304 stainless steel coil tubing chemical component ,The structure of the SPACA6 ectodomain contains a conserved superfamily of proteins associated with gamete fusion.

Thank you for visiting Nature.com. You are using a browser version with limited CSS support. For the best experience, we recommend that you use an updated browser (or disable Compatibility Mode in Internet Explorer). In addition, to ensure ongoing support, we show the site without styles and JavaScript.
Sliders showing three articles per slide. Use the back and next buttons to move through the slides, or the slide controller buttons at the end to move through each slide.

SPACA6 is a sperm-expressed surface protein that is critical for gamete fusion during mammalian sexual reproduction. Despite this fundamental role, the specific function of SPACA6 is poorly understood. We elucidate the crystal structure of the extracellular domain of SPACA6 at a resolution of 2.2 Å, revealing a two-domain protein composed of a four-stranded bundle and Ig-like β-sandwiches joined by quasi-flexible linkers. This structure resembles IZUMO1, another gamete fusion-associated protein, making SPACA6 and IZUMO1 founding members of the superfamily of fertilization-associated proteins referred to herein as the IST superfamily. The IST superfamily is structurally defined by its twisted four-helix bundle and a pair of disulfide-linked CXXC motifs. A structure-based AlphaFold search of the human proteome identified additional protein members of this superfamily; notably, many of these proteins are involved in gamete fusion. The SPACA6 structure and its relationship to other members of the IST superfamily provide the missing link in our knowledge of mammalian gamete fusion.
Every human life begins with two separate haploid gametes: the father’s sperm and the mother’s egg. This sperm is the winner of an intense selection process during which millions of sperm cells pass through the female genital tract, overcome various obstacles1 and undergo capacitation, which enhances their motility and the process of surface components2,3,4. Even if the sperm and oocyte find each other, the process is not over yet. The oocyte is surrounded by a layer of cumulus cells and a glycoprotein barrier called the zona pellucida, through which sperm must pass to enter the oocyte. Spermatozoa use a combination of surface adhesion molecules and membrane-associated and secreted enzymes to overcome these final barriers5. These molecules and enzymes are mainly stored in the inner membrane and acrosomal matrix and are detected when the outer membrane of the sperm is lysed during the acrosomal reaction6. The final step in this intense journey is the sperm-egg fusion event, in which the two cells fuse their membranes to become a single diploid organism7. Although this process is groundbreaking in human reproduction, the necessary molecular interactions are poorly understood.
In addition to the fertilization of gametes, the chemistry of the fusion of two lipid bilayers has been extensively studied. In general, membrane fusion is an energetically unfavorable process that requires a protein catalyst to undergo a structural conformation change that brings two membranes closer together, breaking their continuity and causing fusion8,9. These protein catalysts are known as fusogens and have been found in countless fusion systems. They are required for viral entry into host cells (e.g., gp160 in HIV-1, spike in coronaviruses, hemagglutinin in influenza viruses)10,11,12 placental (syncytin)13,14,15 and gamete-forming fusions in lower eukaryotes (HAP2/GCS1 in plants, protists and arthropods) 16,17,18,19. Fusogens for human gametes have yet to be discovered, although several proteins have been shown to be critical for gamete attachment and fusion. The oocyte-expressed CD9, a transmembrane protein required for the fusion of mouse and human gametes, was the first to be discovered 21,22,23. Although its precise function remains unclear, a role in adhesion, the structure of adhesion foci on egg microvilli, and/or the correct localization of oocyte surface proteins seems likely 24,25,26. The two most typical proteins that are critical for gamete fusion are the sperm protein IZUMO127 and the oocyte protein JUNO28, and their mutual association is an important step in gamete recognition and adhesion prior to fusion. Male Izumo1 knockout mice and female Juno knockout mice are completely sterile, in these models sperm enter the perivitelline space but gametes do not fuse. Similarly, confluence was reduced when gametes were treated with anti-IZUMO1 or JUNO27,29 antibodies in human in vitro fertilization experiments.
Recently, a newly discovered group of sperm-expressed proteins phenotypically similar to IZUMO1 and JUNO20,30,31,32,33,34,35 has been discovered. Sperm acrosomal membrane-associated protein 6 (SPACA6) has been identified as essential for fertilization in a large-scale murine mutagenesis study. Insertion of the transgene into the Spaca6 gene produces nonfusible spermatozoa, although these spermatozoa infiltrate the perivitelline space 36 . Subsequent knockout studies in mice confirmed that Spaca6 is required for gamete fusion 30,32 . SPACA6 is expressed almost exclusively in the testes and has a localization pattern similar to that of IZUMO1, namely within the intima of the spermatozoa before the acrosomal reaction, and then migrates to the equatorial region after the acrosomal reaction 30,32. Spaca6 homologues exist in a variety of mammals and other eukaryotes 30 and its importance for human gamete fusion has been demonstrated by inhibition of human fertilization in vitro by resistance to SPACA6 30 . Unlike IZUMO1 and JUNO, the details of the structure, interactions, and function of SPACA6 remain unclear.
To better understand the fundamental process underlying the fusion of human sperm and eggs, which will enable us to inform future developments in family planning and fertility treatment, we conducted SPACA6 structural and biochemical studies. The crystal structure of the extracellular domain of SPACA6 shows a four-helical bundle (4HB) and an immunoglobulin-like (Ig-like) domain connected by quasi-flexible regions. As predicted in previous studies,7,32,37 the domain structure of SPACA6 is similar to that of human IZUMO1, and the two proteins share an unusual motif: 4HB with a triangular helical surface and a pair of disulfide-linked CXXC motifs. We propose that IZUMO1 and SPACA6 now define a larger, structurally related superfamily of proteins associated with gamete fusion. Using features unique to the superfamily, we conducted an exhaustive search for the AlphaFold structural human proteome, identifying additional members of this superfamily, including several members involved in gamete fusion and/or fertilization. It now appears that there is a common structural fold and superfamily of proteins associated with gamete fusion, and our structure provides a molecular map of this important aspect of the human gamete fusion mechanism.
SPACA6 is a single-pass transmembrane protein with one N-linked glycan and six putative disulfide bonds (Figures S1a and S2). We expressed the extracellular domain of human SPACA6 (residues 27–246) in Drosophila S2 cells and purified the protein using nickel affinity, cation exchange, and size exclusion chromatography (Fig. S1b). The purified SPACA6 ectodomain is very stable and homogeneous. Analysis using size exclusion chromatography combined with polygonal light scattering (SEC-MALS) revealed one peak with a calculated molecular weight of 26.2 ± 0.5 kDa (Fig. S1c). This is consistent with the size of the SPACA6 monomeric ectodomain, indicating that oligomerization did not occur during purification. In addition, circular dichroism (CD) spectroscopy revealed a mixed α/β structure with a melting point of 51.3 °C (Fig. S1d,e). Deconvolution of the CD spectra revealed 38.6% α-helical and 15.8% β-stranded elements (Figure S1d).
The SPACA6 ectodomain was crystallized using random matrix seeding38 resulting in a data set with a resolution of 2.2 Å (Table 1 and Figure S3). Using a combination of fragment-based molecular substitution and SAD phasing data with bromide exposure for structure determination (Table 1 and Figure S4), the final refined model consists of residues 27–246. At the time the structure was determined, there were no experimental or AlphaFold structures available. The SPACA6 ectodomain measures 20 Å × 20 Å × 85 Å, consists of seven helices and nine β-strands, and has an elongated tertiary fold stabilized by six disulfide bonds (Fig. 1a, b). The weak electron density at the end of the Asn243 side chain indicates that this residue is an N-linked glycosylation. The structure consists of two domains: an N-terminal four-helix bundle (4HB) and a C-terminal Ig-like domain with an intermediate hinge region between them (Fig. 1c).
a Structure of the extracellular domain of SPACA6. Strip diagram of the extracellular domain of SPACA6, the color of the chain from N to C-terminus from dark blue to dark red. Cysteines involved in disulfide bonds are highlighted in magenta. b Topology of the extracellular domain of SPACA6. Use the same color scheme as in Figure 1a. c SPACA6 extracellular domain. The 4HB, hinge, and Ig-like domain strip charts are colored orange, green, and blue, respectively. The layers are not drawn to scale.
The 4HB domain of SPACA6 includes four main helices (helices 1–4), which are arranged in the form of a helical helix (Fig. 2a), alternating between antiparallel and parallel interactions (Fig. 2b). A small additional single-turn helix (helix 1′) is laid perpendicular to the bundle, forming a triangle with helices 1 and 2. This triangle is slightly deformed in the helical-twisted packing of the relatively dense packing of helices 3 and 4 (Fig. 2a).
4HB N-terminal strip chart. b Top view of a bundle of four helices, each helix highlighted in dark blue at the N-terminus and dark red at the C-terminus. c Top-down spiral wheel diagram for 4HB, with each residue shown as a circle labeled with a single-letter amino acid code; only the four amino acids at the top of the wheel are numbered. Non-polar residues are colored yellow, polar uncharged residues are colored green, positively charged residues are colored blue, and negatively charged residues are colored red. d Triangular faces of the 4HB domain, with 4HBs in orange and hinges in green. Both insets show rod-shaped disulfide bonds.
4HB is concentrated on an inner hydrophobic core composed mainly of aliphatic and aromatic residues (Fig. 2c). The core contains a disulfide bond between Cys41 and Cys55 that links helices 1 and 2 together in an upper raised triangle (Fig. 2d). Two additional disulfide bonds were formed between the CXXC motif in Helix 1′ and another CXXC motif found at the tip of the β-hairpin in the hinge region (Fig. 2d). A conservative arginine residue with an unknown function (Arg37) is located inside a hollow triangle formed by helices 1′, 1, and 2. Aliphatic carbon atoms Cβ, Cγ, and Cδ Arg37 interact with the hydrophobic core, and its guanidine groups cyclically move between helices 1′ and 1 via interactions between the Thr32 backbone and side chain (Fig. S5a,b). Tyr34 extends into the cavity leaving two small cavities through which Arg37 can interact with the solvent.
Ig-like β-sandwich domains are a large superfamily of proteins that share the common feature of two or more multi-stranded amphipathic β-sheets interacting via a hydrophobic core 39. The C-terminal Ig-like domain of SPACA6 has the same pattern and consists of two layers (Fig. S6a). Sheet 1 is a β-sheet of four strands (strands D, F, H, and I) where strands F, H, and I form an anti-parallel arrangement, and strands I and D take on a parallel interaction. Table 2 is a small anti-parallel double-stranded beta sheet (strands E and G). An internal disulfide bond was observed between the C-terminus of the E chain and the center of the H chain (Cys170-Cys226) (Fig. S6b). This disulfide bond is analogous to the disulfide bond in the β-sandwich domain of immunoglobulin40,41.
The four-strand β-sheet twists along its entire length, forming asymmetrical edges that differ in shape and electrostatics. The thinner edge is a flat hydrophobic environmental surface that stands out compared to the remaining uneven and electrostatically diverse surfaces in SPACA6 (Fig. S6b,c). A halo of exposed backbone carbonyl/amino groups and polar side chains surrounds the hydrophobic surface (Fig. S6c). The wider margin is covered by a capped helical segment that blocks the N-terminal portion of the hydrophobic core and forms three hydrogen bonds with the open polar group of the F chain backbone (Fig. S6d). The C-terminal portion of this edge forms a large pocket with a partially exposed hydrophobic core. The pocket is surrounded by positive charges due to three sets of double arginine residues (Arg162-Arg221, Arg201-Arg205 and Arg212-Arg214) and a central histidine (His220) (Figure S6e).
The hinge region is a short segment between the helical domain and the Ig-like domain, consisting of one antiparallel three-stranded β-layer (strands A, B and C), a small 310 helix, and several long random helical segments. (Fig. S7). A network of covalent and electrostatic contacts in the hinge region appears to stabilize the orientation between 4HB and the Ig-like domain. The network can be divided into three parts. The first part includes two CXXC motifs (27CXXC30 and 139CXXC142) that form a pair of disulfide bonds between the β-hairpin in the hinge and the 1′ helix in 4HB. The second part includes electrostatic interactions between the Ig-like domain and the hinge. Glu132 in the hinge forms a salt bridge with Arg233 in the Ig-like domain and Arg135 in the hinge. The third part includes a covalent bond between the Ig-like domain and the hinge region. Two disulfide bonds (Cys124-Cys147 and Cys128-Cys153) connect the hinge loop to a linker that is stabilized by electrostatic interactions between Gln131 and the backbone functional group, allowing access to the first Ig-like domain. chain.
The structure of the SPACA6 ectodomain and individual structures of 4HB and Ig-like domains were used to search for structurally similar records in protein databases 42 . We identified matches with high Dali Z scores, small standard deviations, and large LALI scores (the latter being the number of structurally equivalent residues). While the first 10 hits from the full ectodomain search (Table S1) had an acceptable Z-score of >842, a search for 4HB or Ig-like domain alone showed that most of these hits corresponded to β-sandwiches only. a ubiquitous fold found in many proteins. All three searches in Dali returned only one result: IZUMO1.
It has long been suggested that SPACA6 and IZUMO1 share structural similarities7,32,37. Although the ectodomains of these two gamete fusion-associated proteins share only 21% sequence identity (Figure S8a), complex evidence, including a conserved disulfide bond pattern and a predicted C-terminal Ig-like domain in SPACA6, allowed early attempts to build a homology model of A a SPACA6 mouse using IZUMO1 as a template37. Our structure confirms these predictions and shows the true degree of similarity. In fact, the SPACA6 and IZUMO137,43,44 structures share the same two-domain architecture (Fig. S8b) with similar 4HB and Ig-like β-sandwich domains connected by a hinge region (Fig. S8c).
IZUMO1 and SPACA6 4HB have common differences from conventional spiral bundles. Typical 4HBs, like those found in SNARE protein complexes involved in endosomal fusion 45,46, have evenly spaced helices maintaining a constant curvature around a central axis 47. In contrast, the helical domains in both IZUMO1 and SPACA6 were distorted, with variable curvature and uneven packing (Figure S8d). The twist, probably caused by the triangle formed by helices 1′, 1 and 2, is retained in IZUMO1 and SPACA6 and stabilized by the same CXXC motif on helix 1′. However, the additional disulfide bond found in SPACA6 (Cys41 and Cys55 covalently linking helices 1 and 2 above) creates a sharper apex at the triangle apex, making SPACA6 more twisted than IZUMO1, with more pronounced cavity triangles. In addition, IZUMO1 lacks Arg37 observed in the center of this cavity in SPACA6. In contrast, IZUMO1 has a more typical hydrophobic core of aliphatic and aromatic residues.
IZUMO1 has an Ig-like domain consisting of a double-stranded and five-stranded β-sheet43. The extra strand in IZUMO1 replaces the coil in SPACA6, which interacts with the F strand to limit backbone hydrogen bonds in the strand. An interesting point of comparison is the predicted surface charge of the Ig-like domains of the two proteins. The IZUMO1 surface is more negatively charged than the SPACA6 surface. An additional charge is located near the C-terminus facing the sperm membrane. In SPACA6, the same regions were more neutral or positively charged (Fig. S8e). For example, the hydrophobic surface (thinner edges) and positively charged pits (wider edges) in SPACA6 are negatively charged in IZUMO1.
Although the relationship and secondary structure elements between IZUMO1 and SPACA6 are well preserved, the structural alignment of the Ig-like domains showed that the two domains differ in their general orientation relative to each other (Fig. S9). The spiral bundle of IZUMO1 is curved about the β-sandwich, creating the previously described “boomerang” shape at about 50° from the central axis. In contrast, the helical beam in SPACA6 was tilted about 10° in the opposite direction. The differences in these orientations are likely due to differences in the hinge region. At the primary sequence level, IZUMO1 and SPACA6 share little sequence similarity at the hinge, with the exception of cysteine, glycine, and aspartic acid residues. As a result, hydrogen bonds and electrostatic networks are completely different. β-sheet secondary structure elements are shared by IZUMO1 and SPACA6, although the chains in IZUMO1 are much longer and the 310 helix (helix 5) is unique to SPACA6. These differences result in different domain orientations for two otherwise similar proteins.
Our Dali server search revealed that SPACA6 and IZUMO1 are the only two experimentally determined structures stored in the protein database that have this particular 4HB fold (Table S1). More recently, DeepMind (Alphabet/Google) has developed AlphaFold, a neural network based system that can accurately predict the 3D structures of proteins from primary sequences48. Shortly after we solved the SPACA6 structure, the AlphaFold database was released, providing predictive structure models covering 98.5% of all proteins in the human proteome48,49. Using our resolved SPACA6 structure as a search model, a structural homology search for the model in the AlphaFold human proteome identified candidates with possible structural similarities to SPACA6 and IZUMO1. Given the incredible accuracy of AlphaFold in predicting SPACA6 (Fig. S10a)—especially the 1.1 Å rms ectodomain compared to our resolved structure (Fig. S10b)—we can be confident that the identified SPACA6 matches are likely to be accurate.
Previously, PSI-BLAST searched for the IZUMO1 cluster with three other sperm-associated proteins: IZUMO2, IZUMO3, and IZUMO450. AlphaFold predicted that these IZUMO family proteins fold into the 4HB domain with the same disulfide bond pattern as IZUMO1 (Figures 3a and S11), although they lack an Ig-like domain. It is hypothesized that IZUMO2 and IZUMO3 are one-sided membrane proteins similar to IZUMO1, while IZUMO4 appears to be secreted. The functions of IZUMO 2, 3, and 4 proteins in gamete fusion have not been determined. IZUMO3 is known to play a role in acrosome biogenesis during sperm development51 and the IZUMO protein has been found to form a complex50. The conservation of IZUMO proteins in mammals, reptiles, and amphibians suggests that their potential function is consistent with that of other known gamete-fusion-associated proteins, such as DCST1/2, SOF1, and FIMP.
Diagram of the domain architecture of the IST superfamily, with 4HB, hinge, and Ig-like domains highlighted in orange, green, and blue, respectively. IZUMO4 has a unique C-terminal region that looks black. Confirmed and putative disulfide bonds are shown by solid and dotted lines, respectively. b IZUMO1 (PDB: 5F4E), SPACA6, IZUMO2 (AlphaFold DB: AF-Q6UXV1-F1), IZUMO3 (AlphaFold DB: AF-Q5VZ72-F1), IZUMO4 (AlphaFold DB: AF-Q1ZYL8-F1), and TMEM95 (AlphaFold DB: AF-Q1ZYL8-F1) : AF-Q1ZYL8-F1) : AF-Q3KNT9-F1) are displayed in the same color range as panel A. Disulfide bonds are shown in magenta. TMEM95, IZUMO2 and IZUMO3 transmembrane helices are not shown.
Unlike the IZUMO protein, other SPACA proteins (i.e., SPACA1, SPACA3, SPACA4, SPACA5, and SPACA9) are thought to be structurally different from SPACA6 (Fig. S12). Only SPACA9 has 4HB, but it is not expected to have the same parallel-anti-parallel orientation or the same disulfide bond as SPACA6. Only SPACA1 has a similar Ig-like domain. AlphaFold predicts that SPACA3, SPACA4 and SPACA5 have a completely different structure than SPACA6. Interestingly, SPACA4 is also known to play a role in fertilization, but to a greater extent than SPACA6, instead facilitating the interaction between sperm and oocyte zona pellucida52.
Our AlphaFold search found another match for IZUMO1 and SPACA6 4HB, TMEM95. TMEM95, a single sperm-specific transmembrane protein, renders male mice infertile when ablated 32,33. Spermatozoa lacking TMEM95 had normal morphology, motility, and the ability to penetrate the zona pellucida and bind to the egg membrane, but could not fuse with the oocyte membrane. Previous studies have shown that TMEM95 shares structural similarities with IZUMO133. Indeed, the AlphaFold model confirmed that TMEM95 is a 4HB with the same pair of CXXC motifs as IZUMO1 and SPACA6 and the same additional disulfide bond between helices 1 and 2 found in SPACA6 (Fig. 3a and S11). Although TMEM95 lacks an Ig-like domain, it has a region with a disulfide bond pattern similar to the SPACA6 and IZUMO1 hinge regions (Fig. 3b). At the time of publication of this manuscript, the preprint server reported the structure of TMEM95, confirming the AlphaFold53 result. TMEM95 is very similar to SPACA6 and IZUMO1 and is evolutionarily conserved already in amphibians (Fig. 4 and S13).
The PSI-BLAST search used the NCBI SPACA6, IZUMO1-4, TMEM95, DCST1, DCST2, FIMP, and SOF1 databases to determine the position of these sequences in the tree of life. Distances between branch points are not shown to scale.
The striking overall structural similarity between SPACA6 and IZUMO1 suggests that they are founding members of a conserved structural superfamily that includes the TMEM95 and IZUMO 2, 3, and 4 proteins. known members: IZUMO1, SPACA6 and TMEM95. Because only a few members possess Ig-like domains, the hallmark of the IST superfamily is the 4HB domain, which has unique features common to all of these proteins: 1) Coiled 4HB with helices arranged in an anti-parallel/parallel alternation (Fig. 5a), 2) the bundle has a triangular face consisting of two helices within the bundle and a third vertical helix (Fig. key area (Fig. 5c). The CXXC motif, found in thioredoxin-like proteins, is known to function as a redox sensor 54,55,56 , while the motif in IST family members can be associated with protein disulfide isomerases such as ERp57 in gamete fusion. Roles are associated 57,58.
Members of the IST superfamily are defined by three characteristic features of the 4HB domain: four helices alternating between parallel and antiparallel orientation, ba-triangular helical bundle faces, and a ca CXXC double motif formed between small molecules. ) N-terminal helixes (orange) and hinge region β-hairpin (green).
Given the similarity between SPACA6 and IZUMO1, the ability of the former to bind to IZUMO1 or JUNO was tested. Biolayer interferometry (BLI) is a kinetic-based binding method that has previously been used to quantify the interaction between IZUMO1 and JUNO. After incubation of the biotin-labeled sensor with IZUMO1 as a bait with a high concentration of JUNO analyte, a strong signal was detected (Fig. S14a), indicating a binding-induced change in the thickness of the biomaterial attached to the sensor tip. Similar signals (i.e., JUNO coupled to the sensor as a bait against IZUMO1 analyte) (Fig. S14b). No signal was detected when SPACA6 was used as an analyte against sensor-bound IZUMO1 or sensor-bound JUNO (Figure S14a,b). The absence of this signal indicates that the extracellular domain of SPACA6 does not interact with the extracellular domain of IZUMO1 or JUNO.
Because the BLI assay is based on biotinylation of free lysine residues on the bait protein, this modification can prevent binding if lysine residues are involved in the interaction. In addition, the orientation of the binding relative to the sensor can create steric hindrances, so conventional pull-down assays were also performed on the recombinant SPACA6, IZUMO1 and JUNO ectodomains. Despite this, SPACA6 did not precipitate with His-tagged IZUMO1 or His-tagged JUNO (Fig. S14c,d), indicating no interaction consistent with that observed in BLI experiments. As a positive control, we confirmed the interaction of JUNO with labeled His IZUMO1 (Figures S14e and S15).
Despite the structural similarity between SPACA6 and IZUMO1, the inability of SPACA6 to bind JUNO is not surprising. The surface of human IZUMO1 has more than 20 residues that interact with JUNO, including residues from each of the three regions (although most of them are located in the hinge region) (Fig. S14f). Of these residues, only one is conserved in SPACA6 (Glu70). While many residue substitutions retained their original biochemical properties, the essential Arg160 residue in IZUMO1 was replaced by the negatively charged Asp148 in SPACA6; previous studies have shown that the Arg160Glu mutation in IZUMO1 almost completely abolishes binding to JUNO43. In addition, the difference in domain orientation between IZUMO1 and SPACA6 significantly increased the surface area of ​​the JUNO-binding site of the equivalent region on SPACA6 (Fig. S14g).
Despite the known need for SPACA6 for gamete fusion and its similarity to IZUMO1, SPACA6 does not appear to have an equivalent JUNO binding function. Therefore, we have sought to combine our structural data with evidence of importance provided by evolutionary biology. Sequence alignment of SPACA6 homologues shows the conservation of the common structure beyond mammals. For example, cysteine ​​residues are present even in distantly related amphibians (Fig. 6a). Using the ConSurf server, multiple sequence alignment retention data of 66 sequences were mapped to the SPACA6 surface. This type of analysis can show which residues have been conserved during protein evolution and can indicate which surface regions play a role in function.
a Sequence alignment of SPACA6 ectodomains from 12 different species prepared using CLUSTAL OMEGA. According to the ConSurf analysis, the most conservative positions are marked in blue. Cysteine ​​residues are highlighted in red. Domain boundaries and secondary structure elements are shown at the top of the alignment, where arrows indicate β-strands and waves indicate helices. The NCBI Access Identifiers containing the sequences are: human (Homo sapiens, NP_001303901), mandrill (Mandrilus leucophaeus, XP_011821277), capuchin monkey (Cebus mimic, XP_017359366), horse (Equus caballus, XP_023506102), killer whale (Orcinus orca3_23 XP_032_034) . ), sheep (Ovis aries, XP_014955560), elephant (Loxodonta africana, XP_010585293), dog (Canis lupus familyis, XP_025277208), mouse (Mus musculus, NP_001156381), Tasmanian devil (Sarcophilus harrisii, XP_03611, XP_0318), Platypus, 8) , 61_89 and Bullfrog (Bufo bufo, XP_040282113). The numbering is based on human order. b Surface representation of the SPACA6 structure with 4HB at the top and Ig-like domain at the bottom, colors based on conservation estimates from the ConSurf server. The best preserved parts are in blue, the moderately preserved parts are in white, and the least preserved are in yellow. purple cysteine. Three surface patches demonstrating a high level of protection are shown in the inset labeled patches 1, 2 and 3. A 4HB cartoon is shown in the inset at top right (same color scheme).
The SPACA6 structure has three highly conserved surface regions (Fig. 6b). Patch 1 spans 4HB and the hinge region and contains two conserved CXXC disulfide bridges, an Arg233-Glu132-Arg135-Ser144 hinge network (Fig. S7), and three conserved outer aromatic residues (Phe31, Tyr73, Phe137)). a wider rim of the Ig-like domain (Fig. S6e), which represents several positively charged residues on the sperm surface. Interestingly, this patch contains an antibody epitope that has been previously shown to interfere with SPACA6 30 function. Region 3 spans the hinge and one side of the Ig-like domain; this region contains conserved prolines (Pro126, Pro127, Pro150, Pro154) and outward-facing polar/charged residues. Surprisingly, most of the residues on the surface of 4HB are highly variable (Fig. 6b), although the fold is conserved throughout the SPACA6 homologue (as indicated by the conservatism of the hydrophobic bundle core) and beyond the IST superfamily.
Although this is the smallest region in SPACA6 with the fewest detectable secondary structure elements, many hinge region remnants (including region 3) are highly conserved among SPACA6 homologues, which may indicate that the orientation of the helical bundle and β-sandwich plays a role. as a conservative. However, despite extensive hydrogen bonding and electrostatic networks in the hinge region of SPACA6 and IZUMO1, evidence of intrinsic flexibility can be seen in the alignment of the multiple allowed structures of IZUMO137,43,44. The alignment of the individual domains overlapped well, but the orientation of the domains relative to each other varied from 50° to 70° from the central axis (Fig. S16). To understand the conformational dynamics of SPACA6 in solution, SAXS experiments were performed (Fig. S17a,b). Ab initio reconstruction of the SPACA6 ectodomain conformed to a rod crystal structure (Fig. S18), although the Kratky plot showed some degree of flexibility (Fig. S17b). This conformation contrasts with IZUMO1, in which the unbound protein assumes a boomerang shape both in the lattice and in solution43.
To specifically identify the flexible region, hydrogen-deuterium exchange mass spectroscopy (H-DXMS) was performed on SPACA6 and compared with data previously obtained on IZUMO143 (Fig. 7a,b). SPACA6 is clearly more flexible than IZUMO1, as evidenced by the higher deuterium exchange throughout the structure after 100,000 s of exchange. In both structures, the C-terminal part of the hinge region shows a high level of exchange, which probably allows limited rotation of 4HB and Ig-like domains relative to each other. Interestingly, the C-terminal part of the SPACA6 hinge, consisting of the 147CDLPLDCP154 residue, is a highly conserved region 3 (Fig. 6b), possibly indicating that interdomain flexibility is an evolutionarily conserved feature of SPACA6. According to the flexibility analysis, CD thermal melt data showed that SPACA6 (Tm = 51.2°C) is less stable than IZUMO1 (Tm = 62.9°C) (Fig. S1e and S19).
a H-DXMS images of SPACA6 and b IZUMO1. The percentage deuterium exchange was determined at the indicated time points. Levels of hydrogen-deuterium exchange are indicated by color on a gradient scale from blue (10%) to red (90%). Black boxes represent areas of high exchange. The boundaries of 4HB, hinge and Ig-like domain observed in the crystal structure are shown above the primary sequence. Deuterium exchange levels at 10 s, 1000 s, and 100,000 s were plotted on a strip chart superimposed on the transparent molecular surfaces of SPACA6 and IZUMO1. Parts of structures with a deuterium exchange level below 50% are colored white. Areas above 50% H-DXMS exchange are colored in a gradient scale.
The use of CRISPR/Cas9 and mouse gene knockout genetic strategies has led to the identification of several factors important for sperm and egg binding and fusion. Apart from the well-characterized interaction of IZUMO1-JUNO and CD9 structure, most of the proteins associated with gamete fusion remain structurally and functionally enigmatic. The biophysical and structural characterization of SPACA6 is another piece of the adhesion/fusion molecular puzzle during fertilization.
SPACA6 and other members of the IST superfamily appear to be highly conserved in mammals as well as individual birds, reptiles, and amphibians; in fact, it is thought that SPACA6 is even required for fertilization in zebrafish 59. This distribution is similar to other known gamete fusion-associated proteins such as DCST134, DCST234, FIMP31, and SOF132, suggesting that these factors are HAP2-deficient (also known as GCS1) proteins that are responsible for the catalytic activity of many protists. , plants, and arthropods. Fertilized fusion proteins 60, 61. Despite the strong structural similarity between SPACA6 and IZUMO1, knockout of genes encoding either of these proteins resulted in infertility in male mice, indicating that their functions in gamete fusion are not duplicated. . More broadly, none of the known sperm proteins required for the adhesion phase of fusion is redundant.
It remains an open question whether SPACA6 (and other members of the IST superfamily) participate in intergametic junctions, form intragametic networks to recruit important proteins to fusion points, or perhaps even act as elusive fusogens. Co-immunoprecipitation studies in HEK293T cells revealed an interaction between full length IZUMO1 and SPACA632. However, our recombinant ectodomains did not interact in vitro, suggesting that the interactions seen in Noda et al. were both deleted in the construct (note the cytoplasmic tail of IZUMO1, which has been shown to be unnecessary for fertilization62). Alternatively, IZUMO1 and/or SPACA6 may require specific binding environments that we do not reproduce in vitro, such as physiologically specific conformations or molecular complexes containing other proteins (known or not yet discovered). Although the IZUMO1 ectodomain is believed to mediate attachment of spermatozoa to the egg in the perivitelline space, the purpose of the SPACA6 ectodomain is unclear.
The structure of SPACA6 reveals several conserved surfaces that may be involved in protein-protein interactions. The conserved part of the hinge region immediately adjacent to the CXXC motif (designated Patch 1 above) has several outward-facing aromatic residues that are often associated with hydrophobic and π-stacking interactions between biomolecules. The broad sides of the Ig-like domain (region 2) form a positively charged groove with highly conserved Arg and His residues, and antibodies against this region have previously been used to block gamete fusion 30 . The antibody recognizes the linear epitope 212RIRPAQLTHRGTFS225, which has three of the six arginine residues and highly conserved His220. It is not clear whether the dysfunction is due to blockage of these specific residues or the entire region. The location of this gap near the C-terminus of the β-sandwich indicates cis-interactions with neighboring sperm proteins, but not with oocyte proteins. Furthermore, the retention of a highly flexible proline-rich tangle (site 3) within the hinge may be the site of a protein-protein interaction or, more likely, indicate the retention of flexibility between the two domains. Gender is important for the unknown role of SPACA6. fusion of gametes.
SPACA6 has properties of intercellular adhesion proteins, including Ig-like β-sandwiches. Many adhesive proteins (eg, cadherins, integrins, adhesins, and IZUMO1) possess one or more β-sandwich domains that extend proteins from the cell membrane to their environmental targets63,64,65. The Ig-like domain of SPACA6 also contains a motif commonly found in β-sandwiches of adhesion and cohesion: doublets of parallel strands at the ends of β-sandwiches, known as mechanical clamps66. It is believed that this motif increases resistance to shear forces, which is valuable for proteins involved in intercellular interactions. However, despite this similarity to adhesins, there is currently no evidence that SPACA6 interacts with egg whites. The SPACA6 ectodomain is unable to bind to JUNO, and SPACA6-expressing HEK293T cells, as shown here, hardly interact with oocytes lacking zona 32 . If SPACA6 does establish intergametic bonds, these interactions may require post-translational modifications or be stabilized by other sperm proteins. In support of the latter hypothesis, IZUMO1-deficient spermatozoa bind to oocytes, demonstrating that molecules other than IZUMO1 are involved in the gamete adhesion step 27 .
Many viral, cellular, and developmental fusion proteins have properties that predict their function as fusogens. For example, viral fusion glycoproteins (classes I, II and III) have a hydrophobic fusion peptide or loop at the end of the protein that is inserted into the host membrane. The hydrophilicity map of IZUMO143 and the structure (determined and predicted) of the IST superfamily showed no apparent hydrophobic fusion peptide. Thus, if any proteins in the IST superfamily function as fusogens, they do so in a manner different from other known examples.
In conclusion, the functions of the members of the IST superfamily of proteins associated with gamete fusion remain a tantalizing mystery. Our characterized SPACA6 recombinant molecule and its resolved structure will provide insight into the relationships between these shared structures and their role in gamete attachment and fusion.
The DNA sequence corresponding to the predicted human SPACA6 ectodomain (NCBI accession number NP_001303901.1; residues 27–246) was codon-optimized for expression in Drosophila melanogaster S2 cells and synthesized as a gene fragment with the sequence encoding Kozak (Eurofins Genomics). , the BiP secretion signal and the corresponding 5′ and 3′ ends for ligation-independent cloning of this gene into a pMT expression vector based on a metallothionein promoter modified for selection with puromycin (pMT-puro). The pMT-puro vector encodes a thrombin cleavage site followed by a 10x-His C-terminal tag (Figure S2).
Stable transfection of the SPACA6 pMT-puro vector into D. melanogaster S2 (Gibco) cells was performed similarly to the protocol used for IZUMO1 and JUNO43. S2 cells were thawed and grown in Schneider’s medium (Gibco) supplemented with a final concentration of 10% (v/v) heat-inactivated fetal calf serum (Gibco) and 1X antimycotic antibiotic (Gibco). Early passage cells (3.0 x 106 cells) were plated in individual wells of 6-well plates (Corning). After 24 hours of incubation at 27°C, cells were transfected with a mixture of 2 mg of the SPACA6 pMT-puro vector and Effectene transfection reagent (Qiagen) according to the manufacturer’s protocol. Transfected cells were incubated for 72 hours and then harvested with 6 mg/ml puromycin. Cells were then isolated from complete Schneider’s medium and placed in serum free Insect-XPRESS medium (Lonza) for large scale protein production. A 1 L batch of S2 cell culture was grown to 8–10 × 106 ml-1 cells in a 2 L ventilated flat-bottomed polypropylene Erlenmeyer flask and then sterilized with a final concentration of 500 µM CuSO4 (Millipore Sigma) and sterile filtered. induced. The induced cultures were incubated at 27° C. at 120 rpm for four days.
Conditioned medium containing SPACA6 was isolated by centrifugation at 5660×g at 4°C followed by a Centramate tangential flow filtration system (Pall Corp) with a 10 kDa MWCO membrane. Apply concentrated medium containing SPACA6 to a 2 ml Ni-NTA agarose resin (Qiagen) column. The Ni-NTA resin was washed with 10 column volumes (CV) of buffer A and then 1 CV of buffer A was added to give a final imidazole concentration of 50 mM. SPACA6 was eluted with 10 ml of buffer A supplemented with imidazole to a final concentration of 500 mM. Restriction class thrombin (Millipore Sigma) was added directly to the dialysis tubing (MWCO 12-14 kDa) at 1 unit per mg SPACA6 vs. 1 L 10 mM Tris-HCl, pH 7.5 and 150 mM NaCl (buffer B) for dialysis. ) at 4°C for 48 hours. The thrombin-cleaved SPACA6 was then diluted threefold to reduce salt concentration and loaded onto a 1 ml MonoS 5/50 GL cation exchange column (Cytiva/GE) equilibrated with 10 mM Tris-HCl, pH 7.5. The cation exchanger was washed with 3 CV of 10 mM Tris-HCl, pH 7.5, then SPACA6 was eluted with a linear gradient of 0 to 500 mM NaCl in 10 mM Tris-HCl, pH 7.5 for 25 CV. After ion exchange chromatography, SPACA6 was concentrated to 1 ml and eluted isocratically from an ENrich SEC650 10 x 300 column (BioRad) equilibrated with buffer B. According to the chromatogram, pool and concentrate fractions containing SPACA6. Purity was controlled by Coomassie-stained electrophoresis on a 16% SDS-polyacrylamide gel. Protein concentration was quantified by absorbance at 280 nm using the Beer-Lambert law and the theoretical molar extinction coefficient.
Purified SPACA6 was dialyzed overnight against 10 mM sodium phosphate, pH 7.4 and 150 mM NaF and diluted to 0.16 mg/mL prior to analysis by CD spectroscopy. Spectral scanning of CDs with a wavelength of 185 to 260 nm was collected on a Jasco J-1500 spectropolarimeter using quartz cuvettes with a 1 mm optical path length (Helma) at 25°C at a rate of 50 nm/min. The CD spectra were baseline corrected, averaged over 10 acquisitions, and converted to mean residual ellipticity (θMRE) in degrees cm2/dmol:
where MW is the molecular weight of each sample in Da; N is the number of amino acids; θ is the ellipticity in millidegrees; d corresponds to the length of the optical path in cm; protein concentration in units.