Vous êtes sur la page 1sur 2

HIGHLIGHTS

G E N O M I C I N S TA B I L I T Y

Translocations, tumours and H2AX


DNA double-strand breaks (DSBs) can be
induced by exogenous agents and during
physiological processes such as lymphocyte
development, but regardless of how they are
induced, lack of their repair can lead to
genomic instability and tumour development.
Now, in two studies from the August 8 issue of
Cell from Frederick Alt and colleagues and
Andre Nussenzweig and colleagues a new
factor has been identified that is important in
suppressing oncogenic translocations initiated
by DSBs.
Histone H2AX has previously been implicated in DSB repair, so both groups generated
mice that were either deleted or heterozygous
for H2AX. Alt reported a modest increase in
cancer predisposition with loss of H2AX,
whereas Nussenzweig did not. However, when
the mice were crossed with a Trp53/ strain,

both groups found that there was a significant


increase in tumour development particularly
of lymphomas. This specific increase in lymphomas would be anticipated if H2AX was
involved in DSB repair as specific DSBs are
created during normal lymphocyte development and its loss would lead to chromosome
translocations.
To further understand the mechanism by
which tumour formation is induced, the two
groups performed spectral karyotyping (SKY)
analysis of the resulting lymphomas.
Nussenzweig found that loss of H2AX resulted
in a general increase in chromosomal translocations, and Alt characterized this further.
They found that Trp53/ thymic lymphomas
did not possess clonal translocations, but one
out of two Trp53/ H2AX+/ and six out of
seven Trp53/ H2AX/ lymphomas displayed

A P O P TO S I S

A deadly combination
p53 is a pro-apoptotic protein that is
required for the programmed death of
tumour cells in response to DNA damage,
and type I interferons (IFN- and IFN-) are
known to be involved in anti-viral immune
responses. Despite the successful use of IFN/ for the treatment of some types of
human cancer, the relationship between p53
and IFN-/ is unknown.
When mouse embryonic fibroblasts
(MEFs) and the hepatic cancer cell line
HepG2 were treated with IFN-/, the level
of p53 protein was increased in a dosedependent manner. IFN treatment did not
affect the half-life of p53, and so does not
increase protein degradation. IFN- induced
the expression of Trp53 mRNA by MEFs,
indicating that gene transcription is
increased. The mouse and human TP53
genes were both shown to contain IFNstimulated response elements (ISREs) in
their promoter or first intron, which are
known to be activated by a transcriptionfactor complex, known as ISGF3, that
contains IFN regulatory-factor 9 (IRF9); p53
induction in response to IFN- was not
observed in Irf9 / MEFs.

634

| SEPTEMBER 2003 | VOLUME 3

IFN- stimulation did not induce serine


phosphorylation and hence did not activate
p53 and had no effect on the induction of
p53 target genes. Therefore, the authors
propose that although type I IFNs do not
activate p53, they increase the sensitivity of
cells to stress stimuli that activate p53 by
increasing the level of p53 protein. They
went on to show that the interaction
between IFN and p53 pathways has
implications for defence against both
tumours and viruses.
The human papilloma virus (HPV)
protein E6 induces the degradation of p53
and, together with another oncoprotein
such as HRAS, can induce the
transformation of primary MEFs. However,
when IFN- was added, the level of p53
protein was restored and there was a marked
decrease in the number of transformed
colonies. Similarly, IFN- enhanced the p53dependent apoptosis of MEFs expressing
adenovirus E1A oncoprotein in response to
X-ray radiation.
In terms of anti-viral responses, MEFs
and HepG2 cells infected with various
viruses were shown to have marked

complex translocations that did not involve


the T-cell receptor loci. The lymphomas that
develop when at least one H2AX allele is
deleted are clearly distinct from those that
arise in a Trp53/ background alone in terms
of the nature of their genomic instability.
Characterization of the B-cell lymphomas
by SKY and fluorescence in situ hybridization
from Trp53/ H2AX/ and Trp53/ H2AX+/
mice revealed further differences. The
pro-B-cell lymphomas that developed in
Trp53/ H2AX/ mice had clonal translocations involving chromosomes 12 and 15,
which caused amplification of IgH and c-Myc,
and Nussenzweig obtained similar results
using comparative genomic hybridization
(CGH). Alt also found that the B-cell lymphomas from Trp53/ H2AX+/ mice all developed from more mature B cells.
The increased tumour incidence in
Trp53/ H2AX+/ compared with Trp53/ mice
indicated that H2AX might be haploinsufficient, and both groups showed this to be correct. Alt showed that, in the B-cell lymphomas, the gene was not deleted or mutated
and that the protein was expressed.

phosphorylation of p53. The apoptosis of


virus-infected cells mediated by p53 was
significantly suppressed in MEFs deficient
for IFN-/ receptor 1 (IFNAR1), whereas
p53 phosphorylation still occurred. This
supports the theory that IFN signalling is
required for enhancement of the p53
response by p53 induction rather than
activation. p53-deficient MEFs infected
with vesicular stomatitis virus gave a higher
virus yield than wild-type MEFs, which
indicates that p53-dependent apoptosis
(enhanced by IFN-/ signalling) is
important for controlling virus replication.
This information could have important
implications for cancer therapy as it indicates
that IFN-treated cells should be more
susceptible to DNA-damaging chemotherapeutic agents such as 5-fluorouracil
(5-FU), allowing lower doses to be used. In
agreement with this, the authors showed that
the death of HepG2 cells was increased by
IFN- at a dose of 5-FU that had minimal
effects on its own.
Kirsty Minton
References and links
ORIGINAL RESEARCH PAPER Takaoka, A. et al.

Integration of interferon-/ signalling to p53 responses in


tumour suppression and antiviral defence. Nature 16 July
2003 (doi: 10.1038/nature01850)
FURTHER READING Vousden, K. H. & Lu, X. Live or let die:
the cells response to p53. Nature Rev. Cancer 2, 592604
(2002) | Katze, M. G., He, Y. & Gale, M. Viruses and
interferons: a fight for supremacy. Nature Rev. Immunol. 2,
675687 (2002)

www.nature.com/reviews/cancer

Nussenzweig showed that the protein was


expressed in the tumours derived from
Trp53/ H2AX+/ mice as well, and both
groups went on to show that the protein level
in H2AX+/ thymocytes was about half that in
H2AX+/+ thymocytes. Both groups also found
that genomic instability in H2AX heterozygotes was intermediate between that in wildtype and null cells. Finally, Nussenzweig found
that the phosphorylation of H2AX on serine
residues 136 and 139 is crucial for its ability to
prevent genomic instability.
So, these mouse models indicate that H2AX
suppresses genomic instability and tumorigenesis. Whether it has the same effect in humans
remains to be established.
Emma Greenwood

TUMORIGENESIS

Stopping the leaks

References and links


ORIGINAL RESEARCH PAPERS Bassing, C. H. et al. Histone

H2AX: a dosage-dependent suppressor of oncogenic


translocations and tumors. Cell 114, 359370 (2003) | Celeste, A.
et al. H2AX haploinsufficiency modifies genomic instability and
tumor susceptibility. Cell 114, 371383 (2003)
WEB SITES

Frederick Alts lab:


http://www.hms.harvard.edu/dms/bbs/fac/alt.htm
Andre Nussenzweigs lab:
http://rex.nci.nih.gov/RESEARCH/basic/eib/nusenzwg.htm

NATURE REVIEWS | C ANCER

Compared with normal blood vessels, tumour


vasculature is extremely leaky. This is thought
to aid tumour progression, as plasma proteins
can enter the surrounding tissue to provide a
fertile environment for tumour growth. In the
July issue of Cancer Cell, William Sessa and
colleagues have now found that plugging the
leaks in tumour blood vessels actually blocks
tumour growth in mice.
The authors have previously shown that
cavtratin a chimeric peptide containing the
scaffolding domain of caveolin-1 linked to a
cellular internalization sequence reduces
vascular permeability in mouse models of
inflammation. Caveolin-1 is the primary coat
protein of caveolae and has been shown to
regulate endothelial nitric-oxide synthase
(eNOS), which controls vascular remodelling
and angiogenesis and is activated by vascular
endothelial growth factor (VEGF). They
began by investigating the effects of cavtratin
on VEGF-mediated vascular leakage by measuring Evans blue-dye extravasation an
indicator of albumin and plasma-protein
leakage after intradermal administration of
VEGF. Mice pretreated with cavtratin showed
a marked reduction in Evans blue-dye
extravasation compared with control mice
pretreated with a control peptide AP-CAV-X.
So, as cavtratin reduces VEGF-mediated vascular permeability, does it have a similar effect
on tumour blood-vessel permeability?
Mice with Lewis-lung-carcinoma (LLC)
tumours were treated with cavtratin or
AP-CAV-X, but only cavtratin inhibited
tumour vascular leakage without affecting vascular permeability in the lungs. As the effects of
cavtratin were tumour specific, they monitored
tumour progression in the LLC model and
found that daily administration of cavtratin
significantly reduced tumour size. Cavtratin,
but not AP-CAV-X, reduced tumour-derived

NOS activity and significantly reduced expression of platelet-endothelial-cell adhesion molecule-1 (PECAM-1) a common marker for
tumour endothelium and the VEGF receptor FLT4 a marker for lymphatic vasculature.
Histological examination identified large areas
of necrosis in cavtratin-treated tumours, and
cavtratin increased the number of apoptotic
cells in non-necrotic regions of the tumour. So,
cavtratin reduces eNOS activity, decreases
tumour permeability and causes apoptosis.
Direct effects of cavtratin on angiogenesis,
endothelial-cell proliferation and tumour-cell
growth were ruled out and the role of eNOS in
vascular leakage was investigated further.
Comparison of LLC tumours in wild-type and
eNOS/ mice showed that tumours from
eNOS/ mice were less permeable to Evans
blue-dye than tumours from wild-type animals. Administration of cavtratin to wild-type
animals reduced tumour permeability to the
same level seen in eNOS/ tumours. In addition, eNOS/ mice have reduced vascular permeability, tumour growth and sensitivity to
cavtratin, compared with wild-type mice,
providing further evidence that eNOS is the
primary molecular target of cavtratin. As treatment of endothelial cells with cavtratin did not
block VEGF-induced autophosphorylation of
the VEGF receptor FLK1 or phosphorylation of
c-SRC, genetic loss of eNOS is comparable to
inhibition by cavtratin.
These results indicate that caveolin-1 is a
novel target for antitumour therapy, as it specifically affects tumour microvasculature.
Emma Croager
References and links
ORIGINAL RESEARCH PAPER Gratton, J.-P. et al. Selective

inhibition of tumor microvasculature permeability by cavtratin


blocks tumor progression in mice. Cancer Cell 4, 3139 (2003)
WEB SITE

William Sessas lab:


http://info.med.yale.edu/pharm/sessa/index.html

VOLUME 3 | SEPTEMBER 2003 | 6 3 5