In mammals, testis development is triggered by the expression of the sex-determining Y-chromosome gene SRY to commit the Sertoli cell (SC) fate at gonadal sex determination in the fetus. Several genes have been identified to be required to promote the testis pathway following SRY activation (i.e., SRY box 9 (SOX9)) in an embryo; however, it largely remains unknown about the genes and the mechanisms involved in stabilizing the testis pathway after birth and throughout adulthood. Herein, we report postnatal males with SC-specific deletion of Raptor demonstrated the absence of SC unique identity and adversely acquired granulosa cell-like characteristics, along with loss of tubular architecture and scattered distribution of SCs and germ cells. Subsequent genome-wide analysis by RNA sequencing revealed a profound decrease in the transcripts of testis genes (i.e., Sox9, Sox8, and anti-Mullerian hormone (Amh)) and, conversely, an increase in ovary genes (i.e., LIM/Homeobox gene 9 (Lhx9), Forkhead box L2 (Foxl2) and Follistatin (Fst)); these changes were further confirmed by immunofluorescence and quantitative reverse-transcription polymerase chain reaction. Importantly, co-immunofluorescence demonstrated that Raptor deficiency induced SCs dedifferentiation into a progenitor state; the Raptor-mutant gonads showed some ovarian somatic cell features, accompanied by enhanced female steroidogenesis and elevated estrogen levels, yet the zona pellucida 3 (ZP3)-positive terminally feminized oocytes were not observed. In vitro experiments with primary SCs suggested that Raptor is likely involved in the fibroblast growth factor 9 (FGF9)-induced formation of cell junctions among SCs. Our results established that Raptor is required to maintain SC identity, stabilize the male pathway, and promote testis development.
Summary Sentence
Raptor deletion induced Sertoli cells into an undifferentiated state and showed some ovarian somatic cell features.
Graphical Abstract
Introduction
In mammals, the testis and ovary differentiate from an embryonic bipotential gonad. However, the testis appears to have a differentiating priority due to the presence of the sex-determining gene SRY on the Y chromosome [1, 2]. In mice, Sry is transiently expressed in XY progenitor cells from embryonic day 10.5 (E10.5)–E12.5, soon after the gonad is first formed [3, 4]. Sry predominantly and directly activates SOX9 [5]. SOX9 stimulates the expression of Pgds [6], which in turn amplifies SOX9 activity synergistically with FGF9 [7–9], further inducing differentiation of supporting cell lineages to the Sertoli cell (SC) fate and establishing the testicular cords. It has been recently demonstrated that expression of Sry is controlled by the epigenetic machinery by specific transcription factors, and epigenetic regulation plays an important role in gonadal sex determination [10, 11]. In the absence of Sry or Sox9 activation, granulosa cell fate and ovary pathway are initiated, mainly involving the R-spondin1 (RSPO1)/WNT4-β-catenin signaling pathway [12–14], in coordination with FOXL2 [15–17] and Follistatin [18]. Other established pro-testis genes include Dhh [19], Pdgfrα [20], Wt1 [21], Cbx2 [22], as well as Nr0b1/Dax1 [23], which was initially classified as a pro-ovary gene [24]. RUNX1 and COUP-TF2 were the two newly identified pro-ovary proteins [25, 26].
The gene expression patterns underlying initial sex differentiation suggest that some testis-specific genes are exclusively activated in the XY gonad, whereas alternately suppressed in the XX gonad at sex determination, and vice versa. It would be important to maintain such a gene activation/suppression mechanism throughout adulthood, as both terminally differentiated Sertoli and granulosa cells have the potential to mutually transform long after sex commitment. In support of this, loss of Foxl2 in adult mouse ovaries led to reprogramming of granulosa cells into Sertoli-like cells along with the formation of a testicular cord-like structure [17]. In contrast, conditional deletion of the pro-testis gene Dmrt1 in adult mouse testes resulted in transdifferentiation of SCs into granulosa cells accompanied by ovarian reorganization [27]. These reports highlight the susceptibility of gonadal sex reversal and indicate that sex-associated genes should be strictly monitored to ensure the destined testis or ovary pathway, even long after commitment in fetal life. However, relative to the initial stage of sex determination in the fetus, it largely remains unknown about gonadal sexual maintenance from the postnatal stage to adulthood, particularly in terms of the genes and mechanisms involved.
The mechanistic target of rapamycin (mTOR) is a highly conserved serine/threonine kinase that nucleates two distinct multiprotein complexes: mTOR complex 1 (mTORC1) and mTORC2 [28]. mTORC1 is a sensitive target of rapamycin and includes mTOR, regulatory associated protein of mTOR (Raptor), and other components, in which Raptor acts as a scaffold for recruiting mTOR substrates. Rheb is a direct upstream activator of mTORC1, which is suppressed by TSC1/2, a functional complex with GTPase-activating protein activity for Rheb [28]. mTORC1 integrates diverse signals, including nutrients, growth factors, energy, and stresses, to regulate cell growth, proliferation, survival, and metabolism [29]. In contrast, mTORC2 is insensitive to rapamycin and mainly contains the rapamycin-insensitive subunit, Rictor [28].
The central roles of the mTOR complex in growth, metabolism, autophagy, aging, and disease have been well established [29–31]. Recently, increasing evidence from us and other groups has demonstrated the critical roles of mTORC1/2 in regulating spermatogonial population homeostasis [32, 33] and SC function [34–37], suggesting its importance in testis development and spermatogenesis. Interestingly, a recent study with conditional knockout of mTOR in primordial follicles reported granulosa-to-SC transdifferentiation and follicle-to-testicular cord conversion in adult mutant ovaries [38], indicating normal mTOR signaling activity is required for the maintenance of granulosa fate and oogenesis. This study showed that mutant mice with conditional knockout of Raptor in SCs (SCRaptorKO) exhibited a profound change in SC morphology, accompanied by disrupted tubular structure and complete loss of testicular cell architecture. To gain more insight into the alteration of molecular networks underlying this change, we performed a genome-wide gene expression analysis by RNA sequencing (RNA-seq) of the control and SCRaptoKO testes. Interestingly, the data indicated dedifferentiation in SCs upon loss of Raptor.
Materials and methods
Mice
The details of generating the mutant mice with SC-specific knockout of Raptor (SCRaptorKO) and their control mates by mating the Amh-Cre and the RaptorLoxP/LoxP mice had been described in our previous study [35]. Eight-week-old wild-type (WT) female C57BL/6J mice were from the Medical Animal Center of the Southern Medical University. All the animal experiments were approved by the Southern Medical University Committee on the Use and Care of Animals and were performed following the Committee guidelines and regulations.
Hematoxylin and eosin, immunofluorescent staining, and transmission electron microscopy
Testes were fixed in modified Davidson fixative, processed in paraffin, and sectioned using standard procedures. At least three sections from each testis (5 µm, taken 100 µm apart) were stained with hematoxylin and eosin (H&E) for regular histological examination.
Immunofluorescent staining was performed according to a standard procedure, using the primary antibodies summarized in Supplementary Table S3 (supplemental_tables_ioac104.pdf) and the Alexa-Fluor-488- or Alexa-Fluor-594-labeled secondary antibodies (Jackson Immunoresearch, West Grove, PA, USA). 4,6-diamidino-2-phenylindole was used to visualize the nuclei. Immunofluorescent images were obtained using a FluoView FV1000 confocal microscope (Olympus, Shinjuku-ku, Tokyo, Japan).
Testis samples were fixed in 2% glutaraldehyde at room temperature for 2 h, and processed under a transmission electron microscope (TEM, H-7500, Hitachi, Chiyoda-ku, Tokyo, Japan).
Quantitative reverse-transcription polymerase chain reaction
Total testicular RNA was purified using the Trizol Reagent (Invitrogen, Carlsbad, CA, USA), and processed to cDNA using the Hifair™ II 1st Strand cDNA Synthesis Kit, followed by amplification and quantification using the Hieff® qPCR SYBR Green Master Mix (all from YEASEN Biotech Co. Ltd, Shanghai, China) with a StepOne Plus Real-Time PCR System (Applied Biosystems, Waltham, MA, USA). Gapdh was used as the endogenous control transcript. Three technical replicates were applied for each transcript. The primer sequences are summarized in Supplementary Table S4 (supplemental_tables_ioac104.pdf).
Enzyme-linked immunosorbent assay
Adult mouse gonads were homogenated in cold saline on ice. After centrifuging at 12 000 rpm at 4°C for 15 min, the supernatant was collected for testosterone and estradiol measurement by using the enzyme-linked immunosorbent assay (ELISA) kits (testosterone, E-EL-0072c, detection level: 0.10–20 ng/ml; estradiol, E-EL-0150c, 9.37–600 pg/ml; both from Elabscience, Wuhan, China). The values are normalized by protein concentrations and present as ng/mg protein and pg/mg protein, respectively.
Ultra performance chromatography-tandem mass spectrometry analysis for steroid measurement
A 50 mg aliquot of each individual sample was weighed and 200 µl distilled water was added, two small steel balls and homogenated three times. After adding 500 µl of extract solution methanol and 40 µl internal standard, the samples were vortexed. Then 600 µl water was added and vortexed again. After centrifugation (5 min, 13 000 rpm, and 4°C), a 700 µl aliquot of the supernatant was transferred to an Eppendorf tube. The sample was further purified with Positive Pressure 96 Processor and then was collected into the 96-hole collecting plate. After being mixed in the micro-hole plate constant temperature vibrator, the sample was subjected to ultra performance chromatography-tandem mass spectrometry (UPLC-MS/MS) analysis. The UPLC separation was carried out using an ACQUITY UPLC-I/CLASS (Waters, Milford, MA, USA), equipped with an XBridge® BEH C8 (100 × 2.1 mm, 2.5 µm, Waters). The auto-sampler temperature was set at 10°C, and the injection volume was 10 µl. MassLynx Work Station Software (Version 4.1) was employed for MRM data acquisition and processing.
Western blotting
Testes were triturated and lysed on ice with radio immunoprecipitation assay (RIPA) lysis buffer. After centrifugation at 12 000 rpm at 4°C, the supernatant was collected and boiled in sodium dodecyl sulfate (SDS) loading buffer. Protein extracts were then subjected to 6–12% SDS-polyacrylamide gel electrophoresis and electrotransferred to nitrocellulose membranes (10 600 001, GE Healthcare Life Sciences). The membranes were then blocked in 5% nonfat dry milk for 1 h at room temperature, washed, and incubated with a primary antibody at 4°C overnight. The membranes were further washed, incubated with horseradish peroxidase-conjugated secondary antibody (111-035-003, 115-035-003, Jackson Immunoresearch) for 1 h at room temperature, washed again, and finally visualized using an enhanced chemiluminescence kit (NEL105001EA, PerkinElmer, Waltham, MA, USA). Quantification was performed by measuring the gray value of bands using the Image J software (1.52a, National Institutes of Health, USA).
RNA sequencing and differential analysis
5-dpp-old control and SCRaptorKO testes were used for total RNA isolation using TRIzol reagent. After quality control, RNA-seq was performed as digital gene expression on an Illumina HiSeq 2500 platform (RiboBioCo. Ltd, Guangzhou, China). Significant differences were analyzed by negative binomial distribution test after read-counts normalization, fold change, and adjusted P value (Q value) by the Benjamini-Hochberg false discovery rate procedure (Q value) were considered as factors for differential analysis. Control and SCRaptorKO testes RNA-seq data have been deposited in the NCBI SRA database (PRJNA767807). WT ovary sequencing data were downloaded from https://www.ncbi.nlm.nih.gov/geo/(GSM5184147, GSM5184148).
Primary SCs isolation, culture, and treatment
Immature SCs were isolated from 5-dpp-old WT C57BL/6J mice as previously described with minor modification [39]. Briefly, decapsulated testes were minced into small fragments and incubated at 37°C for 15 min with collagenase A (1 mg/ml) in Dulbecco's modified Eagle's medium (DMEM)/12 with L-glutamine plus 15 mM 2-[4-(2-Hydroxyethyl)-1-piperazinyl]ethanesulfonicacid (HEPES) and DNase (20 µg/ml). Subsequently, cells were dispersed by incubation with (2 mg/ml) collagenase A, hyaluronidase (2 mg/ml), and DNase (20 µg/ml) (all from Sigma-Aldrich, Shanghai, China) in DMEM/F12 with L-glutamine plus 15 mM HEPES at 34°C for 30 min. Enzymatic digestions were stopped by brief treatment with soybean trypsin inhibitor (400 µg/ml) in DMEM/F12 supplemented with 2 mg/ml BSA. The cell suspension was centrifuged for 45 s at 50 × g. The pellets contain SCs and germ cells and were further digested with (2 mg/ml) collagenase A, (2 mg/ml) hyaluronidase, and (20 µg/ml) DNase in DMEM/F12 for 20 min at 34°C. Cell clusters were gently dispersed by homogenization using a potter. The cell suspension was filtered through a sterile (70 µm pore size) nylon mash. The cells were cultured in DMEM/F12 supplemented with 2 mM L-glutamine, 100 U/ml penicillin, 100 µg/ml streptomycin, 10 ng/ml epidermal growth factor (C600056, Sangon, Shanghai, China), 5 µg/ml human transferrin (T8158, Sigma-Aldrich), 2 µg/ml insulin (91077C, Sigma-Aldrich), 10 nM testosterone (HY-15554A, MedChem Express, Shanghai, China), 100 ng/ml follicle-stimulating hormone (8576-FS-010, R&D, Minneapolis, MN, USA), and 3 ng/ml cytosine arabinoside (C1768, Sigma-Aldrich). Cells were seeded at a density of 1–2 × 107 onto 100 mm2 Matrigel-covered dishes and incubated at 34°C in a humidified atmosphere containing 5% CO2 and 95% air. To remove germ cells and increase the purity of the SCs preparation, after 3 days of culture, the SC monolayer was subjected to hypotonic shock by incubation with 20 mM Tris–HCl (pH 7.5) for 5 min at room temperature. The hypotonic solution was replaced with a medium (lacking cytosine arabinoside). The medium was exchanged every day and the SCs used for experiments after 4 days of cultures. The identity of these cells as SCs was based on immunostaining for vimentin. The purity of the SC cultures was higher than 95%.
For experiments, SCs were seeded at greater than 90% confluent then transfected with negative control or Raptortargeting siRNA via Lipofectamine 3000 (Invitrogen). 12 h later, cells were starvated with DMEM/F12 only without any growth factors. 24 h later, cells were treated with human FGF9 (20 nM in DMEM/F12, C198, Novoprotein, Shanghai, China) for another 24 h and finally collected for indicated analyses.
Results
Deletion of raptor in SCs disrupted testicular organization and cell architecture
Histomorphological examinations showed that the SCRaptor KO tubules displayed age-dependent degeneration. By 60 dpp, most SCRaptorKO tubules were collapsed, characterized by the absence of tubular cell hierarchy and veil-like cytoplasmic extensions pointing toward the lumen and expansion of interstitia (Figure 1A). As SCs are the central organizer of the testicular architecture, we focused on the effects of Raptor deletion on SCs. Immunofluorescence of Wilms tumor 1 (WT1), an SC nuclei marker, showed that Raptor-deficient SCs did not align along the basement membrane but dispersed in the seminiferous epithelia compared with control mature SCs (Figure 1B). Transmission electron microscopy images further confirmed that Raptor-deficient SCs lost typical tripartite nucleoli, veil-like cytoplasmic extensions, and intrinsic cell junctions and had a distinct nucleic morphology; they were detached from the basement membrane (Figure 1C). These changes might indicate an immature configuration in the Raptor-deficient SCs. Notably, gap junction protein GJA1 exhibited a regular and continuous distribution in the control testes, whereas in the SCRaptorKO testes, a discontinuous distribution at 15 dpp and a severely disrupted and dot distribution in adulthood were observed (Figure 1D). This resembled its counterpart among granulosa cells in the adult WT ovary (Figure 1D). When calculating the ratio of the short to the long axis of the adult SCRaptorKO testes, we found that it was close to that of adult WT ovaries ( Supplemental Figure S1 (supplemental_figures_ioac104.pdf)). These observations together suggested that Raptordeficient SCs lost their mature characteristics and may present an immature state, thus disrupting testis development.
Loss of raptor induces SCs into an undifferentiated state
To define the extent of dedifferentiation of the Raptormutant SCs and explore the possibility of undifferentiated SCs reversing into ovarian somatic cells, we performed whole-transcriptome sequencing and analyses of control and SCRaptorKO testes mice at 5 dpp when tubular morphological changes were initiated but without sharp degeneration. This revealed a large gene set upregulated in the Raptor-mutant testes, and surprisingly many of them were expressed in common between Raptor-mutant testes and WT ovaries (Figure 2A). We observed many genes upregulated were involved in sex-related biological processes, including male and female gonad development, sex somatic cell fate determination, response to estrogen, and regulation of canonical WNT signaling (Figure 2B). Of note, the genes characteristic of testicular Sertoli and Leydig cells and essential for testis development and function, i.e., Sox9, Sox8, Amh, Clu, Dhh, Gata1, Inha, Hsd17b3, and Cyp26b1 [19, 40–46], were significantly downregulated, while the genes well-characterized as ovary markers, i.e., Foxl2, Lhx9, Fst, Esr2, and Cyp1b1 [16, 18, 47–49], were significantly upregulated (Figure 2B; Dataset S1). Among the most significantly enriched Gene Ontology (GO) terms associated with the altered transcripts were “cell adhesion”, “regulation of cell proliferation”, “regulation of apoptotic process”, “regulation of gene expression”, and “cell differentiation” (Figure 2C). The top GO term “cell adhesion” was of the highest significance and associated with numerous transcripts. These data demonstrated the loss of unique SC characteristics upon loss of Raptor and simultaneously indicated a general activation of the ovarian genes.
Using single-cell RNA seq of Nr5a1-expressing gonadal somatic cells [50], Stévant et al. deconvoluted the gene expression dynamics during three categorized periods of supporting cell lineage specification, namely: stage a, XY/XX early progenitor cells; stage b, pre-Sertoli/pre-granulosa cells; and stage c, Sertoli/granulosa cells. We then compared all the genes detected in our sequencing data with all in the Stévant, aiming to (1) determine the extent of dedifferentiation in gonadal somatic cell lineages and (2) explore the possibility of gonadal sex reprogramming in the Raptor-mutant testes. We found 204 genes in common (Figure 2D). In the Raptormutant testes, two XX early progenitor cell markers (Ibsp and Cnr1), four pre-granulosa cell markers (Lhx9, Thsd7b, Fgfr2, and Fst), and 43 granulosa cell markers were markedly upregulated, while seven XY early progenitor cell markers, five pre-SC markers, and 114 mature SC markers were markedly downregulated (Figure 2D; Dataset S2). These data globally suggested downregulation of the mature SC markers whereas upregulation of the markers for ovarian mature and somatic progenitor cells in the Raptor-mutant testes, reminiscent of dedifferentiation of Raptor-mutant SCs.
Expression of LHX9, but not FOXL2, was increased in SCRaptorKO testes
To further validate the RNA-seq data and examine the possible gonadal somatic cell fate transition, we performed immunofluorescent staining of LHX9, expressed in progenitor cells for both SCs and granulosa cells [51, 52]. LHX9 was strongly expressed in XX gonads at E12 when supporting cell commitment is undergoing and then largely decreased at E14.5 when cell commitment is complete. In the XY gonad, the LHX9 level was lightly detectable (Figure 3A). These results are consistent with previous reports. Based on this, we characterized expression profiles in paired testes from 5 to 60 dpp, which showed that LHX9 expression was robustly increased in the degenerating/degenerated tubules in the SCRaptorKO testes in an age-dependent manner (Figure 3B). We then profiled the expression of another granulosa cell marker, FOXL2, which is required for ovary commitment, and its deficiency in adult ovaries triggers a testis-like reorganization [16, 17]. Quantitative reverse-transcription polymerase chain reaction (qRT-PCR) analysis revealed a striking elevation of Foxl2 mRNA level in the SCRaptorKO testes, validating its change in the sequencing data. An elevation of Wnt4 while downregulation of Dmrt1 was also observed in the SCRaptorKO testes. Unexpectedly, we only observed a subtle increase in the FOXL2 protein level in the SCRaptorKO testes ( Supplemental Figure S2 (supplemental_figures_ioac104.pdf)), indicating that FOXL2-positive terminal mature granulosa cells were not induced in the SCRaptorKO testes. Alternatively, there might be a post-transcriptional regulation. These results provided evidence that the markers for ovarian somatic progenitor or granulosa cells are upregulated in the SCRaptorKO testes.
Raptor-deficient SCs exhibit some features resembling ovarian granulosa cells and are functional in producing estrogen
We observed that a Leydig cell marker 3β-HSD2, a key enzyme in testosterone synthesis, was largely upregulated in the Raptor-mutant testes ( Supplemental Figures S3 (supplemental_figures_ioac104.pdf) and S4 (supplemental_figures_ioac104.pdf)). We previously thought this might be a negative feedback phenomenon due to impaired SC function. It is noteworthy that both adult Leydig and theca cells are thought to be recruited from surrounding progenitor populations [53, 54]. In the ovary, granulosa cells produce estrogen, and theca cells are induced during follicle growth, likely by signals from granulosa cells. Together with oocytes, these three comprise the functional unit of the ovary. Therefore, we asked whether Raptor-deficient SCs can be transdifferentiated into ovarian somatic cells (e.g., granulosa cell), whether they were functional (e.g., producing estrogen), and whether oocytes were formed. To this end, we performed immunofluorescent staining of aromatase/CYP19α1, an enzyme in granulosa cells important for estrogen synthesis that catalyzes testosterone to estradiol, as well as 3β-HSD2, which is expressed in ovary theca cells ( Supplemental Figure S4 (supplemental_figures_ioac104.pdf)). α-smooth muscle actin (α-SMA) was also labeled to mark the base membrane. As expected, their expressions were robustly enhanced in the Raptor-mutant testes. More importantly, many signals were localized within the basement membrane (Figure 4A), contrary to their counterparts, which were restricted to the interstitium in the control testes. Interestingly, some aromatase-positive and 3β-HSD2-positive intratubular cells in the Raptor-mutant testes slightly resembled the WT ovary granulosa and theca cells (Figure 4A, inserts). We further performed co-immunofluorescence of aromatase and 3β-HSD2 with GATA-4, an SC marker, respectively, and found that aromatase/GATA-4 and 3β-HSD2/GATA-4 each co-localized in cells (Figure 4B, inserts). Some 3β-HSD2/GATA-4-positive cells appeared with a spindle shape analogous to WT theca cells (Figure 4B, arrowhead in the inserts). Furthermore, the progenitor cell marker LHX9 was also co-expressed with GATA-4 (Figure 4B). However, as Leydig cells also express 3β-HSD2 and most of the 3β-HSD2-positive intratubular cells were located in the tubules with disrupted tubular integrity (discontinued SMA expression, asterisks in Figure 4A), they could be derived from expanded Leydig cells that invaded the tubule remnants due to SC dysfunction. Overall, these results suggested some signs that Raptor-mutant SCs can be dedifferentiated into progenitor cells, which then turn to express enzymes required for ovarian somatic cell functions.
We next assessed female steroidogenesis by qRT-PCR analyses of the associated enzymes. The results revealed a significant elevation in the mRNA levels of estrogen associated enzymes ( Supplemental Figure S4 (supplemental_figures_ioac104.pdf)), including estrogen-producing aromatase/Cyp19a1 and Hsd17b1, and estradiol-metabolizing Cyp1a1 and Cyp1b1, as early as 5 dpp; these levels were further elevated at 15 dpp, although Cyp1a1 and Cyp1b1 decreased at 60 dpp (Figure 5A). Notably, when introducing age-match WT ovaries for comparison, we observed a similar expression pattern for most enzymes between the Raptor-mutant testes and the WT ovaries, especially for aromatase, Cyp1a1, and Cyp1b1. Consistent with these enzyme changes, an ELISA assay with whole gonad extracts showed that estradiol level was raised in the adult mutants relative to the adult controls (Figure 5B). The level of androgenic enzyme Hsd17b3 was increased after 15 dpp, and that of Sr5d3, which encodes 5-α-reductase that catalyzes testosterone to generate active dihydrotestosterone [55], was relatively steady in the mutants (Figure 5A). Nevertheless, the testosterone level in the mutant gonads was significantly decreased (Figure 5C). Shortages of 17-OH-progesterone and androstenedione (Figure 5D and E), the two last steroids in the testosterone synthesis pathway ( Supplemental Figure S4 (supplemental_figures_ioac104.pdf)), primarily accounted for this decrease. We did not detect expression of the oocyte-specific zona pellucida protein ZP3 in the mutant XY gonads at any ages detected (Figure 5F), indicating male germ cells were not yet feminized. Overall, these data suggested that Raptor-deficient SCs present some ovarian somatic cell features and can produce estrogen.
Knockdown of raptor attenuates the promoting effects of FGFf9 on cell junctions among SCs
FGF9/FGFR2 signaling plays a crucial role in male sex determination [56]. GO analysis of our RNA-seq data also revealed that FGF signaling was associated with notable changes in transcripts. qRT-PCR analyses with the 5 dpp paired gonads confirmed that mRNA levels of Fgf9 and Flrt1 were downregulated, while those of Fgf7, Fgfr2, Fgfr3, and Fat4 were upregulated and that of Fgf10 was unchanged (Figure 6A). To determine whether Raptor was involved in FGF9 signaling, immature primary SCs were isolated from 5 dpp WT mouse testes, transfected with siRNA targeting Raptor, and treated with FGF9. qRT-PCR results showed that Raptor mRNA expression was effectively eliminated by the siRNA, but Raptor knockdown hardly changed mRNA levels of the detected FGF pathway members under physiological conditions (Figure 6B). Upon FGF9 stimulation, the mRNA levels of FGF pathway members were significantly raised, which were counteracted by Raptor knockdown, although Raptor mRNA levels remained unchanged in response to FGF9 (Figure 6B). This indicated that Raptor downregulation attenuated FGF9 signaling in vitro. Formation of the blood-testis barrier (BTB) is widely accepted as a hallmark of SC maturation [36]. We next asked whether Raptor plays a role in the FGF9-driven formation of cell junctions among SCs. Immature SCs transfected with control or Raptor-siRNA were normalized in a basic medium and treated with FGF9. Western blotting showed that the protein levels of BTB-associated Cingulin, β-catenin, and ZO-2 were decreased in the knockdown group when cultured in a basic medium. FGF9 treatment enhanced their protein expression, while expression was attenuated by Raptor knockdown (Figure 6C). To confirm this further, we performed immunofluorescent staining. Consistent with our previous report showing aberrant Cingulin expression and distribution in the SCRaptorKO testes [35], Raptor knockdown also markedly decreased the cytoplasmic spreading of Cingulin, as well as its accumulation around the interacting boundary among primary SCs in response to FGF9 stimulation (Figure 6D). FGF9 treatment induced nucleus-to-cytoplasm transport of β-catenin and gathering in the contact areas among SCs; this was blocked by Raptor knockdown (Figure 6D). Similar results were observed with ZO-2. FGF9 treatment also induced cytoplasmic spreading of Vimentin, the SC intrinsic cytoskeleton protein, but it appeared unchanged after Raptor knockdown (Figure 6D). These data suggested that Raptor downregulation did not affect the cytoskeletal organization in SCs, but disrupted the establishment of cell junctions among SCs by impairing the gathering and arrangement of BTB-related proteins. This may present a potential mechanism in which Raptor acts downstream of FGF9 signaling to promote SC maturation and maintain their identity.
Discussion
To our knowledge, this is the first report showing that one of the key components of the mTOR complex, more precisely, mTORC1, is essential for the maintenance of SC differentiation identity after birth. Its deficiency induced SCs dedifferentiation into a progenitor state, thus depriving SC maturation characteristics, including specific cell junctions between SCs and between SCs and male germ cells, and unique nucleic morphology and cell location. We also observed that Raptor-mutant SCs exhibited some ovarian somatic cell-like features, including actively producing estrogens. Yet, we did not observe terminally feminized ZP3-positive follicles in the XY Raptor-mutant gonads.
A recent study showed that conditional knockout of mTOR in primordial follicles reversed granulosa cells into Sertoli-like cells, along with several testicular properties in ovary tissues [38]. Though involving alternate gonads, these two studies support each other and together establish the critical role of mTOR or its binding partner in maintaining supporting cell fate and gonad development and function. However, since mTOR is present in both mTORC1 and mTORC2, the deletion of mTOR should block both pathways. Thus it is unclear whether the ovary-to-testis reversal observed in that study was due to single inactivation of either mTORC1 or mTORC2, or both, yet this issue was not clearly addressed in that study. Our study provided evidence that blocking mTORC1 alone (by Raptor deletion) can induce an SC dedifferentiation and a subtle ovary-like reorganization.
Our previous study revealed that mutant mice with SC-specific inactivation of Rheb, a direct upstream activator of mTORC1, displayed almost normal tubular histomorphology and functions [35]. Thus, we deemed that Raptor may function independently of canonical Rheb/mTORC1 signaling to control postnatal SC differentiation and testis development. Evidence supporting this notion also came from another study demonstrating that mTORC1-independent (“free”) Raptor negatively regulated hepatic Akt activity and lipogenesis [57], suggesting Raptor does not simply function as a scaffold in mTORC1. In addition, we previously showed that the ribosomal protein S6, a downstream target of canonical Rheb/mTORC1 signaling, was exclusively activated in differentiating spermatogonia but was scarcely detectable in SCs in postnatal testis [32], suggesting canonical mTORC1 signaling plays a limited role in postnatal SCs. Moreover, it is worthwhile to note that the International Mouse Phenotyping Consortium recently published a resource of targeted mutant mouse lines for 5061 genes [58], which reported that mTOR knockout mice showed preweaning lethality and died as early adults (ID: 1928394), whereas Raptor knockout mice exhibited embryonic lethality before tooth bud stage and died at E12.5 (ID: 1921620). This suggested that Raptor is more important than mTOR for embryo development and controls critical biological processes independently of mTOR. These observations support ascribing SC dedifferentiation and potential cell fate transition (SC to granulosa cell) in the Raptor mutant gonads to inactivating “free” Raptor rather than canonical mTORC1 signaling. Further study of conditional ablation of mTOR in SCs, or alternatively, Raptor in oocytes, will help define whether “free” Raptor exists and what role it plays in gonadal sex differentiation, identity maintenance, and gonad development.
As SC dedifferentiated, the size of adult Raptor mutant gonads was greatly decreased (Figure 1B), with an adult testis mass of only ∼6% of the controls. This was not observed in the Dmrt1 mutant testes or the mTOR or Foxl2 mutant ovaries. We ascribed this specific phenotype mainly to Raptor deletion-reduced proliferation of three types of cells: (1) SCs. Besides maintaining SC identity, we believe that Raptor is also essential for the second intensive proliferation of SCs, which takes place in the first two weeks after birth in mice and is vital for the expansion and mature of testicular cords [59]. Indeed, Raptor was intensively expressed in mouse SCs across embryonic and postnatal periods ( Supplemental Figure S5 (supplemental_figures_ioac104.pdf)). (2) Male germ cells, all biochemical activities of which exclusively rely on SCs; evidence for this came from the observation obtained by excessive germ cell loss in histology and Ki67 staining showing insufficient germ cell proliferation. (3) Granulosa cells. Potentially transdifferentiated granulosa-like cells also likely required Raptor to support the expansion of their population, which is required for follicle growth and ovary development. In addition, increased apoptotic activity was probably a second cause of gonad shrinking, and perhaps altered gene expression activity was the effect of Raptor deletion (Figure 2C).
In summary, we established that Raptor is required for stabilizing postnatal SC characteristics and testicular architecture. Ablation of Raptor converted postnatal SCs into an undifferentiated state, and they presented some ovarian somatic cell-like features and can produce estrogen. Our study extends the current knowledge of maintenance of postnatal male somatic cell characteristics and testis development. Although we provided in vitro evidence suggesting Raptor is involved in FGF9 signaling in the regulation of SC maturation, an in vivo study is needed to unravel further whether Raptor acts downstream of FGF9 signaling or in parallel to stabilize the male route, dependently or independently of mTOR. This would greatly enrich our understanding of gonadal sex differentiation, maintenance, and gonad development and contribute to new ideas for the prevention and therapy for hermaphroditism or infertility.
Data availability
RNA-seq data have been deposited in the NCBI SRA database (PRJNA767807). Other data are available upon reasonable request.
Funding
This work was supported by the National Natural Science Foundation of China [32171113, 31972910, 82071723, 81871162], Guangdong Basic and Applied Basic Research Foundation [2021A1515010774, 2021B1515020069, 2019A1515010755], Guangdong Medical Science and Technology Research Foundation [A2020080], The Science and Technology Project of Guangzhou [202002030063], The Health Science and Technology Project of Guangzhou [20201A011016, 201904010024], and The Scientific Program of Dongguan People's Hospital (K202022).