Computationally optimized anti-staphylococcal biotherapeutics

描述（由申请人提供）：抗生素耐药性使大多数金黄色葡萄球菌（金黄色葡萄球菌）感染复杂化，因为三分之二的医院相关金黄色葡萄球菌感染和大约50%的社区获得性金黄色葡萄球菌感染现在是耐甲氧西林（MRSA）。金黄色葡萄球菌和其他细菌的多重耐药发生率日益增加，这突出表明需要能够对抗这些危险病原体的下一代抗生素。虽然传统的小分子抗生素抑制遗传编码的细胞内酶，但另一种策略是使用天然裂解酶的重组版本，如拟葡萄球菌溶葡萄球菌蛋白（ssLST），它通过催化降解细胞壁起作用，因此可能对进化抗性的易感性较低。不幸的是，作为一种细菌蛋白本身，已知ssLST会驱动有效的免疫反应，这为ssLST治疗的临床开发提供了主要障碍。本研究假设，通过将新的计算去免疫算法与前沿的生物分子工程和免疫原性筛选技术相结合，我们可以在分子水平上重新设计ssLST，从而在保持野生型稳定性和催化功能的同时降低免疫原性。将同时寻求两种互补的方法来开发去免疫的ssLST变体。目的1旨在通过计算设计和实验评估预测同时具有良好活性和降低免疫原性的少数变异。设计算法将采用详细的序列和结构建模，以选择最优的去免疫突变集。工程变异体的杀菌活性将通过测定最小抑制浓度（MIC）、最小杀菌浓度（MBC）和金黄色葡萄球菌裂解动力学来量化。将在转基因小鼠模型中使用抗体滴度、炎症细胞因子分泌和T细胞激活作为读数来评估工程变体的免疫原性。Aim 2旨在通过计算设计预测富含免疫原性降低变异的组合文库，然后采用高通量活性筛选来鉴定活性变异以进行进一步评估。设计算法在选择用于文库构建的突变时将主要针对免疫原性进行优化，留下筛选来识别高度活跃的文库成员，以进行详细的表征，如Aim 1所示。成功实现这些目标将产生强大的算法，用于优化治疗蛋白的个体变异和文库，广泛适用的基于荧光的检测，可以超高通量筛选基因工程抗菌蛋白，以及功能齐全的非免疫原性抗葡萄球菌生物催化剂，可能对治疗耐药金黄色葡萄球菌感染有用。

DESCRIPTION (provided by applicant): Antibiotic resistance complicates the majority of Staphylococcus aureus (S. aureus) infections, as a full two thirds of hospital-associated S. aureus infections and ~50% of those acquired in the community are now methicillin-resistant (MRSA). The increasing incidence of multi-drug resistance in S. aureus and other bacteria underscores the need for next-generation antibiotics capable of combating these dangerous pathogens. While traditional small molecule antibiotics inhibit genetically-encoded intracellular enzymes, an alternative strategy is to employ recombinant versions of natural lytic enzymes such as Staphylococcus simulans lysostaphin (ssLST), which acts by catalytic degradation of the cell wall and may therefore have lower susceptibility to evolved resistance. Unfortunately, as a bacterial protein itself, ssLST is known to drive a potent immune response, providing a major barrier to clinical development of ssLST therapies. This proposal hypothesizes that by integrating novel computational deimmunization algorithms with cutting- edge biomolecular engineering and immunogenicity screening technologies, we can redesign ssLST at the molecular level so as to maintain wild-type stability and catalytic function while simultaneously reducing immunogenicity. Two complementary approaches to developing deimmunized ssLST variants will be pursued in parallel. Aim 1 seeks to computationally design and experimentally evaluate a small number of variants predicted to have simultaneously good activity and reduced immunogenicity. The design algorithms will employ detailed modeling of sequence and structure in order to select optimal sets of deimmunizing mutations. The bactericidal activity of the engineered variants will be quantified by determination of Minimal Inhibitory Concentration (MIC), Minimal Bactericidal Concentration (MBC), and S. aureus lysis kinetics. The immunogenicity of the engineered variants will be assessed in a transgenic mouse model using antibody titers, inflammatory cytokine secretion, and T cell activation as readouts. Aim 2 seeks to computationally design combinatorial libraries predicted to be enriched in variants with reduced immunogenicity, and then employ high-throughput activity screening to identify active variants for further evaluation. The design algorithms will optimize primarily for immunogenicity in selecting mutations for library construction, leaving the screens to identify highly active library members for detailed characterization as in Aim 1. Successfully achieving these aims will result in powerful algorithms for optimizing individual variants and libraries of therapeutic proteins, a broadly applicable fluorescence-based assay enabling ultra-high-throughput screening of genetically-engineered antibacterial proteins, and fully functional, non-immunogenic, anti- staphylococcal biocatalysts potentially useful in treating drug-resistant S. aureus infections.