissue collection for molecular analysis may
MK-2206; however, several lines with PTEN/ become routine in the near future in order to clas-
PIK3CA mutations were MK-2206 resistant. sify patients for alternative treatment options and
siRNA knockdown of PTEN in breast cancer to optimize treatment strategies based on molec-
cells arose Akt phosphorylation consistent with ular structure of the cancer.
increased MK-2206 sensitivity. Stable transfec-
tion of PIK3CA, E545K, or H1047R mutant
plasmids into normal-like MCF10A breast cells References
increased MK-2206 sensitivity. Cell lines that
were less sensitive to MK-2206 had lower ratios 1. Lee SY, Jeong SH, Kim YN, Kim J, Kang DR, Kim
HC, et al. Cost-effective mammography screening in
of Akt1/Akt2 and had reduced growth inhibition
Korea: high incidence of breast cancer in young
with Akt siRNA knockdown. In PTEN-mutant women. Cancer Sci. 2009;100(6):1105–11.
ZR75-1 breast cancer xenografts, MK-2206 doi:10.1111/j.1349-7006.2009.01147.x.
treatment repressed Akt signaling, cell prolifera- 2. Jemal A, Murray T, Ward E, Samuels A, Tiwari RC,
Ghafoor A, et al. Cancer statistics, 2005. CA Cancer J
tion, and tumor growth. In vitro, MK-2206 dem-
Clin. 2005;55(1):10–30.
onstrated a synergistic interaction with paclitaxel 3. Etzioni RN, Urban S, Ramsey M, McIntosh S,
in MK-2206-sensitive cell lines, and this combi- Schwartz B, Reid J, et al. The case for early detection.
nation had significantly further antitumor effi- Nat Rev Cancer. 2003;3(4):243–52. doi:10.1038/
cacy than either agent alone in vivo.
4. Tuszynski GP, Rothman VL, Zheng X, Gutu M, Zhang
As a result, MK-2206 has antitumor activity X, Chang F. G-protein coupled receptor-associated
alone and in combination with chemotherapy. sorting protein 1 (GASP-1), a potential biomarker in
This effect may be greater in tumors with PTEN breast cancer. Exp Mol Pathol. 2011;91(2):608–13.
loss or PIK3CA mutation [59].
5. Kerlikowske K, Grady D, Barclay J, Sickles EA,
Eaton A, Ernster V. Positive predictive value of
screening mammography by age and family history of
Conclusion and Future Perspective breast cancer. JAMA. 1993;270(20):2444–50.
6. Morris SR, Carey LA. Molecular profiling in breast
cancer. Rev Endocr Metab Disord. 2007;8(3):185–98.
In conclusion, genetic profiles of tumors are now doi:10.1007/s11154-007-9035-3.
being informative about clinical outcome, and 7. Galanina N, Bossuyt V, Harris LN. Molecular pre-
some prognostic and predictive indicators have dictors of response to therapy for breast cancer.
3 Oncogenes and Tumor Suppressor Genes as a Biomarker in Breast Cancer 49

Cancer J. 2011;17(2):96–103. doi:10.1097/ 20. McGrogan BT, Gilmartin B, Carney DN, McCann
PPO.0b013e318212dee3. A. Taxanes, microtubules and chemoresistant breast
8. Hudis CA. Trastuzumab – mechanism of action and cancer. Biochim Biophys Acta. 2008;1785(2):96–
use in clinical practice. N Engl J Med. 2007;357(1):39– 132. doi:10.1016/j.bbcan.2007.10.004.
51. doi:10.1056/NEJMra043186. 21. Pusztai L. Markers predicting clinical benefit in
9. Chan KC, Lo YM. Circulating nucleic acids as breast cancer from microtubule-targeting agents.
a tumor marker. Histol Histopathol. 2002;17(3): Ann Oncol. 2007;18 Suppl 12:xii15–20. doi:10.1093/
937–43. annonc/mdm534.
10. Silva J, Silva JM, Garcia V, Garcia JM, Dominguez 22. Andre F, Hatzis C, Anderson K, Sotiriou C, Mazouni
G, Bonilla F. RNA is more sensitive than DNA in C, Mejia J, et al. Microtubule-associated protein-tau
identification of breast cancer patients bearing tumor is a bifunctional predictor of endocrine sensitivity and
nucleic acids in plasma. Genes Chromosomes Cancer. chemotherapy resistance in estrogen receptor-positive
2002;35(4):375–6. doi:10.1002/gcc.10124. breast cancer. Clin Cancer Res. 2007;13(7):2061–7.
11. Rosenthal SI, Depowski PL, Sheehan CE, Ross doi:10.1158/1078-0432.CCR-06-2078.
JS. Comparison of HER-2/neu oncogene amplification 23. Tanaka S, Nohara T, Iwamoto M, Sumiyoshi K,
detected by fluorescence in situ hybridization in lobu- Kimura K, Takahashi Y, et al. Tau expression and effi-
lar and ductal breast cancer. Appl Immunohistochem cacy of paclitaxel treatment in metastatic breast can-
Mol Morphol. 2002;10(1):40–6. cer. Cancer Chemother Pharmacol. 2009;64(2):341–6.
12. Bedard PL, Piccart-Gebhart MJ. Current paradigms doi:10.1007/s00280-008-0877-5.
for the use of HER2-targeted therapy in early-stage 24. Livingston DM. Cancer. Complicated supercom-
breast cancer. Clin Breast Cancer. 2008;8 Suppl plexes. Science. 2009;324(5927):602–3. doi:10.1126/
4:S157–65. doi:10.3816/CBC.2008.s.012. science.1174839.
13. Yin W, Jiang Y, Shen Z, Shao Z, Lu J. Trastuzumab in 25. Irminger-Finger I, Jefford CE. Is there more to BARD1
the adjuvant treatment of HER2-positive early breast than BRCA1? Nat Rev Cancer. 2006;6(5):382–91.
cancer patients: a meta-analysis of published random- doi:10.1038/nrc1878.
ized controlled trials. PLoS One. 2011;6(6):e21030. 26. Grigorova M, Staines JM, Ozdag H, Caldas C,
doi:10.1371/journal.pone.0021030. Edwards PA. Possible causes of chromosome insta-
14. Chia S, Norris B, Speers C, Cheang M, Gilks B, Gown bility: comparison of chromosomal abnormali-
AM, et al. Human epidermal growth factor receptor 2 ties in cancer cell lines with mutations in BRCA1,
overexpression as a prognostic factor in a large tis- BRCA2, CHK2 and BUB1. Cytogenet Genome Res.
sue microarray series of node-negative breast cancers. 2004;104(1–4):333–40. doi:10.1159/000077512.
J Clin Oncol. 2008;26(35):5697–704. doi:10.1200/ 27. Starita LM, Machida Y, Sankaran S, Elias JE, Griffin
JCO.2007.15.8659. K, Schlegel BP, et al. BRCA1-dependent ubiquitina-
15. Guiu S, Liegard M, Favier L, van Praagh I, Largillier tion of gamma-tubulin regulates centrosome number.
R, Weber B, et al. Long-term follow-up of HER2- Mol Cell Biol. 2004;24(19):8457–66. doi:10.1128/
overexpressing stage II or III breast cancer treated MCB.24.19.8457-8466.2004.
by anthracycline-free neoadjuvant chemotherapy. 28. Joukov V, Groen AC, Prokhorova T, Gerson R, White
Ann Oncol. 2011;22(2):321–8. doi:10.1093/annonc/ E, Rodriguez A, et al. The BRCA1/BARD1 het-
mdq397. erodimer modulates ran-dependent mitotic spindle
16. Slamon DJ, Press MF. Alterations in the TOP2A and assembly. Cell. 2006;127(3):539–52. doi:10.1016/j.
HER2 genes: association with adjuvant anthracycline cell.2006.08.053.
sensitivity in human breast cancers. J Natl Cancer 29. Scully R, Chen J, Ochs RL, Keegan K, Hoekstra M,
Inst. 2009;101(9):615–18. doi:10.1093/jnci/djp092. Feunteun J, et al. Dynamic changes of BRCA1 sub-
17. O’Malley FP, Chia S, Tu D, Shepherd LE, Levine nuclear location and phosphorylation state are initi-
MN, Bramwell VH, et al. Topoisomerase II alpha ated by DNA damage. Cell. 1997;90(3):425–35.
and responsiveness of breast cancer to adjuvant che- 30. Daniels MJ, Wang Y, Lee M, Venkitaraman
motherapy. J Natl Cancer Inst. 2009;101(9):644–50. AR. Abnormal cytokinesis in cells deficient in
doi:10.1093/jnci/djp067. the breast cancer susceptibility protein BRCA2.
18. Orlando LB, Del Curto S, Gandini R, Ghisini E, Pietr Science. 2004;306(5697):876–9. doi:10.1126/
R, Torrisi A, et al. Topoisomerase IIalpha gene sta- science.1102574.
tus and prediction of pathological complete remission 31. Yang X, Lippman ME. BRCA1 and BRCA2 in breast
after anthracycline-based neoadjuvant chemotherapy cancer. Breast Cancer Res Treat. 1999;54(1):1–10.
in endocrine non-responsive Her2/neu-positive breast 32. Honrado E, Benitez J, Palacios J. The molecu-
cancer. Breast. 2008;17(5):506–11. doi:10.1016/j. lar pathology of hereditary breast cancer:
breast.2008.03.007. genetic testing and therapeutic implications.
19. Miyoshi Y, Kurosumi M, Kurebayashi J, Matsuura Mod Pathol. 2005;18(10):1305–20. doi:10.1038/
N, Takahashi M, Tokunaga E, et al. Predictive fac- modpathol.3800453.
tors for anthracycline-based chemotherapy for human 33. Foulkes WD. Inherited susceptibility to common can-
breast cancer. Breast Cancer. 2010;17(2):103–9. cers. N Engl J Med. 2008;359(20):2143–53.
doi:10.1007/s12282-009-0152-6. doi:10.1056/NEJMra0802968.
50 E. Uctepe et al.

34. Widakowich C, de Azambuja E, Gil T, Cardoso F, 45. Meindl A, Ditsch N, Kast K, Rhiem K, Schmutzler
Dinh P, Awada A, et al. Molecular targeted therapies RK. Hereditary breast and ovarian cancer: new
in breast cancer: where are we now? Int J Biochem genes, new treatments, new concepts. Dtsch
Cell Biol. 2007;39(7–8):1375–87. doi:10.1016/j. Arztebl Int. 2011;108(19):323–30. doi:10.3238/
biocel.2007.04.015. arztebl.2011.0323.
35. Birgisdottir V, Stefansson OA, Bodvarsdottir SK, 46. Rahman N, Seal S, Thompson D, Kelly P, Renwick A,
Hilmarsdottir H, Jonasson JG, Eyfjord JE. Epigenetic Elliott A, et al. PALB2, which encodes a BRCA2-
silencing and deletion of the BRCA1 gene in sporadic interacting protein, is a breast cancer susceptibility
breast cancer. Breast Cancer Res. 2006;8(4):R38. gene. Nat Genet. 2007;39(2):165–7. doi:10.1038/
doi:10.1186/bcr1522. ng1959.
36. Wei M, Grushko TA, Dignam J, Hagos F, Nanda R, 47. Ripperger T, Gadzicki D, Meindl A, Schlegelberger
Sveen L, et al. BRCA1 promoter methylation in spo- B. Breast cancer susceptibility: current knowl-
radic breast cancer is associated with reduced BRCA1 edge and implications for genetic counselling. Eur
copy number and chromosome 17 aneusomy. Cancer J Hum Genet. 2009;17(6):722–31. doi:10.1038/
Res. 2005;65(23):10692–9. doi:10.1158/0008-5472. ejhg.2008.212.
CAN-05-1277. 48. Stacey SN, Manolescu A, Sulem P, Rafnar T,
37. Antoniou A, Pharoah PD, Narod S, Risch HA, Eyfjord Gudmundsson J, Gudjonsson SA, et al. Common vari-
JE, Hopper JL, et al. Average risks of breast and ovar- ants on chromosomes 2q35 and 16q12 confer suscep-
ian cancer associated with BRCA1 or BRCA2 muta- tibility to estrogen receptor-positive breast cancer. Nat
tions detected in case Series unselected for family Genet. 2007;39(7):865–9. doi:10.1038/ng2064.
history: a combined analysis of 22 studies. Am J Hum 49. Turnbull CS, Ahmed J, Morrison D, Pernet A,
Genet. 2003;72(5):1117–30. doi:10.1086/375033. Renwick M, Maranian S, et al. Genome-wide associa-
38. Chen S, Iversen ES, Friebel T, Finkelstein D, Weber tion study identifies five new breast cancer suscepti-
BL, Eisen A, et al. Characterization of BRCA1 and bility loci. Nat Genet. 2010;42(6):504–7. doi:10.1038/
BRCA2 mutations in a large United States sample. J ng.586.
Clin Oncol. 2006;24(6):863–71. doi:10.1200/ 50. Jarzabek K, Koda M, Kozlowski L, Sulkowski S,
JCO.2005.03.6772. Kottler ML, Wolczynski S. The significance of
39. Consortium, Chek Breast Cancer Case-Control. the expression of ERRalpha as a potential bio-
CHEK2*1100delC and susceptibility to breast marker in breast cancer. J Steroid Biochem Mol
cancer: a collaborative analysis involving 10,860 Biol. 2009;113(1–2):127–33. doi:10.1016/j.
breast cancer cases and 9,065 controls from 10 jsbmb.2008.12.005.
studies. Am J Hum Genet. 2004;74(6):1175–82. 51. Petrarca CR, Brunetto AT, Duval V, Brondani A,
doi:10.1086/421251. Carvalho GP, Garicochea B. Survivin as a predictive
40. Soussi T. The p53 tumor suppressor gene: from biomarker of complete pathologic response to neoad-
molecular biology to clinical investigation. Ann N Y juvant chemotherapy in patients with stage II and
Acad Sci. 2000;910:121–37; discussion 137–9. stage III breast cancer. Clin Breast Cancer.
41. Rigatti MJ, Verma R, Belinsky GS, Rosenberg DW, 2011;11(2):129–34. doi:10.1016/j.clbc.2011.03.002.
Giardina C. Pharmacological inhibition of Mdm2 52. Callagy GM, Webber MJ, Pharoah PD, Caldas
triggers growth arrest and promotes DNA breakage in C. Meta-analysis confirms BCL2 is an independent
mouse colon tumors and human colon cancer cells. prognostic marker in breast cancer. BMC Cancer.
Mol Carcinog. 2012;51(5):363–78. doi:10.1002/ 2008;8:153. doi:10.1186/1471-2407-8-153.
mc.20795. 53. Bast Jr RC, Lilja H, Urban N, Rimm DL, Fritsche H,
42. Tovar C, Rosinski J, Filipovic Z, Higgins B, Gray J, et al. Translational crossroads for biomarkers.
Kolinsky K, Hilton H, et al. Small-molecule Clin Cancer Res. 2005;11(17):6103–8.
MDM2 antagonists reveal aberrant p53 signaling doi:10.1158/1078-0432.CCR-04-2213.
in cancer: implications for therapy. Proc Natl Acad 54. Kolb TM, Lichy J, Newhouse JH. Comparison of the
Sci U S A. 2006;103(6):1888–93. doi:10.1073/ performance of screening mammography, physical
pnas.0507493103. examination, and breast US and evaluation of factors
43. Zhang W, Zeng X, Briggs KJ, Beaty R, Simons B, Chiu that influence them: an analysis of 27,825 patient
Yen RW, et al. A potential tumor suppressor role for evaluations. Radiology. 2002;225(1):165–75.
Hic1 in breast cancer through transcriptional repres- 55. Burnside ES, Park JM, Fine JP, Sisney GA. The use of
sion of ephrin-A1. Oncogene. 2010;29(17):2467–76. batch reading to improve the performance of screen-
doi:10.1038/onc.2010.12. ing mammography. AJR Am J Roentgenol.
44. Nicoll G, Crichton DN, McDowell HE, Kernohan 2005;185(3):790–6.
N, Hupp TR, Thompson AM. Expression of the 56. Papanicolaou GN, Holmquist DG, Bader GM, Falk
Hypermethylated in Cancer gene (HIC-1) is associ- EA. Exioliative cytology of the human mammary
ated with good outcome in human breast cancer. gland and its value in the diagnosis of cancer and
Br J Cancer. 2001;85(12):1878–82. doi:10.1054/ other diseases of the breast. Cancer. 1958;
bjoc.2001.2163. 11(2):377–409.
3 Oncogenes and Tumor Suppressor Genes as a Biomarker in Breast Cancer 51

57. Dalamaga M, Archondakis S, Sotiropoulos G, tor MK-2206 in breast cancer. Clin Cancer Res.
Karmaniolas K, Pelekanos N, Papadavid E, et al. Could 2012;18(20):5816–28. doi:10.1158/1078-0432.
serum visfatin be a potential biomarker for postmeno- CCR-12-1141.
pausal breast cancer? Maturitas. 2012;71(3):301–8. 60. Zhang et al. The Important Molecular Markers
doi:10.1016/j.maturitas.2011.12.013. on Chromosome 17 and Their Clinical Impact in
58. de Kruijf EM, Dekker TJ, Hawinkels LJ, Putter H, Smit Breast Cancer. Int. J. Mol. Sci. 2011;12:5672–83;
VT, Kroep JR, et al. The prognostic role of TGF-beta doi:10.3390/ijms12095672.
signaling pathway in breast cancer patients. Ann Oncol. 61. Meindle et al. Hereditary Breast and Ovarian Cancer.
2013;24(2):384–90. doi:10.1093/annonc/mds333. Dtsch Arztebl Int 2011;108(19).
59. Sangai T, Akcakanat A, Chen H, Tarco E, Wu Y,
Do KA, et al. Biomarkers of response to Akt inhibi-
Breast Cancer Genomics
Birendra Kumar

Breast cancer is a complex disease caused by the progressive accumula-
tion of multiple gene mutations combined with epigenetic dysregulation
of critical genes and protein pathways. There is substantial interindividual
variability in both the age at diagnosis and phenotypic expression of the
disease. With an estimated 1,152,161 new breast cancer cases diagnosed
worldwide per year, cancer-control efforts in the post-genome era should
be focused at both population and individual levels to develop novel risk
assessment and treatment strategies that will further reduce the morbidity
and mortality associated with the disease.
The discovery that mutations in the BRCA1 and BRCA2 genes increase
the risk of breast and ovarian cancers has radically transformed our under-
standing of the genetic basis of breast cancer, leading to improved man-
agement of high-risk women. A better understanding of tumor host biology
has led to improvements in the multidisciplinary management of breast
cancer, and traditional pathological evaluation is being complemented by
more sophisticated genomic approaches. A number of genomic biomark-
ers have been developed for clinical use, and increasingly, pharmacoge-
netic end points are being incorporated into the clinical trial design.
For women diagnosed with breast cancer, prognostic or predictive
information is most useful when coupled with targeted therapeutic
approaches, very few of which exist for women with triple-negative breast
cancer or those with tumors resistant to chemotherapy. The immediate
challenge is to learn how to use the molecular characteristics of an indi-
vidual and their tumor to improve detection and treatment and, ultimately,
to prevent the development of breast cancer.

B. Kumar, PhD
Department of Biochemistry, J.C. Bose Institute
of Life Sciences, Bundelkhand University,
Jhansi, India
e-mail: birendra_bio23@yahoo.co.in,

D. Barh (ed.), Omics Approaches in Breast Cancer: Towards Next-Generation Diagnosis, 53

Prognosis and Therapy, DOI 10.1007/978-81-322-0843-3_4, © Springer India 2014
54 B. Kumar

Genomic • Epigenome • Breast cancer • Prevention • Biomarker • Tumor
dormancy • Personal genomics • Genetic test • Gene expression • Risk
assessment • Targeted therapy • GWAS • Genotype • BRCA1 • BRCA2

Introduction and progression of breast cancer is increasing at a

remarkable rate through application of powerful
Breast cancer is the most common malignancy analysis tools that enable genome-wide analysis
affecting women, with more than one million of DNA sequence and structure, copy number,
cases occurring worldwide annually. There were allelic loss, and epigenomic modification.
412,000 deaths attributed to breast cancer for Application of these techniques to elucidation of
women in the world, representing 1.6 % of all the nature and timing of these events is enriching
female deaths [1]. Breast cancer is mainly a dis- our understanding of mechanisms that increase
ease of the genome, with cancers occurring and breast cancer susceptibility, enable tumor initia-
progressing through accumulation of abnormali- tion and progression to metastatic disease, and
ties that amend the genome—by altering DNA determine therapeutic response or resistance.
sequence, copy number, and construction—in These studies also reveal the molecular differ-
ways that supply to sundry aspects of tumor ences between cancer and normal that may be
pathophysiology. Classic examples of genomic exploited to therapeutic benefit or that provide
events that contribute to breast cancer pathophys- targets for molecular assays that may enable
iology include inherited mutations in BRCA1, early cancer detection and predict individual dis-
BRCA2, TP53, and CHK2 that contribute to the ease progression or response to treatment.
initiation of breast cancer; amplification of In this chapter, I assess present and prospec-
ERBB2 (formerly HER2) and mutations of ele- tive instructions in genome analysis and précis
ments of the PI3-kinase pathway that activate studies that provide insights into breast cancer
aspects of epidermal growth factor receptor pathophysiology or that propose policy to
(EGFR) signaling; and deletion of CDKN2A/B improve management of breast cancer.
that contributes to cell cycle deregulation and
genome instability. It is now apparent that accu-
mulation of these aberrations is a time-dependent Primary Prevention of Breast
process that accelerates with age [2]. Cancer Through a Genomic
Although American women living to an age of Approach
85 have a 1 in 8 chance of developing breast
cancer, the incidence of cancer in women younger High-Penetrance Breast Cancer Genes
than 30 years is uncommon. This is consistent
with a multistep cancer progression model A developing body of evidence credentials the
whereby mutation and selection drive the tumor’s benefits of preventive measures with minimal
development, analogous to traditional Darwinian possibility to women with identifiable highly
evolution [3, 4]. In the case of cancer, the driving penetrant mutations in the BRCA1 and BRCA2
events are changes in sequence, copy number, genes. While other genes, such as PTEN in
and structure of DNA and alterations in chroma- Cowden syndrome and TP53 in Li–Fraumeni
tin structure or other epigenetic marks. Our syndrome, also contribute to a small fraction of
understanding of the genetic, genomic, and epig- hereditary breast cancer, alterations in these
enomic events that manipulate the development genes are rare and account for a relatively
4 Breast Cancer Genomics 55

small percentage of inherited breast cancer families, a method that has been successful in
possibility [5]. identifying many Mendelian disease genes.
About 100 genes for hereditary diseases show- However, this strategy has contributed little to the
ing Mendelian patterns of inheritance in families study of more common or “complex” forms of
are known [6]. These are consistently rare genes disease, mediated by genetic variants with
and associated with high relative risks. Most of reduced penetrance which may interact with
the genes have been identified through linkage environmental and other genetic factors. The
analysis of carefully selected families, followed complexity of this pattern of inheritance greatly
by positional cloning. Within this class are the reduces the power to detect genes through family-
breast cancer BRCA1 and BRCA2 genes, which based studies.
contain over 1,000 alterations. Genetic screening
for the spectrum of significant mutations in these
genes in high-risk families is well established. Rare Cancer Syndromes and Rare
The BRCA1 “breast cancer 1 early-onset” gene Moderate-Penetrance Breast Cancer
[7] is involved in susceptibility to breast and Genes
ovarian cancer at a young age, and tumors can
arise through somatic or germ line mutations. Garcia-Closas and Chanock have provided a
Impaired or lost BRCA1 function underlies sub- comprehensive review of recent work from large
stantial genome instability, including augments consortial studies that have led to the discovery
in the number of mutations, DNA breakage and of additional breast cancer susceptibility loci
chromatid exchanges, increased sensitivity to through candidate gene or whole-genome
DNA damage, and defects in cell cycle check- approaches [13]. These studies suggest that much
point functions. of the genetic component of breast cancer risk
The role of BRCA1 in the DNA damage remains uncharacterized and probably arises
response is that of “caretaker” or “master regula- from combinations of low-penetrance variants
tor” in the genome [8–10]. Jensen et al. isolated that, individually, might be quite common in the
the large protein encoded by the BRCA2 gene population. There are a number of syndromes
and showed it to be a key mediator of homolo- that include breast cancer as a component of the
gous recombination [11]. It is a crucial element disease phenotype. Rare to uncommon mutations
in the DNA repair process, which, if impaired in the PTEN [14] and STK11 [15] genes reason
through mutation, can lead to chromosome insta- Cowden and Peutz–Jeghers syndromes, respec-
bility and cancer. It is known to mediate recombi- tively, and both are associated with considerably
national DNA repair by promoting assembly of increased breast cancer risk [16]. The E-cadherin
RAD51 onto single-stranded DNA. This has a gene (CDH1) encodes a cellular adhesion protein
key role in catalyzing the invasion and exchange and is an influential tumor suppressor of breast
of homologous DNA sequences. Mutations in the cancer [17]. It is particularly implicated in inva-
BRCA2 gene may disrupt this mechanism and sive lobular breast carcinomas. RAD51C is
impair repair of DNA breaks, using homologous another gene involved in the recombinational
sequences from an intact homolog or sister chro- repair of double-stranded DNA breaks. Rare
matid, leading to errors in the repair process and germ line mutations have been shown to confer
chromosome instability. increased risks of breast and ovarian cancer [18].
BRCA1 and BRCA2 are likely to be the only Segregation in families follows Mendelian pat-
major high-penetrance genes underlying breast terns, and the disease phenotype resembles that
cancer. Germ line mutations in the TP53 gene of BRCA1 and BRCA2 mutation carriers.
cause Li–Fraumeni syndrome, a phenotype that There are also a number of gene mutations
includes early-onset breast cancer [12], but these associated with more moderate risks of breast
mutations are far rarer. Both BRCA1 and BRCA2 cancer, which show marked departures from
genes were identified using linkage mapping in Mendelian patterns of inheritance. As a result,
56 B. Kumar

segregation of disease with the mutation may be spots, the nascent polymorphisms are often
unhelpful to confirm relationship with disease. strongly correlated. GWAS have identified more
Genes in this class include germ line mutations in than 100 such low-penetrance loci involved in
the ataxia-telangiectasia (A-T) gene, which are cancer, including at least 17 related to breast can-
associated with increased risk (~2.2-fold) of cer (Table 4.1). These variants have allele fre-
breast cancer in carriers of heterozygous muta- quencies in the range 0.05–0.5, but they confer
tions, with apparently higher risks below the age only small increases in disease risk [24]. Because
of 50 years [19]. Other rare moderate-penetrance of the greatly reduced penetrance and strongly
genes include heterozygous mutations in BRIP1 non-Mendelian patterns of inheritance, there is
(encoding a BRCA1-interacting protein) that often considerable uncertainty about the exact
confers elevated risks of breast cancer and underlying genetic mutation. Not only are the
Fanconi anemia subtype FA-J for bi-allelic muta- most strongly associated SNPs unlikely to be the
tions. The partner and localizer of BRCA2 causal sites, but there also may be uncertainty
(PALB2) gene interacts with BRCA2, and mono- about the gene involved. It has also been sug-
allelic mutations are involved in familial breast gested that multiple rare variants create “syn-
cancer, conferring a 2.3-fold risk. Mutations in thetic association” signals in a GWAS if they
BRCA2 are also known to underlie Fanconi ane- occur more often in association with a common
mia (subtype FA-D1), and bi-allelic mutations of tag SNP. This implies that causal variants could
PALB2 underlie the very similar Fanconi anemia be many mega bases away from variants detected
subtype FA-N [20]. Rare variants in the cell cycle in GWAS [25], although this scenario appears to
checkpoint kinase 2 (CHEK2) genes are known be rare [26].
to underlie an approximately twofold increase in Perhaps one of the unexpected findings from
risk of breast cancer. Products of this gene are these studies is a greater-than-anticipated role for
involved in DNA damage repair, and mutations noncoding variants in common diseases [27].
are found in 1–2 % of unselected women with From the analysis of population sequences [28],
breast cancer [21]. 30 % of common variants associated with disease
are annotated as, or in linkage disequilibrium
with, non-synonymous (coding) variation. This
Common Low-Penetrance Breast supports the view that many of the common dis-
Cancer Genes ease variants have gene regulatory roles. Among
the set of well-established common susceptibility
In recent years the research of low-penetrance genes are variants in intron 2 of the FGFR2 gene
allelic variants was conducted mainly through [29], which, among the common variants, are
genome-wide association studies (GWAS). These likely to make one of the larger contributions to
studies use a large number of common genetic relative risk, at least for postmenopausal disease.
single-nucleotide polymorphisms (SNPs) to iden- Easton et al. [30] found that the rs2981582 SNP
tify associations with disease that rely upon pat- (allele frequency 0.38) contributes odds ratios of
terns of linkage disequilibrium (LD) in the human 1.23 and 1.63 for heterozygote and homozygote
genome [22]. The power of GWAS is to evaluate genotypes, respectively. The FGFR2 gene encodes
the association of genetic variants at different loci a fibroblast growth factor (FGF) receptor. FGFs
on different chromosomes (LD) in a large series and their corresponding receptors are involved in
of cases versus controls, analyzing a panel of regulation of the proliferation, survival, migra-
100,000 SNPs simultaneously, to identify new tion, and differentiation of cells. The considerable
alleles of susceptibility to breast cancer [23]. importance of FGF signaling in a range of tumor
In the human genome, it has been estimated types is now becoming recognized [31].
that there are seven million common SNPs that SNPs within intron 2 are involved in FGFR2
have a minor allele frequency (m.a.f.), 45 %, and upregulation, and aberrant signaling activation
because recombination occurs in different hot induces proliferation and survival of tumor cells
4 Breast Cancer Genomics 57

Table 4.1 Recognized breast tumor vulnerability genes and region

Recognized gene/region Position Mapped by Allele incidence Recognized/feasible role
BRCA1 17q21 Linkage Rare DNA repair/genome stability
BRCA2 13q13.1 Linkage Rare Recombinational repair
TP53 17p13.1 Linkage Rare Li–Fraumeni syndrome, apoptosis
ATM 11q22.3 CS Rare DNA repair
BRIP1 17q23.2 CS Rare DNA repair, associated with
CHEK2 22q12.1 CS Rare DNA repair/cell cycle
PALB2 16p12.2 CS Rare Associated with BRCA2
RAD51C 17q22 CS Rare Homologous recombination repair
PTEN 10q23.3 Linkage Rare Cowden disease, cell signaling
STK1 (LKB1) 19p13.3 Linkage Rare Peutz–Jeghers syndrome, cell
cycle arrest
CDHå1 16q22.1 Linkage Rare Intercellular adhesion: lobular BC
FGFR2 10q26 GWAS Common Fibroblast growth factor receptor
TOX3(TNRC9)//RBL2 16q12 GWAS Common Chromatin structure/cell cycle
MAP3K1 5q11.2 GWAS Common Cellular response to growth
LSP1 11p15.5 GWAS Common Neutrophil motility
8q24 8q24 GWAS Common Intergenic, enhancer of MYC
2q35 2q35 GWAS Common –
CASP8 2q33 GWAS Common Apoptosis
SLC4A7/NEK10? 3p24.1 GWAS Common Cell cycle control?
COX11/STXBP4? 17q22 GWAS Common Transport?
MRPS30? 5p12 GWAS Common Apoptosis?
NOTCH2/FCGR1B? 1p11.2 GWAS Common Signaling/immune response?
RAD51L1 14q24.1 GWAS Common Homologous recombination
CDKN2A/CDKN2B? 9p21 GWAS Common Cyclin-dependent kinase
MYEOV/CCNDL? 11q13 GWAS Common Cell cycle control/fibroblast
growth factors?
ZNF365? 10q21.2 GWAS Common Zinc finger protein gene
ANKRD16/FBXO18? 10p15.1 GWAS Common Helicase?
ZMIZ1? 10q22.3 GWAS Common Regulates transcription factors?
Notes: ? refers to “possible” gene or function in the breast cancer context. There is uncertainty about the exact genes
and their useful roles in breast cancer
Abbreviation: CS candidate resequencing, GWAS genome-wide association studies

[32]. The identification of this gene, which was [34]. The variant rs1045485 is protective, con-
unanticipated as a cancer gene, has prompted tributing odds ratios of 0.89 and 0.74 for hetero-
research into related genes and their potential zygotes and rare homozygotes, respectively.
roles in cancer. Other FGFs (e.g., FGF-8) appear Recently, variants in CASP8 have been shown to
to be involved in breast cancer cell growth alter risks (in a protective direction) in individu-
through stimulation of cell cycle and prevention als with a family history of breast cancer [35].
of cell death [33]. Breast tumors are classified according to whether
Other low-penetrance variants that have been they have receptor proteins that bind to estrogen
identified through GWAS include CASP8 and progesterone. Such cells are termed ER + and
(caspase 8), which encodes an apoptotic enzyme PR + and require estrogen and progesterone to
58 B. Kumar

grow. Conversely, ER − and PR − tumors lack the for ER+ [41]. Other breast cancer associations
protein that allows the hormones to bind. Tumor include signals on 3p24, potentially relating to the
classifications manipulate the choice of treatment genes SLC4A7 or NEK10, and on 17q22, perhaps
regimes for the patient. A further classification related to COX11. These SNPs contribute odds
arises through tumors that overexpress the human ratios of 1.11 and 0.97 for heterozygote and
epidermal growth factor receptor 2 (HER2) genes, homozygote genotypes, respectively [42].
which are termed HER2+ (conversely, HER2−). Additionally, a common variant close to MRPS30
The triple-negative subtypes are ER−, PR−, and on 5p12 was found to confer higher risk of
HER2− and are characterized by aggressive tumors ER + disease [43]. Turnbull et al. described five
and reduced range of effective treatment options. new associations on chromosomes 9, 10 (three
Several common gene variants are more strongly regions), and 11 [44]. Two further signals reported
associated with specific cancer subtypes. These by Thomas et al. [45] include a SNP in the peri-
include the TOX3 gene, formerly called TNRC9 in centromeric part of chromosome 1, within a
which variant rs3803662 contributes a 1.64-fold region containing NOTCH2 and FCGR1B, and a
homozygote risk, specifically in ER + cancer [36]. signal associated with another double-strand
This gene encodes a high-mobility group chromatin- break repair gene (RAD51L1) on 14q24.1. There
associated protein, and increased expression is is evidence that the chromosome 1 locus is more
implicated in bone metastasis [37]. strongly associated with ER + disease.
Fine mapping has shown that hypothesized Considerable additional follow-up investiga-
susceptibility variants lie in an intergenic region tion will be required to establish the relationships
consistent with a gene regulatory function [38]. between many of the SNPs and the actual causal
These authors note there remains uncertainty as to variant(s) and to further elucidate the role in dis-
whether the causal variant is actually involved in ease for many of these common genes.
the regulation of the nearby retinoblastoma-like
gene 2 (RBL2) gene, which is involved in cell
cycle regulation, given gene expression evidence. Genome-Wide Association Studies
The mitogen-activate protein kinase
(MAP3K1) breast cancer gene [30] is a member In modern years, the research of low-penetrance
of the Ras/Raf/MEK/ERK signaling pathway (as allelic variants was conducted mainly through
is FGFR2) and is involved in regulating transcrip- GWAS. These studies use a large number of com-
tion of a number of cancer genes. MAP3K1 has mon genetic SNPs to identify associations with
been found to be more strongly associated with disease that rely upon patterns of linkage disequi-
ER + and PR + tumors than ER−/PR − subtypes. librium in the human genome [22]. The power of
There is also a stronger association with HER2+ GWAS is to evaluate the association of genetic
tumors [39]. The LSP1 gene was identified as variants at different loci on different chromo-
breast cancer susceptibility locus by Easton et al. somes in large series of cases versus controls, ana-
[30] who identified a SNP within the intron as the lyzing a panel of 100,000 SNPs simultaneously,
most strongly associated. LSP1 encodes lympho- to identify new alleles of susceptibility to BC
cyte-specific protein 1, which is an F-actin bind- [23]. In the human genome, it has been estimated
ing cytoskeletal protein. The same study also that there are 7 million common SNPs that have a
identified a breast cancer variant in the 8q24 minor allele frequency (m.a.f.), 45 %, and because
region containing no known genes. This region is recombination occurs in different hot spots, the
also associated with prostate cancer [40]. Stacey nascent polymorphisms are often strongly corre-
et al. identified a SNP on 2q35, a region with no lated. These studies therefore supply a powerful
known genes, as associated with breast cancer in tool to recognize novel markers for susceptibility
Icelandic patients with ER + breast cancer [36]. and prognosis of disease [46–48].
Milne et al. also found an association with In the GWA studies, the accumulation of a
ER − disease, although there was a stronger signal large number of data is crucial. Houlston and
4 Breast Cancer Genomics 59

Peto have estimated the number of cases required additional studies in the Breast Cancer
to identify low-penetrance alleles conferring a Association Consortium. These combined analy-
relative risk of two both in an unselected popula- ses allowed the observation that the SNPs show-
tion and in families with first-degree relatives ing a stronger statistical evidence of association
affected [47]. In an unselected population, the with an increased familial risk were rs2981582 in
identification of a susceptibility allele with a fre- intron 2 of FGFR2, rs12443621 and rs8051542
quency of 5 % requires over 800 cases. In the within TNRC9, rs889312 in a region that con-
same population, the identification of a suscepti- tains the MAP3K1 gene, rs3817198 in intron 10
bility allele with a frequency of 1 % requires over of lymphocyte-specific protein 1 (LSP1), and
3,700 unselected cases, whereas about 700 would rs2107425 within the H19 gene.
be enough if three affected families are selected. In brief, GWAS analyze thousands of cases
Therefore, the power of association studies can and controls (huge numbers are necessary to reach
be significantly increased using selected cases sufficient statistical power) in order to compare
with a family history of cancer because fewer SNPs, which makes it possible to identify genetic
cases are required to demonstrate the association variants associated with disease risk. More than
with the disease [47]. 100 GWAS have related a slight increase in can-
The potential of the association studies of cer risk [51]. In breast cancer, at least 18 variants
cases with a family history to identify low- have been identified, with a 1.1- to 1.5-fold
penetrance alleles conferring a relative risk of 2 increase in risk. Interestingly, these SNPs might
has been demonstrated by the mutation CHEK2 not only provide useful information on the risk of
1100delC in patients with BC. This variant car- breast cancer, but they could also be linked to spe-
ried by 1 % of the population confers an increased cific molecular subtype of breast cancer, such as
risk of 1.7-fold. The frequency was not signifi- FGFR2-RS2981582 and TNRC9-RS3803662
cantly increased in unselected cases (1.4 %), but with positive estrogen receptor [52] and rs8170 in
it was strongly increased in familial cases with- chromosome 19p13 with negative estrogen recep-
out BRCA1 and BRCA2 mutations (5.1 %) [49]. tors [53]. However, the clinical impact of these
In the past several years, several novel risk approaches is not clear, and this could illustrate
alleles for BC were identified by four recent how techniques are developing faster than knowl-
GWA studies: Breast Cancer Association edge. For example, a 1.2-fold risk with one of
Consortium, Cancer Genetic Markers of these variants is comparable to that of delaying
Susceptibility, DeCode Islanda, and Memorial the age of the first pregnancy to more than 35
Sloan–Kettering Cancer Center [29, 30, 43, 50]. years [51]. In addition, these approaches give no
In each of them, the association study was shared information about other complex factors that can
in three phases: the first phase identifies the com- affect risk, and there are no data on the possible
mon SNPs in cases and controls, the second effects of combining two or more variants.
phase evaluates how many of the above SNPs are Consequently, large research consortia are essen-
common to a greater number of cases and con- tial if a high number of cases and controls are to
trols, and, finally, the third phase aims to identify be attained. Finally, direct access to some GWAS
new alleles of susceptibility to BC. Easton et al., via the Internet can generate anxiety and danger-
in their study, identified five independent loci ous misinterpretation of genetic risk.
associated with increased susceptibility to BC
(Po10_7) [30]. This multistage study involved in
the first stage 390 BC cases with a strong family Breast Cancer Genomics Based
history and 364 controls and 3,990 cases and on Biobanks
3,916 controls in the second stage. To define the
risk associated with the 30 most significant SNPs, Endeavors to discover genes’ contributions to
a third stage of the study was conducted involving intricate diseases, such as cancer, require new
21,860 cases and 22,578 controls from 22 study designs that incorporate an efficient use of
60 B. Kumar

population resources and modern genotyping genetic abnormalities in cancer, the general idea
technologies. There are two approaches used for being that tumor genotyping would be valuable in
the study of breast cancer, both of which incorpo- defining cancer phenotypes. In a previous study,
rate the use of biobanks. One uses a cancer regis- we showed that it was possible to delineate sub-
try as a source of case information, which is then sets of breast tumors according to specific combi-
linked to a biobank on blood DNA. The biobank nations of DNA amplifications [61]. The present
also provides samples from matched controls. work allowed us to extend the phenotypic descrip-
After genotyping, clinical data are retrieved from tion to prognostic significance. We show here that
hospital records, and the results can be presented some of the markers tested presented prognostic
for genotype-specific cancer risks or similarly for significance in specific subsets of patients. This
genotype-specific clinical and survival parame- was particularly evident for MDM2 amplification
ters. The second approach uses registered data on and p53 mutations, which showed a strong prog-
cancer in families or among twins. With defined nostic value in the N2 subset of patients, or for the
groups of patients, paraffin tissue is collected by amplification of CCND1, EMS1, and FGFR1 in
contacting the pathology departments of the hos- N1 patients. During the course of this study, we
pitals where the patients were diagnosed. Tumor also made some observations that suggest the
and healthy tissue are prepared and used for muta- existence of correlation clustering in other patient
tion, the loss of heterozygosity, or copy number subsets, such as MYC in patients under 50 years
analysis. In the era of whole-genome genotyping or MDM2 in ER1 patients (data not shown).
technologies, the importance of well-characterized Our data constitute an attempt to delineate
sample sets cannot be overemphasized. Samples, tumor subsets according to their genotypic speci-
rather than technologies, limit the rate of gene dis- ficity. Knowing the complexity of the genetic rear-
covery in complex diseases [54]. rangements in breast cancer, the nine events
studied here probably correspond to a small por-
tion of the genes involved in tumorigenesis.
Linking Genotype to Phenotype Genotyping of breast tumors will involve the anal-
in Breast Cancer ysis of an ever-larger number of parameters and
sorting of the significance of complex combina-
Women with germ line BRCA1 or BRCA2 muta- tions. Because different combinations of genes or
tions are estimated to have a 45–70 % risk of genetic anomalies may bear a meaning in different
breast cancer by age 70 years [55–58]. The iden- populations of patients, the analysis of specific
tification of BRCA1 and BRCA2 mutations was phenotypic subsets will be necessary, thus leading
a major step in personalizing breast cancer risk to an increase of the number of comparisons. This
assessment, screening, and risk reduction strate- will require the analysis of very large cohorts of
gies. Studies are ongoing to determine whether or patients (several thousand) and consequently the
not certain subgroups of BRCA mutation carriers use of high-throughput analytical methods [62] in
may be at a higher risk for breast cancer. It has association with statistical tools, especially devised
been proposed that certain BRCA mutations may for multiple-comparison analyses.
confer a differential risk of future breast cancer
development, suggesting an important genotype–
phenotype linking [59, 60]. Genomics Landscape of Breast
In a recent kin-cohort study in Ontario, Risch Cancer and Comprehensive Atlas
et al. observed a trend of increasing breast cancer of Breast Cancer Genomes
risk associated with increasing downstream loca- for Various Applications
tion of BRCA1 mutation with a continuous linear
trend and a 32 % increase in risk associated with A pilot genome-wide sequencing exertion on
each additional 10 % or 559 nucleotides of down- breast cancers identified a total of 1,137 somati-
stream distance [60]. Over the past few years, a cally mutated genes from 11 breast cancers, with
considerable effort has been made to characterize an average of 52 non-synonymous mutations per
4 Breast Cancer Genomics 61

sample. Using gene mutation incidence as the analysis integrating copy number and gene
principal criterion, 140 genes were identified as expression profiles of ~2,000 breast cancers
candidate cancer genes that require further assess- suggested a novel classification system [72].
ment to confirm their functions as causal contrib- A recent multiplatform study on hundreds of
utors to tumorigenesis [63, 64]. These studies also breast cancers revealed subtype-specific pattern
portrayed the genomic landscape of human breast in numerous tumor characteristics including
cancer that consists of a few repeatedly mutated gene mutations, microRNA expression, DNA
gene “mountains” and a huge number of rarely methylation, copy number changes, and protein
mutated (usually <5 %) gene “hills” [64]. expression. Moreover, in whole-exome sequenc-
Besides the detection of novel candidate genes, ing of more than 500 tumors, this study also
more recent studies have delineated new aspects revealed almost all repeatedly altered pathways
of breast cancer exomes. Interrogating luminal- (PI3K/AKT, TP53, RB) in breast cancer [73].
type breast cancer genomes with clinical data The quickly evolving sequencing technolo-
revealed that somatic mutations in TP53 signaling gies generate massive genomic data at an increas-
pathway, DNA replication, and mismatch repair ing rate with reduced cost. In recent years in
are associated with aromatase inhibitor resistance particular, large-scale analyses of cancer genomes
[65]. Determination of clonal frequencies by deep have produced a prosperity of information, which
sequencing provided new insights into the initiat- greatly expanded our knowledge on human breast
ing events of TNBCs [66]. Massively parallel cancer, summarized in Table 4.2. The launch of
paired-end sequencing technologies enable comprehensive cancer genome projects including
whole-genome detection of gene rearrangements the Cancer Genome Project (CGP) [78], the
at the DNA sequence level [67]. An analysis on Cancer Genome Atlas (TCGA) [79], and the
24 breast cancers revealed more than 2,000 gene International Cancer Genome Consortium
rearrangements, enriched with tandem duplica- (ICGC) [80] facilitates the compilation of an
tions [68]. Analysis of breast cancers across a encyclopedic catalogue of the genomic changes
variety of subtypes revealed that luminal B and involved in cancer. Whole-genome sequencing
HER2-enriched breast tumors harbor many more studies enable observation of genetic alterations
structural rearrangements when compared to the earlier undetectable by protein-coding sequence
luminal A subtype. However, no repeatedly recur- screens, including mutations in noncoding
rent rearrangements have been discovered in regions and large rearrangements. Primary breast
breast cancer by earlier studies except for the cancers were reported to harbor ~7,000–10,000
MAGI3–AKT3 gene fusion detected in 4 % (9 out somatic point mutations per genome [65, 69, 81]
of 257) of breast cancers [69]. in which tens to hundreds reside in the protein-
Like all cancer types, breast cancer progres- coding regions [63, 64, 69, 77], as well as up to
sion is thought to be a dynamic multistep hundreds (average 20–50) of somatic structural
Darwinian evolution process. Independent muta- variants [65, 68, 69, 75, 81]. The minimum num-
tions arise in a stepwise fashion, of which those ber of mutations necessary for tumorigenesis has
conferring selective advantages promote cell been estimated to be around 5–6, according to
proliferation and clonal expansion [70]. Through incidence modeling of solid tumors such as breast
deep whole-genome sequencing of 21 breast and colorectal cancers, and this number would be
cancers and analysis of subclonal genetic smaller in leukemia and childhood cancers [82].
alterations, Nik-Zainal et al. proposed a model However, recent systematic mutational screens of
for clonal evolution that many molecular cancer genomes suggested a higher number of
aberrations accumulate in dormant cell lineages causal gene mutations in each tumor (range
before final expansion of the most recent com- 10–20 genes) [63, 83].
mon ancestor, which triggers diagnosis [71]. Distinguishing the driver mutations from pas-
Integrative breast cancer studies aim at develop- sengers cannot be accomplished by analyzing
ing new definition of breast cancer subtypes with genetic data alone but requires functional valida-
better prognostic and predictive values. A cluster tion of the cancer-relevant activities. Since most
62 B. Kumar

Table 4.2 Summary of genome sequencing studies of breast cancer

Study Cancer type Sequencing target T/N pairsa Findings
Stephens et al. [74] Breast Protein kinome (518 25 Identified diverse patterns of somatic
genes) mutations in breast cancer
Wood et al. [64] Breast, All RefSeqGene 11 + 24 per The first sequencing effort of all
Colorectal (18,191 genes) tumor type coding regions in cancer genomes.
Identified 280 candidate genes and
revealed the mutation landscape of
breast and colorectal cancer genomes
Shah et al. [70] Breast Whole genome whole 1b Demonstrated single-nucleotide
(metastasis) transcriptome mutational heterogeneity and
mutational evolution in breast tumor
Stephens et al. [68] Breast Whole genome 24 The first genomic screen for somatic
rearrangements in tumor samples.
Revealed the genome landscape of
somatic rearrangements in breast
Ding et al. [75] Breast Whole genome 1c Indicated that metastasis may arise
from a minority of cells within the
primary tumor
Edgren et al. [76] Breast Whole transcriptome 4 Discovered novel fusion genes (e.g.,
VAPB-IKZF3) with potential
functional role in breast cancer
Nik-Zainal et al. [71] Breast Whole genome 21 Identified distinct nucleotide
substitution signatures, observed
localized hypermutation, and
constructed a model of breast cancer
Ellis et al. [65] Breast Whole genome 77 + 240 Identified novel significantly mutated
(n = 46), genes (e.g., GATA3, TBX3, ATR,
whole exome (n = 31) RUNX1, LDRAP1, STMN2, AGTR2,
SF3B1) in luminal breast cancer and
revealed pathways (e.g., TP53, DNA
replication, MMR) associated with
aromatase inhibitor response
Shah et al. [66] Breast Whole genome 65 Revealed mutations and structural
(n = 15), alterations with clonal frequency and
whole exome (n = 54) suggested involvement of cytoskeletal
gene mutations in breast cancer
Stephen et al. [77] Breast Whole exome 100 + 250 Revealed multiple mutation
signatures of breast cancers and
identified novel driver mutations
(e.g., AKT2)
Banerji et al. [69] Breast Whole genome 108 + 235 Identified novel recurrent mutations
(n = 22), in CBFB and a recurrent fusion gene
whole exome (n = 130) MAGI3–AKT3
TCGA [73] Breast Whole exome 507 Revealed molecular subtype-specific
patterns of mutations and identified
novel candidate genes
T/N pairs, patient-matched tumor/normal pairs investigated. In some cases, numbers of T/N pairs in discovery screen
and validation screen are indicated before and after the plus sign
DNA from primary tumor and metastasis was analyzed in this study
DNA from blood, primary tumor, metastasis, and xenograft was analyzed in this study
4 Breast Cancer Genomics 63

functional assays are relatively labor- and time- biological studies can fully credential a candidate
intensive, prioritization of the genes for functional gene as a bonafide cancer gene. It is clear from
studies presents a great challenge in cancer cancer genome resequencing efforts that not all
genomic data interpretation. Several measure- cancer genes are mutated at high prevalence. On
ments have been adopted to identify the most the contrary, despite conferring selective advan-
promising driver mutations. First, analyzing the tage, the vast majorities of cancer genes are not
ratio of non-synonymous mutations to synony- frequently mutated and are therefore difficult to
mous mutations of a given gene would indicate identify through sequencing of a limited number
whether the mutations have been under positive of samples. In order to discover these infrequent
selection during tumor development, thus a higher driver mutations, systematic screens of large
than expected ratio always suggests driver muta- cohorts of patients are required. For example, it
tion [84, 85]. Second, assessment of the mutation was estimated that 500 tumor samples of a par-
prevalence in genes also identifies drivers that ticular tumor type are needed in whole-exome
contribute to cancer if they are highly unlikely to sequencing studies to get an ~80 % detection
be mutated by chance [63]. Third, several tools power of genes with ~3 % true mutation fre-
have been employed to predict the effect of non- quency [80].
synonymous single nucleotide variants on protein
function based on phylogenetic conservation and
physical considerations (e.g., Sorting Intolerant Classification of Breast Cancer
From Tolerant (SIFT) [86], Polymorphism Based on Genome Profile
Phenotyping (PolyPhen) [87], Panther [88],
MutationTaster [89], etc.). Last but not least, as Gene expression profiling has proven to be a use-
the number of pathways involved in cancer is ful and reliable tool for classifying breast cancers
much smaller than that of cancer genes and a vari- into subgroups that reflect different histopatho-
ety of mutations in multiple cancer genes from the logical characteristics as well as differential
same pathway would likely to have similar patho- prognostic outcome. It has been suggested that
logical effects [90], evaluation of the combined estrogen receptor-negative and estrogen receptor-
prevalence of somatic alterations at the pathway positive breast cancers can be subdivided into
level provides strategies for identification of can- Her-2 positive basal-epithelial-like, normal
cer-associated processes [91, 92]. breast-like, and luminal-like [95]. The potentially
In the past few years, comprehensive mutation different origins of the tumor cells may signify
interpretation implementing most, if not all, of distinct pathways of tumorigenesis and differ-
these measurements has been introduced into ences in the clinical course of the disease. Germ
cancer genome analyses. For example, Carter and line mutations in the BRCA1 and BRCA2 genes
her colleagues developed a computational pipe- together account for a significant portion of
line for cancer-specific high-throughput annota- hereditary breast cancers. They have been shown
tion of somatic mutations (CHASM), which to leave a characteristic imprint on the panel of
takes a total of 49 predictive features into account genes expressed by the tumors [96], with
for driver identification [93]. Another example is BRCA1-dependent tumors exhibiting a tran-
a package for determination of mutational sig- scriptional profile similar to the basal subtype of
nificance in cancer (MuSiC), designed by Dees tumors [97]. These findings suggest that the cel-
et al. MuSiC is the first software suite that inte- lular origin of BRCA1- and BRCA2-mutation-
grates clinical data with coverage data and data- positive tumors may differ or that these tumors
base references to identify drivers from large traverse down separate pathways in their progres-
mutational discovery sets [94]. sion toward malignancy [96]. Furthermore, the
Although many tools can help to prioritize the molecular subclassification of non-BRCA1/2
candidates of interest for downstream analyses, familial breast cancers into homogeneous sub-
only the evidence from functional assays and groups underscores the potential differences in
64 B. Kumar

cellular origin and/or disease progression due to 113]. This frequently affects the activity in
the presence of multiple diverse underlying oncogenes and tumor suppressor genes, and in
genetic alterations, which is reflected in the phe- this way CNAs are important for the carcino-
notype of the tumors [98]. genic process.
The diversity of breast cancer has been CNAs in tumors are a result of deregulated cell
acknowledged for decades, but recent technologi- cycle control and of DNA maintenance and repair
cal advances in molecular biology have given [114]. Different patterns of copy number altera-
detailed knowledge on how extensive this hetero- tions have been identified with distinct differ-
geneity really is. Traditional classification based ences; simplex profiles are characterized by few
on morphology has given limited clinical value, alterations and complex genomic profiles have
mostly because the majority of breast carcinomas extensive changes [115]. Complex genomic rear-
are classified as invasive ductal carcinomas, which rangements are areas with high-level amplifica-
show a highly variable response to therapy and tions and have prognostic value in breast cancer
outcome [99]. The first molecular subclassifica- even when they do not harbor known oncogenes,
tion with a major impact on breast cancer research suggesting that the phenotype of defect DNA
was proposed by Perou and colleagues where the repair may be associated with more aggressive
tumors were subdivided according to their pattern disease [115, 116]. Alterations in the expression
of gene expression [95, 100]. Five groups were pattern are caused by changes at the genomic
identified and named luminal A, luminal B, basal- level, and a robust classification of breast cancer
like, normal-like, and HER-2-enriched subgroups. for clinical use should probably take these more
These intrinsic subgroups have been shown to be into account. Changes at the genomic level
different in terms of biology, survival, and recur- include point mutations, changes in copy number,
rence rate [95, 97]. The molecular subgroups have and epigenetic events. These are characteristics
been extended to also include a sixth subgroup, that enable and drive carcinogenesis together with
which has been named the claudin-low group, tumor-promoted inflammation [117].
based on its low expression level of tight junction
genes (the claudin genes) [101].
Different methods for the assignment of indi- Risk Assessment and Prognosis
vidual tumors to its molecular subgroup are pro- by Genetic Test: Market Players
posed, each based on the expression levels of in Genomics-Based Personalized
different sets of genes [97, 102, 103]. The agree- Diagnosis, Prognosis, and Therapy
ment between methods on how to classify indi-
vidual tumors are not optimal, and how to Risk Assessment
establish more robust single sample predictors is
actively debated [104–107]. Hereditary alterations have been associated with
Aneuploidy is the presence of an abnormal 10–15 % of all breast cancer cases; however, dis-
number of parts of or whole chromosomes and is ease etiology in the majority of women appears
one feature that clearly separates cancer cells to be sporadic, lacking a momentous family his-
from normal cells. This was proposed as being tory. Because sporadic breast cancer may be
important in cancer nearly a century ago by influenced by a number of lifestyle and environ-
Theodor Boveri. With array-based comparative mental factors as well as common low-risk vari-
genomic hybridization (aCGH), a genome-wide ant in a number of genes, numerous models have
profile of the copy number alterations in the been developed in an endeavor to quantify indi-
tumor can be obtained. These patterns are related vidualized breast cancer risk:
to the molecular subtypes with distinct differ- • The Gail model measures risk based on patient
ences in the number of alterations between the age, age at menarche, number of prior breast
subtypes [108–111]. These copy number biopsies, age at first live birth, and number of
alterations (CNAs) alter the dosage of genes and first-degree relatives affected by breast cancer
highly influence the level of expression [112, [118].
4 Breast Cancer Genomics 65

• The Claus model estimates risk based on the DNA variants in genes from molecular path-
number of affected relatives and their respec- ways believed to be involved in breast cancer
tive age at diagnosis [119]. development between cases with disease and
• The BRCAPRO model calculates risk of healthy matched controls. An association study
developing breast cancer based on the proba- using candidate genes recently identified cas-
bility of carrying a BRCA1 or BRCA2 muta- pase 8 (CASP8) as a low-risk susceptibility
tion [120]. gene where the major (H) allele of the D302H
These models have been widely used to pre- polymorphism had a protective effect on the
dict risk and direct patient care, but each model development of breast cancer [125]. Despite
has limitations, because no model accounts for success in identifying CASP8, candidate gene
the spectrum of risk factors influencing breast approaches have not been widely successful in
cancer. For example, the Gail model considers identifying additional breast cancer suscepti-
only first-degree relatives without regard to age bility genes [126].
at diagnosis or presence of ovarian cancer, thus
potentially underestimating genetic risk. The
Claus and BRCAPRO models only consider Whole-Genome Approaches
family history, potentially underestimating risk
in women with other risk factors [121]. In addi- Candidate gene approaches are rapidly giving
tion, these models were developed 10–20 years way to genome-wide association studies
ago, when incidence of breast cancer in the gen- (GWAS), which evaluate a dense array of genetic
eral population was lower than it is today, and markers representing common variation through-
use of lower baseline risk estimates may con- out the genome. Completion of the human
tribute to an underestimation of recent risk genome sequence and subsequent identification
[122]. More recent models, such as the Tyrer– of single-nucleotide polymorphisms (SNPs) now
Cuzick model, utilize family history, endoge- permits millions of informative SNPs across the
nous estrogen exposure, and presence of benign genome to be assayed simultaneously. GWAS are
disease to model breast cancer risk [123], but useful for mapping genes of interest to small,
contributions from other factors such as mam- localized regions of the genome and for detecting
mographic breast density, weight gain, steroid the effects of common (>5 % minor allele fre-
hormone levels, and susceptibility genes have quency) alleles on disease risk [127]. Moreover,
not been incorporated [124]. GWAS are performed without a priori knowledge
The discovery of the BRCA1 and BRCA2 of the underlying genetic defect(s), which may be
genes advanced risk assessment in families advantageous since many genes identified
affected by hereditary breast and ovarian cancer, through whole genome approaches were not pre-
but identification of molecular markers associ- viously suspected to influence the disease under
ated with increased breast cancer risk in patients investigation [128].
without a family history of breast cancer has Recent GWAS have identified a number of
remained far more challenging. Without a strong loci that appear to be associated with breast can-
family history, linkage approaches involving cer susceptibility. For example, the fibroblast
large pedigrees such as those used to identify growth factor receptor 2 (FGFR2), mitogen-
BRCA1 and BRCA2 are not applicable. activated protein kinase kinase kinase 1
Sporadic breast cancer is not usually associated (MAP3K1), lymphocyte-specific protein 1
with other cancers, such as ovarian or male (LSP1), and trinucleotide repeat-containing 9
breast cancer, and unlike BRCA1-positive carci- (TNRC9/LOC643714) genes, along with a 110-
nomas, which exhibit specific histological char- kb region of chromosome 8q24, have been asso-
acteristics, sporadic breast cancer cases ciated with breast cancer in large studies
comprise a vast array of phenotypes. Early involving thousands of subjects [29, 30].
approaches to identify sporadic breast cancer Associations with other chromosomal regions—
susceptibility genes compared the frequency of 2q35, 5p12, 6q22, and 16q12—also have been
66 B. Kumar

Table 4.3 Leading direct-to-consumer genetic testing companies

Company Genetic Breast tumor
Headquarters Website Price (USD) counseling susceptibility variants
23andMe Mountain View, www.23andme.com $399 No 2 SNPS
deCODEme Reykjavik, www.decodeme.com $985a Yes 11 variantsb
Knome Cambridge, MA www.knome.com Customc Yes DNA sequence
Navigenics Foster City, CA www.navigenics.com $999d Yes Unknown
Complete scan
For women of European descent
KnomeSELECT™ is $24,500 for complete sequence of 20,000 genes; KnomeCOMPLETE™ is $99,500 for complete
genome sequence
Option for ongoing subscription ($199 per year) for updates

reported [36, 43, 50]. Further analysis has shown patients. However, genetic analysis and risk
that allelic variation at FGFR2, TNRC9, 8q24, assessment are available commercially through
2q35, and 5p12 is associated with physiological direct-to-consumer (DTC) testing. A number of
characteristics of breast tumors, such as ER sta- for-profit companies offer personal genetic infor-
tus [36, 43, 129], and specific FGFR2, MAP3K1, mation based on DTC tests—the largest and most
and TNRC9 variants may interact with BRCA1 recognized companies include 23andMe, deCO-
and BRCA2 mutations to increase breast cancer DEme, Navigenics®, and Knome®, Inc.
risk [130]. (Table 4.3). For a fee of $99 to $99,500 consum-
Despite recent success in identifying genetic ers provide a blood, buccal, or saliva sample for
determinants of breast cancer, susceptibility targeted SNP analysis or whole-genome sequenc-
alleles identified through GWAS are believed to ing. Genetic information provided to the con-
account for only ~5 % of breast cancer risk [131]. sumer varies greatly among companies, from
If future studies are to be successful in identify- trivial facts such as earwax type and ancestry
ing additional low-risk susceptibility alleles and information to information on risk for disease
low-frequency, highly penetrant variants [132], [134, 135]. Although DTC tests epitomize “per-
interactions between genes and environmental sonalized genomics” by providing consumers
exposures must be assessed [133], and methods with individual genotypes, critics note that the
must be developed to evaluate mechanisms by clinical utility of such tests is limited and often
which DNA variants in intronic or intergenic incongruent with marketing claims. Because
regions contribute to disease. As risk associated information on family history and environmental
with susceptibility alleles may vary between exposures is usually not accounted for, DTC risk
racial/ethnic populations due to differences in estimates may not be sufficiently accurate to
frequency, patterns of disequilibrium, and inter- enable consumers to make appropriate medical
actions with environmental factors [5, 30, 36], decisions [136, 137].
sufficiently powered genetic studies in women The majority of genetic risk assessments
from various ethnic groups are needed to improve developed thus far focus on DNA variants;
risk reduction strategies for all women. however, a new RNA-based signature has been
developed for noninvasive breast cancer screening
using peripheral blood samples. Although based
Direct-to-Consumer Testing on a small number of cases (n = 24) and controls
(n = 32), a subset of 37 genes in the assay correctly
New susceptibility variants identified by GWAS classified 82 % of patients [138]. Despite a
have not yet been incorporated into genetic tests relatively high misclassification rate, DiaGenic
with beneficial clinical utility for breast cancer (www.diagenic.no) has since developed this gene
4 Breast Cancer Genomics 67

expression signature into a clinical screening survival (patients with negative nodes) from those
tool, currently available only in India as BCtect™ who are likely to have <70 % survival (women
India. with nodal metastasis) [149]. Although these clin-
ical attributes are currently the standard of care
for breast cancer patients, many are imprecise in
Breast Cancer Personalized their ability to accurately predict outcomes.
Pathological Characterization of Breast Molecular markers have the prospective to pro-
Cancer vide additional prognostic information to supple-
Human breast carcinomas exhibit diverse patho- ment traditional pathological assessments for
logical characteristics that are associated with disease management in breast cancer patients. As
different clinical outcomes and thus are routinely mentioned above, traditional immunohistochem-
used to guide treatment options. Accordingly, an istry (IHC) markers routinely used in the classifi-
accurate definition of prognosis is dependent on cation of breast cancer include ER, PR, and
the ability to detect and quantify differences in HER2. Tumors positive for ER and PR expres-
tumor attributes, such as rates of proliferation sion frequently have low cellular proliferation
and propensity to metastasize. Routine tumor rates, tend to exhibit lower histological grade,
evaluation currently includes (1) histopathologi- and are associated with more favorable prognosis
cal classification; (2) grade determination; and [150]. ER and PR expression also is useful for
(3) quantification of tumor size, surgical margin identifying patients who will likely benefit from
status, and lymph node involvement. hormonal therapy, as women with ER- and
Histopathological characterization, based on PR-negative breast cancer do not gain a survival
microscopic cellular morphology, classifies breast benefit from antiestrogen tamoxifen [151].
carcinomas into common subtypes (ductal or lobu- The HER2 gene is a member of the epidermal
lar carcinoma), which tend to have similar progno- growth factor receptor family with tyrosine
ses [139]; or less common forms such as mucinous, kinase activity and is amplified at the DNA level
tubular, and papillary (favorable prognosis) [140]; and/or overexpressed in 15–25 % of breast can-
or inflammatory breast cancer (poor prognosis) cers. Carcinomas with amplified/overexpressed
[141]. Increasing tumor size has long been associ- HER2 exhibit high histological grade and usually
ated with poor prognosis [142], but improved have a poor prognosis [152, 153]. Some patients
mammographic detection of smaller tumors has with positive HER2 status (15–20 %) are eligible
decreased the prognostic utility of tumor size to receive trastuzumab, a monoclonal antibody
[143]. Presence of positive surgical margins has targeting HER2, in combination with standard
been associated with local recurrence, but only chemotherapy [154].
27 % of patients with extensively positive margins Rigorous clinical studies have shown that
will have recurrent disease [144, 145]. evaluating ER, PR, and HER2 status provides
Likewise, the Nottingham Histological Score, additional prognostic information beyond that
widely used for assessing histological grade, is normally achieved by histological assessment
clinically useful for stratifying patients into low- alone. For example, breast carcinomas that are
risk (low-grade disease, 95 % 5-year survival) and ER negative and PR negative and do not have
high-risk (high-grade disease, 50 % 5-year sur- HER2 overexpressed (triple negative) are marked
vival) groups [146, 147], but the reliability of by aggressive behavior, but because women with
breast tumor grade in predicting survival is ham- triple-negative disease are not eligible for tamox-
pered by subjectivity associated with its assess- ifen or trastuzumab treatment, they usually have
ment [148]. Axillary lymph node status is the relatively low long-term survival [155]. Other
most reliable predictor of survival, differentiating markers such as nuclear antigen Ki67 are not rou-
women who are likely to have >90 % 5-year tinely used to guide treatment selection, but hold
68 B. Kumar

Table 4.4 Selected molecular diagnostic tests for breast cancer

Test Number of Study
Company Assay type# genes/proteins Classification
Breast Bioclassifier™ University qRT-PCR 55 Tumor subtype Perou et al. [100]
genomics Therapeutic
MammaPrint™ Agendia Microarray 70 Prognostic van’t Veer et al.
Therapeutic van de Vijver et al.
guidance [161]
MammoStrat® Applied Genomics IHC 5 Prognostic Ring et al. [162]
MapQuant DX™ Ipsogen Microarray 97 Tumor grade Sotiriou et al. [163]
Loi et al. [164]
Oncotype DX™ Genomic Health qRT-PCR 21 Prognostic Paik et al. [165]
Therapeutic Paik et al. [166]
Rotterdam signature Veridex Microarray 76 Prognostic Wang et al. [167]
qRT-PCR quantitative real-time PCR, IHC immunohistochemistry

great promise for monitoring the effectiveness of uses conventional paraffin-embedded tissue to
neoadjuvant chemotherapy and predicting assay five markers by IHC:
recurrence-free survival [156–158]. • Tumor protein p53 (TP53)—known to play a
Individual estimates of outcome using clinical central role in cell cycle regulation
and pathological characteristics of breast tumors, • HpaII tiny fragments locus 9C (HTF9C)—
including age, menopausal status, comorbid con- involved in DNA replication and cell cycle
ditions, tumor size, number of positive lymph control
nodes, and ER status, have been incorporated • Carcinoembryonic antigen-related cell adhe-
into a computer program, Adjuvant! Online sion molecule 5 (CEACAM5)—aberrantly
(www.adjuvantonline.com/index.jsp), which is expressed in some cancers
available over the Internet as a decision aid for • N-myc downstream-regulated gene 1
patients and their physicians [159]. The program (NDRG1)—may function as a signaling protein
estimates the efficacy of endocrine therapy and in growth arrest and cellular differentiation
chemotherapy as well as overall and disease-free • Solute carrier family 7 (cationic amino acid
survival in a user-friendly format that effectively transporter, y + system), member 5
brings patients into the decision-making process (SLC7A5)—mediates amino acid transport
regarding personalized treatments. The MammosStrat® test may have utility for
Although IHC analysis of ER, PR, and HER2 predicting patient outcomes but currently requires
is widely used in the pathological evaluation of five separate slides (one slide per antibody),
breast tumors, additional molecular signatures which has the potential to show variability in
involving multiple genes and/or proteins are des- staining intensity and scoring between patients.
perately needed to more accurately classify
tumors and guide treatment selection. Recently, a Gene Expression Signatures
multigene IHC-based test known as MammoStrat® and Disease Risk
(Applied Genomics, Huntsville, AL; www. Molecular profiles are now being used more fre-
applied-genomics.com/mammostrat.html) was quently as clinical tools to determine treatment
developed to classify breast cancer patients into for certain groups of patients by categorizing
low-, moderate-, or high-risk categories for dis- them into low-risk and high-risk groups. The
ease recurrence (Table 4.4) [162]. MammoStrat® MammaPrint™ assay (Agendia, Amsterdam,
4 Breast Cancer Genomics 69

The Netherlands; www.agendia.com) is a 14.3 % (95 % CI 8.3–20.3); and high risk, 30.5 %
70-gene signature developed using tumor tissue (95 % CI 23.6–37.4). Recurrence scores also cor-
from young women (<55 years of age) with node- related significantly with relapse-free interval
negative disease, who either developed distant and overall survival [165]. In a subsequent study,
metastasis or remained disease-free after 5 years Oncotype DX™ was used to assess the benefit of
[160]. Overall 10-year survival for the “poor- adjuvant chemotherapy in ER-positive, node-
prognosis” signature is ~55 %, while 10-year sur- negative patients. Because the highest benefit
vival in women with the “good-prognosis” was observed in patients with high-risk scores,
signature is 95 %. The probability of being free while women with low-risk recurrence scores did
from distant metastasis after 10 years is 51 % for not benefit from chemotherapy [166], Oncotype
the poor prognosis and 85 % for the good prog- DX™ may be useful in guiding treatment options
nosis profile [161]. in ER-positive, node-negative patients.
A second group of researchers subsequently Clinical trials of the MammaPrint™ and
developed a 76-gene profile (Rotterdam signa- Oncotype DX™ assays are currently in progress.
ture) that could identify breast cancer patients at In the Microarray In Node negative Disease may
high risk for distant recurrence. The signature Avoid ChemoTherapy (MINDACT) trial, 6,000
could identify patients who developed distant node-negative women will be assigned to treat-
metastases within 5 years when traditional prog- ment groups based on risk stratification by tradi-
nostic factors were considered (hazard ratio 5.55, tional clinical–pathological factors (Adjuvant!
95 % CI 2.46–12.5) and could predict metastasis Online) and the MammaPrint™ molecular signa-
in both premenopausal and postmenopausal ture [168]. Patients classified as low risk by both
patients [167]. methods will not receive chemotherapy, while
The gene expression signatures outlined above those considered high risk for relapse by both
were refined from global expression profiling methods will be given the opportunity to receive
experiments involving thousands of genes and adjuvant chemotherapy. Patients of primary inter-
flash-frozen tumor specimens. An alternative est, those with discordant results, will be ran-
approach relied on an extensive literature search domized to treatment based on either Adjuvant!
to identify candidate genes (n = 250) believed to Online or MammaPrint™ to determine which
be involved in disease development based on test is more effective in defining treatment in
known function. Gene expression levels were node-negative patients.
assayed in 447 patients with ER-positive, node- The Trial Assigning Individualized Options
negative breast cancer to identify a small subset for Treatment (TAILORx) is examining whether
of 16 genes (plus five reference genes) amenable hormone receptor-positive patients with an inter-
to analysis by real-time PCR (RT-PCR) on RNA mediate Oncotype DX™ risk recurrence score
isolated from formalin-fixed, paraffin-embedded benefit from chemotherapy. The trial is recruiting
(FFPE) specimens. The resulting 21-gene signa- 10,000 hormone receptor-positive patients with
ture, known as Oncotype DX® (Genomic Health, HER2-negative and lymph-node-negative dis-
Redwood, CA; www.genomichealth.com/), pro- ease. Treatment will be based on the risk recur-
vides a probability of recurrence score for women rence score as follows: <10, hormone therapy
with early-stage (stage I or II), ER-positive, alone; >26, hormone and chemotherapy; and
node-negative breast cancer and categorizes intermediate scores, randomization to either hor-
patients as low, intermediate, or high risk. mone therapy alone or to hormone therapy and
In validation studies using patients from the chemotherapy. The goal is to integrate Oncotype
National Surgical Adjuvant Breast and Bowel DX™ into the clinical decision-making process
Project (NSABP) clinical trial B-14 who received and refine the utility of the assay in clinical prac-
tamoxifen, the probability of distant recurrence at tice [169].
10 years for the three risk categories was low Molecular signatures have improved the abil-
risk, 6.8 % (95 % CI 4.0–9.6); intermediate risk, ity to predict outcome and identify breast cancer
70 B. Kumar

Table 4.5 Selected genetic polymorphisms affecting response to therapy in breast cancer patients
Treatment Response to Study
Gene Variant Functional change treatment
Doxorubicin CBR3 11G > A Decreased enzyme Hematological Fan et al. [174]
activity toxicity
Anthracyclines MnSOD Ala16 Higher levels of Decreased Ambrosone et al.
reactive oxygen mortality [176]
MPO —463GG Higher levels of Decreased Ambrosone et al.
reactive oxygen mortality [176]
GSTP1 313A > G Altered drug Hematological Zárate et al. [177]
transport toxicity
MTHFR 1298A > C Altered drug Non-hematological Zárate et al. [177]
metabolism toxicity
Endocrine therapy
Tamoxifen CYP2D6 *3, *4, *5, Reduced function/ Poor clinical Schroth et al.
*10, *41 nonfunctional outcome [178]
enzyme Goetz et al. [179]
Aromatase CYP19A1 Cys264, Thr364 Decreased enzyme Reduced benefit Ma et al. [180]
inhibitors activity
TP53 Arg72Pro, Decreased Risk of Chang-Claude
PIN3 apoptosis telangiectasia et al. [181].
Targeted therapy
Trastuzumab HER2 Heterodimer Prevents disruption Poor response to Lee-Hoeflich et al.
by trastuzumab treatment [182]
*The different variant alleles of CYP2D6, in this case tamoxifen treatment, is related to variants 3, 4, 5, 10, and 41 of

patients who would most likely benefit from sys- been properly validated and proven to be clini-
temic therapy, thus providing an additional layer cally informative; thus, the degree to which
of personalized medicine. However, no current expression-based tests will alter the course of
molecular signature is 100 % accurate, and patient treatment remains unclear [172, 173].
5–10 % of patients now classified as low risk are
likely to relapse. Furthermore, current classifica-
tion systems were developed to predict only short- Pharmacogenomics of Breast Cancer
term (<5 years) outcomes; thus, there is a need to
develop signatures that identify patients with pro- Pharmacogenomics in breast cancer assesses the
tracted disease progression who may benefit from effect of inherited genomic variation on patient
prolonged therapy [170]. Although outcome pre- response or resistance to treatment. Genetic vari-
diction tends to be similar between gene expres- ability is commonly measured at the DNA level
sion signatures, overlap among genes comprising in the form of chromosomal alterations or DNA
the signatures is relatively low, suggesting that sequence variants (Table 4.5). Conversely,
these profiles assess common biological pathways somatic genomic changes (DNA variants and
but have not identified the actual genes driving gene expression profiles) in breast tumors can
tumor behavior and outcome [171]. influence rates of apoptosis, cell proliferation,
Finally, some multigene predictor assays are and DNA damage repair, which may have direct
being adopted and marketed before they have effects on response to treatment and survival.
4 Breast Cancer Genomics 71

To be most effective, personalized medicine must inhibitors that block the production of estrogen
incorporate information from innate genetic vari- and compounds such as fulvestrant (Faslodex®)
ation as well as somatic mutations in diseased that downregulate and degrade the ER protein.
tissue [183]. Aromatase inhibitors such as anastrozole
(Arimidex®), letrozole (Femara®), and exemes-
Endocrine Therapy tane (Aromasin®) target cytochrome P450 19
Estrogens play an important role in the etiology (CYP19A1 or aromatase), an enzyme involved in
of breast cancer by stimulating growth and prolif- estrogen synthesis in peripheral organs.
eration of ductal epithelial cells in the breast; Premenopausal women with functional ovaries
thus, the status of the estrogen receptor in breast do not receive aromatase inhibitor therapy
carcinomas provided one of the earliest avenues because first- and second-generation aromatase
for personalized medicine. Fortunately, hormone inhibitors did not effectively suppress estrogen
receptor-positive tumors usually are responsive levels and because decreased estrogen levels in
to agents such as tamoxifen that block the func- peripheral tissues could be counteracted by
tion of estrogen. Tamoxifen is a potent antagonist increased estrogen synthesis in the ovaries [189].
of the ER with inhibitory effects on tumor growth In postmenopausal women, aromatase inhibitors
that has become the gold standard for endocrine are well tolerated and improve both disease-free
treatment of estrogen receptor-positive breast and recurrence-free survival [190–192]. Similar
cancer in premenopausal and postmenopausal to CYP2D6, the Cys264 and Thr364 variants in
women [184]. Tamoxifen is associated with side aromatase are associated with decreased activity
effects such as blood clots, stroke, and increased and lower levels of immunoreactive protein,
risk of endometrial and uterine cancer, but 5-year which may contribute to variation among patients
use of tamoxifen has been shown to reduce risk in response to aromatase inhibitor therapy [180].
of cancer recurrence by ~50 % [185]. For most Although directed endocrine therapies provide
patients, the benefit of using tamoxifen for hor- treatments specific for patients with hormone
mone receptor-positive disease outweighs the receptor-positive breast cancer, factors such as
risk of serious side effects; however, a small sub- menopausal status and innate genetic variability
group of hormone receptor-positive patients who may alter the effectiveness of treatment.
carry specific variants in the cytochrome P450
2D6 (CYP2D6) gene do not benefit from tamoxi- Treatment for HER2-Positive Breast
fen. The CYP2D6 gene is a key enzyme in the Cancer
metabolism of tamoxifen to its active metabolite Therapies directed at the HER2 protein provide a
endoxifen. Several DNA variants in CYP2D6 second avenue of targeted treatment for some
result in poor metabolism of tamoxifen and lower patients with breast cancer. Trastuzumab
levels of endoxifen [186]. Patients who carry (Herceptin®, Genentech, South San Francisco,
reduced-function or nonfunctional CYP2D6 CA; www.gene.com/) is a humanized monoclo-
alleles have been found to derive inferior thera- nal antibody that binds to the extracellular
peutic benefit from tamoxifen and thus are at domain of the HER2 protein, blocking tumor cell
increased risk of breast cancer recurrence [178] growth. Trastuzumab is the current standard of
or have significantly shorter disease-free survival care in adjuvant therapy for HER2-positive breast
than noncarriers [179]. Studies are underway to cancer, effective as a single agent or in combina-
determine the utility of CYP2D6 genotyping for tion with chemotherapeutics for the 20–25 % of
making clinical decisions about tamoxifen and patients with HER2-positive cancer [193].
the potential to optimize breast cancer therapy However, many patients with HER2-positive dis-
[187, 188]. ease do not derive tangible benefit from trastu-
Alternate forms of directed antiestrogen thera- zumab. Given that the cost per patient for
pies do exist for patients with hormone receptor- trastuzumab ranges from $20,000 to $80,000 per
positive breast cancer, including aromatase year with the potential for significant adverse
72 B. Kumar

side effects [194], a more precise classification of specific drug combinations [197]. Obviously, the
HER2-positive patients who will derive benefit ability to predict which patients will benefit from
from trastuzumab and improved understanding adjuvant therapy and identify who will respond
of how amplification and/or overexpression of favorably to neoadjuvant regimens would pro-
HER2 contribute to aggressive tumor biology are vide an additional level of personalized care.
critical to improving patient treatment. The major
oncogenic unit in HER2-positive breast cancer Gene Expression
appears to be a heterodimer between the HER2 and Chemotherapeutic Agents
and epidermal growth factor receptor-3 (HER3) Gene expression profiling has been used to study
proteins, where HER3 functions as a necessary the biological responses of human breast carcino-
dimerization partner for HER2 to achieve full mas to optimize chemotherapeutic treatments.
oncogenic signaling potential [195]. Cell lines derived from luminal and basal epithe-
Recent studies have shown that HER2/HER3 lium have been observed to respond differently to
heterodimers promote cellular proliferation in agents commonly used in chemotherapy, such as
both in vitro and in vivo models, suggesting that doxorubicin (DOX) and 5-fluorouracil (5FU). In
HER3 may be an important therapeutic target in culture, luminal cell lines show low levels of
HER2-positive patients [182]. Pertuzumab has expression for genes regulating cellular prolifera-
been shown to bind to the dimerization arm of tion and the cell cycle, while basal cell lines tend
HER2, blocking HER2/HER3 heterodimeriza- to repress genes involved in cellular differentia-
tion and attenuating growth of solid tumors in tion when exposed to DOX and 5FU [198].
model systems [196]. Thus, combining pertu- Similarly, different molecular subtypes of breast
zumab with trastuzumab may augment therapeu- cancer defined by gene expression profiling
tic benefit by blocking HER2/HER3 signaling. respond differently to preoperative chemother-
Monogram Biosciences (South San Francisco, apy, with basal-like and HER2-positive subtypes
CA; www.monogrambio.com/) has developed being more sensitive to paclitaxel and doxorubi-
the commercially available HERmark™ test to cin than luminal and normal-like cancers [199].
measure total HER2 levels and HER2 homodi- Expression signatures also have been used to
mers in FFPE tissue and is developing a predict clinical response of breast cancer patients
VeraTag™ assay to quantify levels of HER2/ receiving either cyclophosphamide–adriamycin
HER3 heterodimers. These assays may allow or epirubicin–5FU as part of their adjuvant che-
patients with HER2-positive breast cancer to motherapy regimen [200] and to distinguish pri-
receive the most efficacious combination of new mary breast tumors that are responsive or resistant
drugs targeting HER2. to docetaxel chemotherapy [201]. These observa-
tions further highlight the vast amount of molec-
Chemotherapeutics ular variability among breast carcinomas and
Chemotherapy involves use of chemical agents as emphasize the need for additional molecular sig-
part of a systemic treatment targeting prolifera- natures to more effectively guide treatment.
tive cancer cells. Adjuvant chemotherapy is used
to reduce risk of recurrence after primary therapy DNA Variation and Chemotherapeutic
in women with localized breast cancer and to Agents
provide palliative care in patients with advanced Clinical responses in breast cancer patients to
(metastatic) disease. In contrast, neoadjuvant commonly used chemotherapeutic agents vary
chemotherapy is normally used to shrink moder- considerably, from optimum therapeutic response
ate- to large-sized breast carcinomas prior to sur- to partial (beneficial) response to severe adverse
gical resection, which permits use of less events. Variation at the DNA level in an increas-
aggressive surgical options, including breast con- ing number of genes is now known to affect the
servation, and may be useful in guiding longer- pharmacokinetics and pharmacodynamics of
term treatment based on tumor response to many chemotherapeutic drugs [202, 203], thus
4 Breast Cancer Genomics 73

influencing toxicity and patient response. To blood)—contribute to decisions regarding

improve the safety and efficacy of current therapy for metastatic breast cancer in con-
treatments, therapies could be tailored to indi- junction with diagnostic imaging, history, and
vidual patients based on their genetic makeup physical examination
[204]. For example, the carbonyl reductase 3 • Carcinoembryonic antigen (CEA)—contrib-
(CBR3) gene contributes to the reduction of utes to decisions regarding therapy for meta-
DOX to doxorubicinol, a less potent metabolite, static breast cancer in conjunction with
and the extent of metabolism is believed to be a diagnostic imaging, history, and physical
source of variability in doxorubicin chemother- examination
apy. The 11G > A variant (rs8133052) in CBR3 • ER/PR—should be measured on every pri-
has been shown to influence tumor tissue expres- mary invasive breast cancer to identify patients
sion of CBR3 and is associated with interindi- most likely to benefit from endocrine therapy
vidual variability in clinical outcomes. Women • HER2—should be measured on every primary
with the 11GG genotype experience greater leu- invasive breast cancer at diagnosis or recur-
kocyte toxicity and are less likely to show a rence to guide trastuzumab therapy
reduction in tumor size than women carrying • Urokinase plasminogen activator (uPA) and
11AA [174]. plasminogen activator inhibitor 1 (PAI-1)—
A number of chemotherapeutics generate measured by ELISA on fresh or frozen tissue
reactive oxygen species that function by damag- for determining prognosis in newly diagnosed,
ing DNA and triggering the apoptotic cascade. node-negative breast cancer patients
Women carrying variants in genes associated • Oncotype DX®—in newly diagnosed patients
with oxidative stress, such as manganese super- with node-negative, ER-positive breast can-
oxide dismutase (MnSOD), catalase (CAT), and cer, can be used to predict risk of recurrence in
myeloperoxidase (MPO) that result in higher lev- women treated with tamoxifen
els of reactive oxygen species, tend to have better Large cancer centers such as Massachusetts
overall survival than women with genotypes General Hospital and Memorial Sloan-Kettering
associated with lower levels of reactive oxygen Cancer Center are now embracing the importance
species when treated with chemotherapy [176]. of genomics in clinical practice, recently imple-
Due to the large number of drug-metabolizing menting policies to routinely assay a number of
enzymes and drug transporters containing poly- breast cancer-related genes: vakt murine thy-
morphisms that affect chemotherapy-related tox- moma viral oncogene homolog 1 (AKT1) and
icity and treatment outcomes in breast cancer HER2 at Memorial Sloan-Kettering, phosphatase
patients, improved pharmacogenetic information and tensin homolog (PTEN) and TP53 at Mass
is needed to identify individuals at risk for toxic- General, and phosphatidylinositol 3-kinase, cata-
ity and poor response. lytic, alpha (PIK3CA) at both institutions [205].
As genomic medicine becomes an integrated part
Genomics in Clinical Practice of health-care delivery, use of personalized
Recent developments in the clinical arena are genomics in the clinical treatment of breast can-
indicative of the emerging importance of per- cer will increase.
sonal genomics in the prevention, surveillance,
and treatment of breast cancer. Professional orga-
nizations such as the American Society of Genomic Studies of Breast Cancer
Clinical Oncology (ASCO) have issued recom- Initiation, Progression,
mendations on the use of molecular markers for and Metastasis
guiding therapy and determining prognosis in
breast cancer patients [175]. Breast cancers progress through multiple
• CA 15-3 and CA 27.29 (assays to detect genomic and epigenomic steps. However, the
circulating MUC-1 antigen in peripheral evolutionary process is unlikely to be the result of
74 B. Kumar

Normal duct Hyperplasia DCIS Invasive cancer

80 140 15 20
(2.2) 120
60 15
40 10
40 5
1220 12 12 12
9 9 9 9
6 6 6 6
0 3 0 3 0 3 3
0 2 0 2 0 2 0 2 0 1c
4 0 1c 4 0 1c 4 0 1c 4 6 8
6 8 6 8 6 8 10 12
10 12 10 12 10 12 20q13.2
20q13.2 20q13.2 20q13.2

Fig. 4.1 Genome instability measured during breast centromere of chromosome 1 and at chromosome 20q13.
tumor progression using FISH [206]. Histological sections The onset of instability at the DCIS stage of evolution is
depicting stages of evolution are illustrated above. apparent (Reprinted from Korkola and Gray [211]. With
Bivariate measures of genome copy number at the permission from Elsevier)

a linear and more or less constant rate of succes- increase in genome instability during breast
sive genomic and epigenomic aberration tumor progression is illustrated in Fig. 4.1.
accretion. It seems more likely that cancer pro- Rarely, a single cell accumulates genomic or
gression varies between individuals and over epigenomic alterations that reactivate telomerase
time within an individual. The breast progenitor and confer a proliferative advantage. This cell
cell in which the tumor arises likely determines might be considered the tumor initiation cell and
the spectrum of aberrations needed to enable pro- will have multiple characteristics that appear
gression. Evidence for this comes from studies “stem cell like.” The extent to which this is
showing that the spectrum of genomic aberra- related to normal stem cells remains unclear.
tions accumulated is strongly influenced by the However, it is likely that the genomic characteris-
normal progenitor cell type in which the cancer tics of this cell—both transcriptional and
arises [108, 206]. In general, the events associ- genomic—will be reflected in subsequent prog-
ated with early aspects of breast tumor appear to eny. This may explain why tumor genomes
be epigenomic in nature, wherein stepwise DNA appear to evolve relatively slowly after telomere
methylation changes enable escape of telomer- crisis and why metastases that develop years after
ase-negative epithelial cells from proliferation immortalization usually retain the genomic char-
barriers [207]. This leads to proliferation in the acteristics of the primary tumor from which they
absence of telomerase and culminates in a period were derived [212].
of high genome instability owing to checkpoint That said, not all breast tumors progress this
deregulation [208, 209] and/or entry into telo- way. A recent analysis of cancer progression
mere crisis when telomeres become critically termed Sector-Ploidy-Profiling (SPP) demon-
short [206, 210]. This barrier is highly effective strates that breast tumor evolution may evolve as
but not perfectly so. As a consequence, most cells a single major clonal subpopulation or as multi-
become genomically unstable and die. The ple clonal subpopulations [213]. The latter model
4 Breast Cancer Genomics 75

may explain why a small percentage of meta- yields a very large number of differentially
static cancers do not resemble the primary tumor expressed genes. The relative position of individ-
from which they were derived. Figure 4.1 also ual genes in a rank-ordered gene list varies greatly,
shows that the cells that survive telomere crisis but the consistency of the gene list membership is
remain genomically unstable. Thus, tumors are fairly high across various datasets [214].
likely to consist of a large number of cells that are Functional annotation indicates that the
not faithful genomic representations of the tumor- majority of these prognostic genes are proliferation-
initiating cell and are likely to be biologically related genes, and the remainder are mostly
compromised. These cells are likely to be much ER-associated and, to a lesser extent, immune-
more sensitive to treatment than the tumor- related genes [215–217]. Because these genes
initiating cells from which they were derived. function together in a coordinated manner in the
This may partially explain why breast tumors ini- regulation and execution of complex biological
tially respond well to treatment but fail to exhibit processes, such as cell proliferation, or originate
a durable response. from a particular cell type, such as immune cell
infiltrate, many of these prognostic genes are also
highly coexpressed with one another. It is there-
Various Genomic Signatures fore expected that a large number of nominally dif-
ferent prognostic signatures can be constructed
Unsupervised molecular classification identified that will all perform equally well. For example, a
three major and robust groups of breast cancers particular gene may be highly significantly dis-
that differ in the expression of several hundred to criminating in two datasets but it is ranked 5th
a few thousand genes. These include basal-like among the most discriminating genes in one data-
breast cancers, which are negative for ER, pro- set (based on P-value or fold difference) but only
gesterone receptor (PR), and human epidermal 35th in another dataset (which is still very high,
growth factor receptor 2 (HER2); low histologi- considering the thousands of comparisons!).
cal grade ER-positive breast cancers (also called In multivariate prediction model building, the
luminal A); and high-grade, highly proliferative top few informative features are usually com-
ER-positive cancers (luminal B). Several smaller bined, and genes are added incrementally to
and less stable molecular subsets (such as normal- increase the predictive performance. However,
like, HER-2-positive, and claudin-low) have also because many of the genes are highly correlated
been proposed but are less consistently seen and with each other, adding genes lower on the list
are distinguished by substantially smaller molec- yields less and less improvement in the model as
ular differences [95, 104]. Importantly, among a result of lack of independence. Therefore, the
the various molecular subsets, one group, the gene in question will be included in a predictor
luminal A class that includes low-grade developed from the first dataset (because it is
ER-positive cancers, stands out with a very favor- ranked as 5th) and will work well on validation in
able prognosis with or without adjuvant endo- the second dataset; but if a new predictor were to
crine therapy. The other groups have worse but be developed from the second dataset, this gene
rather similar prognosis [95, 102]. may not be included in the predictor (because it is
If one understands these close associations ranked 35th).
between clinical phenotype, molecular class, and These three features of the breast cancer prog-
prognosis, it is no longer surprising that compar- nostic gene space—the large number of individu-
ing gene expression profiles of breast cancers that ally prognostic features, the unstable rankings,
recurred (mostly the ER-negative and the high- and the highly correlated expression of informa-
grade, ER-positive cancers) and those that did not tive genes—explain why it is easy to construct
(low-grade, ER-positive cancers) in the absence many different prognostic predictors that perform
of any systemic therapy (or after antiestrogen equally well even if they rely on nominally dif-
therapy alone in the case of ER-positive cancers) ferent genes in the model. However, this does not
76 B. Kumar

mean that all published prognostic gene signa- proliferation rate, and ER status, limits the practi-
tures are equally ready for clinical use. Before cal value of these tests.
adoption in the clinic, a molecular diagnostic Efforts are under way to develop simple mul-
assay has to be standardized, the reproducibility tivariate prognostic models that use routine path-
within and between laboratories and stability of ological variables (such as ER, histologic grade,
results over time have to be demonstrated, and its and HER2 status), and these could eventually
predictive accuracy has to be validated in the rival the performance of the first-generation
right clinical context, preferably in multiple inde- prognostic gene signatures [221, 222]. However,
pendent cohorts of patients. Most importantly, standardization of the pathological assessment of
clinical utility implies that the assay improves breast cancer and reducing the interobserver vari-
clinical decision making and complements or ability remain an important challenge. Predicting
replaces older standard methods, which in turn clinical outcome, such as prognosis or response
leads to better patient outcomes. Few published to chemotherapy, within clinically and molecu-
prognostic predictors have met these criteria larly more homogeneous subsets (such as triple-
[175, 218]. negative breast cancers or high-grade, ER-positive
The predictive performance of a multivariate cancers) would be highly desirable. Unfortunately,
model largely depends on the number of indepen- these prediction problems seem to be more diffi-
dent informative genes included in the model, the cult [223, 224]. It seems that fewer genes are
magnitude of differential expression of the infor- associated with outcome in homogeneous disease
mative genes, and the complexity of the back- subsets and the magnitude of association is mod-
ground. Different clinical prediction problems est when currently available datasets are ana-
show different degrees of difficulty. From the dis- lyzed. This leads to predictors that are specific
cussion above, it should be apparent that predic- for a particular dataset from which they were
tion of ER status, histological grade of breast developed. These prediction models are fitted to
cancer, or better or worse prognosis associated the dataset and rely on features that have no or
with these clinical phenotypes should be rela- limited generalizability. This means that they fail
tively easy when considering all breast cancers to validate when applied to independent data or
together and that such prediction can therefore may demonstrate only nominally significant pre-
yield predictors with good overall accuracy. dictive value (i.e., they may predict outcome
Indeed, prognostic gene signatures developed for slightly better than chance). Also, the discrimi-
breast cancer in general or for ER-positive can- nating value may not be substantial enough to be
cers tend to have good performance characteris- clinically useful [225, 226]. For example, if the
tics [161, 165, 167, 217]. However, the good-prognosis group has a recurrence rate of
first-generation prognostic signatures share some 30 % compared with 50 % in the poor-risk group,
limitations. Because these were invariably devel- these may be significantly different, but the risk
oped by analyzing all subtypes of breast cancers of recurrence in the good-risk group is still too
together, they tend to assign high risk category to high to safely forego adjuvant chemotherapy.
almost all ER-negative cancers (which are almost
always high grade), even though a substantial
majority of these cancers have good prognosis Genomic Mechanism of Breast Cancer
[219, 220]. Similarly, the good- and poor- Dormancy and Recurrence
prognosis ER-positive cancers, as assigned by
gene profiling, tend to correspond to the clini- Factors that determine the length of the dor-
cally low-grade/low-proliferation versus high- mancy period remain unclear [227]. Current data
grade/high-proliferation subsets, respectively. have led to various experimental models that
This strong correlation between prognostic risk address the phenomenon of tumor dormancy. It
as predicted by gene signatures and routine clini- appears that dormant cancer cells can persist
cal variables, such as histological grade, either by completely withdrawing from the cell
4 Breast Cancer Genomics 77

cycle (mitotic arrest) or by continuing to bone marrow microenvironment by conferring

proliferate at a slow rate that is counterbalanced resistance to TRAIL [238].
by cell death [228, 229]. These two types of
dormancy are not mutually exclusive; both forms Micrometastatic Dormancy
of latency could coexist in the entire population In contrast to dormancy due to mitotic arrest, dor-
of disseminated tumor cells (DTC) of a particu- mancy of a micrometastasis seems to be caused
lar cancer patient. by a balance of cell proliferation and apoptosis,
such that the tumor does not increase in size. This
Single-Cell dormancy constant balance is regulated by proangiogenic
In the single-cell dormancy model, isolated proteins and angiogenic inhibitors produced by
tumor cells detached from the primary tumor tumor and stromal cells, as well as immunologic,
arrive at the future metastatic organ and enter a hormonal, or other microenvironmental switches
prolonged state of mitotic arrest. This model of [239]. According to Naumov et al., a failure to
arrested apoptosis contrasts with the micrometa- activate the angiogenic switch can maintain a
static dormancy model, in which proliferation in group of cancer cells in a dormant state [240].
micrometastatic foci is counterbalanced by cell Indraccolo et al. reported that a short-term pertur-
death. Speculations about the metabolic status of bation in the tumor microenvironment, in the
minimal residual disease (MRD) during dor- form of a transient angiogenic burst, could suf-
mancy are not yet sufficiently investigated. Cell fice to interrupt tumor dormancy [239]. Genetic
cycle regulation mechanisms are highly complex, data support the interpretation that minimal
with an assortment of stimuli interacting at residual cancer might be divided into “active”
numerous cell cycle checkpoints to determine the and “dormant” groups, in which an advantageous
proliferative status. The presence of tumor cell mutation is acquired shortly before a highly
dormancy due to a growth arrest of cancer cells is aggressive metastatic clone appears [241]. At
supported by evidence that within tissues where present, there is no definite answer to the ques-
primary tumors are developing or tissues that tion of which model best represents tumor dor-
harbor disseminated cells and have a functional mancy in breast cancer. Hussein and Komarova
vasculature, tumor cells are found to be in a non- hypothesized that indolent breast cancers might
proliferative mode [230–232]. In numerous stud- fit into the single-cell dormancy model, while
ies, dormant cancer cells have been demonstrated more aggressive diseases are linked to the micro-
to be in a G0–1 arrest; this was linked to negative metastatic dormancy model [229]. Indeed, in a
staining for proliferation markers (e.g., Ki67, series of experiments, Barkan et al. demonstrated
PCNA) [233, 234]. Dormant tumor cells may that more aggressive basal-type cell lines, such as
gain a survival advantage by blocking the recep- MDA-MB-231, proliferated readily, while estro-
tors for tumor necrosis factor-related apoptosis- gen receptor (ER)-positive MCF7 remained in a
inducing ligand (TRAIL), which would result in state of mitotic arrest, potentially linking the dor-
arrested apoptosis. Two exemplary mechanisms mancy type associated with arrested growth with
of TRAIL-receptor blocking have been described the less aggressive disease phenotype [242].
and may be of relevance to dormant tumor cells. Genes driving recurrence may be encoded in
TRAIL receptors in cancer cells can be blocked the gene expression profile of primary tumors.
by osteoprotegerin, an important member of the Transcriptomic analysis of a variety of primary
tumor necrosis factor receptor superfamily [235, tumors by Ramaswamy and colleagues identified
236]. Interestingly, bone marrow stromal cells a 17-gene signature present in a wide range of
from breast cancer patients secrete enough osteo- human primary tumors that predicted metastasis
protegerin to inhibit apoptosis in vitro [237]. and poor clinical outcome [243]. Although the
More recently, c-Src (a tyrosine-specific kinase genomes of clinically similar breast cancers are
involved in breast cancer progression) was dem- remarkably different, some regions of the genome
onstrated to support cancer cell survival in the are recurrently aberrant. The number of genes
78 B. Kumar

mapping to regions of significant abnormality is pathophysiology of the disease. Functionally

remarkable. In fact, correlative analyses of important coamplified genes demonstrated so far
genome copy number and gene expression indi- include LSM1, BAG4, and C8orf4 at 8p11–12
cate that 10–15 % of the entire genome is deregu- [246]; MYC and PVT1 at 8q24 [247]; CCND1
lated by recurrent genome copy number and EMSY at 11q13 [248]; and ERBB2 and
abnormalities [108, 109]. Analysis of CGH data GRB7 at 17q21 [249]. Recent studies also sug-
suggests the existence of at least three general gest that amplifications of regions that are well
classes—those with “simple” genomes display- separated in the normal genome are not indepen-
ing relatively few genome aberrations, those with dent events. For example, coamplification of
“complex” genomes displaying many aberra- regions at 8p11–8p12 and 11q12–11q14 [250] or
tions, and those displaying high-level amplifica- 8q24 and 17q21 [251] have been reported to con-
tion [109, 244]. It is likely that these subtypes tribute collectively to breast cancer pathophysiol-
may be caused by differences in DNA repair ogy. These interactions may be just the tip of the
defects. For example, complex genomes typi- iceberg since dozens of genes have been identi-
cally carry p53 mutations while simple genomes fied as overexpressed as a result of recurrent,
do not. high-level amplification and most have not been
CGH profiles for tumors from BRCA1 carri- functionally assessed [109]. Recurrent muta-
ers also display high genome complexity, tions, although relatively infrequent in breast
although the spectrum of recurrent abnormalities cancer, provide particularly strong evidence of
differs between sporadic and heritable breast can- functional importance. Bringing all of these
cers. In particular, tumors from BRCA1 mutation observations together, recent integrative analyses
carriers typically display recurrent aberrations at of homozygous deletions, high-level amplifica-
chromosomes 3 and 5 that sporadic tumors do not tion, and recurrent mutations suggest that
have [245]. Relatively few recurrent mutations genomic deregulation of pathways involving
have been found in breast cancers. High- ERBB2, EGFR, and PI3K signaling and DNA
frequency somatic mutations (i.e., present in topology is particularly important in breast can-
>3 % of reported cases) include PIK3CA, TP53, cer genesis and progression [252].
CDH1, CDKN2A, and AKT1 (http://www. Epigenomic alterations also contribute to breast
sanger.ac.uk/genetics/CGP/cosmic). However, cancer pathophysiology [253, 254]. Targeted and
almost 200 others have been reported as recur- genome-wide analyses of promoter region meth-
rently mutated but at lower frequency [63]. ylation in breast cancer have shown that genes
Prevalent germ line mutations that contribute to with known or suspected tumor suppressor func-
breast cancer genesis involve BRCA1, BRCA2, tion including BRCA1, CCND2, CDKN2A,
CHEK1, and TP53. The mechanisms by which RARβ, and RASSF1A [255, 256] and members of
low-level copy number abnormalities contribute the WNT signaling pathway [257] are recurrently
to cancer pathophysiology have not yet been downregulated by hypermethylation. Consistent
determined definitively. However, we have sug- with this, genes found to be recurrently mutated in
gested that these deregulate transcription of large large-scale genome analysis studies are frequently
numbers of genes that collectively increase meta- targets of hypermethylation, and hypermethyl-
bolic activity [109]. The contributions of regions ation is usually mutually exclusive from genetic
of amplification at 8p11–12, 8q24, 11q13, 12q13, changes [258]. At the chromatin level, modifica-
17q11–12, 17q21–24, and 20q13 as somewhat tion via acetylation and/or phosphorylation
better understood since each encodes a known directly modifies estrogen receptor alpha and other
oncogene (e.g., FGFR1, MYC, CCND1, MDM2, steroid hormone receptors superfamily in breast
ERBB2, PS6K, and ZNF217, respectively). cancer, thereby modifying ligand sensitivity and
However, substantial evidence is now emerg- hormone antagonist responses [259]. Chromatin
ing that demonstrates that multiple genes in each remodeling also has been implicated in growth
region of amplification contribute to the and metastasis. In particular, SATB1 has been
4 Breast Cancer Genomics 79

reported to be a genome organizer that tethers Several sensitive DNA methylation detection
multiple genomic loci and recruits chromatin- techniques (summarized in Table 4.6) were
remodeling enzymes to regulate chromatin struc- developed upon the basis of bisulfite conversion,
ture and gene expression. Overexpression of this including bisulfite sequencing, methylation-
“master chromatin regulator” appears to drive specific PCR (MSP), combined bisulfate restric-
breast cancer progression by upregulating metas- tion analysis (COBRA), and so on [263–265].
tasis-associated genes and downregulating tumor- Among these technologies, MSP is the most pop-
suppressor genes [260]. ular and powerful method for DNA methylation
detection, which needs limited amounts of DNA
material [264]. Since MSP is a gel-based assay, it
Various Genomic Approaches cannot provide quantitative information and is
and Their Outcome in Breast Cancer subjective. Several real-time methylation-specific
Research PCR methods, such as bisulfite treatment in com-
bination with MethyLight™ [266], quantitative
To facilitate understanding of the scope of epi- multiplex MSP (QM-MSP) [267, 268], or pyro-
genetic modifications that occur in normal and sequencing [269], have been developed and used
cancer cells, a range of gene-specific or genome- in DNA methylation studies, which have
wide technological approaches have been devel- improved features to detect minimal amounts of
oped. We present an overview of recent aberrant DNA methylation in a quantitative
technological developments and discuss the mer- manner.
its and the limitations of these approaches with In addition to the gene-locus specific DNA
respect to studies on cancer cells. methylation detection methods mentioned
above, several recently developed genome-wide
techniques have been applied to the study of
DNA Methylation global DNA methylation profiles in normal and
cancer cells (summarized in Table 4.6).
DNA methylation is the first identified epigenetic Restriction landmark genomic scanning (RLGS)
mechanism of gene regulation. The initial meth- is one of the earliest methods to be applied to
ods of detection of methylation were restricted to genome-wide methylation analysis [270]. RLGS
the quantitation of total 5-methylcytosine content has an ability to globally analyze the methyla-
by high-performance liquid chromatography tion status of approximately 1,000 unselected
(HPLC) or high-performance capillary electro- CpG islands. Other important techniques for
phoresis (HPCE) [261] and the study of DNA analyzing altered DNA methylation patterns
methylation of selected sequences using restric- across the genome have relied on an arbitrarily
tion enzymes that can distinguish between meth- primed PCR technique (e.g., AIMS) [271, 285].
ylated and unmethylated recognition sites in Further advances in this field are derived from
genes of interest [262]. For restriction-enzyme- the application of the DNA microarray technol-
based methods, incomplete restriction-enzyme ogy. A widely used example is differential meth-
digestion and the limitation of enzyme- ylation hybridization (DMH) that uses CpG
recognition sites restricted their extensive appli- island and promoter sequence microarrays,
cation. These technical limitations caused delay which enables the simultaneous analysis of the
in advances in the cancer epigenetic field in con- methylation levels of a large number of CpG
trast to the rapid development in cancer genetics islands in the genome [272].
field. The development of a bisulfite-conversion A recently developed, related technique called
technique that reproducibly changes unmethyl- HpaII tiny fragment enrichment by
ated cytosines to uracil but leaves methylated ligation-mediated PCR assay (HELP) uses a mod-
cytosines unchanged [263] was a key develop- ified approach to globally analyze DNA methyla-
ment that drastically sped up progress in the field. tion patterns [273]. Methods (e.g., methylated
80 B. Kumar

Table 4.6 Technologies for assessing and profiling DNA methylation and histone modifications
Technology Description Study
Gene-specific detection of DNA methylation
Bisulfite sequencing Bisulfite-converted DNA is PCR- Fraga and Esteller [261]
amplified to enrich the target templates.
The purified DNA templates are
subjected to sequencing analysis
directly or after cloning into plasmid
Methylation-specific PCR (MSP) This technique takes advantage of the Herman et al. [264]
altered sequence of bisulfite-converted
unmethyated and methylated DNA for
designing primers, which can amplify
DNA in a methylation-state-specific
Combined bisulfite restriction analysis The qualitative and quantitative Xiong and Laird [265]
(COBRA) detection of methylated alleles is
achieved by restriction enzyme-
mediated digestion of PCR amplified
target amplicons from bisulfite-modified
MethyLight™ By including the fluorescent probe Eads et al. [266]
technology (TaqMan®), this method is
able to quantitatively and sensitively
detect methylated alleles
Quantitative multiplex MSP (QM-MSP) This technology is a modified version of Fackler et al. [267]
fluorogenic probe-based quantitative Swift-Scanlan et al. [268]
MSP assay. This method includes the
multiplex PCR step that allows
amplification of multiple target alleles.
The diluted multiple PCR products are
subjected to quantitative MSP assay for
multiple gene detection
Pyrosequencing This method is based on sequencing-by- Tost and Gut [269]
synthesis technology to quantitative
detect methylation levels of individual
CpG site by monitoring the real-time
incorporation of nucleotides through the
enzymatic conversion of released
pyrophosphate into a bioluminometric
Genome-wide profiling of DNA methylation
Restriction landmark genomic scanning In the RLGS technique, restriction- Costello et al. [270]
(RLGS) enzyme digestion, radioactive labeling,
and two-dimensional electrophoresis
combine to quantitatively display DNA
methylation levels of thousands of CpG
Amplification of intermethylated sites The AIMS method combines Frigola et al. [271]
(AIMS) methylation-sensitive restriction enzyme
digestion with the display of
methylation fingerprint of PCR-
amplified DNA fragments, which can
then be isolated and characterized
individually by sequencing
4 Breast Cancer Genomics 81

Table 4.6 (continued)

Technology Description Study
Differential methylation hybridization DMH is a promoter-sequence- Huang et al. [272]
(DMH) microarray- basis method that combines
methylation-sensitive restriction-
enzyme digestion with linker-basis PCR
labeling to serve as probes for array
hybridization, capable of globally
displaying methylated CpG islands
HpaII tiny fragment enrichment by The HELP method adopts differential Khulan et al. [273]
ligation-mediated PCR assay (HELP) digestion with a methylation-sensitive
restriction enzyme or its methylation-
insensitive isoschizomer and ligation
mediate PCR amplification of digested
templates for cohybridization to a
genomic microarray, enabling the
display of genome-wide methylated
CpG islands
Methylated DNA immunoprecipitation The methyl-DIP technique uses Keshet et al. [274]
(methyl-DIP) antibodies against methyl- CpG-binding Weber et al. [275]
domain proteins (MBDs) to Weber et al. [276]
immunoprecipitate sheared genomic
DNA for isolation of methylated DNA
fragments. Methyl-DIP has been
combined with tiling microarrays or
with high-density promoter arrays to
map the human methylome
Microarray-based gene expression Gene expression microarray has been Suzuki et al. [277]
profiling applied to display expression-profile Yamashita et al. [278]
changes in cells treated with epigenetic Hoque et al. [279]
inhibitors for identification of
methylation-targeted genes
Methylation-specific digital karyotyping The MSDK method is a modified serial Hu et al. [280]
(MSDK) analysis of gene expression (SAGE)
assay that combines a methylation-
sensitive restriction enzyme and a
fragmenting restriction enzyme to
generate short sequence tags for
providing information on gene loci and
their methylation levels
Genome-wide profiling of histone modifications
Chromatin immunoprecipitation This method combines ChIP technology Bernstein et al. [281]
(ChIP)-on-chip with high-density microarrays for
measuring and mapping histone-binding
genomic loci
ChIP-SAGE This method adopts a combination of Roh et al. [282]
ChIP and SAGE technologies to
globally quantify and map genomic
binding sites for specifically modified
ChIP-Seq This new method employs a high- Barski et al. [283]
throughput sequencing technique to Mikkelsen et al. [284]
analyze ChIP DNA for genome-wide
mapping of histone-DNA binding
82 B. Kumar

DNA immunoprecipitation [methyl-DIP]) based technique is ChIP-on-chip [281]. In contrast to the

on chromatin immunoprecipitation (ChIP) using ChIP-on-chip method, another new technique for
the ChIP-on-chip technology are other seminal profiling histone modifications at a genomic scale
recent advances in the epigenomic profiling of is ChIP-SAGE, which combines ChIP experiments
cancer cells [274–276]. Finally, gene expression with the SAGE technology [282]. The merit of
profiling using microarrays is another powerful ChIP-SAGE is that unlike ChIP-on-chip that
and widely used technique for assessing genome- requires sequence information of preselected
wide DNA methylation patterns. This approach is genomic regions to manufacture genomic microar-
used to compare gene expression levels from can- rays, no prior genomic sequence information is
cer cell lines before and after treatment with a required in this assay. However, the cost for ChIP-
demethylating drug, an HDAC inhibitor, or a SAGE is higher than ChIP-on-chip due to the use
combination of both drugs [277–279]. As with all of the more expensive traditional sequencing meth-
microarray-based technologies, the identified ods. Besides, the fact that not every region of chro-
candidate genes are further verified, in this case, mosomes contains restriction-enzyme recognition
by quantitative RT-PCR and promoter methyla- sites used to cleave the ChIP DNA limits the capac-
tion analyses. ity of ChIP-SAGE to study the entire genome.
In addition to microarray-based techniques, a More recently, a new technique, chromatin
serial analysis of gene expression (SAGE)- immunoprecipitation combined with high-
technology-based method, known as methylation- throughput sequencing techniques (ChIP-Seq),
specific digital karyotyping (MSDK), has been has been developed for analyzing ChIP DNA
developed recently for the genome-wide analysis using a high-throughput massively parallel signa-
of methylation profiles [280]. The advantage of ture sequencing-like technique developed by
this new technique is that there is no need of prior Solexa [283, 284]. This technique is more power-
sequence information for analysis and the obtained ful and cost-effective than the ChIP-SAGE tech-
data can be used to map the methylated gene loci nique. There are several advantages in this new
and to determine their methylation levels. technique, including the use of less PCR amplifi-
cation of ChIP DNA after ligation to adaptors and
a highly efficient sequencing procedure (sequenc-
Histone Modifications ing by synthesis). In contrast to ChIP-on-chip,
more quantitative data of histone modification
Characterization of posttranslational histone levels at different chromosomal regions can be
modifications is a greater challenge than analysis obtained in ChIP-Seq experiments. In addition to
of DNA methylation and needs special technol- mapping the genome-wide histone-DNA binding
ogy. Currently, the gold standard for accurately patterns, it can be envisioned that this new tech-
assessing global levels of histone modifications is nique has great potential for globally defining the
mass spectrometry [286]. Since antibodies spe- methylome of a particular cell type. With rapid
cifically recognizing the amino acid modifica- and striking technological advancements, global
tions of histone proteins are available, a simple analysis of DNA methylation and histone modifi-
Western blot analysis is also used for detecting cation mapping on chromosomes has become
histone modifications. In addition to determining eminently practical.
the types and relative levels of histone modifica-
tion in cells, characterization of distribution of
each type of histone modification on chromo- Treatment of Breast Cancer
somes also provides very important information. Through a Genomic Approach
The current techniques for genome-wide analy-
sis all adopt the ChIP technology with antibodies Treatment is complicated by the complexity of
against specifically modified histones (summarized cancer genomes and the large and growing num-
in Table 4.6). The first developed and widely used bers of candidate anticancer agents that should be
4 Breast Cancer Genomics 83

considered for use in the treatment. In fact, a cancers cluster within the basal-like subtype [288],
2009 survey by the pharmaceutical industry indi- which suggests that host germ line genetics may
cates that more than 800 experimental therapies determine the tumor subtype. These subtypes have
are now being evaluated clinically, including been shown to correlate with clinical outcomes [97,
106 in breast cancer (http://www.pharma.org/ 103, 171]. The luminal A and B subtypes have the
files/09-046PhRMACancer09_0331.pdf ). best prognosis, whereas HER2-like and basal-like
Identification of genomic and epigenomic fea- have the worst prognosis, although the inclusion of
tures that predict response to these agents is chal- trastuzumab in primary therapy for HER2-positive
lenging in the context of clinical trials because of breast cancer has led to vast improvements in out-
the difficulty of obtaining tumor samples during comes for patients with HER2-positive breast
trials [211]. tumors. Immunohistochemical markers have been
Schneider et al. discuss triple-negative breast used to define these subtypes with similar prognos-
cancer, a distinct molecular subtype of breast tic value [289, 290], which allows for breast cancer
cancer that has emerged from DNA microarray subtype assignment in epidemiologic studies and
gene expression studies [287]. Human mammary clinical practice.
glands contain two distinct subtypes of epithelial The distribution of breast cancer subtypes has
cells, basal (myoepithelial) and luminal, which been reported to vary across populations
can be easily distinguished by the pattern of [291–294]. Estrogen receptor-positive, luminal A
expression of certain cytokeratins. Basal cells lie breast cancers were predominant in Asian, white,
closest to the basement membrane and stain posi- and postmenopausal African American popula-
tive for cytokeratin 5/6. Luminal cells compose tions, about 40 % in premenopausal African
the upper, more differentiated, layer of the mam- Americans, and only 27 % in indigenous
mary gland epithelium and express cytokeratin Africans. In contrast, the proportion of the estro-
8/18. The cytokeratin pattern is largely conserved gen receptor-negative, basal-like subtype was
after transformation of epithelial cells, allowing 27 % in indigenous Africans and premenopausal
determination of the cell-type origin of primary African Americans, about 15 % in postmeno-
cancers. Most breast cancers originate from the pausal African Americans and premenopausal
luminal epithelium and express luminal cell- European Americans, and only about 10 % in
specific cytokeratins. It is estimated, however, other populations. There is also a clear gradient
that 3–15 % of all breast cancers seem to origi- in the proportion of estrogen receptor-negative
nate from basal-like epithelium and express unclassified breast tumors across populations,
basal-specific cytokeratins as discussed by with Africans having the highest proportion.
Schneider et al. [287]. Perou and colleagues, Interestingly, the proportion of HER2-positive
using complementary DNA microarrays, first tumors (luminal B and HER2-positive/estrogen
proposed a molecular classification of breast can- receptor-negative subtypes combined) was about
cers based on variations in global gene expres- 15 % in all populations except the Japanese. As
sion patterns [95, 97, 100]. discussed by Garcia-Closas and Chanock [13],
Following these original reports, several other genomic risk factors for these subtypes are also
groups have confirmed that individual cancers likely to vary across populations, which suggest
could be categorized, based on their gene signa- that an integrative epidemiology approach [295]
ture, to at least five distinct subtypes: luminal A, must be applied to drug development at both the
luminal B, normal-like, HER2-like, and basal-like. population and individual levels. Breast cancer
Normal-like tumors resemble normal breast tissue; clinical trials must be conducted in populations
HER2-like are characterized by HER2 overexpres- in which the drugs are to be used, as one can no
sion; luminal A and B are estrogen receptor-posi- longer assume that drugs that are developed in
tive; and basal-like are triple negative (estrogen predominantly European populations will pro-
receptor-negative, progesterone receptor-negative, duce similar outcomes in populations of Asian or
and HER2-negative). BRCA1-associated breast African ancestry.
84 B. Kumar

Genomic Therapeutic Targets breast cancer are attractive targets because they
and Biomarkers are events for which strong evidence indicates
positive selection, so the tumors may be addicted
Given the heterogeneity of expression and to the aberration. In addition, aberrations are not
genomic composition within the distinct sub- present in normal tissues so that therapies against
classes, it might also be expected that there will them are likely to be relatively nontoxic. ERBB2
be distinct clinical responses to chemotherapeu- is the prototypic genome-based therapeutic target
tic and targeted agents within the different sub- [313, 314]. This receptor tyrosine kinase is highly
classes. If agents can be identified that show amplified in about 30 % of human breast cancers,
enhanced efficacy against specific subclasses, and the antibody, trastuzumab [315], and the
this might represent the first step toward more small-molecule inhibitor, lapatinib [316], have
personalized medicine for the treatment of breast proved to be clinically effective against tumors in
cancer. Such class distinctions have already been which ERBB2 is amplified.
identified, in particular for HER2-positive One of the advantages of aberration-targeted
tumors. Targeted therapeutic agents, such as the therapies is that markers can be readily devel-
monoclonal antibody trastuzumab or the small- oped to identify tumors carrying the aberration
molecule tyrosine kinase inhibitor lapatinib, have that are most likely to respond to therapy. In the
both been approved for use in treating patients case of ERBB2-targeted therapies, FISH with
with HER2-positive breast cancer, leading to probes to ERBB2 readily identify the tumor to be
improved survival in these patients [154, 296, treated [317]. Following this lead, therapies
297]. Given that HER2-directed therapies were directed against tumors with aberrations involv-
developed before the identification of breast can- ing TP53 [318], MDM2 [319], TOP2A [320],
cer subclasses, it is now of interest to see if addi- PI3-kinase mutations (PTEN and PIK3CA)
tional subclass-specific agents can be identified. [321], and BRCA1 [322] are now being devel-
Recent studies have demonstrated that inhibi- oped or tested. Clearly, the development of
tors of poly (ADP-ribose) polymerase (PARP), aberration-targeted therapy is only beginning.
an important transducer of BRCA-mediated Biomarker discovery for breast cancer is still
DNA damage response, selectively target BRCA1 very much in its early phase. Multiple approaches
and BRCA2 mutant breast cancer [298]. have been developed that hold promise for the
Numerous recurrently aberrant genes associated identification of serum biomarkers [323]. A bio-
with reduced survival duration and other aspects marker can be a substance that is introduced into
of breast cancer pathophysiology have been sug- an organism as a means to examine organ func-
gested as therapeutic targets in breast cancer. tion or other aspects of health. Dowsett and
Amplified genes implicated as therapeutic targets Dunbier discuss the important role of biomarkers
include ERBB2 [299], TOP2A [300], CCND1 in the optimal selection of treatment for breast
and EMS1 [301], MYC [302], ZNF217 [303], cancer [324]. For example, two multigene expres-
RAB25 [304], MDM2 [305], TBX2 [306] sion profiles have shown the ability to outdo the
RPS6KB1, and the microRNA mir-21 [307]. traditional prognostic and predictive factors.
More recently, correlative analyses of gene Oncotype Dx, a 21-gene reverse transcription
expression and high-level amplification have PCR-based assay [165], has already proved suc-
identified 66 genes in regions of amplification that cessful in identifying subsets of node-negative,
are associated with reduced survival duration that estrogen receptor-positive patients who will ben-
are candidate therapeutic targets, nine of which efit from the addition of chemotherapy to adju-
(FGFR1, IKBKB, ERBB2, PROCC, ADAM9, vant antiestrogen therapy and those who will not.
FNTA, ACACA, PNMT, and NR1D1) were pre- MammaPrint, a 70-gene signature, was devel-
dicted to be drug able [109]. Recurrent mutations oped from studying tumors of women with node-
or deletions in breast cancers include TP53 [308], negative disease, unselected for estrogen receptor
PIK3CA [309], PTEN [310], BRCA1 [311], and status [325, 326]. Tumors from patients who
BRCA2 [312]. Recurrent genomic aberrations in remained disease-free for 10 years were found to
4 Breast Cancer Genomics 85

have distinct profiles compared with patients who 1.4 million cells [329]. Pharmacogenomics is
experience early relapse. Both Oncotype Dx and particularly attractive in oncology as most che-
MammaPrint have the potential to identify motherapy agents have a narrow therapeutic win-
patients with node-negative tumors who do not dow, with severe drug toxicities that can be
require additional therapy, thus sparing these potentially life-threatening. Although advances
women from unnecessary and potentially toxic in adjuvant chemotherapy for breast cancer have
therapy. Both of these assays are Food and Drug led to marked reductions in recurrence and mor-
Administration-approved and are currently tality, women are still concerned about the short-
undergoing prospective validation to more clearly and long-term toxicities associated with
define their roles in the optimization of therapy treatment. Moreover, improvements in chemo-
for node-negative breast cancer patients. therapy have effects that are much more striking
The next decade will surely witness the fur- in women with estrogen receptor-negative tumors
ther explosion of predictive and prognostic than in women with estrogen receptor-positive
genomic markers, and older biomarkers such as tumors, some of whom can now be spared che-
Ki67 will assume more prominence. Expanding motherapy. For example, the risk of death for
on the success of Oncotype Dx and MammaPrint, dose-dense doxorubicin/cyclophosphamide fol-
investigators are trying to identify signatures that lowed by dose-dense paclitaxel (as in INT
predict response to specific therapies. Hess and C9741) when compared with low-dose cyclo-
colleagues have used transcriptional profiling to phosphamide/doxorubicin/5-fluorouracil was
identify genomic signatures, which predict for reduced by 55 % (38–69 %) in women with
response to the T-FAC adjuvant chemotherapy estrogen receptor-negative tumors versus only
regimen [327]. 23 % (−17 to 49 %) in estrogen receptor-positive
An important advance in the last decade is the tumors [330]. Thus, it is conceivable that given
incorporation of trastuzumab into adjuvant ther- the right combination of highly effective and less
apy for HER2-positive breast cancer. Trastuzumab, toxic chemotherapy, women with hormone
a monoclonal antibody to HER2, in combination receptor-negative breast cancer might be expected
with standard adjuvant chemotherapy, has to have a more favorable outcome. In addition,
decreased the risk of recurrence by more than hormonal therapies might be more effective if
50 % [328]. Targeting HER2 is a true success they are given in appropriate doses, taking into
story and shows how identifying and inhibiting a account the genetic variants that affect metabo-
target can transform a very aggressive phenotype lizing enzymes as discussed in the review by Tan
into one with a favorable outcome. Prior to the use et al. [187]. Dr. Elias Zerhouni, the former direc-
of targeted therapy for HER2-positive breast can- tor of the National Institutes of Health, stated in
cer, this form of the disease was associated with a his “new strategic vision for medicine that the
high risk of relapse and a uniformly poor progno- shift from a late curative paradigm to an early
sis. Understanding the biology of HER2-positive preemptive one is becoming increasingly possi-
breast cancer and the development of a tailored ble” [331]. This is one mechanism of containing
therapeutic approach has led to a cure for many the escalating cost of cancer treatment nation-
women with this form of the disease. wide. Electronic health records are another tech-
nology that can help to identify early onset and
thus preempt the trajectory of cancer pathology
Drug Development Through and the ominous consequences of metastasis.
a Genomic Approach

Pharmacogenomics is the science that investi- Genomic Epidemiology

gates how individuals react to medication.
Although the recent sequencing revealed the Genomic epidemiology is defined as the investi-
99.9 % semblance in human DNA makeup, the gation of the actions of genetic factors in deter-
0.1 % single-nucleotide polymorphism is about mining health and disease onset in families and
86 B. Kumar

populations and the interplay of genetics and however, that drug efficacy and toxicity is a poly-
environmental characteristics of people. Genomic genic trait not attributable to a single gene.
epidemiology is the link uniting the intersection Genome-wide association studies are making
between genetic and molecular epidemiology. It the study of multiple genes and SNPs involved in
focuses on the determinants and distribution of drug toxicity and efficacy possible. Several
diseases and injuries in human populations [332, groups have developed genome-wide approaches
333]. There are interindividual and interethnic to identify germ line polymorphisms that corre-
variabilities of drug pharmacokinetics and phar- late with cytotoxicity. Huang et al. have devel-
macodynamics, which may be caused by com- oped a preclinical model that uses the International
monly occurring genetic polymorphisms of HapMap Project lymphoblastoid cell lines [334].
drug-metabolizing enzymes and transporters The HapMap Project contains lymphoblastoid
[334]. Spitz et al. recently proposed a unifying cell lines from four distinct ethnic populations:
premise of integrative epidemiology and sug- Caucasians (Utah, United States), Africans
gested that the same genes that are implicated in (Yorubas from Nigeria), Japanese (Tokyo), and
cancer risk may also be involved in a person’s Chinese (Han from Beijing). These cell lines are
propensity to carcinogenic exposure and/or to an invaluable resource because they have been
modulation of therapeutic outcome [295]. extensively genotyped. When lymphoblastoid
Examples include glutathione-related transporter cell lines are treated in vitro with a chemothera-
genes and the cytochrome P450 enzymes, such as peutic agent, individual cytotoxicity phenotypes
CYP3A4 and CYP191A variants. Variants in are identified, which can then be combined with
these genes have been implicated in cancer risk cell line-specific genotype data to do genetic
and are important pathways for metabolizing association studies. This information can then be
antineoplastic drugs. correlated to gene expression data, the so-called
Tamoxifen, a selective estrogen receptor mod- “triangular” approach, to allow the identification
ulator, is commonly used for both the chemopre- of SNPs that may explain variation in drug sensi-
vention of breast cancer and the treatment of tivity. Importantly, the approach allows for simul-
early- and advanced-stage estrogen receptor- taneous discovery of multiple SNPs involved in
positive breast cancer. Failure of tamoxifen ther- susceptibility, without bias for any particular
apy has long been attributed to intrinsic or gene.
acquired resistance of the tumor to the effects of
estrogen receptor blockade. Tan and colleagues
show that interindividual genetic variability plays Disparities in Clinical Outcomes
a critical role not only in determining toxicity Through a Genomic Approach
from therapy but also in determining benefit, in
some cases [187]. Tamoxifen undergoes exten- Racial disparities have been demonstrated in
sive metabolism via the CYP pathway to several breast cancer mortality and are associated with
primary and secondary metabolites, some of differences in access to care and biological char-
which exhibit more potent antiestrogenic effects acteristics [335–337]. Poor access to care can
than tamoxifen itself on breast cancer cells. lead to diagnosis at a more advanced stage, delays
CYP2D6 is one of the key enzymes in this path- in starting treatment, failure to complete treat-
way that metabolizes tamoxifen to a more active ment, and receipt of inferior chemotherapy regi-
metabolite, endoxifen. There are several variants mens [338–342]. Differences in tumor biology
in the CYP2D6 gene that result in the poor can result in more aggressive cancers, such as
metabolizer phenotype which, in turn, has been those that are negative for estrogen and proges-
shown to correlate with worse outcomes. The terone receptors, as well as human epidermal
majority of variations in drug-metabolizing growth factor receptor 2 (HER2)/neu, so-called
enzymes identified to date have been through a triple-negative cancers. Black patients are more
“candidate gene” approach. It is unlikely, likely to be diagnosed with breast cancer that is
4 Breast Cancer Genomics 87

hormone receptor-negative and at a more Despite the data, undertreatment is common,

advanced stage than white patients, and they are especially among African American women. In a
more likely to receive inferior care [336, 338, retrospective study of more than 20,000 women
343]. Hispanics fare better than black patients, treated in community practices, Lyman and col-
but their rate of overall survival seems to be lower leagues found that 36.5 % of patients received
than that of non-Hispanic white patients [344]. <85 % of their planned chemotherapy [355]. In a
For more than a decade, a prominent question study of 472 women with early-stage breast can-
in health services research has been how to best cer, Hershman et al. found that a substantial frac-
understand the relative contribution of these dif- tion of women terminated their chemotherapy
ferent causes of disparities in cancer survival prematurely and that early termination was sig-
rates, including breast cancer. Differences in nificantly associated with both poorer survival
treatment patterns have been a major focus of and black race [341]. Suboptimal delivery of che-
these efforts. Studies have shown that black motherapy dose and intensity has been associated
patients are more likely than white patients to with decreased efficacy and poorer survival
receive treatment that leads to poorer clinical out- [356–358]. For these reasons, future drug devel-
comes, including nonstandard adjuvant chemo- opment in breast cancer should incorporate
therapy regimens, first-cycle dose reductions genomic markers to identify interindividual and
(regardless of body mass index) [345], delays in interethnic variability of drug pharmacokinetics
radiation therapy [50, 342], and delays in and and pharmacodynamics. With the development
early termination of chemotherapy [341]. of more efficacious and less toxic drug regimens,
More recently, the list of factors known to one can expect to see a reduction in health dis-
worsen outcomes for minorities has expanded to parities for the most vulnerable patients, espe-
include lower levels of English proficiency in cially women of African ancestry.
some immigrant groups, lower education levels,
and poor health literacy. One manifestation is
that patients with low levels of English profi- Breast Cancer Epigenome
ciency may be less able to participate in health-
care decisions, which may affect the treatment DNA methylation is one of the three known lay-
that they receive [346]. Population differences in ers of epigenetic control of germ line and tissue-
the distribution of variants in drug-metabolizing specific gene expression. Hypermethylation
enzymes and transporters might be relevant in plays an integral role in genomic imprinting,
addressing differences in outcomes in diverse wherein one of the two parental alleles of a gene
populations and specifically in addressing health is silenced in order to establish monoallelic
disparities. For example, women of African expression; X-chromosome inactivation in
ancestry have a lower overall incidence of breast females occurs through a similar imprinting
cancer, but a higher overall mortality compared mechanism [359, 360]. Stated simply, DNA
with white women [347, 348]. The disparity in methylation is a heritable, epigenetic change that
outcomes may be partly related to lower toler- alters gene expression and is confined to the addi-
ance for side effects of treatment. Recent studies tion of a methyl group to the 5-carbon position of
focus on disparities in treatment outcomes cytosine in a CpG dinucleotide. In the vertebrate
because differences in socioeconomic factors and genome, CpG dinucleotide sequences have been
tumor biology do not entirely account for the dis- severely depleted to approximately 20 % of the
parity in clinical outcomes. Several studies have predicted frequency during evolution, and among
documented differences in the receipt of cancer the remaining CpG dinucleotides, over 70 % are
treatment by race [345, 349–351], and large clini- methylated [361]. A study of the human genome
cal trials have established that dose delays, reduc- revealed that the distribution of CpG dinucleo-
tions, or early termination of chemotherapy tides is not random and some of them cluster
greatly reduce treatment benefit [352–354]. together to form CpG-rich DNA regions called
88 B. Kumar

CpG islands. CpG islands are mostly located in in BRCA1-methylated cases resembles that of
the upstream promoter and exon 1 region of over BRCA1-mutated cases rather than that of spo-
half of human genes [362]. In normal cells, CpG radic cancers [371].
islands in actively expressed genes are unmethyl- We and others have offered a model of breast
ated. However, during neoplastic transformation, carcinogenesis in which BRCA1 promoter meth-
DNA methylation in cancer cells exhibits inverse ylation serves as a “first hit,” much like an inher-
profiles compared with normal cells; focal hyper- ited germ line mutation, and the “second hit”
methylation of CpG islands of 5′-end regions of results in reduced BRCA1 copy number and/or
many genes is observed [286, 363–365]. Thus, a chromosome 17 aneusomy. In this model,
change in DNA methylation profile is a hallmark BRCA1 promoter hypermethylation occurs early
of almost all human cancers, including breast and, when complete, causes defects in chromo-
cancer. some structure, cell division, and viability [371].
Much of the focus in breast cancer research A BRCA1-deficient cell must acquire additional
over the past few decades has been on breast can- alterations, such as TP53 mutations or MYC
cer genetics: identifying mutations and character- amplification, that overcome these problems and
izing their roles in carcinogenesis. In the last force tumor progression down the same limited
several years, studies have shown that epigenetics set of molecular pathways, similar to the progres-
plays a key role in tumor progression. Epigenetic sion of hereditary BRCA1 mutated tumors.
changes are defined as heritable and reversible Because the majority of BRCA1-mutated breast
changes in gene expression that are not accompa- cancers are basal-like, Foulkes has hypothesized
nied by changes in DNA sequence. In cancer, the that the key function of BRCA1 is to be a stem
main epigenetic mechanisms underlying abnor- cell regulator and promote the differentiation of
mal gene expression include aberrant CpG island glandular epithelium [372]. For this reason, in a
promoter methylation of specific tumor suppres- BRCA1-deficient cell, this transition can fail or
sor genes, global changes in genomic DNA meth- abort, and the basal cell phenotype gene expres-
ylation, and alterations in histone modification sion pattern would be retained. Interestingly, Liu
(deacetylation and methylation; Fig. 4.1). These et al. have recently confirmed that inactivation of
abnormalities can be reversed by inhibitors of BRCA1 in breast epithelial stem cells restricts
both DNA methyltransferases and histone deacet- subsequent progenitor cells to a basal-like cell
ylases [366]. It has been postulated that promoter subtype [373]. Collectively, these observations
methylation may serve as the second “hit” in the suggest that inactivation of BRCA1 by mutation
Knudson two-hit model in sporadic cancer or methylation promotes breast cancer with basal
through inactivation of the normal allele of a tumor phenotype.
tumor suppressor gene, implicated in hereditary In contrast, loss of BRCA2 expression via
cancer syndrome such as BRCA1 [367]. aberrant promoter methylation does not seem to
Hypermethylation of BRCA1 at promoter CpG occur in sporadic cancers. Functional equiva-
islands occurs in 15–31 % of sporadic breast lency between the effect and significance of the
tumors and may play a critical role in sporadic epigenetic silencing of BRCA1 in sporadic breast
breast cancer by inactivating one BRCA1 allele cancer and genetic suppression of the gene in
followed by loss of the wild-type BRCA1 [368]. BRCA1 mutation carriers has implications for
BRCA1 methylation in sporadic breast cancer clinical practice. Lafarge and colleagues showed
seems to result in a similar tumor phenotype to that decreased expression of BRCA1 in the
that seen in BRCA1 mutation carriers [367–370]. HBL100 breast cancer cell line led to increased
These findings are consistent with our study of sensitivity to the DNA-damaging agent cisplatin
C-MYC amplification in BRCA1-deficient breast [374]. As previously discussed, a number of stud-
cancers (BRCA1-mutated hereditary and ies have also correlated loss of BRCA1 with
BRCA1-methylated sporadic), in which we defects in DNA repair and cell cycle checkpoints.
showed that the pattern of CMYC amplification Tumors lacking functional BRCA1 protein show
4 Breast Cancer Genomics 89

a high frequency of chromosomal aneuploidy, certain genes (e.g., RASSF1A, CYP26A1,

characteristic of a defective G2-M checkpoint. KCNAB1, SNCA, HIN-1, TWIST, and Cyclin
Sudo and colleagues have shown that paclitaxel D2) occurs in both premalignant lesions, such as
sensitivity is dependent on an intact checkpoint atypical hyperplasia, and carcinoma of the breast
function [375], thus implying that any interfer- [385, 386]. These findings suggest that epigene-
ence with the spindle assembly checkpoint would tic changes occur early in breast tumorigenesis
generate paclitaxel resistance. BRCA1-deficient and may serve as potential markers for early
cells secondary to mutation are relatively resis- detection or risk assessment. Moreover, specific
tant to paclitaxel [374, 376–378]. Together, these epigenetic changes may have prognostic and/or
data suggest that loss of normal BRCA1 expres- predictive value [387]. These observations are
sion by mutation or epigenetic regulation may being translated into clinical care. For example,
confer a unique chemosensitivity profile. Thus, the National Cancer Institute is sponsoring a
BRCA1 deficiency secondary to promoter meth- study of women at high risk of developing breast
ylation may represent a novel therapeutic target cancer who, following surgical resection for duc-
for the management of a subset of basal-like or tal carcinoma in situ or stage I, II, or III invasive
triple-negative breast cancers. breast cancer, are treated with simvastatin.
Both inherited and sporadic breast cancers can Simvastatin belongs to the statin family of
also exhibit variable estrogen receptor-a expres- 3-hydroxy-3-methylglutaryl-CoA reductase
sion. Interestingly, estrogen receptor-a coding inhibitors, drugs that lower cholesterol in patients
region mutations seem to be quite rare [379, 380], with cardiovascular disease; these agents have a
although there is convincing evidence that estro- theoretical role in chemoprevention through
gen receptor-a is an epigenetically regulated gene downregulating Ras, upregulating p27, and alter-
that can undergo promoter methylation in a sig- ing estrogen receptor levels. The change in meth-
nificant proportion of breast cancers [381] and ylation status across a panel of genes (estrogen
can strongly associate with BRCA1 promoter receptor-a and estrogen receptor-b, cyclin D2,
methylation [382]. An alternative epigenetic RAR-b, Twist, RASSF1A, and HIN-1) that are
mechanism underlying the loss of estrogen known to be frequently and specifically hyper-
receptor-a expression has been suggested by the methylated in breast cancer will be evaluated and
results of cell-based assays analyzing histone correlated with changes in hsCRP, lipid profile,
function as a determinant of gene expression contralateral breast density, and estrogen
[383]. Restoration of estrogen receptor-a expres- concentration.
sion by histone deacetylase inhibitors suggests
that reorganizing the heterochromatin-associated
proteins, without demethylation per se, can Conclusion and Future Perspective
restore functional estrogen receptor-a expression
[383]. This possibility is being explored clini- The command of contemporary genome analysis
cally in an ongoing phase II trial of a new- tools, especially massively similar sequencing, is
generation histone deacetylases inhibitor, increasing at a remarkable rate. It seems clear
vorinostat. The investigators of this trial will that the “thousand-dollar genome” is nearly at
determine whether or not, following vorinostat hand. Modern genome-wide, high-resolution
treatment, a tumor becomes sensitive to hormone breast cancer analyses so far have focused mostly
therapy and/or exhibits increased expression of on evaluation of aberrations in invasive breast
estrogen receptor-a. tumors. This provides a functioning aberration
In addition, there have recently been a number pieces list and recognizes aberrations that may be
of phase I/II trials initiated to investigate combin- useful as prognostic markers or as therapeutic
ing different classes of histone deacetylases targets. They have not, though, provided consid-
inhibitors with traditional therapies for the treat- erable information about how these aberrations
ment of breast cancer [384]. DNA methylation of occur and evolve throughout development. This
90 B. Kumar

information will come from longitudinal incor- part of cancer management. The challenges will
porated “omic” analyses of genome aberration be to manage the information so that it is gener-
appearance during cancer development. Studies ally accessible to the scientific and medical com-
in mouse models can provide some information, munities and to interpret it in ways that lead to
but studies of changes during the evolution of improved cancer management. This will require
individual human cancers will be invaluable. a substantial investment in functional studies,
This will entail a long-term, continued endeavor since the roles of most genomic and epigenomic
from the breast tumor research community to aberrations in cancer pathophysiology are not
gather the obligatory samples and sustain refine- understood. It will also require full development
ment of large-scale omic investigation machiner- and deployment of large-scale information man-
ies so they are capable of investigating the petite agement and interpretation systems. Success in
quantities of neoplastic tissue that may be obtain- these areas will allow the potential of cancer
able at premature stages of evolution. genome and epigenome in treatment personaliza-
Fewer than 30 % of women with metastatic tion to be fully realized.
breast tumor will live 5 years. This is in compari-
son to the cure of premature disease, where
results have been enhanced greatly over the last References
decade. Dozens of subsequent generation thera-
1. Stewart BW, Kleihues P, editors. World cancer report.
pies are being designed to target these aberra- Lyon: IARC Press; 2003. p. 188–9. Breast cancer.
tions. The best developed of these for breast 2. Wun LM, Merrill RM, Feuer EJ. Estimating lifetime
tumors are trastuzumab and lapatinib, which tar- and age‐conditional probabilities of developing can-
get ERBB2-positive tumors. The responses to cer. Lifetime Data Anal. 1980;4:169–86.
3. Foulds L. The experimental study of tumor progres-
these agents, even when combined with predict- sion: a review. Cancer Res. 1954;14:327–39.
able chemotherapeutic agents, are not, though, 4. Nowell PC. The clonal evolution of tumor cell popu-
durable in patients with metastatic sickness, and lations. Science. 1976;194:23–8.
long-term survival is uncommon. 5. Olopade OI, Grushko TA, Nanda R, Huo D. Advances
in breast cancer: pathways to personalized medicine.
It looks likely that recent cure strategies fail Clin Cancer Res. 2008;14:7988–99.
because these strategies do not take into account 6. Bodmer W, Tomlinson I. Rare genetic variants and the
the genomic and epigenomic aberrations that risk of cancer. Curr Opin Genet Dev. 2010;20(3):
contribute to resistance in metastatic breast 262–7.
7. Hall JM, Lee MK, Newman B, Morrow JE, Anderson
tumors; resistance-related homeostatic, or feed- LA, Huey B, et al. Linkage of early-onset familial
back loops induced by pathway-targeted thera- breast cancer to chromosome 17q21. Science.
pies; and factors unique to the metastatic 1990;250(4988):1684–9.
microenvironments in the bone marrow, brain, 8. Venkitaraman AR. Cancer susceptibility and the func-
tions of BRCA1 and BRCA2. Cell. 2002;108(2):
liver, and lung that contribute to therapeutic resis- 171–82.
tance. Improvement of a meticulous “omic” con- 9. Yun MH, Hiom K. CtIP-BRCA1 modulates the choice
siderate of drug-resistant, metastatic tumors will of DNA double strand- break repair pathway through-
greatly facilitate cure of this important aspect of out the cell cycle. Nature. 2009;459(7245):460–3.
10. Huen MS, Sy SM, Chen J. BRCA1 and its toolbox for
the disease. This will necessitate development of the maintenance of genome integrity. Nat Rev Mol
clinical trials in which samples of metastatic Cell Biol. 2010;11(2):138–48.
breast tissue suitable for large-scale omic investi- 11. Jensen RB, Carreira A, Kowalczykowski SC. Purified
gation are acquired before and after development human BRCA2 stimulates RAD51-mediated recom-
bination. Nature. 2010;467(7316):678–83.
of medicine resistance, as well as the develop- 12. Børresen AL, Andersen TI, Garber J, Barbier-Piraux
ment of experimental model systems that mirror N, Thorlacius S, Eyfjord J, et al. Screening for germ
the aberrations that contribute to resistance. line TP53 mutations in breast cancer patients. Cancer
It looks reasonable to anticipate that compre- Res. 1992;52(11):3234–6.
13. Garcia-Closas M, Chanock S. Genetic susceptibility
hensive cancer scans of genomic and epigenomic loci for breast cancer by estrogen receptor (ER) status.
aberrations will become an important and routine Clin Cancer Res. 2008;14:8000–9.
4 Breast Cancer Genomics 91

14. Nelen MR, Padberg GW, Peeters EA, Lin AY, van den 30. Easton DF, Pooley KA, Dunning AM, Pharoah PD,
Helm B, Frants RR, et al. Localization of the gene for Thompson D, Ballinger DG, et al. Genome-wide
Cowden disease to chromosome 10q22-23. Nat association study identifies novel breast cancer sus-
Genet. 1996;13(1):114–6. ceptibility loci. Nature. 2007;447:1087–93.
15. Boardman LA, Thibodeau SN, Schaid DJ, Lindor 31. Turner N, Grose R. Fibroblast growth factor signal-
NM, McDonnell SK, Burgart LJ, et al. Increased risk ling: from development to cancer. Nat Rev Cancer.
for cancer in patients with the Peutz–Jeghers syn- 2010;10(2):116–29.
drome. Ann Intern Med. 1998;128(11):896–9. 32. Katoh Y, Katoh M. FGFR2-related pathogenesis and
16. Antoniou AC, Easton DF. Models of genetic suscepti- FGFR2-targeted therapeutics. Int J Mol Med. 2009;
bility to breast cancer. Oncogene. 2006;25(43): 23(3):307–11.
5898–905. 33. Nilsson EM, Brokken LJ, Härkönen PL. Fibroblast
17. Berx G, Van Roy F. The E-cadherin/catenin complex: growth factor 8 increases breast cancer cell growth by
an important gatekeeper in breast cancer tumorigene- promoting cell cycle progression and by protecting
sis and malignant progression. Breast Cancer Res. against cell death. Exp Cell Res. 2010;316(5):
2001;3(5):289–93. 800–12.
18. Meindl A, Hellebrand H, Wiek C, Erven V, 34. Cox A, Dunning AM, Garcia-Closas M,
Wappenschmidt B, Niederacher D, et al. Germline Balasubramanian S, Reed MW, Pooley KA, et al.
mutations in breast and ovarian cancer pedigrees A common coding variant in CASP8 is associated
establish RAD51C as a human cancer susceptibility with breast cancer risk. Nat Genet. 2007;39(3):
gene. Nat Genet. 2010;42(5):410–4. 352–8.
19. Thompson D, Duedal S, Kirner J, McGuffog L, Last 35. Latif A, Hadfield KD, Roberts SA, Shenton A, Lalloo
J, Reiman A, et al. Cancer risks and mortality in het- F, Black GC, et al. Breast cancer susceptibility vari-
erozygous ATM mutation carriers. J Natl Cancer Inst. ants alter risks in familial disease. J Med Genet.
2005;97(11):813–22. 2010;47(2):126–31.
20. Rahman N, Seal S, Thompson D, Kelly P, Renwick A, 36. Stacey SN, Manolescu A, Sulem P, Rafnar T,
Elliot A, et al. PALB2, which encodes a BRCA2- Gudmundsson J, Gudjonsson SA, et al. Common vari-
interacting protein, is a breast cancer susceptibility ants on chromosomes 2q35 and 16q12 confer suscep-
gene. Nat Genet. 2007;39(2):165–7. tibility to estrogen receptor positive breast cancer. Nat
21. Robson M. CHEK2, breast cancer, and the under- Genet. 2007;39(7):865–9.
standing of clinical utility. Clin Genet. 2010;78(1): 37. Smid M, Wang Y, Klijn JG, Sieuwerts AM, Zhang Y,
8–10. Atkins D, et al. Genes associated with breast cancer
22. Hirshfield KM, Rebbeck TR, Levine AJ. Germline metastatic to bone. J Clin Oncol. 2006;24(15):
mutations and polymorphisms in the origins of can- 2261–7.
cers in women. J Oncol. 2010;2010:297671. 38. Udler MS, Ahmed S, Healey CS, Meyer K, Struewing J,
23. Orr N, Chanock S. Common genetic variation and Maranian M, et al. Fine scale mapping of the breast
human disease. Adv Genet. 2008;62:1–32. cancer 16q12 locus. Hum Mol Genet. 2010;19(12):
24. Stratton MR, Rahman N. The emerging landscape of 2507–15.
breast cancer susceptibility. Nat Genet. 2008;40(1): 39. Rebbeck TR, DeMichele A, Tran TV, Panossian S,
17–22. Bunun GR, Troxel AB, et al. Hormone-dependent
25. Dickson SP, Wang K, Krantz I, Hakonarson H, effects of FGFR2 and MAP3K1 in breast cancer sus-
Goldstein DB. Rare variants create synthetic genome- ceptibility in a population-based sample of Post-
wide associations. PLoS Biol. 2010;8(1):e1000294. menopausal African-American and European-American
26. Orozco G, Barrett JC, Zeggini E. Synthetic associa- women. Carcinogenesis. 2009;30(2):269–74.
tions in the context of genome-wide association scan 40. Amundadottir LT, Sulem P, Gudmundsson J, Helgason
signals. Hum Mol Genet. 2010;19(R2):R137–44. A, Baker A, Sigurdsson A, et al. A common variant
27. Hindorff LA, Sethupathy P, Junkins HA, Ramos EM, associated with prostate cancer in European and
Mehta JP, Collins FS, et al. Potential etiologic and African populations. Nat Genet. 2006;38(6):652–8.
functional implications of genome-wide association 41. Milne RL, Benítez J, Nevanlinna H, Heikkinen T,
loci for human diseases and traits. Proc Natl Acad Sci Aittomäki K, Blomqvist C, et al. Risk of estrogen
U S A. 2009;106(23):9362–7. receptor-positive and -negative breast cancer and sin-
28. Durbin RM, Abecasis GR, Altshuler DL, Auton A, gle-nucleotide polymorphism 2q35-rs13387042. J Natl
Brooks LD, Gibbs RA, et al, 1000 Genomes Project Cancer Inst. 2009;101(4):1012–8.
Consortium. A map of human genome variation from 42. Ahmed S, Thomas G, Ghoussaini M, et al. Newly dis-
population-scale sequencing. Nature. 2010;467: covered breast cancer susceptibility loci on 3p24 and
1061–73. 17q23.2. Nat Genet. 2009;41(5):585–90.
29. Hunter DJ, Kraft P, Jacobs KB, Cox DG, Yeager M, 43. Stacey SN, Manolescu A, Sulem P, Thorlacius S,
Hankinson SE, et al. A genome-wide association Gudjonsson SA, Jonsson GF, et al. Common variants
study identifies alleles in FGFR2 associated with risk on chromosome 5p12 confer susceptibility to estro-
of sporadic postmenopausal breast cancer. Nat Genet. gen receptor-positive breast cancer. Nat Genet. 2008;
2007;39:870–4. 40(6):703–6.
92 B. Kumar

44. Turnbull C, Ahmed S, Morrison J, Pernet D, Renwick ciated with specific mutations of BRCA1 and BRCA2
A, Maranian M, et al. Genome-wide association study among Ashkenazi Jews. N Engl J Med. 1997;336:
identifies five new breast cancer susceptibility loci. 1401–8.
Nat Genet. 2010;42(6):504–7. 59. Gayther SA, Mangion J, Russell P, Seal S, Barfoot R,
45. Thomas G, Jacobs KB, Kraft P, Yeager M, Wacholder Ponder BA, et al. Variation of risks of breast and ovar-
S, Cox DG, et al. A multistage genome-wide associa- ian cancer associated with different germline muta-
tion study in breast cancer identifies two new risk tions of the BRCA2 gene. Nat Genet. 1997;15:103–5.
alleles at 1p11.2 and 14q24.1 (RAD51L1). Nat Genet. 60. Risch HA, McLaughlin JR, Cole DE, Rosen B,
2009;41(5):579–84. Bradley L, Fan I, et al. Population BRCA1 and
46. Peto J. Breast cancer susceptibility—a new look at an BRCA2 mutation frequencies and cancer penetrances:
old model. Cancer Cell. 2002;1:411–2. a kin-cohort study in Ontario, Canada. J Natl Cancer
47. Houlston RS, Peto J. The search for low-penetrance Inst. 2006;98:1694–706.
cancer susceptibility alleles. Oncogene. 2004;23: 61. Courjal F, Cuny M, Simony-Lafontaine J, Louason G,
6471–6. Speiser P, Zeillinger R, et al. Mapping of DNA ampli-
48. Easton DF, Eeles RA. Genome-wide association stud- fications at 15 chromosomal localizations in 1875
ies in cancer. Hum Mol Genet. 2008;17:R109–15. breast tumors: definition of phenotypic groups.
49. Meijers-Heijboer H, van den Ouweland A, Klijn J, Cancer Res. 1997;57:4360–7.
Wasielewski M, de Snoo A, Oldenburg R, et al. Low- 62. Kononen J, Bubendorf L, Kallioniemi A, Barlund M,
penetrance susceptibility to breast cancer due to Schraml P, Leighton S, et al. Tissue microarrays for
CHEK2(*) 1100delC in noncarriers of BRCA1 or high-throughput molecular profiling of tumor speci-
BRCA2 mutations. Nat Genet. 2002;31:55–9. mens. Nat Med. 1998;4:844–7.
50. Gold B, Kirchhoff T, Stefanov S, Lautenberger J, 63. Sjöblom T, Jones S, Wood LD, Parsons DW, Lin J,
Viale A, Garber J, et al. Genome-wide association Barber TD, et al. The consensus coding sequences of
study provides evidence for a breast cancer risk locus human breast and colorectal cancers. Science. 2006;
at 6q22.33. Proc Natl Acad Sci U S A. 2008;105: 314(5797):268–74.
4340–5. 64. Wood LD, Parsons DW, Jones S, Lin J, Sjoblom T,
51. Stadler ZK, Gallagher DJ, Thom P, Offit K. Genome- Leary RJ, et al. The genomic landscapes of human
wide association studies of cancer: principles and poten- breast and colorectal cancers. Science. 2007;
tial utility. Oncology (Williston Park). 2010;24(27): 318(5853):1108–13.
629–37. 65. Ellis MJ, Ding L, Shen D, Luo J, Suman VJ, Wallis
52. Reeves GK, Travis RC, Green J, Bull D, Tipper S, JW, et al. Whole-genome analysis informs breast can-
Baker K, et al. Incidence of breast cancer and its sub- cer response to aromatase inhibition. Nature.
types in relation to individual and multiple low- 2012;486(7403):353–60.
penetrance genetic susceptibility loci. JAMA. 2010; 66. Shah SP, Roth A, Goya R, Oloumi A, Ha G, Zhao Y,
304(4):426–34. et al. The clonal and mutational evolution spectrum of
53. Antoniou AC, Wang X, Fredericksen ZS, McGuffog primary triple-negative breast cancers. Nature. 2012;
L, Tarrell R, Sinilnikova OM, et al. A locus on 19p13 486(7403):395–9.
modifies risk of breast cancer in BRCA1 mutation 67. Korbel JO, Urban AE, Affourtit JP, Godwin B,
carriers and is associated with hormone receptor- Grubert F, Simons JF, et al. Paired-end mapping
negative breast cancer in the general population. Nat reveals extensive structural variation in the human
Genet. 2010;42(10):885–92. genome. Science. 2007;318(5849):420–6.
54. Forsti A, Hemminki K. Breast cancer genomics based 68. Stephens PJ, McBride DJ, Lin ML, Varela I, Pleasance
on biobanks. Methods Mol Biol. 2011;675:375–85. ED, Simpson JT, Stebbings LA, et al. Complex land-
55. Antoniou A, Pharoah PD, Narod S, Risch HA, Eyfjord scapes of somatic rearrangement in human breast can-
JE, Hopper JL, et al. Average risks of breast and ovar- cer genomes. Nature. 2009;462(7276):1005–10.
ian cancer associated with BRCA1 or BRCA2 muta- 69. Banerji S, Cibulskis K, Rangel-Escareno C, Brown
tions detected in case Series unselected for family KK, Carter SL, Frederick AM, et al. Sequence analy-
history: a combined analysis of 22 studies. Am J Hum sis of mutations and translocations across breast can-
Genet. 2003;72:1117–30. cer subtypes. Nature. 2012;486(7403):405–9.
56. Brose MS, Rebbeck TR, Calzone KA, Stopfer JE, 70. Shah SP, Morin RD, Khattra J, Prentice L, Pugh T,
Nathanson KL, Weber BL. Cancer risk estimates for Burleigh A, et al. Mutational evolution in a lobular
BRCA1 mutation carriers identified in a risk evalua- breast tumour profiled at single nucleotide resolution.
tion program. J Natl Cancer Inst. 2002;18:1365–72. Nature. 2009;461(7265):809–13.
57. Ford D, Easton DF, Stratton M, Narod S, Goldgar D, 71. Nik-Zainal S, Van Loo P, Wedge DC, Alexandrov LB,
Devilee P, et al. Genetic heterogeneity and penetrance Greenman CD, Lau KW, et al. The life history of 21
analysis of the BRCA1 and BRCA2 genes in breast breast cancers. Cell. 2012;149(5):994–1007.
cancer families. The Breast Cancer Linkage 72. Curtis C, Shah SP, Chin SF, Turashvili G, Rueda OM,
Consortium. Am J Hum Genet. 1998;62:676–89. Dunning MJ, et al. The genomic and transcriptomic
58. Struewing JP, Hartge P, Wacholder S, Baker SM, architecture of 2,000 breast tumours reveals novel
Berlin M, McAdams M, et al. The risk of cancer asso- subgroups. Nature. 2012;486(7403):346–52.
4 Breast Cancer Genomics 93

73. TCGA. Comprehensive molecular portraits of human 89. Schwarz JM, Rodelsperger C, Schuelke M, Seelow
breast tumours. Nature. 2012;490:61–70. D. MutationTaster evaluates disease-causing potential of
74. Stephens P, Edkins S, Davies H, Greenman C, Cox C, sequence alterations. Nat Methods. 2010;7(8):575–6.
Hunter C, et al. A screen of the complete protein 90. Van Dyke T, Jacks T. Cancer modeling in the mod-
kinase gene family identifies diverse patterns of ern era: progress and challenges. Cell. 2002;108(2):
somatic mutations in human breast cancer. Nat Genet. 135–44.
2005;37(6):590–2. 91. Kan Z, Jaiswal BS, Stinson J, Janakiraman V, Bhatt
75. Ding L, Ellis MJ, Li S, Larson DE, Chen K, Wallis D, Stern HM, et al. Diverse somatic mutation pat-
JW, et al. Genome remodelling in a basal-like breast terns and pathway alterations in human cancers.
cancer metastasis and xenograft. Nature. 2010; Nature. 2010;466(7308):869–73.
464(7291):999–1005. 92. Sjöblom T. Systematic analyses of the cancer
76. Edgren H, Murumagi A, Kangaspeska S, Nicorici D, genome: lessons learned from sequencing most of
Hongisto V, Kleivi K, et al. Identification of fusion the annotated human protein-coding genes. Curr
genes in breast cancer by paired-end RNA-sequencing. Opin Oncol. 2008;20(1):66–71.
Genome Biol. 2011;12(1):R6. 93. Carter H, Chen S, Isik L, Tyekucheva S, Velculescu
77. Stephens PJ, Tarpey PS, Davies H, Van Loo P, VE, Kinzler KW, et al. Cancer-specific high-
Greenman C, Wedge DC, Nik-Zainal S, et al. The throughput annotation of somatic mutations: compu-
landscape of cancer genes and mutational processes in tational prediction of driver missense mutations.
breast cancer. Nature. 2012;486(7403):400–4. Cancer Res. 2009;69(16):6660–7.
78. Dickson D. Wellcome funds cancer database. Nature. 94. Dees ND, Zhang Q, Kandoth C, Wendl MC,
1999;401(6755):729. Schierding W, Koboldt DC, et al. MuSiC: identify-
79. Collins FS, Barker AD. Mapping the cancer genome. ing mutational significance in cancer genomes.
Pinpointing the genes involved in cancer will help Genome Res. 2012;22(8):1589–98.
chart a new course across the complex landscape of 95. Sørlie T, Perou CM, Tibshirani R, Aas T, Geisler S,
human malignancies. Sci Am. 2007;296(3):50–7. Johnsen H, et al. Gene expression patterns of breast
80. International Cancer Genome Consortium, Hudson carcinomas distinguish tumor subclasses with clini-
TJ, Anderson W, Artez A, Barker AD, Bell C, Bernabé cal implications. Proc Natl Acad Sci U S A. 2001;98:
RR, et al. International network of cancer genome 10869–74.
projects. Nature. 2010;464(7291):993–8. 96. Hedenfalk I, Duggan D, Chen Y, Radmacher M,
81. Nik-Zainal S, Alexandrov LB, Wedge DC, Van Loo P, Bittner M, Simon R, et al. Gene-expression profiles
Greenman CD, Raine K, et al. Mutational processes in hereditary breast cancer. N Engl J Med. 2001;344:
molding the genomes of 21 breast cancers. Cell. 539–48.
2012;149(5):979–93. 97. Sørlie T, Tibshirani R, Parker J, Hastie T, Marron JS,
82. Hornsby C, Page KM, Tomlinson IPM. What can we Nobel A, et al. Repeated observation of breast tumor
learn from the population incidence of cancer? subtypes in independent gene expression data sets.
Armitage and Doll revisited. Lancet Oncol. 2007; Proc Natl Acad Sci U S A. 2003;100:8418–23.
8(11):1030–8. 98. Hedenfalk I, Ringner M, Ben-Dor A, Yakhini Z,
83. Beerenwinkel N, Antal T, Dingli D, Traulsen A, Chen Y, Chebil G, Meltzer P, et al. Molecular classi-
Kinzler KW, Velculescu VE, et al. Genetic progres- fication of familial non-BRCA1/BRCA2 breast can-
sion and the waiting time to cancer. PLoS Comput cer. Proc Natl Acad Sci U S A. 2003;100:2532–7.
Biol. 2007;3(11):e225. 99. Weigelt B, Reis-Filho JS. Histological and molecu-
84. Greenman C, Wooster R, Futreal PA, Stratton MR, lar types of breast cancer: is there a unifying taxon-
Easton DF. Statistical analysis of pathogenicity of omy? Nat Rev Clin Oncol. 2009;6:718–30.
somatic mutations in cancer. Genetics. 2006;173(4): 100. Perou CM, Sørlie T, Eisen MB, van de Rijn M,
2187–98. Jeffrey SS, Rees CA, et al. Molecular portraits of
85. Greenman C, Stephens P, Smith R, Dalgliesh GL, human breast tumours. Nature. 2000;406:747–52.
Hunter C, Bignell G, et al. Patterns of somatic muta- 101. Prat A, Parker JS, Karginova O, Fan C, Livasy C,
tion in human cancer genomes. Nature. 2007; Herschkowitz JI, et al. Phenotypic and molecular
446(7132):153–8. characterization of the claudin-low intrinsic subtype
86. Ng PC, Henikoff S. SIFT: predicting amino acid of breast cancer. Breast Cancer Res. 2010;12:R68.
changes that affect protein function. Nucleic Acids 102. Parker JS, Mullins M, Cheang MCU, Leung S,
Res. 2003;31(13):3812–4. Voduc D, Vickery T, et al. Supervised risk predictor
87. Adzhubei IA, Schmidt S, Peshkin L, Ramensky VE, of breast cancer based on intrinsic subtypes. J Clin
Gerasimova A, Bork P, et al. A method and server for Oncol. 2009;27:1160–7.
predicting damaging missense mutations. Nat 103. Hu Z, Fan C, Oh DS, Marron JS, He X, Qaqish BF,
Methods. 2010;7(4):248–9. et al. The molecular portraits of breast tumors are
88. Thomas PD, Campbell MJ, Kejariwal A, Karlak B, conserved across microarray platforms. BMC
Daverman R, Diemer K, et al. PANTHER: a library of Genomics. 2006;7:96.
protein families and subfamilies indexed by function. 104. Weigelt B, Mackay A, A’hern R, Natrajan R, Tan
Genome Res. 2003;13(9):2129–41. DSP, Dowsett M, et al. Breast cancer molecular
94 B. Kumar

profiling with single sample predictors: a retrospec- 119. Claus EB, Risch N, Thompson WD. Autosomal
tive analysis. Lancet Oncol. 2010;11:339–49. dominant inheritance of early-onset breast cancer.
105. Perou CM, Parker JS, Prat A, Ellis MJ, Bernard Implications for risk prediction. Cancer. 1994;73:
PS. Clinical implementation of the intrinsic subtypes 643–51.
of breast cancer. Lancet Oncol. 2010;11:718–9. 120. Berry DA, Parmigiani G, Sanchez J, Schildkraut J,
106. Sørlie T, Borgan E, Myhre S, Vollan HKM, Russnes Winer E. Probability of carrying a mutation of
H, Zhao X, et al. The importance of gene-centring breast–ovarian cancer gene BRCA1 based on family
microarray data. Lancet Oncol. 2010;11:719–20. history. J Natl Cancer Inst. 1997;89:227–37.
107. Dunning MJ, Curtis C, Barbosa-Morais NL, Caldas 121. Euhus DM. Understanding mathematical models for
C, Tavaré S, Lynch AG. The importance of platform breast cancer risk assessment and counseling. Breast
annotation in interpreting microarray data. Lancet J. 2001;7:224–32.
Oncol. 2010;11:717. 122. Jacobi CE, de Bock GH, Siergerink B, van Asperen
108. Bergamaschi A, Kim YH, Wang P, Sørlie T, CJ. Differences and similarities in breast cancer risk
Hernandez-Boussard T, Lonning PE, et al. Distinct assessment models in clinical practice: which model
patterns of DNA copy number alteration are associ- to choose? Breast Cancer Res Treat. 2009;115:
ated with different clinicopathological features and 381–90.
gene-expression subtypes of breast cancer. Genes 123. Tyrer J, Duffy SW, Cuzick J. A breast cancer predic-
Chromosomes Cancer. 2006;45:1033–40. tion model incorporating familial and personal risk
109. Chin K, DeVries S, Fridlyand J, Spellman PT, factors. Stat Med. 2004;23:1111–30.
Roydasgupta R, Kuo WL, et al. Genomic and tran- 124. Evans DGR, Howell A. Breast cancer risk-assessment
scriptional aberrations linked to breast cancer patho- models. Breast Cancer Res. 2007;9:213.
physiologies. Cancer Cell. 2006;10:529–41. 125. MacPherson G, Healey CS, Teare MD,
110. Chin SF, Teschendorff AE, Marioni JC, Wang Y, Balasubramanian SP, Reed MWR, Pharoah PD,
Barbosa-Morais NL, Thorne NP, et al. High- et al. Association of a common variant of the CASP8
resolution aCGH and expression profiling identifies gene with reduced risk of breast cancer. J Natl
a novel genomic subtype of ER negative breast can- Cancer Inst. 2004;96:1866–9.
cer. Genome Biol. 2007;8:R215. 126. Dunning AM, Healey CS, Pharoah PDP, Teare MD,
111. Chin SF, Wang Y, Thorne NP, Teschendorff AE, Ponder BAJ, Easton DF. A systematic review of
Pinder SE, Vias M, et al. Using array-comparative genetic polymorphisms and breast cancer risk.
genomic hybridization to define molecular portraits of Cancer Epidemiol Biomarkers Prev. 1999;8:843–54.
primary breast cancers. Oncogene. 2007;26:1959–70. 127. Hirschhorn JN, Daly MJ. Genome-wide association
112. Chari R, Coe BP, Vucic EA, Lockwood WW, Lam studies for common diseases and complex traits. Nat
WL. An integrative multidimensional genetic and Rev Genet. 2005;6:95–108.
epigenetic strategy to identify aberrant genes and 128. Frazer KA, Murray SS, Schork NJ, Topol EJ. Human
pathways in cancer. BMC Syst Biol. 2010;4:67. genetic variation and its contribution to complex
113. Jain AN, Chin K, Børresen-Dale A-L, Erikstein BK, traits. Nat Rev Genet. 2009;10:241–51.
Eynstein Lonning P, Kaaresen R, et al. Quantitative 129. Garcia-Closas M, Hall P, Nevanlinna H, Pooley K,
analysis of chromosomal CGH in human breast Morrison J, Richesson DA, et al. Heterogeneity of
tumors associates copy number abnormalities with breast cancer associations with five susceptibility
p53 status and patient survival. Proc Natl Acad Sci U loci by clinical and pathological characteristics.
S A. 2001;98:7952–7. PLoS Genet. 2008;4:e1000054.
114. Aguilera A, Gómez-González B. Genome instabil- 130. Antoniou AC, Spurdle AB, Sinilnikova OM, Healey
ity: a mechanistic view of its causes and conse- S, Pooley KA, Schmutzler RK, et al. Common breast
quences. Nat Rev Genet. 2008;9:204–17. cancer-predisposition alleles are associated with
115. Hicks J, Krasnitz A, Lakshmi B, Navin NE, Riggs breast cancer risk in BRCA1 and BRCA2 mutation
M, Leibu E, et al. Novel patterns of genome rear- carriers. Am J Hum Genet. 2008;82:937–48.
rangement and their association with survival in 131. Pharoah PDP, Antoniou AC, Easton DF, Ponder
breast cancer. Genome Res. 2006;16:1465–79. BAJ. Polygenes, risk prediction, and targeted pre-
116. Russnes HG, Vollan HKM, Lingjaerde OC, Krasnitz vention of breast cancer. N Engl J Med. 2008;358:
A, Lundin P, Naume B, et al. Genomic architecture 2796–803.
characterizes tumor progression paths and fate in 132. Bodmer W, Bonilla C. Common and rare variants in
breast cancer patients. Sci Transl Med. 2010;2:38ra47. multifactorial susceptibility to common diseases.
117. Hanahan D, Weinberg RA. Hallmarks of cancer: the Nat Genet. 2008;40:695–701.
next generation. Cell. 2011;144:646–74. 133. Ambrosone CB. The promise and limitations of
118. Gail MH, Brinton LA, Byar DP, Corle DK, Green genome-wide association studies to elucidate the
SB, Schairer C, Mulvihill JJ. Projecting individual- causes of breast cancer. Breast Cancer Res. 2007;
ized probabilities of developing breast cancer for 9:114.
white females who are being examined annually. 134. Kaye J. The regulation of direct-to-consumer genetic
J Natl Cancer Inst. 1989;81:1879–86. tests. Hum Mol Genet. 2008;17:R180–3.
4 Breast Cancer Genomics 95

135. Lenzer J, Brownlee S. Knowing me, knowing you. carcinomas: a study of interobserver agreement.
BMJ. 2008;336:858–60. Hum Pathol. 1995;26:873–9.
136. Geransar R, Einsiedel E. Evaluating online direct-to- 149. Ries LAG, Melbert D, Krapcho M, Mariotto A, Miller
consumer marketing of genetic tests: informed choices BA, Feuer EJ, et al. SEER cancer statistics review.
or buyers beware? Genet Test. 2008;12:13–23. Bethesda: National Cancer Institute; 2007. Available
137. Hogarth S, Javitt G, Melzer D. The current land- from: http://seer.cancer.gov/csr/1975_2004/.
scape for direct-to-consumer genetic testing: Legal, 150. Elledge RM, Fuqua SA. Estrogen and progesterone
ethical, and policy issues. Annu Rev Genomics Hum receptors. In: Harris JR, editor. Diseases of the
Genet. 2008;9:161–82. breast. Philadelphia: Lippincott, Williams, and
138. Sharma P, Sahni NS, Tibshirani R, Skaane P, Urdal Wilkins; 2000. p. 471–88.
P, Berghagen H, et al. Early detection of breast can- 151. Early Breast Cancer Trialists’ Collaborative Group
cer based on gene-expression patterns in peripheral (EBCTCG). Effects of chemotherapy and hormonal
blood cells. Breast Cancer Res. 2005;7:R634–44. therapy for early breast cancer on recurrence and
139. Sastre-Garau X, Joule M, Asselain B, Vincent- 15-year survival: an overview of the randomised tri-
Salomon A, Beuzeboc P, Dorval T, et al. Infiltrating als. Lancet. 2005;365:1687–717.
lobular carcinoma of the breast. Clinicopathologic 152. Slamon DJ, Godolphin W, Jones LA, Holt JA, Wong
analysis of 975 cases with reference to data on con- SG, Keith DE, et al. Studies of the HER-2-neu proto-
servative therapy and metastatic patterns. Cancer. oncogene in human breast and ovarian cancer.
1996;77:113–20. Science. 1989;244:707–12.
140. Louwman MWJ, Vriezen M, van Beek MWP, 153. Wolff AC, Hammond MEH, Schwartz JN, Hagerty
Nolthenius-Puylaert MC, van der Sangen MJC, KL, Allred DC, Cote RJ, et al. American Society of
Roumen RM, et al. Uncommon breast tumors in per- Clinical Oncology/College of American Pathologists
spective: incidence, treatment and survival in the guideline recommendations for human epidermal
Netherlands. Int J Cancer. 2007;121:127–35. growth factor receptor 2 testing in breast cancer.
141. Kleer CG, van Golen KL, Merajver SD. Molecular J Clin Oncol. 2007;25:118–45.
biology of breast cancer metastasis. Inflammatory 154. Slamon DJ, Leyland-Jones B, Shak S, Fuchs H,
breast cancer: clinical syndrome and molecular Paton V, Bajamonde A, et al. Use of chemotherapy
determinants. Breast Cancer Res. 2000;2:423–9. plus a monoclonal antibody against HER2 for meta-
142. Engel J, Eckel R, Aydemir U, Aydemir S, Kerr J, static breast cancer that overexpresses HER2.
Schlesinger-Raab A, et al. Determinants and prog- N Engl J Med. 2001;344:783–92.
noses of locoregional and distant progression in 155. Irvin Jr WJ, Carey LA. What is triple-negative breast
breast cancer. Int J Radiat Oncol Biol Phys. 2003; cancer? Eur J Cancer. 2008;44:2799–805.
55:1186–95. 156. Urruticoechea A, Smith IE, Dowsett M. Proliferation
143. Ross JS, Harbeck N. Prognostic and predictive fac- marker Ki-67 in early breast cancer. J Clin Oncol.
tors overview. In: Ross JS, Hortobagyi GN, editors. 2005;23:7212–20.
Molecular oncology of breast cancer. Sudbury: 157. Dowsett M, Smith IE, Ebbs SR, Dixon JM, Skene A,
Jones and Bartlett Publishers; 2005. p. 128–41. A’hern R, et al, IMPACT Trialists Group. Prognostic
144. Park CC, Mitsumori M, Nixon A, Recht A, Connolly value of Ki67 expression after short-term presurgical
J, Gelman R, et al. Outcome at 8 years after breast- endocrine therapy for primary breast cancer. J Natl
conserving surgery and radiation therapy for inva- Cancer Inst. 2007;99:167–70.
sive breast cancer: influence of margin status and 158. Jones RL, Salter J, A’hern R, Nerurkar A, Parton M,
systemic therapy on local recurrence. J Clin Oncol. Reis-Filho JS, et al. The prognostic significance of Ki67
2000;18:1668–75. before and after neoadjuvant chemotherapy in breast
145. Kunos C, Latson L, Overmoyer B, Silverman P, cancer. Breast Cancer Res Treat. 2009;116:53–68.
Shenk R, Kinsella T, Lyons J. Breast conservation 159. Ravdin PM, Siminoff LA, Davis GJ, Mercer MB,
surgery achieving >2 mm tumor-free margins results Hewlett J, Gerson N, et al. Computer program to
in decreased local-regional recurrence rates. Breast assist in making decisions about adjuvant therapy for
J. 2006;12:28–36. women with early breast cancer. J Clin Oncol.
146. Elston CW, Ellis IO. Pathological prognostic factors in 2001;19:980–91.
breast cancer. I The value of histological grade in 160. Van’t Veer LJ, Dai H, van de Vijver MJ, He YD, Hart
breast cancer: experience from a large study with long- AA, Mao M, et al. Gene expression profiling pre-
term follow-up. Histopathology. 1991;19:403–10. dicts clinical outcome of breast cancer. Nature.
147. Henson DE, Ries L, Freedman LS, Carriaga 2002;415:530–6.
M. Relationship among outcome, stage of disease, 161. van de Vijver MJ, He YD, van’t Veer LJ, Dai H, Hart
and histologic grade for 22,616 cases of breast can- AA, Voskuil DW, et al. A gene-expression signature
cer: the basis for a prognostic index. Cancer. as a predictor of survival in breast cancer. N Engl J
1991;68:2142–9. Med. 2002;347:1999–2009.
148. Robbins P, Pinder S, de Klerk N, Dawkins H, Harvey 162. Ring BZ, Seitz RS, Beck R, Shasteen WJ, Tarr SM,
J, Sterrett G, et al. Histological grading of breast Cheang MC, et al. Novel prognostic
96 B. Kumar

immunohistochemical biomarker panel for estrogen CAT) and survival after treatment for breast cancer.
receptor-positive breast cancer. J Clin Oncol. Cancer Res. 2005;65:1105–11.
2006;24:3039–47. 177. Zárate R, González-Santigo S, de la Haba J, Bandres
163. Sotiriou C, Wirapati P, Loi S, Harris A, Fox S, E, Morales R, Salgado J, et al. GSTP1 and MTHFR
Smeds J, et al. Gene expression profiling in breast polymorphisms are related with toxicity in breast
cancer: understanding the molecular basis of histo- cancer adjuvant anthracycline-based treatment. Curr
logic grade to improve prognosis. J Natl Cancer Inst. Drug Metab. 2007;8:481–6.
2006;98:262–72. 178. Schroth W, Goetz MP, Hamann U, Fasching PA,
164. Loi S, Haibe-Kains B, Desmedt C, Lallemand F, Tutt Schmidt M, Winter S, et al. Association between
AM, Gillet C, et al. Definition of clinically distinct CYP2D6 polymorphisms and outcomes among
molecular subtypes in estrogen receptor-positive women with early stage breast cancer treated with
breast carcinomas through genomic grade. J Clin tamoxifen. JAMA. 2009;302:1429–36.
Oncol. 2007;25:1239–46. 179. Goetz MP, Rae JM, Suman VJ, Safgren SL, Ames
165. Paik S, Shak S, Tang G, Kim C, Baker J, Cronin M, MM, Visscher DW, et al. Pharmacogenetics of
et al. A multigene assay to predict recurrence of tamoxifen biotransformation is associated with clini-
tamoxifen-treated, node-negative breast cancer. N cal outcomes of efficacy and hot flashes. J Clin
Engl J Med. 2004;351:2817–26. Oncol. 2005;23:9312–8.
166. Paik S, Tang G, Shak S, Kim C, Baker J, Kim W, et al. 180. Ma CX, Adjei AA, Salavaggione OE, Coronel J,
Gene expression and benefit of chemotherapy in Pelleymounter L, Wang L, et al. Human aromatase:
women with node-negative, estrogen receptor-positive gene resequencing and functional genomics. Cancer
breast cancer. J Clin Oncol. 2006;24:3726–34. Res. 2005;65:11071–82.
167. Wang Y, Klijn JG, Zhang Y, Sieuwerts AM, Look 181. Chang-Claude J, Ambrosone CB, Lilla C, Kropp S,
MP, Yang F, et al. Gene expression profiles to predict Helmbold I, von Fournier D, et al. Genetic polymor-
distant metastasis of lymph-node-negative primary phisms in DNA repair and damage response genes
breast cancer. Lancet. 2005;365:671–9. and late normal tissue complications of radiotherapy
168. Cardoso F, Piccart-Gebhart M, Van’t Veer L, Rutgers for breast cancer. Br J Cancer. 2009;100:1680–6.
E, TRANSBIG Consortium. The MINDACT trial: 182. Lee-Hoeflich ST, Crocker L, Yao E, Pham T, Munroe
the first prospective clinical validation of a genomic X, Hoeflich KP, et al. A central role for HER3 in
tool. Mol Oncol. 2007;1:246–51. HER2-amplified breast cancer: implications for tar-
169. Sparano JA, Paik S. Development of the 21-gene geted therapy. Cancer Res. 2008;68:5878–87.
assay and its application in clinical practice and clin- 183. Pusztai L, Stec J, Ayers M, Ross JS, Wagner P,
ical trials. J Clin Oncol. 2008;26:721–8. Rouzier R, et al. Pharmacogenetics, pharmacoge-
170. Desmedt C, Ruíz-García E, André F. Gene expres- nomics, and predicting response to therapy. In: Ross
sion predictors in breast cancer: current status, limi- JS, Hortobagyi GN, editors. Molecular oncology of
tations and perspectives. Eur J Cancer. 2008;44: breast cancer. Sudbury: Jones and Bartlett Publishers;
2714–20. 2005. p. 439–56.
171. Fan C, Oh DS, Wessels L, Weigelt B, Nuyten DSA, 184. Jordan VC. Tamoxifen: a most unlikely pioneering
Nobel AB, et al. Concordance among gene- medicine. Nat Rev Drug Discov. 2003;2:205–13.
expression-based predictors for breast cancer. N 185. Fisher B, Costantino JP, Wickerham DL, Redmond
Engl J Med. 2006;355:560–9. CK, Kavanah M, Cronin WM, et al, National
172. Ross JS, Hatzis C, Fraser Symmans W, Pusztai L, Surgical Adjuvant Breast and Bowel Project
Hortobágyi GN. Commercialized multigene predic- Investigators. Tamoxifen for prevention of breast
tors of clinical outcome for breast cancer. Oncologist. cancer: report of the National Surgical Adjuvant
2008;13:477–93. Breast and Bowel Project P-1 Study. J Natl Cancer
173. Ross JS. Multigene predictors in early-stage breast Inst. 1998;90:1371–88.
cancer: moving in or moving out? Expert Rev Mol 186. Bradford LD. CYP2D6 allele frequency in European
Diagn. 2008;8:129–35. Caucasians, Asians, Africans and their descendants.
174. Fan L, Goh B-C, Wong C-I, Sukri N, Lim S-E, Tan Pharmacogenomics. 2002;3:229–43.
S-H, et al. Genotype of human carbonyl reductase 187. Tan S-H, Lee S-C, Goh B-C, Wong J.
CBR3 correlates with doxorubicin disposition and Pharmacogenetics in breast cancer therapy. Clin
toxicity. Pharmacogenet Genomics. 2008;18:621–9. Cancer Res. 2008;14:8027–41.
175. Harris L, Fritsche H, Mennel R, Norton L, Ravdin P, 188. Higgins MJ, Rae JM, Flockhart DA, Hayes DF,
Taube S, et al. American Society of Clinical Stearns V. Pharmacogenetics of tamoxifen: who
Oncology 2007 update of recommendations for the should undergo CYP2D6 genetic testing? J Natl
use of tumor markers in breast cancer. J Clin Oncol. Compr Canc Netw. 2009;7:203–13.
2007;25:5287–312. 189. Dowsett M, Haynes BP. Hormonal effects of aroma-
176. Ambrosone CB, Ahn J, Singh KK, Rezaishiraz H, tase inhibitors: focus on premenopausal effects and
Furberg H, Sweeney C, et al. Polymorphisms in interaction with tamoxifen. J Steroid Biochem Mol
genes related to oxidative stress (MPO, MnSOD, Biol. 2003;86:255–63.
4 Breast Cancer Genomics 97

190. Breast International Group (BIG) 1-98 Collaborative 204. Hoskins JM, Carey LA, McLeod HL. CYP2D6 and
Group, Thürlimann B, Keshaviah A, Coates AS, tamoxifen: DNA matters in breast cancer. Nat Rev
Mouridsen H, Mauriac L, Forbes JF, et al. A compari- Cancer. 2009;9:576–86.
son of letrozole and tamoxifen in postmenopausal 205. Check HE. Personalized cancer therapy gets closer.
women with early breast cancer. N Engl J Med. Nature. 2009;458:131–2.
2005;353:2747–57. 206. Chin K, de Solorzano CO, Knowles D, Jones A,
191. Boccardo F, Rubagotti A, Puntoni M, Guglielmini P, Chou W, Rodriguez EG, et al. In situ analyses of
Amoroso D, Fini A, et al. Switching to anastrozole genome instability in breast cancer. Nat Genet.
versus continued tamoxifen treatment of early breast 2004;36:984–8.
cancer: preliminary results of the Italian Tamoxifen 207. Novak P, Jensen TJ, Garbe JC, Stampfer MR,
Anastrozole Trial. J Clin Oncol. 2005;23:5138–47. Futscher BW. Stepwise DNA methylation changes
192. Goss PE, Ingle JN, Martino S, Robert NJ, Muss HB, are linked to escape from defined proliferation barri-
Piccart MJ, et al. Randomized trial of letrozole fol- ers and mammary epithelial cell immortalization.
lowing tamoxifen as extended adjuvant therapy in Cancer Res. 2009;69:5251–8.
receptor-positive breast cancer: updated findings 208. Berman H, Zhang J, Crawford YG, Gauthier ML,
from NCIC CTG MA.17. J Natl Cancer Inst. Fordyce CA, McDermott KM, et al. Genetic and epi-
2005;97:1262–71. genetic changes in mammary epithelial cells identify
193. Nahta R, Esteva FJ. Herceptin: mechanisms of action a subpopulation of cells involved in early carcino-
and resistance. Cancer Lett. 2006;232:123–38. genesis. Cold Spring Harb Symp Quant Biol.
194. Harris M. Monoclonal antibodies as therapeutic 2005;70:317–27.
agents for cancer. Lancet Oncol. 2004;5:292–302. 209. McDermott KM, Zhang J, Holst CR, Kozakiewicz
195. Holbro T, Beerli RR, Maurer F, Koziczak M, Barbas BK, Singla V, Tisty TD. p16 (INK4a) prevents cen-
3rd CF, Hynes NE. The ErbB2/ErbB3 heterodimer trosome dysfunction and genomic instability in pri-
functions as an oncogenic unit: ErbB2 requires mary cells. PLoS Biol. 2006;4:e51.
ErbB3 to drive breast tumor cell proliferation. Proc 210. Artandi SE, Depinho RA. Telomeres and telomer-
Natl Acad Sci U S A. 2003;100:8933–8. ase in cancer. Carcinogenesis 2010;31(1):9–18.
196. Agus DB, Akita RW, Fox WD, Lewis GD, Higgins 211. Korkola J, Gray JW. Breast cancer genomes—form
B, Pisacane PI, et al. Targeting ligand-activated and function. Curr Opin Genet Dev. 2010;20(1):
ErbB2 signaling inhibits breast and prostate tumor 4–14. doi:10.1016/j.gde.2009.11.005.
growth. Cancer Cell. 2002;2:127–37. 212. Kuukasjarvi T, Karhu R, Tanner M, Kähkönen M,
197. Moulder S, Hortobagyi GN. Advances in the treat- Schäffer A, Nupponen N, et al. Genetic heterogene-
ment of breast cancer. Clin Pharmacol Ther. 2008; ity and clonal evolution underlying development of
83:26–36. asynchronous metastasis in human breast cancer.
198. Troester MA, Hoadley KA, Sørlie T, Herbert B-S, Cancer Res. 1997;57:1597–604.
Børresen-Dale A-L, Lønning PE, et al. Cell-type- 213. Navin N, Krasnitz A, Rodgers L, Cook K, Meth J,
specific responses to chemotherapeutics in breast Kendall J, et al. Inferring tumor progression from
cancer. Cancer Res. 2004;64:4218–26. genomic heterogeneity. Genome Res. 2010;20(1):68–
199. Rouzier R, Perou CM, Symmans WF, Ibrahim N, 80. Epub 2009 Nov 10.
Cristofanilli M, Anderson K, et al. Breast cancer 214. Subramanian A, Tamayo P, Mootha VK, Mukherjee
molecular subtypes respond differently to preopera- S, Ebert BL, Gillette MA, et al. Gene set enrichment
tive chemotherapy. Clin Cancer Res. 2005;11: analysis: a knowledge-based approach for interpret-
5678–85. ing genome-wide expression profiles. Proc Natl
200. Li L-F, Xu X-J, Zhao Y, Liu Z-B, Shen Z-Z, Jin Acad Sci U S A. 2005;102:15545–50.
W-R, et al. Integrated gene expression profile pre- 215. Reyal F, van Vliet MH, Armstrong NJ, Horlings
dicts prognosis of breast cancer patients. Breast HM, de Visser KE, Kok M, et al. A comprehensive
Cancer Res Treat. 2009;113:231–7. analysis of prognostic signatures reveals the high
201. Chang JC, Wooten EC, Tsimelzon A, Hilsenbeck predictive capacity of the proliferation, immune
SG, Gutierrez MC, Elledge R, et al. Gene expression response and RNA splicing modules in breast can-
profiling for the prediction of therapeutic response to cer. Breast Cancer Res. 2008;10:R93.
docetaxel in patients with breast cancer. Lancet. 216. Bianchini G, Qi Y, Alvarez RH, Iwamoto T, Coutant
2003;362:362–9. C, Ibrahim NK, et al. Molecular anatomy of breast
202. Bosch TM, Meijerman I, Beijnen JH, Schellens cancer stroma and its prognostic value in estrogen
JHM. Genetic polymorphisms of drug-metabolising receptor-positive and -negative cancers. J Clin Oncol.
enzymes and drug transporters in the chemothera- 2010;28:4316–23.
peutic treatment of cancer. Clin Pharmacokinet. 217. Symmans WF, Hatzis C, Sotiriou C, Andre F,
2006;45:253–85. Peintinger F, Regitnig P, et al. Genomic index of
203. Marsh S, Liu G. Pharmacokinetics and pharmacoge- sensitivity to endocrine therapy for breast cancer. J
nomics in breast cancer chemotherapy. Ad Drug Clin Oncol. 2010;28:4111–9.
Deliv Rev. 2009;61:381–7.
98 B. Kumar

218. Goldhirsch A, Ingle JN, Gelber RD, Coates AS, 229. Hussein O, Komarova SV. Breast cancer at bone
Thurlimann B, Senn HJ. Thresholds for therapies: metastatic sites: recent discoveries and treatment tar-
highlights of the St Gallen International Expert gets. J Cell Commun Signal. 2001;5(2):85–99.
Consensus on the primary therapy of early breast 230. Townson JL, Chambers AF. Dormancy of solitary
cancer 2009. Ann Oncol. 2009;20:1319–29. metastatic cells. Cell Cycle. 2006;5(16):1744–50.
219. Bueno-de-Mesquita JM, van Harten WH, Retel VP, 231. White DE, Rayment JH, Muller WJ. Addressing the
van’t Veer LJ, van Dam FS, Karsenberg K, et al. Use role of cell adhesion in tumor cell dormancy. Cell
of 70-gene signature to predict prognosis of patients Cycle. 2006;5(16):1756–9.
with node-negative breast cancer: a prospective 232. Ranganathan AC, Adam AP, Aguirre-Ghiso JA.
community-based feasibility study (RASTER). Opposing roles of mitogenic and stress signaling
Lancet Oncol. 2007;8:1079–87. pathways in the induction of cancer dormancy. Cell
220. Goldstein LJ, Gray R, Badve S, Childs BH, Cycle. 2006;5(16):1799–807.
Yoshizawa C, Rowley S, et al. Prognostic utility of 233. Aguirre-Ghiso JA, Kovalski K, Ossowski L. Tumor
the 21-gene assay in hormone receptor-positive dormancy induced by down regulation of urokinase
operable breast cancer compared with classical clini- receptor in human carcinoma involves integrin and
copathologic features. J Clin Oncol. 2008;26: MAPK signaling. J Cell Biol. 1999;47(1):89–104.
4063–71. 234. Naumov GN, Macdonald IC, Weinmeister PM,
221. Cuzick J, Dowsett M, Wale C, Salter J, Quinn E, Kerkvliet N, Nadkarni KV, Wilson SM, et al.
Zabaglo L, et al. Prognostic value of a combined ER, Persistence of solitary mammary carcinoma cells in
PgR, Ki67, HER2 immunohistochemical (IHC4) a secondary site: a possible contributor to dormancy.
score and comparison with the GHI recurrence Cancer Res. 2002;62(7):2162–8.
score—results from TransATAC. Cancer Res. 235. Fisher JL, Thomas-Mudge RJ, Elliott J, Hards DK,
2009;69:74. Sims NA, Slavin J, et al. Osteoprotegerin overex-
222. Viale G, Regan MM, Dell’Orto P, Mastropasqua pression by breast cancer cells enhances orthotopic
MG, Rasmussen BB, MacGrogan G, et al. Central and osseous tumor growth and contrasts with that
review of ER, PGR and HER2 in BIG 1–98 evaluat- delivered therapeutically. Cancer Res. 2006;66(7):
ing letrozole vs. letrozole followed by tamoxifen vs. 3620–8.
tamoxifen followed by letrozole as adjuvant endo- 236. Holen I, Cross SS, Neville-Webbe HL, Cross NA,
crine therapy for postmenopausal women with hor- Balasubramanian SP, Croucher PI, et al.
mone receptor-positive breast cancer. Cancer Res. Osteoprotegerin (OPG) expression by breast cancer
2009;69:76. cells in vitro and breast tumours in vivo—a role in
223. Shi L, Campbell G, Jones WD, Campagne F, tumour cell survival? Breast Cancer Res Treat.
Wen Z, Walker SJ, et al. The MicroArray Quality 2005;92(3):207–15.
Control (MAQC)-II study of common practices 237. Neville-Webbe HL, Cross NA, Eaton CL, Nyambo
for the development and validation of microarray- R, Evans CA, Coleman RE, et al. Osteoprotegerin
based predictive models. Nat Biotechnol. 2010;28: (OPG) produced by bone marrow stromal cells pro-
827–38. tects breast cancer cells from TRAIL-induced apop-
224. Popovici V, Chen W, Gallas BG, Hatzis C, Shi W, tosis. Breast Cancer Res Treat. 2004;86(3):269–79.
Samuelson FW, et al. Effect of training-sample size 238. Zhang XH, Wang Q, Gerald W, Hudis CA, Norton L,
and classification difficulty on the accuracy of Smid M, Foekens JA, et al. Latent bone metastasis in
genomic predictors. Breast Cancer Res. 2010;12:R5. breast cancer tied to Src-dependent survival signals.
225. Juul N, Szallasi Z, Eklund AC, Li Q, Burrell RA, Cancer Cell. 2009;16(1):67–78.
Gerlinger M, et al. Assessment of an RNA interfer- 239. Indraccolo S, Stievano L, Minuzzo S, Tosello V,
ence screen-derived mitotic and ceramide pathway Esposito G, Piovan E, et al. Interruption of tumor
metagene as a predictor of response to neoadjuvant dormancy by a transient angiogenic burst within the
paclitaxel for primary triple-negative breast cancer: tumor microenvironment. Proc Natl Acad Sci U S A.
a retrospective analysis of five clinical trials. Lancet 2006;103(11):4216–21.
Oncol. 2010;11:358–65. 240. Naumov GN, Bender E, Zurakowski D, Kang SY,
226. Rody A, Holtrich U, Pusztai L, Liedtke C, Gaetje R, Sampson D, Flynn E, et al. A model of human tumor
Ruckhaeberle E, et al. T-cell metagene predicts a dormancy: an angiogenic switch from the nonangio-
favorable prognosis in estrogen receptor-negative genic phenotype. J Natl Cancer Inst. 2006;98(5):
and HER2-positive breast cancers. Breast Cancer 316–25.
Res. 2009;11:R15. 241. Klein CA, Blankenstein TJ, Schmidt-Kittler O,
227. Fehm T, Mueller V, Marches R, Klein G, Gueckel B, Prtronio M, Polzer B, Stoecklein NH, et al. Genetic
Neubauer H, et al. Tumor cell dormancy: implica- heterogeneity of single disseminated tumour cells in
tions for the biology and treatment of breast cancer. minimal residual cancer. Lancet. 2002;360(9334):
APMIS. 2008;116(7–8):742–53. 683–9.
228. Aguirre-Ghiso JA. Models, mechanisms and clinical 242. Barkan D, Kleinman H, Simmons JL, Asmussen H,
evidence for cancer dormancy. Nat Rev Cancer. Kamaraju AK, Hoenorhoff MJ, et al. Inhibition of
2007;11:834–46. metastatic outgrowth from single dormant tumor
4 Breast Cancer Genomics 99

cells by targeting the cytoskeleton. Cancer Res. 257. Klarmann GJ, Decker A, Farrar WL. Epigenetic
2008;68(15):6241–50. gene silencing in the Wnt pathway in breast cancer.
243. Ramaswamy S, Ross KN, Lander ES, Golub TR. A Epigenetics. 2008;3:59–63.
molecular signature of metastasis in primary solid 258. Chan TA, Glockner S, Yi JM, Chen W, Van Neste L,
tumors. Nat Genet. 2003;33:49–54. Cope L, et al. Convergence of mutation and epigen-
244. Fridlyand J, Snijders AM, Ylstra B, Li H, Olshen A, etic alterations identifies common genes in cancer
Segraves R, et al. Breast tumor copy number aberra- that predict for poor prognosis. PLoS Med. 2008;
tion phenotypes and genomic instability. BMC 5:e114.
Cancer. 2006;6:96. 259. Fu M, Wang C, Zhang X, Pestell RG. Acetylation of
245. Wessels LF, van Welsem T, Hart AA, van’t Veer LJ, nuclear receptors in cellular growth and apoptosis.
Reinders MJ, Nederlof PM. Molecular classification Biochem Pharmacol. 2004;68:1199–208.
of breast carcinomas by comparative genomic 260. Han HJ, Russo J, Kohwi Y, Kohwi-Shigematsu T.
hybridization: a specific somatic genetic profile for SATB1 reprogrammes gene expression to promote
BRCA1 tumors. Cancer Res. 2002;62:7110–7. breast tumour growth and metastasis. Nature. 2008;
246. Yang ZQ, Streicher KL, Ray ME, Abrams J, Ethier SP. 452:187–93.
Multiple interacting oncogenes on the 8p11–p12 261. Fraga MF, Esteller M. DNA methylation: a profile of
amplicon in human breast cancer. Cancer Res. methods and applications. Biotechniques. 2002;33:
2006;66:11632–43. 632, 634, 636–49.
247. Guan Y, Kuo WL, Stilwell JL, Takano H, Lapuk AV, 262. Bird AP, Southern EM. Use of restriction enzymes to
Fridlyand J, et al. Amplification of PVT1 contributes study eukaryotic DNA methylation: I The methyla-
to the pathophysiology of ovarian and breast cancer. tion pattern in ribosomal DNA from Xenopus laevis.
Clin Cancer Res. 2007;13:5745–55. J Mol Biol. 1978;118:27–47.
248. Brown LA, Johnson K, Leung S, Bismar TA, 263. Frommer M, McDonald LE, Millar DS, Collis CM,
Benitez J, Foulkes WD, et al. Co-amplification of Watt F, Grigg GW, et al. A genomic sequencing pro-
CCND1 and EMSY is associated with an adverse tocol that yields a positive display of 5-methylcytosine
outcome in ER-positive tamoxifen-treated breast residues in individual DNA strands. Proc Natl Acad
cancers. Breast Cancer Res Treat. 2010;121(2): Sci U S A. 1992;89:1827–31.
347–54. Epub 2009 Jul 28. 264. Herman JG, Graff JR, Myohanen S, Nelkin BD,
249. Hodgson JG, Malek T, Bornstein S, Hariono S, Baylin SB. Methylation-specific PCR: a novel PCR
Ginzinger DG, Muller WJ, et al. Copy number aber- assay for methylation status of CpG islands. Proc
rations in mouse breast tumors reveal loci and genes Natl Acad Sci U S A. 1996;93:9821–6.
important in tumorigenic receptor tyrosine kinase 265. Xiong Z, Laird PW. COBRA: a sensitive and quanti-
signaling. Cancer Res. 2005;65:9695–704. tative DNA methylation assay. Nucleic Acids Res.
250. Kwek SS, Roy R, Zhou H, Climent J, Martinez- 1997;25:2532–4.
Climent JA, Fridlyand J, et al. Co-amplified genes at 266. Eads CA, Danenberg KD, Kawakami K, Saltz LB,
8p12 and 11q13 in breast tumors cooperate with two Blake C, Shibata D, et al. MethyLight: a high-
major pathways in oncogenesis. Oncogene. 2009;28: throughput assay to measure DNA methylation.
1892–903. Nucleic Acids Res. 2000;28:E32.
251. Park K, Kwak K, Kim J, Lim S, Han S. c-myc ampli- 267. Fackler MJ, McVeigh M, Mehrotra J, Blum MA,
fication is associated with HER2 amplification and Lange J, Lapides A, et al. Quantitative multiplex
closely linked with cell proliferation in tissue micro- methylation-specific PCR assay for the detection of
array of nonselected breast cancers. Hum Pathol. promoter hypermethylation in multiple genes in
2005;36:634–9. breast cancer. Cancer Res. 2004;64:4442–52.
252. Leary RJ, Lin JC, Cummins J, Boca S, Wood LD, 268. Swift-Scanlan T, Blackford A, Argani P, Sukumar S,
Parsons DW, et al. Integrated analysis of homozy- Fackler MJ. Two-color quantitative multiplex
gous deletions, focal amplifications, and sequence methylation-specific PCR. Biotechniques. 2006;40:
alterations in breast and colorectal cancers. Proc 210–9.
Natl Acad Sci U S A. 2008;105:16224–9. 269. Tost J, Gut IG. DNA methylation analysis by pyrose-
253. Das R, Hampton DD, Jirtle RL. Imprinting evolution quencing. Nat Protoc. 2007;2:2265–75.
and human health. Mamm Genome. 2009;20(9–10): 270. Costello JF, Fruhwald MC, Smiraglia DJ, Rush LJ,
563–72. Robertson GP, Gao X, et al. Aberrant CpG-island
254. Lo PK, Sukumar S. Epigenomics and breast cancer. methylation has non-random and tumour-type-
Pharmacogenomics. 2008;9:1879–902. specific patterns. Nat Genet. 2000;24:132–8.
255. Kenemans P, Verstraeten RA, Verheijen RH. 271. Frigola J, Ribas M, Risques RA, Peinado MA.
Oncogenic pathways in hereditary and sporadic Methylome profiling of cancer cells by amplification
breast cancer. Maturitas. 2008;61:141–50. of inter-methylated sites (AIMS). Nucleic Acids
256. Zhu W, Qin W, Hewett JE, Sauter ER. Quantitative Res. 2002;30:E28.
evaluation of DNA hypermethylation in malignant 272. Huang TH, Perry MR, Laux DE. Methylation
and benign breast tissue and fluids. Int J Cancer. profiling of CpG islands in human breast cancer
2010;126(2):474–82. cells. Hum Mol Genet. 1999;8:459–70.
100 B. Kumar

273. Khulan B, Thompson RF, Ye K, Fazzari MJ, Suzuki 288. Vander Groep P, Bouter A, vander Zanden R, Menko
M, Stasiek E, et al. Comparative isoschizomer FH, Buerger H, Verheijen RH, et al. Re: Germline
profiling of cytosine methylation: the HELP assay. BRCA1 mutations and a basal epithelial phenotype
Genome Res. 2006;16:1046–55. in breast cancer. J Natl Cancer Inst. 2004;96:712–3.
274. Keshet I, Schlesinger Y, Farkash S, Rand E, Hecht 289. Carey LA, Perou CM, Livasy CA, Dressler LG,
M, Segal E, et al. Evidence for an instructive mecha- Cowan D, Karaca G, et al. Race, breast cancer
nism of de novo methylation in cancer cells. Nat subtypes, and survival in the Carolina Breast Cancer
Genet. 2006;38:149–53. Study. JAMA. 2006;295:2492–502.
275. Weber M, Davies JJ, Wittig D, Oakeley EJ, Haase 290. Nielsen TO, Hsu FD, Jensen K, Cheang M, Karaca G,
M, Lam WL, et al. Chromosome-wide and promoter- Hu Z, et al. Immunohistochemical and clinical charac-
specific analyses identify sites of differential DNA terization of the basal-like subtype of invasive breast
methylation in normal and transformed human cells. carcinoma. Clin Cancer Res. 2004;10:5367–74.
Nat Genet. 2005;37:853–62. 291. Kurebayashi J, Moriya T, Ishida T, Hirakawa H,
276. Weber M, Hellmann I, Stadler MB, Ramos L, Pääbo S, Kurosumi M, Akiyama F, et al. The prevalence of
Rebhan M, et al. Distribution, silencing potential and intrinsic subtypes and prognosis in breast cancer
evolutionary impact of promoter DNA methylation in patients of different races. Breast. 2007;16 Suppl
the human genome. Nat Genet. 2007;39:457–66. 2:S72–7.
277. Suzuki H, Gabrielson E, Chen W, Anbazhagan R, 292. Millikan RC, Newman B, Tse CK, Moorman PG,
van Engeland M, Weijenberg MP, et al. A genomic Conway K, Dressler LG, et al. Epidemiology of
screen for genes upregulated by demethylation and basal-like breast cancer. Breast Cancer Res Treat.
histone deacetylase inhibition in human colorectal 2008;109:123–39.
cancer. Nat Genet. 2002;31:141–9. 293. Yang XR, Sherman ME, Rimm DL, Lissowska J,
278. Yamashita K, Upadhyay S, Osada M, Hoque MO, Brinton LA, Peplonska B, et al. Differences in risk fac-
Xiao Y, Mori M, et al. Pharmacologic unmasking of tors for breast cancer molecular subtypes in a popula-
epigenetically silenced tumor suppressor genes in tion-based study. Cancer Epidemiol Biomarkers Prev.
esophageal squamous cell carcinoma. Cancer Cell. 2007;16:439–43.
2002;2:485–95. 294. Zhang C, Ikpatt F, Khramtsov A, et al. Molecular
279. Hoque MO, Kim MS, Ostrow KL, Liu J, Wisman GB, classification of West African breast tumors demon-
Park HL, et al. Genome-wide promoter analysis strates an overrepresentation of hormone receptor
uncovers portions of the cancer methylome. Cancer negative breast cancer. In: 98th AACR annual meet-
Res. 2008;68:2661–70. ing. Los Angeles: AACR; 2007.
280. Hu M, Yao J, Polyak K. Methylation-specific digital 295. Spitz MR, Wu X, Mills G. Integrative epidemiology:
karyotyping. Nat Protoc. 2006;1:1621–36. from risk assessment to outcome prediction. J Clin
281. Bernstein BE, Kamal M, Lindblad-Toh K, Bekiranov Oncol. 2005;23:267–75.
S, Bailey DK, Huebert DJ, et al. Genomic maps and 296. Cameron DA, Stein S. Drug insight: intracellular
comparative analysis of histone modifications in inhibitors of HER2 -clinical development of lapa-
human and mouse. Cell. 2005;120:169–81. tinib in breast cancer. Nat Clin Pract Oncol. 2008;
282. Roh TY, Ngau WC, Cui K, Landsman D, Zhao K. 5:512–20.
High-resolution genome-wide mapping of histone 297. Geyer CE, Forster J, Lindquist D, Chan S, Romieu
modifications. Nat Biotechnol. 2004;22:1013–6. CG, Pienkowski T, et al. Lapatinib plus capecitabine
283. Barski A, Cuddapah S, Cui K, Roh TY, Schones DE, for HER2-positive advanced breast cancer. N Engl J
Wang Z, et al. High-resolution profiling of histone Med. 2006;355:2733–43.
methylations in the human genome. Cell. 2007;129: 298. Fong PC, Boss DS, Yap TA, Tutt A, Wu P, Mergui-
823–37. Roelvink M, et al. Inhibition of poly (ADP-ribose)
284. Mikkelsen TS, Ku M, Jaffe DB, Issac B, Lieberman E, polymerase in tumors from BRCA mutation carriers.
Giannoukos G, et al. Genome-wide maps of chroma- N Engl J Med. 2009;361:123–34.
tin state in pluripotent and lineage-committed cells. 299. Slamon DJ, Clark GM, Wong SG, Levin WJ, Ulrich
Nature. 2007;448:553–60. A, McGuire WL. Human breast cancer: correlation
285. Toyota M, Ho C, Ahuja N, Jair KW, Li Q, Ohe- of relapse and survival with amplification of the
Toyota M, et al. Identification of differentially meth- HER2/neu oncogene. Science. 1987;235:177–82.
ylated sequences in colorectal cancer by methylated 300. Al-Kuraya K, Novotny H, Bavi P, Siraj AK, Uddin
CpG island amplification. Cancer Res. 1999;59: S, Ezzat A. HER2, TOP2A, CCND1, EGFR and
2307–12. C-MYC oncogene amplification in colorectal can-
286. Esteller M. Cancer epigenomics: DNA methylomes cer. J Clin Pathol. 2007;60:768–72.
and histone-modification maps. Nat Rev Genet. 301. Fantl V, Smith R, Brookes S, Dickson C, Peters
2007;8:286–98. G. Chromosome 11q13 abnormalities in human
287. Schneider BP, Winer EP, Foulkes WD, Garber J. breast cancer. Cancer Surv. 1993;18:77–94.
Triple negative breast cancer: risk factors to poten- 302. Escot C, Theillet C, Lidereau R, Spyratos F,
tial targets. Clin Cancer Res. 2008;14:8010–8. Champeme MH, Gest J, et al. Genetic alteration of
4 Breast Cancer Genomics 101

the c‐myc Protooncogene (MYC) in human primary 317. Kallioniemi OP, Kallioniemi A, Kurisu W, Thor A,
breast carcinomas. Proc Natl Acad Sci U S A. Chen LC, Smith HS, et al. ERBB2 amplification in
1986;83:4834–8. breast cancer analyzed by fluorescence in situ hybrid-
303. Collins C, Rommens JM, Kowbel D, Godfrey T, ization. Proc Natl Acad Sci U S A. 1992;89:5321–5.
Tanner M, Hwag SI, et al. Positional cloning of 318. Heise C, Sampson-Johannes A, Williams A,
ZNF217 and NABC1: genes amplified at 20q13.2 McCormick F, Von Hoff DD, Kim DH. ONYX-015,
and overexpressed in breast carcinoma. Proc Natl an E1B gene-attenuated adenovirus, causes tumor-
Acad Sci U S A. 1998;95:8703–8. specific cytolysis and antitumoral efficacy that can
304. Cheng KW, Lahad JP, Kuo WL, Lapuk A, Yamada K, be augmented by standard chemotherapeutic agents.
Auersperg N, et al. The RAB25 small GTPase deter- Nat Med. 1997;3:639–45.
mines aggressiveness of ovarian and breast cancers. 319. Wang H, Nan L, Yu D, Agrawal S, Zhang R.Antisense
Nat Med. 2004;10:1251–6. anti‐MDM2 oligonucleotides as a novel therapeutic
305. Courjal F, Cuny M, Rodriguez C, Louason G, approach to human breast cancer: in vitro and in vivo
Speiser P, Katsaros D, et al. DNA amplifications at activities and mechanisms. Clin Cancer Res. 2001;7:
20q13 and MDM2 define distinct subsets of evolved 3613–24.
breast and ovarian tumours. Br J Cancer. 1996; 320. Burgess DJ, Doles J, Zender L, Xue W, Ma B,
74:1984–9. McCombie WR, et al. Topoisomerase levels deter-
306. Sinclair CS, Adem C, Naderi A, Soderberg CL, mine chemotherapy response in vitro and in vivo.
Johnson M, Wu K, et al. TBX2 is preferentially Proc Natl Acad Sci U S A. 2008;105:9053–8.
amplified in BRCA1- and BRCA2-related breast 321. Hennessy BT, Smith DL, Ram PT, Lu Y, Mills
tumors. Cancer Res. 2002;62:3587–91. GB. Exploiting the PI3K/AKT pathway for cancer
307. Haverty PM, Fridlyand J, Li L, Getz G, Beroukhim drug discovery. Nat Rev Drug Discov. 2005;4:
R, Lohr S, et al. High‐resolution genomic and expres- 988–1004.
sion analyses of copy number alterations in breast 322. Rottenberg S, Jaspers JE, Kersbergen A, van der
tumors. Genes Chromosomes Cancer. 2008;47: Burg E, Nygren AO, Zander SA, et al. High sensitiv-
530–42. ity of BRCA1-deficient mammary tumors to the
308. Tennis M, Krishnan S, Bonner M, Ambrosone CB, PARP inhibitor AZD2281 alone and in combination
Vena JE, Moysich K, et al. p53 Mutation analysis in with platinum drugs. Proc Natl Acad Sci U S A.
breast tumors by a DNA microarray method. Cancer 2008;105:17079–84.
Epidemiol Biomarkers Prev. 2006;15:80–5. 323. Kumar B, Yadav PR, Singh S. Use of protein bio-
309. Samuels Y, Wang Z, Bardelli A, Silliman N, Ptak J, markers for early detection of breast cancer. In: Barh
Szabo S, et al. High frequency of mutations of the D, Zambare V, Azevedo V, editors. OMICS applica-
PIK3CA gene in human cancers. Science. 2004; tion in biomedical, agricultural, and environmental
304:554. science. Boca Raton: CRC Press Taylor & Francis
310. Li J, Yen C, Liaw D, Podsypanina K, Bose S, Group; 2013. p. 299–313.
Wang SI, et al. PTEN, a putative protein tyrosine 324. Dowsett M, Dunbier AK. Emerging biomarkers and
phosphatase gene mutated in human brain, breast, new understanding of traditional markers in person-
and prostate cancer. Science. 1997;275:1943–7. alized therapy for breast cancer. Clin Cancer Res.
311. Futreal PA, Liu Q, Shattuck-Eidens D, Cochran C, 2008;14:8019–26.
Harshman K, Tavtigian S, et al. BRCA1 mutations in 325. Pusztai L, Cristofanilli M, Paik S. New generation of
primary breast and ovarian carcinomas. Science. molecular prognostic and predictive tests for breast
1994;266:120–2. cancer. Semin Oncol. 2007;34:S10–6.
312. Wooster R, Bignell G, Lancaster J, Swift S, Seal S, 326. Sotiriou C, Piccart MJ. Taking gene-expression pro-
Mangion J, et al. Identification of the breast cancer filing to the clinic: when will molecular signatures
susceptibility gene BRCA2. Nature. 1995;378: become relevant to patient care? Nat Rev Cancer.
789–92. 2007;7:545–53.
313. Ross JS, Fletcher JA. The HER2/neu oncogene in 327. Hess KR, Anderson K, Symmans WF, Valero V,
breast cancer: prognostic Factor, predictive factor, Ibrahim N, Mejia JA, et al. Pharmacogenomic pre-
and target for therapy. Oncologist. 1998;3:237–52. dictor of sensitivity to preoperative chemotherapy
314. Slamon D, Pegram M. Rationale for trastuzumab with paclitaxel and fluorouracil, doxorubicin, and
(Herceptin) in adjuvant breast cancer trials. Semin cyclophosphamide in breast cancer. J Clin Oncol.
Oncol. 2001;28:13–9. 2006;24:4236–44.
315. Vogel CL, Cobleigh MA, Tripathy D, Gutheil JC, 328. Hudis CA. Trastuzumab-mechanism of action and
Harris LN, Fehrenbacher L, et al. First-line Herceptin use in clinical practice. N Engl J Med. 2007;357:
monotherapy in metastatic breast cancer. Oncology. 39–51.
2001;61:37–42. 329. Ebomoyi EW. Genomic epidemiology of BRCA1/
316. Moy B, Goss PE. Lapatinib: current status and future BRCA2; breast cancer associated genes and use of
directions in breast cancer. Oncologist. 2006;11: electronic health record to reduce the escalating cost
1047–57. of treatment. Br J Med Med Res. 2011;1(4):333–45.
102 B. Kumar

330. Citron ML, Berry DA, Cirrincione C, Hudis C, and its impact on survival. Int J Radiat Oncol Biol
Winer EP, Gradishar WJ, et al. Randomized trial of Phys. 2006;65:1353–60.
dose-dense versus conventionally scheduled and 343. Anderson WF, Chatterjee N, Ershler WB, Brawley
sequential versus concurrent combination chemo- OW. Estrogen receptor breast cancer phenotypes in
therapy as postoperative adjuvant treatment of node- the surveillance, epidemiology, and end results data-
positive primary breast cancer: first report of base. Breast Cancer Res Treat. 2002;76:27–36.
Intergroup Trial C9741/Cancer and Leukemia Group 344. Elledge RM, Clark GM, Chamness GC, Osborne CK.
B Trial 9741. J Clin Oncol. 2003;21:1431–9. Tumor biologic factors and breast cancer prognosis
331. Zerhouni EA. Fiscal Year 2009 directors Budget among white, Hispanic, and black women in the
request statement. A new strategic vision for medi- United States. J Natl Cancer Inst. 1994;86:705–12.
cine. Available from: http://www.nih.gov/about/ 345. Griggs JJ, Sorbero ME, Stark AT, Heninger SE, Dick
director/budget/request/FY/2009ector’s/senate/bud- AW. Racial disparity in the dose and dose intensity
get/request. of breast cancer adjuvant chemotherapy. Breast
332. Ebomoyi E. Genomics application in public health Cancer Res Treat. 2003;81:21–31.
across all population, environment and work set- 346. Maly RC, Umezawa Y, Ratliff CT, Leake B. Racial/
tings. Continuing education institute; presented at ethnic group differences in treatment decision-
the American public Health association 133rd making and treatment received among older breast
annual Meeting and Exposition, Philadelphia. 2005; carcinoma patients. Cancer. 2006;106:957–65.
www.apha.Confex.com/apha/133am/techprogram/ 347. Clegg LX, Li FP, Hankey BF, Chu K, Edwards
sessions_16480html_6k. BK. Cancer survival among US whites and minori-
333. Mausner J, Bahn A. Epidemiology: an introductory ties: a SEER (Surveillance, Epidemiology, and End
text. Philadelphia: W. B Saunders; 1987. Results) Program population-based study. Arch
334. Huang RS, Duan S, Shukla SJ, Kistner EO, Clark Intern Med. 2002;162:1985–93.
TA, Chen TX, et al. Identification of genetic variants 348. Joslyn SA, West MM. Racial differences in breast
contributing to cisplatin-induced cytotoxicity by use carcinoma survival. Cancer. 2000;88:114–23.
of a genomewide approach. Am J Hum Genet. 349. Chu KC, Anderson WF, Fritz A, Ries LA, Brawley
2007;81:427–37. OW. Frequency distributions of breast cancer char-
335. Albain KS, Unger JM, Crowley JJ, Coltman Jr CA, acteristics classified by estrogen receptor and pro-
Hershman DL. Racial disparities in cancer survival gesterone receptor status for eight racial/ethnic
among randomized clinical trials patients of the groups. Cancer. 2001;92:37–45.
Southwest Oncology Group. J Natl Cancer Inst. 350. Chu KC, Lamar CA, Freeman HP. Racial disparities
2009;101:984–92. in breast carcinoma survival rates: separating factors
336. Bach PB, Schrag D, Brawley OW, Galaznik A, that affect diagnosis from factors that affect treat-
Yakren S, Begg CB. Survival of blacks and whites ment. Cancer. 2003;97:2853–60.
after a cancer diagnosis. JAMA. 2002;287:2106–13. 351. Gwyn K, Bondy ML, Cohen DS, Lund MJ, Liff JM,
337. Menashe I, Anderson WF, Jatoi I, Rosenberg Flagg EW, et al. Racial differences in diagnosis,
PS. Underlying causes of the black-white racial dis- treatment, and clinical delays in a population-based
parity in breast cancer mortality: a population-based study of patients with newly diagnosed breast carci-
analysis. J Natl Cancer Inst. 2009;101:993–1000. noma. Cancer. 2004;100:1595–604.
338. Amirikia KC, Mills P, Bush J, Newman LA. Higher 352. Budman DR, Berry DA, Cirrincione CT, Henderson
population-based incidence rates of triple-negative IC, Wood WC, Weiss RB, et al. Dose and dose inten-
breast cancer among young African-American sity as determinants of outcome in the adjuvant treat-
women: implications for breast cancer screening ment of breast cancer. The Cancer and Leukemia
recommendations. Cancer. 2011;117:2747–53. Group B. J Natl Cancer Inst. 1998;90(16):1205–11.
339. Gold HT, Do HT, Dick AW. Correlates and effect of 353. Hryniuk W. Importance of chemotherapy schedul-
suboptimal radiotherapy in women with ductal car- ing: pieces of the puzzle. Cancer Invest. 1999;17:
cinoma in situ or early invasive breast cancer. 545–6.
Cancer. 2008;113:3108–15. 354. Wood WC, Budman DR, Korzun AH, Cooper MR,
340. Griggs JJ, Culakova E, Sorbero ME, Poniewierski MS, Younger J, Hart RD, et al. Dose and dose intensity of
Wolff DA, Crawford J, et al. Social and racial adjuvant chemotherapy for stage II, node-positive
differences in selection of breast cancer adjuvant breast carcinoma. N Engl J Med. 1994;330:1253–9.
chemotherapy regimens. J Clin Oncol. 2007;25: 355. Lyman GH, Dale DC, Crawford J. Incidence and
2522–7. predictors of low dose-intensity in adjuvant breast
341. Hershman D, McBride R, Jacobson JS, Lamerato L, cancer chemotherapy: a nationwide study of com-
Roberts K, Grann VR, et al. Racial disparities in munity practices. J Clin Oncol. 2003;21:4524–31.
treatment and survival among women with early- 356. Bonadonna G, Valagussa P. Dose-response effect of
stage breast cancer. J Clin Oncol. 2005;23:6639–46. adjuvant chemotherapy in breast cancer. N Engl J
342. Hershman DL, Wang X, McBride R, Jacobson JS, Med. 1981;304:10–5.
Grann VR, Neugut AI. Delay in initiating adjuvant 357. Bonadonna G, Valagussa P, Moliterni A, Zambetti M,
radiotherapy following breast conservation surgery Brambilla C. Adjuvant cyclophosphamide,
4 Breast Cancer Genomics 103

methotrexate, and fluorouracil in node-positive 374. Lafarge S, Sylvain V, Ferrara M, Bignon YJ. Inhibition
breast cancer: the results of 20 years of follow-up. of BRCA1 leads to increased chemoresistance to
N Engl J Med. 1995;332:901–6. microtubule- interfering agents, an effect that involves
358. Hryniuk WM, Levine MN, Levin L. Analysis of dose the JNK pathway. Oncogene. 2001;20:6597–606.
intensity for chemotherapy in early (stage II) and 375. Sudo T, Nitta M, Saya H, Ueno NT. Dependence of
advanced breast cancer. NCI Monogr. 1986;1:87–94. paclitaxel sensitivity on a functional spindle assem-
359. Ferguson-Smith AC, Surani MA. Imprinting and the bly checkpoint. Cancer Res. 2004;64:2502–8.
epigenetic asymmetry between parental genomes. 376. Chabalier C, Lamare C, Racca C, Privat M, Valette A,
Science. 2001;293:1086–9. Larminat F. BRCA1 down- regulation leads to prema-
360. Reik W, Lewis A. Co-evolution of X-chromosome ture inactivation of spindle checkpoint and confers
inactivation and imprinting in mammals. Nat Rev paclitaxel resistance. Cell Cycle. 2006;5:1001–7.
Genet. 2005;6:403–10. 377. McGrogan BT, Gilmartin B, Carney DN, McCann
361. Widschwendter M, Jones PA. DNA methylation and A. Taxanes, microtubules and chemoresistant breast
breast carcinogenesis. Oncogene. 2002;21:5462–82. cancer. Biochim Biophys Acta. 2008;1785:96–132.
362. Takai D, Jones PA. Comprehensive analysis of CpG 378. Quinn JE, Kennedy RD, Mullan PB, Gilmore PM,
islands in human chromosomes 21 and 22. Proc Natl Carty M, Johnston PG, et al. BRCA1 functions as a
Acad Sci U S A. 2002;99:3740–5. differential modulator of chemotherapy-induced
363. Baylin SB, Ohm JE. Epigenetic gene silencing in apoptosis. Cancer Res. 2003;63:6221–8.
cancer—a mechanism for early oncogenic pathway 379. Haiman C, Greene G. Genetic variation within the
addiction? Nat Rev Cancer. 2006;6:107–16. coding region of steroid hormone receptor co-
364. Jones PA, Baylin SB. The fundamental role of epi- activator and co-repressor genes and breast cancer
genetic events in cancer. Nat Rev Genet. 2002;3: risk 2007. Los Angeles: American Association for
415–28. Cancer Research; 2007.
365. Ting AH, McGarvey KM, Baylin SB. The cancer 380. Herynk MH, Fuqua SA. Estrogen receptor mutations
epigenome—components and functional correlates. in human disease. Endocr Rev. 2004;25:869–98.
Genes Dev. 2006;20:3215–31. 381. Giacinti L, Claudio PP, Lopez M, Giordano A.
366. Bird A. Perceptions of epigenetics. Nature. 2007;447: Epigenetic information and estrogen receptor a expres-
396–8. sion in breast cancer. Oncologist. 2006;11:1–8.
367. Esteller M, Fraga MF, Guo M, Garcia-Foncillas J, 382. Wei M, Xu J, Dignam J, Nanda R, Sveen L,
Hedenfalk I, Godwin AK, et al. DNA methylation Fackenthal J, et al. Estrogen receptor a, BRCA1, and
patterns in hereditary human cancers mimic sporadic FANCF promoter methylation occur in distinct sub-
tumorigenesis. Hum Mol Genet. 2001;10:3001–7. sets of sporadic breast cancers. Breast Cancer Res
368. Wei M, Grushko TA, Dignam J, Hagos F, Nanda R, Treat. 2008;111:113–20.
Sveen L, et al. BRCA1 promoter methylation in spo- 383. Zhou Q, Atadja P, Davidson NE. Histone deacetylase
radic breast cancer is associated with reduced inhibitor LBH589 reactivates silenced estrogen
BRCA1 copy number and chromosome 17 aneu- receptor a (ER) gene expression without loss of DNA
somy. Cancer Res. 2005;65:10692–9. hypermethylation. Cancer Biol Ther. 2007;6:64–9.
369. Birgisdottir V, Stefansson OA, Bodvarsdottir SK, 384. Arce C, Pérez-Plasencia C, González-Fierro A, de la
Hilmarsdottir H, Jonasson JG, Eyfjord JE. Epigenetic Cruz-Hernández E, Revilla-Vázquez A, Chávez-
silencing and deletion of the BRCA1 gene in spo- Blanco A, et al. A proof-of-principle study of epi-
radic breast cancer. Breast Cancer Res. 2006;8:R38. genetic therapy added to neoadjuvant doxorubicin
370. van’t Veer LJ, Dai H, van de Vijver MJ, He YD, cyclophosphamide for locally advanced breast can-
Hart AA, Bernards R, et al. Expression profiling pre- cer. PLoS One. 2006;1:e98.
dicts outcome in breast cancer. Breast Cancer Res. 385. Fackler MJ, Malone K, Zhang Z, Schilling E,
2003;5:57–8. Garrett-Mayer E, Swift-Scanlan T, et al. Quantitative
371. Grushko TA, Dignam JJ, Das S, Blackwood AM, multiplex methylation-specific PCR analysis dou-
Perou CM, Ridderstrale KK, et al. MYC is amplified bles detection of tumor cells in breast ductal fluid.
in BRCA1-associated breast cancers. Clin Cancer Clin Cancer Res. 2006;12:3306–10.
Res. 2004;10:499–507. 386. Yan PS, Venkataramu C, Ibrahim A, Liu JC, Shen RZ,
372. Foulkes WD. BRCA1 functions as a breast stem cell Diaz NM, et al. Mapping geographic zones of cancer
regulator. J Med Genet. 2004;41:1–5. risk with epigenetic biomarkers in normal breast tis-
373. Liu S, Ginestier C, Charafe-Jauffret E, Foco H, sue. Clin Cancer Res. 2006;12:6626–36.
Kleer CG, Merajver SD, et al. BRCA1 regulates 387. Visvanathan K, Sukumar S, Davidson NE. Epigenetic
human mammary stem/progenitor cell fate. Proc biomarkers and breast cancer: cause for optimism.
Natl Acad Sci U S A. 2008;105:1680–5. Clin Cancer Res. 2006;12:6591–3.
Epigenomics of Breast Cancer
Kursat Oguz Yaykasli, Ertugrul Kaya,
and Emine Yaykasli

Breast cancer is the second most common malignant cancer and accounts
for 1.38 million of the total new cancer cases and 458,400 of the total
cancer deaths reported in 2008. Breast cancer with several subtypes is an
extremely heterogeneous disease caused by interaction of both genetic and
environmental risk factors. In order to understand the etiology of this het-
erogeneity, new perspectives like epigenetics are needed.
The term epigenetics was coined by Conrad Hal Waddington in the
early 1940s. It refers to the study of gene function and regulation altera-
tions without changes in the DNA sequence of the genome. The main
epigenetic modifications are DNA methylation, histone modifications, and
small noncoding RNAs (miRNAs). DNA methylation is the first to be
associated with cancer and the most widely studied among epigenetic
modifications. It regulates the gene expression by modifying the accessi-
bility of DNA to the transcriptional machinery.
The importance of histone modification has been realized during the
last 10 years, after identification of the coexistence of histone modifica-
tions. From the dynamically changing pattern of histone modification has
emerged a new concept termed “histone cross talk.” The epigenetic modi-
fications are faster and reversible than mutation and easily affected by

K.O. Yaykasli, PhD (*)

Department of Medical Genetics,
School of Medicine, Duzce University,
Duzce, Turkey
e-mail: kursatyay@yahoo.com
E. Kaya, MD
Department of Medical Pharmacology,
School of Medicine, Duzce University,
Duzce, Turkey
E. Yaykasli, MSc
Department of Medical Genetics and Biology,
Graduate School of Health Sciences, Duzce
University, Duzce, Turkey

D. Barh (ed.), Omics Approaches in Breast Cancer: Towards Next-Generation Diagnosis, 105
Prognosis and Therapy, DOI 10.1007/978-81-322-0843-3_5, © Springer India 2014
106 K.O. Yaykasli et al.

aging, environmental stimuli, and food in heritable manner. These charac-

teristics provide a vital position in the etiology of diseases. After several
investigations, it is well understood that the epigenetic modifications are
involved in not only many biological processes such as X-chromosome
inactivation, genomic imprinting, RNA interference, and programming of
the genome but also several disease like breast cancer. Today we realize
that the accumulation of epigenetic modifications occurs in the develop-
ment of breast cancer. In addition, the epigenetic modifications improve
our knowledge about the biology and heterogeneity of breast cancer by
large-scale methods. Therefore, the researchers focused on epigenetic
alterations-based breast cancer therapy, and it is speculated that epigenetic
modifications may be markers for breast cancer. It is likely that epi-
genetics-based therapy will become a reality in the near future.

Epigenetics • Breast cancer • DNA methylation • Histone modification
• miRNA

Introduction Breast cancer is the second most common

malignant cancer and the leading cause of cancer
Cancer is the leading cause of death in economi- death in women, with increasing age bringing a
cally developed countries and the second leading sharp rise in incidence [1, 8]. Breast cancer con-
cause of death in developing countries. It is esti- stituted 23 % (1.38 million) of the total new can-
mated that about 12.7 million cancer cases and 7.6 cer cases and 14 % (458,400) of the total cancer
million cancer deaths occurred in 2008. Among deaths reported in 2008. About half the breast
them, 56 % of the cases and 64 % of the deaths cancer cases and 60 % of the deaths are estimated
occurred in economically developing countries to occur in economically developing countries. In
[1]. In order to effectively combat it, we must general, incidence rates are high in Western and
understand the basic principles and processes of Northern Europe, Australia/New Zealand, and
cancer. At the cellular level, we must understand North America; intermediate in South America,
the complex circuitries that dictate the cell-divi- the Caribbean, and Northern Africa; and low in
sion cycle, survival, migration, and invasion. At sub-Saharan Africa and Asia [9]. The American
the tissue level, the susceptible target cell popu- Cancer Society estimates that approximately
lation and the interactions between cancer cells 230,480 new cases of invasive breast cancer and
and the microenvironment must be understood. 39,520 breast cancer deaths occurred among US
Finally, the complex features that establish can- women in 2011 [10].
cer “organ” at primary and distant sites, including Breast cancer is an extremely heterogeneous
metabolic and physiological effects and the estab- disease with molecular, histological, and pheno-
lishment of a blood nutrient supply (angiogene- typic diversity caused by interaction of both
sis), must also be clarified [2]. It had traditionally inherited and environmental risk factors (age,
been considered that the underlying foundation obesity, alcohol intake, lifetime estrogen expo-
of the mechanism of cancer development is the sure, and mammographic density). Breast cancer
accumulation of genetic mutations. However, this can be classified into five major subtypes that dif-
paradigm has now been expanded to incorporate fer significantly with regard to both molecular
the distribution of epigenetic regulatory mecha- and clinical features. These subtypes are luminal
nisms that are prevalent in cancer [3–7]. A, luminal B, triple negative/basal like, HER2
5 Epigenomics of Breast Cancer 107

enriched, and normal like [11, 12]. However, the suppressor genes become silenced. Accumulation
main classification of breast cancer is based on of epigenetic modifications is also associated
the presence or absence of the estrogen receptor with oncogenesis. The epigenetic modifications
(ER), and investigations have been done regard- occur early during carcinogenesis as potentially
ing these subtypes [13]. Estrogens, sex steroid initiating events for cancer development, they
hormones, are responsible for the development of have been identified as promising new targets for
sex characteristics like breasts. Estrogens have cancer prevention strategies [24]. Nowadays, the
also been recognized as the major factor in the epigenetic mechanisms are known to be involved
development of breast cancers. The activity of in several cancer types and diseases [25–34]. The
estrogens is mediated by the two main isoforms epigenetic mechanisms also explain how two
of intracellular estrogen receptors (ERs): ERα identical genotypes can give rise to different phe-
and ERβ. The major breast cancer subtypes are notypes in response to the same environmental
ER-positive and ER-negative tumors [14, 15]. stimulus [35].
Our knowledge about breast cancer subtypes has
increased since developing new high-throughput
molecular techniques, such as microarray, next- Epigenomic Markers for Breast
generation sequencing, etc. [16, 17], and new Cancer Diagnosis
perspectives like epigenetics [18–20].
The geographical variation of the incidence of
breast cancer ratio indicates a significant role of
Applications of Epigenomics factors affecting the epigenetic mechanism for
of Breast Cancer the breast cancer risk [36]. Because epigenetic
modifications are significant factors in the devel-
In the early 1940s, Conrad Hal Waddington opment of breast cancers, the assessment of the
coined the term epigenetics as “the causal inter- breast cancer in terms of the epigenetics could
actions between genes and their products, which strongly improve our understanding of the biol-
bring the phenotype into being.” Nowadays, epi- ogy and heterogeneity of breast cancer [37, 38].
genetics refers to the study of gene function and The best-known epigenetic markers are DNA
regulation alterations in heritable manner. Unlike methylation, histone modifications and chroma-
the genotoxic mechanism involving changes in tin remodeling, and miRNAs.
genomic DNA sequences leading to mutations,
the epigenetic modifications modulate the gene DNA Methylation and Breast Cancer
expression directly without changes in the DNA The first and most widely studied epigenetic
sequence of the genome. Epigenetic mecha- modification in mammals is DNA (cytosine)
nisms coordinate biological processes such as methylation [39, 40]. DNA methylation plays a
X-chromosome inactivation, genomic imprint- crucial role in modulating the expression of the
ing, RNA interference, and programming of the genetic information by modifying the accessibil-
genome during differentiation and development ity of DNA to the transcriptional machinery,
leading to gene silencing. Both the genetic and generally resulting in transcriptional gene
epigenetic events change the function and regula- silencing (activation of the gene is also possible
tion of the gene products or lead to gain/loss of in some cases). It involves several mechanisms
function of genes. like imprinting, X-chromosome inactivation,
It is now acknowledged that genetic altera- and inhibition of repeat elements and transpo-
tions are not the only path to gene disruption; sons transcription [41, 42]. DNA methyltrans-
reversible epigenetic modifications are increas- ferases (DNMTs) catalyze the addition of the
ingly being considered in cancer [18, 21–23]. In methyl group (-CH3) using S-adenosyl-
cancer cells, oncogenes are activated by muta- methionine (SAM) as a methyl donor to dC resi-
tions or overexpression, whereas tumor- dues in DNA.
108 K.O. Yaykasli et al.

Although most cytosine methylation occurs MBD2, MBD3, and MeCP2 provide a platform
in the sequence CpG dinucleotides, the cytosine for the DNA methylation [57], and it has already
nucleotide of the CpA and CpT dinucleotides been determined that the mutations in DNMTs
may also be methylated in some cases. CpG and MBDs contribute to diseases like acute
dinucleotides are found throughout the genome, myeloid leukemia (AML) [58, 59].
but largely concentrated in small regions termed There are three types of DNA methylation:
“CpG islands” [43, 44]. CpG islands are short hypermethylation, hypomethylation, and loss of
sequences (length of 0.5 kilobases to several imprint. The CpG islands in the promoter region
kilobases) of genomic DNA with a G + C con- are commonly unmethylated (genes active) in
tent of at least 50 % and a ratio of observed to normal tissues. In hypermethylation the CpG
statistically expected CpG frequencies of at least islands in the promoter region are aberrantly
0.6. It is found in approximately 60–70 % of methylated, leading to gene silencing through the
gene promoters commonly 5′-regulatory (pro- inhibition of transcription via recruitment of
moter) regions of many “housekeeping” genes chromatin remodeling corepressor complexes.
(which are essential for general cell function) The loss of DNA methylation occurs in many
and some tissue-specific genes [45–47]. CpG gene-poor genomic areas including repetitive
islands also can be found in the 3′-region of the elements, retrotransposons, and introns at hypo-
gene and within the body of the genes (refer- methylation. It causes genomic instability and
ring to exonic CpG island) [48]. Recent studies leads to reactivation of the genes. Loss of imprint-
showed that not only methylation of CpGs in ing could be explained as the loss of specific
promoter but also methylation of CpGs within monoallelic expression of genes in a parent-
the gene bodies associate with transcriptional origin-specific manner [12, 35, 37].
activation [49]. In contrast to expectation (the DNA methylation is the first epigenetic mecha-
methylation level negatively correlates with the nism to be associated with cancer after demonstra-
gene expression level), there is a positive cor- tion of global DNA hypomethylation and
relation between gene-body methylation and CpG-island hypermethylation in cancer tissues
gene activity in humans. It is proposed that the compared to normal tissues [60]. Global and gene-
gene-body methylation may repress transcrip- specific DNA hypomethylation and site-specific
tional noise, inhibit antisense transcription, and hypermethylation are common features in tumori-
relate to replication timing [21, 50]. Intragenic genesis [61]. The most extensively studied epigen-
methylation is also found at repetitive sequences etic alteration in cancer is DNA methylation of
in human DNA [51]. To date, several DNMTs CpG islands. When the CpG islands of important
(DNMT1p, DNMT1b, DNMT1o, DNMT1p, genes like tumor-suppressor genes are hypermeth-
DNMT2, DNMTB3a, DNMT3b, and DNMT3L) ylated, the tumor-suppressor genes become inac-
have been identified. Among them only DNMT1, tive and cancer emerges [62–64]. Nowadays, the
DNMTB3a, and DNMT3b have catalytic meth- next-generation sequencing (NGS) platform gives
yltransferase activity [52]. DNMT1 recognizes us enormous data relating genome-wide maps of
established hemimethylated DNA (the one strand CpG methylation. It is demonstrated that 5–10 %
of the CpG dinucleotides methylated, the other of normally unmethylated CpG promoter islands
one not) and then methylates newly synthesized become abnormally methylated in various cancer
CpG dinucleotide whose partners on the parental genomes, and hypermethylation of promoter
strand are already methylated [53, 54]. Besides region also affects expression of various noncod-
the capability of methylating hemimethylated ing RNAs, some of which have a role in malignant
DNA, the primary function of DNMT3a and transformation [5, 64]. DNA hypomethylation is
DNMT3b is capable of de novo methylation pat- observed in several tumor types, such as colorectal
terns (both strands of the CpG dinucleotides are and gastric cancers, melanomas, etc. [65].
not methylated) during embryogenesis [55, 56]. Decreased DNA methylation is thought to pro-
Several methyl-binding proteins such as MBD1, mote chromosomal instability, eventually leading
5 Epigenomics of Breast Cancer 109

to carcinogenesis. Genome-wide DNA hypometh- gland epithelium. Similarly, the loss of E-cadherin
ylation also affects transcription through loss of expression in all tumor stages of breast cancer
imprinting and upregulation of silent genes, all of has been observed due to hypermethylation of
which might induce tumor development [66]. CpG islands. Therefore, epigenetic suppression
During tumor progression, the degree of hypo- of ERα and E-cadherin may occur prior to inva-
methylation of genomic DNA increases as the sion and then increases as cells acquire invasive-
lesion derives from a benign proliferation of cells ness and metastatic potential [18, 78].
to an invasive cancer [67]. Cancer is a disease characterized by uncon-
Because of the high histological and molecular trolled cell division due to checkpoints damaged
heterogeneity, the assessment of breast cancer in by several factors such as chemical, UV, etc. [83,
terms of the epigenetic aspects, especially DNA 84]. Despite the fact that the exact role of the
methylation, helps us to clarify the breast cancer BRCA1 protein is not clarified in detail, BRCA1
mechanism. It is speculated that changed DNA protein is known to be a tumor-suppressor gene. It
methylation pattern of global or specific genes, involves several important biological processes,
such as RASSF1A, GHSR, etc., may be markers such as DNA repair damage, induction of apopto-
for breast cancer, after their appearance in sev- sis, etc. [85–87]. The mutations on BRCA1 and
eral studies [68–71]. To find a reliable biomarker, BRCA2 genes increase the development of famil-
several changed methylation pattern genes in ial breast cancers [88, 89]. The other mechanism
breast cancer have been reported during the last of suppression of BRCA1 expression is hyper-
decade, based on the tumor’s clinicopathological methylation of promoter region of genes. Recent
characteristics, such as hormonal receptor status studies have shown that suppression of BRCA1
[72–76]. The methylated RASSF1A, CCND2, expression by hypermethylation is involved not
GSTP1, and TWIST genes for ER-positive breast only in breast and ovarian cancer but also lung
cancers and PGR, TFF1, and CDH13 genes, pre- and oral cancers [90, 91].
dominantly for ER-negative breast cancers, have Hypermethylation of CpG islands resulting
been linked [12, 37]. from overactivity of DNMTs occurs in many
The involvement ERα in breast cancer is cancers. Several studies reported that DNMTs
already known, and ERα is expressed approxi- are also overexpressed in breast cancer [92, 93].
mately in 65–75 % of diagnosed breast tumors. A recent study in Tunisian breast cancer showed
ERα is encoded by the estrogen receptor 1 overexpression of three hypermethylating
(ESR1) gene. The promoter region and first exon enzymes (DNMT1, DNMT3a, and DNMT3b)
of the ESR1 gene contain five CpG islands by immunohistochemistry. They found that over-
[77, 78]. Several mechanisms relating the lack of expression of various DNA methyltransferases
ERα expression in ER-negative breast cancer might be involved in epigenetic inactivation of
have been proposed to date. Among them, the multiple tumor-suppressor genes, leading to the
suppression of the ESR1 gene by hypermethyl- development of aggressive forms of sporadic
ation of CpG islands has been investigated [79]. breast cancer [94]. However, re-expression of
DNMTs are responsible for this methylation, and promoter-methylated genes can be achieved after
it has been demonstrated that the re-expression of DNMT inhibitor treatment, such as 5-aza-2′-
the ER gene is possible by a DNMT1 inhibitor deoxycytidine treatment [95, 96].
(5-aza-2′-deoxycytidine) or antisense oligonu- The other epigenetic mechanism, hypometh-
cleotide for inhibiting DNMT1 specifically [80, ylation, also is involved in activating genes in
81]. A recent study also showed that ER pro- breast cancer. The promoter region of the MDR1
motes genomic methylation through upregulation gene is always highly methylated in normal con-
of DNMT1 in ER-positive breast cancer cells ditions, while its hypomethylation occurs during
[82]. Another important molecule in breast can- tumorigenesis, and it might be a putative implica-
cer, E-cadherin, is responsible for maintaining tion in biological aggressiveness of tumors [97].
the normal differentiated state of the mammary Several hypermethylated and hypomethylated
110 K.O. Yaykasli et al.

genes are involved in biological functions linked Histone Modifications in Breast Cancer
to breast cancer. The demonstrated genes are in The chromatin is a highly organized structure of
listed Table 5.1. DNA and protein. The organization of DNA in
Moreover, global hypomethylation can be chromatin (euchromatin, active; heterochroma-
seen in breast cancer. It is widely assumed that tin, inactive) has many functions, such as packag-
global hypomethylation activates the gene ing DNA into smaller volume, preventing DNA
expression. However, it might decrease the gene damage, and controlling DNA replication, tran-
expression when accompanied by a gain of scription, and repair [100]. The fundamental unit
repressive chromatin. Taken together, it has been of chromatin is the nucleosome, an octomeric
found that the global hypomethylation silences structure containing two copies each of histones
tumor-suppressor genes via repressive chromatin (H3, H4, H2A, and H2B) around which 147 base
domains in breast cancer [98]. pairs of DNA are wrapped [101]. The states of
Male breast cancers often differ from female chromatin are controlled by chemical modifica-
breast cancers in several respects. Kornegoor tion of histone tail (N-terminus) via posttran-
et al. studied the comparison of male and female scriptional including acetylation, methylation,
breast cancers in terms of the DNA methylation phosphorylation, sumoylation, poly(ADP)-ribo-
patterns. The methylation patterns of the most sylation, and ubiquitination and histone com-
frequently methylated genes (MSH6, WT1, position in conjunction with other nonhistone
PAX5, CDH13, GATA5, and PAX6) were found proteins [102, 103].
to be similar in male and female breast cancer. It was first proposed in 1964 that histone mod-
On the other hand, methylation occurred less ifications may affect the regulation of gene
often in male breast cancer when compared to expression, after demonstrating acetylation of the
female breast cancer [99]. ε-amino group of lysine residues on histones

Table 5.1 Hypermethylated and hypomethylated genes in human breast cancer

Sample Methy.
Gene (description) Function obtained Case # status Marker Reference
14-3-3-σ/stratifin (SFN) Cell cycle Cell lines, 20 Hyper Therapeutic Ferguson et al.
regulation tissue [202]
14-3-3-σ/stratifin (SFN)å Cell cycle Serum 100 Hyper Diagnostic, Mirza et al. [203]
regulation prognostic
ESR1 (estrogen receptor 1) Cell cycle Serum 106 Hyper Diagnostic Martínez-Galán
or 14-3-3-σ/stratifin (SFN) regulation et al. [204]
RASSF1A (ras association Cell cycle Cell lines, 45 Hyper Therapeutic Dammann et al.
domain family protein1) regulation tissue [205]
APC (adenomatous polyposis Inhibitor of Tissue 50 Hyper Therapeutic Jin et al. [206]
of the colon) β-catenin
RASSF1, APC, DAPK1 Serum 34 Hyper Diagnostic Dulaimi et al.
RARβ (retinoic acid Cell cycle Cell lines, 24 Hyper Therapeutic Sirchia et al.
receptor β) regulation tissue [208]
RASSF1A and RARβ Cell cycle Serum 20 Hyper Diagnostic, Shukla et al.
regulation prognostic [209]
RASSF1A or ATM Cell cycle Plasma 50 Hyper Diagnostic Papadopoulou
regulation et al. [210]
RASSF1, RARB, MGMT, Serum, 33 Hyper Prognostic Taback et al.
APC tissue [211]
TMS1 (target of methylation- Involved in Cell lines, 27 Therapeutic Conway et al.
induced silencing-1) apoptosis tissue [212]
5 Epigenomics of Breast Cancer 111

Table 5.1 (continued)

Sample Methy.
Gene (description) Function obtained Case # status Marker Reference
TMS1, BRCA1, ERα, and Serum 50 Hyper Diagnostic Mirza et al. [173]
CCND2 (cyclin D2) Cell cycle Tissue 106 Hyper Diagnostic, Evron et al.
regulation prognostic [213]
CCND2, CDKN2A, and Serum, 36 Hyper Diagnostic, Sharma et al.
SLIT2 tissue prognostic [214]
CDH1 (E-Kadherin) Cell adhesion Tissue 151 Hyper Prognostic Shinozaki et al.
and invasion [215]
CDH1 (E-Kadherin) Cell adhesion Tissue 79 Hyper Prognostic Caldeira et al.
and invasion [216]
CDKN2A (cyclin-dependent Cell cycle Plasma 35 Hyper Diagnostic Silva et al. [217]
kinase inhibitors) regulation
CDKN2A or CDH1 Serum 36 Hyper Diagnostic, Hu et al. [218]
CDH 13 (H-Kadherin) Cell adhesion Cell lines, 55 Hyper Therapeutic Toyooka et al.
and invasion tissue [219]
BRCA1 (breast cancer 1) DNA repair and Tissue 143 Hyper Diagnostic Birgisdottir et al.
recombination [220]
BRCA1, CDKN2A, or Serum 38 Hyper Diagnostic Jing et al. [221]
APC, RASSF1, or ESR1 Serum 79 Hyper Prognostic Van der Auwera
et al. [222]
GSTP1 (glutathione-S- Carcinogen Tissue 77 Hyper Prognostic Esteller et al.
transferase P1) detoxification [223]
GSTP1, RARB, RASSF1, or Plasma 47 Hyper Diagnostic Hoque et al.
APC [224]
TWIST (TWIST homology of Involved in cell Mammary 72 Hyper Therapeutic Vesuna et al.
drosophila) death ducts’ [225]
CCND2, RARB, TWIST1, or Plasma 34 Hyper Diagnostic Bae et al. [226]
RUNX3 (run-related Transcriptional Cell lines, 44 Hyper Diagnostic Lau et al. [227]
transcription factor 3) regulation tissue
RUNX3, CDKN2A, RASSF1, Serum 19 Hyper Diagnostic, Tan et al. [228]
or CDH1 prognostic
MDR1 (multidrug resistance Transmembrane Serum, 100 Hypo Prognostic Sharma et al.
1) efflux pump tissue [97]
CAV1 (Caveolin 1) Cell invasion, Cell line 30 Hypo Prognostic Rao et al. [229]
NAT1 (N-acetyltransferase Cell invasion, Tissue 103 Hypo Prognostic Kim et al. [230]
type 1) metastasis
UPA (Urokinase) Cell invasion, Cell line 1 Hypo Therapeutic Pakneshan et al.
metastasis [231]

[104]. After nearly half a century, it is has been with activation or repression led to the proposal
elucidated that the posttranscriptional modifica- that the modification constitutes a code that could
tions of histone tails determine not only be recognized by transcription factors to deter-
transcriptional activity but also all DNA- mine the transcriptional state of a gene 10 years
templated processes. The identification of the before [105]. However, these patterns appear to
coexistence of histone modifications associated be not static, and a dynamically changing and
112 K.O. Yaykasli et al.

complex landscape via the chromatin signaling polymerase and transcription factors easier to
pathway led to the new concept termed “histone access the promoter region. So histone acetylation
cross talk.” This term represents the influence one facilitates gene expression by allowing transcrip-
or more coexisting histone modifications have on tion factors to access the DNA. In contrast, the
the deposition, interpretation, or erasure of other HDACs remove the acetyl group from histones to
histone modifications [5, 106]. The recent inves- coenzyme A (CoA), resulting in coiling of chro-
tigations showed that histone cross talk mecha- matin, which inhibits transcription [22, 103].
nisms commonly seen and have a great importance At least 25 HATs and 18 HDACs have been
for biological processes in organism [107, 108]. identified in humans [114]. HATs were the first
Histone modifications affect the chromosome enzymes shown to modify histones [115]. There
function via several mechanisms. Generally it is are two major classes of HATs: type A and type
believed that histone modifications cause struc- B. The type A HATs are nuclear proteins and can
tural changes in histone. This structural change be grouped into at least three families—Gcn5/
may act as specific binding sites for protein PCAF, MYST, and p300/CBP—depending on
domains (e.g., bromodomains, chromodomains, amino acid sequence homology [116]. In contrast
tudor domains) [109, 110]. Among the epigenetic to type A HAT, the type B HATs are predomi-
mechanisms, histone modifications have further nantly cytoplasmic and show similar highly con-
grown over the last decade with the discovery served primary structures, with acetylate-free
and characterization of a large number of histone- histones but not those already deposited into chro-
modifying molecules and protein complexes. The matin, and newly synthesized histones H4 at K5
deregulation of these molecules or complexes and K12. This pattern of acetylation is important
may lead to deregulation of the control of for deposition of the histones [117]. The HDACs
chromatin-based processes by changing histone also have critical importance in the regulation of
modifications and may have been associated with expression of genes involving cell survival, prolif-
a large number of human malignancies. Genome- eration, differentiation, and apoptosis and can be
wide studies revealed that the histone modifica- divided into four major groups depending on
tions of malignant cells patterns disrupted when sequence homology and target both histone and
compared to healthy cells [111]. The posttransla- nonhistone proteins. Class I includes HDACs 1, 2,
tional modification at amino acid tail of histone 3, and 8; class II includes HDACs 4, 5, 6, 7, 9, and
protein may result in changed transcription of 10; and class IV includes HDAC 11. In contrast to
important genes such as tumor suppressors. other HDACs, class III HDACs consist of NAD+-
Changed patterns of histone modifications are a dependent sirtuin family 1–7 [5]. HDACs also
hallmark of cancer, and great amount of histone regulate the expression of tumor-suppressor and
modifications have been linked to several cancer specific cell cycle regulatory genes. It has been
types to date [112]. The most well-known histone observed that high HDAC expression level and
modifications types are acetylation/deacetylation hypoacetylation can be seen in several cancers. So
and methylation/demethylation [113]. HDAC inhibitors have been targeted for cancer
therapy [118, 119]. The mechanism of the antip-
Histone Acetylation/Deacetylation roliferative effects of HDAC inhibitors is com-
in Breast Cancer plex. The target of HDAC inhibitors is the zinc
Histone acetylation/deacetylation status regulates cofactor at the active site of the HDACs to change
several important regulatory proteins and tran- chromatin structure and cause re-expression of
scription factors and is controlled by the interplay aberrantly silenced genes [120].
of histone acetyltransferases (HATs) and histone
deacetylases (HDACs), respectively. HATs trans- Histone Methylation/Demethylation
fer acetyl groups from acetyl-CoA to the amino in Breast Cancer
group of lysine residues in histone tail. It removes Besides the gene promoter regions, the methyla-
the positive charges, thereby reducing the affinity tion/demethylation can occur on histone protein
between histones and DNA. This makes RNA residues. DNA methylation at CpG islands of
5 Epigenomics of Breast Cancer 113

promoter regions generates long-term gene described several malignancies, suggesting

silencing and makes the majority chromatin inac- EZH2 was tumor-suppressor gene [134, 135]. In
cessible for transcription, but histone methylation addition, some chemicals like diethylstilbestrol
results in short-term inhibition of gene expres- (DES) or bisphenol A (BPA) contribute to the
sion. Methylation, unlike acetylation and phos- formation of breast cancer by increasing EZH2
phorylation, does not alter the overall charge of expression [136].
the molecule [5, 18]. Histone methylation takes Another study relating EZH2 to breast cancer
place at lysine and arginine residues by histone concluded that the overexpression of EZH2 regu-
methyltransferases (HMTs). HMTs transfer a lates BRACA1 gene expression and genomic
methyl group from the cofactor S-adenosyl instability mediated by PI3K/Akt-1 pathway
methionine to lysine or arginine residues on his- [137]. These investigations suggest that EZH2
tone tails, which play important roles in chroma- histone methyltransferase is involved in breast
tin remodeling and transcriptional activity. The cancer etiology.
methylation at arginine residue of histone tails The HMT G9a methylates at the ε-amino
usually activates the gene transcription, although group of lysine 9 residues of histone 3. It has
it may be involved in transcriptional repression in also been proven that G9a is involved in Snail-
some cases. The methylation at lysine residue of mediated E-cadherin repression by interacting
histone tails can contribute to either activation or with Snail in human breast cancer [138].
repression of transcription, depending on the Another study proposed that G9a contributes to
position of methylation, and adjacent modifica- the estradiol (E2)-dependent induction of some
tions [121, 122]. Some lysine methylases (like endogenous target genes of estrogen receptor
H3K4, H3K36, H3K79) often activate genes in (ER)α in MCF-7 breast cancer cells [139]. Other
euchromatin, while others (like H3K9, H3K27, lysine HMTs (NSD1, NSD3L, and SMYD3) are
and H3K20) are associated with heterochromatin overexpressed in several cancers [125, 140].
regions of the genome. The methylation status Unlike lysine HMT, arginine HMTs have not
(mono-, di-, or trimethylation) also alters gene been as well characterized. Arginine HMTs cat-
expression. For example, the monomethylations alyze methylation of nitrogen of arginine resi-
of H3K27, H3K9, H4K20, H3K79, and H2BK5 dues, called protein arginine methyltransferases
are all linked to gene activation, whereas trimeth- (PRMT). The 10 PRMTs are nearly identified
ylations of H3K27, H3K9, and H3K79 are linked and categorized into two groups based on the
to repression [123]. Histone demethylases type’s methylarginine products they produced
(HDMs), discovered nearly 7 years ago, have [141]. Among PRMTs, the altered PRMT1 gene
been classified into two groups depending on expression has been investigated in breast can-
their mechanism of action [124]. cer [142].
Several HMTs and HDTs relevant to cancer Several types of histone lysine demethylases
development have been identified to date [125]. (HDMs) have been identified, but the pathologi-
The EZH2 one of the HMTs acts mainly as a cal roles of their dysfunction in human disease
gene silencer; it is the major enzyme that methyl- have not been clarified. Among them, lysine-
ates lysine-27 of histone H3 (H3K27). EZH2 can specific demethylase (LSD1) is the first identified
add up three methyl groups to the ε-amino group histone lysine demethylase. LSD1 specifically
of the lysine side chain, leading to chromatin demethylates histone H3 lysine 4, which is linked
condensation [126, 127, 128]. The overexpres- to active transcription [143]. After discovery of
sion of EZH2 is seen in many cancer types, LSD1, the concept of histone methylation
including prostate and melanoma [129, 130]. The changed, and it is understood that histone meth-
elevated EZH2 levels are associated with breast ylation is a dynamically regulated process under
cancer [131]. It also correlates with tumor aggres- enzymatic control rather than chromatin marks
siveness and poor prognosis, which suggests that that could only be changed by histone replacement
EZH2 was an oncogene [132, 133]. However, [19]. It has been reported that the expression
loss-of-function mutations in EZH2 gene have level of LSD1 is elevated in human bladder [144],
114 K.O. Yaykasli et al.

small cell lung, colorectal, and neuroblastoma expression that control both physiological and
cancers, and the mutation of LSD1 gene causes pathological processes, such as DNA methyla-
prostate cancer [145]. tion, development, differentiation, apoptosis, and
In breast cancer, LSD1 expression has been proliferation [155, 156]. miRNAs are synthesized
found to be strongly upregulated in ER-negative and processed in the nucleus, exported to the
breast cancer; it makes LSD1 a putative bio- cytoplasm, and then bind to the target mRNA. The
marker for aggressive tumor biology and a novel regulation of RNA transformation by miRNA is
attractive therapeutic target for treatment of accomplished through RNA-induced silencing
ER-negative breast cancer [146]. It is also dem- complex (RISC). miRNAs can inhibit mRNA
onstrated that LSD1 inhibits the invasion of translation or degrade mRNA [157, 158]. Major
breast cancer cells in vitro and suppresses breast mechanisms of miRNA deregulation include
cancer metastatic potential in vivo [147]. Other genetic and epigenetic alterations as well as
histone demethylase genes GASC1, PLU-1, and defects in the miRNA processing machinery.
JMJD2B are involved in human breast cancers. Each miRNA regulates multiple mRNAs and,
The GASC1 gene may be linked to the stem cell conversely, each mRNA may be targeted by mul-
phenotypes and show oncogene properties in tiple RNAs (several hundreds). They can act as
human breast cancer [148]. PLU-1 is an H3K4 oncogenes or tumor suppressors and have been
demethylase and plays an important role in the implicated in cancer initiation and progression,
proliferative capacity of breast cancer cells and the profiles of miRNA expression differ
through repression of tumor-suppressor genes, between normal and tumor tissues and between
including BRCA1 [149]. The methylation status tumor types [159–161]. To date, several investi-
of histone H3 lysine 4 (H3K4) and of H3K9 is gations relating to miRNA profiling has led to the
mutually exclusive, and H3K9 trimethyl demeth- identification of miRNAs’ changed expression
ylase JMJD2B is an integral component of the level in human breast cancer [162, 163]. The
H3K4-specific methyltransferase MLL2. It has expression level of these miRNAs was correlated
been demonstrated that the JMJD2B/MLL2 com- with specific breast cancer biopathological fea-
plex interacts to define the methylation status of tures, such as estrogen and progesterone receptor
H3K4 and H3K9 in ERα-activated transcription, expression, tumor stage, vascular invasion, or
and JMJD2B itself is transcriptionally targeted proliferation [164]. miRNAs act as tumor sup-
by ERα and may thus form a feed-forward regu- pressors and are oncogenic in breast cancer like
latory loop in promoting hormonally responsive other cancer types. So, tumor formation may
breast carcinogenesis [150]. JMJD2B also func- arise from the overexpression (or amplification)
tions as coregulator of ERα signaling in breast of oncogenic miRNA and/or reduction (or dele-
cancer growth and mammary gland development tion) of a tumor-suppressor miRNA [165].
[151]. And the histone protein LSD1 is able to miRNA-21 is overexpressed in breast cancer
demethylate nonhistone proteins, such as p53 and like in other cancer types [164, 166]. p53 and
DNMT1 [152, 153]. programmed cell death 4 (PDCD4) are tumor-
suppressor proteins, and the deregulation of them
miRNA in Breast Cancer may lead to cancer development. miRNA have
Scientists have long been aware of the existence been linked to breast cancer by targeting these
of noncoding RNAs (ncRNAs). In spite of the proteins in breast cancer cells [167].
great amount of knowledge about the function
and types of ncRNAs, we are still far from fully
knowing the role of large fractions of the tran- Epigenomic Markers for Breast
scriptome that do not encode for proteins [154]. Cancer Prognosis
Among ncRNAs, microRNAs are 18–25
nucleotides-long RNA molecules encoded in the Despite the extreme heterogeneity of breast can-
genome that are transcribed by RNA polymerase cer, global breast cancer survival rates have
II and important regulators of protein of gene increased during the past decades due to advances
5 Epigenomics of Breast Cancer 115

in the central role of genetic alterations in the BRCA1-associated breast cancer, hereditary
diagnosis, treatment, prevention of breast cancer, or nonhereditary, occurs at early age due to
and prognosis [2, 168]. Survival rates should be involvement of the cellular DNA repair machin-
further improved by finding epigenetic molecular ery. The inactivation of the BRCA1 by hyper-
markers associated with risk assessment and/or methylation has been suggested to be the putative
prognosis of breast cancer. The knowledge about prognostic marker in breast cancer [173]. Besides
epigenetic alterations profiles in detail might the methylation, BRCA1 expression level could
prove vital in many respects. First, it might help be regulated by miRNA-335. Overexpression of
us to estimate breast cancer risk and take precau- miR-335 resulted in an upregulation of BRCA1
tions before breast cancer develops. In addition, mRNA expression, suggesting a functional domi-
there are several subtypes of breast cancer and nance of ID4 signaling [174].
corresponding therapies currently used. Each RASSF1A (Ras association domain family 1
subtype, even each individual, has unique molec- isoform A) is a recently discovered tumor-
ular epigenetic characteristics. The elucidating of suppressor gene. The protein encoded by
epigenetic characteristic might contribute to a RASSF1A interact is involved in the regulation
better estimation of breast cancer prognosis and of the cell cycle, apoptosis, and genetic instabil-
lead to the choice of the most useful therapy ity. Thus, loss or altered expression level of the
[169]. In this way, patients will not be exposed to RASSF1A gene has been associated with several
ineffective toxins associated with expensive ther- cancers. After illustrating the association between
apy. Several reports have proposed that hyper- inactivation of the RASSF1A gene and the hyper-
methylation or hypomethylation of specific genes methylation of its CpG-island promoter region,
and global methylation status might be useful the RASSF1A gene has become the attractive
epigenetic markers for breast cancer prognosis. biomarker for early cancer detection, diagnosis,
The recent studies also included miRNAs’ and prognosis in many cancer types [175, 176].
expression profiles into putative epigenetic mark- The increased methylation level of the RASSF1A
ers of breast cancer. gene was observed in tumor size and lymph node
The major breast cancer subtype is status in breast cancer [177]. Similar results have
ER-positive, and it has generally had a more been obtained by a meta-analysis of published
favorable prognosis than ER-negative tumors. It data conducted with 1795 breast cancer patients.
is well established that ERα and E-cadherin are They concluded that RASSF1A promoter hyper-
frequently involved in pathogenesis of breast methylation associates with worse survival in
cancer. The aberrant methylation of these genes breast cancer patients [178]. These findings have
is associated with malignant progression in indicated the great potential for the methylation
human breast cancer [170]. ERα expression level of the RASSF1A gene in terms of the prognostic
is also regulated by miRNAs in the context of value of the breast cancer.
breast cancer. miRNA-206 [171] and miRNA- EZH2, histone-lysine N-methyltransferase
221/222 [77] target and regulate human ERα. acts as gene silencer by methylation and is related
miRNA-206 was upregulated in ERα-negative to several cancers. The overexpression of EZH2
breast cancer. Another study found that miRNA- is associated with aggressive breast cancer
206 inhibits the expression of ESR1 mRNA because of the enhanced cancer cell proliferation
through two binding sites in the ESR1 3′-untrans- and a marker of poor prognosis in many solid
lated region (3′-UTR). The researchers also tumor carcinomas including breast [179–181].
found other miRNAs (miRNA-18a, miRNA-18b, It has been investigated that several miRNAs
miRNA-193b, and miRNA-302c) targeting to are involved in breast cancer pathogenesis like
ESR1 mRNA in breast cancer cells [172]. cell regulation, and it has been proposed to be a
Therefore, the aberrant methylation of the ESR1 prognostic factor for breast cancer. The miRNA-
gene and certain miRNAs altering the ESR1 gene 17-5p and miRNA-17/20 have been reported to
expression might be putative epigenetic markers be involved in breast cancer cell proliferation
for human breast cancer prognosis. [182, 183]. miRNA-21 also could be a molecular
116 K.O. Yaykasli et al.

prognostic marker for breast cancer and disease expression is at significantly high levels in breast
progression because of its association with cancer patients with small tumors measuring less
advanced clinical stage, lymph node metastasis, than 2 cm, with low histological grade, and in estro-
and patient poor prognosis [184]. gen receptor α- and progesterone receptor-positive
Another strategy to clarify the role of miRNA tumors. However, multivariate analysis concluded
in breast cancer is the analysis of DNA methyla- that the mRNA and protein of HDAC-6 were not
tion and expression miRNAs in combination. independent prognostic factors for both overall sur-
Alteration of methylation in the promoters of vival and disease-free survival [189]. These studies
miRNAs has also been linked to transcriptional led to the development of new therapies for breast
changes in cancers. Morita et al. found that DNA cancer by finding suitable HDAC inhibitors. To
methylation in the proximal promoter of miRNAs date, a number of HDAC inhibitors have been
is tightly linked to transcriptional silencing [185]. designed and synthesized based on their chemical
structure and are generally divided into four groups
including hydroxamic acids, benzamides, cyclic
Applications of Epigenomics in Breast peptide, and aliphatic acids (small chain fatty
Cancer Therapy acids). The potential use of these inhibitors for
breast cancer therapy has been investigated, as
Cancer emerges not only because of the accumu- shown in Table 5.2.
lation of genetic mutations, but also because of Among them, some HDAC inhibitors like
the reversible epigenetic changes. The dynamic vorinostat (SAHA) and romidepsin (FK-228)
alterations of the epigenetic mechanisms offer us have already been approved by the US Food
a new field for developing novel cancer drugs that and Drug Administration for clinical treatment
can react to epigenetically silenced tumor- of cutaneous T-cell lymphoma. Vorinostat is the
suppressor genes [186]. So histone deacetylases first HDAC inhibitor and currently under evalu-
and DNA methyltransferases have become the ation in several phase II trials in breast cancer. It
main targets for cancer therapy. In breast cancer, is already shown that vorinostat has profoundly
epigenetic silencing of tumor-suppressor genes antiproliferative activity and inhibits prolifera-
due to alteration in both HATs and HDACs (his- tion of both ER-positive and ER-negative breast
tone modification) in combination with DNA cancer cell lines [190]. Entinostat (MS-275) and
hypermethylation is commonly observed [187]. panobinostat (LBH-589) HDAC inhibitors are in
The clarification of the epigenetic dysregulation phase I and II studies in combination with endo-
mechanism in breast tumorigenesis has great crine therapies, chemotherapeutic agents, or novel
importance in terms of the development of new targeted therapy in women with breast cancer
therapies for breast cancer patients. [12, 120]. A recent phase II study relating to the
Aberrant HDAC activity has been investigated HDAC inhibitor vorinostat combined with tamox-
in several cancer types, especially in breast cancer. ifen for the treatment of patients with ER-positive
HDAC-1 expression and HDAC-3 protein expres- metastatic breast cancer using 43 patients has
sions were analyzed immunohistochemically on a been done. Even though the number of patients
tissue microarray containing 600 core biopsies was small, they concluded that the combination
from 200 patients by Krusche et al. They found that of vorinostat and tamoxifen is well tolerated and
moderate or strong nuclear immunoreactivity for exhibits encouraging activity in reversing hor-
HDAC-1 was observed in 39.8 % and for mone resistance. HDAC inhibitor with tamoxifen
HDAC-3 in 43.9 % of breast carcinomas. HDAC-1 may restore hormone sensitivity by causing re-
and HDAC-3 expressions correlated significantly expression of a silenced ER gene [191].
with estrogen and progesterone receptor expression In addition to phase trials, preclinical investi-
[188]. Another study concentrated on HDAC-6 gations have been widely done. The other idea
expression levels in breast cancer has been done by for treatment of ER-negative breast cancer cells
Zhang et al. They also found that HDAC-6 mRNA is using the synergistic effects of a combination
5 Epigenomics of Breast Cancer 117

Table 5.2 The investigations of HDAC inhibitors in breast cancer

Alternative Study
Agent(s) name Class design Samples Case # Reference
Vorinostat SAHA, Hydroxamic Preclinic Human breast Munster et al.
suberoylanilide acid cancer cells [190]
hydroxamic acid
Vorinostat Phase II Metastatic 14 Luu et al. [232]
breast cancer
Vorinostat + tamoxifen Phase II ER-positive 43 Munster et al.
metastatic [191]
breast cancer
Vorinostat + Phase I–II Metastatic 54 Ramaswamy
paclitaxel + bevacizumab breast cancer et al. [233]
Panobinostat LBH-589 Hydroxamic Preclinic Human breast Chen et al.
acid cancer cells [234]
Panobinostat Preclinic ER-negative Zhou et al.
human breast [194]
cancer cells
Panobinostat Preclinic Human breast Rao et al. [235]
cancer cells
Panobinostat Preclinic Triple-negative Tate et al.
breast cancer [236]
Entinostat MS-275, Benzamide Preclinic Human breast Lee et al. [237]
SNDX-275 cancer cells
Entinostat Preclinic Human breast Huang et al.
cancer cells [238]
Entinostat Preclinic ERα-negative Sabnis et al.
human breast [239]
cancer cells
Entinostat + trastuzumab Preclinic Human breast Huang et al.
cancer cells [120]
Romidepsin Depsipeptide Cyclic Preclinic Human breast Hirokawa et al.
(FK-228), peptide cancer cells [240]
Valproic acid – Aliphatic Preclinic Human breast Jawed et al.
acids cancer cells [241]
Valproic acid + tamoxifen Preclinic Human breast Hodges-
cancer cells Gallagher et al.
Valproic Preclinic Human breast Reid et al.
acid + trichostatin A cancer cells [243]
Valproic acid + retinoic Preclinic Human breast Mongan et al.
acid + cancer cells [244]
Phenylbutyrate – Aliphatic Preclinic Human breast Dyer et al.
acids cancer cells [245]

treatment of HDAC inhibitors and DNMT inhibi- effect. Both studies have shown the reactivate
tors (demethylating agents). Fan et al. and ERα and PR gene expression in ER-negative
Sharma et al. used 5-aza-2′-deoxycytidine (AZA) breast cancer cell lines, which are known to be
as a DNMT1 inhibitor and trichostatin A (TSA) aberrantly silenced in breast cancer [192, 193].
as a HDAC inhibitor to investigate this synergistic Other studies have shown that the HDAC
118 K.O. Yaykasli et al.

inhibitors lead to reactive of ERα and PR expres- References

sion by inhibition of the HDAC activity in breast
cancer cells [194–196]. 1. Jemal A, Bray F, Center MM, Ferlay J, Ward E,
Forman D. Global cancer statistics. CA Cancer J Clin.
The other enzyme families to target for cancer 2011;61(2):69–90.
therapy are HMTs and HDMs, previously impli- 2. DeVita Jr VT, Lawrence TS, Rosenberg SA, DePinho
cated in cancer, inflammation, and diabetes RA, Weinberg RA. CANCER: principles & practice
[197]. The gene expressions level of the histone- of oncology. Philadelphia: Lippincott, Williams &
Wilkins; 2011.
modifying enzymes (HDMs and HTMs) are spe- 3. You JS, Jones PA. Cancer genetics and epigenetics:
cific to cell types and highly correlated with two sides of the same coin? Cancer Cell. 2012;
target gene expression [198]. A recent study 22(1):9–20.
examined the expression profiles of 16 different 4. Lansdorp PM, Falconer E, Tao J, Brind’amour J,
Naumann U. Epigenetic differences between sister
histone-modifier genes including HATs, HDACs, chromatids? Ann N Y Acad Sci. 2012;1266(1):1–6.
and HDMs in breast cancer. They found that sig- 5. Dawson MA, Kouzarides T. Cancer epigenetics: from
nificantly different expression levels of histone- mechanism to therapy. Cell. 2012;150(1):12–27.
modifier genes exist between breast tumors and 6. Carone DM, Lawrence JB. Heterochromatin instabil-
ity in cancer: from the Barr body to satellites and the
normal tissue, and their findings were signifi- nuclear periphery. Semin Cancer Biol. 2013;23(2):
cantly associated with conventional pathological 99–108.
parameters and clinical outcomes. So, it appears 7. Barneda-Zahonero B, Parra M. Histone deacetylases
that histone-modifier enzymes offer utility as and cancer. Mol Oncol. 2012. doi:10.1016/j.
biomarkers and potential for targeted therapeutic 8. Christinat A, Pagani O. Fertility after breast cancer.
strategies [199]. Maturitas. 2012. doi:10.1016/j.
After these recent findings, miRNAs also have maturitas.2012.07.013.
become the target for developing therapies for 9. Ferlay J, Shin HR, Bray F, Forman D, Mathers C,
Parkin DM. Estimates of worldwide burden of cancer
breast cancer. The miRNA-based treatments, in in 2008: GLOBOCAN 2008. Int J Cancer.
combination with traditional chemotherapy, may 2010;127(12):2893–917.
be a new strategy for the clinical management of 10. DeSantis C, Siegel R, Bandi P, Jemal A. Breast cancer
drug-resistant breast cancers in the near future statistics. CA Cancer J Clin. 2011;61(6):409–18.
11. Liedtke C, Kiesel L. Breast cancer molecular sub-
[200]. One of the initial studies has concluded types—modern therapeutic concepts for targeted
that miRNA-221/222 confers breast cancer ful- therapy of a heterogeneous entity. Maturitas.
vestrant resistance by regulating multiple signal- 2012;73(4):288–94.
ing pathways [201]. 12. Huynh KT, Chong KK, Greenberg ES, Hoon
DS. Epigenetics of estrogen receptor-negative pri-
mary breast cancer. Expert Rev Mol Diagn.
Conclusion and Future Perspective 13. Meeran SM, Patel SN, Li Y, Shukla S, Tollefsbol
TO. Bioactive dietary supplements reactivate ER
expression in ER-negative breast cancer cells by
A new field has been opened to developing effec- active chromatin modifications. PLoS One.
tive clinical therapies now that we understand the 2012;7(5):e37748.
importance of epigenetic alterations. In contrast to 14. Ni M, Chen Y, Lim E, Wimberly H, Bailey ST, Imai
genetic code, the epigenetic codes may be easily Y, et al. Targeting androgen receptor in estrogen
receptor-negative breast cancer. Cancer Cell.
affected by aging, environmental stimuli, and food 2011;20(1):119–31.
in heritable manner. Breast cancer is a multifacto- 15. Hatae J, Takami N, Lin H, Honda A, Inoue R. 17beta-
rial disease with molecular, histological, and phe- Estradiol-induced enhancement of estrogen receptor
notypic diversity caused by the interaction of both biosynthesis via MAPK pathway in mouse skeletal
muscle myoblasts. J Physiol Sci. 2009;59(3):181–90.
inherited and environmental risk factors. The 16. Desmedt C, Voet T, Sotiriou C, Campbell PJ. Next-
importance of epigenomics for breast cancer generation sequencing in breast cancer: first take
development has been realized after gaining of home messages. Curr Opin Oncol. 2012;24(6):
great amount of knowledge by large-scale meth- 597–604.
17. Yang Z, Chevolot Y, Géhin T, Solassol J, Mange A,
ods. Epigenetics-based therapy for breast cancer Souteyrand E, et al. Improvement of protein immobi-
will most likely become a reality in the near future. lization for the elaboration of tumor-associated anti-
5 Epigenomics of Breast Cancer 119

gen microarrays: application to the sensitive and 38. Connolly R, Stearns V. Epigenetics as a therapeutic
specific detection of tumor markers from breast can- target in breast cancer. J Mammary Gland Biol
cer sera. Biosens Bioelectron. 2013;40(1):385–92. Neoplasia. 2012;17(3–4):191–204.
18. Vo AT, Millis RM. Epigenetics and breast cancers. 39. Sharma S, Kelly TK, Jones PA. Epigenetics in cancer.
Obstet Gynecol Int. 2012;2012:602720. Carcinogenesis. 2010;31(1):27–36.
19. Huang Y, Nayak S, Jankowitz R, Davidson NE, 40. Lister R, Pelizzola M, Dowen RH, Hawkins RD, Hon
Oesterreich S. Epigenetics in breast cancer: what’s G, Tonti-Filippini J, et al. Human DNA methylomes
new? Breast Cancer Res. 2011;13(6):225. at base resolution show widespread epigenomic dif-
20. Dedeurwaerder S, Fumagalli D, Fuks F. Unravelling ferences. Nature. 2009;462(7271):315–22.
the epigenomic dimension of breast cancers. Curr 41. Jones PA. DNA methylation and cancer. Cancer Res.
Opin Oncol. 2011;23(6):559–65. 1986;46(2):461–6.
21. Hassler MR, Egger G. Epigenomics of cancer— 42. Ehrlich M. DNA methylation in cancer: too much, but
emerging new concepts. Biochimie. 2012;94(11): also too little. Oncogene. 2002;21(35):5400–13.
2219–30. 43. Bird A, Taggart M, Frommer M, Miller OJ, Macleod
22. Gerhauser C. Cancer chemoprevention and D. A fraction of the mouse genome that is derived
nutri-epigenetics: state of the art and future chal- from islands of nonmethylated CpG-rich DNA. Cell.
lenges. Top Curr Chem. 2012;329:73–132. 1985;40(1):91–9.
23. Franco R, Schoneveld O, Georgakilas AG, 44. Goldberg AD, Allis CD, Bernstein E. Epigenetics: a
Panayiotidis MI. Oxidative stress, DNA methylation landscape takes shape. Cell. 2007;128(4):
and carcinogenesis. Cancer Lett. 2008;266(1):6–11. 635–8.
24. Hatziapostolou M, Iliopoulos D. Epigenetic aberra- 45. Bird AP. CpG-rich islands and the function of DNA
tions during oncogenesis. Cell Mol Life Sci. methylation. Nature. 1986;321(6067):209–13.
2011;68(10):1681–702. 46. Esteller M. CpG island hypermethylation and tumor
25. Gigek CO, Chen ES, Calcagno DQ, Wisnieski F, suppressor genes: a booming present, a brighter
Burbano RR, Smith MA. Epigenetic mechanisms in future. Oncogene. 2002;21(35):5427–40.
gastric cancer. Epigenomics. 2012;4(3):279–94. 47. Portela A, Esteller M. Epigenetic modifications
26. Catalano MG, Fortunati N, Boccuzzi G. Epigenetics and human disease. Nat Biotechnol. 2010;28(10):
modifications and therapeutic prospects in human 1057–68.
thyroid cancer. Front Endocrinol (Lausanne). 48. Nguyen C, Liang G, Nguyen TT, Tsao-Wei D,
2012;3:40. Groshen S, Lübbert M, et al. Susceptibility of nonpro-
27. Jerónimo C, Henrique R. Epigenetic biomarkers in moter CpG islands to de novo methylation in normal
urological tumors: a systematic review. Cancer Lett. and neoplastic cells. J Natl Cancer Inst. 2001;
2014;342(2):264–74. 93(19):1465–72.
28. Seeber LM, Van Diest PJ. Epigenetics in ovarian can- 49. Ndlovu MN, Denis H, Fuks F. Exposing the DNA
cer. Methods Mol Biol. 2012;863:253–69. methylome iceberg. Trends Biochem Sci.
29. Liloglou T, Bediaga NG, Brown BR, Field JK, Davies 2011;36(7):381–7.
MP. Epigenetic biomarkers in lung cancer. Cancer 50. Aran D, Toperoff G, Rosenberg M, Hellman
Lett. 2014;342(2):200–12. A. Replication timing-related and gene body-specific
30. Kim WJ, Kim YJ. Epigenetics of bladder cancer. methylation of active human genes. Hum Mol Genet.
Methods Mol Biol. 2012;863:111–8. 2011;20(4):670–80.
31. Khare S, Verma M. Epigenetics of colon cancer. 51. Li Y, Zhu J, Tian G, Li N, Li Q, Ye M, et al. The DNA
Methods Mol Biol. 2012;863:177–85. methylome of human peripheral blood mononuclear
32. Dubuc AM, Mack S, Unterberger A, Northcott PA, cells. PLoS Biol. 2010;8(11):e1000533.
Taylor MD. The epigenetics of brain tumors. Methods 52. Robertson KD. DNA methylation and chromatin—
Mol Biol. 2012;863:139–53. unraveling the tangled web. Oncogene.
33. Gabay O, Sanchez C. Epigenetics, sirtuins and osteo- 2002;21(35):5361–79.
arthritis. Joint Bone Spine. 2012;79(6): 53. Li E, Bestor TH, Jaenisch R. Targeted mutation of the
570–3. DNA methyltransferase gene results in embryonic
34. Udali S, Guarini P, Moruzzi S, Choi SW, Friso lethality. Cell. 1992;69(6):915–26.
S. Cardiovascular epigenetics: from DNA methyla- 54. Pedrali-Noy G, Weissbach A. Mammalian DNA
tion to microRNAs. Mol Aspects Med. 2012. methyltransferases prefer poly(dI-dC) as substrate. J
doi:10.1016/j.mam.2012.08.001. Biol Chem. 1986;261(17):7600–2.
35. Sandoval J, Esteller M. Cancer epigenomics: beyond 55. Okano M, Bell DW, Haber DA, Li E. DNA methyl-
genomics. Curr Opin Genet Dev. 2012;22(1):50–5. transferases Dnmt3a and Dnmt3b are essential for de
36. McPherson K, Steel CM, Dixon JM. ABC of breast novo methylation and mammalian development. Cell.
diseases; breast cancer-epidemiology, risk factors, 1999;99(3):247–57.
and genetics. BMJ. 2000;321(7261):624–8. 56. Cedar H, Bergman Y. Programming of DNA methyla-
37. Nowsheen S, Aziz K, Tran PT, Gorgoulis VG, Yang tion patterns. Annu Rev Biochem. 2012;81:97–117.
ES, Georgakilas AG. Epigenetic inactivation of DNA 57. Klose RJ, Bird AP. Genomic DNA methylation: the
repair in breast cancer. Cancer Lett. 2014;342(2): mark and its mediators. Trends Biochem Sci.
213–22. 2006;31(2):89–97.
120 K.O. Yaykasli et al.

58. Robertson KD. DNA methylation and human disease. associated with malignant biology. Am J Pathol.
Nat Rev Genet. 2005;6(8):597–610. 2011;179:55–65.
59. Ley TJ, Ding L, Walter MJ, McLellan MD, Lamprecht 74. Li L, Lee KM, Han W, Choi JY, Lee JY, Kang GH,
T, Larson DE, et al. DNMT3A mutations in acute et al. Estrogen and progesterone receptor status affect
myeloid leukemia. N Engl J Med. 2010;363(25): genome-wide DNA methylation profile in breast can-
2424–33. cer. Hum Mol Genet. 2010;19:4273–7.
60. Feinberg AP, Vogelstein B. Hypomethylation 75. Ronneberg JA, Fleischer T, Solvang HK, Nordgard
distinguishes genes of some human cancers from SH, Edvardsen H, Potapenko I, et al. Methylation pro-
their normal counterparts. Nature. 1983;301(5895): filing with a panel of cancer related genes: association
89–92. with estrogen receptor, TP53 mutation status and
61. Arasaradnam RP, Commane DM, Bradburn D, expression subtypes in sporadic breast cancer. Mol
Mathers JC. A review of dietary factors and its influ- Oncol. 2011;5:61–76.
ence on DNA methylation in colorectal carcinogene- 76. Sun Z, Asmann YW, Kalari KR, Bot B, Eckel-Passow
sis. Epigenetics. 2008;3(4):193–8. JE, Baker TR, et al. Integrated analysis of gene
62. Herman JG, Baylin SB. Gene silencing in cancer in expression, CpG island methylation, and gene copy
association with promoter hypermethylation. N Engl J number in breast cancer cells by deep sequencing.
Med. 2003;349(21):2042–54. PLoS One. 2011;6:e17490.
63. Jones PA, Baylin SB. The epigenomics of cancer. 77. Di Leva G, Gasparini P, Piovan C, Ngankeu A,
Cell. 2007;128(4):683–92. Garofalo M, Taccioli C, et al. MicroRNA cluster 221-
64. Baylin SB, Jones PA. A decade of exploring the can- 222 and estrogen receptor alpha interactions in breast
cer epigenome—biological and translational implica- cancer. J Natl Cancer Inst. 2010;102(10):706–21.
tions. Nat Rev Cancer. 2011;11(10):726–34. 78. Lapidus RG, Nass SJ, Butash KA, Parl FF, Weitzman
65. Kulis M, Esteller M. DNA methylation and cancer. SA, Graff JG, et al. Mapping of ER gene CpG island
Adv Genet. 2010;70:27–56. methylation-specific polymerase chain reaction.
66. Su LJ. Diet, epigenetics, and cancer. Methods Mol Cancer Res. 1998;58(12):2515–9.
Biol. 2012;863:377–93. 79. Ottaviano YL, Issa JP, Parl FF, Smith HS, Baylin SB,
67. Fraga MF, Herranz M, Espada J, Ballestar E, Paz MF, Davidson NE. Methylation of the estrogen receptor
Ropero S, et al. A mouse skin multistage carcinogen- gene CpG island marks loss of estrogen receptor
esis model reflects the aberrant DNA methylation pat- expression in human breast cancer cells. Cancer Res.
terns of human tumors. Cancer Res. 2004;64(16): 1994;54(10):2552–5.
5527–34. 80. Ferguson AT, Lapidus RG, Baylin SB, Davidson
68. Christensen BC, Kelsey KT, Zheng S, Houseman EA, NE. Demethylation of the estrogen receptor gene in
Marsit CJ, Wrensch MR, et al. Breast cancer DNA estrogen receptor-negative breast cancer cells can
methylation profiles are associated with tumor size reactivate estrogen receptor gene expression. Cancer
and alcohol and folate intake. PLoS Genet. Res. 1995;55(11):2279–83.
2010;6(7):e1001043. 81. Yan L, Nass SJ, Smith D, Nelson WG, Herman JG,
69. Hill VK, Ricketts C, Bieche I, Vacher S, Gentle D, Davidson NE. Specific inhibition of DNMT1 by anti-
Lewis C, et al. Genome-wide DNA methylation pro- sense oligonucleotides induces re-expression of estro-
filing of CpG islands in breast cancer identifies novel gen receptor-alpha (ER) in ER-negative human breast
genes associated with tumorigenicity. Cancer Res. cancer cell lines. Cancer Biol Ther. 2003;2(5):552–6.
2011;71:2988–99. 82. Shi JF, Li XJ, Si XX, Li AD, Ding HJ, Han X, et al.
70. Faryna M, Konermann C, Aulmann S, Bermejo JL, ERα positively regulated DNMT1 expression by
Brugger M, Diederichs S, et al. Genome-wide meth- binding to the gene promoter region in human breast
ylation screen in low-grade breast cancer identifies cancer MCF-7 cells. Biochem Biophys Res Commun.
novel epigenetically altered genes as potential bio- 2012;427(1):47–53.
markers for tumor diagnosis. FASEB. 83. Jackson SP, Bartek J. The DNA-damage response in
2012;J26(12):4937–50. human biology and disease. Nature.
71. Botla SK, Moghaddas Gholami A, Malekpour M, 2009;461(7267):1071–8.
Moskalev EA, Fallah M, Jandaghi P, et al. Diagnostic 84. Bartek J, Lukas C, Lukas J. Checking on DNA dam-
values of GHSR DNA methylation pattern in breast age in S phase. Nat Rev Mol Cell Biol.
cancer. Breast Cancer Res Treat. 2004;5(10):792–804.
2012;135(3):705–13. 85. Wang H, Yang ES, Jiang J, Nowsheen S, Xia F. DNA
72. Holm K, Hegardt C, Staaf J, Vallon-Christersson J, damage-induced cytotoxicity is dissociated from
Jönsson G, Olsson H, et al. Molecular subtypes of BRCA1’s DNA repair function but is dependent on its
breast cancer are associated with characteristic DNA cytosolic accumulation. Cancer Res.
methylation patterns. Breast Cancer Res. 2010;70(15):6258–67.
2010;12:R36. 86. Jiang J, Yang ES, Jiang G, Nowsheen S, Wang H,
73. Killian JK, Bilke S, Davis S, Walker RL, Jaeger E, Wang T, et al. p53-dependent BRCA1 nuclear export
Killian MS, et al. A methyl-deviator epigenotype controls cellular susceptibility to DNA damage.
of estrogen receptor-positive breast carcinoma is Cancer Res. 2011;71(16):5546–57.
5 Epigenomics of Breast Cancer 121

87. Feng Z, Kachnic L, Zhang J, Powell SN, Xia analysis by multiplex ligation-dependent probe
F. DNA damage induces p53-dependent BRCA1 amplification. Breast Cancer Res. 2012;14(4):R101.
nuclear export. J Biol Chem. 2004;279(27): 100. Luger K, Mäder AW, Richmond RK, Sargent DF,
28574–84. Richmond TJ. Crystal structure of the nucleo-
88. King MC, Marks JH, Mandell JB, New York Breast some core particle t 2.8 A resolution. Nature.
Cancer Study Group. Breast and ovarian cancer risks 1997;389(6648):251–60.
due to inherited mutations in BRCA1 and BRCA2. 101. Kouzarides T. Chromatin modifications and their
Science. 2003;302(5645):643–6. function. Cell. 2007;128(4):693–705.
89. Graeser MK, Engel C, Rhiem K, Gadzicki D, Bick 102. Cedar H, Bergman Y. Linking DNA methylation and
U, Kast K, et al. Contralateral breast cancer risk in histone modification: patterns and paradigms. Nat
BRCA1 and BRCA2 mutation carriers. J Clin Oncol. Rev Genet. 2009;10(5):295–304.
2009;27(35):5887–92. 103. Jenuwein T, Allis CD. Translating the histone code.
90. Esteller M, Silva JM, Dominguez G, Bonilla F, Science. 2001;293(5532):1074–80.
Matias-Guiu X, Lerma E, et al. Promoter hyper- 104. Allfrey VG, Faulkner R, Mirsky AE. Acetylation
methylation and BRCA1 inactivation in sporadic and methylation of histones and their possible role in
breast and ovarian tumors. J Natl Cancer Inst. the regulation of RNA synthesis. Proc Natl Acad Sci
2000;92(7):564–9. U S A. 1964;51:786–94.
91. Marsit CJ, Liu M, Nelson HH, Posner M, Suzuki M, 105. Strahl BD, Allis CD. The language of covalent his-
Kelsey KT. Inactivation of the Fanconi anemia/ tone modifications. Nature. 2000;403(6765):41–5.
BRCA pathway in lung and oral cancers: implica- 106. Lee JS, Smith E, Shilatifard A. The language of his-
tions for treatment and survival. Oncogene. tone crosstalk. Cell. 2010;142(5):682–5.
2004;23(4):1000–4. 107. Zippo A, Serafini R, Rocchigiani M, Pennacchini S,
92. Girault I, Tozlu S, Lidereau R, Bièche I. Expression Krepelova A, Oliviero S. Histone crosstalk between
analysis of DNA methyltransferases 1, 3A, and 3B in H3S10ph and H4K16ac generates a histone code
sporadic breast carcinomas. Clin Cancer Res. that mediates transcription elongation. Cell.
2003;9(12):4415–22. 2009;138(6):1122–36.
93. Butcher DT, Rodenhiser DI. Epigenetic inactivation 108. Wang Z, Zang C, Cui K, Schones DE, Barski A,
of BRCA1 is associated with aberrant expression of Peng W, et al. Genome-wide mapping of HATs and
CTCF and DNA methyltransferase (DNMT3B) in HDACs reveals distinct functions in active and inac-
some sporadic breast tumours. Eur J Cancer. tive genes. Cell. 2009;138(5):1019–31.
2007;43(1):210–9. 109. Wolffe AP, Hayes JJ. Chromatin disruption and
94. Ben Gacem R, Hachana M, Ziadi S, Ben Abdelkarim modification. Nucleic Acids Res. 1999;27(3):
S, Hidar S, Trimeche M. Clinicopathologic signifi- 711–20.
cance of DNA methyltransferase 1, 3a, and 3b over- 110. Gardner KE, Allis CD, Strahl BD. Operating on
expression in Tunisian breast cancers. Hum Pathol. chromatin, a colorful language where context mat-
2012;43(10):1731–8. ters. J Mol Biol. 2011;409(1):36–46.
95. Flotho C, Claus R, Batz C, Schneider M, Sandrock 111. Sawan C, Herceg Z. Histone modifications and can-
I, Ihde S, et al. The DNA methyltransferase cer. Adv Genet. 2010;70:57–85.
inhibitors azacitidine, decitabine and zebularine 112. Füllgrabe J, Kavanagh E, Joseph B. Histone onco-
exert differential effects on cancer gene expres- modifications. Oncogene. 2011;30(31):3391–403.
sion in acute myeloid leukemia cells. Leukemia. 113. Munshi A, Shafi G, Aliya N, Jyothy A. Histone
2009;23(6):1019–28. modifications dictate specific biological readouts. J
96. Billam M, Sobolewski MD, Davidson NE. Effects of Genet Genomics. 2009;36(2):75–88.
a novel DNA methyltransferase inhibitor zebularine 114. Fu S, Kurzrock R. Development of curcumin as an
on human breast cancer cells. Breast Cancer Res epigenetic agent. Cancer. 2010;116(20):4670–6.
Treat. 2010;120(3):581–92. 115. Bannister AJ, Kouzarides T. The CBP co-activator is
97. Sharma G, Mirza S, Parshad R, Srivastava A, Datta a histone acetyltransferase. Nature. 1996;384(6610):
Gupta S, Pandya P, et al. CpG hypomethylation of 641–3.
MDR1 gene in tumor and serum of invasive ductal 116. Hodawadekar SC, Marmorstein R. Chemistry of
breast carcinoma patients. Clin Biochem. acetyl transfer by histone modifying enzymes: struc-
2010;43(4–5):373–9. ture, mechanism and implications for effector
98. Hon GC, Hawkins RD, Caballero OL, Lo C, Lister design. Oncogene. 2007;26(37):5528–40.
R, Pelizzola M, et al. Global DNA hypomethylation 117. Bannister AJ, Kouzarides T. Regulation of chroma-
coupled to repressive chromatin domain formation tin by histone modifications. Cell Res. 2011;21(3):
and gene silencing in breast cancer. Genome Res. 381–95.
2012;22(2):246–58. 118. Arts J, de Schepper S, Van Emelen K. Histone
99. Kornegoor R, Moelans CB, Verschuur-Maes AH, deacetylase inhibitors: from chromatin remodeling
Hogenes MC, de Bruin PC, Oudejans JJ, et al. to experimental cancer therapeutics. Curr Med
Promoter hypermethylation in male breast cancer: Chem. 2003;22:2343–50.
122 K.O. Yaykasli et al.

119. Prince HM, Bishton MJ, Harrison SJ. Clinical stud- 134. Nikoloski G, Langemeijer SM, Kuiper RP, Knops
ies of histone deacetylase inhibitors. Clin Cancer R, Massop M, Tönnissen ER, et al. Somatic
Res. 2009;15(12):3958–69. mutations of the histone methyltransferase gene
120. Huang X, Wang S, Lee CK, Yang X, Liu B. HDAC EZH2 in myelodysplastic syndromes. Nat Genet.
inhibitor SNDX-275 enhances efficacy of trastu- 2010;42(8):665–7.
zumab in erbB2-overexpressing breast cancer cells 135. Zhang J, Ding L, Holmfeldt L, Wu G, Heatley SL,
and exhibits potential to overcome trastuzumab Payne-Turner D, et al. The genetic basis of early
resistance. Cancer Lett. 2011;307(1):72–9. T-cell precursor acute lymphoblastic leukaemia.
121. Campagna-Slater V, Mok MW, Nguyen KT, Nature. 2012;481(7380):157–63.
Feher M, Najmanovich R, Schapira M. Structural 136. Doherty LF, Bromer JG, Zhou Y, Aldad TS, Taylor
chemistry of the histone methyltransferases cofactor HS. In utero exposure to diethylstilbestrol (DES) or
binding site. J Chem Inf Model. 2011;51(3):612–23. bisphenol-A (BPA) increases EZH2 expression in
122. Lee YH, Stallcup MR. Minireview: protein arginine the mammary gland: an epigenetic mechanism link-
methylation of nonhistone proteins in transcriptional ing endocrine disruptors to breast cancer. Horm
regulation. Mol Endocrinol. 2009;23(4):425–33. Cancer. 2010;1(3):146–55.
123. Barski A, Cuddapah S, Cui K, Roh TY, Schones 137. Gonzalez ME, DuPrie ML, Krueger H, Merajver
DE, Wang Z, et al. High-resolution profiling of SD, Ventura AC, Toy KA, et al. Histone methyl-
histone methylations in the human genome. Cell. transferase EZH2 induces Akt-dependent genomic
2007;129(4):823–37. instability and BRCA1 inhibition in breast cancer.
124. Mosammaparast N, Shi Y. (Reversal of histone Cancer Res. 2011;71(6):2360–70.
methylation: biochemical and molecular mecha- 138. Dong C, Wu Y, Yao J, Wang Y, Yu Y, Rychahou
nisms of histone demethylases. Annu Rev Biochem. PG, et al. G9a interacts with Snail and is criti-
2010;79:155–79. cal for Snail-mediated E-cadherin repression in
125. Varier RA, Timmers HT. Histone lysine methylation human breast cancer. J Clin Invest. 2012;122(4):
and demethylation pathways in cancer. Biochim 1469–86.
Biophys Acta. 2011;1815(1):75–89. 139. Purcell DJ, Jeong KW, Bittencourt D, Gerke DS,
126. Tsang DP, Cheng AS. Epigenetic regulation of sig- Stallcup MR. A distinct mechanism for coactivator
naling pathways in cancer: role of the histone meth- versus corepressor function by histone methyltrans-
yltransferase EZH2. J Gastroenterol Hepatol. ferase G9a in transcriptional regulation. J Biol
2011;26(1):19–27. Chem. 2011;286(49):41963–71.
127. Cao R, Wang L, Wang H, Xia L, Erdjument- 140. Zhou Z, Thomsen R, Kahns S, Nielsen AL. The
Bromage H, Tempst P, et al. Role of histone H3 NSD3L histone methyltransferase regulates cell
lysine 27 methylation in Polycomb-group silencing. cycle and cell invasion in breast cancer cells. Biochem
Science. 2002;298(5595):1039–43. Biophys Res Commun. 2010;398(3):565–70.
128. Simon JA, Lange CA. Roles of the EZH2 histone 141. Bedford MT. Arginine methylation at a glance.
methyltransferase in cancer epigenetics. Mutat Res. J Cell Sci. 2007;120(Pt 24):4243–6.
2008;647(1–2):21–9. 142. Le Romancer M, Treilleux I, Leconte N, Robin-
129. Varambally S, Dhanasekaran SM, Zhou M, Barrette Lespinasse Y, Sentis S, Bouchekioua-Bouzaghou K,
TR, Kumar-Sinha C, Sanda MG, et al. The poly- et al. Regulation of estrogen rapid signaling through
comb group protein EZH2 is involved in progression arginine methylation by PRMT1. Mol Cell.
of prostate cancer. Nature. 2002;419(6907): 2008;31(2):212–21.
624–9. 143. Shi Y, Lan F, Matson C, Mulligan P, Whetstine JR,
130. Bachmann IM, Halvorsen OJ, Collett K, Stefansson Cole PA, et al. Histone demethylation mediated by
IM, Straume O, Haukaas SA, et al. EZH2 expression the nuclear amine oxidase homolog LSD1. Cell.
is associated with high proliferation rate and aggres- 2004;119(7):941–53.
sive tumor subgroups in cutaneous melanoma and 144. Hayami S, Kelly JD, Cho HS, Yoshimatsu M, Unoki
cancers of the endometrium, prostate, and breast. J M, Tsunoda T, et al. Overexpression of LSD1 con-
Clin Oncol. 2006;24(2):268–73. tributes to human carcinogenesis through chromatin
131. Kleer CG, Cao Q, Varambally S, Shen R, Ota I, regulation in various cancers. Int J Cancer.
Tomlins SA, et al. EZH2 is a marker of aggressive 2011;128(3):574–86.
breast cancer and promotes neoplastic transforma- 145. Rotili D, Mai A. Targeting histone demethylases: a
tion of breast epithelial cells. Proc Natl Acad Sci U new avenue for the fight against cancer. Genes
S A. 2003;100(20):11606–11. Cancer. 2011;2(6):663–79.
132. Croonquist PA, Van Ness B. The polycomb group 146. Lim S, Janzer A, Becker A, Zimmer A, Schüle R,
protein enhancer of zeste homolog 2 (EZH 2) Buettner R, et al. Lysine-specific demethylase 1
is an oncogene that influences myeloma cell (LSD1) is highly expressed in ER-negative breast
growth and the mutant ras phenotype. Oncogene. cancers and a biomarker predicting aggressive biol-
2005;24(41):6269–80. ogy. Carcinogenesis. 2010;31(3):512–20.
133. Chase A, Cross NC. Aberrations of EZH2 in cancer. 147. Wang Y, Zhang H, Chen Y, Sun Y, Yang F, Yu W,
Clin Cancer Res. 2011;17(9):2613–8. et al. LSD1 is a subunit of the NuRD complex and
5 Epigenomics of Breast Cancer 123

targets the metastasis programs in breast cancer. 164. Iorio MV, Ferracin M, Liu CG, Veronese A, Spizzo
Cell. 2009;138(4):660–72. R, Sabbioni S, et al. MicroRNA gene expression
148. Liu G, Bollig-Fischer A, Kreike B, van de Vijver deregulation in human breast cancer. Cancer Res.
MJ, Abrams J, Ethier SP, et al. Genomic amplifica- 2005;65(16):7065–70.
tion and oncogenic properties of the GASC1 histone 165. O’Day E, Lal A. MicroRNAs and their target gene
demethylase gene in breast cancer. Oncogene. networks in breast cancer. Breast Cancer Res.
2009;28(50):4491–500. 2010;12(2):201.
149. Yamane K, Tateishi K, Klose RJ, Fang J, Fabrizio 166. Asaga S, Kuo C, Nguyen T, Terpenning M, Giuliano
LA, Erdjument-Bromage H, et al. PLU-1 is an H3K4 AE, Hoon DS. Direct serum assay for microRNA-21
demethylase involved in transcriptional repression concentrations in early and advanced breast cancer.
and breast cancer cell proliferation. Mol Cell. Clin Chem. 2011;57(1):84–91.
2007;25(6):801–12. 167. Frankel LB, Christoffersen NR, Jacobsen A, Lindow
150. Shi L, Sun L, Li Q, Liang J, Yu W, Yi X, et al. M, Krogh A, Lund AH. Programmed cell death 4
Histone demethylase JMJD2B coordinates H3K4/ (PDCD4) is an important functional target of the
H3K9 methylation and promotes hormonally microRNA miR-21 in breast cancer cells. J Biol
responsive breast carcinogenesis. Proc Natl Acad Chem. 2008;283(2):1026–33.
Sci U S A. 2011;108(18):7541–6. 168. Peto R, Davies C, Godwin J, Gray R, Pan HC, Clarke
151. Kawazu M, Saso K, Tong KI, McQuire T, Goto K, M, et al. Comparisons between different polychemo-
Son DO, et al. Histone demethylase JMJD2B func- therapy regimens for early breast cancer: meta-
tions as a co-factor of estrogen receptor in breast analyses of long-term outcome among 100,000
cancer proliferation and mammary gland develop- women in 123 randomised trials. Lancet.
ment. PLoS One. 2011;6(3):e17830. 2012;379:432–44.
152. Huang J, Sengupta R, Espejo AB, Lee MG, Dorsey 169. Visvanathan K, Sukumar S, Davidson NE. Epigenetic
JA, Richter M, et al. p53 is regulated by the lysine biomarkers and breast cancer: cause for optimism.
demethylase LSD1. Nature. 2007;449(7158):105–8. Clin Cancer Res. 2006;12(22):6591–3.
153. Wang J, Hevi S, Kurash JK, Lei H, Gay F, Bajko J, 170. Nass SJ, Herman JG, Gabrielson E, Iversen PW, Parl
et al. The lysine demethylase LSD1 (KDM1) is FF, Davidson NE, et al. Aberrant methylation of the
required for maintenance of global DNA methyla- estrogen receptor and E-cadherin 5′ CpG islands
tion. Nat Genet. 2009;41(1):125–9. increases with malignant progression in human
154. Guil S, Esteller M. DNA methylomes, histone codes breast cancer. Cancer Res. 2000;60(16):4346–8.
and miRNAs: tying it all together. Int J Biochem 171. Adams BD, Furneaux H, White BA. The micro-
Cell Biol. 2009;41(1):87–95. ribonucleic acid (miRNA) miR-206 targets the
155. Calin GA, Croce CM. MicroRNA signatures in human estrogen receptor-alpha (ERalpha) and
human cancers. Nat Rev Cancer. 2006;6(11):857–66. represses ERalpha messenger RNA and protein
156. Meltzer PS. Cancer genomics: small RNAs with big expression in breast cancer cell lines. Mol
impacts. Nature. 2005;435(7043):745–6. Endocrinol. 2007;21(5):1132–47.
157. Winter J, Jung S, Keller S, Gregory RI, Diederichs 172. Leivonen SK, Mäkelä R, Ostling P, Kohonen P,
S. Many roads to maturity: microRNA biogenesis Haapa-Paananen S, Kleivi K, et al. Protein lysate
pathways and their regulation. Nat Cell Biol. microarray analysis to identify microRNAs regulat-
2009;11(3):228–34. ing estrogen receptor signaling in breast cancer cell
158. Chen H, Hardy TM, Tollefsbol TO. Epigenomics of lines. Oncogenen. 2009;28(44):3926–36.
ovarian cancer and its chemoprevention. Front 173. Mirza S, Sharma G, Prasad CP, Parshad R, Srivastava
Genet. 2011;2:67. A, Gupta SD, et al. Promoter hypermethylation of
159. Brait M, Sidransky D. Cancer epigenetics: above and TMS1, BRCA1, ERalpha and PRB in serum and
beyond. Toxicol Mech Methods. 2011;21(4):275–88. tumor DNA of invasive ductal breast carcinoma
160. Lu J, Getz G, Miska EA, Alvarez-Saavedra E, Lamb patients. Life Sci. 2007;81(4):280–7.
J, Peck D, et al. MicroRNA expression profiles clas- 174. Heyn H, Engelmann M, Schreek S, Ahrens P,
sify human cancers. Nature. 2005;435(7043):834–8. Lehmann U, Kreipe H, et al. MicroRNA miR-
161. Zhang B, Pan X, Cobb GP, Anderson TA. microR- 335 is crucial for the BRCA1 regulatory cas-
NAs as oncogenes and tumor suppressors. Dev Biol. cade in breast cancer development. Int J Cancer.
2007;302(1):1–12. 2011;129(12):2797–806.
162. Farazi TA, Horlings HM, Ten Hoeve JJ, Mihailovic 175. Donninger H, Vos MD, Clark GJ. The RASSF1A
A, Halfwerk H, Morozov P, et al. MicroRNA tumor suppressor. J Cell Sci. 2007;120
sequence and expression analysis in breast tumors (Pt 18):3163–72.
by deep sequencing. Cancer Res. 2011;71(13): 176. Hesson LB, Cooper WN, Latif F. The role of
4443–53. RASSF1A methylation in cancer. Dis Markers.
163. Davoren PA, McNeill RE, Lowery AJ, Kerin MJ, 2007;23(1–2):73–87.
Miller N. Identification of suitable endogenous con- 177. Sebova K, Zmetakova I, Bella V, Kajo K,
trol genes for microRNA gene expression analysis in Stankovicova I, Kajabova V, et al. Cancer Biomark.
human breast cancer. BMC Mol Biol. 2008;9:76. 2011;10(1):13–26.
124 K.O. Yaykasli et al.

178. Jiang Y, Cui L, Chen WD, Shen SH, Ding LD. The therapy-resistant breast cancer. Br J Cancer.
prognostic role of RASSF1A promoter methylation 2011;104(12):1828–35.
in breast cancer: a meta-analysis of published data. 192. Sharma D, Saxena NK, Davidson NE, Vertino
PLoS One. 2012;7(5):e36780. PM. Restoration of tamoxifen sensitivity in estrogen
179. Fujii S, Ito K, Ito Y, Ochiai A. Enhancer of zeste receptor-negative breast cancer cells: tamoxifen-
homologue 2 (EZH2) down-regulates RUNX3 by bound reactivated ER recruits distinctive corepressor
increasing histone H3 methylation. J Biol Chem. complexes. Cancer Res. 2006;66(12):6370–8.
2008;283(25):17324–32. 193. Fan J, Yin WJ, Lu JS, Wang L, Wu J, Wu FY, et al.
180. Collett K, Eide GE, Arnes J, Stefansson IM, Eide J, ER alpha negative breast cancer cells restore
Braaten A, et al. Expression of enhancer of zeste response to endocrine therapy by combination treat-
homologue 2 is significantly associated with ment with both HDAC inhibitor and DNMT inhibi-
increased tumor cell proliferation and is a marker of tor. J Cancer Res Clin Oncol. 2008;134(8):
aggressive breast cancer. Clin Cancer Res. 883–90.
2006;12(4):1168–74. 194. Zhou Q, Atadja P, Davidson NE. Histone deacety-
181. Pal B, Bouras T, Shi W, Vaillant F, Sheridan JM, Fu lase inhibitor LBH589 reactivates silenced estrogen
N, et al. Global changes in the mammary epigenome receptor alpha (ER) gene expression without loss of
are induced by hormonal cues and coordinated by DNA hypermethylation. Cancer Biol Ther.
Ezh2. Cell Rep. 2013. doi:10.1016/j.celrep.2012. 2007;6(1):64–9.
12.020. 195. Keen JC, Yan L, Mack KM, Pettit C, Smith D,
182. Hossain A, Kuo MT, Saunders GF. Mir-17-5p regu- Sharma D, et al. A novel histone deacetylase inhibi-
lates breast cancer cell proliferation by inhibiting tor, scriptaid, enhances expression of functional
translation of AIB1 mRNA. Mol Cell Biol. estrogen receptor alpha (ER) in ER negative human
2006;26(21):8191–201. breast cancer cells in combination with 5-aza
183. Yu Z, Wang C, Wang M, Li Z, Casimiro MC, Liu M, 2′-deoxycytidine. Breast Cancer Res Treat.
et al. A cyclin D1/microRNA 17/20 regulatory feed- 2003;81(3):177–86.
back loop in control of breast cancer cell prolifera- 196. Yang X, Ferguson AT, Nass SJ, Phillips DL, Butash
tion. J Cell Biol. 2008;182(3):509–17. KA, Wang SM, et al. Transcriptional activation of
184. Yan LX, Huang XF, Shao Q, Huang MY, Deng L, estrogen receptor alpha in human breast cancer cells
Wu QL, et al. MicroRNA miR-21 overexpression in by histone deacetylase inhibition. Cancer Res.
human breast cancer is associated with advanced 2000;60(24):6890–4.
clinical stage, lymph node metastasis and patient 197. Gauthier N, Caron M, Pedro L, Arcand M, Blouin J,
poor prognosis. RNA. 2008;14(11):2348–60. Labonté A, et al. Development of homogeneous
185. Morita S, Takahashi RU, Yamashita R, Toyoda A, nonradioactive methyltransferase and demethylase
Horii T, Kimura M, et al. Genome-wide analysis of assays targeting histone H3 lysine 4. J Biomol
DNA methylation and expression of microRNAs in Screen. 2012;17(1):49–58.
breast cancer cells. Int J Mol Sci. 2012;13(7): 198. Islam AB, Richter WF, Jacobs LA, Lopez-
8259–72. Bigas N, Benevolenskaya EV. Co-regulation of
186. Laird PW. Cancer epigenetics. Hum Mol Genet. histone-modifying enzymes in cancer. PLoS One.
2005;14(Spec No 1):R65–76. 2011;6(8):e24023.
187. Stearns V, Zhou Q, Davidson NE. Epigenetic regula- 199. Patani N, Jiang WG, Newbold RF, Mokbel
tion as a new target for breast cancer therapy. Cancer K. Histone-modifier gene expression profiles are
Invest. 2007;8:659–65. associated with pathological and clinical out-
188. Krusche CA, Wülfing P, Kersting C, Vloet A, Böcker comes in human breast cancer. Anticancer Res.
W, Kiesel L, et al. Histone deacetylase-1 and -3 pro- 2011;31(12):4115–25.
tein expression in human breast cancer: a tissue 200. Kutanzi KR, Yurchenko OV, Beland FA, Checkhun
microarray analysis. Breast Cancer Res Treat. VF, Pogribny IP. MicroRNA-mediated drug resis-
2005;90(1):15–23. tance in breast cancer. Clin Epigenetics.
189. Zhang Z, Yamashita H, Toyama T, Sugiura H, 2011;2(2):171–85.
Omoto Y, Ando Y, et al. HDAC6 expression is cor- 201. Rao X, Di Leva G, Li M, Fang F, Devlin C, Hartman-
related with better survival in breast cancer. Clin Frey C, et al. MicroRNA-221/222 confers breast
Cancer Res. 2004;10(20):6962–8. cancer fulvestrant resistance by regulating multiple
190. Munster PN, Troso-Sandoval T, Rosen N, Rifkind R, signaling pathways. Oncogene. 2011;30(9):1082–97.
Marks PA, Richon VM. The histone deacetylase 202. Ferguson AT, Evron E, Umbricht CB, Pandita TK,
inhibitor suberoylanilide hydroxamic acid induces Chan TA, Hermeking H, et al. High frequency of
differentiation of human breast cancer cells. Cancer hypermethylation at the 14-3-3 sigma locus leads to
Res. 2001;61(23):8492–7. gene silencing in breast cancer. Proc Natl Acad Sci
191. Munster PN, Thurn KT, Thomas S, Raha P, Lacevic U S A. 2000;97(11):6049–54.
M, Miller A, et al. A phase II study of the histone 203. Mirza S, Sharma G, Parshad R, Srivastava A, Gupta
deacetylase inhibitor vorinostat combined with SD, Ralhan R. Clinical significance of Stratifin,
tamoxifen for the treatment of patients with hormone ERalpha and PR promoter methylation in tumor and
5 Epigenomics of Breast Cancer 125

serum DNA in Indian breast cancer patients. Clin 215. Shinozaki M, Hoon DS, Giuliano AE, Hansen NM,
Biochem. 2010;43(4–5):380–6. Wang HJ, Turner R, et al. Distinct hypermethylation
204. Martínez-Galán J, Torres B, Del Moral R, Muñoz- profile of primary breast cancer is associated with
Gámez JA, Martín-Oliva D, Villalobos M, et al. sentinel lymph node metastasis. Clin Cancer Res.
Quantitative detection of methylated ESR1 and 2005;11(6):2156–62.
14-3-3-sigma gene promoters in serum as candidate 216. Caldeira JR, Prando EC, Quevedo FC, Neto FA,
biomarkers for diagnosis of breast cancer and evalu- Rainho CA, Rogatto SR. CDH1 promoter hyper-
ation of treatment efficacy. Cancer Biol Ther. methylation and E-cadherin protein expression in
2008;7(6):958–65. infiltrating breast cancer. BMC Cancer. 2006;6:48.
205. Dammann R, Yang G, Pfeifer GP. Hypermethylation 217. Silva JM, Dominguez G, Villanueva MJ, Gonzalez
of the cpG island of Ras association domain family R, Garcia JM, Corbacho C, et al. Aberrant DNA
1A (RASSF1A), a putative tumor suppressor gene methylation of the p16INK4a gene in plasma DNA
from the 3p21.3 locus, occurs in a large percentage of breast cancer patients. Br J Cancer.
of human breast cancers. Cancer Res. 1999;80(8):1262–4.
2001;61(7):3105–9. 218. Hu XC, Wong IH, Chow LW. Tumor-derived aber-
206. Jin Z, Tamura G, Tsuchiya T, Sakata K, Kashiwaba rant methylation in plasma of invasive ductal breast
M, Osakabe M, et al. Adenomatous polyposis coli cancer patients: clinical implications. Oncol Rep.
(APC) gene promoter hypermethylation in primary 2003;10(6):1811–5.
breast cancers. Br J Cancer. 2001;85(1):69–73. 219. Toyooka KO, Toyooka S, Virmani AK,
207. Dulaimi E, Hillinck J, Ibanez de Caceres I, Sathyanarayana UG, Euhus DM, Gilcrease M, et al.
Al-Saleem T, Cairns P. Tumor suppressor gene pro- Loss of expression and aberrant methylation of the
moter hypermethylation in serum of breast cancer CDH13 (H-cadherin) gene in breast and lung carci-
patients. Clin Cancer Res. 2004;10(18 Pt nomas. Cancer Res. 2001;61(11):4556–60.
1):6189–93. 220. Birgisdottir V, Stefansson OA, Bodvarsdottir SK,
208. Sirchia SM, Ferguson AT, Sironi E, Subramanyan S, Hilmarsdottir H, Jonasson JG, Eyfjord JE. Epigenetic
Orlandi R, Sukumar S, et al. Evidence of epigenetic silencing and deletion of the BRCA1 gene in spo-
changes affecting the chromatin state of the retinoic radic breast cancer. Breast Cancer Res.
acid receptor beta2 promoter in breast cancer cells. 2006;8(4):R38.
Oncogene. 2000;19(12):1556–63. 221. Jing F, Zhang J, Tao J, Zhou Y, Jun L, Tang X, et al.
209. Shukla S, Mirza S, Sharma G, Parshad R, Gupta SD, Hypermethylation of tumor suppressor genes
Ralhan R. Detection of RASSF1A and RARbeta BRCA1, p16 and 14-3-3sigma in serum of sporadic
hypermethylation in serum DNA from breast cancer breast cancer patients. Oncology.
patients. Epigenetics. 2006;1(2):88–93. 2007;30(1–2):14–9.
210. Papadopoulou E, Davilas E, Sotiriou V, 222. Van der Auwera I, Elst HJ, Van Laere SJ, Maes H,
Georgakopoulos E, Georgakopoulou S, Koliopanos Huget P, van Dam P, et al. The presence of circulat-
A, et al. Cell-free DNA and RNA in plasma as a new ing total DNA and methylated genes is associated
molecular marker for prostate and breast cancer. with circulating tumour cells in blood from breast
Ann N Y Acad Sci. 2006;1075:235–43. cancer patients. Br J Cancer. 2009;100(8):1277–86.
211. Taback B, Giuliano AE, Lai R, Hansen N, Singer 223. Esteller M, Corn PG, Urena JM, Gabrielson E,
FR, Pantel K, et al. Epigenetic analysis of body flu- Baylin SB, Herman JG. Inactivation of glutathione
ids and tumor tissues: application of a comprehen- S-transferase P1 gene by promoter hypermethylation
sive molecular assessment for early-stage breast in human neoplasia. Cancer Res.
cancer patients. Ann N Y Acad Sci. 1998;58(20):4515–8.
2006;1075:211–21. 224. Hoque MO, Feng Q, Toure P, Dem A, Critchlow
212. Conway KE, McConnell BB, Bowring CE, Donald CW, Hawes SE, et al. Detection of aberrant methyla-
CD, Warren ST, Vertino PM. TMS1, a novel pro- tion of four genes in plasma DNA for the detection
apoptotic caspase recruitment domain protein, is a of breast cancer. J Clin Oncol. 2006;24(26):4262–9.
target of methylation-induced gene silencing in 225. Vesuna F, Lisok A, Kimble B, Domek J, Kato Y, van
human breast cancers. Cancer Res. der Groep P, et al. Twist contributes to hormone
2000;60(22):6236–42. resistance in breast cancer by downregulating estro-
213. Evron E, Umbricht CB, Korz D, Raman V, Loeb gen receptor-α. Oncogene. 2012;31(27):3223–34.
DM, Niranjan B, et al. Loss of cyclin D2 expression 226. Bae YK, Shim YR, Choi JH, Kim MJ, Gabrielson E,
in the majority of breast cancers is associated with Lee SJ, et al. Gene promoter hypermethylation in
promoter hypermethylation. Cancer Res. tumors and plasma of breast cancer patients. Cancer
2001;61(6):2782–7. Res Treat. 2005;37(4):233–40.
214. Sharma G, Mirza S, Prasad CP, Srivastava A, Gupta 227. Lau QC, Raja E, Salto-Tellez M, Liu Q, Ito K, Inoue
SD, Ralhan R. Promoter hypermethylation of M, et al. RUNX3 is frequently inactivated by dual
p16INK4A, p14ARF, CyclinD2 and Slit2 in serum mechanisms of protein mislocalization and promoter
and tumor DNA from breast cancer patients. Life hypermethylation in breast cancer. Cancer Res.
Sci. 2007;80(20):1873–81. 2006;66(13):6512–20.
126 K.O. Yaykasli et al.

228. Tan SH, Ida H, Lau QC, Goh BC, Chieng WS, Loh inhibitor, selectively induces transforming growth
M, et al. Detection of promoter hypermethylation in factor beta type II receptor expression in human
serum samples of cancer patients by methylation- breast cancer cells. Cancer Res. 2001;61(3):931–4.
specific polymerase chain reaction for tumour sup- 238. Huang X, Gao L, Wang S, Lee CK, Ordentlich P, Liu
pressor genes including RUNX3. Oncol Rep. B. HDAC inhibitor SNDX-275 induces apoptosis in
2007;18(5):1225–30. erbB2-overexpressing breast cancer cells via down-
229. Rao X, Evans J, Chae H, Pilrose J, Kim S, Yan P, regulation of erbB3 expression. Cancer Res.
et al. CpG island shore methylation regulates caveo- 2009;69(21):8403–11.
lin-1 expression in breast cancer. Oncogene. 2012. 239. Sabnis GJ, Goloubeva O, Chumsri S, Nguyen N,
doi:10.1038/onc.2012.474. Sukumar S, Brodie AM. Functional activation of the
230. Kim SJ, Kang HS, Chang HL, Jung YC, Sim HB, estrogen receptor-α and aromatase by the HDAC
Lee KS, et al. Promoter hypomethylation of the inhibitor entinostat sensitizes ER-negative tumors to
N-acetyltransferase 1 gene in breast cancer. Oncol letrozole. Cancer Res. 2011;71(5):1893–903.
Rep. 2008;19(3):663–8. 240. Hirokawa Y, Arnold M, Nakajima H, Zalcberg J,
231. Pakneshan P, Szyf M, Farias-Eisner R, Rabbani Maruta H. Signal therapy of breast cancers by the
SA. Reversal of the hypomethylation status of uroki- HDAC inhibitor FK228 that blocks the activation of
nase (uPA) promoter blocks breast cancer growth PAK1 and abrogates the tamoxifen-resistance.
and metastasis. J Biol Chem. Cancer Biol Ther. 2005;4(9):956–60.
2004;279(30):31735–44. 241. Jawed S, Kim B, Ottenhof T, Brown GM, Werstiuk
232. Luu TH, Morgan RJ, Leong L, Lim D, McNamara ES, Niles LP. Human melatonin MT1 receptor
M, Portnow J, et al. A phase II trial of vorinostat induction by valproic acid and its effects in combina-
(suberoylanilide hydroxamic acid) in metastatic tion with melatonin on MCF-7 breast cancer cell
breast cancer: a California Cancer Consortium study. proliferation. Eur J Pharmacol. 2007;560(1):17–22.
Clin Cancer Res. 2008;14(21):7138–42. 242. Hodges-Gallagher L, Valentine CD, Bader SE,
233. Ramaswamy B, Fiskus W, Cohen B, Pellegrino C, Kushner PJ. Inhibition of histone deacetylase
Hershman DL, Chuang E, et al. Phase I–II study evi- enhances the anti-proliferative action of antiestro-
dence for vorinostat-induced tubulin acetylation and gens on breast cancer cells and blocks tamoxifen-
Hsp90 inhibition in vivo. Breast Cancer Res Treat. induced proliferation of uterine cells. Breast Cancer
2012;132(3):1063–72. Res Treat. 2007;105(3):297–309.
234. Chen S, Ye J, Kijima I, Evans D. The HDAC inhibi- 243. Reid G, Métivier R, Lin CY, Denger S, Ibberson D,
tor LBH589 (panobinostat) is an inhibitory modula- Ivacevic T, et al. Multiple mechanisms induce tran-
tor of aromatase gene expression. Proc Natl Acad scriptional silencing of a subset of genes, including
Sci U S A. 2010;107(24):11032–7. oestrogen receptor alpha, in response to deacetylase
235. Rao R, Nalluri S, Kolhe R, Yang Y, Fiskus W, Chen inhibition by valproic acid and trichostatin
J, et al. Treatment with panobinostat induces A. Oncogene. 2005;24(31):4894–907.
glucose-regulated protein 78 acetylation and endo- 244. Mongan NP, Gudas LJ. Valproic acid, in combina-
plasmic reticulum stress in breast cancer cells. Mol tion with all-trans retinoic acid and 5-aza-2′-
Cancer Ther. 2010;9(4):942–52. deoxycytidine, restores expression of silenced
236. Tate CR, Rhodes LV, Segar HC, Driver JL, Pounder RARbeta2 in breast cancer cells. Mol Cancer Ther.
FN, Burow ME, et al. Targeting triple-negative 2005;4(3):477–86.
breast cancer cells with the histone deacetylase 245. Dyer ES, Paulsen MT, Markwart SM, Goh M, Livant
inhibitor panobinostat. Breast Cancer Res. DL, Ljungman M. Phenylbutyrate inhibits the inva-
2012;14(3):R79. sive properties of prostate and breast cancer cell
237. Lee BI, Park SH, Kim JW, Sausville EA, Kim HT, lines in the sea urchin embryo basement membrane
Nakanishi O, et al. MS-275, a histone deacetylase invasion assay. Int J Cancer. 2002;101(5):496–9.
Nutrigenomics in Breast Cancer
Shailendra Dwivedi, Shailja Shukla, Apul Goel,
Praveen Sharma, Sanjay Khattri,
and Kamlesh Kumar Pant

Environmental factors and genetic makeup play a central role in phenotypic
appearance of a trait via central dogma of biology. Epigenetic, transcrip-
tomics, and proteomics are the key players of the expression biology which
make the difference in structure and function. Environmental factors
include the various exposure factors by oral intake, air, and via the skin
commonly. Several bioactive food components, including both essential
and nonessential nutrients, can regulate gene expression patterns. Thus,
nutrigenomics is providing the effects of ingested nutrients and other food
components on gene expression and gene regulation, i.e., diet–gene
interaction in order to spot the dietetic components having beneficial or
detrimental health effects. Nutritional genomics (nutrigenomics), the junc-
tion between health, diet, and genomics, is influenced via epigenetic, tran-
scriptomics, and proteomics processes of biology. Thus, it will help in
determining the individual nutritional requirements based on the genetic
makeup of the person (personalized diet) as well as the association between
diet and chronic diseases like cancer, opening new vistas to understanding
the complexity of breast cancer and leading to its better management.

Nutrigenomics • Breast cancer • Carcinogenesis • Isoflavones • Vitamins •

S. Dwivedi, MSc, PhD • P. Sharma, MSc, PhD

Department of Biochemistry,
All India Institute of Medical Sciences, Introduction
Jodhpur, India
S. Shukla, BDS, PhD • S. Khattri, MBBS, MD Worldwide, breast cancer accounts for 22.9 % of
K.K. Pant, MBBS, MD (*) all cancers (excluding nonmelanoma skin can-
Department of Pharmacology and Therapeutics,
cers) in women. In 2008, breast cancer caused
King George Medical University, Lucknow, India
e-mail: pharmacsmmu@gmail.com 458,503 deaths worldwide (13.7 % of cancer
deaths in women) [1]. Among all breast cancer
A. Goel, MS, Mch
Department of Urology, King George Medical cases, only 5–10 % are believed to be due to
University, Lucknow, India inherited susceptibility, and thus environmental

D. Barh (ed.), Omics Approaches in Breast Cancer: Towards Next-Generation Diagnosis, 127
Prognosis and Therapy, DOI 10.1007/978-81-322-0843-3_6, © Springer India 2014
128 S. Dwivedi et al.

factors including one’s eating habits are likely The ultimate goal of nutrigenomics is that of
major contributors to risk. developing genomics-based biomarkers that help
Current cancer models include those that are in the early detection and prevention of diet-
inherited through the germ line and represent related diseases, including cancer. To reach this
only ∼5 % of total cases of human cancers. goal, it is essential to develop tissue-specific
These tumors initiate because of mutational dietary responses that can be used as signatures
events. The remaining ∼95 % originate as spo- or fingerprints to estimate risk [5]. The availabil-
radic events and grow as a result of exposure ity of nutritional biomarkers at early stages (e.g.,
to the environment, which includes exposure to initiation) may be used as prognostic tools.
both environmental contaminants and dietary A complicating factor is that the diet contains a
agents. The etiology of cancers, and particu- large number of compounds and that each nutri-
larly breast cancer, is still not very clear. Several ent has different gene targets and affinities. An
researchers have provided various mechanisms example is the cross talk of estrogens and isofla-
of carcinogenesis, but since it is very complex vones with estrogen receptors and how this inter-
and interwoven with multiple factors, so the action may affect the development and prevention
current knowledge is still in infancy [2]. It is of breast cancer [6].
estimated that a third of all cancer deaths can Nutritional genomics (nutrigenomics), the
be attributed to dietary factors. This is due to junction between health, diet, and genomics, can
both a lack of intake of protective natural com- be seen as the combination of molecular nutri-
ponents in an individual’s diet, such as poly- tion and genomics. The diverse tissue- and organ-
phenols, sterols, flavonoids, and carotenoids, specific effects of bioactive dietary components
as well as exposure to natural carcinogens in include gene expression patterns (transcriptome);
the diet, such as aflatoxin, fumonisin, and organization of the chromatin (epigenome);
heavy metals. Bioactive food components can protein-expression patterns, including post-
influence a number of physiological processes: translational modifications (proteome); as well
apoptosis, metabolism, cell differentiation as metabolite profiles (metabolome) as repre-
and growth, DNA repair, hormone regulation, sented in Fig. 6.1, which shows bioactive food
inflammation, etc. components and their integration with various
Nutrigenomics includes the determination molecular streams in phenotype development
of individual nutritional requirements based on and progression. Table 6.1 represents the fac-
the genetic makeup of the person, as well as the tors affecting nutrigenomics and phenotype.
association between diet and chronic disease. Nutrigenomics will promote an increased under-
Nutrigenomics is part of a broader movement standing of how nutrition influences metabolic
toward personalized medicine, focusing on a per- pathways and homeostatic control and how
sonalized diet. Nutrigenomics is linked to nutri- this regulation is concerned in the early phases
genetics, which studies the genetic basis of the of diet-related disease and the extent to which
different individual response to the same nutri- individual sensitizing genotypes contribute to
tional stimulus. This phenomenon arises from such diseases. Nutrigenomics will also identify
gene polymorphism. As a consequence, genes are the genes involved in physiological responses to
important in determining a function, but nutrition diet and the genes in which small changes, called
is able to modify the degree of gene expression polymorphisms, may have significant nutritional
[3]. The risk of certain cancers such as breast consequences and the influence of environmental
cancer increases in association with Western factors on gene expression [7].
diets as compared to the Mediterranean or native Exploitation of this genomic information,
Mexican diets. Therefore, nutrition strategies along with high-throughput “omic” technolo-
need to be developed to prevent the effects of gies, allow the acquisition of new knowledge
carcinogenic agents, target and eliminate prema- aimed at obtaining a better understanding of
lignant lesions at the early stages, and antagonize nutrient–gene interactions depending on the gen-
(i.e., induce apoptosis) the proliferation of clonal otype, with the ultimate goal of developing per-
neoplastic populations [4]. sonalized nutrition strategies for optimal health
6 Nutrigenomics in Breast Cancer 129

Fig. 6.1 Bioactive food Nutritional Epigenomics

components and its integration
with various molecular streams Methylated Proteomics (post-translational)
in phenotype progression and Nutritional transcriptomics
DNA m RNA Proteins


Bioactive Food

Table 6.1 Factors affecting nutrigenomic research and phenotype

Approach Definition and factors
Transcriptomics The branch of molecular biology that deals with the study of messenger RNA molecules produced
in an individual or population of a particular cell type
Identification of transcription factors that respond to nutrients and gene targets
RNA amplification and procedure (quantity, quality, replicates, real-time PCR, high-density analysis)
Quantity of starting tissue/cell material
Fold change in expression
Intraindividual and interindividual variations in healthy and diseased subjects and Identification of
single genes or group of genes that are regulated (up or down) in a particular disease or nutritional
Heterogeneity of cell populations and single cell gene expression profiling
Combination of gene variants (SNPs)
Data processing and interpretation
Epigenetics It is defined as heritable changes in gene expression, which are not due to any alteration in the
DNA sequence
Characterization of chromatin modifications that influence gene expression and impact of nutrients
Histone modifications
DNA methylation
Nucleosome organization
Order, interdependence and intradependence, and reversibility of histone modifications
Cross talk and mutual dependency between histone modifications, DNA methylation, and
methyl-binding proteins
Help in exploring the evolutionary origin of cell differentiation and change in cancer cells
Proteomics It is the large-scale study of proteins, particularly their structures and functions
Linking gene expression studies with protein functions
Tissue and cellular localization
Plasma levels
Expression levels
Posttranslational modifications
Protein–protein interactions
Cellular function
Bioinformatics and data interpretations
Metabolomics It is the scientific study of chemical processes involving metabolites
Linking exposure to biological effects induced by metabolites
Interindividual differences in metabolisms and disposition
Measurements of metabolite in specimens
Recovery methods from tissue/plasma
Adapted from Milner et al. [2]
130 S. Dwivedi et al.

and disease prevention [8]. There are three characteristics depending on life stage, dietary
central factors that underpin nutrigenetics and preferences, and health status.
nutrigenomics as a central science. First there is Genomic and epigenomic processes likely do
great diversity in the inherited genome between not utterly account for the ability of dietary fac-
ethnic groups and individuals which affects tors to influence phenotypic changes, since
nutrient bioavailability and metabolism. Second, changes in the rate of transcription of genes (tran-
people differ greatly in their food/nutrient avail- scriptomics) can also be fundamental to cellular
ability and choices depending on cultural, eco- processes [11]. Multiple pathways appear to
nomic, geographical, and taste perception overlap as a cause of multiple diseases [12].
differences. Third, malnutrition (deficiency or Thus, the examination of these pathways via
excess) itself can influence gene expression and transcriptomic profiles may simultaneously pro-
genome stability; the latter leading to mutations vide important hints about multiple disease risks.
at the gene sequence or chromosomal level, Noteworthy, several bioactive food components,
which may cause abnormal gene dosage and including both essential and nonessential nutri-
gene expression leading to adverse phenotypes ents, can control gene expression patterns. Their
during the various life stages [9]. influence on gene transcription and translation is
The decisive goal is to (1) match the nutriome not only concentration dependent but also time
(i.e., nutrient intake combination) with the cur- dependent [13]. Yet these changes may provide
rent genome status (i.e., inherited and acquired significant insights about the specificity of indi-
genome) so that genome maintenance, gene vidual food components to influence one or more
expression, metabolism, and cell function can biological processes, including those involved in
occur normally and in a homeostatically sustain- the risk of cancer development and/or tumor
able manner [8] and (2) provide better interpreta- behavior.
tion of data from epidemiological and clinical This chapter will try to clarify the current
intervention studies regarding health impacts of research updates on interaction of diet with
dietary factors that may help to revise recommen- genetic backgrounds and how diet is involved in
dations for personalized nutrition [10]. the development of one of the most common
The fundamental hypotheses reinforcing the cancers afflicting females today globally.
science of nutrigenetics and nutrigenomics are Simultaneously we will pinpoint the various diets
the following [9]: that can be used in curing or decreasing the risks,
• Nutrition may apply its impact on health although the nature of these interactions is indeed
outcomes by directly affecting expression of very complex.
genes in critical metabolic pathways and/or
indirectly by affecting the incidence of genetic
mutation at the base sequence or chromo- Nutrigenomic Diseases
somal level which in turn causes alterations in
gene dosage and gene expression. Diseases that are known to be associated with the
• The health effects of nutrients and nutriomes interactions between multiple genetic and envi-
(nutrient combinations) depend on inherited ronmental factors such as diet include many can-
genetic variants that alter the uptake and cers, diabetes, heart disease, obesity, and some
metabolism of nutrients and/or the molecular psychiatric disorders. Therefore, both disciplines
interaction of enzymes with their nutrient aim to unravel diet–genome interactions; how-
cofactor and hence the activity of biochemical ever, their approaches and immediate objectives
reactions. are distinct. Nutrigenomics will unravel the
• Better health outcomes can be achieved if optimal diet from within a series of nutritional
nutritional requirements are customized for alternatives, whereas nutrigenetics will yield
each individual, taking into consideration critically important information that will assist
both his/her inherited and acquired genetic clinicians in identifying the optimal diet for a
6 Nutrigenomics in Breast Cancer 131

given individual, i.e., personalized nutrition [10]. polymorphisms may modify the chance of
The following five tenets of nutritional genomics contact between carcinogens and target cells,
serve as a conceptual basis for understanding the thus acting at the stage of cancer initiation.
focus and promise of this budding field: [7] Influences of polymorphisms of gene encoding
1. Under certain conditions and in some factors involved in hormonal regulation are most
individuals, diet can be a serious risk factor strongly manifested in hormone-dependent
for a number of diseases. tumors such as breast, prostate, ovarian, and
2. Universal dietary chemicals can act on the endometrial cancers. Polymorphisms in sex hor-
human genome, either directly or indirectly, to mone receptor genes comprising those encoding
alter gene expression or structure. estrogen receptor, progesterone receptor, and
3. The degree to which diet influences the bal- androgen receptor have been shown to be associ-
ance between healthy and disease states may ated with cancer risk modulation [18]. Dietary
depend on a person’s genetic makeup. factors can undoubtedly interact with hormonal
4. Some diet-modulated genes (and their normal, regulation. Obesity strongly affects hormonal
common variants) probably play a role in the status. At the same time, some food components,
onset, incidence, progression, and/or severity such as phytoestrogens, are known to be pro-
of chronic diseases. cessed by the same metabolic pathways as sex
5. Dietary intervention based on the knowledge hormones [19]; thus their cancer-preventive
of nutritional requirements, nutritional status, effect can be modulated by the polymorphisms.
and genotype (i.e., personalized nutrition) can
be employed to prevent, mitigate, or cure
chronic disease. Epigenetic Link with Nutrigenomics

An important emerging part of nutrient–gene

Nutrigenomics and Carcinogenesis interaction studies with the potential for both
intra- and transgenerational effects is epigenetics
Cancer is a process composed of multiple stages [20]. Epigenetics refers to the processes that reg-
in which gene expression, and protein and metab- ulate how and when certain genes are turned on
olite function, begins to operate aberrantly [14]. and off, while epigenomics pertains to analysis of
In the post-genomic era, the cellular events medi- epigenetic changes in a cell or the entire organ-
ating the onset of carcinogenesis, in addition to ism. Epigenetic processes have a sturdy influence
their modulation by dietary factors, have yielded on normal growth and development, and this pro-
significant information in understanding of this cess is deregulated in diseases such as cancer.
disease [15]. Inherited mutations in genes can Diet on its own or by interaction with other envi-
increase one’s susceptibility for cancer. Evidences ronmental factors can cause epigenetic changes
of genome and epigenome damage biomarkers, that may turn certain genes on or off. Table 6.2
in the absence of overt exposure of genotoxins, provides a brief description of the nutrients and
are themselves sensitive indicators of deficiency chemicals involved in DNA methylation.
in micronutrients required as cofactors or as com- Epigenetic silencing of genes that would usually
ponents of DNA repair enzymes, for maintenance protect against a disease, as a result, could make
methylation of CpG sequences and prevention of people more susceptible to developing that dis-
DNA oxidation and/or uracil incorporation into ease later in life. The epigenome, which is heri-
DNA [16]. Diet is considered as a source of either table and modifiable by diet, is the global
carcinogens (intrinsic or cooking generated) epigenetic pattern determined by global and
present in certain foods or constituents acting in a gene-specific DNA methylation, histone modifi-
protective manner (vitamins, antioxidants, detox- cations, and chromatin-associated proteins that
ifying enzyme-activating substances, etc.) [17]. control expression of housekeeping genes and
It is clear that carcinogen metabolism-affecting restrains the expression of parasitic DNA such as
132 S. Dwivedi et al.

transposons. Table 6.3 represents the dietary the P21 and BAX genes, and higher expression
chemicals, DNA methylation, and its mechanism of p21Cip1/Waf1 and BAX proteins [33].
of action. Importantly, sulforaphane has been reported to
One study has demonstrated that sulfora- reduce HDAC activity in humans [33]. Future
phane, butyrate, and allyl sulfur are effective research likely needs to relate HDAC changes in
inhibitors of histone deacetylase (HDAC). HDAC humans to a change in a cancer-related process.
inhibition was associated with global increases in Furthermore, since acetylation is only one
histone acetylation, enhanced interactions of method to regulate histone homeostasis [34],
acetylated histones with the promoter regions of greater concentration needs to be given to how
nutrition might influence the other types of his-
Table 6.2 Nutrients and chemicals involved in DNA tone modifications.
(hyper-/hypo-) methylation The field of nutrigenomics harnesses multiple
Nutrient Chemicals disciplines and includes dietary effects on
Alcohol Genistein genome stability (DNA damage at the molecular
Arsenic Methionine and chromosome level), epigenome alterations
Betaine Nickel (DNA methylation), RNA and micro-RNA
Cadmium Polyphenol expression (transcriptomics), protein expression
Choline Selenium (proteomics), and metabolite changes (metabolo-
Coumestrol Vitamin A mics), all of which can be studied independently
Equol Vitamin B6 or in an integrated manner to diagnose health
Fiber Vitamin B12 status and/or disease trajectory. However, of
Folate Zinc these biomarkers, only DNA damage is a clear
Adapted from Trujillo et al. [21], Davis and Uthus [22] biomarker of fundamental pathology that may

Table 6.3 Dietary chemicals, DNA methylation, and its mechanism of action
Dietary chemicals Mechanism of action Phenotype/outcome Reference
Alcohol Affects folate metabolism, altering DNA methylation Cancer susceptibility [23]
Arsenic Compete with cytosine, DNA methyl transferase and Global hypomethylation [24]
selenium for methyl donation from in the liver, cancer
S-adenosil-1-methionine susceptibility
Choline Deficiency in diets has been associated with decreased Hepatic tumorigenesis, [25]
tissue S-adenosil-1-methionine cancer susceptibility
Folate Its deficiency has complex effect on DNA methylation Cancer susceptibility [26]
depending on cell type, organ, and development stage
Genistein Dietary genistein can migrate tumorigenic process via Mitigates tumorigenesis [27]
promoter modulation of gene expression
Lycopene It has direct DNA demethylating activity. It migrates Mitigates tumorigenesis [28]
tumorigenic processes via promoter methylation
modulation of gene expression
Methionine Its deficiency decreases tissue SAM resulting in global HCC [29]
DNA hypomethylation and HCC in rodents
Nickel Environmental carcinogen, induce de novo methylation Cancer susceptibility [30]
of tumor-suppressor genes Suppressive effect on histone
H4 acetylation in mammalian cells
Selenium Its deficiency decreases DNA methylation. Low intake Cancer susceptibility [31]
influences the activity of selenoproteins, causing changes
in mRNA levels for the encoding genes
Vitamins Vitamins (B2, B6, and B12) are necessary cofactors in Affect several metabolic [32]
one carbon (methyl metabolism) pathways, cancer
6 Nutrigenomics in Breast Cancer 133

be mitigated by promotion of apoptosis of Advances in genomics, transcriptomics,

genetically aberrant cells or by reducing the rate proteomics, and metabolomics have enabled a
of DNA damage accumulation. Changes at the more rapid and comprehensive understanding of
epigenome, transcriptome, proteome, and metab- how bioactive compounds affect human health.
olome levels may simply reflect modifiable Dietary bioactive compounds can be tested for
homeostatic responses to altered nutritional their potential health-promoting properties by
exposure and on their own may not be sufficient applying these different technologies to cell cul-
to indicate definite irreversible pathology at the ture, and animal or human studies. Each experi-
genome level. mental approach offers unique strengths and has
DNA damage can be diagnosed in a number of certain limitations.
complementary ways as follows: (1) damage to
single bases (e.g., DNA adducts such as the addi-
tion of a hydroxyl radical to guanine caused by Current Updates of Nutrigenomic
oxidative stress), (2) basic sites in the DNA Studies in Breast Cancer
sequence (measurable by use of the aldehyde-
reactive probe), (3) DNA strand breaks (com- Several studies of sporadic breast cancers have
monly measured using the Comet assay), shown that fruits and vegetables [36], fish, mono-
(4) telomere shortening (measured by terminal unsaturated and polyunsaturated fatty acids [37],
restriction fragment length analysis, quantitative vitamin D, calcium, and phytoestrogens may
PCR, or flow cytometry), (5) chromosome break- reduce the risk of breast cancer, although there
age or loss (usually measured using micronucleus are inconsistencies in the literature. High intake
cytome assays or metaphase chromosome analy- of meat, poultry, total energy, and total fat and
sis), and (6) mitochondrial DNA damage (usually saturated fatty acids has been reported to be
measured as deletions or base damage in the cir- associated with increased risk for breast cancer
cular mitochondrial DNA sequence). These DNA [38]. Malmö Diet and Cancer cohort examined
damage biomarkers are presently at different lev- the association between dietary folate equiva-
els of validation based on evidence relating to the lents (DFE) and breast cancer among carri-
association with nutrition (cross-sectional epide- ers of two genetic polymorphisms for MTHFR
miological and intervention studies) and disease gene (MTHFR 677C/T and 1298A/C). A posi-
(cross-sectional epidemiology and prospective tive association between DFE and breast can-
cohort studies) [35]. The micronucleus assay in cer among women carriers of MTHFR 677CT/
cytokinesis-blocked lymphocytes is currently the TT-1298AA occurred while an inverse associa-
best validated biomarker for nutritional genomic tion was observed in 677CT-1298 AC women
studies of DNA damage. [39]. In a nested case-control study, Maruti et al.
Given the advances in diagnostic technologies reported that postmenopausal women with two
assessing DNA damage, it has now become fea- copies of variant T alleles (TT genotype) had
sible to determine dietary reference values for increased risk of breast cancer. In addition, the
DNA damage prevention and to start translating intake of other B vitamins may influence the rela-
into practice the Genome Health Clinic concept tionship between the MTHFR genetic variants
of DNA damage prevention [35]. The latter is and breast cancer risk. It has been found that the
based on the recognition that damage to the most pronounced MTHFR-breast cancer risk was
genome is the most elementary cause of develop- observed among women with the lowest intakes
mental and degenerative diseases, which can be of dietary folate and vitamin B6 [40].
accurately diagnosed and prevented by appropri- In a nested case-control study within the
ate diet and lifestyle intervention at a genetic sub- Singapore Chinese Health Study, it has been
group and personalized level. The ability of diet observed that there is an inverse relationship
to affect the flow of genetic information can between breast cancer risk and low folate intake
occur at multiple sites of regulation [10]. and weekly/daily green tea intake compared with
134 S. Dwivedi et al.

reduced green tea intake. Also, women carrying genotype [45], and not necessarily be beneficial
the high-activity MTHFR/TYMS genotypes with to all women.
0–1 variant allele and weekly/daily green tea Finally, the response to marine n-3 fatty acids
intake had a lower breast cancer risk, especially with breast cancer risk may depend on genetics.
women who also had low folate intake. No asso- Women with genetic variants that encode lower or
ciation was observed among women carriers of no enzymatic activity of GSTT1 have a 30 %
two variant alleles. These findings suggest one of lower breast cancer risk from the marine n-3 fatty
the mechanisms through which green tea can pro- acids, when compared with women with high-
vide protection against breast cancer is through activity genotypes. These data suggest that the
folate regulation [41]. The anticancer effect of peroxidation products of n-3 fatty acids may be
the EGCG (tea polyphenol (−)-epigallocatechin- involved in the protection against breast cancer
3-gallate) may be mediated by the regulation of [46]. This is not that unusual since other food
epigenetic processes. It was found that EGCG components have been reported to inhibit tumors
can lead to ERα reactivation in the ERα-negative by generating free radicals [47]. Watercress, a rich
breast cancer cell lines by remodeling the chro- source of phenethyl isothiocyanates (PEITC), has
matin structure of the ERα promoter by the alter- been proposed to have anticancer activity. A crude
ation of the histone acetylation and methylation watercress extract was reported to inhibit cancer
status [42]. Further, it has been reported that cell growth and hypoxia-inducible factor (HIF)
EGCG inhibits the telomerase by decreasing the activity and reduced angiogenesis by decreasing
hTERT promoter methylation and ablating the the phosphorylation of the translation regulator
histone H3 Lys9 acetylation in the MCF-7 cell 4E binding protein 1 (4E-BP1) [48].
lines [43]. In a breast case-control study that examined
Another study reported the inverse relation- the association of the 5-lipoxygenase gene
ship between green tea intake and breast cancer (ALOX) and 5-lipoxygenase-activating protein
risk among Asian-Americans [44] as a function gene (ALOX5AP) polymorphisms and dietary
of the catechol-O-methyltransferase (COMT) linoleic acid intake with breast cancer risk, it was
genotype. This enzyme is recognized to be found that women carriers of two variant alleles
involved in the metabolism of the tea polyphe- for the ALOX5AP 4900 A/G who had a diet rich
nols. More specifically, a reduced risk of breast in linoleic acid had a greater breast cancer risk
cancer was observed only among tea (green and compared with women carrying AG or GG geno-
black) drinker carriers of at least one low-activity types. These results propose that genetic predis-
COMT allele. These findings of reduced breast position related to n-6 polyunsaturated fatty acid
cancer risk with tea catechins, especially in metabolism should be taken into account when
women who had the low-activity COMT alleles, the relation between dietary fat and breast cancer
suggest that these women were less efficient in risk is examined [49]. Furthermore, dietary
eliminating tea catechins, therefore optimizing omega-3 polyunsaturated fatty acids downregu-
the benefits from the tea and its associated bioac- late the expression of the polycomb group (PcG)
tive constituent [45]. protein, enhancer of zeste homologue 2 (EZH2)
The Shanghai Breast Cancer Study examined in breast cancer cells. This study reported a
the relationship between breast cancer risk, decrease in histone 3 lysine 27 trimethylation
GSTP1 genetic variants, and other diet compo- (H3K27me3) activity of EZH2 and upregulation
nents, the cruciferous vegetables. The GSTP1 of E-cadherin and insulin-like growth factor-
Val/Val genotype was associated with increased binding protein 3. The treatment with omega-3
breast cancer risk, especially in premenopausal PUFAs led to decrease in the invasion capacity of
women with low intake of cruciferous vegeta- the breast cancer cells [50].
bles. Thus, cruciferous vegetable intake with Additionally, a nested case-control study of
high isothiocyanates may reduce breast cancer postmenopausal women examined the interac-
risk and modify the effect of the GSTP1 tion between oxidative stress-related genes
6 Nutrigenomics in Breast Cancer 135

including catalase (CAT) C262T, myeloperoxi- Dietary berries have been suggested to
dase (MPO) G463A, endothelial nitric oxide influence breast cancer risk (AICR Report),
synthase (NOS3) G894T, and heme oxygen- although considerable variability in response is
ase-1 (HO-1) GT (n) dinucleotide length poly- observed. Preclinical studies demonstrate that
morphism and level of vegetable and fruit intake tumor formation is suppressed as a result of the
on breast cancer risk. The study found that levels of E(2)-metabolizing enzymes during the
women with low intake of vegetables and fruits early phase of E(2) carcinogenesis [55].
and the low-risk CAT CC genotypes appeared to A nested case-control study within the
be associated with increased breast cancer risk, Singapore Chinese Health Study Cohort showed
especially those women with four or more low- a significant interaction between the level of
risk alleles, suggestive of the role of endoge- green tea drinking and the activity of the
nous and exogenous antioxidants in breast angiotensin-converting enzyme (ACE) with
carcinogenesis [51]. respect to breast cancer risk depending on ACE
Iwasaki et al. examined the effect of four gene polymorphism [56]. Even the BRCA breast
SNPs in cytochrome P450c17alpha (CYP17), cancer-associated gene is affected by diet. A diet
aromatase (CYP19), 17beta-hydroxysteroid rich in fruits and vegetables protects a woman
dehydrogenase type I (17beta-HSD1), and sex from the BRCA gene becoming activated [57].
hormone-binding globulin (SHBG) genes on the Olive oil is an integral ingredient of the
association between isoflavone intake and breast “Mediterranean diet” and accumulating evidence
cancer risk. The study identified an inverse asso- suggests that it may have a potential role in low-
ciation between isoflavone intake and breast can- ering the risk of several types of cancers. A num-
cer risk among women with at least one variant ber of epidemiological studies have linked
allele for the 17beta-HSD1 polymorphism and consumption of olive oil with a reduced risk of
among postmenopausal Japanese women with cancer, and researchers are increasingly
GG genotype for the SHBG gene, indicating that investigating this association further in labora-
genetic variants of the 17beta-HSD1 and SHBG tory studies. The mechanisms by which the can-
genes may modify the relationship between iso- cer-preventing effects of olive oil as having novel
flavone intake and breast cancer risk [52]. The anticancer actions may relate to the ability of its
association between isoflavones and breast can- monounsaturated fatty acid (MUFA) oleic acid
cer risk may be explained by the antiestrogenic (OA; 18:1n-9) to specifically regulate cancer-
effect of the isoflavones and the effect on the related oncogenes. Exogenous supplementation
DNA methylation. It has been reported that the of cultured breast cancer cells with physiological
daily administration of isoflavones to healthy concentrations of OA was found to suppress the
premenopausal women led to dose-specific overexpression of HER2 (Her-2/neu, erbB-2),
changes in RARbeta2 and CCND2 gene pro- a well-characterized oncogene playing a key role
moter methylation, changes that correlated with in the etiology, progression, and response to che-
genistein levels. Additionally, an inverse corre- motherapy and endocrine therapy in approxi-
lation between estrogenic marker complement mately 20 % of breast carcinomas.
C3 and genistein was observed, suggesting an OA treatment was also found to synergisti-
antiestrogenic effect [53]. Furthermore, genis- cally enhance the efficacy of trastuzumab
tein has been associated with specific epigenetic (Herceptin) a humanized monoclonal antibody
changes. For example, the long-term exposure to binding with high affinity to the ectodomain
genistein of the MCF-7 breast cancer cell lines (ECD) of the Her2-coded p185(HER2) oncopro-
has been found to lead to reduced expression tein. Moreover, OA exposure significantly
of the acetylated histone 3 (H3). Additionally, diminished the proteolytic cleavage of the ECD
this exposure was associated with alteration of HER2 and, consequently, its activation status,
in growth responses to mitogenic factors and a crucial molecular event that determines both
histone deacetylase inhibitors [54]. the aggressive behavior and the response to
136 S. Dwivedi et al.

trastuzumab of Her2-overexpressing breast Table 6.4 Transcription factor pathways mediating

nutrient–gene interactions [60]
carcinomas. Recent findings further reveal that
OA exposure may suppresses HER2 at the tran- Macronutrients Compound Transcription factor
scriptional level by upregulating the expression Fats Fatty acids PPARs, SREBPs,
Cholesterol LXR, HNF4,
of the Ets protein PEA3 -a DNA-binding protein
that specifically blocks HER2 promoter activity FXR
in breast, ovarian, and stomach cancer cell lines. Carbohydrates Glucose USFs, SREBPs,
This anti-HER2 property of OA offers a previ- ChREBP
ously unrecognized molecular mechanism by Proteins Amino acids C/EBPs
which olive oil may regulate the malignant Micronutrients
behavior of cancer cells. From a clinical perspec- Vitamins Vitamin A RAR, RXR, VDR,
Vitamin D PXR
tive, it could provide an effective means of influ-
encing the outcome of Her-2/neu-overexpressing Vitamin E
human carcinomas with poor prognosis. Indeed, Minerals Calcium Calcineurin/
OA-induced transcriptional repression of HER2
Iron IRO1, IRP2
oncogene may represent a novel genomic expla- Zinc MTF-1
nation linking olive oil and cancer, as it seems to Other food components
equally operate in various types of Her-2/neu- Soy Flavonoids ER, NF-Kb, AP1
related carcinomas [58]. Xenobiotics CAR, PXR
In another study, OA treatment in Her-2/neu-
overexpressing cancer cells was found to induce
upregulation of the Ets protein polyomavirus thus what constitutes an ideal diet for health pro-
enhancer activator 3 (PEA3), a transcriptional motion [2]. For example, the apoptosis or pro-
repressor of Her-2/neu promoter. Also, an intact grammed cell death that is essential in the fight
PEA3 DNA-binding site at endogenous Her-2/ against cancer through two pathways—either the
neu gene promoter was essential for OA-induced intrinsic, mitochondrial-mediated pathway, or the
repression of this gene. Moreover, OA treatment extrinsic, death receptor-mediated pathway—
failed to decrease Her-2/neu protein levels in could be a target for dietary bioactive agents
MCF-7/Her2-18 transfectants, which stably including genistein, curcumin, resveratrol, luteo-
express full-length human Her-2/neu cDNA con- lin, lupeol, indole 3-carbinol, etc. [4]. Dietary
trolled by a SV40 viral promoter. OA-induced components can modulate apoptosis through
transcriptional repression of Her-2/neu occurs effects at different levels, all of which culminate
through the action of PEA3 protein at the pro- in changes in gene expression. Table 6.4 provides
moter level [59]. details on transcription factor pathways mediat-
ing nutrient–gene interactions.
Compounds such as Japanese knotweed
Various Diet Components (Polygonum c.), 20 % resveratrol, ginger, (Zingiber
and Their Cellular/Molecular off.) 5 % gingerols, Rosemary (Rosemarinus
Effects on Breast Cancer off.), 6 % carnosic acid, 1 % rosemarinic acid,
1.5 % ursolic aicd, etc have demonstrated broad-
All main signaling pathways are deregulated in spectrum, multi-targeting, anticancer effects,
cancer, including cell proliferation, apoptosis, as well as disease-preventive and health-pro-
DNA repair, carcinogen metabolism, inflamma- moting benefits. The other important aspect of
tion, immunity, differentiation, and angiogenesis. these compound-rich foods, spices, and herbs is
Increasingly, evidence points to each of these as that have been regularly used by many cultures
molecular targets for cancer prevention. Since sev- throughout the world. Cancer prevention stud-
eral of these sites appear to be tailored by multiple ies have exposed that all of the major signaling
dietary components, it becomes challenging to pathways deregulated in different types of can-
tease apart nutrient–nutrient interactions and cer are affected by nutrients. Pathways studied
6 Nutrigenomics in Breast Cancer 137

include carcinogen metabolism, DNA repair, cell the G1 to the S phase of the cell cycle [68]. This
proliferation/apoptosis, differentiation, inflam- effect was mediated by increased expression of
mation, oxidant/antioxidant balance, and angio- cyclin A1 in ovarian cancer cells [69], whereas
genesis [61]. breast cancer cells had increased expression of
So far, more than 1,000 different phytochemi- cyclin D2 [70]. It has also been suggested that the
cals have been identified with cancer-preventive enzyme responsible for the degradation of vita-
activities [62]. Dietary fibers have a protective min D metabolites, CYP24, can also be influ-
effect against bowel cancer. Long-chain polyun- enced by cancer. The CYP24 gene was amplified
saturated fatty acids (LC-PUFA) beneficially in breast tumors [71].
affect physiological processes, including growth; The majorities of established breast cancer
neurological development; lean and fat mass cell lines express transcriptionally active VDR
accretion; reproduction; innate and acquired and undergo growth inhibition in response to
immunity; infectious pathologies of viruses, bac- 1,25D [72]. In general, VDR expression and sen-
teria, and parasites; and the incidence and sever- sitivity to 1,25D-mediated growth arrest is higher
ity of virtually all chronic and degenerative in the less aggressive, estrogen receptor (ER)-
diseases including cancer, atherosclerosis, stroke, positive breast cancer cell lines such as MCF-7
arthritis, diabetes, osteoporosis, neurodegenera- than in ER-negative cell lines. Tumor cells
tive, inflammatory, and skin diseases [63]. Fish derived from VDR null mice were used to con-
oil, rich in omega-3 fatty acids, inhibits the clusively demonstrate that 1,25D mediates effects
growth of colonic tumors in both in vitro and in breast cancer cells via the nuclear VDR [73].
in vivo systems [64]. Screening for molecular changes induced by
Bioactive components present in fruits and 1,25D or vitamin D analogs in various breast can-
vegetables can prevent carcinogenesis by several cer cells has identified scores of VDR regulated
mechanisms, such as blocking metabolic activa- genes and proteins in diverse pathways, indicat-
tion through increasing detoxification. Plant ing a broad range of downstream [74] involved in
foods can modulate detoxification enzymes as cell cycle (cyclins, cyclin-dependent kinases and
flavonoids, phenols, isothiocyanates, allyl sulfur their inhibitors), apoptosis/autophagy (bcl-2 fam-
compounds, indoles, and selenium [65]. As a ily, caspases, cathepsins), and inflammation
result of carcinogen activation, covalent adducts (NFkB, prostaglandins, cox-2). The net effect of
with the individual nucleic acids of DNA or RNA these changes is to block mitogenic signaling,
are formed. It has also been found that reactive including that of estrogen, EGF, IGF-1, and KGF,
oxygen species (ROS) such as superoxide anions, and to enhance the effects of negative growth fac-
hydrogen peroxide, and hydroxyl radicals attack tors such as TGFb. In many breast cancer cell
DNA bases, resulting in potential mistranscrip- lines, 1,25D-mediated growth arrest is associated
tion of DNA sequence [66]. Such disruptions can with the induction of differentiation markers such
interfere with DNA replication and thus produce as casein, lipid droplets, and adhesion proteins
mutations in oncogenes and tumor-suppressor [75]. Notably, 1,25D exerts additive or synergis-
genes. ROS can also result in breakage of DNA tic effects in combination with other triggers of
strand, resulting in mutations or deletions of apoptosis, such as ionizing radiation and chemo-
genetic material [67]. therapeutic agents [76].
Collectively, these studies indicate that a wide
variety of signaling pathways, cell cycle and
Vitamin D apoptotic regulatory proteins, and proteases con-
tribute to the antiproliferative, pro-differentiating,
Few findings demonstrate that primary circulat- and apoptotic effects of 1,25D depending on the
ing form of vitamin D 1,25(OH)2D acts in a cell specific breast cancer cell line and/or context.
type and tissue-specific manner. For example, In primary cultures of normal human mammary
1,25(OH)2D inhibits cell growth of both normal epithelial (HME) cells, vitamin D signaling
and tumor cells by inhibiting the transition for also mediates growth arrest and induction of
138 S. Dwivedi et al.

differentiation markers such as E-cadherin, but breast cancer, although more studies are needed.
apoptosis has not been observed [77]. In contrast High levels of angiotensin II have been associ-
to breast cancer cells, non-transformed mam- ated with an increased risk of breast cancer
mary cells retain expression of CYP27B1 and development in humans [56]. The angiotensin
generate 1,25D when incubated with physiologi- I-converting enzyme (ACE) gene encodes or acti-
cal concentrations of 25D. Many breast cells also vates the enzyme that converts angiotensin I to
express the megalin–cubilin complex, which the active angiotensin II. A low conversion rate is
mediates internalization of 25D bound to the linked with lower rates of breast cancer in women
vitamin D binding protein [78]. Autocrine metab- than in those who have a high conversion rate. In
olism of 25D triggers chemopreventive effects in those with the high-activity genotype, a high con-
breast epithelial cells including growth inhibi- sumption of green tea has resulted in a dramatic
tion, differentiation, and protection from various drop of one-third in the risk of developing breast
cellular stresses [77]. cancer. The authors concluded that the antioxi-
In the intact mammary gland, the epithelium dant properties (particularly of the EGCG) are
is surrounded by stromal fibroblasts and adipo- protective against the reactive oxygen species (or
cytes, which provide critical growth factor sig- free radicals) generated by the high levels of
nals for development and also impact on angiotensin II. No such association was made in
carcinogenesis. Recent evidence suggests that women with the low levels of angiotensin.
breast adipocytes express CYP27B1 and gener- Epigallocatechin gallate (ECGC) and other green
ate 25D, which signals via adipocyte VDR to and black tea polyphenols inhibit cancer cell sur-
release inhibitory factors that regulate mammary vival. EGCG suppressed androgen receptor
epithelial cell growth [79]. Since vitamin D expression and signaling via several growth fac-
metabolites are stored in fat tissue, the contribu- tor receptors. Cell cycle arrest or apoptosis
tion of adipocyte signaling to the tumor- involved caspase activation and altered Bcl-2
suppressive actions of vitamin D in mammary family member expression. EGCG inhibited
gland are likely of physiological importance and telomerase activity and led to telomere fragmen-
require further study. As in colon cancer, acquisi- tation. While at high concentrations polyphenols
tion of the transformed phenotype in breast cells had pro-oxidative activities, at much lower lev-
is associated with deregulation of the vitamin D els, antioxidative effects occurred [83].
pathway [80]. In HME cells, introduction of Another study conducted at the University of
SV40 large T antigen and/or oncogenic ras Southern California looked at green tea ingestion
induces transformation and reduces responsive- and activity of the catechol-O methyl transferase
ness to 25D in association with downregulation (COMT) gene [45]. Dr Wu’s group also found an
of VDR and CYP27B1 [81]. Oncogenes and association between green tea intake and a
tumor-suppressor genes that impact on VDR cancer-protective effect in those individuals with
expression in breast cells include ras, p53, and at least one copy of the low conversion COMT
slug, which act via diverse mechanisms including gene. This means that the beneficial catechins
transcriptional regulation and mRNA instability remained in circulation for a longer time period
[82]. The mechanism by which transformation and reduced the risk of breast cancer. It is impor-
abrogates CYP27B1 expression in breast cells is tant to note that the study was only conducted in
not yet known. Asian-Americans and needs to be reproduced in
a wider population group.

Green Tea
Soy and Isoflavones
From a nutrigenomic perspective, the use of
green tea as a nutraceutical or functional food has The interest on isoflavones in breast cancer
shown anticancer potential, in particular for prevention derives from the fact that breast
6 Nutrigenomics in Breast Cancer 139

Table 6.5 Various studies on isoflavones and nutrige- receptor (ER/PR) status of breast cancer with a
nomic approaches in breast cancers
dose–response relationship. The protective effect
Experimental Dietary bioactive Reference was more pronounced for women with ER+/
model compounds PR + and ER–/PR– breast tumors [93].
Human MCF-7 Natural estrogens [85]
Several factors may be accountable for the
breast cancer cells (1,7 beta estradiol,
estriol, estrone, inconsistent effects of soy-related diets on cancer
genistein) outcome. These include age, reproductive his-
Human MCF-7 Isoflavones [86] tory, genetic background, dose and timing of
breast cancer cells (genistein, daidzein, exposure, and dietary patterns. For example,
glycitein, biochanin
because of their binding affinity for the ER, iso-
A, and ipriflavone),
flavones (chrysin, flavones may function as agonists or antagonists
luteolin, and depending on the concentration. The differential
apigenin), flavonols binding of isoflavones to the ER may interfere
(kaempferol and
with or activate the genomic actions of the
quercetin), and
coumestan, ER. Moreover, the agonist/competing effects of
flavanone and isoflavones for the ER may be modified by inter-
chalcone actions with polymorphisms for the ER [94]. For
example, polymorphisms in the ERβ have been
naringenin, and
phloretin, shown to modify the association between isofla-
respectively) vone intake and breast cancer risk [95]. Given the
Human MCF-7 Genistein [87] role of cross talk between ER and isoflavones in
breast cancer cells breast cancer risk, genome-wide studies are
Human MCF-7, Genistein [88] required to examine the effects of isoflavones and
T47D breast
cancer cells exposure levels on promoter sequences that are
FVB female mice Isoflavones, of [89] targeted by the ER. DNA microarray technolo-
which 66.5 % was gies have been used to monitor genome-wide
genistein, 32.3 % effects by isoflavones.
daidzein, and 1.2 %
Soy and its processed products (tofu, tempeh,
miso, natto, soy milk, and soy-based yogurts and
desserts) are the only sources providing high
cancer risk for women residing in geographical quantities of isoflavones in the human diet. Soy
areas of high consumption of soy products dur- intake has been reported in some cases to be
ing puberty is lower compared to that of women linked to a reduction in breast cancer risk.
living in Western countries, and Asian women Recently, Satih et al. identified 278 and 334 dif-
who had a low soy intake [84]. Table 6.5 gives ferentially expressed genes after treatment with
details on various studies on isoflavones and two soy constituents, genistein and daidzein,
nutrigenomic approaches in breast cancers. respectively, in estrogen-positive (MCF-7) and
However, clinical trials reported small [90] or no estrogen-negative (MDA-MB-231, MCF-10a)
effect of supplementation with isoflavones on cells [96]. Isoflavone intake has been estimated to
breast cancer risk [91], and administration of iso- be 25–50 mg/day in Asian countries, and there is
flavones elicited in some cases an estrogen-like a lower risk of breast cancer in areas of high soy
effect. Other studies indicated that the reduction and isoflavone intake, especially in Asia [97].
in breast cancer risk due to soy intake was lim- A smaller risk reduction for breast cancer (odds
ited to Asian populations [92]. A case-control ratio 0.86, 95 % CI 0.75–0.99), stronger for pre-
study conducted in Southeast China in 2004– menopausal women, was found in a meta-
2005 reported that premenopausal and post- analysis compiling 6 cohort and 12 case-control
menopausal women in the highest quartile of studies [90]. Asian women whose soy intake was
total isoflavone intake had a reduced risk for all high during puberty experienced lower risk for
140 S. Dwivedi et al.

breast cancer than women who did not consume treated with GLA, it not only helps suppress the
soy products or did so only as adults [98]. cancer-causing gene but also causes up to a
Messina et al. recently discussed the published 40-fold increase in response to the drug Herceptin
and ongoing clinical breast cancer studies [91]. (trastuzumab), which is used as part of breast
Three double-blind randomized controlled trials cancer treatment. GLA also selectively affects
reported no effect of a 1–2-year isoflavone sup- cancer cells without damaging normal cells. This
plementation on mammographic density used as is especially good news because patients who
a marker of breast cancer risk [99]. A 2-week possess the HER2/neu gene also typically have
administration of a soy supplement (45 mg/day an aggressive form of the disease and a poor
isoflavones) increased epithelial cell proliferation prognosis. GLA is one of two essential fatty
and progesterone receptor (PR) expression in acids, which are necessary for the normal func-
normal breast tissue, suggesting an estrogen ago- tioning and growth of cells, nerves, muscles, and
nist effect [100]. Mechanism of isoflavones as organs [104].
anticancerous have been reported to modulate Data derived from epidemiological and exper-
steroid biosynthesis, transport and metabolism, imental studies suggest that alpha-linolenic acid
as well as carcinogen activation and detoxifica- (ALA; 18:3n-3), the main omega-3 polyunsatu-
tion, to inhibit cell proliferation induced by rated fatty acid (PUFA) present in the Western
growth factors, to induce cell cycle arrest or diet, may have protective effects in breast cancer
apoptosis, to favor cell differentiation, to reduce risk and metastatic progression. A recent pilot
oxidative stress, or to inhibit angiogenesis, cell clinical trial assessing the effects of ALA-rich
invasiveness, and metastasis [101]. They may act dietary flaxseed on tumor biological markers in
through modulation of cell signaling (direct bind- postmenopausal patients with primary breast
ing to nuclear receptors, modification of the cancer demonstrated significant reductions in
phosphorylation state of some signal transduc- tumor growth and in HER2 (erbB-2) oncogene
tion proteins), regulation of gene expression, and/ expression. The molecular mechanism by which
or specific inhibition of some key enzyme activi- ALA inhibits breast cancer cell growth and
ties. In addition to the inhibition of cell prolifera- metastasis formation may involve a direct regula-
tion, isoflavones can induce apoptosis in the tion of HER2, a well-characterized oncogene
human breast and prostate cells at concentrations playing a key role in the etiology, progression,
over 25 and 20 mM, respectively [102]. Catechins, and response to some chemo- and endocrine ther-
for example, belong to the flavonoid family, apies in approximately 20 % of breast carcino-
which are polyphenolic compounds available in mas. In a recent study, ALA exposure was found
foods of plant origin, and there is much research to dramatically repress the activity of HER2/neu.
into their beneficial effects as well as multi- Moreover, the nature of the cytotoxic interaction
mechanisms. Several epidemiological studies between ALA and trastuzumab revealed a signifi-
have reported that consumption of flavonoids, cant synergism. Omega-3 fatty acids suppress
and especially catechins, might function as che- overexpression of the HER2 oncogene at the
mopreventive agents against cancer [103]. transcriptional level, which, in turn, interacts
synergistically with anti-HER2 trastuzumab-
based immunotherapy [59].
Gamma-Linolenic Acid (GLA)

GLA is an essential omega-6 fat that is found in Caffeine

evening primrose, black currant seed, borage oil,
and pine seed oil and can inhibit the action of the Knowledge gained by incorporating genetic vari-
cancer gene Her-2/neu. This gene is responsible ation into a nutrition study not only provides a
for almost 30 % of all breast cancers. When can- more rational basis for giving personalized
cer cells that overexpress the Her-2/neu gene are dietary advice but will also improve the quality of
6 Nutrigenomics in Breast Cancer 141

evidence used for making population-based additional drink per week consumed. Binge
dietary recommendations for the prevention of drinking of 4–5 drinks increases the risk by 55 %
specific diseases. Caffeine is metabolized primar- [112]. Moreover, alcohol consumption is related
ily by the cytochrome P450 1A2 (CYP1A2) to promoter methylation of E-cadherin in breast
enzyme, and a polymorphism in the CYP1A2 cancer [113].
gene determines whether individuals are “rapid” There is growing evidence to indicate that
caffeine metabolizers (those who are homozy- dietary fiber, in particular digestion-resistant
gous for the −163 A allele) or “slow” caffeine starch, promotes bowel health, and one of the
metabolizers (carriers of the −163 C allele) [105]. areas of focus for experimental research is its
A similar concept was utilized in an observa- potential protection against the development of
tional study of coffee and breast cancer. The colorectal cancer [114]. Additional studies have
study associated a lower risk of breast cancer shown that butyrate, one of the predominant
among slow metabolizers [106]. No protective short-chain fatty acids produced from the fermen-
effect was observed among fast metabolizers, tation of resistant starch by the gut bacteria, may
implicating caffeine as the protective component be responsible for its physiological effects [115].
of coffee. This study also suggested that caffeine While the cellular effects of butyrate are well doc-
protects against breast cancer in women with a umented, numerous studies have been conducted
BRCA1 mutation and illustrated the importance in order to explain the mechanisms by which
of integrating individual genetic variability when butyrate may elicit its antitumorigenic effects.
assessing diet–disease associations. This is con- There is substantial evidence that alcohol con-
sistent with findings from animal studies show- sumption increases breast cancer risk. In a pooled
ing that caffeine inhibits the development of analysis of the six largest cohort studies with data
mammary tumors [107]. on alcohol and dietary factors [116], the risk for
breast cancer increased monotonically with
increasing intake of alcohol. For a 10 g/day
Alcohol increase in alcohol, breast cancer risk increased
by 9 % (95 % CI = 4–13 %). Adjustment for other
Alcohol has long been considered a risk factor breast cancer risk factors had little impact. Beer,
for breast cancer in women [108]. The wine, and liquor all contribute to the positive
International Agency for Research on Cancer has association, strongly suggesting that alcohol per
declared that there is enough scientific evidence se is responsible for the increased risk. In an
to classify alcoholic beverages a group 1 carcino- intervention study, it is reported that consump-
gen that causes breast cancer in women [109]. tion of approximately one to two alcoholic drinks
Group 1 carcinogens are the substances with the per day increased estrogen levels in premeno-
clearest scientific evidence that they cause can- pausal and postmenopausal women [117], sug-
cer, such as smoking tobacco. A study of more gesting a mechanism by which alcohol may
than 1,280,000 middle-aged British women con- increase breast cancer risk.
cluded that for every additional drink regularly In several large prospective studies, high
consumed per day, the incidence of breast cancer intake of folic acid appeared to mitigate com-
increases by 1.1 % [110]. Approximately 6 % pletely the excess risk for breast cancer due to
(between 3.2 and 8.8 %) of breast cancers alcohol [118, 119]. This relationship was recently
reported in the UK each year could be prevented confirmed using plasma folic acid levels [120].
if drinking was reduced to a very low level (i.e., The public health recommendations for alcohol
less than 1 unit/week). Among women, breast are complicated because consumption of one to
cancer comprises 60 % of alcohol-attributable two alcoholic beverages per day probably pro-
cancers [111]. A study of 17,647 nurses found tects against cardiovascular disease. Because car-
that high drinking levels more than doubled risk diovascular disease is the leading cause of death
of breast cancer with 2 % increase risk for each among women, moderate drinking is associated
142 S. Dwivedi et al.

overall with a modest reduction in total mortality breast cancer cells by preventing p27Kip1-medi-
[121]. However, avoiding alcohol appears to be ated growth arrest [126].
one of relatively few methods for reducing breast
cancer risk, whereas many methods exist to
reduce risk for cardiovascular disease. For Vitamin E
women choosing to consume alcohol regularly,
use of a multivitamin to ensure adequate folic Vitamin E has inhibited mammary tumors in
acid intake may decrease breast cancer risk. rodents in some experiments [127]. In1984, Wald
et al. demonstrated a prospective study of 5,004
women. Thirty-nine of these women who devel-
Lycopene oped breast cancer had lower levels of plasma E
in the blood that had been collected between
Using pangenomic array technology, it has been 1968 and 1975 than did 78 controls [128].
demonstrated that lycopene supplementation Researchers declared a clear association of the
(10 μM–48 H) modulates many molecular path- lower vitamin E levels with a higher risk of breast
ways by affecting the expression of apoptosis cancer. A similar prospective study by Willett
and cell cycle-related genes [122], as well as et al. claimed no such association with cancer
xenobiotic metabolism, fatty acid biosynthesis, when the slightly lower plasma vitamin E levels,
and gap junctional intercellular communication. in those who later developed cancer, were
Further, one other study emphasized that lyco- adjusted for serum cholesterol levels [129].
pene may be a significant dietary element A strong request to support studies to validate the
involved in breast cancer prevention. This study hypothesis that vitamin E can reduce a woman’s
observed upregulation of apoptosis-related genes risk for development of breast cancer has been
such as PIK3C3 and Akt1. PIK3C3 belongs to made by London et al. [130]. It has been sug-
the phosphoinositide (PI)3-kinase family gested that γ-TmT, γ-tocopherol, and δ-tocopherol
involved in both receptor-mediated signal trans- may be involved in inhibiting tumor formation.
duction and intracellular trafficking. PI3K pro- Probable mechanism of actions in inhibiting
teins generate specific inositol lipids involved in breast cancer could be inducing PPARγ expres-
the regulation of cell growth, proliferation, sur- sion and consequently reducing the expression of
vival, differentiation, and cytoskeletal changes. ERα; inducing Nrf2, which consequently reduces
This study also showed upregulation of MAPK- inflammation and oxidative stress; and inhibiting
related genes such as heat shock protein HSPA1B, cell proliferation while inducing apoptosis [131].
the fibroblast growth factor FGF2, and FOS,
a major component of the activator protein-1
(AP-1) transcription factor complex, which Vitamin A
includes members of the JUN family [123]. The
AP-1 transcription factor is a regulator of pro- Vitamin A consists of preformed vitamin A from
cesses essential for normal growth and develop- animal sources, and carotenoids found mostly in
ment as well as carcinogenesis. One of the fruits and vegetables. Many carotenoids are
best-characterized targets of PI3K lipid products potent antioxidants and may provide a defense
is the protein kinase Akt, or protein kinase B against reactive oxygen species that damage
(PKB) [124]. Serine/threonine protein kinase DNA. Vitamin A also regulates cell differentia-
Akt mediates signals from epidermal growth fac- tion and may thus prevent carcinogenesis. In a
tor receptor to the apoptosis-related genes cohort of Canadian women (519 cases) [132],
including BRCA1 through activated Ras and a marginally significant protective association
PI3K in the estrogen and other signaling path- between total vitamin A intake, preformed vita-
ways [125]. Moreover, Viglietto et al. demon- min α- and β-carotene, and breast cancer was
strated that Akt regulates cell proliferation in seen. With 14 years of follow-up in the Nurses’
6 Nutrigenomics in Breast Cancer 143

Health Study (2,697 cases) [133], an inverse in premenopausal women in the Nurses’ Health
association with total vitamin A was seen only Study II [139], intake of animal fat and high-
among premenopausal women. This inverse fat dairy foods was associated with a 33–36 %
association was mainly accounted for by intakes increase in risk for breast cancer for the highest
of β-carotene and lutein/zeaxanthin and was compared with the lowest quintile of intake. Total
strongest among women with a family history of fat per se was not associated with breast cancer
breast cancer. However, in an extended follow-up risk, suggesting that other constituents of dairy
of the Canadian cohort (1,452 cases) and in a foods consumed early in adult life may increase
Swedish cohort (1,271 cases), little overall asso- breast cancer risk.
ciation was seen between intake of carotenoids
and breast cancer [134, 135]. Total Fat
An alternative to the dietary assessment of Preclinical and human ecological studies have
vitamin A intake is the measurement of vitamin suggested an association between increased
A compounds in the blood. In the two largest dietary fat intake and breast cancer risk [140],
studies based on blood samples collected before while cohort studies revealed less consistent
diagnosis [136, 137], low levels of β-carotene effects [141]. For instance, a case-control study
and other carotenoids were associated with an (414 cases and 429 controls) established no asso-
approximately twofold increase in risk for breast ciation between breast cancer and dietary fat
cancer. Thus, available data from observational intake [142]. Similarly, observational studies on
studies suggest a possible protective effect of the influence of dietary fat on breast cancer recur-
vitamin A intake, particularly carotenoids, on rence have produced mixed results [143]. The
breast cancer risk, particularly in premenopausal variable associations may be due to differences
women. Ideally, the effect of vitamin A supple- of fat intake in the study population, difficulty in
ments should be evaluated in randomized trials. accurately measuring fat intake with diet-
However, the β-carotene arm of the Women’s assessment methods, and high correlation
Health Study (a breast cancer-prevention trial between dietary fat and other diet and lifestyle
conducted in 40,000 women) was terminated in variables [144]. Most investigators assume that
1996 after reports that β-carotene supplements any observed tumor-enhancing effect of dietary
appeared to increase the risk for lung cancer fat needs to be adjusted statistically for energy
among smoking men. Thus, data from random- intake. However, this is not a clear-cut issue
ized trials on specific carotenoids and breast can- because a change in fat composition of the diet
cer risk may never be available. may cause alterations in energy intake. Thus,
higher energy intake resulting from changes in
fat intake may be considered as one of the mecha-
Types of Fat nisms by which fat affects tumor development.
Dietary fat may also play a role in the devel-
Specific types of fat could differentially influ- opment of breast cancer via hormone metabo-
ence the risk for breast cancer. In most animal lism. This may be particularly relevant for
studies, diets high in polyunsaturated fat, but ER-positive cancers, as an elevation of endoge-
typically at levels beyond human exposure, have nous estrogen levels with increased fat intake is
evidently increased the occurrence of mammary thought to be related to breast cancer [145].
tumors. A positive association has not been found Alternatively, any role for dietary fat in breast
in prospective epidemiologic studies [138]. In a cancer may be less direct. For example, high-fat
pooled analysis of cohort studies [138], saturated diets may lead to greater body mass or obesity, a
fat (compared with carbohydrate) was weakly probable risk factor for postmenopausal breast
associated with higher risk for breast cancer cancer. In postmenopausal women, high-fat
(RR for 5 % of energy = 1.09; 95 % CI = 1.00– intake may increase levels of bioavailable estro-
1.19). In a recent prospective study conducted gens, thus elevating the risk of breast cancer.
144 S. Dwivedi et al.

Furthermore, higher-fat intake in childhood or study noted a similar absence of association

adolescence may promote faster growth and ear- [139]. Diets high in PUFAs may not be associ-
lier onset of menarche, both established risk fac- ated with breast cancer risk independently of any
tors for breast cancer [145]. contribution to total fat intake. Thus, contrary to
data from animal experiments, human studies do
Saturated Fat not show an increase of breast cancer risk with
Giving more weight to prospective studies, Wakai PUFA intake. Estimating the risk associated with
et al. reported no relationship between saturated PUFA intakes remains difficult as food composi-
fat intake and breast cancer risk in 26,291 sub- tion tables for these fatty acids are incomplete.
jects from the Japan Collaborative Cohort Study
[146]. On the other hand, combined analysis of
12 case-control studies revealed an increased risk Zinc
of postmenopausal breast cancer with higher sat-
urated fat intake, giving an overall OR of 1.57 Zinc (Zn) is an essential trace element required
(p < 0.0001) for the uppermost quintile of intake; for maintaining both optimal human health and
this estimate was adjusted for total fat intake, genomic stability. Zn plays a critical role in the
which was also associated with increased risk regulation of DNA repair mechanisms, cell pro-
[147] and in view of the relationship between liferation, differentiation, and apoptosis involv-
breast cancer and foods high in saturated fat, such ing the action of various transcriptional factors
as meat and dairy products [141, 147]. On the and DNA or RNA polymerases. Zn is an essential
other hand, the observed associations for meat component for more than 1,000 proteins includ-
consumption may reflect a true effect of saturated ing copper/Zn superoxide dismutase (SOD) as
fat [147]. well as a number of other Zn finger proteins. Zn
is an essential cofactor or structural component
Monounsaturated Fatty Acids (MUFAs) for important antioxidant defense proteins and
Olive oil is a rich source of MUFAs, and in case- DNA repair enzymes, such as Cu/Zn SOD,
control studies, it has been revealed that the risk OGG1, APE, and PARP. Thus, it may play an
of breast cancer is decreased with the consump- important role in breast cancer prevention.
tion of more than one teaspoon of olive oil per Recently, it was shown that the migration
day (OR =0.75; 95 % CI: 0.57–0.98) [148]. potential of MDA-MB-231 cells on fibronectin,
Antioxidants present in olive oil, such as vitamin demonstrated in the control, was not affected by
E, have been considered to be one of the protec- a low level of zinc (2.5 μM), but was significantly
tive constituents [149]. However, a meta-analysis inhibited by higher levels of zinc (5–50 μM). Zinc
of 17 case-control and eight cohort studies found at 5–50 μM also reduced magnesium-dependent
no association between MUFAs and breast can- cell adhesion to fibronectin, likely through inter-
cer risk [141]. The relationship between MUFAs fering with magnesium-dependent integrin acti-
intake and breast cancer risk appears to depend vation, and induced cell rounding in the normally
on the contributing foods. elongated, irregular-shaped MDA-MB-231 cell
lines of human breast carcinoma [151].
Polyunsaturated Fatty Acids (PUFAs)
Nkondjock et al. conducted a case-control study
of 414 cases and 429 population-based controls Selenium
and observed no overall association between
PUFAs and breast cancer risk [150]. Similarly, Studies conducted with methylseleninic acid
a combined analysis of 12 case-control studies (MSA), a synthetic mammary cancer chemopre-
indicated no statistically significant association ventive agent, in the rat mammary tumor model
between postmenopausal breast cancer risk and point out that this form of selenium is able to
PUFA intake [147]. Furthermore, another cohort block clonal expansion of premalignant lesions
6 Nutrigenomics in Breast Cancer 145

and induce apoptosis [152]. Similar cellular modifications are epigenetic events that medi-
responses are replicated with human premalig- ate heritable changes in gene expression and
nant breast cells grown in culture using MSA chromatin organization in the absence of
[153]. These investigators characterized the pro- changes in the DNA sequence. Examples of
file of gene expression changes after 4 weeks in large-scale breast cancer studies based on mod-
the whole mammary tissue of rats treated with ern highly sensitive and specific instruments
methylnitrosourea (MNU) and fed MSA [153]. like MS spectrometry microarray, real-time
In a follow-up study, Dong et al. further exam- PCR on related genes like BRCA1 and other
ined the cellular and molecular effects of MSA in single nucleotide polymorphisms and apoptotic
premalignant human breast cells (MCF10AT1 pathways, cell cycle, epigenetic modifications
and MCF10AT3B) [153]. MSA inhibited growth with nutritive compounds that may be negative,
of both cell lines in a dose- and time-dependent positive, or neutral response in their result of
manner, induced apoptosis, and blocked cell cancer treatment, risk prediction, prognosis,
cycle progression at the G1 phase. These and diagnosis.
organoselenium compounds altered additional No single lab will be able to manage the con-
genes (BCL-2, BAD, CYCLIN D1, P27, APO1, cept of a personalized nutrition alone; rather, a
P21, CASPASE-3, CMYC, PCNA) thereby lead- collective effort by the scientific community to
ing to inhibition of cell proliferation and induc- adhere to guidelines put forth regarding experi-
tion of apoptosis. As stated above, modulations mental designs, analysis, and data storage will
of apoptosis and cell proliferation by selenium generate a database that is readily available to
can account for chemoprevention during the researchers and clinicians alike. Thus, all records
post-initiation phase of mammary carcinogene- and databases of nutritional strategies targeted to
sis. Clearly, using mammary adenocarcinomas, the prevention of breast cancer may open new
the results of this study showed that selenium has vistas in better management of breast cancer
an impact on genes that are involved in the multi- patients.
step carcinogenesis process [154]. It is becoming increasingly evident that nutrig-
Recently, human cellular glutathione peroxi- enomics is taking a central stage in the investiga-
dase I was found not only to be a selenium- tion of the effect of nutrition on health outcomes
dependent enzyme that protects against oxidative like breast cancer and that impacts of nutrients
damage and its peroxidase activity but also to be can be evaluated comprehensively by a multitude
associated with cancer risk in the lung and breast of modern “omic” technologies and biomarkers.
[155]. In this view, the future lies not with the technolo-
gies, but with the storage, management, and
interpretation of the immense quantity of metab-
Conclusion and Future Perspective olomic data. The realization that genetic back-
ground, gender, and life stage can have an impact
From the data discussed in this chapter, it is evi- on nutritional requirements is becoming increas-
dent that elucidations of the mechanisms of ingly evident; thus, we are progressing toward
action of bioactive compounds are complex and the personalize diet concept. Translation of this
entail simultaneous examination of alternations knowledge into recommendations based on gen-
in gene expression (transcriptomics), study of otype or at the individual level is only practical in
molecular relationships between nutrients and those few cases (as galactosemia), when the
genes (nutrigenetics), influence changes in the effect of genotype clearly overpowers the impact
profile of proteins (proteomics), study of mul- of any other factor and is the ultimate determin-
tiple signaling and metabolic pathways (metab- ing factor of the nutritional and health status for
olomics), and associations of different nutritive an individual or genetic subgroup. Thus, nutrig-
compounds that exert synergistic, additive, or enomic research is now providing a new ray of
opposing effects. DNA methylation and histone hope in better management of breast cancer.
146 S. Dwivedi et al.

References 18. Loktionov A. Common gene polymorphisms and

nutrition: emerging links with pathogenesis of multi-
factorial chronic diseases. J Nutr Biochem. 2003;14:
1. Boyle P, Levin B. World cancer report 2008.
International agency for research on cancer. http://
19. Adlercreutz H. Phyto-oestrogens and cancer. Lancet
Oncol. 2002;3:364–73.
php. Accessed 26 Feb 2011.
20. Sharma S, Kelly TK, Jones PA. Epigenetics in cancer.
2. Milner JA, Romagnolo DF. Bioactive compounds and
Carcinogenesis. 2010;31:27–36.
cancer. Springer New York Dotrecht Heidelberg
21. Trujillo E, Davis C, Milner J. Nutrigenomics,
London; 2010. doi:10.1007/978-1-60761-627-6_16.
proteomics, metabolomics, and the practice of
3. Miggiano GA, De Sanctis R. Nutritional genomics:
dietetics. J Am Diet Assoc. 2006;106(3):403–13.
toward a personalized diet. Clin Ter. 2006;157(4):
22. Davis CD, Uthus EO. DNA methylation, cancer
susceptibility, and nutrient interactions. Exp Biol
4. Martin KR. Using nutrigenomics to evaluate apopto-
Med. 2004;229:988–95.
sis as a preemptive target in cancer prevention. Curr
23. Munaka M, Kohshi K, Kawamoto T, Takasawa S,
Cancer Drug Targets. 2007;7(5):438–46.
Nagata N, Itoh H, et al. Genetic polymorphisms of
5. Afman L, Muller M. Nutrigenomics: from molecular
tobacco- and alcohol-related metabolizing enzymes
nutrition to prevention of disease. J Am Diet Assoc.
and the risk of hepatocellular carcinoma. J Cancer Res
Clin Oncol. 2003;129:355–60.
6. Magee PJ, Rowland IR. Phyto-oestrogens, their
24. Davis CD, Uthus EO, Finley JW. Dietary selenium
mechanism of action: current evidence for a role in
and arsenic affect DNA methylation in vitro in Caco-2
breast and prostate cancer. Br J Nutr. 2004;91(4):
cells and in vivo in rat liver and colon. J Nutr.
7. DeBusk RM, Fogarty CP, Ordovas JM, Komman KS.
25. Steinmetz KL, Pogribny IP, James SJ, Pitot HC.
Nutritional genomics in practice: where do we begin?
Hypomethylation of the rat glutathione S-transferase
J Am Diet Assoc. 2005;105:589–98.
pi (GSTP) promoter region isolated from methyl
8. Simopoulos AP. Nutrigenetics/nutrigenomics. Annu
deficient livers and GSTP-positive liver neoplasm.
Rev Public Health. 2010;31:53–68.
Carcinogenesis. 1998;19:1487–94.
9. Fenech M, El-Sohemy A, Cahill L, Ferguson LR,
26. Giovannucci E, Chen J, Smith-Warner SA, Rimm EB,
French TA, Tai ES, et al. Nutrigenetics and
Fuchs CS, Palomeque C, et al. Methylenete-
Nutrigenomics: viewpoints on the current status and
trahydrofolate reductase, alcohol dehydrogenase,
applications in nutrition research and practice.
diet, and risk of colorectal adenomas. Cancer
J Nutrigenet Nutrigenomics. 2011;4(2):69–89.
Epidemiol Biomarkers Prev. 2003;12:970–9.
10. Ordovas JM, Corella D. Nutritional genomics. Annu
27. Day JK, Bauer AM, Des Bordes C, Zhuang Y,
Rev Genomics Hum Genet. 2004;5:71–118.
Kim BE, Newton LG, et al. Genistein alters methyla-
11. Lackner DH, Bähler J. Translational control of gene
tion patterns in mice. J Nutr. 2002;132(8 Suppl):
expression from transcripts to transcriptomes. Int Rev
Cell Mol Biol. 2008;271:199–251.
28. Kune G, Watson L. Colorectal cancer protective
12. Goh KI, Cusick ME, Valle D, Childs B, Vidal M,
effects and the dietary micronutrients folate, methio-
Barabási AL. The human disease network. Proc Natl
nine, vitamins B6, B12, C, E, selenium, and lycopene.
Acad Sci U S A. 2007;104:8685–90.
Nutr Cancer. 2006;56:11–21.
13. Legg RL, Tolman JR, Lovinger CT, Lephart ED,
29. Cooney CA, Dave AA, Wolff GL. Maternal methyl
Setchell KD, Christensen MJ. Diets high in sele-
supplements in mice affect epigenetic variation and
nium and isoflavones decrease androgen-regulated
DNA methylation of offspring. J Nutr. 2002;132(8
gene expression in healthy rat dorsolateral prostate.
Reprod Biol Endocrinol. 2008;6:57. doi:10.1186/
30. Lee YW, Klein CB, Kargacin B, Salkinow K,
Kitahara J, Dowjat K, et al. Carcinogenic nickel
14. Go VL, Butrum RR, Wong DA. Diet, nutrition, and
silences gene expression by chromatin condensation
cancer prevention: the postgenomic era. J Nutr.
and DNA methylation: a new model for epigenetic
carcinogens. Mol Cell Biol. 1995;15:2547–57.
15. Anderle P, Farmer P, Berger A, Roberts MA.
31. Hesketh J. Nutrigenomics and selenium gene expres-
Nutrigenomic approach to understanding the mecha-
sion patterns, physiological targets, and genetics.
nisms by which dietary long-chain fatty acids induce
Annu Rev Nutr. 2008;28:157–77.
gene signals and control mechanisms involved in car-
32. Fuchs CS, Willett WC, Colditz GA, Hunter DJ,
cinogenesis. Nutrition. 2004;20:103–8.
Stampfer MJ, Speizer FE, et al. The influence of
16. Fenech M. The genome health clinic and genome
folate and multivitamin use on the familial risk of
health nutrigenomics concepts: diagnosis and epig-
colon cancer in women. Cancer Epidemiol Biomarkers
enome damage on an individual basis. Mutagenesis.
Prev. 2002;11:227–34.
33. Myzak MC, Hardin K, Wang R, Dashwood RH,
17. Sugimura T. Nutrition and dietary carcinogens.
Ho E. Sulforaphane inhibits histone deacetylase
Carcinogenesis. 2000;21:387–95.
6 Nutrigenomics in Breast Cancer 147

activity in BPH-1, LnCaP and PC-3 prostate epithelial binding protein 1 (4E-BP1) phosphorylation by
cells. Carcinogenesis. 2006;27:811–9. watercress: a pilot study. Br J Nutr. 2010;104:
34. Ross SA. Diet and DNA methylation interactions in 1288–96.
cancer prevention. Ann N Y Acad Sci. 2003;983: 49. Wang J, John E, Ingles SE. 5-lipoxygenase and
197–207. 5-lipoxygenase- activating protein gene polymor-
35. Fenech MF. Dietary reference values of individual phisms, dietary linoleic acid, and risk for breast can-
micronutrients and nutriomes for genome damage cer. Cancer Epidemiol Biomarkers Prev. 2008;17:
prevention: current status and a road map to the future. 2748–54.
Am J Clin Nutr. 2010;91:1438S–54. 50. Dimri M, Bommi PV, Sahasrabuddhe AA,
36. Fink BN, Gaudet MM, Britton JA, Abrahamson PE, Khandekar JD, Dimri GP. Dietary omega-3 polyun-
Teitelbaum SL, Jacobson J, et al. Fruits, vegetables, saturated fatty acids suppress expression of EZH2 in
and micronutrient intake in relation to breast cancer breast cancer cells. Carcinogenesis. 2010;31:489–95.
survival. Breast Cancer Res Treat. 2006;98:199–208. 51. Li Y, Ambrosone CB, McCullough MJ. Oxidative
37. Chlebowski RT, Blackburn GL, Thomson CA, stress-related genotypes, fruit and vegetable con-
Nixon DW, Shapiro A, Hoy MK, et al. Dietary fat sumption and breast cancer risk. Carcinogenesis.
reduction and breast cancer outcome: interim efficacy 2009;30(5):777–84.
results from the Women’s Intervention Nutrition 52. Iwasaki M, Hamada GS, Nishimoto IN, Netto MM,
Study. J Natl Cancer Inst. 2006;98:1767–76. Motola Jr J, Laginha FM, et al. Dietary isoflavone
38. Bissonauth V, Shatenstein B, Ghadirian P. Nutrition intake, polymorphisms in the CYP17, CYP19, 17beta-
and breast cancer among sporadic cases and gene HSD1, and SHBG genes, and risk of breast cancer in
mutation carriers: an overview. Cancer Detect Prev.