Subinhibitory chloramphenicol exposure drives B. subtilis mobilization.

A. Sliding motility is induced in B. subtilis spots (vertical) when cocultured near S. venezuelae (horizontal spots) due to Cm exposure (imaged at 24 hours after inoculation, 30°C). B. Direct plating of B. subtilis on an agar plate with 1µM Cm leads to mobilization, and not in the control without Cm. C. Transplantation of B. subtilis cells from the mobilized population to different agar plates with the following concentrations of Cm: 0 μM, 1 μM (sub-MIC) and 16 μM (MIC). Bars, 1 cm.

Subinhibitory chloramphenicol exposure supresses protein synthesis.

The extent of protein synthesis was reduced in the presence of 1 μM subinhibitory concentration of Cm by Click-iT assay. C1 is a control without HPG. C2 is a control without Alexa Fluor 488. 16 µM Cm is equal to 1-fold MIC for Cm. The decrease of Cm-treated group was calculated based on the average of 0 µM Cm group. Each group contains three biological replicates (p value < 0.05 between 0 µM Cm and 1 µM Cm group). The cells were incubated with Cm for 5 minutes prior to sample fixation and processing.

Patterns of gene expression reflect regulated changes in metabolism within the mobilized population.

Functional classification of genes that change ≥2-fold with adjusted p value<0.05 in response to chloramphenicol at 6 h (A) and 24 h (B). a. Purine biosynthesis and salvage; b. Purine catabolism; c. Pyrimidine metabolism; d. Utilization of nitrogen sources; e. Carbon metabolism; f. Amino acid metabolism; g. Ribosomal biogenesis and translation; h. Secondary metabolism; i. Metabolism of cofactors; j. Stress response; k. Sporulation and germination; l. Others; m. unknown. Red: ≥ 2-fold increase, Blue: ≥ 2-fold decrease relative to control samples.

A network view of regulation highlights pathways of control for adaptive colony surface expansion.

Cytoscape networks of regulators with ≥ 30% engagement of regulated genes at either: A. 6 h following chloramphenicol exposure (gene expression ≥ 2-fold difference from untreated controls with adjusted p-value < 0.05), or B. 24 h following chloramphenicol exposure (gene expression ≥ 2-fold difference from untreated controls with adjusted p-value < 0.05). In both A and B, the size of each node represents weighted gene number in each regulon (weighted by percentage of differentially expressed genes in each regulon) and the edge width represents weighted overlapped gene number between two regulons (weighted by percentage of differentially expressed genes among overlapped genes). Blue: downregulation; Red: upregulation.

Genetic analysis of different regulators in the network orchestrating the sliding response.

Strains with deletion of codY, abrB, purR, pucR, tnrA, glnR, scoC, pyrR, sigB, respectively, were analyzed in the absence (-Cm) and presence of Cm (+Cm). Pictures were taken at 24 h. Bar, 1 cm.

Comparison of major network regulators in wild type exposed to chloramphenicol and ΔcodY.

Venn diagram of prominent metabolic functions from regulon expression patterns in the Cytoscape networks for wild type exposed to chloramphenicol and ΔcodY at A. 6 h or B. 24 h.

Metabolomics analysis underscores the pattern of shifting metabolism reflected by transcriptional analysis.

Metabolic profiles of wild type at 6 h (A) and 24 h (B). Metabolites that change ≥1.5-fold (6 h) and ≥2-fold (24 h) upon exposure to 1 μM Cm in wild-type strain are listed (D1). The profiles of corresponding metabolites in ΔcodY (D2) relative to wild-type strain at 6 h and 24 h are shown here for comparison.

Spatial metabolism in the sliding population supports colony migration.

Reporter strains with luciferase operon luxABCDE fused to pdhA or pucA promoter in the wild type B. subtilis (A) and ΔcodY (B) strains were spotted on the agar plate without (-) or with Cm (+). Pictures were taken with phase contrast (top) and chemiluminescence (bottom) mode at different time points: 6 h, 24 h and 48 h. Bar, 1 cm.

pdhA is required for colony expansion on the Cm plate.

Wild type, ΔpdhA, and ΔpucM strains were spotted on the agar plate in the absence (-Cm) or presence of chloramphenicol (+Cm). Pictures were taken at 24 h. Bar, 1 cm.

Model for chloramphenicol-induced mobilization using spatiotemporal control of metabolism compared to loss of CodY regulation.

Wildtype B. subtilis growing on solid agar media develops a colony with typical size and morphology. Chloramphenicol exposure stresses translation, leading to a metabolic adaptation for colony expansion via sliding motility (mobilization). During outward growth of the population, spatiotemporal separation of different metabolic populations is visible using reporters for glycolysis and nitrogen metabolism respectively. A deletion of codY disrupts regulation of mobilization, resulting in a constitutively mobilized population. The elevation of glycolytic activity, which is required for mobilization, persists at the leading edge of the population. However, the ΔcodY strain loses control of nitrogen metabolism in the lagging population, revealing that the changes in nitrogen metabolism are not essential for mobilizing the population.

The colony of Δspo0A strain was unable to expand in the presence of chloramphenicol.

Wild type (WT) and Δspo0A B. subtilis NCIB3610 were spotted on the GYM7 plate in the absence (-Cm) and presence (+Cm) of chloramphenicol. Pictures were taken at 24 h. Bar, 1 cm.

Disruption of purine de novo biosynthesis pathway or catabolism pathway has different effects on B. subtilis growth under different conditions.

Wild type, ΔpurD, ΔpurH, and ΔpurM B. subtilis NCIB3610 were spotted on the agar plate (1.5% agar) with GYM7 (top), Msgg (middle), and CM (bottom) media in the absence (-Cm) and presence (+Cm) of chloramphenicol. Pictures were taken at 24 h. Bar, 1 cm.

The colony of ΔguaC strain was able to expand in the presence of chloramphenicol.

Wild type (WT) and ΔguaC B. subtilis NCIB3610 were spotted on the GYM7 plate in the absence (-Cm) and presence (+Cm) of chloramphenicol. Pictures were taken at 24 h. Bar, 1 cm.

A network view ΔcodY regulation in a constitutively mobile population.

Cytoscape networks of regulators with ≥ 40% engagement of regulated genes at either: A. 6 h ΔcodY gene expression ≥ 2-fold difference from wildtype controls with adjusted p-value < 0.05, or B. 24 h ΔcodY gene expression ≥ 2-fold difference from untreated controls with adjusted p-value < 0.05. In both A and B, the size of each node represents weighted gene number in each regulon (weighted by percentage of differentially expressed genes in each regulon) and the edge width represents weighted overlapped gene number between two regulons (weighted by percentage of differentially expressed genes among overlapped genes). Blue: downregulation; Red: upregulation.

Metabolomics analysis underscores the pattern of shifting metabolism reflected by transcriptional analysis.

The same data used for Figure 7 is presented here with ΔcodY as the D1 and Cm-treated as the D2 for comparison. Metabolic profiles of ΔcodY (D1) strain at 6 h (A) and 24 h (B). Metabolites that change ≥1.5-fold (6 h) and ≥2-fold (24 h) in ΔcodY compared with wild-type strain are listed. The profiles of corresponding metabolites in wild-type strain (D2) upon exposure to 1 μM Cm at 6 h and 24 h.

Supplementation of BCAAs was unable to inhibit sliding motility in the presence of chloramphenicol.

Wild type B. subtilis NCIB3610 was spotted on the GYM7 plate and GYM7 supplemented with 10 mM each of isoleucine, leucine and valine in the absence (-Cm) and presence (+Cm) of chloramphenicol. Pictures were taken at 24 h. Bar, 1 cm.