Germ Cell Nuclear Factor (GCNF/RTR) Regulates Transcription of Gonadotropin-Regulated Testicular RNA Helicase (GRTH/DDX25) in Testicular Germ Cells—The Androgen Connection

Abstract

Gonadotropin-regulated testicular RNA helicase (GRTH) (GRTH/DDX25), is a testis-specific protein essential for completion of spermatogenesis. Transgenic mice carrying 5′-flanking regions of the GRTH gene/green fluorescence protein (GFP) reporter revealed a region (−6.4/−3.6 kb) which directs its expression in germ cells (GCs) via androgen action. This study identifies a functional cis-binding element on the GRTH gene for GC nuclear factor (GCNF) (GCNF/RTR) required to regulate GRTH gene expression in postmeiotic testis GCs and explore the action of androgen on GCNF and GRTH transcription/expression. GCNF expression decreased in mice testis upon flutamide (androgen receptor antagonist) treatment, indicating the presence of an androgen/GCNF network to direct GRTH expression in GC. Binding studies and chromatin immunoprecipitation demonstrated specific association of GCNF to a consensus half-site (−5270/−5252) of the GRTH gene in both round spermatids and spermatocytes, which was abolished by flutamide treatment in round spermatids. Moreover, flutamide treatment of wild-type mice caused selective reduction of GCNF and GRTH in round spermatids. GCNF knock-down in seminiferous tubules from GRTH-transgenic mice (dark zone, round spermatid rich) caused decreased GFP expression. Exposure of tubules to flutamide caused decrease in GCNF and GFP expression, whereas androgen exposure induced significant increase. Our studies provide evidence for actions of androgen on GCNF cell-specific regulation of GRTH expression in GC. GRTH associates with GCNF mRNA, its absence caused increase on GCNF expression and mRNA stability indicative of a negative autocrine regulation of GCNF by GRTH. These in vivo/in vitro models link androgen actions to GC through GCNF, as regulated transfactor that controls transcription/expression of GRTH.

Gonadotropin-regulated testicular RNA helicase (GRTH) (GRTH/Ddx25) is a testis-specific member of the DEAD-box protein family of RNA helicase present in Leydig cells and germ cells (GCs) (spermatocytes, round and elongated spermatids) (1–3). GRTH is a multifunctional protein and a posttranscriptional regulator of genes essential for completion of spermatogenesis. GRTH null mice are sterile, with azoospermia caused by failure of round spermatids to elongate and complete arrest at step 8 of spermiogenesis (4). It is an integral component of mRNP particles acting as shuttle protein in the transport of specific mRNAs to cytoplasmic sites (chromatoid body of round spermatids) presumably for storage before their translation at specific times during spermatogenesis (4, 5). Phosphorylated GRTH associates with actively translated polyribosomes and regulates translation of specific target genes like H4, HMG2, TP1, Tp2, and protamines 1 and 2 (5, 6). GRTH is the only member of the helicase family known to be hormone regulated by LH through androgen at the transcriptional and translational levels in Leydig cells and GCs of the testis where its expression is both cell and stage specific (6, 7). It displays a novel negative autocrine control of the androgen production in Leydig cells by preventing overstimulation of the gonadotropin-induced androgen pathway through enhanced degradation of StAR protein (8).

GRTH is a TATA-less gene with multiple transcriptional start sites, and GC-rich sequences at the promoter located within −205/+63 bp of the gene. The basal transcriptional activity of GRTH gene is driven by an Sp1/Sp3 functional binding site (9). Our previous studies demonstrated that transgenic mice carrying 5′-flanking sequences of the GRTH gene differentially directs its cell-specific expression both in Leydig cells and GCs (7, 10). The 1085-bp 5′-UTR to the ATG of GRTH gene was found to contain a functional androgen-response element (ARE) to direct GFP expression in Leydig cells (11). Androgen regulates GRTH in Leydig cells through its cognate receptor at a nonconsensus ARE half-site which resides at −827 relative to the GRTH translational start site via short range chromosomal looping between androgen receptor (AR)/ARE2 and the core transcriptional machinery at the promoter (11). We have shown that transgenic mice carrying the full-length 6.4-kb 5′-flanking region of the GRTH gene (−6.4 kb/+63 bp) revealed the expression of reporter gene GFP in both testicular cell compartments and deletion of −3.6 kb/−205-bp sequences distal to the promoter domain in the 6.4-kb 5′-flanking region of GRTH gene directed specific expression of GFP solely to GCs. Further, the reduction in GFP expression in both Leydig and GCs directed by the 5′ 6.4-kb sequences in transgenic mice after flutamide treatment provided intrinsic physiological evidence of direct (in Leydig cells)/indirect (in GCs) action of androgen (10).

We commenced searching for a GC-specific transcriptional factor (TF)(s) that regulates GRTH expression in GCs. Our in silico analysis revealed a putative TF binding site for GC nuclear factor (GCNF)/RTR at -5273/-5264 bp upstream of the GRTH gene. GCNF originally cloned from mouse heart cDNA library (12) is a member of orphan nuclear receptor superfamily essential for embryonic survival, development of anterior/posterior axis and organogenesis (13, 14). GCNF is relevant to female and male reproduction and regulates gametogemesis in developing gonads. In the ovary, it is expressed in developing oocytes and regulates paracrine communication between oocytes and somatic cells (15). In the testis, it is expressed in spermatocytes undergoing meiotic prophase and in round spermatids and down-regulates the transcription of chromatin remodeling genes protamine 1 and protamine 2 (16, 17). In the present study, we investigated a role for GCNF in the regulation of GRTH gene in GCs and the participation of androgens in this regulation. This is the first study to our knowledge describing the impact of androgen on genes essential in spermatogenesis by concerted transcriptional regulation of GCNF and GRTH in GCs.

Materials and Methods

Generation of GRTH transgenic mice used for in vitro studies

The nucleotides covering the transgene of −6400 to −3472/−205 to +63 bp (6.4Kb/del-Tg) tailed with XhoI/SalI were generated by the direct DNA synthesis (GeneScript Corp) and subsequently subcloned into XhoI/SalI site of pEGFP-1. DNA fragment was first isolated by restriction digestion of fusion construct to produce 4.2 kb of XhoI/AfI2 fragment for GRTH 6.4Kb/del-Tg. Transgenic mice were generated by microinjection of purified transgene construct into fertilized eggs in Transgenic Core Facility of National Institute Mental Health (NIH). Mice carrying GRTH-EGFP fusion genes were identified by PCR based screening to produce 420-bp fragment using isolated DNA from mice tails as described previously (7). Four separate mouse lines were created and out of these, 2 lines (GRTH-Tg) with similar high levels of transgene expression were used as founders and maintained in C57/BL6 mouse strain. Animals were housed in pathogen-free, temperature- and light-controlled conditions (22°C), with an alternating light-dark cycle with 14 hours of light and 10 hours of darkness. All studies were approved by National Institute of Child Health and Human Development Animal Care and Use Committee.

Treatment of adult mice

Adult male mice (90 d old) were treated with 0.5 mg of AR antagonist flutamide (2-methyl-N-(4-nitro-3-(trifluoromethyl)-phenyl) propanamide) in 1,2-propanediol. (Sigma-Aldrich). Animals were given 2 sc injections of flutamide with a 12-hour interval for 3 days. This therapy was proven effective to obliterate the action of endogenous circulating testosterone in mice (11). Animals were killed by asphyxiation with CO₂ and decapitated after flutamide or vehicle (controls) treatment. Testes were removed for histological and Western blot analyses.

Preparation of testicular GCs and seminiferous tubule culture from adult mice

Testicular GCs (spermatocytes and round spermatids) from 90-day-old mice were prepared by collagenase/trypsin dispersion and purified by centrifugal elutriation (5). After collagenase dispersion, seminiferous tubules were minced and incubated in medium 199 containing 0.1% BSA, 0.1% trypsin (Sigma) for 15 minutes in a rotary water bath (100 rpm, 35°C). After the addition of 0.02% trypsin inhibitor (Sigma), the sample was filtered through a 300-, 90-, or 40-μm mesh screen and glass wool, and the cells were pelleted and resuspended in elutriation buffer containing 2-μg/mL DNase. The round spermatids were subsequently separated and purified by centrifugal elutriation using Beckman Avanti 21B centrifuge with elutriator rotor model J 5.0. The first 3 fractions were collected with flow rates of 13.5, 31.5, and 41.4 mL/min at 3000 rpm, and 2 additional fractions (fractions 4 and 5) were obtained with flow rates of 23.2 and 40 mL/min at 2000 rpm. Fractions 3 and 5 containing round spermatids and spermatocytes at purity of 84% and 86%, respectively (1), were used for protein analyses. For preparation of seminiferous tubules for in vitro culture experiments, testes from transgenic mice 90 days old (see above) were decapsulated and dispersed with forceps in Petri dished containing medium 199 for 10 minutes. The dispersed seminiferous tubules were segmented. Fractions containing the dark zone (stages VII and VIII of spermatogenesis) under transillumination-assisted microdissection technique (6, 18) were used immediately after dissection and segment isolation for culture in DMEM/F12 containing 5% FBS (see Figure 5, top).

Figure 5.

GCNF regulates GRTH expression in GCs. Top panel (left) region of seminiferous tubule dark zone rich in RS (in red) were used in cultures (right). A, Western blotting showing effect of knock-down of endogenous GCNF protein (upper panel) on GFP and GRTH protein expression (middle panels) in seminiferous tubules (culture prepared from GRTH transgenic mice with GFP as reporter) transfected with 2 different sets of GCNF siRNAs using RNAi technique. Lower panel, β-Actin used as loading control. B, Western blotting showing effect of DHT or flutamide (Flu) or combination of both on the expression of GCNF and GFP in seminiferous tubules. Right side panel, Densitometry analysis of protein bands from 3 independent experiments in each group quantified and normalized to β-actin. Mean ± SE of 3 independent experiments done in triplicates. *, significant difference (P < .05) in GCNF and GFP protein expression levels compared with control.

Western blot analysis

Protein extracts from adult mice testis and transgenic mice were prepared by homogenization in RIPA lysis buffer (Upstate) containing protease inhibitor cocktail (Roche Applied Sciences). The protein lysates were centrifuged at 1500g for 10 minutes. The supernatant was collected, and protein concentration was determined by the Bradford assay (Bio-Rad Laboratories). The protein extracts (50 μg) were separated using 4%–12% Bis-Tris gel (Invitrogen) and transferred onto nitrocellulose membrane. The membrane was blocked with 5% skimmed milk powder in phosphate buffer saline and then incubated either with specific rabbit anti-GFP antibody at 1:1000 dilution (ab13970; Abcam), rabbit anti-GRTH-peptide antiserum at 1:500 (5) or β-actin (sc69879; Santa Cruz Biotechnology, Inc) antibody at 1:1000 dilution. After the incubation of primary antibody and washing steps, the membranes were incubated with the respective secondary antibodies conjugated with HRP at 1:2000 dilution. Immunosignals were detected by a supersignal chemiluminescence system (Pierce).

Immunohistochemistry

Immunocytochemical studies were performed to localize the GFP expression in testicular sections of adult mice. Testes were fixed in 4% paraformaldehyde and embedded in paraffin. Serial sections were incubated overnight with 1:250 dilution of rabbit anti-GFP antibody (ab13970; Abcam) or 1:100 dilution of rabbit anti-AR antibody (sc13062; Santa Cruz Biotechnology, Inc) or rabbit IgG isotype (negative control). Subsequently, these were developed with horseradish peroxidase polymer conjugated to the secondary antibody (Invitrogen). Antigen retrieval was performed using Tris-EDTA buffer (pH 9.0) before the primary antibody incubation step. This was performed in a DAEWOO 800W Microwave at approximate 100% power for 5 minutes, and subsequently, the slides were let stand in the buffer for an additional 15 minutes. The immunocomplex was visualized with diaminobenzidine as chromagen and hydrogen peroxide as substrate to produce brown color. The sections were counterstained with hematoxylin. In parallel studies, periodic acid/Schiff reagent was used for acrosome determination. Image J software (Media Cybernetics, Inc) was used for quantification of immunosignal. The same conditions and exposure times were used on all samples when comparing different slides. All the sections analyzed/quantitated were processed in the same run.

Electrophoretic mobility shift assay

Nuclear extracts were prepared from GCs isolated from mice testis using NE-PER Nuclear and Cytoplasmic extraction kit (Thermo Fisher Scientific) according to the manufacturer's instructions. EMSA was performed with the LightShift chemiluminescent EMSA kit (Thermo Fisher Scientific). We used 5′-end biotin-labeled sense and antisense oligonucleotides containing putative binding site for GCNF (wild type [WT], 5′-TGA AAC TAG ATC TGT AGA CCA CGC TGA CCT TGA ATT CAG AGA TCC ACC TGC CT-3′ and mutant, 5′ TGA AAC TAG ATC TGT AGA CCA CGC CTA CCT GTA ATT CAG AGA TCC ACC TGC CT3′). The complementary biotinylated and unlabeled oligonucleotide pairs were annealed to generate double-stranded biotin-labeled probes. Biotinylated oligonucleotide pairs carrying the mutation in the GCNF putative binding site (5′-TGA CCT TGA-3′) were annealed to generate biotin-labeled mutant probe. The biotin-labeled probes were incubated at room temperature for 30 minutes with 1–2 μg of nuclear extracts in the reaction buffer containing 10mM Tris (pH 7.5), 50mM KCl, 1mM DTT, 10mM MgCl₂, 0.1% NP-40, and 5% glycerol in 20-μL reaction. For competition assay unlabeled probe (200× concentration) was allowed to bind the nuclear extracts before the addition of labeled probes. The samples were subjected to electrophoresis on a 5% TBE gel in 0.5× TBE buffer for 1 hour and transferred onto a nylon membrane in 0.5× TBE buffer for 1 hour. The nylon membrane was UV cross-linked at 1200 mJ/cm² for 1 minute, blocked, and incubated with streptavidin-HRP solution at RT for 15 minutes. The membrane was washed for 3 times for 5 minutes, and immunosignals were detected by a supersignal chemiluminescence system (Pierce).

Chromatin immunoprecipitation (ChIP) assay

ChIP assays were performed using MAGnify ChIP system from Invitrogen following the manufacturer's protocol as previously described (19) The relative binding of GCNF protein to the GRTH promoter was quantitatively analyzed by real-time PCR assay of the precipitated DNA and input DNA using SYBER Green Master Mix in an ABI 7500 sequence detection system. The primers used for amplification of the GRTH gene promoter sequence span the GCNF binding site (forward, 5′-CTC TGT GTA GCC CTG GTT AAC CTG A-3′ and reverse, 5′-GCC TTT AAT CCC AGT TGG GAG CCA G-3′).

RNA interference (RNAi)

Silencers select predesigned and validated siRNA against GCNF and scramble control siRNA were purchased from Ambion (Life Technologies). The seminiferous tubules in the culture were transfected with 25nM siRNA using siPORTNeoFX reagent (Life Technologies) as described in the manufacturer's protocol. After 48 hours of transfection, tubules were harvested for preparation of total protein extracts.

Statistical analysis

The significance of the differences between groups was determined by Tukey's multiple comparison test (one-way ANOVA analysis) using the Prism software program (GraphPad Software, Inc). Data are presented as the mean ± SE of 3 independent experiments.

Results

GCNF expression in different stages of spermatogenic cycle of mice testis

Sections from adult mice testis were immunostained with the GCNF antibody to assess the temporal expression of GCNF during different stages of spermatogenesis and determine its subcellular localization. Our immunocytochemical analysis revealed that GCNF expression was present in GCs in all stages of the spermatogenic cycle of mice testis and was undetectable in Sertoli and Leydig cells. Its expression was specifically localized in the nucleus of spermatocytes and round spermatids (Figure 1A). The expression was maximum in spermatocytes at stages VII–VIII and in spermatids at stages VII–IX (Figure 1, A and B, above), which is coincident with levels of GRTH expression previously reported in mouse (Figure 1B, below) (10).

Figure 1.

A, Immunohistochemical analysis of the GCNF expression in different stages (I–XII) of spermatogenic cycle of mice testis. Shown below for each stage are IgG negative controls. B, upper panel, Quantitative analysis of immunosignals (DAB staining) in spermatocytes and spermatids at different stages of the spermatogenic cycle. Immunosignal intensities in different cell types were quantified as optical density using ImageJ software. Results are mean ± SE of 3 independent experiments performed. Lower panel, Different stages of spermatogenesis in seminiferous tubules of mice testis (50). The vertically boxed areas present the type of GCs at the different stages corresponding to the expression of GCNF protein shown above. The red lines indicate the GRTH protein expression profile revealed by immunostaining of spermatocytes and spermatids from WT mice testis and nonimmune IgG controls (in yellow) from our previous study (10). P, pachytene spermatocyte; D, diplotene; RS, round spermatid; ES, elongated spermatid; SC, Sertoli cell; SM, spermatocytes in the metaphase of meiosis. Scale bar, 50 μm.

Binding of GCNF to its putative DNA sequence in the GRTH promoter

To determine whether GCNF bind to the putative DNA extended half-site consensus element (TCAAGGTCA) present in the GRTH promoter, EMSAs were performed. Gel shift assays revealed a specific band (DNA protein complex) when GCNF probe was added to the reaction mixture containing nuclear extracts prepared from GCs (Figure 2A, lane 2). This complex is completely inhibited by excess of unlabeled probe (as competitor). The probe containing the mutation in the GCNF consensus binding site failed to bind GCNF, resulting in a complete lack of the specific band observed with the WT probe (Figure 2, lanes 3 and 4). The complex was supershift by GCNF antibody. Results from ChIP assay showed recruitment of GCNF to the GRTH promoter both in spermatocytes and round spermatids, indicating that GCNF binds to its putative sequence present in the GRTH gene (Figure 2B).

Figure 2.

A, EMSA showing binding of GCNF protein to the putative binding site of the GRTH promoter. Labeled WT GCNF probe was incubated with nuclear extracts prepared from mice testis (lane 2) or in absence of nuclear extracts (lane 1) in a DNA-binding reaction mixture for 30 minutes. Similarly labeled mutant probe was incubated with nuclear extract (lane 4). The DNA-protein complex in lane 2 is indicated by arrow. For competition experiments, a 200-fold molar excess of unlabeled GCNF WT probe (lane 3) was added in DNA-binding reactions. For supershift assay, rabbit polyclonal anti-GCNF was added into the binding reaction (lane 5). B, ChIP assay showing recruitment of GCNF to the GRTH promoter in spermatocytes (SPs) and round spermatids (RSs) of mice testis. The sequence of GRTH 5′-upstream region (-5264/-5273 bp) containing GCNF putative binding site was analyzed by qPCR after immunoprecipitation (IP) with the GCNF antibody. IgG was the negative control.

Effect of flutamide on GCNF and GRTH expression in GCs

Flutamide (AR antagonist) treatment of adult male mice resulted in significant decrease by 80%–90% in GCNF protein in round spermatids when compared with sham control. Similarly, a marked decrease in GRTH expression was found round spermatids (Figure 3A). However, no significant change was observed in spermatocytes (Figure 3A), where expression levels of GCNF and GRTH were as those observed in controls. Moreover, our ChIP data revealed a significant reduction in the recruitment of GCNF protein to the GRTH promoter in round spermatids of flutamide-treated mice by 80%–90% (Figure 3B). In contrast no significant reduction on the recruitment of GCNF protein to the GRTH promoter was observed in spermatocytes. The decrease in recruitment observed in spermatids but not in spermatocytes is consistent with differential reduction of GCNF expression in the two GC types. These results correlate also with immunohistochemical analysis of mice testis section prepared from flutamide-treated mice where loss of GCNF immunoreactivity was observed in most of the round spermatids (Figure 4).

Figure 3.

Effect of flutamide (Flu) (AR antagonist) treatment on GCNF and GRTH expression in GCs of mice testis. A, Western blotting showing the GCNF and GRTH protein expression bands in spermatocytes and spermatids isolated from WT and Flu-treated mice testis. Actin was used as a loading control. Densitometry analysis of protein bands from 3 independent experiments in each group were quantified and normalized to β-actin. Mean ± SE of 3 independent experiments done in triplicates. *, significant decrease in GCNF (P < .001) and GRTH (P < .05) protein expression in Flu samples compared with WT samples in round spermatids. B, ChIP assay showing the effect of Flu in the recruitment of GCNF to the GRTH promoter in SP and RS of mice. Mean ± SE of 3 independent experiments done in triplicates. *, significant decrease (P < .001) in the recruitment of GCNF to the GRTH promoter in flutamide mice compared with WT mice.

Figure 4.

Effect of flutamide (AR antagonist) on GCNF immunoreactivity in spermatocytes (SPs) and spermatids of mice testis using immunohistochemistry. Upper panel, GCNF immunostaining (dark brown staining as DAB signal) in the nucleus of SPs and round spermatids (RSs) as indicated by arrows in the WT and flutamide mice testicular sections at stage VII of spermatogenesis. Lower panel, Quantitative analysis of GCNF immunosignals in SPs and spermatids different stages of spermatogenic cycle. Immunosignal intensities in different cell types were quantified as optical density using ImageJ software. Results are mean ± SE of 3 independent experiments performed. *, P < .05.

GCNF regulates GRTH expression in GCs

To find whether GCNF plays a role in regulating GRTH expression in GCs, GCNF was knocked down using RNAi interference in cultured seminiferous tubules prepared from GRTH 6.4kb/Del-Tg mice with reporter GFP. This GRTH construct of 5′-upstream region which contains the GCNF functional binding element and sequences of the promoter was previously validated to specifically direct express expression to GCs (10). The tubules fractions dark zones from stages VII–VIII (round spermatid rich) were transfected with scramble siRNA and siRNA against GCNF mRNA (Figure 5). Round spermatids in dark zone were previously shown to contain a high level of GRTH expression in the three cellular compartments, including nucleous, cytoplasm, and chromatoid body (6). The efficiency of siRNA was checked by GCNF protein expression in Western blottings. Tranfection of tubules with the GCNF siRNA produced at least 80% reduction in expression of GCNF protein (Figure 5A). Also, was observed a marked reduction in the expression of GFP and GRTH (endogenous) in tubules transfected with GCNF siRNA compared with scramble siRNA samples further demonstrating a role of GCNF on GRTH transcription and consequent expression in GCs.

Androgen effect on GCNF expression and its transcriptional activity in GCs in vitro

Other studies were conducted to explore the role of androgen in GCNF expression in GCs. Incubation of tubules with dihydroxy testosterone (DHT) show comparable increases in the expression of GCNF and GFP protein (reflective of GRTH transcription/expression), which were negated by preincubation of the tubules with flutamide to levels below control values (Figure 5B). In addition, flutamide treatment of tubules reduced significantly the basal levels of GCNF and GFP, which reflects the action of the inhibitor on the residual effect of the in vivo androgen in the in vitro system. Androgen (DHT) increases GCNF and in turn GFP (resulting from activation of GRTH gene by GCNF), presumably via Sertoli cells within the tubular compartment.

Reciprocal regulation of GRTH and GCNF

We found that GRTH associates with GCNF mRNA as revealed by real-time PCR analysis of immunoprecipitated testicular GRTH complexes in WT mice (Figure 6A). This association was not observed in testis of GRTH KO or in immunoprecipitates with IgG used to validate the findings. A significant increase in half-life of GCNF mRNA was observed in KO mice compared with WT mice (12.62 ± 0.4 h in KO vs 5.04 ± 1.3 h in WT, P < .01) (Figure 6B). Interestingly, a significant increase in GCNF protein expression was observed in round spermatids of GRTH-KO mice compared with WT, whereas no difference was observed in spermatocytes (Figure 6C). Similar results were observed with increase in GCNF protein expression when endogenous GRTH protein was depleted using GRTH siRNA transfection in seminiferous tubules of WT mice (Figure 6D). These results indicate that GRTH, which is transcriptional regulated by GCNF, down-regulates GCNF expression through its negative control of GCNF message stability.

Figure 6.

Reciprocal regulation of GRTH and GCNF in GCs. A, Real-time PCR analysis of the GCNF message associated with immunoprecipitated total testicular GRTH from WT and GRTH knock out mice. *, significant (P < .001) reduction in association of GCNF mRNA with GRTH protein in GRTH-KO mice compared with WT. B, GCNF mRNA stability assay showing GCNF mRNAs in GCs from WT and GRTH-KO mice using real-time PCR. Seminiferous tubule cultures were incubated with 10-μg/mL actinomycin D (Act D) for 1–10 hours. Data were presented as relative to WT mice at 0 hours (n = 6 mean ± SE). C, Western blotting showing the expression of GCNF protein in spermatocytes and round spermatids of WT and GRTH-KO mice. *, significant increase (P < .01) in GCNF expression in GRTH-KO mice compared with WT mice. D, left, Western blotting showing the effect of depletion of endogenous GRTH protein (above) on the expression of GCNF protein (middle) in seminiferous tubules of WT mice transfected with GRTH siRNAs. β-Actin used as loading control (below). D, right. Densitometry analysis of GCNF protein bands from 3 independent experiments were quantified and normalized to β-actin. *, significant increase (P < .05) in GCNF expression in GRTH siRNA sample compared with scramble siRNA sample. Mean ± SE of 3 independent experiments done in triplicates.

Discussion

The in vivo and in vitro studies have demonstrated an indirect action of androgen on GCNF expression. This increase in turn up-regulates GRTH transcription/expression of GRTH gene primarily in round spermatids. The actions of GCNF are exerted through a distal extended half-site consensus response element residing at −5273/−5264 of the 5′-flanking region of the GRTH gene.

Our studies have shown a nuclear localization profile of GCNF expression in spermatocytes and spermatids of adult mice. Gradual increases in staining intensity were observed from stages I to VI/VII reaching maximal levels at VIII–IX of spermatocytes and at stage VIII (step 7/8) in spermatids (round spermatids and elongated spermatids) with mark reductions thereafter. These were more pronounced in elongated spermatids, where undetectable levels were observed at stages X–XII (round spermatids elongate from stage IX, in stages X–XII round spermatids are not present). The GCNF protein profiles are also consistent with GCNF RNA expression profile reported in previous studies (20) but differ from earlier immunocytochemical studies, where nuclear immunostaining was less intense in stages VII and VIII round spermatids (21). Of much relevance is the GCNF protein profile derived from our studies resembling those reported for protein expression of GRTH in our previous study (10). GCNF binds specifically to an extended half-site consensus sequence single core motif TCAAGGTCA (AGTTCCAGT) residing in the 5′-flanking region of the GRTH. This was shown in EMSA by binding of a complex to the labeled DNA probe upon incubation with nuclear extracts that was prevented by excess unlabeled probe. Moreover, mutation of critical bases of the element prevented binding and formation of the complex and complete supershift of the complex was attained using an specific GCNF antibody. Further evidence was provided by the association of GCNF to GCNF putative binding site on GRTH gene as shown by ChIP assay which validated the specificity of its interaction with the GRTH gene. GCNF generally binds with high affinity as homodimer to a direct-repeat (DR) motif of AGGTCA half-site with DR0 (22, 23) unlike to other orphan receptors, including SF1, NGFI-B, and ROR/RZR, which bind as monomers to extended half-sites (24, 25). GCNF was also found to bind strongly as homodimer to a consensus element with extended half-site TCAAGGTCA (26). In contrast, other studies indicated this type of interaction to have an apparent lower affinity than that for the homodimeric form without or with extension (22). In our study, GCNF, which binds to a single core motif preceded by TCA in the GRTH gene, its effective functional binding is indicated by the significant decrease of GRTH gene directed expression in GCs resulting from knock-down of GCNF (Figure 5A). Thus, our studies revealed that GCNF functions as positive regulator of transcription of the GRTH gene through its binding to the extended half-site element. In contrast, in previous studies GCNF was found to inhibit through binding DR0 element transcription/expression of protamine 1 and protamine 2 genes, which are essential for chromatin compactation in spermatogenesis (27).

Treatment of mice with flutamide, which blocks androgen action at the receptor level, caused marked inhibition of GCNF protein expression in round spermatids but not in spermatocytes, and immunocytochemical studies corroborated these findings. This differential effect of androgen was also observed in the recruitment of GCNF to the GRTH promoter, indicating a transcriptional nature of the regulation of GRTH by GCNF and of androgen in GCNF expression. The actions of androgen in GCNF expression are presumably mediated through ARs present in Sertoli cells through yet unidentified mediators or activators acting in a paracrine fashion on round spermatids at tubular sites. There is an intimate cooperation between Sertoli cells and GCs (meiotic and haploid GCs), and these depend on the stage of the cycle (28). Moreover, GCs can modulate Sertoli cell responses to regulators (29). Thus, it reasonable to foresee the contribution of a subset of Sertoli cells mediating the actions of androgen in round spermatids at stages VI/VII and VIII, where both GCNF and GRTH were found to be maximally expressed.

The evidence of GRTH association with GCNF mRNA and its inhibitory effect on GCNF message stability indicates that this helicase has a role in the regulation of its own transcriptional regulator in GCs. In addition, the increases on GCNF expression observed in round spermatids of GRTH null mice (Figure 6C) could be compounded by the increase androgen formation in Leydig cells resulting from the lack of GRTH (8). Thus, the combinatorial effects from significant increased message stability and paracrine androgen action (from Sertoli cells) in GCNF provide posttranscriptional and transcriptional avenues for the observed significant increase in GCNF expression to explain the autocrine regulation by GRTH of its own transfactor. In the posttranscriptional regulation of GCNF expression the small RNA regulatory pathway could be involved in the control GCNF message degradation (Figure 6B) (2). In our previous studies in Leydig cells, androgen through its receptor stimulates GRTH transcription, which in turn by association with StAR message controls StAR mRNA stability/levels and hence protein expression posttranscriptionally. This is similar to the case of GCNF in the present study, because StAR protein is required for androgen formation which is essential for GRTH transcription in Leydig cells (8).

Testosterone produced in the Leydig cells is essential for the initiation, progression and maintenance of spermatogenesis beyond meiosis. Loss of androgen production by hypophysectomy causes degeneration of GCs and androgen treatment promotes spermatogenesis (30). Complete loss of AR function with marked resistance to androgen due to mutation in the X-linked Tfm mouse or targeted disruption of the AR gene cause infertility due to failure of GCs to progress beyond meiosis (31). Many studies indicate that ARs are not present in GCs of adult testes (32, 33); however, other studies point their presence (34, 35). Our immunocytochemical studies using AR antibody revealed AR expression in Sertoli and Leydig cells but not in GCs of mice testis (Supplemental Figure 1). Furthermore, mice with specific disruption of AR in GCs preserved fertility. Thus, if present in this location, those ARs seem not to be essential for spermatogenesis and male fertility in mice and/or direct targets for androgen action. In addition, those present in peritubular cells when disrupted similarly do not alter fertility despite the observed reduction of sperm count. However, it is generally accepted the relevance and function of AR receptors present in Sertoli cells and Leydig cells to male fertility, because their disruption cause spermatogenic arrest in meiotic and postmeiotic stages, respectively (36). A number of early studies using microarray analysis of Sertoli cells from AR gene-targeted mice vs WT showed little correlation among different genes regulated by androgen and most of these were down-regulated (37). More recently, two studies based on Sertoli cell models using RNA-seq have unveiled transcriptomes with a significant proportion of androgen up-regulated genes (38, 39). However, these studies have not provided a corresponding functional proteomic analysis in Sertoli cells or direct/indirect connection to androgen intermediates or targets in GCs, and this area will require much further development to understand the relevance of these genes in the regulation of GC target genes.

We identified putative binding sites for TFs with the highest conserved core elements in upstream of the GCNF gene promoter region by MatInspector (Genomatrix). All of these TFs that are expressed in testicular cell types play a crucial role in spermatogenesis (Table 1) (36, 40–45) and are potential candidates for the regulation of GCNF transcription/expression. These include CREB, RHOX6, GATA1, STAT1, DMRT3, and CTCF. GATA1 factor which is up-regulated in Sertoli cells during first wave of spermatogenesis has been identified as one of the AR-regulated gene in a transcription profile analysis, and STAT1 is at the center of a large network of potential TFs regulators of the Sertoli cell transcriptional landscape (39). CREB, which is AR regulated via SRC and MAPK signaling cascade by phosphorylation, could induce specific spermatogenic genes (46). Moreover, activation of the nonclassical pathway has been shown to induce the expression of CREB-mediated gene expression (47), and such modality could activate GCNF transcription/expression in GCs.

Table 1.

Putative Binding Elements for TFs in 5′-UTR Region of GCNF Gene

Gene	Gene Description	Cell Type in Testis	Nucleotide Position	Function in Testis	Reference
CREB1	cAMP-response element-binding protein	Sertoli cells, GCs	−201/−208	CREB plays an important role in cell proliferation and differentiation in the early phase of postnatal development and spermatogenesis of mouse testis.	Kim et al (40)
Rhox6	Reproductive homeobox 6	GCs	−390/−396	Expressed in primordial GCs in the developing gonad and plays an important role in GC differentiation.	Liu et al (41)
AR (ARE 2.0 half site)	ARE	Leydig cells, Sertoli cells	−790/−795	It plays an important role in male spermatogenesis and fertility. Functions as a ligand-dependent TF, regulating expression of an array of androgen-responsive genes.	Wang et al (36)
GATA1	GATA-binding protein 1	Sertoli cells	−1628/−1632	It is predominantly expressed in the Sertoli cells lineage in testis. It is a developmental stage- and spermatogenic cycle-specific regulator of gene expression in Sertoli cells.	Yomogida et al (42)
STAT1	Signal transducers and activators of transcription	Sertoli cells	−895/−887	Stat1 is activated during capacitation and the acrosomal reaction.	Bastian et al (43)
DMRT3	Doublesex and mab-3-related TF 3	GCs	−424/−430	Dmrts are required for testicular differentiation in vertebrates. It is expressed in the differentiating male genital ridges and also adult testis.	Raymond et al (44); Kim et al (36)
CTCF	CCCTC-binding factor	GCs	−512/−524	It is a DNA binding protein acting as epigenetic regulator where it modifies the transcription of genes by altering their location within the nucleus. It is expressed in testicular GCs.	Hore et al (45)

Our studies clearly indicate that GCNF expression in round spermatids is regulated in a paracrine manner by androgens/AR acting on the Sertoli cells. In response to androgens, GCNF within round spermatids up-regulates GRTH transcription. Androgen action may be mediated by classical or nonclassical signaling pathways (48, 49). Both pathways could lead to activation of kinases and/or phosphatases or other modifiers for recruitment of coactivators or release of repressors from GCNF to induce transcription/expression of GRTH. Moreover, we demonstrated autocrine regulation of GCNF by GRTH at posttranscriptional level.

We have clearly provided to our knowledge for the first time a connection of androgen action to two relevant GC genes, GCNF and GRTH, which are essential for the progress for spermatogenesis and clearly established their regulatory interrelationship (Figure 7). Studies encompassing molecular regulatory aspects of androgen on GCNF transcription/expression and function could provide valuable links an clearly facilitate what could be a quite difficult search for identification of androgen/AR-mediated regulated gene product(s) in Sertoli cells affecting germinal cell function and spermatogenesis.

Figure 7.

Schematic diagram showing GCNF regulation of GRTH transcription/expression in testicular GCs linking androgen action and GC gene activation. Androgen (A) synthesized and released from LCs binds to ARs present on Sertoli cells. A/AR complex enters into nucleus, where it activates AR-responsive genes/factors (classical pathway). In addition, the nonclassical androgen pathway should be considered as the mediator of this GRTH activation by TF GCNF (see discussion text). Such Sertoli cells' mediated events/signals is/are in turn passed onto GCs, where activate GCNF, a GC-specific TF, that binds the upstream 5′-region of GRTH gene and promotes its transcription/expression. Association of GRTH with GCNF mRNA and its inhibitory effect on GCNF message stability suggest that this helicase has a role in the regulation of its own transcriptional regulator in GCs.

Additional material

Supplementary data supplied by authors.