Meiotic gatekeeper STRA8 regulates cell cycle by interacting with SETD8 during spermatogenesis

Abstract STRA8 (Stimulated By Retinoic Acid Gene 8) is a retinoic acid (RA) induced gene that plays vital roles in spermatogonial proliferation, differentiation and meiosis. The SETD8 and STRA8 protein interaction was discovered using the yeast two‐hybrid technique using a mouse spermatogonial stem cell (SSC) cDNA library. The interaction of these two proteins was confirmed using co‐immunoprecipitation and identification of key domains governing the protein: protein complex. STRA8 and SETD8 showed a mutual transcriptional regulation pattern that provided evidence that SETD8 negatively regulated transcriptional activity of the STRA8 promoter. The SETD8 protein directly bound to the proximal promoter of the STRA8 gene. STRA8 increased the transcriptional activity of SETD8 promoter in a dose‐dependent manner. For the first time, we have discovered that STRA8 and SETD8 display a cell cycle‐dependent expression pattern in germline cells. Expression levels of SETD8 and H4K20me1 in S phase of STRA8 overexpression GC1 cells were different from that previously observed in tumour cell lines. In wild‐type mice testis, SETD8, H4K20me1 and PCNA co‐localized with STRA8 in spermatogonia. Further, our studies quantitated abnormal expression levels of cell cycle and ubiquitination‐related factors in STRA8 dynamic models. STRA8 and SETD8 may regulate spermatogenesis via Cdl4‐Clu4A‐Ddb1 ubiquitinated degradation axis in a PCNA‐dependent manner.

molecular diagnosis methods and possible gene therapy for male infertility.
Retinoic acid (RA), an active metabolite of vitamin A in vivo, is a crucial signalling molecule for cell proliferation, differentiation, meiosis and sperm release. 1,2 Stimulated by Retinoic Acid Gene 8 (STRA8) is induced by RA and plays vital roles in spermatogenesis, primarily in spermatogonial proliferation, differentiation and meiosis. In the Han-Chinese population, STRA8 mutation is associated with azoospermia and oligozoospermia. [3][4][5][6] Male STRA8 knockout (KO) mice are infertile, displaying a lack of meiotic and post-meiotic germ cells. As a specific meiosis gatekeeper, STRA8 displayed a stage and cell specificity expression pattern in male spermatogenesis. 7,8 However, the molecular mechanism of STRA8 in spermatogenes remains unclear.
STRA8 protein was used as bait to screen for interacting proteins from the mouseSSC cDNA library though the yeast two-hybrid technique. SETD8 and H4K20me1 are integral to many key physiological processes, including transcriptional regulation, 10,11 cell cycle progression, [11][12][13] DNA replication 14 and regulation of PCNA 15 to name a few. SETD8 was identified as a novel protein interacting with STRA8, the expression and function of which during spermatogenesis has not been previously studied.
In this study, we verified the protein-protein interaction and determined the critical domains required for STRA8 and SETD8 interaction. The SETD8 protein directly binds to the proximal promoter of STRA8 gene and negatively regulates the transcriptional activity of STRA8. STRA8 increased the transcriptional activity of SETD8 promoter in a dose-dependent manner. STRA8 is critical to the cell cycle transition from mitosis to meiosis; for the first time, our study revealed a STRA8 and SETD8 cell cycle-dependent expression pattern in germ cells. In wild-type testis, STRA8 co-localized with SETD8, H4K20me1 and PCNA in spermatogonia. In STRA8 dynamic models, cell cycle-related molecules and ubiquitination-related factors displayed significant abnormal expression patterns.
Our findings indicate that STRA8 and SETD8 may regulate spermatogenesis via Cdl4-Clu4A-Ddb1 ubiquitinated degradation axis in a PCNA-dependent manner.

| Mice
Heterozygous B6.Cg-STRA88 tm1Dcp/J mice (STRA8 +/− mice) were purchased from Jackson Laboratories (https ://www.jax.org/) with a C57BL/6 genetic background. Homozygous (STRA8 −/− , STRA8 KO) and wild-type (WT) male mice were obtained from the mating pair of STRA8 +/− female and male mice. Two additional models were derived, a vitamin A deficient (VAD) male mouse model and a vitamin A recovery (VAR) male mouse model on C57BL/6 background. 16 WT female and male mice were mated after being fed a vitamin A deficient diet for two weeks. The vitamin A deficient diet was maintained. The VAD mouse model was obtained after newborn mice were fed for 13-14 weeks on the VAD diet. The VAD mice were then fed a normal vitamin A diet for 40 days to establish the VAD recovery mouse model. All animal feed was provided by the Trophic Animal Feed High-tech Co., Ltd. All mice were maintained at the Laboratory Animal Center of Yangzhou University (Yangzhou, China). The day of birth was defined as 0 days post-partum (dpp). All animal experiments of this study were approved by the Animal Ethics Committee of Yangzhou University.

| Dual-luciferase reporter (DLR) assay
We constructed mouse STRA8 promoter luciferase reporter plasmid by restriction digest with Xhol-HindⅢ and SETD8 promoter reporter plasmids containing different fragments by restriction digest with KpnI-HindⅢ. Respectively, they were named pGL3-STRA8Pro, pGL3-SETD8ProF1R (−1980~+1 bp), pGL3-SETD-8ProF2R (−1499~+1 bp), pGL3-SETD8ProF3R (−1000~+1 bp) and pGL3-SETD8ProF4R (−495~+1 bp). The primers sequences are provided in Table 1. Luciferase activity was quantified using the Dual-Luciferase Reporter assay system (Promega). Firefly (Photinus pyralis) luciferase was used as an reporter gene and Renilla (Renilla reniformis) luciferase (pRL-CMV) as the internal reference reporter gene. The specific fragment of the target promoter was inserted in front of the luciferase expression sequence. pGL3-Basic were null reporter vectors without transcriptional activity and acted as a negative control. pGL4 luciferase vectors were used as positive controls as they contained numerous configurations. These included configurations with the synthetic Firefly Luc2 and Renilla hRluc luciferase genes, which have been codon optimized for more efficient expression and risk of anomalous transcription. Approximately 0.998 μg of reporter gene (pGL3-Basic, pGL3-target promoter or pGL4) was cotransfected with 0.0014 μg pRL-CMV into 293 cells and GC1 cells by

| Chromatin immunoprecipitation (ChIP) assay
A minimum of 4 × 10 6 F9 cells were prepared for each chromatin immunoprecipitation. The following experiments were performed according to the instructions of SimpleChIP ® Enzymatic Chromatin IP Kit (Cell Signaling Technology). Formaldehyde and glycine were used to crosslink proteins to DNA, and 0.5 μL micrococcal nuclease per IP prep was used to digest DNA and 0.5 mol/L EDTA to stop digestion. Nuclei were completely lysed after 3 sets of 20-seconds pulse sonication.
After purification, DNA concentration was between 50 and 200 μg/ mL and DNA was digested to a length of approximately 150-900 bp.
Each sample IP contained 5 to 10 μg of digested, crosslinked chromatin and an appropriate amount of antibody and incubated for 4 hours at 4°C with rotation. About 30 μL of Protein G Magnetic Beads was immediately added to each IP reaction and incubated for 2 hours at 4°C with rotation. After the elution of chromatin from antibody/protein G magnetic beads and reversal of crosslinks, DNA was purified using spin columns and analysed by quantitative reverse transcription PCR (qRT-PCR) using specific primers (Table 1). IP efficiency was using the Percent Input Method using the Equation 1. With this method, signals obtained from each IP are expressed as a per cent of the total input chromatin. Absolute value: % Input: positive H3 > 1%, negative IgG < 0.1%, histone 1%-50%, TF/cofactor: 0.2%-1%.
Calculation of CHIP signal (%Input): CT = Threshold cycle of PCR reaction.

| Cell culture and RA induction
The spermatogonia germ cell line (GC1 spg) was a gift from Professor Sha, State Key Laboratory of Reproductive Medicine, Nanjing Medical University, China. GC1 spg cell line was established by using SV40 large T antigen from 10 dpp BALB/c mouse testis. It is a cell type that intermediates between type B spermatogonia and pre-leptotene spermatocyte. It exhibited some characteristics of early spermatogonia which

| Cell cycle synchronization
In order to further study the mechanism of interaction between

| Western blot
Cells and tissue proteins were extracted using RIPA buffer. Protein extracts (50 μg) were subjected to SDS-polyacrylamide gel electrophoresis and transferred to a polyvinylidene difluoride (PVDF) membrane which was activated by methanol. Post-blocking with Tris-Buffered Saline (TBS) containing 5% non-fat milk for 1 hour at RT, the primary antibody was incubated using the anti-sticking method (anti-STRA8, 1:500; anti-SETD8, 1:500, anti-H4k20me1, 1:10 000) overnight at 4°C. The membranes were then incubated with secondary antibodies (ZSbio) at room temperature for an hour.
Proteins were detected by ECL chemiluminescence.

| Histology
Mice testes were fixed in Bouin's solution or 4% paraformaldehyde and dehydrated using a gradient alcohol series and embedded in paraffin for analysis. Sections (5-μm thick) were prepared for haematoxylin and eosin (HE) staining and immunohistofluorescence following standard procedures. Cell types were identified by their location, morphological structure and chromatin. Images were obtained using a Nikon epifluorescence microscope.

| Statistical analysis
Statistical analysis was performed using SPSS18.0 (SPSS Inc). Data are expressed as mean ± standard deviation (SD) and Student's t test.
All experiments were repeated independently a minimum of three times. P value < .05 represents a statistically significant difference.

| Mutual transcriptional regulation of STRA8 and SETD8
Previously, we have reported the SETD8 and STRA8 protein interaction, but the mechanism of how this protein: protein combination may regulate inter-transcriptional regulation during spermatogenesis remains unknown. To examine the transcriptional regulation of SETD8 on the STRA8 promoters, we co-transfected the pCMV-HA, pCMV-HA-SETD8 with the recombinant luciferase reporter plasmid pGL3-STRA8Pro into HEK-293T and GC1 spg, respectively, finding that the luciferase activity of the SETD8 eukaryotic expression plasmid was significantly lower than that of the pCMV-HA plasmid transfected group (P < .05). We then varied the quantity of eukaryotic expression plasmid pCMV-HA-SETD8, 0.0625 μg, 0.125 μg, 0.25 μg and 0.5 μg, which were added into the pGL3-STRA8Pro transfection group. We found different concentrations of pCMV-HA-SETD8 had no obvious affect on STRA8 promoter activity ( Figure 1A,B).
Western blot results verified that the expression of SETD8 protein increases with DNA concentration ( Figure 1C). These results suggest that SETD8 protein inhibits the transcriptional activity of the STRA8 promoter but not in a dose-dependent manner.
Subsequently, we constructed reporter plasmids containing different length fragments of the SETD8 promoter. Luciferase analysis demonstrated that all these SETD8 promoters had luciferase activity, and the promoter located upstream of SETD8 (−1499 ～+1 bp, F2R) reported the strongest transcriptional activity. From these studies, we concluded the SETD8 promoter F2R would be an ideal candidate for subsequent experiments ( Figure 1D). pCMV-MYC-STRA8 and pGL3-SETD8 ProF2R were co-transfected into HEK-293T and GC1 cells. Luciferase activity of STRA8 eukaryotic expression plasmid was significantly higher than that of pCMV-MYC plasmid transfection group (P < .05). We then scaled the DNA concentration of pCMV-MYC-STRA8 0, 0.0625 μg, 0.125 μg, 0.25 μg and 0.5 μg, respectively. These studies found that the SETD8 promoter activity was significantly increased (P < .05) when the dose of pCMV-MYC-STRA8 increased, especially, at 0.25 μg and 0.5 μg plasmid concentrations ( Figure 1E). Western blot analysis confirmed the expression of STRA8 protein was increased as DNA concentration ramped up ( Figure 1F). These results suggest that STRA8 protein enhances the transcriptional activity of SETD8 promoter in a dose-dependent pattern. Taken together, the above studies indicate that STRA8 and SETD8 are involved in spermatogenesis by mutual transcriptional regulation.

| SETD8 directly binds to the promoter of STRA8
Deficient levels of SETD8 lead to embryonic lethality, 20 while the absence of STRA8 results in no abnormalities except for reproductive defects. 16 Knockout phenotypes indicate that SETD8 might be an upstream regulator of STRA8. To verify this hypothesis,  Figure 1H). These results show that the SETD8 protein directly binds to the proximal promoter of STRA8 to regulate its transcription.

| Key interaction domains between STRA8 and SETD8
In order to investigate the mechanism of interaction between STRA8 and SETD8, we constructed different fragments of STRA8 and performed self-activation activity analysis in our previous work. These studies identified a glutamic acid (GA)-rich region (143-193 aa) that had no self-activation and toxicity. We reported that SETD8 acted as a interacting protein with STRA8, by yeast two-hybrid technique in a mouse SSC cDNA library. 9 Consistent with the results of the yeast two-hybrid experiments, the interaction between STRA8 and SETD8 was confirmed  Figure 2A). In addition, both truncated fragments F1R2 and F2R1 of SETD8 could interact with truncated fragment F1R2 of STRA8, but neither SETD8 F1R2 nor F2R1 could interact with STRA8 truncated fragment F1R1 (Figure 2A).  spermatogenesis through the germ cell cycle, but the mechanism still remains unclear.

| Construction of STRA8 dynamic expression model
It has been shown in germ cells that SETD8 and STRA8 have similar  Figure 4D). Overall, we successfully constructed a STRA8 dynamic expression model, in which STRA8 expression levels were decreased by the deficiency of vitamin A and increased when vitamin A recovered. These analyses provide a foundation for further studies on the mechanism of interaction between STRA8 and SETD8 during spermatogenesis.

| Expression of SETD8 and H4K20me1 in STRA8 dynamic models
In the testis of adult WT mice, STRA8 was expressed in spermatogonia and pre-leptotene spermatocytes. SETD8, in WT testis, was highly expressed in Sertoli cells and Leydig cells, and expression was also observed in spermatogonia, spermatocytes and spermatozoa.
H4K20me1 was highly expressed in spermatogonia, spermatocytes and spermatozoa of WT testis. PCNA was expressed in proliferating spermatogonia of WT testis. Immunofluorescence co-localization studies showed that STED8, H4K20me1 and PCNA could co-localize with STRA8 in spermatogonia ( Figure 4E).
To further explore the mechanism of interaction between STRA8 and SETD8, we examined the expression patterns of SETD8, H4K20me1 and PCNA in the STRA8 dynamic expression models.
The mRNA level of SETD8 decreased with the reduction in vitamin A and increased when returned to a normal vitamin A diet ( Figure 5A).
However, the protein expression of SETD8 showed an opposite expression pattern. SETD8 expression was significantly increased during the prolonged period of vitamin A deficiency and rapidly returned to normal levels after vitamin A recovery. These results were obtained from wild-type testis. The expression of STED8 was significantly increased in 11 dpp STRA8 knockout testis which is consistent with that in STRA8 dynamic models ( Figure 5B,C). The opposite expression pattern was observed for H4K20me1 protein ( Figure 5E).
The subcellular localization of SETD8 was examined in different stages of the model group testes. During different stages of WT mice development, SETD8 was expressed in the nuclei of spermatogonia, spermatocytes, round spermatids and Sertoli cells ( Figure 5D). H4K20me1 is involved in spermatogenesis with a similar distribution pattern with SETD8 ( Figure 5F); the only observed difference was H4K20me1 high expression in spermatogonia and spermatids in stage IX-XII. In the testis of VAD mice, even with the quantity of germ cells decreased significantly, SETD8 was highly overexpressed in the remaining spermatogonia and Sertoli cells.
As spermatogenesis gradually recovered, SETD8 expression rebounded in spermatogonia, spermatocytes, round spermatid and Sertoli cells. However, H4K20me1 was expressed only in the cytoplasm of spermatogonia at VAD61D and 93D. There was no observed expression difference of H4K20me1 in WT and VAR testis.

| Interaction studies between STRA8 and SETD8 during Spermatogenesis
We have demonstrated that STRA8 and SETD8 play a role in reg- Previous studies have reported that SETD8 protein levels require precise regulation using many regulatory routes to appropriately regulate cell cycle progression and its degradation in S phase is dependent on ubiquitination and phosphorylation. For these reasons, the expression levels of cell cycle-related molecules while Cdt1 displayed an opposite trend in these mouse model studies ( Figure 6D). In addition, we performed RNA-sequencing in a previous study on STRA8 KO and WT testis at 11 dpp, the time that no significant differences existed in morphology and number of germ cells of WT and KO testis. The differentially expressed genes detected by RNA-seq included SETD8, PCNA, Cyclin E2 and DDB1, which were the same as those detected in STRA8 dynamic model. qRT-PCR was performed to verify the sequencing results.
Compared with WT mice, the expression level of SETD8 was significantly increased in KO testis while PCNA, Cyclin E2 and DDB1 were significantly decreased. The expression trends of above molecules are consistent with what we detected in STRA8 dynamic model ( Figure 6E).

| D ISCUSS I ON
Spermatogenesis is a complex process involving precise regulation of cell division and differentiation. Previous studies have shown the STRA8:SETD8 interaction, both of which may act as transcription factors to regulate the functions of other genes. 10,11,23,24 SETD8 knockout mice are embryonic lethal, 20   F I G U R E 5 Expression analysis of SETD8 and H4K20me1 in VAD and VAR mouse models A, Relative mRNA expression analysis of SETD8 in WT, VAR and VAD mouse models. *P < .05 represented statistical differences. B, SETD8 protein expression analysis in VAD and VAR mouse models by Western blot. C, SETD8 protein expression analysis in STRA8 knockout testis. *P < .05 represented statistical difference. D, Immunofluorescence assay of SETD8 (red) and DAPI (blue) in WT testis and STRA8 dynamic expression mouse model. Scale bar = 50 μm. E, H4K20me1 protein expression analysis in STRA8 knockout testis. *P < .05 represented statistical difference. F, Immunofluorescence assay of H4K20me1 (red) and DAPI (blue) in WT testis and STRA8 dynamic expression mouse model. Scale bar = 50 μm gene was specifically expressed in male mice testis and restricted to type B spermatogonia and pre-meiotic germ cells. Although STRA8-deficient testis initiated meiosis, we discovered aberrant expressions of Cyclin A2 [28][29][30][31] and Cyclin E2. 31  processes. In future studies, we will establish and culture spermatogonial stem cell lines (SSCs) to induce differentiation, which may help to determine the exact period of their interaction.
STRA8 is a RA induced gene. In juvenile C57BL/6 males lacking STRA8 gene function, the early mitotic development of germ cells appears to be undisturbed. However, these cells then fail to undergo the morphological changes that define meiotic prophase, and they do not display the molecular hallmarks of meiotic chromosome cohesion, synapsis and recombination. The Anderson EL group concludes that STRA8 regulates meiotic initiation in spermatogenesis. 7 The Endo T group reported that undifferentiated spermatogonia accumulated in unusually high numbers as early as 10 d after birth in mice lacking STRA8, whereas differentiating spermatogonia was depleted. They thus conclude that STRA8 promotes (but is not strictly required for) spermatogonial differentiation. 3 In summary, STRA8 promotes spermatogonial differentiation and is required for meiotic initiation.
STRA8 has been proved to be involved in the process that leads to stable commitment to the meiotic cell cycle but does not affect DNA replication in pre-meiosis. 25 In our study, STRA8, as a meiotic gatekeeper, co-localized with STED8 in spermatogonia, is considered to regulate meiosis cell cycle by interacting with SETD8. We also found the differentially expressed genes related to spermatogonia differentiation in the sequencing results of STRA8 knockout mice, 45 which further suggesting its role in differentiation. However, whether it is related to the role of SETD8 needs further verification.
In the next step, we will establish and culture spermatogonial stem cell lines (SSCs) to induce differentiation, which may help to determine whether the interaction is required for differentiation. Fully understanding the molecular mechanism of STRA8 and SETD8 in spermatogenesis requires further study.

| CON CLUS ION
The interaction between SETD8 and STRA8 has been previously described. STRA8 and SETD8 showed a similar transcriptional regulation pattern. The SETD8 protein directly binds to the proximal promoter of the STRA8 gene. STRA8 and SETD8 expression patterns are cell cycle-dependent in germline cells. STRA8 plays an important role in the transition from mitosis to meiosis cell cycle and co-localized with SETD8, H4K20me1 and PCNA in WT spermatogonia. In STRA8 dynamic expression models, cell cycle-related molecules and ubiquitination-related factors displayed abnormal expression. STRA8 and SETD8 may regulate spermatogenesis via Cdl4-Clu4A-Ddb1 ubiquitinated degradation axis in S phase by a PCNA-dependent manner. In this study, we explored the interaction between meiosis gatekeepers STRA8 and SETD8 during spermatogenesis and laid a foundation for further studies on the mechanism of interaction between them. The study of these spermatogenesis-related proteins will lead to a further understanding of spermatogenesis, provide a basis for the gene F I G U R E 7 Meiotic gatekeeper STRA8 and SETD8 regulate spermatogenesis via Cdl4-Clu4A-Ddb1 ubiquitinated degradation axis in S phase by a PCNAdependent manner. Yellow arrow: SETD8 protein interacts with PCNA, regulating S phase progress of cell cycle. STRA8 protein can interact with SETD8 and PCNA protein, respectively. Blue solid line represents that SETD8 protein inhibits the promoter activity of STRA8 gene. Blue dotted line arrow represents STRA8 protein promotes the promoter activity of SETD8 gene regulation of spermatogenesis, and molecular diagnostic methods and gene therapy possibilities for male infertility. We thank Dr Sha (Nanjing Medical University, China) for kindly providing GC1 spg cell line.

CO N FLI C T O F I NTE R E S T
All authors declare no conflict of interest.

AUTH O R CO NTR I B UTI O N S
CN and JG designed and carried out experiments. XS carried out cell experiments. SM built the mice models. MX and JX collected and analysed the data. YZ designed the study. All authors checked and approved the final manuscript.

DATA AVA I L A B I L I T Y S TAT E M E N T
All data generated during the study are available from the corresponding author (Dr YZ) on request.