MOLECULAR AND CELLULAR BIOLOGY,
Copyright © 1999, American Society for Microbiology. All Rights Reserved.
July 1999, p. 4843–4854Vol. 19, No. 7
BRCA1 Is Phosphorylated at Serine 1497 In Vivo at a
Cyclin-Dependent Kinase 2 Phosphorylation Site
HEINZ RUFFNER, WEI JIANG, A. GREY CRAIG, TONY HUNTER, AND INDER M. VERMA*
The Salk Institute, La Jolla, California 92037
Received 6 January 1999/Returned for modification 25 February 1999/Accepted 26 March 1999
BRCA1 is a cell cycle-regulated nuclear protein that is phosphorylated mainly on serine and to a lesser
extent on threonine residues. Changes in phosphorylation occur in response to cell cycle progression and DNA
damage. Specifically, BRCA1 undergoes hyperphosphorylation during late G1and S phases of the cell cycle.
Here we report that BRCA1 is phosphorylated in vivo at serine 1497 (S1497), which is part of a cyclin-
dependent kinase (CDK) consensus site. S1497 can be phosphorylated in vitro by CDK2-cyclin A or E. BRCA1
coimmunoprecipitates with an endogenous serine-threonine protein kinase activity that phosphorylates S1497
in vitro. This cellular kinase activity is sensitive to transfection of a dominant negative form of CDK2 as well
as the application of the CDK inhibitors p21 and butyrolactone I but not p16. Furthermore, BRCA1 coim-
munoprecipitates with CDK2 and cyclin A. These results suggest that the endogenous kinase activity is
composed of CDK2-cyclin complexes, at least in part, concordant with the G1/S-specific increase in BRCA1
While breast cancer occurs mostly as a sporadic disease,
genetic predisposition accounts for 5 to 10% of all breast
cancer cases. However, the contribution of hereditary factors
to breast cancer in women under 30 years of age may be as high
as 25%, since familial breast cancer often occurs at an early age
(17). Mutations in the BRCA1 gene account for about half of
the families with high breast cancer incidence and at least 80%
of families predisposed to both early-onset breast and ovarian
cancer (18). It has been proposed that the BRCA1 gene en-
codes a tumor suppressor protein, since tumors from BRCA1
mutation carriers display loss or inactivation of the remaining
wild-type allele (32, 47). Although somatic mutations in the
BRCA1 gene are rarely found in sporadic breast and ovarian
cancers, BRCA1 may still play a role in these forms of cancer
(20, 21, 25, 38, 44, 55). Loss of heterozygosity is frequently
observed in the region of chromosome 17q that harbors the
BRCA1 gene (20, 33, 34), and recent evidence suggests that
BRCA1 levels are reduced in sporadic breast cancers (53, 65,
77). It also has been proposed that BRCA1 is aberrantly lo-
calized in sporadic breast cancers (14).
The BRCA1 gene encodes a 1,863-amino-acid (aa) protein
whose primary sequence offers few clues about its function
(45). It contains a RING finger at its amino (N)-terminal
region and a domain termed BRCT at its carboxy (C) terminus
(36). Proteins harboring BRCT domains have been implicated
to participate in DNA damage-responsive checkpoint and cell
cycle control functions (6, 10, 36). BRCA1 is a nuclear protein
whose RNA and protein levels are cell cycle regulated (14, 15,
26, 57, 62, 68, 73, 76). Moreover, BRCA1 is phosphorylated,
and its phosphorylation state also undergoes cell cycle-specific
alterations (16, 57, 60, 69).
The molecular function of BRCA1 has not yet been deter-
mined. Evidence that the C-terminal domain of BRCA1 (aa
1528 to 1863) has transcription activation activity and that
BRCA1 is linked to the RNA polymerase II holoenzyme via
RNA helicase A has been presented (3, 11, 46, 59, 64). Fur-
thermore, BRCA1 can regulate gene expression in concert
with p53 or CBP/p300 (52, 54, 78). The finding that BRCA1
associates with Rad51 in mitotic and meiotic cells implies a
role for BRCA1 in the control of recombination and genome
integrity and underlines the proposed function of BRCA1 as a
caretaker (35, 61). Moreover, the subnuclear localization and
the phosphorylation state of BRCA1 change in response to
DNA damage, suggesting that BRCA1 participates in a DNA
damage-dependent replication checkpoint response (60, 69). A
recent report demonstrates a role for BRCA1 in transcription-
coupled repair of oxidative DNA damage (23). BRCA1 has
also been proposed to regulate cell proliferation, differentia-
tion, and apoptosis (24, 27, 37, 41, 42, 63, 70).
Since BRCA1 phosphorylation responds to cell cycle pro-
gression and DNA damage, one can assume that phosphory-
lation regulates the activity of the protein, as is the case for
other tumor suppressors, e.g., p53 and Rb (4, 22). It is there-
fore crucial to understanding the biology of BRCA1 that the
specifics of BRCA1 phosphorylation be investigated. This in-
cludes the identification of the sites of phosphorylation and the
responsible protein kinase(s). Such information would link
BRCA1 to a cellular pathway(s) and facilitate investigation of
its function at the molecular level. Candidate protein kinases
that may be responsible at least in part for the cell cycle-
specific phosphorylation changes of BRCA1 include the cyclin-
dependent kinases (CDKs).
Members of the CDK family are important regulators of the
eukaryotic cell cycle (29, 50, 56). The activity of CDKs is tightly
regulated by association with other polypeptides and by addi-
tion or removal of phosphate moieties. The intrinsically inac-
tive CDK catalytic subunit requires association with a positive
regulatory cyclin partner, and multiple regulatory phosphory-
lation and dephosphorylation events occur on both CDK and
cyclin subunits. The activity of the CDK-cyclin complexes can
be further modulated by association with other polypeptides,
such as inhibitors of CDKs (CDIs) (19, 39, 49). Vertebrates
possess multiple CDK and cyclin subunits. D-type cyclins in
association with CDK4 and CDK6 function early in the cell
cycle, during G1-phase progression, whereas CDK2-cyclin E
* Corresponding author. Mailing address: Laboratory of Genetics,
The Salk Institute, 10010 North Torrey Pines Rd., La Jolla, CA 92037.
Phone: (619) 453-4100, ext. 1462. Fax: (619) 558-7454. E-mail: verma
and CDK2-cyclin A act later on, at the G1/S transition and
during S-phase progression, respectively. CDC2-cyclin B1
functions at the G2/M transition. The CDIs are divided into
two categories, based on differences in structure, mechanism of
inhibition, and specificity. Members of the p21 family (com-
prising p21, p27, and p57) preferentially inhibit CDKs of the
G1and S phases (CDK2, -3, -4, and -6), whereas members of
the INK4 (inhibitor of CDK4) family (comprising p15, p16,
p18, and p19) are selective for CDK4 and CDK6 (28). The
CDKs display substrate specificity toward proteins containing
the motif S/T-P-(X)-K/R (S, serine; T, threonine; P, proline; X,
any amino acid; K, lysine; R, arginine) or minimally S/T-P
(followed by K/R) (40, 48).
Here we show that serine residue 1497 (S1497), which con-
stitutes one of the four CDK consensus sites in BRCA1, is
phosphorylated in vivo and can be phosphorylated by recom-
binant CDK2-cyclin complexes in vitro. BRCA1 coimmuno-
precipitates with an endogenous serine-threonine protein ki-
nase activity that phosphorylates this particular serine residue
in a manner sensitive to p21 and butyrolactone I but not to p16.
Moreover, a dominant negative CDK2 mutant (CDK2 dn)
inhibits BRCA1 phosphorylation in vivo. We propose that cel-
lular CDK2 is responsible, at least in part, for the G1/S-depen-
dent increase in BRCA1 phosphorylation.
MATERIALS AND METHODS
Plasmids. To make subsequent cloning easier, full-length BRCA1 cDNA
derived from pCL-MFG-BRCA1 (57) was engineered in two consecutive steps
into pCL-MFG-MCS (64a) to generate pCL-MFG?-BRCA1. BRCA1?772-1050
was generated in pCL-MFG?-BRCA1 by ligation of KpnI (blunted by 3? over-
hang removal) to ScaI-cut cDNA in a three-way ligation. BRCA1 1051-1863 was
generated by inserting a composite of ScaI-ApaI and ApaI-BamHI from pUHD-
P1-BRCA1 (57) into NcoI (blunted by fill in) and BamHI-cut pCL-MFG-MCS.
To generate Myc-tagged, full-length wild-type BRCA1 (Myc-BRCA1 wt), a tag
containing five Myc epitopes derived by PCR from 6x myc BRCA1 1314-1863
(54) was inserted into the NcoI site (at the start methionine) of pCL-MFG?-
BRCA1. BRCA1 1314-1652 was cloned by deleting aa 1653 to 1863 by PCR in
6x myc BRCA1 1314-1863. Mutants S1497A and S1497T were generated first in
BRCA1 1051-1863, using a Clontech transformer site-directed mutagenesis kit,
and then lifted into Myc-BRCA1 (yielding Myc-BRCA1 S1497A and S1497T);
mutants T967S and T967D were first generated in pCL-MFG?-BRCA1 lacking
aa 2 to 473 before being cloned into Myc-BRCA1 (yielding Myc-BRCA1 T967S
and T967D). Mutagenic primers for S1497A, S1497T, T967S, and T967D were
TTGACCTTTCCACTCCTG, GTTTATTTGGAGAAATGAGTCCAG, and
In addition, we used plasmids expressing CDK2 and CDK2 dn (71), cyclin E
(29a), and cyclin A (25a).
Cell culture, transfections, and in vivo labeling. 293T and HBL-100 cells were
cultivated in Dulbecco modified Eagle medium (DMEM) (Cellgro; Mediatech)
containing 10% fetal bovine serum (FBS) (HyClone) at 37°C in 10% CO2and in
McCoy’s 5A medium (Gibco BRL) containing 10% FBS at 37°C in 5% CO2,
Transfections of 293T cells were performed by an adaptation of the calcium
phosphate method described by Wigler et al. (75). We usually obtained trans-
fection efficiencies of 70 to 100%, as determined by green fluorescent protein
staining due to cotransfection of trace amounts of a green fluorescent protein-
expressing plasmid in some experiments (data not shown).
incubated in phosphate-free DMEM containing 10% dialyzed FBS (Gibco
BRL), 0.1 mM MEM nonessential amino acids (Gibco), 2 mM L-glutamine
(Gibco), and 1 mM MEM-sodium pyruvate (Gibco) for about 45 min at 37°C
before addition of 0.7 mCi of32P (H3PO4; ICN Pharmaceuticals, Inc.) per ml.
After 4 h of labeling, cells were washed with phosphate-buffered saline (without
Ca2?and Mg2?), scraped from the plates, pelleted by centrifugation, and frozen
on dry ice.
Western blot analysis. Western blot analysis was performed as described
elsewhere (57) except that cells were lysed under slightly modified inhibitor
conditions: 50 mM HEPES (pH 7.4)–150 mM NaCl–10% glycerol–1% Triton
X-100–15 mM MgCl2–10 mM EGTA–1 mM dithiothreitol containing inhibitors
pepstatin A (1 ?g/ml), phenylmethylsulfonyl fluoride (1 mM), leupeptin (10
?g/ml), aprotinin (21 ?g/ml), 20 mM NaF, 10 mM 4-nitrophenyl phosphate, 1
mM Na3VO4, 20 mM ?-glycerophosphate, 10 mM Na2MoO4, and 100 nM
calyculin A. The following antibodies were used: BRCA1 Ab-D (57), Myc-
32P labeling, cells were washed twice in phosphate-free DMEM and
specific antibody 9E10 (Jill Meisenhelder, The Salk Institute), and mouse CDK2-
specific monoclonal antibody (D-12; Santa Cruz).
Immunoprecipitations. For immunoprecipitations of in vivo-labeled overex-
pressed and endogenous proteins and of unlabeled BRCA1, cells were lysed in
radioimmunoprecipitation assay (RIPA) buffer (57) containing 2 mM EDTA and
the proteinase/phosphatase inhibitors mentioned above in the presence of 0.5%
sodium dodecyl sulfate (SDS) before being heated for 6 to 10 min at 99°C. After
the SDS concentration was adjusted to 0.1% by addition of 4 volumes of the
above buffer without SDS (and addition of RNase A to 100 ?g/ml in the case of
labeled proteins), the lysates were passed 10 times through a 22G1 1/2 needle
(Becton Dickinson & Co.) and cleared by centrifugation at 4°C and 12,000 rpm
for 30 min. To reduce nonspecific binding, the supernatants were incubated with
protein A-Sepharose (Pharmacia) for 1 h at 4°C while rotating; after removal of
the beads by centrifugation, the lysates were incubated with BRCA1 antibody
Ab-D or Ab-C plus Ab-D (57) at a total of 5 ?l/ml for at least 2 h at 4°C while
rotating. Finally, protein A-Sepharose beads were added for 1 h while tumbling
at 4°C before the beads were washed five to seven times with RIPA buffer–0.1%
SDS. The immunoprecipitates were dissociated from the beads by heating at
99°C for 6 to 8 min and then separated on SDS-polyacrylamide gels (5% gels for
full-length BRCA1 and 8% for the fragment spanning aa 1051 to 1863 [BRCA1
1051-1863]) before they were transferred to polyvinylidene difluoride mem-
branes (Immobilon-P; Millipore) and visualized by autoradiography or subjected
to Western blot analysis.
For in vitro kinase assays of Myc-tagged BRCA1, cells were lysed and pro-
cessed in RIPA buffer as described above, without phosphatase inhibitors, SDS,
and heat denaturation, using 10 ?g of antibody 9E10 per ml for immunoprecipi-
tation in the presence of protein G-Sepharose (Pharmacia). After precipitation,
beads were washed twice with RIPA buffer and three times with 50 mM HEPES
(pH 7.4)–10 mM MgCl2. In some experiments, bacterial alkaline phosphatase
(BAP) (BAPF; Worthington) was added to the immunoprecipitation reactions
(together with the first addition of protein G-Sepharose) in order to efficiently
dephosphorylate BRCA1 before performance of in vitro kinase reactions; how-
ever, tryptic phosphopeptide maps of BAP-treated and untreated BRCA1 were
similar (data not shown).
For coimmunoprecipitations, lysates were prepared and processed as for in
vitro kinase assays, using the following rabbit antibodies (Santa Cruz) in the
presence of protein A-Sepharose: anti-CDK2 (M2; where indicated, the antibody
was preincubated with an equal amount of the corresponding blocking peptide
on ice for 35 min), anti-cyclin A (H-432), anti-cyclin E (C-19), and anti-NF-?B
p65 (A). The immunocomplexes were washed six times with RIPA buffer con-
taining 2 mM EDTA and proteinase inhibitors before analysis.
In vitro kinase assays, phosphopeptide mapping, and phosphoamino acid
analysis (PAA). In vitro kinase assays were performed on the immunoprecipi-
tates bound to the beads in 25 ?l of a mixture containing 50 mM HEPES (pH
7.4), 10 mM MgCl2, 1 mM dithiothreitol, 25 ?M ATP, and 10 ?Ci of [?-32P]ATP
at 30°C for 30 min. Where indicated, the corresponding baculovirus-expressed,
purified CDK-cyclin complexes (50 to 100 ng/reaction ) were added to the
reactions; for the inhibition experiments, 5 ?g of either recombinant p16 or p21
(Tim Mayall, The Salk Institute) or 62 ?M butyrolactone I (Calbiochem) (31)
was added. The reaction products were separated on SDS–5% polyacrylamide
gels before autoradiography.
Two-dimensional phosphopeptide mapping and PAA were performed as pre-
viously described (57, 72).
Mass spectrometric analysis. Matrix-assisted laser desorption measurements
were carried out on a Bruker Reflex (Bruker Instruments Inc., Manning Park,
Billerica, Mass.) reflectron time-of-flight mass spectrometer utilizing a nitrogen
UV laser. The instrument was operated with an accelerating voltage of ?31 kV
and a reflector potential of ?30 kV. The mass spectrum represents the accumu-
lation of approximately 20 laser shots. The mass accuracy of the instrument was
typically ?200 ppm. Micro-high-performance liquid chromatography analysis
was performed on a UMA (Michrom Bioresources, Auburn, Calif.) using a 0.5-
by 150-mm column (Michrom Bioresources) packed with 5-?m (300-Å) C18
reverse-phase material (Vydac, Hesperia, Calif.) to separate the Lys-C (Boehr-
inger Mannheim, Mannheim, Germany) digest.
Phosphorylation pattern of endogenous and overexpressed
BRCA1. BRCA1 is a cell cycle-regulated nuclear phosphopro-
tein that is phosphorylated at multiple amino acids, mainly on
serine and to a lesser extent on threonine residues (57). Be-
cause human BRCA1 contains 224 serine and 111 threonine
residues (corresponding to 12 and 6%, respectively, of the
entire protein ), we decided to narrow our search to re-
gions of the protein that are phosphorylated before identifying
individual phosphorylated residues. To investigate the phos-
phorylation of BRCA1, we set out to characterize the phos-
phorylation of the transiently overexpressed protein, which
would allow us to introduce mutations and test their effects on
4844 RUFFNER ET AL.MOL. CELL. BIOL.
phosphorylation. We transiently transfected a series of over-
lapping BRCA1 fragments encompassing the entire BRCA1
protein into human embryonic kidney cells (293T) that were
then labeled with [32P]phosphoric acid. The tryptic phos-
phopeptide maps generated from the immunopurified frag-
ments after separation on SDS-polyacrylamide gels were then
compared to the map of full-length BRCA1 (72). The presence
of the same phosphopeptide in the maps derived from a frag-
ment and wild-type protein indicated that the phosphorylated
amino acid is located within that particular fragment. BRCA1
was found to be phosphorylated in several regions, and some
were located within the C-terminal half of the protein (57a).
We then investigated candidate consensus phosphorylation
sites within these phosphorylated regions, with particular em-
phasis on CDK consensus sites, because of the cell cycle-spe-
cific changes in phosphorylation of BRCA1.
Human BRCA1 contains four CDK consensus sites; three
[S(896)PK, T(967)PNK, and S(1009)PER] are located within
exon 11, and one [S(1497)PSK] is located close to the C ter-
minus. All of these sites could be target sites in vivo for CDKs
based on their primary sequences. To determine whether
BRCA1 is phosphorylated at these sites, we transiently ex-
pressed in 293T cells either full-length BRCA1 protein or
fragments harboring mutations at the CDK sites and compared
their resulting phosphopeptide maps to that of wild-type full-
length BRCA1 protein. A difference in the map of a particular
mutant indicated that the sequence surrounding that amino
acid played a role in phosphorylation in vivo.
To demonstrate that in vivo-overexpressed BRCA1 is in-
deed correctly phosphorylated and reflects the phosphoryla-
tion of the endogenous protein, we compared the tryptic phos-
phopeptide maps of endogenous BRCA1 derived from human
breast epithelial (HBL-100 [Fig. 1A]), 293T (Fig. 1B), and
malignant glioma (M059J [Fig. 1C]) cells as well as from cer-
vical carcinoma cells (HeLa [data not shown]) to the map of
overexpressed protein (pCL-MFG?-BRCA1 [Fig. 1D; see also
reference 57]). The majority of tryptic phosphopeptides
present in the map of the endogenous protein are identical to
those present in the map of the overexpressed protein (p1 to
p4, p9, p11, p12, p18, and p19), verifying that the phosphory-
lation of the overexpressed protein occurs at physiological
sites. Phosphopeptides p7 and p20 found in the map of over-
FIG. 1. Endogenous and overexpressed BRCA1 reveal similar phosphorylation patterns in vivo. Shown are the two-dimensional tryptic phosphopeptide maps of
endogenous BRCA1 from HBL-100 (A), 293T (B), and M059J (C) cells and of overexpressed full-length (D) and truncated BRCA1 lacking aa 772 to 1050 (E) from
transiently transfected 293T cells. BRCA1 was immunoprecipitated with Ab-D (for maps B to E) or Ab-C plus Ab-D (for map A) from lysates of [32P]phosphoric
acid-labeled cells. The immunoprecipitates were separated by SDS-polyacrylamide gel electrophoresis, and the labeled BRCA1 protein species were enzymatically
hydrolyzed with trypsin. The resulting peptides were separated in the first and second dimensions by electrophoresis and chromatography, respectively, as indicated by
the arrows. Diamonds mark sample origins; circles (labeled p1 to p7, p9, p11 to p13 , and p17 to p20) indicate phosphopeptides detected in the individual maps.
Map D represents a combination of phosphopeptide maps of endogenous and overexpressed BRCA1 protein from 293T cells because Ab-D immunoprecipitated both
proteins, which have similar mobilities on SDS-polyacrylamide gels (unlike the faster-running truncated BRCA1 lacking aa 772 to 1050 [data not shown]); however,
since overexpressed BRCA1 is far more abundant than the endogenous protein (data not shown), the contribution of endogenous BRCA1 in map D is negligible.
VOL. 19, 1999 BRCA1 PHOSPHORYLATION AT SERINE 14974845
expressed BRCA1 (Fig. 1D), although not detected in the
maps of endogenous BRCA1 from HBL-100 and 293T cells
(Fig. 1A and B), are present in the map of endogenous BRCA1
from M059J cells (Fig. 1C). Phosphopeptides p5, p6, p13, and
p17 are present in the maps of endogenous BRCA1 from
HBL-100 and 293T cells, although at relatively low levels (Fig.
1A and B), and of overexpressed BRCA1 (Fig. 1D), but they
are below detection levels in the map of endogenous BRCA1
from M059J cells (Fig. 1C). We note that some of the spots
differ in relative intensities between (i) endogenous BRCA1
from different cell lines and (ii) endogenous and overexpressed
protein. The former may reflect differences in phosphorylation
characteristic for each cell line. The latter observation may be
due to an alteration in the ratio of protein kinase(s) to BRCA1
and/or improper subcellular localization of overexpressed
BRCA1, since it is expressed in large excess compared to the
endogenous protein. Immunofluorescence analysis of tran-
siently overexpressed BRCA1 protein revealed nuclear and
cytoplasmic staining (data not shown). Alternatively, overex-
pressed BRCA1 could have an effect on the cellular machinery
which leads to an imbalanced phosphorylation of BRCA1.
There are a few additional phosphopeptides detected in the
maps of endogenous and overexpressed BRCA1, e.g., the two
spots below p4 (Fig. 1; spots are not numbered). The variation
of their relative abundance among different experiments as
well as the fact that these spots are also present in the tryptic
peptide map of the BRCA1 fragment lacking aa 772 to 1050
(Fig. 1E [see below]), which has a different mobility on a
SDS-polyacrylamide gel compared to wild-type protein, sug-
gests that they are partially digested BRCA1 phosphopeptides.
We conclude that we are able to study the characteristics of
BRCA1 phosphorylation by using the transiently overex-
BRCA1 is phosphorylated at S1497 in vivo. We then com-
pared the phosphopeptide map of an in vivo-labeled overex-
pressed fragment of BRCA1 protein from which aa 772 to
1050, encompassing the three CDK sites in exon 11, were
deleted (Fig. 1E) to the map of the full-length protein (Fig.
1D). The two maps are virtually identical, revealing that no
tryptic phosphopeptides were derived from the region between
aa 772 and 1050. Furthermore, a Myc-tagged, full-length
BRCA1 protein in which T967 was replaced by aspartic acid
(see below) revealed the same tryptic phosphopeptide map as
the corresponding wild-type protein (data not shown). There-
fore, the three CDK sites in exon 11 are not major phosphor-
ylation sites in vivo (see Discussion).
To check whether the C terminus of BRCA1 containing the
fourth CDK consensus site (S1497) is phosphorylated, we over-
expressed the C-terminal half of BRCA1 (aa 1051 to 1863,
untagged) in 293T cells followed by immunoprecipitation using
Ab-D (Fig. 2A, lane 1) (57) and performed tryptic phos-
phopeptide analysis as described above. Several phosphopep-
tides in full-length BRCA1 are derived from this fragment
(Fig. 2B, left; compare to Fig. 1D), in accordance with this
fragment appearing as several closely migrating bands that
probably represent distinctly phosphorylated forms (Fig. 2A,
lane 1). Substituting alanine for S1497 (S1497A) results in
reduced phosphorylation of the fragment, as evidenced by a
decrease in the slower-migrating species compared to the fast-
est-migrating phosphorylated form (Fig. 2A; compare lanes 1
and 2). This reduction is due to the loss of phosphopeptides p4
and p20 (Fig. 2B, right), suggesting that this CDK site of
BRCA1 is phosphorylated in vivo. The fact that two phos-
phopeptides, of which at least p4 is detected in maps of en-
dogenous BRCA1 protein (Fig. 1A to C), disappear in the
two-dimensional peptide map upon mutation of one amino
acid implies that these two peptides are related, suggesting that
peptide p20 is a partial tryptic hydrolysis fragment encompass-
ing peptide p4. This result is also obtained in the context of
full-length BRCA1 protein, both untagged and Myc tagged
(data not shown). BRCA1 fragment 1051-1863 described
above lacks the BRCA1 nuclear localization signals (12, 67, 76)
and localizes primarily to the cytoplasm, as determined by
indirect immunofluorescence analysis and biochemical frac-
tionation (data not shown). Since CDK2-cyclins are predomi-
nantly nuclear, a cytoplasmic localization might be expected to
preclude phosphorylation of this fragment. However, it has
been reported that in transformed cells, including 293 cells, a
significant fraction of CDK-cyclin complexes including cyclin
A, cyclin E, and CDK2 are present in the cytoplasm (51). Such
cytoplasmic CDK-cyclin complexes could phosphorylate the
1051-1863 fragment localized in the cytoplasm.
Based on the results described above and the predicted
relative phosphopeptide mobilities in the electrophoresis and
chromatographic dimensions (reference 7 and data not shown),
we propose that the amino acid sequences of peptides p4 and
p20 are S(1496)SPSK(1500) and N(1488)KEPGVERSSPSK
(1500), respectively (the phosphorylated serine residue is un-
derlined; trypsin does not efficiently cleave the sequences K-E
and R-X-phospho-S ).
We note that it is formally possible that rather than S1497,
a neighboring residue is the phosphorylated amino acid within
peptide p4, in which case the S1497A substitution inhibits the
phosphorylation of the neighboring residue. This amino acid
would have to be a serine residue, since phosphopeptide p4
contains only phosphoserine (57). To identify S1497 as the
phosphorylated amino acid, we carried out the following bio-
chemical characterization. First, we showed that the radiola-
beled amino acid within phosphopeptide p4 is the second res-
idue, as predicted for the peptide SSPSK (aa 1496 to 1500). A
Myc-tagged BRCA1 fragment consisting of aa 1314 to 1652
was transiently transfected into 293T cells and subsequently
labeled with [32P]phosphoric acid. The fragment was immuno-
precipitated with the Myc-specific antibody 9E10, separated by
SDS-polyacrylamide gel electrophoresis, and subjected to two-
dimensional tryptic peptide mapping, revealing the presence of
phosphopeptide p4 within this fragment (data not shown). The
phosphopeptide was isolated from the thin-layer chromatog-
raphy plate and subjected to manual Edman degradation (72).
The major release of radiolabeled phosphate occurred after
the second cycle (Fig. 2C), consistent with the hypothesis that
phosphopeptide p4 is phosphorylated at S1497. Second, mass
spectrometric analysis was performed on peptides derived
from the same in vivo-labeled C-terminal BRCA1 fragment
described above. The fragment was enzymatically hydrolyzed,
purified by micro-high-performance liquid chromatography,
and subjected to mass spectrometric analysis. To optimize the
analysis, we chose to hydrolyze the BRCA1 fragment with
endoproteinase Lys-C rather than trypsin, generating a peptide
fragment larger than the 5-aa peptide p4. A peptide with a
mass-to-charge ratio of 1,252.8 was observed (Fig. 2D), con-
sistent with the expected monoisotopic [M?H]?of 1,252.56
Da for the singly phosphorylated form of the Lys-C peptide aa
1490 to 1500 (EPGVERSSPSK). Thus, considering the results
described above combined with the Edman degradation data
and mass spectrometric analysis, we conclude that S1497 is the
phosphorylated serine residue.
BRCA1 can be phosphorylated in vitro by CDK2-cyclin A
and CDK2-cyclin E. To investigate which cyclin-dependent
protein kinase might be responsible for phosphorylating
BRCA1 at aa S1497, we performed in vitro kinase reactions on
BRCA1 protein, using four different recombinant CDK-cyclin
4846RUFFNER ET AL.MOL. CELL. BIOL.
complexes. Myc-BRCA1 wild type (wt) was transiently over-
expressed in 293T cells and subsequently immunoprecipitated
under native conditions using antibody 9E10. The immunopre-
cipitate was incubated with baculovirus-expressed, purified
CDK2-cyclin A, CDK2-cyclin E, CDK6-cyclin D1, or CDC2-
cyclin B1 in the presence of [?-32P]ATP (30). The reaction
products were separated on SDS-polyacrylamide gels and
transferred to a membrane before autoradiography. Figure 3A
FIG. 2. BRCA1 serine residue 1497 is phosphorylated in vivo. (A) The C-terminal half of BRCA1 (aa 1051 to 1863), either wild type (lane 1) or with the S1497A
substitution (lane 2), was expressed in 293T cells that were labeled with [32P]phosphoric acid. The fragments were immunoprecipitated with Ab-D and separated on
SDS-polyacrylamide gels. Protein marker bands (200 and 97.4 kDa) are indicated on the left. (B) Two-dimensional tryptic phosphopeptide maps of the BRCA1
fragments shown in panel A. The asterisk (*) in the left panel denotes a fraction of phosphopeptide p7 (§) having an abnormal mobility in the chromatographic
dimension, possibly due to a thin-layer chromatography plate artifact. (C) Manual Edman degradation of phosphopeptide p4. The procedure was performed as
described elsewhere (72), and the reaction products were analyzed on a PhosphorImager (Molecular Dynamics). Numbers above the panel indicate cycles of
degradation (0 denotes starting material). M, free [32P]phosphate applied as a marker. The ratio of free [32P]phosphate to phosphopeptide p4 was determined by using
the ImageQuaNT software (Molecular Dynamics). About 50% of [32P]phosphate was liberated after the second cycle (compared to the fifth cycle), whereas most of
the remaining [32P]phosphate was gradually released over cycles 3 to 5, probably due to incomplete reactions. (D) Matrix-assisted laser desorption mass spectrum of
the purified Lys-C peptide 1490-1500 derived from BRCA1 aa 1314 to 1652. x axis, mass-to-charge ratio (m/z); y axis, relative intensity.
VOL. 19, 1999 BRCA1 PHOSPHORYLATION AT SERINE 14974847
shows that both CDK2-cyclin E (lane 2) and CDK2-cyclin A
(lane 4) were able to phosphorylate BRCA1 in vitro, as evi-
denced by increased incorporation of radioactivity compared
to the reaction where no CDK-cyclin was added (lane 1). In
contrast, BRCA1 was a poor substrate for CDK6-cyclin D1
and CDC2-cyclin B1 in vitro (Fig. 3A, lanes 3 and 5, respec-
tively). All four CDK-cyclin complexes efficiently phosphory-
lated glutathione S-transferase–Rb in vitro (30). Two-dimen-
sional tryptic peptide maps of CDK2-cyclin A-treated BRCA1
(see below [Fig. 4B, map d]) and CDK2-cyclin E-treated
BRCA1 (data not shown) were very similar, both showing
highly increased phosphorylation of peptides p4 and p20, com-
pared to the map of untreated BRCA1 (see below [Fig. 4B,
map a]). Both CDK6-cyclin D1 and CDC2-cyclin B1 treatment
led to a comparably modest increase in phosphorylation of
peptides p4 and p20 (Fig. 3B). In conclusion, BRCA1 is a good
substrate for CDK2-cyclin A and CDK2-cyclin E in vitro.
S1497 and T967 are in vitro substrates for an endogenous
protein kinase activity and for CDK2. Since we have shown
above by in vivo labeling experiments that the C-terminal CDK
site of BRCA1 is phosphorylated, and since BRCA1 can be
phosphorylated by CDK2 in vitro, we examined whether CDK2
can specifically phosphorylate S1497 in vitro. We replaced
S1497 with alanine (S1497A) or threonine (S1497T) in the
context of Myc-BRCA1 wt and transfected each construct into
293T cells. The Myc tag allows immunoprecipitation of the
overexpressed but not endogenous BRCA1, allowing us to
investigate an individual mutant without interference of the
endogenous wild-type protein. The overexpressed proteins
were immunoprecipitated from cell lysates by using antibody
9E10 under native conditions as mentioned above in order to
coimmunoprecipitate associated protein kinases. The immu-
noprecipitates were subjected to in vitro kinase reactions, both
in the presence and in the absence of CDK2-cyclin A. All three
forms of the protein (wild type, S1497A, and S1497T) were
phosphorylated by CDK2-cyclin A in vitro (Fig. 4A, lanes 1 to
6). The labeled BRCA1 species were then subjected to tryptic
phosphopeptide analysis. The majority of phosphopeptides de-
rived from in vivo-labeled BRCA1 were also detected in the
case of in vitro-labeled untreated wild-type BRCA1 (data not
shown), among them peptides p4 (as judged by the same rel-
ative migration of phosphopeptide p4 derived from both
sources [Fig. 4C]; see below) and p20 (detected on longer
autoradiographic exposures [data not shown]). For the S1497A
mutant, the peptide map obtained was like that for wild-type
BRCA1 except that phosphopeptides p4 and p20 were absent
(Fig. 4B, map b). In the case of the S1497T mutation, PAA
(72) on phosphopeptide p4 revealed that its identity had
changed accordingly from phosphoserine (wild type; Fig. 4B,
map a) to phosphothreonine (Fig. 4B, map c). These results
demonstrate that an endogenous serine-threonine protein ki-
nase activity coimmunoprecipitates with and phosphorylates
BRCA1 at S1497 in vitro.
On the other hand, in vitro-phosphorylated BRCA1 re-
vealed a few phosphopeptides that were not detected in a map
of in vivo-labeled protein, including peptide pA (Fig. 4B, maps
a to c). Tryptic hydrolysates of in vitro-phosphorylated Myc-
tagged BRCA1 and in vivo-labeled endogenous BRCA1 from
HBL-100 cells or overexpressed BRCA1 from 293T cells were
compared by two-dimensional phosphopeptide analysis either
individually (Fig. 4C, maps a, b, and d) or in combination (Fig.
4C, maps c and e). Whereas phosphopeptides p4 derived from
in vitro-phosphorylated and in vivo-labeled BRCA1 run iden-
tically, pA is unique to in vitro-phosphorylated BRCA1 and
runs slightly distinct from p18 derived from in vivo-labeled
BRCA1 (Fig. 4C, maps c and e; note the ellipse-like spot
resulting from partial overlap of pA and p18 and its migration
relative to other phosphopeptides, e.g., p3, p11, and p9). As
was the case for wild-type BRCA1, peptide pA derived from
the S1497A and S1497T mutants consisted of phosphothreo-
nine (Fig. 4B, maps a to c). Another phosphopeptide that
appears only in maps of in vitro-phosphorylated but not in
vivo-labeled BRCA1 is peptide pB (Fig. 4B and C). It contains
phosphoserine (Fig. 4B, maps a and h) and is probably derived
from the BRCA1 sequence from aa 772 to 1050, since pB is not
present in a tryptic phosphopeptide map of an in vitro-phos-
phorylated BRCA1 fragment lacking this region (data not
shown). To locate pA within BRCA1, we divided BRCA1 into
four consecutive fragments, each of them Myc tagged (54).
These fragments were expressed in 293T cells, immunoprecipi-
tated with antibody 9E10, and subjected to in vitro kinase
reactions. Phosphopeptide mapping revealed that phos-
FIG. 3. BRCA1 is phosphorylated by CDK2-cyclin complexes in vitro. (A) Myc-BRCA1 wt was transiently transfected into 293T cells and immunoprecipitated with
antibody 9E10. The immunoprecipitate was phosphorylated in vitro in the absence (lane 1) or presence of a recombinant CDK-cyclin complex: CDK2-cyclin E (lane
2), CDK6-cyclin D1 (lane 3), CDK2-cyclin A (lane 4), or CDC2-cyclin B1 (lane 5). The reaction products were visualized by autoradiography following SDS-
polyacrylamide gel electrophoresis and transfer to a membrane. The 200-kDa protein marker is indicated on the left. (B) Two-dimensional tryptic phosphopeptide maps
derived from BRCA1 protein depicted in panel A. a, b, and c, maps of BRCA1 from lanes 1, 3, and 5, respectively. Phosphopeptides p4, pA, and pB are depicted in
4848 RUFFNER ET AL.MOL. CELL. BIOL.
phopeptide pA is derived from a BRCA1 fragment spanning
aa 772 to 1314 (data not shown). Since this sequence contains
the threonine CDK consensus site T(967)PNK, we hypothe-
sized that T967 may be the phosphorylated residue that gives
rise to phosphopeptide pA. We therefore mutated T967 to
aspartic acid (T967D) or to serine (T967S) in the context of
Myc-BRCA1 wt and analyzed these mutants as described
above (Fig. 4A, lanes 7 to 10). As predicted, the T967D mu-
tation abolished phosphopeptide pA (Fig. 4B, map g), whereas
its phosphoamino acid content changed accordingly to phos-
phoserine in the case of the T967S mutant (Fig. 4B, map h).
These results demonstrate that T967 of BRCA1 is phosphor-
ylated in vitro (but not in vivo) by a serine-threonine protein
kinase activity that coimmunoprecipitates with BRCA1.
In the case of wild-type BRCA1, treatment with CDK2-
cyclin A led to an increased phosphorylation of several pep-
tides, such as phosphopeptides p4 and p20 (Fig. 4B, map d). As
for untreated BRCA1, both phosphopeptides were abolished
by the S1497A mutation (Fig. 4B, map e), whereas the S1497T
mutation led to a change in their phosphoamino acid identities
from phosphoserine (Fig. 4B, map d) to phosphothreonine
(Fig. 4B, map f). Peptide pA also displayed increased phos-
phorylation due to CDK2 treatment in vitro (Fig. 4B, maps d
to f), but this effect was abrogated as a result of the T967D
mutation (Fig. 4B, map i). Moreover, pA of the CDK2-cyclin
A-treated T967S mutant consisted of phosphoserine (Fig. 4B,
map j). The three spots marked with an asterisk in Fig. 4B,
maps d to f, which probably represent incompletely digested
tryptic fragments containing phosphopeptide pA [predicted
sequence of pA is G(960)NETGLITPNK(970), where the
FIG. 4. An endogenous protein kinase activity as well as recombinant CDK2-
cyclin complexes phosphorylate BRCA1 residues T967 and S1497 in vitro. (A)
293T cells were transiently transfected with Myc-BRCA1 that was either wild
type (Wt; lanes 1 and 2) or mutated as indicated. The different BRCA1 species
were immunoprecipitated with antibody 9E10 and subjected to in vitro kinase
reactions in the absence (lanes 1, 3, 5, 7, and 9) or presence (lanes 2, 4, 6, 8, and
10) of the recombinant CDK2-cyclin A complex. The 200-kDa protein marker is
shown on the left. (B) Two-dimensional tryptic phosphopeptide maps derived
from the overexpressed BRCA1 protein species depicted in panel A. a to j, maps
of BRCA1 from lanes 1, 3, 5, 2, 4, 6, 7, 9, 8, and 10, respectively. (C) Two-
dimensional maps of mixtures of tryptic phosphopeptides derived from in vitro-
phosphorylated and in vivo-labeled BRCA1. a, in vitro-phosphorylated BRCA1
alone; b, in vivo-labeled endogenous BRCA1 from HBL-100 cells alone; c,
mixture of in vitro-phosphorylated BRCA1 and in vivo-labeled endogenous
BRCA1 from HBL-100 cells; d, in vivo-labeled overexpressed BRCA1 in 293T
cells alone; e, mixture of in vitro-phosphorylated BRCA1 and in vivo-labeled
overexpressed BRCA1 in 293T cells. Numbered circles indicate phosphopeptides
p3, p4, p9, p11, p18, p20, pA, and pB (for simplicity, only a few phosphopeptides
were circled); unlabeled circles indicate the lack of the respective peptides;
circles labeled by an asterisk mark phosphopeptides related to pA (see text); T
(threonine) and S (serine) indicate the phosphoamino acid content of the cor-
VOL. 19, 1999 BRCA1 PHOSPHORYLATION AT SERINE 14974849
phosphorylated T967 is underlined], since they all contain
phosphothreonine (as determined for CDK2-cyclin A-treated
wild-type BRCA1 [Fig. 4B, map d]), were eliminated by the
T967D mutation (Fig. 4B, map i), and at least the two peptides
that migrate faster in the chromatography dimension were
converted to phosphoserine as a consequence of the T967S
mutation (Fig. 4B, map j) (the phosphoamino acid identity of
the third peptide was not determined). In conclusion, S1497
and T967 of BRCA1 are phosphorylated by CDK2-cyclin A or
E in vitro.
Is CDK2 responsible for phosphorylation of BRCA1 at
S1497 in vivo? As described above, S1497 of BRCA1 is phos-
phorylated in vivo and can be phosphorylated by CDK2 in
vitro. Is the in vivo phosphorylation due to CDK2 action? To
address this question, we first tested whether BRCA1 and
CDK2 interact in vivo. Myc-BRCA1 wt was overexpressed
alone or in combination with CDK2 in the presence or absence
of cyclin A or cyclin E in 293T cells, and cell lysates were
subjected to anti-CDK2 immunoprecipitations under native
conditions. As shown in Fig. 5A, parts a and b, Myc-BRCA1 wt
coimmunoprecipitated with CDK2 in a manner proportional
to the amount of CDK2 (e.g., more Myc-BRCA1 wt was im-
munoprecipitated from lysates where CDK2 was overex-
pressed [Fig. 5A, parts a and b, lanes 2, 3, and 6] than from a
lysate where only endogenous CDK2 was present [lane 1]).
These results suggest that BRCA1 and CDK2 interact. This
interaction was not dependent on cotransfected cyclin A or E,
presumably because the levels of endogenous cyclin A and/or
E are sufficient to mediate interaction. Interestingly, when
cyclin A or E was cotransfected with CDK2 (Fig. 5A, part c,
lanes 2 and 3), BRCA1 migrated slightly slower than BRCA1
expressed alone or in combination with only CDK2 (lanes 1
and 6), suggesting that overexpression of CDK2-cyclin A or
CDK-cyclin E led to increased phosphorylation of BRCA1.
Furthermore, BRCA1 and CDK2 were coimmunoprecipitated
from HeLa cells that stably express untagged, full-length wild-
type BRCA1 protein (57a). As shown in Fig. 5B, BRCA1 was
coimmunoprecipitated by a CDK2-specific antibody (lane 2)
but not by an antibody that had been preincubated with the
cognate CDK2 blocking peptide (lane 3). A p65-specific con-
trol antibody was not able to precipitate BRCA1 (lane 4).
Similarly, a cyclin A-specific antibody coimmunoprecipitated
BRCA1 as well as CDK2 (lane 5), suggesting that the CDK2-
cyclin A complex interacts with BRCA1 in vivo. Under our
experimental conditions, a cyclin E-specific antibody was able
to coimmunoprecipitate CDK2 but not BRCA1 (lane 6).
Whether this means that CDK2-cyclin E does not interact with
BRCA1 in vivo remains to be investigated.
To test the effect of CDK2 on BRCA1 phosphorylation in
vivo, we transiently overexpressed CDK2 dn (with an
Asp1453Asn mutation that renders the kinase inactive but
does not affect its ability to associate with cyclins ) in 293T
cells and measured the phosphorylation state of endogenous
BRCA1 by Western blot analysis. CDK2 dn caused a decrease
in BRCA1 phosphorylation compared to the mock-transfected
sample, as judged by an increase in BRCA1 mobility (Fig. 6A,
lanes 1 and 2). BRCA1 from both sources was then immuno-
precipitated and treated with BAP as previously described
(57). Phosphatase treatment led to a BRCA1 species with
increased mobility in both cases (compare lanes 3 and 4 and
lanes 5 and 6), showing that the CDK2 dn had caused partial
BRCA1 dephosphorylation. These results demonstrate that
inhibition of CDK2 activity leads to a decrease in BRCA1
phosphorylation. Furthermore, we cotransfected a Myc-tagged
BRCA1 expression plasmid with either CDK2 dn or a control
plasmid into 293T cells. Antibody 9E10 was used to immuno-
precipitate BRCA1, and the immunoprecipitates were sub-
jected to in vitro kinase reactions as described above. Expres-
sion of CDK2 dn led to decreased levels of cotransfected
Myc-BRCA1 wt protein compared to expression of Myc-
BRCA1 wt alone (Fig. 6B, top, lanes a and b). Importantly,
peptide mapping revealed that the intensities of spots p4 and
pA were markedly reduced relative to other phosphopeptides
by the CDK2 dn, e.g., pB (Fig. 6B, maps a and b), suggesting
FIG. 5. BRCA1 coimmunoprecipitates with CDK2 and cyclin A. (A) Myc-
BRCA1 wt coimmunoprecipitates with CDK2. 293T cells were transfected with
expression plasmids for the following proteins: lane 1, Myc-BRCA1 wt; lane 2,
Myc-BRCA1 wt, cyclin A, and CDK2; lane 3, Myc-BRCA1 wt, cyclin E, and
CDK2; lane 4, cyclin A and CDK2; lane 5, cyclin E and CDK2; lane 6, Myc-
BRCA1 wt and CDK2; and lane 7, CDK2. CDK2-containing complexes were
immunoprecipitated (I.P.) from each sample with a rabbit CDK2 antibody (?-
CDK2) and analyzed after separation on SDS-polyacrylamide gels by Western
blot analysis for the presence of Myc-BRCA1 wt, using antibody 9E10 (a, bottom
row), and CDK2, using a mouse anti-CDK2 monoclonal antibody (b, bottom
row). The two top rows in panels a and b (Input) represent the relative amounts
of Myc-BRCA1 wt and CDK2 in the cell lysates before immunoprecipitation.
Panel c shows that in a duplicate experiment, Myc-BRCA1 wt from cell lysates
overexpressing proteins as indicated in lanes 1, 2, 3, and 6 (in panels, a and b) was
further resolved on an SDS-polyacrylamide gel (compared to panel a, top row)
in order to increase resolution. (B) Untagged, full-length wild-type BRCA1
coimmunoprecipitates with CDK2 and cyclin A. Lysates from a HeLa cell line
stably overexpressing BRCA1 (57a) were subjected to immunoprecipitations
using the following antibodies: lane 1, an aliquot before immunoprecipitation;
lane 2, anti-CDK2; lane 3, anti-CDK2, after preincubation with the blocking
peptide; lane 4, anti-NF-?B p65 as a control; lane 5, anti-cyclin A; and lane 6,
anti-cyclin E. Western blot analysis revealed the presence of BRCA1 and CDK2,
as shown in the upper and lower panels, respectively. The 250- and 42-kDa
markers are indicated on the left.
4850 RUFFNER ET AL.MOL. CELL. BIOL.
that the level of the associated in vitro protein kinase activity
was decreased by the CDK2 dn.
The above results are consistent with a role of CDK2 in
phosphorylating BRCA1 in vivo. However, they do not prove
that CDK2 directly phosphorylates BRCA1, since CDK2 dn
blocks cells in the G1phase of the cell cycle (71). Therefore,
the decrease in BRCA1 phosphorylation could be due to an
effect of the cell cycle block on BRCA1 rather than to the
mutant CDK2 directly. We therefore performed in vitro kinase
reactions on Myc-BRCA1 wt immunoprecipitated from lysates
of transiently transfected 293T cells, in the presence of the
CDI p16 or p21 or in the presence of butyrolactone I, which is
an inhibitor of CDC2, CDK2, and CDK5 but not CDK4 and
CDK6 (43). As shown in Fig. 6C, p21 and butyrolactone I but
not p16 decreased the relative amount of phosphopeptides p4
and pA compared to mock-treated BRCA1. These results sug-
gest that the cellular protein kinase activity that phosphory-
lates BRCA1 at S1497 and T967 in vitro is in fact CDK2 and
that CDK2 phosphorylates BRCA1 at S1497 in vivo.
The results presented here demonstrate that S1497 of
BRCA1 is phosphorylated in vivo. They further establish
BRCA1 as an in vivo substrate of CDK2-cyclin complexes, by
virtue of at least partial fulfillment of the following criteria, as
proposed elsewhere (48). First, BRCA1 is phosphorylated by
CDK2-cyclin A or CDK2-cyclin E in vitro. Second, as previ-
ously reported by us and others, BRCA1 is phosphorylated in
vivo, and the G1/S-specific increase in phosphorylation is con-
sistent with the cell cycle-dependent activation of CDK2-cyclin
complexes. Third, S1497 is phosphorylated in vivo and by
CDK2-cyclin A or E in vitro. An endogenous serine-threonine
protein kinase activity that is sensitive to p21, to butyrolactone
I, and to a dominant negative form of CDK2 phosphorylates
BRCA1 at S1497 in vitro. The fourth condition, that S1497
phosphorylation should change the properties of BRCA1 in a
way consistent with a corresponding G1/S-phase event, cannot
be addressed yet since the molecular function of BRCA1 is
unknown. However, determining the cellular effects of express-
ing the mutant forms S1497A (which cannot be phosphory-
lated) and S1497D or S1497E (the negative charge may mimic
the phosphate moiety on the serine residue) may provide in-
sights into normal BRCA1 function during cell cycle progres-
FIG. 6. Phosphorylation of BRCA1 is decreased by inhibition of CDK2. (A)
CDK2 dn decreases the phosphorylation state of endogenous BRCA1. 293T cells
were transiently transfected with a control plasmid (lane 1) or a plasmid express-
ing CDK2 dn (lane 2). Endogenous BRCA1 from both sources was immunopre-
cipitated with Ab-D prior to treatment with (lanes 4 and 6) or without (lanes 3
and 5) BAP before analysis by immunoblotting with Ab-D. Filled arrow, BRCA1
from mock-transfected cells; open arrow, faster-migrating BRCA1 species due to
expression of CDK2 dn; asterisk, BAP-treated BRCA1. (B) BRCA1 is phos-
phorylated in vitro by a cellular kinase activity that is sensitive to CDK2 dn. 293T
cells were cotransfected with a plasmid expressing Myc-BRCA1 wt and either a
control plasmid (lane a) or CDK2 dn (lane b). BRCA1 was immunoprecipitated
under native conditions with antibody 9E10, and in vitro kinase reactions were
performed on the immunoprecipitates. a and b, two-dimensional tryptic phos-
phopeptide maps of the BRCA1 protein species shown in lanes a and b, respec-
tively, from a duplicate experiment. (C) p21 and butyrolactone I but not p16
inhibit the endogenous protein kinase activity that phosphorylates BRCA1 at
T967 and S1497 in vitro. Myc-BRCA1 wt was subjected to in vitro kinase reac-
tions as described above. Lane a, no CDI added; lanes b, c, and d, p16, p21, and
butyrolactone I (Butyr), respectively, added. a to d, phosphopeptide maps of
BRCA1 protein shown in lanes a to d, respectively. The top portions of panels B
and C represent the amount of (Western blot analysis using antibody 9E10
[bottom rows]) and the in vitro kinase activity on (radioactive incorporation onto
BRCA1 [top rows]) BRCA1. The 250- and 200-kDa markers are shown on the
VOL. 19, 1999 BRCA1 PHOSPHORYLATION AT SERINE 14974851
sion. We are currently investigating these mutants toward this
Phosphorylation of S1497 by CDK2-cyclin complexes is con-
cordant with increased phosphorylation of BRCA1 in late G1
and S phases. We propose that CDK2-cyclin E and/or CDK2-
cyclin A phosphorylate BRCA1 during late G1and/or S phase,
respectively. However, we cannot rule out the possibility that
other protein kinases also phosphorylate BRCA1 at the G1/S
transition, since we have previously detected that BRCA1 un-
dergoes extensive hyperphosphorylation during this cell cycle
phase (57). For S1497 to be solely responsible for the observed
BRCA1 hyperphosphorylation at G1/S, one would expect that
the sum of phosphorylation at all the other phosphorylated
sites is comparably marginal. We have not formally demon-
strated by32P metabolic labeling experiments that phosphor-
ylation of S1497 increases relative to that of other phosphor-
ylation sites during this phase, since our standard labeling
conditions themselves induce increased phosphorylation of
BRCA1 (possibly due to
would therefore interfere with cell cycle-specific phosphoryla-
tion events (57a). Alternatively, we could generate antibodies
that specifically recognize phospho-S1497 or phospho-T967 to
study cell cycle-specific phosphorylation by using nonradioac-
tive methods. We are currently generating inducible cell lines
that express the full-length BRCA1 mutants S1497A and
T967D in a regulated manner; these cell lines will be useful to
study the effect of the lack of these phosphorylation sites on
the overall G1/S-specific phosphorylation of BRCA1.
Two previous reports had shown that BRCA1 associates
with endogenous CDK2 and cyclin A but not with cyclin E in
lysates of HBL-100 or CAL-51 (a human breast cancer cell
line) cells (16, 74), in agreement with our data obtained from
the BRCA1-overexpressing HeLa cell line. These findings and
our in vitro phosphorylation studies suggest that BRCA1 is a
substrate for CDK2-cyclin A (but possibly not for CDK2-cyclin
E) in vivo. Moreover, both groups provided data showing that
BRCA1 also interacts with cyclin D (cyclin D1), suggesting that
BRCA1 is phosphorylated by a kinase(s) associated with cyclin
D. At least in vitro, CDK6-cyclin D1 was able to only weakly
phosphorylate BRCA1 (Fig. 3). The same was true for CDC2-
cyclin B1, in agreement with the notion that BRCA1 is not
phosphorylated by CDC2 or kinases associated with cyclin B in
lysates from HBL-100 cells (16). Nevertheless, CDK6-cyclin
D1, CDC2-cyclin B1, and other CDK-cyclin complexes could
possibly phosphorylate BRCA1 in vivo, at least in certain cell
types. For example, in lysates of CAL-51 cells, BRCA1 has
been reported to associate with cyclin B1 and CDC2 (74).
Other kinases may also be involved in phosphorylating
BRCA1. Two groups have reported that BRCA1 is a tyrosine
phosphoprotein (74, 79). However, our previous studies re-
vealed that BRCA1 is predominantly phosphorylated on serine
and threonine residues, at least in the cell lines analyzed (57).
Furthermore, a kinase activity that associates with and phos-
phorylates a BRCA1 fragment containing aa 329 to 435 in vitro
was identified (9).
What is the consequence of phosphorylating S1497 on
BRCA1? The addition or removal of the phosphate group may
affect the interaction with other proteins. So far, no protein
that interacts with BRCA1 in the domain encompassing S1497
has been identified. Two proteins that may be considered as
candidates are CBP and BRCA2, which interact with BRCA1
sequences from aa 1314 to 1863 and 1314 to 1756, respectively
(13, 54). However, in vitro CBP binding occurs irrespective of
BRCA1 phosphorylation (54). Whether the same is true for
BRCA2 is now under investigation, in particular with regard to
S1497 phosphorylation. It is also possible that a change in
32P-induced DNA damage) and
phosphorylation leads to a change in BRCA1’s subnuclear
localization, which changes in response to cell cycle progres-
sion (from G1to S phase) and DNA damage, two events that
cause phosphorylation of BRCA1. Alternatively, the S1497
phosphorylation state may regulate a so far undetected intrin-
sic activity of BRCA1, possibly by modulating the protein’s
The fact that the SPXK motif (where the S is S1497 in
human BRCA1) is conserved among human, mouse, and rat
BRCA1 proteins underlines the functional importance of this
motif (1, 5) (in dogs, SP but not the K is conserved ).
However, no tumor-associated mutation has been found at
either of these residues (8). A mutation in close proximity has
been compiled as an unclassified variant, where a conserved
arginine is changed to methionine (R1495M ). If this mu-
tation turns out to be associated with cancer development, one
could investigate whether R1495 influences phosphorylation of
the neighboring CDK site, such as by determining the speci-
ficity for CDK2-cyclin complexes. On the other hand, the
TPNK motif (where the T is T967 in human BRCA1) is not
conserved, and no mutations have yet been identified either
within this motif or at conserved amino acids in the vicinity (1,
5, 8, 66). Consistently, in vivo labeling has not revealed T967
phosphorylation of either endogenous or overexpressed
BRCA1 (Fig. 1 and 4C). However, T967 is phosphorylated in
vitro by a cellular serine-threonine protein kinase activity. Why
is T967 phosphorylation detected in vitro but not in vivo? One
explanation is that T967 is not in fact phosphorylated in vivo
but rather is phosphorylated in vitro by the same protein kinase
activity that is associated with S1497. This scenario would re-
quire that phosphorylation of T967 in vivo is impaired by
BRCA1’s secondary structure, or a protein that masks this site,
and that these constraints are lost upon cell lysis and/or im-
munoprecipitation in vitro kinase procedures. Consequently,
both S1497 and T967 are excellent substrates for CDK2-cyclin
complexes in vitro. Alternatively, phosphorylation of T967 in
vivo could occur very transiently or at a low stoichiometric
level or turnover rate and would therefore escape detection by
in vivo labeling.
The results presented above are consistent with a direct
action of CDK2-cyclin complexes on BRCA1. We note, how-
ever, that it cannot be ruled out that CDK2 and BRCA1 exist
in a complex with another serine-threonine protein kinase
which is phosphorylated by CDK2 and in turn phosphorylates
BRCA1, as opposed to a direct action of CDK2 on BRCA1. A
subset of CDK-cyclin substrates contain an RXL (L designates
leucine) motif that is required for docking of the CDK-cyclin
complex via the cyclin and for phosphorylation (2, 58). BRCA1
contains nine such motifs, seven in the N-terminal half and two
at the C terminus (aa 1699 to 1701 and 1762 to 1764), which
may be critical for CDK2-cyclin A or E-mediated phosphory-
lation of endogenous BRCA1. The fact that S1497 was also
phosphorylated in the overexpressed BRCA1 fragment 1314-
1652 (which contains no RXL motif) suggests either that S1497
phosphorylation of endogenous BRCA1 occurs in an RXL-
independent manner or that overexpressed BRCA1 does not
rely on the RXL motif due to an increased substrate to CDK2-
cyclin ratio. Furthermore, it is possible that other protein ki-
nases contribute to the phosphorylation of S1497 in vivo, for
example, mitogen-activated protein (MAP) kinases. MAP ki-
nases and CDKs display similar proline-directed substrate
specificities (48), although MAP kinases generally require an L
or P and an intervening residue preceding S/T-P (consensus
L/P-X-S/T-P) and are therefore probably not involved in S1497
In summary, we present evidence that BRCA1 is a physio-
4852RUFFNER ET AL.MOL. CELL. BIOL.
logical substrate of CDK2-cyclin complexes. Elucidation of the
functional consequences of S1497 phosphorylation will shed
light on the role that BRCA1 plays during cell cycle progres-
sion and may help in understanding why cells become cancer-
ous in the absence of the functional protein.
We thank Tim Mayall for providing recombinant p16 and p21, Nik
Somia for providing pCL-MFG-MCS and for excellent suggestions, Jill
Meisenhelder for providing 9E10 antibody and for technical advice,
Mirta Grifman and Matthew Weitzman for the pRK5-cyclin A expres-
sion construct, Lamya Shihabuddin for technical advice, and Brian
Spain, Chris Larson, Tal Kafri, and other members of the Verma
laboratory for valuable discussions. We thank Jean E. Rivier for his
interest in our work.
H.R. is supported by consecutive funds from the Schweizerische
Nationalfonds fu ¨r wissenschaftliche Forschung, grant 823A-046698,
and the California Breast Cancer Research Program of the University
of California, grant 4FB-0102. W.J. is supported by a postdoctoral
fellowship from the American Cancer Society. I.M.V., T.H., and
A.G.C. are supported by grants from the National Institutes of Health.
I.M.V. is an American Cancer Society Professor of Molecular Biology,
and he is also supported by the Elsa Pardee Foundation. T.H. is a
Frank and Else Schilling American Cancer Society Research Profes-
1. Abel, K. J., J. Xu, G. Y. Yin, R. H. Lyons, M. H. Meisler, and B. L. Weber.
1995. Mouse Brca1: localization, sequence analysis and identification of
evolutionarily conserved domains. Hum. Mol. Genet. 4:2265–2273.
2. Adams, P. D., W. R. Sellers, S. K. Sharma, A. D. Wu, C. M. Nalin, and W. G.
Kaelin, Jr. 1996. Identification of a cyclin-cdk2 recognition motif present in
substrates and p21-like cyclin-dependent kinase inhibitors. Mol. Cell. Biol.
3. Anderson, S. F., B. P. Schlegel, T. Nakajima, E. S. Wolpin, and J. D. Parvin.
1998. BRCA1 protein is linked to the RNA polymerase II holoenzyme
complex via RNA helicase A. Nat. Genet. 19:254–256.
4. Beijersbergen, R. L., and R. Bernards. 1996. Cell cycle regulation by the
retinoblastoma family of growth inhibitory proteins. Biochim. Biophys. Acta
5. Bennett, L. M., H. A. Brownlee, S. Hagevik, A. Haugen-Strano, and R. W.
Wiseman. 1996. Evolutionarily conserved domains of the rat BRCA1 pro-
tein. Proc. Annu. Meet. Am. Assoc. Cancer Res. 37:514–515.
6. Bork, P., K. Hofmann, P. Bucher, A. F. Neuwald, S. F. Altschul, and E. V.
Koonin. 1997. A superfamily of conserved domains in DNA damage-respon-
sive cell cycle checkpoint proteins. FASEB J. 11:68–76.
7. Boyle, W. J., P. van der Geer, and T. Hunter. 1991. Phosphopeptide mapping
and phosphoamino acid analysis by two-dimensional separation on thin-layer
cellulose plates. Methods Enzymol. 201:110–149.
8. Breast cancer information core. 12 March 1999, revision date. www.nhgri
.nih.gov/Intramural_research/Lab_transfer/bic/. [Online.] [5 May 1999, last
9. Burke, T. F., K. S. Cocke, S. J. Lemke, E. Angleton, G. W. Becker, and R. P.
Beckmann. 1998. Identification of a BRCA1-associated kinase with potential
biological relevance. Oncogene 16:1031–1040.
10. Callebaut, I., and J. P. Mornon. 1997. From BRCA1 to RAP1: a widespread
BRCT module closely associated with DNA repair. FEBS Lett. 400:25–30.
11. Chapman, M. S., and I. M. Verma. 1996. Transcriptional activation by
BRCA1. Nature 382:678–679.
12. Chen, C. F., S. Li, Y. Chen, P. L. Chen, Z. D. Sharp, and W. H. Lee. 1996.
The nuclear localization sequences of the BRCA1 protein interact with the
importin-alpha subunit of the nuclear transport signal receptor. J. Biol.
13. Chen, J., D. P. Silver, D. Walpita, S. B. Cantor, A. F. Gazdar, G. Tomlinson,
F. J. Couch, B. L. Weber, T. Ashley, D. M. Livingston, and R. Scully. 1998.
Stable interaction between the products of the BRCA1 and BRCA2 tumor
suppressor genes in mitotic and meiotic cells. Mol. Cell 2:317–328.
14. Chen, Y., C. F. Chen, D. J. Riley, D. C. Allred, P. L. Chen, D. Von Hoff, C. K.
Osborne, and W. H. Lee. 1995. Aberrant subcellular localization of BRCA1
in breast cancer. Science 270:789–791.
15. Chen, Y., P.-L. Chen, D. J. Riley, W.-H. Lee, D. C. Allred, and C. K. Osborne.
1996. Location of BRCA1 in human breast and ovarian cancer cells. Science
16. Chen, Y., A. A. Farmer, C. F. Chen, D. C. Jones, P. L. Chen, and W. H. Lee.
1996. BRCA1 is a 220-kDa nuclear phosphoprotein that is expressed and
phosphorylated in a cell cycle-dependent manner. Cancer Res. 56:3168–
17. Claus, E., N. Risch, and W. Thompson. 1991. Genetic analysis of breast
cancer in the cancer and steroid hormone study. Am. J. Hum. Genet. 48:
18. Easton, D. F., D. T. Bishop, D. Ford, G. P. Crockford, and The Breast
Cancer Linkage Consortium. 1993. Genetic linkage analysis in familial
breast and ovarian cancer: results from 214 families. Am. J. Hum. Genet.
19. Fisher, R. P. 1997. CDKs and cyclins in transition(s). Curr. Opin. Genet.
20. Futreal, P. A., Q. Liu, D. Shattuck-Eidens, C. Cochran, K. Harshman, S.
Tavtigian, L. M. Bennett, A. Haugen-Strano, J. Swensen, Y. Miki, et al. 1994.
BRCA1 mutations in primary breast and ovarian carcinomas. Science 266:
21. Garcia-Patino, E., B. Gomendio, M. Provencio, J. M. Silva, J. M. Garcia, P.
Espana, and F. Bonilla. 1998. Germ-line BRCA1 mutations in women with
sporadic breast cancer: clinical correlations. J. Clin. Oncol. 16:115–120.
22. Gottlieb, T. M., and M. Oren. 1996. p53 in growth control and neoplasia.
Biochim. Biophys. Acta 1287:77–102.
23. Gowen, L. C., A. V. Avrutskaya, A. M. Latour, B. H. Koller, and S. A. Leadon.
1998. BRCA1 required for transcription-coupled repair of oxidative DNA
damage. Science 281:1009–1012.
24. Gowen, L. C., B. L. Johnson, A. M. Latour, K. K. Sulik, and B. H. Koller.
1996. Brca1 deficiency results in early embryonic lethality characterized by
neuroepithelial abnormalities. Nat. Genet. 12:191–194.
25. Greenman, J., S. Mohammed, D. Ellis, S. Watts, G. Scott, L. Izatt, D.
Barnes, E. Solomon, S. Hodgson, and C. Mathew. 1998. Identification of
missense and truncating mutations in the BRCA1 gene in sporadic and
familial breast and ovarian cancer. Genes Chromosomes Cancer 21:244–249.
25a.Grifman, M., and M. Weitzman (The Salk Institute). Unpublished data.
26. Gudas, J. M., T. Li, H. Nguyen, D. Jensen, F. J. Rauscher III, and K. H.
Cowan. 1996. Cell cycle regulation of BRCA1 messenger RNA in human
breast epithelial cells. Cell Growth Differ. 7:717–723.
27. Hakem, R., J. L. de la Pompa, C. Sirard, R. Mo, M. Woo, A. Hakem, A.
Wakeham, J. Potter, A. Reitmair, F. Billia, E. Firpo, C. C. Hui, J. Roberts,
J. Rossant, and T. W. Mak. 1996. The tumor suppressor gene Brca1 is
required for embryonic cellular proliferation in the mouse. Cell 85:1009–
28. Harper, J. W., and S. J. Elledge. 1996. Cdk inhibitors in development and
cancer. Curr. Opin. Genet. Dev. 6:56–64.
29. Hunter, T., and J. Pines. 1994. Cyclins and cancer II: cyclin D and CDK
inhibitors come of age. Cell 79:573–582.
29a.Jiang, W. Unpublished data.
30. Jiang, W., G. Jimenez, N. J. Wells, T. J. Hope, G. M. Wahl, T. Hunter, and
R. Fukunaga. 1998. PRC1: a human mitotic spindle-associated CDK sub-
strate protein required for cytokinesis. Mol. Cell 2:877–885.
31. Kanemitsu, M. Y., W. Jiang, and W. Eckhart. 1998. Cdc2-mediated phos-
phorylation of the gap junction protein, connexin43, during mitosis. Cell
Growth Differ. 9:13–21.
32. Kelsell, D. P., D. M. Black, D. T. Bishop, and N. K. Spurr. 1993. Genetic
analysis of the BRCA1 region in a large breast/ovarian family: refinement of
the minimal region containing BRCA1. Hum. Mol. Genet. 2:1823–1828.
33. Kelsell, D. P., N. K. Spurr, D. M. Barnes, B. Gusterson, and D. T. Bishop.
1996. Combined loss of BRCA1/BRCA2 in grade 3 breast carcinomas. Lan-
34. Kerangueven, F., F. Eisinger, T. Noguchi, F. Allione, V. Wargniez, C. Eng, G.
Padberg, C. Theillet, J. Jacquemier, M. Longy, H. Sobol, and D. Birnbaum.
1997. Loss of heterozygosity in human breast carcinomas in the ataxia tel-
angiectasia, Cowden disease and BRCA1 gene regions. Oncogene 14:339–
35. Kinzler, K. W., and B. Vogelstein. 1997. Cancer-susceptibility genes. Gate-
keepers and caretakers. Nature 386:761–763.
36. Koonin, E. V., S. F. Altschul, and P. Bork. 1996. BRCA1 protein products...
functional motifs... Nat. Genet. 13:266–268.
37. Lane, T. F., C. Deng, A. Elson, M. S. Lyu, C. A. Kozak, and P. Leder. 1995.
Expression of Brca1 is associated with terminal differentiation of ectoder-
mally and mesodermally derived tissues in mice. Genes Dev. 9:2712–2722.
38. Langston, A. A., K. E. Malone, J. D. Thompson, J. R. Daling, and E. A.
Ostrander. 1996. BRCA1 mutations in a population-based sample of young
women with breast cancer. N. Engl. J. Med. 334:137–142.
39. Lees, E. 1995. Cyclin dependent kinase regulation. Curr. Opin. Cell Biol.
40. Lewin, B. 1990. Driving the cell cycle: M phase kinase, its partners, and
substrates. Cell 61:743–752.
41. Liu, C. Y., A. Flesken-Nikitin, S. Li, Y. Zeng, and W. H. Lee. 1996. Inacti-
vation of the mouse Brca1 gene leads to failure in the morphogenesis of the
egg cylinder in early postimplantation development. Genes Dev. 10:1835–
42. Marquis, S. T., J. V. Rajan, A. Wynshaw-Boris, J. Xu, G. Y. Yin, K. J. Abel,
B. L. Weber, and L. A. Chodosh. 1995. The developmental pattern of Brca1
expression implies a role in differentiation of the breast and other tissues.
Nat. Genet. 11:17–26.
VOL. 19, 1999 BRCA1 PHOSPHORYLATION AT SERINE 1497 4853
43. Meijer, L. 1996. Chemical inhibitors of cyclin-dependent kinases. Trends
Cell Biol. 6:393–397.
44. Merajver, S. D., T. M. Pham, R. F. Caduff, M. Chen, E. L. Poy, K. A. Cooney,
B. L. Weber, F. S. Collins, C. Johnston, and T. S. Frank. 1995. Somatic
mutations in the BRCA1 gene in sporadic ovarian tumours. Nat. Genet.
45. Miki, Y., J. Swensen, D. Shattuck-Eidens, P. A. Futreal, K. Harshman, S.
Tavtigian, Q. Liu, C. Cochran, L. M. Bennett, W. Ding, et al. 1994. A strong
candidate for the breast and ovarian cancer susceptibility gene BRCA1.
46. Monteiro, A. N., A. August, and H. Hanafusa. 1996. Evidence for a tran-
scriptional activation function of BRCA1 C-terminal region. Proc. Natl.
Acad. Sci. USA 93:13595–13599.
47. Neuhausen, S. L., and C. J. Marshall. 1994. Loss of heterozygosity in familial
tumors from three BRCA1-linked kindreds. Cancer Res. 54:6069–6072.
48. Nigg, E. A. 1993. Cellular substrates of p34cdc2and its companion cyclin-
dependent kinases. Trends Cell Biol. 3:296–301.
49. Nigg, E. A. 1995. Cyclin-dependent protein kinases: key regulators of the
eukaryotic cell cycle. Bioessays 17:471–480.
50. Norbury, C., and P. Nurse. 1992. Animal cell cycles and their control. Annu.
Rev. Biochem. 61:441–470.
51. Orend, G., T. Hunter, and E. Ruoslahti. 1998. Cytoplasmic displacement of
cyclin E-cdk2 inhibitors p21Cip1 and p27Kip1 in anchorage-independent
cells. Oncogene 16:2575–2583.
52. Ouchi, T., A. N. Monteiro, A. August, S. A. Aaronson, and H. Hanafusa.
1998. BRCA1 regulates p53-dependent gene expression. Proc. Natl. Acad.
Sci. USA 95:2302–2306.
53. Ozcelik, H., M. D. To, J. Couture, S. B. Bull, and I. L. Andrulis. 1998.
Preferential allelic expression can lead to reduced expression of BRCA1 in
sporadic breast cancers. Int. J. Cancer 77:1–6.
54. Pao, G., R. Janknecht, H. Ruffner, T. Hunter, and I. M. Verma. CBP/p300
interact functionally with BRCA1 as transcriptional co-activators. Submitted
55. Papa, S., D. Seripa, G. Merla, C. Gravina, M. Giai, P. Sismondi, M. Rinaldi,
A. Serra, G. Saglio, and V. M. Fazio. 1998. Identification of a possible
somatic BRCA1 mutation affecting translation efficiency in an early-onset
sporadic breast cancer patient. J. Natl. Cancer Inst. 90:1011–1012.
56. Pines, J. 1993. Cyclins and cyclin-dependent kinases: take your partners.
Trends Biochem. Sci. 18:195–197.
57. Ruffner, H., and I. M. Verma. 1997. BRCA1 is a cell cycle-regulated nuclear
phosphoprotein. Proc. Natl. Acad. Sci. USA 94:7138–7143.
57a.Ruffner, H. Unpublished data.
58. Schulman, B. A., D. L. Lindstrom, and E. Harlow. 1998. Substrate recruit-
ment to cyclin-dependent kinase 2 by a multipurpose docking site on cyclin
A. Proc. Natl. Acad. Sci. USA 95:10453–10458.
59. Scully, R., S. F. Anderson, D. M. Chao, W. Wei, L. Ye, R. A. Young, D. M.
Livingston, and J. D. Parvin. 1997. BRCA1 is a component of the RNA
polymerase II holoenzyme. Proc. Natl. Acad. Sci. USA 94:5605–5610.
60. Scully, R., J. Chen, R. L. Ochs, K. Keegan, M. Hoekstra, J. Feunteun, and
D. M. Livingston. 1997. Dynamic changes of BRCA1 subnuclear location
and phosphorylation state are initiated by DNA damage. Cell 90:425–435.
61. Scully, R., J. Chen, A. Plug, Y. Xiao, D. Weaver, J. Feunteun, T. Ashley, and
D. M. Livingston. 1997. Association of BRCA1 with Rad51 in mitotic and
meiotic cells. Cell 88:265–275.
62. Scully, R., S. Ganesan, M. Brown, J. A. De Caprio, S. A. Cannistra, J.
Feunteun, S. Schnitt, and D. M. Livingston. 1996. Location of BRCA1 in
human breast and ovarian cancer cells. Science 272:123–125.
63. Shao, N., Y. L. Chai, E. Shyam, P. Reddy, and V. N. Rao. 1996. Induction of
apoptosis by the tumor suppressor protein BRCA1. Oncogene 13:1–7.
64. Somasundaram, K., H. Zhang, Y. X. Zeng, Y. Houvras, Y. Peng, G. S. Wu,
J. D. Licht, B. L. Weber, and W. S. El-Deiry. 1997. Arrest of the cell cycle by
the tumour-suppressor BRCA1 requires the CDK-inhibitor p21WAF1/CiP1.
64a.Somia, N. Unpublished data.
65. Sourvinos, G., and D. A. Spandidos. 1998. Decreased BRCA1 expression
levels may arrest the cell cycle through activation of p53 checkpoint in
human sporadic breast tumors. Biochem. Biophys. Res. Commun. 245:75–
66. Szabo, C. I., L. A. Wagner, L. V. Francisco, J. C. Roach, R. Argonza, M. C.
King, and E. A. Ostrander. 1996. Human, canine and murine BRCA1 genes:
sequence comparison among species. Hum. Mol. Genet. 5:1289–1298.
67. Thakur, S., H. B. Zhang, Y. Peng, H. Le, B. Carroll, T. Ward, J. Yao, L. M.
Farid, F. J. Couch, R. B. Wilson, and B. L. Weber. 1997. Localization of
BRCA1 and a splice variant identifies the nuclear localization signal. Mol.
Cell. Biol. 17:444–452.
68. Thomas, J. E., M. Smith, B. Rubinfeld, M. Gutowski, R. P. Beckmann, and
P. Polakis. 1996. Subcellular localization and analysis of apparent 180-kDa
and 220-kDa proteins of the breast cancer susceptibility gene, BRCA1.
J. Biol. Chem. 271:28630–28635.
69. Thomas, J. E., M. Smith, J. L. Tonkinson, B. Rubinfeld, and P. Polakis.
1997. Induction of phosphorylation on BRCA1 during the cell cycle and after
DNA damage. Cell Growth Differ. 8:801–809.
70. Thompson, M. E., R. A. Jensen, P. S. Obermiller, D. L. Page, and J. T. Holt.
1995. Decreased expression of BRCA1 accelerates growth and is often
present during sporadic breast cancer progression. Nat. Genet. 9:444–450.
71. van den Heuvel, S., and E. Harlow. 1993. Distinct roles for cyclin-dependent
kinases in cell cycle control. Science 262:2050–2054.
72. van der Geer, P., K. Luo, B. M. Sefton, and T. Hunter. 1994. Phosphopeptide
mapping and phosphoamino acid analysis on cellulose thin-layer plates, p.
422–448. In J. E. Celis (ed.), Cell biology: a laboratory handbook. Academic
Press, New York, N.Y.
73. Vaughn, J. P., P. L. Davis, M. D. Jarboe, G. Huper, A. C. Evans, R. W.
Wiseman, A. Berchuck, J. D. Iglehart, P. A. Futreal, and J. R. Marks. 1996.
BRCA1 expression is induced before DNA synthesis in both normal and
tumor-derived breast cells. Cell Growth Differ. 7:711–715.
74. Wang, H., N. Shao, Q. M. Ding, J. Cui, E. S. Reddy, and V. N. Rao. 1997.
BRCA1 proteins are transported to the nucleus in the absence of serum and
splice variants BRCA1a, BRCA1b are tyrosine phosphoproteins that asso-
ciate with E2F, cyclins and cyclin dependent kinases. Oncogene 15:143–157.
75. Wigler, M., A. Pellicer, S. Silverstein, R. Axel, G. Urlaub, and L. Chasin.
1979. DNA-mediated transfer of the adenine phosphoribosyltransferase lo-
cus into mammalian cells. Proc. Natl. Acad. Sci. USA 76:1373–1376.
76. Wilson, C. A., M. N. Payton, G. S. Elliott, F. W. Buaas, E. E. Cajulis, D.
Grosshans, L. Ramos, D. M. Reese, D. J. Slamon, and F. J. Calzone. 1997.
Differential subcellular localization, expression and biological toxicity of
BRCA1 and the splice variant BRCA1-delta11b. Oncogene 14:1–16.
77. Wilson, C. A., L. Ramos, M. R. Villasenor, K. H. Anders, M. F. Press, K.
Clarke, B. Karlan, J. J. Chen, R. Scully, D. Livingston, R. H. Zuch, M. H.
Kanter, S. Cohen, F. J. Calzone, and D. J. Slamon. 1999. Localization of
human BRCA1 and its loss in high-grade, non-inherited breast carcinomas.
Nat. Genet. 21:236–240.
78. Zhang, H., K. Somasundaram, Y. Peng, H. Tian, D. Bi, B. L. Weber, and
W. S. El-Deiry. 1998. BRCA1 physically associates with p53 and stimulates
its transcriptional activity. Oncogene 16:1713–1721.
79. Zhang, H. T., X. Zhang, H. Z. Zhao, Y. Kajino, B. L. Weber, J. G. Davis, Q.
Wang, D. M. O’Rourke, H. B. Zhang, K. Kajino, and M. I. Greene. 1997.
Relationship of p215BRCA1 to tyrosine kinase signaling pathways and the
cell cycle in normal and transformed cells. Oncogene 14:2863–2869.
4854 RUFFNER ET AL.MOL. CELL. BIOL.