Solution structure of sperm lysin yields novel insights into molecular dynamics of rapid protein evolution Damien B. Wilburna,1, Lisa M. Tuttleb, Rachel E. Klevitb, and Willie J. Swansona aDepartment of Genome Sciences, University of Washington, Seattle, WA 98195; and bDepartment of Biochemistry, University of Washington, Seattle, WA 98195 Edited by Andrew G. Clark, Cornell University, Ithaca, NY, and approved December 19, 2017 (received for review June 1, 2017) Protein evolution is driven by the sum of different physiochemical Fig. 1). The abalone VE is largely composed of a gigantic ∼1-MDa and genetic processes that usually results in strong purifying glycoprotein, termed the VE receptor for lysin (VERL) which selection to maintain biochemical functions. However, proteins contains 23 tandem repeats of a ZP-N polymerization domain. that are part of systems under arms race dynamics often evolve at These ZP-N repeats are thought to self-assemble through hydro- unparalleled rates that can produce atypical biochemical proper- gen bonding to form intermolecular β-sheets (9), and lysin dissolves ties. In the marine mollusk abalone, lysin and vitelline envelope the VE by competing for these hydrogen bonds (10). Molecular ∼ receptor for lysin (VERL) are a pair of rapidly coevolving proteins evolutionary analysis revealed that 25 of 134 residues in lysin are that are essential for species-specific interactions between sperm under positive selection (11), which may play a role in maintaining and egg. Despite extensive biochemical characterization of lysin— species boundaries. Off the North American Pacific coast, seven including crystal structures of multiple orthologs—it was unclear sympatric species of abalone are found with overlapping breeding how sites under positive selection may facilitate recognition of seasons; while viable hybrids can be produced in the laboratory, they are rarely observed in nature (4). Additionally, lysin- VERL. Using a combination of targeted mutagenesis and multidi- mediated VE dissolution is species-specific (10) and helps mensional NMR, we present a high-definition solution structure of Haliotis rufescens maintain species boundaries. Lysin is secreted as a dimer, sug- sperm lysin from red abalone ( ). Unapparent gesting an equivalent stoichiometry to bind dimerized VERL ZP-N from the crystallography data, multiple NMR-based analyses con- repeats; however, prior biochemical experiments suggested the ly- ducted in solution reveal clustering of the N and C termini to form sin monomer is the bioactive unit (12). The majority of the VERL a nexus of 13 positively selected sites that constitute a VERL bind- repeats are >98% identical at the nucleotide level within a species, ing interface. Evolutionary rate was found to be a significant pre- homogenized by concerted evolution (13). The two most N- dictor of backbone flexibility, which may be critical for lysin terminal ZP-N repeats are not subjected to the same homogeni- bioactivity and/or accelerated evolution. Flexible, rapidly evolving zation and are rapidly coevolving with lysin, suggestive of arms race segments that constitute the VERL binding interface were also the dynamics (5, 14, 15). The specific molecular interactions between most distorted regions of the crystal structure relative to what lysin and VERL that facilitate VE dissolution remain unknown. was observed in solution. While lysin has been the subject of ex- Consistent with protein structure generally being more highly tensive biochemical and evolutionary analyses for more than conserved than sequence (16), multiple structures determined by 30 years, this study highlights the enhanced insights gained from X-ray crystallography support that the lysin tertiary structure is applying NMR approaches to rapidly evolving proteins. generally conserved despite its extraordinarily rapid evolution. Nearly identical crystal structures were observed for lysin dimers fertilization | sexual selection | nuclear magnetic resonance Significance he innovation of elaborate male ornaments and female pref- Terences by sexual selection has been of interest to evolutionary The fertilization of eggs by sperm is a critical biological process for biologists since Darwin (1). Compared with other coevolving traits, nearly all sexually reproducing organisms to propagate their ge- sexual ornaments and preferences often evolve via runaway selec- netic information, yet the molecules that mediate egg−sperm in- tion to produce exacerbated traits (2, 3). A similar phenomenon has teractions evolve at extraordinary rates, and their biochemical been observed in genes coding for reproductive proteins that evolve mechanisms are poorly understood. In the marine mollusk abalone, at extraordinary rates, and are usually among the fastest evolving sperm lysin interacts with egg vitelline envelope receptor for lysin genes in any genome (4). This pattern has been observed across (VERL) in a species-specific manner to facilitate fertilization. In this diverse taxa (microbes, plants, and animals) and different types of report, we characterized the solution structure and molecular reproductive proteins (sex pheromones, seminal fluid proteins, and evolution of sperm lysin from red abalone (Haliotis rufescens), and gamete recognition proteins) (5). The rapid molecular evolution of identified that the VERL binding interface has experienced in- reproductive genes is thought to be driven by the continual co- cessant sexual coevolution. Furthermore, increased dynamics in the evolution between interacting pairs of male and female proteins. lysin fold has facilitated this exacerbated evolution, and may re- However, there presently exist few systems where the interacting flect a common molecular basis for accelerated evolution in other proteins from both sexes have been identified, creating a barrier to rapidly evolving proteins, such as immune genes. understanding the structural and biochemical interactions of re- productive proteins within an evolutionary framework. Author contributions: D.B.W. and W.J.S. designed research; D.B.W. and L.M.T. performed A textbook example of such a rapidly evolving reproductive pro- research; D.B.W. and L.M.T. analyzed data; D.B.W., R.E.K., and W.J.S. wrote the paper; tein is abalone sperm lysin (6), a highly positively charged 16-kDa and R.E.K. and W.J.S. supervised the project. protein that is essential for dissolving the vitelline envelope (VE) of The authors declare no conflict of interest. abalone oocytes (7). The VE—termed the zona pellucida (ZP) in This article is a PNAS Direct Submission. mammals—is a glycoprotein envelope that serves as a barrier to Published under the PNAS license. the sperm and the external environment (8). In many taxa, sperm Data deposition: The NMR chemical shifts have been deposited in the BioMagResBank, secrete proteins that nonenzymatically dissociate the VE, creating www.bmrb.wisc.edu (accession no. 30246) and the Research Collaboratory for Structural a hole through which sperm can pass to access the plasma mem- Bioinformatics Protein Data Bank, www.wwpdb.org (RCSB PDB ID code 5UTG). brane and fuse to complete fertilization (4). While the molecular 1To whom correspondence should be addressed. Email: [email protected]. mechanisms and essential sperm proteins are unknown in mam- This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10. mals, the process is well characterized in abalone (illustrated in 1073/pnas.1709061115/-/DCSupplemental. 1310–1315 | PNAS | February 6, 2018 | vol. 115 | no. 6 www.pnas.org/cgi/doi/10.1073/pnas.1709061115 Downloaded by guest on September 24, 2021 VERL repeat 1 produced similar levels of binding relative to controls (Fig. 2D), suggesting minimal difference in VERL af- finity between dimeric and monomeric lysin. Therefore, while the lysin monomer can bind and recognize VERL repeats, the lysin dimer is required for efficient VE dissolution. The VERL Binding Interface Comprises Several Clustered Positively Selected Sites. To understand how the N terminus may be involved in interactions with individual VERL repeats—despite being “extended” in the available crystal structures—we determined the solution structure of F104A lysin using a suite of multidimensional heteronuclear NMR and restrained molecular dynamics simula- Fig. 1. Proposed model of lysin-mediated VE dissolution. VERL is a gigantic tion (Fig. 3). An ensemble of the 20 lowest energy models was well molecule (∼1 MDa) that includes 23 repeats of a ZP-N polymerization defined by the NMR restraints (backbone rmsd = 1.6 Å; Table module. The VE supramolecular structure is supported by intermolecular S1), with regions of backbone flexibility at the N terminus (resi- hydrogen bonds between VERL ZP-N repeats. The two most N-terminal re- dues 1 to 12) and in the turns that link α-helices (residues 39 to 43, peats are rapidly coevolving with lysin, while the remaining repeats are 77 to 81, and 95 to 97). However, comparison of the solution and homogenized through concerted evolution (15). Upon sperm contact with crystal structures revealed several discrepancies (Fig. 4 A and B). the VE, lysin will disrupt the hydrogen bonds of the first repeat, forming a Although extended and flexible in the crystal structure, the N ter- − pair of lysin VERL repeat heterodimers, and expose the next pair of repeats. minus forms a nexus with the C terminus and residues 60 to 65 that Lysin can then disrupt the second pair of hydrogen-bonded VERL repeats, appears to be driven largely