Fig. 2: Strain displacement occurs as predicted with engineered E. coli strains.
From: Strain displacement in microbiomes via ecological competition
a, Scheme of invasion experiments using isogenic E. coli strains. LB medium is supplemented with 4% sorbitol and then inoculated with either wild-type (WT) E. coli or an isogenic ∆srlAEB mutant that cannot use sorbitol. A WT E. coli with or without colicin E2 is inoculated 8 h later. Populations are enumerated using selective plating. b, Ecological invasion experiment using the scheme in a for scenarios modelled in Fig. 1c–f. Lines connect medians, and dashed black lines indicate the detection limits of selective plating. Here the invader does not have a private nutrient (resident is WT E. coli; kanamycin resistant; open blue circles) and does not have a toxin (invader is WT E. coli; chloramphenicol resistant; open red circles). N = 9 independent biological replicates, each from independent experiments. c, Invader has a private nutrient (resident is E. coli ∆srlAEB; kanamycin resistant; blue circles with grey fill) and does not have a toxin (invader is WT E. coli). N = 8 independent biological replicates each from independent experiments. d, Invader lacks a private nutrient (resident is WT E. coli) but has a toxin (invader is E. coli with colicin E2; chloramphenicol resistant; red circles with black fill). N = 5 independent biological replicates each from independent experiments. e, Invader has both a private nutrient (resident is E. coli ∆srlAEB) and a toxin (invader is E. coli with colicin E2). N = 5 independent biological replicates each from independent experiments. f,g, Same experiment as in b–e, but only the end-point values at 72 h are plotted for both the invader (f) and the resident (g). See descriptions above for sample size. Two-tailed Mann–Whitney U tests are used to compare population sizes when the invader does not use a toxin with when the invader uses either colicin E2 or colicin K. Population dynamics of ecological invasion experiments using colicin K are shown in Extended Data Fig. 6i,j (N = 5 independent biological replicates each from independent experiments for experiments with colicin K). Black lines indicate medians; dashed lines indicate the detection limits of selective plating.
