Evidence is accumulating to suggest that some of the diverse functions associated with BRCA1 may relate to its ability to transcriptionally regulate key downstream target genes. Here, we identify S100A7 (psoriasin), S100A8, and S100A9, members of the S100A family of calcium-binding proteins, as novel BRCA1-repressed targets. We show that functional BRCA1 is required for repression of these family members and that a BRCA1 disease–associated mutation abrogates BRCA1-mediated repression of psoriasin. Furthermore, we show that BRCA1 and c-Myc form a complex on the psoriasin promoter and that BRCA1-mediated repression of psoriasin is dependent on functional c-Myc. Finally, we show that psoriasin expression is induced by the topoisomerase IIα poison, etoposide, in the absence of functional BRCA1 and increased psoriasin expression enhances cellular sensitivity to this chemotherapeutic agent. Therefore, we identified a novel transcriptional mechanism that is likely to contribute to BRCA1-mediated resistance to etoposide.
- DNA damage
The BRCA1 tumor suppressor gene was identified by positional cloning as one of the genes conferring genetic predisposition to breast and ovarian cancer ( 1). BRCA1 mutation carriers have recently been reported to have an 82% risk of developing breast cancer and a 54% risk of developing ovarian cancer by the age of 80 ( 2). Although the exact function of BRCA1 remains to be defined, roles in DNA damage detection, DNA damage repair ( 3), cell cycle checkpoint control ( 4– 6), transcriptional regulation ( 7), and ubiquitination ( 8) have been inferred. A substantial body of evidence has accumulated over the last number of years indicating a role for BRCA1 in transcriptional regulation. BRCA1 copurifies with the RNA polymerase II holoenzyme complex through an association with RNA helicase A ( 9), suggesting that BRCA1 is a component of the core transcriptional machinery. A number of other reports have shown that BRCA1 associates with a range of different transcription factors, including p53 ( 10), c-Myc ( 11), activating transcription factor 1 ( 12), and signal transducers and activators of transcription 1 ( 13), modulating their activity. These findings suggest that BRCA1 functions as either a coactivator or corepressor of transcription, an effect that may involve its ability to recruit both the basal transcription machinery and proteins implicated in chromatin remodeling such as the histone deacetylases HDAC1 and HDAC2 ( 14) or the SWI/SNF-related chromatin-remodeling complex ( 15).
Preclinical and clinical studies have identified BRCA1 as being important in modulating the response to therapeutic DNA damage ( 16). We have previously shown that functional BRCA1 inhibits apoptosis in breast cancer cell lines following exposure to DNA-damaging chemotherapeutic agents. In particular, we observed that functional BRCA1 resulted in a significant (over 1,000-fold) increase in resistance to the topoisomerase IIα poison, etoposide, when compared with BRCA1-deficient cells ( 17). In the current study, we have further investigated this interaction by performing microarray-based expression profiling to identify novel BRCA1 transcriptional targets that may be involved in regulating the cellular response to etoposide treatment. Through a comparison of the expression profiles between the BRCA1 mutant HCC1937 cells reconstituted with functional BRCA1 and vector-only reconstituted controls, we have identified the S100A family of proteins as being etoposide inducible and potently repressed by BRCA1. We have shown that BRCA1-dependent regulation of psoriasin (S100A7) is dependent on functional c-Myc and that psoriasin overexpression sensitizes cells to etoposide.
Materials and Methods
Generation of cell lines. HCC-EV1, HCC-BR116, HP-1, and MDA-MB-231-zeo breast cancer cell lines and the pRc/CMV-BRCA1 construct were generated as previously described ( 17– 19). The BG1-Neo and AS4 ovarian cancer cell lines were gifts from Dr. Lois Annab (Laboratory of Molecular Carcinogenesis, National Institute of Environmental Health Sciences, Research Triangle Park, NC).
Maintenance of cell lines. The HCC1937 and T47D breast cancer cell lines were maintained as previously described ( 17). BG1-Neo, AS4, MDA-MB-231-zeo, and HP-1 cell lines were maintained in DMEM supplemented with 10% FCS, 1 mmol/L sodium pyruvate, 2 mmol/L l-glutamine, and 50 μg/mL penicillin-streptomycin. For etoposide treatment, cells were grown in a 90-mm-diameter tissue culture dishes to 80% confluence and treated with etoposide “etopo-phos” (Bristol Myers Squibb, London, United Kingdom) in regular antibiotic-free medium at a concentration of 1 μmol/L for 24 to 48 hours continuously.
Northern blot assays. Total RNA was extracted from cells and analyzed by Northern blotting as previously described ( 6). Northern cDNA probes were generated from IMAGE consortium clones using the appropriate restriction enzymes (Invitrogen, Paisley, United Kingdom): psoriasin IMAGE:4751685 (EcoR1 and Not1), S100A8 IMAGE:4246359 (Sal1 and Not1), S100A9 IMAGE:1332797 (EcoR1 and Not1). The BRCA1 and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) oligonucleotide probes were generated using PCR as previously described ( 6).
Western blot assays. Western blot assays were done as previously described ( 17). Blots were probed with antibodies to psoriasin (Imgenex, Buckinghamshire, United Kingdom), S100A8, S100A9 (Bachem, St. Helens, United Kingdom), c-Myc 9E10 (Santa Cruz, Santa Cruz, CA), and GAPDH (Sigma, Dorset, United Kingdom). BRCA1 immunoprecipitation Western blots were done using the rabbit polyclonal AB-4 (Calbiochem, Nottingham, United Kingdom) and the mouse monoclonal antibodies AB1 (Oncogene Research Products, Nottingham, United Kingdom).
Small interfering RNA transfection. Small interfering RNA (siRNA) transfection was done using OligofectAMINE (Invitrogen) as per instructions of the manufacturer. Each oligonucleotide was used at a concentration of 40 nmol/L. Transfections were done twice 24 hours apart and the protein was extracted 72 hours later. BRCA1-specific and scrambled oligonucleotide sequences were described previously ( 17). A commercially available c-Myc–specific oligonucleotide was purchased from Qiagen (Crawley, United Kingdom). Two psoriasin-specific oligonucleotides were used simultaneously (Qiagen). The target sequences were AACTTCCTTAGTGCCTGTGAC and AAAGGGCACAAATTACCTCGC.
Electrophoretic mobility shift assay. The oligonucleotide sequence containing a CACATG sequence was synthesized by Qiagen. Nuclear extracts were prepared as described by Lee et al. ( 20). Six milligrams of nuclear extract in a final volume of 25 μL binding buffer [15 mmol/L HEPES (pH 7.9), 6% glycerol, 4% Ficoll, 220 mmol/L NaCl, 0.36 mmol/L MgCl2, 1.0 mmol/L EDTA, 0.04 mg/mL poly(dI-dC) poly(dI-dC), 0.1 mg/mL bovine serum albumin, 14 mmol/L DTT, and 0.12 mmol/L phenylmethylsulfonyl fluoride] were preincubated for 2 hours at 4°C with 0.5 μg poly(dIdC) and 4 μg c-Myc or BRCA1-specific antibody. One microliter of 32P-labeled (30,000 cpm) oligonucleotide probe was added and the reaction mixture was incubated at ∼22°C for 30 minutes before electrophoresis on a 6% polyacrylamide gel in 50 mmol/L Tris-borate EDTA buffer at 140 V for 2 hours.
Re-chromatin immunoprecipitation assays. Chromatin was cross-linked using 1.5% formaldehyde for 15 minutes at room temperature. The cells were then collected after two washings with PBS in 1 mL collection buffer [100 mmol/L Tris-HCL (pH 9.4) and 100 mmol/L DTT]. The cell suspension was then incubated on ice for 15 minutes. Cells were then lysed sequentially by resuspension and 5-minute centrifugation at 3,000 × g at 4°C with 1 mL buffer A (10 mmol/L EDTA, 0.5 mmol/L EGTA, 10 mmol/L HEPES, and 0.25% Triton X-100) and 1 mL buffer B (1 mmol/L EDTA, 0.5 mmol/L EGTA, 10 mmol/L HEPES, and 200 mmol/L NaCl), and sonicated thrice for 10 seconds at maximum settings in 250 μL lysis buffer (10 mmol/L EDTA, 50 mmol/L Tris-HCl, 1% SDS, and 0.5% Empigen BB). After 15-minute centrifugation, 10 μL of the supernatants were taken as inputs and the remainder was diluted 5-fold in immunoprecipitation buffer (2 mmol/L EDTA, 100 mmol/L NaCl, 20 mmol/L Tris-HCl, and 0.5% Triton X-100). This was then subjected to immunoprecipitation overnight with specific antibodies after preclearing with preimmune IgG, 2 μg salmon sperm DNA, and 60 μL protein A/G Sepharose bead slurry. Precipitate complexes were serially washed with 300 μL washing buffer I (2 mmol/L EDTA, 20 mmol/L Tris-HCl, 1% SDS, 01% Triton X-100, and 150 mmol/L NaCl), washing buffer II (2 mmol/L EDTA, 20 mmol/L Tris-HCl, 1% SDS, 01% Triton X-100, and 250 mmol/L NaCl), washing buffer III (1 mmol/L EDTA, 10 mmol/L Tris-HCl, 1% NP40, 1% deoxycholate, and 0.25 mol/L LiCl) then twice with 1 mmol/L EDTA and 10 mmol/L Tris-HCl. Complexes were removed from the beads through subsequent 15-minute incubations, vortexing, and 5-minute centrifugations with 50 μL of 1% SDS, 0.1 mol/L NaHCO3. For re-chromatin immunoprecipitation (re-chIP) assays, complexes were eluted in 25 μL of 10 mmol/L DTT at 37°C for 30 minutes, and then subjected to the chIP procedure. Cross-linking was reversed overnight at 65°C; the DNA was purified with QIAquick columns (Qiagen).
Luciferase reporter assay. The dual luciferase kit from Promega (Southampton, United kingdom) was used as per the instructions of the manufacturer. Transfections in HCC-EV and HCC-BR116 cell lines were done with Genejuice (Novagen, Nottingham, United Kingdom) in complete medium as recommended. A 10-fold excess of promoter-firefly plasmid construct was added over transfection control plasmid expressing renilla luciferase. Cells were harvested 24 hours posttransfection and luciferase activity was assayed. Reporter activity was normalized by renilla luciferase activity.
Etoposide sensitivity studies. The HCC-EV1 cell line was treated with scrambled or psoriasin-specific siRNA and after 24 hours seeded at 100,000 cells per well on 24-well tissue culture plates. The MDA-MB-231-zeo and HP1 cell lines cells were seeded onto 24-well tissue culture plates at a density of 100,000 cells per well. Cells were then incubated in regular medium supplemented with etoposide in a concentration range from 100 pmol/L to 100 μmol/L. After 72 hours of continuous drug exposure, cells were detached with trypsin and counted using a Z2 particle and size analyzer (Coulter, Miami, FL). Each dose inhibition study was repeated thrice. IC50 values were calculated from the respective sigmoidal dose-response curves using Prism software.
Microarray analysis. Total RNA (10 μg) derived from the HCC-BR116 and HCC-EV1 cells was biotin labeled and fragmented according to standard protocols. Labeled cRNA was hybridized to Affymetrix Human Genome U133 Plus 2.0 arrays. The hybridized arrays were scanned using a GeneChip Scanner 3000. Data were converted to expression measures using the MAS5.0 algorithim (GeneChip operating software version 1.1) and raw data imported into the GeneSpring data analysis program version 6.1 (Silicon Genetics, Palo Alto, CA). Normalization was carried out to remove variability in the data by dividing each measurement by the 50th percentile of all measurements in that sample and then by dividing each gene by the median of its measurements in all samples. Following normalization, data were log transformed using the natural log. Genes were considered for further analysis based on standard GeneSpring filtering methods. Genes isolated for further analysis were required to have acceptable Affymetrix flag calls (present or marginal in all samples), >2-fold differential expression compared with control (in at least one sample), a t test P value below 0.05 (in all samples), and passed an error filter based on the Rocke-Lorenzato model (in all samples). The list of genes that satisfied the above specifications and the samples were clustered using a hierarchical clustering algorithim using Pearson correlation as the similarity measure, giving rise to a two-dimensional dendrogram.
Identification of psoriasin as a BRCA1-repressed target. To identify novel BRCA1-regulated target genes involved in the response to DNA damage, we used a microarray-based expression profiling approach. We compared the gene expression profiles of cells containing functional or mutant BRCA1 in the presence and absence of the topoisomerase IIα poison, etoposide. The cell lines used were derived from the BRCA1 mutant HCC1937 cell line by stable expression of either wild-type exogenous BRCA1 (HCC-BR116) or an empty vector (HCC-EV1; ref. 17). The parental cell line contains only one copy of the BRCA1 gene that encompasses a COOH-terminal deletion rendering the transcribed protein transcriptionally inactive. Total RNA was extracted in triplicate from each cell line in independent experiments, biotinylated, and hybridized to the Affymetrix human genome U133 Plus 2.0 representing 47,000 known transcripts and expressed sequence tags. Analysis of the microarray data revealed 303 genes that exhibited BRCA1-dependent expression changes. Unsupervised hierarchical clustering of genes based on their expression profiles, and samples based on their expression signature, was done using Pearson correlation as the similarity metric resulting in a two-dimensional dendrogram. The clustering of samples or condition tree is shown vertically and the clustering of genes or gene tree is shown horizontally ( Fig. 1A ). The samples diverged into two distinct clusters representing the HCC-BR116 and HCC-EV1 samples, respectively. The gene clusters also diverged into two main groupings at the metanode of the tree with 188 genes induced, and 115 genes repressed, by BRCA1 ( Fig. 1A; Supplementary Fig. S1). Of particular interest were the S100A gene family members, including psoriasin (S100A7), S100A8, and S100A9, that were markedly repressed by BRCA1 (5.4-fold, 8.3-fold, and 24.6-fold, respectively) and were displayed as a subset of the original tree ( Fig. 1B). Northern and Western blot analyses confirmed that psoriasin, S100A8, and S100A9 were potently repressed by BRCA1 in agreement with the microarray data ( Fig. 2A and B ). These genes have been implicated in a variety of different cancer types. Psoriasin, in particular, has been correlated to invasive breast tumors and ductal carcinomas in situ that show estrogen receptor negativity, poor differentiation, and the presence of lymphocytic infiltration ( 21, 22). As these pathologic features are also observed in BRCA1 mutant tumors, we postulated that psoriasin may represent a functionally important BRCA1 transcriptional target.
To ensure that the observed BRCA1-mediated repression of psoriasin expression was not a clonal artifact of the HCC-EV1 and HCC-BR116 cell line models, we repeated these experiments transiently. The HCC1937 cell line was transiently transfected with either pRcCMV-BRCA1 or an empty pRcCMV vector and protein lysates were extracted after 48 hours. Transient expression of exogenous BRCA1 dramatically reduced psoriasin expression, indicating that the observed effect in the HCC-BR116 cell line was not due to a clonal artifact ( Fig. 2C). To further emphasize the significance of this effect, we extended this analysis to additional cell line model systems. To study the effect of inhibiting endogenous BRCA1 function on psoriasin expression, we used the BG1-Neo ovarian cell line that constitutively expresses a BRCA1 antisense construct (AS4 cells) and an isogenic vector-transfected control cell line (BG1-Neo cells; ref. 23). In accordance with our previous data, the BRCA1 antisense-transfected AS4 cell line showed increased expression of psoriasin at protein level when compared with the vector-only control cell line ( Fig. 2D). To confirm that BRCA1 was required to suppress psoriasin in a third model system, we used a BRCA1-specific siRNA oligonucleotide to abrogate protein expression in the T47D breast cancer cell line. In agreement with the BG1 ovarian cancer cell line model, reduction of BRCA1 protein expression in the T47D cell line resulted in up-regulation of psoriasin expression ( Fig. 2E). Finally, we investigated if a BRCA1 disease-associated mutation would abrogate repression of psoriasin. To do this, we stably transfected BRCA1 mutant HCC1937 cells with full-length exogenous BRCA1 encoding a disease-associated mutation, A1708E, and evaluated its effect on psoriasin expression. Mutant BRCA1 completely failed to repress psoriasin, suggesting that psoriasin represents a physiologically important BRCA1 target ( Fig. 2F).
BRCA1-induced repression of psoriasin is mediated by c-Myc. To identify the mechanism by which BRCA1 represses expression of psoriasin, we analyzed the psoriasin promoter sequence for transcription factor binding sites and identified a c-Myc consensus sequence, CACATG (Genbank accession no. AF050167 bases 500-505). BRCA1 has previously been reported to bind to c-Myc, thereby repressing its ability to activate specific gene targets ( 11). In accordance with this observation, we showed a strong interaction between wild-type BRCA1 and c-Myc by coimmunoprecipitation in the HCC-BR116 compared with HCC-EV1 cells ( Fig. 3A ). Similarly, a clear interaction between BRCA1 and c-Myc was observed by coimmunoprecipitation in the BRCA1 functional BG1-Neo cell line but not in the AS4 cell line that stably expresses a BRCA1 antisense construct ( Fig. 3B). To determine if BRCA1 and c-Myc formed a complex on the psoriasin promoter, we did electrophoretic mobility shift assays (EMSA) using an oligonucleotide designed to include the CACATG c-Myc consensus sequence from the psoriasin promoter. Nuclear extracts derived from the BRCA1-reconstituted HCC-BR116 cells showed enhanced binding to the c-Myc consensus oligonucleotide compared with extracts derived from vector-transfected HCC-EV1 controls, suggesting that wild-type BRCA1 could stabilize c-Myc on this promoter sequence ( Fig. 3C). Consistent with these data, binding to the c-Myc consensus sequence was dramatically reduced when extracts derived from the AS4 cells was used compared with the BRCA1 functional control BG1-Neo–derived nuclear extract ( Fig. 3D). Preincubation of the extracts with the anti-BRCA1 antibody AB1, recognizing residues 1 to 304, abrogated the interaction in both cell line model systems. It has been previously shown that amino acids 175 to 303 and 443 to 511 of BRCA1 are required for the interaction with c-Myc ( 24); this, therefore, suggests that BRCA1 is present in the complex bound to the oligonucleotide ( Fig. 3C and D). Consistent with this proposal, preincubation of the extracts with an anti-c-Myc antibody resulted in a shift in the complex confirming the presence of c-Myc ( Fig. 3C and D). In contrast, incubation with anti-BRCA1 antibody D20 that recognizes residues 1 to 20 does not interfere with this interaction. Data presented suggest that wild-type BRCA1 mediates repression of psoriasin with c-Myc, possibly by acting as a corepressor, and that wild-type BRCA1 is recruited to the psoriasin promoter in a c-Myc–dependent manner. To examine this concept in greater detail, we carried out a chIP assay ( Fig. 4A ). ChIP patterns obtained on chromatin prepared from the HCC-BR116 cells showed that both wild-type BRCA1 and c-Myc are engaged on the psoriasin promoter along with the general transcription factors TATA box binding protein (TBP) and RNA Polymerase II. In contrast, the pattern obtained from the HCC-EV cells showed engagement of RNA polymerase II, TBP, and c-Myc, but not BRCA1, suggesting that loss of the last 33 amino acids of BRCA1 abrogates its ability to engage the psoriasin promoter. To confirm that BRCA1 and c-Myc are simultaneously complexed on the psoriasin promoter in BRCA1-proficient cells, a re-chIP assay was done ( Fig. 4B). The data show that BRCA1 and c-Myc are recruited concomitantly on chromatin prepared from HCC-BR116 cells but not on chromatin prepared from BRCA1-mutated HCC-EV cells. To confirm the importance of c-Myc in BRCA1-mediated repression of psoriasin expression, we transfected the HCC-BR116 cells with a firefly luciferase reporter construct containing either the wild-type psoriasin promoter or constructs containing point mutations (SDM1, SDM2, and SDM3) of the c-Myc consensus site that have previously been shown to inhibit c-Myc binding ( 25). The activity of the promoter was reflected in the level of luciferase activity measured using a luciferase assay ( Fig. 5A ). The wild-type construct was repressed in the HCC-BR116 cell line compared with the control HCC-EV cell line. In contrast, there was no inhibition of promoter activity in HCC-BR116 cells expressing the c-Myc–mutated constructs, demonstrating that this region of the promoter is required for BRCA1-mediated repression of psoriasin expression. Finally, inhibition of endogenous c-Myc expression in the HCC-BR116 cell line using a c-Myc–specific siRNA oligonucleotide resulted in reexpression of psoriasin despite the presence of functional BRCA1 ( Fig. 5B). Collectively, these data suggest that BRCA1 and c-Myc function together as a repressor complex and that repression is dependent on functional c-Myc and BRCA1.
Psoriasin mediates sensitivity to etoposide. The microarray data also indicated that S100A8 was etoposide inducible in the absence of functional BRCA1 (data not shown). As BRCA1-deficient cells have been reported to be highly sensitive to etoposide ( 16), we postulated that the S100 family of proteins may play a role in mediating the cellular response to DNA damaging–based chemotherapeutic agents. To explore this further, we studied psoriasin protein expression levels in HCCEV1 and HCCBR-116 cells upon treatment with a range of DNA-damaging agents, including etoposide, cisplatin, bleomycin, and 5-fluorouracil. Psoriasin expression was induced by all agents in the BRCA1-deficient HCC-EV1 cells compared with the BRCA1-reconstituted HCC-BR116 cells with highest levels being induced by etoposide ( Fig. 6A ). Subsequently, we studied the expression of psoriasin mRNA and protein in response to etoposide treatment in the HCC-EV1 and HCC-BR116 cells. Northern and Western blot analyses confirmed enhanced expression of psoriasin mRNA and protein in BRCA1-deficient cells following etoposide treatment ( Fig. 6B and C).
To examine if psoriasin expression was directly associated with response to etoposide, we carried out dose inhibition assays after siRNA inhibition of endogenous psoriasin in the BRCA1-deficient, etoposide-sensitive HCC-EV cells. Abrogation of psoriasin expression induced a >20-fold increase in resistance (IC30-Scr 4.88 × 10−6; psoriasin siRNA >10−4) to etoposide ( Fig. 6D). In addition, we carried out dose inhibition assays in MDA-MB231 cells stably expressing exogenous psoriasin (HP1) compared with vector-only transfected controls (MDA-MB231-neo). In agreement with our previous observation, expression of psoriasin in MDA231 cells induced a 40-fold increase in sensitivity (IC30-HP1, >6.9 ×10−5; MDA-MB231-zeo, 1.8 ×10−6) to etoposide ( Fig. 6E). Collectively, these data suggest that loss of BRCA1-mediated repression of psoriasin results in enhanced sensitivity to etoposide, implicating psoriasin as a functionally important BRCA1 target.
To identify BRCA1 and etoposide-regulated transcriptional targets, we did a microarray analysis comparing the gene expression profiles between the BRCA1 wild-type HCC-BR116 cell line and BRCA1 mutant HCC-EV1 cell line untreated or treated with etoposide. BRCA1 was observed to differentially regulate over 303 genes, including 189 that were up-regulated and 114 that were down-regulated. To further define genes that may be of particular interest, we examined subsets of targets that were also regulated by etoposide treatment. Of particular interest were the S100A gene family members, including psoriasin (S100A7), S100A8, and S100A9, that were markedly repressed by BRCA1 and induced by etoposide in the absence of functional BRCA1. We confirmed the microarray data by a variety of approaches and showed that BRCA1-mediated repression of psoriasin was dependent on both functional c-Myc and BRCA1. Consistent with its identification as a functionally important BRCA1 transcriptional target, we showed that psoriasin augments cellular sensitivity to etoposide.
Psoriasin is a member of the S100A family of proteins that is characterized by the presence of two calcium-binding EF-hand helix-loop-helix domains. These proteins have been implicated in cell growth, differentiation, and organization of the cytoskeleton and have been associated with a number of human conditions, such as Parkinson's disease, Alzheimer's diseases, hypertension, cystic fibrosis, and cancer ( 26). The present study has focused on psoriasin, which was initially identified as a protein highly expressed in the skin disease psoriasis, a condition characterized by epidermal proliferation and inflammation ( 27, 28). Importantly, as regards the present study, psoriasin has also been reported to be overexpressed in ∼80% of high-grade ductal carcinomas in situ and ∼50% of estrogen receptor–negative invasive breast cancers ( 22, 29, 30). Expression of psoriasin in human breast tumors represents a bad prognostic marker and correlates with lymphocyte infiltration and high-grade morphology ( 21, 30, 31). These pathologic features are also characteristic of BRCA1 mutant tumors, which raises the possibility that psoriasin may be responsible, at least in part, for the BRCA1 mutant phenotype. Our observations that wild-type BRCA1 interacts with c-Myc, that both proteins are in a complex bound to the c-Myc–responsive element in the psoriasin promoter, and that BRCA1-mediated repression of psoriasin is dependent on functional c-Myc suggest that both proteins cooperate to transcriptionally repress psoriasin expression. This is consistent with previous studies that have shown that BRCA1 and c-Myc interact to form a repressor complex that, for example, functions to repress the H-TERT gene ( 11). Recently, it has been reported that c-Myc is specifically amplified in BRCA1-associated breast cancers and in sporadic breast cancers where BRCA1 is epigenetically inactivated ( 32). These observations add further support to the suggestion that c-Myc deregulation may represent an important factor in the development and progression of BRCA1-deficient tumors.
Psoriasin expression has recently been reported to be a marker of basal (myoepithelial) differentiation in breast tissue ( 33), a phenotype that is also associated with BRCA1 mutant tumors ( 34). Basal breast tumors have been shown to possess a distinct gene expression profile that has been associated with a worse prognosis when compared with breast tumors that undergo ductal differentiation ( 35). Our data support a model where loss of BRCA1 function may result in expression of basal phenotype–associated genes, such as psoriasin, thereby conferring a worse prognosis.
We have also shown that expression of exogenous psoriasin in a breast cancer cell line confers sensitivity to the topoisomerase IIα poison, etoposide. This observation suggests that although psoriasin expression is a poor prognostic marker in breast tumors ( 21, 36), it may also represent a favorable predictive marker of response to topoisomerase IIα poisons. Our data suggest that loss of BRCA1 function may, at least in part, result in etoposide sensitivity through a psoriasin-mediated mechanism. Although the pathway by which psoriasin sensitizes cells to etoposide is unclear, it may be related to the putative role of psoriasin in regulating the p27KIP1 DNA damage checkpoint protein through JAB1 ( 19).
In conclusion, our findings show that BRCA1 and c-Myc function together to repress expression of psoriasin, a gene associated with the basal differentiation phenotype and poor prognosis in breast cancer. Furthermore, our data suggest that psoriasin expression may be a useful marker of loss of BRCA1 function and increased topoisomerase IIα poison sensitivity in human breast tumors.
Grant support: Research and Development Office of Northern Ireland (R.D. Kennedy), Breast Cancer Campaign UK (J.E. Quinn), Medical Research Council UK (G.E. Stewart), Cancer Research UK (J.J. Gorski, C.R. James, P.B. Mullan, P.G. Johnston, and D.P. Harkin), National Cancer Institute of Canada Studentship Award Grant (E.D. Emberley), and Canadian Institutes of Health Research Scientist Award (P.H. Watson).
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Note: R.D. Kennedy, J.J. Gorski, and D.P. Harkin contributed equally to this work.
Supplementary data for this article are available at Cancer Research Online (http://cancerres.aacrjournals.org/).
- Received May 26, 2005.
- Revision received July 11, 2005.
- Accepted September 2, 2005.
- ©2005 American Association for Cancer Research.