A novel hemizygous missense variant in the BEND2 gene is associated with nonobstructive azoospermia

Abstract

Nonobstructive azoospermia (NOA), the most severe form of male infertility, frequently arises from genetic defects that disrupt spermatogenesis. In this study, a novel hemizygous missense variant (NM_001184767.2 [c.G1069A; p.V357I]) is identified in the X-linked BEN domain-containing 2 (BEND2) gene of a patient with NOA characterized by spermatocyte maturation arrest. Whole-exome sequencing and Sanger validation confirmed that this rare variant is absent in fertile controls and that no pathogenic variants were detected in established NOA genes. Computational analysis predicted potential structural alterations via AlphaFold modeling, leading to the hypothesis that the ability of BEND2 to recognize genomic targets may be compromised. The patient’s phenotype phenocopies the meiotic arrest observed in Bend2-knockout mice. Expression profiling confirmed predominant BEND2 transcription in human and mouse testes, peaking in early spermatocytes and coinciding with meiotic initiation, with reduced transcript levels detected in the proband’s peripheral blood compared with those in an obstructive azoospermia control. This study reports a pathogenic BEND2 variant associated with NOA with spermatocyte arrest, highlighting its critical role in human meiosis and expanding the genetic etiology of male infertility.

INTRODUCTION

Male infertility represents a major global health concern, accounting for nearly 30%–50% of infertility cases among the couples of reproductive age.1 Clinically, it manifests through diverse phenotypes, including azoospermia, oligozoospermia, asthenozoospermia, and teratozoospermia. Among these, nonobstructive azoospermia (NOA), the most severe form, is defined by a lack of sperm in the ejaculate due to defective spermatogenesis and affects 10%–20% of infertile men.2,3 Unlike obstructive azoospermia (OA), which results from reproductive tract blockage, NOA arises from testicular dysfunction.4 The histopathological spectrum of NOA is heterogeneous, encompassing Sertoli cell-only syndrome (SCOS), maturation arrest (MA), and hypospermatogenesis.5,6 MA can be subdivided into early MA (characterized by arrest at the spermatogonial and spermatocyte stages) and late MA (characterized by spermatid arrest), with spermatocyte arrest notably linked to a higher prevalence of genetic abnormalities than other NOA subtypes.7–9 Molecular diagnostics have become fundamental in the clinical assessment of severe male factor infertility owing to the substantial incidence of genomic anomalies in affected individuals.8,10–12 The implementation of whole-exome sequencing (WES) has revolutionized this field by revealing new monogenic causes of NOA, with pathogenic variants identified in genes such as helicase for meiosis 1 (HFM1), syntaxin 2 (STX2), testis-expressed 11 (TEX11), KASH domain-containing 5 (KASH5), and heat shock transcription factor 5 (HSF5).13–19 Despite remarkable advances in genomic technologies that expand our knowledge of infertility pathogenesis, the catalog of validated disease-causing mutations remains far from exhaustive, leaving many NOA-associated genetic factors uncharacterized. Therefore, persistent investigations of novel genetic contributors are warranted to increase diagnostic precision, optimize patient counseling, and facilitate the development of targeted treatment approaches.

BEN domain-containing 2 (BEND2), an X-linked gene encoding a BEN domain-containing protein, is predominantly expressed in germ cells (Genotype-Tissue Expression [GTEx] Portal: https://www.gtexportal.org/home/gene/BEND2; last accessed on May 5, 2025). Although BEND2 is frequently identified as a fusion partner in neurodevelopmental disorders and is suspected to be potentially associated with postmeiotic arrest in a WES study of NOA, no pathogenic BEND2 variants have been systematically documented in monogenic human diseases.20–23 In 2022, Ma et al.24 demonstrated that mouse Bend2 is specifically expressed in spermatogenic cells during meiotic initiation and is essential for meiotic progression. Genetic ablation of Bend2 in male mice resulted in meiotic arrest at the zygotene-to-pachytene transition, accompanied by synaptic disruption, impaired DNA double-strand break (DSB) repair, and nonhomologous chromosomal pairing.24

In this study, we identified a novel hemizygous missense variant in BEND2 (c.G1069A; p.V357I) in a NOA patient through WES. Functional characterization confirmed its potential association with azoospermia, expanding the genetic landscape of male infertility and supporting its inclusion in future diagnostic screening panels for individuals with azoospermia.

PARTICIPANTS AND METHODS

Study participants

A 31-year-old Han Chinese male with a 2-year history of infertility was enrolled in September 2023 from the Department of Urology and Andrology at Sir Run Run Shaw Hospital, Zhejiang University School of Medicine (Hangzhou, China). Peripheral blood samples were obtained from this patient. Karyotype analysis, serum reproductive hormone analysis, scrotal content palpation, scrotal ultrasound, and testicular histology were performed in local clinical laboratories. In accordance with the 6^th edition of World Health Organization (WHO) guidelines, the proband received an idiopathic NOA diagnosis.25 An OA patient exhibiting normal spermatogenesis but with epididymal obstruction served as the control. Microsurgical testicular sperm extraction (micro-TESE) performed on the proband revealed no spermatozoa. The proband provided written informed consent prior to study initiation. This research received approval from the ethics committee of Sir Run Run Shaw Hospital, Zhejiang University of Medicine (Approval No. 2022-726-01). Written informed consent was obtained from the participant, and permission for publication of the data was secured.

Semen parameter analysis

Semen analyses were performed at the relevant laboratories in accordance with the 6^th edition of WHO guidelines.25 Samples from the proband were collected through masturbation after a period of 2–7 days of sexual abstinence. Following 20 min of liquefaction at 37°C, a routine semen examination was conducted.

WES and Sanger sequencing

Genomic DNA (gDNA) was isolated from the proband’s peripheral blood using a QIAamp DNA Blood Mini Kit (Qiagen, Hilden, Germany) following standard protocols. Library preparation for targeted panel sequencing was performed on the samples using the IDT xGen Exome Research Panel version 1.0 (Integrated DNA Technologies, Coralville, IA, USA), followed by sequencing on an Illumina NovaSeq platform (Illumina, San Diego, CA, USA). Reads were aligned to the human reference genome (GRCh38/hg38) using Burrows–Wheeler Aligner software.26 Alignment against the GRCh38 reference assembly revealed a novel rare BEND2 variant. This BEND2 mutation was validated by the polymerase chain reaction (PCR) with the following primers: forward, 5’-CAGCAATTTCACTCTTAGGTGCTT-3’; and reverse, 5’-ACTCACCATACTTGGGAGGGAT-3’. The resulting PCR product was then subjected to Sanger sequencing using the forward primer.

In silico analysis

Computational predictions of the functional impact of the variant were performed using three algorithms: Sorting Intolerant From Tolerant (SIFT; https://sift.bii.a-star.edu.sg/, last accessed on December 6, 2024), Functional Analysis Through Hidden Markov Models (FATHMM; http://fathmm.biocompute.org.uk/, last accessed on December 6, 2024), Mutation Taster (http://www.mutationtaster.org/, last accessed on December 6, 2024), and the Likelihood Ratio Test (LRT), with LRT predictions retrieved from the dbNSFP database (version 5.1; https://sites.google.com/site/jpopgen/dbNSFP, last accessed on December 6, 2024). The evolutionary conservation of the mutated amino acid residue was evaluated using Molecular Evolutionary Genetics Analysis (MEGA) software.27 To further assess structural consequences, wild-type (WT) and mutant polypeptide structures were modeled with AlphaFold (https://alphafold.com/, last accessed on May 16, 2025), and protein folding alterations were visualized using PyMol Software (version 2.0). Analysis of BEND2 expressions in spermatogenic cells followed our established single-cell RNA sequencing (scRNA-seq) workflow.16 Briefly, we implemented an established scRNA-seq pipeline integrating published mouse and human testicular datasets.16,28 Data normalization and batch corrections were performed with Seurat’s Harmony algorithm, with cell-type annotation based on canonical marker genes. Expression patterns were visualized through Seurat’s DotPlot and FeaturePlot functions.

Histological and immunofluorescence (IF) analysis

Testicular biopsy specimens from azoospermia patients were subjected to histopathological characterization by hematoxylin-eosin (H&E) staining and IF staining. Fine-needle aspiration biopsies were fixed overnight in 4% paraformaldehyde (PFA) at 4°C, dehydrated with ethanol, embedded in paraffin, and sectioned at 4–5 μm thickness. For histological evaluation, deparaffinized sections were stained with hematoxylin for 3 min, rinsed in tap water (10 min), differentiated in 0.1% HCl/70% ethanol, counterstained with eosin (20 s), dehydrated, and mounted using HistoMount (Solarbio, Beijing, China). Stained sections were examined with an Olympus BX43 microscope (Olympus, Tokyo, Japan) under standardized magnification. For IF, antigen retrieval was performed with sodium citrate buffer (pH 6.0, 98°C, and 20 min). The sections were blocked with 5% donkey serum before being incubated overnight at 4°C with a primary antibody against phosphorylated histone H2AX (Ser139), as γH2AX (ab11174; 1:300; Abcam, Cambridge, UK). After the cells were washed with phosphate-buffered saline (PBS), the sections were incubated with Alexa Fluor 488-conjugated anti-rabbit IgG (A-11008; 1:500; Invitrogen, Carlsbad, CA, USA) and peanut agglutinin (PNA; CL-1073-1; 1:500; Vector Labs, Burlingame, CA, USA) for 2 h at room temperature, after which the nuclei were counterstained with Hoechst 33342 (H3570; 1:5000; Invitrogen). Imaging was performed on a Zeiss LSM 980 confocal system (ZEISS, Oberkochen, Germany).

mRNA expression analysis

Total RNA was isolated from the peripheral blood of the NOA proband and OA control, after which first-strand complementary DNA (cDNA) synthesis was performed using a Takara PrimeScript RT Reagent Kit (6110A; Takara, Otsu, Japan) with random hexamers. Reverse transcription quantitative polymerase chain reaction (RT-qPCR) was performed on a Bio-Rad CFX96 system with Takara TB Green Premix Ex Taq II (RR420A; Takara), with the following thermal cycling conditions: initial denaturation (95°C for 30 s), 40 cycles of 95°C for 5 s; and 60°C for 30 s. Relative BEND2 mRNA expressions were quantified via the 2^−ΔCt method and normalized to that of glyceraldehyde-3-phosphate dehydrogenase (GAPDH) using the following primers: BEND2 forward, 5’-CCAATTTGGAATCTGGCCCAC-3’, and BEND2 reverse, 5’-GGCCACTGTCACTGTCCAAT-3’; and GAPDH forward, 5’-GGAGCGAGATCCCTCCAAAAT-3’, and GAPDH reverse, 5’-GGCTGTTGTCATACTTCTCATGG-3’.

Statistical analyses

All the statistical analyses were performed using GraphPad Prism version 10 (GraphPad Software, San Diego, CA, USA). The data are presented as the means ± standard deviation (s.d.). Student’s t-test was used for between-group comparisons, with statistical significance defined as P < 0.05.

RESULTS

Clinical characteristics of the proband

A 31-year-old index patient sought evaluation at Sir Run Run Shaw Hospital for primary infertility after 2 years of unprotected marital intercourse. The family history revealed no heritable fertility disorders or chronic diseases and no parental consanguinity. The patient had no personal history of cryptorchidism, hypogonadism, malignancy, varicocele, or substance use. Physical examination confirmed the normal development of the external genitalia and accessory glands, whereas transrectal and scrotal ultrasound revealed no obstructive pathology. Both testes exhibited reduced volumes (left: 7.2 ml; right: 5.8 ml) relative to those of the normative standards (12–25 ml), as shown in Table 1. A semen analysis performed according to the 6^th edition of WHO guidelines25 revealed that the ejaculate volume was within the normal reference range, but complete azoospermia was identified. Laboratory studies revealed slightly elevated luteinizing hormone (LH) levels and borderline low testosterone levels. Follicle-stimulating hormone (FSH) and prolactin levels were within reference ranges, and genetic testing revealed a normal 46,XY karyotype without Y chromosome microdeletions (Table 1). Histopathological evaluation following micro-TESE at Sir Run Run Shaw Hospital revealed spermatogenic arrest at the spermatocyte level, characterized by seminiferous tubules containing solely Sertoli cells, spermatogonia, and spermatocytes without observable spermatids (Figure 1).

Table 1.

Clinical data of the analyzed BEND2 mutation-bearing azoospermic patient

Variable	Patient	Reference value a
Testosterone (ng ml−1)	1.58	1.75–7.81
LH (IU l−1)	8.74	1.24–8.62
FSH (IU l−1)	8.9	1.27–19.26
Prolactin (ng ml−1)	8.46	2.64–13.13
Testis (ml)
Left	7.2	12–25
Right	5.8	12–25
Semen volume (ml)	3	≥1.5
Semen pH	6.8	≥7.2
Sperm concentration (106 ml−1)	0	≥15
Sperm motility (%)	0	≥40
Y chromosome microdeletion	Negative	Negative
Karyotype	46,XY	46,XY
Histology	Spermatogenic arrest at the spermatocyte level	Normal spermatogenesis

^aReference values were established by Sir Run Run Shaw Hospital (Hangzhou, China) based on the normal population. BEND2: BEN domain-containing 2; LH: luteinizing hormone; FSH: follicle-stimulating hormone

Figure 1.

NOA with a spermatocyte arrest phenotype in the proband. (a) H&E-stained testicular biopsy sections from an OA control (upper panel) and the proband (lower panel). (b) IF detection of γH2AX (red; DNA double-strand break marker) and PNA (green; acrosome marker) in testicular tissues from an OA control (upper panel) and the proband (lower panel). NOA: nonobstructive azoospermia; OA: obstructive azoospermia; H&E: hematoxylin−eosin; SPG: spermatogonia; PS: pachytene spermatocytes; RS: round spermatids; ES: elongated spermatids; SC: Sertoli cells; γH2AX: phosphorylated histone H2AX (Ser139); PNA: peanut agglutinin; IF: immunofluorescence.

Identification of a novel hemizygous BEND2 variant in an azoospermic patient by WES

To investigate the genetic determinants of the patient’s azoospermia, WES was conducted (Figure 2a). Further integrated bioinformatic analysis revealed a rare hemizygous missense variant in the X-linked BEND2 gene (NM_001184767.2 [c.G1069A; p.V357I]), as shown in Table 2. The variant was validated by Sanger sequencing and was absent in the OA control (Figure 2b); furthermore, it demonstrated a minimal population frequency in the genome aggregation database (gnomAD), as shown in Table 2. Notably, comprehensive screening of the proband excluded pathogenic variants in established NOA and male infertility-associated genes.

Figure 2.

Identification of a hemizygous BEND2 variant associated with NOA. (a) Pedigree chart of the proband. (b) Confirmation of the BEND2 variant through Sanger sequencing. (c) Orthologous sequence alignment demonstrating that the affected amino acid residues are not highly conserved. (d) Predicted structural models of BEND2 p.V357I by AlphaFold. The magnified views illustrate the structural differences between the wild-type BEND2 and the mutant p.V357I. BEND2: BEN domain-containing 2; NOA: nonobstructive azoospermia; WES: whole-exome sequencing; WT: wild-type; MT: mutant; Y: Y chromosome.

Table 2.

Detailed information of the BEND2 variant carrying by the patient

Subject	Patient
Gene	BEND2
Position (hg38)	chrX:18180597
Transcript ID	NM_001184767.2
Genomic mutation	c.G1069A
Amino change	p.V357I
Mutation type	Hemizygous missense
Allele frequency
Genome Aggregation Database	0.00002686
Genome Aggregation Database XY	0.00005896
Genome Aggregation Database XY East Asia	0.0002
Pathogenicity prediction
SIFT	Tolerated
FATHMM	Tolerated
Mutation Taster	Neutral
LRT	Neutral

BEND2: BEN domain-containing 2; SIFT: Sorting Intolerant From Tolerant; FATHMM: Functional Analysis Through Hidden Markov Models; LRT: likelihood ratio test

The missense variant was subjected to in silico analysis employing multiple bioinformatics tools (e.g., SIFT, FATHMM, Mutation Taster, and LRT). Although these predicted the mutation as tolerated or neutral (Table 2), their underlying methodologies share significant similarity, being primarily based on sequence conservation. Consequently, for variant assessment, the results from these tools are integrated as a single evidentiary component rather than as distinct lines of evidence. Compounding this, the BEND2 gene exhibits relatively low evolutionary conservation across species (Figure 2c), a factor that previously contributed to its misclassification as noncoding RNA in mice. While the computational predictions did not strongly indicate pathogenicity, they do not preclude a deleterious effect.

Predictive structure modeling of the BEND2 p.V357I variant

Next, we characterized the impact of the p.V357I mutation on the BEND2 protein structure. This amino acid substitution site resides within the representative region of BEN family proteins, although it is not located in the BEN domain (https://www.ebi.ac.uk/interpro/protein/UniProt/Q8NDZ0/?isoform=Q8NDZ0-2, last accessed on May 16, 2025). To further understand the effect of this mutation on the protein’s three dimensional (3D) structure, we investigated the conformational changes and folding alterations in BEND2 that were caused by this variant using AlphaFold (https://alphafold.com/, last accessed on May 16, 2025) and PyMOL software (version 2.0). AlphaFold predictions indicate that the p.V357I mutation may compromise the geometric integrity of the hydrophobic core surrounding Val357, potentially triggering conformational shifts in the functional domains of BEND2 (Figure 2d). This perturbation possibly induces conformational rearrangements within functional domains that are essential for DNA binding, potentially compromising BEND2’s ability to recognize genomic targets and regulate transcription. However, to rule out global protein destabilization as the underlying mechanism, we assessed the change in free energy (ΔΔG) using I-Mutant (version 2.0). The obtained value of −0.9 kcal mol⁻¹ indicates that the mutation has a minimal effect on overall protein stability.

BEND2 is robustly expressed in testes

Before investigating BEND2 expression patterns in testicular germ cells, we assessed its pan-tissue distribution using the GTEx database. GTEx analysis revealed pronounced testis-specific enrichment of BEND2 transcripts in humans, although low-level expression was detected in peripheral blood (Figure 3a). This peripheral blood expression was corroborated by our RT-qPCR assays. Using primers spanning the mutation site, RT-qPCR demonstrated reduced BEND2 transcript levels in peripheral blood from the NOA proband harboring the variant (c.G1069A; p.V357I) compared with those in the OA control (Figure 3b). However, the absence of a specific antibody prevented us from performing protein-level validation of this observation. Collectively, these data indicate that while BEND2 is predominantly expressed in the testis in humans, its transcripts are present in peripheral blood at lower levels in NOA patients than in OA controls.

Figure 3.

BEND2 expression profiling in human and mouse testes. (a) GTEx database analysis showing the pan-tissue distribution of BEND2 in humans, demonstrating predominant testis-specific expression. (b) Reduced BEND2 transcript levels in peripheral blood from the proband carrying the p.V357I variant versus the OA control. (c) UMAP (a dimensionality-reduction technique for visualizing high-dimensional single-cell data) revealed enrichment of Bend2 expression in early meiotic germ cells within murine testicular subpopulations. (d) UMAP projection of human testicular subpopulations; the dashed circle denotes BEND2-expressing germ cell clusters during early meiosis. BEND2: BEN domain-containing 2; GTEx: genotype-tissue expression; TPM: transcripts per million; OA: obstructive azoospermia; UMAP: uniform manifold approximation and projection; SPG: spermatogonia; pL: preleptotene; L: leptotene; Z: zygotene; eP: early pachytene; mP: middle pachytene; lP: late pachytene; D: diplotene; M: metaphase; RS1–4: steps 1–4 spermatids; RS5-7: steps 5–7 spermatids; SCytes: meiotic spermatocytes; Stids: postmeiotic haploid round spermatids; PTM: peritubular myoid; VSM: vascular smooth muscle.

We next characterized BEND2 expression dynamics during spermatogenesis and its cell-type specificity. Previously published mouse testicular scRNA-seq data from our laboratory and public human testicular scRNA-seq datasets were analyzed.16,28 BEND2 expressions were overwhelmingly confined to germ cells within the testis, with negligible somatic cell detection. In mouse germ cells, the expressions peaked in spermatogonia and early spermatocytes but were absent in round spermatids and subsequent stages (Figure 3c), which is consistent with prior immunofluorescence studies.24 Similarly, human germ cells displayed marked BEND2 upregulation in early spermatocytes, which coincided with meiotic initiation, followed by rapid downregulation (Figure 3d). These conserved expression patterns across species suggest that BEND2 plays a critical role in the initial phases of spermatogenesis, particularly during meiosis.

DISCUSSION

This study identifies a novel hemizygous missense variant (NM_001184767.2 [c.G1069A; p.V357I]) in the X-linked BEND2 gene that is associated with NOA characterized by possible spermatocyte maturation arrest. The proband’s spermatocyte-stage arrest phenocopies the meiotic arrest observed in male Bend2-knockout mice at the zygotene-pachytene transition,24 providing compelling phenotypic evidence for pathogenicity. Although in silico predictors (e.g., SIFT, FATHMM, Mutation Taster, and LRT) classified p.V357I as tolerated/neutral, this likely reflects algorithmic limitations from overreliance on evolutionary conservation, notably low for BEND2 across species. AlphaFold modeling led to the hypothesis that the p.V357I substitution causes subtle perturbations to the local hydrophobic core, potentially affecting the structure of its DNA-binding and transcriptional regulatory domains. Comprehensive screening of the proband excluded pathogenic variants in established NOA genes, with this variant absent in the fertile OA control. The reduced BEND2 transcript levels in the proband’s peripheral blood compared with those in the OA control (P < 0.01) suggest the presence of mutant mRNA instability/dysregulation, although the absence of a specific antibody precluded protein-level validation. Conserved BEND2 expression dynamics peaking during meiotic initiation in mouse and human spermatogenesis then sharply decreases, which aligns with its critical role in early meiosis.

On the basis of multiple lines of evidence, we cautiously established this novel hemizygous BEND2 variant as a possible pathogenic cause of the patient’s azoospermia. First, Bend2 is a well-characterized male infertility gene, with knockout mouse models demonstrating spermatocyte-stage arrest that matches the proband’s phenotype.24 Second, comprehensive genetic screening of the proband revealed no pathogenic variants in known NOA-associated genes; the identified BEND2 variant was absent in fertile controls and exhibited a very low population frequency in public databases. Finally, germ cell-specific expressions of BEND2 correspond to the developmental stage of the observed meiotic defect. Taken together, these observations are consistent with the contributory role of the BEND2 variant in spermatogenic failure.

However, several important questions remain for consideration and further exploration. First, owing to the scarcity of patient samples and the lack of an available specific antibody against human BEND2, we were unable to investigate the impact of this mutation at the protein level. For instance, we could not determine whether it affects the interaction of BEND2 with transcriptional suppressors (such as histone deacetylase 1 [HDAC1] and lysine demethylase 1A [KDM1A]) or its consequent role in repressing meiotic gene expression in a mouse model.24 Second, we attempted to explore this possibility in vitro by generating C-terminal enhanced green fluorescent protein (EGFP)-tagged WT and mutant constructs and transfecting them into Human Embryonic Kidney 293 cells expressing SV40 large T antigen (HEK293T). However, western blot analysis revealed no detectable expression for either construct, despite a robust signal from the empty vector control. After multiple transfection methods were systematically tested, consistent BEND2 expression remained unachievable. Given the testis-specific expression pattern of BEND2, we propose that its functional production likely requires transcriptional coregulators unique to germ cells in vivo. Third, the amino acid affected by this mutation (p.V357I) is not conserved across many species, such as mice. This lack of conservation limits our ability to further study the functional impact of this specific mutation using gene-edited mouse models.

In conclusion, we identified a novel pathogenic variant in the BEND2 gene that is potentially associated with NOA. These findings expand the known genetic architecture of male infertility and provide valuable information for future diagnostic and functional studies.

AUTHOR CONTRIBUTIONS

CHL performed the functional experiments and wrote the manuscript. CHL and ZQY analyzed the WES data and performed the bioinformatics analysis. HYW and HZ helped carry out the histological experiments. YHW and LFS provided clinical information. HCL, JFZ, and CX collected the samples. SWK and BBZ conceived the study and participated in its design and coordination. All authors read and approved the final manuscript.

COMPETING INTERESTS

All authors declare no competing interests.