Rapid Accumulation of Motility-Activating Mutations in Resting Liquid Culture of Escherichia coli

Abstract

Expression of motility genes is a potentially beneficial but costly process in bacteria. Interestingly, many isolate strains of Escherichia coli possess motility genes but have lost the ability to activate them under conditions in which motility is advantageous, raising the question of how they respond to these situations. Through transcriptome profiling of strains in the E. coli single-gene knockout Keio collection, we noticed drastic upregulation of motility genes in many of the deletion strains compared to levels in their weakly motile parent strain (BW25113). We show that this switch to a motile phenotype is not a direct consequence of the genes deleted but is instead due to a variety of secondary mutations that increase the expression of the major motility regulator, FlhDC. Importantly, we find that this switch can be reproduced by growing poorly motile E. coli strains in nonshaking liquid medium overnight but not in shaking liquid medium. Individual isolates after the nonshaking overnight incubations acquired distinct mutations upstream of the flhDC operon, including different insertion sequence (IS) elements and, to a lesser extent, point mutations. The rapidity with which genetic changes sweep through the populations grown without shaking shows that poorly motile strains can quickly adapt to a motile lifestyle by genetic rewiring. IMPORTANCE The ability to tune gene expression in times of need outside preordained regulatory networks is an essential evolutionary process that allows organisms to survive and compete. Here, we show that upon overnight incubation in liquid medium without shaking, populations of largely nonmotile Escherichia coli bacteria can rapidly accumulate mutants that have constitutive motility. This effect contributes to widespread secondary mutations in the single-gene knockout library, the Keio collection. As a result, 49/71 (69%) of the Keio strains tested exhibited various degrees of motility, whereas their parental strain is poorly motile. These observations highlight the plasticity of gene expression even in the absence of preexisting regulatory programs and should raise awareness of procedures for handling laboratory strains of E. coli.

Spurious regulatory connections dictate the expression-fitness landscape of translation termination factors

“…S1d). The global gene expression changes resulting from these perturbations were probed with RNA-seq (Parker et al, 2019) (Methods). Expression levels were converted to fractions of the proteome using ribosome profiling (Ingolia et al, 2009; Li et al, 2014) (Methods for details and assumptions).…”

Global coordination of metabolic pathways in Escherichia coli by active and passive regulation

“…The vast majority (88%) of proteins were within twofold of the wild-type concentrations, confirming that deletion of cyaA and cpdA has a limited impact on the proteome independently of cAMP (see Appendix Text 3 and Appendix Fig S7A and B). The few exceptions included proteins belonging to flagella and chemotaxis, which notably are under the control of promoters that frequently acquire mutations during strain development (Parker et al, 2019), and proteins whose operons are located close to the deletion loci (i.e., IlvB and RbsB), which could be attributed to genetic differences in the donor strain used to generate the gene deletions (Lyons et al, 2011) (see Appendix Text 2). Second, we titrated the cAMP concentration below and above the level required to yield wild-type-like growth rate.…”

Spurious regulatory connections dictate the expression‐fitness landscape of translation factors

“…The range of these tunable expression systems spanned 31-to 111-fold and was centered near their respective endogenous levels (Fig EV1D). The global gene expression changes resulting from these perturbations were probed with RNA-seq (Parker et al, 2019) (Materials and Methods), with highly reproducible approaches (Appendix Fig S1). Expression levels were converted to fractions of the proteome using ribosome profiling (Ingolia et al, 2009;Li et al, 2014) (Materials and Methods for details and assumptions).…”

“…As our primary method to quantify gene expression genome-wide, we used a previously reported RNA-seq strategy (Parker et al, 2019). Briefly, 10 ml of culture at OD 600 = 0.1 were collected and mixed with 1 ml of 10:1 ethanol/phenol stop solution by rapid inversion.…”