PROJECT SUMMARY Bacterial vaginosis (BV) is the most common vaginal dysbiosis in women of reproductive age and is associated with a number of serious health complications. Gardnerella vaginalis is believed to play a critical role in the etiology of BV, functioning as a pioneering organism that can displace beneficial lactobacilli and initiate BV biofilm development. As G. vaginalis proliferates, the pH of the vaginal environment increases, leading to further decreases in lactobacilli and enabling the growth of additional BV-associated bacteria within these biofilms. Standard antibiotic treatment for BV suffers from high relapse rates, the development of antibiotic resistant bacteria, and collateral damage to the healthy lactobacilli in the vaginal microbiome. Given these drawbacks, novel antimicrobial agents that can provide alternatives to antibiotics and can selectively target BV-associated bacteria without damaging beneficial bacteria are urgently needed. Endolysins are phage-encoded enzymes that can degrade bacterial cell walls. Exogenously added endolysins can quickly lyse their target bacteria and because they bind very specific epitopes in target cell walls, they can have lytic specificity down to a single species or even sub-species. Given these properties, endolysins hold enormous potential as high-specificity microbiome modulators. Recently, endolysins targeting G. vaginalis have been identified and demonstrated to kill G. vaginalis. However, the endolysins that have been characterized thus far are not well suited for commercial development. Endolysins that are more active, more thermostable, and function at lower pH are needed to be broadly useful for treating BV. At Topaz Biosciences, we have developed a proprietary metagenomic platform for the discovery and optimization of endolysins. We have previously leveraged this platform to develop endolysins against Staphylococcus aureus that are more active, more thermostable, and have a broader pH range than benchmark enzymes. In this Phase I proposal, we will leverage this platform to expand the diversity of endolysins known to have activity against G. vaginalis and then exploit this diversity to develop chimeric enzymes with improved properties. To accomplish this, we will take advantage of the modularity of endolysins to build a library of endolysin “parts” – enzymatic domains (EADs) and cell wall binding domains (CBDs) that we will systemically characterize for anti-Gardnerella activity, thermostability, pH tolerance, biofilm reduction, and genus specificity. These parts will be sourced both from recently identified endolysins from Gardnerella prophages as well from diverse endolysins/domains from proprietary and public metagenomic databases that we predict to have anti-Gardnerella activity. Finally, we will leverage sequence-function insights gained from our “parts” development to design chimeric endolysins composed of EADs and CBDs with the most promising properties to generate enzymes...