Tyrosine Phosphorylation of Sam68 by Breast Tumor Kinase Regulates Intranuclear Localization and Cell Cycle Progression*

The breast tumor kinase (BRK) is a growth promoting non-receptor tyrosine kinase overexpressed in the majority of human breast tumors. BRK is known to potentiate the epidermal growth factor (EGF) response in these cells. Although BRK is known to phosphorylate the RNA-binding protein Sam68, the specific tyrosines phosphorylated and the exact role of this phosphorylation remains unknown. Herein, we have generated Sam68 phospho-specific antibodies against C-terminal phosphorylated tyrosine residues within the Sam68 nuclear localization signal. We show that BRK phosphorylates Sam68 on all three tyrosines in the nuclear localization signal. By indirect immunofluorescence we observed that BRK and EGF treatment not only phosphorylates Sam68 but also induces its relocalization. Tyrosine 440 was identified as a principal modulator of Sam68 localization and this site was phosphorylated in response to EGF treatment in human breast tumor cell lines. Moreover, this phosphorylation event was inhibited by BRK small interfering RNA treatment, consistent with Sam68 being a physiological substrate of BRK downstream of the EGF receptor in breast cancer cells. Finally, we observed that Sam68 suppressed BRK-induced cell proliferation, suggesting that Sam68 does indeed contain anti-proliferative properties that may be neutralized in breast cancer cells by phosphorylation.

Sam68, initially identified as a c-Src-associated substrate during mitosis of 68 kDa (1,2), has since been reported to be a substrate of other Src family tyrosine kinases as well as BReast tumor kinase (BRK) 5 (3) and ZAP70 (4). Sam68 is a prototype of STAR proteins, so named for their implication in signal transduction and activation of RNA metabolism (5). STAR proteins contain the GSG (GRP33, Sam68, and GLD-1) or STAR domain, an evolutionarily conserved protein module initially identified by aligning the first three members of this family (6). Although the precise role of Sam68 is unknown, its multifunctionality is revealed by the presence of polyproline sequences, an RNA-binding K homology module shared by all GSG/STAR proteins and multiple potential tyrosine phosphorylation sites (5). The proline-rich sequences mediate interaction with SH3 domains of signaling molecules (7) and the WW domain containing proteins (8). The K homology domain of Sam68 has been shown to bind homopolymeric RNA poly(U), poly(A) (2,9,10), and synthetic RNA sequences with a core UAAA (11). However, its in vivo RNA targets have just begun to be identified (12).
Several reports have indicated that phosphorylation regulates key cellular roles of Sam68. Both tyrosine phosphorylation and SH3 binding severely hamper the RNA binding capability of Sam68 (13,14). Sam68 was found to interact and colocalize with the splicing-associated factor YT512-B, to synergize with the human immunodeficiency virus Rev protein, enhancing export of unspliced viral RNA, and to increase protein expression from RNA containing the constitutive transport element of some retroviruses (3,(15)(16)(17). On the other hand, activation of the Ras-ERK pathway generates ERK-threonine-phosphorylated Sam68 that in turn enhances inclusion of exon v5 of the CD44 pre-RNA (18). Thus phosphorylation is an essential regulatory mechanism of the RNA-binding activity of Sam68.
Recently elevated phosphorylation of Sam68 and its correlation to increased acetylation was reported in breast cancer cell lines Hs578T, MDA-435, and MDA-468 (19), implying that tyrosine phosphorylation of Sam68 may be required for the invasiveness of these cancer cells. In fact, the involvement of Sam68 in tumor progression is supported by the finding that depletion of Sam68 is associated with neoplastic transformation (20). Consistent with this observation, it was demonstrated that overexpression of Sam68 blocks cell cycle progression (21). In contrast, overexpression of the RNA binding-defective splice variant (Sam68⌬KH), but not wild-type Sam68 was shown to suppress cell growth (22). Furthermore, the targeted disruption of Sam68 in DT20 cells was shown to cause growth retardation (23), which is inconsistent with earlier studies (20). Therefore, additional studies are required to define the role of Sam68 in cell cycle progression.
BRK is a non-receptor tyrosine kinase that belongs to the classified BRK family tyrosine kinases that include Frk, Srm, and Src42A (24). BRK contains an SH3, an SH2, a kinase domain, and a C-terminal regulatory tyrosine in a similar arrangement as Src family kinases, but lacks the myristoylation signal conserved within the Src family. We previously identified BRK as the first nuclear kinase that can phosphorylate Sam68 and negatively regulate its RNA-binding activity (3) and recently showed that both the Sam68-like proteins, SLM-1 and SLM-2, are also substrates of BRK (25). Other substrates of BRK comprise the adaptor proteins BKS (26) and paxillin (27). Furthermore, BRK has been shown to associate with the epidermal growth factor (EGF) receptor, GAPassociated p65 protein (28), erbB3/HER3 (29), and protein kinase B/Akt (29,30). BRK is overexpressed in more than 65% of breast tumors (31). Small interfering RNA (siRNA) silencing of BRK was shown to specifically suppress BRK expression, resulting in significant inhibition of the proliferation of the T-47D breast cancer cell line (32). Besides, BRK was shown to potentiate the mitogenic effects of EGF stimulation (29,33). These studies indicate that BRK indeed has growth promoting activities and is able to stimulate cell cycle progression. Wild-type BRK is known to reside throughout the cell and the overexpression of Sam68 promotes its nuclear localization (25). The nuclear expression of BRK occurs in normal prostate epithelial cells as well as in 70% of benign prostate hyperplasia, but not in high grade prostate intraepithelial neoplasia (34).
We have previously shown that both BRK and Sam68 reside in Sam68 nuclear bodies (SNBs), a novel nuclear structure that we discovered in 1999 (35). The C terminus of Sam68 is clustered with 16 tyrosine residues, all potential tyrosine phosphorylation sites. This region also harbors an unconventional nuclear localization signal (NLS) comprising the last 24 amino acids ( 420 RPSLKAPPARPVKGAYREHPYGRY 443 ) of the protein with tyrosine residues at positions 435, 440, and 443 (36). Mapping the phosphorylated tyrosines in the C-terminal of Sam68 is challenging, as the tyrosine residues are often located in clusters. Besides, traditional tryptic mapping is hampered by the scarcity of lysine and arginine residues in the C terminus. We have overcome this problem by generating Sam68 phospho-specific antibodies. We demonstrate that BRK phosphorylates Sam68 on all three tyrosines in the NLS. We show that tyrosine 440 dictates the localization of Sam68, as mutating it to phenylalanine completely blocks nuclear localization. Our data also identify Sam68 as a downstream substrate of BRK in breast cancer cells in response to EGF treatment. We also show that Sam68 inhibits cell cycle progression, whereas BRK has a stimulatory effect on cell growth. Co-infection of Sam68 and BRK in astrocytes prevents cell cycle progression as fewer cells were able to transverse the S phase of the cell cycle compared with mock-infected cells. We propose that phosphorylation of Sam68 may be a repressor of BRK-induced cell cycle progression.
Enzyme-linked Immunosorbent Assay (ELISA)-ELISA plates (Costar, Cambridge, MA) were coated with the indicated quantity of peptide in 50 l of 50 mM carbonate buffer, incubated at 37°C for 30 min, and blocked with blocking buffer (1% bovine serum albumin, 5% sucrose in phosphate-buffered saline (PBS)). Primary antibodies were diluted at 1:1000 in dilution buffer (1% bovine serum albumin, 0.5% ovalbumin, 10 mM Tris, pH 7.4, 150 mM NaCl) and added to the corresponding well followed by incubation at 37°C for 30 min. The plate was washed extensively with PBS containing 0.1% Tween. Goat anti-rabbit antibodies covalently coupled to horseradish peroxidase (Cappel Laboratories, Durham, NC) were incubated at 1:1000 in dilution buffer at 37°C for 15 min. The plate was washed and developed using BM Blue POD substrate (Roche Diagnostics) and quantitated using spectrophotometry at 405 nm.
Expression Constructs-Green fluorescent protein (GFP)-Sam68 was constructed previously by subcloning the EcoRI fragment of myc-Sam68 into pEGFP-C1 (35). GFP-Sam68 mutants were generated by PCR using the forward primer 5Ј-TCCTGCTGGAGTTCGTGACC-3Ј and the following reverse primers: GGTGAATTCCTTTTATGCTCC-CTTCACTGGCCTAGC (for ⌬423-443); GGTGAATTCCTTTTAA-TAACGTCCATATGGGTGCTC (for 435Y-F); GGTGAATTCCTT-TTAATAACGTCCAAATGGGTGCTC (for 440Y-F), and GGTGAA-TTCCTTTTAAAAACGTCCATATGGGTGCTC (for 443Y-F). The EcoRI site is underlined. Wild-type BRK and BRK-YF constructs in the vector pRcCMV (33) were generously provided by Dr. Mark Crompton (School of Biological Sciences, Royal Holloway, University of London, London, United Kingdom). To introduce the myc epitope at the N termini, the constructs were digested with SmaI and XbaI to remove the BRK inserts. BRK inserts were then cloned into pcDNA3myc digested with EcoRI (blunt ended with Klenow large fragment) and XbaI. The resulting constructs were sequenced for verification. Myc-tagged BRK adenoviruses and Sam68 adenoviruses were constructed by first PCR amplification of the cDNAs encoding myc-tagged BRK YF (described above) and myc-tagged Sam68 described previously (35) using T7 primer and a reverse primer containing a BamHI overhang (5Ј-GGAA-GATCTACCATGGTGTCTTGGGACA-3Ј). After digestion with BamHI, the fragment was cloned into the BglII site of pADTR5-K7-GFPq (37). These plasmids express myc-BRK and myc-Sam68 under the regulation of a tetracycline-inducible promoter and a second cassette that constitutively expresses GFP. The latter cassette serves as a marker for transduction. Wild-type c-Src was a gift from Stéphane A. Laporte (McGill University, Montréal, Québec, Canada) and has been previously described (38). Recombinant adenoviruses were generated, purified, and titered as described previously (39).
Protein Expression and Immunoprecipitation-The day before transfection, HeLa cells were plated on glass coverslips at a density of 10 5 cells per 22-mm 2 coverslip (Fisher Scientific). The cells were transfected with Lipofectamine Plus (Invitrogen) according to the manufacturer's protocol. For immunoprecipitations, cell lysates were incubated on ice with the primary antibody for 1 h. Then 30 l of 50% protein A-Sepharose slurry was added and incubated at 4°C for 30 min with constant endover-end mixing. The beads were washed twice with lysis buffer and once with PBS. Protein samples were analyzed on SDS-polyacrylamide gels and transferred to nitrocellulose membranes. Immunoblotting was performed using the anti-Sam68 antibodies, anti-phospho-Sam68 antibodies, anti-BRK, and anti-myc (9E10) antibodies. Immunoreactive proteins were visualized using either goat anti-mouse or goat anti-rab-bit antibodies conjugated to horseradish peroxidase (ICN Pharmaceuticals) and the chemiluminescence (ECL) detection kit (DuPont).
Dephosphorylation of BRK-phosphorylated GFP-Sam68 by Proteintyrosine Phosphatase 1B (PTP1B)-Immunoprecipitated phosphorylated GFP-Sam68 on beads was washed twice with lysis buffer and once with PBS, drained, and incubated with 20 l of PTP1B buffer (10 mM HEPES, 0.1 mM EGTA, 0.1 mM EDTA, 1 mM dithiothreitol, 0.5 mg/ml bovine serum albumin, pH 7.5) and 1 g/ml of recombinant PTP1B (R&D Systems) at 37°C for 30 min. The reaction was quenched by the addition of Laemmli sample buffer and subjected to SDS-gel electrophoresis.
Immunofluorescence on Cultured Cells and Tissue Array-HeLa cells were cultured directly on glass coverslips in a 6-well dish. Transfection of HeLa cells for immunofluorescence was achieved using Lipofectamine Plus (Invitrogen) according to the manufacturer's protocol using 2 g of DNA. The cells were fixed with 1% paraformaldehyde in 1ϫ PBS, pH 7.4, for 5 min and permeabilized with 0.5% Triton X-100 in PBS for 5 min at room temperature. If the cells were to be visualized only for GFP, then the coverslips were mounted onto glass slides with glycerol containing 3 mg/ml 4Ј,6-diamidino-2-phenylindole to stain the nuclei. If the cells required antibody staining, the permeabilized cells were first blocked with 10% calf serum (Hyclone) in PBS for 30 min, then incubated with the primary antibodies (1:200) for 1 h in PBS at room temperature. The cells were washed with 0.1% Triton X-100 in PBS and incubated with the appropriate secondary antibodies (1:200) in PBS for 30 min. Goat anti-rabbit coupled to Alexa 488 (Molecular Probes) and goat anti-mouse coupled to Alexa 488 (Molecular Probes) were used as secondary antibodies. The cells were washed again, mounted onto glass slides, and visualized with a confocal microscope (Carl Zeiss, Thornwood, NY). LandMark TM High Density Breast Tissue MicroArray on slides (catalog number 3190) was purchased from Ambion Inc. (Austin, TX). The tissue were permeabilized as described above and stained with the appropriate antibodies.
Cell Cycle Analysis of Astrocytes-Primary cultures of astrocytes were generated from newborn rat brains as described by McCarthy and de Vellis (40). Proliferating astrocytes were infected for 24 h with adenoviruses at a multiplicity of infection of 50 of the indicated adenovirus (AdGFP control, AdSam68, AdBRK) co-expressing the transgene and GFP from two different promoters as described previously (37,41). Astrocytes were either left untreated or pulsed for 14 h with 10 mM BrdUrd (Sigma). The BrdUrd-labeled cells were collected, fixed in 70% ethanol for 20 min, pelleted, denatured with 1 N HCl for 30 min at room temperature, and neutralized with NaBO 4 (0.1 M). The cells were pelleted by centrifugation, resuspended in PBS/bovine serum albumin/ Tween 20, and incubated for 1 h with a monoclonal antibody to BrdUrd (Chemicon). This was followed by incubation with a Cy5-labeled secondary antibody to mouse (1:500, Molecular Probes), resuspended in PBS containing 10 g/ml propidium iodide, and analysis by flow cytometry. The cells were analyzed using a FACScalibur flow cytometer (BD Biosciences). For BrdUrd labeling on astrocytes cultured on coverslips, infected astrocytes were pulsed for 8 h with 10 mM BrdUrd (Sigma) and fixed with 4% paraformaldehyde in phosphate-buffered saline and visu-alized by immunofluorescence using an Alexa 546-conjugated antibody (Molecular Probes, Inc.).

Generation of Sam68
Phospho-NLS-specific Antibodies-The nuclear localization of Sam68 is imposed, in part, by its NLS located at its C terminus between amino acids 420 and 443 (36). To investigate whether any of the tyrosine residues within the Sam68 NLS are tyrosine phosphorylated, we raised phospho-specific antibodies against each tyrosine within this region (Fig. 1A). The peptides that were synthesized contained phosphorylated tyrosine 435 (␣-435pY), 440 (␣-440pY), or 443 (␣-443pY) (Fig. 1A). The peptides were coupled to keyhole limpet hemocyanin and used to immunize rabbits to generate polyclonal antibodies. The serum was applied to a column containing the non-phosphorylated peptide coupled to beads and the flow-through subsequently affinity purified with the phosphorylated immunogenic peptide. By ELISA we determined that all 3 antibodies recognized the corresponding phosphorylated immunogenic peptide and not the unphosphorylated backbone peptide (supplemental Fig. S1). These findings demonstrate that we have successfully generated Sam68 phospho-NLS specific antibodies.
BRK-phosphorylated Sam68 Is Recognized by New Anti-phospho-Sam68 Antibodies-Sam68 is the first substrate identified for BRK (3).
To confirm that tyrosine residues are phosphorylated in GFP-Sam68, we subjected immunoprecipitates to dephosphorylation assays with recombinant PTP1B followed by immunoblotting with anti-Sam68 antibodies (Fig. 1D). In the presence of the PTPase, the phosphotyrosine content of phospho-GFP Sam68 was significantly reduced for each of the phospho-Sam68 antibodies (compare, lanes 2 and 4). These results demonstrate that tyrosines 435, 440, and 443 are indeed phosphorylated by BRK and confirm that we have generated phospho-Sam68 antibodies.
Phospho-Sam68 Antibodies Are Site-specific-To demonstrate site specificity of the phospho-Sam68 antibodies, we generated a C-terminal-truncated Sam68 protein that deletes the C-terminal 9 amino acids from the GFP-Sam68 fusion protein (⌬432-443). In addition, each tyrosine within the NLS was separately replaced with a phenylalanine and expressed as GFP fusion proteins. Expression vectors encoding the mutant GFP-Sam68 fusion proteins were cotransfected with the plasmid expressing BRK YF in HeLa cells. The cell lysates were separated by SDS-PAGE and immunoblotted with the indicated phospho-specific antibodies (Fig. 1E). Deletion of the C-terminal 9 amino acids completely abolished the recognition of all 3 Sam68 phospho-specific antibodies (Fig. 1E, lane 2). The GFP-Sam68 Y435F, Y440F, and Y443F proteins were not recognized by their respective antibodies (Fig. 1E,  lanes 3-5). The fact that, for example, ␣-440pY recognized wild-type GFP-Sam68 (lane 1) as well as the Y435F (lane 3) and Y443F (lane 5) mutants demonstrates that the absence of phosphorylation at Tyr 435 and Tyr 443 does not influence the recognition of Tyr 440 by the anti-440pY antibody. These results further show that the phospho-Sam68 antibodies are site-specific.
BRK Regulates Nuclear Distribution of SNBs-As shown in Fig. 1, B and C, BRK phosphorylates all 3 tyrosines within the Sam68 NLS. We therefore reasoned that distribution of SNBs, shown previously to be highly prevalent in cancer cells (35), may be regulated by phosphorylation. As such, the GFP-Sam68 chimeric protein was expressed either alone or in combination with vectors encoding constitutively active BRK YF (Fig. 2). GFP-Sam68 alone localized in the nucleoplasm and within 2-3 SNBs as described previously (35). These SNBs were not stained by phospho-Sam68 antibodies ( Fig. 2A, top panels). Remarkably, co-transfection of BRK YF caused a dramatic relocalization and multiplication of GFP-Sam68 from ϳ3 to ϳ40 SNBs that we termed the multiple SNBs phenotype (or mSNBs) ( Fig. 2A, middle panel). Each Tyrosine phosphorylation of GFP-Sam68 was detectable with novel Sam68 phosphospecific antibodies ␣-435pY, ␣-440pY, and ␣-443pY as well as the general anti-phosphotyrosine antibody 4G10 in Western blot analysis. Controls depicting the expression of transfected proteins are detected by antibodies against GFP and myc are represented. C, essentially as indicated above in B, except that Sam68 was immunoprecipitated from HeLa cell lysates using anti-Sam68 antibody and each of Sam68 phosphospecific antibodies, ␣-435pY, ␣-440pY, ␣-443pY. The proteins were separated by SDS-PAGE and immunoblotted with ␣-440pY. D, tyrosine dephosphorylation of Sam68 by recombinant PTP1B. HeLa cells transfected with GFP-Sam68 in the absence (lanes 1, 3, and 5) or presence of active myc-BRK YF (lanes 2, 4, and 6) were lysed and immunoprecipitated with ␣-Sam68 antibodies and the immunoprecipitates were split equally for analysis. Lanes 1 and 2, immunoprecipitates were resolved on SDS-PAGE and transferred to nitrocellulose membranes and immunoblotted with ␣-Sam68 antibodies (top panel, lanes 1 and 2) or with ␣-435pY, ␣-440pY, and ␣- 443pY (bottom panel, lanes 1 and 2). Lanes 3 and 4, immunoprecipitates were incubated with 1 g/ml recombinant PTPase (PTP1B), subjected to SDS-PAGE, transferred to a nitrocellulose membrane, and immunoblotted with ␣-Sam68 antibodies or ␣-440pY. Expressed constructs in the total cell lysates were detected with antibodies recognizing the epitope tags. E, phospho-Sam68 antibodies are site-specific. Expression vectors encoding mutant GFP-Sam68 fusion proteins were cotransfected with the plasmid expressing myc-BRK YF in HeLa cells. Total cell lysates were divided equally, separated by SDS-PAGE, and immunoblotted with the indicated phospho-Sam68 antibodies as well as antiphosphotyrosine antibodies, 4G10. Anti-GFP antibodies were used to demonstrate equal loading of the expressed Sam68 mutants and immunoblotting with anti-actin antibodies served as a loading control.
nuclear focus was indeed recognized with the anti-440pY antibodies demonstrating the phosphorylation of Sam68 in mSNBs. As demonstrated recently, activated BRK YF has a predominantly nuclear localization (25), however, cotransfection with GFP-Sam68 caused its redistribution within mSNBs (Fig. 2B).
Tyr 440 Is Essential for Nuclear Localization of Sam68-We have established that BRK phosphorylates Sam68 in the NLS resulting in the induction of mSNBs. Thus, we set out to investigate whether one or all of the NLS tyrosines are essential in localization and generation of mSNBs. The localization of GFP-Sam68 mutants Y435F, Y440F, Y443F, and Sam68 ⌬432-443 was therefore examined in HeLa cells in the presence or absence of BRK (Fig. 3A). Sam68 ⌬432-443 was expressed in the cytoplasm and also in the perinuclear compartment, as expected with the removal of the NLS (Fig. 3A). Substitution of Y440F caused the mutant GFP-Sam68 Y440F to localize within the cytoplasm, whereas the substitution of Sam68 Y435F and Y443F had no effect on the nuclear localization of Sam68 (Fig. 3A). These data identify Tyr 440 , and neither Tyr 435 nor Tyr 443 , as critical residues for the nuclear localization of Sam68. BRK, however, appears to alter the cytoplasmic pattern of GFP-Sam68 Y440F and Sam68 ⌬432-443 causing the appearance of multiply cytoplasmic foci (Fig. 3A, right panel). Moreover, the expression of BRK YF caused the relocalization of Y435F and Y443F GFP-Sam68 to multiple SNBs, further demonstrating that GFP-Sam68 harboring amino acid substitutions of Tyr 435 and Tyr 443 behave like wild-type GFP-Sam68. To rule out the possibility that tyrosine-phosphorylated Sam68 form foci, we co-expressed GFP-Sam68 with active c-Src and we did not observe a redistribution of SNBs, suggesting that BRK has a unique ability to redistribute Sam68 (Fig. 3B).
Sam68 Is Phosphorylated in Breast Cancer Cell Lines-Thus far we have performed experiments with active BRK YF and overexpression. We wished to reproduce our findings with cell lines and tumor tissues endogenously expressing BRK and Sam68. Phosphorylation of endogenous Sam68 in breast cancer cell lines Hs578T, MDA-435, and MDA-468 has been recently reported (19). As the highest level of Sam68 phosphorylation was observed in MDA-468, we therefore surveyed this cell line and breast tumor cell line BT20, as well as the invasive breast tumor cell line MDA-231, for Sam68 phosphorylation using the ␣-440 antibodies. MDA-231 and BT20 have been previously reported to express moderate and high levels of BRK expression, respectively (31). Cell extracts were prepared, separated by SDS-PAGE, and the phosphorylation of Sam68 Tyr 440 was visualized by immunoblotting. Comparable phosphorylation of Sam68 was detected in breast cancer cells MDA-231, MDA-468, and BT-20 but not in HeLa cells derived from human epithelial cervical carcinoma (Fig. 4A, top panel). Each cell lysate contained equivalent levels of Sam68, as detected with anti-Sam68 antibodies. Anti-BRK antibodies confirmed the presence of BRK in the breast tumor cell lines and weakly expressed in HeLa cells (Fig. 4A, middle  panel). To determine whether all tyrosines in the NLS of Sam68 are phosphorylated, cell lysates from MDA-MB-468 breast cancer cells were immunoprecipitated with control IgG or with the anti-Sam68 antibody. The immunoprecipitated proteins were separated by SDS-PAGE and immunoblotted with anti-435pY, anti-440pY, and anti-443pY antibodies. Immunoprecipitated Sam68 was easily detected by all phosphoantibodies (Fig. 4B). These data concur with the overexpression experiments in HeLa cells that Sam68 is indeed tyrosine phosphorylated by BRK on tyrosines 435, 440, and 443.  We next stained human breast tumor tissue arrays (Ambion) with anti-BRK, anti-440pY, and anti-Sam68 antibodies (TABLE ONE). The anti-BRK staining detected the presence of BRK and the anti-440pY detected phosphorylated Sam68 at Tyr 440 . The anti-Sam68 antibodies stained all the cells and tissues as expected for a ubiquitously expressed protein (supplemental Fig. S2). We focused our results on the infiltrating ductal carcinomas (IDCs) because the array contained 175 different samples of this particular tumor type. 20/175 (11.4%) IDCs stained positively with anti-440pY antibodies, whereas 29/175 (16.6%) IDCs stained positive for BRK expression and 8 of the BRK-positive IDCs contained phosphorylated Tyr 440 . The predicted coincidence of BRK and phospho-Sam68 in the represented sample set is 3.3/175 (1.9%) and our observed overlap of 8/175 (4.6%) is therefore statistically relevant. Thus, the expression of BRK may be, in part, responsible for Sam68 phosphorylation in IDCs.
BRK Is Required for the Tyrosine Phosphorylation of Sam68 in Response to EGF Treatment-Because BRK has been shown to potentiate the mitogenic signals of EGF (33), we investigated first whether endogenous Sam68 was phosphorylated in response to EGF treatment. We stimulated serum-starved MDA-231 cells with EGF (100 ng/ml) and observed a time-dependent activation of Sam68 phosphorylation (Fig. 5A). The level of Sam68 Tyr 440 phosphorylation peaked at ϳ30 min stimulation compared with ϳ15 min for EGF receptor (pp180) phosphorylation. These data indicate the phosphorylation of Sam68 is enhanced in EGF signaling in MDA-231 cells. We further showed that    Sam68 phosphorylated on Tyr 440 relocalized to perinuclear structures in response to EGF treatment (Fig. 5B). To determine whether or not BRK played a role in EGF-induced phosphorylation of Sam68, we depleted BRK from MDA-231 cells by treating cells with BRK or control siRNA (Fig. 6A). We achieved ϳ60% knockdown of BRK expression in MDA-231 using the BRK-specific siRNA. Treatment of these cells with EGF diminished the EGF-induced phosphorylation of Sam68 by ϳ50% (Fig. 6B). These data demonstrate that Sam68 is a physiological substrate of the EGF receptor signaling pathway that requires BRK in the MDA-231 breast cancer cell line.

Sam68 Suppresses BRK-induced Cell Proliferation in Primary
Astrocytes-The expression of BRK is known to be up-regulated in human breast tumors suggesting that BRK induces cell cycle progression (31). To investigate the interplay between BRK and Sam68 in cell proliferation, we generated adenoviruses expressing each protein. In addition, the adenoviruses also express GFP under a separate promoter and GFP serves as a marker of infection. Primary rat astrocytes are devoid of BRK and express low levels of Sam68 (42) and are suitable primary cells to examine the role of Sam68 and BRK in cell proliferation. Astrocytes were infected for 24 h with control (AdGFP), AdSam68, AdBRK, or AdSam68/BRK adenoviruses and cell proliferation was examined by BrdUrd incorporation. The BrdUrd incorporation was visualized by immunofluorescence staining (Fig. 7A) and expressed as the mean Ϯ S.D. (Fig. 7B, n Ͼ 400 cells). Approximately 95% of the astrocytes were transduced with the adenoviruses (Fig. 7A). Fewer astrocytes transduced with AdSam68-incorporated BrdUrd (ϳ10%) compared with the control AdGFP (ϳ30%) (Fig. 7B). In contrast, the infection of astrocytes with AdBRK increased the number of cells incorporating BrdUrd from ϳ30% with AdGFP to ϳ60%. However, co-infection of AdSam68 and AdBRK also resulted in a reduction of BrdUrd incorporation from ϳ60 to ϳ25% (Fig. 7B). To further define what stage of the cell cycle accounts for the observed defects, we carried out flow cytometry analysis using propidium iodide and BrdUrd labeling (Fig.  7C). In control infected cultures, ϳ19.4% of the cells were in S phase, as compared with ϳ25.2% for cells infected with AdBRK and ϳ13.4% in the case of AdSam68 (Fig. 7C). However, co-infection with AdSam68 and AdBRK resulted in few cells in S phase of the cell cycle, as only 9.3% of cells were observed in the S phase. These findings demonstrate that the Sam68 cell cycle inhibition overcomes the BRK-induced cell cycle progression.

DISCUSSION
In the present study we report the generation of new phospho-specific antibodies against three phosphorylated tyrosines in the NLS of FIGURE 7. Phosphorylated Sam68 inhibits BRK-increased cell cycle progression in astrocytes. A, proliferating astrocytes were infected for 24 h with AdBRK, AdSam68, and AdSam68/AdBRK adenoviruses and incubated with BrdUrd for 8 h. Incorporation of BrdUrd was detected with anti-BrdUrd followed by staining with Alexa 546-conjugated goat anti-mouse antibody (red). B, the percentage of GFP-positive cells that incorporated BrdUrd was quantified from n Ͼ 400 cells and is expressed as mean Ϯ S.D. C, cell cycle analysis was performed using flow cytometry. A total of 10,000 cells were selected in each set of experiments. The DNA of the AdGFP-, AdSam68-, AdBRK-, and AdSam68/AdBRK-infected cells was stained with anti-BrdUrd antibodies followed by a Cy5-labeled anti-mouse secondary antibody (x axis) and propidium iodide (PI) (y axis). Cells in G 0 /G 1 , S phase, and G 2 /M were counted and expressed as a percentage using a cytometer. STAR protein Sam68. We provide evidence that BRK phosphorylates Sam68 on all three tyrosines in the NLS and demonstrate that tyrosine 440 is essential in the nuclear localization of Sam68. Furthermore, we observe that phosphorylation of Sam68 by BRK induces multiple SNBs and also suppresses BRK-induced cell proliferation. We show that EGF stimulation of breast cancer cell line MDA-231 enhanced the tyrosine phosphorylation of Sam68 that was inhibited by BRK siRNA treatment. Finally, we reveal the prevalence of phosphorylated Sam68 in breast cancer cells and infiltrating ductal carcinomas. Our findings position Sam68 downstream of BRK in breast cancer cells.
The NLS of Sam68 contains an RXHPY(Q/G)R motif located at its distal end. This motif, originally identified and mapped in QKI-5, is also conserved in other QKI homologs as well as mammalian Sam68, SLM-1, SLM-2, and non-STAR protein Drosophila HNF-4 homologs, all of which are nuclear (43). The RXHPY(Q/G)R motif carries a conserved tyrosine (Tyr 440 in Sam68). Indeed tyrosine 440 was critical for the nuclear localization of Sam68 as substituting this tyrosine with phenylalanine blocked nuclear entry Tyr 440 (Fig. 3). Consistent with tyrosine phosphorylation altering the localization of Sam68 we noted that the induction of phosphorylation Tyr 440 in response to EGF treatment led to relocalization of Sam68. The localization of Sam68 is known to be regulated by virus infection (44,45) and arginine methylation (46), but this is the first report showing that tyrosine phosphorylation regulates its cellular localization.
We first reported the localization of Sam68 within dynamic nuclear structures that we called SNBs (35). These SNBs were observed in immortalized and transformed cells, and the prevalence of SNBs correlated with the tumorigenicity of some cancer cell lines (35). SNBs are preferentially positioned in proximity to the nucleoli and seemed to be the site of RNA metabolism as they were also found to contain heterogeneous nuclear ribonucleoprotein K (47), scaffold attachment factor-B (SAF-B)/HAP (48), and splicing factor YT521 (15). Also, we previously identified BRK in SNBs in the HT29 colon adenocarcinoma cell line (3). In its unphosphorylated state, endogenous and transfected Sam68 show a general diffuse nuclear localization and usually three distinct SNBs (35). Upon BRK phosphorylation, Sam68 relocalizes to multiple SNBs (Fig. 2). The BRK-induced effect was specific to SNBs, as the localization of Cajal bodies and PML bodies were unaltered with BRK overexpression (data not shown). Interestingly, SNBs were barely found in Src 3T3 cells (35), a cell type that contains tyrosine-phosphorylated Sam68 (1, 2), suggesting that the relocalizing effect is BRK-specific.
The tumor suppressor activities of Sam68 have been previously reported (20). Taylor et al. (21) used an inducible expression system to show that overexpression of Sam68 induces a G 1 cell cycle arrest, an effect that was independent of the RNA binding properties of Sam68. Furthermore, random homozygous knock-out of Sam68 in NIH3T3 cells has been associated with neoplastic transformation of mouse fibroblasts and tumorigenesis (20). We confirmed by adenoviral delivery into primary rat astrocytes that Sam68 and BRK have opposing effects on cell cycle progression (Fig. 7). Sam68 inhibited cell cycle progression, whereas BRK promoted cell proliferation of primary rat astrocytes. The intracellular localization of BRK is nuclear in normal prostate epithelial cells and benign prostate hyperplasia, and cytoplasmic in high grade prostate intraepithelial neoplasia (34). Our data support the hypothesis that Sam68 may function as a tumor suppressor, a characteristic shared by other STAR proteins such as SLM-2 (49), GLD-1 (50), and Quaking (51).
The ability of BRK to induce cell proliferation is not surprising. BRK is overexpressed in ϳ60% of breast carcinomas, but not in normal mammary tissue (24). Knockdown of BRK in the breast carcinoma cell line T47D resulted in significant suppression of cell proliferation (32). Moreover, BRK has been shown to associate with and sensitize EGF to proliferative responses and also to potentiate anchorage-independent growth of mammary epithelial cells (33). This BRK-dependent mitogenic sensitization may be achieved through its enhancement of erbB3 phosphorylation and subsequent activation of phosphatidylinositol 3-kinase/Akt pathway (29). These reports show that BRK confers a proliferative advantage on cells. In the present study, we show elevated expression of phosphorylated Sam68 in MDA-231, MDA-468, and BT20 breast cancer cell lines (Figs. 4 -6). These cell lines have been shown to overexpress the EGF receptor (52). Moreover, we show for the first time that Sam68 is tyrosine phosphorylated downstream of the EGF receptor with slower kinetics than the EGF receptor implying that there is an intermediate protein.
Because BRK siRNA treatment attenuated the Sam68 phosphorylation of Tyr 440 , this implies that BRK is the tyrosine kinase that mediates this phosphorylation. Therefore, the tyrosine phosphorylation of Sam68 in breast cancer cell lines may be part of a mechanism of inactivating its anti-proliferative functions, by altering its RNA binding properties and stimulating its adaptor functions (3,7).
In conclusion, this study demonstrates that Sam68 is phosphorylated by BRK on tyrosine residues in the NLS in breast cancer cell lines and primary tumors. We also show that tyrosine 440 is a key modulator of Sam68 localization by BRK and EGF treatment. Our study also identifies phospho-Sam68 in BRK-positive breast cancer cell lines and tissues and reveals that Sam68 attenuates the ability of BRK to stimulate cell proliferation. These findings suggest that overexpression of Sam68 or blocking Sam68 Tyr 440 phosphorylation may be valuable therapeutic targets for suppression of BRK-induced tumor progression.