Novel biallelic DNHD1 variants associated with male infertility with severe MMAF phenotype

Abstract

Multiple morphological abnormalities of the sperm flagella (MMAF), characterized by severe morphological sperm defects, such as absent, short, irregular caliber, and coiled flagella with extreme asthenoteratozoospermia, are the most prevalent cause of human male infertility. Previous studies have identified several genes linked to MMAF; however, the increasing incidence of infertility indicates that most affected individuals remain undiagnosed, prompting further investigation to uncover novel mutations and genes. Whole-exome sequencing (WES) was conducted on a consanguineous infertile family from Pakistan to investigate the potential monogenic inheritance pattern in individuals affected by asthenoteratozoospermia. WES identified novel homozygous variants (c.A4457G; p.K1486R, and c.C10624T; p.R3542*) in dynein heavy chain domain 1 (DNHD1) in the proband and his affected brother. Semen analysis revealed a low progressive motility and severe MMAF in both siblings. Hematoxylin and eosin staining, immunofluorescence, and transmission electron microscopy unveiled an abnormal axoneme structure characterized by missing central pairs, disorganized microtubule duplets, and severe mitochondrial sheath defects, which led to the low sperm progressive motility and infertility in the affected siblings. This study identified a novel biallelic nonsense variant in DNHD1 that caused MMAF in a Pakistani population, endorsing previous findings and expanding the spectrum of potential DNHD1 variants in the pathogenesis of asthenoteratozoospermia.

INTRODUCTION

Asthenoteratozoospermia, which refers to sperm with morphological abnormalities and reduced motility, is a leading cause of male infertility, accounting for approximately 19% of cases.1,2 The condition is also evident in individuals with primary ciliary dyskinesia (PCD), a multisystem disorder resulting from the malfunctioning of motile flagella and cilia that ultimately results in chronic bronchiectasis, heterotaxis, or rhinosinusitis.3 However, some patients present with similar sperm abnormalities without showing any other symptoms associated with PCD, a phenomenon widely recognized as multiple morphological abnormalities of the sperm flagella (MMAF).4,5,6,7,8,9 MMAF have been characterized as a condition of severe morphological defects in spermatozoa, including absent, shortened, irregular caliber, and coiled flagella, which characterize extreme asthenozoospermia. At the ultrastructural level, profound disorganization of axonemal and periaxonemal structures is evident, with deficiencies in dynein arms and frequent loss of peripheral microtubule doublets or central pair (CP).10

Genetic factors are the major causes of MMAF, with dynein axonemal heavy chain 1 (DNAH1) identified in 2014 as the first gene mutation linked to male sterility and the MMAF phenotype in humans.11,12 Since 2014, advances in next-generation sequencing, particularly the popularity and decreased cost of whole-exome sequencing (WES), and the increasing number of phenotypically well-characterized cohorts from various ethnic backgrounds have led to a significant increase in the identification of numerous pathogenic genes associated with MMAF. However, identifying monogenic variants in WES data is challenging due to the large number of variant differences in each individual compared to the reference genome, many of which are of uncertain significance.13 Alternatively, for investigating genetically heterogeneous conditions, such as infertility, analyzing consanguineous kindreds has proven to be an effective approach in which WES of small families has been successfully used to pinpoint single deleterious genes in conditions such as sperm fertilization defects and nonobstructive azoospermia.14,15,16

Dynein heavy chain domain 1 (DNHD1), also known as coiled-coil domain-containing 35 (CCDC35), is an evolutionarily conserved protein in primates and non-primates that consists of 4750 amino acids and is characterized by five coiled-coil (CC) domains. DNHD1 is expressed predominantly in the sperm flagellum. The CC domain represents a specialized conserved protein structural domain that features one or more stretches of alpha-helical peptides that form a supercoil by twisting around each other.17 Dynein contributes significantly to ciliary and flagellar motility and facilitates intercellular movement and cytoskeletal remodeling in the cytoplasm, where it is a main constituent of microtubule-associated motor protein complexes fueled by adenosine triphosphate (ATP) to propagate these cellular movements.18 Two previous independent studies19,20 correlated mutations in DNHD1 to impaired sperm motility and found that its deficiency led to asthenoteratozoospermia with periaxoneme and axoneme structural defects. One study validated its findings using mouse models, which successfully recapitulated the phenotype observed in eight unrelated patients in a Chinese population.19 The second study identified DNHD1 mutations in three unrelated patients among a cohort of 167 men diagnosed with MMAF in a French population.20

In this study, WES was conducted on a consanguineous infertile family (family 7) from Pakistan. We identified a novel biallelic DNHD1 variant in the proband, and the same homozygous variant was identified in his infertile brother. Both brothers had biallelic DNHD1 mutations and exhibited reduced sperm progressive motility with severe MMAF. Our study identified a novel biallelic DNHD1 variant in a Pakistani population, endorsing the findings that suggest that DNHD1 is a potential biomarker of asthenoteratozoospermia.

PARTICIPANTS AND METHODS

Participants and sample collection

In this study, we recruited a consanguineous infertile family (family 7). The proband was initially identified through a preclinical data collection conducted in a local community of Swat Pakistan and subsequently contacted. In accordance with the ethical principles outlined in the Helsinki Declaration, the patients and controls were fully informed about the study and carefully completed written consent forms prior to participation in the study. Furthermore, a comprehensive data pro forma was completed for each patient, including additional details, such as name, age, height, hobbies, profession, medical history, and pedigrees, which were meticulously drawn as part of the comprehensive data collection process. A semen analysis was promptly performed, that is, within 60 min of sample collection, in the nearest local laboratory, adhering to the guidelines outlined in the 5^th edition of the World Health Organization (WHO) Manual.21 Blood samples were obtained from patients and available controls, stored in 3-ml ethylenediaminetetraacetic acid (EDTA) tubes, and preserved at 4°C for subsequent analysis.

Ethical statement

This study, which involved human participants, was approved by the Biomedical Research Ethics Committee of The First Affiliated Hospital of the University of Science and Technology of China (Hefei, China; Approval No. 2019-N[H]-031) and the Institutional Review Board of the Institute of Biotechnology and Genetic Engineering (IBGE) at The University of Agriculture (Peshawar, Pakistan; Approval No. IBGE/988). Before the commencement of this study, an informed consent form was meticulously filled out for each participant. This research is part of the collaborative project titled “Functional studies of pathogenic mutations causing sperm deficiency and male infertility in Pakistani families”, conducted by Bao Lab, Division of Life Sciences and Medicine, The First Affiliated Hospital of the University of Science and Technology of China, and the Discipline of Physiology, Faculty of Animal Husbandry and Veterinary Sciences, The University of Agriculture.

Semen parameter analysis

A semen parameter analysis was performed in a local laboratory as part of routine biological examinations, strictly adhering to the WHO guidelines to ensure a standardized assessment.22 The sample was collected from each participant through masturbation after 2–7 days of sexual abstinence, followed by 30 min of liquefaction at 37°C before evaluation. Standard analyses, including sperm concentration, semen volume, and sperm motility, were performed during routine examinations. For the morphological assessment of sperm cells, hematoxylin and eosin (H&E) staining was performed to allow for a meticulous examination of at least 210 spermatozoa for each patient to determine the percentage of morphologically abnormal sperm cells.

Genomic DNA extraction

Genomic DNA from whole blood was extracted with the PureLinkTM Genomic DNA Mini Kit (Catalog #K182001 M/s; Thermo Fisher Scientific, Carlsbad, CA, USA), using the standard protocol prescribed by the manufacturer. The quality of the DNA was determined using agarose gel electrophoresis after assessment with a gel documentation system (Tanon-2500R; Tanon, Shanghai, China), and the quantity of DNA was measured using NanoDrop (NanoPhotometer N50 Touch; Implen, Munich, Germany). The extracted DNA was stored at −20°C for downstream analysis.

WES and bioinformatics analysis

WES was performed on the patients and controls, and the data analysis was performed using the Illumina NovaSeq X Plus Platform Library Prep protocol (San Diego, CA, USA), according to our reported method.23 Initial profiling of the DNA samples was performed using the Agilent SureSelect Exome Capture and IDT xGen Exome Capture techniques. Sequencing libraries were prepared and subsequently sequenced using two platforms, NextSeq 500 and NovaSeq 6000, both manufactured by Illumina. Sequencing was performed using 150-base pair (bp) paired-end reads, achieving an average depth of 100× per sample and generating Fast Quality Score (FASTQ) sequencing reads. The Burrows-Wheeler Aligner (BWA) was then used to align the FASTQ reads to the GRCh38/hg38 human reference genome.24 Duplicate polymerase chain reaction (PCR) products were identified and marked using the Picard Toolkit (http://broadinstitute.github.io/picard/). Subsequently, BAM files were sorted and indexed using SAMtools to ensure efficient data processing and analysis.25 The Genome Analysis Toolkit (GATK4) was used for base quality score recalibration (BQSR) using the ApplyBQSR tool to improve the accuracy of variant calls. Variant calling was performed using HaplotypeCaller, a core component of the GATK pipeline, which generated gVCF files containing both single-nucleotide variants and insertions/deletions (indels) for each sample.26 The gVCF files were annotated using annotation variation (ANNOVAR), focusing on identifying variants associated with diseases in familial cases. Priority was given to variants with potential functional impact on protein sequences, including missense, splice site, frameshift, and nonsense mutations.27 Variants with a minor allele frequency (MAF) >0.01 in publicly available databases, including the 1000 Genomes Project, ExAC, and ESP6500, were excluded. In addition, variants that appeared homozygous in the control group (32 fertile controls) were filtered out to focus on potentially pathogenic mutations (Supplementary Figure 1 ^{(136.3KB, tif)} ).28 Candidate variants identified through this process were further validated using the traditional Sanger sequencing method, adhering to established protocols.

Sanger sequencing

Sanger sequencing was performed to confirm and validate the target variants in both the patients and controls. Primers were designed at least 200 bp away from the target site to ensure specificity and were used for PCR amplification with the 2× Phanta Max Master Mix (Catalog P525), generating amplicons of up to 1000 bp in length. The PCR products were then sequenced using a 3730 XL sequencer (Applied Biosystems, Foster City, CA, USA), following the manufacturer’s instructions. The optimized PCR conditions and the list of primers used in this study are provided in Supplementary Table 1 and 2.

Supplementary Table 1.

List of PCR primers for sanger sequencing

Primer name	Primer sequence	Annealing temperature (℃)	Product size (bp)
F1-DNHD1	5’- GGAGTCAAGCCCAAACACAC-3’	58	492
R1-DNHD1	5’- GCACTGAGAAGGCTGGTCTG-3’
F2-DNHD1	5’- TCCTTTTATGCCGTTCGCCT-3’	58	703
R2-DNHD1	5’- GCTGAGTGTTGGGCAGTTCA-3’

MMAF: multiple morphological abnormalities of the flagella; DNHD1: dynein heavy chain domain 1; F: forward; R: reverse.

Supplementary Table 2.

PCR method optimized for Sanger sequencing

Cycles	Temperature	Time
Denaturation/activation
1	95°C	5 min
Amplification program
35	95°C	30 s
	60°C	30 s
	72°C	30-60 s
Final elongation
1	72°C	5 min

The amplification time for <500 bp and 700 bp has been optimized for 30 s and 40 s, respectively. PCR: polymerase chain reaction; bp: base parir.

H&E staining

H&E staining was performed for the spermatozoa samples from the fertile control subjects and patients, in accordance with our reported procedure, with minor changes.29 Sperm smear slides were prepared and initially fixed in a 5% paraformaldehyde (PFA) solution for 5 min. The slides were then washed three times with 1× phosphate-buffered saline (PBS), followed by sequential fixation in 100%, 90%, and 80% ethanol for 9 min (3 min each). Subsequently, the slides were stained with hematoxylin solution (Catalog G1005-1; ServiceBio, Wuhan, China) for 6 min, rinsed with tap water, and air-dried. A brief dip in 0.5% HCl-ethanol was followed by a quick rinse with ammonium water and then tap water. The slides were then immersed in 95% ethanol for 1 min before staining with an eosin solution (Catalog G1005-2; ServiceBio) for 30 s. After another rinse with tap water, the slides underwent gradual dehydration using 80%, 90%, 95%, and 100% ethanol (2 min each). After dehydration, the slides were treated with xylene–ethanol and then with pure xylene and air-dried. Finally, the slides were mounted with a mounting medium, covered with coverslips, and examined under a microscope (BX53; Olympus, Tokyo, Japan).

Transmission electron microscopy (TEM)

TEM was used to examine the spermatozoa of the patients and controls according to the reported method, with slight modifications.30 The spermatozoa samples were fixed with 2.5% glutaraldehyde overnight at 4°C. After fixation, the samples were rinsed three times with 1× PBS (pH 7.2) and put in a 1% osmium acid solution for 2 h. Afterward, the acid solution was carefully removed, and the samples were rinsed three times with 1× PBS, with each rinse lasting 10 min. The samples were then dehydrated with a series of 30%, 50%, 70%, 80%, and 100% ethanol for 15 min each, followed by 100% acetone dehydration (three times, 20 min each). After dehydration, the spermatozoa samples were immersed in an acetone/Epon 812 solution at 1:1 (v/v) and 1:2 (v/v) for 2 h each. The samples were subsequently soaked in pure Epon 812 twice for 4 h each. The samples were placed in an embedding plate containing Epon 812 and incubated at 37°C for 48 h, followed by 45°C for 12 h. Subsequently, the embedding plate was transferred to an oven and incubated at 60°C for 48 h before being removed for later use. Ultrathin slices of 60–70 nm were cut through an ultra-thin slicer (EM UC7; Leica Microsystem, Wetzlar, Germany) with a diamond slicing knife (Ultra 45°; DiATOME, Nidau, Switzerland), and a copper net was used to capture the slices. The slices were stained with uranyl acetate for 25 min, followed by washing and subsequent staining with lead citrate for 7 min. After staining, the slices were washed and dried, and images were captured using a transmission electron microscope (HT-7800; Hitachi, Tokyo, Japan) for further analysis.

Immunofluorescence analysis (IF)

For IF analysis, smear slides were prepared in accordance with the WHO guidelines22 and our reported procedure, with minor modifications.31 The slides were fixed in 5% PFA for 5 min and washed three times with 1× PBS. Next, the samples were permeabilized with 1× PBS containing 0.2% Triton X-100 for 45 min, followed by blocking with a blocking buffer (3% skim milk in 1× PBS containing 0.1% Triton X-100) for 1 h. After overnight incubation at 4°C with primary antibodies, the slides were incubated with secondary antibodies at 37°C for 1 h. Finally, the slides were mounted using 4’,6-diamidino-2-phenylindole (DAPI) fluorescence stain (1:500 dilution; C0060; Solarbio, Beijing, China) diluted in an antifluorescence quenching mounting solution (P0126; Beyotime, Beijing, China) and covered with coverslips. Images of spermatozoa were captured using NEXCOPE NCF950 equipped with an NIB950 Full Motorized Inverted Microscope and NIS60 Infinite Optical System (F200), utilizing the NOMIS Advanced C Display/Image Processing/Multi-Channel Color Confocal Image software (NCF950; NEXCOPE, Ningbo, China). The primary antibodies used included anti-α-tubulin (1:100 dilution; 11224-1-AP; Proteintech, Des Plaines, IL, USA), anti-β-tubulin (1:200 dilution; 66240-1-lg; Proteintech), anti-DNHD1 (1:100 dilution; PAB15797; Abnova, Taipei, China), anti-sperm-associated antigen 6 (anti-SPAG6; 1:100 dilution; 12642-1-AP; Proteintech), anti-sperm flagellar protein 2 (anti-SPEF2; 1:100 dilution; HPA040343; Sigma, Toronto, Canada), and anti-translocase of outer mitochondrial membrane 20 (anti-TOMM20; 1:100 dilution; AB56783; Abcam, Cambridge, UK). The secondary antibodies used were the Alexa Fluor 647 goat antimouse IgG (1:500 dilution; A0473; Beyotime), Alexa Fluor 647 goat antirabbit (1:200 dilution; A04668; Beyotime), and CoraLite 488 goat antirabbit antibodies (1:100 dilution; SA00013-2; Proteintech). Fluorescein isothiocyanate-labeled peanut agglutinin (FITC-PNA; 1:400 dilution; 16802256; Thermo Fisher Scientific) was used to stain the spermatozoa acrosome.

Reverse transcription quantitative PCR and western blot analyses

For total RNA extraction, semen samples from the patients and control subjects were used to isolate total messenger RNA (mRNA) using the RNAiso Plus reagent (Catalog 9108; Takara, San Jose, CA, USA).32 For complementary DNA (cDNA) synthesis, 1 µg of RNA was reverse-transcribed into cDNA using Hifair V One-Step RT-genomic DNA (gDNA) Digestion SuperMix for quantitative PCR (Catalog 11142; Yeasen, Gaithersburg, MD, USA). The reaction components were gently mixed with the designed PCR primers (Supplementary Table 3) and run on a thermocycle, in accordance with the standard procedure (Supplementary Table 4). For the western blot analysis, human sperm protein extracts were prepared from the sperm samples of the patients and control subjects, stored in TRIzol reagent, and subsequently analyzed with western blot, as previously described.32 The primary antibodies used were rabbit anti-SPAG6 (1:2000 dilution; 12462-1-AP; Proteintech) and rabbit anti-α-tubulin (1:3000 dilution; 11224-1-AP; Proteintech). The secondary antibody used was horseradish peroxidase (HRP)-conjugated goat antirabbit IgG (1:2000 dilution; AS126; Abclonal, Woburn, MA, USA).

Supplementary Table 3.

List of qPCR primers for DNHD1 and TUBA4A mRNA transcripts

Primer name	Primer sequence	Annealing Temperature (℃)	Product size (bp)
qF1-DNHD1	5’- TGTCAAGACCTCTGCCTTGC-3’	58	164
qR1-DNHD1	5’- GGCCACAGAAGATGCTCACA-3’
qF1-TUBA4A	5’- GAGACCTGTCACCCCGACTC-3’	58	174
qR1-TUBA4A	5’- AATGGTCTTGTCACTGGGCA-3’

TUBA4A: Tubulin Alpha 4a; qPCR: quantitative PCR; DNHD1: dynein heavy chain domain 1; bp: base parir.

Supplementary Table 4.

Optimized RT-qPCR method

Cycles	Temperature	Time
Denaturation/activation
1	95°C	5 min
Amplification program
40	95°C	10 s
	60°C	20 s
	72°C	20 s
Final elongation
1	72°C	5 min

RT-qPCR: reverse transcription-quantitative polymerase chain reaction

Data availability

The raw WES data presented in this paper were deposited in the Genome Sequence Archive (GSA-Human) at the National Genomics Data Center (NGDC), China National Center for Bioinformation/Beijing Institute of Genomics, Chinese Academy of Sciences.33,34 The data were publicly accessible under the accession number HRA010380 at https://ngdc.cncb.ac.cn/gsa-human.

RESULTS

Clinical characteristics of the infertile patients

In this cohort study, family 7 hosted four brothers and three sisters, among whom two brothers (IV:3 and IV:5) were diagnosed with primary infertility, and their wives were confirmed to be free of fertility-related disorders (Figure 1a). Physical examinations of the male patients showed normal external genital development, testicular size, and bilateral spermatic veins. Neither patient had a history of alcohol or drug abuse, exposure to hazardous environments, respiratory disorders, or situs inversus. Semen analysis revealed sufficient sperm concentration but abnormal motility. Specifically, sperm motility was significantly lower in IV:5 than that in IV:3, with 3.0% ± 0.5% and 5.0% ± 2.8% motility, respectively, as shown in Supplementary Table 5. Both patients exhibited normal serum testosterone, follicle-stimulating hormone, luteinizing hormone levels, and a 46,XY karyotype. H&E staining was used to assess the sperm morphology, revealing no discernible difference in the percentage of abnormalities in the sperm head between the fertile control subjects and the patients. However, a significantly higher percentage of severe morphological abnormalities was observed in the spermatozoa flagella of the patients, including short, absent, bent, coiled, and irregular caliber patterns, than that in the fertile control subjects (Figure 1b and 1c). To assess the acrosome deformity, we stained the acrosome with a FITC-PNA probe. Our results showed an intact and cap-shaped texture of the PNA signal in both the patients and control subjects, indicating a normal head structure and morphology (Figure 1d). These observations indicate that the affected individuals had a severe MMAF phenotype.

Figure 1.

Pedigree chart and morphological analysis of spermatozoa samples from the controls and patients. (a) Pedigree chart of family 7. Circles and squares represent males and females, respectively, while rhombus represents individuals without specified sex information. Black arrows mark individuals chosen for whole-exome sequencing. Parallel horizontal lines denote first-degree consanguineous marriage. Black-filled squares represent individuals affected by asthenoteratozoospermia, while slash symbols indicate deceased family members. (b) Hematoxylin-eosin (H&E) staining and (c) morphological statistical analysis of spermatozoa from the fertile control subjects and patients with biallelic DNHD1 mutations. Scale bars = 5 mm. Data are presented as mean ± standard deviation. In the control group, 115 ± 10 spermatozoa, while for the patients, 230 ± 15 (IV:3) and 215 ± 13 (IV:5) spermatozoa were counted in triplicate. ^**P < 0.01, ^***P < 0.001, ^****P < 0.0001, Tukey’s multiple comparison test. (d) Fluorescein isothiocyanate-labeled peanut agglutinin (FITC-PNA; green) along with anti-β-tubulin (red) immunofluorescence staining for spermatozoa from fertile control subjects and patients with biallelic DNHD1 mutations. Scale bars = 5 mm. DNHD1: dynein heavy chain domain 1; NS: not significant; WES: whole-exome sequencing.

Supplementary Table 5.

Clinical characteristics of patients diagnosed with MMAF

Characteristic	Reference	IV: 3	IV: 5
Age (year)*	--	55	37
Marriage duration*	--	33	13
Height (cm)/weight (kg)	--	162/72	165/70
Karyotype	--	46, XY	46, XY
Semen analysis**
Semen volume (ml)	>1.4	2.5±0.3	3.0±0.5
Sperm concentration (×106 ml-1)	>16	22.0±3.7	29.5±5.3
Motility (%)	>40	15.0±4.2	10.0±3.8
Progressive motility (%)	>32	5.0±2.8	3.0±2.1
Non-motile spermatozoa (%)		85.0±4.2	88.0±6.3
Normal flagella (%)		7.4±5.0	4.0±3.0
Absent flagella (%)		7.4±2.0	13.0±2.0
Short flagella (%)		15.3±1.0	20.5±1.0
Bent flagella (%)		46.0±3.0	31.0±2.0
Coiled flagella (%)		18.1±2.0	21.5±3.0
Irregular caliber flagella (%)		5.0±6.0	10.0±2.3
Affected spermatozoa (%)		93.6±2.8	95±3.6
Hormones analysis***
FSA (U l-1)	0.80-5.10	2.6	4.2
LH (U l-1)	1.24-8.62	2.1	7.2
Prolactin (ng ml-1)	2.64-13.13	3.5	11.5
Testosterone (mmol ml-1)	4.6-20	14.2	15.5

*Age and marriage duration at the time of sample collection, i.e., 2023. **Reference values were published by World Health Organization in 2021. ***Reference values were suggested by the local hospital. Data are presented as the mean±standard deviation. FSH: follicle-stimulating hormone; LH: luteinizing hormone

WES identified novel variants in DNHD1 in the patients with MMAF

We performed WES on the patient (IV:5), his father (III:1), and his fertile brother (IV:4) to identify the genetic cause of MMAF in this family (Figure 1a). Biallelic deleterious variants (c.A4457G; p.K1486R, and c.C10624T; p.R3542*) in DNHD1 were identified in the patient (Table 1 and Supplementary Figure 1 ^{(136.3KB, tif)} ). According to the GTEx and NCBI databases, DNHD1 is highly expressed in the testis. It is located on chromosome 11p15.4, with an approximate size of 74.74 kb, and consists of 40 exons. It encodes the DNHD1 protein, which is predicted to be 4753 amino acids long and includes five CC domains along with domains containing acidic and basic residues (Figure 2a and 2b).

Table 1.

Candidate dynein heavy chain domain 1 variants identified in patients IV:3 and IV:5 from family 7 with multiple morphological abnormalities of the sperm flagella

Patient	Gene	NT genomic change (GRCh38)	Consequence	Variant	Variant zygosity	MAF 1000 genomes project	MAF gnomAD	SIFT/PolyPhen-2/Mutation Taster
Patient IV: 3 and Patient IV: 5	DNHD1	chr11: 6545396_A>G	Missense	NM_144666: exon21: c.A4457G: p.K1486R	Homozygous	0	0.0006	T/B/B
	DNHD1	chr11: 6564672_C>T	Stop-gain	NM_144666: exon32: c.C10624T: p.R3542*	Homozygous	0.00019	0.0005	D/D/D

Variant chromosomal location, consequences, identification ID, exon number, cDNA and protein notation, zygosity, MAF in the 1000 Genomes Project, and gnomAD database, and three in silico prediction algorithms (SIFT, PolyPhen-2, and Mutation Taster) have been shown. T: tolerated; B: benign; D/D/D: damaging/deleterious/disease causing; MAF: multiple allelic frequency; DNHD1: dynein heavy chain domain 1; cDNA: complementary DNA

Figure 2.

Detailed sequence and structural information for DNHD1 variants. (a) Schematic showing the DNHD1 location on chromosome (chr) 11. The mutation site is depicted by red arrowheads in the transcript and protein. (b) Expression profiling of DNHD1 across 27 different tissues, as defined by bulk RNA sequencing by the GTEx project. (c) Sanger sequencing chromatograms of the patients with biallelic DNHD1 mutations (IV:3 and IV:5), the father (III:1), and the fertile brother (IV:4). The mutation site is indicated with a dotted rectangle with a red arrowhead. Multiple sequence alignment conservation analysis of the altered amino acids; (d) p.K1486R and (e) p.R3542*, across different organisms. Mutant amino acid is indicated with a red rectangle. The alignment was accomplished online through NCBI BLAST. A: adenine; G: guanine; C: cytosine; T: thymine; aa: amino acid; DNHD1; dynein heavy chain domain 1; UTR: untranslated region; DHC: dynein heavy chain; TPM: transcripts per million; GTEx: genotype-tissue expression; NCBI: National Center for Biotechnology Information; BLAST: Basic Local Alignment Search Tool.

Sanger sequencing was performed for validation, and segregation analysis was used to detect variants in the patient, his father, and his fertile brother. Our findings revealed that the same homozygous mutations, c.A4457G and c.C10624T (DNHD1^MT/MT), were found in the affected brother; however, the father and fertile brother were heterozygous (DNHD1^WT/MT) for these variants (Figure 2c). Although we did not confirm these variants in the mother, the fertile brother and father showed heterozygosity, which strongly suggests that they inherited these variants from their heterozygous carrier parents. This pattern of inheritance aligns with an autosomal recessive mode of inheritance in DNHD1, resulting in the MMAF in these patients.

DNHD1 mutations impaired the stability and loss of function of the DNHD1 protein

To further assess the pathogenicity of the identified DNHD1 variants, we used combinatorial bioinformatics tools to evaluate their potential impact on the DNHD1 protein. Owing to their rare existence, that is, MAF <0.001 in public databases, such as 1000 Genomes Project and gnomAD, several bioinformatics tools, including PolyPhen-2, MutationTaster, and SIFT, predicted the nonsense variant to be deleterious/damaging (Table 1). Conservation analysis revealed that both variants have a significant impact on DNHD1 protein, as the mutations are located within highly conserved regions. The missense variant site is conserved across primates (Figure 2d), whereas the nonsense variant site is conserved across both primates and non-primates (Figure 2e and Supplementary Figure 2 ^{(136.8KB, tif)} ). To further validate and confirm these predictions, we investigated the distribution of DNHD1 in sperm cells from individuals who exhibited DNHD1 mutations and fertile control subjects, using IF staining against DNHD1 and β-tubulin. Our results showed that DNHD1 is colocalized along the sperm flagella with the β-tubulin microtubule protein, particularly prominent in the mid and principal regions in the fertile control subjects. However, almost complete absence of DNHD1 was observed in the spermatozoa of the patients with biallelic DNHD1 mutations (Figure 3a), which corroborated and confirmed earlier predictions. Furthermore, quantitative polymerase chain reaction (qPCR) analysis of DNHD1 mRNA revealed a significant reduction (P < 0.0001) in the patients compared with the controls (Supplementary Figure 3 ^{(117.7KB, tif)} ), strongly indicating that the mutation induces nonsense-mediated mRNA decay (NMD). This mRNA degradation likely prevents the synthesis of a truncated or nonfunctional protein, ultimately leading to the loss of protein expression and function in affected individuals.

Figure 3.

TEM and IF analyses of the spermatozoa samples from the fertile control subjects and patients. (a) Immunofluorescence (IF) staining for anti-DNHD1 (green) with anti-β-tubulin (red) in spermatozoa samples from the fertile control subjects and patients with biallelic DNHD1 mutations. Scale bars = 5 mm. (b) Transmission electron microscopy (TEM) image showing intact (“9 + 2”) and disorganized (“9 + 0”) axoneme structures of spermatozoa from the fertile control subjects and patient with biallelic DNHD1 mutations, respectively. Scale bars = 100 nm. Immunofluorescence staining for (c) anti-SPAG6 (red) and anti-α-tubulin (green), and (d) anti-SPEF2 (red) and anti-α-tubulin (green) in spermatozoa samples from the fertile control subjects and patients with biallelic DNHD1 mutations. Scale bars = 5 mm. CP: central pair; MTD: microtubule duplet; ODF: outer dense fibers; DAPI: 4’,6-diamidino-2-phenylindole; SPEF2: sperm flagellar protein 2; DNHD1: dynein heavy chain domain 1; SPAG6: sperm-associated antigen 6.

DNHD1 variants are associated with severe axonemal defects

To assess the impact of the DNHD1 variants on spermatozoa flagella in the patients with MMAF, TEM was performed on the spermatozoa samples from the patients and fertile control subjects. Typically, in normal spermatozoa, the midpiece and principal piece feature a “9 + 2” microtubule arrangement, comprising nine peripheral microtubule doublets (MTDs) aligned with nine outer dense fibers (ODF), along with a CP. In our study, the spermatozoa samples from the control subjects displayed a well-preserved, intact, and complete “9 + 2” axoneme microtubule organization. By contrast, the spermatozoa from the individuals who harbored the DNHD1 variants exhibited abnormalities in the cross-section, including the absence of the CP “9 + 0” and the disorganization of the “9 + 0” structure, and anomalies in the MTDs (Figure 3b). To investigate in more detail, immunostaining was performed using antibodies targeting SPAG6 and SPEF2, which are essential CP components, along with α-tubulin. Our results showed that both CP structural components, SPAG6 and SPEF2, were almost missing in the patients compared with the controls, where an intact signal was found with α-tubulin along the whole flagella (Figure 3c and 3d, and Supplementary Figure 4 ^{(51.2KB, tif)} ). Collectively, these findings suggest that the DNHD1 variants resulted in severe axonemal defects by disrupting the CP structure of flagella owing to the loss of SPAG6 and SPEF2, which are crucial CP components.

DNHD1 variants lead to defective mitochondrial sheath

Initially, we conducted TEM for ultrastructural analysis and found the complete absence of CP. During the comprehensive examination of the sperm structure, we observed that the spermatozoa carrying the DNHD1 variants exhibited a swollen midpiece and disrupted mitochondrial sheath (MS; Figure 4a). To confirm the integrity of the MS along the entire midpiece and assess for vacuolation or swelling, we conducted IF against the TOMM20, a mitochondrial marker, along with α-tubulin, on the spermatozoa samples from the patients and fertile controls to validate the TEM findings. Our findings revealed that in the controls, the TOMM20 signal persisted throughout the length of the midpiece; however, in the patients, a broadened, misshapen, and shortened signal was evident in the midpiece and showed a positive correlation with the TEM analysis result (Figure 4b). In addition, staining for the fibrous sheath marker A-kinase anchor protein 4 (AKAP4) indicated that the structure was intact and comparable with the control (Supplementary Figure 5 ^{(185.4KB, tif)} ). Collectively, these findings indicate that the DNHD1 variants lead to a defective MS due to mitochondrial disarrangement, which is likely caused by the curling behavior of the sperm flagella. This is accompanied by the absence of CP, ultimately resulting in the MMAF phenotype in the affected individuals.

Figure 4.

TEM and immunofluorescence analysis of mitochondrial sheath. (a) Transmission electron microscopy (TEM) analysis result showing intact midpiece with arranged mitochondria (red arrowhead) in a spermatozoa sample from a fertile control subject compared with the widened midpiece with disorganized mitochondria (red arrowhead) in a spermatozoa sample from a patient with biallelic DNHD1 mutations. Scale bars = 500 nm. (b) Immunofluorescence analysis results of anti-TOMM20 in spermatozoa samples from the fertile control subjects and patients with biallelic DNHD1 mutations. Scale bars = 5 mm. DNHD1: dynein heavy chain domain 1; TOMM20: translocase of outer mitochondrial membrane 20; DAPI: 4’,6-diamidino-2-phenylindole.

DISCUSSION

Asthenozoospermia, one of the most prevalent forms of infertility, affecting approximately 19% of infertile men, is phenotypically diverse, with genetic associations primarily linked to conditions such as PCD and MMAF.8,35,36 MMAF was initially introduced in 2014 and represents a distinct subtype of asthenoteratozoospermia marked by various abnormal flagellar phenotypes, such as short, absent, coiled, bent, and irregular caliber flagella, and severely compromised sperm motility. However, in previous studies, the terms used to describe similar conditions included “short tails”, “dysplasia of the fibrous sheath”, or “stump tails”.5,7,37 The absence of CP is considered a hallmark of the MMAF phenotype.37

In this study, WES identified novel homozygous variants in DNHD1 in two patients diagnosed with MMAF from a consanguineous Pakistani family. These biallelic variants are extremely rare in the general population, having MAF <0.001 in databases such as 1000 Genomes Project and gnomAD. Their pathogenic nature was further validated through conservation and in silico analyses, including SIFT, PolyPhen-2, and MutationTaster. In addition, data from the H&E staining, IF, and TEM analyses revealed abnormalities in sperm flagella, including axoneme disorganization, loss of CP, and a widened and misshapen MS in the affected men harboring DNHD1 variants compared with the controls. However, the head morphology was normal, as indicated by the intact acrosome signal observed in both the patients and controls. The intact, cap-shaped acrosome, which covers 40%–70% of the sperm head, is crucial for proper nuclear formation and head integrity. Abnormal acrosome biogenesis can lead to nuclear abnormalities and head deformities, as seen in cases of human teratozoospermia.38,39,40,41 Together, all these data strongly suggest the presence of the MMAF phenotype in these affected individuals, which distinguishes it from other forms of asthenozoospermia, such as primarily MS defects, defects related to PCD, dysfunction of the annulus, and defects associated with ion channels. For instance, MS defects are typically localized to the midpiece, characterized by the absence or reduction of mitochondria and the absence of an annulus, which is a ring-like structure marking the boundary between the midpiece and the principal piece of the sperm tail.42,43,44 In our study, MS malformation was observed on TEM and further confirmed by IF staining against TOMM20, which revealed a broad and widened signal in the patients’ spermatozoa compared with those of the controls, correlating with the TEM findings. The MS defect is an atypical characteristic observed in patients with MMAF, with only a few gene mutations, such as cilia and flagella-associated protein 58 (CFAP58) and CFAP65, reported to result in MS abnormalities.45,46 Such ultrastructural defects may be linked to intraflagellar transport (IFT) defects.46 However, previous study has shown that mutations in DNHD1 do not affect CFAP65 and IFT.47 Moreover, IF staining for AKAP4, a major component of the fibrous sheath, also displayed a normal signal in the patients’ spermatozoa. Therefore, the MS defect observed in this study was strongly attributed to the unique curling behavior of flagella due to the disrupted axoneme structure, differentiating it from the typical MS defect because TOMM20 expression remains consistent. This observation highlights the structural impact of flagellar defects on mitochondrial organization despite the preservation of TOMM20 expression. In addition, we did not observe any respiratory problems in the affected individuals that were highly correlated to MMAF other than PCD. PCD and MMAF share similarities in the axonemal structures of cilia and flagella. However, some patients with PCD do not experience infertility, which strongly suggests that MMAF represents a distinct cause of asthenoteratozoospermia, independent of PCD.48

The identified mutations include a missense mutation (c.A4457G; p.K1486R), and a stop-gain mutation (c.C10624T; p.R3542*). The missense variant is localized to exon 21, which is one of the largest exons of the DNHD1 gene. Two previous studies independently reported 5/14 (c.5560C>T, c.6498T>G, c.4072C>T, c.4141C>T, and c.5347delC) and 2/4 variants (c.5989G>A and c.6031C>T), respectively, located at exon 21.19,20 These observations suggest that exon 21 is a mutation hotspot, crucially important for DNHD1 function, and is often mutated in individuals with DNHD1-associated asthenoteratozoospermia. However, owing to reliance on bioinformatic predictions and the lack of functional validation, the missense variant identified in our study remains classified as a variant of uncertain significance (VUS). The second stop-gain mutation occurs at amino acid 3542 and is predicted to result in a truncated protein. However, our study confirms that this mutation triggers NMD, ultimately leading to the loss of DNHD1 protein expression. It is strongly suggested that nonsense variants cause deleterious effects or pathogenicity, which leads to MMAF in these individuals, while the missense variant is classified as a VUS. Both variants were found to be homozygous (DNHD1^MT/MT) in the patients and heterozygous (DNHD1^WT/MT) in the father and fertile brother. Although blood samples from the mother were not available, the fertile brother and father showed heterozygosity, which strongly suggests that they (IV:3 and IV:5) inherited homozygous variants from their heterozygous carrier parents. This pattern of inheritance aligns consistently with an autosomal recessive mode of inheritance in DNHD1, resulting in the MMAF in these patients. Given the lack of effective treatments to improve semen parameters and fertility outcomes in these individuals, intracytoplasmic sperm injection is strongly recommended as an effective option for achieving pregnancy.19,49

In conclusion, we identified a novel biallelic nonsense variant through WES in DNHD1 in an infertile individual from a consanguineous Pakistani family diagnosed with MMAF. Our experimental results in the patients showed that a nonsense mutation in DNHD1 can induce MMAF-associated asthenoteratozoospermia (Figure 5). These findings expanded the spectrum of genetic mutations in the DNHD1 gene linked to MMAF and asthenozoospermia. Further functional analyses of specific DNHD1 mutations could provide valuable insights into the genetic basis of male infertility, helping genetic counselors and clinicians in designing more personalized treatment strategies.

Figure 5.

DNHD1 is crucial for sperm flagellar integrity and male fertility. We identified a novel biallelic nonsense mutation in DNHD1 that triggered nonsense-mediated mRNA decay, leading to loss of DNHD1 protein. This resulted in impaired sperm motility and a severe MMAF phenotype, underscoring the essential role of DNHD1 in flagellar structure and function. DNHD1: dynein heavy chain domain 1; mRNA: messenger RNA; MMAF: multiple morphological abnormalities of the sperm flagella.

AUTHOR CONTRIBUTIONS

JQB and LMW conceived the project. IU collected samples with the help of MIK, carried out all the experiments and designed the figures. IZ and CLX performed WES data bioinformatics analysis. XMX, WQL, JQZ, and HT helped in experimentation. IU and JQB wrote the manuscript. All authors read and approved the final manuscript.

COMPETING INTERESTS

All authors declare no competing interests.

Supplementary Figure 1

WES data analysis pipeline. Whole-exome sequencing data analysis pipeline. Variants with a minor allele frequency >0.01 in databases such as the 1000 Genomes Project, ExAC, or ESP6500 were excluded, as were variants that appeared homozygous in control groups (our in-house 32 fertile control). GATK: Genome Analysis Toolkit-4; MAF: multiple allelic frequency; DNHD1: dynein heavy chain domain 1; SNP: single-nucleotide polymorphisms; INDELs: insertion and deletion; FATHMM: Functional Analysis through Hidden Markov Models.

Supplementary Figure 2

Conservation analysis of DNHD1 protein. Multiple sequence alignment was performed using NCBI BLAST to analyze the conservation of the DNHD1 protein within the specified amino acid region across both primates and non-primates. The target amino acid is indicated by a red arrow, while amino acids highlighted in red represent differences across species. DNHD1: dynein heavy chain domain 1; BLAST: Basic Local Alignment Search Tool.

Supplementary Figure 3

RT-qPCR analysis of DNHD1 mRNA. (a) RT-qPCR analysis for DNHD1, normalized with TUBA4A. Statistical analysis was performed using t-test to compare the datasets (GraphPad Prism 10.1.1). ^****P < 0.0001. (b) Gel electrophoresis of RT-qPCR of DNHD1 (164 bp) and TUBA4A (174 bp). M: marker; C: control, IV:3 and IV:5: patients; NTC: no template control; RT-qPCR: reverse transcription-quantitative polymerase chain reaction; DNHD1: dynein heavy chain domain 1; mRNA: messenger RNA; TUBA4A: Tubulin Alpha 4a.

Supplementary Figure 4

Immunoblot (IB) analysis of SPAG6 from fertile controls and patients with DNHD1 mutation, with α-tubulin used as an internal control. SPAG6: sperm-associated antigen 6; TUBA: Tubulin Alpha.

Supplementary Figure 5

Immunofluorescence staining of AKAP4. Immunofluorescence staining of anti-AKAP4 (red) and anti-α-tubulin (green) of spermatozoa from fertile control subject and patients with biallelic DNHD1 mutations. Scale bars = 5 mm. AKAP4: A-kinase anchoring protein 4; DNHD1: dynein heavy chain domain 1.