Fig 1.
E. coli kills Staphylococcus spp in vitro.
(A) Enumeration of S. aureus USA300 grown alone or co-cultured with E. coli UTI89 in macrocolonies for 0, 6, 8, 16, 24 and 48 hours. N = 3 independent experiments. (B) Enumeration of different strains of S. aureus grown alone or together with E. coli UTI89 in macrocolonies for 24 h. N = 6 independent experiments. (C) Enumeration of S. aureus HG001 from single species or mixed species macrocolonies containing the indicated strain of E. coli, at 24 h. N = 6 independent experiments. Statistical significance was determined by one-way ANOVA with Dunnett’s test for multiple comparison. (D) Staphylococci from single species or mixed species macrocolonies co-cultured with E. coli UTI89 for 24 h. N = 6 independent biological experiments. (E) Enumeration of S. aureus after planktonic growth alone or mixed with either E. coli UTI89 or MG1655 for 24 or 48 hours. N = 6 independent experiments (A-E) Data from single species macrocolonies or planktonic culture are indicated with open bars, and data from mixed species (all inoculated at a ratio of 1EC:1SA) macrocolonies are indicated with checked bars. Individual data points from each biological replicate are indicated with closed circles. (B, D, E) Statistical significance was determined by two-way ANOVA with Sidak’s (A and C) and Tukey’s (D) test for multiple comparisons. ****p< 0.0001. Error bars represent SD from the mean. All statistical tests were performed on log-transformed CFU data. (See S1 Fig for paired E. coli CFU to match the S. aureus data shown here).
Fig 2.
Co-culture of E. coli and S. aureus promotes differential gene expression.
Transcription comparison between single species S. aureus strain HG001 or E. coli strain UTI89 macrocolonies and mixed species macrocolonies (1EC:1SA) after 6 h incubation, a time point at which sufficient live S. aureus could be recovered. Vertical black lines represent median values for each gene category. Each circle represents a gene that is differentially regulated (p<0.05, FDR<0.05) in the mixed species macrocolony compared to the single species macrocolony in the respective functional categories, with blue color indicating a functional category where the median value shows increased expression in the mixed species macrocolony and red color indicating decreased expression. Closed magenta circles represents recA and uvrAB genes. Data represent ≥2 biological replicates.
Fig 3.
The BarA-UvrY two component system (TCS) and the pks island are required for E. coli-mediated killing of S. aureus.
(A) Enumeration of S. aureus USA300 LAC and mixed (1EC:1SA) macrocolonies with either UTI89 wild type, knockout mutants of the pks island, barA or uvrY or complemented strains. The vertical dotted line indicates that data collected on either side were collected from separate experiments. Data from single species macrocolonies are indicated with open bars, and data from mixed species macrocolonies are indicated with checked bars. N = 6 independent biological experiments. Statistical significance was determined by one-way ANOVA with Dunnett’s test for multiple comparison. (B) Enumeration of S. aureus USA300 LAC from 8 h macrocolonies. Wild type S. aureus USA300 LAC and uvrABC transposon mutants were mixed 1EC:1SA with either E. coli UTI89 or knockout mutants of the pks island. N = 6 independent experiments. (C) Enumeration of S. aureus from 24 h macrocolonies. Wild type S. aureus USA300 LAC was transformed with pJC-2343 (pEmpty) or pJC-2343-ClbS (pClbS) and mixed 1:1 with either E. coli UTI89 or knockout mutants of the pks island. N = 6 independent experiments. Individual data points are indicated with closed circles. (D) Enumeration of S. aureus from 24 h macrocolonies. Wild type S. aureus USA300 LAC or USA300 pClbS was mixed 1:1 with either E. coli UTI89, UTI89 Δpks island, pks+ or pks- clinical isolate. N = 6 independent experiments. (B-D) Statistical significance was determined by two-way ANOVA with Tukey’s test for multiple comparisons. ****p< 0.0001, error bars represent SD from the mean. All statistical tests were performed on log-transformed CFU data.
Fig 4.
Co-culture of E. coli and S. aureus induces pks island expression in a BarA-UvrY TCS dependent manner.
(A) RT-qPCR of E. coli single species macrocolonies and E. coli ΔuvrY single species macrocolonies at 24 h. (B) RT-qPCR of E. coli single species macrocolonies and E. coli mixed species macrocolonies at 24 h. (C) RT-qPCR of E. coli mixed species macrocolonies and E. coli ΔuvrY mixed species macrocolonies at 24 h. (D) RT-qPCR of E. coli single species macrocolonies and E. coli ΔuvrY mixed species macrocolonies at 24 h. N = 3 independent experiments, each the average of 4 technical replicates. Gene expression was normalized to the gyrA housekeeping gene. Individual data points from each biological replicate are indicated with closed circles. Statistical significance was determined by Bonferroni’s multiple comparisons test for two-way ANOVA, ****p< 0.0001, error bars represent SD from the mean.
Fig 5.
The BarA-UvrY TCS regulates the pks island via the Csr system.
(A) RT-qPCR of 16 h macrocolonies of E. coli pTrc99a (pEmpty) and E. coli pTrc99a-CsrA (pCsrA). (B) RT-qPCR of 16 h macrocolonies of E. coli pEmpty and E. coli pCsrB. N = 5–6 independent experiments, each the average of 2 technical replicates. Gene expression was normalized to the gyrA housekeeping gene. Statistical significance was determined by two-way ANOVA with Bonferroni’s test for multiple comparison, **p< 0.01, ***<p< 0.001, ****p< 0.0001, error bars represent SD from the mean. (C) Enumeration of S. aureus from 16 h mixed macrocolonies (1EC:1SA) with either E. coli pEmpty, E. coli pCsrA or E. coli pCsrB. N = 6 independent experiments. (D) Enumeration of S. aureus mixed macrocolonies (1EC:1SA) from 16 h with either E. coli WT, E. coli ClbRmut or E. coli pks deletion mutant. N = 6 independent experiments. Individual data points are indicated with closed circles. (C and D) Statistical significance was determined by Ordinary One-way ANOVA with Dunnett’s test for multiple comparison. ****p< 0.0001, error bars represent SD from the mean. Statistical tests were performed on log-transformed data.
Fig 6.
E. coli antagonizes S. aureus growth during wound infection and antagonism is dependent on the pks island and the BarA-UvrY TCS.
Mice were co-infected with E. coli UTI89 and S. aureus USA300 LAC at 1–2 x 106 CFU/wound. Wound CFU were enumerated at 24 h post infection. S. aureus single species infection or co-infection with (A) E. coli UTI89 WT, (B) E. coli pks mutant, or (C) BarA-UvrY TCS mutants. Each black circle represents one mouse, horizontal lines represent the median. N = 2 independent experiments, each with 5–6 mice per group. Statistical analysis was performed using Kruskal-Wallis test with Dunn’s post-test to correct for multiple comparisons. *p< 0.05, **p< 0.01, ****p< 0.0001. (See S6 Fig for paired E. coli CFU to match the S. aureus data shown here).
Fig 7.
Proposed model of E. coli and S. aureus polymicrobial interactions.
(A) In the absence of a signal, the BarA-UvrY two component system (TCS) is inactive. CsrA regulates the pks island negatively by binding the mRNA of ClbR, leading to the downregulation of pks island genes. (B) When E. coli and S. aureus are in proximity, the BarA-UvrY TCS is activated by the presence of specific signals, such as short-chained fatty acids. Activation of the TCS leads to upregulated expression of sRNA CsrB, which is a negative regulator of CsrA. CsrB binds to CsrA and relieves the suppression on the pks island, resulting in the upregulation of pks island genes and increased synthesis of pre-colibactin. Maturation of pre-colibactin leads to the release of both NMDA and colibactin into the environment. Entry of colibactin into S. aureus, which possibly occurs through NMDA-mediated membrane disruptions, ultimately leads to DNA damage and growth inhibition of S. aureus. The exact mechanism of colibactin entry into S. aureus remains to be determined. Created with BioRender.com.
Table 1.
List of bacterial strains and plasmids used in this study.