WDR62 is the second most common genetic alteration associated with microcephaly. It has been shown that Wdr62 is required for germ cell meiosis initiation in mice, and the majority of male germ cells are lost in the meiotic defect of first wave spermatogenesis in Wdr62 mutants. Strikingly, in this study, we found that the initiation of meiosis following spermatogenesis was not affected and the germ cells were gradually repopulated at later developmental stages. However, most germ cells were arrested at metaphase of meiosis I and no mature sperm were detected in epididymides. Further, this study demonstrated that metaphase I arrest of Wdr62-deficient spermatocytes was caused by asymmetric distribution of the centrosome and aberrant spindle assembly. Also, mechanistic studies demonstrated that WDR62 interacts with centrosome-associated protein CEP170, and deletion of Wdr62 causes downregulation of the CEP170 protein, which in turn leads to the aberrant spindle assembly. In summary, this study indicates that the meiosis of first wave spermatogenesis and the following spermatogenesis started from spermatogonium is probably regulated by different mechanisms. We also demonstrated a new function of WDR62 in germ cell meiosis, through its interaction with CEP170.
Spermatogenesis is a complex process of cellular differentiation that is involved in the proliferation and differentiation of spermatogonia (Barroca et al., 2009; Brinster and Zimmermann, 1994; de Rooij and Russell, 2000), meiosis of spermatocytes and postmeiotic development of spermatids (Blobel, 2000). Meiosis is a program of two successive cell divisions preceded by one round of DNA replication that is required for the generation of haploid gametes (Braun, 2000; Yu et al., 2003). Meiotic recombination in prophase of meiosis I allows the exchange of genetic materials between paternal and maternal homologous chromosomes (Baudat et al., 2013; Kohl and Sekelsky, 2013). Homologous recombination is based on the formation of programmed DNA double-strand breaks (DSBs) and DSB repair (Ma et al., 2017; Romanienko and Camerini-Otero, 2000; Smirnova et al., 2006; Tarsounas et al., 1999). Defects in homologous recombination and DSB repair lead to meiosis arrest at the pachytene stage in spermatogenesis (Haines et al., 2015; Keeney et al., 2014; Xu et al., 2016). The spindle assembly checkpoint (SAC) is a key regulator of chromosome segregation in both mitosis and meiosis (Gorbsky, 2015). The SAC is activated when spindle assembly is disturbed or the interaction between the spindle and kinetochore of chromosomes is ruined, and prevents progression from metaphase to anaphase (Hartwell and Weinert, 1989). Disruption of the SAC results in metaphase arrest or aneuploidy gametes (Jia et al., 2013; Sacristan and Kops, 2015; Subramanian and Hochwagen, 2014; Woglar and Jantsch, 2014; Yamamoto, 2014).
The WD repeat domain 62 (Wdr62) gene contains 33 exons and encodes a protein containing 13 WD40-domain repeats that belongs to the WD-repeat protein family (Stirnimann et al., 2010). There are five key functional domains of WDR62 among human paralogous and orthologous proteins in various mammalian species (Cohen-Katsenelson et al., 2011). WD40 repeats at the N terminus are prominent features within proteins that mediate diverse protein-protein interactions (Stirnimann et al., 2010). The WD40-repeat domain is also required for microtubule association (Lim et al., 2015a). The MKK7β1-binding domain (MBD) is responsible for the association of WDR62 with MKK7β1 through direct protein-protein interaction and is present at the carboxyl terminus of human WDR62 (1212-1284). The JNK-binding domain (JBD), which is located at the carboxyl terminus (1291-1301), is required for c-Jun (also known as Jun) N-terminal kinase (JNK)-mediated phosphorylation of the WDR62 C terminus, which is required for spindle maintenance and mitotic progression (Bogoyevitch et al., 2012; Cohen-Katsenelson et al., 2011; Wasserman et al., 2010). The loop helix domain (LHD) is located at the carboxyl terminus of human WDR62 (1414-1520) and is necessary for homodimerization (Cohen-Katsenelson et al., 2013).
Mutation of WDR62 is associated with microcephaly (Kousar et al., 2011; Nicholas et al., 2010; Wollnik, 2010) and with defects of neural progenitor cell development. Patients with WDR62 mutations show a disrupted centrosome and spindle association (Farag et al., 2013). It has been reported that WDR62 is a spindle pole protein (Nicholas et al., 2010; Yu et al., 2010) and is required for the recruitment of JNK1 (Mapk8) and Aurora A (Aurka) to the spindle pole. Phosphorylation of WDR62 by JNK1 or Aurora A (Bogoyevitch et al., 2012; Xu et al., 2014) plays an opposing role in spindle maintenance and timely mitotic progression (Lim et al., 2015b). A recent study showed that WDR62 is also phosphorylated by polo-like kinase 1 (PLK1) and it is required to maintain the orientation of the spindle in mitosis during the development of the astral spindle (Miyamoto et al., 2017).
Our recent study found that Wdr62 plays important roles in germ cell meiotic initiation by activating JNK signaling (Zhou et al., 2018). Inactivation of Wdr62 causes a defect in meiosis initiation and most germ cells in ovaries are lost during embryonic stage. The majority of male germ cells are lost postnatally and very few germ cells are noted at postnatal day (P)10. We also found that the defect of germ cell loss could be partially rescued by overexpression of JNK1 (Zhou et al., 2018). Interestingly, in this study, we find that the germ cells in male mice in a mouse model with only exon 2 deleted (Wdr62e2−⁄e2−) are gradually repopulated at later developmental stages. However, most germ cells are arrested at the metaphase stage of meiosis I, with an asymmetrically distributed centrosome and aberrant spindle assembly. Mechanistic studies demonstrate that WDR62 interacts with centrosome-associated protein CEP170 and is required for its stability. Our study reveals a new function of Wdr62 in spermatogenesis by regulating spindle assembly during meiosis.
Inactivation of Wdr62 results in male infertility
Wdr62e2−⁄e2− mice were born at a normal Mendelian ratio, and no obvious developmental abnormalities were noted until 6 months of age. However, male Wdr62e2−⁄e2− mice were completely infertile, and the size of testes was significantly smaller than that of control mice (data not shown). To explore the reasons of male infertility, germ cells were examined at different developmental stages. MVH (DDX4)-positive germ cells were detected in both control (Fig. 1A) and Wdr62e2−⁄e2− (Fig. 1H) testes at P1, and no difference was noted. Compared with control mice (Fig. 1B-G), germ cell loss was first observed in Wdr62e2−⁄e2− mice at P3 (Fig. 1I), and very few germ cells were observed in Wdr62e2−⁄e2− mice at P5 (Fig. 1J) and P10 (Fig. 1K). Interestingly, germ cells in Wdr62e2−⁄e2− mice were gradually recovered from 2 weeks (Fig. 1L) to 3 weeks (Fig. 1M) and a large number of germ cells were observed at 6 weeks (Fig. 1N). Although germ cells were repopulated, very few round and elongated spermatids were observed in Wdr62e2−⁄e2− mice (Fig. S1C) compared with control mice (Fig. S1A), and only cell debris was observed in the epididymides of Wdr62e2−⁄e2− mice (Fig. S1D). Histological studies showed that most of the germ cells in Wdr62e2−⁄e2− mice were at the spermatocyte stage (Fig. S1C). This result was further confirmed by immunostaining of SYCP3. Only one to two layers of SYCP3-positive spermatocytes were noted in the seminiferous tubules of control mice (Fig. S1E), whereas most of the germ cells in Wdr62-deficient mice were SYCP3 positive (Fig. S1F).
DNA DSB repair and homologous recombination are not affected in germ cells of Wdr62e2−⁄e2− mice
To examine whether spermatocyte arrest in Wdr62e2−⁄e2− mice was due to the defect of DSB repair or homologous recombination in meiosis, immunostaining of SYCP3 was conducted to visualize the chromosome axes of primary spermatocytes. As shown in Fig. 2A, primary spermatocytes at the leptotene, zygotene, pachytene, diplotene and diakinesis (data not shown) stages were observed in both control and Wdr62e2−⁄e2− mice. The quantitative results showed that the percentage of spermatocytes in the pachytene stage was increased in Wdr62e2−⁄e2− mice (Fig. 2B). To examine whether the DSBs were properly formed, the expression of γH2AX was examined by immunofluorescence. As shown in Fig. 2C, the γH2AX signal was scattered along the chromosomes in the zygotene stage and restricted to the XY body in the pachytene and diplotene stages in both control and Wdr62e2−⁄e2− mice. Expression of DSB repair-related genes was also examined by immunostaining. Numerous RAD51-positive foci (Fig. 2D) were observed in spermatocytes in the zygotene stage, and the number was significantly reduced in the late zygotene and pachytene stages in Wdr62e2−⁄e2− mice, similar to in control mice. Statistical analysis showed that in Wdr62e2−⁄e2− spermatocytes, RAD51 signals at the pachytene stage were not affected, in comparison with those in control spermatocytes (Fig. S2K). The foci of the RPA70 (RPA1) protein (green) were located on the chromosomes (red) of both control (Fig. S2A,B) and Wdr62-deficient (Fig. S2C,D) germ cells at late zygotene (Fig. S2A,C) and pachytene (Fig. S2B,D) stages. A large number of DMC1-positive foci were detected on the chromosomes of both control (Fig. S2E,F) and Wdr62-deficient (Fig. S2G,H) germ cells at the late zygotene stage (Fig. S2E,G), and the number was dramatically reduced in the pachytene stage (Fig. S2F,H) in both control and Wdr62-deficient germ cells, and no difference was noted (Fig. S2L). MLH1 is a marker that indicates the completion of homologous recombination by interfering with the crossover pathway at the pachytene stage. MLH1-positive foci were observed in both control (Fig. S2I) and Wdr62e2−⁄e2− (Fig. S2J) germ cells at the pachytene stage and no difference was noted (Fig. S2M). These results indicated that the process of DSB repair and DNA homologous recombination during meiosis I were not affected in the germ cells of Wdr62e2−⁄e2− mice.
Germ cells in Wdr62-deficient mice are arrested at the metaphase I stage of meiosis
Spermatocytes at metaphase I of meiosis were observed (Fig. 3B, inset) and quantified (Fig. 3I) in the seminiferous tubules of Wdr62e2−⁄e2− mice at 5 weeks of age compared with control (Fig. 3A). To test whether germ cells in Wdr62-deficient mice were arrested at metaphase I of meiosis, the expression of BUBR1 (BUB1B) and anti-centromere antibody (ACA) was examined by immunofluorescence. BUBR1 is the core protein of the SAC and is recruited to the kinetochores when the spindle is improperly assembled (Fan, 2010; Wei et al., 2010). ACA protein was located at the kinetochore in both control (Fig. 3D) and Wdr62-deficient (Fig. 3G) spermatocytes at the metaphase I stage. Many BUBR1-positive foci were detected in Wdr62-deficient (Fig. 3F) spermatocytes at metaphase I stage and co-localized with ACA protein (Fig. 3H). Very few BUBR1-positive foci were detected in control germ cells at this stage (Fig. 3C,E). Quantitative analysis showed that the percentage of BUBR1 and ACA co-labeled spermatocytes in metaphase I was dramatically increased in Wdr62e2−⁄e2− mice (Fig. 3J). These results indicated that Wdr62-deficient germ cells were arrested at the metaphase I stage.
Aberrant spindle assembly in spermatocytes of Wdr62e2−⁄e2− mice at the metaphase I stage
It has been reported that activation of the SAC is closely related to aberrant assembly of the spindle (Xiong et al., 2008; Xu et al., 2012; Yuan et al., 2010). The morphology of the spindle was examined by immunostaining with an anti-α-tubulin antibody. As shown in Fig. 4, the spindles in spermatocytes from control mice in metaphase I were well organized, with symmetric spindle poles (Fig. 4A,C). By contrast, spindles in spermatocytes from Wdr62e2−⁄e2− mice were severely disrupted, with asymmetrically located spindle poles (Fig. 4B,D). The average length and width of the spindles were dramatically reduced in Wdr62e2−⁄e2− spermatocytes compared with those of control mice (Fig. 4E). In addition, the relative intensity of α-tubulin was also significantly reduced in Wdr62e2−⁄e2− mice. The spindle polarity axis of spermatocytes during the metaphase I stage in Wdr62-deficient mice exhibited a substantial rotation from the planar orientation (as noted by γ-tubulin and α-tubulin staining). The intersection angle between the spindle polarity axis and planar orientation was significantly increased in Wdr62-deficient spermatocytes (Fig. 4F). Very few spermatocytes at anaphase I were observed in Wdr62e2−⁄e2− mice compared with control (Fig. S3A), but the intermediate tubules were severely disrupted (Fig. S3B). High magnification images showed that homologous chromosomes moved to bipolarity in control mice (Fig. S3C) during anaphase I; however, owing to the destruction of the spindle in Wdr62e2−⁄e2− mice (Fig. S3D), the separation of homologous chromosomes was perturbed. These results indicate that the assembly and positioning of the spindle during metaphase I were disrupted in Wdr62-deficient germ cells, which most likely caused metaphase I arrest.
The positioning of the centrosome during metaphase I is disturbed in spermatocytes of Wdr62e2−⁄e2− mice
To examine the polarity and morphology of the centrosome, pericentriolar material (PCM) was labeled with γ-tubulin. As shown in Fig. 5, an asymmetric location of γ-tubulin-positive PCM was observed in most Wdr62-deficient germ cells (Fig. 5C,D) compared with control (Fig. 5A,B). Also, very little multipolar PCM was observed in a small proportion of Wdr62-deficient germ cells (Fig. 5E,F). The results of quantitative analysis showed that the number of germ cells with asymmetric PCM was dramatically increased in Wdr62e2−⁄e2− mice (Fig. 5G).
To determine whether Wdr62 was also involved in centrosome polarity and spindle assembly in somatic cells, mouse embryonic fibroblast (MEF) cells from control and Wdr62e2−⁄e2− mice were cultured in vitro, the proliferation (Fig. S4A) and spindle morphology (Fig. S4B) were examined. The results of the MTT assay showed that the proliferative rate of Wdr62e2−⁄e2− MEF cells was comparable with that of control MEF cells. No defect of centrosome positioning or spindle assembly was observed in Wdr62e2−⁄e2− MEF cells (Fig. S4Bg-Bl). These results suggest that the long isoform of Wdr62 is not expressed in somatic cells and thus not involved in spindle assembly of somatic cells.
WDR62 is abundantly expressed in spermatocytes/round spermatids and interacts with CEP170
To explore the regulatory mechanisms of WDR62 in germ cell meiosis, a hemagglutinin (HA)-tagged Wdr62 knock-in mouse model (hereafter referred to as Wdr62-HA) was generated using CRISPR/Cas9 technology (Fig. S5A). The HA signal was detected in the cytoplasm of spermatocytes and round spermatids (Fig. S5C, arrowheads and arrows, respectively) in Wdr62-HA mice, but not in the control testes (Fig. S5B). A ∼200-kD band was detected in the testes of Wdr62-HA mice by western blotting using both anti-HA and anti-WDR62 antibodies (Fig. S5D). These results indicate that the endogenous long isoform of WDR62 protein in Wdr62-HA mice was recognized by the anti-HA antibody. Then, the subcellular localization of WDR62 in spermatocytes was examined by double-staining of HA (Fig. S5E,I), γ-tubulin (Fig. S5F) and α-tubulin (Fig. S5J) at metaphase I (Fig. S5G,K). We found that WDR62 was co-localized with the γ-tubulin-positive centrosome (Fig. S5H) and α-tubulin-positive spindle (Fig. S5L).
We further examined the expression of WDR62 protein in other tissues (Fig. S6B). Interestingly, the ∼200-kD band was only detected in testes with anti-HA antibody, not in other tissues, including the brain. By contrast, a smaller band (∼170-kD) was detected in the brain and ovary by an anti-Wdr62 antibody and was absent in adult testes. Then, we examined the expression of WDR62 protein in testes at different developmental stages (Fig. S6A). Both the small and large isoforms were detected in the control testes at 3 weeks, whereas only the large band was absent in Wdr62e2−⁄e2− testes. Only the large isoform was expressed in testes at 3 months and was absent in Wdr62e2−⁄e2− testes. All of these results suggest that different isoforms of WDR62 are expressed in testes and other tissues. The large isoform (∼200-kD) was expressed in adult testes, and the small isoform (∼170-kD) was expressed in the brain and other tissues.
To further explore the mechanistic basis of WDR62 in regulating the polarity of the centrosome and spindle assembly, WDR62-interacting proteins were screened by co-immunoprecipitation (Co-IP) and iTRAQ mass spectrometry analyses using testes from adult control and Wdr62-HA mice. Notably, several classes of proteins were pulled down by the anti-HA antibody, including centrosome-related protein, PIWI-like protein, motor-related protein and APC-related protein (Fig. 6A). Among others, CEP170 is a centrosome-associated protein that has been reported to be involved in spindle assembly (Welburn and Cheeseman, 2012). For a more detailed characterization of the subcellular localization of WDR62 and CEP170, immunofluorescence was performed in CEP170- and WDR62-overexpressing HeLa cells. WDR62 (Fig. 6Cg) was co-localized with γ-tubulin (Fig. 6Ci), but the signal was broader than that of γ-tubulin (Fig. 6Ch). The centrin signal was surrounded by the WDR62 protein, but these two proteins were not overlapped (Fig. 6Cf). The CEP170 protein partially overlapped with the centrin protein (Fig. 6Co) and WDR62 (Fig. 6Cc).
To further confirm the interaction between WDR62 and CEP170 proteins, Flag epitope-tagged CEP170 and HA epitope-tagged WDR62 were overexpressed in 293 cells. Co-IP experiments showed that the CEP170 protein was pulled down by the anti-HA antibody and that WDR62 was pulled down by the anti-Flag antibody (Fig. 6B). The interaction of WDR62 and CEP170 protein was also confirmed by Co-IP experiments using testes from Wdr62-HA mice. As shown in Fig. 7A, endogenous CEP170 protein was pulled down by the anti-HA antibody. The results of immunostaining showed that CEP170 protein was localized at the centrosome in spermatocytes of control testes (Fig. 7Ca, arrowheads), whereas it was virtually absent in spermatocytes of Wdr62-deficient testes (Fig. 7Cb). We also found that the protein level of CEP170 was dramatically reduced in Wdr62-deficient testes (Fig. 7B). To determine whether Wdr62 depletion destabilizes CEP170 positioning and disrupts the spindle formation in mitotic cells, the CEP170 positioning (Fig. S7B,F) and spindle morphology (Fig. S7A,E) were examined in MEF cells from control (Fig. S7C,D) and Wdr62e2−⁄e2− mice (Fig. S7G,H). No defect in CEP170 expression (Fig. S7F) and spindle assembly (Fig. S7E) was observed in Wdr62e2−⁄e2− MEF cells (Fig. S7H). These results indicated that the inactivation of Wdr62 does not cause instability of CEP170 protein in somatic cells. Given that CEP170 is an important centrosome protein, the defect of centrosome positioning and aberrant spindle assembly in Wdr62-deficient germ cells were most likely due to downregulation of the CEP170 protein in germ cells.
The defect in germ cell survival in Wdr62e2−⁄e2− mice is partially rescued by JNK1 overexpression, but the defect in spermatogenesis is not rescued
It has been reported that WDR62 is involved in spindle assembly in neural progenitor cells via activating JNK signaling (Shohayeb et al., 2017). And our recent study also finds that WDR62 is involved in germ cell meiosis initiation via activating JNK signaling (Zhou et al., 2018). To test whether WDR62 involvement in spindle assembly in germ cell meiosis is also JNK signaling dependent, Wdr62e2−⁄e2−; CAJNK1+/flox; Tnap-Cre mice (hereafter referred to as Tnap-Cre rescued mice) were generated. In this mouse model, constitutively activated JNK1 was specifically expressed in Wdr62-deficient germ cells from approximately embryonic day (E)8.5. As shown in Fig. 8, a few MVH and PLZF (Zbtb16) double-positive germ cells were observed in a small portion of seminiferous tubules in Wdr62e2−⁄e2− mice at P10 (Fig. 8D) compared with control (Fig. 8A), and the number of MVH and PLZF double-positive germ cells was significantly increased in the testes of Tnap-Cre rescued mice (Fig. 8G,J) at this stage. At two months of age, MVH-positive (green) and PLZF-positive (red) germ cells were noted in testes of control (Fig. 8B), Wdr62e2−⁄e2− (Fig. 8E) and Tnap-Cre rescued (Fig. 8H) mice. The number of MVH-positive germ cells was significantly increased in Tnap-Cre rescued mice (Fig. 8H,J) compared with Wdr62e2−⁄e2− mice (Fig. 8E,J). Mature sperm were observed in the epididymides of control mice (Fig. 8C), but not in Wdr62e2−⁄e2− (Fig. 8F) and Tnap-Cre rescued (Fig. 8I) mice. Aberrant spindle assembly was also observed in most spermatocytes at metaphase I in Tnap-Cre rescued mice (Fig. S8C,F), which was consistent with Wdr62e2−⁄e2− (Fig. S8B,E) mice. The quantified analysis (Fig. S8G) showed that the average length and width of the spindles or the relative intensity of α-tubulin were dramatically reduced in Wdr62e2−⁄e2− and Tnap-Cre rescued spermatocytes compared with those of control mice (Fig. S8A,D). To further test whether the defect of spindle assembly in Wdr62e2−⁄e2− mice can be rescued by activating JNK signaling at a later developmental stage, Wdr62e2−⁄e2−; CAJNK1+/flox; Stra8-Cre mice were generated. In this mouse model, JNK1 was specifically overexpressed in germ cells of Wdr62e2−⁄e2− mice at ∼P5. The results of Hematoxylin and Eosin (H&E) staining showed that the spermatogenesis defect was not rescued in this mouse model (Fig. S9) at 2 months of age. We propose that the lost MVH-positive cells at P10 were the germ cells which directly start meiosis from gonocytes, and the remaining PLZF and MVH co-labeled cells were the spermatogonium transformed from the small number of gonocytes, which will start the following wave of spermatogenesis. All of these results indicate that the survival of germ cells in Wdr62e2−⁄e2− mice was dependent on JNK signaling, whereas the function of WDR62 in spindle assembly during spermatogenesis was independent of JNK signaling.
WDR62 was originally identified as a JNK signaling scaffold protein that plays important roles in brain development (Chen et al., 2014; Cohen-Katsenelson et al., 2013, 2011; Verloes et al., 1993). The function of WDR62 in brain development has been investigated using a gene-trap mouse model. Homozygous mutant mice exhibited growth retardation and reduced brain size with spindle instability and mitotic arrest in neural progenitors and MEF cells. A mechanistic study demonstrated that WDR62 interacted with Aurora A to regulate spindle formation (Chen et al., 2014). Surprisingly, in the current study, we found that Wdr62e2−⁄e2− mice were born at a normal Mendelian ratio, and no obvious developmental defects were observed. Moreover, the proliferation and viability of MEF cells were not affected. This discrepancy is probably caused by the different strategies used to generate the mouse models. The gene trap mouse model was generated by inserting a β-geo reporter in the intron region between exons 14 and 15, which only caused a downregulation of WDR62 expression (Chen et al., 2014). By contrast, exon 2 was deleted in our mouse model, which caused a frame-shift of the Wdr62 gene. Interestingly, a large isoform (∼200-kD) of the WDR62 protein was detected in the testes, whereas it was absent in brain and MEF cells. By contrast, a small isoform (∼170-kD) was detected in the brain, which was similar to the findings of a previous study (Lim et al., 2015b; Bogoyevitch et al., 2012). However, this small isoform was not recognized by the anti-HA antibody in Wdr62-HA mice. The large isoform (∼200-kD) of the WDR62 protein has not been reported in previous work. Based on these results, we speculated that different isoforms of the Wdr62 gene were expressed in the testes and brain. The small isoform probably terminates earlier, and thus, the HA tag was not translated. In our mouse model, the large isoform of WDR62 was deleted, but the small isoform was intact. The large and small isoforms of WDR62 most likely have different functions. The small isoform is involved in the development of neural progenitor cells and MEF cells, whereas the large isoform is important for germ cell meiosis. However, detailed information about the functions of these two isoforms requires further investigation.
It has been reported that WDR62 is required for spindle assembly and timely mitotic progression as a spindle pole protein which can be phosphorylated (Garcia-Higuera et al., 1996; Nawrot et al., 2006; Nicholas et al., 2010). An in vitro study also demonstrated that inactivation of WDR62 in HeLa cells caused the defect of spindle orientation and delayed mitotic progression. A mechanistic study found that WDR62 was involved in spindle assembly in mitosis by interacting with JNK1 (Lim et al., 2015b; Bogoyevitch et al., 2012). A uterus electroporation study also found that WDR62 was involved in spindle assembly of rat neural progenitor cells by activating JNK signaling (Chen et al., 2014; Xu et al., 2014). Our previous study also found that the defect in female germ cell development was partially rescued by overexpression of JNK1 (Zhou et al., 2018). In this study, we found that inactivation of Wdr62 also caused a defect in spindle assembly and centrosome positioning in germ cells, which in turn led to metaphase arrest of male germ cells. However, defective germ cell meiosis was not rescued by JNK1 overexpression, suggesting that the involvement of Wdr62 in spindle assembly during meiosis of male germ cells is independent of JNK signaling.
The centrosome is the major microtubule organizing center and directs spindle assembly at metaphase in both meiosis and mitosis. It is composed of a pair of centrioles and embedded PCM (Gönczy, 2012; Sir et al., 2011). Centrosome proteins, such as CEP152, CEP63, CEP135 and CEP192, among others, are required for centriole duplication and spindle assembly in mitosis (Gomez-Ferreria and Sharp, 2008; Inanç et al., 2013; Kitagawa et al., 2011; Tang et al., 2011; Zhu et al., 2008). Inactivation of Cep63 causes destruction of centriole duplication and impaired bipolar spindle assembly in meiosis (Marjanović et al., 2015). CEP170 is also a centrosome protein and localizes in subdistal appendages of the mature centriole to regulate spindle assembly and cell morphology (Guarguaglini et al., 2005). CEP170 interacts with the centrosome-associated kinesins KIF2A, KIF2C and KIFC3, and this complex plays an important role in metaphase spindle size control (Maliga et al., 2013). It also has been reported that WDR62 is a binding partner of CEP170 and other centrosomal proteins, and mutation of these genes is associated with microcephaly (Yu et al., 2010). In this study, we found that WDR62 was partially co-localized with CEP170, and the interaction between these two proteins was confirmed by Co-IP experiments. Most importantly, we found that the expression of CEP170 protein was dramatically reduced in Wdr62-deficient germ cells. Given that CEP170 is a centrosome protein and is involved in spindle assembly, we speculated that WDR62 is involved in spindle assembly, most likely by interacting with CEP170. Loss of Wdr62 results in instability of the CEP170 protein, which in turn causes a defect of centrosome positioning and aberrant spindle assembly. However, we could not exclude the possibility that WDR62 is involved in spindle assembly by also interacting with other centrosome proteins or that it directly interacts with spindle proteins.
In mouse testis, a large number of germ cells directly start meiosis from gonocytes at ∼P3. This is considered the first wave of spermatogenesis. By contrast, a small number of gonocytes are transformed into spermatogonium, and spermatogenesis starts from these cells. Whether the meiosis of the first wave and the following spermatogenesis is regulated by similar mechanisms is unclear. The results of this and our previous study demonstrate that Wdr62 has different roles in these two mechanisms. The meiosis of first-wave spermatogenesis is more similar to that seen in female germ cells. JNK signaling is involved in meiosis in female germ cells and the first wave of spermatogenesis, whereas it is not required for the following spermatogenesis. Another interesting finding was that the defect in germ cell survival in Wdr62-deficient mice was rescued by JNK1 overexpression, but the defect in spermatogenesis in juveniles and adults was not rescued. These results suggest that Wdr62 plays different functions in germ cells at different developmental stages. In summary, we demonstrated that WDR62 played important roles in germ cell meiosis by interacting with CEP170. Loss of Wdr62 resulted in downregulation of CEP170, which in turn caused an asymmetric distribution of the centrosome and aberrant spindle assembly. We also demonstrated that a large isoform of the Wdr62 gene was specifically expressed in the testes and absent in the brain and other organs. This study provides new information that allows a better understanding of the regulation of germ cell meiosis progression and the functions of the Wdr62 gene in different cell types.
MATERIALS AND METHODS
All mice studies were carried out in accordance with the principles approved by the Institutional Animal Care and Use Committee at the Institute of Zoology, Chinese Academy of Sciences. All mice were maintained in a C57BL/6; 129/SvEv mixed background. Wdr62e2−⁄e2− mice, Wdr62-HA mice, Wdr62e2−⁄e2−;CAJNK1+/flox;Tnap-Cre mice and Wdr62e2−⁄e2−;CAJNK1+/flox;Stra8-Cre mice were used in this research. See supplementary Materials and Methods for further details.
Tissue collection and histological analysis
Testes were dissected from control and mutant mice immediately after euthanasia, fixed in 4% paraformaldehyde (PFA) for up to 24 h, stored in 70% ethanol and embedded in paraffin. Next, 5 μm-thick sections were prepared using a rotary microtome (Leica) and mounted on glass slides. After deparaffinization, sections were stained with H&E for histological analysis.
Immunohistochemistry (IHC) analysis
IHC was performed with at least three mice for each genotype using a Vectastain ABC (avidin-biotin-peroxidase) kit (Vector Laboratories) as recommended by the manufacturer and as described previously (Han et al., 2018). The slides were examined using a Nikon Microscope, and images were captured with a Nikon DS-Ril CCD camera.
Immunofluorescence (IF) analysis
After deparaffinization and antigen retrieval, the sections were blocked with 5% donkey serum at room temperature (RT) for 1 h, followed by incubating with a primary antibody for 1 h at RT or overnight at 4°C. After washing three times in 1× PBS buffer, the slides were incubated with the corresponding secondary antibody, fluorescent dye-conjugated-FITC or TRITC (1:150, Jackson) for 1 h at RT. The nucleus was stained with DAPI. All images were captured with a confocal laser scanning microscope (Leica SP8 fitted with hybrid detectors).
Preparation of synaptonemal complex
The testes were dissected and seminiferous tubules were washed in 1× PBS buffer. After incubation with hypo extraction buffer (HEB) for 30 min, the tubules were disrupted in 0.1 M sucrose to make a cell suspension. The cell suspension was mounted on slides covered with 1% PFA. Slides were placed in a humidified box for at least 6 h. Then, the slides were air dried at RT for 30 min, washed in 0.04% Photo-Flo (Equl, 1464510) for 4 min and set at an angle for 10 min to dry. The air-dried-slides were then prepared for synaptonemal complex staining of SYCP3. Slides were incubated with 1× antibody dilution buffer (ADB) at RT for 30 min, followed by incubating with SYCP3 antibody overnight at 37°C. After washing three times in 1× Tris buffer, saline (TBS), the slides were blocked with 1× ADB overnight at 4°C. Washing three times in cold 1× TBS buffer, the slides were incubated with the corresponding secondary antibody, fluorescent dye-conjugated-TRITC for 3 h at 37°C. All images were captured using a confocal laser scanning microscope (Leica SP8 fitted with hybrid detectors).
The HA tag TACCCATACGATGTTCCAGATTACGCT was inserted in front of the stop codon, TGA, of the Wdr62 gene. See supplementary Materials and Methods for further details.
Wdr62-HA cDNA was amplified from adult Wdr62-HA mouse testes by RT-PCR and cloned into Xho1 and Not1 restriction sites to generate pCMS.EGFP-Wdr62-HA. Cep170-Flag cDNA was amplified from adult C57 mouse testes by RT-PCR and cloned into Mlu1 and Not1 restriction sites to generate pCMS.EGFP.Flag-Cep170-Flag.
Cell culture and transfection
HeLa cells (3111C0001CCC000011) and 293 cells (3111C0001CCC000010) were purchased from National Infrastructure of Cell Line Resource. They were authenticated and tests showed no contamination. HeLa cells and 293 cells were cultured in DMEM (Invitrogen) supplemented with 10% fetal bovine serum and 1% penicillin/streptomycin and grown in a humidified 5% CO2 environment at 37°C, and showed a healthy state (i.e. adherent and normal proliferation). Transfection was mediated with Lipofectamine® 3000 (Invitrogen) and fetal bovine serum-free DMEM medium (Invitrogen) according to the manufacturer's instructions.
Isolation of MEF cells
Cell proliferation assay
The relative number and viability of MEF cells were evaluated by MTT assays at 1, 2, 3, 4, 5 or 6 days. See supplementary Materials and Methods for further details.
Plasmids (2.5 μg of each plasmid) were transfected into 293 cells. Forty-eight hours after transfection, cell extracts were prepared using lysis buffer [50 mM Tris-HCl (pH 7.5), 100 mM NaCl, 2 mM EDTA (pH 8.0), 1% (v/v) NP40, 2 mM PMSF] supplemented with protease inhibitors and phosphatase inhibitors. For Co-IP, ∼1000 μg protein was incubated for 2-4 h in vitro (and 12 h in vivo) at 4°C with 15 μl indicated monoclonal HA-agarose (Sigma-Aldrich, A2095) or Flag-agarose (Sigma-Aldrich). The immunoprecipitates were washed five times using cell lysis buffer and the bound protein complex eluted by 2× SDS-PAGE loading buffer and analyzed with western blotting as described previously (Fan et al., 2017).
The primary antibodies used were: anti-BUBR1 (1:200, Abcam, ab28193), anti-SYCP3 (1:200, Abcam, ab97672), anti-γ-H2AX (1:400, Millipore, 05-636), anti-RAD51 (1:25, Santa Cruz Biotechnology, sc-839), anti-RPA1 (1:300, Abcam, ab87272), anti-DMC1 (1:25, Santa Cruz Biotechnology, sc-22768), anti-MLH1 (1:200, Abcam, ab92312), anti-α-tubulin-FITC (1:100, Sigma-Aldrich, F2168), anti-γ-tubulin (1:400, Abcam, ab11316), anti-centrin (1:100, Millipore, 04-1624), anti-HA (1:400, MBL International, M180-3) and anti-FLAG (1:400, MBL International, M185-3L).
We cut 5-μm-thick sections of testes using a rotary microtome (Leica, MR2235). Sections were immunostained with anti-α-tubulin antibody (as in the IF analysis) and the spindles were captured using a confocal laser scanning microscope (Leica, TCS SP8). We then counted a similar number of spermatocytes with α-tubulin-positive spindles in control and Wdr62e2−⁄e2−testes and measured the parameters of the spindles, including the length, width and the intensity of α-tubulin.
Data and statistical analysis
All images were processed with Photoshop CS6 (Adobe) and ImageJ (National Institutes of Health). All statistics were analyzed using Microsoft Excel and Prism software (GraphPad Software). All experiments were confirmed with at least three independent experiments, and three to five control or mutant testes were used for immunostaining. The quantitative results were presented as the mean±s.e.m. The significant difference was evaluated with t-test and χ2 test. P-value < 0.05 was considered as significant.
We thank Shuguang Duo, from the Institute of Zoology, Chinese Academy of Sciences, for help with CRISPR/Cas9 experiments. We thank the proteomics facility of the Institute of Genetics and Developmental Biology, Chinese Academy of Sciences, for help with mass spectrometry analyses.
Conceptualization: F.G.; Methodology: Y.Q., Y.Z., Z.S., B.X., M.C. (0000-0001-5074-2174), Y.L., A.B., X.Q., Y.W.; Software: Y.Z., M.C. (0000-0001-6577-6705), J.Z.; Formal analysis: Y.Q., W.M., Y.W.; Investigation: Y.Q.; Resources: F.G.; Data curation: Y.Q., Z.S.; Writing - original draft: Y.Q.; Writing - review & editing: F.G.; Supervision: F.G.; Project administration: Y.Q.; Funding acquisition: F.G.
This work was supported by the National Key R&D program funded by the Ministry of Science and Technology of the People's Republic of China (2018YFA0107700); the National Science Fund for Distinguished Young Scholars from the Chinese Academy of Sciences (81525011); the Strategic Priority Research Program of the Chinese Academy of Sciences (XDB19000000); and the National Natural Science Foundation of China (31601193 and 31671496).
The authors declare no competing or financial interests.