BCAR1, a Human hom*ologue of the Adapter Protein p130Cas, and Antiestrogen Resistance in Breast Cancer Cells (2024)

Search for other works by this author on:

Article history

The steroid hormone estradiol plays a pivotal role in breastcancer initiation and cell proliferation (1,2). Estradiol andits receptor participate in a multiprotein complex that acts as atranscription regulator for various genes (3-6). Inapproximately two thirds of all breast tumors, the estrogen receptor(ER) is present and functionally active (7-9). The presenceof the ER in breast carcinoma cells and their responsiveness toestrogens is clinically exploited by the administration of antagonistsof estrogens or antiestrogens (10-12). On binding,antiestrogens induce an aberrant conformation of the ER, resulting inan inactive transcription complex (13). As a consequence, theantiestrogen-receptor complex may bind DNA but can no longer modulatethe expression of its target genes (14). This inhibition ofthe estrogen response pathway (15) may block cellularproliferation (19,20).

During the last 2 decades, antiestrogens, particularly tamoxifen, have proved to be effective inthe treatment of hormone-responsive breast cancer (21-23). Adjuvanttreatment with tamoxifen reduces tumor recurrence and increases survival of patients withER-positive breast cancer (24,25). In metastatic breast cancer, tamoxifenleads to objective response in nearly one half of patients with ER-positive primary tumors (26). Resistance to antiestrogens, however, is a serious obstacle in themanagement of breast cancer. About 40% of ER-positive tumors fail to respond toantiestrogen therapy (intrinsic resistance) (27), whereas eventually most,if not all, breast tumors that initially respond to antiestrogens develop resistant metastases(acquired resistance).

The mechanisms underlying intrinsic or acquired antiestrogen resistance of breast tumors arestill poorly understood. Altered pharmacology of the antiestrogens (30-32), andchanges in the interactions between tumor cells and their environment (paracrine interactions) (33) have been proposed to contribute to the development of antiestrogenresistance. In addition, genetic or epigenetic changes in the tumor cells that promote thedevelopment of antiestrogen resistance have been postulated (34,35).Although each of these possibilities may account for or contribute to the resistant phenotype inindividual patients, none so far has been shown to explain antiestrogen resistance in a majority ofpatients. It is likely that progression to antiestrogen resistance is a multifactorial process.

We attempted to identify the genetic factors that may lead to antiestrogen resistance.Therefore, we applied retroviral-insertion mutagenesis to the human estrogen-dependent breastcancer cell line ZR-75-1. We demonstrated that random integration of a defective murineretrovirus in the genome of ZR-75-1 cells transformed the cells from an estrogen-dependent to atamoxifen-resistant phenotype (36). Mapping of common integration sitesand cell fusion-mediated gene transfer experiments so far have identified three independent locithat are involved in antiestrogen resistance (BCAR1, BCAR2, and BCAR3) (37). Here we report the isolation and characterization of the target gene of the BCAR1locus and its functional involvement in antiestrogen resistance in vitro.

Materials and Methods

Cell Lines

The human breast carcinoma cell line ZR-75-1 and derivatives thereof were cultured asdescribed elsewhere (36,38). Hybrid cell lines and BCAR1-transfected celllines were cultured in RPMI-1640 medium supplemented with 10% heat-inactivated bovinecalf serum and 1 nM 17β-estradiol. For BCAR1-transfected cell lines, 500μg/mL of the neomycin analogue G418 (Life Technologies, B.V., Breda, The Netherlands)was added to the culture medium.

The tamoxifen-resistant cell line XI-13, with two copies of the defective LN retroviralgenome (36,39), one located within the BCAR1 locus, was lethally γirradiated (40 Gy) and subsequently fused to hygromycin B-resistant ZR-75-1 (ZH3D7) cells withpolyethylene glycol as described previously (37). G418 plus hygromycinselection was performed in estradiol-containing culture medium (37).

Molecular Biology Techniques

A human genomic cosmid library in pJB8 (40) was obtained from W.J. M. van de Ven (University of Leuven, Leuven, Belgium) and screened by use of standardprotocols (41).

Exon trapping was performed according to the manufacturer's instructions (LifeTechnologies, Inc). In short, genomic DNA of the cosmid clones HC26 and HC34 was partiallydigested with the enzyme MboI. Fragments of approximately 5-10 kilobases (kb)(average, 7.5 kb) were isolated and cloned in the exon-trap vector pSPL3. Batches of pooledpurified plasmid DNA from five clones each were transfected into the simian kidney cell line(COS-1) (41). After transient expression, RNA was isolated andcomplementary DNA (cDNA) was synthesized. Following two rounds of polymerase chainreaction (PCR), the trapped sequences were analyzed, cloned, and sequenced.

RNA was isolated from the somatic hybrid D4E5 (containing the BCAR1 locus) and culturedfor 2 days with 4-hydroxy-tamoxifen by use of the CsCl-guanidine thiocyanate procedure (41). Poly(A) RNA was selected with the use of the PolyATtract mRNAIsolation System (Promega Corp., Madison, WI) and converted into cDNA (Zap Express cDNAand cloning kit; Stratagene Cloning Systems, La Jolla, CA) as recommended by the manufacturer.A cDNA library of approximately 2 × 10⁶ phages was created and screened.Twenty-one phage clones hybridizing with a BCAR1 exon-trap probe were isolated andsequenced by use of standard procedures (41). DNA and protein sequenceanalyses and alignments were performed with the use of DNASIS (Hitachi Software EngineeringAmerica Ltd., Brisbane, CA), Blast (www.ncbi.nlm.gov/blast), and Clustalw(www2.ebi.ac.uk/clustalw) software. The BCAR1 sequence has been assigned GenBankAccession No. AJ 242987.

Northern blots were prepared by use of 1% agarose-formaldehyde gels as describedpreviously (41). Blots were hybridized with random-primer-labeled probesfor BCAR1, the estrogen-regulated PS2 gene, and the glyceraldehyde-3-phosphatedehydrogenase gene for loading control as described (42,43)

Mapping of the BCAR1 locus was performed on genomic DNA from a panel of 30 somaticcell hybrids obtained from D. F. Callen (Women's and Children's Hospital, Adelaide,Australia). PCR was performed with the locus-specific primers 11405′-CCCCACATACCCAGCACA-3′ and 11425′-CCCAGCTCCTCCTTATTCA-3′. The amplified product was verified byhybridization with the BCAR1 locus-specific probe 18/1 (36).

Expression Constructs and Transfection of BCAR1 in ZR-75-1 Cells

The full-length BCAR1 cDNA was inserted into the EcoRI-XhoI site of thelong terminal repeat-promoted pLXSN expression vector (39) and into amodified version of the episomal vector LZRS-IRES-Neo (LZRSpBMN-IRES) (44). In the latter vector, the BCAR1 cDNA and the neomycin resistance gene (to conferselectability) were separated by an IRES sequence, so that both genes were under control of asingle long terminal repeat (LTR) promoter and transcribed as a polycistronic messenger.

Stable transfectants were generated with the pLXSN/BCAR1 construct by the use of calciumphosphate precipitation (41). Transfection of ZR-75-1 cells with theepisomal LZRSpBMN-IRES/BCAR1 vector construct was performed by use of Lipofectinreagent (Life Technologies, Inc.), according to the manufacturer's instructions.Transfectants were selected for neomycin resistance (500 μg of G418/mL) inestradiol-containing medium.

Proliferation Assay

Somatic hybrid cells, BCAR1, and mock-transfected cells were grown in culture mediumsupplemented with 1 nM estradiol. Cells were harvested at 80% confluence bytrypsinization. The cells (0.7 × 10⁶) were seeded in 25-cm² flasks, in triplicate, and grown in the presence of either 1 μM OH-tamoxifen or 100nM of the pure antiestrogen ICI 182,780 (Zeneca Pharma, Ridderkerk, TheNetherlands). After 4-6 days, the cells were again trypsinized, counted using a Coulter Z1 cellcounter (Coulter Electronics Ltd, Luton, U.K.), and reseeded in triplicate as above. To follow theproliferation of the cells, the procedure was repeated at several time points and for each cell line.In case of limited recovery of cells at the end of the assay period, only two flasks were reseededor the culture was terminated. Proliferation was determined as the fold multiplication (mean of atleast two flasks) relative to day 0.

Drug-sensitivity tests on ZR-75-1-derived cell lines were performed in complete mediumcontaining estradiol in 96-well plates. Five thousand cells were plated with threefold dilutions(four wells each) of the drug. After 9 days of culture in the presence of the drug, viable cells weremeasured by use of the MTT assay as described (38)

Bcar1 Protein Detection

BCAR1 and mock-transfected cells were trypsinized, rinsed with phosphate-buffered saline,sonicated for 10 seconds, and lysed for 10 minutes at 100 °C in 1% sodium dodecylsulfate (SDS) and 10 mM Tris-HCl (pH 7.5). The protein concentration of the crudelysates was determined (BCA™, Protein assay reagent; Pierce Chemical Co., Rockford, IL).Equal amounts (3.3 μg) of protein were used for SDS-polyacrylamide gel electrophoresis(PAGE) (8% acrylamide) and subjected to western blot analysis as described (45). Bcar1 was visualized with the mouse monoclonal antibody to rat p130Cas(Transductional Laboratories, Lexington, KY). In addition, polyclonal antibodies raised againstthe N-terminus and C-terminus of rat p130Cas (Cas N-17 and Cas C-20; Santa CruzBiotechnology Inc., Santa Cruz, CA) were used to identify Bcar1 in an independent controlexperiment.

Results

Transfer of the BCAR1 Locus by Cell Fusion

We have used retrovirus insertion-mediated mutagenesis on the humanestrogen-dependent breast cancer cell line ZR-75-1 to generate 80 celllines showing resistance to 4-hydroxy-tamoxifen (36).Integration of the defective murine retrovirus (LN) into the host cellgenome could transform these cells to a tamoxifen-resistant phenotype.With the use of integration site-specific probes, we previously found acommon integration locus for the retrovirus in a subset of this panelof cell lines. Four independent cell clones were identified by Southernblot analysis, with a viral integration in the same orientation atdifferent positions within a 2.2-kb genomic region (36)(see also Fig. 2, A). We termed this locus BCAR1, for breastcancer antiestrogen resistance 1 locus—i.e., the first BCAR locusidentified.

To establish whether retroviral integration in the BCAR1 locus induces tamoxifen resistancein ZR-75-1 cells, we transferred the genomic region encompassing the BCAR1 locus fromtamoxifen-resistant cells into estrogen-dependent ZR-75-1 cells by cell fusion. XI-13 cells carrytwo copies of the defective retroviral genome, one copy of which is located within the BCAR1locus. On lethal γ irradiation, the cells were fused to hygromycin B-resistant ZR-75-1 cells(ZH3D7) with polyethylene glycol as described previously (37). Theneomycin resistance gene in the integrated retrovirus allowed the isolation of somatic cell hybridsby dual selection with G418 plus hygromycin in estradiol-containing culture medium. Twelve cellhybrids containing either one of the integration loci in the ZR-75-1 background were rescued andanalyzed. Southern blotting analysis by use of a specific probe for the neomycin resistance generevealed that five hybrids had retained the BCAR1 integration locus (a 15-kb BglIIfragment), whereas the remaining seven carried the other integration locus (a 10-kb BglII fragment). Both the virus-induced BCAR1 cell line XI-13 and the BCAR1 locus-containing cellhybrids derived therefrom (e.g., D4E5 and D4E6) showed enhanced proliferation when culturedin the presence of the antiestrogen ICI 182,780 (Fig. 1, A). Hybrid celllines carrying the other locus, as well as the ZH3D7 recipient cells, were growth inhibited. In theabsence of antiestrogens, BCAR1 locus-containing hybrids also demonstrated increasedproliferation compared with the control cell lines (Fig. 1, B). Note that theproliferation of most cell lines is more efficient in the absence of antiestrogen, which is likelybecause of some residual estrogens in the serum-containing medium. In the presence ofantiestrogens, the ER pathway is more effectively blocked. The lack of the ER in XI-13 cells (36) can explain the comparable growth levels of these cells in the presenceas well as in the absence of antiestrogens (Fig. 1). These fusionexperiments demonstrated unambiguously that the BCAR1 locus induced antiestrogen-resistantproliferation of the somatic cell hybrids and that the BCAR1 phenotype is dominant.

Cloning of the BCAR1 Locus

A human cosmid library was screened with genomic probes that havebeen employed to identify the BCAR1 locus (36). The probesincluded a 500 base-pair (bp) ApaLI-PstI fragment(14B2) and a 450-bp SphI fragment (18/1). Both probes resideat the center of the viral integration site in the BCAR1 locus (Fig. 2, A). Cosmids were isolated and analyzed by restrictionmapping, resulting in a continuous contig of approximately 80 kb (Fig.2, A). Unique probes derived from the 80-kb contig identified nofurther proviral integrations events in the 80 antiestrogen-resistantcell lines from our panel.

Identification of the BCAR1 Gene

Two of 10 overlapping cosmid clones covering the major part of theBCAR1 locus—HC26 and HC34—were selected for identification oftranscripts by exon trapping. A total of 31 trapped sequences wererecovered, of which 16 were related to known genes. Eight clonescorresponded to exons 2-7 of the chymotrypsin gene (accession No.NM_001906), and the remaining eight appeared to be hom*ologous tosequences of the rat p130Cas gene (accession No. D29766) (46).Of the remaining 15 clones, two seemed to be generated because ofcryptic splicing events in frequently occurring alu-type repetitivesequences in the BCAR1 locus, while the other clones did not showhom*ology to any known sequences in available databases.

The trapped exons were hybridized to polyA messenger RNAs (mRNAs) isolated from theantiestrogen-resistant BCAR1 cell lines (XII-13, XI-14, and XI-13) and from the parental cell lineZR-75-1. The chymotrypsin gene sequences did not hybridize in a northern blot analysis,indicating that this gene is not expressed in our ZR-75-1-derived cell lines. In contrast, exonsequences that were hom*ologous to the rat p130Cas gene hybridized to a single mRNA of 3.2 kb(Fig. 2, B). This transcript appeared to be increased in theantiestrogen-resistant cell lines but not in the parental cell line. As shown in Fig. 2, B, ZR-75-1 exhibits only low levels of an equally sized transcript. The differentialexpression of this transcript suggested that the rat p130Cas-hom*ologous gene was the candidatebreast cancer antiestrogen resistance gene, BCAR1. Conclusive evidence for its role inantiestrogen resistance required the isolation of the BCAR1 cDNA and its functional transfer toZR-75-1 cells by transfection (see below).

Primary Structure of the BCAR1 Gene

A cDNA library from the antiestrogen-resistant BCAR1 cell hybridD4E5 was screened with p130Cas hom*ologous sequences retrieved from exontrapping. Twenty-one positive cDNA clones were identified and assembledby sequence alignment into a continuous cDNA of 3208 bp. The BCAR1 cDNAencloses an open-reading frame of 2610 nucleotides. At position 122, aninitiation codon is flanked by sequences matching Kozak's criteria(47). The open-reading frame is flanked at the 3′ end by atranslation termination codon (TGA) at position 2732 and a 3′untranslated sequence of 468 nucleotides that contains multiple stopcodons. A canonic polyadenylation site is located 14 bp in front of thestart of the polyA tail in the cDNAs. The GenBank Accession No. for theBCAR1 sequence is AJ 242987.

The open-reading frame has a coding capacity for a protein of 870 amino acid residues, with acalculated molecular mass of approximately 93 kd. The predicted protein features a Srchom*ology 3 (SH3) domain in the N-terminal part and multiple potential tyrosine phosphorylationsites (Tyr-Ala/Pro-Xxx-Pro) in the central part of the protein (Fig. 3). TheC-terminal part encloses a proline-rich stretch (Arg-Pro-Leu-Pro-Ser-Pro-Pro) that may interactwith the SH3-binding site of the Src protein (48). Sequence alignment ofBcar1 revealed extensive hom*ology (91%) with rat and mouse p130Cas protein (Fig. 3).

When hybridized with BCAR1 cDNA probes, RNA preparations of BCAR1 cell linesdisplayed a single transcript of approximately 3.2 kb (Fig. 2, B). The sizeof the mRNA is in agreement with that of the cDNA, suggesting that the cDNA represents thefull-length transcript. Northern blots of mRNAs derived from a panel of multiple human tissueshybridized to the BCAR1 cDNA probe showed that the BCAR1 transcript is widely expressed(Fig. 4). The 3.2-kb mRNA was present in all tissues tested, being mostabundant in testis and with only a low representation in liver, thymus, and peripheral bloodleukocytes.

Additionally, in vitro transcription and translation of the BCAR1 cDNAdemonstrated that the cDNA encodes a protein with an apparent molecular mass of approximately116 kd on SDS-PAGE. The discrepancy in observed and calculated (93-kd) molecular mass isprobably because of folding of the molecule, causing aberrant gel mobility. Both rat p130Cas andthe Bcar1 protein exhibit the same size. As expected, a monoclonal antibody directed against ratp130Cas identified the BCAR1 gene product (Fig. 5, A) as well aspolyclonal antibodies that are directed against either the N-terminus or the C-terminus of the ratp130Cas (data not shown).

Genomic Organization of BCAR1

The genomic organization of the BCAR1 gene was delineated by use ofthe BCAR1 cDNA as a template. Restriction mapping and partialsequencing of cosmid clones HC26 and HC3 revealed that the transcribedregion of the BCAR1 gene consists of seven exons (I-VII) spanningapproximately 25 kb of genomic DNA (Fig 2, B). The size of the exonsvaries from 76 to 1101 bp, and the size of the introns varies from 182bp to approximately 8.5 kb. Sequencing through the exon-intronboundaries demonstrated that all splice junctions conform to the GT/AGrule for exon-intron junctions. Southern blot analysis showed that allrestriction fragments from the cosmid clones that hybridized to BCAR1cDNA were also present in genomic DNA. More important, the codingsequence of the BCAR1 gene was found to be identical to that of thecDNA, excluding cloning artifacts.

Exon 1 of BCAR1 contained the first 12 bp of the coding sequence. It is of interest that thisposition of the exon-intron boundary exactly matched the position where alternative splicing hasbeen observed in rat as well as mouse p130Cas (46,49). Exons 2-6 allmapped within the coding sequence, and exon 7 contained the last 509 coding nucleotides and thecomplete 3′ nontranslated region. The 5′ flanking region of the BCAR1 gene had ahigh G + C content (80%) and lacks suppression of CpG dinucleotides, which ischaracteristic for CpG islands. Such a configuration, which is often found for housekeeping genes,is consistent with the ubiquitous expression of the BCAR1 gene (Fig. 4).

In situ hybridization with the complete cosmid HC26 suggested that the BCAR1gene was located at human chromosome 16q22-23 (Van Agthoven T: personal communication).Primers specific for the integration region were used to analyze a panel of radiation hybridscontaining fragments of human chromosome 16. PCR analysis on genomic DNA derived fromthese hybrid cell lines determined the position of the BCAR1 locus to 16q23.1. Hybrids lackingthe region 16q23.2 to telomere (e.g., CY145) gave no BCAR1-specific product, whereas hybridsthat retained the 16q23.1 region (e.g., CY116 and CY117) gave specific products (50). Evidence for this position of the BCAR1 gene was further supported by theconcurrent trapping of exons of the chymotrypsin gene that had been assigned to chromosome16q23.1 (50). Hybridization with chymotrypsin gene probes on theBCAR1 cosmid clones confirmed the position of the chymotrypsin gene 4-5 kb downstream of theBCAR1 gene (Fig. 2, A).

Considering the vast hom*ology of the BCAR1 cDNA with rat and murine p130Cas sequencesand the strict conservation of the first exon-intron boundary, we postulate that the BCAR1 gene isthe human p130Cas hom*ologue, and we will further refer to this gene as BCAR1/p130Cas.

Transfection of BCAR1/p130Cas cDNA and Estrogen Independence in ZR-75-1Cells

To demonstrate that Bcar1/p130Cas is functionally involved inantiestrogen resistance, the full-length BCAR1/p130Cas cDNA wasintroduced in antiestrogen-sensitive ZR-75-1 cells by transfection.Five independent LXSN/BCAR1/p130Cas transfectants were generated,carrying an intact transgene integrated in the host genome asdetermined by Southern blot analysis. Similarly, eight independent celllines carrying the BCAR1/p130Cas cDNA in an episomal vector wereestablished. In addition, more than 22 vector-control cell lines havebeen raised.

All transfected cell lines were generated in estrogen-containing medium and subsequentlytested for proliferation in the presence of either 4-hydroxy-tamoxifen or the pure antiestrogen ICI182,780. A typical example of such an assay with the use of 4-hydroxy-tamoxifen is shown in Fig.5, A. Aliquots of the cultured cells have been used to measure the amountof Bcar1/p130Cas protein with the use of western blotting and immunodetection (Fig. 5, A). All cell lines exhibiting high expression of Bcar1/p130Cas (e.g.,4A12, B4, B6, and C4 and the somatic cell hybrid D4E6) showed net proliferation in the presenceof 4-hydroxy-tamoxifen. In contrast, ZR-75-1 and all vector control cell lines (e.g., C2, C6, D2,and D4) expressed very low levels of or no Bcar1/p130Cas and failed to proliferate in thepresence of the antiestrogen. Similar results were obtained in the assays by the use of the pureantiestrogen ICI 182,780 (data not shown). Of interest, Bcar1/p130Cas-transfected cellsunderwent a morphologic change when cultured in the presence of antiestrogens (data notshown). The cells displayed a typical flattened shape with multiple protrusions (similar to thevirus-induced cell lines), which may be suggestive of a reorganization of the cytoskeleton. Furtherexperiments with the use of immunofluorescence are needed to verify these observations. Toinvestigate the possibility that Bcar1/p130Cas overexpression also generates resistance to otherdrugs, we performed cell survival assays by use of doxorubicin, 5-fluorouracil, and methotrexate.These pilot experiments showed no relation between sensitivity to these drugs and overexpressionof Bcar1/p130Cas; i.e., comparable ID₅₀ (dose showing 50% inhibition) valueswere obtained for overexpressing cell lines and cell lines without overexpression (data notshown).

Northern blotting analysis of Bcar1/p130Cas-transfected cells grown under antiestrogenselection revealed an impaired expression of the estrogen-regulated PS2 gene (43) similar to the parental cells (Fig. 5, B), suggesting that theproliferation of these cells was independent of ER function. Because integrin and growthfactor-receptor pathways may signal through p130Cas to the mitogen-activated protein kinase(MAPK) pathways (51-54), we have performed preliminary experimentsto monitor the activation state of various MAPKs in BCAR1/p130Cas-transfected cell lines.Specific antibodies directed against activated MAPKs (ERK, p38, and JNK) failed to documentconstitutive activation of these kinases in the presence of antiestrogens (not shown). Because ofthe slow growth and response kinetics of the transfected cells in antiestrogen-supplementedcultures, the levels of activated MAPK may have been below the threshold of detection in ourtests.

Discussion

The mechanisms underlying antiestrogen resistance in breast cancerare the subject of intensive study. Our working hypothesis was based onthe assumption that genetic or epigenetic alterations in the tumorcells contribute to the failure of the therapy. To identify targetgenes, we have employed a model system of estrogen dependenceresembling that seen in human breast cancer. In this article, wedescribe the cloning and characterization of the BCAR1 gene. On thebasis of our findings, BCAR1 appears to be the human hom*ologue of therat and mouse p130Cas genes; their function is described below.Overexpression of Bcar1/p130Cas in human estrogen-dependent ZR-75-1breast cancer cells is sufficient to drive cell proliferation in thepresence of antiestrogens. No difference in structure or size of theBCAR1/p130Cas transcript was seen in the virus-infected cell lines.This suggests that overexpression of the BCAR1/p130Cas gene is achievedby increased activity of the BCAR1/p130Cas gene promoter, likely due tothe presence of strong enhancer elements within the viral LTRs(55). This mechanism of transcription enhancement is differentfrom the observed viral promoter-insertion activation of the BCAR3 genethat resulted in an altered BCAR3 transcript and protein (43).It is possible that such events are selected against, in the case ofthe BCAR1/p130Cas gene, to ensure an intact and functional protein.

Our experiments indicate that BCAR1/p130Cas-mediated antiestrogen resistance is notassociated with resistance to other drugs relevant to the treatment of breast cancer. In addition,the ER does not play a crucial role in the resistant phenotype. Cell lines obtained after viralintegration in the BCAR1 locus (e.g., XI-13) completely lack ER expression and function (36) and show comparable growth rates in the absence or presence ofantiestrogens (Fig. 1). The somatic cell hybrids containing the BCAR1locus and the BCAR1/p130Cas transfectants are estrogen responsive with respect to growth andregulation of PS2 expression. These variants exhibit antiestrogen-resistant cell proliferation butare not dependent on antiestrogen for growth (Fig. 1, B). Theseobservations suggest a cell proliferation regulatory mechanism involving Bcar1/p130Cas, whichbypasses the ER-regulated pathway.

The observation that the BCAR1 gene is highly hom*ologous to the well-characterized rat andmouse p130Cas genes may help to understand the mechanism of antiestrogen-resistant breastcancer cell proliferation. Rat p130Cas has been identified as the major tyrosine-phosphorylatedprotein in v-Src- and v-Crk-transformed rat cells (46,56). Subsequentreports have implicated p130Cas in various processes in different cell types, including celltransformation (57,58), linking the extracellular matrix with the actincytoskeleton (59,60), integrin signaling (51,61-64), growth factor-receptor signaling (53,54,65-68), antigen-receptorsignaling (73,74), and essential cardiovascular development (58). This varied role of p130Cas may be explained by the presence ofseveral protein-protein interaction domains (46,75), which is the hallmarkof this novel class of adapter proteins. Family members are HEF1 (also designated CasL) (76,77) and Efs (also termed Sin) (78,79), whichmay have distinct functions (80). The SH3 domain, in particular, is wellconserved between these family members and has been shown to interact with proline-rich targetsequences from FAK (49), RAFTK (81),PTP-PEST (82), CAKbeta (83), and PTP1B (52). The central part of p130Cas contains multiple potentialtyrosine-phosphorylation sites capable of interacting with SH2 domain-containing proteins likeCrk (84,85), Src (48), and Nck (51). The C-terminal part of the p130Cas family members is well conserved but has notyet been implicated in a particular function in mammalian cells. Together, the numerousdocumented interactions and functions of p130Cas in various biologic processes and in differentcell types suggest a dynamic role for p130Cas. The outcome of these complex interactions willdepend on the availability of binding partners (including their activation status and affinity) andthe specific process in that cell type.

So far, little information is available regarding the role of Bcar1/p130Cas in regulation ofproliferation of breast epithelial cells. The observation that expression of Bcar1/p130Cas inprimary breast cancer predicted disease prognosis and response to tamoxifen therapy (45) suggests a genuine role for Bcar1/p130Cas in growth control of breast cancercells. Further elucidation of the pathway involved in Bcar1/p130Cas-mediated cell proliferation inbreast cancer cells may identify in vitro key regulators and possibly novel clinical targets.The recognition that both BCAR1/p130Cas and BCAR3 (42) can controlproliferation of breast cancer cells further supports the validity of our model system to identifybreast cancer antiestrogen resistant genes.

Supported by grant DDHK 94-847 from the Dutch Cancer Foundation.

We thank Drs. M. Look, J. Veldscholte, and M. Schutte for critically reviewing thismanuscript. We also thank Drs. J. A. Foekens and T. van Agthoven for their stimulatingdiscussions, Dr. K. Nooter and K. van Wingerden for their contribution to the testing of drugs,Dr. J. G. Collard for the gift of the episomal expression vector, and Dr. A. M. Cleton-Jansen forher assistance with the analysis of the somatic cell hybrid panel.

Oxford University Press

Issue Section:

Articles

Email alerts

Citing articles via

More from Oxford Academic