Systems biology approach to elucidate complex metabolic dependencies in the evolution of antibiotic resistance

项目摘要 我们提出了一个系统生物学的方法来调查之间的联系代谢和出现 两种主要的人类病原体，铜绿假单胞菌和 金黄色葡萄球菌，在生理相关环境中。两个最严重的 AMR、铜绿假单胞菌和S.金黄色葡萄球菌导致超过30，000和300，000耐药感染 在美国，每年[4]。此R 01应用程序的目标是构建计算 铜绿假单胞菌和S.金黄色葡萄球菌和 实验验证模型预测的代谢基因和途径，支持他们的进化， 阻力为了对抗机会性感染和医院感染，已经开发了用于临床的抗微生物剂。 使用;然而，随着每一种新的抗菌剂的引入，上述病原体很快就产生了耐药性。 [12，13]。因此，迫切需要新的方法来限制这些地区AMR的出现和扩大。 物种本申请集中于革兰氏阴性菌（铜绿假单胞菌）和革兰氏阳性菌（S。aureus） 细菌，我们将比较支持抗菌素耐药性的代谢功能，以评估 观察到的关键代谢基因和途径是生物体所特有的或更普遍的机制 代谢适应性。我们认为，表型分析、计算建模和 对临床抗菌药物耐药菌株的分析将揭示细菌的代谢过程， 对抗微生物剂抗性微生物的生长至关重要，并且可以被调节以选择对抗生长 对抗生素有抗药性的人群。我们实验室和其他实验室的数据最近发现，抗菌剂- 耐药铜绿假单胞菌和S.金黄色葡萄球菌在代谢中显示系统水平的差异[2，14-16]。此外，委员会认为， 从实验中收集的数据将使：新的方法来整合基因表达数据与代谢 复杂环境的网络模型，更好地解释代谢不确定性的集成方法， 网络重建和机器学习方法来描述与AMR相关的代谢状态 在临床分离物中。所提出的工作的意义在于机械地理解本质 基因、反应和底物偏好是两种临床重要的AMR发展的基础， 病原体此外，代谢状态将在临床分离株中表征，以支持代谢状态与临床分离株的相关性。 AMR发展过程中代谢依赖性的系统研究。知识差距这项工作 将解决的是代谢和AMR的发展之间的机制联系。随着成功 实施拟议的项目，我们将确定几个潜在的治疗目标，在临床上重要的 病原体以及建立一个框架，如何计算模型驱动的方法可以 应用于其他抗微生物耐药病原体。

Project Summary We propose a systems biology approach to investigate the connection between metabolism and the emergence of antimicrobial resistance (AMR) in two prominent human pathogens, Pseudomonas aeruginosa and Staphylococcus aureus, within the context of physiologically-relevant environments. Two of the most serious threats for AMR, P. aeruginosa and S. aureus cause more than 30,000 and 300,000 drug-resistant infections per year in the United States, respectively [4]. The objective of this R01 application is to construct computational metabolic network models of antimicrobial-resistant and -sensitive P. aeruginosa and S. aureus and to experimentally validate model predictions of metabolic genes and pathways supporting their evolution towards resistance. To combat opportunistic and nosocomial infections, antimicrobials have been developed for clinical use; yet with every new antimicrobial introduced, the aforementioned pathogens have quickly evolved resistance [12, 13]. Therefore, new approaches are urgently needed to limit emergence and expansion of AMR in these species. With a focus in this application on Gram-negative (P. aeruginosa) and Gram-positive (S. aureus) bacteria, we will compare metabolic functions supportive of antimicrobial resistance to evaluate whether the metabolic genes and pathways observed to be critical are unique to an organism or a more general mechanism of metabolic adaptation. We posit that a combination of phenotyping, computational modeling, and analysis of clinical antimicrobial-resistant strains will reveal bacterial metabolic processes that are central to the growth of antimicrobial-resistant microbes and can be modulated to select against growth of antimicrobial-resistant populations. Data from our lab and others have recently found that antimicrobial- resistant P. aeruginosa and S. aureus display systems-level differences in metabolism [2, 14-16]. Furthermore, data collected from experiments will enable: novel approaches to integrate gene expression data with metabolic network models for complex environments, ensemble methods that better account for uncertainty in metabolic network reconstructions, and machine learning methods to delineate metabolic states that correlate with AMR in clinical isolates. The significance of the proposed work lies in the mechanistic understanding of essential genes, reactions, and substrate preferences that underlie the development of AMR in two clinically important pathogens. Further, metabolic states will be characterized in clinical isolates to support the relevance of the systematic interrogation of metabolic dependencies in the development of AMR. The knowledge gap this work will address is the mechanistic link between metabolism and the development of AMR. With the successful implementation of the proposed project, we will identify several potential therapeutic targets in clinically important pathogens as well as establish a framework for how such computational model-driven approaches can be applied to other antimicrobial resistant pathogens.