Follicle-stimulating hormone-mediated decline in miR-92a-3p expression in pubertal mice Sertoli cells is crucial for germ cell differentiation and fertility

Abstract

Sertoli cells (Sc) are the sole target of follicle-stimulating hormone (FSH) in the testis and attain functional maturation post-birth to significantly augment germ cell (Gc) division and differentiation at puberty. Despite having an operational microRNA (miRNA) machinery, limited information is available on miRNA-mediated regulation of Sc maturation and male fertility. We have shown before that miR-92a-3p levels decline in pubertal rat Sc. In response to FSH treatment, the expressions of FSH Receptor, Claudin11 and Klf4 were found to be elevated in pubertal rat Sc coinciding with our finding of FSH-induced decline in miR-92a-3p levels. To investigate the association of miR-92a-3p and spermatogenesis, we generated transgenic mice where such pubertal decline of miR-92a-3p was prevented by its overexpression in pubertal Sc under proximal Rhox5 promoter, which is known to be activated specifically at puberty, in Sc. Our in vivo observations provided substantial evidence that FSH-induced decline in miR-92a-3p expression during Sc maturation acts as an essential prerequisite for the pubertal onset of spermatogenesis. Elevated expression of miR-92a-3p in post-pubertal testes results into functionally compromised Sc, leading to impairment of the blood–testis barrier formation and apoptosis of pre-meiotic Gc, ultimately culminating into infertility. Collectively, our data suggest that regulation of miR-92a-3p expression is crucial for Sc-mediated induction of active spermatogenesis at puberty and regulation of male fertility.

Graphical abstract

Supplementary Information

The online version contains supplementary material available at 10.1007/s00018-022-04174-9.

Introduction

During the past few decades, the semen quality has progressively deteriorated globally with an alarming decline in sperm counts [1–3]. Recent data indicate 1 in 20 men currently suffer from reduced fertility [4]. At least 30% of male infertility is idiopathic in nature with unknown aetiologies, thereby presently incurable by conventional hormonal supplementations [5]. Thus, there is an urgent need to understand the cellular and molecular events in the testes regulating sperm production [6–8]. The division and differentiation of male germ cells (Gc) to sperm occur inside the testicular seminiferous tubules under the tight control of gonadotropins—follicle-stimulating hormone (FSH) and luteinizing hormone (LH) [9, 10]. LH acts on the interstitial Leydig cells (Lc) to produce the testicular androgen—testosterone (T) [11]. Sertoli cells (Sc) are the major somatic cells of the seminiferous tubules that mature after birth and express the receptors for both FSH (FSH receptor, FSHR) as well as T (androgen receptor, AR), thus providing the micro-environment for Gc nourishment and differentiation [12, 13]. Sc offers structural support to all stages of the developing Gc and aids their movement from the basal compartment to the adluminal region of the seminiferous tubules [14]. This is achieved through the continuous remodelling of the tight junction proteins that form the immunological blood–testis barrier (BTB) [15]. Sc also provides the essential nutrients and nourishment necessary for Gc development [14]. Failure of Sc to perform any of the above functions negatively impacts male fertility.

FSH directs the expansion of Sc population during the neonatal/infantile period which determines the maximal spermatogenic capacity of the adult testes [16]. However during infancy, proliferative Sc remain incapable of responding adequately to the hormonal cues, which is necessary for the robust initiation of Gc differentiation [17–19]. Sc undergo functional maturation during pubertal testicular development with remarkable changes in hormonal signalling networks and transcriptome, leading to the establishment of BTB and robust initiation of Gc differentiation [20–22]. In pubertal Sc, FSH upregulates genes such as Kit ligand (Kitlg) or stem cell factor (Scf), glial cell-derived neurotropic factor (Gdnf), androgen-binding protein (Abp), Kruppel-like factor 4 (Klf4), etc., which are essential for spermatogenesis [20–25]. Mature Sc also produce inhibin B which inhibits FSH synthesis and secretion from the anterior pituitary [26]. Therefore, any defect in such hormonal cues during maturation of Sc adversely impacts the sperm production.

MicroRNAs (miRNA) are small non-coding RNAs (~ 22bps) that regulate the transcriptomic fate during cellular differentiation [27]. After being transcribed by RNA polymerase II, these small RNA fragments are processed by DICER and DROSHA to produce mature miRNA that binds to the 3’ UTRs of the target transcripts, leading to their degradation or translational inhibition [28, 29]. Sc- or Gc-specific ablation of the DICER and/or DROSHA in mice has shown severe testicular defects leading to infertility [30–33]. Several studies have shown miRNAs such as miR-34/449, miR-10a, miR-100-3p, and miR-383 to be crucial for mammalian spermatogenesis [34–36]. Furthermore, patients with Sertoli cell only syndrome (SCOS) show an altered miRNA profile in the testis as compared to healthy fertile individuals [37]. The miRNA profiles differ substantially in immature and mature stages of testicular development in various mammalian species [38–40]. Since the causes of idiopathic male infertility, where hormones and signalling defects are ruled out in addition to obstructive causes, remain unknown, it is crucial to investigate this issue with new angles so that declining sperm count and male infertility both can be addressed appropriately.

In this present study, we intended to identify a microRNA which regulates the expression of various genes known to be essential for the functional maturation of Sc. In the semi-high-throughput screening of miRNAs in Sc reported previously by us, miR-92a-3p showed up as a promising candidate that is abundantly expressed in infant (5 days old) rat Sc, but declined naturally in pubertal (19 days old) rat Sc [41]. MiR-92a-3p belongs to the miR-17/92 cluster, which is one of the most studied miRNA clusters. Comprising a total of 15 miRNAs and four ‘seed’ families, these miRNAs have been studied for their role in cell proliferation, tumorigenesis, immunology, ageing, cardiovascular disease and neurodegenerative diseases [42, 43]. MiR-92a-3p targets genes such as Klf4, Pten, Sirt1, and Rora that play an important role in Sc maturation. However, there are no reports yet on miR-92a-3p-mediated regulation of Sc maturation or male fertility. Here, we have reported that miR-92a-3p directly targets genes such as Cldn11 and Fshr that are crucial for Sc functioning. To evaluate the association between the developmentally altered expression pattern of miR-92a-3p in Sc and spermatogenesis, we generated a novel transgenic mouse model with persistent expression of miR-92a-3p in adult Sc without any decline. These results demonstrate the critical role of miR-92a-3p in FSH-mediated Sc maturation and male fertility.

Results

Differential expression of miR-92a-3p in infant (5 days old) and pubertal (19 days old) rat Sc and regulation of its expression in pubertal Sc by FSH

MiR-92a-3p levels were found to be significantly (p < 0.05) declined in cultured pubertal (19 days old) rat Sc as compared to cultured infant (5 days old) Sc, both of which were given a pulsatile co-treatment with FSH and testosterone (Fig. 1A). We attempted to understand the regulation of decline in miR-92a-3p expression in pubertal Sc by treating them with FSH. MiR-92a-3p expression levels declined significantly (p < 0.05) in response to FSH in cultured pubertal (19 days old) Sc as compared to untreated pubertal Sc (Fig. 1B).

Fig. 1.

Validation of miR-92a-3p-predicted target genes and effect of miR-92a-3p overexpression in pubertal (19 days old) rat Sc. (A) q-RT-PCR data showing the levels of miR-92a-3p in cultured infant (5 days old) and pubertal (19 days old) rat Sc, (B) q-RT-PCR data showing the levels of miR-92a-3p in FSH-treated cultured pubertal (19 days old) rat Sc as compared to untreated Sc, (C) MiR-92a-3p sequence (3′ to 5') and its binding site on the 3’UTR of the predicted target genes, (D) luciferase analysis of WT and mutated 3’UTRs of predicted target genes of miR-92a-3p in HEK293T cells using synthetic Rno-miR-92a-3p mimic, (E) mRNA expression for direct targets of miR-92a-3p in cultured pubertal (19 days old) rat Sc as compared to scramble controls, (F) q-RT-PCR data showing change in levels of genes involved in Sc functional maturation, in miR-92a-3p mimic transfected cultured rat pubertal (19 days old) Sc as compared to scramble transfected Sc. All values are mean ± SEM of at least three independent biological replicates. Paired Student’s t test was used for determining statistical significance. p < 0.05 was considered to be statistically significant

In vitro validation of genes predicted to be targeted by miR-92a-3p

We performed luciferase assay to validate the targets of miR-92a-3p in HEK293T (human embryonic kidney cells) which were used as host cells for conducting luciferase assay as they are easy to transfect. For validating the predicted targets of miR-92a-3p, we co-transfected synthetic Rno-miR-92a-3p mimic with the WT or mutated 3’ UTR (of target genes) bound luciferase in HEK293T cells. Genes like Klf4, Cldn11 and Fshr were found to be directly targeted by miR-92a-3p (Fig. 1C, D). We then sought to validate these direct targets of miR-92a-3p by q-RT-PCR in pubertal (19 days old) Sc transfected with commercial Rno-miR-92a-3p mimics (miRNA overexpression) or scrambled controls. MiR-92a-3p overexpression significantly (p < 0.05) downregulated the transcript levels of Klf4, Cldn11 and Fshr in cultured 19-day rat Sc (Fig. 1E). The transcript levels of other reported targets of miR-92a-3p such as Pten, Rora, Sirt1 and Esr1, and predicted targets such as Lats1 and Tead1 were also significantly (p < 0.05) downregulated in miR-92a-3p mimic-transfected pubertal rat Sc as compared to scrambled miRNA-transfected Sc (Supp. Fig. 1A).

MiR-92a-3p overexpression in cultured pubertal Sc compromised its functional maturation

The expression of genes such as Ar, Gja1, Kitlg, Gdnf and InhbB, which are critically essential for spermatogenesis, were significantly (p < 0.05) downregulated in miR-92a-3p-overexpressing pubertal rat Sc as compared to scrambled miRNA-transfected Sc (Fig. 1F).

Generation of transgenic mice with Sc-specific overexpression of miR-92a-3p

Since the in vitro overexpression of miR-92a-3p in pubertal rat Sc resulted in a prominent decline in the expression of functional maturation markers of the cell, we generated a transgenic mouse model overexpressing miR-92a-3p (miR-92a-Tg mice) in Sc. The expression of miRNA was driven under the proximal Rhox5 promoter, which ensured the overexpression of miR-92a-3p specifically in Sc at the onset of puberty (Supp. Fig. 2A). Slot blot analysis of the F1 generation from the electroporated fore-founder mice identified the transgene-positive animals (Supp. Fig. 2B). MiR-92a-3p levels were significantly (p < 0.05) upregulated in the transgenic testes as compared to age-matched wild-type (WT) controls (Fig. 2A). Despite a similar body weight in miR-92a-3p-overexpressing transgenic mice and WT controls, there was a significant (p < 0.05) decline in the testis weight of the transgenic mice (Supp. Fig. 2C, D). The testis size (Fig. 2B) was evidently reduced in the miR-92a-3p transgenic mice from F1 generation and F2 generation (Supp. Fig. 2E). The percentage ratio of testis weight to body weight, gonadosomatic index (GSI %) also showed a sharp decline in the miR-92a-3p-overexpressing transgenic mice as compared to age-matched WT mice (Fig. 2C).

Fig. 2.

miR-92a-3p in Sc-overexpressing transgenic mice had reduced testis size and low epididymal sperm count. (A) q-RT-PCR data showing the levels of miR-92a-3p in transgenic mice testis as compared to age-matched WT control mice testis. (B) Representative image of testis size and seminal vesicles of mir-92a-3p-overexpressing transgenic mice from F2 generation as compared to age-matched WT controls. (C) Gonadosomatic index (testis weight/body weight × 100) of WT and transgenic mice overexpressing miR-92a-3p. (D) Epididymal sperm count (million per mL) in control and miR-92a-3p-overexpressing transgenic mice. (E) Average litter produced by miR-92a-3p transgenic mice as compared to age-matched WT controls. (F) Representative images of haematoxylin and eosin staining of testicular paraffin sections from miR-92a-3p transgenic mice from F2 generation and age-matched WT controls. Images captured at 20X (scale bar 50 µm) and 40 X (scale bar 20 µm) objective magnifications. Black arrows point at vacuoles observed towards the basal lamina. Asterisk marks represent tubules with very low numbers of mature sperms. (G) Tubule diameter as measured in the testis from miR-92a transgenic and WT mice. At least three different sections from each animal (each with at least 5 seminiferous tubules at 20X magnification) were analysed to measure the tubule diameter. All values are mean ± SEM of at least five WT and five transgenic mice. Unpaired Student’s t test was used for determining statistical significance. p < 0.05 was considered to be statistically significant

MiR-92a-3p-overexpressing transgenic mice were infertile showing testicular atrophy

The total epididymal sperm count was lower than one million per ml in miR-92a-Tg mice as compared to age-matched WT mice that had an average sperm count of more than 5 million per ml (Fig. 2D). The miR-92a-Tg mice were completely infertile, as they failed to produce any litter despite successful coitus, whereas the WT mice consistently produced 10–12 pups in each litter (Fig. 2E). This was confirmed by checking the vaginal plugs in the female mice that were mated with either miR-92a-Tg male mice or WT males. In contrast to the WT testis, the seminiferous tubules in the testis of miR-92a-Tg mice from both F1 and F2 generation had significantly (p < 0.05) reduced tubule size and discernible histological differences with disrupted seminiferous tubule architecture as observed with haematoxylin and eosin staining on tissue paraffin sections (Fig. 2F, G and Supp. Fig. 2E). The tubules also had a vacuolated phenotype towards the basal lamina in the transgenic mice testis (Fig. 2F).

MiR-92a-3p overexpression impaired Sc maturation and increased germ cell apoptosis

The testicular mRNA levels of Klf4, Cldn11 and Fshr that are directly targeted by miR-92a-3p were significantly (p < 0.05) declined in the transgenic mice as compared to age-matched WT mice (Fig. 3A). The functional maturation markers of Sc such as Ar, Gja1, Gdnf, Kitlg, Abp and InhbB were also significantly downregulated in miR-92a-Tg mice testis as compared to that of the age-matched WT control animals (Fig. 3B). These results were concordant with the findings obtained in our in vitro experiments with cultured pubertal (19 days old) Sc from rats. To ensure that miR-92a-3p overexpression specifically causes repression of its gene targets, and not a global repression in transcription, we compared the transcript levels of genes that are not predicted to be targeted by miR-92a-3p. The testicular mRNA levels of Creb1, Runx2, Pitx2 and Ctnnb failed to show any significant changes in their expression (Supp. Fig. 3), thus indicating that overexpression of miR-92a-3p had a direct impact on its target genes and Sc functional maturity markers. The decline in the Sc maturation markers was also associated with an increase in the number of apoptotic Gc. The number of apoptotic cells per tubule as detected by TUNEL assay was significantly (p < 0.05) higher in the transgenic mice testis as compared to WT testis (Fig. 3C, D).

Fig. 3.

miR-92a-3p overexpression in transgenic mice impaired Sc maturity and was associated with germ cell apoptosis. (A) mRNA expression data showing levels of genes directly targeted by miR-92a-3p in miR-92a-Tg mice testis as compared to age-matched WT mice testis. (B) mRNA expression data showing the levels of Sc functional maturation marker genes in miR-92a-Tg mice testis as compared to age-matched WT mice testis. (C) Representative image of testicular sections (7 µm) of miR-92a-3p transgenic and WT mice showing apoptotic germ cells as detected by TUNEL assay. Arrows indicate TUNEL-positive cells. Images captured at 20X (scale bar 50 µm) objective magnification, (D) Quantification of TUNEL-positive cells per tubule in transgenic mice testis as compared to WT testis. At least three different sections from each animal (each with at least 10 seminiferous tubules at 10X magnification) were analysed to determine the extent of germ cell apoptosis. Nuclei were stained with Hoechst. All values are mean ± SEM of at least five WT and five transgenic mice. Unpaired Student’s t test was used for determining statistical significance. p < 0.05 was considered to be statistically significant

MiR-92a-3p-overexpressing transgenic mice exhibited low number of mitotically active Gc and an increased number of Sc

We intended to detect any proliferation of Sc in adult miR-92a-Tg mice, which in turn would explain the maturational status of Sc in transgenic mice that fail to undergo timely maturation. To this end, we performed co-immunostaining for proliferating cell nuclear antigen (PCNA) and Sc-specific SOX9 on tissue sections from adult transgenic and age-matched WT mice (Fig. 4A). The total number of PCNA-positive mitotically active cells per tubule was significantly reduced, while SOX9-positive Sertoli cells per tubule were significantly (p < 0.05) higher in the transgenic mice as compared to age-matched WT controls (Fig. 4B, C). Interestingly, there were no double-positive (PCNA and SOX9) Sc (Fig. 4A), indicating that adult Sc in Tg mice were not proliferative in nature and that there was a significantly (p < 0.05) lesser number of mitotically active Gc in the Tg mice. In addition to this, there was a significant (p < 0.05) increase in the testicular transcript levels of Sox9 in the miR-92a-Tg mice as compared to age-matched WT control mice (Fig. 4D).

Fig. 4.

miR-92a-3p transgenic mice had an increased number of Sc. (A) Representative image of co-immunostaining for PCNA and Sc-specific SOX9 on testis paraffin section (7 µm) from transgenic and WT mice. Nuclei were stained with Hoechst. Images captured at 20X (scale bar 50 µm) objective magnifications. Inset shows magnified image of a part of the tubule. Yellow arrows indicate PCNA-positive cells and red arrow indicates SOX9-positive cells, (B) Quantification of PCNA-positive cells per tubule in transgenic mice testis as compared to WT testis, (C) Quantification of SOX9-positive cells per tubule in transgenic mice testis as compared to WT testis. At least three different sections from each animal (each with at least 10 seminiferous tubules at 10X magnification) were analysed to count the number of PCNA/SOX9-positive cells. Nuclei were stained with Hoechst, (D) Testicular mRNA levels of Sox9 in miR-92a-Tg mice as compared to age-matched WT testis. All values are mean ± SEM of at least five WT and five transgenic mice. Unpaired Student’s t test was used for determining statistical significance. p < 0.05 was considered to be statistically significant

MiR-92a-3p overexpression in transgenic mice led to disruption of the blood– testis barrier (BTB)

Since the testicular mRNA levels of Cldn11 (claudin11) and Gja1 (connexin43) were found to be reduced in miR-92a-Tg mice as compared to age-matched WT mice (as shown in Fig. 3A, B), we further investigated the protein level expression of these BTB components. Immunohistochemistry for claudin11 and connexin43 on frozen testicular sections (7 µm) showed that the transgenic mice testis expressed very low levels of these proteins as compared to age-matched WT testis (Fig. 5A–D). Moreover, testicular protein level of Cx43 (encoded by Gja1) was also found to be significantly reduced in immunoblot analysis (Supp. Fig. 4). Transmission electron microscopy (TEM) of freshly collected testis from miR-92a-Tg and age-matched WT mice revealed a discontinuous and relatively narrow BTB in the Tg mice as compared to the intact BTB in age-matched WT controls (Fig. 5E).

Fig. 5.

miR-92a-3p overexpression in transgenic mice led to a defective BTB. (A) Representative image of immunostaining for claudin11 on testis frozen sections (7 µm) from miR-92a-3p-overexpressing transgenic mice and WT mice. Nuclei were stained with Hoechst. Arrows indicate expression of claudin11. Images captured at 40X (scale bar 20 µm) objective magnifications, (B) Quantification of mean fluorescence intensity per tubule in WT and miR-92a-Tg immuno-stained tissue sections, (C) Representative image of immunostaining for connexin43 (coded by Gja1) on testis frozen sections (7 µm) from transgenic mice and age-matched WT mice. Nuclei were stained with Hoechst. Arrows indicate expression of connexin43. Images captured at 40X (scale bar 20 µm) objective magnifications, (D) Quantification of mean fluorescence intensity per tubule in WT and miR-92a-Tg immunostained tissue sections, (E) TEM images from transgenic mice testis and WT testis showing the BTB near the Sc nucleus (marked with red arrows and braces, scale bar of 5 µm). Higher magnification (scale bar of 1 µm) of marked area to show discontinuous BTB in transgenic testis as compared to an intact continuous BTB in WT controls

Discussion

FSH signalling via testicular Sc plays a critical role in spermatogenesis [13, 44]. FSH acts as a mitogen in foetal and neonatal Sc and such proliferation of Sc ceases during pubertal testicular development setting the upper limit of spermatogenic output of the adult testes [16, 45–48]. A developmental shift in FSH signalling has been observed in maturing rat Sc with an enhanced FSH binding to FSHR, leading to the rapid transition of spermatogonia A to B [18]. MicroRNAs are critical trans-acting factors regulating developmental gene expression during testicular differentiation [37, 49–51]. Both FSH and T induce a series of miRNAs to regulate the transcriptome of Sc [52–54]. We have found in the past that expression of miR-92a-3p declines at puberty and in the present study we have deciphered its role on male fertility [41].

We overexpressed miR-92a-3p in pubertal Sc in vitro, using commercial miRNA mimics. The validated direct targets of miR-92a-3p included genes such as Fshr, Klf4, Cldn11, Lats1 and Tead1. Additionally, previously reported targets of miR-92a-3p such as Pten, Rora, Sirt1 and Esr1 were also validated in pubertal Sc, as positive controls [55–61]. Our in vitro data showed that overexpression of miR-92a-3p in pubertal Sc led to significant decline in critical FSH-responsive genes such as Ar, Gja1, Kitlg, Gdnf and InhbB, which are known to nurture Gc division and differentiation to support spermatogenesis. For example, Kit ligand (coded by Kitlg) or stem cell factor (SCF) and glial cell-derived neurotropic factor (coded by Gdnf) are reported to be essential for renewal of SSC, whereas claudin11 (coded by Cldn11) or connexin43 (coded by Gja1) are the major Sc–Sc and Sc–Gc junctional proteins that are essential components of the BTB [62–65]. Reduced expression of these genes is often associated with azoospermia or oligozoospermia.

Intriguingly, FSH also significantly downregulated the levels of miR-92a-3p in cultured pubertal rat Sc. This led us to investigate the probable role of FSH-responsive miR-92a-3p in Sc maturation and functionality. We therefore prevented the natural decline of miR-92a-3p in pubertal testes by generating a novel transgenic mouse model (miR-92a-Tg). The uninterrupted overexpression of miR-92a-3p in pubertal and post-pubertal transgenic mice was achieved by driving the expression of pri-miR-92a-3p under proximal Rhox5 (Pem) promoter, which gets activated in Sc at around 14 days of post-natal age and continues to express the tagged gene during adulthood in all stages of seminiferous cycles, in particular stages VI–VIII in an FSH-independent manner [66–68]. The Rhox5 promoter is reported to be expressed in the epididymis [69]. However, changes in epididymal miRNA levels would not affect the testicular architecture, which was critically damaged in the miR-92a-Tg mice. The histological analysis of testicular sections from the transgenic mice revealed the presence of few sperms in some tubules. This indicated that the defects in fertility due to overexpression of miR-92a-3p originated within the testes. Overexpression of miR-92a-3p in the transgenic mice resulted in severe dysregulation of the Sc functional maturation, compromising its ability to support Gc division and differentiation, thereby leading to a significant reduction in the testis size, with acute oligozoospermia and complete infertility in these mice. The Lc function in adult transgenic testes was, however, not affected due to overexpression of miR-92a-3p. This was evident from the size of the seminal vesicles, which were similar in transgenic and age-matched WT control mice. The distorted testicular architecture of the miR-92a-Tg mice also lead to massive Gc apoptosis.

Furthermore, a significant decline in the number of mitotically active Gc was found in the transgenic testes as compared to age-matched WT controls, indicating that mostly pre-meiotic Gc (mitotic Gc) were undergoing apoptosis. Although there has been a significant increase in the Sc number (as evidenced by an increased number of SOX9 immunopositive cells) in adult miR-92a-Tg testes, Sc were mitotically inactive. This can be justified, as PCNA-positive cells fail to overlap with the SOX9-positive cells in adult miR-92a-Tg testes. Our data indicated that the number of mitotic Gc were significantly low in adult transgenic testes, corroborating with the reduced epididymal sperm count. Furthermore, our data showed a delayed pubertal onset of Sc maturation in miR-92a-Tg mice, with poor expression of Ar, Klf4, Cldn11, Gja1, Kitlg, Gdnf and InhbB transcripts, thereby having compromised functionality to support robust Gc division and development. FSH and T play critical role in functional maturation of Sc during puberty, which is a prerequisite for establishing male fertility [70, 71]. Persistent presence of immature Sc in adult testes has been reported to be associated with male infertility in human and rodent models [24, 41, 67, 72–76]. It is essential to note here that despite such rise in Sc number in adult transgenic testes, Fshr mRNA expression in whole testicular extract was found to be significantly low in miR-92a-Tg mice probably due to a substantial decline in FSHR in individual Sc, suggesting poor FSH signalling in adult transgenic testes.

Our results demonstrated that preventing the natural decline of miR-92a-3p levels in the mice testis severely affected BTB formation. Besides T, FSH has also been shown to directly upregulate these junction molecules in primary Sc culture [77, 78]. Poor expression of these junction proteins was further justified by low level of Fshr and Ar found in the adult miR-92a-Tg mice as compared to WT controls. The expression of AR is induced by FSH in Sc [79–81]. The AR-responsive tight junction protein Claudin11 and gap junction protein Connexin43 are the major components of the BTB, which spatially protects the neoantigens, expressed during the development of new Gc from the reach of immune surveillance [15]. Consistent with the immunohistological data of claudin11 or connexin43, the ultra-architecture of BTB was found to be discontinuous and non-uniform in the transgenic testis as evidenced by TEM imaging. The disruption of BTB is known to cause male infertility as reported in Sc-specific ablation of Cldn11 or Gja1, where the polarity of Sc gets impaired leading to delayed and disturbed functional maturation of Sc [82, 83]. The disruption of the BTB may be responsible for the severe oligozoospermia observed in the miR-92a-3p-overexpressing transgenic mice.

Kruppel-like factor 4 (KLF4) is a pleiotropic zinc finger transcription factor that critically regulates cellular differentiation and cell cycle control and has been previously reported to be directly targeted by miR-92a-3p [84–88]. Our miR-92a-Tg mice showed significant decline in Klf4 transcript level in adult testes as compared to age-matched WT mice. KLF4 is known to be highly responsive to FSH and we found similar upregulated expression of Klf4 by FSH in normal pubertal rat Sc. Sc-specific selective ablation of Klf4 in mice has been shown to affect the terminal differentiation of Sc, leading to compromised vesicular transport and secretory capacity of these cells [85, 87, 89]. MiR-92a-3p-mediated downregulation of KLF4 might have played a crucial role in limiting Sc maturation.

In summary, we for the first time have demonstrated a critical role of miR-92a-3p, in regulating Sc function and male fertility. Our results suggested that preventing the natural decline of miR-92a-3p in pubertal Sc by its uninterrupted presence in post-pubertal cells leads to a compromised maturation and functioning of Sc, disruption of the BTB and Gc apoptosis, resulting in acute oligozoospermia and complete male infertility. These findings can be attributed to reduced expression of genes such as Fshr, Cldn11 and Klf4, which are directly targeted by miR-92a-3p and genes such as AR, Gdnf, Kitlg, Gja1 and Inhbb, which are critical for Sc functions. We therefore infer that the FSH-induced pubertal decline in miR-92a-3p is critical for the adequate FSH signalling to occur at puberty, which in turn is necessary for the timely maturation of Sc and subsequent spermatogenic onset. However, more detailed studies are required to reveal the regulatory network of miR-92a-3p and its impact on FSH-induced transcriptome in Sc. This study showed that deregulated expression of miR-92a-3p due to genetic or lifestyle factors may be one of the important aetiologies for some forms of idiopathic male infertility. Our data can be helpful to diagnose such cases appropriately using miR-92a-3p as a potential marker of fertility and such individuals can be cured by manipulating the expression of miR-92a-3p in the testis using transient transfections with viral vectors or in cultures using testicular biopsies.

Conclusion

This study has established the role of miR-92a-3p in maturation of Sc and male fertility by directly targeting genes such as Klf4, Cldn11 and Fshr. Our observations suggested that the FSH-induced downregulation of miR-92a-3p in Sc is critical for pubertal onset of robust Gc differentiation at the onset of puberty. Persistent expression of this microRNA in adult transgenic testes resulted in compromised maturation of Sc and impaired spermatogenic progression. Sc-specific overexpression of miR-92a-Tg mice was associated with disruption of BTB and massive Gc death resulting in infertility. Therefore, it is reasonable to conclude that the sustained expression of miR-92a-3p in adult testes may be a putative cause of idiopathic male infertility, and manipulation of miR-92a-3p expression ex vivo or in vivo may serve as potential therapy for treatment of infertility/sub-fertility.

Materials and methods

Animals and reagents used

Wistar rats and FVB/J mice used in this study were procured from the Small Animal Facility of the National Institute of Immunology, New Delhi, India. All animals were housed and used as per the guidelines laid down by the Committee for the Purpose of Control and Supervision of the Experiments on Animals (CPCSEA). All animal-related experimental protocols followed were approved by the Institutional Animal Ethics Committee (IAEC). All reagents were purchased from Sigma (St. Louis, USA) unless stated otherwise.

Cell culture

Rat Sertoli cell culture

Testes were obtained from 20 to 22 infant (5 days old) and 5–6 pubertal (19 days old) Wistar rats for Sc isolation and culture as described previously [18, 90]. Briefly, the animals were euthanized by CO₂ asphyxiation and the testes were dissected out and decapsulated using fine forceps to unpack the seminiferous tubules which were then chopped with a sterile scalpel. These finely chopped tubules were subjected to sequential enzymatic digestion with collagenase (3 mg/20 ml, for 30–35 min) and pancreatin (3 mg/20 ml, for ~ 2 min). This ensured the removal of other testicular cells, namely interstitial cells and peritubular myoid cells, respectively. The cells so obtained were plated in 12-well cell culture plates at a density of 8–9 × 10⁵ cells/clusters per well in DMEM/HAM’s F-12 media with 1% fetal bovine serum (FBS) and kept in a humidified cell culture incubator maintained at 34 °C with 5% CO_2. After 24 h, cells were replenished with fresh serum-free media supplemented with 1% growth factor (5 μg/ml sodium selenite, 10 μg/ml insulin, 5 μg/ml transferrin, and 2.5 ng/ml epidermal growth factor). On the third day, cells were subjected to a brief hypotonic shock using 20 mM Tris–HCl dissolved in culture media, to remove any contaminating germ cells. The cells were subjected to various treatments on day 4 of culture. Sc culture purity was assessed by SOX9 immunostaining along with Hoechst (Supp. Fig. 4).

HEK-293 T cell culture

HEK cells were cultured in DMEM high glucose media supplemented with 10% FBS and maintained in humidified cell culture incubator at 37 °C with 5% CO₂. All experiments were performed at the third or fourth passage of culture, at a seeding density of 0.05 × 10⁶ in 24-well cell culture plates and transfected at ~ 70% cell confluency.

Plasmid cloning

Cloning of UTRs in PmirGlo

The 3′ UTRs of the predicted targets of miRNA-92a-3p were obtained from the MiRDB database (mirdb.com). The miRNA binding site on the 3′ UTR along with its flanking sequences (25 bp each) was taken from the MiRDB database. This sequence was then synthesized as a pair of complementary single-stranded DNA oligos (Sigma Aldrich). The 3′ UTRs were mutated for use as control plasmids, where the seed sequence binding region was replaced by Ts. All oligos were designed such that upon annealing, they generated overhangs compatible with those generated by NheI (3′) and SalI (5′) digestion. The oligos were annealed and the product was checked on 4% agarose gel to ensure the annealing efficiency. The annealed insert was then ligated to the PmirGlo Dual Luciferase plasmid digested with the same set of restriction enzymes (NheI and SalI). To assist the screening of the bacterial colony harbouring the desired clone, a unique restriction enzyme site (EcoRI) was introduced at the 3′ end of the UTR sequences during its design. The ligated product was transformed into E.coli DH5α competent cells and positive clones were identified by double digesting the isolated plasmid samples with EcoRI and HindIII. The plasmid was then used for transfecting HEK cells. The list of UTRs cloned has been provided in Table 1.

Table 1.

List of UTRs used for microRNA target validation by luciferase assay

Name	Sequence
KLF4-UTR-WT-FP	CTAGCTAGAATTCTATGCACTGTGGTTTCAGATGTGCAATAATTTGTACAATGGTTTAG
KLF4-UTR-WT-RP	TCGACTAAACCATTGTACAAATTATTGCACATCTGAAACCACAGTGCATAGAATTCTAG
KLF4-UTR-MUT-FP	CTAGCTAGAATTCTATGCATTTTTTTTTTTTTTTTTTTTATTTGTACAATGGTTTAG
KLF4-UTR-MUT-RP	TCGACTAAACCATTGTACAAATAAAAAAAAAAAAAAAAAAAATGCATAGAATTCTAG
CLDN11-UTR-WT-FP	CTAGCTAGAATTCTTGTTCAGGTAAAATCTGTGCAATAAAATAACAAACTGTCTCCAAAGCG
CLDN11-UTR-WT-RP	TCGACGCTTTGGAGACAGTTTGTTATTTTATTGCACAGATTTTACCTGAACAAGAATTCTA
CLDN11-UTR-MUT-FP	CTAGCTAGAATTCTTGTTTTTTTTTTTTTTTTTTTTTTAAATAACAAACTGTCTCCAAAGCG
CLDN11-UTR-MUT-RP	TCGACGCTTTGGAGACAGTTTGTTATTTAAAAAAAAAAAAAAAAAAAAAACAAGAATTCTAG
FSHR-WT-FP	CTAGCTAGAATTCCTGAATTATTGGTAACAGCAATTAAAACAATGCAATACCGCTATGTGTTTGG
FSHR-WT-RP	TCGACCAAACACATAGCGGTATTGCATTGTTTTAATTGCTGTTACCAATAATTCAGGAATTCTAG
FSHR-MUT-FP	CTAGCTAGAATTCGCTATAACATGTTTTTTTTTTTTTTTTTTTTACTTTGTTTAAACTACTAAAG
FSHR-MUT-RP	TCGACTTTAGTAGTTTAAACAAAGTAAAAAAAAAAAAAAAAAAAACATGTTATAGCGAATTCTAG
LATS1-UTR-WT-FP	CTAGCTAGAATTCTAGTGGTTTTACACCTGGATTACTGTGCAATACAGGAGAAAACCAGCTTTGG
LATS1-UTR-WT-RP	TCGACCAAAGCTGGTTTTCTCCTGTATTGCACAGTAATCCAGGTGTAAAACCACTAGAATTCTAG
LATS1-UTR-MUT-FP	CTAGCTAGAATTCTAGTGGTTTTATTTTTTTTTTTTTTTTTTTTTCAGGAGAAAACCAGCTTTGG
LATS1-UTR-MUT-RP	TCGACCAAAGCTGGTTTTCTCCTGAAAAAAAAAAAAAAAAAAAAATAAAACCACTAGAATTCTAG
TEAD1-UTR-WT-FP	CTAGCTAGAATTCTAACTTTAATACCCATGACAGTTAAGTGCAATTATTTCATCACTCTAAAAG
TEAD1-UTR-WT-RP	TCGACTTTTAGAGTGATGAAATAATTGCACTTAACTGTCATGGGTATTAAAGTTAGAATTCTAG
TEAD1-UTR-MUT-FP	CTAGCTAGAATTCTAACTTTAATATTTTTTTTTTTTTTTTTTTTTATTTCATCACTCTAAAAG
TEAD1-UTR-MUT-RP	TCGACTTTTAGAGTGATGAAATAAAAAAAAAAAAAAAAAAAAATATTAAAGTTAGAATTCTAG
PTEN-WT-FP	CTAGCTAGAATTCGTTTAGTTTTAGAAAATTTGTGCAATATGTTCATAACGATGGCTGTGGTTG
PTEN-WT-RP	TCGACAACCACAGCCATCGTTATGAACATATTGCACAAATTTTCTAAAACTAAACGAATTCTAG
PTEN-MUT-FP	CTAGCTAGAATTCGTTTAGTTTTTTTTTTTTTTTTTTTTTTGTTCATAACGATGGCTGTGGTTG
PTEN-MUT-RP	TCGACAACCACAGCCATCGTTATGAACAAAAAAAAAAAAAAAAAAAAAACTAAACGAATTCTAG
SIRT1-WT-FP	CTAGCTAGAATTCAGTTGTGAGCTTAAGTGAAGTCTGTGCAATTGCCTGAAGTCCTGTTTCACG
SIRT1-WT-RP	TCGACGTGAAACAGGACTTCAGGCAATTGCACAGACTTCACTTAAGCTCACAACTGAATTCTAG
SIRT1-MUT-FP	CTAGCTAGAATTCAGTTGTGAGCTTTTTTTTTTTTTTTTTTTTTGCCTGAAGTCCTGTTTCACG
SIRT1-MUT-RP	TCGACGTGAAACAGGACTTCAGGCAAAAAAAAAAAAAAAAAAAAAGCTCACAACTGAATTCTAG
ESR1-WT-FP	CTAGCTAGAATTCCAAGAACCACAAATGGAAAGTGGATGTGCAATAAGTATTTTGTATGAAAAG
ESR1-WT-RP	TCGACTTTTCATACAAAATACTTATTGCACATCCACTTTCCATTTGTGGTTCTTGGAATTCTAG
ESR1-MUT-FP	CTAGCTAGAATTCCAAGAACCACAAATTTTTTTTTTTTTTTTTTTTAGTATTTTGTATGAAAAG
ESR1-MUT-RP	TCGACTTTTCATACAAAATACTAAAAAAAAAAAAAAAAAAAATTTGTGGTTCTTGGAATTCTAG
RORA-WT-FP	CTAGCTAGAATTCTCTTTTTAATTCTTACCTAGTGCAATATCTGTACATAGAGCACTTGCGGGG
RORA-WT-RP	TCGACCCCGCAAGTGCTCTATGTACAGATATTGCACTAGGTAAGAATTAAAAAGAGAATTCTAG
RORA-MUT-FP	CTAGCTAGAATTCTCTTTTTTTTTTTTTTTTTTTTTTTTTTCTGTACATAGAGCACTTGCGGGG
RORA-MUT-RP	TCGACCCCGCAAGTGCTCTATGTACAGAAAAAAAAAAAAAAAAAAAAAAAAAAGAGAATTCTAG

Cloning of miR-92a-3p overexpression plasmid in dual PEM (Rhox5) vector

The primary microRNA (pri-miRNA) sequence for Mmu-miR-92a was obtained from the miRBase database (www.mirbase.com). The pri-miRNA sequence and the upstream/downstream 200 bp flanking sequences were PCR amplified from mouse genomic DNA. The forward and reverse primers were designed to have EcoRI and SalI restriction sites, respectively, to facilitate its cloning in dual PEM (Rhox5) plasmid vector double digested with the same set of enzymes. The dual PEM vector was designed on an IRES backbone. The insert was cloned downstream of the first PEM promoter which drives the expression of the gene of interest, while the second PEM promoter drives the expression of the GFP reporter gene. The vector and insert were ligated and transformed into E.coli. The colonies obtained were screened by colony PCR with specific primers on the vector backbone and the insert.

In vitro treatments

Pulsatile hormone treatment of rat Sc

The cultured rat Sc were subjected to hormone treatment on day 4 of the culture in a pulsatile manner [91]. The cells were treated together with ovine follicle-stimulating hormone (50 ng/ml o-FSH) and testosterone (10⁻⁷ M) for 30 min, followed by 2.5 h incubation with hormone-free media for four cycles. The cells were harvested in TRI reagent at the end of the fourth cycle for RNA isolation.

Mimic transfection in rat Sc

On day 4, cultured pubertal (19 days old) Sc were transfected in a 12-well plate with Rno-miR-92a-3p mimic (200 pmol) procured from Dharmacon, using Lipofectamine 2000 (Invitrogen) as per the manufacturer’s protocol. Scrambled mimic was used as a negative control. Briefly, the miRNA mimic and Lipofectamine were separately diluted (quick vortex) in 50 µL of Opti-MEM media and mixed together upon 5 min of incubation at room temperature (RT) to make the transfection mixture. This mixture was incubated further for 20–30 min at RT. Meanwhile, the Sc were drained of its existing media and gently washed once with 1× phosphate-buffered saline (PBS) and kept in 150 µLOpti-MEM (Invitrogen) media and the transfection mixture was added dropwise to the Sc. The cells were then incubated at 34 °C for 9 h, after which the media was changed back to DMEM/Ham’s F-12 (with 1% growth factors). The cells were finally harvested in TRI reagent or as cell pellets 24 h post-media change for RNA and protein isolation, respectively.

Plasmid transfection in HEK-293 T

HEK293T cells were cultured and at ~ 70% confluency the cells were co-transfected with Rnu-miR-92a-3p mimic (20 pmol) and PmirGlo Dual Luciferase plasmid containing wild-type (WT) or mutated 3’UTR of the target gene (500 ng), using Lipofectamine 2000 (Invitrogen) as per the manufacturer’s protocol. Briefly, the mimic and the plasmid were mixed in Opti-MEM and incubated to make the transfection mixture. The existing media of the cells were replaced with the transfection mixture and incubated for 6 h, after which the media were changed back to complete media (DMEM high Glucose + 10% FBS). The cells were harvested at 24 h post-media change and the cell pellet was stored at − 20 °C for luciferase assay.

RNA isolation and c-DNA preparation

Total RNA was isolated from Sc frozen in TRI reagent (Sigma Aldrich) using chloroform as described previously [92]. Additionally, sodium acetate (0.3 M) and glycogen (0.2 mg/ml) were added for enhanced precipitation of small RNAs. The quantity and quality (260/280) of RNA were determined using Nanodrop 2000c spectrophotometer (Thermo scientific, USA). 1 μg of RNA was treated with 0.5U DNaseI (ThermoScientific, USA) to remove any contaminating genomic DNA fragments. This was followed by single-strand c-DNA synthesis using M-MLV Reverse Transcriptase (Promega, USA,) as per the manufacturer's protocol.

For the c-DNA synthesis of microRNAs, 1 μg RNA was treated with E.coli PolyA Polymerase (NEB) to generate a poly(A)-tail, followed by reverse transcription using 0.5 mg/ml oligo-dT and M-MLV RT. A 32 bp unique sequence was added to the 5’ end of the oligo-dT (GCGAGCACAGAATTAATACGACTCACTATAGGTTTTTTTTTTTT), which served as a template for a universal reverse primer for the downstream qPCR analysis of all the miRNA. The entire microRNA sequence was used as the forward primer. Universal reverse primer sequence has been mentioned in Table 2.

Table 2.

List of primers used for q-RT-PCR of genes

Gene	Mouse forward primer	Mouse reverse primer	Rat forward primer	Rat reverse primer
18S	GCAATTATTCCCCATGAACG	GGCCTCACTAAACCATCCAA	-do-	-do-
AR	GCCTCCGAAGTGTGGTATCC	GGTACTGTCCAAACGCATGT	CTTATGGGGACATGCGTTTGG	GAGCTCCGTAGTGACAACCA
Gja1	TAAGTGAAAGAGAGGTGCCCAG	CCCAGGAGCAGGATTCTGAAA	GTCTACCCCTCTGGGTGTGA	AGGACCAGTCGAGGATGAT
InhbB	CTTCGTCTCTAATGAAGGCAACC	CTCCACCACATTCCACCTGTC	TCCTAGTGCCCTGCTGAGAT	ACCCACAGGGACAACTTCTG
Klf4	CTTGGCCCCGGAAAAGAAC	TTCTCGCCTGTGTGAGTTCG	GTGCCCCGACTAACCGTTG	GTCGTTGAACTCCTCGGTCT
Fshr	CCTTGCTCCTGGTCTCCTTG	CTCGGTCACCTTGCTATCTTG	AACGCCATTGAACTGAGGTTTG	GGTTGGAGAACACATCTGCCT
Cldn11	CTCCTTATTCTGCTGGCTCTCT	CATCACAGCACCGATCCAAC	ACGGTTGCGTATGCTTTGA	ACACCCATGAAGCCAAATT
KITLG	TCTGCGGGAATCCTGTGACT	CGGCGACATAGTTGAGGGTTAT	GTGGATGACCTCGTGGCATGTA	TCAGATGCCACCATGAAGTCC
GDNF	CCGGATCCGAGGTGCC	CCGAGGGAGTGGTCTTCAG	CGCTGACCAGTGACTCCAAT	CGCCGCTTGTTTATCTGGTG
ABP	GAACGGGGATTCACTGCTG	GCAGGCACGAGCGGAAA	GACGGACCCTGAGACACATT	AGGGTTTGCTGATTTTGGTG
Sox9	AAGAACGGACAAGCGGAGG	AGATTGCCCAGAGTGCTCG	–	–
PTEN	–	–	GCAGAGTTGCACAGTATCC	CCGTCCTTTCCCAGCTTTACA
ESR1	–	–	CGCTCTGCCTTGATCACACA	CGGATGAGCCACCCTGC
Sirt1	–	–	GTTCCAGCCGTCTCTGTGT	GCTGTTGCAAAGGAACCATGA
Lats1	–	–	TGGTGACTCTGGGGATAAAGAA	GGGAGTAACTCTGAATCCGAGAC
Tead1	–	–	GTCCACCAACTCATCACCCG	AACAACCAAACGTGTAGGCAG
Rora	–	–	GTGGAGACAAATCGTCAGGAAT	TGGTCCGATCAATCAAACAGTTC
miR-92a-OE	TTTTTTTTGAATTCAGCACTTCTAGTACTCCTGGATCAAC	TTTTTTTTGTCGACGCTAAAGGATTTTACAATCTTACTATAGCACTG	–	–
miR-92a-3p	TCATTCACGGACAACACTTTTT	GCGAGCACAGAATTAATACGAC	TATTGCACTTGTCCCGGCCTG	GCGAGCACAGAATTAATACGAC

Quantitative real-time PCR

Q-RT PCR was performed in Realplex⁴ master cycler (Eppendorf, Germany) using Kapa Sybr mix as per the manufacturer’s protocol. 1 μL of the c-DNA preparation was taken for each reaction along with 5 μL Kapa Sybr, 0.5 μM each primer (forward and reverse) and 3 μL nuclease-free water (NFW). Melting curve was analysed to detect single amplification peak and the differential expression of gene or miRNA in terms of fold change was calculated using the 2^(−ΔΔCt) method as described previously [93]. Each reaction was set in three technical replicates and at least four biological replicates. The expression level of 18S rRNA and let-7a was used for normalizing genes and miRNA expression level, respectively [94]. The list of primers used in the study has been provided in Table 2.

Luciferase assay for target validation

The harvested HEK cells (co-transfected with miRNA mimic and PmirGlo Dual Luciferase plasmid) were re-suspended in 150 μL PBS and 10 μL of the suspension was taken to perform luciferase assay using Dual-Glo Luciferase Assay Kit (Promega, E2920) following the manufacturer’s protocol. Briefly, the luciferase assay reagent (LAR) was added to the cell suspension. The LAR comprises the lysis solution and the substrate for firefly luciferase protein. After measuring the firefly luciferase activity, a Stop and Glo solution was added which stalled the firefly activity and induced the Renilla luciferase activity. The ratio of firefly/Renilla luciferase activity was calculated and compared between mutated and wild-type (WT) UTR samples for estimating the decline in firefly activity, which reflects the translational repression of the transcripts and thus the miRNA-mediated suppression. Each reaction was set in technical duplicates and at least four biological replicates.

Generation of transgenic mice

Transgenic mice were generated using non-surgical testicular electroporation as described previously by our laboratory [95]. Briefly, the miR-92a overexpression plasmid DNA which used Rhox5 promoter was prepared in high quantity using GenElute^™ HP Plasmid Maxi prep (Sigma Aldrich) and was linearized using StuI restriction enzyme that cuts the plasmid in the backbone. The linearized plasmid was purified by ethanol precipitation and quantified using Nanodrop. The integrity of the linearized DNA was verified by running it on 1% agarose gel. 15 μg of the linearized plasmid DNA at a concentration of 500 ng/μL was injected into each testis of anaesthetized 30 days old FVBJ male mice, followed by electroporation (60 V for 50 ms, 4 forward + 4 reverse pulses) using tweezer electrodes (Electro Square Porator, ECM 830, BTX, USA). The fore-founder animals were housed for 30 days and then mated with age-matched WT females. The pups born (F1 generation) were screened for transgene integration using slot blot analysis (as described below). Several individuals from the F1 generation were transgenic. Infertility was confirmed in F1 generation transgenic males using experiments described below. To continue the transgenic line, WT males and transgenic females from the F1 generation were mated to generate an F2 generation and all experiments were performed on transgenic males from the F2 generation at around 80–90 days of age (Supp. Fig. 2). A control transgenic animal was previously generated by expressing a stem loop-structured shRNA against bacterial LacZ to mimic the pri-microRNA structure, in the same vector backbone. LacZ mice were compared to WT mice to check for any probable effects of transgene overexpression in the animals. As reported previously, no such changes were found between the LacZ control Tg mice and WT mice, and WT was thereafter used as control for all experiments [41].

Genomic DNA isolation and slot blot analysis

Genomic DNA was isolated from 3–4 mm of tail snippets using salt precipitation method [96]. The isolated DNA was quantified using Nanodrop and 1 μg of DNA of each sample was used for slot blot analysis as described previously in our laboratory [95]. Briefly, each genomic DNA sample was denatured at 95 °C for 10 min and immobilized and cross-linked on a positively charged nylon-66 transfer membrane (MDI Membrane Technologies, Ambala Cantt, India, CL-1000 UV crosslinker, UVP). A radioactive DNA probe complementary to the reporter GFP sequence in the plasmid was synthesized by PCR amplifying the GFP coding region. The clean PCR product was denatured and incubated with αP³²-CTP to produce the radioactive probe. The membrane was hybridized with a αP³²-CTP radiolabelled probe in a rotating chamber, overnight at 60 ºC (Amersham). The membrane was washed gently with low stringency and high stringency 2X SSC (3 M NaCl, 0.3 M sodium citrate, pH 7.0 for 20X SSC) buffer twice for 10 min each and kept in exposure cassettes for 12–24 h, which was then scanned using Phosphorimager Typhoon 9400 (GE Healthcare). Genomic DNA from WT mice was used as negative control and 10 ng of the plasmid (which contained the probe region) was used as a positive control.

Immunohistochemistry

The freshly collected testes from WT and Tg mice were fixed in 4% paraformaldehyde for 48 h, washed well to remove the fixative and then used for paraffin embedding [97]. 7 µm sections were cut from paraffin embedded blocks and used for immunostaining. The sections were deparaffinized using xylene–alcohol and subjected to antigen unmasking by boiling the sections for 10 min in antigen unmasking solution (Vector Laboratories, H3300) and cooled to room temperature under tap water. The sections were permeabilized by treating them with 0.1% Triton X-100 for 5 min at room temperature, followed by washing with 1X PBS thrice, for 5 min each, and then blocked with 3% BSA solution for 1 h. The sections were then incubated overnight with the primary antibody in 1% BSA at 4 °C in a moist chamber. The primary antibody was removed and sections were washed with 1X PBS to remove any residual antibody and then kept in secondary antibody solution for 3–4 h, washed again and finally stained with 20 μg/mL Hoechst-3342 to stain the nuclei. The sections were mounted on glass slides using prolong gold antifade mounting media (Life Technologies, USA) and viewed under fluorescent microscope (Nikon Eclipse TE2000-E). Primary antibody dilutions are mentioned in Table 3. The fluorescence intensity of the images was quantified using ImageJ. Each tubule was manually outlined using the draw tool and the intensity was measured. This was done for at least four tubules per section (from 20X images) and at least four sections were analysed from three wild-type and three miR-92a-Tg mice.

Table 3.

List of antibodies used and their dilution

Antibody	Company	Cat. No	Dilution for IHC
Connexin43	CST	3512	1:100a
Claudin11	Thermo Fisher Scientific	36–4500	1:50a
PCNA	Thermo Fisher Scientific	MA511358	1:100a
SOX9	Abcam	ab185230	1:100a
Beta Actin	CST	4967	–
Anti-mouse alexa 488	Thermo Fisher Scientific	A11001	1:500
Anti-rabbit alexa 488	Thermo Fisher Scientific	A11008	1:500

^aWith 1% BSA

Fertility assessment

The fertility assessment was done by measuring parameters such as gonadosomatic index (GSI%), sperm count and litter size in the transgenic and age-matched wild-type mice [41, 98]. The body weight and testis weight for each mice (WT or Tg) were noted and the gonadosomatic index (body weight/ testis weight X 100) was calculated. The testis size was also observed for any visible morphological changes. Transgene-positive animals were euthanized at around 80–90 days of age along with their age-matched WT counterparts. The entire intact epididymis was surgically removed and transferred into a 1.5 ml microcentrifuge tube containing 1 ml 1X PBS and ruptured thereafter for sperm count analysis using a haemocytometer. The litter size for each transgene-positive animal was also noted for at least five mating cycles with at least three age-matched females.

TUNEL assay

Paraffin sections of WT and Tg testis were taken and deparaffinized using xylene and subjected to antigen unmasking using 20 µg/ml proteinase K solution. Apoptosis was detected by performing the TUNEL assay using Promega DeadEnd^™ Fluorometric TUNEL System (G3250) as per the manufacturer’s protocol [99]. Briefly, the sections were fixed in 4% paraformaldehyde for 20–30 min and then permeabilized using 0.2% Triton X-100 in 1X PBS for 5 min. The sections were then equilibrated and incubated with the enzyme mix containing the fluorophore. The enzyme reaction was stopped using 2X SSC buffer and the sections were stained with 20 μg/mL Hoechst for 5 min, washed and finally preserved using antifade (Prolong Gold antifade, Invitrogen). Slides were viewed under fluorescent microscope (Nikon Eclipse TE2000-E) to detect fluorescent TUNEL-positive cells.

Transmission electron microscopy

Whole testis from WT and transgene-positive animals were fixed in 2% paraformaldehyde + 2.5% glutaraldehyde in 0.1 m phosphate buffer solution, for 6–8 h and cut into small pieces of approximately 2 X 2 mm dimension. These pieces were washed once with 1X PBS and incubated for 1 h in osmium tetroxide for secondary fixation. The tissue was dehydrated using ethanol and infiltrated using epoxy resin, which was allowed to settle and polymerize at 60 ºC overnight. Ultrathin sections of approximately 70 nm were cut at Sophisticated Analytical Instrumentation Facility, All India Institute of Medical Science, New Delhi. TEM imaging was done (Tecnai G2 20 Twin) at the National Institute of Immunology, New Delhi.

Protein extraction and immunoblot analysis

Protein was extracted from the tissue by homogenizing it in RIPA lysis buffer (with 1X Protease and phosphatase inhibitor cocktail). Protein was quantified using Bradford solution [100] and 20–30 μg total protein was run on 12% resolving SDS PAGE gel at 100 mV. The protein was transferred onto PVDF membranes and incubated with blocking solution (5% skimmed non-fat milk in 1X TBST) for 1 h at room temperature. The membrane was briefly washed with 1 X TBST and incubated overnight with the primary CX43 antibody (CST-3512, at 1:1000) dissolved in 1X TBST. The next day, the membrane was washed thrice with 1X TBST (5 min each at RT under rapid shaking condition) and incubated for an hour with the HRP-labelled secondary antibody (1:5000) in 1X TBST. The membrane was then developed using BioRad Clarity Western ECL substrate and imaged in a chemi-doc (ImageQuant Las 500, GE Healthcare). The protein bands obtained were quantified using ImageJ. β-actin was used as loading control (CST-4967, 1:1000).

Statistical analysis

Statistical analysis was performed using GraphPad Prism 8.2 software. Data (mean ± SEM) from at least three independent experiments was used for calculating statistical significance. Details of the statistical tests are provided in the figure legends. p value < 0.05 was considered to be statistically significant, where * = p < 0.05, ** = p < 0.01, *** = p < 0.001, **** = p < 0.0001.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

AG, KM and SSM conceptualized the study. AG, KM and SSM designed the experiments. AG performed most experiments with help from AV, AGH and RS. AG, KM, IB and SSM analysed the experimental data. AG, IB and SSM wrote the manuscript. All authors contributed to the article and approved the submitted version.

Funding

We acknowledge the financial support provided by the Department of Biotechnology, Government of India, under a core fund through the National Institute of Immunology. We also acknowledge financial support from JC Bose fellowship of SSM (SERB-JCB/2017/000027) for this work.

Data availability

Not applicable.

Code availability

Not applicable.

Declarations

Conflict of interest

The authors declare no competing interest.

Ethical approval

Included in “Materials and methods”.

Consent to participate

Not applicable.