Identification of the male infertility causal gene, DNAH11 and DNAH5
In this study, we performed WES and bioinformatics analysis of samples from multiple male infertile patients, following the previously established protocol. Among the analyzed samples, seven DNAH11 variants were identified in four unrelated individuals, and two DNAH5 variants were found in one individual. The DNAH11 (NM_003777.3) variants identified in individuals included a homozygous stop-gain variant c.9050G > A (p.Trp3017*) in AY0524, hailing from a consanguineous family. Additionally, compound heterozygous variants were observed: c.3470T > G (p.Leu1157Arg)/c.9790C > T (p.Pro3264Ser) in AY0540, c.100_101delinsTT (p.Glu34Leu)/c.6766A > G (p.IIe2256Val) in XFY03, and c.2419G > C (p.Asp807His)/c.8316 + 4A > C (splicing) in XFY04. All these variants were validated via PCR-Sanger sequencing and were inherited from their heterozygous parents (Fig. 1A), in accordance with an autosomal recessive inheritance pattern. As indicated in Table 1, all these seven variants were at low allele frequencies or unavailable in control sequence databases (1000G, gnomAD_exome_eas, gnomAD_genome_eas). They were also predicted to be deleterious by the SIFT, PolyPhen-2, MutationTaster, and CADD tools. Additionally, patient AY0435 harbored compound heterozygous variants of c.7915C > T (p.Arg2639*/c.4703G > A (p.Arg1568His) in DNAH5 (NM_001369). Both variants were identified as infrequent or absent within the general population and were predicted to be deleterious, as summarized in Table 1 and Fig. 1C.
Table 1
Bi-allelic Variants of DNAH11 and DNAH5 Variants Identified in Chinese Men
| DNAH11 Variants | DNAH5 Variant |
Subjects | AY0524 | AY0540 | XFY03 | XFY04 | AY0435 |
cDNA alteration | c.9050G > A | c.3470T > G | c.9790C > T | c.6766A > G | c.100_101GA > TT | c.2419G > C | c.8316 + 4A > C | c.7915C > T | c.4703G > A |
Variant allele | Hom | Het | Het | Het | Het | Het | Het | Het | Het |
Protein alteration | p.Trp3017* | p.Leu1157Arg | p.Pro3264Ser | p.IIe2256Val | p.Glu34Leu | p.Asp807His | - | p.Arg2639* | p.Arg1568His |
Variant type | Stop-gained | Missense | Missense | Missense | Missense | Missense | Splicing | Stop-gained | Missense |
Allele Frequency in Human Population |
1000G_eas | 0 | 0.001 | 0.002 | NA | 0 | NA | NA | 0 | 0 |
gnomAD_exome_eas | 0 | 0.0017 | 0.0009 | 0.0001 | 0 | 0.0046 | 0.0024 | 0 | 0 |
gnomAD_gnome_eas | 0 | 0.0025 | 0.0006 | - | 0 | 0.0070 | 0.0032 | 0 | 0 |
Function Prediction |
SIFT | NA | D | T | D | NA | D | NA | NA | D |
PolyPhen-2 | NA | D | D | P | NA | P | NA | NA | D |
MutationTaster | D | D | D | D | NA | D | NA | A | D |
CADD | 54 | 26.7 | 23.9 | 27 | NA | 13.53 | 18.36 | 43 | 27.6 |
RefSeq accession number of DNAH11 is NM_003777.3 and RefSeq accession number of DNAH5 is NM_001369. Abbreviations: NA, not available; 1KGP, 1000 Genomes Project; ExAc_all, all the data of Exome Aggregation Consortium; gnomAD, the Genome Aggregation Database; D, damaging, P, possibly damaging; T, tolerated. |
The DNAH11 gene encompasses 82 exons and is responsible for encoding the axonemal heavy chain dynein type 11 protein, primarily characterized by the presence of DHC_N1-2, P-loop_NTPase, MT, AAA_9, and Dynein_heavy domains. The identified stopgain variant (M1) and frameshift variant (M4) within DNAH11, as observed in this study, were forecasted to instigate premature termination codons, resulting in truncated proteins that adversely impact the majority of these domains. Furthermore, the affected amino acids in the remaining four missense variants exhibited a high degree of conservation across various species and were predicted to confer damage (Fig. 1B). The DNAH5 gene encodes the axonemal heavy chain dynein type 5 protein with 4624 amino acids, including DHC_N1-2, P-loop_NTPase, MT, AAA_9, and Dynein_heavy domains. The rare variant of DNAH5 identified in this study altered an evolutionarily conserved amino acid. They may result in a putative disruptive expression of the DNAH5 protein or disruption of the structure of critical domains (Fig. 1D).
To evaluate the pathogenicity of the identified DNAH11 variants, we investigated the localization and abundance of the DNAH11 protein in spermatozoa from both normal controls and individuals harboring DNAH11 variants through immunofluorescence staining. As depicted in Fig. 2A, DNAH11 signals were ubiquitously distributed along the entire sperm flagella in fertile controls, with a predominant concentration observed in the mid-piece. In contrast, the immunostaining pattern revealed that the predicted truncated DNAH11 proteins exhibited accumulation at the microtubule-organizing centers while being notably scarce throughout the flagella compartment in sperm from a subject (AY0524) carrying a homozygous stop-gain variant of DNAH11. Conspicuous deficiencies in DNAH11 protein levels were further discerned in samples obtained from another DNAH11-afflicted patient (AY0540). Simultaneously, DNAH11 signals were detected in samples from subjects harboring bi-allele variants in DNAH5 (AY0435), DNAH9, HYDIN, and DNALI1. Notably, DNAH11 staining was entirely absent in DNAH5−/− sperm. In contrast, no discernible differences in staining patterns were observed in sperm samples from subjects with DNAH9, HYDIN, and DNALI1 variants compared to normal sperm flagella (Fig. 2A and Fig S1).
Subsequently, we examined the DNAH5 staining pattern in both normal sperm and those deficient in DNAH5, DNAH11, and DNAH9. Notably, DNAH5 signals were almost absent in sperm deficient in both DNAH5 and DNAH11. While these signals were approximately comparable in the sperm flagella from the subject harboring DNAH9 variants with that in control. These findings suggest that the variants of DNAH5 and DNAH11 genes are pathogenic and cause male infertility. Moreover, the deficiency of DNAH11 results in the absence of DNAH5. The existence of DNAH11 protein in the sperm flagella was also influenced by DNAH5 deletion, not other ODA-related components, such as DNAH9 and DNALI1.
Biallelic variants of DNAH11 and DNAH5 variants cause oligoasthenoteratozoospermia
Semen analyses were carried out in the laboratories during routine examination according to World Health Organization (WHO) guidelines for the individuals. Notably, the spermatozoa of subjects AY0524, AY0540, XFY03, XFY04 and AY0435 presented with very low or almost no progressive motility and the total numbers of spermatozoa were also decreased in these patients (Table 2).
Table 2
Semen Routine Parameters and Sperm Morphology in Men Harboring DNAH11 and DNAH5 Variants
Genes | DNAH11 Variants | DNAH5 Variant | Reference Limits |
Subjects | AY0524 | AY0540 | XFY03 | XFY04 | AY0435 |
Semen Parameter |
Semen volume (mL) | 4.8 | 4.0 | 2.2 | 3.5 | 1.9 | > 1.5 |
Semen concentration (106/mL) | 7.5 | 14.1 | 9.8 | 6.4 | 5.6 | > 15.0 |
The total number of sperm (106) | 36 | 56.4 | 21.6 | 22.4 | 10.6 | > 39.0 |
Progressive motility (%) | 28.9 | 4.5 | 0 | 5.0 | 0 | > 32.0 |
Sperm Morphology |
Sperm Tail |
Normal tail(%) | 30.5 | 32 | 28.5 | 35 | 22.2 | > 23.0 |
Absent flagella (%) | 2.8 | 8 | 10.5 | 4.5 | 7.4 | < 5.0 |
Short flagella(%) | 8.5 | 9.3 | 7.5 | 12 | 38.9 | < 1.0 |
Coiled flagella(%) | 51.3 | 45.5 | 40 | 42.5 | 28.1 | < 17.0 |
Angulation (%) | 3.5 | 3.9 | 8.5 | 3 | 3.4 | < 13.0 |
Irregular caliber (%) | 3.4 | 1.3 | 5 | 3 | 0 | < 2.0 |
PCD-related phenomenon |
Rhinosinusitis | No | No | No | No | No | |
Wet cough | Yes | No | No | No | Yes | |
Otitis media | No | No | No | No | No | |
Bronchiectasis | Yes | No | No | No | Yes | |
Situs inversus | No | No | No | No | No | |
Congenital heart disease | No | No | No | No | No | |
The morphology of sperm samples was evaluated using Hematoxylin-eosin (H&E) staining and SEM. Spermatozoa from fertile males exhibited a typical morphology characterized by a normal head and long, smooth flagella. Conversely, spermatozoa from individuals harboring DNAH11 and DNAH5 variants predominantly displayed coiled and shortened flagella. Notably, various abnormalities were evident in the mid-pieces of these spermatozoa, including the accumulation or severe reduction of disordered material, bare axonemes devoid of a surrounding mitochondrial sheath, or mid-pieces with shortened mitochondrial sheaths. Furthermore, SEM analysis revealed that abnormal necks and coiled flagella were the most frequently observed defects in these cases (Fig. 3).
Additionally, we investigated the ultrastructure of spermatozoa from both fertile men and individuals carrying DNAH11 and DNAH5 variants, including AY0540, AY0524, and AY0435, utilizing TEM. In the spermatozoa of normal controls, the ultrastructure of flagella exhibited a well-organized "9 + 2" microtubule arrangement, comprising a central pair of microtubules surrounded by nine peripheral doublets, outer dense fiber (ODF), and either a mitochondrial sheath (MS) in the mid-piece or a fibrous sheath (FS) in the principal piece, with only bare "9 + 2" structures at the end-piece. Conversely, in the spermatozoa of patients, cross-sectional analysis revealed significant defects, including the absence of portions of the MS, CP, radial spoke (RS), and ODAs. This was concomitant with pronounced axonemal disorganization, coupled with structural defects in the peri-axonemal region (Fig. 4). Moreover, our previous study demonstrated that the ultrastructures of the mitochondrial sheath and flagellum in spermatozoa from a subject with DNAH9 variants were nearly normal [26]. Collectively, these findings suggest a crucial role for DNAH11 and DNAH5 in the normal assembly process of the sperm flagellar structure.
Confirmation of spermatozoa flagella component defects in subjects with DNAH11 and DNAH5 variants
In the TEM observations, we found various defects in the sperm neck and flagellum structure in patients harboring DNAH11 and DNAH5 variants. To delineate the molecular implications of these variants in human sperm, we performed IF and WB to investigate the presence, localization, and levels of specific components within the substructures of the sperm flagellum. These components included SPAG6 (a marker for CP), RSPH4A (a constituent of the RS complex), DNAI2 (a component of the ODAs), and TOMM20 as a marker of mitochondrial sheath (MS). As illustrated in Fig. 5A, B, and C, the signals of DNAI2, SPAG6, and RSPH4A were predominantly concentrated along the mid-piece and principal piece of normal sperm flagella. In stark contrast, DNAI2 and SPAG6 immunostainings were almost entirely absent in the sperm flagella of all three subjects with bi-allelic DNAH11 or DNAH5 variants (AY0540, AY0524, and AY0435). RSPH4A signal was accumulated at the sperm neck instead of the whole sperm flagella in patients. In normal sperm, TOMM20 staining signals were localized outside the mid-piece, roughly equivalent in length to the sperm head. Conversely, these corresponding staining signals were nearly absent in spermatozoa from AY0540 and AY0435, and shortened or absent in sperm from AY0524.We also detected the signals of DNAI2, SPAG6, RSPH4A and TOMM20 in DNAH9-deficient spermatozoa collected from our previous study [26]. Remarkably, these signals closely resembled those in fertile controls. Furthermore, we performed WB to examine the levels of DNAI2, SPAG6, RSPH4A and TOMM20 proteins in spermatozoa from subjects harboring DNAH11 (AY0540, AY0524, XFY03 and XFY04), DNAH5 (AY0435) and DNAH9 variants. As shown in Fig. 5E, the levels of these proteins were down-regulated dramatically in spermatozoa from subjects with DNAH11 and DNAH5 variants, except for the level of SPAG6 in XFY03. However, the corresponding proteins levels in DNAH9-deficient sperm were not affected. Last, we also examined the presence of inner dynein arms (IDA) related proteins DNAH1 and DNAH6 in sperm from these subjects and their distribution ware not affected by the absence of DNAH11, DNAH5 and DNAH9 (Fig. S2).
In summary, these findings substantiate the potential contribution of DNAH11 and DNAH5 variants to structural defects in sperm flagella, particularly affecting the ODAs and the adjacent structures in the vicinity of sperm flagella. This suggests a plausible pathological mechanism underlying male infertility. In contrast, our observations indicate that DNAH9 is dispensable for the assembly of these flagellar structures.
ICSI treatment outcomes of subjects harboring DNAH11 and DNAH5 variants
ICSI treatment has proven to be an efficacious intervention for addressing male infertility. To assess the viability of ICSI as a solution for males affected by DNAH11 and DNAH5 variants, four out of six couples underwent ICSI procedures. As outlined in Table 3, the rates of fertilization, cleavage, eight-cell formation, and blastocyst formation following ICSI were almost normal for couples with DNAH11 deficient, resulting in clinical pregnancies for two of them.
Table 3
The ICSI clinical outcomes of subjects with DNAH11/DNAH5 variants.
Genes | DNAH11 Variants | DNAH5 Variant |
No. of couples | AY0524 | AY0540 | XFY03 | XFY04 | AY0435 |
Mean male age (years) | 31 | 34 | 28 | 28 | 33 |
Mean female age (years) | 29 | 29 | 30 | 29 | 30 |
No. of ICSI cycles | 1 | 1 | / | 1 | 1st | 2nd |
No. of oocytes retrieved | 38 | 17 | / | 21 | 23 | 16 |
No. of oocytes injected | 23 | 11 | / | 21 | 15 | 16 |
Fertilization rate(%) | 65.2(15/23) | 63.6(7/11) | / | 100(21/21) | 0 | 68.8(11/16) |
Cleavage rate (%) | 100(15/15) | 100(7/7) | / | 100(21/21) | 0 | 90.9(10/11) |
8-Cell formation rate(%) | 60 (9/15) | 85.7(6/7) | / | 28.6 (6/21) | 0 | 36.4(4/11) |
Blastocyst formation rate (%) | 60(9/15) | 85.7(6/7) | / | 28.6(6/21) | 0 | 36.4(4/11) |
High quality blastocyst rate(%) | 77.8(7/9) | 66.7(4/6) | / | 28.6(6/21) | 0 | 18.2(2/11) |
No. of transfer cycles | 1 | 1 | / | 1 | / | 2 |
Number of embryos transferred per cycle | 1 | 1 | / | 1 | / | 2 |
Implantation rate (%) | 100(1/1) | 100(1/1) | / | 0 | / | 0 |
Clinical pregnancy rate per transfer cycle (%) | 100(1/1) | 100(1/1) | / | 0 | / | 0 |
Miscarriage rate(%) | 0 | 0 | / | 100 | / | 100 |
The partner of AY0435 harboring DNAH5 variants underwent two ICSI treatment cycles. In the first cycle, 23 oocytes were retrieved, with 15 metaphase II (MII) oocytes subjected to microinject, yet none successfully fertilized. Subsequently, in the second ICSI treatment cycle, 16 oocytes underwent microinjection, resulting in successful fertilization of 11 oocytes. Four blastocysts were obtained and transferred in two cycles, but unfortunately, none of them resulted in implantation. Regrettably, data regarding the assisted reproductive cycles of XFY03 were unavailable. Therefore, our study suggests that ICSI is a viable clinical option for patients with DNAH11 variants. However, for DNAH5-associated male infertility, conclusive evidence necessitates a larger cohort of DNAH5-deficient patients to provide definitive answers.