MORC2B is essential for meiotic progression and fertility

The microrchidia (MORC) family proteins are chromatin-remodelling factors and function in diverse biological processes such as DNA damage response and transposon silencing. Here, we report that mouse Morc2b encodes a functional germ cell-specific member of the MORC protein family. Morc2b arose specifically in the rodent lineage through retrotransposition of Morc2a during evolution. Inactivation of Morc2b leads to meiotic arrest and sterility in both sexes. Morc2b-deficient spermatocytes and oocytes exhibit failures in chromosomal synapsis, blockades in meiotic recombination, and increased apoptosis. Loss of MORC2B causes mis-regulated expression of meiosis-specific genes. Furthermore, we find that MORC2B interacts with MORC2A, its sequence paralogue. Our results demonstrate that Morc2b, a relatively recent gene, has evolved an essential role in meiosis and fertility.


Author summary
In sexually reproducing organisms, meiosis, a process unique to germ cells, produces haploid gametes. Abnormalities in meiosis can lead to infertility, loss of pregnancy, or genetic diseases such as Down syndrome. The meiotic processes are tightly regulated by a large number of genes including many meiosis-specific ones. The majority of meiosis-specific factors are conserved, however, species-specific factors have evolved. Here we report functional studies of a rodent lineage-specific gene named Morc2b. Morc2b belongs to a family of chromatin-remodelling factors. Morc2b is specifically expressed in germ cells. Disruption of Morc2b causes meiotic arrest and infertility in both sexes. Notably, MORC2B regulates the expression of a number of meiosis-specific genes. Interestingly, MORC2B interacts with its sequence homologue MORC2A. These functional studies have

Introduction
The microrchidia (MORC) protein family forms a conserved class of chromatin remodeling factors found in diverse species from Arabidopsis to human [1]. MORC proteins contain GHKL-type (Gyrase, Hsp90, histidine kinase, MutL) ATPase domain and PHD zinc finger domain, implying functions related to DNA metabolism and epigenetic regulation. Arabidopsis AtMORC1 and AtMORC6 repress transposable elements in a methylation-independent manner and are essential for heterochromatin formation and gene silencing [2]. In mammals, four different MORC proteins (MORC1-4) have been identified [1]. Human MORC2 recruits histone deacetylases to promoter regions, causing local histone H3 deacetylation and transcriptional repression [3,4]. MORC2 also modulates chromatin relaxation in response to DNA damage [5,6]. MORC3 binds to H3K4me3 (trimethylated histone H3 lysine 4) in vitro and localizes to H3K4me3-marked genomic sites [7]. Collectively, these studies reveal a conserved role for MORC proteins in the regulation of high-order chromatin organization. Mutations in MORC2 cause axonal Charcot-Marie-Tooth disease (CMT) in humans [8]. CMT is a neural disorder characterized by muscle weakness and atrophy, and changes in the sensation in the body periphery. In neuronal cells, MORC2 is recruited to heterochromatin by the HUSH (human silencing hub) complex to compact chromatin and thus is required for epigenetic silencing [9]. The HUSH complex mediates H3K9me3 deposition in heterochromatin by SETDB1 (H3K9 trimethyltransferase) to maintain transcriptional silencing [10]. In addition, MORC2 promotes breast cancer invasion/metastasis and gastric tumorigenesis [6]. These studies demonstrate the critical role of MORC2 in human diseases.
Genetic requirements of Morc1 and Morc3 in mouse have been reported. Mouse Morc1, the founding member of the MORC family, is specifically expressed in the male germline and its ablation results in male sterility with meiotic arrest [11,12]. Morc1 deficiency is associated with de-silencing of transposable elements in the male germline [13]. Morc3 -/mice die at birth or within one day after birth [14]. Study of mice heterozygous for a Morc3 mutation reveals a role in bone homeostasis [15]. However, the physiological functions of Morc2 and Morc4 are not known.
In mouse, two paralogues of Morc2 are present: Morc2a and Morc2b. Mouse Morc2b was reported to be a transcriptional target of PRDM9, a histone H3 trimethyltransferase required for meiotic progression and involved in speciation [16]. PRDM9 is the only known mammalian speciation gene [17]. PRDM9 specifies sites of preferred meiotic recombination (i.e. hotspots) and drives recombination away from functional genomic elements such as gene promoter regions [18]. Following sequence-specific DNA binding through its array of zinc fingers, PRDM9 catalyzes trimethylation of H3K4 (H3K4me3) and H3K36 (H3K36me3) [19][20][21][22][23]. In Prdm9-deficient testes, Morc2b is not expressed and H3K4me3 at the Morc2b promoter is low, suggesting that PRDM9 normally induces Morc2b expression via H3K4me3 [16]. However, the function of Morc2b remains unknown. Here, we report that Morc2b is required for chromosomal synapsis and meiotic recombination in both sexes. Inactivation of Morc2b causes mis-expression of a number of genes including meiosis-specific genes. We find that MORC2B interacts with MORC2A. The Morc2b-null mouse mutant exhibits meiotic defects similar to the Prdm9-null mutant, suggesting that MORC2B may be a key downstream effector of PRDM9 in meiosis.

Morc2b is a retrotransposed homologue of Morc2a
Sequence comparison of the five murine MORC members revealed that MORC2A and MORC2B exhibited the highest sequence homology within the family, with 73% amino acid sequence identity (S1A Fig). MORC2A and MORC2B were also the closest homologues according to phylogenetic analysis (S1B Fig). Both MORC2A and MORC2B contain the conserved GHKL-type ATPase and PHD zinc finger domains shared by MORC proteins (S2 Fig). The gene structure of Morc2a and Morc2b differs fundamentally: Morc2a contains 26 introns, whereas Morc2b lacks introns in the coding region ( Fig 1A). This gene structure implies that Morc2b is a retrotransposed homologue of Morc2a that arose from reverse transcription of a processed transcript followed by integration into the genome. Although most retrotransposition events produce truncated or otherwise non-functional pseudogenes, a small number of retrotransposed genes have retained functionality [24,25]. An annotated Morc2a gene (referred to as Morc2 in non-rodent species) is found in more than 100 mammalian species in the NCBI database. In contrast, Morc2b is only present in mouse and rat but not in other nonrodent eutherians, suggesting that the Morc2b retrotransposition event occurred 12-24 million years ago prior to the radiation of mouse and rat ( Fig 1B) [26][27][28]. Western blot analysis with a MORC2B-specific polyclonal antibody showed that MORC2B migrates at the predicted size of 120 kDa (Fig 1C). MORC2B protein was abundant in testes but not detected in other adult mouse tissues examined, whereas MORC2A, migrating at the predicted size of~120 kDa, was highly expressed in both testis and skeletal muscle ( Fig 1C). Thus, the Morc2b gene encodes a bona fide testis-expressed protein and represents a functional retrotransposed gene rather than a pseudogene.

Expression and localization of MORC2B in germ cells
We next examined the spatiotemporal localization pattern of MORC2B in adult testis. MORC2B was detected in germ cells with a distinct developmental-specific expression pattern but not in somatic cells such as Sertoli cells (Fig 1D and 1E). MORC2B was present in meiotic spermatocytes, abundant in post-meiotic haploid round spermatids, and absent from elongated spermatids (Fig 1D). MORC2B localized to the nucleus in spermatocytes and strongly to the nucleus in round spermatids. Absence of immunofluorescence signals in Morc2b-deficient spermatocytes supported the specificity of the MORC2B antibody ( Fig 1E). The observed tissue-and cell type-specific expression of MORC2B and its stage-specific subcellular localization suggest a germ cell-specific nuclear function of Morc2b.
We further examined the expression of Morc2b using juvenile testes (day 8 through day 20) (Fig 1F and 1G). The first wave of spermatogenesis is synchronized [29]. At postnatal day 8, testes contain spermatogonia but no spermatocytes. Pre-leptotene and leptotene spermatocytes first appear at day 10, zygotene spermatocytes at day 12, pachytene spermatocytes at day 14, and round spermatids at day 20. Morc2b expression was absent prior to day 12, was detected at a low level at day 12, and increased significantly at day 14 and beyond ( Fig 1F and  1G). This expression pattern was consistent with the immunofluorescence analysis of MORC2B in adult testis ( Fig 1D). In conclusion, Morc2b is not expressed in spermatogonia, begins to express in zygotene spermatocytes at a low level, and increases expression from pachytene spermatocytes through round spermatids.
The ovaries of adult Morc2b -/female mice were much smaller than those from heterozygous littermates and were devoid of oocytes ( Fig 3A). To determine the time point of oocyte loss, we performed immunofluorescence analysis of ovaries with anti-YBX2 antibodies to label oocytes [32]. Oocytes were present in Morc2b -/ovaries at birth (Fig 3B) but disappeared by postnatal day 2 ( Fig 3C). TUNEL analysis showed increased apoptosis of oocytes in Morc2b -/ovaries at birth ( Fig 3D). Perinatal loss of oocytes was observed in several recombinationdefective mouse mutants (Dmc1, Msh5, Atm, Meiob, or Prdm9) [16,33,34]. The early postnatal loss of oocytes in Morc2b -/mice therefore suggests severe defects in female meiosis. This data is consistent with the expression of Morc2b during meiosis in embryonic ovaries [16]. The Morc2b (previously referred to as 4932411A10Rik) transcript is only present in embryonic ovaries at E13.5 and E14.5, but not at E15.5 and beyond including adulthood [16]. Collectively, our results show that Morc2b is essential for meiosis and fertility in both sexes.

Failure of chromosomal synapsis in Morc2b -/germ cells
We assessed chromosomal synapsis by immunofluorescence analysis of spread nuclei using antibodies against SYCP2, a component of synaptonemal complex (SC) axial elements, and SYCP1, a component of SC transverse elements [35,36]. SC axial elements are formed at the leptotene stage, initiate synapsis through physical juxtaposition at the zygotene stage, achieve full synapsis on autosomes at the pachytene stage, and subsequently separate at the diplotene stage [37]. Wild-type pachytene spermatocytes contained fully synapsed autosomes, whereas the most advanced spermatocytes from Morc2b -/males were at a pachytene-like stage, characterized by apparent chromosome pairing and formation of SC axial elements (SYCP2) but absence of full chromosomal synapsis ( Fig 4A). We quantified spermatocytes at different stages from juvenile wild type and Morc2b -/males and found that diplotene spermatocytes were absent in Morc2b -/males, indicating meiotic arrest at the pachytene-like stage (Fig 5).
We identified similar defects in meiotic progression in Morc2b -/oocytes ( Fig 4B). Female germ cells enter meiosis shortly after sex determination during embryogenesis. At embryonic day 17.5 (E17.5), wild type pachytene oocytes had all 20 chromosome pairs fully synapsed, whereas Morc2b -/ovaries did not contain normal pachytene stage oocytes. The most advanced oocytes were at a pachytene-like stage as characterized by pairing and alignment of chromosomes and absence of extensive synapsis (Fig 4B). The defects in chromosomal synapsis were strikingly similar between Morc2b-deficient spermatocytes and oocytes. These results demonstrate that MORC2B is required for chromosomal synapsis during meiosis in both sexes.

MORC2B is essential for meiotic recombination
To monitor meiotic recombination in Morc2b-deficient spermatocytes, we evaluated the formation of DNA double strand breaks (DSBs) and localization of recombination nodules in spread nuclei of Morc2b -/spermatocytes. During meiosis, following PRDM9-mediated chromatin changes at recombination hotspots, the SPO11 protein catalyses the formation of DSBs,  which elicits the DNA damage response [19,21,[41][42][43], leading to activation of the ATM kinase and subsequent phosphorylation of H2AX (termed γH2AX). At the leptotene and zygotene stages, γH2AX is present on autosomal chromatin and distributed widely throughout the nucleus (Fig 5A). The presence of strong γH2AX signals suggested that DSBs are formed in Morc2b -/leptotene and zygotene spermatocytes (Fig 5B). In normal spermatocytes, γH2AX disappears from the autosomes following meiotic DSB repair and becomes restricted to the XY chromatin during the pachytene and diplotene stages, concomitant with meiotic sex chromatin inactivation (Fig 5A). However, the pachytene-like Morc2b -/spermatocytes remained γH2AX-positive throughout the nucleus and failed to form a sex body (Fig 5B), suggesting a failure in the repair of meiotic DSBs.
Meiotic recombination is executed through coordinated actions of a large number of DNA repair proteins [37]. We examined four single-stranded DNA-binding proteins: RPA, MEIOB, RAD51, and DMC1 (Fig 6). These recombination proteins form distinct foci on meiotic chromosomes. The RPA heterotrimer consists of RPA1, RPA2, and RPA3. RPA binds to ssDNA ends of the meiotic DSBs [44]. MEIOB forms a heterodimer with SPATA22 and interacts with RPA [34,45,46]. RAD51 and DMC1 are recombinases. RAD51 and DMC1 form filaments on RPA-coated ssDNA and direct strand invasion into the homologous chromosome, which is required for crossover formation, homologue pairing, and chromosomal synapsis [47]. Consistent with the formation of DSBs, unsynapsed chromosomes in Morc2b -/spermatocytes contained abundant foci of these recombination proteins (Fig 6). The initial number of RPA2, MEIOB, and RAD51 was similar between Morc2b +/and Morc2b -/spermatocytes at the leptotene stage. With the progression of meiotic recombination, the number of RPA2 and RAD51 foci decreased progressively in control (Morc2b +/-) spermatocytes, however, the number of RPA2 foci was sharply higher in Morc2b -/spermatocytes at both zygotene-like and pachytenelike stages (Fig 6A) and the number of RAD51 foci was higher at the pachytene-like stage (Fig 6C). The number of MEIOB foci also increased in Morc2b -/spermatocytes at the zygotene-like and pachytene-like stages (Fig 6B). These defects further suggested that meiotic DSBs were not repaired in the absence of MORC2B. In contrast, the number of DMC1 foci decreased significantly in Morc2b -/spermatocytes, suggesting defects in strand invasion and/ or stabilization of homologue pairing (Fig 6D). Such defects were consistent with the failure in chromosomal synapsis in Morc2b-deficient germ cells (Fig 4). Furthermore, we did not detect MLH1 foci, representing sites of future crossovers, in Morc2b -/spermatocytes. These results demonstrate that MORC2B is required for meiotic recombination.

Lack of change in the levels of retrotransposons and H3K4me3 in Morc2b -/testes
Loss of MORC1 causes upregulation of retrotransposons in male germ cells [13]. We examined the expression of LINE1 and IAP retrotransposons in Morc2b -/testes. In contrast with the upregulation of LINE1 and IAP in Mov10l1 -/testes (positive control) [48], retrotransposons were not de-silenced in Morc2b -/testes (S3 Fig), implying functional divergence of these two MORC family members.
PRDM9 catalyzes trimethylation of H3K4 and consequently loss of PRMD9 reduces the level of H3K4me3 in male meiotic germ cells [16]. We confirmed the reduced level of H3K4me3 in spermatocytes from Prdm9 -/testes (S4 Fig). Morc2b was reported to be a PRDM9 target gene [16]. We found that the H3K4me3 level was comparable in spermatocytes between wild type and Morc2b -/males (S4 Fig), suggesting that loss of MORC2B is not responsible for reduced H3K4me3 in Prdm9 -/testes.

Misregulated expression of meiosis-specific genes in Morc2b -/testes
Since the MORC family proteins are involved in chromatin remodelling, we sought to examine the transcriptome in Morc2b -/testes by RNA-seq. We chose testes at postnatal day 12 for two reasons. First, Morc2b begins its expression at day 12 ( Fig 1G). Second, the histology of testes is comparable between wild type and Morc2b -/males at day 12, when the most advanced germ cells are at the zygotene stage. Analysis of RNA-seq data identified 71 differentially expressed genes: 57 downregulated and 14 upregulated in Morc2b -/testes (Fig 7A and S1 Table). Seven genes (six downregulated and one upregulated) were chosen for validation by real-time PCR. The differential expression was confirmed for all seven genes at day 12 ( Fig 7B). As expected, their expression was comparable between wild type and Morc2b -/testes at day 10 ( Fig 7B). Gene ontology analysis identified meiotic cell cycle as the most affected biological process ( Fig  7C). Interestingly, two meiosis-specific genes Msh5 and Ccnb3 are downregulated and upregulated respectively. MSH5, a DNA repair protein, is required for chromosomal synapsis [49,50]. CCNB3 is a meiosis-specific cyclin [51]. These data strongly suggest that mis-regulated expression of meiosis-specific genes may contribute to the meiotic defects in Morc2b -/mice.

MORC2B interacts with MORC2A
To identify MORC2B-interacting partners, we performed immunoprecipitation (IP) with testicular protein extracts using the MORC2B antibody. Two protein bands were present in immunoprecipitated proteins from wild type testis but not detected in Morc2b -/testis IP ( Fig  8A). The upper band had the same apparent molecular weight as MORC2B and contained both MORC2A and MORC2B and three other proteins (Fig 8A and S2 Table). Mass spectrometry of the lower unique band identified three proteins: DDX41, HSP72, and ARID3B (S3 Table). We verified the association of MORC2A and MORC2B in testes by co-IP and western blot analyses and confirmed that MORC2B was present in the immunoprecipitated proteins from wild type testes with anti-MORC2A antibody (Fig 8B). Additionally, co-expression and co-IP in HEK 293T cells also validated the association between MORC2A and MORC2B ( Fig  8C). Our results suggest that MORC2B may regulate meiosis through interaction with MORC2A.

Discussion
The MORC family proteins are involved in chromatin remodelling, transcriptional regulation, and transposon silencing. Here we find that Morc2b is required for meiosis and fertility in both males and females. Interestingly, the Morc2b gene evolved in the rodents via retrotransposition from Morc2a. Strikingly, the relatively young Morc2b gene has evolved an essential role in meiosis and fertility, suggesting a strong selection pressure.
Among the Morc gene family, Morc1 and Morc2b are germ cell-specific but exhibit distinct functions. This is evident from differences in the phenotype of the corresponding mouse mutants. The fertility of Morc1 mutant is sexually dimorphic: males are sterile but females are fertile [11,12], whereas inactivation of Morc2b causes sterility in both sexes. The Morc1 mutant phenotype is similar to that of piRNA (Piwi-interacting RNA) pathway mutants: male-only sterility and de-repression of transposable elements (LINE1 and IAP) in male germ cells [13,52]. Since the piRNA pathway appears to be intact in Morc1 mutant germ cells, MORC1 most likely protects genome integrity in male germ cells by silencing transposable elements through Identification of MORC2B-associated proteins in lysates from PND20 mouse testes by IP and mass spectrometry. The gel was stained with SYPRO Ruby. Gel slices corresponding to the two unique protein bands present in wild-type but not in Morc2b -/testes ( Ã~1 10 kDa; ÃÃ~6 5 kDa) were cut from both lanes and subject to mass spectrometry. Proteins with at least three unique peptides are provided in S2 and S3 Tables. (B) Co-IP analysis of MORC2A and MORC2B from PND20 testicular protein extracts. (C) Co-IP analysis of MORC2A and MORC2B in HEK 293T cells. Expression constructs for tagged MORC2A and MORC2B were transfected into HEK 293T cells alone or together as indicated, followed by co-IP with anti-V5 antibody and western blotting with antibodies as indicated. Input was 10% of the lysate used for IP. https://doi.org/10.1371/journal.pgen.1007175.g008

Requirement of MORC2B in meiosis
a different yet unknown mechanism [13]. In contrast, MORC2B deficiency does not cause desilencing of LINE1 and IAP retrotransposons in testes (S3 Fig) but leads to a failure in chromosomal synapsis and meiotic recombination in both sexes.
Human MORC2 plays a critical role in chromatin remodelling in DNA damage response and transcriptional gene silencing [3][4][5][6]. MORC2 acts as a transcriptional repressor of the CAIX gene (carbonic anhydrase IX) by decreasing histone H3 acetylation at the CAIX promoter. MORC2 binds to the CAIX promoter and recruits HDAC4 (histone deacetylase 4) to deacetylate histone H3, which is associated with a repressed chromatin state [3]. Similarly, MORC2 represses p21 in gastric cancer cells by recruiting HDAC1 to the p21 promoter [4]. MORC2 also modulates chromatin configuration during the DNA damage response [5,6]. Upon DNA damage, MORC2 becomes phosphorylated by p21-activated kinase 1 (PAK1), exhibits DNA-dependent ATPase activity, and facilitates chromatin relaxation [5]. Given the known function of MORC proteins in chromatin remodelling, MORC2B might play a role in the regulation of high-order chromatin structure during meiosis.
Loss of MORC2B results in mis-expression of 71 genes in testes. Out of the 71 genes, 30 genes have been disrupted in mice (S1 Table). Twelve knockout mice exhibit sterility or impaired fertility: Adam2, Adam3, Clgn, Crisp1, Fmr1, Krt8, Msh5, Piwil1, Rsph1, Tnp1, Tnp2, and Ybx2. The remaining 18 knockout mice exhibit lethality, or no defects, or somatic defects but normal fertility. Several affected genes are known to play critical roles in meiosis: Msh5, Fmr1, and Ccnb3. Msh5 is downregulated in Morc2b -/testes, whereas Fmr1 and Ccnb3 are upregulated in Morc2b -/testes. MSH5 forms a heterodimer with MSH4 and functions in meiotic recombination. Inactivation of Msh5 causes a failure in chromosomal synapsis and thus meiotic arrest [49,50]. FMR1 localizes to chromatin and regulates DNA damage response [53]. CCNB3 (cyclin B3) is specifically expressed in leptotene and zygotene spermatocytes [51]. Strikingly, mis-expression of the human CCNB3 transgene in mouse spermatocytes after the zygotene stage disrupts spermatogenesis [54]. Therefore, mis-expression of these meiosis genes could contribute to the meiotic defects in Morc2b -/mice. It is possible that MORC2B regulates the transcription of these genes through chromatin relaxation or epigenetic modifications.
Our biochemical studies demonstrate that MORC2B interacts with MORC2A. The interaction among MORC proteins is also present in Arabidopsis. AtMORC6 interacts with AtMORC1 and AtMORC2 in two mutually exclusive protein complexes [55]. Both AtMORC1 and AtMORC2 are needed to repress the set of genomic loci silenced by AtMORC6. The interaction between MORC proteins and the non-redundant nature of their functions are conserved between Arabidopsis and mouse and possibly so in other species. In addition, MORC2B may function through other interacting proteins such as ARID3B -a member of the ARID (AT-rich interaction domain) family of DNA-binding proteins (S3 Table) [56,57]. Genetic studies of MORC2A and ARID3B in germ cells are not available yet but are necessary to determine their functional requirement for meiosis and the physiological significance of their interaction with MORC2B. As DNA-dependent ATPases, MORC proteins have been found to modulate chromatin superstructure in DNA damage response, heterochromatin formation, and gene silencing. Further studies are necessary to elucidate a possible role of MORC2B in chromatin remodelling in the regulation of meiosis.
PRDM9, a meiosis-specific histone H3 methyltransferase, is a major determinant of meiotic recombination hotspots in mice, primates, and humans [19][20][21]. PRDM9 binds to recombination hotspots in a sequence-specific manner through its variable number of zinc fingers. Disruption of Prdm9 results in meiotic failure and sterility in both males and females [16]. The cause of meiotic failure in Prdm9-null mice has not been identified, and the relationship between control of meiotic recombination hotspots and meiotic progression remains unclear.
While PRDM9 catalyses H3K4me3 at hotspots, it also affects the expression of one target gene-Morc2b. In Prdm9-deficient testes, Morc2b is not expressed [16]. Morc2b expression is also nearly absent in sterile hybrids of mouse subspecies [17]. Furthermore, both Morc2b and Prdm9 mutants exhibit a failure in chromosomal synapsis and meiotic recombination. The similar phenotype of these two mutants raises the intriguing possibility that the absence of Morc2b might be responsible for or at least contribute to the meiotic failure in Prdm9-null mice.

Ethics statement
Mice were maintained and used for experimentation according to the protocol approved by the Institutional Care and Use Committee of the University of Pennsylvania.

Targeted inactivation of the Morc2b gene
In the targeting vector, the 3.2-kb Morc2b coding exon was replaced with the PGKNeo selection cassette (Fig 2A). The two homologous arms were amplified from a Morc2b-containing BAC clone (RP24-63E7) by high-fidelity PCR. The HyTK negative selection cassette was cloned after the right arm. V6.5 embryonic stem (ES) cells (on a C57BL/6 x 129S4/SvJae hybrid background) were electroporated with the ClaI-linearized targeting vector. ES cells were cultured in the presence of 350 μg/ml G418 and 2 μM ganciclovir. Out of 96 double-resistant ES cell clones, nine targeted clones were identified by long-distance PCR. Clone 1F3 was injected into blastocysts. Germline transmission of the knockout allele was obtained through breeding of chimera males with C57BL/6 females. Offspring of intercrosses of Morc2b +/mice were used for all the analyses. Wild-type allele (220 bp) was assayed by PCR with primers TGCACTGAA CCCGACACTAC and GGTAGGAGCGGCAGAGATTC. The Morc2b mutant allele (415 bp) was assayed by PCR with primers ATAGCAGGCATGCTGGGGATGCGGT and TGCACC TACACCAGGCAGCTCAGG.

Histological, immunofluorescence, and surface nuclear spread analyses
For histological analysis, testes and ovaries were fixed in Bouin's solution, embedded with paraffin, and sectioned. Sections were stained with haematoxylin and eosin. Color histological images were captured on a Leica DM5500B microscope with a DFC450 digital color camera (Leica Microsystems). For immunofluorescence and TUNEL analysis, testes and ovaries were fixed in 4% paraformaldehyde for 3 h or overnight at 4˚C, dehydrated, embedded, and sectioned using a cryostat. TUNEL assays were performed with the ApopTag Fluorescein In Situ Apoptosis Detection Kit (Catalogue number S7110, EMD Millipore). Nuclear spread analysis of spermatocytes and oocytes was performed as previously described [58,59]. The following antibodies were used for immunofluorescence: SYCP1 (1:50, catalogue number ab15090, Abcam), SYCP2 [35] [16]. Images were acquired under the same condition. The relative intensity of H3K4me3 fluorescence signal was quantified using the ImageJ software. One pachytene or pachytene-like spermatocyte and one Sertoli cell were randomly selected from each tubule cross-section (10 tubules/genotype). The relative H3K4me3 signal intensity in the spermatocyte was normalized to that in the Sertoli cell.

RNA-seq, DESeq analysis, and real-time PCR validation
Total RNA was isolated from eight pairs of postnatal day 12 mouse testes (~16 mg/pair; 4 pairs of wild type and 4 pairs of Morc2b -/-) using TRIzol reagents (Thermo Fisher Scientific). The RNA concentration was determined using a NanoDrop 2000 Spectrophotometer (Thermo Fisher Scientific). Equal amounts (1 μg) of total RNA from each sample were used to generate RNA-seq libraries using TruSeq Stranded mRNA Library Preparation Kit Set A (Cat. No. RS-122-2101, Illumina) according to the manufacturer's instruction. The concentration of DNA library templates was determined using a Qubit 3.0 Fluorometer (Thermo Fisher Scientific). The quality of libraries was evaluated using the Agilent 4200 TapeStation (Agilent Technologies). Eight individual libraries (4 wild type and 4 Morc2b -/-) were pooled in equal amounts for sequencing using the Illumina NextSeq 500/550 High Output v2 kit (Illumina, 75 cycles, FC-404-2005) and the NextSeq 500 system (Illumina). The RNA-seq data are available under GEO accession no: GSE103127.
After trimming the adapter sequences and removing the low-quality reads, the clean reads were mapped to the mouse reference genome (NCBI37/mm9) using Tophat with default parameters. Mapped reads were subjected to Cufflinks to estimate gene expression levels [60]. The expression of each gene was normalized by calculating fragments per kilobase of exon per million fragments mapped (FPKM). The FPKM values of each gene for both wild type and Morc2b -/group were used to assess the differential expression with Cuffdiff. The expression cutoff of ! 1FPKM in either wild type or Morc2b -/testes was applied. Differentially expressed genes were determined by an adjusted P value (false discovery rate, FDR) < 0.05 based on Benjamini and Hochberg multiple testing correction. A volcano plot was constructed to illustrate the differentially expressed genes by plotting log2 of the fold change on the X axis and the negative log10 of the p value on the Y axis (Fig 7A).
The expression of seven differentially expressed genes was analyzed using independent testis samples from postnatal day 10 and 12 mice (3 testis samples per genotype per time point) by real-time PCR. Expression of LINE1 and IAP retrotransposons in wild type and Morc2b -/testes at postnatal day 14 was assayed by real-time PCR (S3 Fig). Postnatal day 14 Mov10l1 -/testes were used as a positive control for de-silencing of LINE1 and IAP [48,61]. Real-time PCR primers are listed in S4 Table. Each sample was assayed in triplicates. Quantification was normalized to Actb using the Ct method (ABI Prism 7700 Sequence Detection System, Applied Biosystems).

Co-immunoprecipitation, mass spectrometry, and transfection constructs
Co-immunoprecipitation was performed with postnatal day 20 mouse testes using affinitypurified MORC2B antibodies as previously described [34]. Immunoprecipitated proteins were resolved by SDS-PAGE. The protein bands specific to the wild type testis sample were subjected to mass spectrometry for protein identification.

Statistics
Statistical analysis was performed with Student's t-test or χ 2 test. and Sertoli cells were quantified using Image J. The Y axis shows the relative intensity (spermatocyte/Sertoli cell). Ã , statistically significant (p < 0.05, Student's t-test); ns, non-significant. (TIF) S1