Large metabolic rewiring from small genomic changes between strains of Shigella flexneri

Abstract

The instability of Shigella genomes has been described, but how this instability causes phenotypic differences within the Shigella flexneri species is largely unknown and likely variable. We describe herein the genome of S. flexneri strain PE577, originally a clinical isolate, which exhibits several phenotypic differences compared to the model strain 2457T. Like many previously described strains of S. flexneri, PE577 lacks discernible, functional CRISPR and restriction-modification systems. Its phenotypic differences when compared to 2457T include lower transformation efficiency, higher oxygen sensitivity, altered carbon metabolism, and greater susceptibility to a wide variety of lytic bacteriophage isolates. Since relatively few Shigella phages have been isolated on 2457T or the previously characterized strain M90T, developing a more universal model strain for isolating and studying Shigella phages is critical to understanding both phages and phage-host interactions. In addition to phage biology, the genome sequence of PE577 was used to generate and test hypotheses of how pseudogenes in this strain-whether interrupted by degraded prophages, transposases, frameshifts, or point mutations-have led to metabolic rewiring compared to the model strain 2457T. Results indicate that PE577 can utilise the less-efficient pyruvate oxidase/acetyl-CoA synthetase (PoxB/Acs) pathway to produce acetyl-CoA, while strain 2457T cannot due to a nonsense mutation in acs, rendering it a pseudogene in this strain. Both strains also utilize pyruvate-formate lyase to oxidize formate but cannot survive with this pathway alone, possibly because a component of the formate-hydrogen lyase (fdhF) is a pseudogene in both strains.Importance Shigella causes millions of dysentery cases worldwide, primarily affecting children under five years old. Despite active research in developing vaccines and new antibiotics, relatively little is known about the variation of physiology or metabolism across multiple isolates. In this work, we investigate two strains of S. flexneri that share 98.9% genetic identity but exhibit drastic differences in metabolism, ultimately affecting the growth of the two strains. Results suggest additional strains within the S. flexneri species utilize different metabolic pathways to process pyruvate. Metabolic differences between these closely-related isolates suggest an even wider variety of differences in growth across S. flexneri and Shigella in general. Exploring this variation further may assist the development or application of vaccines and therapeutics to combat Shigella infections.

FIG 1

Genome maps and features of the S. flexneri PE577 chromosome and plasmids, with coding sequences (CDS), tRNA, rRNA, and GC characteristics colored as indicated. GC skew refers to the GC richness of a given region of the genome.

FIG 2

Growth of strains CFS100 and PE577 in nutrient-rich (LB) or nutrient-poor (M9) medium in aerobic conditions. (A and B) Measurements are based on CFU per milliliter over time or optical density at 600 nm (B). (C) The relationship between these two measurements.

FIG 3

(A) Pathways of pyruvate oxidation in S. flexneri. (B) Comparison of genes involved in the PoxB/Acs bypass or in formate detoxification between PE577 and CFS100. In the latter, PoxB is grayed out because CFS100 encodes a truncated RpoS. Since RpoS controls poxB expression, this may affect protein abundance. Similarly, FdhF is grayed out because it is a pseudogene in both backgrounds, but a different formate dehydrogenase may associate with the FHL complex.

FIG 4

CFU per milliliter (A) and concentrations of lactate (B), acetate (C), and formate (D) in ΔhycC strains with and without complementation after overnight growth. The strain is indicated on the x axis. Error bars represent the standard error from three biological replicates.

FIG 5

CFU per milliliter (A) and concentrations of lactate (B), acetate (C), and formate (D) in strains lacking various components of pyruvate oxidation pathways after overnight growth. The strain is indicated on the x axis. Error bars represent the standard error from three biological replicates.