Adapting clinical practice guidelines to local context and assessing barriers to their use

Abstract Extensive efforts have been made to understand the phenotypic diversity of lipopolysaccharide (LPS) structures through deletion and complementation experiments. However, this approach likely underestimates the available phenotypic diversity. To better explore LPS diversity, we generate LPS mutants in Escherichia coli C by selecting for resistance to ΦX174, a bacteriophage that relies solely on binding to core LPS to infect its host. An analysis of 31 E. coli C mutants that are resistant to ΦX174 reveals that each mutant carries at least one mutation in genes linked to core LPS biosynthesis or assembly. Based on which genes are mutated, we predict the core LPS structures of each bacterial mutant, and test our predictions by evolving phages to recognize each evolved LPS structure. We find that phages that evolved to infect the same predicted LPS structure were not always able to cross-infect each other’s host, suggesting that core LPS structure diversity is higher than predicted. Similarly, phage genotype-phenotype maps can be constructed using the bacterial LPS mutant classes. For example, we demonstrate that a combination of two phage mutations leads to loss of the ability to infect wildtype E. coli C. Our results show that phages are a useful tool to study LPS structures, and conversely that the study of LPS structures helps to understand phage evolution and biology.