DESC-1680; No of Pages 9

Journal of Dermatological Science (2007) xxx, xxx—xxx

www.intl.elsevierhealth.com/journals/jods

Reference gene selection for real-time rtPCR in

human epidermal keratinocytes
Danny Allen *, Eleanor Winters, Paul F. Kenna, Pete Humphries,
G. Jane Farrar

Department of Genetics, Trinity College Dublin, Dublin 2, Ireland

Received 5 June 2007; received in revised form 21 September 2007; accepted 11 October 2007

KEYWORDS Summary
RNA interference;
Background: RNA interference represents a powerful tool with which to achieve
Human epidermal
suppression of specific target mRNA. Real-time rtPCR is a useful technique for
keratinocytes;
assessing levels of mRNA expression. Critically, for real-time rtPCR to yield meaningful
Real-time rtPCR;
results, it is necessary to normalise expression of the gene of interest to stably
Gene therapy;
expressed endogenous control genes.
Epidermolysis bullosa
Objectives: The study involved establishing expression profiles for 11 housekeeping
genes in human epidermal keratinocyte cell lines determining their relative stability.
Furthermore, the effect of the presence of shRNA on these expression profiles has
been established.
Methods: Keratinocytes were transfected using lipid-based transfection or AMAXA
nucleofection. Real-time rtPCR was used to establish RNA expression levels. Data
analysis was carried out using geNORM.
Results: When using HaCaT or adult NHEK cells any combination of 8 of the house-
keeping genes would be appropriate for normalisation. In contrast, with juvenile
NHEK cells only 4 of the housekeeping genes were found to be sufficiently stable to be
deemed appropriate for normalisation of expression data. Furthermore data demon-
strated that the expression of housekeeping genes may sometimes be affected by the
induction of the RNAi pathway.
Conclusions: The data obtained highlight the importance of characterising house-
keeping gene expression profiles in each specific cell type prior to choosing a set of
housekeeping genes for expression studies. The results from this study are applicable
to researchers working with human epidermal keratinocytes and the experimental
approach is more broadly applicable to any researcher carrying out real-time rtPCR.
# 2007 Japanese Society for Investigative Dermatology. Published by Elsevier Ireland
Ltd. All rights reserved.

* Corresponding author. Tel.: +353 1 8962482; fax: +353 1 8963848.

E-mail address: danny.allen@tcd.ie (D. Allen).

0923-1811/$30.00 # 2007 Japanese Society for Investigative Dermatology. Published by Elsevier Ireland Ltd. All rights reserved.
doi:10.1016/j.jdermsci.2007.10.001

Please cite this article in press as: Allen D, et al., Reference gene selection for real-time rtPCR in human epidermal
keratinocytes, J Dermatol Sci (2007), doi:10.1016/j.jdermsci.2007.10.001
DESC-1680; No of Pages 9
2 D. Allen et al.
1. Introduction
Use of double-stranded RNA (dsRNA) to modulate
gene expression, or RNA interference (RNAi), has
been adopted extensively subsequent to the break-
through paper of Fire et al. in 1998 demonstrating
potent sequence specific gene silencing in C. ele-
gans [1]. RNA interference is a post-transcriptional,
targeted gene-silencing technology that utilises
double-stranded RNA molecules to degrade messen-
ger RNA (mRNA) containing the same sequence as
the dsRNA [2]. In May 2001 Tuschl and co-workers
demonstrated that the provision of small interfering
RNAs (siRNA) could induce a sequence-specific sup-
pression of gene expression without activating a
general inhibition of translation in mammalian cells
[3]. These siRNAs, in principle, represent a potent
molecular tool to suppress expression of a target
gene [4—10]. Much of the research undertaken in
the field of RNAi in mammalian cells and animals has
utilised chemically synthesised siRNAs. However,
several methods have been developed allowing
for stable production of short hairpin RNA (shRNA)
molecules in cells and in animals [11—14].
Analysis of gene expression profiles has become
an integral part of biological research. Changes in
mRNA expression may be central to developmental
processes, tumorigenesis, molecular diagnosis and
both drug and gene therapeutic treatments
[15—17]. There are several tools available for RNA
detection including, inter alia, northern blotting, in
situ hybridisation, RNase protection assays, and
microarray analysis. Reverse transcription (rt) fol-
lowed by polymerase chain reaction (PCR) is
another, particularly powerful, tool for the detec-
tion of mRNA. Real-time PCR combines amplification
and detection to allow data collection throughout
the PCR process. This is accomplished using one of
several fluorescent systems to correlate PCR pro-
duct concentration to fluorescence intensity [18].
The combining of these two technologies, in the
form of real-time rtPCR, is superior to other meth-
ods of detecting RNA for several reasons including its
sensitivity, specificity and reproducibility [19—21].
Real-time rtPCR requires no post-amplification
manipulation and can produce quantitative data
with a large dynamic range [22]. It is 1000-fold more
sensitive than dot blot hybridisation, 10,000—
100,000-fold more sensitive than RNase protection
assays and in principle can detect a single copy of a
specific transcript [23—25]. Furthermore real-time
rtPCR can be used to distinguish between nearly
identical mRNA [26].
For real-time rtPCR to yield meaningful results, it
is necessary to normalise expression of the gene of
interest to a stably expressed endogenous control
Please cite this article in press as: Allen D, et al., Reference gene selection for real-time rtPCR in human epidermal
keratinocytes, J Dermatol Sci (2007), doi:10.1016/j.jdermsci.2007.10.001
gene [27,28]. While commonly known housekeeping
genes such as Beta-actin (ACTB) and glyceralde-
hyde-3-phosphate dehydrogenase (GAPDH) are still
widely used, there is a growing trend towards utilis-
ing the expression profiles of a wider range of genes
in cell types or tissue types of interest to normalise
expression data [29—35].
The current study involved establishing expres-
sion profiles for 11 putative housekeeping genes
(PPIA, GUSB, TBP, B2M, UBC, ALAS, GAPDH, ACTB,
YWHAZ, HPRT and RPLP) in human epidermal kera-
tinocytes thus determining their relative stability.
Primary juvenile and adult normal human epidermal
keratinocytes (NHEK) were used in this study as
were HaCaT cells. HaCaT cells are a spontaneously
immortalised human epidermal keratinocyte line
[36]. They are one of the most widely studied
keratinocyte cell lines and are capable of prolifer-
ating, differentiating and forming organised epider-
mis both in organotypic culture (in vitro) and when
transplanted onto nude mice (in vivo) [37—39].
Furthermore, the effect of the presence of shRNA
on these expression profiles has been established
demonstrating that the expression of frequently
used housekeeping genes such as GAPDH and ACTB,
in keratinocytes, may be affected by the induction
of the RNAi pathway. The data from this study will be
valuable to researchers working with keratinocytes.
The data obtained provides a platform of informa-
tion for those interested in evaluating expression
profiles in keratinocytes and is a prerequisite to any
RNAi studies in keratinocytes.
2. Materials and methods
2.1. HaCaT cell transfection
HaCaT cells were cultured in DMEM+ (500 ml DMEM,
50 ml foetal calf serum (FCS), 5 ml 100 mM sodium
pyruvate and 5 ml 200 mM L-glutamine) using stan-
dard procedures. All transfections were carried out in
6-well plates. Twenty-four hours prior to transfection
cells were plated in each well of a 6-well plate and
grown under standard conditions. Transfections were
carried out using Lipofectamine 2000 according to
the manufacturer’s instructions (Invitrogen).
2.2. Primary NHEK transfection
Adult NHEK cells (from 30-year-old Caucasian female
breast) and juvenile NHEK cells (from 5-year-old male
foreskin) were cultured in keratinocyte growth med-
ium (500 ml growth media containing 0.4% (v/v)
bovine pituitary extract, 0.125 ng/ml epidermal
growth factor (rec. human), 5 mg/ml human insulin,
DESC-1680; No of Pages 9

Effects of shRNA on housekeeping gene expression 3

0.33 mg/ml hydrocortisone, 10 mg/ml human trans-
ferring, 0.39 mg/ml epinephrine and 0.15 mM CaCl2)
using standard procedures. All nucleofections were
carried out in 6-well plates. Twenty-four hours prior
to nucleofection cells were plated in each well of a 6-
well plate and grown under standard conditions.
Nuclefections were carried out using the AMAXA
nucleofector according to the manufacturer’s
instructions (Amaxa biosystems).
2.3. RNA extraction, quantification and
purity
Twenty-four hours post-transfection, total RNA was
isolated from cells using TRI reagent (MRC) accord-
ing to the manufacturer’s instructions. RNA was then
treated with RNase-free DNase (Promega). RNA
extracted from cells was resuspended in 200 ml of
nuclease free water and stored at 80 8C in 10 ml
aliquots. The RNA concentration and purity (260/
280 and 260/230 ratios) was analysed using a ND-
1000 Spectrophotometer (NanoDrop Technologies).
2.4. Real-time rtPCR
RNA levels were analysed using real-time rtPCR
carried out on a Lightcycler (Roche Diagnostics)
using the QuantiTect SYBR Green rtPCR kit (Qiagen).
Samples were run under the following conditions; a
reverse transcription step (50 8C, 20 min), followed
by an activation step (95 8C, 15 min) and 35 cycles of
an amplification step (94 8C, 15 s; 55 8C, 20 s; 72 8C,

Table 1 Primer sequence of housekeeping genes

Gene
Beta-2-microglobulin
5-Aminolevulinate synthase
Hypoxanthine guanine
phosphoribosyl transferase
Tyrosine 3-monooxygenase/ YWHAZ sense ACTTTTGGTACATTGTGGCTTCAA
tryptophan 5-monooxygenase
activation protein
(zeta polypeptide) Antisense
Ubiquitin C
TATA box binding protein
Large ribosomal protein P0
Beta-glucuronidase
Peptidylprolyl isomerase
A (cyclophilin A)
Glyceraldehyde-3-phosphate
dehydrogenase
Beta-actin
a
Refs. [41,42].
Primer Sequence Function a
B2M sense GATGAGTATGCCTGCCGTGTG Secreted protein involved in
Antisense CAATCCAAATGCGGCATCT MHC class I receptor activity
ALAS sense CTGCAAAGATCTGACCCCTC First and rate-limiting enzyme
Antisense CCTCATCCACGAAGGTGATT of heme biosynthesis
HPRT sense TGACACTGGCAAAACAATGCA Enzyme involved in purine
metabolism where it salvages
Antisense GGTCCTTTTCACCAGCAAGCT purines from degraded DNA
Adapter protein implicated
in several signal transduction
pathways
CCGCCAGGACAAACCAGTAT
UBC sense ATTTGGGTCGCGGTTCTTG Regulatory protein involved
in maintenance of chromatin
structure, ATP-dependent
selective degradation and
Antisense TGCCTTGACATTCTCGATGGT regulation of gene expression
TBP sense CACGAACCACGGCACTGATT Binds specifically to the
TATA box is a subunit of
the eukaryotic transcription
Antisense TTTTCTTGCTGCCAGTCTGGAC factor TFIID
RPLP sense GGCGACCTGGAAGTCCAACT Gene codes for the 60S
Antisense CCATCAGCACCACAGCCTTC acidic ribosomal protein P0
GUSB sense CTCATTTGGAATTTTGCCGATT Plays an important role in
the degradation of dermatan
Antisense CCGAGTGAAGATCCCCTTTTTA and keratan sulfates
PPIA sense CCCACCGTGTTCTTCGACAT Accelerates the folding
Antisense CCAGTGCTCAGAGCACGAAA of proteins
GAPDH sense GAAGGTGAAGGTCGGAGTC Catalyses an important
Antisense GAAGATGGTGATGGGATTTC energy-yielding step in
carbohydrate metabolism
ACTB sense TCACCCACACTGTGCCCATCTACGA An actin isoform involved
Antisense CAGCGGAACCGCTCATTGCCAATGG in cell motility, structure
and integrity

Please cite this article in press as: Allen D, et al., Reference gene selection for real-time rtPCR in human epidermal
keratinocytes, J Dermatol Sci (2007), doi:10.1016/j.jdermsci.2007.10.001
DESC-1680; No of Pages 9

4 D. Allen et al.

4 s). All real-time rtPCR primer sequences for
reverse transcription and activation are detailed
in Table 1.
2.5. Data analysis
The stability of each housekeeping gene was
assessed using geNORM version 3.4. [40]. Analysis
using geNORM was carried out according to the
manual. In short, for every control gene geNORM
calculates the pair wise variation with all other
control genes. The internal control gene-stability
measure, M, is defined as the average pair wise
variation of a particular gene with all other control
genes. Genes with the lowest M values have the
most stable expression. Genes with the highest M
value are then excluded allowing stepwise ranking
of each gene from least to most stable.
3. Results
Cells were cultured in flasks and plated in 6-well
plates 24 h prior to transfection. Treatments were
as follows, 24 wells contained cells only, 24 wells
contained cells transfected with pBlueScript (Stra-
tagene) (between 0.05 and 10 mg) and 24 wells
contained cells transfected with pBluescript con-
taining a luciferase-targeting shRNA (between 0.05
and 10 mg) (non-targeting control shRNA). Twenty-
four hours post-transfection total RNA was isolated
from cells and RNA concentrations and purities
analysed. RNA levels of the housekeeping genes
(Table 1) were analysed using real-time rtPCR. Eva-
luation of the stability of the gene expression pro-
files for each of the housekeeping genes was carried
out using geNORM version 3.4. [40]. The geNORM
applet for MS Excel calculates the gene expression
stability (M) of each reference gene from a data set.
M is a measure of how stable each housekeeping
gene is with respect to the other housekeeping
genes in the data set. A lower M value indicates
the gene expression is more stable and hence is
more reliable for normalisation. The cut-off M value
is typically set to 0.5; genes with M values above this
are considered unreliable for normalisation [40].
Expression of nine of the housekeeping genes
remained stable across the samples, in untrans-
fected HaCaT cells, HaCaT cells transfected with
pBluescript and HaCaTcells treated with pBluescript
containing a luciferase-targeting shRNA (Fig. 1). In
contrast to these nine, the expression of GAPDH and
ACTB was relatively unstable. The data suggests
that GAPDH or ACTB alone, or in fact even if used
together, would be unsuitable for the normalisation
of real-time rtPCR data. However, expression of any
of the other nine housekeeping genes would be
suitable for normalisation of RNA levels.
Expression of all eleven housekeeping genes
remained stable across the samples, in untrans-

Fig. 1 geNORM analysis (HaCaT cells): gene expression stability of the 11 candidate reference genes analysed by the
geNORM program in HaCaT cells. The graph presented is adapted from the program output file. Average expression
stability values (M) of the 11 control genes are plotted from least stable (left) to most stable (right). As shown above
expression of nine of the housekeeping genes remained stable across the samples, in HaCaT cells only (blue), HaCaT
cells + pBlueScript (red) and HaCaT cells + luciferase-targeting shRNA (green). In contrast to these nine the expression of
GAPDH and ACTB were relatively unstable.

Please cite this article in press as: Allen D, et al., Reference gene selection for real-time rtPCR in human epidermal
keratinocytes, J Dermatol Sci (2007), doi:10.1016/j.jdermsci.2007.10.001
DESC-1680; No of Pages 9

Effects of shRNA on housekeeping gene expression 5

Fig. 2 geNORM analysis (adult NHEK cells): gene expression stability of the 11 candidate reference genes analysed by
the geNORM program in adult NHEK cells. The graph presented is adapted from the program output file. Average
expression stability values (M) of the 11 control genes are plotted from least stable (left) to most stable (right). As shown
above expression of all eleven of the housekeeping genes remained stable across the samples, in adult NHEK cells only
(blue). In adult NHEK cells + pBlueScript (red) the expression of ACTB was slightly unstable and in Adult NHEK
cells + luciferase-targeting shRNA (green) the expression of ALAS, GAPDH and ACTB were relatively unstable.

fected adult NHEK cells (Fig. 2). Like HaCaT cells
GAPDH and ACTB were the least stable genes, how-
ever, in this case the M values were below the
typical cut-off value of 0.5. Adult NHEK cells trans-
fected with pBluescript (Fig. 2) were not signifi-
cantly different from untransfected cells with the
exception that ACTB was slightly above the typical
cut-off value of 0.5. In adult NHEK cells treated with
pBluescript containing a luciferase-targeting shRNA
(Fig. 2) expression of eight of the housekeeping
genes remained stable across the samples. In con-
trast to these eight, the expression of GAPDH, ACTB
and ALAS was relatively unstable.
Expression of ten of the housekeeping genes
remained stable across the samples, in untransfected
juvenile NHEK cells (Fig. 3). In contrast to these ten,
the expression of PPIA was relatively unstable. In
juvenile NHEK cells transfected with pBluescript
(Fig. 3) only five of the housekeeping genes remained
stable across the samples. In contrast to these five,
the expression of ACTB, PPIA, UBC, HPRT, RPLP and
GAPDH were relatively unstable. Finally, in NHEK
cells treated with pBluescript containing a lucifer-
ase-targeting shRNA (Fig. 3) expression of only four of
the housekeeping genes remained stable across the
samples. In contrast to these four, the expression of
ACTB, PPIA, UBC, HPRT, RPLP, GAPDH and ALAS was
relatively unstable.
GAPDH and ACTB are unsuitable as housekeeping
genes in untransfected HaCaT cells due to variation
in their levels of expression between the 24 inde-
pendent samples. In HaCaT cells transfected with
pBluescript, at various concentrations, GAPDH and
ACTB expression levels again vary between the 24
independent samples. This variation is across all the
treatments irrespective of plasmid concentration
(data not shown). In contrast, in HaCaTcells treated
with pBluescript containing a luciferase-targeting
shRNA (Fig. 4) there is an increase in expression of
three genes (GAPDH, ACTB and ALAS) as the con-
centration of shRNA increases. With ALAS this
increase in expression is statistically significant with
a treatment of 5.0 mg ( p = 0.0163) and 10.0 mg
( p = 0.0166) of pBluescript containing a lucifer-
ase-targeting shRNA but not at concentrations of
1.0 mg or less ( p = 0.8146). With ACTB this increase
in expression is statistically significant with treat-
ments of 1.0 mg ( p = 0.0209), 5.0 mg ( p = 0.014) and
10.0 mg ( p = 0.0021) of pBluescript containing a
luciferase-targeting shRNA but not at concentra-
tions of 0.5 mg or less ( p = 0.6111). The greatest
change in expression is with GAPDH expression; here
the increase in expression is statistically significant
with treatments of 0.5 mg ( p = 0.0032), 1.0 mg
( p = 0.0312), 5.0 mg ( p = 0.0025) and 10.0 mg
( p = 0.0002) of pBluescript containing a lucifer-
ase-targeting shRNA but not at concentrations of
0.1 mg or less ( p = 0.7727).
In sharp contrast to the change in expression seen
with ALAS, ACTB and GAPDH (Fig. 4) is the stability

Please cite this article in press as: Allen D, et al., Reference gene selection for real-time rtPCR in human epidermal
keratinocytes, J Dermatol Sci (2007), doi:10.1016/j.jdermsci.2007.10.001
DESC-1680; No of Pages 9

6 D. Allen et al.

Fig. 3 geNORM analysis (juvenile NHEK cells): gene expression stability of the 11 candidate reference genes analysed by
the geNORM program in juvenile NHEK cells. The graph presented is adapted from the program output file. Average
expression stability values (M) of the 11 control genes are plotted from least stable (left) to most stable (right). As shown
above expression of ten of the housekeeping genes remained stable across the samples, in juvenile NHEK cells (blue). In
contrast to these ten the expression of PPIA was relatively unstable. As shown above expression of five of the
housekeeping genes remained stable across the samples, in juvenile NHEK cells + pBlueScript (red). In contrast to these
five the expression of ACTB, PPIA, UBC, HPRT, RPLP and GAPDH was relatively unstable. In juvenile NHEK cells + lucifer-
ase-targeting shRNA (green) four of the housekeeping genes remained stable across the samples, in juvenile NHEK cells
(24 wells). In contrast to these four the expression of ACTB, PPIA, UBC, HPRT, RPLP, ALAS and GAPDH were relatively
unstable.

Fig. 4 Relative expression of housekeeping genes in HaCaT cells transfected with various concentrations of luciferase-
targeting shRNA. Graphed above on the X-axis are mean threshold cycle values (Ct value)  standard deviation. Threshold
cycle value refers to the real-time rtPCR cycle number in which the amount of fluorescence begins to increase rapidly,
typically a few standard deviations above background. As a plot of Ct versus template is linear the Ct value is dependent
on the starting amount of RNA in the sample. With increasing amounts of starting RNA fewer cycles are required to
achieve this specific fluorescence value. As shown above there is an increase in expression of the three genes as the
concentration of shRNA increases. With ALAS this increase in expression is significant with a treatment of 5.0 mg of
luciferase-targeting shRNA. With ACTB and GAPDH this increase in expression is significant with a treatment of greater
than 1.0 and 0.5 mg (respectively) of luciferase-targeting shRNA. In contrast to this the level of expression stability of the
other eight housekeeping genes appears to be irrespective of plasmid concentration.

Please cite this article in press as: Allen D, et al., Reference gene selection for real-time rtPCR in human epidermal
keratinocytes, J Dermatol Sci (2007), doi:10.1016/j.jdermsci.2007.10.001
DESC-1680; No of Pages 9
Effects of shRNA on housekeeping gene expression
of expression, irrespective of treatment, of the
other eight housekeeping genes chosen, further
demonstrating that these housekeeping genes
should be used in evaluating expression profiles in
HaCaT cells.
Data from adult NHEK cells again reflects the
expression profiles of HaCaT cells, with variation
in ALAS, ACTB and GAPDH expression again increas-
ing compared to untransfected or pBluescript trans-
fected treatments (data not shown). However, in
adult NHEK cells the change in expression with
GAPDH expression at varying concentrations of
pBluescript containing a luciferase-targeting shRNA
is not quite statistically significant ( p = 0.0879).
The data with juvenile NHEK cells differs from that
seen with adult NHEK cells and HaCaTcells. Six of the
eleven tested housekeeping genes were deemed
unsuitable for normalisation in both sets of trans-
fected juvenile NHEK cells. In this case the instability
is seen in both the pBluescript transfected NHEK cells
and the pBluescript containing a luciferase-targeting
shRNA transfected NHEK cells suggesting an effect
caused by the exogenous DNA or the nucleofection
procedure. Furthermore the actual concentration of
luciferase-targeting shRNA appeared to have no sig-
nificant effect on expression of any of the house-
keeping genes in juvenile NHEK cells (data not
shown), although potential concentration dependent
effects may have been masked by the large variations
in expression across the samples.
4. Discussion
From this study it can be seen that putative ‘house-
keeping genes’ cannot be assumed to be stably
expressed in all cell and tissue types. However it
is necessary to normalise expression data with a
stably expressed endogenous control gene for
real-time rtPCR to yield results that are meaningful.
Hence, analyses such as that provided here involving
semi-quantitative RNA analysis are essential at the
outset of any study in a new cell or tissue type. The
study allows researchers working with the same cell
type (i.e. human epidermal keratinocytes) to select
or exclude the use of housekeeping genes based on
the data presented. From the data it is clear that
any combination of PPIA, GUSB, TBP, B2M, UBC,
YWHAZ, HPRT and RPLP would be appropriate to
use as housekeeping genes when using HaCaT or
adult NHEK cells. In contrast, juvenile NHEK cells
appear more susceptible to unstable expression of
housekeeping gene subsequent to plasmid transfec-
tion and/or nucleofection. Indeed only the expres-
sion of TBP, B2M, YWHAZ and GUSB was found to be
sufficiently stable (under these experimental con-
Please cite this article in press as: Allen D, et al., Reference gene selection for real-time rtPCR in human epidermal
keratinocytes, J Dermatol Sci (2007), doi:10.1016/j.jdermsci.2007.10.001
7
ditions) to be deemed to be appropriate for normal-
isation of expression data. Notably these four
housekeeping gene were deemed to be suitable as
housekeeping genes to normalise expression data in
all three cells types under all conditions. The data
obtained from juvenile NHEK cells highlights the
importance of characterising housekeeping gene
expression profiles in each specific cell type prior
to choosing a set of housekeeping genes for expres-
sion studies.
Furthermore, it was observed that the presence
of shRNA effects the expression of some housekeep-
ing genes in both adult NHEK and HaCaT cells.
Notably this change was only statistically significant
in HaCaT cells. This may reflect the homogeneity in
the spontaneously immortalised HaCaT cell line
compared to the relative heterogeneity in a popula-
tion of harvested adult human epidermal keratino-
cytes. This has implications for shRNA suppression
studies where normalisation using a housekeeping
gene that is in fact fluctuating in expression may
yield results that are inaccurate. It is worth noting
that the trend observed in adult NHEK and HaCaT
cells was significantly different to that observed in
juvenile NHEK cells. Whilst the presence of plasmid
and/or the nucleofection procedure altered expres-
sion profiles in juvenile NHEK cells (Fig. 3), the
specific alterations in expression profiles observed
in the presence of shRNA in both adult NHEK and
HaCaT cells was not mirrored in juvenile NHEK cells
or was possibly masked by the changes in gene
expression profiles that occurred due to presence
of pBluescript and/or the nucleofection procedure.
The results from this study are applicable to
researchers working with human epidermal kerati-
nocytes and the experimental approach is more
broadly applicable to any researcher carrying out
real-time rtPCR.
Acknowledgements
Special thanks to Prof. Fusenig at the DKFZ for the
gift of the HaCaT cells. The research was supported
by DEBRA Ireland.
References
[1] Fire A, Xu S, Montgomery MK, Kostas SA, Driver SE, Mello CC.
Potent and specific genetic interference by double-stranded
RNA in Caenorhabditis elegans. Nature 1998;391(6669):806—
11.
[2] Ullu E, Djikeng A, Shi H, Tschundi C. RNA interference:
advances and questions. Philos Trans R Soc Lond B Biol Sci
2002;357:65—70.
DESC-1680; No of Pages 9
8 D. Allen et al.
[3] Elbashir SM, Harborth J, Lendeckel W, Yalcin A, Weber K,
Tuschl T. Duplexes of 21-nucleotide RNAs mediate RNA
interference in cultured mammalian cells. Nature
2001;411:494—8.
[4] Caplen NJ, Parrish S, Imani F, Fire A, Morgan RA. Specific
inhibition of gene expression by small double-stranded RNAs
in invertebrate and vertebrate systems. Proc Natl Acad Sci
2001;98:9742—7.
[5] Paddison PJ, Caudy AA, Bernstein E, Hannon GJ, Conklin DS.
Short hairpin RNAs (shRNAs) induce sequence-specific silen-
cing in mammalian cells. Genes Dev 2002;16:948—58.
[6] Krichevsky AM, Kosik KS. RNAi functions in cultured mam-
malian neurons. Proc Natl Acad Sci 2002;99:11926—9.
[7] Hasuwa H, Kaseda K, Einarsdottir T, Okabe M. Small inter-
fering RNA and gene silencing in transgenic mice and rats.
FEBS 2002;532:227—30.
[8] Palfi A, Ader M, Kiang AS, Millington-Ward S, Clark G, O’Reilly
M, et al. RNAi-based suppression and replacement of rds-
peripherin in retinal organotypic culture. Hum Mutat
2006;27(3):260—8.
[9] Kiang AS, Palfi A, Ader M, Kenna PF, Millington-Ward S, Clark
G, et al. Toward a gene therapy for dominant disease:
validation of an RNA interference-based mutation-indepen-
dent approach. Mol Ther 2005;12(3):555—61.
[10] Millington-Ward S, McMahon HP, Allen D, Tuohy G, Kiang AS,
Palfi A, et al. RNAi of COL1A1 in mesenchymal progenitor
cells. Eur J Hum Genet 2004;12(10):864—6.
[11] Allen D, Kenna PF, Palfi A, McMahon HP, Millington-Ward S,
O’Reilly M, Humphries P, Farrar GJ. Development of strate-
gies for conditional RNA interference. J Gene Med
2007;9(4):287—98.
[12] Brummelkamp TR, Bernards R, Agami R. A system for stable
expression of short interfering RNAs in mammalian cells.
Science 2002;296:550—3.
[13] Sui G, Soohoo C, Affar el B, Gay F, Shi Y, Forrester WC, et al. A
DNA vector-based RNAi technology to suppress gene expres-
sion in mammalian cells. Proc Natl Acad Sci 2002;99:5515—
20.
[14] Yu JY, DeRuiter SL, Turner DL. RNA interference by expres-
sion of short-interfering RNAs and hairpin RNAs in mamma-
lian cells. Proc Natl Acad Sci 2002;99:6047—52.
[15] Mackay IM, Arden KE, Nitsche A. Real-time PCR in virology.
Nucleic Acids Res 2002;30(6):1292—305.
[16] Mackay IM. Real-time PCR in the microbiology laboratory.
Clin Microbiol Infect 2004;10(3):190—212.
[17] Radonic A, Thulke S, Mackay IM, Landt O, Siegert W, Nitsche
A. Guideline to reference gene selection for quantitative
real-time PCR. Biochem Biophys Res Commun 2004;313(4):
856—62.
[18] Higuchi R, Fockler C, Dollinger G, Watson R. Kinetic PCR
analysis: real-time monitoring of DNA amplification reac-
tions. Biotechnology 1993;11(9):1026—30.
[19] Bustin SA. Absolute quantification of mRNA using real-time
reverse transcription polymerase chain reaction assays. J
Mol Endocrinol 2000;25(2):169—93.
[20] Bustin SA. Quantification of mRNA using real-time reverse
transcription PCR (RT-PCR): trends and problems. J Mol
Endocrinol 2002;29(1):23—39.
[21] Ginzinger DG. Gene quantification using real-time quanti-
tative PCR: an emerging technology hits the mainstream.
Exp Hematol 2002;30(6):503—12.
[22] Morrison TB, Weis JJ, Wittwer CT. Quantification of low-
copy transcripts by continuous SYBR Green I monitoring
during amplification. Biotechniques 1998;24(6):954—8.
960, 962.
[23] Wang T, Brown MJ. mRNA quantification by real time Taq-
Man polymerase chain reaction: validation and comparison
Please cite this article in press as: Allen D, et al., Reference gene selection for real-time rtPCR in human epidermal
keratinocytes, J Dermatol Sci (2007), doi:10.1016/j.jdermsci.2007.10.001
with RNase protection. Anal Biochem 1999;269(1):198—
201.
[24] Malinen E, Kassinen A, Rinttila T, Palva A. Comparison of
real-time PCR with SYBR Green I or 50 -nuclease assays and
dot-blot hybridization with rDNA-targeted oligonucleotide
probes in quantification of selected faecal bacteria. Micro-
biology 2003;149(Pt. 1):269—77.
[25] Palmer S, Wiegand AP, Maldarelli F, Bazmi H, Mican JM, Polis
M, et al. New real-time reverse transcriptase-initiated PCR
assay with single-copy sensitivity for human immunodefi-
ciency virus type 1 RNA in plasma. J Clin Microbiol
2003;41(10):4531—6.
[26] Wong ML, Medrano JF. Real-time PCR for mRNA quantitation.
Biotechniques 2005;39(1):75—85.
[27] Bustin SA, Benes V, Nolan T, Pfaffl MW. Quantitative real-
time RT-PCR–—a perspective. J Mol Endocrinol 2005;34(3):
597—601.
[28] Huggett J, Dheda K, Bustin S, Zumla A. Real-time RT-PCR
normalisation; strategies and considerations. Genes Immun
2005;6(4):279—84.
[29] O’Gorman GM, Park SD, Hill EW, Meade KG, Mitchell LG,
Agaba M, et al. Cytokine mRNA profiling of peripheral blood
mononuclear cells (PBMC) from trypanotolerant and trypa-
nosusceptible cattle infected with Trypanosoma congolense.
Physiol Genomics 2006;28(1):53—61.
[30] de Brouwer AP, van Bokhoven H, Kremer H. Comparison of 12
reference genes for normalization of gene expression levels
in Epstein-Barr virus-transformed lymphoblastoid cell lines
and fibroblasts. Mol Diagn Ther 2006;10(3):197—204.
[31] Bogaert L, Van Poucke M, De Baere C, Peelman L, Gasthuys F,
Martens A. Selection of a set of reliable reference genes for
quantitative real-time PCR in normal equine skin and in
equine sarcoids. BMC Biotechnol 2006;6:24.
[32] Ohl F, Jung M, Radonic A, Sachs M, Loening SA, Jung K.
Identification and validation of suitable endogenous refer-
ence genes for gene expression studies of human bladder
cancer. J Urol 2006;175(5):1915—20.
[33] Gilsbach R, Kouta M, Bonisch H, Bruss M. Comparison of in
vitro and in vivo reference genes for internal standardization
of real-time PCR data. Biotechniques 2006;40(2):173—7.
[34] Taylor DL, Thomson PC, de Silva K, Whittington RJ. Valida-
tion of endogenous reference genes for expression profiling
of RAW264.7 cells infected with Mycobacterium avium
subsp. paratuberculosis by quantitative PCR. Vet Immunol
Immunopathol 2007;115(1—2):43—55.
[35] Robinson TL, Sutherland IA, Sutherland J. Validation of
candidate bovine reference genes for use with real-time
PCR. Vet Immunol Immunopathol 2007;115(1—2):160—5.
[36] Boukamp P, Petrussevska RT, Breitkreutz D, Hornung J,
Markham A, Fusenig NE. Normal keratinization in a sponta-
neously immortalized aneuploid human keratinocyte cell
line. J Cell Biol 1988;106(3):761—71.
[37] Schoop VM, Mirancea N, Fusenig NE. Epidermal organization
and differentiation of HaCaT keratinocytes in organotypic
coculture with human dermal fibroblasts. J Invest Dermatol
1999;112(3):343—53.
[38] Maas-Szabowski N, Starker A, Fusenig NE. Epidermal tissue
regeneration and stromal interaction in HaCaT cells is
initiated by TGF-alpha. J Cell Sci 2003;116(Pt 14):2937—48.
[39] Breitkreutz D, Schoop VM, Mirancea N, Baur M, Stark HJ,
Fusenig NE. Epidermal differentiation and basement mem-
brane formation by HaCaT cells in surface transplants. Eur J
Cell Biol 1998;75(3):273—86.
[40] Vandesompele J, De Preter K, Pattyn F, Poppe B, Van Roy N,
De Paepe A, et al. Accurate normalization of real-time
quantitative RT-PCR data by geometric averaging of multiple
internal control genes. Genome Biol 2002;3(7).
DESC-1680; No of Pages 9

Effects of shRNA on housekeeping gene expression 9

[41] Rebhan M, Chalifa-Caspi V, Prilusky J, Lancet D. GeneCards: [42] Maglott D, Ostell J, Pruitt KD, Tatusova T. Entrez Gene: gene-
encyclopedia for genes, proteins and diseases. Rehovot, centered information at NCBI. Nucleic Acids Res
Israel: Weizmann Institute of Science, Bioinformatics Unit 2005;33:D54—8 [database issue].
and Genome Center; 1997URL:http://www.genecards.org/.

Available online at www.sciencedirect.com

Please cite this article in press as: Allen D, et al., Reference gene selection for real-time rtPCR in human epidermal
keratinocytes, J Dermatol Sci (2007), doi:10.1016/j.jdermsci.2007.10.001