Introduction

Infertility is a global public health and social problem that affects approximately one in six couples worldwide (1). Male infertility, which accounts for half of infertile cases, is a multifactorial disease with common phenotypes, including oligo/azoospermia (poor sperm count or absence of spermatozoa); teratozoospermia (aberrant sperm morphology); asthenozoospermia (weakened sperm motility); and a combination of these phenotypes, such as asthenoteratozoospermia, oligoasthenozoospermia, oligoteratozoospermia and oligoasthenoteratozoospermia (2, 3).

Asthenoteratozoospermia is one of the most common phenotypes of male infertility, and genetic factors have been established as the predominant cause of asthenoteratozoospermia. Multiple morphological abnormalities of the flagella (MMAF), a subtype of asthenoteratozoospermia, characterized by a mosaic of abnormalities of the flagellar morphology, including absent, short, coiled, bent and/or irregular flagella, is almost always caused by genetic defects (4, 5). To date, more than 40 genes have been identified as pathogenic genes of MMAF, but these genes can only explain approximately 60% of MMAF-affected cases (6–9). Therefore, the genetic basis of the remaining cases is still unknown.

The motility of a sperm is driven by its rhythmically beating flagella, and at the center of the flagella lies a conserved axonemal structure, containing the “9 + 2” microtubular arrangement: a ring of nine microtubule doublets (MTDs) surrounding a central pair (CP) of singlet microtubules. Each MTD consists of an A tubule and a B tubule, and the outer (ODA) and inner (IDA) dynein arms are anchored along the A tubule (10). The ODA and IDA are ATPase-based protein complexes that drive the movement between the A tubule and the neighboring B tubule of the next doublet, producing the original force for sperm motility (11, 12). Structural and functional abnormalities of the ODA and IDA have been demonstrated to cause male infertility associated with asthenozoospermia and/or asthenoteratozoospermia (4, 13, 14).

The dynein axonemal heavy chain (DNAH) family comprises a series of proteins (DNAH1–3, DNAH5–12, and DNAH17) that are precisely assembled with other axonemal dynein motor proteins in the ODAs and IDAs of sperm flagella and motile cilia (15–17). In humans, DNAH1, DNAH2, DNAH6, DNAH7, DNAH8, DNAH10, DNAH12 and DNAH17 are highly expressed in the testis, and deficiency of these proteins has been demonstrated to cause MMAF-associated asthenoteratozoospermia (18–25). DNAH3 is an evolutionarily conserved IDA-associated protein and is highly expressed in testes of humans and mice (26). Deficient DNAH3 has been shown to impair sperm motility in Drosophila and cattle (27, 28). In humans, DNAH3 has been identified as a novel breast cancer candidate gene (29). However, the role of DNAH3 in male reproduction in humans and mice remains largely unknown.

In the present study, we identified four biallelic variations in DNAH3 in four unrelated Han Chinese patients with asthenoteratozoospermia using whole-exome sequencing (WES). The spermatozoa of the patients showed extremely reduced sperm motility and a high proportion of sperm tail defects characterized by the MMAF phenotype. We further generated Dnah3 knockout (KO) mice, and the male KO mice expectedly showed aberrations in sperm movement, flagellar IDA, and mitochondrion. Moreover, the absence of DNAH3 led to decreased expression of other IDA-associated proteins, including DNAH1, DNAH6 and DNALI1. Importantly, good outcomes of intracytoplasmic sperm injection (ICSI) treatment were observed in DNAH3-deficient patients and Dnah3 KO mice. This study revealed DNAH3 as a novel pathogenic gene of asthenoteratozoospermia, and the findings provide valuable suggestions for the clinical diagnosis and treatment of male infertility.

Results

Identification of biallelic pathogenic variants of DNAH3 in four unrelated infertile men

In the present study, we employed whole-exome sequencing (WES) to identify potential candidate variants associated with primary asthenoteratozoospermia. After comprehensive filtering and screening, we identified 98, 101, 67 and 91 candidate variants for Patient 1, Patient 2, Patient 3 and Patient 4, respectively (Table S1). To refine these candidate variants, we excluded those whose corresponding genes were not expressed in the human or mouse testis, were associated with diseases unrelated to male infertility, or were monoallelic variants. Ultimately, only bi-allelic variants in DNAH3 (NG_052617.1, NM_017539.2, NP_060009.1) remained, suggesting as the pathogenic variants responsible for the infertility of the patients : a compound heterozygous mutation of c.3590C>T (p.Pro1197Leu) and c.3590C>G (p.Pro1197Arg) in Patient 1, a homozygous missense mutation of c.4837G>T (p.Ala1613Ser) in Patient 2, a compound heterozygous mutation of c.5587del (p.Leu1863*) and c.10355C>T (p.Ser3452Leu) in Patient 3 and a compound heterozygous mutation of c.2314C>T (p.Arg772Trp) and c.4045G>A (p.Asp1349Asn) in Patient 4 (Figure 1A). Importantly, routine semen analysis revealed that all patients showed extremely reduced sperm motility and a high proportion of sperm tail defects (Table 1). These variants either were not recorded or had an extremely low frequency in East Asian population in multiple public population databases, including the ExAC browser, GnomAD and the 1000 Genomes Project, and were predicted to be potentially deleterious by SIFT (https://sift.bii.a-star.edu.sg/), PolyPhen-2 (http://genetics.bwh.harvard.edu/pph2/), MutationTaster (https://www.mutationtaster.org/), and CADD (https://cadd.gs.washington.edu/) (Table 2) (30–33). Next, Sanger sequencing confirmed these variants in the probands, and their fertile parents carried the heterozygous variants (Figure 1A). Moreover, the variant sites are localized in several domains of the DNAH3 protein and are highly conserved across species (Figure 1B).

Figure 1.

Table 1.

Table 2.

Strikingly, immunofluorescence staining revealed that DNAH3 was exclusively resided in the tail and concentrated in the midpiece of control sperm. However, the fluorescence signal of DNAH3 was hardly detected in the patients’ spermatozoa (Figure 1C). Additionally, subsequent western blotting analysis yielded consistent results with immunofluorescence staining, indicating that these variants led to disrupted expression of DNAH3 (Figure 1D). These results suggested that biallelic variants in DNAH3 disrupted DNAH3 expression and might be responsible for the infertility of the four patients.

Asthenoteratozoospermia phenotype is observed in patients with DNAH3 variants

We next investigated the aberrant sperm morphology of the patients using Papanicolaou staining and SEM analysis. Notably, the tails of sperm from the patients exhibited a typical phenotype associated with MMAF, including coiled, short, bent, irregular, and/or absent flagella (Figure 2A and B, Figure S1A). In addition, a fraction of defects in the sperm head were also present in the patients’ sperm (Figure 2B).

Figure 2.

TEM was employed to determine the ultrastructure of the sperm from the patients. Compared to the integrated and well-organized “9L+L2” axonemal arrangement of the sperm flagella from the normal control, spermatozoa from the patients showed absent or disordered CPs, MTDs, and outer dense fibers (ODFs) in different regions of the flagella (Figure 3A, Figure S1B). Interestingly, the IDAs of sperm flagella of the patients were hardly captured compared to the control (Figure 3A). Additionally, in the midpiece of sperm flagella of the patients, dissolved mitochondrial material was also observed evidently under TEM (Figure 3A). We next conducted immunofluorescence staining to label the mitochondria of patients’ sperm with TOM20, a subunit of the mitochondrial import receptor. Remarkably, in contrast to the robust TOM20 signals observed in the normal control, the TOM20 signals in the sperm from the patients were considerably diminished, indicating a disrupted mitochondrial function (Figure 3B). Together, these data suggested that DNAH3 may function in sperm flagellar development, and loss-of-function variants were associated with MMAF in humans.

Figure 3.

DNAH3 is exclusively expressed in the sperm flagella of humans and mice

To further understand the function of DNAH3 in male reproduction, we explored the expression pattern of DNAH3 in humans and mice. qPCR results revealed that Dnah3 was predominantly expressed in the mouse testis (Figure S2A). Moreover, when observing the expression of Dnah3 in testes from mice at different postnatal days, we found that Dnah3 expression was significantly elevated beginning on postnatal Day 22, peaked at postnatal Day 30, and maintained a stable expression level thereafter (Figure S2B). In addition, germ cells at different stages were isolated from the testes of humans and mice and were stained with anti-DNAH3 antibody. The results showed that DNAH3 was expressed in the cytoplasm of spermatocytes and spermatogonia and then obviously in the flagellum of early and late spermatids (Figure S3A and B). These expression data suggest that DNAH3 may play an important role in sperm flagellar development during spermatogenesis in humans and mice.

Deletion of Dnah3 causes male infertility in mice

Considering the absent expression of DNAH3 in the patient sperm, we generated Dnah3 KO mice using CRISPRLCas9 technology to further confirm the essential role of DNAH3 in spermatogenesis (Figure S4A). PCR, qPCR, and immunofluorescence staining were used to confirm that Dnah3 was null in KO mice (Figure S4B-E). The Dnah3 KO mice survived without any evident abnormalities in development and behavior. H&E staining further revealed that there were no histological differences in the lung, brain, eye, or oviduct between wild-type (WT) and Dnah3 KO mice (Figure S5A). In addition, no obvious abnormalities in ciliary development were observed in these organs in KO mice compared to WT mice (Figure S5B). The Dnah3 KO female mice were fertile with normal oocyte development (Figure S6A). However, the Dnah3 KO male mice were completely infertile (Figure 4A). We next examined the testis and epididymis of Dnah3 KO male mice to elucidate the etiology of infertility. There was no detectable difference in the testis/body weight ratio of Dnah3 KO mice when compared to WT mice (Figure S6B).

Figure 4.

Moreover, subsequent computer-assisted sperm analysis (CASA) also showed that sperm isolated from the cauda epididymis were slightly decreased, and nearly all sperm were completely immobile (Table 3, Movie S1 and Movie S2). Papanicolaou staining and SEM analysis revealed morphological defects in partial spermatozoa from Dnah3 KO mice, including coiled, bent, and irregular flagella, as well as aberrant heads and acephalic spermatozoa (Figure S7A and B).

Table 3.

TEM was further utilized to evaluate the sperm flagellar ultrastructure of Dnah3 KO mice. There were no obvious abnormalities of “9 + 2” microtube arrangement in most sperm from the Dnah3 KO mice when compared to WT mice (Figure 4B, Figure S7C). However, in contrast to the clear display of an IDA and an ODA on the A-tube of each microtubule doublet in the sperm flagella of WT mice, the sperm flagella of Dnah3 KO mice exhibited an absence of almost all the IDAs (Figure 4B, Figure S7D). In addition, the disrupted mitochondria of spermatozoa from Dnah3 KO mice were also observed under TEM, as manifested by the dilated intermembrane spaces and dissolved mitochondrial material (Figure 4B and C, Figure S7E). We next performed immunofluorescence staining to label SLC25A4, which is responsible for the exchange of ATP and ADP across the mitochondrial inner membrane. Strikingly, compared to the bright fluorescence signals in the midpiece of WT sperm, the signals Dnah3 KO were significantly diminished (Figure 4D), indicating impaired mitochondrial function. Collectively, DNAH3 is essential for spermatogenesis, and its deficiency seriously damages the sperm motility and IDAs in both humans and mice.

DNAH3 deficiency impairs IDAs related to the reduction of IDA-associated proteins

Considering the disrupted IDAs revealed by TEM analysis in both our patients and Dnah3 KO mice, we speculated whether the defective IDAs were attributed to the decreased expression of the key IDA-associated proteins. The immunofluorescence data showed that DNAH1/DNAH6 and DNALI1, corresponding to the heavy and light intermediate chains of the IDAs (16), respectively, were almost invisible along the sperm flagella of the patients when compared to control (Figure 5A-C). Consistent results were obtained in our subsequent western blotting analysis of sperm lysates from the patients (Figure 5D-F), indicating that DNAH3 may manipulate the assembly of IDA through regulating the expression of IDA-associated proteins. In contrast, DNAH8/DNAH17 and DNAI1, corresponding to the heavy and intermediate chains of ODAs (25), were readily detectable in the patients’ sperm flagella and were comparable to the control (Figure S8A-C), suggesting that DNAH3 may not regulate the expression of ODA-associated proteins. We also performed immunofluorescence staining and western blotting analysis of DNAH1, DNAH6, DNALI1, DNAH8, DNAH17 and DNAI1 on sperm from Dnah3 KO mice, and the results observed were consistent with those of the patients (Figure 6A-F, Figure S9A-C). These findings suggested that other IDA-associated proteins might be downstream effectors of DNAH3, which needs more future research.

Figure 5.

Figure 6.

ICSI treatment of humans with DNAH3 variants and Dnah3 KO mice

ICSI treatment has been reported to be effective in asthenoteratozoospermia-associated infertility (34, 35). ICSI cycles were attempted for the patients after written informed consent was obtained. The female partners all had normal basal hormone levels and underwent a long gonadotrophin-releasing hormone agonist protocol (Table 4). The wife of Patient 1 underwent one ICSI attempt. A total of 21 metaphase II (MII) oocytes were retrieved and microinjected, of which 17 oocytes were successfully fertilized (17/21, 80.95%) and cleaved (17/17, 100%). Thirteen Day 3 (D3) embryos were formed, six of which developed into blastocysts (8/13, 61.54%) after standard embryo culture. Two blastocysts were transferred, one of which was implanted. She eventually achieved clinical pregnancy, and the pregnancy is ongoing (Table 4). The partner of Patient 2 underwent two ICSI attempts. In her first ICSI attempt, six MII oocytes were retrieved, of which three were fertilized (3/6, 50%) and cleaved (3/3, 100%). After standard embryo culture, two D3 embryos were formed and transferred. However, this ICSI failed because no embryos were implanted. In her second ICSI attempt, all five MII oocytes were fertilized and cleaved (5/5, 100%). Five D3 embryos were obtained, of which two were transferred, but no embryos were implanted. The remaining three D3 embryos were cultured continuously, and two available blastocysts were formed and kept to be transferred in the future (Table 4). The partner of Patient 3 underwent one ICSI attempt. Of the 20 MII oocytes retrieved, 19 oocytes were fertilized (19/20, 95.0%) and cleaved (19/19, 100%). Fifteen D3 embryos were obtained, and 10 developed into available blastocysts (10/15, 66.7%). One blastocyst was transferred and implanted. She achieved clinical pregnancy, and the pregnancy is ongoing (Table 4). The wife of Patient 4 underwent four failed ICSI attempts. In her first two ICSI attempts, 13 and 12 MII oocytes were retrieved, of which five (5/12, 41.67%) and six (6/13, 46.15%), respectively, were fertilized and cleaved. Two available D3 embryos were obtained and transferred in both ICSI attempts, but no embryos were implanted. In her third ICSI attempt, of the eight MII oocytes retrieved, four (4/8, 50%) were fertilized and cleaved (4/4, 100%). However, no available D3 embryos were acquired. In her last ICSI attempt, seven MII oocytes were retrieved, of which three were fertilized (3/7, 42.68%) and two were cleaved (2/3, 66.7%), but no available D3 embryos were formed (Table 4). The vivid embryonic development of the partner of Patient 1 and Patient 3 after ICSI treatment was shown in Figure 7A.

Figure 7.

Table 4.

We also carried out ICSI treatment on Dnah3 KO male mice. Strikingly, favorable outcomes of ICSI were obtained in Dnah3 KO male mice. After injection of spermatozoa from Dnah3 KO male mice, pronuclei were observed in most embryos in both the KO and WT groups, indicating a normal fertilization rate (Figure 7B). There was no difference in the percentage of 2-cell and blastocyst-stage embryos between the KO and WT groups (Figure 7B). Collectively, we observed successful ICSI outcomes in two out of four DNAH3-deficient patients and Dnah3 KO male mice and therefore suggested ICSI as an optional treatment for infertile men harboring biallelic pathogenic variants in DNAH3, and the additional female risk factors for infertility should not be excluded in the failed patients.

Discussion

In the present study, we identified pathogenic variants in DNAH3 in unrelated infertile men with asthenoteratozoospermia. These variations resulted in the almost absence of DNAH3 and sharply decreased the expression of other IDA-associated proteins, including DNAH1, DNAH6 and DNALI1. Combined with similar findings in Dnah3 KO mice, we demonstrated that DNAH3 is fundamental for male fertility. Moreover, we suggest that ICSI might be a favorable treatment for male infertility caused by DNAH3 deficiency. Our findings identify a function for DNAH3 in male reproduction in humans and mice and may provide a new view on the clinical practice of male infertility.

Recently, Meng et al. reported DNAH3 mutations in asthenoteratozoospermia affected patients, revealing multiple morphological defects in sperm tail (36). Moreover, ultrastructural abnormalities of the flagellar axoneme in the patients were evident in these patients, characterized by a disrupted ’9+2’ arrangement and the notable absence of IDAs (36). Additionally, they generated Dnah3 KO mice, which were infertile and exhibited moderate morphological abnormalities (36). While the ’9+2’ microtubule arrangement in the flagella of their Dnah3 KO mice remained intact, the IDAs on the microtubules were partially absent (36). In our study, we observed similar phenotypic differences between DNAH3-deficient patients and Dnah3 KO mice. Both studies suggest that DNAH3 may play crucial yet distinct roles in human and mouse male reproduction.

However, there are notable differences between the two studies. Firstly, the phenotypes of Dnah3 KO mice showed slight differences. Meng et al. generated two Dnah3 KO mouse models (KO1 and KO2), and both of which exhibited significantly higher sperm motility and progressive motility than in our study (36), where nearly all sperm were completely immobile. Furthermore, their Dnah3 KO2 mice displayed motility comparable to WT mice and retained partial fertility (36). We speculate that these differences may be attributed to variations in mouse genetic background or the presence of a truncated DNAH3 protein resulting from specific knockout strategies. Secondly, we conducted additional research and uncovered novel findings. We revealed that male infertility caused by DNAH3 mutations follows an autosomal recessive inheritance pattern, as confirmed through Sanger sequencing of the patients’ families. We also discovered the dynamic expression and localization of DNAH3 during spermatogenesis in humans and mice through immunofluorescent staining. Initially, DNAH3 was expressed in the cytoplasm of spermatogonia and spermatocytes, and then it clearly transferred into the flagellum of early and late spermatids. We further found that DNAH3 deficiency had no impact on ciliary development in the oviduct or on oogenesis in mice, resulting in normal female fertility. Moreover, in the absence of DNAH3 in both humans and mice, the expression of IDA-associated proteins, including DNAH1, DNAH6 and DNALI1, was decreased, while the expression of ODA-associated proteins remained unaffected, indicating that DNAH3 is involved in sperm axonemal development, specifically through its role in the assembly of IDAs. Collectively, our study corroborates the findings of Meng et al., and provides additional unique insights, comprehensively elucidating the critical role of DNAH3 in human and mouse spermatogenesis.

Primary ciliary dyskinesia (PCD, MIM: 244400) is a genetic disorder affecting at least one in 7554 individuals (37). The most common symptoms of PCD are recurrent infections in airways due to malfunction of the motile cilia that are responsible for mucus clearance (38). It has been suggested that male infertility associated with sperm defects is highly prevalent (up to 75%) among individuals with PCD (39). Axonemal defects caused by variants within DNAH family members, including DNAH5, DNAH6, DNAH7, DNAH9 and DNAH11, are causative factors for PCD (40–42). Moreover, deficiency in these PCD-causing DNAHs has also been associated with male infertility (9, 14, 20, 21, 43–45). Additionally, other DNAHs, such as DNAH1, DNAH2, DNAH8, DNAH10, DNAH12 and DNAH17, are suggested to be pathogenic genes of isolated male infertility (18, 19, 22, 24, 25). These phenotype–genotype correlations may be attributed to the fact that ciliary and flagellar axonemes have cell type-specific or cell type-enriched DNAHs (46). DNAH3 resides in the IDA and is expressed in testis and ciliary tissues, including the lung, brain, eye, and oviduct. However, despite its presence in these tissues, the relationship between deficient DNAH3 and disease is unclear to date. Intriguingly, in our study, none of the patients with DNAH3 deficiency reported experiencing any of the principal symptoms associated with PCD. Additionally, our Dnah3 KO mice exhibited normal ciliary development in the lung, brain, eye, and oviduct. Similarly, Meng et al. did not mention any PCD symptoms in their DNAH3-deficient patients, and their Dnah3 KO mice also demonstrated normal ciliary morphology in the trachea and brain (36). These combined observations suggest that DNAH3 may play a more important role in sperm flagellar development than in other motile cilia functions. Given that DNAH3 is expressed in ciliary tissues, its role in these tissues remains intriguing and could be elucidated through sequencing of larger cohorts of individuals with PCD.

ICSI has been an efficient treatment for male infertility (47, 48). However, the outcomes of ICSI for male infertility caused by variants in different DNAH genes are variable. It has been demonstrated that infertile males with variants in DNAH1, DNAH2, DNAH7, and DNAH8 have a favorable prognosis (22, 49–53), while patients with variants in DNAH17 have poor outcomes after ICSI treatment (25, 54). Meanwhile, the ICSI outcomes in male infertility caused by DNAH6 variants may depend on the specific mutation or be controversial (20, 55, 56). The patients with DNAH3 mutations in our study experienced different clinical outcomes of ICSI treatment. The partners of Patient 1 and Patient 3 achieved clinical pregnancy. The wives of Patient 2 and Patient 4 obtained favorable fertilization and cleavage rates but experienced no clinical pregnancy due to the nonimplantation of the transferred embryos. Remarkably, despite the diverse variants within DNAH3 observed in the four patients, all variants led to a complete absence of DNAH3 expression. Additionally, we did not identify any pathogenic variants that associated with fertilization failure and early embryonic development in the two patients with failed ICSI outcomes. Therefore, these different ICSI outcomes might be attributed to additional unexplained factors from the female partners. Importantly, in the study from Meng et al., one patient carrying DNAH3 variants received ICSI treatment, and the partner obtained clinical pregnancy (36). Combined with the successful ICSI outcomes observed in Dnah3 KO mice, we suggest ICSI as an optimized treatment for infertile men carrying variants in DNAH3. More cases are needed to precisely estimate the prevalence of DNAH3 mutations and determine a prognosis for ICSI treatments.

In conclusion, our study revealed an unexplored role of DNAH3 in male reproduction in humans and mice, suggesting DNAH3 as a novel causative gene for human asthenoteratozoospermia. Moreover, ICSI is as an optimized treatment for infertile men with DNAH3 variants. This study expands our knowledge of the relationship between DNAH proteins and disease, facilitating genetic counseling and clinical treatment of male infertility in the future.

Methods

Human subjects

Four unrelated Han Chinese infertile men and their family members were recruited from West China Second University Hospital of Sichuan University and Women and Children’s Hospital of Chongqing Medical University. All patients exhibited a normal karyotype (46 XY) without deletion of the azoospermia factor (AZF) region in the Y-chromosome. All of the participants were provided informed consent, and the study was approved by the ethics committee of West China Second University Hospital and The First Affiliated Hospital of Chongqing Medical University.

Genetic analysis

Peripheral blood samples were obtained from the subjects to extract genomic DNA using a DNA purification kit (TIANGEN, DP304). For WES, 1 μg of genomic DNA was utilized for exon capture using the Agilent SureSelect Human All Exon V6 Kit and sequenced on the Illumina HiSeq X system (150-bp read length). The quality of WES, including clean reads, sequencing depth, sequencing coverage, and mapping quality are listed in Table S1. The variants identified through WES were annotated and filtered using Exomiser. Next, the variants were screened to obtain candidate variants based on the following criteria: (1) the allele frequency in the East Asian population was less than 1% in any database, including the ExAC Browser, gnomAD, and the 1000 Genomes Project; (2) the variants affected coding exons or canonical splice sites; (3) the variants were predicted to be possibly pathogenic or damaging. The remain genes were then analyzed using the Human Protein Atlas (HPA) database (https://www.proteinatlas.org/) and Mouse Genome Informatics (MGI) database (https://informatics.jax.org/) to access their expression in human and mouse testis. Additionally, OMIM database (https://www.omim.org/) and relevant literature were used to understand their relationship with human infertility. Given the assumption of a recessive inheritance pattern, monoallelic variants were excluded from consideration. The remained candidate pathogenic variants were verified by Sanger sequencing on DNA from the patients’ families. The primer pairs used for PCR amplification are listed in Table S2.

Electron microscopy

For scanning electron microscopy (SEM), sperm samples were fixed in glutaraldehyde (2.5%, w/v) and dehydrated using an ethanol gradient (30, 50, 75, 85, 95, and 100% ethanol). The samples were dried using a CO₂ critical-point dryer (Eiko HCP-2, Hitachi) and observed under SEM (S-3400, Hitachi).

For transmission electron microscopy (TEM), sperm samples were fixed in glutaraldehyde (3%, w/v) and osmium tetroxide (1%, w/v) and dehydrated with an ethanol gradient. The samples were embedded in Epon 812. Ultrathin sections were stained with uranyl acetate and lead citrate and analyzed under TEM (Tecnai G2 F20).

STA-PUT velocity sedimentation

Single testicular cells from obstructive azoospermia and 8-week-old C57BL male mice were obtained using the STA-PUT velocity sedimentation method as described previously (57, 58). In brief, total spermatogenic cells were harvested by digesting seminiferous tubules with collagenase (Invitrogen, 17100017), trypsin (Sigma, T4799) and DNase (Promega, M6101) for 15 min each at 37 °C. Cells were diluted in bovine serum albumin (BSA, 3%, w/v) and filtered through an 80 mm mesh to remove fragments. Then, the cells were resuspended in BSA (3%, w/v) and loaded into an STA-PUT velocity sedimentation cell separator (ProScience) to obtain germ cells at different stages.

RNA isolation and quantitative PCR (qPCR)

Total RNA of mouse tissues was extracted using TRIzol reagent (Invitrogen,15596026,) and reverse-transcribed using the 1st Strand cDNA Synthesis Kit (Yeasen, HB210629) according to the manufacturer’s instructions. qPCR was carried out on an iCycler RTLPCR Detection System (Bio-Rad Laboratories) using SYBR Green qPCR Master Mix (Bimake, B21202). Primer sequences are listed in Table S2.

Immunofluorescence staining

Sperm samples were fixed in paraformaldehyde (4%, w/v), permeabilized with Triton X-100 (0.3% v/v) and blocked with BSA (3%, w/v) at room temperature. Samples were incubated with primary antibodies, including DNAH1 (Cusabio, CSB-PA878961LA01HU, 1:100), DNAH3 (Cusabio, CSB-PA823461LA01HU, 1:100), DNAH6 (Proteintech, 18080-1-AP, RRID: AB_2878493, 1:50), DNAH8 (Atlas, HPA028447, RRID: AB_10599600, 1:200), DNAH17 (Proteintech, 24488-1-AP, RRID: AB_2879568, 1:50), DNAI1 (Proteintech, 12756-1-AP, RRID: AB_10643244, 1:50), DNALI1 (Proteintech, 17601-1-AP, RRID: AB_2095372, 1:50), TOM20 (Proteintech, 11802-1-AP, RRID: AB_2207530, 1:50), SLC25A4 (Signalway, 32484, RRID: AB_2941094, 1:100), lectin PNA (Invitrogen, L-32460, 1:50) and alpha tubulin (Abcam, ab7291, RRID: AB_2241126, 1:500), overnight at 4 °C. The next day, the samples were washed and incubated with the secondary antibody Alexa Fluor 488 (Invitrogen, A11008, RRID: AB_143165, 1:1000) or Alexa Fluor 594 (Invitrogen, A11005, RRID: AB_141372, 1:1000), and the nuclei were labeled with 4′,6-diamidino-2-phenylindole (DAPI, SigmaLAldrich, D9542). Image capture was performed by a laser scanning confocal microscope (Olympus, FV3000).

For staining of mouse tissues, samples were first fixed in paraformaldehyde (4%, w/v) and dehydrated with an ethanol gradient. Then, the samples were embedded in paraffin and sliced into 5-μm sections. After deparaffinization and rehydration, sections were processed with 3% hydrogen peroxide and incubated in sodium citrate for antigen repair. Subsequently, sections were blocked with goat serum and incubated with primary antibodies against DNAH3 (Cusabio, CSB-PA823461LA01HU, 1:100) or ac-Tubulin (Abcam, ab24610, RRID: AB_448182, 1:500) at 4 °C overnight. The next day, the sections were incubated with the secondary antibody Alexa Fluor 488, followed by labeling the nuclei with DAPI. Image capture was performed using a fluorescence microscope (Zeiss, Ax10).

Western Blotting

Sperm samples were lysed in RIPA buffer (Beyotime, P0013B) to extract the total protein. For analysis of DNALI1, the protein samples were mixed with SDS loading buffer (P0015, Beyotime, China), boiled at 95 °C for 5 minutes, and separated by 12.5% SDS-PAGE. For analysis of DNAH1, DNAH3, DNAH6, the protein samples were mixed with NuPAGE™ LDS sample buffer (Invitrogen, NP0007), denatured at 70°C for 10 minutes, and separated by 3–8% NuPAGE™ Tris-Acetate gels (EA0375BOX, Invitrogen). Then the resolved proteins were transferred to 0.45 μm PVDF membranes (Merck Millipore, IPVH00010). The membranes were blocked, incubated with primary antibodies, including DNALI1 (Proteintech, 17601-1-AP, RRID: AB_2095372, 1:150), DNAH3 (Cusabio, CSB-PA823461LA01HU, 1:200), DNAH6 (Proteintech, 18080-1-AP, RRID: AB_2878493, 1:150) and alpha tubulin (Proteintech, 11224-1-AP, RRID: AB_2210206, 1:1000) at 4 L overnight. The following day, membranes were washed and incubated with HRP-conjugated secondary antibody (Proteintech, SA00001-2, RRID: AB_2722564, 1:5000). Protein bands were visualized using enhanced chemiluminescence reagents (Millipore, WBKLS0500).

Histology hematoxylin-eosin (H&E) staining

Tissue samples from mice were fixed with 4% paraformaldehyde (w/v) overnight. Following dehydration by ethanol, the samples were embedded in paraffin and sliced into 5-μm sections. The sections were stained with hematoxylin and eosin and observed under a microscope (Zeiss, Axio Imager 2).

Generation of the Dnah3 KO mouse model

Animal experiments in this study were approved by the Experimental Animal Management and Ethics Committee of West China Second University Hospital, Sichuan University, and complied with the Animal Care and Use Committee of Sichuan University. A Dnah3 knockout mouse model was generated by the CRISPRLCas9 system. Briefly, Cas9 and signal-guide RNAs (5’-GTATCAAGTGGATGTAAACC-3’) were transcribed using T7 RNA polymerase in vitro and comicroinjected into the cytoplasm of single-cell C57BL/6J mouse embryos to generate frameshift mutations by nonhomologous recombination through introduction of a 1 bp insertion in exon 13. Then, the embryos were cultured and transferred into the oviducts of pseudopregnant female mice at 0.5 days post-coitum. A mutation of Dnah3 in the founder mouse and their offspring was confirmed using PCR and Sanger sequencing. The primers used for the generation of animal models are listed in Table S2.

Intracytoplasmic sperm injection (ICSI)

ICSI was carried out using standard techniques. In brief, one-month-old female KM mice were injected with 5 IU of equine chorionic gonadotropin (eCG) (ProSpec, HOR-272) to induce superovulation. Metaphase II-arrested (MII) oocytes were acquired through another injection of 5 IU human chorionic gonadotropin after 48 hours. MII oocytes were incubated with Chatot-Ziomek-Bavister medium (Easycheck, M2750) at 37.5 °C and 5% CO₂ until use. Mouse cauda epididymal spermatozoa were incubated in human tubal fluid (HTF) medium (Easycheck, M1150) and then frozen and thawed repeatedly to remove sperm tails. For ICSI, a single sperm head was microinjected into an MII oocyte by using a NIKON inverted microscope and a Piezo (PrimeTech, Osaka, Japan) in Whitten’s-HEPES medium containing 0.01% polyvinyl alcohol (Gibco,12360-038) and cytochalasin B (3.5 g/ml; SigmaLAldrich, C-6762). The successfully injected oocytes were transferred into G1-Plus medium (Vitrolife, 10132) and incubated at 37.5 °C and 5% CO₂. The animal experiments were approved by the Experimental Animal Management and Ethics Committee of West China Second University Hospital, Sichuan University.

Statistical analysis

Prism (version 8.4.0, GraphPad, Boston, MA, USA) and SPSS (version 18.0, IBM Corporation, Armonk, NY, USA) were used to perform statistical analyses. All data are presented as the means ± SEMs. Data from two groups were compared using an unpaired, parametric, two-sided Student’s t test, and a p value less than 0.05 was considered statistically significant.

Ethics approval

This study was approved by Ethical Review Board of West China Second University Hospital, Sichuan University. Informed consent was obtained from each participate in this study before taking part.

Data availability

The published article includes all datasets generated or analyzed during this study. The whole exome-sequencing data were deposited in the National Genomics Data Center (NGDC) (https://ngdc.cncb.ac.cn/, accession number: HRA007467).

Competing interests

The authors declare that they have no conflict of interest.

Author contributions

Xiang Wang: Data curation; formal analysis; investigation; methodology; writing – original draft. Gan Shen: Data curation; formal analysis; Yihong Yang: Resources; investigation; methodology. Chuan Jiang: Data curation; formal analysis. Tiechao Ruan: Formal analysis; methodology. Xue Yang: Validation; investigation. Liangchai Zhuo: Investigation; Yingteng Zhang: Investigation, methodology. Yangdi Ou: Investigation. Xinya zhao: Investigation. Shunhua Long: Methodology. Xiangrong Tang: Investigation. Tingting Lin: Conceptualization; funding acquisition; project administration. Ying Shen: Conceptualization; supervision; project administration; writing – review and editing.

Acknowledgements

The authors thank the patient and his family members for their voluntary participation. We are grateful to Guiping Yuan from Analytical and Testing Center of Sichuan University and Yan Liang from Research Core Facility of West China Hospital, Sichuan University for their help with TEM images and preparing histology slides. This work was supported by National Natural Science Foundation of China (82301807).

Supporting information

Figure S1.

Figure S2.

Figure S3.

Figure S4.

Figure S5.

Figure S6.

Figure S7.

Figure S8.

Figure S9.