ATAD2 deficiency leads to subfertility by impairing spermatogenesis in mice

Abstract

During spermatogenesis, chromatin remodeling regulated by histone modification is essential for spermatogenic cell development, and multiple epigenetic regulators are involved in this process. Recent studies reported that ATAD2 was a newly discovered cancer/testis factor that functioned in chromatin remodeling in somatic cells by binding acetylated histone. However, the physiological role of ATAD2 in spermatogenesis is largely unknown. In this study, we characterized the expression pattern of ATAD2 in mouse testes and found that the highly expressed ATAD2 in the gonads was lowly expressed in meiotic spermatocytes with distinct localization in nucleus, but highly expressed in round spermatids. By generating Atad2 knockout (KO) mice using CRISPR/Cas9 technology, we revealed that ATAD2 deletion leads to failure of DSB repair and chromosome synapsis in spermatocytes and impairs spermiogenesis in spermatids. Atad2 KO mice were subfertile, as characterized by reduced sperm count, impaired motility, and abnormal morphology. RNA-Seq analysis showed that hundreds of genes were dysregulated in Atad2-KO round spermatids. As revealed by GSEA analysis, the gene set related to spermatid development was downregulated, while gene sets related to chromatin binding and positive and negative DNA-templated transcription were upregulated. In conclusion, our results indicate that ATAD2 contributes to meiotic progression and participates in spermiogenesis by regulating RNA transcription in spermatids.

Spermatogenesis, which is dedicated to the generation of sperm cells, is a complicated process that includes three main steps taking place in the seminiferous epithelium: 1) spermatogonial renewal and proliferation via mitosis; 2) spermatocyte meiosis from type B spermatogonia to preleptotene spermatocytes and to diplotene spermatocytes followed by meiosis I and II; and 3) development of round spermatids to elongated spermatids, and then spermatozoa via spermiogenesis. During germ cells transition from the mitotic cycle into meiosis, chromatin reorganization is required for chromosome condensation and homolog pairing and genetic recombination [1,2,3]. During spermiogenesis, spermatids undergo extensive chromatin reorganization for chromosomal condensation and replacement of histones by protamines, i.e., 90%–99% of nuclear histones are removed and replaced by small, basic protamines [4, 5]. Histone acetylation is the regulator of chromatin reorganization and occurs in different stages of male germ cells [6, 7]. Histone H4 hyperacetylation, which occurs in elongating spermatids to remove histones and regulate the transcription of spermatogenic genes, appears to be essential for the progress of spermiogenesis [8, 9]. Recently, several regulatory proteins involved in histone H4 hyperacetylation were found. Knockout (KO) of histone acetyltransferase Gcn5 resulted in abnormal chromatin dynamics, leading to increased sperm histone retention and disruption of male fertility [10]. EPC1 and TIP60, two components of the nucleosome acetyltransferase of H4 (NuA4) complexes, are required for histone hyperacetylation and compaction of the spermatid genome during spermiogenesis [11]. Histone hyperacetylation reader BRDT, which contains bromodomains, is expressed specifically in late prophase I spermatocytes and spermatids [12]. Loss of bromodomain 1 results in defects in spermiogenesis due to impaired chromatin organization in the spermatids [13, 14].

ATPase family AAA domain-containing protein 2 (ATAD2), containing AAA (“ATPases Associated with diverse cellular Activities”) domains and a bromodomain for binding acetylated histones in chromatin, is widely recognized as a newly emerging oncoprotein due to its importance in a wide range of human cancers [15]. In embryonic stem cells, ATAD2 binds to chromatin by recognizing histone acetylation, resulting in an overall increase in chromatin accessibility and histone dynamics [16]. In addition to in many cancer cells, ATAD2 is also highly expressed in testes and is therefore also recognized as a new cancer/testis factor [17]. However, the role of ATAD2 in spermatogenesis has not been investigated yet, which warrants a systematic study of ATAD2 during spermatogenesis and its influence on chromatin structure, nucleosome retention, and fertility.

Here, we generated an Atad2 KO mouse using CRISPR/Cas9 technology to investigate its role in spermatogenesis. It was found that KO of Atad2 impairs meiosis and the post-meiotic spermatid development and that Atad2 KO mice suffered from oligo-astheno-teratozoospermia and thus subfertility. Combined, our results demonstrate the critical role of ATAD2 in male germline development and fertility.

Materials and Methods

Animals

An Atad2 KO mouse was generated by using the CRISPR/Cas9 techniques. The mouse genotypes were determined by genotyping PCR based on the extracted DNA from tail tissue (primers can be found in Supplementary Table 1). All animal experiments were approved and conducted according to guidelines and regulatory standards from the Institutional Animal Care and Use Committee of Wuhan University. All mice were housed in the Wuhan University animal vivarium with standard 12 h light/dark cycles with ad libitum access to standard laboratory chow and water. All experimental animals were previously untreated, and were considered healthy at the time of experimental use. All animal experiments in the study complied with the regulatory standards approved by the Institutional Animal Care and Use Committee of Wuhan University.

Fertility assessment

For fertility tests, three adult Atad2 KO males and control males were separately mated with two sexually mature wild-type (WT) females for at least two months. The number of pups per litter was counted and recorded.

Sperm count, motility, and morphology analysis

To quantify sperm numbers, 10-week-old male mice were anesthetized using CO₂ and then euthanized via cervical dislocation. The cauda epididymis from each mouse was excised and minced into small pieces, followed by incubation in 1 ml human tubal fluid (HTF) medium (MR-070-D; Merck Millipore, Darmstadt, Germany) at 37°C for 30 min. For sperm parameter analysis, a 20-µl aliquot of the sperm suspension was analyzed using a computer-aided sperm analysis (CASA) system to obtain sperm parameters. Briefly, sperm were observed under a phase-contrast microscope (Leica, Wetzlar, Germany), and analyzed with a sperm quality analysis system (BEION V4.20, Beionmed, Beijing, China) to extract key velocity-related parameters alongside sperm count and overall motility. For the assessment of sperm morphology, sperm spreads were prepared, fixed, and stained following the manufacturer's instructions using a Diff-Quick Stain Kit (G1540, Solarbio, Beijing, China). The percentage of morphologically normal sperm was evaluated under a bright-field optical microscope (Axio Imager 2, Zeiss, Oberkochen, Germany).

Western blot

Total proteins were extracted from tissues of adult mice using RIPA buffer supplemented with a protease inhibitor cocktail (PIC) (P6730, Solarbio). Protein samples were separated on a 10% SDS-PAGE gel, followed by transfer to a piece of polyvinylidene fluoride (PVDF) membrane (IPVH00010; Merck Millipore). The membrane was blocked with 5% non-fat milk (1172, BioFroxx, Einhausen, Germany) for 1 h and then incubated with primary antibodies at 4°C overnight (Supplementary Table 2), followed by incubation with a goat anti-rabbit/mouse IgG-HRP secondary antibody (1:8000 dilution, SA00001-2/SA00001-1; Proteintech, Rosemont, IL, USA) for 1 h at room temperature after washing. Finally, protein bands were visualized with SuperSignal^TM West Pico PLUS (34580, Thermo Fisher Scientific, Waltham, MA, USA) and pictures were captured with G: Box Chemi-XRQ GENE System (Syngene, Cambridge, UK).

Histological analysis, immunofluorescence staining, and imaging

Histological analysis, frozen section immunofluorescence staining, and spermatocyte spreads were performed as previously described [18]. Briefly, for histological analysis, testis tissue was fixed in Bouin’s solution (HT10132, Sigma, St. Louis, MO, USA) overnight at room temperature, embedded with paraffin, and sectioned at 5–8 µm in thickness. Then, the sections were stained with hematoxylin and eosin. For immunofluorescence staining, the tissues were fixed in 4% paraformaldehyde (PFA) overnight at 4°C, dehydrated, embedded in optimal cutting temperature compound, and sectioned at 5–8 µm in thickness using a cryostat (HM525NX, Thermo Fisher Scientific). Frozen sections were permeabilized with TBST solution (1% Trition X-100 in TBS), blocked with 5% bovine serum albumin, incubated with primary antibody (Supplementary Table 2) at 4°C overnight, and stained with a CoraLite488- or CoraLite594-conjugated secondary antibody (Supplementary Table 2). TUNEL assays were performed using a TMR (Red) Tunel Cell Apoptosis Detection Kit (G1502, Servicebio, Wuhan, China) according to the manufacturer’s instructions. Briefly, following permeabilization treatment, the samples were incubated in TUNEL detection solution. For visualization of nuclei, the samples were stained with DAPI (0100-20, SouthernBiotech, Birmingham, MA, USA). Images were captured with an Axio Imager 2 microscope (Zeiss, Baden-Württemberg, Germany).

Round spermatid purification by fluorescence-activated cell sorting (FACS)

Round spermatids were isolated from adult male mice using FACS as previously described [19]. In brief, testes were obtained from 3-month-old mice and seminiferous tubules were enzymatically digested with collagenase I (C0130, Sigma) at a concentration of 250 µg/ml in 10 ml of 1x DMEM (L110KJ, Basalmedia, Shanghai, China) for approximately 15 min at 37°C. Digested tissues were then treated with 10 µg/ml DNase I (D4263, Sigma) and 250 µg/ml Trypsin (T9201, Sigma) for an additional 15 min at 37°C. The enzymatic digestion was halted by adding 1 ml of fetal bovine serum (FBSSA500-S, Gold Coast, QLD, Australia). Following filtration through a 70 µm cell strainer (352350, Corning Incorporated, Corning, NY, USA), the flow-through was centrifuged at 300× g for 10 min. The cell pellet was resuspended in 1× DMEM supplemented with 10% FBS, and the cell concentration was adjusted to (5–8) × 10⁶ cells/ml. Next, the cells were stained with Hoechst 33342 (H1399, Invitrogen, Thermo Fisher Scientific) at a concentration of 5 µg/ml for 30–40 min at 32°C. After centrifugation, the pellets were resuspended in PBS containing 2% FBS. Cell sorting and data analysis were conducted with an ARIA III cell sorter (Becton Dickinson, Franklin Lakes, NJ, USA).

RNA-seq

RNA-sequencing analysis was conducted at Benagen Company Ltd., Wuhan, China (www.benagen.com). Total RNA was extracted separately from isolated round spermatids, with three biological replicates for each genotype. mRNA libraries were prepared using a NEBNext® UltraTM RNA Library Prep Kit for Illumina®, followed by sequencing on a Novaseq 6000 platform (Illumina, San Diego, CA, USA). For gene expression analysis, the false discovery rate (FDR) was computed and used for filtering significantly differentially expressed genes (DEGs), which were determined using R (v3.5.1) with the DESeq2 package, applying a threshold value of FDR < 0.05 and fold change > 1.5 or < 0.5. Data are deposited in the Sequence Read Archive (SRA) under BioProject ID/Accession No. PRJNA1082210 (https://www.ncbi.nlm.nih.gov/bioproject/PRJNA1082210/).

In vitro fertilization

Sperm were collected from the epididymides of male mice and allowed to undergo capacitation for 1 h in human tubal fluid (HTF) medium (MR-070-D; Merck Millipore). Cumulus-oocyte complexes (COCs) were isolated from the oviductal ampullae of female mice. Following capacitation, the sperm were introduced into HTF medium droplets containing COCs for fertilization, which was carried out in a 37°C, 5% CO₂ incubator (MCO-18AC; Panasonic, Osaka, Japan). After 6 h of co-incubation, the COCs were washed to remove cumulus cells and excess sperm. The 2PN rate was calculated as the number of 2PN zygotes divided by the total number of retrieved eggs.

Statistical analysis

All data are expressed as mean ± standard deviation (SD). Statistical comparisons between groups were conducted by Student's t test using GraphPad Prism v8.0.1 (GraphPad Software, Boston, MA, USA). A P value < 0.05 was considered to be statistically significant.

Results

ATAD2 is highly expressed in mouse testes

To investigate the role of ATAD2 in fertility, we first analyzed its expression profile in multiple tissues of adult mice by Western blot. The results showed that it was highly expressed in the testes and the spleen (Fig. 1A). Next, we determined the dynamic ATAD2 protein levels in developing testes from postnatal day 3 (P3) through P35, when the first wave of spermatogenesis happened, and the adult stage P56 (Fig. 1B). A previous study reported that another longer Atad2 transcript besides the canonical one existed in testes, encoding a longer ATAD2 with an additional 324 amino acids at the N-terminal region (ATAD2-L) [17]. In our study, ATAD2-L was detectable in testes from P3 to P56, while the canonical ATAD2 encoded by the shorter transcript was almost undetectable until P35, with an extremely low expression before P35 (Fig. 1B). Immunofluorescence staining of testicular cryosections demonstrated that ATAD2 expression was weak in spermatogonia, spermatocytes and early stages of the post-meiotic round spermatids of tubules at stage I-VI, but peaked in late stages of round spermatids of tubules at stage VII and VIII, and returned lower in the elongated spermatids after stage IX (Fig. 1C). In order to determine the expression of ATAD2 in spermatocytes in detail, we assayed chromosome spreads of spermatocytes by immunofluorescence staining. The results indicated that ATAD2 is lightly expressed in the nucleus of spermatocytes from leptonema to diplonema, except in the XY body (Fig. 1D). Taken together, ATAD2 is preferentially expressed in mouse testes and enriched in round spermatids at step 7-8, suggesting that it may play an important role in spermatogenesis.

Fig. 1.

Expression patterns of ATAD2 in mice. (A) Western blot showing the expression levels of ATAD2 in nine different organs/tissues of WT adult mice. GAPDH served as a loading control.ATAD2-L (long isoform) and ATAD2-S (short isoform) are indicated. (B) Western blot illustrating the ATAD2 levels in developing testes. Testes on postnatal day 3 (P3), P6, P12, P14, P18, P21, P35 and P56 were analyzed. GAPDH served as a loading control. (C) Immunofluorescence staining to determine the subcellular localization of ATAD2 in adult testis. Scale bar, 50 µm. (D) Localization of ATAD2 in spermatocytes determined by co-immunostaining of spermatocyte spreads using antibodies against ATAD2 and SYCP3. SYCP3 shows the stage of spermatocytes in prophase meiosis I. Nuclei were stained with DAPI. Scale bar, 20 µm. (E) Two different microarrays (GDS2696 and GDS2697) captured Atad2 transcript levels in sperm samples from normal and teratozoospermic men. GDS2696: 5 normal individuals and 8 teratozoospermia patients; GDS2697: 13 normal individuals and 8 teratozoospermia patients. We retrieved two GEO datasets (GSE6872 and GSE6967) [20] and extracted the data of Atad2 transcription levels in seminal fluid from normal individuals and males with teratozoospermia. ** P < 0.01, *** P< 0.001.

In addition, we retrieved two GEO datasets (GSE6872 and GSE6967) [20] and extracted the data of Atad2 transcription levels in seminal fluid from normal individuals and males with teratozoospermia. Microarray analysis of 21 human sperm samples (GDS2697, 13 normal ones and 8 teratozoospermia ones) revealed that the Atad2 transcript levels were significantly lower in the teratozoospermia samples than in the normal ones (Fig. 1E, P < 0.001). Similarly, microarray data (GDS2696) also revealed significantly decreased Atad2 mRNA levels in the teratozoospermia samples (Fig. 1E; P < 0.01). In conclusion, the Atad2 transcript level is lower in semen samples from patients with teratozoospermia, further suggesting a potential role of ATAD2 in spermiogenesis.

Generation of Atad2 knockout mice

To elucidate the physiological role of ATAD2 in spermatogenesis, we constructed an Atad2 KO mouse model by targeting the genomic fragment between exon 2 and exon 13 of Atad2 gene with two sgRNAs using CRISPR/Cas9 technology (Fig. 2A). The Atad2 KO mouse lines carried a 21,240 bp deletion in the Atad2 gene, as detected by Sanger sequencing, finally causing a frameshift prematured termination mutation (Fig. 2B). WT, heterozygous (HET) and homozygous mutant Atad2 alleles were confirmed by genotyping PCR of mouse tail DNA (Fig. 2C). Western blot confirmed the absence of the ATAD2 protein in Atad2 KO testes (Fig. 2D). Furthermore, double immunofluorescence staining of ATAD2 and γH2AX (a marker for DNA double-strand breaks) in mouse testicular sections also reported that ATAD2 was absent from the Atad2 KO testes (Fig. 2E). Taken together, these results indicated the successful generation of Atad2 KO mice.

Fig. 2.

Generation of Atad2 knockout mice using CRISPR/Cas9 technology. (A) Schematic representation of the targeting strategy for generating Atad2 knockout (KO) mice using CRISPR/Cas9. (B) Sanger sequencing results showing a 21,240 bp deletion in the Atad2 gene. (C) Representative genotyping PCR results of Atad2 alleles showing that the PCR product of WT allele is 350 bp long (primers F2 and R2) and the product of mutant allele is 352 bp long (primers F1 and R1). M, marker; Het, heterozygous; H₂O, negative control. (D) Western blot showing the ATAD2 protein levels in WT and KO testes on P30. GAPDH served as a loading control. (E) Co-immunofluorescence staining of ATAD2 and γH2AX in the testes of adult wild-type (WT) control and KO mice. Nuclei were stained with DAPI. Scale bars, 50 µm.

Loss of Atad2 in mice results in spermatogenic defects

All Atad2 KO mice exhibited normal somatic growth without significant developmental defects. However, the testes of Atad2 KO mice were markedly smaller than those of WT males. Meanwhile, the epididymides of Atad2 KO mice also showed a noticeable reduction in size compared to WT controls (Figs. 3A, B). Fertility test reported that females inseminated by KO males produced smaller litter sizes when compared with those mated with WT males (Fig. 3C), indicating an attenuated fertility in Atad2 KO males. Evaluation of epididymal sperm parameters by CASA demonstrated that KO male mice had a significant reduction in both sperm count and motility when compared with WT males (Figs. 3D, E). Further microscopic examination of epididymal sperm morphology revealed a significantly higher incidence of sperm head malformations in KO mice (Fig. 3F). In order to determine the fertilizing ability of Atad2 KO sperm, we conducted an in vitro fertilization experiment. Our results indicated that the sperm-fertilizing ability of Atad2 KO mice in vitro was lower than that of the wild-type control mice, but there was no significant difference (Supplementary Fig. 1). These findings collectively indicated that ATAD2 deficiency led to oligo-astheno-teratozoospermia (OAT) in male mice and affects the fertilizing ability of sperm.

Fig. 3.

Loss of ATAD2 in mice results in spermatogenic defects. (A) Testis tissue from WT and Atad2 KO mice aged 10 weeks. (B) Testis/body weight ratio of 10-week-old control and Atad2 KO male mice(n = 3/group). (C) Fertility analysis of WT and Atad2 KO male mice (n = 3/group). (D) Sperm number per epididymis from 10-week-old WT and Atad2 KO male mice(n = 3/group). (E) Motility and morphology of epididymal sperm from control and Atad2 KO males(n = 3/group). (F) Percentage of abnormal sperms in WT and Atad2 KO mice(n = 3/group). (G) TUNEL analysis of testicular tissue sections from control and Atad2 KO testes. Scale bars, 50 µm. (H) Quantitative analysis of TUNEL positive cells in TUNEL positive seminiferous tubules from WT and Atad2 KO testes (n = 3/group). (I) Hematoxylin and eosin staining of testis and epididymis sections from WT and Atad2 KO mice. a, presence of multinucleated giant cells; b, absence of spermatocytes; c, absence of elongated spermatids. Red arrow indicates the immature germ cells detached from seminiferous epithelium. Scale bar, 50 µm. Bar graphs represent mean ± SD. * P < 0.05, ** P < 0.01, *** P < 0.001, **** P < 0.0001.

TUNEL staining of testicular cryosections demonstrated a relatively larger number of apoptotic cells in the seminiferous tubules of Atad2 KO mice, with these TUNEL-positive cells predominantly localized to the middle regions of the seminiferous epithelium (Figs. 3G, H). To further investigate the impact of ATAD2 deficiency on spermatogenesis, we conducted histological analyses on testes and epididymides. Hematoxylin-eosin (HE) staining of testicular sections showed normal spermatogenic cycles in most seminiferous tubules of Atad2 KO mice. However, a subset of tubules exhibited abnormalities, manifested by absence of spermatocytes, lack of elongated spermatids, or presence of multinucleated giant cells (Fig. 3I). HE staining of epididymides revealed reduced sperm numbers in the lumen of Atad2 KO mice when compared with WT controls, along with occasional presence of immature germ cells detached from seminiferous epithelium (Fig. 3I). Taken together, these results suggest that ATAD2 deficiency resulted in defective spermatogenesis and OAT with attenuated fertility in male mice.

ATAD2 deficiency impairs DSB repair and chromosome synapsis during meiosis

Because Atad2 is expressed in spermatocytes and its deletion resulted in defective spermatogenesis, we then investigated its role in meiosis. Double-strand break (DSB) repair and chromosomal synapsis between homologous chromosomes are two critical events for proper meiotic progression in spermatocytes. To determine the influence of ATAD2 deficiency on DSB repair, we immunostained chromosome spreads of spermatocytes with antibodies against γH2AX, a marker of DSB, and SYCP3, a component of the synaptonemal complex (SC) axial elements. In WT pachytene spermatocytes, γH2AX signals were exclusively localized to the XY body and absent from the autosomes (Fig. 4A). However, γH2AX signals were still observed on autosomes in a small proportion of Atad2 KO pachytene spermatocytes, indicating that ATAD2 deficiency impairs DSB repair (Fig. 4A).

Fig. 4.

Defective DSB repair and chromosome synapsis in Atad2 KO spermatocytes. Chromosome spreads of pachytene spermatocytes from WT and Atad2 KO testes were co-stained with SYCP3 (green) and γH2AX (red)(A), SYCP3 (red) and SYCP1 (green)(B), SYCP3 (red) and MLH1 (green)(C). Nucleus was stained with DAPI. Scale bar, 20 µm. (D) Quantitative analysis of the number of MLH1 foci per pachytene spermatocyte. n = 3/group (5-7 pachytene spermatocytes counted per mouse). **** P < 0.0001.

To evaluate the chromosome synapsis between homologs in pachytene spermatocytes, we performed immunofluorescence staining of spread nuclei using antibodies against SYCP3 and SYCP1 (a component of the SC transverse elements). The results showed that homologs were completely synapsed in autosomes, and sex chromosomes were synapsed in the PAR region in WT spermatocytes (Fig. 4B). However, in some Atad2 KO pachytene spermatocytes, autosomes were not in full synapsis, and sex chromosomes were dissociated (Fig. 4B). As a result of meiotic recombination, crossover takes place between homologous chromosomes. We visualized MLH1 foci on pachytene spermatocytes to determine the influence of Atad2 deletion on crossover formation and found that the number of MLH1 foci was significantly decreased in KO spermatocytes when compared with WT pachytene cells (Figs. 4C, D).

ATAD2 deficiency impairs spermiogenesis

As established above, ATAD2 is predominantly expressed in round spermatids, and its deletion induced OAT in mice. Therefore, we postulate that ATAD2 deficiency impairs spermiogenesis. To test this hypothesis, we performed immunofluorescence staining on cryosectioned testes from 2-month-old mice to evaluate their spermatid development.

Seminiferous tubules were staged (I-XII) according to the murine spermatogenic cycle, which is subdivided into 16 steps [21]. Using these criteria, we systematically analyzed spermatid maturation in WT and Atad2 KO mice. At 2 months after birth, the seminiferous tubules of Atad2 KO males exhibited reduced round spermatid counts. Ectopic presence of round spermatids was observed in stage VII tubules, and aberrant round spermatid-like (RSL) germ cells appeared in stage IX-X and XI-XII tubules (Fig. 5A). Misshaped acrosomes and spermatids were found in stage IX-X tubules, and only few spermatids were observed in stages IX-X tubules (Fig. 5A). Elongating spermatids (steps 9–16) displayed incomplete nuclear condensation in KO tubules at stages II-IV, V-VI, and VIII (Fig. 5A). Consistently, quantitative analysis further demonstrated reduction​ in both elongated spermatids (P < 0.01) and mature spermatozoa (P < 0.001) in KO seminiferous tubules (Figs. 5B, C). Therefore, we conclude that ATAD2 deficiency impairs spermiogenesis.

Fig. 5.

ATAD2 deficiency leads to defective spermiogenesis. Frozen sections of control and Atad2 KO testes from 2-month-old mice were stained with PNA and DAPI. Scale bar, 50 µm. (B) Representative image of immunofluorescence-stained testis tissue by PNA and DAPI showing decreased elongating spermatids (ES) and spermatozoa (SP) in 2-month-old Atad2 KO tubules. (C) Comparison of the number of ES in X-XI tubules and SP in VII-VIII tubules between 2-month-old Control and KO mice (n = 3/group). ** P < 0.01, *** P< 0.001.

ATAD2 deficiency alters the transcriptome of round spermatids

To investigate molecular defects underlying spermiogenesis in Atad2 KO mice, we purified the round spermatids and performed RNA-seq on them. Comparison of the mRNA level revealed 421 upregulated and 317 downregulated genes in Atad2 KO testes compared with WT testes (FDR-adjusted P value < 0.05 and absolute log2(fold change) > 1) (Fig. 6A). To explore the biological functions of the differentially transcribed genes, we carried out GO enrichment analysis and found that multiple GO functional terms are significantly enriched (Fig. 6B). For biological processes, genes associated with reproduction and germ cell development were significantly changed and for molecular functions, genes involved in chromatin binding and histone binding were enriched. We performed further analysis of the expression patterns of DEGs via gene set enrichment analysis (GSEA). Spermatid development was the top enriched gene set and was downregulated. These genes were involved in histone-protamine transition, SEPTINS, the ubiquitin-proteasome-system, and testis-specific serine kinases such as Tnp1, Prm1, Prm2, Septin4, Tssk1, Ube2j1, and Cdyl. Chromatin binding and positive and negative DNA-templated transcription were the top upregulated gene sets (Fig. 6C). Chromatin binding gene sets were involved in chromatin remodeling, chromatin modifying enzymes, and SUMO E2 ligase SUMOylate target proteins such as EP300, PBRM1, KMT5C, CHD8, RNF20 ATRX, KMT2A, and L3MBTL2. In conclusion, gene expression is dysregulated in Atad2 KO round spermatids.

Fig. 6.

Transcriptome analysis of WT and Atad2 KO round spermatids. (A) Volcano plot of differentially expressed genes in Atad2 KO round spermatids, with a significance threshold of FDR < 0.05 and fold change > 1.5. (B) Bubble plot of Gene Ontology (GO) enrichment analysis. The color of the bubbles represents the P value, and the size of the bubbles represents the number of genes enriched in corresponding term (count). (C) Enrichment results for spermatid development, chromatin binding, and positive and negative regulation of DNA-templated transcription as obtained via GSEA.

Discussion

A previous study has reported that ATAD2 is a new cancer/testis factor and is highly expressed in testes [17]. However, its role in spermatogenesis remained unknown. Here, we established a global Atad2 KO mouse line for the first time and demonstrated that knockout of Atad2 led to OAT by impairing spermatogenesis in mice.

Using Western blot, we identified two protein isoforms of ATAD2 in testicular tissues, consistent with previous reports [17]. Through comparative analysis of Atad2 expression patterns during postnatal development, we observed low ATAD2 protein levels from P3 to P14, followed by a gradual increase from P18 to P56. Notably, testicular tissues at P25 predominantly produced the long isoform (ATAD2-L), whereas the canonical short isoform became dominant on P35 (coinciding with round spermatid elongation), suggesting stage-specific isoform switching during spermiogenesis. Given that spermatogenic cells at P25 develop into secondary spermatocytes with few round spermatids, while massive round spermatids undergo spermiogenesis at P35, we speculated that the canonical ATAD2 mainly played a role in spermiogenesis, whereas ATAD2-L functioned primarily in meiosis and only to a lesser extent in spermiogenesis. Immunofluorescence staining localized ATAD2 primarily to step 7–9 round spermatids, further supporting its functional importance in sperm morphogenesis, as knockout of Atad2 impaired spermatid development, resulting in OAT in male mice.

There are two possible explanations for ATAD2’s role in regulating spermiogenesis. First, it may participate in regulating chromatin remodeling by binding acetylated histone H4. Our experiments support this because ATAD2 was highly expressed in steps 7–9 round spermatids, which is a critical stage of spermatid elongation and histone H4 hyperacetylation. Also, in human cells, ATAD2 recognizes acetylated H4 and regulates the level of histone H4 acetylation [22]. According to the RNA-Seq analysis of round spermatids, the chromatin binding gene set was upregulated. Among these genes, EP300 is a histone acetyltransferase and regulates chromatin remodeling [23]; CHD8 is a chromatin remodeling factor and required for spermatogonia proliferation and early meiotic progression [24]; ATRX is a chromatin remodeling protein localized in somatic and spermatogenic cells of human testis and may play a role in spermatogenesis by regulating chromatin [25]; KMT2A is a lysine-specific histone N-methyltransferase and promotes transcription by inducing an open chromatin conformation [26]; and L3MBTL2 functions in chromatin remodeling during meiosis and spermiogenesis [27]. Therefore, ATAD2 deletion may lead to chromatin remodeling during spermiogenesis. Second, ATAD2 is a transcription factor that is possibly involved in regulating gene expression during spermiogenesis. In this study, transcription was active in round spermatids at steps 1–8. ATAD2 approaches chromatin by binding acetylated histones, resulting in increased chromatin accessibility, which is required for the proper activity of the highly expressed gene [16]. As expected, the spermatid development genes were downregulated as a result of ATAD2 deletion. Among these genes, Tnp1, Prm1, and Prm2 are critical for the histone-protamine transition during spermiogenesis. Septin4 deletion results in male sterility with sperm shape abnormality [28], Tssk1 KO mice have lower sperm numbers and motility and are infertile [29], and Cdyl^cKO male mice suffer from defects in spermatozoon morphogenesis, demonstrating teratozoospermia [30]. Therefore, these downregulated genes may have contributed to the observed defects in spermiogenesis.

Intriguingly, we also detected Atad2 expression in meiotic spermatocytes, with distinct localization patterns during prophase I, enriched on autosomes but absent from sex chromosomes in pachytene and diplotene spermatocytes. As a result, ATAD2 deficiency led to failure in DSB repair and chromosome synapsis in partial pachytene spermatocytes. The similar expression pattern of ATAD2 and BRDT, both containing a bromodomain for recognizing acetylated histones, suggests that ATAD2 possibly functions as an epigenetic regulator in meiosis, similar to BRDT [2].

In summary, the current study demonstrates that ATAD2 likely has two main functions in spermatogenesis: one is in meiosis and the other is in spermiogenesis. These functions are related to the expression pattern in spermatogenic cells – distinct localization in meiotic spermatocytes and enrichment in round spermatids with stage-specific isoforms. In conclusion, the role of ATAD2 in regulating spermatogenesis is probably associated with chromatin remodeling and histone modification, which we will further evaluate in future studies.

Conflict of interests

The authors declare no conflict of interest.

Data availability statement

The datasets generated and/or analyzed during the current study are available in the NCBI (National Center for Biotechnology Information) repository under the accession number: PRJNA1082210. They can be accessed via the following link: https://www.ncbi.nlm.nih.gov/.