| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Epidemiology and Prevention |
Department of Biomolecular Sciences, University of Manchester Institute of Science and Technology, Manchester M60 1QD, United Kingdom [F. H.]; Laboratory of Molecular Oncology, Strang Cancer Prevention Center, New York, New York 10021 [C. C., A. N. A. M.]; Sir William Dunn School of Pathology, University of Oxford, Oxford OX1 3RE, United Kingdom [D. B.]; and Department of Cell Biology, Weill Medical College of Cornell University, New York, New York 10021 [A. N. A. M.]
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The BRCA1 COOH terminus (amino acids 1560–1863) consists of two BRCTspreceded by a region likely to be nonglobular. We combined site-directed and random mutagenesis, followed by a functional transcription assay in yeast: (a) error-prone PCR-induced random mutagenesis generated eight unique missense mutations causing loss of function, six of which targeted hydrophobic residues conserved in canine, mouse, rat, and human BRCA1; (b) random insertion of a variable pentapeptide cassette generated 21 insertion mutants. All pentapeptide insertions NH2-terminal to the BRCTs retained wild-type activity, whereas insertions in the BRCTs were, with few exceptions, deleterious; and (c) site-directed mutagenesis was used to characterize five known germ-line mutations and to perform deletion analysis of the COOH terminus. Deletion analysis revealed that the integrity of the most COOH-terminal hydrophobic cluster (I1855, L1854, and Y1853) is necessary for activity. We conclude that the integrity of the BRCT domains is crucial for transcription activation and that hydrophobic residues may be important for BRCT function. Therefore, the yeast-based assay for transcription activation can be used successfully to provide tools for structure-function analysis of BRCA1 and may form the basis of a BRCA1 functional assay.
|
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
45% of families with high incidence of breast cancer and up to 80%
of families with both breast and ovarian cancer (2)
. After
an extensive search, BRCA1 was mapped to the long arm of
chromosome 17 by linkage analysis (3)
and was cloned by
positional cloning techniques (4)
. Human BRCA1
codes for a 1863-amino acid protein with no detectable similarity to
known proteins, with the exception of a zinc-binding RING finger domain
located in the NH2-terminal region
(4)
, and two BRCT4 domains found in a variety
of proteins involved in cell cycle control and DNA repair
(5, 6, 7)
.
Recent evidence points to the involvement of BRCA1 in two basic
cellular processes: DNA repair and transcriptional regulation. BRCA1 is
present in a complex containing Rad51 (8)
and BRCA2
(9)
, and DNA damage may control BRCA1 phosphorylation and
subnuclear location (10
, 11)
, strongly suggesting its
involvement in the maintenance of genome integrity. Additional evidence
for the role of BRCA1 in maintenance of genome integrity is provided by
targeted disruption of Brca1 in the mouse. Mouse embryos
lacking Brca1 are hypersensitive to 
-irradiation, and
cells display numerical and structural chromosomal aberrations
(12)
.
We and others have shown that the BRCA1 COOH terminus has the ability to activate transcription in mammalian and yeast cells and that the introduction of germ-line disease-associated mutations, but not benign polymorphisms, abolishes this activity (13, 14, 15) . BRCA1 can be copurified with the RNA polymerase II holoenzyme, supporting the idea that BRCA1 is involved in transcription regulation (16 , 17) . In addition, BRCA1 causes cell cycle arrest via transactivation of p21WAF1/CiP1 (18) and regulates p53-dependent gene expression, acting as a coactivator for p53 (19 , 20) . In all of these studies, the COOH-terminal region was necessary for activity. It is still not clear whether BRCA1 is a multifunctional protein with repair and transcription regulation functions or whether the role of BRCA1 in repair is mediated through transcription activation. In either case, these functions are not necessarily mutually exclusive.
The dearth of knowledge concerning the precise biochemical function of BRCA1 is a major hurdle in developing a functional test to provide reliable presymptomatic assessment of risk for breast and ovarian cancer. The available data derived from linkage analysis indicate that all mutations that cause premature termination (even relatively subtle mutations such as the deletion of 11 amino acids from the COOH terminus) will confer high risk (21) . However, a considerable number of mutations result in amino acid substitutions that, in the absence of extensive population-based studies or a functional assay, do not allow assessment of risk. Two related yeast-based assays designed to characterize mutations in the BRCA1 COOH terminal region have generated results that provide an excellent correlation with genetic linkage analysis (13 , 14 , 22) . This led us to propose the general use of a yeast-based assay to provide functional information and a more reliable risk assessment (23) .
In this report, we use site-directed and random mutagenesis to generate mutations in the BRCA1 COOH terminal region that disrupt transcription activation with the intention of both defining critical residues for BRCA1 function and deriving general rules to predict the impact of a particular mutation.
|
MATERIALS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, ura3, trp1, his3, 6
lexA operator-LEU2] (25)
. HF7c has an
HIS3 reporter gene under the control of the GAL1
upstream activating sequence, responsive to GAL4 transcription
activation. The vectors used for expression confer growth in the
absence of tryptophan (see below). The SFY526 strain has a
lacZ reporter under the control of GAL1 upstream
activating sequence and was transformed with the GAL4 DBD fusion. EGY48
cells were cotransformed with the LexA fusion vectors and plasmid
reporters of lacZ under the control of LexA operators (see
below). If the fusion proteins activate transcription, EGY48 and SFY526
yeast transformants will produce ß-galactosidase, and HF7c
transformants will grow in medium lacking histidine.
Yeast Expression Constructs.
The GAL4 DBD fusion of the wild-type human BRCA1 COOH terminal region
(amino acids 1560–1863) was described previously (13)
.
Alternatively, this fragment was subcloned into the yeast expression
vector pLex9 (25)
in-frame with the DBD of LexA. Both
plasmids have TRP1 as a selectable marker, allowing growth
in the absence of tryptophan. We noticed that our previously described
BRCA1 (amino acids 1560–1863) construct (13)
was made
with a 3' primer lacking a termination codon. This introduces 16
exogenous amino acids to the COOH-terminal region of BRCA1. We have
corrected this by using primer 24ENDT
(5'-GCGGATCCTCAGTAGTGGCTGTGGGGGAT-3'). We compared both constructs and
ascertained that qualitatively and quantitatively, they have the same
activity (not shown).
BRCA1 deletion mutants were generated by PCR on a BRCA1 (amino acids 1560–1863) context using pcBRCA1–385 (a gift from Michael Erdos, National Human Genome Research Institute) as a template and the following primers: H1860X (S9503101, 5'-CGGAATTCGAGGGAACCCCTTAC-CTG-3'; S970074, 5'-GCGGATCCTCAGGGGATCTGGGG-3'); P1856X (S9503101, S970073, 5'-GCGGATCCTCATATCAGGTAGGTGTCC-3'); I1855X (S9503101, 1855STOP, 5'-GCGGATCCTCACAGGTAGGTGTCC-3'); and L1854X (S9503101, 1854STOP, 5'-GCGGATCCTCAGTAGGTGTCCAGC-3'). Mutant Y1853X corresponds to a germ-line mutation and has been described previously (13) . The constructs were sequenced to verify the mutations. The PCR products were digested with EcoRI and BamHI and subcloned into similarly digested pGBT9 vectors. Alternatively, the PCR fragments were subcloned into a vector, pAS2–1 (Clontech), with higher expression levels. Introduction of additional mutations was made using the Quick-Change method. Briefly, a pair of primers encoding each mutation flanked by homologous sequence on each side was added to the wild-type plasmid pLex9 BRCA1 (amino acids 1560–1863) prepared in a methylation-competent strain. The plasmid was amplified using Pfu polymerase (one cycle at: 96°C for 30 s; 12 cycles at: 96°C for 30 s; 50°C for 1 min; and 68°C for 12 min), and DpnI was added at the end of the reaction to digest the parental plasmid. The mixture was then transformed into bacteria. The following oligonucleotide primers were used: T1561I (T1561IF, 5'-CTGGAATTCGAGGGAATCCCTTACCTCGAGTCTGG-3'; T1561IR, 5'-CCAGACTCGAGGTAAGGGATTCCCTCGAATTCCAG-3'); L1564P (L1564PF, 5'-GGGTACCCCTTACCCGGAATCTGGAATCAG-3'; L1564PR, 5'-CTGA-TTCCAGATTCCGGGTAAGGGGTACCCTC-3'); D1733G (D1733GF, 5'-GAAAAATGCTCAATGAGCATGGTTTTGAAGTCCGCGGAG-3'; D1733-GR, 5'-CTCCGCGGACTTCAAAACCATGCTCATTCAGCATTTTTC-3'); G1738E (G1738EF, 5'-GAGCATGATTTTGAAGTCAGAGAAGATGTG-GTTAACGGAAG-3'; G1738ER, 5'-CTTCCGTTAACCACATCTTCTCTGACTTCAAAATCATGCTC-3'); P1806A (P1806AF, 5'-GGTACCGGTGTCCACGCAATTGTGGTTGTGCAGC-3'; and P1806AR, 5'-GCTGCA-CAACCACAATTGCGTGGACACCGGTACC-3').
Yeast Plasmid Reporters.
Plasmid pSH18-34 (25)
, a kind gift of Erica Golemis (Fox
Chase Cancer Center, Philadelphia, PA), was used as a reporter
in the LexA fusion assays. This vector has lacZ under the
control of eight LexA operators, conferring low stringency of gene
expression (26)
.
Yeast Transformation.
Competent yeast cells were obtained using the yeast transformation
system (Clontech) based on lithium acetate, and cells were transformed
according to the manufacturer’s instructions.
Filter ß-Galactosidase Assay.
SFY526 and EGY48 transformants (several clones for each construct) were
streaked on a filter overlaid on solid medium lacking tryptophan
(SFY526) or tryptophan and uracil (EGY48) and allowed to grow
overnight. Cells growing on the filter were lysed by freeze/thawing in
liquid nitrogen, and each filter was incubated in 2.5 ml of Z buffer
(16.1 g/liter
Na2HPO4·7H2O,
5.5 g/l
NaH2PO4·H2O,
0.75 g/l KCl, and 0.246 g/l
MgSO4·7H2O, pH 7.2)
containing 40 µl of X-gal solution (20 mg/ml of X-gal in
N,N-dimethylformamide) at 30°C for up to 16 h.
Liquid ß-Galactosidase Assay.
Liquid assays were performed as described previously (27)
.
At least three separate transformants were assayed, and each was
performed at least in duplicate.
Growth Curves.
HF7c transformants (several clones) containing different pGBT9 or pAS2
constructs were grown overnight in synthetic medium plus 2% dextrose
(SD medium) lacking tryptophan. The saturated cultures were used to
inoculate fresh medium lacking tryptophan or tryptophan and histidine
to an initial A600 of 0.0002.
Cultures were grown at 30°C in the shaker, and the absorbance was
measured at different time intervals starting at 12 h, then every
4 h up to 36 h after inoculation.
Plasmid Recovery from Yeast Cells.
EGY48 transformants were grown to saturation in liquid medium lacking
uracil (but in the presence of tryptophan). Cells were harvested and
treated with yeast lysis solution [2% Triton X-100, 1% SDS, 100
mM NaCl, 10 mM Tris (pH 8.0), and 1
mM EDTA], phenol:chloroform:isoamyl alcohol (25:24:1), and
0.3 g of acid-washed beads. The sample was vortexed for 2 min and
centrifuged, and the supernatant precipitated with one-tenth volume of
3 M NaOAc (pH 5.2) and 2.5 volumes of ethanol.
Alternatively, plasmid rescue was performed as suggested by Strathern
and Higgins (28)
.
Screening in X-gal Plates.
To allow direct screening of the clones with loss of activity, EGY48
cells transformed with the mutagenized cDNAs were plated on
X-gal-containing plates: 2% galactose, 1% raffinose, 80 mg/l X-gal,
and 1x BU salts (1 liter of 10x BU salts: 70 g
Na2HPO4·0.7H2O,
30 g NaH2PO4).
Error-prone PCR Mutagenesis.
A 60-cycle PCR reaction (94°C denaturation; 55°C annealing; 72°C
extension) was performed using Taq polymerase and p385-BRCA1 plasmid as
a template and oligonucleotide primers (S9503101,
5'-CGGAATTCGAGGGAACCCCTTACCTG-3'; S9503098,
5'-GCGGATCCGTAGTGGCTGTGGGGGAT-3'). The PCR product was gel purified and
cotransformed in an equimolar ratio with a NcoI-linearized
wild-type pLex9 BRCA1 (amino acids 1560–1863) plasmid and pSH18-34.
After transformation, cells were plated on X-gal plates and incubated
for 5 days. Eighty-one white and four control blue clones were
recovered and restreaked on master plates. White clones were screened
again on a filter assay, and the 62 clones that were consistently white
were analyzed further. Plasmid DNA was recovered from the yeast cells
and transformed into Escherichia coli. Miniprep DNAs from
each of two bacterial transformants from the 62 candidates were
retransformed into yeast cells and tested again for ß-galactosidase
production. The BRCA1 inserts in plasmid DNAs generating white clones
were subjected to direct sequencing using dye terminators.
Pentapeptide Scanning Mutagenesis.
Pentapeptide scanning mutagenesis is a technique whereby 5-amino acid
insertions are introduced at random in a target protein
(29)
. Briefly, an E. coli donor strain
containing the target plasmid and pHT385, a conjugative delivery vector
for transposon Tn4430, is mated with a plasmid-free E.
coli recipient strain. By plating the mating mix simultaneously on
antibiotics selecting for the recipient, the target plasmid, and
Tn4430, transconjugants containing pHT385::target
plasmid cointegrates are isolated. This cointegrate resolves rapidly
in vivo, regenerating pHT385 and the target plasmid into
which a copy of Tn4430 has been inserted. Tn4430
contains KpnI restriction enzyme sites located 5-bp from
both ends of the transposon and duplicates 5-bp of target site sequence
during transposition. By digesting the target
plasmid::Tn4430 hybrid with KpnI and
religating the digested DNA, the bulk of the transposon is deleted to
generate a target plasmid derivative containing a 15-bp insertion. If
the insertion is in a protein-encoding sequence, this will result in a
5-amino acid insertion in the target protein.
Tn4430 insertions in the COOH-terminal region of BRCA1 were identified either by genetic or physical means. In the former case, 30 separate matings were performed as detailed previously (30) using appropriate antibiotic selections and in which the target plasmid was pLex9 containing the BRCA1 COOH terminal region fused to LexA DBD. Transconjugant colonies were harvested by washing from the mating plates, and plasmid DNA was isolated from the pooled colonies. The plasmid preparations were pooled further and transformed into Saccharomyces cerevisiae EGY48 harboring the pSH18-34 reporter plasmid. Transformants were tested for transcription activation by replica-plating to plates containing X-gal. Plasmid DNA was recovered from white colonies and transformed into E. coli XL1-Blue selecting on X-gal-containing plates. Plasmid DNA was isolated from white colonies (which contain only pLex9::BRCA1 COOH-terminal::Tn4430), and the insertion of Tn4430 into the BRCA1 COOH terminal region was confirmed by restriction enzyme mapping. For the identification of Tn4430 insertions by physical means, pooled plasmid DNA from E. coli consisting of the target plasmid into which Tn4430 was inserted was digested with EcoRI and BamHI, enzymes which liberate the BRCA1 insert but do not cut Tn4430. This digestion of pooled plasmid DNA generates four fragments: the pLex9 vector backbone, the pLex9 vector containing Tn4330 insertions, the BRCA1 COOH-terminal fragment, and the BRCA1 COOH terminal fragment containing Tn4430 insertions. The latter fragment was recovered from an agarose gel and recloned in EcoRI-BamHI-digested pLex9 to produce a library of pLex9::BRCA1 COOH terminal domain plasmids containing Tn4430 insertions in the BRCA1 COOH terminal region. In the case of Tn4430 insertions identified by either genetic or physical means, following further restriction mapping the bulk of Tn4430 was deleted from selected clones by digestion with KpnI and religation. The positions of the 15-bp insertions were determined by sequence analysis. Twenty-one plasmids harboring the BRCA1 COOH terminal region with 15-bp insertions were analyzed for transcription activation in S. cerevisiae EGY48 containing pSH18–34.
Western Blots.
Yeast cells were grown in selective media to saturation, and
A600 was measured. Cells were
harvested and lysed in cracking buffer [8 M
urea, 5% SDS, 40 mM Tris-HCl (pH 6.8), 0.1
mM EDTA, and 0.4 mg/ml bromphenol blue; used 100
µl/7.5 total A600] containing
protease inhibitors. The samples were boiled and separated on a 10%
SDS-PAGE. The gel was electroblotted on a wet apparatus to a
polyvinylidene difluoride membrane. The blots were blocked overnight
with 5% skim milk using TBS-Tween and incubated with the -pLexA
(for LexA constructs) or -GAL4 DBD (for GAL4 constructs) monoclonal
antibodies (Clontech) using 0.5% BSA in TBS-Tween. After four washes,
the blot was incubated with the -mouse IgG horseradish peroxidase
conjugate in 1% skim milk in TBS-Tween. The blots were developed using
an enhanced chemiluminescent reagent (DuPont NEN, Boston, MA).
|
RESULTS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Mutagenesis Strategies.
To shed light on the critical residues and regions necessary for
function, we used four complementary strategies: (a)
error-prone PCR mutagenesis followed by a screen for loss of function;
(b) pentapeptide insertion mutagenesis; (c)
site-directed mutagenesis to analyze documented mutations; and
(d) deletion analysis of the COOH terminus.
Error-prone PCR Mutagenesis Reveals Critical Residues for
Activation.
Approximately 105 yeast clones were screened for
loss of transcription activation function. Sixty-two clones were
isolated that had lost activity, most of which contained small
insertions or deletions causing frameshift mutations and premature
termination of the BRCA1 protein, as subsequently confirmed by SDS-PAGE
and Western blot analysis (not shown). Two independent clones displayed
the same nonsense mutation (Y1769X). Four clones had two mutations
(E1660G/M1689K, K1727R/L1786P, S1722P/N1774Y, and S1715N/Q1811L),
limiting their further characterization. The 10 remaining clones each
had a single missense mutation (one clone also had a silent mutation)
and corresponded to eight distinct mutations (Table 1)
. Interestingly, the screen revealed that hydrophobic residues were the
major targets of mutation (six of eight). Furthermore, all of the
targeted residues are perfectly conserved in canine, mouse, and rat
Brca1 (Table 1)
. Even conservative mutations may not be well accepted
in residues that are perfectly conserved in all species. This is
illustrated by mutation F1761I, where a smaller hydrophobic residue is
not tolerated in place of a bulkier one. Loss-of-function mutations
were located primarily in the BRCT domains. In particular, mutations
that occur in BRCT-C [the most COOH-terminal BRCT (amino acids
1756–1855); BRCT-N (amino acids 1649–1736) is located
NH2-terminally to BRCT-C] are in residues that
constitute the hydrophobic clusters conserved in the BRCT superfamily.
Western blot analysis of the mutant clones (three independent clones of
each) revealed that all of the mutants were expressed at levels
comparable with the wild type, ruling out the possibility that loss of
function was attributable to instability of the protein (Fig. 1)
. It is important to stress, however, that protein levels are
relatively variable in different yeast clones carrying the same
constructs and should only be taken as a rough estimate.
|
|
groups the insertions by location: the first group includes mutations
in the region NH2-terminal to the BRCT domains
(amino acids 1560–1649); the second group contains mutations in
BRCT-N; and the third group includes mutations in the intervening
region between BRCT-N and BRCT-C. The last group includes mutations in
BRCT-C. None of the insertions NH2-terminal to
the BRCT domains had a negative effect on transcription activation.
Also, insertions in the interval between the BRCT domains or at its
boundary (1723RGTPI) had generally less drastic effects. In contrast,
all insertions within BRCT-N and several within BRCT-C had a more
severe effect. It is clear that BRCT-C tolerates insertions better
(only three of five showed loss of activity) than BRCT-N (all mutations
reduced activity with six of seven showing drastic impairment). The
difficulty in predicting the outcome of mutations can be well
exemplified by mutations 1824GGTPI and 1822GVPLH. Both of these
mutations target residues at the end of BRCT-C -helix 2, do not
change the net charge of the protein, and are only two residues apart.
However, 1822GVPLH has 6% of the wild-type activity, whereas
1824GGTPI has an activity 80% higher than wild type. Interestingly,
the 1793GVPLK insertion increased transcriptional activation 4-fold,
suggesting that this region of BRCA1 may directly contact a component
of the transcription machinery. The pentapeptide mutagenesis results
demonstrated that, in addition to substitution mutations, insertion
mutagenesis in the COOH-terminal region, particularly in the BRCT
domains, can profoundly alter transcriptional activity by BRCA1.
|
. Mutations T1561I and L1564P are both located in the region preceding
the BRCT domains and displayed wild-type activity. L1564P was expected
to be a polymorphism because proline is the residue found in the rat
Brca1 sequence. The three remaining variants are localized to the BRCT
domains. Two variants, D1733G and P1806A, displayed wild-type activity
and are suggested to be benign polymorphisms. D1733G introduces a
glycine that probably does not affect BRCT structure. P1806A involves a
conservative change, and it is important to note that the rat Brca1
sequence has leucine in that position. Only one of the variants tested,
G1738E, displayed a loss of function phenotype. Thus, we propose that
G1738E is a disease-predisposing variant.
|
shows the several deletion mutants analyzed aligned to mouse, rat, dog,
and human BRCA1 wild-type sequences. Mutant H1860X introduces a stop
codon but maintains all of the conserved amino acids in canine and
human BRCA1. P1856X maintains the hydrophobic residues, which are
conserved in all of the BRCT domains described in several species.
I1855X and L1854X delete one and two conserved hydrophobic residues,
respectively. Y1853X is a mutation found in the germ-line of breast and
ovarian cancer patients in high-risk families (21)
. These
constructs were transformed into SFY526 and HF7c and analyzed for their
ability to activate different reporters (Fig. 2b)
. Activity
comparable with the wild-type was obtained with mutants H1860X and
P1856X. However, mutations that disrupted the conserved hydrophobic
residues (I1855X and L1854X) at the end of the BRCT domain abolished
activity. Therefore, we define the MTD in BRCA1 as amino acids
1760–1855. To determine whether the loss of activity by the mutants
correlated with the stability of the protein, yeast cells were
transformed with the same mutated alleles in a vector conferring high
expression (pAS2-1). Transcriptional activity using these constructs
(in pAS2-1 backbone) was measured, and results were similar with I1855X
showing some residual activity. Expression was highly variable, and
mutants were in general expressed at lower levels than wild type (Fig. 2c)
. There was no correlation between loss of activity and
lower levels of expression because the transcriptionally active mutant
H1860X was expressed at levels lower or comparable with
transcriptionally inactive mutants I1855X and Y1853X (Fig. 2c)
.
|
|
DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. We discuss the general conclusion of each strategy and then we analyze
the possible structural outcome of the mutations based on the crystal
structure of XRCC1 BRCT (36)
.
|
. Six of eight unique PCR-generated mutations were in hydrophobic residues conserved in human, canine, mouse, and rat Brca1 (6 , 7) , supporting the notion that hydrophobic residues are important for the stability of the BRCT domains and BRCA1 function in vivo.
Pentapeptide Scanning Mutagenesis.
Pentapeptide scanning mutagenesis is a method by which a variable
5-amino acid cassette is introduced at random into a target protein
(29
, 30 , 37)
. This approach differs from error-prone
mutagenesis because clones are not selected for loss of activity but
rather mutations are analyzed only after they have been generated.
Therefore, mutants with gain of function, loss of function, and novel
activities can be produced (30
, 37)
. Moreover, it has been
shown that insertion is essentially random (29)
. The
results obtained are in agreement with the PCR-mediated mutagenesis in
that the region NH2-terminal to the BRCT domains
(amino acids 1560–1649) seems to be more tolerant of mutation; none of
six different pentapeptide insertions in this region affected
transcription activation. The fact that derivatives containing
insertion mutations in this region retained wild-type activity suggests
that this region is nonglobular and is probably a flexible part of the
COOH-terminal region without many critical secondary structure
elements. In fact, the region encompassing amino acids 1524–1661 is
predicted to be nonglobular (5)
. The pentapeptide
mutagenesis results also suggest that changing the net charge of the
protein does not necessarily correlate with an alteration in
transcription activity, as would be expected for classical acidic
activators (38)
, because 1793GVPLK (which adds a positive
charge) shows a 4-fold increase in activity. Interestingly, only 4 of
the 63 COOH-terminal germ-line variants involve nonconservative
substitutions in acidic residues, thought to be important for
activation, suggesting that, contrary to initial predictions, BRCA1 may
not be a classical acidic activator (4)
. The 1793GVPLK
mutation, which is hyperactive for transcription activation, may define
a point of contact between the BRCA1 COOH-terminal region and the
transcription machinery.
Deletion Analysis.
Our analysis demonstrates that residues COOH-terminal to amino acids
1855 are dispensable for activation, consistent with the extreme
evolutionary divergence of those residues (Fig. 2
; Refs.
39
and 40
). The results also underscore the
importance of the last hydrophobic cluster in the sequence (YLI for
human and canine; YLV for mouse and rat) and provide a plausible
explanation for the complete loss of function (in vitro and
in vivo) of Y1853X alleles.
Site-directed Mutagenesis.
Only one of five germ-line mutations analyzed displayed loss of
function, suggesting that a large part of variants in the COOH-terminal
region will probably be benign polymorphisms, including some variants
found in the BRCT domains. Very little data are available at this
moment to confirm or contradict the results obtained. In particular,
T1561I illustrates the difficulties involved in predicting outcome from
population data. T1561I was found in one affected individual but not in
control individuals (41)
. This could suggest that T1561I
is a disease-predisposing variant. However, although found as a
germ-line mutation, it was absent from the tumor from the same patient
(41)
, indicating that this mutation is a benign
polymorphism.
Structural Basis for Effects of BRCT Domain Mutations.
The COOH-terminal BRCT domain of XRCC1 consists of a four-stranded
parallel ß-sheet (ß1–ß4) surrounded by three -helices
(1–3; Ref. 36
). The ß-sheet forms the core of the
structure with a pair of -helices (1 and 3) on one side of the
ß-sheet and the remaining -helix (2) on the other side. A model
of the more COOH-terminal BRCT domain of BRCA1 has been constructed
based on the crystal structure of the BRCT domain of XRCC1
(36)
. This model allows an interpretation of the effect of
some of the mutations described in this study (Tables 1
2
3)
on BRCT
domain structure (Fig. 3)
.
The position of the M1652K mutation corresponds to a position (Asp4) in the XRCC1 structure that is thought to form a salt bridge at the BRCT dimer interface (36) . Although M1652 would not be expected to be involved in salt bridge formation at neutral pH, residues in this region nevertheless may also be involved in homo- or heterodimer formation in BRCA1.
Missense mutations at positions 1702, 1703, and 1705 of the BRCT-N
domain and a pentapeptide insertion at position 1822 of the BRCT-C
domain abolish transcription activation by the BRCA1 COOH terminus
(Tables 1
2)
. These mutations are predicted to occur in a region of
highly variable length and composition that encompasses helix 2 in
BRCT domains (36)
. It was suggested that this variability
indicated that this region was not involved in formation of the core
fold of the BRCT domain (36)
. Nevertheless, the mutations
isolated here reveal that this region of the BRCT domain is critical
for the transcription activation function of the BRCA1 COOH terminus.
Residue F6 forms part of a highly conserved hydrophobic pocket centered
on residue W74 in helix 3 in the COOH-terminal BRCT domain of XRCC1
(36)
. Mutations at the corresponding position (F1761) in
the BRCT-C domain of BRCA1 abolish transcription activation (Table 1)
.
By analogy with XRCC1, residue F1761 of BRCA1 is also predicted to form
part of a hydrophobic pocket, the disruption of which by mutation may
compromise correct BRCT domain folding. In contrast, residue L25 is
implicated in the interactions between helices 1 and 3, which
form a paired helical bundle in the three-dimensional structure of the
BRCT domain of XRCC1 (36)
. A missense mutation of the
corresponding residue (L1780) or a pentapeptide insertion at this
position in the BRCT-C domain of BRCA1 abolishes transcription
activation by the BRCA1 COOH terminus region (Tables 1
2)
. These
mutations are likely to affect the interactions between helices 1
and 3, thereby destabilizing the BRCT domain structure. Two other
missense mutations in the BRCT-C domain, P1806A and V1833E, were shown,
respectively, to display wild-type activity and to abolish
transcription activation (Tables 1
2
3)
. Interestingly, P1806A is
predicted to have no obvious effect on the structure, whereas a less
drastic mutation at position V1833 (to methionine) has been predicted
to destabilize the fold of the domain (36)
, suggesting
that V1833E will behave similarly.
Pentapeptide insertions in many of the predicted secondary
structure elements in the COOH-terminal region of BRCA1 abolish
transcription activation (Table 1
and Fig. 3
). Some of these insertions
are likely to disrupt formation of the correct BRCT domain core fold,
e.g., insertions in strand ß2 (1676RGTPL) and in helices
2 (1822GVPLH) and 3 (1717WGTPF). In contrast, the 1780GVPQL
insertion in helix 1 is predicted to be at the BRCT dimer interface
and thereby may affect the association of this domain with another
protein, e.g., RNA helicase A, which interacts with BRCA1
through residues in helix 1 (17)
.
Different Roles of BRCT-N and BRCT-C.
Our insertion mutagenesis results suggest that BRCT-C can tolerate
insertions better than BRCT-N without affecting transcription
activation function. In addition, BRCT-N is more highly conserved in
other species than is BRCT-C (39
, 40)
, suggesting a higher
constraint for function. The BRCT-N seems to be very important for
binding to RNA helicase A (17)
, although it seems to lack
an independent activation domain (mutant Y1769X is inactive). The
borders of BRCT-C coincide well with the limits of the MTD, but only in
combination with BRCT-N are high levels of activation achieved
(13)
. It is tempting to speculate that BRCT-N is involved
in the interaction of BRCA1 with RNA helicase A and is responsible for
presenting BRCT-C in a correct way to obtain a transcriptionally
competent activator.
Functional Assay.
We have performed an extensive analysis of the BRCA1 COOH terminal
region (amino acids 1560–1863) and have found that there is a
correlation between loss of transcription activation function and the
human genetic data, suggesting that the assay could be used to predict
the effect of missense mutations in this region. Although the effects
of mutations on transcriptional activity have been found to be
comparable in yeast and mammalian cells (13
, 15)
, it is
possible that the effect of some mutations may be evident only in
mammalian cells, e.g., because of an interaction with
mammalian-specific regulators, raising the possibility of a
misinterpretation of the data obtained in yeast.
In the results presented here for substitution mutations, we have used a reporter gene with relatively low stringency (eight Lex operators; Ref. 26 ). The rationale for this choice was to recover only mutants that cause dramatic reduction or complete loss of activity. Mutations that partially disrupt the function would still activate the reporter. In the absence of knowledge of the minimum in vivo threshold of transcription activity needed for tumor suppression, it would be inappropriate to make decisions on whether a particular mutation would represent a wild-type or a cancer predisposing allele. For example, a particular mutation that shows 50% loss of activity in yeast could still be perfectly functional in breast and ovarian cells.
In conclusion, the data presented here suggest that the yeast assay for monitoring transcription activation by BRCA1 will provide a wealth of functional information in a research setting. That includes identifying protein-protein interaction regions, defining critical residues for activity, and providing tools to identify possible regulators. A general use of the assay to help in risk assessment and providing information for clinical decisions must await further confirmation from population-based studies.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
1 This work was supported by the Iris and B.
Gerald Cantor Foundation, the Irving Weinstein Foundation, and the Lee
Kaplan Cancer Foundation. Work by F. H. was initiated while the
recipient of Wellcome Trust Research Career Development Fellowship
040822/Z/94/Z in the Department of Biochemistry, University of Oxford,
Oxford. 
2 These authors contributed equally to this
work.
3 To whom requests for reprints should be
addressed, at Laboratory of Molecular Oncology, Strang Cancer
Prevention Center, The Rockefeller University, Box 231, 1230 York
Avenue, New York, NY 10021. Phone: (212) 734-0567, extension 225; Fax:
(212) 472-9471; E-mail: monteia{at}rockvax.rockefeller.edu
4 The abbreviations used are: BRCT, BRCA1 COOH
terminal domain; DBD, DNA binding domain; X-gal,
5-bromo-4-chloro-3-indolyl-ß-D-galactopyranoside; MTD,
minimal transactivation domain.
5 Breast Cancer Information Core,
http://www.nhgri.nih.gov/Intramural_research/Lab_transfer/Bic/.
Received 11/12/99. Accepted 3/ 7/00.
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
-irradiation hypersensitivity and genetic instability. Oncogene, 17: 3115-3124, 1998.[Medline]
-loop of TEM-1 ß-lactamase. J. Biol. Chem., 272: 28833-28836, 1997.This article has been cited by other articles:
|
|
 |
|
  A. Kumar, W. S. Joo, G. Meinke, S. Moine, E. N. Naumova, and P. A. Bullock Evidence for a Structural Relationship between BRCT Domains and the Helicase Domains of the Replication Initiators Encoded by the Polyomaviridae and Papillomaviridae Families of DNA Tumor Viruses J. Virol., September 1, 2008; 82(17): 8849 - 8862. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Subbaramaiah, C. Hudis, S.-H. Chang, T. Hla, and A. J. Dannenberg EP2 and EP4 Receptors Regulate Aromatase Expression in Human Adipocytes and Breast Cancer Cells: EVIDENCE OF A BRCA1 AND p300 EXCHANGE J. Biol. Chem., February 8, 2008; 283(6): 3433 - 3444. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Pajunen, H. Turakainen, E. Poussu, J. Peranen, M. Vihinen, and H. Savilahti High-precision mapping of protein protein interfaces: an integrated genetic strategy combining en masse mutagenesis and DNA-level parallel analysis on a yeast two-hybrid platform Nucleic Acids Res., August 15, 2007; (2007) gkm563v1. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. A. Carvalho, S. M. Marsillac, R. Karchin, S. Manoukian, S. Grist, R. F. Swaby, T. P. Urmenyi, E. Rondinelli, R. Silva, L. Gayol, et al. Determination of Cancer Risk Associated with Germ Line BRCA1 Missense Variants by Functional Analysis Cancer Res., February 15, 2007; 67(4): 1494 - 1501. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Chenevix-Trench, S. Healey, S. Lakhani, P. Waring, M. Cummings, R. Brinkworth, A. M. Deffenbaugh, L. A. Burbidge, D. Pruss, T. Judkins, et al. Genetic and Histopathologic Evaluation of BRCA1 and BRCA2 DNA Sequence Variants of Unknown Clinical Significance Cancer Res., February 15, 2006; 66(4): 2019 - 2027. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C M Phelan, V Dapic, B Tice, R Favis, E Kwan, F Barany, S Manoukian, P Radice, R B van der Luijt, B P M van Nesselrooij, et al. Classification of BRCA1 missense variants of unknown clinical significance J. Med. Genet., February 1, 2005; 42(2): 138 - 146. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. A. Figge and L. Blankenship Missense Mutations in the BRCT Domain of BRCA-1 from High-Risk Women Frequently Perturb Strongly Hydrophobic Amino Acids Conserved among Mammals Cancer Epidemiol. Biomarkers Prev., June 1, 2004; 13(6): 1037 - 1041. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. Mirkovic, M. A. Marti-Renom, B. L. Weber, A. Sali, and A. N. A. Monteiro Structure-Based Assessment of Missense Mutations in Human BRCA1: Implications for Breast and Ovarian Cancer Predisposition Cancer Res., June 1, 2004; 64(11): 3790 - 3797. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. S. Williams, D. I. Chasman, D. D. Hau, B. Hui, A. Y. Lau, and J. N. M. Glover Detection of Protein Folding Defects Caused by BRCA1-BRCT Truncation and Missense Mutations J. Biol. Chem., December 26, 2003; 278(52): 53007 - 53016. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Lingaas, K. E. Comstock, E. F. Kirkness, A. Sorensen, T. Aarskaug, C. Hitte, M. L. Nickerson, L. Moe, L. S. Schmidt, R. Thomas, et al. A mutation in the canine BHD gene is associated with hereditary multifocal renal cystadenocarcinoma and nodular dermatofibrosis in the German Shepherd dog Hum. Mol. Genet., December 1, 2003; 12(23): 3043 - 3053. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. A. Fleming, J. D. Potter, C. J. Ramirez, G. K. Ostrander, and E. A. Ostrander Understanding missense mutations in the BRCA1 gene: An evolutionary approach PNAS, February 4, 2003; 100(3): 1151 - 1156. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V. Robert, F. Hayes, A. Lazdunski, and G. P. F. Michel Identification of XcpZ Domains Required for Assembly of the Secreton of Pseudomonas aeruginosa J. Bacteriol., March 15, 2002; 184(6): 1779 - 1782. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Vallon-Christersson, C. Cayanan, K. Haraldsson, N. Loman, J. T. Bergthorsson, K. Brondum-Nielsen, A.-M. Gerdes, P. Moller, U. Kristoffersson, H. Olsson, et al. Functional analysis of BRCA1 C-terminal missense mutations identified in breast and ovarian cancer families Hum. Mol. Genet., February 1, 2001; 10(4): 353 - 360. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Houvras, M. Benezra, H. Zhang, J. J. Manfredi, B. L. Weber, and J. D. Licht BRCA1 Physically and Functionally Interacts with ATF1 J. Biol. Chem., November 10, 2000; 275(46): 36230 - 36237. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Miyake, Y.-F. Hu, D. S. Yu, and R. Li A Functional Comparison of BRCA1 C-terminal Domains in Transcription Activation and Chromatin Remodeling J. Biol. Chem., December 15, 2000; 275(51): 40169 - 40173. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
| Molecular Cancer Research | Cancer Prevention Research |
| Cancer Prevention Journals Portal | Cancer Reviews Online |
| Annual Meeting Education Book | Meeting Abstracts Online |
