EIF4G1 is a novel candidate gene associated with severe asthenozoospermia

Abstract Background Asthenozoospermia (AZS), also known as asthenospermia, is characterized by reduced motility of ejaculated spermatozoa and is detected in more than 40% of infertile patients. Because the proportion of progressive spermatozoa in severe AZS is <1%, severe AZS is an urgent challenge in reproductive medicine. Several genes have been reported to be relevant to severe asthenospermia. However, these gene mutations are found only in sporadic cases and can explain only a small fraction of severe AZS, so additional genetic pathogenies need to be explored. Methods and results By screening the variant genes in a patient with severe AZS using whole exome sequencing, we identified biallelic mutations c.2521C>T: p.(Pro841Ser) (NC_000003.11: g.184043412C>T) in exon13 and c.2957C>G: p.(Ala986Gly) (NC_000003.11: g.184045117C>G) in exon17 in the eukaryotic translation initiation factor 4 gamma 1 gene (EIF4G1, RefSeq: NM_004953.4, OMIM: 600495) of the patient. Both of the mutation sites are rare and potentially deleterious. Transmission electron microscopy analysis showed a disrupted axonemal structure with mitochondrial sheath defects. The EIF4G1 protein level was extremely low, and the mitochondrial marker cytochrome c oxidase subunit 4I1 (COXIV, OMIM: 123864) and mitochondrially encoded ATP synthase 6 (ATP6, OMIM: 516060) protein levels were also decreased in the patient's spermatozoa as revealed by WB and IF analysis. This infertility associated with this condition was overcome by intracytoplasmic sperm injections, as his wife became pregnant successfully. Conclusion Our experimental findings indicate that the EIF4G1 gene is a novel candidate gene that may be relevant to severe AZS.

Here, we found biallelic mutations in EIF4G1 from an infertile patient with severe asthenozoospermia and revealed that these mutations may be the key factor responsible for infertility due to this condition.

| Patient and control subject
The proband (30 years of age, II:4) was recruited from the Xiamen Maternity and Child Care Hospital. He was healthy and had a normal sexual ability, but he has not been able to make his wife pregnant. His testes and accessory glands were normal; the reproductive hormones were in normal ranges (FSH 2.25 mIU/mI, LH 1.92 mIU/ml, T 3.94 ng/ ml, E2 27 pg/ml, and PRL 7.01 ng/ml), and the biochemical test indexes of the seminal plasma were normal. The semen examination results were as follows: semen volume, 2.1 ml; semen pH, 7.2; sperm density, 58.8 million/ml; percentage of motile sperm, 11.3%; and percentage of progressive sperm, 0.2%. Sperm morphological analysis showed that the normal sperm morphology amounted to 6%. The chromosomal karyotype of the patient was normal: 46, XY. No microdeletions were found in the Y chromosome. Based on these results, the patient was diagnosed with severe asthenozoospermia. Five milliliters of peripheral blood was collected from the patient, his elder brother and his parents. The control subject was a healthy male aged 28 with normal fertility.

| Ethical compliance
This study was approved by the Ethics Committee of Xiamen Maternity and Child Care Hospital. Written informed consent was obtained from each participant.

| Whole exome sequencing and sanger sequencing validation
Whole exome sequencing (WES) was performed as described previously . Briefly, a DNA library was prepared, and exomes were concentrated with the TruSeq Exome Enrichment kit (Illumina, San Diego, CA) following the manufacturer's protocol. Then, high-throughput sequencing was performed using the Illumina Hiseq 2000 sequencer. The reads were aligned against UCSC hg19 with the Burrows-Wheeler Aligner (http://biobwa.sourc eforge. net/). Variants were annotated by the ANNOVAR (http:// www.openb ioinf ormat ics.org/annov ar/) database, Mutation Taster (http://www.mutat ionta ster.org/) database, the Exome Aggregation Consortium (ExAC) (http://exac.broad insti tute. org/), and the 1,000 Genomes project (http://www.1000g enomes.org/data) databases for pathopoiesia, novelty, and frequency. Mutations that met the following criteria were retained for subsequent analyses: missense, nonsense, frameshift, or splice site variants and variants that were absent or rare (variants with a minor allele frequency <1% in the ExAC and 1,000 Genomes datasets by considering the rare prevalence of severe asthenozoospermia). Sanger sequencing was used to validate the mutation of the EIF4G1 gene in the patient and his parents. The primers used for Sanger sequencing are listed in Table S1.

| Papanicolaou staining
Papanicolaou staining of the spermatozoa was performed according to the World Health Organization standards for human semen examination and processing (5th ed.) with modification to confirm morphological changes in sperm tails as described previously (Sha, Yang, et al., 2017). Briefly, slides were fixed in 95% ethanol for 15 min, followed by immersion in a graded alcohol series from 80% to 50%. Then, the slides were rinsed with distilled water for 2 min and stained with hematoxylin for 5 min. After soaking with distilled water and ethanol hydrochloride, the slides were inserted into Bluing Reagent for 4 min. Then, the slides were dehydrated in a graded alcohol series from 50% to 90% and dyed with Orange G6 and EA50. Then, the slides were dehydrated with 95% and absolute alcohol. Subsequently, slides were washed in xylene and mounted with permanent mounting medium.

| Transmission electron microscopy
Sperm samples were examined following a previously published procedure for subcellular structural changes in sperm (Sha, Yang, et al., 2017). Briefly, prepared spermatozoa were immobilized with 2.5% phosphate-buffered glutaraldehyde. Then, the samples were washed with 0.1 M phosphate buffer (pH 7.2) three times and postfixed with 1% osmium tetroxide. Dehydration was performed using a graded alcohol series and 100% acetone sequentially, followed by infiltration with 1:1 acetone and SPI-CHEM resin. After infiltration, samples were embedded and polymerized. Ultrathin 70-nm thick sections were cut with diamond knives using an Ultramicrotome Leica EM UC7 (Leica, Wetzlar, Germany). The sections were collected on 200 mesh transmission electron microscopy (TEM) copper grids and counterstained with uranyl acetate and lead citrate. The ultrastructure of the sample was observed and photographed using a Tecnai G2 Spirit transmission electron microscope (FEI, Oregon) at 80 kV.

| Western blot analysis
Sperm samples from the patient and control subject were prepared as described previously (Martinez-Heredia, Estanyol, Ballesca, & Oliva, 2006). Each sample was centrifuged at 800 g for 20 min at 4°C in a 50% step Percoll gradient to remove the seminal plasma and other potentially contaminating cells in the semen. Spermatozoa protein was extracted as per the method by Liu et al., (2015), separated by 6% (w/v) SDS-PAGE for EIF4G1 protein and 10% (w/v) SDS-PAGE for acetylated tubulin, and then transferred to a polyvinylidene difluoride membrane (Millipore, USA). The membrane was blocked for 1 hr at room temperature (RT) with 5% skimmed milk in Tris-buffered saline solution (pH 7.4) containing 0.05% Tween-20 (TBST) and then incubated with either rabbit anti-EIF4G1 (15704-1-AP, Proteintech, USA), anti-COXIV (11242-1-AP, Proteintech, USA), anti-ATP6 (55313-1-AP, Proteintech, USA), or anti-acetylated tubulin (66200-1-Ig, Proteintech, USA) primary antibody overnight at 4°C. After three washes with TBST, the membranes were incubated with goat anti-rabbit IgG secondary antibody, HRP (31460, Thermo Fisher, USA) or goat anti-mouse IgG secondary antibody, HRP (31430, Thermo Fisher, USA) for 1 hr and washed three times with TBST at RT. The signals were developed using an ECL (enhanced chemiluminescence) kit (K-12045-D50, Advansta, USA) and visualized and recorded with the use of an ImageQuant LAS 4000 mini machine (GE Healthcare Life Sciences, USA). The specific antibodies used for the Western blots analysis are listed in Table S2.

| Intracytoplasmic sperm injection, embryo transfer, and pregnancy follow-up
Intracytoplasmic sperm injection (ICSI) was performed as previously described (Sha, Zhang, Ding, & Li, 2017). The couple chose to use a long-acting GnRHa protocol. In brief, long-acting GnRHa was injected subcutaneously on day 23 of the menstrual cycle. On the 5th day of menstruation, blood hormones levels were measured, and ultrasound imaging was performed to confirm the success of the protocol. Afterward, based on the ovarian reserve function, 150-300 U of Gn was injected daily until at least 1-2 follicles had developed to 18-20 mm and until the E2 level reached >200 pg/ ml. Then, 5,000-10,000 U of HCG was injected subcutaneously at night at 10 o'clock, and eggs were retrieved 36 hr later. ICSI was performed for the couple, and fertilization was assessed 18-19 hr later for the presence of two pronuclei and two polar bodies. The fertilized oocytes were cultured individually in G1 medium (Vitrolife) until the day of transfer. Cleaved embryos were evaluated for cell number, blastomere appearance and fragmentation rate. The best embryos were selected for transfer, and extra good-quality embryos were frozen. Two embryos were transferred by ultrasound guidance. The patient received luteal support, including 600 mg of vaginally administered micronized progesterone (Utrogestan; Besins Laboratories). Serum hCG levels were measured 14 days after embryo transfer. Clinical pregnancy was defined as a visible sac with a fetal heart beat 7 weeks after embryo transfer.

| Morphological defects in the sperm of the patient with severe asthenozoospermia
The proband and his family were recruited for this study ( Figure 1A). The patient's elder brother (II:1) had two sons (III:1 and III:2). The spermatozoa of the patient had low motility, and the percentages of progressive sperm and nonprogressive sperm were 0.2% and 1.1%, respectively. Therefore, he was diagnosed with severe asthenozoospermia.
As indicated by the black arrow, the sperm from the patient showed defects in the mitochondrial sheath (MS) and flagellum compared to the normal sperm by Papanicolaou staining ( Figure 1B). To further characterize these defects, TEM was performed, and spermatozoa from the patient showed numerous ultrastructural defects in the MS and flagellum ( Figure 1C). As indicated by the arrow, control sperm had a normal ultrastructure in the longitudinal section in the midpiece (MP) (Figure 1C, a). However, in the patient's sperm, the flagellum appeared severely disorganized, and the MS of the spermatozoon was incomplete, misassembled and much shorter than that of the normal MS ( Figure 1C, b). The ultrastructure of the cross-section showed the typical '"9 + 2"' microtubule structure surrounded by a well-organized MS in the normal spermatozoon from the control ( Figure 1C, c). However, the sperm flagellum from the patient showed a complete absence of central pairs (CPs) and doublet microtubules (DMTs), mitochondria were vacuolated, and the MS showed defective assembly ( Figure 1C, d). Clearly, the CPs were surrounded by nine DMTs, and the FS was composed of two longitudinal columns (LCs) connected by circumferential ribs (CR) in cross-sections of the principal piece (PP) from the control ( Figure 1C, e). In contrast, cross-sections of sperm flagellum from the patient showed disorder. The FS was dysplastic, thickened, and dramatically disorganized. Various axonemal anomalies were observed, including a lack of CPs and DMTs ( Figure 1C, f). These results show that the infertility of the patient was mainly due to vacuolated mitochondria, a short misarranged MS and an abnormal flagellum.

| Biallelic mutations of the EIF4G1 gene were identified in the patient with severe asthenozoospermia
To determine the origin of this disease in the patient, genomic DNA was extracted from the whole blood of the patient, his elder brother and his parents, and the patient's genomic DNA was subjected to WES analysis. The results
were analyzed by bioinformatics to exclude irrelevant or meaningless mutations. We obtained a list of rare and potentially pathogenic variants, including homozygous or compound heterozygous mutations (Table S3). We did not find any known genetic mutations that could cause asthenospermia on this list. After ruling out the mutations that  show extremely low expression in the testis or that have low pathogenicity based on protein expression databases (https ://www.prote inatl as.org) and bioinformatic analysis, we obtained biallelic mutations of EIF4G1 (RefSeq NM_004953.4, genome build GRCh37/hg19) because this gene was the only gene that was potentially closely associated with severe asthenozoospermia based on the previous literature. A second round of Sanger sequencing for validation of the biallelic mutations in the EIF4G1 gene was performed in the patient, his elder brother and his parents (Figure 2a). Specifically, the patient carried the following mutations: c.2521C>T: p.(Pro841Ser) (NC_000003.11: g.184043412C>T) in exon13 and c.2957C>G: p.(Ala986Gly) (NC_000003.11: g.184045117C>G) in exon17 (Figure 2a, line 3). Furthermore, we found c.2521C>T: p.(Pro841Ser) in exon13 mutation in his father ( Figure 2a, line 1, left), c.2957C>G: p.(Ala986Gly) in exon17 mutation in his mother (Figure 2a, line 2, right) and c.2957C>G: p.(Ala986Gly) in exon17 mutation in his elder brother (Figure 2a, line 3, right), implying that the biallelic mutations were inherited from his parents. The biallelic mutation sites are located in the coding exons (Figure 2b, upper), and the amino acids influenced by these mutations are in the functional domains of the EIF4A-binding sites (Figure 2b, under). In addition, the amino acids encoded by mutation sites are highly conserved between different species (Figure 2c). Furthermore, the impact of these mutations on protein function was confirmed by bioinformatics analysis using Mutation Taster databases. The mutations were highly pathogenic. The frequencies of the observed mutations in the general population were assessed using ExAC and 1,000 Genomes databases, and these mutations are rare mutation sites (Table 1).

| EIF4G1 protein was absent in the patient with severe asthenozoospermia
To assess the effects of the biallelic mutations on the EIF4G1 protein, we measured the EIF4G1 protein levels in the sperm of the patient by Western blotting and immunofluorescence (Figure 3). The results showed that the expression of EIF4G1 protein in the sperm of the patient was very weak compared with that in the sperm of the control (Figure 3a). The quantification of the Western blotting results showed that protein expression was decreased significantly in the patient's sperm (Figure 3b). The protein expression level and localization of EIF4G1 were examined by immunofluorescence. EIF4G1 protein was mainly located in the head of the normal sperm, and the signal was very weak and barely detectable in the patient's sperm compared to the normal sperm (Figure 3c).

| COXIV and ATP6 protein levels were decreased in the patient with the EIF4G1 mutations
The MS of the sperm from the patient was severely destroyed by TEM analysis. Therefore, we examined the mitochondrial marker COXIV and ATP synthesis enzyme ATP6 protein levels in the sperm of the patient by Western blotting and immunofluorescence. Western blotting showed that the expression of COXIV was decreased, and that of the ATP6 protein was decreased significantly in the sperm of the patient (Figure 4a). Quantification analysis of the Western blot band signal further confirmed the results (Figure 4b). Consistently, the immunofluorescence data also displayed the descending expression range of COXIV ( Figure 4c) and the sharply decreased expression of ATP6 in the mutant sperm compared with that in the control sperm (Figure 4d).

| Pregnancy outcome by ICSI with the patient's sperm
Assisted fertilization by ICSI was performed for this patient. The patient's wife underwent a long-acting GnRHa protocol in our hospital in December 2017. Fifteen eggs were collected from the female, and all of them were at the MII stage. After ICSI, 11 of the fertilized eggs reached 2PN, and four of the D3 embryos, including two high-quality embryos, were frozen. The other four embryos were grown to blastocysts ( Figure S1), but the transfer was canceled due to endometrial problems. In March 2018, two high-quality D3 embryos were transferred, but the patient's wife failed to become pregnant. Two months later, two more D3 embryos were transferred. The hCG level reached 411.27 mIU/ml in the blood 10 days after transfer, and the woman was pregnant clinically in the end.

| DISCUSSION
Our present study identified biallelic mutations in the EIF4G1 gene of a patient with severe asthenozoospermia, in which the EIF4G1 protein level was significantly decreased, the spermatozoa had lost their motility and the spermatozoa showed a variety of morphological and ultrastructural defects. These findings revealed that EIF4G1 is a novel candidate gene associated with severe asthenozoospermia. EIF4G1 is a component of eukaryotic initiation factor 4F (eIF4F), which is critical for recognizing the mRNA cap, unwinding the 5'-terminal secondary structure and recruiting the mRNA to the ribosome (Aitken & Lorsch, 2012;Merrick, 2015). EIF4G1, a scaffold protein, is mainly expressed in the cytoplasm and activates translation initiation by binding eIF4E, eIF4A, eIF3, poly(A)-binding protein, eIF4E kinase, and Mnk (Akabayov, Akabayov, & Wagner, 2014;Korneeva, Lamphear, Hennigan, & Rhoads, 2000;Zakowicz et al., 2005). Depletion of initiation factor eIF4G1 can selectively inhibit translation of specific mRNAs, but this depletion only modestly reduces overall protein synthesis in cells and impairs cell proliferation, bioenergetics, and mitochondrial activity (Ramirez-Valle et al., 2008). In Paip2a/Paip2b-DKO mice, flagellum formation is impaired, and the MS is absent in the middle piece. Further research proved that the expression of PABP was significantly increased and that abundant PABP competes with EIF4G1 and leads to translational inhibition (Delbes et al., 2012;Yanagiya et al., 2010). These results suggest that EIF4G1 may play an important role in spermatogenesis.
We identified biallelic mutations in EIF4G1 (NC_000003.11:g.184043412C>T and g.184045117C>G). These changes may affect the stability of the EIF4G1 protein F I G U R E 3 EIF4G1 protein level in the patient and control. (a) EIF4G1 protein levels were determined with Western blots. (b) The density of each band was quantified with ImageJ. Ac-tubulin was used as the loading control. The results were expressed as the mean ± SD of three independent experiments. Data were analyzed with spss 18.0 software. ***p < 0.001. (c) EIF4G1 protein expression was determined with an immunofluorescence assay. Multiple images were taken, and representative images are presented. Scale bar: 10 μm.
due to a drastic change in polarity from proline (a hydrophobic amino acid) to serine (a hydrophilic amino acid). As expected, we observed that EIF4G1 protein was significantly decreased in the sperm of the patient. Morphological analysis showed abundant defects in the MS and flagellum of the patient's sperm and that the spermatozoa lost their motility. The damaged mitochondria could not form a completed MS, which in turn damaged the structure of the sperm flagellum. In addition, we found vacuolated mitochondria in the chaotic MS and found that the mitochondrial marker COXIV and ATP synthesis enzyme ATP6 F I G U R E 4 COXIV and ATP6 protein levels in the patient and control. (a) COXIV and ATP6 protein levels were determined with Western blots. (b) The density of each band was quantified with ImageJ. Ac-tubulin was used as the loading control. The results were expressed as the mean ± SD of three independent experiments. Data were analyzed with spss 18.0 software. *p < 0.05, **p < 0.01. (c) COXIV and ATP6 protein expressions were determined by immunofluorescence assay. Multiple images were taken, and representative images are presented. Scale bar: 10 μm levels also decreased. This evidence suggests that these biallelic mutations may affect the function of mitochondria, leading to ultrastructural defects in the MS and flagellum.
Our findings are the first to suggest that EIF4G1 may be a novel candidate gene that is relevant to severe asthenozoospermia. However, severe asthenozoospermia is a rare disease. Due to the limitations of sample capacity, we only found mutations in this sporadic case. A larger sample size should be investigated to further confirm the role of the EIF4G1 mutation in severe asthenospermia. In addition, as a translation initiation factor, EIF4G1 is widely expressed and plays an important role in various tissues; thus, only the specific conditional knockout system Cre/loxP can be used to construct EIF4G1-absent sperm in mice to validate our findings.
Hence, our study confirms that the EIF4G1 gene mutation may be a novel pathogeny of severe asthenozoospermia and provides more information about severe asthenozoospermia for researchers and clinicians.