Molecular basis of egg coat cross-linking sheds light on ZP1-associated female infertility

Abstract

Interaction between sperm and the egg zona pellucida (ZP) is the first step of mammalian fertilization, and ZP component ZP1 is important for fertility by covalently cross-linking ZP filaments into a matrix. Like ZP4, a structurally-related subunit absent in the mouse, ZP1 is predicted to contain an N-terminal ZP-N domain of unknown function. Characterization of ZP1 proteins carrying mutations from infertile patients suggests that, unlike in the mouse, filament cross-linking by ZP1 is crucial for human ZP assembly. We map the function of ZP1 to its ZP-N1 domain and determine crystal structures of ZP-N1 homodimers from a chicken homolog of ZP1.These reveal that ZP filament cross-linking is highly plastic and can be modulated by ZP1 fucosylation and, potentially, zinc sparks. Moreover, we show that ZP4 ZP-N1 forms non-covalent homodimers in chicken but not human. Together, these data identify human ZP1 cross-links as a promising target for non-hormonal contraception.

Mammalian egg coat modifications and the block to polyspermy

“…Biochemical and electron microscopy (EM) analysis of ZP filament material treated with reducing agents or chymotrypsin indicated that, by forming intermolecular disulfide bond(s), ZP1 acts as a cross‐linker of individual ZP filaments (Greve & Wassarman, 1985). This suggestion was supported by a recent crystallographic study of the ZP1 cross‐link that, together with mutagenesis experiments, revealed that a single ZP‐N1 domain Cys conserved among species is essential for ZP1 homodimerization in chicken, mouse, and human (Nishimura et al, 2019). Notably, although ZP4 shares the same domain architecture as ZP1 (Figure 1) and their genes are considered paralogous (Conner et al, 2005; Hughes & Barratt, 1999), the cross‐linking function of the two proteins appears to be conserved in chicken (where ZP4 ZP‐N1 makes noncovalent homodimers that can presumably functionally substitute ZP1 in the germinal disk region of the VE) but not in human (where ZP4 ZP‐N1 is a monomer; Nishimura et al, 2019).…”

“…This suggestion was supported by a recent crystallographic study of the ZP1 cross‐link that, together with mutagenesis experiments, revealed that a single ZP‐N1 domain Cys conserved among species is essential for ZP1 homodimerization in chicken, mouse, and human (Nishimura et al, 2019). Notably, although ZP4 shares the same domain architecture as ZP1 (Figure 1) and their genes are considered paralogous (Conner et al, 2005; Hughes & Barratt, 1999), the cross‐linking function of the two proteins appears to be conserved in chicken (where ZP4 ZP‐N1 makes noncovalent homodimers that can presumably functionally substitute ZP1 in the germinal disk region of the VE) but not in human (where ZP4 ZP‐N1 is a monomer; Nishimura et al, 2019). Because human ZP1 mutations can cause lack of a ZP and sterility (Huang et al, 2014), this suggests that ZP filament cross‐linking plays a much more important role in humans than in mouse (Nishimura et al, 2019).…”

“…The precursor forms of mammalian ZP proteins are composed of a signal peptide (SP), a ZP “domain” module consisting of two immunoglobulin‐like domains (ZP‐N/ZP‐C; Bokhove & Jovine, 2018; Bork & Sander, 1992; Jovine, Qi, Williams, Litscher, & Wassarman, 2004), a consensus furin cleavage site (CFCS; Litscher, Qi, & Wassarman, 1999; Yurewicz, Hibler, Fontenot, Sacco, & Harris, 1993), and a C‐terminal propeptide that contains an external hydrophobic patch (EHP; Jovine et al, 2004), a single‐spanning transmembrane domain (TM; Yurewicz et al, 1993) and a short cytoplasmic tail. The N‐terminal regions of ZP1, ZP2, and ZP4 also contain additional isolated copies of the ZP‐N domain (Callebaut, Mornon, & Monget, 2007; Monné, Han, Schwend, Burendahl, & Jovine, 2008; Nishimura et al, 2019; Raj et al, 2017); moreover, a trefoil (P‐type) domain invariantly precedes the ZP modules of ZP1 and ZP4 (Bork, 1993; Figure 1).…”

“…Scheme of the domain architecture of human and mouse ZP components. The positions of the conserved ZP1‐specific Cys essential for filament cross‐linking (Nishimura et al, 2019) and the post‐fertilization cleavage site of ZP2 important for ZP hardening (Gahlay, Gauthier, Baibakov, Epifano, & Dean, 2010) are marked by magenta arrowheads and black arrows, respectively. Dark gray rectangle, SP; light yellow diamond, trefoil/P domain; red rectangle, CFCS; dark yellow rectangle, EHP; brown rectangle, transmembrane domain.…”

“…Notably, although ZP4 shares the same domain architecture as ZP1 (Figure 1) and their genes are considered paralogous (Conner et al, 2005; Hughes & Barratt, 1999), the cross‐linking function of the two proteins appears to be conserved in chicken (where ZP4 ZP‐N1 makes noncovalent homodimers that can presumably functionally substitute ZP1 in the germinal disk region of the VE) but not in human (where ZP4 ZP‐N1 is a monomer; Nishimura et al, 2019). Because human ZP1 mutations can cause lack of a ZP and sterility (Huang et al, 2014), this suggests that ZP filament cross‐linking plays a much more important role in humans than in mouse (Nishimura et al, 2019). Despite this significant difference, a structurally comparable three‐dimensional network of glycoprotein filaments (5–7‐µm‐thick in mice and up to 13–19‐µm‐thick in humans) surrounds the oocytes of both species.…”