Caseinolytic mitochondrial matrix peptidase X is essential for homologous chromosome synapsis and recombination during meiosis of male mouse germ cells

Abstract

Meiosis is the process of producing haploid gametes through a series of complex chromosomal events and the coordinated action of various proteins. The mitochondrial protease complex (ClpXP), which consists of caseinolytic mitochondrial matrix peptidase X (ClpX) and caseinolytic protease P (ClpP) and mediates the degradation of misfolded, damaged, and oxidized proteins, is essential for maintaining mitochondrial homeostasis. ClpXP has been implicated in meiosis regulation, but its precise role is currently unknown. In this study, we engineered an inducible male germ cell-specific knockout caseinolytic mitochondrial matrix peptidase X (Clpx^cKO) mouse model to investigate the function of ClpX in meiosis. We found that disrupting Clpx in male mice induced germ cell apoptosis and led to an absence of sperm in the epididymis. Specifically, it caused asynapsis of homologous chromosomes and impaired meiotic recombination, resulting in meiotic arrest in the zygotene-to-pachytene transition phase. The loss of ClpX compromised the double-strand break (DSB) repair machinery by markedly reducing the recruitment of DNA repair protein RAD51 homolog 1 (RAD51) to DSB sites. This dysfunction may be due to an insufficient supply of energy from the aberrant mitochondria in Clpx^cKO spermatocytes, as discerned by electron microscopy. Furthermore, ubiquitination signals on chromosomes and the expression of oxidative phosphorylation subunits were both significantly attenuated in Clpx^cKO spermatocytes. Taken together, we propose that ClpX is essential for maintaining mitochondrial protein homeostasis and ensuring homologous chromosome pairing, synapsis, and recombination in spermatocytes during meiotic prophase I.

INTRODUCTION

Caseinolytic mitochondrial matrix peptidase X (ClpX) is a component of the mitochondrial protease complex (ClpXP), which localizes to the mitochondrial matrix and is highly conserved among bacteria and eukaryotes. It primarily eliminates incorrect or abnormal proteins produced due to oxidative damage and stress in an ATP-dependent manner, and activates the mitochondrial unfolded protein response to maintaining mitochondrial protein homeostasis.1,2,3 The human caseinolytic protease P (ClpP) proteasome, a core component of the ClpXP complex, is associated with oxidative damage and protein degradation. Its dysfunction is linked to Perrault syndrome, cancer, primary acute myeloid leukemia, and other diseases.1,3 A deficiency in Clpp caused an overabundance of misfolded/unfolded proteins in the mitochondria, which decreased oxidative phosphorylation (OXPHOS) activity, increased mitochondrial oxidative damage, and resulted in proteotoxicity and cell death.2,4,5 ClpX recognizes misfolded or damaged proteins, binds and extends substrates, and transfers them to the ClpP core for hydrolysis. Although male Clpp-deficient mice were reported to be sterile,5 the role of ClpX in mouse spermatogenesis is unknown, especially how it affects homologous chromosome pairing, synapsis, and recombination during meiosis.

Meiotic prophase I is a prolonged cell cycle G2 stage specific to meiosis, involving the pairing, synapsis, recombination, and segregation of homologous chromosomes. Initially, meiotic recombination protein SPO11 (SPO11) mediates the formation of double-strand breaks (DSBs) in the meiotic DNA, the repair of which is immediately effected by the MRE11-RAD50-NBS1 (MRN) complex, C-terminal-binding protein interacting protein (CtIP), and the 5´–3´-human exonuclease 1.6,7 The DNA ends are resected to form an unstable 3´ single-stranded DNA (ssDNA) overhang, which is quickly coated and protected by the replication protein A (RPA) complex.8 The BRCA2–MEILB2 (HSF2BP)–BRME1 ternary complex, consisting of breast cancer susceptibility gene 2 (BRCA2), meiotic localizer of BRCA2 (MEILB2/HSF2BP), and BRCA2 and MEILB2-associating protein 1 (BRME1), was shown to recruit two homologous recombinases, DNA meiotic recombinase 1 (DMC1) and DNA repair protein RAD51 homolog 1 (RAD51). This complex forms the RAD51/DMC1–ssDNA nucleoprotein filament with ssDNA, which mediates the search for the homologous strand, invasion of the 3´-single strand into the homologous duplex, and pairing with one of its strands, resulting in the formation of a D-loop.9,10,11,12,13 When the invaded 3´-ssDNA serves as a primer for DNA synthesis through DNA polymerase, RAD51/DMC1 dissociates from the double-stranded DNA, completing the process of DNA homologous recombination. In humans and rodents, an increasing number of genes have been implicated in DSB formation, synapsis, and homologous recombination during meiosis I. Errors in any of these processes can result in meiotic arrest and sterility.14

Meiotic silencing of unsynapsed chromatin is critical for meiotic chromosomal pairing, synapsis, and formation of the XY body.15,16,17,18 Homolog asynapsis could trigger an ataxia telangiectasia and Rad3-related protein-mediated DNA damage response (DDR) pathway centered on phosphorylated H2A histone family member X Ser139 (γH2AX) that recognizes autosomal asynapsis and eliminates germ cells with chromosomal abnormalities.16,17,18 Meiotic silencing is also responsible for inactivating XY chromosomes, referred to as meiotic sex chromosome inactivation (MSCI), during the pachytene of male meiotic prophase I. The ubiquitin–proteasome system controls protein homeostasis, ensures the occurrence of MSCI,19 and regulates meiotic recombination factors as well as SUMOylation modifications.19,20,21,22,23 For instance, the germ cell-specific Scm polycomb group protein-like 2 forms a complex with ubiquitin-specific protease 7 to remove monoubiquitination of histone H2A at lysine119 (H2AK119ub) from the XY chromosome to facilitate its silencing.20,24 Deletion of the gene encoding H2B ubiquitin E3 ligase ring finger protein 20 leads to persistently tight chromosomal structures during the initiation of DSB repair, which prevents the proper recruitment of DNA damage factors, affects the synapsis and recombination repair of homologous chromosomes, and eventually causes infertility in male mice.25 Although ubiquitination and proteasomal degradation have been reported to regulate meiosis, the mitochondrial proteasome’s influence on meiosis is yet unknown.

To understand the role of ClpX in the first meiotic division, particularly in homolog synapsis and recombination, we used inducible DEAD-Box Helicase 4 Cre (Ddx4-iCre) transgenic mice to inactivate Clpx in the spermatocytes at specific stages. Clpx-deficient (Clpx^cKO) mice exhibited apoptosis of germ cells in seminiferous tubules, sperm loss in the epididymis, altered development of the XY body, and increased percentage of spermatocytes with homologous asynapsis. The early-pachytene spermatocytes were suppressed to initiate meiotic silencing of unsynapsed chromatin. Asynapsed chromosome axes failed to load the ATP-dependent RAD51 recombinase in the Clpx^cKO zygotene and pachytene-like spermatocytes, resulting in homologous chromosomal synapsis, recombination failure, and apoptosis. Furthermore, XY chromosomes could not be condensed to form the XY body, and the ubiquitination of XY chromosomes and autosomes was significantly attenuated. These results demonstrate that ClpX is indispensable for homologous pairing, synapsis, and recombination during meiotic prophase I because it maintains mitochondrial protein homeostasis.

MATERIALS AND METHODS

Animals

C57BL/6 mice were housed in a pathogen-free environment with a controlled temperature in the range of 20°C–22°C, a 12/12-h light/dark cycle, 50%–70% relative humidity, and enough food and water. After being euthanized, they were dissected immediately to collect their testes and epididymis. The ages of the male mice analyzed were between 1 week and 8 weeks. All animal protocols were approved by the Ethics Committee of The University of Hong Kong-Shenzhen Hospital (Shenzhen, China; Approval No. hkuszh2020031).

Targeted inactivation of the Clpx gene

The Clpx-floxed mouse model (stock No. S-CKO-09886) was generated by Cyagen Biosciences (Suzhou, China). The loxP sites flanking exon 3 were constructed using the clustered regularly interspaced short palindromic repeats (CRISPR)/CRISPR-associated protein 9 (Cas9) system. Clpx^fl/+ females were crossed with Ddx4-inducible Cre/estrogen receptor-binding domain fusion protein 2 (Ddx4-Cre^ERT2) males (024760, Jackson Laboratory, Bar Harbor, ME, USA) to generate Clpx^fl/+ Ddx4-Cre^ERT2 mice, which were then bred with Clpx^fl/fl mice to produce Clpx^fl/fl Ddx4-Cre^ERT2 mice for experiments. Mice were genotyped by polymerase chain reaction (PCR) analysis of genomic DNA extracted from their tails. The wild-type allele (119 bp) and Clpx floxed allele (184 bp) were analyzed by PCR using the following primers: forward 5´-GTAAGTCTCCTGATCCAACCT-3´ and reverse 5´-GTCCAGCTAAGGATTCTCATT-3´. The Ddx4-Cre^ERT2 allele (approximately 240 bp) was assayed by PCR using the following primers: forward 5´-CACGTGCAGCCGTTTAAGCCGCGT-3´ and reverse 5´-TTCCCATTCTAAACAACACCCTGAA-3´.

Tamoxifen (T5648, Sigma-Aldrich, Burlington, MA, USA) was dissolved in corn oil and heated at 50°C for 3–4 h to a final concentration of 20 mg ml⁻¹ in a water bath. To induce germ cell-specific Clpx deletion, tamoxifen was intraperitoneally injected into Clpx^fl/fl Ddx4-Cre^ERT2 male pups on postnatal day (PD) 12 at a dose of 80 mg kg⁻¹ body weight for 4 consecutive days. Littermates of Clpx^fl/fl or Clpx^fl/+ genotype were simultaneously treated and used as controls. According to the half-life of ClpX upon gene deletion in spermatocytes, testis and/or epididymis samples were collected on PD20 (5 days posttreatment [dpt]), PD25 (10 dpt), PD30 (15 dpt), PD35 (20 dpt), and PD56 (41 dpt).

Histological and nuclear surface spreading analyses

Fresh samples of testicles and epididymis were fixed in Bouin’s solution (HT10132-1L, Sigma-Aldrich) for 16 h at room temperature (RT) or in 4% paraformaldehyde (P1110, Solarbio, Beijing, China) for 20 h at 4°C and embedded into paraffin blocks. Slices, 5-μm-thick, were stained with hematoxylin and mounted with resins for microscopic examination. Spermatocyte chromosome spreads were prepared as described by Peters et al.26 Briefly, the seminiferous tubules were separated with a pair of fine forceps, subjected for 25 min to a hypotonic treatment buffer (30 mmol l⁻¹ Tris-HCl, 17 mmol l⁻¹ trisodium citrate dihydrate, 5 mmol l⁻¹ ethylenediaminetetraacetic acid [EDTA], and 50 mmol l⁻¹ sucrose [pH 8.2–8.4]), and immediately transferred to 100 mmol l⁻¹ sucrose buffer for 5 min. The cell suspension was obtained by gently squeezing the tubules using forceps, added to a slide containing 50–80 μl of fixative buffer (1% paraformaldehyde, 1 mol l⁻¹ NaOH, and 0.15% Triton X-100 [pH 9.2]), incubated overnight in a humidified box at 4°C, air-dried for 2–3 h, and finally stored at −80°C until staining.

Transmission electron microscopy

The testis samples from PD56 control and Clpx^cKO mice were cut into 1 mm3 pieces, placed immediately in 2.5% glutaraldehyde solution (G1102, Servicebio, Wuhan, China) at RT in the dark for 2 h, washed thrice with 0.1 mol l⁻¹ phosphate buffer (pH 7.4), fixed in 1% osmic acid prepared in 0.1 mol l⁻¹ phosphate buffer (pH 7.4) at RT in the dark for 2 h, rinsed thrice with 0.1 mol l⁻¹ phosphate buffer (pH 7.4), dehydrated using a graded alcohol solution at RT, embedded in resin, cut into 60–80-nm-wide ultra-thin slices using an ultra-microtome, placed on 150-mesh formvar-coated cuprum grids, stained in 2% uranium acetate-saturated alcohol solution in the dark for 8 min, washed with 70% alcohol and ultra-pure water, stained with 2.6% lead citrate for 8 min, and air-dried overnight. Images were acquired and analyzed using a transmission electron microscope (HT7800, Hitachi, Hitachinaka, Japan).

Immunofluorescence staining

Paraffin-embedded sections were dewaxed, rehydrated, subjected to antigen retrieval by heating in Tris-EDTA buffer (pH 9.0; ZLI-9069, ZSGB-BIO, Beijing, China) for immunostaining, blocked in a 5% bovine serum albumin solution for 60 min, incubated with primary antibodies overnight at 4°C, washed thrice with phosphate-buffered saline (PBS), and incubated with the appropriate secondary antibodies at RT for 1 h. For immunostaining of chromosome spreads, the slides were washed with PBS containing 0.15% Triton X-100 for 5 min, blocked for 30 min with 5% bovine serum albumin, incubated with primary antibodies overnight at 4°C in a humidified chamber, washed thrice for 15 min with PBS containing 0.15% Triton X-100, probed with the corresponding secondary antibodies at RT for 1 h, and mounted using VECTASHIELD medium containing 4´,6-diamidino-2-phenylindole (DAPI; H-1200, Vector Laboratories, Burlingame, CA, USA). For immunofluorescence analyses, the following antibodies were used: rabbit anti-cleaved poly(ADP-ribose) polymerase (c-PARP) Asp214 (1:200, #94885, Cell Signaling Technology, Boston, MA, USA), mouse anti-synaptonemal complex protein 3 (SYCP3; 1:200, ab97672, Abcam, Cambridge, UK), rabbit anti-Lamin B1 (1:200, 12987-1-AP, Proteintech, Wuhan, China), rabbit anti-centromere antibody (ACA; 1:200, homemade), rabbit anti-synaptonemal complex protein 1 (SYCP1; 1:200, ab15090, Abcam), rabbit anti-ClpP (1:200, #15698-1-AP, Proteintech), rat anti-testis-specific histone variant (H1t; 1:500, homemade), rabbit anti-RPA2 (1:200, ab76420, Abcam), mouse anti-γH2AX (1:400, #05636, Millipore, Burlington, MA, USA), mouse anti-telomere repeat-binding factor 1 (TRF1; 1:200, ab10579, Abcam), mouse anti-ubiquity-histone H2A clone E6C5 (1:500, #05678, Millipore), mouse anti-ubiquitinylated proteins clone FK2 (1:500, #04263, Millipore), rabbit anti-ubiquitin antibody Lys48-specific clone Apu2-07 (1:100, #051307, Millipore), rabbit anti-RAD51 (1:200, PA5-27195, Invitrogen, Carlsbad, MA, USA), rabbit anti-HORMA domain-containing protein 1 (HORMAD1; 1:400, 13917-1-AP, Proteintech), rabbit anti-ClpX (1:200, ab168338, Abcam), mouse anti-translocase of the mitochondrial outer membrane 70 (TOM70; 1:200, 66593-1-IG, Proteintech), mouse anti-DEAD box polypeptide 4/mouse vasa homolog (DDX4/MVH; 1:200, ab27591, Abcam), Alexa Fluor® 488- and 594-Conjugate (1:200, ab150117 and ab150120, respectively, Abcam), Alexa Fluor® 488- and 594-Conjugate (1:200, ab150119 and ab150081, respectively, Abcam), and Alexa Fluor® 488- and 594-Conjugate (1:200, ab150084 and ab150083, respectively, Abcam). Immunolabeled chromosome spreads and immunostained testis sections were imaged using an LSM 900 microscope (Zeiss, Oberkochen, Germany).

Western blotting

Mouse testes were homogenized on ice and lysed in lysis buffer (50 mmol l⁻¹ Tris-HCl pH 8.0, 120 mmol l⁻¹ NaCl, 1 mmol l⁻¹ EDTA, 6 mmol l⁻¹ ethylene glycol-bis[β-aminoethyl ether]-N,N,N´,N´-tetraacetic acid [EGTA], 1% Nonidet P-40, 1 mmol l⁻¹ dithiothreitol, 10 mmol l⁻¹ NaF, 0.25 mmol l⁻¹ Na₃VO₄, and 50 mmol l⁻¹ β-glycerophosphate) supplemented with complete protease inhibitors (4693159001, Roche, Basel, Switzerland). The lysate was centrifuged at 20 000g (5804R, Eppendorf, Hamburg, Germany) for 30 min at 4°C, and the supernatant was used for western blotting. The following primary antibodies were used: rabbit anti-ClpX (1:3000, ab168338, Abcam), mouse anti-total OXPHOS (1:1000, ab110413, Abcam), and rabbit anti-β-actin (1:5000, #ab227387, Abcam). The secondary antibodies used were goat anti-mouse IgG-horseradish peroxidase (HRP)-conjugated (1:5000, FDM007, Fdbio Science, Hangzhou, China) and goat anti-rabbit IgG-HRP-conjugated (1:5000, FDR007, Fdbio Science). Relative protein levels, obtained after normalization to β-actin, were used to standardize the loading variations.

Statistical analyses

All statistical analyses and graphs were realized using GraphPad Prism 9 (version 9.0.0; GraphPad Software, Inc., San Diego, CA, USA). The means and standard deviation for all data were displayed. An unpaired two-tailed Student’s t-test was used to compare the outcomes of two experimental groups. All experiments were repeated at least thrice, and the results of one representative experiment are shown. The data were considered significant when P < 0.05.

RESULTS

Depletion of Clpx in spermatocytes led to impaired spermatogenesis

We first examined the expression pattern of ClpX during spermatogenesis. Immunofluorescence results showed that ClpX was expressed in most spermatogenic cells, including spermatogonia, spermatocytes, and postmeiotic round spermatids. Notably, ClpX was significantly upregulated in zygotene to meiosis I spermatocytes (Supplementary Figure 1 ^{(294.7KB, tif)} ).

To study the role of ClpX in meiosis, we generated a Clpx-floxed mouse, in which exon 3 was flanked by loxP sites (Supplementary Figure 2a ^{(261.2KB, tif)} ), and crossed it with a germ cell-specific, tamoxifen-inducible Ddx4-Cre^ERT2 strain to engineer Clpx^fl/fl Ddx4-Cre^ERT2 mice (Clpx^cKO mice) for experiments (Supplementary Figure 2b ^{(261.2KB, tif)} ). Without tamoxifen treatment, Clpx^cKO mice were fertile. Originally, we aimed to target the first wave of spermatogenic cells. We selected PD12 for tamoxifen treatment because germ cells in most testicular tubules were entering meiosis at this time (Supplementary Figure 2c ^{(261.2KB, tif)} ). Intraperitoneal injection of tamoxifen caused no apparent change in testis morphology at 5–15 dpt, as revealed by hematoxylin staining of histological sections, suggesting that ClpX had a long half-life (Supplementary Figure 2d ^{(261.2KB, tif)} ). Strikingly, at 20 dpt, the testes of Clpx^cKO mice exhibited severe loss of germ cells, indicating the successful knockout of Clpx as well as a potential role for it in spermatogenesis (Supplementary Figure 2d ^{(261.2KB, tif)} ).

Given the long half-life of ClpX, we examined the adult testes. Clpx^cKO male mice developed normally without any apparent abnormalities, but the testes of PD56 Clpx^cKO males were considerably smaller in size and less in weight than those of control littermates (Figure 1a and 1b). Immunofluorescence staining of testis sections confirmed the elimination of ClpX in Clpx^cKO spermatocytes (Figure 1c). In control mice, ClpX colocalized with translocase of the mitochondrial outer membrane 70, a mitochondrial protein marker (Supplementary Figure 2e ^{(261.2KB, tif)} ). Consecutive tissue sections showed that the expression of ClpP was similar in Clpx^cKO and control spermatocytes, implying that its function was unperturbed (Supplementary Figure 3 ^{(263.1KB, tif)} ). Histological analyses showed that the testes of adult Clpx^cKO mice suffered from severe loss of germ cells in most seminiferous tubules and elongating spermatids were scarcely observed (Figure 1d). No sperm was detected in the epididymis of Clpx^cKO mice (Figure 1d). These findings indicate that ClpX is essential for germ cell development in male mice.

Figure 1.

Depletion of Clpx in spermatocytes results in impaired spermatogenesis with defects in meiotic progression. (a) Significant size reduction in Clpx^cKO testes at 8 weeks. (b) Testis weight/body weight ratio quantification in control and Clpx^cKO mice. ^***P < 0.001, n = 3. The data are shown as mean ± s.d. Scale bar=2 mm. (c) Immunofluorescence staining demonstrating the tamoxifen-induced successful elimination of ClpX protein in PD56 Clpx^cKO spermatocytes. The right panels are enlarged figures for the square regions on the left, and the dashed lines separate the different types of spermatocytes. The various stages of seminiferous tubules are depicted. PreL: preleptotene; Z: zygotene; P: pachytene; LP: late pachytene; D: diplotene spermatocytes. Scale bars = 20 μm. (d) Hematoxylin-stained sections of testes and epididymises from 8-week-old control and Clpx^cKO males. Asterisk indicates the loss of germ cells. Z-like stands for zygotene-like spermatocytes, and D stands for diplotene spermatocytes. Scale bars = 50 μm. ClpX: caseinolytic mitochondrial matrix peptidase X; s.d.: standard deviation; PD: postnatal day; DAPI: 4’,6-diamidino-2-phenylindole; SYCP3: synaptonemal complex protein 3.

Loss of ClpX disrupted synapsis and the progression of meiotic prophase I

To determine if ClpX knockout affected the progression of meiotic prophase I, we compared the composition of germ cells within the seminiferous tubules of control and Clpx^cKO mice. γH2AX serves as a marker for DSBs distributed throughout the nucleus during the leptotene/zygotene stages and in the sex body (XY chromosome region) during the pachytene/diplotene stages.27 Coimmunostaining of SYCP3, the axial/lateral elements of the synaptonemal complex (SC), and γH2AX showed that testicular tubules from Clpx^cKO mice contained a significantly lower number of leptotene/zygotene and pachytene/diplotene spermatocytes compared with tubules from control mice (P < 0.01; Figure 2a and 2b). Interestingly, in tubules from Clpx^cKO mice that had no spermatids, we observed several pachytene-like spermatocytes wherein the γH2AX signal was partially aggregated on the sex body but still retained on the autosomes, indicating that sex body formation might be impaired. Furthermore, most cells lacked H1t, a histone variant that is incorporated into chromatin during the mid-late pachytene stage (Figure 2c). Coimmunostaining with anti-SYCP3 and anti-H1t revealed that H1t was absent in the pachytene-like spermatocytes from Clpx^cKO mice (Figure 2d), implying that they might be arrested during the zygotene-to-pachytene transition.

Figure 2.

ClpX is essential for spermatogenesis in male mice. (a) Immunofluorescence staining of testis sections from PD56 control and Clpx^cKO males with SYCP3 in red, γH2AX in green, and DAPI in blue. Asterisk indicates loss of germ cells. Scale bars = 50 μm. (b) The number of spermatocytes with γH2AX signals per tubule. n: the pooled number of seminiferous tubules analyzed from three animals for each genotype. ^**P < 0.01, ^****P < 0.0001. (c) Testis sections from PD56 control and Clpx^cKO mice were stained for H1t (green) and DAPI (blue). (d) Immunofluorescent labeling of testicular sections derived from PD56 control and Clpx^cKO males. SYCP3 (red), H1t (green), DAPI (blue), and merged images are shown. Scale bars = 50 μm. ClpX: caseinolytic mitochondrial matrix peptidase X; PD: postnatal day; Z: zygotene; P: pachytene; D: diplotene; PreL: preleptotene; EP: early pachytene; P-like: pachytene-like spermatocytes; RS: round spermatid; H1t: testis-specific histone variant; SYCP3: synaptonemal complex protein 3; DAPI: 4’,6-diamidino-2-phenylindole; γH2AX: phosphorylated H2A histone family member X Ser139; Lep/Zyg: leptotene/zygotene spermatocytes; Pac/Dip: pachytene/diplotene spermatocytes.

To examine SC formation in meiosis, we immunostained chromosome spreads using anti-SYCP3, anti-SYCP1 (SC transverse filament), and anti-centromere antibodies. Some homologous chromosomes in Clpx^cKO spermatocytes failed to pair or synapse completely, exhibiting a pachytene-like pattern, whereas autosomes fully synapsed and the sex body was formed in control pachytene spermatocytes (Figure 3a). The asynapsis in Clpx^cKO spermatocytes may partly be due to misassociation between nonhomologous chromosomes (Figure 3b). Synapsis occurred between nonhomologous chromosomes at either the centromeric or the noncentromeric ends (Figure 3c).

Figure 3.

ClpX is required for synapsis of homologous chromosomes in meiosis. (a) Immunostaining of spermatocyte spreads from PD35 control and Clpx^cKO testes for SYCP1 (green), SYCP3 (red), and ACA (blue). (b) The white dashed box in the middle panel indicates the asynapsed chromosomes. (c) Higher magnification of the white dashed box in B. The right panel illustrates asynapsed chromosomes. A–A’ represents the pairing of homologous chromosomes partially synapsed at the noncentromeric end. A’–B represents the pairing of nonhomologous chromosomes at the centromeric end. (d) Immunostaining of spermatocyte spreads from PD35 control and Clpx^cKO testes for HORMAD1 (green) and SYCP3 (red). Scale bars = 5 μm. ClpX: caseinolytic mitochondrial matrix peptidase X; PD: postnatal day; SYCP1: synaptonemal complex protein 1; SYCP3: synaptonemal complex protein 3; ACA: anti-centromere antibodies; HORMAD1: HORMA domain-containing protein 1.

At the beginning of meiosis, HORMAD1 is recruited to the unsynapsed chromosome axes to ensure efficient DSB formation. SC formation, in turn, depletes HORMAD1 from the axes.28 Therefore, HORMAD1 also serves as a marker for the completion of synapsis. SYCP3-HORMAD1 coimmunostaining of nuclear surface spreads demonstrated that the HORMAD1 signal was retained on the unsynapsed axes, confirming the existence of chromosome asynapsis in Clpx^cKO pachytene-like spermatocytes (Figure 3d). HORMAD1 was restricted to the unsynapsed axes of X and Y chromosomes in control pachytene spermatocytes. Taken together, these observations suggest that the mitochondrial ClpX plays an essential role in meiotic prophase I by regulating the synapsis of homologous chromosomes.

Meiotic recombination was compromised in Clpx^cKO spermatocytes

Since recombination is a hallmark event in meiosis, we examined its progression in Clpx^cKO spermatocytes using recombination markers. In control spermatocytes, the γH2AX signal was distributed throughout the nucleus during leptonema and zygonema but restricted to the XY body during pachynema and diplonema. In contrast, the γH2AX signal decreased but persisted on most chromosomes in Clpx^cKO pachytene-like spermatocytes that lacked the XY body (Figure 4a).

Figure 4.

Meiotic recombination is impaired in Clpx^cKO spermatocytes. (a) Immunostaining of spermatocyte surface spreads from PD35 control and Clpx^cKO males for SYCP3 (red) and γH2AX (green). (b) Immunostaining for SYCP3 (red) and RPA2 (green). (c) Comparison of RPA2 foci numbers between control and Clpx^cKO spermatocytes at the zygotene and pachytene/pachytene-like stages, respectively. (d) Immunostaining for SYCP3 (red) and RAD51 (green). (e) Comparison of RAD51 foci numbers between control and Clpx^cKO spermatocytes at the zygotene (Zyg) and pachytene/pachytene-like (Pac/Pac-like) stages, respectively. n: the pooled number of seminiferous tubules analyzed from three animals for each genotype. Scale bars = 5 μm. ^***P < 0.001, ^****P < 0.0001. ClpX: caseinolytic mitochondrial matrix peptidase X; PD: postnatal day; SYCP3: synaptonemal complex protein 3; NS: no significant difference; γH2AX: phosphorylated H2A histone family member X Ser139; RAD51: DNA repair protein RAD51 homolog 1; RPA2: replication protein A2.

RPA is a heterotrimer comprising RPA1/2/3 that binds to and prevents the degradation of ssDNA generated after DSBs. In control spermatocytes, RPA2 localized as foci on DSB sites during zygonema and early pachynema (Figure 4b). The number of RPA2 foci remained high in Clpx^cKO pachytene-like spermatocytes (Figure 4b and 4c and Supplementary Figure 4 ^{(149.8KB, tif)} ), which suggests that γH2AX persisted on most chromosomes in these spermatocytes because of unrepaired DSBs. Indeed, the number of RAD51 foci was significantly reduced in Clpx^cKO zygotene and pachytene-like spermatocytes when compared with same-stage controls (Figure 4d and 4e). These findings indicate that the DSB repair machinery in meiosis may be impaired in Clpx^cKO spermatocytes.

The attachment of telomeres to the inner nuclear membrane facilitates homologous pairing and DSB repair, while the lack of telomere-binding proteins leads to failed homologous pairing, synapsis, and recombination.29,30 To assess whether telomere-binding proteins were involved in the impaired synapsis and recombination observed in Clpx^cKO spermatocytes, we immunostained testis sections to detect the shelterin protein telomere repeat-binding factor 1, SYCP3, and the nuclear envelope component Lamin B1 (Supplementary Figure 5 ^{(82.3KB, tif)} ). The percentages of control pachytene and Clpx^cKO pachytene-like spermatocytes displaying defective telomere anchoring to the nuclear envelope were comparable, implying that the telomere complex functioned normally during synapsis in Clpx^cKO spermatocytes. Together, these results show that ClpX plays a vital role in meiotic recombination in male mice.

XY chromosomes and autosomes were aberrantly ubiquitinated in Clpx^cKO pachytene-like spermatocytes

Ubiquitinated proteins regulate the stabilization and degradation of recombinant proteins to ensure correct meiotic synapsis and recombination.21,23,31 To investigate whether DSB repair dysfunction in Clpx^cKO spermatocytes was associated with the aberrant ubiquitination of sex chromosomes and autosomes, we measured the levels of histone ubiquitination using the antibody clone E6C5 (referred to as Ub-E6C5), which recognizes monoubiquitination of H2A at K119, and FK2 (referred to as Ub-FK2), which recognizes mono- and K29-, K48-, and K63-linked polyubiquitination. In control spermatocytes, Ub-E6C5 and Ub-FK2 were concentrated on pachytene XY bodies, whereas in Clpx^cKO pachytene-like spermatocytes, they were incorrectly loaded onto the sex chromosomes, with abnormal signaling patterns on asynaptic chromosomes (Supplementary Figure 6a ^{(94.6KB, tif)} and 6b ^{(94.6KB, tif)} ). In control pachytene spermatocytes, K48 (Apu2.07)-linked polyubiquitin chains (referred to as Ub-K48) were expressed on autosomes, where they target the proteolytic degradation of K48 ubiquitin chains.6,13 Comparatively, the K48 signals on autosomal chains were significantly attenuated in Clpx^cKO pachytene-like spermatocytes (Supplementary Figure 6c ^{(94.6KB, tif)} ). These findings suggest that the deletion of Clpx markedly reduces the ubiquitination of XY bodies and induces aberrant ubiquitination of autosomes in male germ cells, which may be the underlying reason for impaired XY body formation in Clpx^cKO pachytene-like spermatocytes.

ClpX maintained the number and structure of mitochondria in spermatocytes

We compared the number and distribution of mitochondria in control and Clpx^cKO spermatocytes using electron microscopy. In control spermatocytes that had a clear sex body (Figure 5a), mitochondria were similar both in shape and size and evenly distributed throughout the cytoplasm. In contrast, the sex body was absent in the nuclei of Clpx^cKO spermatocytes, and the number of mitochondria in the cytoplasm was significantly lower (P < 0.001; Figure 5b). The loss of Clpp has been shown to damage OXPHOS proteins.5 Therefore, we used western blotting to detect the expression of all OXPHOS subunits in Clpx^cKO testes, namely complex I subunit NADH:ubiquinone oxidoreductase subunit B8 (NDUFB8), complex II subunit succinate dehydrogenase B (SDHB), complex III subunit ubiquinone reductase C2 (UQRC2), complex IV subunit mitochondrial cytochrome c oxidase I (MTCOI), and complex V subunit adenosine triphosphate 5A (ATP5A). Compared with the control, the expression of OXPHOS subunits, especially complexes I and IV, was significantly reduced in Clpx^cKO testes (Figure 5c). Importantly, the mitochondria in Clpx^cKO cells varied in size and were abnormally aggregated. The marked difference in the mitochondrial morphology within control and Clpx^cKO spermatocytes indicates that ClpX was required for mitochondrial homeostasis and meiotic progression.

Figure 5.

ClpX is responsible for maintaining mitochondrial function in spermatocytes. (a) The structures of mitochondria in adult control (A and A’) and Clpx^cKO (B and B’) spermatocytes. Electron microscopic images of mitochondria (arrowheads) in spermatocytes of PD56 control and Clpx^cKO testes. Enlarged views of the marked mitochondria are shown in green dashed squares. The red dotted line marks the distribution of mitochondria in Clpx^cKO spermatocytes. The yellow arrow indicates the XY body. Red arrowheads indicate mitochondria. Scale bars = 1 μm. (b) Quantification of mitochondria numbers in spermatocytes. n: the number of spermatocytes analyzed. ^***P < 0.001 by two-tailed Student’s t-test. (c) Western blots showing the levels of all respiratory complex subunits. ClpX: caseinolytic mitochondrial matrix peptidase X; PD: postnatal day; C: complex; ATP5A: adenosine triphosphate 5A; UQRC2: ubiquinone reductase C2; MTCOI: mitochondrial cytochrome c oxidase I; SDHB: succinate dehydrogenase B; NDUFB8: NADH:ubiquinone oxidoreductase subunit B8.

ClpX was required for the survival of pachytene and diplotene spermatocytes

We hypothesized that the defective spermatocytes in Clpx^cKO testes would be eliminated via apoptosis. Expectedly, the seminiferous tubules of Clpx^cKO mice had a significantly higher number of c-PARP-positive cells than the tubules of control mice (P < 0.01; Figure 6a-6c). To further characterize the spermatocytes destined to die, we coimmunostained testis sections with anti-SYCP3 and anti-c-PARP (Figure 6d). In Clpx^cKO testes, pachytene/pachytene-like and diplotene spermatocytes were c-PARP-positive in stage VIII and XI seminiferous tubules, whereas those in control testes appeared normal. Clpx^cKO testes had a significantly higher number of c-PARP-positive pachytene/pachytene-like spermatocytes than control testes (P < 0.0001; Figure 6e), suggesting that the loss of Clpx induced the death of spermatocytes in their pachytene stage. Overall, we conclude that the localization of ClpX to the mitochondria of mouse spermatocytes is essential for homologous chromosome synapsis and recombination during meiotic prophase I (Figure 6f).

Figure 6.

(a) Clpx deletion induces spermatocyte apoptosis during the pachytene stage. Testis sections from 8-week-old control and Clpx^cKO males were stained for c-PARP (green) and DAPI (blue). Scale bar = 100 μm. (b) Quantification of the numbers of c-PARP-positive seminiferous tubules. (c) Numbers of c-PARP-positive cells per tubule. The mean was calculated for each male and data from three males for each genotype were used. n: the pooled number of seminiferous tubules analyzed from three animals for each genotype. The data are shown as mean ± s.d. ^**P < 0.01, ^****P < 0.0001. (d) Testis sections from 8-week-old control and Clpx^cKO males stained for c-PARP (green), SYCP3 (red), and DAPI (blue). The regions within the dashed line of the square are enlarged on the right, and the dashed lines separate the different types of cells. PreL: preleptotene; Z: zygotene; P-like: pachytene-like spermatocytes; D: diplotene spermatocytes. Scale bar = 20 μm. (e) Numbers of c-PARP-positive spermatocytes in different stages. n: the pooled number of seminiferous tubules containing c-PARP-positive spermatocytes. The data are shown as mean±s.d. ^****P < 0.0001. (f) Schematic summary of Clpx^cKO male phenotypes. c-PARP: cleaved poly(ADP-ribose) polymerase; ClpX: caseinolytic mitochondrial matrix peptidase X; SYCP3: synaptonemal complex protein 3; DAPI: 4’,6-diamidino-2-phenylindole; Pac: pachytene spermatocytes; Dip: diplotene spermatocytes; s.d.: standard deviation.

DISCUSSION

The mitochondrial caseinolytic protease complex ClpXP maintains the homeostasis of mitochondrial proteins in organisms.32,33 However, little is known about the function of the protease ClpX in meiosis. In this study, we used an inducible inactivation strategy to reveal the role of ClpX in homologous chromosome pairing, synapsis, and recombination during meiotic prophase I. We found that ClpX participated in mitochondrial protein quality control in spermatocytes, and the growth of Clpx^cKO spermatocytes was arrested in the zygotene-to-pachytene transition with impaired homologous pairing, synapsis, and recombination.

DSB formation is a requisite for the accurate assembly of SCs during meiosis. Therefore, the incomplete assembly of SCs observed in Clpx^cKO spermatocytes could be due to dysfunctional DSB repair. Mice carrying recombinant mutations in genes such as BRCA2,34 DMC1,35 and SPO11³⁶ exhibited asynapsis or nonhomologous chromosome synapsis. We noted abundant RPA2 foci and intense γH2AX staining, which indicate the existence of unrepaired DSBs in Clpx^cKO spermatocytes. Consistently, we found that the deletion of Clpx significantly impacted the recruitment of the recombinant protein RAD51 to DSBs during the zygotene and pachytene-like stages, corroborating that the RAD51-DNA presynaptic complex was essential for synapsis and recombination. Although γH2AX phosphorylation promotes the early repair of DSBs by increasing RAD51 recruitment in meiosis, we detected the γH2AX signal throughout the asynaptic axis in Clpx^cKO pachytene-like spermatocytes, but RAD51 recruitment was not higher. Clpx-deficient spermatocytes may suffer from mitochondrial damage, which precludes the recruitment of the ATP-dependent RAD51 to the DSBs. Thus, we conclude that the knockout of Clpx disrupts RAD51 recruitment to DSBs in spermatocytes, which hinders the search for homologous chromosomes, synapsis, and DSB repair.

Synapsis and recombination of the X and Y chromosomes occur in the pseudoautosomal region of mammalian spermatocytes during meiotic prophase I. In response to asynapsis of the XY chromosome, DDR factors such as BRCA1, ATR, and γH2AX are restricted to the lateral element regions to achieve MSCI.21 In this study, γH2AX staining revealed that a typical sex body failed to form within Clpx^cKO spermatocytes, reflecting the failure of the DSB repair machinery and meiotic arrest at the pachytene-like stage. Ubiquitination and SUMOylation have been shown to contribute toward repression of sex chromosome transcription downstream of the DDR pathway.21,24,37,38 We found that the expressions of Ub-K48, Ub-E6C5, and Ub-FK2 were significantly reduced on autosomes and sex chromosomes, suggesting that asynapsis and associated MSCI were more prevalent in Clpx-deficient spermatocytes. Accordingly, we hypothesized that abnormal histone ubiquitination in Clpx^cKO spermatocytes may be associated with mitochondrial OXPHOS damage. Indeed, we found that the expression of OXPHOS subunits was significantly mitigated in Clpx^cKO testes. The link between mitochondrial OXPHOS damage and the extent of aberrant ubiquitination remains unknown. The regulation of mitochondrial homeostasis and chromosomal behavior inside the nucleus is intricate and requires further investigation.

CONCLUSION

These results demonstrate that the mitochondrial protein ClpX was required for meiosis in male mice. By maintaining mitochondrial homeostasis, ClpX may provide sufficient energy for recruiting the nuclear recombinase RAD51 to DSB sites, which in turn successfully executes DSB repair and synapsis of homologous chromosomes, leading to the completion of meiosis. This study provides a new paradigm to understand the relationship between cytoplasmic mitochondria and nuclear chromosome behavior during meiosis.

AUTHOR CONTRIBUTIONS

HWF, DTL, and LJH were responsible for the study conception and data discussion. HWF, YZ, and YLG were responsible for data collection and the accuracy of the data. HWF drafted the manuscript, and all authors revised the manuscript. DTL and LJH supervised the project. All authors read and approved the final manuscript.

COMPETING INTERESTS

All authors declare no competing interests.

Supplementary Figure 1

CLPX is expressed in pachytene spermatocytes at high levels. Immunofluorescent labeling of testis sections from control males at PD56 with SYCP3 (red), CLPX (green), DAPI (blue), and enlarged images are shown. Seminiferous epithelium at stages VIII–XII displayed the highest CLPX expression on pachytene spermatocytes. PreL, preleptotene. Z, zygotene. P, pachytene. D, diplotene. Dia, diakinesis. MI, metaphase I. RS, round spermatids. Scale bar, 50 μm. ClpX: caseinolytic mitochondrial matrix peptidase X; SYCP3: synaptonemal complex protein 3; DAPI: 4’,6-diamidino-2-phenylindole.

Supplementary Figure 2

Tamoxifen successfully induces Cre-mediated deletion of Clpx in male mice. (a) Schematic of the Clpx conditional knockout allele. (b) Tamoxifen treatment regimen. Intraperitoneal injections (80 mg/kg) were administered to PD12 male pups for 4 consecutive days. Samples were collected at PD20 (5 dpt), PD25 (10 dpt), PD30 (15 dpt), PD35 (20 dpt), and PD56 (41 dpt). (c) Genotyping results to distinguish the control and Clpx^cKO animals. #1, Clpx^fl/+ Ddx4-Cre mouse. #2, Clpx^fl/fl Ddx4-Cre mouse. Primer pair of F1/R1 for Clpx-floxed allele, F1/R2 for Clpx^cKO allele, and F3/R3 for the Ddx4-Cre allele were used. (d) Histological sections of control and Clpx^cKO testes stained with hematoxylin. Scale bar, 25 μm. (e) Immunofluorescence staining of testicular sections from control and Clpx^cKO mice with antibodies against TOM70 (green) and CLPX (red). The nuclei were stained with DAPI (blue). Scale bar, 5 μm.

Supplementary Figure 3

Deletion of Clpx does not affect CLPP expression in mitochondrial spermatocytes. Immunofluorescent staining of CLPX (green), CLPP (yellow), SYCP3 (red), and DAPI (blue) in adjacent testis sections from control and Clpx^cKO males at PD56. Stages of seminiferous tubules are indicated. PreL, preleptotene. Z, zygotene. P, pachytene. D, diplotene spermatocytes. Scale bars, 50 μm. ClpX: caseinolytic mitochondrial matrix peptidase X; SYCP3: synaptonemal complex protein 3; DAPI: 4’,6-diamidino-2-phenylindole.

Supplementary Figure 4

Failure in RPA2 degradation in Clpx^cKO spermatocytes. Analysis of adjacent histological sections of adult control and Clpx^cKO with fluorescent RPA2 (green), SYCP3 (red) and DAPI (blue). The region in the dashed box is magnified in its right panels. PreL, preleptotene. L, leptotene. P, pachytene. D, diplotene spermatocytes. Scale bars, 50 μm. SYCP3: synaptonemal complex protein 3; DAPI: 4’,6-diamidino-2-phenylindole.

Supplementary Figure 5

Normal telomere-nuclear envelope connection in Clpx^cKO spermatocytes. Equator images of spermatocytes at pachytene and pachytene-like, stained with TRF1 (red), SYCP3 (green), and Lamin B (blue). Numbers on the graph indicate the percentage of cells deficient in telomere anchoring to NE in control pachytene and Clpx^cKO pachytene-like spermatocytes, respectively. Scale bars, 10 μm. SYCP3: synaptonemal complex protein 3.

Supplementary Figure 6

Ubiquitination of XY chromosomes and autosomes is aberrant in Clpx^cKO pachytene-like spermatocytes. Immunostaining of spermatocyte surface spreads for SYCP3 (red) and ubiquitin (green). (a) Ub-H2A (Clone E6C5). (b) Ubiquitin (Clone FK2). (c) Ub-K48. Scale bars, 5 μm. SYCP3: synaptonemal complex protein 3.