Defects in efferent duct multiciliogenesis underlie male infertility in GEMC1-, MCIDAS- or CCNO-deficient mice

Spermatogenesis is a highly regulated developmental process that generates haploid sperm. To become capable of fertilization, spermatozoa must detach from the seminiferous epithelium, enter the tubule lumen and travel through the epididymis, a process that promotes maturation and motility. The efferent ducts (ED) connect the rete testes to the epididymis and are important for sperm concentration, reabsorbing most of the luminal seminiferous fluid (Clulow et al., 1998). The ED epithelia contain poorly characterized multiciliated cells (MCCs) that mobilize luminal fluids through the action of hundreds of motile cilia on the apical surface of these cells (Joseph et al., 2011). The luminal turbulence that is generated by the MCCs of the ED has been proposed to prevent the agglutination of spermatozoa and promote fluid reabsorption by non-MCCs in the ED (Yuan et al., 2019).

Differentiation of airway MCCs is initiated in part through the action of the miR-449/34 family of miRNAs that downregulate numerous genes, including Cp110 (Ccp110), an inhibitor of cilia assembly, as well as Notch genes and Dll1, that together inhibit MCC differentiation (Kyrousi et al., 2015; Lafkas et al., 2015; Marcet et al., 2011; Song et al., 2014; Tsao et al., 2009; Zhou et al., 2015). This is followed by the activation of a transcriptional program by the geminin family members GEMC1 (also known as GMNC) and MCIDAS that interact with E2F4/5-DP1 and are required for the generation of MCCs in fish, frogs and mammals (Arbi et al., 2016; Balestrini et al., 2010; Boon et al., 2014; Chong et al., 2018; Danielian et al., 2007; Kyrousi et al., 2015; Lu et al., 2019; Ma et al., 2014; Stubbs et al., 2012; Terré et al., 2016; Zhou et al., 2015). In addition, the transcription factors MYB (Pan et al., 2014; Tan et al., 2013), FOXJ1 (You et al., 2004; Yu et al., 2008), RFX2/3 (Chung et al., 2012; Didon et al., 2013; El Zein et al., 2009) and TAp73 (transcriptionally active isoform encoded by the Trp73 gene) (Marshall et al., 2016; Nemajerova et al., 2016), as well as the atypical cyclin CCNO (Funk et al., 2015; Núnez-Ollé et al., 2017; Wallmeier et al., 2014), are necessary to promote deuterosome-mediated centriole amplification and the generation of multiple motile cilia. Mutations in MCIDAS or CCNO underlie reduced generation of multiple motile cilia (RGMC), a rare ciliopathy that is characterized by hydrocephalus, mucus accumulation in the respiratory system and reduced fertility, all presumably due to defects in MCC differentiation (Amirav et al., 2016; Boon et al., 2014; Funk et al., 2015; Wallmeier et al., 2014).

Male and female infertility occurs in a number of mice that are mutant for genes involved in MCC development, including Ccno, the miR-34b/c and miR-449a/b/c (miR-dKO) loci, Trp73, E2f4/5, Mcidas and Gemc1. When this has been addressed in females, it appears to be because of the loss of MCCs in the oviducts (Chen et al., 1998; Lu et al., 2019; Marshall et al., 2016; Núnez-Ollé et al., 2017; Terré et al., 2016; Wu et al., 2014). However, in males, the origin of the defect has not been clearly established in all cases. Both miR-34/449 and Trp73 are expressed in the testes, and miR-dKO mice are impaired in meiosis and spermiogenesis and exhibit a ‘nearly empty’ seminiferous tubule phenotype (Comazzetto et al., 2014; Holembowski et al., 2014; Wu et al., 2014; Yuan et al., 2015). Trp73-deficient mice showed a similar seminiferous tubule phenotype (Tomasini et al., 2008), as well as degeneration of the Sertoli cells (SCs) that support spermatid development and ensure integrity of the blood-testes barrier (Holembowski et al., 2014). However, a conditional knockout of 3 out of 4 E2f4 and E2f5 alleles in the EDs, but not spermatogonia or spermatocytes, as well as the Foxj1-Cre (MCC specific) mediated miR-dKO deletion, phenocopied the seminiferous tubule phenotype (Danielian et al., 2016; Yuan et al., 2019). This indicates that ED defects, particularly in the MCC population, are likely to be sufficient to cause testicular atrophy and male infertility.

Here, we show that male mice lacking Gemc1, Mcidas or Ccno exhibited a testes phenotype similar to miR-dKO, Trp73 deletion or loss of multiple E2f4 and E2f5 alleles (Comazzetto et al., 2014; Danielian et al., 2016; Holembowski et al., 2014; Inoue et al., 2014; Terré et al., 2016; Wu et al., 2014; Yuan et al., 2015). We found that Gemc1-, Mcidas- or Ccno-deficient mice exhibited luminal dilation of the seminiferous tubules, atrophy of the germinal epithelium, rete testes dilation, SC degeneration and failure of spermatozoa to enter the epididymis, instead accumulating in the EDs. In each case, defects in MCC maturation were clearly evident. Moreover, we show that, similar to Foxj1, Trp73 expression is high in the EDs and is dependent on GEMC1 but not on MCIDAS or CCNO, further establishing the distinct temporal roles of these factors. Our results demonstrate that GEMC1, MCIDAS and CCNO are required for ED MCC differentiation, and this further underscores that these defects are likely to be the primary cause of male infertility in several mouse lines with MCC defects, and potentially in human RGMC patients with mutations in MCIDAS or CCNO.

We analyzed adult testes of Gemc1^−/− mice over the first three months and found no consistent changes in size and weight compared with wild-type (Wt) or Gemc1^+/− littermates when normalized to body size (Fig. 1A). Histological evaluation during the first semi-synchronous wave of spermatogenesis revealed no overt differences between Wt, Gemc1^+/− or Gemc1^−/− testes during the first 20 days post partum (p0-p20) (Fig. 1B,C). However, by p27-p35, the thinning of the seminiferous germinal epithelia became obvious, corresponding to the first appearance of elongating spermatids (ES) (Fig. 1B,D). Despite the reduction in cellularity, numbers of mitotic cells, dead cells, meiotic progression and levels of hormonal gene expression were normal in Gemc1^−/− mice (Fig. S1A-E).

Consistent with histological observations, Gemc1 messenger RNA (mRNA) expression peaked around p27, although peak levels were considerably lower than in the trachea, which contains a large number of MCCs (Fig. 1E,F). As the peak of Gemc1 expression and appearance of seminiferous tubule dilation correlated with late stages of spermatogenesis (Fig. 1E), we isolated and quantified enriched populations of testicular cell types [leptotene-zygotene (LZ), pachytene-diplotene (PD), round spermatids (RS) and ES] by fluorescence-activated cell sorting (FACS) (Fig. S1F). Gemc1 mRNA was enriched in RS and ES populations compared with the germ cell pellet (Fig. 1G) and a significant reduction in RS and ES populations was observed in testes from Gemc1^−/− mice (Fig. 1H), compared with similar numbers of prophase cells (LZ and PD). This suggested that GEMC1 may support late stages of spermatogenesis through the control of transcription, but the prominent role of GEMC1 in MCC differentiation and the phenotypic similarity to miR-dKO mice prompted us to consider that these effects may be secondary to defects in MCC function in the EDs (Comazzetto et al., 2014; Yuan et al., 2019).

In conditional E2f4^−/− E2f5^+/− or miR-dKO mice generated with Cre transgenes active in the EDs of the epididymis and not the testes, defects in MCC formation in the EDs and signs of fluid backpressure, namely seminiferous tubule and rete testes dilation, have been described (Danielian et al., 2016; Yuan et al., 2019). We examined the transcriptional activation of GEMC1 target genes in the testes and, in contrast to tissues containing MCCs, we did not observe any significant alterations in Ccno, Mcidas, Foxj1, Trp73 and Cdc20b in the absence of GEMC1 (Fig. 2A). This was further confirmed at the protein level for TP73 (Fig. 2B). As male mice that are mutant for Mcidas and Ccno, which, like GEMC1, play key roles in MCC development, were also infertile (Lu et al., 2019; Núnez-Ollé et al., 2017), we histologically examined the testes of these animals in parallel to Gemc1^−/−. In each case they exhibited a similar seminiferous tubule phenotype characterized by reduced cellularity and luminal dilation (Fig. 2C,D). Moreover, extensive rete testes dilation was apparent (Fig. 2E), indicative of fluid backpressure.

Previous work has reported seminiferous tubule dilation accompanied by extensive SC degeneration and spermatid detachment due to loss of Trp73, which is also required for MCC generation (Holembowski et al., 2014; Inoue et al., 2014; Marshall et al., 2016; Nemajerova et al., 2016; Tomasini et al., 2008). Immunostaining of vimentin-containing intermediate filaments, one of the main components of the SC cytoskeleton (Aumüller et al., 1988), revealed marked structural abnormalities in the SCs of Gemc1-, Mcidas- and Ccno-deficient testes (Fig. 2F,G). This was characterized by abnormally shorter and thinner cytoplasmic projections and the appearance of SC-only seminiferous tubules (Fig. 2H) and was accompanied by extensive spermatid detachment in Gemc1^−/− mice (Fig. 2I). In contrast, the SCs of Wt mice extended long cytoplasmic arms in contact with germ cell populations (Fig. 2F,G,I). Thus, Gemc1, Mcidas and Ccno mutants phenocopied the loss of Trp73, E2f4/5 and miR-dKO, exhibiting dilated seminiferous tubules and SC degeneration, with the loss of GEMC1 having little impact on the expression of known MCC transcriptional targets in the testes.

Spermatozoa enter the caput epididymis but remain immature until reaching the cauda epididymis, where they acquire motility and fertilization competency (Fig. 3A). Gemc1^−/− and Ccno^−/− epididymides appeared to be paler, smaller and thinner than Wt (Fig. 3B). Histological analysis of Gemc1^−/−, Mcidas^−/− and Ccno^−/− revealed no detectable sperm in the caput, corpus or cauda epididymides in any case, in contrast to Wt, in which sperm were abundant in all sections (Fig. 3C). Consistent with this, the lumen of Gemc1^−/−, Mcidas^−/− and Ccno^−/− epididymides was often filled with amorphous periodic acid–Schiff stain (PAS)-positive material, a phenotype that is typically observed in the absence of spermatozoa (Abe and Takano, 1988). Gemc1^−/−, Mcidas^−/− and Ccno^−/− males thus have an apparently complete block in transit of spermatozoa to the epididymis.

As the EDs are crucial for sperm transit and contain MCCs, we examined Gemc1, Mcidas and Ccno expression in the testes, cauda epididymis and EDs. The expression of Gemc1 was considerably higher in the EDs than in testes (Fig. 3D), and similar to its levels in the trachea. A similar pattern was observed for its target gene Mcidas (Fig. 3E), as well as Trp73 (Fig. 3F). In contrast, Ccno was marginally upregulated in the ED compared with the testes and further increased in the cauda epididymis (Fig. 3G), suggesting that it may play roles independently of GEMC1 and MCIDAS, which both appear to be restricted to MCCs in other tissues (Kyrousi et al., 2015).

We next examined the expression levels of key MCC regulators in Gemc1^−/− EDs. Although Ccno levels were moderately increased, other known target genes including Mcidas, Cdc20B and Foxj1, as well as Trp73, were strongly downregulated (Fig. 3H), consistent with what has been observed in the trachea (Terré et al., 2016). Similar to what has been reported for miR-dKO, or Trp73- and E2f4/5-deficient mice, PAS staining of histological sections revealed sperm agglutinations in the EDs of Gemc1^−/−, Mcidas^−/− and Ccno^−/− mice; they are normally not detected in Wt animals because of the rapid transit through this region (Fig. 3I). Therefore, regardless of gene expression patterns in the testes, spermatozoa were unable to enter the epididymis of Gemc1^−/−, Mcidas^−/− and Ccno^−/− males, corresponding with defects in ED formation or function. In the case of GEMC1, this appeared to reflect its transcriptional role, similar to other MCC-containing tissues.

Staining of motile cilia with anti-acetylated tubulin (ac-tub) antibodies revealed that columnar MCCs were absent in Gemc1^−/− EDs (Fig. 4A). A less severe phenotype was observed in both Mcidas^−/− and Ccno^−/− mice, as some cells with MCC morphology were apparent, but few cilia visible (Fig. 4A). In all three mutants, ac-tub staining was observed in spermatozoa agglutinations in the central cavity of the ED.

The ED epithelium of Gemc1^−/− mice was thinner than that of Wt, Mcidas^−/− and Ccno^−/− mice (Fig. 4A), so we examined the expression of the FOXJ1 and TP73 transcription factors that are crucial for MCC formation in other tissues (Marshall et al., 2016; Nemajerova et al., 2016; You et al., 2004). Cells lining the EDs were strongly immunopositive for FOXJ1 in Wt and Ccno^−/−, but staining was completely absent in Gemc1^−/− (Fig. S2A). Similarly, TP73-positive cells were absent in Gemc1^−/−, but readily identifiable in both Mcidas^−/− and Ccno^−/− mice (Fig. 4B). We, and others, have previously shown that the transient transfection of GEMC1 can activate early MCC factors, including both MCIDAS and FOXJ1, in AD293 cells (Arbi et al., 2016; Lu et al., 2019; Terré et al., 2016). Transient overexpression of GEMC1 led to a strong increase in Trp73 mRNA and protein levels, in contrast to CCNO overexpression that did not influence its levels (Fig. 4C), demonstrating that GEMC1 can activate TP73 expression.

Collectively, our experiments demonstrated that male mice lacking Gemc1, Mcidas or Ccno were infertile and exhibited a testes phenotype closely resembling that of Trp73-deficient mice, or ED-specific miR-dKO or E2f4^−/−E2f5^+/− deletions (Comazzetto et al., 2014; Holembowski et al., 2014; Yuan et al., 2019). The impairment of the MCC transcriptional program and agglutination of spermatozoa in the EDs appears to be sufficient to generate backpressure that prevents spermatozoa from entering the epididymis (Yuan et al., 2019). We propose that impaired MCC function is likely the dominant cause of infertility in Gemc1^−/−, Mcidas^−/− and Ccno^−/− mice, and potentially in male RGMC patients that harbor mutations in MCIDAS or CCNO (Fig. 4D). The seminiferous tubule and rete testes dilation, degeneration of the SC support structures and failure of spermatozoa to enter the epididymis is most likely a secondary effect of backpressure caused by loss of MCC function that results in the agglutination of spermatozoa in the EDs. We cannot, however, rule out a role for these factors in other cell types of the ED that are associated with reabsorption of seminal vesicle fluids, as has been proposed for both Esr1 and E2f4/5 mutants (Danielian et al., 2016; Joseph et al., 2011). However, we observed normal staining of the E2F4/5-dependent AQP1 marker of secretory cells in CCNO and GEMC1 mutants, suggesting that E2F4/5-mediated transcription is not severely affected in non-MCC cell types of the ED in these mice (Fig. 4E and Fig. S2B) (Danielian et al., 2016).

The roles of GEMC1 and MCIDAS in multiciliogenesis are primarily related to transcription and our data lend further support to them acting in a stepwise manner, with GEMC1 playing a key role in specification and MCIDAS required for ciliogenesis (Fig. 4B). However, the function of CCNO remains enigmatic. Recent work demonstrated that the mitotic oscillator and fine-tuning of CDK activity is required for stepwise deuterosome-mediated centriole amplification in MCCs (Al Jord et al., 2017; Vladar et al., 2018). Given that CCNO interacts with CDK1 and CDK2 (Roig et al., 2009), and CCNO-deficient mice exhibit abnormal deuterosomes (Funk et al., 2015), it appears plausible that one of its primary functions is CDK regulation during deuterosome formation. Future work will be needed to understand precisely how GEMC1, MCIDAS and CCNO regulate different aspects of the transcriptional response and deuterosome formation, as well as their potential roles in other tissues.

Gemc1^−/−, Mcidas^−/− and Ccno^−/− mice on a mixed C57BL/6-129SvEv background have been described previously (Lu et al., 2019; Núnez-Ollé et al., 2017; Terré et al., 2016). Animals were maintained in accordance with the European Community (86/609/EEC) guidelines in the specific-pathogen free (SPF) facilities of the Barcelona Science Park (PCB) and in SPF facilities following guidelines of the Biological Resource Center of the Agency for Science, Technology and Research (A*STAR) of Singapore. Protocols were approved by the Animal Care and Use Committee of the PCB (IACUC; CEEA-PCB) in accordance with applicable legislation (Law 5/1995/GC; Order 214/1997/GC; Law 1201/2005/SG) and ethical guidelines of the Singapore National Advisory Committee on Laboratory Animal Research. All efforts were made to minimize use and suffering. Sample sizes were not defined to detect pre-determined effect size. Animals were not randomized, were identified by genotyping for analysis and were all males, and ages are indicated in the figure legends.

Testes and epididymis were harvested and fixed in 4% paraformaldehyde (PFA) or Bouin's solution (Electron Microscopy Sciences) overnight at 4°C, dehydrated using a graded series of ethanols, transferred to xylene and embedded in paraffin wax. Sections were cut at 3-5 µm thickness, dewaxed and stained with Hematoxylin and Eosin (H&E) and PAS (Sigma-Aldrich) using a CoverStainer (Dako-Agilent) following the manufacturer’s instructions. For immunohistochemistry (IHC), epitope retrieval was performed using citrate pH 6 buffer in an autoclave at 121°C for 12 min. Washings were performed using the Wash Solution AR (AR10211-2, Dako, Agilent). Quenching of endogenous peroxidase was performed by 10 min of incubation with Peroxidase-Blocking Solution at room temperature (RT) (S2023, Dako, Agilent). Non-specific unions were blocked using 5% of goat normal serum (16210064, Life Technologies) mixed with 2.5% bovine serum albumin (BSA) diluted in wash buffer for 60 min at RT. Ac-tub was also blocked with M.O.M.™ Blocking Reagent (MK-2213, Vector Laboratories) following the manufacturer’s instructions. IHC was performed using an Autostainer Plus (Dako, Agilent) for p73 antibody (90 min at RT) and manually for ac-tub staining (overnight at 4°C). The secondary antibodies used were a BrightVision poly-HRP-anti-rabbit IgG biotin-free (DPVR-110HRP, Immunologic), ready to use for 45 min at RT, and a polyclonal goat anti-mouse immunoglobulins/HRP (P0447, Dako) diluted at 1:100 for 30 min at RT. Antigen–antibody complexes were revealed with 3-3′-diaminobenzidine (K346811, Dako), with the same time exposure (10 min). Sections were counterstained with Hematoxylin (CS700, Dako, Agilent) and mounted with toluene-free mounting medium (CS705, Dako, Agilent) using a Dako CoverStainer. Specificity of staining was confirmed by omission of the primary antibody or staining with rabbit IgG, polyclonal-isotype control (ab27478, Abcam) or a mouse IgG-isotype control (ab37355, Abcam). For γH2AX and pH3S10, sections were incubated with primary antibody overnight at room temperature after quenching endogenous peroxidase using 0.6% H₂O₂ (vol/vol) in methanol for 10 minutes. Slides were washed and labeled with the Vectastain Elite kit (PK-6100, Vector Labs) and immunoreactive signals visualized with the VIP substrate kit (SK-4600, Vector Labs) using the manufacturer's protocol. Sections were counterstained with 0.1% (wt/vol) Methyl Green (67060, Merck), dehydrated and mounted in DPX mountant (06522, Merck). Brightfield images were acquired using a NanoZoomer-2.0 HT C9600 digital scanner (Hamamatsu) equipped with a 20× objective. All images were visualized using the NDP.view2 U123888-01 software (Hamamatsu Photonics). All images were visualized with a gamma correction set at 1.8 in the image control panel of the NDP.view2 U12388-01 software. For TUNEL staining, testes sections were labeled with In Situ Cell Death Detection Kit (TUNEL) (11684795910, Roche) according to the manufacturer's instructions.

Primary antibodies used in the publication were: mouse anti-ac-tub (6-11B-1, T6793, Sigma-Aldrich, 1:1000); rabbit anti-vimentin (EPR3776, ab92547, Abcam, 1:2000); rabbit anti-vinculin (EPR8185, ab129002, Abcam, 1:2000); rabbit anti-TP73 (EP436Y, ab40658, Abcam, 1:50); rabbit anti-Myc (ab9132, Abcam, 1:1000); mouse anti-FLAG (M2, F3165, Sigma-Aldrich, 1:1000); rabbit anti-AQP1 (AB2219, Merck, 1:1000).

All antibodies have been previously validated and were subjected to additional controls. The ac-tub antibody has been used widely in the field and recognizes multiciliated cells that are absent in Gemc1-deficient mice (Terré et al., 2016). The AQP1 and vimentin antibodies have been validated in knockouts and extensively in publications (see Merck and Abcam, respectively). The TP73 antibody was validated in knockout animals (Marshall et al., 2016) and the signal is in the expected cell types and correlates with the mRNA level in our experiments. The vinculin antibody recognizes a band of the correct size and was used as a loading control. Myc and FLAG epitope antibodies have been validated extensively and recognize proteins of the correct size only when tagged complementary DNAs (cDNAs) are expressed. SCP1/3, γH2AX, FOXJ1 and pH3S10 antibodies (Table S1) have been validated (see manufacturer or CiteAb.com) and we have previously reported their use for meiotic staging or testes characterization (Marjanović et al., 2015).

Histological sections from testes, epididymis and EDs were imaged using the digital slide scanner Nanozoomer 2.0HT (Hamamatsu) and analyzed using the NDP.view2 free software (Hamamatsu). Image analysis and quantification of IHC were performed with the TMARKER free software (GitHub). For quantitation of percentage of lumen space, tubule area and lumen area were analyzed for ten tubules of PAS-stained sections (two sections per animal). The length of cytoplasmic SC arms was determined in µm using vimentin staining. Vimentin marks intermediate filaments that are seen as cytoplasmic arms surrounding the nucleus and extending from the basal region towards the tubular lumen. SC arm length was calculated in ten tubules of vimentin-stained sections (two sections per animal). Vimentin was also used to detect SC-only tubules, characterized by intense staining of the entire seminiferous tubule combined with the lack of or dramatic decrease in the number of germ cells.

Testes, epididymis, EDs or pituitary gland were carefully dissected and collected on ice, washed in PBS and frozen. Testes were disrupted in Tri-Reagent (Sigma-Aldrich) using zirconium beads in a mechanical tissue disruptor (Precellys 24, Bertin Instruments). Total RNA was isolated according to the manufacturer’s recommendations (PureLink RNA Mini Kit, Ambion) and 1 μg of RNA was treated with DNase I (Thermo Fisher Scientific) before cDNA synthesis. cDNA was generated using 0.5-1 μg of total RNA and a High Capacity RNA-to-cDNA Kit (Applied Biosystems). RT-qPCR was performed using the comparative CT method and a StepOne Real-Time PCR System (Applied Biosystems). Amplification was performed using Power SYBR Green PCR Master Mix (Applied Biosystems) or TaqMan Universal PCR Master Mix (Applied Biosystems). All assays were performed in duplicate. For TaqMan assays, ActB (mouse and human) probe was used as an endogenous control for normalization and a specific Taqman probe was used for mouse or human Gemc1 (Mm02581229_m1), Mcidas (Mm01308202_m1), Ccno (Mm01297259_m1), Foxj1 (Mm01267279_m1), Trp73 (Mm00660220_m1) and TP73 (Hs01056231_m1). Primers used for SYBR Green assays (Sigma-Aldrich) were mCdc20b_FW (5′-TATCTTAGGAGATTCAGCGG-3′) and mCdc20b_RV (5′-ATGTATACAGAGGTTCCGAG-3′). See also Table S2.

Testes from adult male mice were isolated, decapsulated and processed together in a 15 ml falcon tube. Testes were first digested in EKRB medium [120 mM NaCl, 4.8 mM KCl, 25.2 mM NaHCO₃, 1.2 mM KH₂PO₂, 1.2 mM MgSO₄, 1.3 mM CaCl₂, 11 mM glucose, non-essential amino acids (Invitrogen), penicillin-streptomycin (Invitrogen)] with collagenase (0.5 mg/ml, Sigma-Aldrich, T6763) in a shaking water bath at 32°C for 10 min. Seminiferous tubules were allowed to sediment by gravity for 1-3 min and washed twice with EKRB. The tubules were further dissociated in EKRB with trypsin (1 mg/ml, Sigma-Aldrich, T6763) and DNase (5 µg/ml, Sigma-Aldrich, D4263) under shaking at 32°C for 15 min. Cells were resuspended thoroughly using a pipette until a single cell suspension was obtained and 1 ml of fetal bovine serum (FBS) was added to neutralize the trypsin. The cell suspension was filtered with a 70 µm cell strainer and total cell number was counted using the Neubauer chamber. One million cells/ml were resuspended in EKRB supplemented with 10% FBS and Hoechst 33342 (10 µg/ml, Life Technologies) under shaking at 32°C for 30-60 min. Propidium iodide (PI, 30 µg/ml, Life Technologies) was added to identify dead cells. Flow cytometric experiments were carried out using a FacsAria I SORP cell sorter (Beckton Dickinson), using a 70-micron nozzle at 60 PSI. Excitation of the sample was carried out using a blue (488 nm) laser for forward scatter (FSC) parameter, a green-orange laser (561 nm) was used for the excitation of PI and side scatter (SSC) signal, and a UV laser (350 nm) was used for Hoechst 33342 excitation. Cells were gated according to their scatter (FSC versus SSC) parameters; fluorescence of Hoechst 33342 was measured on live (not stained with PI) non-aggregated cells. Red emission (660/40 nm) versus blue emission (395/25 nm) from the UV laser was used on a dot plot in order to discriminate populations. Results were analyzed using the FlowJo software.

Slides containing meiotic squashes were washed three times in PBS followed by washing in PBS-T (0.4% Triton X-100 in PBS) and blocked with 5% goat serum and 1% BSA in PBS-T for 1 h. Slides were incubated overnight at 4°C with primary antibody, washed three times in PBS-T and stained with Alexa Fluor-conjugated antibodies (594, A-11032 and A-11037; 488, A-11029 and A-11034, Life Technologies, 1:200) for 1 h at RT. After the final wash, DNA was counterstained with DAPI. Slides were mounted using Vectashield Antifade reagent (Vector Laboratories) and imaged using a Leica TCS SP5 confocal microscope equipped with 63× NA 1.40 oil immersion objective and HyD detectors. For paraffin sections, heat-induced epitope retrieval was performed in a Rodent Decloaker (Biocare Medical). Tissue sections were blocked with 10% normal sera in PBS-0.1% Triton X-100 for 1 h at RT. After three washes with PBS, tissue sections were sequentially probed with primary antibodies and fluorescently labeled secondary antibodies (Jackson ImmunoResearch).

AD293 cells (Stratagene) were cultured in Dulbecco's Modified Eagle Medium (Gibco) with 10% FBS (Hyclone) and routinely tested for mycoplasma and found negative. For transient transfections, AD293 cells were seeded in 10 cm plates at 70% confluence and 10 µg of plasmid were transfected with polyethylenimine (Polysciences) the day after. The medium was changed 12 h post-transfection and cells were collected 48 h after with RIPA buffer [50 mM Tris-HCl (pH 8), 150 mM NaCl, 1% NP-40, 0.1% SDS, 0.5% sodium deoxycholate]. For tissue samples, testes were disrupted in RIPA buffer using zirconium beads in a mechanical tissue disruptor (Precellys 24). Samples were incubated for 20 min on ice and sonicated using a Bioruptor XL sonication device (Diagenode) for 15 min with 15 s intervals and centrifuged at 4°C for 20 min at 1300 rpm.

The expression construct for FLAG-tagged human GEMC1 (pcDNA5-FRT/TO-flag-hGemc1) was previously described in Terré et al. (2016). Human CCNO (hCCNO) cDNA was obtained from Expressed Sequence Tag EST IMAGE: 6421733 (GenBank accession number BQ917277.1) as a template and amplified using 5′-GCGAATTCCATGGTGACCCCCTGTCCCACCAGCC-3′ and 5′-GCTCTAGATTATTTCGAGCTCGGGGGCAGG-3′ primers. hCCNO cDNA was cloned into the pBluescript (I) SK(+) vector (Stratagene) at the EcoRI and XbaI restriction sites and then into the mammalian expression vector pCDNA3.1 (Invitrogen) modified with an N-terminal myc-tag to produce proteins N-terminally fused to myc under the control of the constitutive CMV promoter (pcDNA3.1-myc-hCcno). The pCMX-flag plasmid was used as an empty-vector control (gift from Ron Evans, the Salk Institute for Biological Studies, USA).

Spermatocyte squashes were performed as previously described (Marjanović et al., 2015). In brief, tubules from adult mice were placed in fixative solution (2% formaldehyde, 0.05% Triton X-100 in PBS) for 10 min. To release spermatocytes from tubules, a small aliquot of the tubules was minced with tweezers on a glass slide pre-treated with poly-L lysine. Cells were covered with a coverslip and gently squashed by applying pressure with the thumb. The slide was snap-frozen in liquid nitrogen, the coverslip was removed and the slide was processed for immunofluorescence.

We thank M. Lize, H. Omran, J. Wallmeier and R. Hess for discussing unpublished data and experimental suggestions; M. Di Giacomo and Ignasi Roig for technical input; A. Nebreda, X. Salvatella and E. Batlle for reagents; L. Palenzuela for mouse colony management; the Institute for Research in Biomedicine (IRB) Barcelona histopathology facility and the Advanced Molecular Pathology Laboratory of the Institute of Molecular and Cell Biology, Singapore, for assistance with IHC and protocol establishment; and O. Reina of the IRB Biostatistics facility for help with data analysis.