ª
Oncogene (2002) 21, 8994 – 9007
2002 Nature Publishing Group All rights reserved 0950 – 9232/02 $25.00
www.nature.com/onc
The cancer connection: BRCA1 and BRCA2 tumor suppression in mice and
humans
Mary Ellen Moynahan*,1
1
Department of Medicine, Memorial Sloan-Kettering Cancer Center, 1275 York Avenue, New York, NY 10021, USA
BRCA1 and BRCA2 mutations are estimated to be
responsible for the great majority of familial breast and
ovarian cancers. Much progress has been made toward
the understanding of the function of these proteins
through genetic, biochemical, and structural studies.
The embryonic lethality encountered in the knockout
mouse initially hindered the development of mouse
models aimed at studying tumor suppression. However,
mice that harbor hypomorphic Brca1 and Brca2 alleles
and cre-mediated tissue-specific deletions for Brca1 and
Brca2 have been generated. Mice deficient for either
Brca1 or Brca2 sustain a wide range of carcinoma and
mammary epithelium deleted for Brca1 or Brca2 is
highly susceptible to mammary tumorigenesis. Mammary
(and other) tumors occur at long latency as compared to
oncogene-induced mouse tumors. p53 deficiency is highly
cooperative with both Brca1 and Brca2 in promoting
tumorigenesis. Analysis of Brca1-associated mammary
tumors reveals significant similarities to BRCA1-associated breast cancer in regard to high tumor grade,
hormone receptor negativity, a high incidence of p53
mutations and genetic instability.
Oncogene (2002) 21, 8994 – 9007. doi:10.1038/sj.onc.
1206177
Keywords: BRCA1; BRCA2; tumor suppression; genetic instability
Introduction
Tumor suppression by BRCA1, located on chromosome 17q21, was established by linkage analysis of
human cancer families, prior to positional cloning of
the gene (Easton et al., 1993; Hall et al., 1990; Miki et
al., 1994). A similar, albeit more expeditious, approach
was used to identify the BRCA2 tumor suppressor on
chromosome 13q12 (Wooster et al., 1994). The
observation that BRCA1 and BRCA2 form nuclear
foci with Rad51 during S phase and following DNA
damage was the key to forming a link with DNA
repair (Chen et al., 1998; Scully et al., 1997). Repair
assays of Brca1 and Brca2 mutant cells confirmed
defects in the homologous repair of chromosomal
double-strand breaks (DSBs) (Moynahan et al., 1999,
*Correspondence: ME Moynahan; E-mail: [email protected]
2001b). Homologous recombination is a fundamental
cellular process for generating diversity during meiosis
and precise repair of chromosomal breaks in mitotic
cells; therefore, it was somewhat disquieting that
homologs were not identified in yeast. The recent
observations that homologs of particularly conserved
domains of BRCA1 and BRCA2 are found in lower
eukaryotes suggest that the functions of these proteins
are fundamental for genomic stability (reviewed by
Jasin, current issue).
Although BRCA1 and BRCA2 are both involved in
homology-directed repair (HDR), BRCA1 appears
omnipresent with a wealth of biochemical data
describing multi-protein interactions (Welcsh and King,
2001). BRCA1 has a broader role in cellular functions
than BRCA2, despite a similarity in cellular phenotype.
Functions ascribed to BRCA1 include transcriptional
regulation that includes both p53-dependent and
independent responses (MacLachlan et al., 2002;
Monteiro, 2002; Welcsh et al., 2002), ubiquitin ligase
activity when dimerized to BARD1 (Baer and Ludwig,
2002), and damage associated phosphorylation of
BRCA1 by multiple kinases that precedes repair –
complex formation (Cortez et al., 1999; Lee et al.,
2000; Tibbetts et al., 2000). The function of BRCA2
appears more straightforward; BRCA2 is central to
HDR repair through BRC-mediated Rad51 interactions and direct binding to single-stranded DNA
(Davies et al., 2001; Yang et al., 2002).
A precise determination as to the function most
likely to safeguard against tumorigenesis is not yet
proven. One common hypothesis relates to a striking
phenotype, genetic instability, which is observed in cells
deficient in BRCA1 or BRCA2 as well as in cells
mutant for any gene involved in the homologous repair
of DSBs. This instability manifests as spontaneous
chromosome aberrations and hypersensitivity to agents
that produce DNA DSBs such as ionizing radiation
and interstrand cross-linking agents (reviewed in Jasin,
current issue). Although genetic instability is observed
in most human solid tumors, the rapidity with which
cells deficient for Brca1 and Brca2 accumulate
chromosomal aberrations is striking (Turner et al.,
2002; Tutt et al., 2002; Xu et al., 2001b), and the extent
of chromosomal aberrations in Brca1-associated
mammary tumors exceeds that of oncogene-induced
tumors (Weaver et al., 2002). That accumulating
genetic change, in the form of chromosomal aberra-
Brca1 and Brca2 tumor suppression
ME Moynahan
8995
tions and resultant loss and gain of function of other
genes, is the trigger for BRCA-associated tumorigenesis
is supported by animal models with late-onset tumorigenesis, tumor multiclonality and relatively few
cancers arising per cell at risk. These mouse models
of tumorigenesis and their parallels to human BRCA1
and BRCA2-associated tumorigenesis will be reviewed.
Mouse models and Brca-associated tumorigenesis
Brca1
Deficiency of Brca1 results in early embryonic lethality
in the majority of mutations predicted to result in
protein truncations that delete the C-terminal half of
the protein (Deng and Brodie, 2001). Mid-embryonic
lethality was noted in the Brca1D223-763 and Brca1D11
embryos which express a Brca1 splice product that
deletes exon 11 but maintains the N-terminal RING
domain and the C-terminal BRCT repeats and in the
Brca11700T embryos which express a Brca1 product
truncated in the C-terminus before the second BRCT
repeat (Deng and Brodie, 2001; Hohenstein et al.,
2001). p53 mutation extended embryonic viability in
the Brca1 null mice by 1 – 2 days and permitted live
births in mice that express the Brca1 exon 11 deleted
splice product (Cressman et al., 1999a; Xu et al.,
2001b) (Table 1; Figure 1).
Prolonged breeding of Brca1D223-763/+ and p53+/7
heterozygote mutant mice resulted in three viable
homozygous double knockout mice that survived to
adulthood. These mice had severe growth defects and
acquired thymic lymphomas and a hemangiosarcoma,
tumors typical of p537/7 mice (Cressman et al., 1999a).
Tumor latency and life span were decreased in the
homozygous double knockout as compared to p537/7
mice. Interestingly, post-natal viability was accomplished in the Brca1D11/D11 mice by Brca1 cre-mediated
deletion in germ cells by crossing Brca1flox11 mice with
EIIA-cre mice (Xu et al., 2001b). The cre-mediated
Brca1D11/D11 deleted mice had 1% viability; however,
further details regarding the life span and development
of the few animals that survived were not reported.
Crossing to p537/7 and p53+/7 mice resulted in near
complete rescue of viability. Twenty-two female
Brca1D11/D11/p53+/7 mice were followed, with most
developing mammary carcinoma at a latency of 6 – 10
months. Additionally, in this group of mice, seven
lymphomas and two ovarian carcinomas were identified.
Recently, a viable Brca1-deficient homozygous
mouse (Brca1Tr/Tr) line was generated on both 129/Sv
and 129/Sv/MF1 genetic backgrounds; whereas the
same mutation generated on the C57Bl/6 genetic
background resulted in embryonic lethality (Ludwig
et al., 2001a). The 50 bp exon 11 insertion was
predicted to result in a truncated product at amino
acid 924; however, the truncated transcript could not
be detected. There was no detectable full-length Brca1
product, but the endogenous Brca1 splicing variant
lacking exon 11 was expressed at wild-type levels.
Typically, this splice variant is weakly expressed. Like
the Brca1D223-763/D223-763/p537/7 mice, Brca1Tr/Tr males
were infertile. A wide spectrum of tumors arose in both
male and female adult Brca1Tr/Tr mice at a significant
frequency (85%), with the majority occurring in mice
with an MF1 genetic background. This was significantly different from the 26% incidence of spontaneous
tumors in the control mice. Lymphoma occurred with
variable latency, ranging from 1 – 24 months, whereas
non-lymphoid tumors began occurring after 9 months,
with an average latency of 18 months for all nonlymphoid cancers and a 15-month breast-cancerspecific latency. Carcinomas of multiple epithelial
tissues occurred; most common were mammary and
lung, but also included endometrial and colon. In
addition, sarcomas, hepatocellular carcinomas and
adenomas were observed. As with the Brca1D223-763/
D223-763
/p537/7 mice, the double knockouts of Brca1Tr/
7/7
Tr
/p53
resulted in a shorter life span and decreased
tumor latency as compared to p537/7 single mutant
mice (Cressman et al., 1999a).
Developmental analysis in the Brca1D223-763/D223-763/
p537/7 was notable for male infertility with defective
spermatogenesis. Visceral development appeared
normal. However, developmental abnormalities were
noted in the epithelium of mammary, parotid, and
prostate glands. Mammary development was the most
significantly altered, with decreased ductal branching,
fewer primary ducts, and underdeveloped end-bud
formation. Fewer developmental defects were observed
in the parotid and prostate gland, but the findings were
notable for atrophy of salivary acinar epithelium and
cytomegaly in both glands. Skin abnormalities were
also apparent with alopecia, graying, and atrophy of
dermal appendages (Cressman et al., 1999a). In the
female Brca1D11/D11/p53+/7 and the Brca1Tr/Tr mice,
normal mammary gland development was reported
(Ludwig et al., 2001a; Xu et al., 2001b).
Brca2
Similar to Brca1, Brca2 deficiency results in early
embryonic lethality in the mouse when disrupting
mutations produce truncations that occur 5’ to exon
11 (Ludwig et al., 1997; Sharan et al., 1997). However,
partial viability is obtained when a truncated Brca2
product retains Rad51 interacting BRC repeat
sequences (Connor et al., 1997; Friedman et al.,
1998). The Brca2tm1Cam and Brca2Tr2014 homozygous
mice have mutations that result in partial viability with
truncating Brca2 mutations occurring within the BRC
repeat domain. These mice had similar phenotypes. All
succumb to thymic lymphoma with markedly decreased
life spans. Both males and females are infertile (Table
2; Figure 2).
Brca2 mutant mice that lack exon 27, the Brca2D27/D27,
have been recently reported (McAllister et al., 2002).
These mice have a subtle degree of embryonic/peri-natal
lethality, appear developmentally normal, and have a
shortened life span. The shortened life span was
determined to be related to an increase in tumorigenesis.
Overall, there was a modest 2.5-fold increase in the
Oncogene
Brca1 and Brca2 tumor suppression
ME Moynahan
8996
Table 1 Homozygous deficiency of mouse Brca1
Mutation
position/
effect
Viability
(EL=embryonic
lethality)
Exon 2
insertion
Exon 5
truncation
Exon 11
insertion
Exon 11
spliced
product
Exon 11
insertion
EL 7.5d
Brca1D223-763/p537/7
Exon 11
spliced
product
3 viable mice
(2 males,
1 female)
Brca1D11/D11
Exon 11
spliced
product
EL 12.5+
1% viability
EL
EL
Brca1 mutation –
homozygous
ex2
Brca1
5-6
Brca1
Brca1ko
D223-763
Brca1
Brca1117/7
Brca1D11/D11/Bax7/7
Brca1D11/D11/Cdkn1a7/7
Brca1D11/D11/p537/7
Brca1D11/D11/p53+/7
Brca1Tr/Tr
Life
span
Fertility
EL 7.5d
EL 7.5 – 8.5d
EL 10 – 13.5d
EL 8.5 – 9.5d
EL – C57BL/6;
4% viable
129/C57BL;
100% 129/129;
100% 129MF1
10 – 12
weeks
Male
infertility
shorter
Male
infertility
EL 10.5d
Cressman
et al., 1999a
Severe growth
retardation
Xu et al.,
2001b
Normal mammary
development
85% tumorigenesis
(76/89); lymphoma;
sarcoma; carcinoma of
mammary, lung, liver,
endometrial and colon
Mild growth defect;
kinky tails; skin
pigmentation
defect; normal
mammary
development
thymic lymphoma
(8 mice analysed)
Truncation
after
BRCT1
Severe growth
retardation; defects
in mammary
development and
spermatogenesis
Mammary carcinoma;
few lymphoma/ovary
Brca1Tr/Tr/p537/7
incidence of spontaneous tumorigenesis with tumors
occurring in 61% of the Brca2D27/D27 mice as compared
to 24% of heterozygote and wild-type mice. Similar to
the Brca1Tr/Tr mice, a wide spectrum of tumors was
observed including carcinomas, lymphomas, sarcomas,
and adenomas. Most notable was the finding that
carcinoma only arose in the mutant Brca2D27/D27 mice
(11 tumors in nine mice out of 41 analysed), whereas no
carcinomas occurred in the heterozygote or wild-type
mice. The 22% incidence of carcinoma occurred in
mammary, gastric, endometrial, and lung epithelial cells.
There was long tumor latency, with tumors occurring in
the 12th to 17th month and only one-third of the tumors
were detected prior to animal sacrifice.
Although mammary tissue was not highly susceptible
to tumorigenesis, this observation may be due to the
129/C57Bl/6 genetic background as these strains rarely
develop spontaneous or carcinogen-induced mammary
carcinogenesis (Kuperwasser et al., 2000). Supporting
this hypothesis is the Brca1Tr/Tr mouse model,
Oncogene
Thymic lymphomas;
1 hemangiosarcoma
lymphoma, sarcoma,
carcinoma of breast and
lung (7 mice analysed)
Brca1
Reference
Shen et al.,
1996
Brca1Tr/Tr/p53+/7
1700T
Development
Ludwig et al.,
1997
Hakem et al.,
1996
Liu et al.,
1996
Gowen et al.,
1996
65 – 80%
viability
Exon 11
insertion
Tumorigenesis
Ludwig et al.,
2001a
Hohenstein
et al., 2001
discussed above, where tumors were predominantly
seen on the 129/Sv/MF1 outbred genetic background
as compared to the 129/Sv strain (Ludwig et al.,
2001a).
Analysis of mammary development in the Brca2D27/D27
mice was notable for a dramatic increase in ductal
side branching and an overall decreased density of
ducts. However dams were able to nurse their pups
despite diminished development of the mammary
gland.
Tumorigenesis in conditional Brca1 and Brca2 mutations
Several mouse models have been created which mutate
either Brca1 or Brca2 by cre-mediated recombination
in specific tissues. Brca1 and Brca2 deletion restricted
to mammary epithelium was achieved using the highly
restricted WAP (whey acidic protein) promoter and the
somewhat less specific MMTV-LTR (mouse mammary
tumor virus-long terminal repeat) to direct expression
Brca1 and Brca2 tumor suppression
ME Moynahan
8997
Figure 1 Wild-type Brca1, the Brca1 mouse mutations and the predicted protein products. A 92kd exon 11 deleted Brca1 product
is expressed in many of the mutant mouse models by alternative splicing. aa encoded by exon 11 are depicted in blue
Table 2 Homozygous deficiency of mouse Brca2
Brca2 mutation
homozygous
Mutation
position
Viability
Brca2Brdml
exon 10
EL 7.5d
Brca2ex11
5’exon 11
EL 8.5d
129B27/7
BALB/cB27/7
exon 10
EL 8.5d
EL 10.5d
Brca2Tm1Cam
truncation
after BRC3
10%
50% (outbred)
no
14 weeks
thymic lymphoma
not assessed
Friedman et al.,
1998
Brca2Tr2014
truncation
after BRC7
20% average;
6% 129/B10,
adults 0%;
30% 129/B6/DBA
adults 17%
no
longest
5.5 months;
avg 15
weeks
thymic lymphoma;
latency 11 – 22
weeks
small, kinky tail,
underdeveloped
organs, no male
germ cells or
ovary follicles
Connor et al.,
1997
Brca2F11
deletion
of exon 11
EL 8.5 – 9.5d
Brca2D27/D27
exon 27
deletion
67%
Fertility
Life
span
Tumorigenesis
Mammary/other
development
Reference
Sharan et al.,
1997
Ludwig et al.,
1997
Bennett et al.,
2000a
Jonkers et al.,
2001
yes
shortened
of cre (Ludwig et al., 2001b; Xu et al., 1999a). In
addition, mouse models with tissue-restricted cremediated deletion of Brca1 in thymocytes directed by
the Lck promoter and of Brca2 exon 11 in epithelium
directed by the human keratin 14 (K14) gene promoter
were generated (Jonkers et al., 2001; Mak et al., 2000)
(Table 3).
All three models that delete Brca1 or Brca2 in
mammary epithelium sustain mammary tumors.
increased
lymphoma,
sarcoma, carcinoma
latency 65+weeks
decreased sidebranching and
overall decreased
density
McAllister et al.,
2002
However, mammary tumorigenesis in the K14cre/
Brca2F11/F11 mice was dependent on the loss of p53
(Jonkers et al., 2001). An important feature to keep in
mind in these conditional animal models is that cremediated deletion of Brca1 or Brca2 is dependent on
the activity of the promoter that is driving cre
expression and can be quite variable. In addition,
cre-mediated deletion is active only in the mammary
epithelial cells and not in surrounding stromal tissue.
Oncogene
Brca1 and Brca2 tumor suppression
ME Moynahan
8998
Figure 2 Wild-type Brca2, the Brca2 mouse mutations and the predicted protein products. The Rad51 interacting BRC repeats are
within exon 11, and the DSS1 interacting and DNA binding domain is in the C-terminal end
In the Brca1Ko/Co/Wap-cre and Brca1Ko/Co/MMTV-cre
mice, the extent of cre-mediated recombination was
analysed during pregnancy when both promoters
would be expected to be most active. Brca1 transcripts
were variably decreased ranging from 10 – 70% as
compared to controls (Xu et al., 1999a). In the
Brca2flox/7/Wap-cre mice, reported by Ludwig et al.,
32% of the alleles underwent cre-mediated recombination following the first pregnancy, however with
subsequent pregnancies the presence of the floxed allele
increased (Ludwig et al., 2001b). In the K14cre/
Brca2F11/F11 mice, expression of cre by the K14
promoter resulted in recombination in only 5 – 30%
of mammary epithelium (Jonkers et al., 2001). The
timing and strength of cre expression in the conditional
models may significantly alter the incidence and latency
of tumorigenesis as well as the potential for developmental abnormalities.
Analysis of mammary gland development in the
Brca1Ko/Co/MMTV-cre mice during pregnancy revealed
incomplete ductal and alveolar development, with
incomplete fat pad filling and increased apoptosis.
The magnitude of the developmental defect was
correlated to the amount of cre-mediated deletion,
whereas those glands that had extensive apoptosis and
filling of less than 50% of the fat pad had transcripts
that were reduced by 60 – 80%. Surprisingly, mothers
were able to nurse their litters (Xu et al., 1999a). No
developmental defects were noted in the K14cre/
Brca2F11/F11 or the Brca2flox/7/Wap-cre mice (Jonkers
et al., 2001; Ludwig et al., 2001b).
The first mouse model to directly demonstrate the
tumor suppressor activity of Brca1 was reported by
Xu et al., with the creation of the Brca1Ko/Co/
Oncogene
MMTV-cre and Brca1Ko/Co/Wap-cre mice which
deletes exon 11 upon cre expression in mammary
epithelium (Xu et al., 1999a). As noted above, this
deletion can be alternatively spliced, resulting in
expression of an exon 11 deleted Brca1 product.
The initial report on a small number of mice was
notable for five mammary tumors occurring in 23
mice at a latency of 10 – 13 months. Molecular
analysis of the tumors revealed reduced intensity of
the unrecombined Brca1 allele, suggesting that most
of the cells had undergone cre-mediated deletion.
Analysis of additional mice was reported in a followup study, confirming an incidence of mammary
tumorigenesis of 25% in 150 mice followed over a
2-year period with an average tumor latency of 11
months (Brodie et al., 2001). Generation of Brca1Ko/Co/
p53+/7 revealed both a striking increase in tumor
incidence, from 25 to 100%, and a decrease in tumor
latency. The number of mice with multifocal tumors or
metastatic sites was not altered by p53 heterozygosity.
Two Brca1Co/Co mice and three Brca1Co/Co/p53+/7 mice
had metastasis to lung, liver, and lymph node.
The first conditional knockout of Brca2 in mammary
epithelium, Brca2flox/7/Wapcre mice, created a mutation in exons 3 and 4. Twenty-six experimental and 12
control multiparous female mice were analysed for
mammary tumor formation over a 20-month period.
Twenty out of 26 Brca2flox/7/Wapcre mice developed
mammary tumors (77%), compared to no mammary
tumor development in the control animals. Palpable
tumors were first observed at 13 months. Analysis of
the sacrificed mice revealed significant tumor multifocality, with 10 of the 20 mice having 2 – 4 tumors
(Ludwig et al., 2001b).
Brca1 and Brca2 tumor suppression
ME Moynahan
8999
Tissue-restricted deletion of Brca1 or Brca2
Table 3
Conditional
mutation
Cre
promoter
Tissue-restricted
expression
Brca1Ko/Co
(Brca1D11)
Wap-Cre
Mammary
epithelium-pregnancy
MMTVCre
Mammary and
salivary gland
Lck-Cre
Thymocytes and
peripheral T cells
Brca15-6
Crosses
Development
Tumorigenesis
Decreased mammary
ductal development
and increased
apoptosis
Breast (22%; 5/23); no
difference between
Wapcre and MMTVcre
mice;
Breast (25%; 35/150);
24% multifocal
Breast (73%; 8/11)
Breast (100%; 56/56)
24% multifocal
p53+/7
90% decrease in
thymocyte #,
increase in
apoptosis, decrease
in proliferation
Thymoma
p537/7
Normal thymocyte
#; (increase in
chromosome breaks
and instability)
Thymoma
p217/7
Persistent decrease
in thymocyte #
None
Em-Bcl-2-36
Tg
Normal thymocyte
#; (increase in
chromosome breaks
and instability)
None
Tumor
latency
10 – 15
months; 11
months avg
Reference
Brodie et al.,
2001;
Xu et al.,
1999a
6 – 8 months
Hakem and
Mak, 2001;
Mak et al.,
2000
Decreased
latency
Brca2flox
Wap-Cre
Mammary
epithelium-pregnancy
Not reported
Breast (77%; 20/26);
50% multifocal
508 days
Ludwig et al.,
2001b
Brca2F11/F11
(exon 11
deleted)
K14-Cre
Skin, mammary and
salivary gland
epithelial cells
(also tongue,
esophagus, forestomach and thymus)
Fertile, normal
mammary
development
Few non-epithelial
tumors (no diff.
control mice)
800 days
Jonkers et al.,
2001
Brca2F11/F11
p53F2-10/F2-10
Mammary (72%)
multifocal in situ; skin
181 days
Brca2F11/+
p53F2-10/F2-10
Mammary (55%); skin
298 days
Mammary (31%); skin
360 days
F11/F11
Brca2
F2-10/+
p53
A second conditional mutation was created by cremediated deletion of Brca2 exon 11 (Jonkers et al.,
2001). The predicted Brca2 product would be devoid of
the Rad51 interacting BRC domains in exon 11 but
should retain N-terminal and C-terminal sequences
that may possess other functional domains, such as
DNA binding (Yang et al., 2002). This mutation
resulted in loss of essential Brca2 functions, as cre
expression in the germ-line resulted in embryonic
lethality. As noted above, the K14 promoter used in
these mice to drive cre expression was weakly expressed
in mammary epithelium. The K14cre/Brca2F11/F11
female mice did not incur mammary tumors; however,
breeding to p53F2-10/F2-10 mice resulted in mammary
tumors with an incidence of 77% for double-mutant
mice and 55% for K14cre/Brca2F11/F11/p53+F2-10
heterozygote mice. Additionally, Brca2F11/+/p53F2-10/
F2-10
mice containing a wild-type Brca2 allele had a
31% incidence of mammary tumors.
Confirmation that both p53 and Brca2 are tumor
suppressors for breast cancer was shown by molecular
analysis of the 21 Brca2F11/F11/p53F2-10/F2-10 mammary
tumors. All four conditional alleles had undergone
cre-mediated deletion. However, in the Brca2F11/+/
p53F2-10/F2-10 mammary tumors, several had lost the
wild-type and conditionally deleted Brca2 alleles, but
nearly half had retained both alleles. The decreased
incidence and increased latency of tumorigenesis
observed in this conditional model, when only one of
these tumor suppressor genes is mutated, supports the
hypothesis that multiple genetic changes are required
for Brca2-associated tumorigenesis. No significant
metastatic potential was observed in the Brca2flox/7
or Brca2F11/F11 mice.
Oncogene
Brca1 and Brca2 tumor suppression
ME Moynahan
9000
Tumor suppression by Brca1 and Brca2 in the mouse
appears to be a more general phenomenon than in
humans, although predisposition to multiple cancers
has been associated with Brca2 (discussed below).
Carcinoma, lymphoma and sarcoma arose in the
hypomorphic Brca2ex27 and the Brca1Tr mouse models.
In the conditional models, the K14 promoter directs
expression of cre in skin and salivary gland epithelium
as well as upper alimentary tract epithelium and
thymus. Squamous cell carcinoma of skin was
frequently observed, as were rare salivary gland
tumors. In the Lck-cre/Brca1f5-6 T-cell specific deletion,
a predisposition to thymoma was observed (Hakem
and Mak, 2001).
Pathology of Brca-associated mouse mammary tumors
Several reports that have documented an increased
predisposition to cancer in the Brca-deficient mouse
models have provided detailed pathologic descriptions,
assessments of genomic instability, and analysis of genes
commonly associated with human breast cancers. The
pathology of human BRCA1 and BRCA2-associated
breast cancer reveals significant differences between the
two types and in particular for BRCA1-associated
breast cancers as compared to sporadic breast cancers
(reviewed in Phillips, 2000). Distinctive gene expression
profiles for both BRCA1 and BRCA2 associated breast
and ovary cancers have been reported (Hedenfalk et al.,
2001; Jazaeri et al., 2002; van’t Veer et al., 2002). One
significant purpose of mouse models of disease is to
recapitulate the human disease process, as an understanding of the pathogenesis of the tumors arising from
these mouse models may yield important biologic
pathways and therapeutic strategies that can be
extended to human disease prevention and treatment.
The pathology of human BRCA1-associated breast
cancers is most commonly described as high grade,
hormone-receptor negative, Her2 non-overexpressing
infiltrating ductal adenocarcinomas. The initial
description of the mammary tumors arising from the
Brca1Ko/Co conditional mouse model was of histologic
diversity and genetic instability (Xu et al., 1999a). The
tumors arising from this model have undergone
subsequent molecular analysis (Brodie et al., 2001;
Weaver et al., 2002). In addition, Ludwig et al.
(2001a,b) have provided a comprehensive pathologic
and immunohistochemical analysis on mammary
tumors arising in both the Brca1Tr mouse and the
conditional Brca2flox mouse.
Breast tumors arising in the Brca1Tr mouse were
notable for extremely heterogenous histologic patterns
and variability in nuclear atypia. Both expansile and
invasive tumors were observed. Immunohistochemical
analysis for human breast cancer related genes were
notable for uniformly positive expression of cyclinD1
and p21, whereas a majority of the cancers stained
positive for p53 and negative for the Her2, estrogen
and progesterone receptors (Ludwig et al., 2001a). The
follow-up study reporting on larger numbers of the
Oncogene
conditionally deleted Brca1Co mouse confirmed diverse
histology, nuclear atypia, a high mitotic index and
negative staining for the estrogen receptor (Brodie et
al., 2001). No histologic differences were noted to be
attributable to tumors arising from a p53-deficient
genetic background as compared to Brca1-deficiency
alone.
A candidate gene analysis of Her2 (erbB2), cMyc
and CyclinD1, was specifically undertaken, since these
genes are often over-expressed and/or amplified in
human breast cancers. Immunohistochemial (IHA)
and/or Western blot analysis for expression was shown
for nine Brca1Co tumors, three of which were derived
from the p53+/7 cross (Brodie et al., 2001). All tumors
were notable for demonstrable but variable expression
of cyclinD1. The authors report over-expression of
Her2 and cMyc in the majority of tumors; however,
the evidence for over-expression by Western analysis
was modest in most and quite variable. In human
breast tumors, it has become clear that the degree to
which Her2 is overexpressed is important for selecting
those patients most likely to benefit from therapies
directed against the Her2 pathway. Typically, only
those patients who have intense (3+) Her2 immunostaining of HER2 gene amplification benefit from antiHer2 therapies (Seidman et al., 2001). Expression of
cell cycle regulators was notably absent for p16 and
present for cdc2, p21 (weak), and p27 expression.
BRCA2-associated human breast tumor and sporadic breast cancer have similar histology and
immunohistochemical profiles, with a significant
proportion being hormone receptor positive. However,
BRCA2 breast tumors are more likely to be of
moderate to high grade (Phillips, 2000). Unlike
Brca1-associated mouse mammary tumors, the Brca2associated tumors were much less heterogenous. The
Brca2flox conditional mammary deletion gave rise to
grossly palpable tumors, with only two histological
patterns (solid carcinoma and adenosquamous carcinoma), with an occasional tumor containing both types
(Ludwig et al., 2001b). The tumor cells were uniform
with mild nuclear pleomorphism, with the exception of
two tumors that had significant nuclear atypia. IHA of
the tumors revealed 50% estrogen receptor positive,
greater than 90% progesterone receptor negative,
100% Her2 negative, 100% cyclinD1 positive, 100%
p21 positive and 70% p53 positive. Sequence analysis
of p53 was performed on five tumors, three of which
contained point mutations.
Pathology from the mammary tumors arising in the
K14cre/Brca2F11/F11/p53F2-10/F2-10 female mice was
notable for multiple in situ lesions, high nuclear grade,
and an expansive growth pattern (Jonkers et al., 2001).
One difference in this conditional model, that is likely
attributable to the cell specificity of the keratin 14
promoter, was that carcinoma arose in the myoepithelial/basal cell types. Basal epithelial and myoepithelial
cells are typically positive for keratin 14 staining (Jones
et al., 2001). The most common type of tumor was a
‘baseloid’ carcinoma with a monomorphic cell type,
and with pseudo-glandular and occasional squamous
Brca1 and Brca2 tumor suppression
ME Moynahan
9001
differentiation. Baseloid cancer is seen infrequently in
humans. However it was recently reported to be a
variant of adenoid cystic carcinoma with a propensity
for lymph node metastases (Shin and Rosen, 2002).
Few mammary cancers arose in the Brca2ex27 mouse,
although adenosquamous carcinoma was observed,
similar to one of the dominant histologies observed
in the Brca2flox mammary cancers (McAllister et al.,
2002).
Genomic instability
A phenotypic hallmark in cells mutated for genes
involved in DNA double-strand break repair is
spontaneous chromosome instability. Genetic instability is commonly observed in human tumors. However,
in cells with defective chromosome break repair, this is
an early and likely precipitating event. Dramatic
chromosome aberrations were noted in Brca1 and
Brca2-deficient embryonic tissue (Tutt et al., 2002; Xu
et al., 2001b) and in cells deficient in Brca1 and Brca2
(Moynahan et al., 2001a; Shen et al., 1998; Turner et
al., 2002; Tutt et al., 1999; Yu et al., 2000).
Tumor cells were derived from six primary Brca2flox
tumors to assess for genetic instability by metaphase
karyotype analysis. The tumor cells demonstrated
various degrees of aneuploidy, chromosome aberrations,
and fragments. Notably, the variability observed in
chromosome instability did not give rise to varying
histology (Ludwig et al., 2001b). Three primary p53+/7/
Brca1Co/Co tumor cell lines were aneuploid with
associated structural aberrations (Brodie et al., 2001).
The observed genomic instability in initial metaphase
spreads was further investigated using a newer and more
sensitive cytogenetic technique called spectral karyotyping (SKY) (Weaver et al., 2002). This method uses
multicolor chromosome paints, that can be used to more
accurately assess chromosome aberrations and allow the
identification of the chromosomes contributing to the
rearrangements. Seven tumors (5-Brca1Ko/Co and 2p53+/7/Brca1Ko/Co) were analysed, six at early passage
number. All the tumors displayed numerical and
structural chromosome aberrations, including chromatid
and chromosome breaks, insertions, deletions, and rearrangements. Two were largely clonal, whereas other
tumors had different degrees of multiclonality, indicating
ongoing genomic instability.
Comparative genomic hybridization (CGH) was used
to evaluate chromosomal gains and losses in 11
Brca1Ko/Co and four Brca1Ko/Co/p53+/7 mammary
tumors (Weaver et al., 2002). An average of eight
chromosome copy alterations/tumor genome were
observed despite diverse histology. CGH performed
on the Brca1Ko/Co/p53+/7 tumors revealed no increase
in chromosomal instability above that observed for
Brca1Ko/Co tumors. This degree of genomic alteration
is greater than that measured for oncogene-driven
mouse mammary tumors, and in the range of that seen
with aneuploid human breast cancers. The hypothesis
that loss of BRCA functions results in genomic
instability arising from unrepaired or misrepaired
chromosomal breaks suggests that a random pattern
of chromosome alterations might be observed.
However, a pattern of non-random, recurring chromosome alterations was noted in the 15 Brca1 mammary
tumors with a gain of distal chromosome 11 (9/15), a
partial gain of chromosome 15 (8/15), loss of distal
chromosome 14 (6/15), and gain of the X chromosome
(4/15). Four of the 15 also had a loss of proximal
chromosome 11, which is a region orthologous to
human 17p and may include the p53 gene. Specific
genes important in human carcinogenesis are mapped
to these regions of chromosomal gain (chr 15-cMyc)
and chromosomal loss (chr 14-Rb). Thus far, the
pathology features and the genetic alterations observed
in the Brca1Co mouse mammary tumor appears to be a
close, if not true, ‘model’ for human BRCA1associated breast cancers and may yield important
additional insights into this disease with further study.
Modifiers of tumorigenesis – the p53 gene
The embryonic lethality originally described in the
Brca1 and Brca2 knockouts could be modulated by
p53. Loss of p53 allowed the embryos to survive a day
or two longer. This was thought to be due to both the
abrogation of apoptosis and the relief from a
proliferation block caused by p53-dependent p21
activation of the G1-S checkpoint. It is likely that
unrepaired chromosome breaks that arise in proliferating cells trigger p53 activation. However, the activation
and outcome of p53-dependent processes may be celltype dependent. Breeding of the Brca1D11/+ mice to
Bax7/7 and Cdkn1a7/7 mice could not rescue the
lethality of Brca1D11/D11 embryos, as did the cross to
p53-deficient mice, which implies a role for multiple
p53 pathways in the rescue of the Brca1-deficient
embryos (Xu et al., 2001b). However, restoration of the
T-cell population in the conditionally deleted Lck-cre/
Brca1f5-6 mouse occurred by intercross to either p53deficient or Bcl-2 transgenic mice, but was not rescued
by breeding to the Cdkn1a7/7 mice (Mak et al., 2000).
The abrogation of one or both of these processes
may be important in permitting tumorigenesis. As was
noted in the Brca1 and Brca2 mouse models that have
been crossed to p53-deficient mice, p53 loss accelerated
tumor formation and increased tumor incidence. The
effect of p53 heterozygosity was dramatic, with
typically greater than a doubling of tumor incidence.
A fourfold increase in mammary tumorigenesis was
observed when the conditional Brca1Co mouse was
bred to a p53 heterozygote mouse, whereby the tumor
incidence increased from 25 – 100% and average tumor
latency decreased by three months (Brodie et al., 2001).
The majority of tumors had lost the wild-type p53
allele. In both Brca1D11/+ and Brca1D223-763/+ mutations, p53-deficiency was required for post-natal
survival of the Brca1 homozygous mutant mouse;
therefore, the role of p53 in promoting tumorigenesis
cannot be measured. The Brca1Tr/Tr mouse mutation
had a high incidence of tumorigenesis even in the
Oncogene
Brca1 and Brca2 tumor suppression
ME Moynahan
9002
absence of a p53 intercross; however, tumor latency
was decreased by approximately 6 months in the
Brca1Tr/Tr/p53+/7 mice (Ludwig et al., 2001a).
Mammary tumors occurred only with p53 deficiency
in the K14-cre/Brca2F11/F11 mice (Jonkers et al., 2001).
In all tumors that arose in the Brca2F11/F11/p53F2-F10
background the wild-type p53 allele had been lost. The
effect of genetic p53-deficiency has not yet been
reported for the conditional Brca2flox mammary
mutation or the Brca2ex27 deleted mouse model. It will
be of interest to see if the magnitude of the p53 effect
with these mutations is as striking as those currently
described.
Loss of wild-type p53 or positive p53 expression by
immunohistochemical staining has been reported in the
majority of Brca-associated mammary tumors
analysed. However, this was not a universal finding
and therefore cannot be considered a prerequisite, per
se, to Brca-associated mouse mammary tumorigenesis.
Positive immunostain for p53 expression assumes a p53
stabilizing point mutation. There is limited sequence
analysis data to identify p53 point mutations. When it
was examined in two studies, only 3/5 tumors
examined confirmed a point mutation in p53 (Ludwig
et al., 2001b; Weaver et al., 2002).
The significance of p53 status in human BRCA1associated breast cancer is difficult to ascertain. Some
studies report that mutated TP53 is strongly
associated with BRCA-associated breast cancer,
whereas others have not found the frequency of p53
alterations to be different from that found in sporadic
breast cancers as measured by IHA (Phillips, 2000).
TP53 is reported to be mutated in 20 – 50% of
sporadic breast cancer. However, other studies report
that 85% of human breast cancers have p53
dysfunction as determined by the presence of TP53
allelic loss and/or p53 protein overexpression
(Janocko et al., 2001). A recent re-analysis of a data
set of BRCA-associated breast cancers in comparison
to control cancers revealed a complex pattern of p53
immunostaining. A similar percentage of all cancers
were negative for p53 (39 – 45%). However, more
BRCA1 tumors demonstrated strong p53 immunoreactivity as measured by both intensity and number
of cells stained (BRCA1, 34%; control, 16%) and
fewer BRCA1 tumors demonstrated weak p53
immunostaining (BRCA1, 8%; control, 26%).
BRCA2 breast cancers were not significantly different
from control tumors for p53 immunostaining (Lakhani et al., 2002).
A provocative early study demonstrated novel TP53
mutations arising in BRCA1-associated breast tumors
(Crook et al., 1998). Functional studies performed
with these novel TP53 mutations revealed a separation of function where transactivation, checkpoint,
and apoptosis remained intact. However, the mutants
were unable to suppress transformation (Smith et al.,
1999). A recent analysis pooled data from p53
mutations reported in 12 studies in BRCA1 and
BRCA2-associated breast cancer and compared the
TP53 mutations in these cancers to those reported to
Oncogene
the TP53 somatic mutation data base (Greenblatt et
al., 2001). Both p53 overexpression and TP53
mutation were significantly more frequent in BRCAassociated cancers than in sporadic cancers. However,
out of 82 TP53 mutations, 58 were missense
mutations, 19 of which had never been reported in
breast cancer and nine had never been reported in
any cancer, confirming the initial report of novel
TP53 mutations in BRCA1-associated breast cancer.
Over 80% of TP53 missense mutations reported to
the database are in hot-spot codons. These hot-spot
mutations abrogate wild-type p53 function, and the
majority are located in important p53 structural
domains. However, over a third of the BRCAassociated TP53 mutations were in non hot-spot
codons. Computational modeling based on the solved
p53 crystal structure found that a majority of these
mutations were located outside these important
structural domains.
p53 and BRCA1 are intimately linked in the DNA
damage response through transcriptional regulation
(MacLachlan et al., 2002), common signaling pathways
and protein stability (Xu et al., 2001b). Given these
interactions and the frequency with which novel
BRCA-associated TP53 mutations are observed, the
precise contribution of p53 to BRCA-associated
tumorigenesis may be somewhat different from that
occurring in tumors where the DNA damage response
is less perturbed.
Haploinsufficient Brca1 and Brca2 mouse models
Lack of a phenotype in heterozygote knockout mice
has been reported by most of the groups that
attempted to derive null animals (Tables 4 and 5).
However, the length of follow-up and the degree to
which the heterozygote animals were examined varied.
There have been no reports of tumor predisposition in
animals followed through 1 – 2 years of age. Attempts
to induce tumor formation by cross-breeding with
other mice and exposure to DNA-damaging agents
have been performed. No tumors were detected 1 year
after full-body irradiation of the Brca1+/1700T mice as
compared to wild-type controls. Crosses to p53+/7,
p537/7, Msh2D7n and Apc1638N mutant mice also
revealed no difference in tumor incidence between
wild-type and Brca11700T mice (Hohenstein et al.,
2001). The Brca1+/D223-763 mouse was crossed to
p53+/7 and p537/7 mice. Expectantly, thymic
lymphomas arose in both the p537/7/Brca1+/D223-763
and the p537/7/Brca1+/+ mice with no difference in
tumor number, latency, or metastatic potential (Cressman et al., 1999b). Crosses obtained with p53
heterozygous mice revealed no difference in life span
or overall tumor incidence. Four mammary cancers
occurred in the p53+/7/Brca1+/D223-763 mice as
compared to one mammary cancer in the p53+/7/
Brca1+/+ mice; however, molecular analysis of three of
the mammary cancers revealed that none had lost the
wild-type Brca1 allele.
Brca1 and Brca2 tumor suppression
ME Moynahan
9003
Table 4
Brca1 mutation –
heterozygote
Length of
follow-up
Development
Brca15-6/+
11 months
normal
ex2/+
Brca1
15 months
normal
Brca1ko/+
10 months
normal
3 months
normal
5 months
normal
6 months
no difference
mammary/
reproductive
development
none
mammary –
decreased ductal
branching;
increased ovarian
atrophy, follicular
arrest and ovary
inflammation
normal
no mammary
tumors
Brca1D11/+
vs Brca1+/+
Crosses
ApcMin/+
Carcinogen
exposure
Brca1 haploinsufficiency in mouse
ENU
Brca1D223-763/+
Brca1D223-763/+
vs Brca1+/+
no DES
DES
Brca1D223-763/+
vs Brca1+/+
2 years
5 Gy XRT
7/7
p53
p53+/7
4 months
2 years
normal
normal
5 Gy XRT
Brca1D11/+
p53+/7
Brca11700T/+
5 Gy XRT
p53+/7
p537/7
Msh2D7n
Apc1638N
Life
span
Tumor
development
Reference
none
Hakem et al.,
1996
Ludwig et al.,
1997
Liu et al.,
1996
mammary
squamous ca
(irrespective of
Brca1 genotype)
Karabinis et al.,
2001
Gowen et al.,
1996
no difference
no tumors
no difference
no tumors
no difference
no difference
thymic lymphoma
various tumors (all
retain wt Brca1)
Bennett et al.,
2000b
Cressman et al.,
1999b
5 vs 0 mammary
cancers (3/5 lost wt
Brca1 allele)
1 year
normal
none
2 years
normal
no
1 year
normal
no
A further analysis examined the effect of ionizing
radiation damage on the p53+/7/Brca1+/D223-763 and
p53+/7/Brca1+/+ mice. No difference in survival was
observed. Interestingly, no mammary cancers arose in
17 control mice, but five mammary cancers arose in 21
p53+/7/Brca1+/D223-763 mice, three of which had lost
the wild-type Brca1 allele. No statistical significance
could be ascertained due to the small sample size. All
tumors had lost the wild-type p53 allele. In ENUtreated ApcMin/+ mice, a mouse line that is highly
predisposed to developing mammary cancers, Brca1
heterozygosity had no effect on tumor incidence,
latency, or multiplicity (Karabinis et al., 2001).
Although the repair of ENU-damaged DNA, which
produces DNA base modifications without DSBs, may
not be a particularly suitable model in which to
examine the effects of Brca1 haploinsufficiency.
Xu et al.,
2001b
Hohenstein et al.,
2001
Medina et al. (2002) employed a mammary transplant approach to examine the effects of p53 deficiency
and Brca2 haploinsufficiency on tumor incidence and
latency by transplantation of p53 null mammary cells
on a BALB/c genetic background. As compared to
Brca2 wild-type, Brca2 haploinsufficiency resulted in a
modest increase in mammary tumor incidence in glands
arising from p53 null transplanted mammary cells. A
decrease in tumor latency was not observed (Medina et
al., 2002).
Mammary and reproductive tract development in
Brca1 and Brca2 heterozygote mice were carefully
analysed with and without DES exposure (Bennett et
al., 2000b). Subtle decreases were found in the
complexity of mammary development in untreated
Brca2 heterozygous mice as compared to wild-type
mice. Following DES administration, similar differOncogene
Brca1 and Brca2 tumor suppression
ME Moynahan
9004
Table 5
Brca2 mutation –
heterozygote
Crosses
Carcinogen
exposure
Brca2Brdml/+
ex11/+
Brca2
129B2+/7
vs 129+/+
no DES
Brca2 haploinsufficiency in mouse
Length of
follow-up
Development
8 months
normal
8 months
normal
1 sq cell of skin
6 months
variable; mammary –
decreased side
branching and alveolar
bud formation
none
subtle decreased side
branching and alveolar
bud formation; uterine
adenomyosis;
increased ovary
follicular arrest
no mammary
tumors
DES
BALB/cB2+/7
mammary transplant
experiment
Brca2+/7/p537/7
vs
Brca2+/+/p537/7
Brca2Tr2014/+
12 months
normal
Brca2Tm1Cam/+
10 months
Normal mammary and
ovary development
Brca2+/7
vs Brca2+/+
Apc+/7
ENU
Tumor
development
Reference
Sharan et al.,
1997
Ludwig et al.,
1997
Bennett et al.,
2000b
small increase in
mammary
carcinogenesis
Medina et al.,
2002
none
Connor et al.,
1997
Friedman et al.,
1998
no difference in
mammary tumor
development
Bennett et al.,
2001
Figure 3 A graphic depiction of gender specific cancer risk by age 70 in carriers of BRCA1 or BRCA2 mutations. The diagrams
are representative of the risks reported from both population based and cancer family databases. Other cancers for BRCA1 female
carriers include colon, gastric and pancreas, although the increases in these cancers are small and not uniformly observed. In male
carriers of BRCA1 mutations, there is discrepancy as to if there is any increase in cancer risk, although in some studies small increases in breast, prostate and pancreas cancer are reported. A wider spectrum of cancer risk for carriers of BRCA2 is clearly seen
in both males and females. Other cancers for BRCA2 carriers include stomach, melanoma and gall bladder
Oncogene
Brca1 and Brca2 tumor suppression
ME Moynahan
9005
ences with defects in ductal branching were seen in
both Brca1 and Brca2 heterozygote mice. In addition,
an increased incidence of ovarian follicular arrest in the
heterozygote mice was noted. Mammary or ovary
tumor development was not observed in the DEStreated mice.
BRCA1 and BRCA2 tumor suppression
In mice or humans, the loss of either BRCA1 or
BRCA2 predisposes to tumorigenesis. In humans, the
tissues at risk to develop tumors were initially thought
to be restricted to mammary and ovary epithelium. For
BRCA2, this is clearly not the case, although the
penetrance of disease for breast and ovary cancer is not
as high as for a carrier of a BRCA1 mutation (Begg,
2002). Carriers of BRCA2 mutation have an increased
risk of developing male breast cancer, pancreas cancer,
gastrointestinal cancers and prostate cancer (Figure 3)
(BCLC, 1999). An even broader tissue range of
BRCA2-associated cancers has been reported in some
cancer families, strongly suggesting the role of
modifiers in the penetrance of BRCA2-associated
disease. The cancer risk associated with BRCA1
germline mutation is less clear for tumors other than
breast, ovary, and fallopian tube cancers (BCLC, 1999;
Brose et al., 2002; Thompson and Easton, 2002). Large
studies have reported both a lack of association and
small association with BRCA1 mutation in male breast
and prostate cancer (Basham et al., 2002; Brose et al.,
2002; Frank et al., 2002; Thompson and Easton, 2002).
A recent editorial commenting on reports from two
large and overlapping databases brings to light the
difficulties in estimating/establishing other small cancer
risks due to BRCA1 mutation (Gruber and Petersen,
2002). Additional differences between BRCA1 and
BRCA2-associated cancers include gender-specific risk
(Thompson and Easton, 2002), average age at
diagnosis (Risch et al., 2001), and the effect of genetic
modifiers on cancer risk (Levy-Lahad et al., 2001;
Wang et al., 2001). As compared to BRCA1 mutation
carriers, BRCA2 mutation carriers present with disease
at an older age and have an overall lower penetrance of
cancer risk but are at risk of developing more types of
cancer in both males and females.
BRCA1 and BRCA2-associated breast cancers are
significantly distinct, from each other, by histology, IHA
and gene expression profiles. It is less clear as to whether
BRCA2 mammary tumors are significantly different
from sporadic breast cancers, as pathology features are
not dissimilar but differences in gene expression profiles
have been reported. These differences underscore the
complexity underlying BRCA1-associated cancers.
Tumor suppression by BRCA1 may possibly be due to
the interrelationship of defective DNA repair and an
overall dysregulation of the DNA damage response,
whereas BRCA2-deficiency results in a homologydirected DSB repair defect that may be attenuated by
an overlapping function of other DSB repair pathways.
Certainties and uncertainties in BRCA-associated tumor
suppression
Human BRCA1 breast cancers and Brca1 mammary
tumor models have similar pathologic features, such as
high tumor grade, hormone receptor negativity, a high
incidence of p53 mutations and genetic instability. The
recurring chromosome aberrations observed in the
Brca1 mammary tumors mapped to chromosome
regions and gene locations commonly altered in human
breast cancer. The similarities between human BRCA2
breast cancer and reported murine Brca2 models are
less apparent, although both are more likely to be
hormone receptor positive, Her-2 negative and less
genetically unstable as compared to BRCA1-associated
tumors. The viable hypomorphic Brca1Tr and Brca2Dex27
mouse models sustain a wide spectrum of
carcinomas. No gender difference as pertaining to
tumor incidence was noted.
The observed pattern of multiple chromosome gains
and losses and progressive genomic instability in
Brca1-associated tumors is highly suggestive of Brca1
as a caretaker of the genome through DNA damage
response and repair. Early genetic instability, long
tumor latency and multiclonality support the hypothesis that loss of Brca1 promotes further genetic change,
some of which provide additional growth advantages.
p53 is markedly cooperative with both Brca1 and
Brca2 in tumor suppression of carcinoma. However,
loss of p53 function does not appear to be required for
tumorigenesis. Novel TP53 mutations identified in
human breast cancer suggest that a significant
proportion of the mutations may not abrogate most
wild-type p53 functions, the clinical significance of
these mutations in either tumorigenesis or tumor
behavior is unknown.
Caretakers of the genome act as safeguards against
genetic change during routine cellular metabolism and
damage from external stresses. Several mechanisms
have been attributed to the genetic instability found in
BRCA1 deficiency and include defective DSB repair by
the homologous repair pathway, defective DSB repair
which may possibly result in an increase in the
proportion of error prone repair events by NHEJ
(reviewed in Jasin), defective G2-M checkpoint (Xu et
al., 1999b, 2000a; Yarden et al., 2002) and centrosome
abnormalities (Deng, 2002). Does BRCA-associated
tumorigenesis occur when all or only a subset of these
functions are abrogated? Additional hypomorphic or
conditionally deleted animal models that allow for a
separation of function may help refine the role of
BRCA1 and BRCA2 in tumor suppression.
Acknowledgments
I would like to thank Larry Norton and the Breast Cancer
Research Fund for support and Maria Jasin and Mark
Robson for helpful comments and review of the manuscript.
Oncogene
Brca1 and Brca2 tumor suppression
ME Moynahan
9006
References
Baer R and Ludwig T. (2002). Curr. Opin. Genet. Dev., 12,
86 – 91.
Basham VM, Lipscombe JM, Ward JM, Gayther SA, Ponder
BA, Easton DF and Pharoah PD. (2002). Breast Cancer
Res., 4, R2.
Breast Cancer Linkage Consortium. (1999). J. Natl. Cancer
Inst., 91, 1310 – 1316.
Begg CB. (2002). J. Natl. Cancer Inst., 94, 1221 – 1226.
Bennett LM, McAllister KA, Blackshear PE, Malphurs J,
Goulding G, Collins NK, Ward T, Bunch DO, Eddy EM,
Davis BJ and Wiseman RW. (2000a). Mol. Carcinog., 28,
174 – 183.
Bennett LM, McAllister KA, Malphurs J, Ward T, Collins
NK, Seely JC, Gowen LC, Koller BH, Davis BJ and
Wiseman RW. (2000b). Cancer Res., 60, 3461 – 3469.
Bennett LM, McAllister KA, Ward T, Malphurs J, Collins
NK, Seely JC, Davis BJ and Wiseman RW. (2001).
Toxicol. Pathol., 29, 117 – 125.
Brodie SG, Xu X, Qiao W, Li WM, Cao L and Deng CX.
(2001). Oncogene, 20, 7514 – 7523.
Brose MS, Rebbeck TR, Calzone KA, Stopfer JE,
Nathanson KL and Weber BL. (2002). J. Natl. Cancer
Inst., 94, 1365 – 1372.
Chen J, Silver DP, Walpita D, Cantor SB, Gazdar AF,
Tomlinson G, Couch FJ, Weber BL, Ashley T, Livingston
DM and Scully R. (1998). Mol. Cell, 2, 317 – 328.
Connor F, Bertwistle D, Mee PJ, Ross GM, Swift S,
Grigorieva E, Tybulewicz VL and Ashworth A. (1997).
Nat. Genet., 17, 423 – 430.
Cortez D, Wang Y, Qin J and Elledge SJ. (1999). Science,
286, 1162 – 1166.
Cressman VL, Backlund DC, Avrutskaya AV, Leadon SA,
Godfrey V and Koller BH. (1999a). Mol. Cell. Biol., 19,
7061 – 7075.
Cressman VL, Backlund DC, Hicks EM, Gowen LC,
Godfrey V and Koller BH. (1999b). Cell Growth Differ.,
10, 1 – 10.
Crook T, Brooks LA, Crossland S, Osin P, Barker KT,
Waller J, Philp E, Smith PD, Yulug I, Peto J, Parker G,
Allday MJ, Crompton MR and Gusterson BA. (1998).
Oncogene, 17, 1681 – 1689.
Davies AA, Masson JY, McIlwraith MJ, Stasiak AZ, Stasiak
A, Venkitaraman AR and West SC. (2001). Mol. Cell, 7,
273 – 282.
Deng CX. (2002). Oncogene, 21, 6222 – 6227.
Deng CX and Brodie SG. (2001). Semin. Cancer Biol., 11,
387 – 394.
Easton DF, Bishop DT, Ford D and Crockford GP. (1993).
Am. J. Hum. Genet., 52, 678 – 701.
Frank TS, Deffenbaugh AM, Reid JE, Hulick M, Ward BE,
Lingenfelter B, Gumpper KL, Scholl T, Tavtigian SV,
Pruss DR and Critchfield GC. (2002). J. Clin. Oncol., 20,
1480 – 1490.
Friedman LS, Thistlethwaite FC, Patel KJ, Yu VP, Lee H,
Venkitaraman AR, Abel KJ, Carlton MB, Hunter SM,
Colledge WH, Evans MJ and Ponder BA. (1998). Cancer
Res., 58, 1338 – 1343.
Gowen LC, Johnson BL, Latour AM, Sulik KK and Koller
BH. (1996). Nat. Genet., 12, 191 – 194.
Greenblatt MS, Chappuis PO, Bond JP, Hamel N and
Foulkes WD. (2001). Cancer Res., 61, 4092 – 4097.
Gruber SB and Petersen GM. (2002). J. Natl. Cancer Inst.,
94, 1344 – 1345.
Oncogene
Hakem R, de la Pompa JL, Sirard C, Mo R, Woo M, Hakem
A, Wakeham A, Potter J, Reitmair A, Billia F, Firpo E,
Hui CC, Roberts J, Rossant J and Mak TW. (1996). Cell,
85, 1009 – 1023.
Hakem R and Mak TW. (2001). Annu. Rev. Genet., 35, 209 –
241.
Hall JM, Lee MK, Newman B, Morrow JE, Anderson LA,
Huey B and King MC. (1990). Science, 250, 1684 – 1689.
Hedenfalk I, Duggan D, Chen Y, Radmacher M, Bittner M,
Simon R, Meltzer P, Gusterson B, Esteller M, Kallioniemi
OP, Wilfond B, Borg A and Trent J. (2001). N. Engl. J.
Med., 344, 539 – 548.
Hohenstein P, Kielman MF, Breukel C, Bennett LM,
Wiseman R, Krimpenfort P, Cornelisse C, van Ommen
GJ, Devilee P and Fodde R. (2001). Oncogene, 20, 2544 –
2550.
Janocko LE, Brown KA, Smith CA, Gu LP, Pollice AA,
Singh SG, Julian T, Wolmark N, Sweeney L, Silverman JF
and Shackney SE. (2001). Cytometry, 46, 136 – 149.
Jazaeri AA, Yee CJ, Sotiriou C, Brantley KR, Boyd J and
Liu ET. (2002). J. Natl. Cancer Inst., 94, 990 – 1000.
Jones C, Nonni AV, Fulford L, Merrett S, Chaggar R,
Eusebi V and Lakhani SR. (2001). Br. J. Cancer, 85, 422 –
427.
Jonkers J, Meuwissen R, van der Gulden H, Peterse H, van
der Valk M and Berns A. (2001). Nat. Genet., 29, 418 – 425.
Karabinis ME, Larson D, Barlow C, Wynshaw-Boris A and
Moser AR. (2001). Carcinogenesis, 22, 343 – 346.
Kuperwasser C, Hurlbut GD, Kittrell FS, Dickinson ES,
Laucirica R, Medina D, Naber SP and Jerry DJ. (2000).
Am. J. Pathol., 157, 2151 – 2159.
Lakhani SR, Van De Vijver MJ, Jacquemier J, Anderson TJ,
Osin PP, McGuffog L and Easton DF. (2002). J. Clin.
Oncol., 20, 2310 – 2318.
Lee JS, Collins KM, Brown AL, Lee CH and Chung JH.
(2000). Nature, 404, 201 – 204.
Levy-Lahad E, Lahad A, Eisenberg S, Dagan E, Paperna T,
Kasinetz L, Catane R, Kaufman B, Beller U, Renbaum P
and Gershoni-Baruch R. (2001). Proc. Natl. Acad. Sci.
USA, 98, 3232 – 3236.
Liu CY, Flesken-Nikitin A, Li S, Zeng Y and Lee WH.
(1996). Genes Dev., 10, 1835 – 1843.
Ludwig T, Chapman DL, Papaioannou VE and Efstratiadis
A. (1997). Genes Dev., 11, 1226 – 1241.
Ludwig T, Fisher P, Ganesan S and Efstratiadis A. (2001a).
Genes Dev., 15, 1188 – 1193.
Ludwig T, Fisher P, Murty V and Efstratiadis A. (2001b).
Oncogene, 20, 3937 – 3948.
MacLachlan TK, Takimoto R and El-Deiry WS. (2002).
Mol. Cell. Biol., 22, 4280 – 4292.
Mak TW, Hakem A, McPherson JP, Shehabeldin A,
Zablocki E, Migon E, Duncan GS, Bouchard D, Wakeham
A, Cheung A, Karaskova J, Sarosi I, Squire J, Marth J and
Hakem R. (2000). Nat. Immunol., 1, 77 – 82.
McAllister KA, Bennett LM, Houle CD, Ward T, Malphurs
J, Collins NK, Cachafeiro C, Haseman J, Goulding EH,
Bunch D, Eddy EM, Davis BJ and Wiseman RW. (2002).
Cancer Res., 62, 990 – 994.
Medina D, Ullrich R, Meyn R, Wiseman R and Donehower
L. (2002). Environ. Mol. Mutagen., 39, 178 – 183.
Miki Y, Swensen J, Shattuck-Eidens D, Futreal PA,
Harshman K, Tavtigian S, Liu Q, Cochran C, Bennett
LM and Ding W et al. (1994). Science, 266, 66 – 71.
Brca1 and Brca2 tumor suppression
ME Moynahan
9007
Monteiro AN. (2002). Cancer Biol. Ther., 1, 187 – 188.
Moynahan ME, Chiu JW, Koller BH and Jasin M. (1999).
Mol. Cell, 4, 511 – 518.
Moynahan ME, Cui TY and Jasin M. (2001a). Cancer Res.,
61, 4842 – 4850.
Moynahan ME, Pierce AJ and Jasin M. (2001b). Mol. Cell, 7,
263 – 272.
Phillips KA. (2000). J. Clin. Oncol., 18, 107S – 112S.
Risch HA, McLaughlin JR, Cole DE, Rosen B, Bradley L,
Kwan E, Jack E, Vesprini DJ, Kuperstein G, Abrahamson
JL, Fan I, Wong B and Narod SA. (2001). Am. J. Hum.
Genet., 68, 700 – 710.
Scully R, Chen J, Plug A, Xiao Y, Weaver D, Feunteun J,
Ashley T and Livingston DM. (1997). Cell, 88, 265 – 275.
Seidman AD, Fornier MN, Esteva FJ, Tan L, Kaptain S,
Bach A, Panageas KS, Arroyo C, Valero V, Currie V,
Gilewski T, Theodoulou M, Moynahan ME, Moasser M,
Sklarin N, Dickler M, D’Andrea G, Cristofanilli M,
Rivera E, Hortobagyi GN, Norton L and Hudis CA.
(2001). J. Clin. Oncol., 19, 2587 – 2595.
Sharan SK, Morimatsu M, Albrecht U, Lim DS, Regel E,
Dinh C, Sands A, Eichele G, Hasty P and Bradley A.
(1997). Nature, 386, 804 – 810.
Shen SX, Weaver Z, Xu X, Li C, Weinstein M, Chen L, Guan
XY, Ried T and Deng CX. (1998). Oncogene, 17, 3115 –
3124.
Shin SJ and Rosen PP. (2002). Am. J. Surg. Pathol., 26, 413 –
420.
Smith PD, Crossland S, Parker G, Osin P, Brooks L, Waller
J, Philp E, Crompton MR, Gusterson BA, Allday MJ and
Crook T. (1999). Oncogene, 18, 2451 – 2459.
Thompson D and Easton DF. (2002). J. Natl. Cancer Inst.,
94, 1358 – 1365.
Tibbetts RS, Cortez D, Brumbaugh KM, Scully R,
Livingston D, Elledge SJ and Abraham RT. (2000). Genes
Dev., 14, 2989 – 3002.
Turner BC, Ottey M, Zimonjic DB, Potoczek M, Hauck
WW, Pequignot E, Keck-Waggoner CL, Sevignani C,
Aldaz CM, McCue PA, Palazzo J, Huebner K and Popescu
NC. (2002). Cancer Res., 62, 4054 – 4060.
Tutt A, Gabriel A, Bertwistle D, Connor F, Paterson H,
Peacock J, Ross G and Ashworth A. (1999). Curr. Biol., 9,
1107 – 1110.
Tutt AN, van Oostrom CT, Ross GM, van Steeg H and
Ashworth A. (2002). EMBO Rep., 3, 255 – 260.
van’t Veer LJ, Dai H, van de Vijver MJ, He YD, Hart AA,
Mao M, Peterse HL, van der Kooy K, Marton MJ,
Witteveen AT, Schreiber GJ, Kerkhoven RM, Roberts C,
Linsley PS, Bernards R and Friend SH. (2002). Nature,
415, 530 – 536.
Wang WW, Spurdle AB, Kolachana P, Bove B, Modan B,
Ebbers SM, Suthers G, Tucker MA, Kaufman DJ, Doody
MM, Tarone RE, Daly M, Levavi H, Pierce H, Chetrit A,
Yechezkel GH, Chenevix-Trench G, Offit K, Godwin AK
and Struewing JP. (2001). Cancer Epidemiol. Biomarkers
Prev., 10, 955 – 960.
Weaver Z, Montagna C, Xu X, Howard T, Gadina M,
Brodie SG, Deng CX and Ried T. (2002). Oncogene, 21,
5097 – 5107.
Welcsh PL and King MC. (2001). Hum. Mol. Genet., 10,
705 – 713.
Welcsh PL, Lee MK, Gonzalez-Hernandez RM, Black DJ,
Mahadevappa M, Swisher EM, Warrington JA and King
MC. (2002). Proc. Natl. Acad. Sci. USA, 99, 7560 – 7565.
Wooster R, Neuhausen SL, Mangion J, Quirk Y, Ford D,
Collins N, Nguyen K, Seal S, Tran T and Averill D et al.
(1994). Science, 265, 2088 – 2090.
Xu B, Kim S and Kastan MB. (2001a). Mol. Cell. Biol., 21,
3445 – 3450.
Xu X, Qiao W, Linke SP, Cao L, Li WM, Furth PA, Harris
CC and Deng CX. (2001b). Nat. Genet., 28, 266 – 271.
Xu X, Wagner KU, Larson D, Weaver Z, Li C, Ried T,
Hennighausen L, Wynshaw-Boris A and Deng CX.
(1999a). Nat. Genet., 22, 37 – 43.
Xu X, Weaver Z, Linke SP, Li C, Gotay J, Wang XW, Harris
CC, Ried T and Deng CX. (1999b). Mol. Cell, 3, 389 – 395.
Yang H, Jeffrey PD, Miller J, Kinnucan E, Sun Y, Thoma
NH, Zheng N, Chen PL, Lee WH and Pavletich NP.
(2002). Science, 297, 1837 – 1848.
Yarden RI, Pardo-Reoyo S, Sgagias M, Cowan KH and
Brody LC. (2002). Nat. Genet., 30, 285 – 289.
Yu VP, Koehler M, Steinlein C, Schmid M, Hanakahi LA,
van Gool AJ, West SC and Venkitaraman AR. (2000).
Genes Dev., 14, 1400 – 1406.
Oncogene