The male germline-specific protein MAPS is indispensable for pachynema progression and fertility

Significance

During meiosis prophase I in mammals, a series of events occur, such as homologous recombination, DSB repair, crossover formation, XY body formation, and meiotic sex chromosome inactivation (MSCI). We report here that a male pachynema-specific (MAPS) protein is essential for prophase I progression in mouse spermatocytes. Maps knockout causes the demise of pachytene spermatocytes throughout early to mid- and late pachynema, resulting in male but not female infertility. Moreover, protein ubiquitination was dramatically increased in Maps^−/− pachytene spermatocytes, likely contributing to the dysregulated global gene expression, including failed MSCI. In summary, MAPS is an indispensable protein for male fertility.

Abstract

Meiosis is a specialized cell division that creates haploid germ cells from diploid progenitors. Through differential RNA expression analyses, we previously identified a number of mouse genes that were dramatically elevated in spermatocytes, relative to their very low expression in spermatogonia and somatic organs. Here, we investigated in detail 1700102P08Rik, one of these genes, and independently conclude that it encodes a male germline-specific protein, in agreement with a recent report. We demonstrated that it is essential for pachynema progression in spermatocytes and named it male pachynema-specific (MAPS) protein. Mice lacking Maps (Maps^−/−) suffered from pachytene arrest and spermatocyte death, leading to male infertility, whereas female fertility was not affected. Interestingly, pubertal Maps^−/− spermatocytes were arrested at early pachytene stage, accompanied by defects in DNA double-strand break (DSB) repair, crossover formation, and XY body formation. In contrast, adult Maps^−/− spermatocytes only exhibited partially defective crossover but nonetheless were delayed or failed in progression from early to mid- and late pachytene stage, resulting in cell death. Furthermore, we report a significant transcriptional dysregulation in autosomes and XY chromosomes in both pubertal and adult Maps^−/− pachytene spermatocytes, including failed meiotic sex chromosome inactivation (MSCI). Further experiments revealed that MAPS overexpression in vitro dramatically decreased the ubiquitination levels of cellular proteins. Conversely, in Maps^−/− pachytene cells, protein ubiquitination was dramatically increased, likely contributing to the large-scale disruption in gene expression in pachytene cells. Thus, MAPS is a protein essential for pachynema progression in male mice, possibly in mammals in general.

Meiosis is a conserved and specialized form of cell division that produces haploid gametes for sexual reproduction. Meiosis prophase I is a relatively long period during which a series of orderly events occur. Programmed DNA double-strand breaks (DSBs) catalyzed by the topoisomerase-like protein SPO11 are formed at leptonema (1). The DSBs then undergo 3′ end resection to generate 3′ single-stranded DNA overhangs that are protected by the heterotrimeric RPA (replication protein A) complex. This event is followed by loading of the RAD51 recombinase and DNA meiotic recombinase 1 along with other proteins to enable strand invasion into duplex DNA for homolog pairing and recombination intermediate formation (2–4). All DSBs on autosomes are repaired, but only a subset leads to crossovers (5). When autosomes complete synapsis at pachynema, the X and Y chromosomes remain largely unsynapsed and compartmentalized into a nuclear subdomain called the XY body or sex body (6). In response to asynapsis, the X and Y chromosomes are subjected to chromatin modifications leading to transient transcriptional silencing in a process termed meiotic sex chromosome inactivation (MSCI) (6, 7).

Over the past decades, the underlying molecular aspects of meiosis prophase I have been widely studied. Various mutant mouse models have been developed for studying the key events during prophase I progression through the pachytene stage (for reviews, see refs. 8–10). However, the machineries regulating this progression are still not fully understood (11, 12). Transcriptomes during meiosis in males have been reported, but the regulatory factors remain poorly described (13, 14). Nevertheless, epigenetic regulation is thought to be of great importance for the progression of meiosis prophase I, and the role of ubiquitination represents a topic of interest (15–17). For instance, a recent study reported that ubiquitinated proteins localized on the chromosome axis might regulate the stabilization and degradation of recombination factors in order to enable proper synapsis, DSB repair, and crossover formation (16).

Wu et al. (18) have reported that the gene 1700102P08Rik is specifically expressed in mouse testes, and the knockout male mice were infertile. To identify novel genes regulating meiosis prophase I progression in spermatocytes, we previously performed RNA sequencing of spermatogonia, leptotene/zygotene spermatocytes, pachytene spermatocytes, and round spermatids isolated by bovine serum albumin gradient sedimentation (19, 20). A number of genes were found to be highly expressed in spermatocytes relative to spermatogonia and somatic tissues, including genes, such as 4930432K21Rik/Mamerr (male meiosis recombination regulator) and Spdya (speedy/RINGO cell cycle regulator family, member A), both of which we had further studied (21, 22). In this report, we selected 1700102P08Rik for detailed analyses and independently concluded that it is a male germline-specific protein. We found that its RNA and protein were primarily expressed in high levels in pachytene spermatocytes. We named this protein MAPS (male pachynema-specific) protein because it is indispensable for the progression of pachynema.

This protein is of 252 amino acids but has no predicted functional domain through analyses by an online protein domain predictor, ThreaDom (23). To investigate its function, we independently knocked out this gene and generated polyclonal antibodies. Indirect immunofluorescence (IF) examination confirms that MAPS protein is expressed specifically during pachynema progression. Maps knockout caused pachytene arrest in spermatocytes. The defects occur at different stages during the pachynema progression in pubertal as opposed to adulthood mice. Nonetheless, they all lead to germ cell death and male infertility. Finally, by detailed molecular studies, we showed that Maps^−/− spermatocytes exhibited extensive dysregulation in gene expression from both sex chromosomes and autosomes. We conclude that MAPS plays an indispensable role in regulating gene expression during meiosis pachytene progression in male mice. We observed a substantial reduction in ubiquitination when MAPS was ectopically expressed in vitro. Conversely, in Maps^−/− spermatocytes, protein ubiquitination was greatly elevated. We propose that alternations in ubiquitination of cellular proteins are likely to have contributed to these changes in gene expression.

Results

Maps Is a Testis‐Specific Gene That Is Highly Expressed in Meiotic Spermatocytes at the Pachytene Stage.

First, we isolated spermatocytes form pubertal mice PD17 (postnatal day 17) and adulthood PD60–90. Using RT-PCR (reverse transcription polymerase chain reaction) and qRT-PCR (quantitative real-time PCR) targeting exon 1 to exon 3 of Maps, we confirmed its high expression in pachytene spermatocytes relative to other stages of sperm development (Fig. 1A and SI Appendix, Fig. S1A). RT-PCR and qRT-PCR analyses also established that Maps mRNA (messenger RNA) was specifically expressed in mouse testes but not in embryonic day 17.5 (E17.5) or adult ovaries, nor in other organs (SI Appendix, Fig. S1 B and C).

Fig. 1.

MAPS expression and localization in meiotic germ cells. (A) RT-PCR analysis of Maps expression in isolated male germ cells, including two types of spermatogonia (A and B), preleptotene, leptotene/zygotene, PD17 and PD90 pachytene spermatocytes, and round spermatids, demonstrating that Maps was highly expressed in pachytene spermatocytes. Maps^−/− testes were negative for Maps mRNA. β-actin was used as an internal RT-PCR control. (B and C) IF analysis of MAPS (red) and γH2AX (green) in testis sections from Maps^+/+ (B) and Maps^−/− (C) mice. MAPS was specifically expressed in pachytene cells (B, arrows; with γH2AX accumulated on XY body) but not in leptotene cells (B, arrowheads; with γH2AX expressed in the nuclei). MAPS (red) signal was absent in Maps^−/− testis sections (C), confirming the specificity of the MAPS antibody. (Scale bars, 50 μm.) (D–I) chromosome spreads of leptotene to diplotene spermatocytes from PD35 Maps^+/+ testes stained for MAPS (green) and SYCP3 (red), showing that MAPS was diffusely expressed in the nuclei from early pachytene to diplotene stages (F–I), and particularly high signals were seen in the XY body from mid-pachytene to diplotene stages (G–I, dashed circles). (Scale bars, 5 μm.) (J and K) A Maps^+/+ spermatocyte at late pachytene was stained for MAPS (green), γH2AX (blue), and SYCP3 (red), showing that MAPS colocalized with γH2AX in the XY body (dashed circles). (Scale bars, 5 μm.) (L) No MAPS signal was detected in a Maps^−/− spermatocyte at pachytene that was stained for MAPS (green) and SYCP3 (red). (Scale bars, 5 μm.)

We generated a polyclonal antibody against MAPS protein. Immunoblots showed that the MAPS protein was present in both nucleus and cytoplasm fractions in mouse testes (SI Appendix, Fig. S1D). We then conducted indirect IF assay on paraffin sections and again found that MAPS (in red) was not detected in testicular somatic cells, nor in leptotene spermatocytes where diffused γH2AX (phosphorylation of histone H2AX at serine 139, in green) marked DSBs were seen throughout the entire nucleus (Fig. 1B, arrowheads). Rather, MAPS was detected in spermatocytes at the pachytene stages (Fig. 1B, arrows) when condensed γH2AX signals were found in the XY body area (Fig. 1J).

We next performed chromosome spreading to observe chromosomal behaviors throughout meiotic prophase I. We stained MAPS (in green) along with synaptonemal complex protein 3 (SYCP3), which marks chromosome axes (in red), and confirmed that MAPS was not observed in the nuclei of leptotene or zygotene spermatocytes (Fig. 1 D and E). It was first detected in early pachytene spermatocytes (Fig. 1F) with a widespread localization in the whole nuclei, and this pattern persisted to the diplotene stage (Fig. 1 G–I). Also, well-defined and condensed MAPS signals were seen in the XY body area from mid-pachytene to diplotene stage (Fig. 1 G–I, dashed circles). Notably, MAPS and γH2AX were colocalized in the XY body (Fig. 1 J and K, dashed circles). This colocalization can be better visualized in the three-dimensional movie (Movie S1). Collectively, these results suggest that Maps is a male-specific gene related to reproduction.

Maps Is Essential for Mouse Spermatogenesis and Male Fertility.

The mouse Maps gene is located on chromosome 9. The MAPS protein is highly conserved in mice, rats, and humans, and protein sequence comparison showed a 75% identity between mice and humans and 91% identity between mice and rats (SI Appendix, Fig. S1E). To determine the functional roles of MAPS in germ cells, we used the CRISPR/Cas9 method to generate a Maps knockout mouse model by deleting a genomic DNA fragment of 1,841 bp (base pair) spanning exon 3 (Methods and SI Appendix, Fig. S2A). RT-PCR did not detect Maps mRNA (Fig. 1A, Top), and indirect IF did not detect MAPS protein in Maps^−/− testis sections (Fig. 1C, arrows, arrowheads). Additionally, chromosome spreading confirmed the absence of MAPS in the nuclei of PD35 Maps^−/− spermatocytes at the pachytene stage (Fig. 1L compared with Fig. 1 F–H). Immunoblot analysis also confirmed the absence of MAPS in Maps^−/− testes (Fig. 2A).

Fig. 2.

MAPS is required for mouse spermatogenesis and maintenance of male fertility. (A) Immunoblot analysis of MAPS in Maps^+/+ and Maps^−/− testes, showing that MAPS was absent from PD19 Maps^−/− testes. β-actin was used as a loading control. (B) Testes from PD30 and PD67 Maps^+/+ and Maps^−/− mice. Maps^−/− males had reduced testis size. (Scale bars, 2 mm.) (C–H) Histological analyses of testes from PD28 (C and D), PD60 (E and F), and PD90 (G and H) Maps^+/+ and Maps^−/− mice. Arrowheads indicate round and elongated spermatids in Maps^+/+ testis, whereas Maps^−/− spermatocytes were arrested at the pachytene stage (arrows). (Scale bars, 50 μm.) (I and J) Histological analyses of epididymides from PD90 Maps^+/+ and Maps^−/− mice. Arrows indicate sperms in Maps^+/+ male, whereas no sperm was observed in Maps^−/− male. (Scale bars, 100 μm.) (K–Q) Chromosome spreads of spermatocytes from testes of PD30 Maps^+/+ (K–N) and Maps^−/− (O–Q) males immunostained for SYCP1-N (green) and SYCP3 (red). Maps^−/− spermatocytes showed seemingly normal synapsis but were arrested at the pachytene stage. Arrows indicate synapsed regions of chromosome axes. (Scale bars, 5 μm.)

Maps^−/− males were sterile, in agreement with the report by Wu et al. (18), and their testes were smaller than those of Maps^+/+ mice at PD30 and PD67 (Fig. 2B). Histological analysis showed that compared with PD28, PD60, and PD90 Maps^+/+ testes, where various stages of differentiated spermatocytes, round spermatids, and elongated spermatids were seen (Fig. 2 C, E, and G, arrowheads), the most advanced cells found in Maps^−/− seminiferous tubules were likely pachytene cells (Fig. 2 D, F, and H, arrows), and no spermatids were observed. At PD90, normal sperms were seen in the epididymides of Maps^+/+ mice (Fig. 2I, arrows), whereas Maps^−/− mice showed a complete absence of sperms (Fig. 2J).

Maps Knockout Causes Pachytene Arrest with Seemingly Intact Synapsis in Pubertal Mice.

To characterize the defects in Maps^−/− spermatocytes, we performed chromosome spreading and systematically analyzed prophase I progression using PD30 testes. As the cell cycle progressed from the leptotene to zygotene stage in Maps^+/+ mice, the signals of SYCP1 (N terminus) started to appear at the midline of the SYCP3-stained axes at zygotene stage (Fig. 2 K and L, arrow). Autosomes completed synapsis at the pachytene stage (Fig. 2M, arrow), and spermatocytes proceeded to the diplotene stage where autosomes began to desynapse (Fig. 2N). The Maps^−/− spermatocytes exhibited a similar pattern of cell cycle progression from the leptotene to pachytene stage, where the assemblies of N-terminal ends of SYCP1 dimers were observed in zygotene and pachytene spermatocytes (Fig. 2 P and Q, arrows). However, PD30 Maps^−/− spermatocytes failed to proceed to the diplotene stage and remained at the pachytene stage (Fig. 2Q).

We quantified the cell cycle progression of pubertal PD30 spermatocytes and found that Maps^+/+ spermatocytes contained 63% pachytene and 16% diplotene cells, whereas Maps^−/− spermatocytes contained only 39% pachytene cells and no diplotene cells (Fig. 3A). The percentages of cells in Maps^−/− mice at the leptotene (25%) and zygotene (36%) stages were higher than those in Maps^+/+ mice (8% and 13%, respectively) (Fig. 3A). Thus, the meiotic cell cycle progression appeared retarded and arrested at the pachytene stage in pubertal Maps^−/− spermatocytes.

Fig. 3.

Maps knockout causes pachytene arrest with defective crossover formation, DSB repair, and XY body formation in pubertal mice. (A) Analyses of meiotic stages in Maps^+/+ and Maps^−/− spermatocytes from PD30 testes. n = 462 for Maps^+/+ spermatocytes, n = 360 for Maps^−/− spermatocytes. Percentages of cells at indicated meiosis stages were shown. For each genotype, three mice were analyzed. (B and C) Maps^+/+ and Maps^−/− spermatocytes immunostained for MLH1 (green, arrows) and SYCP3 (red), showing that Maps^−/− spermatocytes lack the MLH1 signal. (Scale bars, 5 μm.) (D) Scatter plot of MLH1 foci numbers per cell in Maps^+/+ (black) and Maps^−/− (red) spermatocytes. Maps^−/− spermatocytes showed no foci of MLH1 (red). Solid lines showed the mean ± SD of foci number in each group of spermatocytes. P values were calculated by Student’s t test. (E–H) Chromosome spreads of spermatocytes from testes of PD30 Maps^+/+ (E) and Maps^−/− (F–H) males immunostained for SYCP3 (red) with a focus of observing XY bodies (arrows). Based on the morphologies of XY body, cells were classified into four categories, including normal XY body (E) where PAR indicates the pseudoautosomal region, XY asynapsis (F), XY chromosomes synapse beyond PAR (G), and extended X and Y chromosomes (H). (Scale bars, 2 μm.) (I–L) γH2AX detection in Maps^+/+ (I) and Maps^−/− (J–L) pachytene spermatocytes from PD30 testes immunostained for SYCP3 (red) and γH2AX (green), showing abnormal γH2AX signals in sex body area (J and K, dashed circles), or retained and mislocalized γH2AX signals as flocs or foci on autosomes (J and L, arrows) of Maps^−/− spermatocytes. (Scale bars, 5 μm.) (M–R) IF analyses of testes from PD20 Maps^+/+ (M–O) and Maps^−/− (P–R) males immunostained for cleaved-PARP (red) and H1t (green). H1t signals were strongly positive in Maps^+/+ testes (M–O, arrows), and cleaved-PARP signals were observed only in Maps^−/− spermatocytes (P–R, arrows). (Scale bars, 50 μm.)

Maps^−/− Spermatocytes in Pubertal Mice Lack Crossover and Exhibit Defective DSB Repair and XY Body Formation.

To determine whether the developmental arrest in Maps^−/− spermatocytes was caused by defects in homologous recombination, we next probed the distribution of recombination proteins RPA2 and RAD51. We did not find any significant difference in either signals on chromosomes at the zygotene or pachytene stages between Maps^+/+ and Maps^−/− mice (SI Appendix, Fig. S2 B–F and G–K). Thus, the initial recruitment of these proteins was normal in Maps^−/− cells.

During recombination, DSBs are repaired by either crossover or noncrossover pathways. We conducted IF assays for MLH1 (mutL homolog 1) (green) and SYCP3 (red) (Fig. 3 B and C) in PD30 spermatocytes. Compared with Maps^+/+ pachytene spermatocytes (Fig. 3B, marked by arrows, and Fig. 3D), there was a complete absence of MLH1 signal in Maps^−/− spermatocytes (Fig. 3 C and D) indicating that crossover did not occur.

In PD30 Maps^+/+ pachytene cells, autosomes were fully synapsed, and the X and Y chromosomes were paired in the PAR (pseudoautosomal region) to form the XY body (Fig. 3E, arrows). We found that 80.2% (85 out of 106 cells) of Maps^−/− pachytene spermatocytes exhibited abnormal bivalent XY. We observed 1) asynapsed X and Y chromosomes, with the two chromosomes being far away from each other, representing 6.6% of observed PD30 pachytene cells (7/106) (Fig. 3F, arrows); 2) X and Y chromosomes synapsed beyond the PAR, and in some cases, the entire Y chromosome axis synapsed with the X chromosome, representing 27.4% of observed pachytene cells (29/106) (Fig. 3G, arrows); and 3) synapsed XY bivalent chromosomes remained extended instead of condensing to form the XY body, representing 46.2% of observed pachytene cells (49/106) (Fig. 3H, arrows). We speculated that MAPS may have a function in XY body formation and MSCI. In addition, since pubertal Maps^−/− spermatocytes were arrested at early pachytene stage (Fig. 3 M–R), the abnormal morphologies of XY chromosomes might also be a result of cell cycle arrest.

At the molecular level, γH2AX signals in Maps^+/+ pachytene cells disappeared from autosomes but were observed in a well-defined area of the XY body (Fig. 3I, dashed circle). In contrast, in most Maps^−/− spermatocytes, we observed abnormal γH2AX signals on XY chromosomes. Specifically, the γH2AX signals were poorly defined and irregular in the XY body area (Fig. 3J, dashed circle), or as weaker γH2AX signals alongside the XY chromosomes (Fig. 3K, dashed circle). Furthermore, the abnormal γH2AX signals were retained on Maps^−/− autosomes and were mislocalized as flocs (Fig. 3J, arrow) or foci (Fig. 3L, arrows). These results strongly support the notion that MAPS is important for proper XY body formation in pachytene spermatocytes in pubertal mice.

Maps Deletion Is Associated with the Arrest of Spermatocytes at the Early Pachytene Stage and with Significant Apoptosis in Pubertal Mice.

During meiosis prophase I, pachytene stage lasts the longest and is subdivided into three stages according to the chromosomal behaviors: the “early,” “mid-,” and “late” pachynema. As XY body was formed at mid-pachytene stage, we speculated that pubertal Maps^−/− pachytene spermatocytes were arrested at early pachytene and could not reach mid-pachytene stage. H1t (testis-specific histone variant) is a useful marker for pachytene stage as it became detectable in the entire nuclear region from mid-pachytene stage and increases at late pachytene and diplotene stages (24).

To determine the fate of spermatocytes arrested at the pachytene stage, we performed IF assays on testis sections in PD20 using antibodies against H1t and cleaved-PARP [poly (ADP ribose) polymerase], a marker of apoptosis. In Maps^+/+ testes, almost all tubules contained many H1t-positive spermatocytes but no apoptotic cells (Fig. 3 M–O, arrows). On the contrary, in Maps^−/− testes, the spermatocytes expressed very weak H1t signals (Fig. 3 P and R, arrows), while numerous apoptotic germ cells were seen as indicated by cleaved-PARP signals (Fig. 3 Q and R, arrows). We concluded that, in young pubertal testes, Maps^−/− spermatocytes were arrested at early pachytene stage and then perished.

Adult Maps^−/− Spermatocytes Show Normal XY Body Formation and Partly Correct the Defects in Crossover Formation Observed in Pubertal Spermatocytes but Still Exhibit Pachytene Arrest.

We compared the percentages of spermatocytes at different stages of prophase 1 progression. In adult (PD60) Maps^−/− mice, percentages of cells in leptotene, zygotene, and pachytene stages were all higher than those in Maps^+/+ cells (SI Appendix, Fig. S3A). In contrast, there were only 5% diplotene cells, in comparison to the 39% diplotene cells in the Maps^+/+ mice (SI Appendix, Fig. S3 A, E, and I). At PD78, chromosome spreading showed that the synapsis process appeared normal in Maps^−/− spermatocytes (SI Appendix, Fig. S3 F–H, compared with SI Appendix, Fig. S3 B–D in Maps^+/+ spermatocytes).

We then systematically analyzed the homologous recombination and pachytene progress in adult (PD80) Maps^−/− and Maps^+/+ mice by staining for RPA2 (SI Appendix, Fig. S4 A–G, in green), RAD51 (SI Appendix, Fig. S5 A–G, in green), and H1t (SI Appendix, Figs. S4 C and F and S5 C and F, in blue, SI Appendix, Figs. S4 A, B, D, and E and S5 A, B, D, and E, H1t-negative) in adult spermatocytes. We found that the foci numbers and distributions of RPA2 and RAD51 were comparable. H1t signals also appeared similar in Maps^+/+ and Maps^−/− mid- to late pachytene cells.

For MLH1 staining (Fig. 4 A and B, in green, arrows), in comparison to around 20 crossover foci in Maps^+/+ pachytene cells, we observed a mean of 13 MLH1 foci in PD130 Maps^−/− spermatocytes (Fig. 4C). Notably, 19.2% of Maps^−/− spermatocytes (10/52) completely lacked MLH1 signals, 25.0% of Maps^−/− spermatocytes (13/52) exhibited fewer than 15 MLH1 foci, and only 48.1% of Maps^−/− spermatocytes (25/52) formed crossovers that were comparable to Maps^+/+ cells. Thus, the defect in crossover formation observed in pubertal Maps^−/− spermatocytes is partly corrected in adult mice. These observations imply that the meiosis progression is progressively halted during the mid- to late pachytene stage in Maps^−/− spermatocytes. We then examined DSB repair in PD90 pachytene spermatocytes by costaining γH2AX, SYCP3, and H1t. Interestingly, relative to Maps^+/+ cells, in mid- to late Maps^−/− pachytene cells, the distribution pattern of γH2AX in the sex body was comparable to the wild-type cells (Fig. 4 D and E).

Fig. 4.

MAPS deletion is associated with the arrest and apoptosis of spermatocytes at the pachytene stage in adult male testes, with defective crossover but seemingly normal DSB repair and XY body formation. (A and B) Chromosome spreads of spermatocytes from the testes of PD130 Maps^+/+ and Maps^−/− mice immunostained for MLH1 (green), SYCP3 (red), and H1t (blue). Representative images are shown. Arrows indicate MLH1 foci on the chromosome axes. (Scale bars, 5 µm.) (C) Quantifications of the foci numbers of MLH1 in Maps^+/+ (black dots) and Maps^−/− (red dots) spermatocytes. Solid lines show the mean ± SD of foci numbers in each group. P values were calculated by Student’s t test. (D and E) Chromosome spreads of spermatocytes from the testes of PD90 Maps^+/+ and Maps^−/− mice immunostained for γH2AX (green), SYCP3 (red), and H1t (blue). The XY body with positive γH2AX signals appears normal in Maps^−/− spermatocytes. Representative images of spermatocytes at the mid- to late pachytene stage are shown. (Scale bars, 5 µm.) (F–H) IF analyses from testes of PD134 Maps^+/+ males immunostained for cleaved-PARP (red) and H1t (green). Most cells appeared H1t-positive in the seminiferous tubules (arrows), and apoptosis was rarely observed (arrows). (I–N) IF analysis from testes of PD134 Maps^−/− males immunostained for cleaved-PARP (red) and H1t (green). (I–K) Twenty-four percent of Maps^−/− tubules remained H1t-negative, and these H1t-negative spermatocytes showed signs of apoptosis, with positive cleaved-PARP signals (arrows). (L–N) Seventy-six percent of Maps^−/− tubules contained numerous H1t-positive cells (arrows), and many of these H1t-positive cells also appeared to undergo apoptosis (arrows). (Scale bars, 50 μm.)

Adult Maps^−/− Spermatocytes Undergo Subsequent Apoptosis.

Since adult Maps^−/− spermatocytes develop normal XY bodies, and a fraction of early pachytene cells can progress to mid-pachytene stage, albeit with a delay, then what is the cause for male infertility? To address this question, we performed IF assay to assess H1t and cleaved-PARP staining patterns on paraffin-embedded sections of PD134 testes. In Maps^+/+ testes, most cells appeared H1t-positive in the seminiferous tubules (14/14) (Fig. 4 F and H, arrows), and apoptosis was rarely observed (Fig. 4 G and H, arrows). In Maps^−/− testes, however, around 24% seminiferous tubules observed (6/25) appeared H1t-negative (Fig. 4 I and K), suggesting that a fraction of early pachytene cells were still arrested in adult mice. For these tubules, many H1t-negative spermatocytes showed signs of apoptosis, being positive for the cleaved-PARP (Fig. 4 J and K, arrows). Although 76% of tubules (19/25) contained numerous H1t-positive cells (Fig. 4 L and N, arrows), many of these cells also appeared to undergo apoptosis, as suggested by the colocalized cleaved-PARP (Fig. 4 M and N, arrows).

Adult Maps^−/− Pachytene Spermatocytes Exhibit Retarded Progression from Early to Mid- and Late Pachynema.

To investigate why the defects in Maps^−/− spermatocytes occur at different phases of pachytene progression in pubertal and adult mice, we systematically examined pachytene progression in Maps^+/+ and Maps^−/− mice of different age. We used H1t expression as a valid marker for progression from early to mid- and late pachynema. We found that in Maps^+/+ pachytene cells, at PD17, 70.0% of the cells were H1t-positive. This proportion increased to 81.0% in PD30 and 82.0% in PD40 (Fig. 5A), and to 88.4% in PD50–80 (Fig. 5A), while H1t-negative cells correspondingly decreased. Spermatogenesis is a progressive process from spermatogonia to elongated spermatids in waves throughout the reproductive lifetime in male mice, and there exist different waves and stages of spermatocytes in different seminiferous tubules of the same testis, including cells at early pachytene stage (H1t-negative), and mid- to late pachytene cells (H1t-positive) (25, 26). Therefore, although all pachytene cells in Maps^+/+ mice would eventually reach late pachytene stage (i.e., H1t-positive), they would do so in an unsynchronized manner (25, 26). Thus, starting from PD17 onwards, the majority of the first-wave pachytene spermatocytes had reached mid-pachytene stage.

Fig. 5.

Adult Maps^−/− pachytene spermatocytes exhibit retarded progression from early to mid- and late pachynema. (A and B) Percentages of H1t-negative and H1t-positive pachytene spermatocytes from testes of Maps^+/+ (A) and Maps^−/− (B) mice at a range of ages. The numbers of total cells counted were also indicated. The white bars indicate the percentages of H1t-negative cells, and the striped (A) or black bars (B) indicate the percentages of H1t-positive cells. For each genotype and age, at least three mice were analyzed.

In sharp contrast, very few H1t-positive (i.e., mid-pachytene) spermatocytes were found in PD17, PD30, and PD40 Maps^−/− testes (Fig. 5B); they eventually reached 33.4% in PD50–80 (Fig. 5B). These observations indicate that MAPS plays a significant role in driving the transition from early pachynema to mid-pachynema for the first wave of meiosis.

Female Maps^−/− Mice Have Normal Meiosis Prophase I Progression and Fertility.

We also performed chromosome spreading with E17.5 ovaries and found that Maps^−/− female germ cells formed normal synaptonemal complexes relative to Maps^+/+ cells (SI Appendix, Fig. S6 A–F), and neither expressed MAPS protein (SI Appendix, Fig. S6 G–J). Also, crossovers in prophase I of Maps^−/− female germ cells were normal, with comparable MLH1 foci to Maps^+/+ pachytene cells (SI Appendix, Fig. S6 K and L, arrows). These results confirm that MAPS is not involved in female meiosis prophase I. Indeed, the fertility of Maps^−/− females was normal.

Maps^−/− Spermatocytes Fail to Establish MSCI at Both Pubertal and Adult Ages due to Transcription Dysregulation of Sex Chromosomes.

To investigate the underlying mechanisms for pachytene arrest in Maps^−/− spermatocytes, we isolated pachytene cells from Maps^+/+ and Maps^−/− testes of PD17 and PD90 mice and performed RNA-sequencing analyses.

In PD17, 89.0% (483 out of 543) of the X-linked protein-coding genes and 86.7% (13 out of 15) of the Y-linked protein-coding genes were differentially expressed between Maps^−/− and Maps^+/+ pachytene spermatocytes (Table 1 and SI Appendix, Fig. S7A). For the X chromosome, 89.6% of the differentially expressed genes (433 out of 483) were up-regulated in Maps^−/− spermatocytes, and for the Y chromosome, all of the genes (13 out of 13) were up-regulated in Maps^−/− spermatocytes. The median expression of genes from the X and Y chromosomes in Maps^−/− cells relative to Maps^+/+ cells was increased by a mean of 1.7-fold and 2.1-fold, respectively.

Table 1.

Differentially expressed (DE) genes in Maps^−/− vs. Maps^+/+ spermatocytes

	Chromosomes	DE genes/total genes		Total	Over 2 folds	Over 4 folds	Over 8 folds	Over 16 folds
PD17	X	483/543	Up	433	417	144	6	2
			Down	50	48	47	45	42
	Y	13/15	Up	13	13	7	2	/
			Down	/	/	/	/	/
	Autosome	8,349/12,262	Up	3,509	2,698	854	54	7
			Down	4,840	4,702	2,588	1,808	1,333
PD90	X	455/463	Up	441	441	428	363	125
			Down	14	13	12	10	7
	Y	6/6	Up	6	6	6	6	1
			Down	/	/	/	/	/
	Autosome	10,669/11,887	Up	5,434	4,273	2,958	1,525	350
			Down	5,235	3,929	2,331	1,239	431

The numbers of DE and total genes and up-regulated and down-regulated genes in isolated pachytene spermatocytes from PD17 and PD90 Maps^−/− mice relative to Maps^+/+ mice were listed. Only protein-coding genes having an average FPKM (fragments per kilobase million) > 1 in either Maps^+/+ or Maps^−/− group were included. The / indicates 0.

In PD90 pachytene spermatocytes, 98.3% (455 out of 463) of the X-linked protein-coding genes and 100% (6 out of 6) of the Y-linked protein-coding genes were differentially expressed between Maps^−/− and Maps^+/+ testes (Table 1 and SI Appendix, Fig. S7B). Of the differentially expressed X chromosome genes, 96.9% (441 out of 455) were up-regulated in Maps^−/− spermatocytes. One hundred percent (6 out of 6) of the differentially expressed Y chromosome genes were up-regulated in Maps^−/− spermatocytes. Notably, 389 up-regulated and 11 down-regulated genes overlapped between the PD17 and PD90 groups (SI Appendix, Fig. S7 C and D), implying a common mechanism that regulates the transcription of the X and Y chromosomes. These data indicate that MSCI establishment in Maps^−/− spermatocytes was significantly disrupted and that MAPS may contribute to the initiation and maintenance of MSCI on X and Y chromosomes.

Maps^−/− Pachytene Spermatocytes Exhibit Dysregulation in Autosomal Gene Transcription.

Substantial dysregulation of autosomal transcription was also observed in Maps^−/− pachytene cells relative to Maps^+/+ pachytene cells. In PD17, 68.0% (8,349 out of 12,262) autosomal genes were differentially regulated, among which, 42.0% (3,509 out of 8,349) genes were significantly up-regulated and 58.0% (4,840 out of 8,349) genes were significantly down-regulated in Maps^−/− cells (Table 1 and SI Appendix, Fig. S8A). Of the genes with expression levels changed over eightfold, 1,808 genes were down-regulated and only 54 genes were up-regulated, suggesting a large-scale suppression of autosomal gene expression. In PD90 Maps^−/− pachytene spermatocytes, 89.8% (10,669 out of 11,887) genes on autosomes were differentially regulated (Table 1 and SI Appendix, Fig. S8B). Among them, 50.9% (5,434 out of 10,669) was up-regulated and 49.1% (5,235 out of 10,669) was down-regulated in Maps^−/− cells. The analyses also revealed that 3,062 up-regulated and 4,359 down-regulated genes overlapped between the PD17 and PD90 groups (SI Appendix, Fig. S8 C and D), suggesting that MAPS regulates autosomal transcription through a similar mechanism in pubertal and adult pachytene spermatocytes.

Importantly, gene ontology analysis showed that, at both PD17 and PD90, genes of the male gamete generation pathway and spermatogenesis pathway were at the top of down-regulated populations (SI Appendix, Fig. S9 A and B).

MAPS Regulates Cellular Protein Ubiquitination In Vitro and In Vivo.

Protein ubiquitination is an important posttranslational modification marking proteins for degradation or functional alternation. Ubiquitin can be conjugated to substrate proteins either as monoubiquitination or polyubiquitination. Monoubiquitination is the addition of one ubiquitin molecule to substrates (27), altering its function. Polyubiquitination refers to the addition of a chain of ubiquitin moiety to a residue of substrate protein. The most prominent polyubiquitination through lysine 48 linkage marks the protein for degradation by proteasomes (28). Previous reports suggested that ubiquitin localizes on the axis along surface-spread chromosomes and ubiquitination may regulate synapsis, DSB repair, recombination, and crossover formation through stabilization and degradation of recombination factors (16). In this study, we used FK2 antibody, which recognizes both mono- and poly-ubiquitin conjugates (29), to detect the levels of ubiquitination of cellular proteins.

To determine whether MAPS levels affect cellular protein ubiquitination, we ectopically expressed MAPS-GFP (green fluorescent protein) or MAPS-HA (human influenza hemagglutinin) fusion proteins in transformed human embryonic kidney (HEK) 293T cells that were negative for MAPS. As in pachytene spermatocytes, MAPS-GFP is localized in both nuclei and cytoplasm (Fig. 6A). Immunoblots with FK2 antibody revealed that the levels of ubiquitination of cellular proteins were dramatically reduced in MAPS-expressing cells (Fig. 6 B and C and SI Appendix, Fig. S10A). Interestingly, the levels of ubiquitinated H2A, a prominent target of monoubiquitination modification, likely represented by the bands around 20 kD, were not obviously affected. This result is better discerned by fractionation of MAPS-expressing cells into soluble and pellet (chromatin-associated proteins) fractions. Immunoblots for FK2 antibody revealed that the levels of ubiquitinated proteins were dramatically reduced in chromatin fractions of MAPS-expressing cells. Immunoblots with HA antibody revealed that MAPS-HA is predominately localized in the soluble fractions with little in the chromatin fractions (Fig. 6D and SI Appendix, Fig. S10B).

Fig. 6.

MAPS regulates cellular protein ubiquitination in vitro. (A) Subcellular localization of MAPS. HEK 293T cells were transfected with a MAPS-GFP expression vector and observed by fluorescent microscopy. Nuclei were stained with DAPI. (Scale bars, 50 μm.) (B) Ectopic overexpression of MAPS reduces cellular ubiquitination levels. HEK 293T cells were transfected with a MAPS-HA expression vector or an empty control vector (CTL), and total cellular lysates were analyzed by immunoblots using antibodies FK2 (first panel) to detect ubiquitinated proteins, HA (second panel) to detect MAPS-HA overexpression, and GAPDH (third panel) to show equal loading. (C) Quantification of changes in protein ubiquitination of cells overexpressing MAPS-HA. Analyses were conducted on B loaded with 20 µg of total protein. (D) Effects of MAPS-HA overexpression on ubiquitinated proteins in chromatin and soluble fractions. Immunoblots for FK2 antibody revealed that the levels of ubiquitinated proteins were dramatically reduced in chromatin fractions of MAPS-expressing cells. Immunoblots with HA antibody revealed that MAPS-HA is predominately localized in the soluble fractions and little in the chromatin fractions. (E) MAPS is not associated with chromatin. A stable HeLa cell line, which expresses Flag-H2B, was transfected with an expression vector of MAPS-HA. The lysates were separated as soluble and chromatin fractions. Anti-Flag immunoprecipitation was used to purify the chromatin. Proteins in In (input), Ft (flow through), and E (eluate) were then immunoblotted with Flag (Top) or MAPS (Bottom) antibodies. MAPS-HA largely remained in the soluble fractions. (F) HeLa cells stably expressing HA-Cul4A were transfected with an MAPS-GFP expressing plasmid. Cul4-DDB-ROC1 complex was purified by anti-HA immunoprecipitation. Immunoblots revealed that MAPS (MAPS-GFP) does not physically associate with the ubiquitin ligase. Parallel gels for B, D, E, and F stained with Coomassie brilliant blue to demonstrate equal loading are shown in SI Appendix, Fig. S10.

To investigate the mechanism by which MAPS affects cellular protein ubiquitination, we tested whether MAPS associates with chromatin. We transfected HeLa cell line expressing Flag-H2B with the MAPS-HA expressing plasmid. Although chromatin was efficiently immunoprecipitated in the pellet fraction by anti–Flag-H2B immunoprecipitation, MAPS (MAPS-HA) was not detected in the immumoprecipitates (Fig. 6E and SI Appendix, Fig. S10C). We then transfected the MAPS-GFP expressing plasmid into HeLa cell line, which stably expresses HA-Cullin 4A (Cul4A), a component of a Cullin family ubiquitin ligase complex that has both cytoplasmic and nuclear substrates (30, 31). We did not find an association between MAPS-GFP with HA-Cul4A by coimmunoprecipitation (Fig. 6F and SI Appendix, Fig. S10D). These results indicate that MAPS does not stably bind to chromatin-associated substrates and did not interact directly with the ubiquitin ligase investigated.

The in vitro data led us to probe for ubiquitinated proteins in pachytene spermatocytes from Maps^+/+ and Maps^−/− mice. As in HEK 293T cells, MAPS expression was primarily observed in soluble fractions of pachytene spermatocytes, but not in chromatin-associated protein (pellet) fractions (Fig. 7A). Histone H3 was used as a marker for chromatin fractions. Immunoblots with FK2 antibody showed that the levels of ubiquitination of cellular proteins were dramatically increased in Maps^−/− cells relative to Maps^+/+ cells, except that a low-molecular-weight band at around 17 kD showed decreased levels (Fig. 7B). These data are in accordance with our in vitro studies (Fig. 6 B and D) and demonstrate that MAPS regulates cellular protein ubiquitination.

Fig. 7.

MAPS regulates cellular protein ubiquitination in pachytene spermatocytes. (A) Immunoblots of MAPS in soluble and pellet (chromatin-associated proteins) fractions of isolated pachytene spermatocytes from Maps^+/+ and Maps^−/− males at PD80, indicating MAPS was primarily present in soluble fractions and only in Maps^+/+ cells. (B) Immunoblot analysis of ubiquitinated proteins using antibody FK2 in isolated pachytene spermatocytes from PD80 Maps^+/+ and Maps^−/− mice. Cellular proteins in Maps^−/− pachytene spermatocytes have elevated levels of ubiquitination. β-actin was used as a loading control.

Discussion

During meiosis prophase I in mouse spermatocytes, the pachytene stage lasts for about 6 d (32). Prominent features of pachytene spermatocytes are established at mid-pachynema, including the crossover formation, the formation of the XY body, and MSCI (8–10). In this work, we have identified a male germline-specific protein MAPS, which is essential for the pachynema progression in meiosis prophase I in mice. Knockout of Maps causes a progressive cessation of progression from early to mid- and late pachynema in spermatocytes, leading to cell demise and male infertility.

In pubertal Maps^−/− pachytene spermatocytes, spermatocytes were arrested at early pachytene stage with defective DSB repair, lack of crossovers, and defective XY body formation. However, in adult Maps^−/− mice, there was seemingly normal DSB repair and normal formation of the XY body, but crossover formation was somewhat defective, and the progression of pachynema was delayed or arrested. Eventually, a fraction of spermatocytes reached the mid- or late pachytene stage. Regardless, spermatocyte demise was a common fate for both pubertal and adult Maps^−/− mice. Collectively, these results suggest that MAPS did not seem to be directly involved in the homologous recombination process. Rather, Maps deficiency retards and arrests pachytene progression, resulting in cell death.

It should be mentioned that we also generated a Maps knock-in mouse model in the Maps^−/− background. The mice overexpressed HA-tagged MAPS and restored spermatogenesis (see SI Appendix, Methods). These results rule out any off-target effects during the generation of the Maps^−/− mice.

It is worth noting that the observed wide range of defects in pachytene progression in Maps^−/− spermatocytes are distinguishable from the previously reported mutant mouse models with pachytene arrest caused by defects in certain single events. For example, deletion of Mlh1 caused crossover deficiency and resulted in pachytene arrest (33). H2ax^−/− and Mdc1^−/− (mediator of DNA damage checkpoint 1) spermatocytes showed mid-pachytene arrest caused by a failure in XY body formation (34, 35).

Our results of the global dysregulation of gene expression from autosomes and sex chromosomes in Maps^−/− spermatocytes were extraordinary. In particular, genes related to male gamete generation and spermatogenesis pathway were dramatically down-regulated. These striking transcriptional changes are likely the reasons for the observed phenotypes, including the defects in DSB repair, crossover, sex body formation, and MSCI in the pubertal and adult Maps^−/− spermatocytes, ultimately leading to male infertility. The scale and magnitude of changes are distinct from previous mutant mouse models with MSCI defects, such as mice deficient in H2ax (H2ax^−/−) (34), Mdc1 (Mdc1^−/−) (35) and mice with meiotic deletion of Setdb1 (SET domain, bifurcated 1) (36), where only expression of a small number of autosomal genes was altered.

The underlying mechanisms for the demise of Maps^−/− spermatocytes in adult and pubertal mice are similar, as indicated by the large number of overlapping perturbed gene transcription in pubertal and adult Maps^−/− mouse spermatocytes. The slightly different phenotypes observed in pubertal and adult Maps^−/− mice may be attributed to differences in the first wave of pubertal spermatogenesis and the adult steady-state spermatogenesis (37). The first wave of spermatocytes comes from NGN-3 (neurogenin-3)-negative differentiating spermatogonia derived from a particular subset of gonocytes in the prepubertal testes. In contrast, spermatocytes in adults are cells differentiated from NGN-3–positive, undifferentiated self-renewing spermatogonia (38). Also, there are known differences in recombination and crossover maturation between the pubertal waves of spermatogenesis and the adult waves (39). In this study, the pubertal spermatocytes seem more vulnerable to the absence of MAPS and were immediately arrested at early pachynema. In adult Maps^−/− spermatocytes, although some of these cells can progress to late pachytene stage, or even diplotene stage, most of them appear to be lost to apoptosis.

In our first effort to shed light on the mechanism that mediates MAPS function, we expressed MAPS ectopically in cultured cells. We observed that cellular ubiquitination was dramatically reduced. These in vitro observations suggest that MAPS may also affect cellular protein ubiquitination in pachytene spermatocytes. Indeed, this was the case. We found that in Maps^−/− spermatocytes, ubiquitinated proteins in chromatin fractions were significantly increased. We propose that the elevated ubiquitinated proteins likely have contributed to the dramatic dysregulation of gene expression from both sex chromosomes and autosomes, ultimately leading to male infertility. In summary, we propose that MAPS is of great importance for controlling diverse processes involved in the proper pachytene progression in spermatocytes. Since MAPS does not contain any predicted functional motif, it appears to affect protein ubiquitination through a novel mechanism. Further studies are needed to test the hypothesis that MAPS affects cellular proteolysis pathway in spermatocytes and to understand why MAPS plays such essential roles in mammalian meiosis.

Nonobstructive azoospermia (NOA) patients carrying loss-of-function mutations of meiosis-related genes have been reported to show spermatocyte development arrest. This condition has been linked to mutations in the meiosis genes, such as TEX11 (testis expressed 11) and SYCP3 (40, 41). It will be interesting to determine whether mutations in MAPS can also lead to NOA or similar conditions. Future studies to reveal whether MAPS has a role in the epigenetic controls of human spermatogenesis is also a subject of interest for the diagnosis and potential treatment of certain types of male infertility.

Methods

The Maps^−/− mouse model in a C57BL/6 genetic background was generated by Cyagen Biosciences by deleting an 1,841-bp genomic DNA fragment spanning exon 3 using the CRISPR/Cas9-mediated genome-editing system (SI Appendix, Fig. S2A). The founder mice were genotyped by PCR and DNA sequencing analysis of DNA extracted from mouse tails. One selected F0 founder mouse was bred to wild-type mice to obtain F1 heterozygous mutant (Maps^+/−) mice, and the Maps^+/− mice were used to backcross to C57/B6 wild-type mice to get rid of possible off-target mutations. Breeding pairs of Maps^+/− male and female mice were used for obtaining wild-type (Maps^+/+) and homozygous-deficient (Maps^−/−) mice. Additional details and the characterization of the Maps^−/− mice and the Maps knock-in mice, as well as all other materials and methods, are described in detail in SI Appendix, Methods.

Data Availability

All study data are included in the article and/or SI Appendix.