Novel frameshift mutation in STK33 is associated with asthenozoospermia and multiple morphological abnormalities of the flagella

Abstract

Serine/threonine kinases domain-containing proteins are known to play important functions in sperm flagella and male fertility. However, the roles of these proteins in human reproduction remain poorly understood and whether their variants are associated with human asthenozoospermia have not been reported. Here, we recruited a Pakistani family having four infertile patients diagnosed with idiopathic asthenozoospermia without any ciliary-related symptoms. Whole-exome sequencing identified a novel homozygous frameshift mutation (c.1235del, p.T412Kfs^*14) in serine/threonine kinase 33 (STK33), which displays a highly conserved and predominant expression in testis in humans. This variant led to a dramatic reduction of STK33 messenger RNA (mRNA) in the patients. Patients homozygous for the STK33 variant presented reduced sperm motility, frequent morphological abnormalities of sperm flagella and completely disorganized flagellar ultrastructures, which are typical for multiple morphological abnormalities of the flagella (MMAF) phenotypes. Overall, these findings present evidence establishing that STK33 is an MMAF-related gene and provide new insights for understanding the role of serine/threonine kinases domain-containing proteins in human male reproduction.

Introduction

Asthenozoospermia is defined as less than 32% of spermatozoa with progressively motility in the ejaculate (1). Multiple morphological abnormalities of the flagella (MMAF) is a severe form of asthenozoospermia and is characterized by immotile spermatozoa showing a mosaic of severe flagellar abnormalities, including absent, short, coiled, bent and irregular flagella (2). MMAF is mainly caused by ultrastructural defects in the flagellum, including the missing and/or disorganized microtubule doublets, dynein arms, outer dense fibers, fiber sheath or mitochondrial sheath. To date, 26 genes have been reported responsible for MMAF, including CFAP43 (MIM: 617558), CFAP44 (MIM: 617559), CFAP58 (MIM: 619129), DNAH1 (MIM: 603332), DNAH8 (MIM: 603337), DNAH17 (MIM: 610063), QRICH2 (MIM: 618304) and so on (3–12), indicating that MMAF is a genetically heterogeneous disorder. However, these genetic findings account for approximately 60% of the MMAF cases (7,8), and more genetic factors need to be studied to understand the pathogenesis of MMAF thoroughly.

Serine/threonine kinase proteins, which phosphorylate specific sites on a broad spectrum of target substrates, play important roles in regulating a variety of cellular processes. Several studies have reported that serine/threonine kinases domain-containing proteins play important roles in spermatogenesis and male fertility in mice. Tssk1, Tssk2, Tssk4 and Sstk (also known as Tssk6) are specifically expressed in mouse testis (13), and disruptions of these genes in mice cause seriously decreased sperm motility and marked morphological abnormalities, resulting in male infertility or subfertility (14–16). STK33 (MIM: 607670) is another serine/threonine kinases domain-containing protein that is required for male fertility in mice (17,18). Unlike the testis-specific expression pattern for those proteins mentioned earlier, STK33 is widely expressed with predominant expression in testis, ovary, lung, trachea and embryonic organs (19,20). Stk33 knockout in mouse testis results in malformation of the manchette, subsequent severely disorganized tail structures and sperm morphological abnormalities, consequently leading to male infertility (21). These findings indicate that serine/threonine kinases domain-containing proteins are important for functional sperm flagella and male fertility. However, the roles of these proteins in human reproduction remain poorly understood and whether their variants are associated with human asthenozoospermia has not been reported.

In this study, through genetic analyses of a Pakistani family having four infertile patients exhibiting asthenozoospermia without any ciliary-related symptoms, we identified a novel homozygous frameshift variant in STK33 (c.1235del, p.T412Kfs^*14) which led to dramatic reductions of STK33 mRNA in the patients. Patients homozygous for this variant exhibited multiple morphological and ultrastructural anomalies in spermatozoa, which are typical for MMAF phenotypes. These findings provide first genetic evidence that STK33 is an MMAF-related gene and that its homozygous frameshift variant is associated with asthenozoospermia and male infertility in humans.

Results

Four asthenozoospermia patients from a Pakistani family

This study was conducted on a Pakistani family with idiopathic male infertility (Fig. 1A). Among the six brothers, one had three children; the other five, III:1 (41-year-old), III:3 (36-year-old), III:4 (31-year-old), III:5 (32-year-old) and III:6 (30-year-old), had been married for 17, 14, 4, 5 and 6 years, respectively, but all were infertile. All of the five brothers had normal karyotypes (46; XY) and had no large-scale deletions in Y chromosomes. Patient III:4 is an infertile case with normal semen parameters, including semen volume and sperm concentration, morphology, motility and progressive motility (Table 1 and Supplementary Material, Fig. S1A) (1). The semen volume and sperm concentration for III:3, III:5 and III:6 were within the normal ranges according to the World Health Organization (WHO) guidelines, and III:1 had normal semen volume and low sperm concentration (Table 1). The spermatozoa of patients III:1, III:3, III:5, and III:6 all presented low motility (10.2 ± 1.6, 34, 25.3 ± 2.6 and 40%; respectively) and progressive motility (2.8 ± 1.6, 20, 9.3 ± 0.7 and 10%; respectively; Table 1). Hence, patients III:1, III:3, III:5 and III:6 were diagnosed with asthenozoospermia.

Figure 1

A homozygous STK33 variant identified in a Pakistani family with asthenozoospermia. (A) Pedigrees of the family (PK-INF-785) with four infertile brothers with asthenozoospermia (marked by black filled boxes) and one infertile brother with unknown cause (marked by a gray filled box). Arrowheads point to the five individuals for whom WES was performed. Slashes denote deceased family members. (B) Representative chromatograms of the STK33 variant (NC_000011.9:g.8435151del) in genomic DNA from all the available family members. M, male; F, female. (C) Schematic showing the position of the STK33 variant at transcript level (NM_001352388.1:c.1235del) and protein level (NP_001339317.1:p.T412Kfs^*14). (D) Quantitative real-time PCR analysis of STK33 mRNA expression in peripheral blood samples from all the available family members. β-Actin was used as an internal control. WT, wild-type allele; MT, the mutant allele.

Table 1

Clinical characteristics of patients

	Reference values	III:1	III:3	III:4	III:5	III:6
Fertility	—	Infertile	Infertile	Infertile	Infertile	Infertile
Infertility phenotype		Asthenozoospermia and MMAF	Asthenozoospermia and MMAF	Unknown	Asthenozoospermia and MMAF	Asthenozoospermia
Mutation	—	MT/MT	MT/MT	WT/MT	MT/MT	MT/MT
Height/weight (cm/kg)	—	176.0/72.0	176.0/72.0	177.0/73.0	172.0/65.0	174.0/68.0
Age (years)a	—	41	36	31	32	30
Years of marriagea	—	17	14	4	5	6
Semen parameters
Semen volume (ml)	>1.5	2.2 ± 0.4	2.5	2.8 ± 0.3	3.0 ± 1.0	2.5
Sperm concentration (106/ml)	>15	6.6 ± 1.6	20.0	46.0 ± 16.0	43.7 ± 18.1	40.0
Normal sperm morphology (%)	>4	0	5.3	57.5	3.3	—
Motile sperm (%)	>40	10.2 ± 1.6	34.0	67.5 ± 22.5	25.3 ± 2.6	40.0
Progressively motile sperm (%)	>32	2.8 ± 1.6	20.0	47.5 ± 17.5	9.3 ± 0.7	10.0
Sperm flagella
Morphologically normal (%)	—	0	12.0	61.5	3.3	—
Absent (%)	—	31.8	5.3	2.0	40.0	—
Short (%)	—	4.5	4.7	10.0	30.0	—
Coiled (%)	—	36.4	26.0	25.0	3.4	—
Bent (%)	—	18.2	32.0	0	10.0	—
Irregular caliber (%)	—	9.1	20.0	1.5	13.3	—

Reference values were published by WHO (2010). For III:1, III:4 and III:5, semen volume, sperm concentration and sperm motility were examined at least two times. Sperm morphology was assessed with at least 200 spermatozoa examined for III:1, III:3, III:4 and III:5. Data are presented as mean ± SEM. WT, the wild-type allele; MT, the mutant allele (c.1235del, p.T412Kfs^*14).

^aAt manuscript preparation (2021).

These four asthenozoospermia-affected patients in our family claimed not having any ciliary-related clinical features, tumors, cancer or any other diseases, and thus they did not wish to participate in any further related examination. Besides, the wives of all the patients declared having a normal onset of puberty and normal menstrual cycles, with no history of ovarian injury or abortion.

A homozygous SKT33 frameshift variant identified in patients

To uncover any potential genetic cause of asthenozoospermia in this family, we performed whole-exome sequencing (WES) for patients III:1, III:5 and III:6, their fertile brother III:2 and their mother II:2. After filtering of variants and Sanger sequencing of all available family members, we identified a homozygous frameshift variant in STK33, NM_001352388.1:c.1235del, cosegregating with asthenozoospermia in this family (Fig. 1B and Supplementary Material, Fig. S2). This variant occurred in exon 11 and was predicted to cause amino acid alteration from position 412 and introduce premature translational termination of STK33 at amino acid position 425 (Fig. 1C), which would either lead to nonsense-mediated decay or produce a truncated protein.

According to the GeneCards database, STK33 messenger RNA (mRNA) is expressed in human bloods, we thus examined the levels of STK33 transcripts in the peripheral blood samples from available family members by quantitative real-time PCR assays. The STK33 mRNA levels in the STK33 variant homozygous carriers III:3, III:5 and III:6 were reduced to 0.2–22.7% of those in wild-type members (Fig. 1D), revealing nonsense-mediated mRNA decay triggered by premature translational termination. Given that targeted disruption of Stk33 results in reduced sperm number, immotile sperm and male infertility in mice (21), the homozygous frameshift variant in SKT33 is likely to be pathogenic for asthenozoospermia in this family.

The SKT33 variant leads to MMAF and severe disorganized flagellar structure

The morphological abnormalities of spermatozoa were assessed by hematoxylin and eosin (H&E) staining on semen smears. The percentages of morphologically normal sperm in III:1 and III:5 (0 and 3.3%, respectively) fell below the reference value suggested by WHO (1), while 5.3% of spermatozoa in III:3 had normal morphology (Table 1). Multiple morphological defects of the sperm flagella were identified, including absent, short, coiled and bent flagella and flagella of irregular caliber, which are typical morphological presentations of MMAF, collectively accounting for 88.0–96.7% of the spermatozoa examined from homozygous STK33 variant carriers, III:1, III:3 and III:5 (Fig. 2A and Table 1). Scanning electronic microscopy (SEM) analyses of patients’ spermatozoa revealed similar MMAF phenotypes (Fig. 2B), which are consistent with those of light microscopy.

Figure 2

Sperm morphology in the patients carrying the homozygous STK33 variant. (A) Representative images of spermatozoa showing sperm morphological abnormalities, including absent (ii), short (iii), coiled (iv) and bent (v) flagella observed in patients III:1 and III:5 after H&E staining. A representative spermatozoon with normal morphology from the fertile control was shown (i). Scale bars: 5 μm. (B) Representative SEM micrographs of spermatozoa from the fertile control and patient III:5. Scale bars: 5 μm.

We next investigated the ultrastructure of spermatozoa from patient III:5 by transmission electronic microscopy (TEM). Typical ‘9 + 2’ axoneme structure was observed in the fertile control (Fig. 3A). However, in patient III:5, 100 and 95% of all cross-sections of sperm flagella at midpiece and principal piece/end piece, respectively, displayed disorganization in axonemal or peri-axonemal ultrastructures, including abnormal, dispositioned and/or missing peripheral microtubule doublets, inner/outer dynein arms, central pair and outer dense fibers (Fig. 3), which is consistent with the flagellar defects detected after Stk33 knockout in testes.

Figure 3

Sperm flagella ultrastructures in the patients carrying the homozygous STK33 variant. (A) Representative images of flagellar cross-sections from the fertile control and patient III:5. Scale bars: 500 nm. ODF, outer dense fibers (red arrows); MTD, microtubule doublets (blue arrows); CP, central pair of microtubules (green arrows); ODA, outer dynein arm (yellow arrows); IDA, inner dynein arms (pink arrows). (B) Quantification of flagellar cross-sections with abnormal ultrastructures. n, the number of cross-sections analyzed.

In summary, patients homozygous for the variant in STK33 displayed low sperm motility, abnormal morphology and disorganized flagellar ultrastructure, which is consistent with the findings after Stk33 knockout in testes (21), proving that the STK33 variant identified in our patients is indeed pathogenic for MMAF and male infertility in this family.

Discussion

In this study, based on the genetic analysis of a Pakistani family with MMAF, we report a novel homozygous frameshift variant in STK33 as a pathogenic variant for sperm morphological and flagellar ultrastructural anomalies, thus first establishing STK33 as an MMAF-related gene.

Recently, Martins et al. (21) reported that knockout Stk33 in mouse testes results in severely malformed and immotile spermatozoa and thus led to male infertility. In our family, patients carrying a homozygous STK33 variant presented reduced sperm motility, multiple sperm malformations and totally disorganized flagella, which are consistent with the phenotypes in mice with Stk33 knockout in testes. The pathogenicity of the homozygous STK33 variant was further supported by the observations in the infertile brother (III:4), who was heterozygous for the STK33 variant but had normal sperm motility. Moreover, 61.3% of sperm flagella from III:4 were morphologically normal and TEM analysis revealed that 11 out of 26 flagellar cross-sections examined showed normal ultrastructure (Table 1 and Supplementary Material, Fig. S1), indicating the STK33 variant recessively cosegregated with the asthenozoospermia/MMAF phenotype in this family. These findings also indicate that the etiology of infertility for patient III:4 is not likely same as that for the other four asthenozoospermia/MMAF patients but may result from defects in fertilization or embryo implantation or development.

In both humans and mice (21) mutant for STK33, typical MMAF phenotypes were observed, indicating the essential role of STK33 in sperm flagella. However, the phenotype of the knockout mouse model appears to be more severe than that in patients. In the Stk33^−/− mice, the sperm count was significantly reduced, all the sperm were immotile and morphologically abnormal (short, often coiled and abnormally shaped tails) with disorganized flagellar ultrastructures (21). Whereas, among the MMAF patients examined in this study, only one patient displayed sperm concentration and sperm morphology which were lower than the reference values suggested by WHO. Motile or progressively motile sperm were detected in all of the four MMAF patients, though their percentages were lower than the reference values. One potential explanation is that, in patients, there may be some remnant mutant STK33 proteins, which were produced by the remaining minimum mRNA, resulting in a milder phenotype than that in the Stk33 knockout mice.

Besides the typical morphological anomalies under light microscopy, ultrastructural defect in the flagellar axoneme is the other typical presentation of MMAF. Recent studies revealed that MMAF-related proteins [except ARMC2 (MIM: 618424), the localization of which is unknown] are either structural components of sperm flagella, such as ODA component DNAH17 and DNAH8, CP component SPEF2 (MIM: 610172), or proteins localizing to sperm flagella, such as CFAP47 (MIM: 301057) and CFAP58, intraflagellar transport complex TTC21A (MIM: 618429) or protein degradation complex QRICH2 (5,7–9,22–29). It is not surprising that disruption in these proteins cause sperm flagella anomaly. However, STK33 protein is not reported in any dataset from sperm flagellum proteome in humans and mice (30,31). We did not detect the expression of STK33 in human and mouse spermatozoa by western blotting in sperm lysates. A previous study has revealed that STK33 protein localizes to the cytoplasm and partially co-localizes with the caudal end of the manchette in elongating spermatids but could not be detected on the sperm flagella in mice (21). The manchettes in Stk33-null spermatids were abnormally elongated and straightened and lacked the basket-like distal termination, indicating that STK33 proteins play a role in manchette formation or function. Therefore, our findings suggest that although some proteins are neither structural components of the flagella nor located to the sperm flagella, their disruptions could also impair the generation of a functional sperm tail and could cause an MMAF phenotype.

Since sperm flagella and cilia share a similar axonemal structure, mutations identified in many MMAF-related genes, which encode axonemal components present in both cilia and flagella, were also found to cause primary ciliary dyskinesia (PCD; MIM: 244400) (27,32–34). However, many of these genes were also found dispensable for cilia function, and not all mutations in PCD genes are associated with male infertility (5,7,8,22,28,35), indicating that the functions of these axonemal genes are likely different in cilia and sperm tails. STK33 expression is highest in the testes among various human tissues, but it is also detected in ciliated tissues (19). However, all of the four patients homozygous for the STK33 variant in our family claimed not having any clinical symptoms related to PCD, except male infertility, indicating that the homozygous STK33 frameshift variant probably does not cause PCD but only a sperm-specific phenotype. Nonetheless, we cannot conclude that STK33 is dispensable for cilia function, as some remnant mutant STK33 mRNAs, which retain the functional domain, were detected in patients which may be enough for sustaining normal STK33 functions in cilia. In addition, whether MMAF-related genes could cause PCD appears to be different across different populations. Mutations in genes which are expressed in both ciliated tissues and testes, for example, SPEF2, DNAH1, DNAH6, etc. were found to cause MMAF without any obvious PCD symptoms in some studies, but their mutations were also reported in infertile men with PCD or PCD-like phenotypes (2,25,27,32–34,36–38). Since our findings are based on four patients originated from only one family, whether STK33 is required in cilia needs to be ascertained by knockout animal models and in more patients.

In conclusion, our current investigation revealed a novel frameshift STK33 variant in a Pakistani family diagnosed with male infertility owing to asthenozoospermia and MMAF, establishing STK33 as an MMAF-related gene. These results offer a further understanding on the complex etiology and pathogenesis of MMAF, which would be important for genetic counseling, diagnosis and prenatal screening of MMAF.

Materials and Methods

Editorial policies and ethical considerations

This study was approved by the Ethical Committee of University of Science and Technology China (USTC). Written informed consent forms from all participants were obtained at the beginning of the study.

Participants and clinical examination

In this study, we recruited a Pakistani family (registered number at the Human Reproductive Disease Resource Bank at USTC, PK-INF-785), including four infertile brothers diagnosed with asthenozoospermia and one infertile brother with unknown cause. According to the WHO guidelines (1), semen volume, sperm concentration and sperm motility were examined at least two times for III:1, III:4 and III:5. During their medical consultation, all patients answered a questionnaire regarding clinical presentations of PCD symptoms.

H&E staining was performed to assess sperm morphology. Briefly, semen smears were fixed in 4% paraformaldehyde at room temperature for 5 min. After being washed in PBS, the samples were stained with hematoxylin, followed by dehydrated and eosin staining. At least 200 spermatozoa were examined for III:1, III:3, III:4 and III:5.

WES and variant filtering

Genomic DNA was extracted from peripheral blood samples of all the available family members using the FlexiGene DNA kit according to the manufacturer’s protocol (QIAGEN, 51206, Valencia, CA). WES of patients III:1, III:5 and III:6, their fertile brother III:2 and their mother II:2 was conducted as we previously described (39). Variants following the recessive inheritance were kept for further screening according to the strategy we described previously (23). Supplementary Material, Figure S2 illustrates the bio-informatic pipeline of the filtering process. Sequences of primers used for Sanger sequencing are: forward 5′-GGAGACTGTCTCGCTAAATG-3′, reverse 5′-TGCCCTACATAATCACCAAC-3′.

EM analysis

Spermatozoa were fixed in 0.1 M phosphate buffer (pH 7.4) containing 4% glutaraldehyde, 4% paraformaldehyde and 0.2% picric acid at 4°C for at least overnight. For SEM analysis, samples were post-fixed in 1% osmic acid for 1.5 h, dehydrated with gradient ethanol at 30, 50, 70, 90, 95 and 100%, dripped on specimen stage, air-dried, sputter-coated and then analyzed by the Field Emission Scanning Electron Microscope (Nova NanoSEM 450, Thermo Fisher Scientific, Waltham, MA). TEM analysis was performed as we previously described (23,24).

RNA extraction and quantitative real-time PCR

Total RNA extraction from peripheral blood samples, cDNA synthesis and quantitative real-time PCR were conducted as described previously (23). Sequences of primers are as follows: for STK33, forward 5′-TCCTGCTCACAGAATCACAGC-3′ and reverse 5′-CAACTTTTCTTCAGTGGACGGC-3′; for ACTB, forward 5′- AATGAGCTGCGTGTGGCTC-3′ and reverse 5′-ATAGCACAGCCTGGATAGCAAC-3′.

Acknowledgements

We are grateful to all the participants for their cooperation. We thank Li Wang and Dandan Song at the Center of Cryo-Electron Microscopy, Zhejiang University, for their technical assistance on SEM and TEM, respectively. We also thank the Bioinformatics Center of the USTC, School of Life Science for providing supercomputing resources.

Conflict of Interest statement. None declared.

Funding

This study was supported by the National Key Research and Developmental Program of China (2018YFC1003900 and 2019YFA0802600 to H.M., and 2016YFC1000600 to Q.S.), the National Natural Science Foundation of China (31630050, 31890780 and 32061143006 to Q.S., 31871514 to X.J.), the Strategic Priority Research Program of the Chinese Academy of Sciences (XDB19000000 to Q.S.).

Author notes

These authors contributed equally to this work.