Skip to main content
PMC home page
eLife logoLink to eLife
. 2020 Jun 2;9:e53913. doi: 10.7554/eLife.53913

TMEM95 is a sperm membrane protein essential for mammalian fertilization

Ismael Lamas-Toranzo 1,†, Julieta G Hamze 2,†, Enrica Bianchi 3, Beatriz Fernández-Fuertes 4,5, Serafín Pérez-Cerezales 1, Ricardo Laguna-Barraza 1, Raúl Fernández-González 1, Pat Lonergan 4, Alfonso Gutiérrez-Adán 1, Gavin J Wright 3, María Jiménez-Movilla 2,✉, Pablo Bermejo-Álvarez 1,✉
Editors: Polina V Lishko6, Anna Akhmanova7
PMCID: PMC7295574  PMID: 32484434

Abstract

The fusion of gamete membranes during fertilization is an essential process for sexual reproduction. Despite its importance, only three proteins are known to be indispensable for sperm-egg membrane fusion: the sperm proteins IZUMO1 and SPACA6, and the egg protein JUNO. Here we demonstrate that another sperm protein, TMEM95, is necessary for sperm-egg interaction. TMEM95 ablation in mice caused complete male-specific infertility. Sperm lacking this protein were morphologically normal exhibited normal motility, and could penetrate the zona pellucida and bind to the oolemma. However, once bound to the oolemma, TMEM95-deficient sperm were unable to fuse with the egg membrane or penetrate into the ooplasm, and fertilization could only be achieved by mechanical injection of one sperm into the ooplasm, thereby bypassing membrane fusion. These data demonstrate that TMEM95 is essential for mammalian fertilization.

Research organism: Mouse

Introduction

In sexually reproducing species, life begins with the fusion of two gametes during fertilization. Mammalian fertilization requires the sperm to pass through a glycoprotein coat that surrounds the egg, termed the zona pellucida, and once in the perivitelline space it must fuse its membrane with that of the egg to deliver the paternal genetic material into the ooplasm. Despite its importance in sexual reproduction, the molecular mechanisms behind the fusion of gamete membranes remain largely unknown. Currently, only three proteins have been demonstrated to be essential for sperm-egg membrane interaction: the sperm proteins IZUMO1 (Inoue et al., 2005) and SPACA6 (Lorenzetti et al., 2014), and the egg protein JUNO (Bianchi et al., 2014). The ablation of another egg protein, CD9 (Kaji et al., 2000; Le Naour et al., 2000; Miyado et al., 2000), also leads to a significant impairment in sperm-egg membrane fusion, but its ablation does not cause complete infertility. Similarly, very recently, the ablation of the sperm protein FIMP has been reported to cause severe subfertility due to gamete fusion defects (Fujihara et al., 2020).

IZUMO1 was the first protein to be shown to be critical for gamete fusion (Inoue et al., 2005). Sperm lacking IZUMO1 exhibit normal morphology and motility and are able to pass through the zona pellucida, but incapable of fusing their membrane with that of the egg. The search for the egg binding partner of IZUMO1 lasted almost a decade, until the protein JUNO was shown to be required for gamete membrane binding through direct interaction with IZUMO1 (Bianchi et al., 2014). Here, we demonstrate by gene editing in mice that besides these two proteins, another sperm protein, TMEM95, plays an essential role in fertilization.

Results

TTMEM95 is a sperm protein dispensable for embryo development that relocalizes to the equatorial region after acrosomal reaction

TMEM95 coding region was located in a 1386 kb segment of extended homozygosity in 40 sub-fertile bulls identified in a Genome Wide Association Study carried out on the Fleckvieh breed population (Pausch et al., 2014). Within this segment, containing 80 transcripts, a non-sense mutation in TMEM95 coding region that reduces protein length to 151 amino acids was identified (Pausch et al., 2014). The semen of bulls carrying two of these partially truncated alleles exhibited normal motility parameters but a dramatic reduction in egg penetration was evident following in vitro fertilization (Fernandez-Fuertes et al., 2017). In silico protein folding analysis shows that TMEM95 protein contains a single transmembrane domain and a secondary structure formed by β-hairpin and α-helix which is remarkably similar to that found on IZUMO1 protein and termed ‘IZUMO1 domain’ (Ellerman et al., 2009), thereby suggesting a possible role in sperm-egg fusion (Figure 1A). Similar bioinformatics predictions have been reported for bovine IZUMO1 (Zhang et al., 2016).

Figure 1. Generation of TMEM95 deficient mice.

(A) Structure prediction of TMEM95 protein (left) using IZUMO1 (right) as template, created by SWISS-MODEL software. (B) Tmem95 KO allele generated following CRISPR-mediated edition. CRISPR target sequence and PAM are depicted in blue and purple letters, respectively. (C) The deletion of 10 bp altered Tmem95 ORF. Large letters indicate the aminoacid sequence corresponding to the codons (DNA sequence) shown in smaller letters below. (D) Western Blot images for TMEM95, IZUMO1 and β-tubulin proteins from protein extracts from WT or KO sperm. Graph on right indicates the abundance of IZUMO1 in WT and KO extracts. (E) Immunocytochemistry images of KO and WT sperm stained with an antibody against TMEM95 and the acrosomal stain PNA. TMEM95 localized to the acrosomal cap in acrosome intact sperm and in the equatorial segment after acrosome reaction. (F) Immunocytochemistry images of acrosome intact (upper images) or reacted (lower images) WT sperm stained against IZUMO1 and TMEM95. Both proteins relocalize to the equatorial segment following acrosome reaction.

Figure 1.

Figure 1—figure supplement 1. Offtarget analysis.

Figure 1—figure supplement 1.

None of the 5 most probable off-target sites for the sgRNA used were edited, showing the same sequence as the WT. Location of the possible off-target is coloured in gray.
Figure 1—figure supplement 2. WB images related to Figure 1D.

Figure 1—figure supplement 2.

(A) Immunoblotting of protein extracts obtained from WT and KO epididymal sperm samples following two lysis protocols. In lysis protocol #1, sperm were re-suspended in 4X reducing SDS Sample Buffer and boiled for 10 min. In lysis protocol #2, sperm were re-suspended in 1% Octyl β-D-glucopyranoside solution in PBS and incubated on ice for 30 min (Nishimura et al., 2011). Supernatants were probed with anti-TMEM95 (MyBiosource MBS7004333), anti-IZUMO1 (Abcam ab211623) or anti-β-TUBULIN (Sigma T8328) antibodies. (B) Gel electrophoresis of PCR products amplified from cDNA obtained from testis, seminal vesicle (S.V.) prostate (prost.), epididymis (epid.) or a negative control testis RNA not retrotranscribed (RT-) to detect Gapdh and Tmem95 transcripts. (C) Immunoblotting of protein extracts obtained from WT epididymal sperm, testis or accessory glands (seminal vesicle and prostate). Same antibodies than (A). (D) Uncropped images of the WB used to generate Figure 1D. (E) WB images used for the quantification of IZUMO1 and β-TUBULIN shown in Figure 1D. Four samples were used for quantification (marked with asterisk), as β-TUBULIN band on samples #4 was dispersed, leading to inaccurate quantification.
Figure 1—figure supplement 3. Acrosomal cap localization of TMEM95 and IZUMO1 and IZUMO1 relocalization in WT and KO sperm.

Figure 1—figure supplement 3.

(A) Immunocytochemistry images of WT sperm stained with PNA and IZUMO1 antibody (upper row) or PNA and TMEM95 antibody (lower row) in the absence of permeabilizing agents. Despite the absence of permeabilizing agents (30 min fixation in 4% PFA without Triton X-100), TMEM95 and inner acrosomal (PNA) and acrosomal membrane (IZUMO1) markers were detected. (B) IZUMO1 relocates to the equatorial segment following acrosome reaction in TMEM95 KO sperm. Upper row shows one acrosome intact spermatozoon. Lower rows show WT and KO acrosome reacted sperm where IZUMO1 has relocated to the equatorial segment.
Open in a new tab

To elucidate the role of TMEM95 in fertilization, we generated Tmem95-deficient mice by CRISPR-directed mutagenesis upon microinjection of mouse embryos (CBAXC57BL6 F1 hybrids) at the zygote stage (Wang et al., 2013). sgRNA was designed against the first exon of Tmem95 using bioinformatic tools to minimize the chances of off-target genome editing (https://crispr.mit.edu). Genome editing was confirmed in the pups derived from CRISPR-injected embryos by clonal sequencing, and one edited founder female harbouring a frame-disrupting allele and an in-frame indel was selected to obtain heterozygous offspring by crossing with wild-type males of the same genetic background. Frame-disrupting indel (knock-out allele) consisted of a 4 bp substitution and a 10 bp deletion (Figure 1B) resulting in a predicted truncated peptide of 53 amino acids sharing a region of only 33 amino acids homologous to the N-terminus of TMEM95 protein (Figure 1C). The absence of off-target mutations in the founder female was confirmed by sequencing the five most probable off-target sites (Figure 1—figure supplement 1). F1 heterozygous pups showed similar frequencies for both in-frame and frame-disrupting indels (9:11, respectively) and those pups harbouring the frame-disrupting allele (KO allele) were selected and intercrossed to obtain an F2 generation composed of Tmem95+/+ (wild-type, WT), Tmem95+/- (heterozygous, Hz) and Tmem95-/- (knock-out, KO) pups. The intercross of the heterozygous F1 generation resulted in normal litter sizes, and alleles segregated in a Mendelian distribution (20:43:26 for WT:Hz:KO), indicating that Tmem95 haploinsufficiency did not impair reproductive performance and that Tmem95 bi-allelic ablation did not result in embryonic developmental failure.

Further confirmation of gene ablation was achieved by Western blotting in 3 samples of WT and KO sperm. A ~ 20 kDa band detected in WT samples was absent in KO samples (Figure 1D, Figure 1—figure supplement 2A and C–D). The size of the band was compatible with TMEM95 predicted size and peptide sequencing identified TMEM95 protein on that band (Table 1). IZUMO1 protein was expressed exclusively in testis and sperm (Figure 1—figure supplement 2B–C) and at similar levels in WT and KO sperm (Figure 1D, Figure 1—figure supplement 2E). Immunocytochemistry also demonstrated the absence of TMEM95 protein in sperm obtained from KO males (Figure 1E). In WT acrosome-intact sperm, TMEM95 protein was present in the acrosomal cap, in agreement with previous findings in cattle (Fernandez-Fuertes et al., 2017; Pausch et al., 2014). The specific localization of TMEM95 within the acrosomal cap could not be determined as paraformaldehyde (PFA) fixed sperm untreated with Triton-X100 displayed signals for the inner acrosomal markers PNA and IZUMO1, suggesting that acrosomal membrane was permeable to antibodies (Figure 1—figure supplement 3A). Following acrosome reaction, TMEM95 relocalized to the equatorial segment (Figure 1F), the place where gamete membrane fusion occurs. To test the possibility that TMEM95 ablation may impair IZUMO1 translocation, we determined IZUMO1 translocation in acrosome-reacted WT and KO sperm, observing that IZUMO1 translocation also occurred in the absence of TMEM95 (Figure 1—figure supplement 3B).

Table 1. Details of peptides identified by MS/MS from 20 kDa gel band.

Protein name Score Spi % Sequence
TMEM95 5.33 68.8 LLLCIFGIVLLsGVVSLQ
TMEM95 5.17 59.8 LLLCIFGIVLLsGVVSLQ
TMEM95 4.12 46.1 LLSGVVSLQVEY
TMEM95 3.42 53 KTRYP
Open in a new tab

Tmem95-deficient sperm cannot fuse with eggs

Tmem95 mutant males were healthy, grossly normal and exhibited normal mating behaviour evidenced by the presence of copulatory plugs in females. However, no offspring were obtained from the cross of KO males and fertile WT females after the observation of 24 copulatory plugs from 8 pairs. In contrast, Hz males exhibited normal fertility, comparable to WT counterparts, indicating that haploinsufficiency had no effect on reproductive performance. To determine whether ejaculated sperm from Tmem95-null males were able to fertilize in vivo, we mated 4 WT, Hz or KO males with superovulated WT females and recovered the embryos 1.5 days after mating by oviductal flushing. Over 90% of the embryos recovered from females mated with WT or Hz males were cleaved at the time of recovery, whereas no divided embryos were recovered from females mated with KO males (Figure 2A). Similar results were obtained after in vitro fertilization (IVF); sperm recovered from KO males were apparently unable to fertilize cumulus-oocyte complexes as no embryonic cleavage occurred. By contrast, sperm recovered from Hz males yielded similar rates of embryo cleavage than those obtained from WT counterparts (~70%, Figure 2B). Following in vivo and in vitro fertilization with sperm lacking TMEM95, uncleaved eggs displayed a similar appearance to those non-fertilized by IZUMO1-deficient sperm (Inoue et al., 2005): the perivitelline space contained multiple sperm, probably due to the absence of sperm penetration into the egg and associated zona hardening, allowing multiple sperm to travel across the zona pellucida (Figure 2C, Video 1). TMEM95-null sperm were in close contact with the oolemma but were still unable to penetrate into the ooplasm. This result indicates that TMEM95-deficient sperm are able to pass through the zona pellucida and exhibit an overtly normal motility. To confirm that TMEM95 ablation did not impair sperm motility, we performed computer aided sperm analysis (CASA), observing no differences in any of the kinetics parameters analysed (Figure 2D). TMEM95-deficient sperm were indistinguishable from WT sperm by contrast microscopy (Figure 1E), and the normal morphology of TMEM95-deficient sperm was also assessed by Transmission Electron Microscopy, which showed no detectable abnormalities at the submicroscopic level (Figure 2E).

Figure 2. Reproductive performance of TMEM95-null male mice.

(A) Developmental outcomes following in vivo fertilization of WT eggs with WT, Hz or KO males. TMEM95-deficient males were unable to fertilize eggs, whereas no statistical differences were found in developmental rates following fertilization with sperm obtained from Hz or KO males (ANOVA p>0.05). (B) Developmental outcomes following in vitro fertilization of WT eggs with WT, Hz or KO males. TMEM95-deficient males were unable to fertilize eggs, whereas no statistical differences were found in developmental rates following fertilization with sperm obtained from Hz or KO males (ANOVA p>0.05).(C) Representative image of a non-fertilized egg following co-incubation with TMEM95-disrupted sperm. Egg penetration failure prevents zona hardening leading to the accumulation of sperm in the perivitelline space. (D) Motility parameters of WT, Hz or KO sperm analysed by CASA; no differences were found between groups in any of the parameters analysed (ANOVA p>0.05). (E) Transmission Electron Microscopy images of WT or KO sperm showing an overtly normal morphology in TMEM95-deficient sperm. (F) Average number of WT or KO sperm bound to each oocyte following binding assay. TMEM95 disruption did not impaired sperm binding to the egg membrane (ANOVA p>0.05). (G) Sperm-egg fusion assay. WT sperm fused with Hoechst pre-loaded zona-free eggs, which transferred the stain to them upon membrane fusion. In contrast TMEM95-depleted sperm were unable to fuse, exhibiting no Hoechst staining (only the egg DNA, marked by an arrow, is stained). (H) Development of WT eggs microinjected with WT or KO sperm following Intracytoplasmic Sperm Injection (ICSI). Similar developmental rates were obtained using WT or KO sperm (ANOVA p>0.05), indicating that TMEM95 null sperm were able to fertilize eggs when the sperm-egg membrane fusion step is bypassed by ICSI.

Figure 2.

Figure 2—figure supplement 1. TMEM95 null sperm undergo acrosome reaction and bind to the egg membrane.

Figure 2—figure supplement 1.

(A) Unfertilized eggs co-incubated with TMEM95 null sperm and stained with DAPI and PNA. All sperm that penetrated into the perivitelline space underwent the acrosome reaction (PNA negative) but were still unable to penetrate into the ooplasm. (B) Schematic representation of TMEM95:Cherry fusion proteins. Cherry (highlighted in red) was inserted near the C-terminus of TMEM95 (TMEM95-Ccherry) or just downstream of the signal peptide of TMEM95 (TMEM95-NCherry). (C) Immunoblotting of TMEM95:Cherry fusion proteins expressed in CHO cells (left). Media was separated by SDS-PAGE and analysed by WB using an anti-FLAG antibody. Zona-free eggs were incubated with purified proteins at 0.9 mg/ml for 1 hr to determine protein binding. No protein signal was detected at the egg membrane.
Open in a new tab

Video 1. Confocal z-sections of an egg fixed in PFA following IVF with TMEM95-null sperm.

Download video file (24.5MB, mp4)
Open in a new tab

Multiple sperm are found in the perivitelline space, unable to fuse with the egg membrane.

The absence of zona hardening strongly suggested a failure in sperm penetration into the egg. To further test this failure, unfertilized eggs co-incubated with Tmem95-deficient sperm were stained with the DNA-binding stain DAPI. These samples were also stained with the acrosome staining PNA to test whether the TMEM95-null sperm that had penetrated the zona pellucida had successfully undergone the acrosome reaction, an essential step prior to membrane fusion (Satouh et al., 2012). No sperm heads or male pronuclei were detected in the ooplasm of eggs co-incubated with sperm lacking TMEM95, and those sperm accumulated in the perivitelline space were PNA negative (Figure 2—figure supplement 1A). These results confirm that TMEM95-null sperm had undergone the acrosome reaction but failed to fuse with the egg membrane. Subsequently, we tested whether TMEM95 was involved in sperm binding by counting the number of KO or WT sperm bound to the egg membrane following sperm incubation with zona free eggs. The number of sperm bound to the egg membrane did not differ between both groups, suggesting that TMEM95 is dispensable for sperm binding to the oolema (Figure 2F). We also incubated zona-free eggs with two recombinant Cherry-tagged TMEM95 proteins, failing to detect TMEM95 binding to the egg membrane (Figure 2—figure supplement 1B–C). Together, these results indicate that TMEM95 is not involved on the initial binding but probably on post-binding events required for membrane fusion.

To further test the inability of TMEM95-disrupted sperm to fuse with the egg membrane, a sperm-egg fusion assay was performed preloading zona-free eggs with the DNA-binding dye Hoechst 33342 prior to fertilization with WT or KO sperm. In this assay, the stain loaded into the egg is transferred to the sperm only if gamete fusion occurs (Inoue et al., 2005). Despite the fact that multiple TMEM95-deficient sperm were bound to the oolema, none were stained by the dye stored in the eggs, confirming that sperm-egg membrane fusion does not occur in the absence of TMEM95 (Figure 2G). Subsequently, we tested whether gamete membrane fusion was the only process involved in the infertility of Tmem95-deficient male mice. TMEM95-null sperm were mechanically introduced inside the egg by intracytoplasmic sperm injection (ICSI), thereby artificially bypassing gamete membrane fusion. Eggs injected with sperm obtained from KO or WT males cleaved and developed at similar rates irrespective of the male genotype (Figure 2G). Heterozygous blastocysts obtained using KO sperm were transferred to pseudopregnant recipients, and the resulting pups were genotyped confirming that they were fathered by a Tmem95-/- male. Together, these results demonstrate that the fertility disruption caused by TMEM95 ablation can only be attributed to failure in sperm-egg fusion and thus, subsequent sperm incorporation into the ooplasm.

TMEM95 does not interact with JUNO or IZUMO1

To assess whether TMEM95 was able to bind IZUMO1 or JUNO, we used an assay designed to detect extracellular protein interactions called avidity-based extracellular interaction screening (AVEXIS) (Bushell et al., 2008; Kerr and Wright, 2012). This assay detects direct interactions between receptor ectodomains expressed as soluble recombinant proteins in mammalian cells and uses highly-avid oligomeric forms to detect even very weak binding events that are a feature of this class of interactions (Wright, 2009). The entire ectodomains of JUNO, TMEM95 and IZUMO1 were expressed as monomeric biotinylated ‘baits’ and systematically tested for their ability to bind pentameric beta-lactamase tagged IZUMO1 and JUNO ‘preys’. This approach confirmed the interaction of JUNO and IZUMO1 but no interaction between the extracellular regions of TMEM95 with either JUNO or IZUMO1 was detected (Figure 3A).

Figure 3. TMEM95 does not interact with JUNO nor IZUMO1.

(A) Binding analysis using the AVEXIS assays shows that the soluble recombinant TMEM95 ectodomain does not interact with JUNO nor with IZUMO1. The entire ectodomains of the named proteins were expressed in HEK293-6E cells either as biotinylated baits or as pentameric beta-lactamase-tagged preys. Bait proteins were immobilised on streptavidin-coated plates and captured prey proteins quantified by measuring the absorbance of a colorimetric reaction product of the beta-lactamase substrate, nitrocefin. The CD200R (bait)-CD200 (prey) binding pair was used as positive control. The same prey, CD200R, was tested against TMEM95 and is shown as negative control. Bars represent means + s.d.; n = 3. (B) HEK293 cells stably expressing the N-terminal half of GFP (GFP1-7) and mouse JUNO stained with a highly avid IZUMO1 probe. (C) HEK293 cells stably expressing the C-terminal half of GFP (GFP8-11) and mouse IZUMO1 stained with a highly avid JUNO. (D) TMEM95 does not induce fusion when expressed in HEK293T cells in the presence of JUNO and IZUMO1 using a GFP-complementation cell fusion assay. HEK293T cells expressing either half of GFP and either JUNO or IZUMO1 were mixed and their fusogenic ability visualized by GFP fluorescence. The IZUMOI-expressing cells were either mock transfected prior to mixing (Control), transfected with Syncitin a, as a positive fusion control, or Tmem95. By contrast to the cells transfected with Syncytin a, Tmem95 did not induce cell fusion. Cell nuclei are stained with DAPI and scale bar represents 20 µm.

Figure 3.

Figure 3—figure supplement 1. Split GFP fusion assay.

Figure 3—figure supplement 1.

(A) Amino acid sequence of FKBP1A underlined in blue fused via a short linker to the N-terminal region of GFP (GFP1-7) (green underline). (B) Amino acid sequence of the C-terminal region of GFP (GFP8-11) highlighted in green, fused to RB, underlined in blue. (C, C´, D and D´) The cells expressing the GFP1-7 stably express mouse Juno and the cells expressing the GFP8-11 stably express mouse Izumo1. To establish functional cell surface expression of both Juno and Izumo1 on the cells, noth cell lines were stained with the corresponding pentameric FLAG-tagged Izumo1 and Juno preys. The cells expressing mouse Juno were stained with the Izumo1 prey but not Juno, and the cells expressing mouse Izumo1 stained with the Juno prey but not Izumo1. Images are representative of at least three independent replicates. Nuclei are stained with DAPI (blue) and scale bars represents 10 µm.
Figure 3—figure supplement 2. Sequence alignments of TMEM95 and IZUMO1.

Figure 3—figure supplement 2.

(A) Structure prediction of TMEM95 created by SWISS-MODEL web-based integrated service. Model TMEM95-P0DJF3 (left) was built using Template 5f4v.1.A (right) from Izumo1 sperm-egg fusion protein 1 (Ohto et al., 2016). Secondary structure alignment shows coloured residues by secondary structure elements: green for β-hairpin and blue for α-helix. (B) Sequence alignment of IZUMO1 from Mus musculus (Q9D9J7) (top) and TMEM95 from Mus musculus (P0DJF3), Sus scrofa (F1ST45), Oryctolagus cuniculus (G1TIM7), Macaca mulatta (F7HLM7) and Homo sapiens (Q3KNT9) are shown. Residues are colored by identity to top sequence (IZUMO1) and to indicate physicochemical properties: dark-gray for mismatch, yellow for cysteine, green for hydrophobic, dull-blue for small alcohol, bright-blue for negative charge, red for positive charge, and purple for polar. The residues of the IZUMO1-JUNO interface binding (Ohto et al., 2016) are indicated by red arrows. cov, percentage of coverage; pid, percentage of identity with top sequence. The sequence identity between mouse and human TMEM95 is 67%.
Open in a new tab

To further investigate the role of TMEM95 during fertilization, we developed a cellular fusion assay based on a split-GFP complementation approach. We first established two cell lines: one expressing mouse JUNO and the N-terminal fragment of GFP (GFP1-7), and another cell line expressing mouse IZUMO1 and a C-terminal fragment of GFP (GFP8-11) (Figure 3—figure supplement 1A and B). It is known that the two GFP fragments are able to reassemble spontaneously and complement fluorescence activity (Kondo et al., 2010; Pedelacq and Cabantous, 2019). The expression of functional JUNO and IZUMO1 on the surface of the cells was demonstrated by their ability to specifically bind to IZUMO1 and JUNO soluble protein probes respectively (Figure 3B and C; Figure 3—figure supplement 1C, C', D and D'). GFP8-11 Izumo1 cells transiently transfected with an expression plasmid encoding the full length Tmem95 cDNA were mixed with the GFP1-7 Juno cells and fusion events were scored by the presence of green cells (Figure 3D). By contrast to Syncytin-a transfected cells which served as a positive control, cells transfected with Tmem95 did not induce any detectable cell fusion events even in the presence of JUNO and IZUMO1.

Discussion

The search for candidate sperm proteins involved in fertilization has been particularly challenging and to date only another sperm protein, IZUMO1 was proved to be involved in this process (Inoue et al., 2005). Highlighting these difficulties, a recent article has reported that 30 genes that have a strong bias of expression in the testis are dispensable for reproduction (Lu et al., 2019). Also, on the sperm side, SPACA6 was found to play a role in sperm-egg membrane fusion based on the reproductive phenotype observed in BART97b mutant male mice. The random mutation harboured by these infertile males consists of a large insertion that introduced an 11 kb deletion disrupting Spaca6 gene expression (Lorenzetti et al., 2014). During the revision of this manuscript, the essential role of SPACA6 on fertilization has been confirmed by targeted gene deletion by two independent groups (Barbaux et al., 2020; Noda et al., 2020). The discovery of TMEM95, a novel sperm protein essential for gamete fusion highlights that IZUMO1-JUNO interaction alone is not sufficient to mediate this process and suggests that more complex molecular interactions are required. Interestingly, IZUMO1-JUNO alone are not able to mediate fusion, as their interaction appears to mediate sperm-egg adhesion, not acting as fusogens (Bianchi et al., 2014).

The lack of interaction between TMEM95 and JUNO was an unexpected outcome, as the predicted structure of TMEM95 contains an α-helix equivalent to that found in IZUMO1 domain and the β-hairpin secondary structure, which provides the main platform for JUNO binding (Ohto et al., 2016). However, the residues on TMEM95 predicted β-hairpin differ from those present at IZUMO1 at the physicochemical level (Figure 3—figure supplement 1), suggesting that they can be responsible for JUNO recognition. Indeed, Trp-mediated interactions described for IZUMO1-JUNO binding ensure the conserved binding mode, but variable regions in the interface have been suggested to determine species-specificity (Bianchi and Wright, 2015; Ohto et al., 2016). It is possible that the structural domains shared by TMEM95 and IZUMO1 constitute a common feature of sperm proteins involved in fertilization, serving as mediators of protein-to-protein interactions. Another possibility would be that TMEM95 plays an architectural role on the sperm membrane similar to CD9. The mechanism by which the egg membrane protein CD9 intervenes in membrane fusion is not yet understood, with most evidence suggesting that it plays an architectural role that does not require a binding partner (Chalbi et al., 2014; Jégou et al., 2011). However, in contrast to CD9 ablation, which only causes complete infertility when combined with CD81 deficiency (Rubinstein et al., 2006), TMEM95 ablation alone completely blocks gamete fusion and thus, subsequent sperm entry into the ooplasm without generating any obvious structural defect in the sperm, thereby suggesting that TMEM95 may not play a major architectural role.

The role of TMEM95 in post-binding events may require one or several binding partners as, alone, it is unable to promote cell fusion following IZUMO1-JUNO binding in HEK293T cells. The residues located at TMEM95 predicted β-hairpin differing from those present in IZUMO1 β-hairpin could mediate affinity to other sperm or oocyte protein/s required for post-binding events. Given that the AVEXIS assay is performed exclusively using the extracellular domains of cell surface proteins, it is also possible that IZUMO1 and TMEM95 might interact via their transmembrane domains following IZUMO1-JUNO binding. While our manuscript was undergoing review, another article confirming our results on the essential role of TMEM95 in fertilization was submitted and published (Noda et al., 2020). This article uncovered the essential role of a sperm protein termed SOF1 on mammalian fertilization, suggesting that TMEM95, SPACA6 and SOF1 may function cooperatively with IZUMO1. However, the nature of this cooperation remains unclear, as the ablation of any of these three genes does not affect the amount and localization of IZUMO1 (Barbaux et al., 2020; Noda et al., 2020 and our results). Besides, whereas co-immunoprecipitation analysis performed on cells overexpressing TMEM95, SPACA6, SOF1 and FIMP suggested interaction of these proteins with IZUMO1, this interaction could not be confirmed on testis lysates (Noda et al., 2020), and we failed to detect any IZUMO1:TMEM95 interaction on a high avidity binding assay capable of detecting subtle protein interactions. In contrast to IZUMO1, TMEM95, SPACA6 and SOF1 do not seem to exert a relevant role on sperm-oocyte binding (Barbaux et al., 2020; Noda et al., 2020 and our results); therefore, their fundamental roles must be played in post-binding events. Together, these recent findings demostrate that gamete fusion is a complex fascinating puzzle involving proteins that have evolved for this specific task.

The discovery of the crucial function of TMEM95, a novel sperm membrane protein required for mammalian gamete fusion, provides a relevant insight into this process, arguably the most important membrane fusion event in a mammal´s life. This finding also paves the way for the investigation of the involvement of TMEM95 alterations in male infertility in humans and for the development of potential immune- or molecular-based contraceptive methods.

Materials and methods

Key resources table.

Reagent type
(species) or
resource		 Designation Source or
reference		 Identifiers Additional
information
Strain, strain background (Mus musculus) Tmem95 KO This article Generated by CRISPR (M and M), available upn request
Antibody Rabbit Anti-TMEM95 MyBioSource MBS7004333
Antibody Rabbit Anti-IZUMO1 AbCam Ab211623; RRID:AB_2650506
Antibody Rat Anti-Izumo Gift from Dr. Ikawa Clone KS139-34
Antibody Mouse Anti-Tubulin Sigma T8328, RRID:AB_1844090
Antibody Rabbit anti-Flag Sigma F7425, RRID:AB_439687
Recombinant DNA reagent Plasmid pMJ920 (Jinek et al., 2013) RRID:Addgene_42234
Recombinant DNA reagent Plasmid px330 vector (Yang et al., 2014) RRID:Addgene_42230
Sequence-based reagent Primers This article Detailed on M and M (ordered from Sigma)
Open in a new tab

Animal models

All experimental procedures were approved by the INIA Animal Care Committee and Madrid Region Authorities (PROEX 040/17) in agreement with European legislation. Tmem95 KO mice were generated by CRISPR technology as detailed below.

In silico protein structure determination

Models were computed by the SWISS-MODEL server homology modelling pipeline (Guex et al., 2009; Waterhouse et al., 2018), which relies on ProMod3, an inhouse comparative modelling engine based on OpenStructure. The SWISS-MODEL template library (SMTL version 2019-03-20, PDB release 2019-03-15) was searched with BLAST and HHBlits for evolutionary related structures matching the target sequence TMEM95 (P0DJF3). The global and per-residue model quality was assessed using the QMEAN scoring function (Benkert et al., 2011); the score value of −3.11 was highlighted by a change of the ‘thumbs-up’ by the software.

sgRNA design and in vitro transcription sgRNA was designed against the first exon of Tmem95 using bioinformatic tools to minimize the chances of offtarget genome edition (https://crispr.mit.edu). A sgRNA targeting the first exon of TMEM95 (ATTTCCACTTCGGCTTCGTC, NC_000077.6, Score 94) was synthesized and purified using Guide-it sgRNA In Vitro Transcription Kit (Takara). Capped polyadenylated Cas9 mRNA was generated by in vitro transcription (mMESSAGE mMACHINE T7 ULTRA kit, Life Technologies) using as template the plasmid pMJ920 (Addgene 42234) linearized with BbsI (NEB). mRNA was purified using MEGAClear kit (Life Technologies).

Generation of Tmem95 KO

C57CBAF1 female mice (7–8 weeks old) were superovulated by intraperitoneal injections of 5 IU of pregnant mare serum gonadotropin (PMSG, Folligon, MSD Animal Health) and an equivalent dose of human chorionic gonadotropin (hCG, Sigma) at a 48 hr interval. Superovulated female mice were mated with C57CBAF1 stud males and zygotes were recovered from oviducts.

Microinjections were performed with a micromanipulation system (Narishige MMO-202ND, MM-88) equipped on a Nikon Diaphot TMD inverted microscope. A mixture of 150 ng/µl of Cas9 mRNA and 50 ng/µl of sgRNA was delivered into the cytoplasm of zygotes (3–5 pl) using a filament needle (Bermejo-Álvarez et al., 2015).

Following microinjection, embryos were cultured in EmbryoMax KSOM Mouse Embryo Media (Millipore) at 37°C under 5% CO2 for 4 days until they reached the blastocyst stage. Six blastocysts were transferred to a pseudopregnant Swiss recipient 2.5 days post-coitum (dpc), resulting in the birth of two pups: one male carrying in-frame mutations and one founder female carrying an in-frame indel and a frameshift allele (KO allele). This F0 female was crossed with C57BL/6 males to obtain heterozygous mutants. Heterozygous F1 individuals harbouring the KO allele were intercrossed to produce WT, Hz or KO individuals used for the experiments.

Mice genotyping

Genomic DNA samples from tail biopsies were prepared using FavorPrep Tissue Genomic DNA Extraction Mini Kit (Favorgen). PCR was performed using primers spanning the target sequence (F: 5’-CCCCCTTAGGATCCAGTGGT-3’, R: 5’-ACTCCTTCCATACCCCAGCA-3’, 255 bp product in WT) under the following conditions: 95°C for 2 min; 35 x (94°C for 20 s, 60°C for 30 s, 72°C for 30 s); 72°C for 5 min; hold at 8°C. The PCR products from F0 mice were purified using FavorPrep PCR Purification Kit (Favorgen), cloned into pMD20 T-vectors (Takara) using Blunt TA ligase (NEB), and transformed into DH5-α competent cells (ThermoFisher Scientific). Ten positive plasmid clones from each transformation were purified (Favorgen) and Sanger sequenced (Stabvida) to uncover the alleles generated following CRISPR-mediated edition harboured by each individual (Lamas-Toranzo et al., 2019).

Off-target analysis in the founder female was performed by analysing the 5 most probable off-target sequences detected by the sgRNA design bioinformatics tool (https://crispr.mit.edu). PCR were performed using the same conditions that above and the following primers: OT1 (F: 5´- ACTCTGTTCATCACCATGAGTCAA-3, R: 5´- TGGCTTCCTTGTCTATGGTGT-3´, 264 bp product), OT2 (F: 5´- AAACCAATGAGATCCGCCGT-3´, R: 5´-TAGTTGCCAGGTTCGACCAC-3´, 223 bp product), OT3 (F: 5´- ACACACACGAGACTCGACAA-3´, R: 5´- TGCAAGATCTACACACGATCCA-3´, 233 bp product), OT4 (F: 5´- TGCAACAGAAGCAGGATGGT-3´, R: 5´- AACCCAGGCAGAAGAAGTGG-3´, 331 bp product) and OT5 (F: 5´- GTGCTGTGTCTGTTGCTTTT-3´, R: 5´- TGTAGTTTGGCCAGTTGTGC-3´, 220 bp product). PCR products were purified as above, Sanger sequenced and compared with the sequence of a WT individual of the same background.

Subsequent generations were genotyped by Sanger sequencing of PCR products until we developed a quantitative PCR high-resolution melting (qPCR-HRM) curve analysis that allowed us an easy detection of WT, Hz and KO individuals. qPCRs were performed using and Mic qPCR cycler (Bio Molecular Systems) with primers flanking the target sequence (F 5’-GGGAAGCCCATTCCTTCCAT-3’, R 5’-GATCATGCCTTGGCAAACCG-3’, 79 bp product in WT). Reaction conditions were as follows: 40x (94°C for 15 s, 56°C for 30 s, 72°C for 20 s). Melting curves were visualized using Mic qPCR hardware (Bio Molecular Systems) and contrasted with those obtained from known WT, Hz and KO samples confirmed by Sanger sequencing.

Protein extraction and immunobloting

WT and KO sperm were collected from cauda epididymis in PBS supplemented with 0.1% PVP and centrifuged at 3,000 g for 7 min. Pellets and tissues samples (testis and accessory glands) were snap frozen in liquid nitrogen and kept at −80°C until analysis. For the analysis, frozen pellets were re-suspended in reducing SDS Sample Buffer (4X) (Millipore) and boiled for 10 min (lysis protocol #1) or in 1% Octyl β-D-Glucopyranoside (Sigma) solution in PBS and incubated on ice for 30 min (Nishimura et al., 2011). Cell suspensions from testis and accessory glands were obtained by thoroughly mincing tissues with razor blades. The cell pellet was re-suspended in 500 µl of 50 mM Tris-HCl ph 7.5, 1 mM EDTA, 1% Igepal, 0.1 mM PMSF, 10 mM iodoacetamide, 10 mM N-ethylmaleimide, phosphatase inhibitor and protease inhibitor, homogenized and incubated for 30 min on strong agitation at 4°C and then centrifuged at 20,000 g for 20 min at 4°C. After incubation, samples were centrifuged at 20,000 g for 10 min at 4°C, supernatants were separated in 16% SDS-PAGE and proteins were transferred to PVDF membranes. Subsequently, membranes were blocked with 5% BSA in TBST 1X for 1 hr at RT and incubated overnight in primary antibody solution in TBST 1 × 1% BSA. The primary antibodies used were anti-TMEM95 (1:1000 v/v, MyBioSource MBS7004333), anti-IZUMO1 (1:1000 v/v, Abcam ab211623) and anti-β-Tubulin (1:5000 v/v, Sigma T8328). Membranes were washed on the following day three times for 10 min with TBST 1X, incubated with corresponding peroxidase-conjugated secondary antibody and washed three times for 10 min with TBST 1X prior to visualization by chemiluminescence (Pierce ECL-Plus, Thermo Fisher Scientific).

The ~20 kDa band present on TMEM95 blot for WT samples but absent in KO samples was cut out and processed for proteomic analysis to confirm TMEM95 identity. Data processing was performed with Data Analysis program for LC/MSD Trap Version 3.3 (Bruker Daltonik) and Spectrum Mills MS Proteomics Workbench (Rev A.03.02.060B, Agilent Technologies) by the Molecular Biology Section, Service of Support to the Experimental Sciences (SACE), University of Murcia.

The bands in IZUMO and β-Tubulin blots were quantified by densitometric scanning and analysed with the ImageQuant TL software v2005 (GE Healthcare). Izumo1/Tubulin ratio was calculated in arbitrary units with data from sperm from 4 WT and 4 KO animals. The obtained data was subjected to Student’s t-test and the level of significance was set at p<0.05. The software used was IBM SPSS Statistics (v22.0).

Transcriptional analysis in reproductive tissues

Transcriptional analysis was performed as previously described (Bermejo-Alvarez et al., 2010). Briefly, total RNA was collected from testis, seminal vesicle, prostate and epididymis samples (3 samples/tissue) obtained from WT males using Trizol (Invitrogen). Following DNAse treatment (Promega), RNA was retrotranscribed (qScript, Quantabio) to cDNA (Bermejo-Álvarez et al., 2015). Tmem95 and Gapdh transcripts were detected on cDNA by PCR using the amplification cycle described in the genotyping section. Primers to detect Tmem95 were the same as those used for HRM-based genotyping (79 bp product). Primers used for Gapdh were F 5’-ACCCAGAAGACTGTGGATGG-3’, R 5’-ATGCCTGCTTCACCACCTTC-3’ (247 bp product). DNAse-treated non-retrotranscribed RNA obtained from testis served as negative control for DNAse treatment and PCR.

Sperm immunocytochemistry

Sperm from WT or KO individuals was recovered from the cauda epididymis in PBS supplemented with 0.1% PVP. For IZUMO1 and TMEM95 relocalization analysis, sperm were acrosome reacted by 20 min incubation in HTF medium supplemented with 20 µM calcium ionophore. Following centrifugation (3,000 g for 7 min), sperm were fixed in 4% PFA in PBS for 5 min and washed twice in PBS. Samples were then permeabilized with 0.1% Triton X-100 in PBS for 10 min, and blocked with 5% FCS in PBS for 45 min at 4°C. Next, samples were incubated with anti-TMEM95 antibody (1:100, MyBioSource) or anti-IZUMO1 antibody (1:100, a gift by Dr. Ikawa, clone KS139-34) overnight at 4°C. For double TMEM95 and IZUMO1 immunocytochemistry, samples were first incubated overnight with anti-TMEM95 followed by another overnight incubation with anti-IZUMO1. To determine TMEM95 localization within the acrosomal cap, the permeabilization step (10 min incubation in 0.1% Triton X-100) was skipped. As TMEM95, IZUMO1 and PNA were still detected in non-permeabilizing conditions, the fixation protocol was substituted with 30 min incubation in 4% PFA at 4°C (Nishimura et al., 2011), which still yielded similar results. Alexa Fluor 596 goat anti-rabbit IgG antibody (1:500, Invitrogen, for TMEM95) or Alexa Fluor 488 or 594 anti-rat IgG antibodies (1:500, Invitrogen, for IZUMO1) were used as a secondary antibody and incubation was performed for 2 hr at RT. Finally, samples were incubated for 5 min with 1 µg/ml FITC-PNA conjugate (Sigma) and 1 µg/ml Hoechst 33342 (Sigma). Samples were mounted and subsequently observed under an epifluorescence inverted microscope (Nikon Eclipse TE300) and NIS software (Nikon).

In vivo fertilization analysis

C57CBAF1 female mice (7–8 weeks old) were superovulated as described above and mated with WT, Hz or KO individuals (4 individuals/group). Embryos were recovered from the oviduct on 1.5 dpc and cultured in vitro as described above. Cleavage rate was assessed on 1.5 dpc and blastocyst rate on 4.5 dpc. Statistical differences were analysed by ANOVA (SigmaStat package) and the level of significance was set at p<0.05.

In vitro fertilization (IVF)

Sperm from WT, Hz or KO individuals was recovered from the cauda epididymis in HTF medium and placed in the bottom of a previously equilibrated 300 µl drop of HTF covered with mineral oil for 2 hr prior to IVF. Following this pre-incubation time, the upper 150 µl of the drop were collected, and sperm concentration was analysed. Cumulus-oocytes complexes (COCs) were recovered from the oviducts of superovulated female mice 14 hr after hCG injection and placed in a 4-well dish with 400 µl of Human tubal fluid (HTF) medium in groups of ~40 COCs per well. Previously prepared sperm were immediately added to the well containing COCs at a final concentration of 106 sperm/ml. After 4 hr of co-incubation, presumptive zygotes were sequentially washed in M2 and KSOM medium and cultured as described above. Statistical differences were analysed by ANOVA (SigmaStat package) and the level of significance was set at p<0.05.

Sperm motility analysis

Motility was analysed in sperm recovered from the cauda epididymis of WT, Hz and KO males (3 individuals/group). For CASA analysis, 20 µL of sperm suspension (2 × 106 sperm/ml) were placed on a pre-warmed slide placed on a stage heated to 37°C and observed on an inverted microscope (Nikon Eclipse 50i) fitted with a digital camera (Basler A312f) capable of recording 25 frames/s. Five 1 s videos (20–60 moving sperm) were recorded in different fields and analyzed using the Integrated Semen Analysis System (ISAS). The parameters analysed (Mortimer, 2000) were straight-line velocity (VSL; time-averaged velocity of the sperm head along a straight line from its first position to its last position, expressed in µm/s); curvilinear velocity (VCL; time-averaged velocity of the sperm head along its actual curvilinear path, expressed in µm/s); average path velocity (VAP; velocity over an average path generated by a roaming average between frames, expressed in µm/s); linearity (LIN) (defined as (VSL/VCL)×100); straightness (STR) (defined as (VSL/VAP)×100); wobble (WOB) (defined as (VAP/VCL)×100); amplitude of lateral head (ALH) displacement (width of the lateral movement of the sperm head, expressed in µm) and beat cross frequency (BCF; number of times the sperm head crosses the direction of movement per second, expressed in Hz).

Transmission electron microscopy analysis

WT and KO sperm were collected from the cauda epididymis in PBS supplemented with 0.1% PVP and centrifuged at 3000 g for 7 min. Pellet was resuspended in 2% glutaraldehyde in PBS and incubated for 2 hr at 4°C. After fixation, sperm were post-fixed in potassium ferrocyanide reduced osmium tetroxide for 1 hr. Following extensive washing, the samples were then dehydrated through a graded series of ethanol and processed for embedding in Epon 812. Ultrathin sections were obtained with an ultramicrotome (Microm International GmbH) and mounted on formvar coated nickel grids. Ultrathin sections were counterstained with uranyl acetate followed by lead citrate and imaged in a Jeol JEL-1011 Transmission Electron Microscope.

Sperm penetration and binding assays

Sperm penetration and sperm and TMEM95:Cherry egg binding assays were performed as described previously (Bianchi et al., 2014; Inoue et al., 2005). Briefly, COCs were collected from superovulated females 14 hr after hCG injection as described above. Cumulus cells were removed by incubation in 300 µg/ml hyaluronidase (Sigma) solution in M2 medium. Zona pellucida was removed by brief incubation in Acidic Tyrode´s medium. For the sperm penetration assay, zona-free mouse eggs were pre-incubated in HTF with Hoechst 33342 1 µg/ml for 10 min and washed before sperm addition. After 30 min of gametes co-incubation, the eggs were fixed in a 0.25% glutaraldehyde solution in PBS. For the sperm binding assay, eggs were also incubated for 30 min with sperm at a concentration of 1 million sperm/ml and stained with Hoechst after fixing; sperm from 3 KO and 3 WT males were tested on 12 eggs/male.

Two recombinant Cherry-tagged TMEM95 proteins were produced for protein-egg binding assay. Expression plasmids (pcDNA3.1[+]) were designed and constructed (GeneArt) to encode mouse TMEM95 protein (UniProt P0DJF3) fused to mCherry and FLAG tags as depicted in Figure 2—figure supplement 1B. TMEM95 was fused in-frame to mCherry, with mCherry inserted near the C-terminus of TMEM95 (TMEM95-Ccherry) or just downstream of the signal peptide of TMEM95 (TMEM95-Ncherry), separated in both cases by a 10 amino acid linker (GGGGSGGGGS). FLAG-tag was added to the C-terminus of both fusion proteins. Following DNA sequence verification, TMEM95-Ccherry and TMEM95-NCherry expression plasmids were amplified in DH5α competent cells (ThermoFisher Scientific) and purified using the GenEluted Plasmid Kit. Chinese Hamster Ovary cells (CHO cells, ECACC, The European Collection of Authenticated Cell Cultures) were grown (37°C, 5% CO2 and 95% humidity) for 48–72 hr to 80–90% confluence using F-12 medium (Biowest) supplemented with 10% fetal bovine serum (Biowest) and 100 U/mL penicillin-streptomycin (Gibco). Transient transfections were performed with Lipotransfectina (Solmeglas). For each transfection, 4 µL Lipotransfectina transfection reagent were added to a final volume of 200 µL Opti-MEM reduced-serum medium (Gibco) pre-dissolved with 2 µg template plasmid and incubated for 15 min at room temperature (RT). The complex was diluted by adding 2 ml Opti-MEM and overlaid on growing cells (37°C, 5% CO2 and 95% humidity). The medium containing the secreted proteins was collected after 48 hr, centrifuged at 4,000 g for 5 min at 4°C to remove cell debris, and concentrated in Vivaspin Turbo 4 of 10,000 Da (Sartorius). A final volume of 200–300 µL of concentrated proteins was obtained in 20 mM sodium phosphate buffer, pH 7.4 with protease inhibitor (EDTA-free EASYpack, Roche). Cell growing medium containing concentrated proteins was separated by SDS-PAGE and transferred to PVDF membranes which were probed with the primary antibody anti-Flag (Sigma F7425) at 1:1000 v/v in TBST 1X, 1% BSA, prior to visualization by chemiluminescence (Pierce ECL-Plus, Thermo Fisher Scientific). For the recombinant TMEM95:Cherry protein binding assay, zona-free eggs were exposed for 30 min to serial concentrations of recombinant proteins (0 to 0.9 mg/ml) diluted in protein-free medium; following incubation they were briefly washed in PBS and immediately observed under fluorescence microscopy. All specimens were observed under an epifluorescence inverted microscope (Nikon Eclipse TE300) and NIS software (Nikon).

Intracytoplasmic sperm injection

Cauda epididymis sperm were collected in M2 medium from KO and WT males (3 individuals/group). Sperm was mixed with 5 volumes of a 10% solution of polyvinyl-pyrrolidone in M2. ICSI was performed in M2 medium at room temperature (Fernández-Gonzalez et al., 2008). Following sperm head injection, eggs were allowed to recover for 15 min in M2 at room temperature. Surviving eggs were cultured until the blastocyst stage in the same conditions as described above.

Recombinant protein production and protein interaction screening by AVEXIS

The regions encoding the entire extracellular domains of IZUMO1 (Q9D9J7), Juno (Q9EQF4) and Tmem95 (P0DJF3) flanked by unique NotI and AscI sites were made by gene synthesis (Invitrogen, GeneArt Gene Synthesis) and cloned into protein expression vectors that encoded either biotinylated ‘baits’ or pentameric ‘preys’ tagged at the C-terminus with the beta-lactamase enzyme, the FLAG epitope and 6 histidines. All the ectodomains were expressed as soluble secreted proteins by transient transfections of HEK293-6E cells as described (Bushell et al., 2008). The rat Cd200 and Cd200R were used as positive controls.

Bait and prey proteins were normalized to activities suitable for the AVEXIS assay as previously described (Kerr and Wright, 2012). Biotinylated baits that had been dialyzed against HBS were immobilized on streptavidin-coated 96-well microtiter plates (Greiner Bio-One). Preys were then incubated for one hour, washed with HBS/0.1% Tween-20 three times and once with HBS. Hydrolysis of the beta-lactamase substrate Nitrocefin (Cayman Chemical) was quantified by reading absorbance values at 485 nm with a TECAN Spark plate reader. Nitrocefin was added at 125 µg/ml. The assay was repeated more than three times using independent protein preparations.

Cell fusion assay: design of the expression vectors and establishment of stably transfected cell lines

To assess the ability of membrane proteins to induce cell fusion in vitro, we developed a complementation assay similar to the system described previously (Kondo et al., 2010) which exploits the ability of two GFP fragments to reassemble spontaneously and to reconstitute the fluorescent protein. We generated two cell lines each expressing one part of the GFP; one fragment is named GFP1-7 and contains the first 7 beta sheets of the protein while the second fragment is namedGFP8-11 and is made of the last 4 beta sheets.

To force a quicker reassembly of the GFP fragments, in case it was needed, we exploited the oligomer-formation system inducible by rapamycin. We tagged FKBP1A and FRB to the split GFP because the two proteins interact with high affinity only in the presence of rapamycin (Banaszynski et al., 2005). The DNA encoding for FKBP1A linked to the N-terminus of GFP1-7 and that of FRB linked to the C-terminus of GFP8-11 were obtained by gene synthesis (Invitrogen, GeneArt Gene Synthesis), digested and ligated into the pIRESHyg3 vector (Clontech) using NheI and BamHI restriction sites.

Clonal cell lines stably expressing the split GFP were established by transfecting HEK293T with the Lipofectamine 200 (ThermoFisher Scientific) and culturing them in selecting medium (61965–026 DMEM with Glutamax from ThermoFisher Scientific supplemented with 10% FBS and 250 µg/ml Hygromycin B). Single cells were sorted with the MoFlo XDP Cell sorter (Beckman Coulter), expanded and functionally tested. GFP1-7 and GFP8-11 cell clones were mixed together and transfected with a plasmid encoding for the mouse gene syncytin A (Origene untagged clone MC219753); the formation of green syncytia containing more than two nuclei was scored 48 hr after the transfection.

Mouse Juno was amplified by PCR from the cDNA of immature oocytes obtained from the cDNA library NIH_MGC_257_N (Bianchi et al., 2014). Mouse Izumo1 was amplified from the ORF clone MG222708 (OriGene Technologies, Inc). Both were ligated in the pIRESPuro3 vector (Clonetech) using EcoRV-NotI and EcoRI-NotI restriction sites, respectively. Finally, the plasmid encoding Juno was transfected in the GFP1-7 cell line and Izumo1-encoding plasmid was transfected in the GFP8-11 cell line. Establishment of stable clones was obtained by selection with 10 µg/ml puromycin and single cell sorting. The expression of JUNO and IZUMO1 was assessed by staining with IZUMO1 and JUNO soluble recombinant pentamers, as previously described (Bianchi et al., 2014). Cells were incubated with the pentameric probes (preys) for 30 min at 37°C, fixed in 4% PFA and stained with an anti-FLAG monoclonal antibody Cy3 conjugated (A9594, clone M2, Sigma-Aldrich) for 1 hr at room temperature. Finally, cells were transferred to a microscope glass in Slowfade Gold mountant with DAPI (S36938 ThermoFisher Scientific) and images were acquired with the Leica TCS SP5 confocal microscope.

Acknowledgements

IZUMO1 antibody was a generous gift by Dr Masahito Ikawa (Osaka University, Japan). Funding was provided by the Spanish Ministry of Science and Competitiveness (RYC-2012–10193, AGL2014-58739-R and AGL2017-84908-R to PBA, AGL2015-70159-P to MJM, RTI2018-093548-B-I00 to AGA and the network project AGL2016-71890-REDT), European Research Council (StG 757886-ELONGAN to PBA), Fundación Seneca-Agencia de Ciencia y Tecnología de Murcia (20887/PI/18 to MJM), the Irish Department of Agriculture, Food and The Marine (11/S/104), and the United Kingdom Medical Research Council (MR/M012468/1 to EB and GJW). ILT and SPC were supported by FPI and Ramón y Cajal contracts, respectively, from the Spanish Ministry of Science and Competitiveness. BFF is supported by a Marie Curie Fellowship from the European Commission.

Funding Statement

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

Contributor Information

María Jiménez-Movilla, Email: mariajm@um.es.

Pablo Bermejo-Álvarez, Email: bermejo.pablo@inia.es.

Polina V Lishko, University of California, Berkeley, United States.

Anna Akhmanova, Utrecht University, Netherlands.

Funding Information

This paper was supported by the following grants:

  • Ministerio de Economía y Competitividad RYC-2012-10193 to Pablo Bermejo-Álvarez.

  • Ministerio de Economía y Competitividad AGL2014-58739-R to Pablo Bermejo-Álvarez.

  • Ministerio de Economía y Competitividad AGL2017-84908-R to Pablo Bermejo-Álvarez.

  • Ministerio de Economía y Competitividad AGL2015-70159-P to María Jiménez-Movilla.

  • Ministerio de Economía y Competitividad RTI2018-093548-B-I00 to Alfonso Gutiérrez-Adán.

  • Ministerio de Economía y Competitividad AGL2016-71890-REDT to Alfonso Gutiérrez-Adán, María Jiménez-Movilla, Pablo Bermejo-Álvarez.

  • H2020 European Research Council StG 757886-ELONGAN to Pablo Bermejo-Álvarez.

  • Fundación Séneca 20887/PI/18 to María Jiménez-Movilla.

  • Department of Agriculture, Food and the Marine 11/S/104 to Pat Lonergan.

  • Ministerio de Economía y Competitividad FPI fellowship to Ismael Lamas-Toranzo.

  • Ministerio de Economía y Competitividad Ramón y Cajal contract to Serafín Pérez-Cerezales.

  • European Union Seventh Framework Programme Marie Curie fellowship to Beatriz Fernández-Fuertes.

  • Medical Research Council MR/M012468/1 to Enrica Bianchi, Gavin J Wright.

Additional information

Competing interests

No competing interests declared.

Author contributions

Writing - original draft, Writing - review and editing, Generation of KO model, breeding and genotyping of the colony, in vivo and in vitro fertility tests, immunocytochemistry, sperm fusion and binding assays.

Writing - original draft, Writing - review and editing, Immunocytochemistry, Western-Blots, electron microscopy analysis, protein-egg binding assay.

Formal analysis, Writing - original draft, Writing - review and editing, AVEXIS and cell fusion assays.

Writing - review and editing, Immunocytochemistry.

Sperm motility analysis.

Intra-cytoplasmic sperm injection.

Intra-cytoplasmic sperm injection.

Supervision, Funding acquisition, Writing - review and editing.

Supervision, Funding acquisition.

Supervision, Funding acquisition, Writing - original draft, Writing - review and editing.

Formal analysis, Supervision, Funding acquisition, Writing - original draft, Writing - review and editing, In silico protein structure determination, Western-Blots, electron-microscopy analysis, protein-egg binding assay, inter-species sequence analysis.

Conceptualization, Resources, Formal analysis, Supervision, Funding acquisition, Writing - original draft, Writing - review and editing, Project coordinator, generation of KO model, off-target analysis, breeding and genotyping of the colony, in vivo and in vitro fertility tests, immunocytochemistry, transcriptional analysis, protein-egg binding assay.

Ethics

Animal experimentation: All experimental procedures were approved by INIA Animal Care Committee and Madrid Region Authorities (PROEX 040/17) in agreement with European legislation.

Additional files

Transparent reporting form

Data availability

All data generated or analysed during this study are included in the manuscript and supporting files.

References

  1. Banaszynski LA, Liu CW, Wandless TJ. Characterization of the fkbp.rapamycin.frb ternary complex. Journal of the American Chemical Society. 2005;127:4715–4721. doi: 10.1021/ja043277y. [DOI] [PubMed] [Google Scholar]
  2. Barbaux S, Ialy-Radio C, Chalbi M, Dybal E, Homps-Legrand M, Do Cruzeiro M, Vaiman D, Wolf JP, Ziyyat A. Sperm SPACA6 protein is required for mammalian Sperm-Egg adhesion/Fusion. Scientific Reports. 2020;10:5335. doi: 10.1038/s41598-020-62091-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Benkert P, Biasini M, Schwede T. Toward the estimation of the absolute quality of individual protein structure models. Bioinformatics. 2011;27:343–350. doi: 10.1093/bioinformatics/btq662. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Bermejo-Alvarez P, Rizos D, Rath D, Lonergan P, Gutierrez-Adan A. Sex determines the expression level of one third of the actively expressed genes in bovine blastocysts. PNAS. 2010;107:3394–3399. doi: 10.1073/pnas.0913843107. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Bermejo-Álvarez P, Ramos-Ibeas P, Park KE, Powell AP, Vansandt L, Derek B, Ramirez MA, Gutiérrez-Adán A, Telugu BP. Tet-mediated imprinting erasure in H19 locus following reprogramming of spermatogonial stem cells to induced pluripotent stem cells. Scientific Reports. 2015;5:13691. doi: 10.1038/srep13691. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Bianchi E, Doe B, Goulding D, Wright GJ. Juno is the egg izumo receptor and is essential for mammalian fertilization. Nature. 2014;508:483–487. doi: 10.1038/nature13203. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Bianchi E, Wright GJ. Cross-species fertilization: the hamster egg receptor, juno, binds the human sperm ligand, Izumo1. Philosophical Transactions of the Royal Society B: Biological Sciences. 2015;370:20140101. doi: 10.1098/rstb.2014.0101. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Bushell KM, Söllner C, Schuster-Boeckler B, Bateman A, Wright GJ. Large-scale screening for novel low-affinity extracellular protein interactions. Genome Research. 2008;18:622–630. doi: 10.1101/gr.7187808. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Chalbi M, Barraud-Lange V, Ravaux B, Howan K, Rodriguez N, Soule P, Ndzoudi A, Boucheix C, Rubinstein E, Wolf JP, Ziyyat A, Perez E, Pincet F, Gourier C. Binding of sperm protein Izumo1 and its egg receptor juno drives Cd9 accumulation in the intercellular contact area prior to fusion during mammalian fertilization. Development. 2014;141:3732–3739. doi: 10.1242/dev.111534. [DOI] [PubMed] [Google Scholar]
  10. Ellerman DA, Pei J, Gupta S, Snell WJ, Myles D, Primakoff P. Izumo is part of a multiprotein family whose members form large complexes on mammalian sperm. Molecular Reproduction and Development. 2009;76:1188–1199. doi: 10.1002/mrd.21092. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Fernandez-Fuertes B, Laguna-Barraza R, Fernandez-Gonzalez R, Gutierrez-Adan A, Blanco-Fernandez A, O'Doherty AM, Di Fenza M, Kelly AK, Kölle S, Lonergan P. Subfertility in bulls carrying a nonsense mutation in transmembrane protein 95 is due to failure to interact with the oocyte vestments. Biology of Reproduction. 2017;97:50–60. doi: 10.1093/biolre/iox065. [DOI] [PubMed] [Google Scholar]
  12. Fernández-Gonzalez R, Moreira PN, Pérez-Crespo M, Sánchez-Martín M, Ramirez MA, Pericuesta E, Bilbao A, Bermejo-Alvarez P, de Dios Hourcade J, de Fonseca FR, Gutiérrez-Adán A. Long-term effects of mouse intracytoplasmic sperm injection with DNA-fragmented sperm on health and behavior of adult offspring. Biology of Reproduction. 2008;78:761–772. doi: 10.1095/biolreprod.107.065623. [DOI] [PubMed] [Google Scholar]
  13. Fujihara Y, Lu Y, Noda T, Oji A, Larasati T, Kojima-Kita K, Yu Z, Matzuk RM, Matzuk MM, Ikawa M. Spermatozoa lacking fertilization influencing Membrane protein (FIMP) fail to fuse with oocytes in mice. PNAS. 2020;117:9393–9400. doi: 10.1073/pnas.1917060117. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Guex N, Peitsch MC, Schwede T. Automated comparative protein structure modeling with SWISS-MODEL and Swiss-PdbViewer: a historical perspective. Electrophoresis. 2009;30:S162–S173. doi: 10.1002/elps.200900140. [DOI] [PubMed] [Google Scholar]
  15. Inoue N, Ikawa M, Isotani A, Okabe M. The immunoglobulin superfamily protein izumo is required for sperm to fuse with eggs. Nature. 2005;434:234–238. doi: 10.1038/nature03362. [DOI] [PubMed] [Google Scholar]
  16. Jégou A, Ziyyat A, Barraud-Lange V, Perez E, Wolf JP, Pincet F, Gourier C. CD9 tetraspanin generates fusion competent sites on the egg membrane for mammalian fertilization. PNAS. 2011;108:10946–10951. doi: 10.1073/pnas.1017400108. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Jinek M, East A, Cheng A, Lin S, Ma E, Doudna J. RNA-programmed genome editing in human cells. eLife. 2013;2:e00471. doi: 10.7554/eLife.00471. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Kaji K, Oda S, Shikano T, Ohnuki T, Uematsu Y, Sakagami J, Tada N, Miyazaki S, Kudo A. The gamete fusion process is defective in eggs of Cd9-deficient mice. Nature Genetics. 2000;24:279–282. doi: 10.1038/73502. [DOI] [PubMed] [Google Scholar]
  19. Kerr JS, Wright GJ. Avidity-based extracellular interaction screening (AVEXIS) for the scalable detection of Low-affinity extracellular Receptor-Ligand interactions. Journal of Visualized Experiments. 2012;61:e3881. doi: 10.3791/3881. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Kondo N, Miyauchi K, Meng F, Iwamoto A, Matsuda Z. Conformational changes of the HIV-1 envelope protein during membrane fusion are inhibited by the replacement of its membrane-spanning domain. Journal of Biological Chemistry. 2010;285:14681–14688. doi: 10.1074/jbc.M109.067090. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Lamas-Toranzo I, Fonseca Balvís N, Querejeta-Fernández A, Izquierdo-Rico MJ, González-Brusi L, Lorenzo PL, García-Rebollar P, Avilés M, Bermejo-Álvarez P. ZP4 confers structural properties to the zona pellucida essential for embryo development. eLife. 2019;8:e48904. doi: 10.7554/eLife.48904. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Le Naour F, Rubinstein E, Jasmin C, Prenant M, Boucheix C. Severely reduced female fertility in CD9-deficient mice. Science. 2000;287:319–321. doi: 10.1126/science.287.5451.319. [DOI] [PubMed] [Google Scholar]
  23. Lorenzetti D, Poirier C, Zhao M, Overbeek PA, Harrison W, Bishop CE. A transgenic insertion on mouse chromosome 17 inactivates a novel immunoglobulin superfamily gene potentially involved in sperm-egg fusion. Mammalian Genome. 2014;25:141–148. doi: 10.1007/s00335-013-9491-x. [DOI] [PubMed] [Google Scholar]
  24. Lu Y, Oura S, Matsumura T, Oji A, Sakurai N, Fujihara Y, Shimada K, Miyata H, Tobita T, Noda T, Castaneda JM, Kiyozumi D, Zhang Q, Larasati T, Young SAM, Kodani M, Huddleston CA, Robertson MJ, Coarfa C, Isotani A, Aitken RJ, Okabe M, Matzuk MM, Garcia TX, Ikawa M. CRISPR/Cas9-mediated genome editing reveals 30 testis-enriched genes dispensable for male fertility in mice†. Biology of Reproduction. 2019;101:501–511. doi: 10.1093/biolre/ioz103. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Miyado K, Yamada G, Yamada S, Hasuwa H, Nakamura Y, Ryu F, Suzuki K, Kosai K, Inoue K, Ogura A, Okabe M, Mekada E. Requirement of CD9 on the egg plasma membrane for fertilization. Science. 2000;287:321–324. doi: 10.1126/science.287.5451.321. [DOI] [PubMed] [Google Scholar]
  26. Mortimer ST. CASA-practical aspects. Journal of Andrology. 2000;21:515–524. doi: 10.1002/j.1939-4640.2000.tb02116.x. [DOI] [PubMed] [Google Scholar]
  27. Nishimura H, Gupta S, Myles DG, Primakoff P. Characterization of mouse sperm TMEM190, a small transmembrane protein with the trefoil domain: evidence for co-localization with IZUMO1 and complex formation with other sperm proteins. Reproduction. 2011;141:437–451. doi: 10.1530/REP-10-0391. [DOI] [PubMed] [Google Scholar]
  28. Noda T, Lu Y, Fujihara Y, Oura S, Koyano T, Kobayashi S, Matzuk MM, Ikawa M. Sperm proteins SOF1, TMEM95, and SPACA6 are required for sperm-oocyte fusion in mice. PNAS. 2020;117:11493–11502. doi: 10.1073/pnas.1922650117. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Ohto U, Ishida H, Krayukhina E, Uchiyama S, Inoue N, Shimizu T. Structure of IZUMO1-JUNO reveals sperm-oocyte recognition during mammalian fertilization. Nature. 2016;534:566–569. doi: 10.1038/nature18596. [DOI] [PubMed] [Google Scholar]
  30. Pausch H, Kölle S, Wurmser C, Schwarzenbacher H, Emmerling R, Jansen S, Trottmann M, Fuerst C, Götz KU, Fries R. A nonsense mutation in TMEM95 encoding a nondescript transmembrane protein causes idiopathic male subfertility in cattle. PLOS Genetics. 2014;10:e1004044. doi: 10.1371/journal.pgen.1004044. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Pedelacq J-D, Cabantous S. Development and applications of superfolder and split fluorescent protein detection systems in biology. International Journal of Molecular Sciences. 2019;20:3479. doi: 10.3390/ijms20143479. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Rubinstein E, Ziyyat A, Prenant M, Wrobel E, Wolf JP, Levy S, Le Naour F, Boucheix C. Reduced fertility of female mice lacking CD81. Developmental Biology. 2006;290:351–358. doi: 10.1016/j.ydbio.2005.11.031. [DOI] [PubMed] [Google Scholar]
  33. Satouh Y, Inoue N, Ikawa M, Okabe M. Visualization of the moment of mouse sperm-egg fusion and dynamic localization of IZUMO1. Journal of Cell Science. 2012;125:4985–4990. doi: 10.1242/jcs.100867. [DOI] [PubMed] [Google Scholar]
  34. Wang H, Yang H, Shivalila CS, Dawlaty MM, Cheng AW, Zhang F, Jaenisch R. One-step generation of mice carrying mutations in multiple genes by CRISPR/Cas-mediated genome engineering. Cell. 2013;153:910–918. doi: 10.1016/j.cell.2013.04.025. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Waterhouse A, Bertoni M, Bienert S, Studer G, Tauriello G, Gumienny R, Heer FT, de Beer TAP, Rempfer C, Bordoli L, Lepore R, Schwede T. SWISS-MODEL: homology modelling of protein structures and complexes. Nucleic Acids Research. 2018;46:W296–W303. doi: 10.1093/nar/gky427. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Wright GJ. Signal initiation in biological systems: the properties and detection of transient extracellular protein interactions. Molecular BioSystems. 2009;5:1405–1412. doi: 10.1039/b903580j. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Yang H, Wang H, Jaenisch R. Generating genetically modified mice using CRISPR/Cas-mediated genome engineering. Nature Protocols. 2014;9:1956–1968. doi: 10.1038/nprot.2014.134. [DOI] [PubMed] [Google Scholar]
  38. Zhang S, Cai H, Yang Q, Shi T, Pan C, Lei C, Dang R, Chen H, Lan X. Identification of novel alternative splicing transcript and expression analysis of bovine TMEM95 gene. Gene. 2016;575:531–536. doi: 10.1016/j.gene.2015.09.026. [DOI] [PubMed] [Google Scholar]

Decision letter

Editor: Polina V Lishko1
Reviewed by: Polina V Lishko2

In the interests of transparency, eLife publishes the most substantive revision requests and the accompanying author responses.

Acceptance summary:

The manuscript reports an important discovery: the identification of the protein required for sperm and egg fusion, TMEM95. Spermatozoa deficient in TMEM95 are devoid of sperm-egg recognition and fail to fuse with the egg. Overall, given murine genetics and in vitro fertilization experiments presented in this manuscript, as well as previously reported data regarding the importance of TMEM95 for bovine fertility and human genetics, TMEM95 could emerge is an important sperm surface receptor required for fertilization. Therefore, the significance of this research is exceptionally high, not only for the basic science and reproductive biology but also for translational applications, such as fertility diagnostics and contraception research.

Decision letter after peer review:

Thank you for submitting your article "TMEM95 is required for sperm-oocyte membrane fusion" for consideration by eLife. Your article has been reviewed by three peer reviewers, including Polina V Lishko as the Reviewing Editor and Reviewer #1, and the evaluation has been overseen by Anna Akhmanova as the Senior Editor.

The reviewers have discussed the reviews with one another and the Reviewing Editor has drafted this decision to help you prepare a revised submission.

While the goal is to provide the essential revision requirements as a single set of instructions, in this particular case we will provide a complete set of the reviews, so you can address all concerns as it is essential for the revision that is necessary for us to publish your work.

Summary:

The manuscript by Lamas-Toranzo et al., reports an important discovery: the identification of the protein required for sperm and egg fusion, TMEM95. Spermatozoa deficient in TMEM95 lacks sperm-egg recognition/fusion abilities much similar to IZUMO1-disrupted spermatozoa, and its requirement in the process is documented in this manuscript. However, the paper lacks detailed evidence regarding exactly how TMEM95 contributes to the process. Additionally, one of the major claims in the manuscript (absence of interaction between TMEM95 and other sperm/egg binding proteins) is not supported by the data. These data are omitted from the manuscript. Overall, given murine genetics and IVF experiments presented in this manuscript, as well as previously reported data regarding the importance of TMEM95 for bovine fertility and human genetics, TMEM95 could emerge is an important sperm surface receptor required for fertilization. Therefore, the significance of this research is exceptionally high, not only for the basic science and reproductive biology but also for translational applications, such as fertility diagnostics and contraception research.

Essential revisions:

The reviewers raise a number of concerns that must be adequately addressed before the paper can be accepted. Some of the required revisions will likely require further experimentation within the framework of the presented studies and techniques.

Specifically:

1) Experiments showing that TMEM95 recombinant protein acquires appropriate physiological activity and retains the ability to bind to the zona-free egg surface directly must be performed.

2) Immunoprecipitation experiments with sperm lysates using anti-IZUMO1 and/or anti-TMEM95 antibodies need to be shown.

3) While it was shown that TMEM95-disrupted spermatozoa express an IZUMO1 protein, it is unclear whether IZUMO1-disrupted sperm still possess a TMEM95 protein.

4) Additional characterization of TMEM95 expression in testes, epididymis, accessory glands etc on the protein level), as well as the tissue specificity should be performed.

5) The authors should carry out the localization studies in permeabilized and non-permeabilized intact WT sperm in order to detect in which exact cellular compartment TMEM95 is found.

6) The analysis of the number of sperm that bind to the egg surface should be performed.

7) As suggested all experiments must be performed with positive controls that confirm that the technique employed. Data not shown is not permitted.

8) Many parts of the manuscript (Abstract, Title, Results and Discussion, as well as terminology used: "egg" vs "oocyte") need to be rewritten and clarified as suggested by reviewers. Overall, the manuscript would benefit from thorough language editing, as well as additional data and experiments.

Reviewer #1:

The manuscript by Lamas-Toranzo et al., reports an important discovery: the identification of the protein required for sperm and egg fusion, TMEM95. The data provided are of good quality, and in vitro fertilization, experiments are very convincing. However, one of the major claims in the manuscript (absence of interaction between TMEM95 and other sperm/egg binding proteins) is not supported by the data. These data were either accidentally omitted from the manuscript or were not intended to be shown. Overall, given murine genetics and IVF experiments presented in this manuscript, as well as previously reported data regarding the importance of TMEM95 for bovine fertility and human genetic data, this reviewer has confidence that TMEM95 is an important sperm surface receptor required for fertilization. Therefore, the significance of this research is exceptionally high, not only for the basic science and reproductive biology but also for translational applications, such as fertility diagnostics and contraception research. However, the manuscript is less polished in regards to its writing style, the data representation, and overall would benefit from thorough language editing.

1) The first paragraph of the Introduction and throughout the text. Sperm usually do not fertilize an oocyte, they fertilize an egg or ovum. An oocyte is an immature female gamete still inside the follicle (for mammals), once it ovulates, it becomes the egg/ovum. Hence, terminology, such as "sperm-oocyte fusion" is technically incorrect and confusing, and should be replaced with sperm-egg fusion, including in the title.

2) Introduction: the last paragraph describing bovine genetics seems to be out of place. The authors first describe gene-editing in mice, and immediately jump to bovine genetics- it is hard to follow why such a transition was made.

3) Figure 1D: the entire western blot for TMEM95 should be shown to evaluate the specificity of the antibody used in the study.

4) Figure 3. The authors claim that the pull-down assay was performed to explore whether TMEM95 interacts with either IZUMO or Juno. However, the only experimental data that are actually shown is either heterologous expression of TMEM and its truncated form in CHO cells or the pull-down assays probed with anti-TMEM antibody. There is NO panel that shows western blot with either anti-IZUMO or anti-Juno antibodies. Moreover, membrane protein extraction from mature sperm is a difficult task, since most of these proteins are covalently attached to the underlying cytoskeletal structures. To rule out TMEM95/IZUMO/Juno interaction, authors must express all three heterologous in CHO cells, and then perform pull-down assay to explore their potential interaction.

5) Introduction first paragraph "perivitellin", should be perivitelline.

6) A triplicate of pull-down experiments is needed.

Reviewer #2:

Through the gene-modified animals, IZUMO1 and SPACA6 on the sperm side, and JUNO (IZUMO1R) and CD9 on the ovum side, have been identified as essential factors for sperm-egg fusion. In this study, the authors newly found a third sperm factor, TMEM95, which was originally identified by genome-wide screening from subfertile bulls, and is indispensable for gamete fusion in mice as well. Spermatozoa lacking TMEM95 are all missing of sperm-egg recognition/fusion abilities much similar to IZUMO1-disrupted spermatozoa, and its requirement in the process is carefully documented in this manuscript.

However, I think that the paper lacks detailed evidence regarding exactly how TMEM95 contributes to the process. Thus, in my opinion, this manuscript in its current form is not yet ready for publication in eLife.

The authors have concluded that TMEM95 is not involved in the IZUMO1-JUNO system due to results obtained via combination of co-immunoprecipitation (His-tag pull down) and HPLC-MS/MS analyses, however, I am concerned that these experiments have not been validated sufficiently. Particularly, I doubt if TMEM95 recombinant protein acquires appropriate physiological activity. Does this recombinant have the ability to bind to the zona-free oocyte surface directly? Also, in terms of IZUMO1, did you perform immunoprecipitation experiments with sperm lysate using anti-IZUMO1 and/or anti-TMEM95 antibodies? In addition, Figure 1D shows that TMEM95-disrupted spermatozoa possess an IZUMO1 protein; however, conversely, does IZUMO1-disrupted spermatozoa possess a TMEM95 protein? Mutual experiments are required to elucidate the detailed molecular relationship between these factors.

I think that those verifications need to be addressed.

Reviewer #3:

The manuscript by Lama-Toranzo et al. describes the identification of TMEM95, a sperm protein essential for mouse gamete fusion. Considering that, so far, only one sperm protein (i.e. IZUMO) has been found to be critical for this process, these results indicate that TMEM95 and IZUMO are both necessary but not sufficient for gamete fusion, representing a very interesting contribution to the field of mammalian fertilization. Nevertheless, the high structural similarity between TMEM95 and IZUMO 1 domain together with the same localization of these two proteins in intact and acrosome-reacted sperm supports the idea that TMEM95 likely participates in gamete fusion through the same ligand-receptor mechanisms involved in IZUMO-JUNO mediated gamete fusion. In this regard, the experiments showing that TMEM95 does not interact with JUNO are not convincing as they lack of appropriate positive controls. Moreover, considering that the only successful approach to identify JUNO has involved the use of a pentameric IZUMO protein due to the very low affinity between the two molecules (Bianchi and Wright et al., 2016), it is not clear why the authors analyzed TMEM95-JUNO interaction using just a regular co-immunoprecipitation assay. Thus, from a mechanistic point of view, the results on the participation of TMEM95 in gamete fusion does not seem to provide novel information. The molecular mechanisms underlying TMEM95 role in gamete fusion should be better analyzed and discussed.

There are also several other points that the authors should take into consideration before the manuscript can be accepted for publication in eLife.

Specific comments:

Abstract:

1) The authors mentioned that "TMEM95-deficient sperm were unable to fuse with or penetrate the oocyte membrane". Sperm bind to and then fuse with the egg plasma membrane and, finally, they penetrate into (or are incorporated into) the ooplasm. It is not correct to say that sperm cannot "penetrate" the oocyte membrane. This expression should be modified here as well as in other sections of the manuscript.

2) It is not clear which is the meaning of "kinetically" normal sperm. Are the authors referring to sperm motility and/or hyperactivity? This should be more clearly indicated.

3) A final sentence briefly indicating the significance of the findings should be added at the end of this section.

Results:

4) the authors should include in this section additional information regarding the characterization of TMEM95 protein which could contribute to a better interpretation of the results obtained. i.e. tissue expression (testes, epididymis, accessory glands etc) and tissue specificity.

5) Based on immunocytochemistry studies, the authors indicate that TMEM95 is present in the acrosome membrane of WT intact sperm. They should be aware that acrosome membrane is not the same than acrosomal region. If sperm are intact, the presence of TMEM95 in the acrosome membrane would imply this is an internal rather that a superficial protein. The authors should carry out the localization studies in permeabilized and non-permeabilized intact WT sperm in order to discriminate between the two possibilities.

Figure 1E legend indicates that "TMEM95 is localized around the acrosomal region of acrosome-intact sperm and in the sperm head after the acrosome reaction". The term "sperm head" is very vague. The authors should indicate where in the sperm head TMEM95 is localized. According to Figure 1F, TMEM95 should be located in the equatorial segment but the image in Figure 1E does not seem to show that localization. In addition to this, Figure 1E does not include any description about what the arrows are showing in each case.

The authors indicate that after the acrosome reaction, TMEM95 translocates to the equatorial segment. A protein can relocalize to the equatorial segment as a consequence of either its exposure due to the release of decapacitation factors or its migration from the acrosomal region. Based on the studies included in this manuscript, the authors cannot state that TMEM95 "translocates" from one region to the other.

6) In addition to the evaluation of gamete fusion by the Hoechst technique, the authors should analyze the number of sperm that bind to the egg surface. This analysis will provide information on whether or not TMEM95 participates in the first stage of gamete fusion i.e. sperm-egg binding

7) Figure 3: as mentioned above, the co-immunoprecipitation assay lacks appropriate positive controls that confirm that the technique employed is in fact capable of detecting a potential interaction between TMEM95 and JUNO (i.e a control showing IZUMO/JUNO interaction).

Figure 3 / Table 2: the authors carried out a series of Pull-Down experiments to identify potential TMEM95 partners in either the egg or the sperm. However, the explanation of these experiments and the interpretation of the results obtained are very poorly described.

Discussion:

The discussion about the potential molecular mechanisms underlying TMEM95 role in gamete fusion is rather confusing. The authors first suggest that TMEM95 may not require an oocyte partner. Then, they propose TMEM95 may have affinity to a yet undescribed receptor and then, based on the fact that TMEM95 acts independently from JUNO and IZUMO1 (which is not really demnstarted), they propose that TMEM95 may play an essential role rather than by single protein-to-protein interactions. The authors should more clearly conclude whether their observations support that TMEM95 participates in the first stage of gamete fusion (i.e sperm-egg binding) through a ligand-receptor mechanism or whether the protein could be acting as a fusogenic molecule capable of being intercalated into the egg lipid bilayer.

[Editors' note: further revisions were suggested prior to acceptance, as described below.]

Thank you for resubmitting your work entitled "ITMEM95 is a sperm protein essential for mammalian fertilization" for further consideration by eLife. Your revised article has been evaluated by Anna Akhmanova (Senior Editor) and a Reviewing Editor.

The manuscript has been significantly improved but there are some remaining issues that need to be addressed before acceptance, as outlined below. We understand that due to the current pandemic situation additional experiments might not be feasible, therefore, we ask authors to focus on the textual revisions as suggested by reviewers. Here we provide a full review as it contains many suggestions authors might find helpful.

Reviewer #1:

The authors addressed my concerns and the revision resulted in an improved manuscript.

Reviewer #2:

I am almost satisfied with the modifications provided by the authors, however, I still remain some concerns before publication in eLife.

1) The authors should mention SPACA6 as a gamete fusion related factor not only in Discussion but also in Abstract and Introduction, because I think SPACA6 is an unneglectable factor for this process (Barbaux et al., 2020).

2) In Figure 2—figure supplement 1C, there is no positive control (e.g. IZUMO1) in this system.

3) I recommend a rat anti-mouse IZUMO1 monoclonal antibody clone 125 (abcam, ab211626 or BioAcademia, 73-045 or anticorps-eligne, ABIN2452040) for detection in western blotting instead of polyclonal antibody because of high reliability.

Reviewer #3:

The revised version submitted by Lozano et al. includes new important studies that address the two main concerns raised by this (and other) reviewers i.e. 1) whether or not TMEM95 interacts with IZUMO1 and/or Juno, and 2) the mechanism underlying TMEM95 involvement in gamete fusion.

Regarding the first point, the authors decided to use the same strategy that had been employed to identify Juno, the partner of IZUMO, having now the appropriate controls. Moreover, the studies were performed in collaboration with the researchers that reported the identification of Juno who are now co-authors of the paper. In contrast to the previous immunoprecipitation and pull-down experiments, this new study clearly shows that TMEM95 interacts neither with IZUMO1 nor with Juno, representing a key incorporation to the paper.

To investigate the mechanisms underlying TMEM95 involvement in gamete fusion i.e whether the protein participates in the first stage of sperm binding to the oolemma or in the subsequent stage of membrane fusion itself, the authors 1) carried out an in vitro assay to evaluate the ability of TMEM95 mutant sperm to interact with the egg plasma membrane observing no differences with the controls, 2) analyzed the ability of TMEM95 to interact with the surface of zona-free eggs, detecting no binding of the protein to the egg, and 3) performed transfection studies to analyze the ability of TMEM95 to promote membrane fusion, observing no detectable fusion events in the analyzed cells. According to these observations, the authors concluded that TMEM95 does not participate in sperm-egg binding neither is a fusogenic protein and proposed that the protein must be involved in a post sperm-egg binding event previous to membrane fusion, as reported for IZUMO.

These studies have certainly fully answered the two main concerns of this reviewer supporting the publication of the manuscript in the journal. Nevertheless, there are some aspects that the authors should take into consideration in order to improve the quality of the manuscript.

– The authors indicated that the exact localization of the protein (internal or external) could not be analyzed because the employed fixation itself produced a permeabilization of the cells. Whereas this is true for many fixation conditions, the authors should consider carrying out immunolocalization studies in live, non-fixed sperm in suspension. This is an easy approach that will allow the authors to define the protein localization in the cell. This is not a minor point as it is known that the localization of a molecule provides important information regarding the mechanisms underlying its functional role, an issue still not fully resolved for TMEM95. In this regard, the author's claim that TMEM95 likely localizes to the acrosomal membrane because both it relocalizes to the equatorial segment after the acrosome reaction and it has a TM domain are not correct as plasma membrane proteins with or without TM domains do so as well.

– Whereas the authors have carried out the tissue expression and specificity studies requested by this reviewer, I could not find the description of the results obtained within the text but just in the figure. This should be corrected.

– I would remove the first section within Results entitled "TMEM95 protein in silico folding suggests a role in gamete fusion" as it describes just an in silico analysis that does not really provide any new information compared to the bull protein and does not deceives an independent paragraph. I would join the results of this paragraph that describes just Figure 1A with the results of the following paragraph describing the rest of Figure 1. The authors may need to change the title of that second paragraph too.

– The authors indicated that the lack of TMEM95 interaction with IZUMO together with the normal relocalization of IZUMO1 in TMEM95 KO sperm makes it unlikely that IZUMO1 KO would lead to the ablation of TMEM95 and that performing that experiment would delay the publication of these findings. In this regard, I think the authors could contemplate the possibility of sending their anti-TMEM95 antibody to Dr Ikawa who could perform this easy localization study in his lab. Of note in this regard, the authors already have contact with Dr Ikawa's lab which has provided the anti-IZUMO antibody to them.

– In my opinion, there are several observations which are not sufficiently discussed. The authors may speculate about several issues such as the finding that TMEM95 and IZUMO are both necessary but not sufficient for gamete fusion, how the protein reaches the equatorial segment, potential ways to identify TMEM95 partners, whether TMEM95 is present or not in IZUMO KO cells, among others, which will certainly enrich the discussion.

eLife. 2020 Jun 2;9:e53913. doi: 10.7554/eLife.53913.sa2

Author response

Essential revisions:

The reviewers raise a number of concerns that must be adequately addressed before the paper can be accepted. Some of the required revisions will likely require further experimentation within the framework of the presented studies and techniques.

Specifically:

1) Experiments showing that TMEM95 recombinant protein acquires appropriate physiological activity and retains the ability to bind to the zona-free egg surface directly must be performed.

We have conducted new experiments to determine a possible role of TMEM95 in sperm binding to the egg, concluding that TMEM95 is dispensable for sperm binding based on the following evidence: 1) TMEM95 ablation did not reduce sperm binding to the oocyte (sperm binding assay, Essential revision 6, Figure 2F), and 2) TMEM95:Cherry proteins did not bind to zona-free eggs (Figure 2—figure supplement 1C). These results suggest that TMEM95 is involved in post-binding events required for fertilization.

2) Immunoprecipitation experiments with sperm lysates using anti-IZUMO1 and/or anti-TMEM95 antibodies need to be shown.

We have substituted the previous pull-down assay with an assay designed to detect extracellular protein interactions (AVEXIS), which was the one responsible for the discovery of IZUMO1-JUNO interaction. Neither pull-down nor immunoprecipitation assays are able to detect such a subtle interaction (Figure 3A). This highly sensitive assay failed to detect any interaction between TMEM95 and JUNO or IZUMO.

3) While it was shown that TMEM95-disrupted spermatozoa express an IZUMO1 protein, it is unclear whether IZUMO1-disrupted sperm still possess a TMEM95 protein.

We believe that given TMEM95 is not closely related to IZUMO1 given that: 1) in the absence of TMEM95, IZUMO1 relocates normally from the acrosomal cap to the equatorial section following the acrosome reaction (Figure 1—figure supplement 1B), 2) IZUMO1 expression is not affected by TMEM95 disruption (Figure 1D), 3) TMEM95 does not interact with IZUMO1 (Figure 3A), and 4) TMEM95 does not interact with JUNO (Figure 3A). Therefore, there is no solid rationale to believe that IZUMO1 disruption would lead to TMEM95 disruption. Generating a colony of IZUMO1 KO to test an improbable co-disruption of TMEM95 would take ~6 months, an excessive and, we believe unnecessary delay to reporting our findings.

4) Additional characterization of TMEM95 expression in testes, epididymis, accessory glands etc on the protein level), as well as the tissue specificity should be performed.

We have performed new experiments analysing TMEM95 expression at mRNA and protein levels observing that the gene is expressed exclusively to the testis (Figure 1—figure supplement B-C).

5) The authors should carry out the localization studies in permeabilized and non-permeabilized intact WT sperm in order to detect in which exact cellular compartment TMEM95 is found.

We have conducted new experiments to determine TMEM95 localization within the acrosomal region of intact WT sperm. Previous experiments showed that TMEM95 was present in the equatorial segment of acrosome reacted sperm, and in the acrosomal cap of acrosome intact sperm, as the TMEM95 signal co-localized with PNA (Figure 1E-F). We have now performed IHC in WT sperm following permeabilizing and non-permeabilizing conditions as requested, observing TMEM95, IZUMO1 (acrosome membrane marker, Inoue et al., 2012) and PNA (internal acrosomal marker, Lybaert et al., 2009 Histol. Histopathol.) signals in both conditions (Figure 1—figure supplement 3A). This result indicates that fixation did not prevent antibody access into the acrosome in the absence of permeabilization (i.e., exposure to Triton X-100) during the whole IHC procedure. Further attempts to “seal” sperm and acrosome membranes, including increasing fixation time in 4 % PFA to 30 min at 4 ºC (following the protocol of Nishimura et al., 2011), were also unsuccessful: both IZUMO1 and PNA signals were clearly visible. These results suggest that signal detection following IHC in non-permeabilizing conditions may not be a solid criteria to exclude acrosomal localization. In agreement, whereas absence of IHC signal following non-permeabilizing conditions proved the intracellular localization of a protein in sperm, the opposite (i.e., IHC signal in non-permeabilizing conditions proving external membrane localization) is not conclusive: PNA, which specifically bounds to the acrosomal content, was also detected following non-permeabilizing conditions (Nishimura et al., 2011). We do not have solid evidence to claim that TMEM95 localizes to the acrosomal membrane, albeit that may be a likely location given 1) its co-localization with PNA and IZUMO1 on intact sperm, 2) its relocalization following acrosome reaction and 3) the presence of a transmembrane domain. We have modified the text to state that it localizes to the acrosomal cap in acrosome intact sperm.

6) The analysis of the number of sperm that bind to the egg surface should be performed.

We have performed this analysis observing no difference in sperm binding to the egg surface in the absence of TMEM95 (Figure 2F), suggesting that TMEM95 is not required for sperm binding.

7) As suggested all experiments must be performed with positive controls that confirm that the technique employed. Data not shown is not permitted.

We have included positive controls.

8) Many parts of the manuscript (Abstract, Title, Results and Discussion, as well as terminology used: "egg" vs "oocyte") need to be rewritten and clarified as suggested by reviewers. Overall, the manuscript would benefit from thorough language editing, as well as additional data and experiments.

We have extensively rewritten the manuscript and hope that is now suitable for publication.

Reviewer #1:

[…] 1) The first paragraph of the Introduction and throughout the text. Sperm usually do not fertilize an oocyte, they fertilize an egg or ovum. An oocyte is an immature female gamete still inside the follicle (for mammals), once it ovulates, it becomes the egg/ovum. Hence, terminology, such as "sperm-oocyte fusion" is technically incorrect and confusing, and should be replaced with sperm-egg fusion, including in the title.

We used oocyte and ovulated cumulus-oocyte-complex (COC) as they are commonly employed terms in mammalian embryology literature. In order to avoid any confusion for a wider audience, we have changed “oocyte” to “egg” throughout the manuscript.

2) Introduction: the last paragraph describing bovine genetics seems to be out of place. The authors first describe gene-editing in mice, and immediately jump to bovine genetics- it is hard to follow why such a transition was made.

We wanted to acknowledge these references early in the manuscript. We have moved them to the beginning of the Results section in the revised version.

3) Figure 1D: the entire western blot for TMEM95 should be shown to evaluate the specificity of the antibody used in the study.

WB images are now shown (Figure 1—figure supplement 2). TMEM95 WB show unspecific bands but a band of the expected size is present in WT samples and absent in KO samples. Peptide sequencing of that band confirmed TMEM95 identity (Table 1).

4) Figure 3. The authors claim that the pull-down assay was performed to explore whether TMEM95 interacts with either IZUMO or Juno. However, the only experimental data that are actually shown is either heterologous expression of TMEM and its truncated form in CHO cells or the pull-down assays probed with anti-TMEM antibody. There is NO panel that shows western blot with either anti-IZUMO or anti-Juno antibodies. Moreover, membrane protein extraction from mature sperm is a difficult task, since most of these proteins are covalently attached to the underlying cytoskeletal structures. To rule out TMEM95/IZUMO/Juno interaction, authors must express all three heterologous in CHO cells, and then perform pull-down assay to explore their potential interaction.

We have now conducted new experiments which demonstrate that TMEM95 does not bind to either IZUMO1 or JUNO using the same technique (AVEXIS, Kerr and Wright, 2012) that was used to identify IZUMO1 and JUNO interaction (Bianchi et al., 2014, Figure 3A). Given the absence of interaction we hypothesized that TMEM95 may be required to induce fusion following IZUMO1 and JUNO interaction. To test this possibility we have performed a GFP-complementation cell fusion assay in HEK293T cells observing that TMEM95 expression did not induce fusion on cells expressing JUNO and IZUMO (Figure 3D). These results suggest that other proteins may be required to achieve cell fusion.

5) Introduction first paragraph "perivitellin", should be perivitelline.

Changed, thank you for the correction.

6) A triplicate of pull-down experiments is needed.

As mentioned above, we have substituted those experiments by AVEXIS, a more sensitive protein interaction assay, performed in triplicate.

Reviewer #2:

[…] The authors have concluded that TMEM95 is not involved in the IZUMO1-JUNO system due to results obtained via combination of co-immunoprecipitation (His-tag pull down) and HPLC-MS/MS analyses, however, I am concerned that these experiments have not been validated sufficiently. Particularly, I doubt if TMEM95 recombinant protein acquires appropriate physiological activity. Does this recombinant have the ability to bind to the zona-free oocyte surface directly? Also, in terms of IZUMO1, did you perform immunoprecipitation experiments with sperm lysate using anti-IZUMO1 and/or anti-TMEM95 antibodies? In addition, Figure 1D shows that TMEM95-disrupted spermatozoa possess an IZUMO1 protein; however, conversely, does IZUMO1-disrupted spermatozoa possess a TMEM95 protein? Mutual experiments are required to elucidate the detailed molecular relationship between these factors. I think that those verifications need to be addressed.

As detailed in the general comments above, we have conducted new experiments (protein interaction and sperm-egg binding assays) concluding that TMEM95 is not directly involved in binding. Using a highly sensitive protein interaction assay able to detect IZUMO1-JUNO interaction (i.e., more sensitive than immunoprecipitation or pull-down assays), we have observed that IZUMO1 does not interact with TMEM95. This new evidence together with the normal relocalization of IZUMO1 in TMEM95 KO sperm makes it unlikely that IZUMO1 KO would lead to the ablation of TMEM95. Performing that experiment would lead to an unjustified delay in the publication of these findings.

Reviewer #3:

The manuscript by Lama-Toranzo et al. describes the identification of TMEM95, a sperm protein essential for mouse gamete fusion. Considering that, so far, only one sperm protein (i.e. IZUMO) has been found to be critical for this process, these results indicate that TMEM95 and IZUMO are both necessary but not sufficient for gamete fusion, representing a very interesting contribution to the field of mammalian fertilization. Nevertheless, the high structural similarity between TMEM95 and IZUMO 1 domain together with the same localization of these two proteins in intact and acrosome-reacted sperm supports the idea that TMEM95 likely participates in gamete fusion through the same ligand-receptor mechanisms involved in IZUMO-JUNO mediated gamete fusion. In this regard, the experiments showing that TMEM95 does not interact with JUNO are not convincing as they lack of appropriate positive controls. Moreover, considering that the only successful approach to identify JUNO has involved the use of a pentameric IZUMO protein due to the very low affinity between the two molecules (Bianchi and Wright et al., 2016), it is not clear why the authors analyzed TMEM95-JUNO interaction using just a regular co-immunoprecipitation assay. Thus, from a mechanistic point of view, the results on the participation of TMEM95 in gamete fusion does not seem to provide novel information. The molecular mechanisms underlying TMEM95 role in gamete fusion should be better analyzed and discussed.

We agree that the conventional pull-down assay performed may fail to detect a subtle TMEM95-Juno interaction. This pull-down assay performed with recombinant proteins is more sensitive than immunoprecipitation, but still may still fail to uncover subtle protein interactions. We have now used the same highly sensitive protein interaction assay employed by Bianchi et al., 2014, proving that TMEM95 does not bind to JUNO or IZUMO1.

There are also several other points that the authors should take into consideration before the manuscript can be accepted for publication in eLife.

Specific comments:

Abstract:

1) The authors mentioned that "TMEM95-deficient sperm were unable to fuse with or penetrate the oocyte membrane". Sperm bind to and then fuse with the egg plasma membrane and, finally, they penetrate into (or are incorporated into) the ooplasm. It is not correct to say that sperm cannot "penetrate" the oocyte membrane. This expression should be modified here as well as in other sections of the manuscript.

We have corrected the Abstract as suggested.

2) It is not clear which is the meaning of "kinetically" normal sperm. Are the authors referring to sperm motility and/or hyperactivity? This should be more clearly indicated.

We have corrected the Abstract as suggested.

3) A final sentence briefly indicating the significance of the findings should be added at the end of this section.

We have added a sentence.

Results:

4) the authors should include in this section additional information regarding the characterization of TMEM95 protein which could contribute to a better interpretation of the results obtained. i.e. tissue expression (testes, epididymis, accessory glands etc) and tissue specificity.

We have performed mRNA and protein analysis on different tissues demonstrating testis-specific expression of TMEM95.

5) Based on immunocytochemistry studies, the authors indicate that TMEM95 is present in the acrosome membrane of WT intact sperm. They should be aware that acrosome membrane is not the same than acrosomal region. If sperm are intact, the presence of TMEM95 in the acrosome membrane would imply this is an internal rather that a superficial protein. The authors should carry out the localization studies in permeabilized and non-permeabilized intact WT sperm in order to discriminate between the two possibilities.

As previously mentioned (essential revision 5), we have conducted IHC in permeabilizing and non-permeabilizing conditions observing TMEM95, IZUMO1 and PNA signals in non-permeabilizing conditions. Given this result, we cannot demonstrate that TMEM95 is present in the acrosome membrane. As the signal is restricted to the acrosomal cap, co-localizing with both IZUMO and PNA in intact sperm, we have modified the text using the term acrosomal cap.

Figure 1E legend indicates that "TMEM95 is localized around the acrosomal region of acrosome-intact sperm and in the sperm head after the acrosome reaction". The term "sperm head" is very vague. The authors should indicate where in the sperm head TMEM95 is localized. According to Figure 1F, TMEM95 should be located in the equatorial segment but the image in Figure 1E does not seem to show that localization. In addition to this, Figure 1E does not include any description about what the arrows are showing in each case.

We now indicate also in that figure legend that TMEM95 relocalized to the equatorial segment. Specific localization details are more difficult to spot on Figure 1E, as it is taken with a lower magnification objective, however, a wider distribution compatible with equatorial segment location can also be observed on the PNA negative sperm compared with the other PNA positive sperm on that figure.

The authors indicate that after the acrosome reaction, TMEM95 translocates to the equatorial segment. A protein can relocalize to the equatorial segment as a consequence of either its exposure due to the release of decapacitation factors or its migration from the acrosomal region. Based on the studies included in this manuscript, the authors cannot state that TMEM95 "translocates" from one region to the other.

We have changed translocation to relocalization through the manuscript.

6) In addition to the evaluation of gamete fusion by the Hoechst technique, the authors should analyze the number of sperm that bind to the egg surface. This analysis will provide information on whether or not TMEM95 participates in the first stage of gamete fusion i.e. sperm-egg binding

We have now conducted this analysis concluding that TMEM95 is not essential for sperm-egg binding.

7) Figure 3: as mentioned above, the co-immunoprecipitation assay lacks appropriate positive controls that confirm that the technique employed is in fact capable of detecting a potential interaction between TMEM95 and JUNO (i.e a control showing IZUMO/JUNO interaction). Figure 3 / Table 2: the authors carried out a series of Pull-Down experiments to identify potential TMEM95 partners in either the egg or the sperm. However, the explanation of these experiments and the interpretation of the results obtained are very poorly described.

We have conducted a new analysis using a more sensitive technique (AVEXIS) capable of detecting IZUMO1-JUNO interaction.

Discussion:

The discussion about the potential molecular mechanisms underlying TMEM95 role in gamete fusion is rather confusing. The authors first suggest that TMEM95 may not require an oocyte partner. Then, they propose TMEM95 may have affinity to a yet undescribed receptor and then, based on the fact that TMEM95 acts independently from JUNO and IZUMO1 (which is not really demnstarted), they propose that TMEM95 may play an essential role rather than by single protein-to-protein interactions. The authors should more clearly conclude whether their observations support that TMEM95 participates in the first stage of gamete fusion (i.e sperm-egg binding) through a ligand-receptor mechanism or whether the protein could be acting as a fusogenic molecule capable of being intercalated into the egg lipid bilayer.

We now provide solid evidence regarding the independence of TMEM95 from the JUNO-IZUMO1 interaction (Figure 3A). As TMEM95 ablation did not disrupt sperm binding (Figure 2F), TMEM95 must be involved in post-binding events required for fertilization. To test a possible role of TMEM95 as a cell fusion promoter, we have performed a fusion assay with cells expressing TMEM95 in the presence of JUNO and IZUMO1. We have not observed cell fusion, suggesting that either TMEM95 is not involved in fusion or that, being involved, it requires the intervention of other yet unknown proteins.

[Editors' note: further revisions were suggested prior to acceptance, as described below.]

The manuscript has been significantly improved but there are some remaining issues that need to be addressed before acceptance, as outlined below. We understand that due to the current pandemic situation additional experiments might not be feasible, therefore, we ask authors to focus on the textual revisions as suggested by reviewers. Here we provide a full review as it contains many suggestions authors might find helpful.

We thank the reviewers and editors for their positive comments and appreciate the transparent reviewing process that distinguishes eLife. We have modified the article to fulfil the requests and have included a very recent article published in PNAS this week (Noda et al., 2020) which reports some aspects of the TMEM95 KO phenotype. We would like to stress that we were unaware of this article during the preparation, submission and revision of our current submitted article. As you will see, Our article was submitted to eLife one month earlier than the Noda paper was submitted to PNAS, and 5 months earlier it was submitted to PNAS, where it was editorially rejected (not sent for review) as the finding of an essential protein for gamete fusion was “not considered to have the broad appeal” required for that journal. We honestly thank eLife for the opportunity of publishing our results.

Reviewer #2:

I am almost satisfied with the modifications provided by the authors, however, I still remain some concerns before publication in eLife.

1) The authors should mention SPACA6 as a gamete fusion related factor not only in Discussion but also in Abstract and Introduction, because I think SPACA6 is an unneglectable factor for this process (Barbaux et al., 2020).

We agree. We have added Barbaux et al., 2020 and Noda et al., 2020, both of which were published during the second revision of our manuscript.

2) In Figure 2—figure supplement 1C, there is no positive control (e.g. IZUMO1) in this system.

We apologize for not using IZUMO1 as positive control for binding. The oocytes used for this study were processed as those used for sperm binding assays, proving that the procedure used (both timing after hCG injection and zona removal) did not impair sperm binding ability to the oolema.

3) I recommend a rat anti-mouse IZUMO1 monoclonal antibody clone 125 (abcam, ab211626 or BioAcademia, 73-045 or anticorps-eligne, ABIN2452040) for detection in western blotting instead of polyclonal antibody because of high reliability.

Thank you for your valuable suggestion.

Reviewer #3:

[…]

These studies have certainly fully answered the two main concerns of this reviewer supporting the publication of the manuscript in the journal. Nevertheless, there are some aspects that the authors should take into consideration in order to improve the quality of the manuscript.

– The authors indicated that the exact localization of the protein (internal or external) could not be analyzed because the employed fixation itself produced a permeabilization of the cells. Whereas this is true for many fixation conditions, the authors should consider carrying out immunolocalization studies in live, non-fixed sperm in suspension. This is an easy approach that will allow the authors to define the protein localization in the cell. This is not a minor point as it is known that the localization of a molecule provides important information regarding the mechanisms underlying its functional role, an issue still not fully resolved for TMEM95. In this regard, the author's claim that TMEM95 likely localizes to the acrosomal membrane because both it relocalizes to the equatorial segment after the acrosome reaction and it has a TM domain are not correct as plasma membrane proteins with or without TM domains do so as well.

We acknowledge the reviewer’s concerns about TMEM95 localization, but regretfully we cannot provide more solid evidence regarding the precise localization within the acrosomal cap. Unfortunately, immunolocalization in live sperm is not an option: as TMEM95 antibody required overnight incubation, sperm would die during the incubation time in antibody solution and their membranes would become permeable. We acknowledged that TMEM95 localization within the acrosomal cap is not fully resolved and consequently we did not claim that Tmem95 localizes to the acrosomal membrane in both the previously revised version and this latest version.

– Whereas the authors have carried out the tissue expression and specificity studies requested by this reviewer, I could not find the description of the results obtained within the text but just in the figure. This should be corrected.

We have included it in the text.

– I would remove the first section within Results entitled "TMEM95 protein in silico folding suggests a role in gamete fusion" as it describes just an in silico analysis that does not really provide any new information compared to the bull protein and does not deceives an independent paragraph. I would join the results of this paragraph that describes just Figure 1A with the results of the following paragraph describing the rest of Figure 1. The authors may need to change the title of that second paragraph too.

We have now joined both sections as requested.

– The authors indicated that the lack of TMEM95 interaction with IZUMO together with the normal relocalization of IZUMO1 in TMEM95 KO sperm makes it unlikely that IZUMO1 KO would lead to the ablation of TMEM95 and that performing that experiment would delay the publication of these findings. In this regard, I think the authors could contemplate the possibility of sending their anti-TMEM95 antibody to Dr Ikawa who could perform this easy localization study in his lab. Of note in this regard, the authors already have contact with Dr Ikawa's lab which has provided the anti-IZUMO antibody to them.

During this second revision of this manuscript, Ikawa´s group has published an article reporting roles of TMEM95, SPACA6 and SOF1 in fertilization (Noda et al., 2020). In that article they found that the relocalization of IZUMO1 was not impaired by the ablation of any of these three genes (in agreement with our results and Barbaux et al., 2020). Unfortunately, they did not test the opposite (i.e., whether IZUMO1 ablation impaired the other proteins), as sperm localization was only shown for SPACA6 and this analysis was not performed in Izumo1 KO sperm. As the antibody we used in this article is commercially available, we believe that they will be able to perform this experiment in subsequent publications.

– In my opinion, there are several observations which are not sufficiently discussed. The authors may speculate about several issues such as the finding that TMEM95 and IZUMO are both necessary but not sufficient for gamete fusion, how the protein reaches the equatorial segment, potential ways to identify TMEM95 partners, whether TMEM95 is present or not in IZUMO KO cells, among others, which will certainly enrich the discussion.

We have now extended the Discussion while avoiding to be too speculative. For instance, we cannot say much about TMEM95 relocalization, as the precise location within the acrosomal cap could not be determined.

Associated Data

    This section collects any data citations, data availability statements, or supplementary materials included in this article.

    Supplementary Materials

    Transparent reporting form
    elife-53913-transrepform.docx (246.4KB, docx)

    Data Availability Statement

    All data generated or analysed during this study are included in the manuscript and supporting files.

    Articles from eLife are provided here courtesy of eLife Sciences Publications, Ltd

    RESOURCES