In Sertoli cells of testis, androgen receptor-regulated gene transcription plays an indispensable role in maintaining spermatogenesis. Androgen receptor activity is modulated by a number of coregulators which are associated with the androgen receptor. Non-POU-domain-containing, octamer binding protein (NONO), a member of the DBHS-containing proteins, complexes with androgen receptor and functions as a coactivator for the receptor. Paraspeckle protein 1 alpha isoform (PSPC1, previously known as PSP1) and Splicing factor, proline- and glutamine-rich (SFPQ, previously known as PSF), other members of the DBHS-containing proteins, are also found in androgen receptor complexes, suggesting that these DBHS-containing proteins may cooperatively regulate androgen receptor-mediated gene transcription. We demonstrated that PSPC1, NONO, and SFPQ are coexpressed in Sertoli cell line TTE3 and interact reciprocally. The effect of the DBHS-containing proteins on the transcriptional activity was assessed using the construct containing androgen-responsive elements followed by a luciferase gene. The results showed that all the DBHS-containing proteins activate androgen receptor-mediated transcription, and PSPC1 is the most effective coactivator among them. Furthermore, we confirmed the presence of PSPC1, NONO, and SFPQ proteins in Sertoli cells of adult mouse testis sections. These observations suggest that PSPC1, NONO, and SFPQ form complexes with each other in Sertoli cells and may regulate androgen receptor-mediated transcriptional activity.
Spermatogenesis is a multistep process leading to the generation of highly specialized spermatozoa. The developmental process begins with spermatogonia that are committed to further differentiation by undergoing two meiotic divisions, resulting in haploid round spermatids. During spermiogenesis, haploid spermatids undergo drastic morphological changes, including formation of the acrosome and the sperm flagellum, decrease of the nuclear size due to the unique DNA packaging, and exclusion of most of the cytoplasm . The complexity of the differentiation process requires a highly specialized program of gene expression of male germ cells [2, 3]. For example, cAMP responsive element modulator (CREM) has been shown to play an important role in germ cell-specific transcription by binding to CRE sequences [4, 5], and Poly (A) polymerase beta (PAPOLB) is known to adjust the timing of haploid-specific translation by controlling the cytoplasmic mRNA polyadenylation .
However, the endogenous gene expression program of male germ cells is not sufficient for spermatogenesis, and support from nearby Sertoli cells is indispensable. Throughout spermatogenesis, Sertoli cells interact directly with germ cells within the seminiferous tubules. Sertoli cells regulate highly organized and precisely synchronized germ cell development by nourishing germ cells via their secretion products [7–9]. These controls by Sertoli cells are also regulated by external stimuli. Androgen and androgen receptor (AR)-mediated gene transcription are important for the Sertoli cell functions [2, 10]. Mice lacking AR in Sertoli cells show spermatogenic arrest, which results in azoospermia and infertility [11, 12]. Thus, AR-mediated transcription in Sertoli cells plays an indispensable role in spermatogenesis.
AR belongs to the nuclear receptor superfamily that includes receptors for thyroid hormone, retinoic acid, estrogen, progesterone, glucocorticoid, and other hormones . AR forms a homodimer and binds to androgen-responsive elements in promoters/enhancers of AR-driven genes. AR is composed of N-terminal transactivation domain (NTD), DNA-binding domain (DBD), and ligand-binding domain (LBD) at the C-terminus. NTD possesses an activation function domain 1 (AF-1) and is involved in making contact with the general transcriptional machinery [14–16]. The transcriptional activity of AR is modulated by coregulators, which include coactivators that enhance AR transactivation and corepressors that suppress AR transactivation [17–19].
Non-POU-domain-containing, octamer binding protein (NONO) is known as one of the coregulators of AR. NONO interacts directly with the AR AF-1 domain and acts as a coactivator [20, 21]. NONO contains a DBHS (Drosophila behavior, human splicing) domain characterized by two tandem RNA recognition motifs (RRMs) and a helix-turn-helix (HTH) DNA binding domain. In mammals, two other DBHS-containing proteins, Paraspeckle protein 1 (PSPC1, previously known as PSP1) and Splicing factor, proline- and glutamine-rich (SFPQ, previously known as PSF), have been reported [22, 23]. PSPC1 has two isoforms, alpha (referred to as PSPC1 in this paper) and beta, produced by alternative splicing. It has been reported that DBHS-containing proteins regulate nuclear receptors, such as progesterone receptor  and thyroid hormone receptor , and also participate in mRNA regulation in the nucleus, including splicing [26, 27], 3′-end cleavage [27, 28], and nuclear retention of edited RNA . Since DBHS-containing proteins are identified in RNA-transporting granules , it is likely that these proteins also participate in RNA metabolism in the cytoplasm. In addition, DBHS-containing proteins are reported to be involved in activation of DNA topoisomerase I  and DNA double- strand break rejoining . Therefore, these proteins are regarded as multifunctional proteins involved in various aspects of gene expression. In many cases, NONO and SFPQ are copurified [20, 27–40], suggesting that these proteins may control gene expression as a complex. We have previously confirmed the direct interaction between PSPC1 and NONO, and SFPQ by coimmunoprecipitation experiments and yeast two-hybrid assays , indicating that the DBHS-containing proteins interact reciprocally.
We have sought to define expression and function of DBHS-containing proteins in testis, and to elucidate the biological significance of these proteins in spermatogenesis. In this paper, we have shown the expression of the DBHS-containing proteins and reciprocal complex formations in the Sertoli cell line TTE3. These proteins enhanced AR-mediated transactivation, and we confirmed the expression of the DBHS-containing proteins in the Sertoli cells of adult mouse testis. These observations suggest that the DBHS-containing proteins may be involved in spermatogenesis by regulating AR-mediated transcription in the Sertoli cells.
Materials and Methods
Plasmids, Cells, and Mice
The expression plasmids for Myc-tagged DBHS-containing proteins (pMyc-CMV-2-Pspc1, pMyc-CMV-2-Nono, and pMyc-CMV-2-Sfpq) were prepared by subcloning mouse Pspc1alpha, mouse Nono, and mouse Sfpq cDNAs (GenBank accession numbers: NM_025682, NM_023144 and NM_023603) into the pMyc-CMV-2 vector (Clontech). A PSPC1 RRM mutant (F118A, F120A, K197A, F199A), which did not bind to RNA, was generated as described previously  and cloned into pMyc-CMV-2. By using the same method, the expression plasmids for a NONO RRM mutant (pMyc-CMV-2-Nono RRM mutant, F113A, F115A, K192A, I194A) and a SFPQ RRM mutant (pMyc-CMV-2-Sfpq RRM mutant, F326A, F328A, K405A, I407A) were also generated. The expression vector for androgen receptor pSG5-hAR and p(ARE)2-luc plasmid containing two consensus androgen-responsive elements were described earlier .
COS-1 cells were cultured in Dulbecco modified Eagle medium (DMEM)(Sigma) supplemented with 10% Donor Calf Serum (DCS) (Thermotrace) and maintained at 37°C in an atmosphere of 5% CO2. TTE3 cells were cultured in DMEM containing 10% DCS on collagen type I pre-coated dishes (Celltight C-1, Sumitomo Bakelite) at the permissive temperature of 33°C or nonpermissive temperature of 37°C in an atmosphere of 5% CO2.
Nine-week-old male BALB/cAJcl mice were purchased from CLEA Japan. Animal experiments were conducted in accordance with the National Institutes of Health standards established in the Guidelines for the Care and Use of Experimental Animals.
Mouse Monoclonal Antibody Production
Synthetic peptides corresponding to amino acids 505–523 in mouse PSPC1 (CFGRGSQGGNFEGPNKRRRY), 453–464 in mouse NONO (CPPAFNRPAPGAE), and 681–699 in mouse SFPQ (CAGYGRGREEYEGPNKKPRF) were purchased from BIO SYNTHESIS. These peptides, conjugated with KLH (Pierce), were injected twice at a 1-wk interval into BALB/cAJcl mice. Three days after the second boost, the lymph node cells were fused with the myeloma line P3U1. The culture media were screened by ELISA and following immunoblotting. The cells from the positive wells were cloned by the standard limiting dilution technique. Anti-PSPC1 (clone 1L4), anti-NONO (clone NC5), and anti-SFPQ (clone FC23) mouse monoclonal antibodies were used in this study.
Fifteen μg of TTE3 cell lysates prepared as described previously  were analyzed on 8% SDS-polyacrylamide gels and blotted onto Immobilon-P membrane (Millipore). After saturation with 5% skim milk (Difco) in Tris-buffered saline (20 mM Tris-HCl pH 7.6, 137 mM NaCl) containing 0.1% Tween-20 for 1 h at room temperature, membranes were incubated with a 1:40 dilution of antibodies to either PSPC1, NONO, or SFPQ followed by an incubation with a 1:10000 dilution of HRP-conjugated goat anti-mouse IgG (ZYMED). Enzymatic activities were detected by ECL substrate (Amersham Biosciences).
TTE3 cells were fixed in 4% paraformaldehyde in PBS for 20 min and for an additional 30 min in methanol. After washing with PBS, coverslips were blocked with 10% goat serum. Cells were incubated with a 1:10 dilution of anti-PSPC1, anti-NONO, or anti-SFPQ antibody respectively for 1.5 h at room temperature. After washing three times with PBS, the cells were incubated with a 1:100 dilution of Alexa Fluor 546 goat anti-mouse IgG (Invitrogen), for 30 min. The expression of proteins was visualized by fluorescence microscopy (Olympus).
For immunoprecipitation analysis, TTE3 cells (1.8 × 107 cells) were harvested and lysed in 4.8 ml of lysis buffer (50 mM Tris-HCl pH 7.4, 150 mM NaCl, 5 mM EDTA, 50 mM NaF, 0.5% NP-40, 1 mM Na3VO4, 1 mM PMSF, 1 μg/ml aprotinin, 1 μg/ml pepstatin, 1 μg/ml leupeptin, 1 mM DTT) with or without 10 μg/ml RNase A and incubated at 4°C for 30 min followed by centrifugation at 15,000 rpm for 15 min. The resulting supernatants were decanted into fresh tubes and preabsorbed with Protein-A agarose beads (SIGMA) for 30 min followed by incubation with anti-PSPC1, anti-NONO, anti-SFPQ, or anti-Myc (9E10) antibody coupled to Protein A-agarose beads for 2 h. The immunoprecipitates were washed three times with lysis buffer. Immunoprecipitates were immunoblotted with each antibody.
COS-1 cells were transiently transfected by the calcium phosphate precipitate method with 0.2 μg of p(ARE)2-luc plasmid, 0.1 μg of pSG5-hAR, 0.3 μg of PSPC1, NONO, or SFPQ expression vector. The total amount of vector added to each well was adjusted by adding an empty vector pMyc-CMV-2. Six hours after transfection, the media were replaced with fresh media containing 0.2% DCS. Dihydrotestosterone (DHT, 10−9 M) ligand was added and cells were incubated for an additional 12 h. Luciferase activities were determined as described previously  using a TD-20/20 luminometer (Turner Designs). Each plasmid was assayed in triplicate at least three different times. pRL-tk (Promega) was cotransfected to normalize transfection efficiencies. All the data were analyzed by Student t-test using Microsoft Excel. A P value of less than 0.05 was considered to be statistically significant.
COS-1 cells were transfected and treated with DHT as done in the luciferase assay with the exception that cells were cultured in 90 mm dishes. Cells were harvested, lysed in IP buffer (50 mM Tris-HCl pH 7.5, 200 mM NaCl, 5 mM EDTA, 1% Triton-X100, 0.5% NP-40, 250 mM sucrose, 1 mM PMSF, 1 μg/ml aprotinin, 1 μg/ml pepstatin, 1 μg/ml leupeptin, 50 mM NaF, 1 mM Na3VO4, 1 mM DTT, 10 U/ml RNase inhibitor (TOYOBO), and incubated at 4°C for 20 min followed by centrifugation at 15,000 rpm for 15 min. The supernatants were decanted into fresh tubes and preabsorbed with Protein G-Sepharose 4 Fast Flow (Amersham Biosciences) for 30 min followed by incubation with anti-Myc monoclonal antibody (9E10) or anti-HA monoclonal antibody (12CA5) coupled to Protein G-Sepharose 4 Fast Flow for 1 h. After washing the immunoprecipitates five times with IP buffer, RNA was purified with an RNeasy Mini Kit (QIAGEN). Eluted RNA was reverse-transcribed using a Sensiscript RT Kit (QIAGEN) with the luciferase rv primer, (5′-CGAGTGTAGTAAACATTCCAAAACCGTGATGG-3′), and amplified using SP-Taq DNA polymerase (Hokkaido System Science) with the luciferase fw primer, (5′-CTAAAACGGATTACCAGGGATTTCAGTCGATG-3′), and the luciferase rv primer.
Testes were removed from BALB/cAJcl mice deeply anesthetized with Ketalar 50 (Sankyo) and fixed overnight in Bouin fixative. The testes were embedded in paraffin and cut into 8-μm sections. After deparaffinization and rehydration by xylene and serial dilutions of aqueous ethanol, the slides were immersed in sodium citrate buffer pH 6.0 and heated for 10 min at 121°C for antigen retrieval. The slides were washed in PBS and permeabilized with 0.2% Triton X-100 in PBS for 30 min at room temperature. After blocking with 10% goat serum and 1% BSA, cells were incubated with a 1:10 dilution of anti-PSPC1, anti-NONO, or anti-SFPQ antibody or a 1:300 dilution of anti-WT1 rabbit polyclonal antibody (Santa Cruz Biotechnology), in 2% goat serum and 1% BSA in PBS overnight at 4°C. Subsequently, slides were washed, incubated with a 1:100 dilution of Alexa Fluor 488 F(ab′)2 fragment of goat anti-mouse IgG (H+L) (Invitrogen) and goat anti-rabbit IgG (H+L)-TRITC (Zymed) in 10% goat serum and 1% BSA in PBS for 2 h at room temperature, and counterstained with DAPI (4′,6-diamidino-2-phenylindole, dihydrochloride).
Monoclonal Antibody Production
Previously we reported that rabbit polyclonal antibody raised against bacterially expressed recombinant PSPC1 recognizes both of PSPC1alpha and its splicing variant PSPC1beta, and kidney expresses PSPC1beta exclusively . However, mass spectrometric analysis showed that the immunoprecipitates from kidney actually contained one isoform of SEPT4 (previously known as M-Septin) , but not PSPC1beta (data not shown). These two proteins have a similar molecular mass (PSPC1beta; 45 kDa, SEPT4; 44 kDa) and share antigenic determinant EELRRXQE in PSPC1 (369–376) and SEPT4 (359–366). To avoid cross-reaction, we designed a new synthetic peptide specific to PSPC1, immunized BALB/cAJcl with the peptide, and developed a mouse anti-PSPC1 monoclonal antibody. On Western blotting, this antibody recognized a single band of 59-kDa protein in testis and kidney (seehttp://www.biolreprod.org). Similarly, we developed anti-NONO and anti-SFPQ mouse monoclonal antibodies. We performed a peptide competition study to verify the specificities, and the results indicated that these antibodies specifically recognized PSPC1, NONO, or SFPQ and had no cross-reactivity with other DBHS-containing proteins (see http://www.biolreprod.org). Moreover we carried out mass spectrometric analysis of the peptides immunoprecipitated from mouse testis extracts by the antibodies. MALDI profiles revealed that anti-PSPC1 antibody immunoprecipitated PSPC1 protein, anti-NONO antibody immunoprecipitated NONO protein, and anti-SFPQ antibody immunoprecipitated SFPQ protein (see http://www.biolreprod.org). These data confirmed the antigenic specificities of monoclonal antibodies, and these antibodies were used for the following analysis.
DBHS-Containing Proteins Are Expressed in Sertoli Cell Line TTE3
NONO interacts with AR and enhances transcriptional activity, and PSPC1 and SFPQ, other members of the DBHS-containing proteins, coprecipitate with AR . These data suggest that PSPC1 and SFPQ may also modulate AR activity. Androgen and AR-dependent gene transcription in Sertoli cells are essential for maintaining normal spermatogenesis. We investigated whether DBHS-containing proteins are expressed in Sertoli cell line TTE3, which is a conditionally immortalized testicular Sertoli cell line from transgenic mice bearing the temperature-sensitive simian virus 40 large T antigen gene . Whole-cell extracts from TTE3 cells were separated by SDS-PAGE, transferred to polyvinylidene fluoride membrane, and probed with anti-PSPC1, anti-NONO, or anti-SFPQ antibody. PSPC1, NONO and SFPQ were expressed abundantly in TTE3 cells (Fig. 1). Next, we examined the intracellular distributions of DBHS-containing proteins in TTE3 cells. TTE3 cells were cultured on coverslips, fixed, and immunostained with each antibody. Strong expressions of PSPC1, NONO, and SFPQ were detected in the nucleus while faint signals for PSPC1 and NONO and an intensive signal for SFPQ were also observed in the cytoplasm (Fig. 2).
Endogenous DBHS-Containing Proteins in TTE3 Cells Form Complexes with Each Other
Complex formation among DBHS-containing proteins is a controversial issue. Fox et al. reported that PSPC1-NONO and NONO-SFPQ complexes were observed in HeLa cells, but PSPC1-SFPQ complex was not . In our previous paper, all reciprocal interactions were shown by coimmunoprecipitation assays of overexpressed DBHS-containing proteins, and yeast two-hybrid assays . Therefore, we investigated complex formation among endogenous DBHS-containing proteins in TTE3 cells. DBHS-containing proteins were immunoprecipitated with anti-PSPC1, anti-NONO, or anti-SFPQ antibody from TTE3 cells, and the immunoprecipitants were subjected to Western blotting with appropriate antibodies. PSPC1 coimmunoprecipitated with NONO and SFPQ, and NONO and SFPQ behaved similarly (Fig. 3). Addition of RNase to the cell extracts did not alter the immunoprecipitation, indicating that the interaction between DBHS-containing proteins is a direct protein-protein interaction. DAZAP1, which is abundantly expressed in testis, did not coimmunoprecipitate with DBHS-containing proteins [47, 48], showing that the interactions are significant. Interestingly, the amount of PSPC1 precipitated with anti-PSPC1 antibody was about the same as the amounts of NONO and SFPQ coimmunoprecipitated with PSPC1. Immunoprecipitations by anti-NONO antibody and anti-SFPQ antibody showed similar tendencies.
DBHS-Containing Proteins Regulate AR-Mediated Transcription
To investigate whether DBHS-containing proteins affect AR-mediated transcription, we planned to perform a luciferase assay with TTE3 cells. Although we tried various transfection reagents and transfection protocols repeatedly, transfection efficiency was extremely low. Therefore the luciferase assay with COS-1 cells using a reporter plasmid containing androgen-responsive elements was performed. COS-1 cells were transfected with expression plasmids for AR and DBHS-containing proteins and reporter plasmids, and treated with DHT. NONO enhanced the transactivation function of AR as previously reported [20, 21], and PSPC1 and SFPQ also enhanced the AR function. PSPC1 showed the highest transactivation of the reporter gene while NONO and SFPQ showed equally lower transactivation (Fig. 4a). DBHS-containing proteins are RNA binding proteins that share RNA recognition motifs (RRMs). We tested whether DBHS-containing proteins bind to luciferase mRNA. COS-1 cells were transfected, treated with DHT as done in the luciferase assay, and harvested. Myc-PSPC1, Myc-NONO, or Myc-SFPQ was immunoprecipitated from cell lysates with anti-Myc antibody. RT-PCR was performed for the immunoprecipitant as template with luciferase-specific primers, and weak binding of DBHS-containing proteins to luciferase mRNA was observed (Fig. 4b, lane 4). Next, to address the question whether the enhancement of luciferase activity is influenced by post-transcriptional regulation, the RRM mutants, in which the RNA binding activity was abolished without disrupting the overall structure of the RRM domain, were constructed according to the procedure of Fox et al. , and luciferase assays were performed. These mutants did not show RNA binding activity (Fig. 4b, lane 2) but did show activated AR-mediated transcription as well as the wild type DBHS-containing proteins in the luciferase assay (Fig. 4a). These data demonstrate that the enhancement of luciferase activity is mediated by transcriptional regulation, not by post-transcriptional regulation.
DBHS-Containing Proteins Are Expressed in Sertoli Cells of Testis
DBHS-containing proteins were expressed in Sertoli cell line TTE3, along with activated androgen receptor-mediated transcription. We next examined the expression of DBHS-containing proteins in Sertoli cells of testis. Adult mouse testis sections were immunostained with anti-PSPC1, anti-NONO, or anti-SFPQ antibody. Anti-WT1 antibody was used as a marker to identify the Sertoli cells . The expression of each DBHS-containing protein in seminiferous tubules displayed a distinct profile. The germ cells expressed PSPC1 and SFPQ, but not NONO (Fig. 5b, 5l and 5g, respectively). The signals of all DBHS-containing proteins were detected in the cells adjacent to the basal membrane of the seminiferous tubule. These cells also expressed WT1 (Fig. 5a, 5f and 5k), demonstrating that all the DBHS-containing proteins are expressed in the Sertoli cells.
In this study, we showed that all of the three DBHS-containing proteins are expressed in mouse Sertoli cell line TTE3, and interact reciprocally. These proteins enhanced AR-mediated transactivation. Expression of the DBHS-containing proteins in Sertoli cells of adult mouse testis suggests that the DBHS-containing proteins may play roles in spermatogenesis by regulating AR-mediated transcription in Sertoli cells.
First, we showed the abundant and equivalent expressions of all the DBHS-containing proteins in Sertoli cell line TTE3. However, in germ cells, PSPC1 and SFPQ were expressed abundantly, but NONO was not detected. In HeLa cells, Fox et al. reported that NONO is expressed at higher levels than that for PSPC1 . Although expression of SFPQ is confirmed in various tissues and cell lines [50, 51], expression levels of NONO are different among tissues and cell lines . PSPC1 is expressed abundantly in mouse testis . These data indicate that at least expression levels of PSPC1 and NONO are different depending on tissues and cell types.
The transcriptional activity of AR is modulated by various coregulators, which include coactivators and corepressors. Luciferase assay using a reporter plasmid containing androgen-responsive elements indicated that all the DBHS-containing proteins enhance the transcription mediated by AR. PSPC1 showed the highest transactivation of the reporter gene, and NONO and SFPQ showed equal transactivation. These data suggest that DBHS-containing proteins are coactivators of AR. The coregulators can be divided into two major types. Type I coregulators, such as HMGB1 and HMGB2  and CMTM2A , function primarily with the nuclear receptor at the target gene promoter to facilitate DNA occupancy, chromatin remodeling, or the recruitment of general transcription factors associated with the RNA polymerase II holocomplex. Type II coregulators, such as FLNC  and PAK6 , contribute to AR protein stability in the presence of ligands or influence the subcellular distribution of AR [17–19]. Although the molecular mechanism of AR transactivation by the DBHS-containing proteins remains elusive, the DBHS-containing proteins are known to have a DNA binding domain [23, 41], and we confirmed the expression of these proteins in the nucleus of the TTE3 cell line. It was previously reported that NONO enhances the association of transcription factors to their target DNA , suggesting that the DBHS-containing proteins enhance the association of AR to their targets, like HMGB1 and HMGB2 enhance the binding of AR, progesterone receptor, and glucocorticoid receptors to their target DNA and enhance the transactivation . Alternatively, DBHS-containing proteins might activate transactivation of AR by stimulating DNA topoisomerase I to relieve torsional strain, and this is supported by the observation that NONO and SFPQ interact with DNA topoisomerase I and activate its enzymatic activity . Thus we speculate that DBHS-containing proteins belong to the Type I coregulators that functions primarily at the target gene promoter site.
DBHS-containing proteins are known not only to regulate transcription but also to bind directly to RNA for post-transcriptional regulation. DBHS-containing proteins are also known to be involved in splicing, polyadenylation, nuclear retention of edited RNA, and transport of mRNA [23, 26–30]. In TTE3 cells, SFPQ was expressed not only in the nucleus but also in the cytoplasm, and weak expression of PSPC1 and NONO was observed in the cytoplasm. We also found weak but significant expression of PSPC1 in the cytoplasm of germ cells (data not shown), suggesting that DBHS-containing proteins function also in the cytoplasm. From these findings, we speculate that DBHS-containing proteins may regulate not only transcription but also RNA metabolism such as splicing, mRNA export, and RNA nuclear retention. The DBHS-containing proteins may maintain spermatogenesis by activating AR-mediated transcription and regulating the transcribed mRNA both in the nucleus and cytoplasm of Sertoli cells.
Fox et al. reported that PSPC1 interacts with NONO, and not with SFPQ in HeLa cells . However, we previously showed the interaction between overexpressed PSPC1 and NONO, and also between PSPC1 and SFPQ . In this study, we investigated the complex formation among the endogenous DBHS-containing proteins in TTE3 cells, and observed all combinations of the interactions among PSPC1, NONO, and SFPQ. It has been reported that all the DBHS-containing proteins are found in the protein complex with the AR AF-1 domain . Moreover, our preliminary proteomic analysis demonstrated that anti-PSPC1 antibody immunoprecipitates NONO and SFPQ together with PSPC1 in testis extract, and yeast two-hybrid assays using PSPC1 as a bait protein showed the interaction of PSPC1 with NONO and SFPQ (data not shown). Therefore PSPC1 actually interacts with both NONO and SFPQ. We assume that the inconsistency between the data by Fox et al. and by us may come from the difference in the interacting partner in the cell line used, not only from the difference in expression levels of the DBHS-containing proteins. If we assume that PSPC1 forms only heterodimers with NONO or SFPQ, PSPC1 should immunoprecipitate most abundantly among these proteins. However, the amount of PSPC1 precipitated with anti-PSPC1 antibody was almost the same as those of NONO and SFPQ coimmunoprecipitated with PSPC1. Immunoprecipitations by anti-NONO antibody and anti-SFPQ antibody also showed similar tendencies. These results raise the possibility that the DBHS-containing proteins may form multimers. It was previously reported that Hrp65 protein, Chironomus tentans homolog of the PSPC1/NONO/SFPQ family, can self-associate, and most of the Hrp65 protein are in complexes that consist of three to six Hrp65 protein molecules , suggesting that PSPC1, NONO, and SFPQ also form large complexes by interacting reciprocally. DBHS-containing proteins are known not only to regulate transcription but also to be involved in multiple regulations including RNA metabolism [23, 26–30] and DNA metabolism [31, 32]. It is likely that the differences in the composition of the DBHS-containing protein complex bring about functional diversity of DBHS-containing proteins.
We confirmed the expressions of all the DBHS-containing proteins in Sertoli cells of adult mouse testis, and these proteins exhibited enhancement of transactivation of AR. PSPC1 and SFPQ were expressed also in germ cells. Because germ cells do not express AR, PSPC1 and SFPQ expressed in germ cells may function in an androgen-independent mechanism. It is known that a number of coactivators function as general coactivators for transcription mediated by nuclear receptors. For example RBM14 is known to activate glucocorticoid receptor-, estrogen receptor-, and thyroid hormone receptor-mediated transcription . It has been reported that SFPQ regulates progesterone receptor, thyroid hormone receptor and retinoic acid receptor, suggesting that SFPQ is a general coregulator of nuclear receptors [24, 25]. In germ cells, several specific nuclear receptors are expressed [60, 61]. PSPC1 and SFPQ may regulate transcription mediated by another nuclear receptor in germ cells. Sertoli cells regulate highly organized and precisely synchronized germ cell development by nourishing the germ cells via their secretion products. Activity of AR-mediated transcription in Sertoli cells is regulated by multiple coregulators. Our present study suggests that the DBHS-containing proteins are coactivators of AR transactivation in Sertoli cells and may be determinants of androgen activity during spermatogenesis. In conclusion, PSPC1, NONO, and SFPQ may support spermatogenesis by regulating androgen receptor-mediated transcription in Sertoli cells.
We thank Dr. Naohito Nozaki for antibody production, and M.S. Atsushi Kawaguchi for his helpful discussion.