Novel mutations in FSIP2 cause male infertility through multiple morphological abnormalities of the sperm flagella

Abstract

Infertility is a global concern, and oligoasthenoteratozoospermia (OAT) is the most severe form of male infertility, characterized by reduced sperm count, decreased motility, and increased abnormal morphology. Multiple morphological abnormalities of the sperm flagella (MMAF) characterize the most severe type of OAT and are usually caused by loss-of-function mutations in the genes essential for vital aspects of sperm biology, including concentration, motility, and morphology. The fibrous sheath interacting protein 2 (FSIP2) plays an essential role in sperm flagellar structure and function by regulating such processes as intraflagellar transport and acrosome formation. The present study, employing whole-exome sequencing (WES), identified two FSIP2 mutations in one patient (patient 1), a homozygous missense (c.262C>A, p.P88T) and a homozygous frameshift mutation (c.10948_10951del, p.N3653Nfs*22), as well as a homozygous FSIP2 frameshift mutation (c.15982_15982del, p.I5328Lfs*33) in another patient (patient 2). The results of bioinformatics analysis indicate that the identified missense mutation (c.262C>A) is rare and predicted to have a deleterious effect on FSIP2. Transmission electron microscopy analysis of sperm revealed several abnormalities, including a disorganized mitochondrial sheath, absence of the central pair and some doublets of microtubules, and significant dysplasia of the fibrous sheath. Reverse transcription-polymerase chain reaction (RT-PCR) indicated significantly reduced FSIP2 messenger RNA (mRNA) levels in sperm lysate of the affected individuals. Immunofluorescence staining revealed a complete absence of FSIP2, A-kinase anchor protein 4 (AKAP4), sperm-associated antigen 6 (SPAG6), intraflagellar transport 20 (IFT20) and actin-like 7A (ACTL7A) proteins in the spermatozoa of patients. Thus, the novel FSIP2 variants identified in patient 1 and patient 2 are recognized as pathogenic mutations responsible for MMAF, providing valuable insights for genetic counseling and reproductive decision-making in affected males.

INTRODUCTION

The World Health Organization (WHO) estimates that the global incidence of infertility ranges from 8% to 15%, with male factors contributing to approximately 50% of all infertile cases. In many such cases, genetic factors play a significant role.1,2 One of the most severe clinical phenotype of male infertility, oligoasthenoteratozoospermia (OAT), is characterized by reduced sperm concentration (<16×10⁶ ml⁻¹), less than 30% progressively motile sperm, and less than 4% morphologically normal sperm.3,4,5

The human sperm flagellum, a highly specialized structure essential to sperm motility and male fertility, is divided into three distinct regions based on accessory structures surrounding the axoneme: the midpiece, principal piece, and end piece.6 Genetic defects in these axonemal and periaxonemal structures can lead to various sperm flagellar abnormalities, often resulting in conditions like asthenoteratozoospermia, characterized by reduced sperm motility and abnormal morphology. Studies show that disruptions in the genes encoding axonemal components of the sperm flagella, such as those in the dynein axonemal heavy chain (DNAH) and cilia and flagella associated protein (CFAP) gene families, are frequently associated with structural and functional abnormalities of the sperm flagella.7,8,9,10

The fibrous sheath interacting protein 2 (FSIP2) gene, located on chromosome 2q32.1, encodes a 6907-amino acid protein critical to the structural integrity of the fibrous sheath of sperm flagella.11 In 2018, FSIP2 mutation was first associated with MMAF and male infertility.12 Furthermore, FSIP2 interacts with A-kinase anchor protein 4 (AKAP4), a pivotal protein in fibrous sheath assembly that regulates sperm motility and structural stability. Moreover, similar defects in sperm motility and morphology have been observed in cases with FSIP2 and AKAP4 mutations underscoring their interconnected roles.12,13 A recent study has revealed compound heterozygous variants in FSIP2 that disrupt fibrous sheath assembly and lead to axonemal defects, indicating the role of FSIP2 in the structural organization of sperm flagellum.14

Whole-exome sequencing (WES) is a high-throughput method that focuses on the exonic regions of the genome, where most protein-coding genes are located. This technology efficiently identifies genetic polymorphisms and mutations, providing a valuable tool for uncovering genetic variations associated with diseases.15 Multiple morphological abnormalities of the sperm flagella (MMAF) is a condition categorized by significant morphological aberrations in the sperm flagella, such as flagella that are short, absent, coiled, bent, or of irregular caliber, which lead to male infertility.16 These flagellar defects significantly impair sperm motility, thereby decreasing the chances of successful fertilization. MMAF is mainly associated with genetic changes that affect the proper formation and function of the flagella. Several genes, including AKAP3 (Online Mendelian Inheritance in Man [OMIM]: 604689) and AKAP4 (OMIM: 300185), are crucial to sperm flagellar structure and function, particularly in maintaining the integrity of the sperm fibrous sheath. Additionally, mutations in other genes, such as DNAH1, CFAP43, DNAH17, CFAP44, CFAP69, CFAP61, CFAP91, CFAP47, CFAP251, CFAP65, CFAP70, Serine/threonine kinase 33 (STK33), Armadillo repeat containing 2 (ARMC2), and ARMC3, have been associated with MMAF, which significantly impairs sperm motility and male fertility.7,10,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31 However, genetic studies explain only around 60% of MMAF cases, so further research is needed to discover new genetic factors that contribute to MMAF.32

Herein, genetic analysis identified the homozygous FSIP2 mutations (c.262C>A, p.P88T; and c.10948_10951del, p.N3653Nfs^*22) in patient 1 (P1) and the homozygous frameshift mutation c.15982_15982del (p.I5328Lfs^*33) in patient 2 (P2). Significantly reduced FSIP2 mRNA levels were detected in the patients’ sperm lysate, indicating the association of FSIP2 mutations with MMAF and male infertility. This study expands our understanding of the genetic basis of MMAF by identifying novel FSIP2 variants, thereby extending the mutational spectrum associated with male infertility.

PARTICIPANTS AND METHODS

Clinical investigation

Two consanguineous Pakistani families (ID: PK-INF-1062 and PK-INF-1154, registered in the Human Reproductive Disease Resource Bank at University of Science and Technology of China [USTC], Hefei, China) were recruited for this study. Patients from both families were diagnosed with primary infertility and exhibited oligoasthenoteratozoospermia. The patients, despite being married and engaging in unprotected sexual intercourse with ejaculation, were unable to conceive after several years of marriage. During sampling, the testicular volume of patients was measured using ultrasound to provide additional clinical data regarding their condition. Written informed consent was obtained from all available family members. This study was conducted with the approval of the institutional ethical committee of the USTC (Approval No. USTCEC202000003) and Gomal University (Dera Ismail Khan, Pakistan; Approval No. 04/ERB/GU).

Semen and sperm morphology analysis

Semen analysis was performed following a minimum of 5 days of sexual abstinence. Sperm concentration was determined using a counting chamber, while motility was assessed through direct microscopic observation. Both sperm concentration and motility were evaluated according to the guidelines provided by the 6^th edition of the World Health Organization (WHO) laboratory manual.3 Spermatozoa samples from fertile volunteers were obtained from The First Affiliated Hospital of USTC for use as control samples. Fertility was confirmed based on normal semen parameters according to 6^th edition of the WHO laboratory manual (Table 1). To evaluate sperm morphology, the sperm smear slides of patients and a control were fixed in 4% paraformaldehyde for 5 min, rinsed with phosphate buffered saline (PBS), and stained with hematoxylin and eosin as described previously.33 A minimum of 200 spermatozoa were analyzed to determine the percentage of morphologically abnormal sperm. Images were captured using a Nikon ECLIPSE 80i microscope (Nikon, Tokyo, Japan) at 100× magnifications with oil immersion. Sperm flagella morphology was assessed based on the 6^th edition of the WHO laboratory manual.

Table 1.

A detailed clinical overview of patients with oligoasthenoteratozoospermia

Parameter	Patient 1	Patient 2	Reference valuesa
Basic information
Genotype of FSIP2	MT/MT	MT/MT	-
Age (year)b	34	30	-
Years of marriagec	10	9	-
Karyotype	46,XY	46,XY	-
Height/weight (cm/kg)	165/65	166/70	-
Diagnose of disease	OAT	OAT	-
Physical examination
Left testicular volume (ml)	13	14.5	12.5–19
Right testicular volume (ml)	13	14.5
External genitalia	Normal	Normal	-
Secondary characteristics	Normal	Normal	-
Semen parameter
Semen volume (ml)	2.0±0.7	2.0±0	>1.4
Semen pH	Alkaline	Alkaline	Alkaline
Sperm concentration (×106 ml−1)	5.0±0	3.5±0.7	>16
Immotile sperm (%)	60.0±10.0	92.5±2.5	-
Progressive motility (%)	10.0±0	5.0±0	>30
Sperm morphology
Normal flagella (%)	6.0	7.0	>23
Absent flagella (%)	10.5	4.0	-
Short flagella (%)	54.5	64.0	-
Coiled flagella (%)	20.0	14.4	-
Bent flagella (%)	6.0	7.0	-
Irregular caliber flagella (%)	3.0	5.6	-

^aReference values are based on the 6^th edition of the WHO. ^bAges at the time of manuscript submission. ^cYears of marriage till submission of this manuscript. For patient 1 and patient 2, semen volume, sperm concentration, and motility were evaluated at least twice. Data are presented as the mean±s.e.m. MT: mutant; OAT: oligoasthenoteratozoospermia; FSIP2: fibrous sheath interacting protein 2; WHO: World Health Organization; s.e.m: standard error of the mean; -: not applicable

The Student’s t-test was used to assess the significance of differences in sperm flagellar morphologies between the fertile control and the patients. Data are presented as the mean ± standard error of the mean (s.e.m.). Statistical significance was defined as ^*P < 0.05, ^**P < 0.01, and ^***P < 0.001. All statistical analyses were performed using GraphPad Prism Software version 6.01 (GraphPad, San Diego, CA, USA).

WES and variant filtration

WES was conducted using the AIExome Enrichment Kit V1 (iGeneTech, Beijing, China) to capture libraries, following the manufacturer’s instructions. WES was performed on P1 (II:2) from family 1, as well as P2 (II:1) and his mother (I:2) from family 2. Sequencing was carried out on the HiSeq 2000 platform (Illumina, San Diego, CA, USA), and the raw data were processed as described previously.34 To validate the identified mutations, Sanger sequencing was performed on P1 (II:2) and his brothers (P1 [II:1] and P1 [II:3]) from family 1, as well as on P2 (II:1) and his mother (P2 [I:2]) from family 2. The primers targeting the mutant sites used for polymerase chain reaction (PCR) are listed in Supplementary Table 1. The FSIP2 gene and protein sequences were retrieved from the Ensembl Genome Browser 37 (http://grch37.ensembl.org/; last accessed on 15 June 2024) and UniProt (https://www.uniprot.org/; last accessed on 15 June 2024). The functional impact of the FSIP2 missense variant (c.262C>A, p.P88T) was assessed using PolyPhen, SIFT, MutPred, CADD, and Fathmm-MKL bioinformatics tools. Protein structural stability was evaluated using mCSM, DUET, and MUpro (http://mupro.proteomics.ics.uci.edu/; last accessed on 17 June 2024). Evolutionary conservation of the affected amino acids was analyzed via multiple sequence alignments using Mega11 (version 11.0.13, Molecular Evolutionary Genetics Analysis software; https://www.megasoftware.net) and Clustal X (version 2.1; https://www.clustal.org/).

Supplementary Table 1.

Primers used

Primer name	Sequence (5’–3’)	Purpose	Temperature (°C)	Cycles
FSIP2-MT1-F	GGCTGTTTTTCCCCATTTCA	Used to amplify the human FSIP2 gDNA mutation site of P1 for sanger sequencing	55	38
FSIP2-MT1-R	GTTAAAAGTGACACAGGCCC
FSIP2-MT2-F	AGTTTCTGGTGGCTTTGATG	Used to amplify the human FSIP2 gDNA mutation site of P1 for sanger sequencing	54	40
FSIP2-MT2-R	CTAGGCATTCATCTGGTGAA
FSIP2-MT3-F	ACACAGCCTTCTCTCTATTCAG	Used to amplify the human FSIP2 gDNA mutation site of P2 for sanger sequencing	55	38
FSIP2-MT3-R	GCAGCAATCATCCCTATAGC
FSIP2-MT-F	GACTCCACAGAAGCAGCATT	Used to amplify the human FSIP2 cDNA for RT-PCR	56	35
FSIP2-MT-R	GTTGACTTGGAAAGCCACAGC
ACTB-RT-F	CACCATTGGCAATGAGCGGTTC	Used to amplify the human TUBB cDNA for RT-PCR	56	35
ACTB-RT-R	AGGTCTTTGCGGATGTCCACGT

FSIP2: fibrous sheath interacting protein 2; MT: mutant; cDNA: complementary DNA; RT: reverse transcription; PCR: polymerase chain reaction; P2: patient 2; P1: patient 1

Transmission electron microscopy (TEM) analysis

TEM was performed following previously published protocols.27,35 Sperm samples from the patients and a fertile control were collected and immediately immersed in 0.1 mol l⁻¹ phosphate buffer containing 0.2% picric acid, 8% glutaraldehyde, and 4% paraformaldehyde, then incubated overnight at 4°C. The following day, the samples were washed with 0.1 mol l⁻¹ phosphate buffer and stained with 1% osmium tetroxide. Spermatozoa were dehydrated in graded alcohol solutions (30%, 60%, 90%, and 100%; 10 min for each step) and embedded in a mixture of epon resin and acetone. Ultra-thin sections, approximately 70 nm thick, were cut and stained with lead citrate and uranyl acetate. The ultrastructure of spermatozoa cross-sections was examined using Tecnai 10 or 12 microscopes at 120 kV or 100 kV (Philips CM10; Philips Electronics, Eindhoven, The Netherlands).

RNA extraction and reverse transcription-PCR (RT-PCR)

Sperm lysates of FSIP2-affected patients and a fertile control were used to extract total RNA using TRIzol reagent (TAKARA, Kusatsu, Japan) as previously described.36 Briefly, 1 μg of extracted RNA was reverse-transcribed into complementary DNA (cDNA) using the PrimeScript RT Reagent Kit (TAKARA). RT-PCR was performed using Phanta Flash Master Mix (Vazyme Biotech Co., Ltd., Nanjing, China) following the manufacturer’s protocol, with β-actin (ACTB) as the internal control. Primer details used for RT-PCR are provided in Supplementary Table 1.

Immunofluorescence staining

Immunofluorescence staining was performed on spermatozoa collected from both patients and a fertile control, as previously described.27,35 Briefly, sperm samples were spread onto clean slides and fixed with 4% paraformaldehyde. The sperm smear slides were washed three times with PBS each for 5 min, next permeabilizing the sperms with PBST (PBS with 0.1% Triton X-100) for 30 min. Then, slides were blocked with 3% skim milk. Primary antibodies were applied and incubated overnight at 4°C. The following day, the slides were washed with PBST and incubated with secondary antibodies for 90 min at 37°C. After three additional washes with PBST, the slides were mounted using Hoechst (category No. H21492; Invitrogen, Carlsbad, CA, USA) and Vectashield (category No. H-1000; Vector Laboratories, Burlingame, CA, USA). Images were acquired using a Nikon ECLIPSE 80i microscope (Nikon, Tokyo, Japan). Details of the primary and secondary antibodies used and their dilutions are provided in Supplementary Table 2.

Supplementary Table 2.

Antibody list

Antibody	Host species	Catalog number	Company	Dilution
Immunofluorescence analysis
anti-FSIP2	Rabbit	1	Red Bioss China	1:100
anti-AKAP4	Rabbit	HPA020046	Sigma	1:100
anti-SPAG6	Rabbit	12462-1-AP	Proteintech	1:100
anti-IFT20	Rabbit	13615-1-AP	Proteintech	1:100
anti-ACTL7A	Rabbit	17355-1-AP	Proteintech	1:100
anti-α-tubulin	Mouse	F2168	Sigma	1:100
Alexa Fluor 488 Donkey anti-Mouse IgG	Mouse	A-21202	Invitrogen	1:100
Alexa Fluor 488 Donkey anti-Rabbit IgG	Rabbit	A-21206	Invitrogen	1:100
Alexa Fluor 555 Donkey anti-Rabbit IgG	Rabbit	A-31572	Invitrogen	1:200
Alexa Fluor 568 Lectin PNA	-	L32458	Invitrogen	1:100
Immunoblot analysis
anti-AKAP4	Rabbit	HPA020046	Sigma	1:1000
anti-ACTL7A	Rabbit	17355-1-AP	Proteintech	1:1000
anti-GAPDH	Mouse	60004-1-Ig	Proteintech	1:3000

FSIP2: fibrous sheath interacting protein 2; AKPA4: A-kinase anchor protein 4; SPAG6: sperm-associated antigen 6; IFT20: intraflagellar transport 20; ACTL7A: actin-like 7A; PNA: peptide nucleic acid

Western blot

To prepare protein lysates, semen samples were lysed using a lysis buffer (50 mmol l⁻¹ Tris–HCl [pH 7.5], 150 mmol l⁻¹ NaCl, 2.5 mmol l⁻¹ ethylenediamine tetraacetic acid [EDTA], and 0.5% Triton X-100) as described previously.37 The proteins were then separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to a nitrocellulose membrane (category No. 10600002; GE Healthcare, Chicago, IL, USA) via electrophoresis. The membrane was blocked for 1 h in TBST buffer (50 mmol l⁻¹ Tris [pH 7.4], 150 mmol l⁻¹ NaCl, and 0.1% Tween-20) containing 5% skimmed milk. Primary antibodies were applied and incubated with the membrane overnight at 4°C. The next day, the membrane was washed three times with TBST for 10 min each. Subsequently, the membrane was incubated for 90 min with secondary antibodies. Protein detection was performed using a chemiluminescent substrate (Thermo Fisher Scientific, Waltham, MA, USA), and images were captured using the Image Quant LAS 4000 Imaging System (GE Healthcare).

RESULTS

Clinical profile of patients from consanguineous Pakistani families

This research examined two infertile men from unrelated consanguineous Pakistani families. Both patients (P1 and P2) were born to first cousin parents (Figure 1a and 1b) and exhibited normal testicular size, external genitalia, and secondary sexual characteristics. Cytogenetic analyses confirmed a normal male karyotype (46,XY) in both patients, with no deletions observed on the Y chromosome. Moreover, routine semen analyses of both affected individuals found normal semen volume, yet their sperm concentration and motility were lower than the reference values and met the diagnostic criteria of 6^th edition of the WHO laboratory manual for OAT.38 Table 1 describes in detail the physical and semen characteristics of P1 and P2.

Figure 1.

Identification of novel FSIP2 variants in infertile men with oligoasthenoteratozoospermia. (a) Family pedigree of family 1 with one infertile patient (P1; II:2). (b) Family pedigree of family 2 with one infertile patient P2 (II:1). Red arrows indicate the individuals for whom WES was performed. Slashes denote deceased family members. Squares represent males and circles represent females. Filled squares denote affected males, while clear symbols indicate unaffected individuals. (c) Representative micrographs showing normal sperm morphology from a fertile control and abnormal spermatozoa from P1 and P2, exhibiting short, bent, irregular caliber, absent, and coiled flagella under light microscopy. Scale bars = 5 µm. (d) Mutation sites of three novel homozygous FSIP2 variants identified in family 1 and family 2, along with previously reported mutations. The variants highlighted in blue represent those identified in our study. The positions of FSIP2 variants are illustrated at both the transcript level (GRCh37, transcript ID: ENST00000424728.1) and protein level (Q5CZC0). WES: whole-exome sequencing; FSIP2: fibrous sheath interacting protein 2; UTR: untranslated region.

Morphological analysis of sperm flagella from P1 and P2 revealed that over 90% of their sperm exhibited flagellar abnormalities, such as flagella that were absent, short, coiled, bent, or of irregular caliber, which are characteristics of the MMAF phenotype. Notably, we observed significantly high percentage of short flagella in both of the patients as illustrated in Figure 1c and Supplementary Figure 1 ^{(80.5KB, tif)} . Neither patient showed primary ciliary dyskinesia symptoms nor other systemic health issues, and both patients’ wives reported regular menstrual cycles, normal onset of puberty, and an absence of ovarian pathologies or miscarriage history, indicating no reproductive health concerns on their part.

Identification of novel homozygous variants in FSIP2

To elucidate the genetic basis of MMAF in P1 and P2, WES analysis was employed on P1 from family 1, and P2 and his mother (I:2) from family 2. We used a detailed variant filtration pipeline (Supplementary Figure 2a ^{(282.5KB, tif)} ) to analyze the variants identified by WES, prioritizing those likely to be pathogenic on the basis of functional annotations, population allele frequencies, and predicted deleterious effects. Moreover, homozygosity mapping revealed the identified FSIP2 variants are located within extended homozygous regions in affected individuals from consanguineous families, supporting a recessive mode of inheritance (Supplementary Figure 2b ^{(282.5KB, tif)} and 2c ^{(282.5KB, tif)} ). This analysis identified homozygous FSIP2 mutations in P1, comprising the missense variant c.262C>A (p.P88T) in exon 3 and the frameshift variant c.10948_10951del (p.N3653Nfs^*22) in exon 17, whereas P2 was found to carry the homozygous frameshift variant c.15982_15982del (p.I5328Lfs^*33) in exon 17 of FSIP2 (Figure 1d). Subsequently, Sanger sequencing validated the novel FSIP2 variants identified in P1 and P2 (Figure 2a). NCBI RNA-sequence data revealed a testis-enriched FSIP2 gene expression (Supplementary Figure 3 ^{(90.2KB, tif)} ), and a previous study reports that FSIP2 knock-in mice exhibit sperm morphological defects characteristic of MMAF,39 which led us to hypothesize that the novel mutations identified in FSIP2 could potentially be pathogenic in patients with MMAF. The FSIP2 missense variant c.262C>A (p.P88T) is extremely rare, its allele frequency in the ExAC is 0.000058, and it is absent from the GnomAD East Asian populations. Multiple bioinformatics tools, including PolyPhen, SIFT, MutPred, CADD, and Fathmm-MKL, consistently predict it to be deleterious. Furthermore, mCSM, DUET, and MUpro indicate that the missense variant decreases the structural stability of the FSIP2 protein (Supplementary Table 3), while conservation analysis highlights that the affected amino acids are highly conserved across species (Supplementary Figure 4a ^{(307.2KB, tif)} ), supporting the functional significance of the FSIP2 missense variant.

Figure 2.

Sanger sequencing and ultrastructural analysis of flagellar cross sections. (a) Sanger sequencing chromatograms of FSIP2 variants at the DNA level, obtained from all available family members (b) Representative images of flagellar cross-sections from a fertile control and P1. Scale bars = 500 nm. The MS is indicated by blue arrows, the FS by red arrows, the DMT by yellow arrows, and the CP by orange arrows. (c) RT-PCR analysis of FSIP2 mRNA levels in sperm lysate from P1 and P2 compared with fertile control. The results show significantly reduced of FSIP2 mRNA expression in both patients. (d) Representative images of spermatozoa from a fertile control, co-stained with anti-α-tubulin and anti-FSIP2 antibodies. Scale bars = 10 µm. WT: wild-type; MT: mutant; RT-PCR: reverse transcription-polymerase chain reaction; MS: mitochondrial sheath; FS: fibrous sheath; DMT: doublet microtubules; CP: central pair; FSIP2: fibrous sheath interacting protein 2; P1: patient 1; P2: patient 2; Ref: reference; mRNA: messenger RNA; ACTB: β-actin.

Supplementary Table 3.

In-silico analyses of fibrous sheath interacting protein 2 missense variant identified in patient 1 from family 1

Characteristics	P1	Prediction
Gene name	FSIP2	-
cDNA alteration	c.262C>A	-
Protein alteration	p.P88T	-
Variant type	Missense	-
Allele frequency in human population
ExAC browser
Total	0.000058
East Asians	0.00016	Rare
GnomAD
Total	0.00007082
East Asians	0.000	Absent
Function prediction
PolyPhen	-	Probably damaging
SIFT	-	Deleterious
MutPred	0.64	Deleterious
CADD	24	Deleterious
Fathmm-MKL	-	Deleterious
Protein stability (ΔΔG, kcal mol−1)a
mCSM	−0.57	Destabilizing
DUET	−0.57	Destabilizing
MUpro	−0.9	Destabilizing

^aChange in Gibbs free energy (ΔΔG) represents stability change upon mutation, ΔΔG ≤0.5 kcal mol⁻¹ was considered destabilizing. FSIP2: fibrous sheath interacting protein 2; GnomAD: genome aggregation database; PolyPhen: polymorphism phenotyping; SIFT: sorting intolerant from tolerant; MutPred: mutation prediction; CADD: combined annotation-dependent depletion; Fathmm-MKL: functional analysis through hidden Markov models-multiple kernel learning; mCSM: mutation cutoff scanning matrix; DUET: deletion using ensembles of trees; P1: patient 1; cDNA: complementary DNA

Notably, a total of 45 FSIP2 variants have been identified to date, including three novel variants found in this study and 42 previously reported ones. Exon 16 accounts for 48.9% (22 of 45) of these variants, while exon 17 accounts for 31.1% (14 of 45), suggesting that these two exons represent high-frequency FSIP2 mutation sites (Supplementary Table 4).

Supplementary Table 4.

Previously identified fibrous sheath interacting protein 2 gene variants and those identified in this study

FSIP2 variants (cDNA level)	Amino acid alteration	Exon	Hom/het	Reference
c.910delC	p.Gln304Lys12ter13	8	Hom	2
c.[1606_1607insTGT; 1607_1616delAAAGATTGCA]	p.Lys536Metfster1	16	Hom
c.2282dupA	p.Asn761llefster4	16	Hom
c.16389_16392delAATA	p.Glu5463Glufster7	17	Hom
c.1907C>A	p.S636ter	16	Hom	1
c.8030_8031insA	p.T2680Nfster9	16	Hom
c.16246_16247insCCCAAATACACC	p.T5416fster7	16	Het	3
c.17323C>T	p.Q5774ter	17	Het
c.1750T>A	p.C584S	16	Het	4
c.13600A>G	p.I4534V	17	Het
c.8368_8369insC	p.L2790Sfs*22	16	Hom	5
c.19981 C>T	p.R6661ter	18	Hom	6
c.18448G>A	p.V6150I	17	Het
c.5238-5240del	p.E1746del	16	Het
c.5480A>T	p.D1827V	16	Het
c.9056T>C	p.I3019T	16	Het
c.10823T>C	p.S5933F	17	Het
c.17798C>T	p.D1827V	17	Het
c.272_275delinsAGGTTTTTAA	p.L92Vfster74	3	Het	7
c.16788_16791del	p.E5596fs	17	Het
c.8104dup	p.L2702Pfs*44	16	Het	8
c.4574C>A	p.S1525*	16	Het
c.8038dup	p.T2680Nfs*9	16	Het
c.9234dup	p.L3079Tfs*6	16	Het
c.10247dup	p.E3417Gfs*12	16	Het
c.9043G>A	p.E3015K	16	Het
c.8909C>T	p.Pro2970Leu	16	Het	9
c.11171G>T	p.Arg3724Ile	17	Het
c.9234dupA	p.Leu3079ThrfsTer6	16	Het
c.5539G>T	p.Val1847Phe	16	Het
c.19925G>C	p.Arg6642Pro	18	Het
c.302A>G	p.His101Arg	3	Het
c.14107delA	p.Ser4703AlafsTer18	17	Het
c.232C>T	p.Arg78Ter	3	Het
c.12329C>A	p.Pro4110His	17	Het
c.16346C>T	p.Thr5449Ile	17	Het
c.20146A>C	p.Ser6716Arg	18	Het
c.3351G>A	p.Met1117Ile	16	Het
c.4345G>A	p.Gly1449Arg	16	Het
c.488C>G	p.Ala163Gly	5	Het
c.10726G>C	p.Ala3576Pro	17	Het
c.2137C>T	p.Arg713Ter	16	Hom
c.262C>A	p.P88T	3	Hom	This study
c.10948_10951del	p.N3653Nfs*22	17	Hom
c.15982_15982del	p.I5328Lfs*33	17	Hom

FSIP2: fibrous sheath interacting protein 2; cDNA: complementary DNA

1 Liu W, Wu H, Wang L, Yang X, Liu C, et al. Homozygous loss-of-function mutations in FSIP2 cause male infertility with asthenoteratospermia. J Genet Genomics Yi Chuan Xue Bao 2019; 46: 53–6.

2 Martinez G, Kherraf ZE, Zouari R, Fourati Ben Mustapha S, Saut A, et al. Whole-exome sequencing identifies mutations in FSIP2 as a recurrent cause of multiple morphological abnormalities of the sperm flagella. Human Reprod 2018; 33: 1973–84.

3 Liu M, Sun Y, Li Y, Sun J, Yang Y, et al. Novel mutations in FSIP2 lead to multiple morphological abnormalities of the sperm flagella and poor ICSI prognosis. Gene 2021; 781: 145536.

4 Yuan Y, Xu WQ, Chen ZY, Chen Y, Zhang L, et al. Successful outcomes of intracytoplasmic sperm injection – embryo transfer using ejaculated spermatozoa from two Chinese asthenoteratozoospermic brothers with a compound heterozygous FSIP2 mutation. Andrologia 2022; 54: e14351.

5 Fang X, Gamallat Y, Chen Z, Mai H, Zhou P, et al. Hypomorphic and hypermorphic mouse models of Fsip2 indicate its dosage-dependent roles in sperm tail and acrosome formation. Development 2021; 148: dev199216.

6 Zheng R, Wang Y, Li Y, Guo J, Wen Y, et al. FSIP2 plays a role in the acrosome development during spermiogenesis. J Med Genet 2023; 60: 254–64.

7 Gao F, Ye F, Zhang Q, Du Y, Xu W, et al. Compound heterozygous mutations in FSIP2 cause morphological abnormalities in sperm flagella leading to male infertility. Andrologia 2023; 2023: 9222954.

8 Lv M, Tang D, Yu H, Geng H, Zhou Y, et al. Novel FSIP2 variants induce super-length mitochondrial sheath and asthenoteratozoospermia in humans. Int J Biol Sci 2023; 19: 393–411.

9 Dai CL, Yin XY, Peng ZY, Lin H, Zhang P, et al. Novel variants of FSIP2 and SPEF2 cause varying degrees of spermatozoa damage in MMAF patients and favorable ART outcomes. J Assist Reprod Genet 2025; 42: 1–13.

Disorganization of axonemal structures and fibrous sheath observed in P1 with OAT

A previous study reports that patients with FSIP2 mutations exhibit severe anomalies in sperm flagellar structure.14 To investigate the impact of the FSIP2 homozygous mutations identified in P1, we used TEM to analyze the ultrastructural organization of spermatozoa in both P1 and a fertile control. In the control, we observed that the axoneme displayed an evolutionary conserved “9+2” arrangement of nine doublets of microtubules (DMTs) arranged circumferentially around a central pair (CP) of microtubules. In P1, in contrast, we observed severe disorganization of the sperm axoneme and periaxonemal structures, characterized by absence of the CPs, missing peripheral DMT, and a disorganized mitochondrial sheath (MS). Additionally, significant dysplasia of the fibrous sheath (DFS) was evident in the spermatozoa of P1 (Figure 2b).

To further investigate the ultrastructural defect identified by TEM, sperm-associated antigen 6 (SPAG6), a marker of the CP of microtubules, was examined in semen smears from P1, P2 and a fertile control. In the control, immunofluorescence staining showed strong SPAG6 signals along the sperm flagella. In contrast, SPAG6 was completely absent from the spermatozoa of both patients (Supplementary Figure 4b ^{(307.2KB, tif)} ). These findings, along with the morphological and ultrastructural anomalies of the sperm flagella, strongly suggest that the FSIP2 variants identified in these patients are pathogenic and associated with MMAF phenotype.

Effect of identified mutations on FSIP2 gene expression

FSIP2 expression is enriched in the testes of both humans and mice, playing an essential role in normal sperm function.12 To assess the effects of FSIP2 mutations, we performed RT-PCR to analyze FSIP2 expression and immunofluorescence assays to assess FSIP2 protein localization in P1 and P2 using sperm lysates and sperm smear slides, respectively. The RT-PCR results show that FSIP2 mRNA expression was significantly lower in both FSIP2-affected patients than that in a fertile control (Figure 2c). Immunofluorescence analysis confirmed the localization of the FSIP2 protein prominently in the principal piece of the sperm flagella of a fertile control. In contrast, spermatozoa from the FSIP2-affected patients exhibited a loss of FSIP2 signals, demonstrating an absence of FSIP2 in the spermatozoa of both patients (Figure 2d). Notably, the FSIP2 antibody epitope is located downstream of the mutations identified in P1 and upstream of P2 variant. This suggests that the lack of detection in P2 could result from either a truncated protein that escapes antibody recognition or nonsense-mediated mRNA decay triggered by premature translation termination. Collectively, these variations negatively impact on FSIP2 protein, potentially disrupting sperm flagella development and contributing to the OAT with MMAF phenotype in P1.

Based on these observations, we further investigated the potential impact of FSIP2 variants on the assembly of axonemal protein complexes of the sperm flagella, focusing specifically on intraflagellar transport (IFT). In the fertile control, immunofluorescence analysis of IFT20 showed that it localized along the sperm flagella. In the men carrying FSIP2 variants, IFT20 was completely absent from the sperm flagella (Supplementary Figure 4c ^{(307.2KB, tif)} ). These findings collectively suggest that FSIP2 protein plays an essential role in the proper assembly of the axoneme and the transport of flagella-associated proteins to the assembly site, underscoring its importance for normal sperm flagellar function.

FSIP2 deficiency disrupts AKAP4 and ACTL7A expression, impairing fibrous sheath and acrosome development in sperm

The Search Tool for the Retrieval of Interacting Genes (STRING) protein–protein interaction database showed that the FSIP2 protein functionally interacts with AKAP4 (Supplementary Figure 5a ^{(290.4KB, tif)} ). Such interactions between FSIP2 and AKAP4 compelled us to examine effect of FSIP2 variants on the expression of fibrous sheath components. We analyzed the localization and expression levels of AKAP4, a key protein in the fibrous sheath. In a fertile control, AKAP4 was localized along the sperm flagella, but AKAP4 signals were absent in sperm from P1 and P2 (Supplementary Figure 5b ^{(290.4KB, tif)} ). This finding was further substantiated by western blot assays, which confirmed a lack of AKAP4 expression in the FSIP2-affected patient (Supplementary Figure 5d ^{(290.4KB, tif)} ). We also assessed the localization of actin-like7A (ACTL7A) and peptide nucleic acid (PNA), which are essential proteins for sperm acrosomal development. IF signals of ACTL7A and PNA were present in a fertile control but absent in sperm from the FSIP2-affected patients (Supplementary Figure 5c ^{(290.4KB, tif)} ), additionally western blot analysis confirmed lack of ACTL7A expression in P2 (Supplementary Figure 5d ^{(290.4KB, tif)} ). These results indicate that FSIP2 protein plays a critical role in maintaining the integrity of the fibrous sheath and acrosome development by interacting with both AKAP4 and ACTL7A.

DISCUSSION

In the present study, we performed genetic analysis using WES to identify the genetic causes of male infertility in two patients (P1 and P2) from two unrelated consanguineous families. In P1, we identified two novel FSIP2 homozygous mutations: a missense mutation, c.262C>A (p.P88T) and a frameshift mutation, c.10948_10951del (p.N3653Nfs^*22). P2 was found to carry a homozygous FSIP2 frameshift mutation, c.15982_15982del (p.I5328Lfs^*33). TEM analysis revealed several periaxonemal as well as axonemal abnormalities, including DFS, a disorganized MS, the absence of the CP, and missing DMTs. However, these fibrous sheath defects have not been observed in the sperm cells of MMAF individuals carrying variants in axonemal genes, such as TTC21A, DNAH1, and SPAG6.7,40,41 These mutations are associated with MMAF in our study.

The fibrous sheath is one of the major structural components of sperm flagella, and dysplasia of this component is associated with male infertility.42 FSIP2 is specifically expressed in spermatogenic cells and most likely binds with the sperm tail through AKAP4, which is a component of the fibrous sheath. However, we found that AKAP4 was not present in any sperm samples of the patients with FSIP2 mutations;12,14 consistent with previous studies, we observed an absence of AKAP4 signals along the sperm flagella in the FSIP2-affected patients. Additionally, western blot analysis did not detect AKAP4 protein in P2, suggesting that FSIP2 protein plays a role in the proper targeting and functioning of AKAP4 within the sperm flagella.

Studies on mice have also established that FSIP2 is a testis enriched gene,43,44 and its protein is detectable from 21 days of age and progressively increasing until adulthood. Its mRNA is transcribed from the late spermatocyte stage, and the protein presents at different stages of spermatogenesis.39,43 In our analysis, RT-PCR and immunofluorescence assays confirmed these findings. RT-PCR analysis confirmed a substantial reduction in FSIP2 mRNA levels in the affected patients, suggesting that these mutations disrupt FSIP2 at both the transcriptional and functional levels, leading to defective sperm flagella development. Immunofluorescence staining clearly demonstrated the presence of FSIP2 proteins within the sperm flagella from a fertile control, especially in the principal piece. However, spermatozoa from both patients in our study lacked FSIP2 signals entirely. Thus, our research further corroborates the strong association between FSIP2 and MMAF.12,43

Recent studies have suggested that FSIP2 mutations contribute to acrosomal abnormalities, possibly through interactions with acrosomal proteins such as dpy-19-like 2 (DPY19L2), sperm acrosome-associated 1 (SPACA1), and acrosomal vesicle protein 1 (Acrv1).39,45 However, whether the FSIP2 protein directly regulates acrosome formation remains unclear. In our study, we observed the absence of ACTL7A and PNA in the sperm smears of the FSIP2-affected patients. This finding was confirmed using western blot analysis, which showed a lack of ACTL7A in the patients’ sperm. Our results, together with those of previous studies, support the involvement of the FSIP2 protein in acrosome development, though the precise molecular mechanism requires further investigation.

We identified two distinct mutations in P1: a missense variant (c.262C>A, p.P88T) in exon 3, resulting in threonine replacing proline at position 88; and a frameshift variant (c.10948_10951del, p.N3653Nfs^*22) in exon 17, leading to a premature stop codon and a truncated FSIP2 protein. Although P1 exhibited the MMAF phenotype, the pathogenic contributions of these two mutations remain uncertain. Based on our analysis, we speculate that because of its disruptive nature, the frameshift mutation, which truncates the FSIP2 protein, is more likely to cause MMAF. Conversely, the role of the missense mutation (p.P88T) is less clear and may not significantly contribute to the MMAF phenotype.

The FSIP2 gene consists of 23 exons, with the mutations under study located in exons 3 and 17, possibly leading to mRNA decay via nonsense-mediated decay due to the introduction of premature stop codons. While this process may prevent the production of a fully functional protein, a small amount of truncated FSIP2 (spanning exons 1 to 17) could possibly still be produced, though it may be undetectable or nonfunctional.

Patients with MMAF have immotile sperm due to severe flagellar defects, making natural conception impossible. Intracytoplasmic sperm injection (ICSI) has emerged as a viable assisted reproductive technology (ART) for such patients, but the success of ICSI in patients with FSIP2 mutations remains variable. For instance, in a case of FSIP2 mutation, the transfer of two blastocysts after ICSI did not result in a pregnancy.43 A recent study reports that a patient with a novel compound heterozygous FSIP2 mutation (c.1750T>A and c.13600A>G) achieved a successful clinical pregnancy following the transfer of embryos on day 3.46 While ICSI offers hope to patients with obstructive azoospermia, oligospermia, or asthenospermia, concerns persist regarding the genetic transmission of pathogenic variants through this technique, particularly to male offspring.47 Given the limited number of reported FSIP2 mutation cases, further studies are necessary to evaluate the long-term outcomes of ART in these patients.

In summary, this study identified FSIP2 mutations (specifically an OAT phenotype) in two infertile patients from two Pakistani families, and these mutations significantly reduced FSIP2 protein stability, thereby impairing its function. Ultrastructure analysis of sperm from P1 revealed severe axonemal and periaxonemal disorganization, evidenced by the absence of CPs, missing DMTs, a disorganized MS, and DFS. Our findings provide further genetic evidence supporting the role of FSIP2 mutations in the pathogenesis of OAT and expand the spectrum of FSIP2 mutations in humans. This knowledge will assist clinicians in developing targeted treatment options and offering reproductive counseling for affected individuals and their families.

AUTHOR CONTRIBUTIONS

MH conducted experiments and composed the manuscript. AM, MU, GM, MA, US, MS, FR, NA, AZ, and TA enrolled the patients and collected patients’ samples. HZ performed the WES analysis and bioinformatics analyses. QHS and WS reviewed and revised the manuscript and conceived and supervised the study. All authors participated in the preparation and review of the manuscript and read and approved the final manuscript.

COMPETING INTERESTS

All authors declare no competing interests.

Supplementary Figure 1

Percentages of abnormal sperm flagella morphologies in patients P1 and P2, compared with fertile control. Data are presented as mean, with more than 200 sperm were counted per slide. ^*P<0.05; ^**P<0.01; ^***P<0.001. P1: patient 1; P2: patient 2; ns: not significant.

Supplementary Figure 2

Bioinformatic analysis of candidate variants. (a) Candidate pathogenic variants were filtered through 6 steps. (b) Homozygous FSIP2 mutations are within runs of homozygosity (RoH) regions in P1 (II:2) from family 1, comprising a missense variant and a frameshift variant. (c) Homozygous FSIP2 frameshift variant is within runs of homozygosity (RoH) in P2 (II:1) from family 2. P1: patient 1; P2: patient 2; SNV: single-nucleotide variant; MAF: minor allele frequency; GnomAD: Genome Aggregation Database; MGI: Mouse Genome Informatics.

Supplementary Figure 3

Expression of FSIP2 in different tissues of human. FSIP2 gene expression is significantly higher in the testis than in any other tissues.

FSIP2: fibrous sheath interacting protein 2.

Supplementary Figure 4

Evolutionary conservation and immunoflurecence analysis of spermatozoa in FSIP2-affected patients. (a) Conservation analysis of the mutant amino acids across multiple species, showing the evolutionary conservation of the mutation sites. Arrowheads indicate the exact positions of the mutations. Representative images of spermatozoa from fertile control and patients (P1 and P2), co-stained with anti-α-tubulin and (b) anti-SPAG6 antibodies, and (c) anti-IFT20 antibodies, Scale bars = 10 µm. P1: patient 1; P2: patient 2; FSIP2: fibrous sheath interacting protein 2; SPAG6: sperm-associated antigen 6; IFT20: intraflagellar transport 20.

Supplementary Figure 5

FSIP2 deficiency disrupts AKAP4 and ACTL7A expression, impairing fibrous sheath and acrosome development in sperm. (a) Representative image of STRING protein interaction data showing interacting partners of FSIP2. Representative images of spermatozoa from P1 (family 1), P2 (family 2), and a fertile control, stained with (b) anti-AKAP4 and (c) anti-ACTL7A antibodies. The anti-α-tubulin antibody was used to mark microtubules. Scale bars = 10 µm. (d) Immunoblotting assay showing the absence of AKAP4 and ACTL7A in spermatozoa from P2, who carries the FSIP2 variant. GAPDH was used as a loading control. P1: patient 1; P2: patient 2; FSIP2: fibrous sheath interacting protein 2; AKPA4: A-kinase anchor protein 4; ACTL7A: actin-like 7A.