Distinct type I and type II toxin-antitoxin modules control Salmonella lifestyle inside eukaryotic cells

Abstract

Toxin-antitoxin (TA) modules contribute to the generation of non-growing cells in response to stress. These modules abound in bacterial pathogens although the bases for this profusion remain largely unknown. Using the intracellular bacterial pathogen Salmonella enterica serovar Typhimurium as a model, here we show that a selected group of TA modules impact bacterial fitness inside eukaryotic cells. We characterized in this pathogen twenty-seven TA modules, including type I and type II TA modules encoding antisense RNA and proteinaceous antitoxins, respectively. Proteomic and gene expression analyses revealed that the pathogen produces numerous toxins of TA modules inside eukaryotic cells. Among these, the toxins HokST, LdrAST, and TisBST, encoded by type I TA modules and T4ST and VapC2ST, encoded by type II TA modules, promote bacterial survival inside fibroblasts. In contrast, only VapC2ST shows that positive effect in bacterial fitness when the pathogen infects epithelial cells. These results illustrate how S. Typhimurium uses distinct type I and type II TA modules to regulate its intracellular lifestyle in varied host cell types. This function specialization might explain why the number of TA modules increased in intracellular bacterial pathogens.

Figure 1. TA modules identified in S. Typhimurium strain SL1344.

(a) Workflow depicting the in silico analyses that led to the identification of 27 putative TA modules in S. Typhimurium strain SL1344. The study of Fozo et al. is focused in type I modules whereas that of Shao et al. describes the TADB web resource that compiles type II modules from different organisms; (b) distribution of the 27 TA loci identified encoding seven type I (orange) and 20 type II (red) modules in the chromosome and plasmids of strain SL1344; (c) conservation of the 27 TA loci identified in S. Typhimurium SL1344 in the genome of other Salmonella species and S. enterica serovars. Homolog search was performed using the tBLASTn tool and as query the protein sequences of toxins and antitoxins identified in S. Typhimurium SL1344. The results of this search were contrasted with recent data reported by Nuccio and Baumler, which compared genome sequences of S. enterica serovars causing gastrointestinal and extraintestinal pathologies. Abbreviations: STm-SL1344, S. Typhimurium SL1344; STm-LT2, S. Typhimurium LT2; SNew, S. Newport SL254; SHei, S. Heildeberg SL476; SSch, S. Schwarzengrund CVM19633; SAgo, S. Agona SL483; SEnt, S. Enteritidis P125109; SPB, S. Paratyphi B SP87; STy1; S. Typhi CT18; STy2; S. Typhi Ty2; SPA1, S. Paratyphi A ATCC9150; SPA2, S. Paratyphi A AKU_12601; SDub, S. Dublin CT_02021853; SGal, S. Gallinarum 287/91; SCho, S. Choleraesuis SC-B67; SPC, S. Paratyphi C RKS4594; S. bongori-1, strain N268-08; S. bongori-2, strain NCTC12419. Colour code: green, both T and A homologs identified; blue, only T homolog identified; yellow: only A homolog identified; black, no homolog identified. Only hits with e-values ≤ 10⁻⁵ and pairing encompassing a minimum of 30 amino acids were considered significant.

Figure 2. Most toxins and antitoxins predicted in S. Typhimurium SL1344 are encoded by bona fide functional TA modules.

Toxin and antitoxin genes were cloned under P_lac and P_BAD promoters in compatible plasmids. These genes were expressed either independently or combined in the natural host, i.e. S. Typhimurium strain SL1344, by plating bacteria onto plates containing IPTG and/or arabinose. Those TA modules highlighted in either red (type II) or orange (type I) boxes correspond to bone fide modules. (a) type II modules; (b) type I modules. Assays were repeated in a minimum of three independent experiments. ‘None' means no inducer (IPTG or arabinose) added.

Figure 3. S. Typhimurium up-regulates inside fibroblasts bona fide toxins encoded by type I and type II TA modules.

S. Typhimurium was chromosomally tagged with 3xFLAG epitope in genes encoding toxins of type II TA modules detected by proteomics (see Table 2). Protein extracts were prepared from intracellular bacteria at 24 h post-infection of BJ5ta human fibroblasts and extracellular bacteria grown to stationary phase in LB medium. (a) toxins of type II modules detected in extracellular bacteria. Note the distinct relative levels of the toxins examined; (b) synthesis of toxins of type II modules in intracellular bacteria. Note that the four toxins shown (T2_ST, T4_ST, T5_ST, and VapC2_ST) are produced at higher relative levels inside the eukaryotic cells. The inner membrane IgaA was used as loading control. Numbers below the toxin bands correspond to the relative values determined by densitometry, referred to values measured in extracellular bacteria and corrected by those obtained for the loading control, the IgaA protein; (c) quantitative RT-PCR (qRT-PCR) assays showing the relative expression of five genes encoding toxins of type II modules. Note that three of them displayed significant induction in intracellular bacteria. Data are the means and standard deviations from three independent experiments. *, P = 0.01 to 0.05; **, P = 0.01 to 0.001, by one-way ANOVA with Dunnett's post-test.

Figure 4. Up-regulation of toxins encoded by type II modules by S. Typhimurium is restricted to VapC2_ST when growing inside HeLa epithelial cells.

3xFLAG-tagged S. Typhimurium strains were used to infect human HeLa epithelial cells for 16 h. At this time, protein extracts were prepared from intracellular bacteria and tested for toxin protein by Western blotting assay. Numbers below the toxin bands correspond to the relative values determined by densitometry, referred to values measured in extracellular bacteria and corrected by those obtained for the loading control, the IgaA protein. Note that among the four toxins tested only VapC2_ST is produced by bacteria at higher relative levels inside the eukaryotic cells. The Western blots shown are representative of a total of three independent experiments.

Figure 5. Distinct type I and type II TA modules impact fitness of intracellular S. Typhimurium in fibroblasts and epithelial cells.

Shown are the survival rates of S. Typhimurium mutants lacking the indicated type I (grey bars) and type II (white bars) TA modules. (a) ratio of viable intracellular bacteria at 24 h and 2 h post-infection in human BJ5ta fibroblasts; (b) ratio of viable intracellular bacteria at 16 h and 2 h post-infection in human HeLa epithelial cells. “Δ5” refers to the Δhok-sok_ST ΔtisB-istR_ST ΔldrA-ldrA_ST Δta4_ST ΔvapBC2_ST mutant. Data are the means and standard deviations from three independent experiments. *, P = 0.01 to 0.05; **, P = 0.01 to 0.001; ***, P = < 0.001, by one-way ANOVA with Dunnett's post-test.