icaR and icaT are Ancient Chromosome Genes Encoding Substrates of the Type III Secretion Apparatus in Shigella flexneri

Abstract

Shigella is an Escherichia coli pathovar that colonizes the cytosol of mucosal cells in the human large intestine. To do this, Shigella uses a Type III Secretion Apparatus (T3SA) to translocate several proteins into host cells. The T3SA and its substrates are encoded by genes of the virulence plasmid pINV or by chromosomal genes derived thereof. We recently discovered two chromosomal genes, which seem unrelated to pINV, although they are activated by MxiE and IpgC similarly to some of the canonical substrates of the T3SA. Here, we showed that the production of the corresponding proteins depended on the conservation of a MxiE box in their cognate promoters. Furthermore, both proteins were secreted by the T3SA in a chaperone-independent manner through the recognition of their respective amino-terminal secretion signal. Based on these observations, we named these new genes icaR and icaT, which stand for invasion chromosome antigen with homology for a transcriptional regulator and a transposase, respectively. icaR and icaT have orthologs in commensal and pathogenic E. coli strains belonging mainly to phylogroups A, B1, D and E. Finally, we demonstrated that icaR and icaT orthologs could be activated by the coproduction of IpgC and MxiE in strains MG1655 K-12 (phylogroup A) and O157:H7 ATCC 43888 (phylogroup E). In contrast, the coproduction of EivF and YgeG, which are homologs of MxiE and IpgC in the E. coli T3SS 2 (ETT2), failed to activate icaR and icaT. IMPORTANCEicaR and icaT are the latest members of the MxiE regulon discovered in the chromosome. The proteins IcaR and IcaT, albeit produced in small amounts, are nonetheless secreted by the T3SA comparably to canonical substrates. The high occurrence of icaR and icaT in phylogroups A, B1, D, and E coupled with their widespread absence in their B2 counterparts agree with the consensus E. coli phylogeny. The widespread conservation of the MxiE box among icaR and icaT orthologs supports the notion that both genes had already undergone coevolution with transcriptional activators ipgC and mxiE- harbored in pINV or a relative- in the last common ancestor of Shigella and of E. coli from phylogroups A, B1, D, and E. The possibility that icaR and icaT may contribute to Shigella pathogenesis cannot be excluded, although some of their characteristics suggest they are fossil genes.

Keywords: Escherichia coli; Shigella; phylogeny; transcription regulation; type III secretion system.

FIG 1

MxiE and IpgC are required to produce IcaR and IcaT. (A and B) Detection of IcaR and IcaT by immunoblotting with a FLAG-antibody in the total cell lysate (TCL) of the indicated S. flexneri M90T strains harboring plasmid-born icaR and icaT placed under the control of their endogenous promoters and grown in TSB at 37°C. IpaH and RecA, for which production is dependent and independent of MxiE and IpgC, respectively, were used as controls. (C) Alignment of icaR and icaT putative MxiE boxes with the consensus MxiE box. The nucleotides in bold are mutated to validate the putative MxiE boxes. (D and E) Detection of IcaR and IcaT, using the method described above, in strains harboring plasmid-borne icaR and icaT in which the indicated mutations were inserted into their MxiE box. In the experimental conditions used here the MxiE regulon is activated in ipaB4 and its derivatives, but not in the WT strain. The results are representative of three independent experiments.

FIG 2

IcaR and IcaT are secreted in a T3SA-dependent fashion. (A and B) Detection of IcaR and IcaT by immunoblotting with a FLAG-antibody in the total cell lysate (TCL) and the secreted fraction (SF) of the indicated S. flexneri 5a str. M90T strains harboring plasmid-born icaR and icaT placed under the control of the lacZ promoter and grown in TSB at 37°C. IpaC, which is a T3SA substrate, and RecA, a housekeeping protein, were used as controls. (C and D) Detection of IcaR and IcaT by immunoblotting with a FLAG-antibody in the TCL and SF, as described above, and in the presence or absence of the secretion inducer Congo red (CR). ΔmxiD is T3SA-deficient. T3SA of the WT strain can be activated by CR whereas they are constitutively active in ΔipaD. The results are representative of three independent experiments.

FIG 3

The secretion of IcaR and IcaT depended on their amino terminus and was chaperone-independent. (A and B) Detection of IcaR and IcaT N-terminal truncation mutants (0, 5, 10, or 20 residues truncation) by immunoblotting with a FLAG-antibody in the total cell lysate (TCL) and the secreted fraction (SF) of the indicated S. flexneri 5a str. M90T strains harboring plasmid-born icaR and icaT placed under the control of the lacZ promoter and grown in TSB at 37°C in the presence or absence of the secretion inducer Congo red. IpaC, which is a T3SA substrate, and RecA, a housekeeping protein, were used as controls. (C) Nitrocefin assays performed on the TCL and SF of ΔmxiD and ΔipaD strains producing IcaR (residues 1 to 20), IcaT (residues 1 to 20) or OspD1 (residues 1 to 80) N terminus and their corresponding deletion mutants fused to bla_TEM3 M182T. Student’s t tests for unpaired data with a 95% confidence interval are shown., NS, not significant; *, P ≤ 0.05; **, P ≤ 0.01; ***, P ≤ 0.001; ****, P ≤ 0.0001. (D and E) Detection of full-length IcaR and IcaT by immunoblotting with a FLAG-antibody in the TCL and SF of S. flexneri 2a str. 2457T harboring plasmid born icaR and icaT under the control of the lacZ promoter and cultivated as above. The ΔΔΔ is a triple-knockout strain devoid of the three chaperones (e.g., IpgA, IpgE, and Spa15) required for the secretion of some of the late substrates in Shigella. The results are representative of three independent experiments.

FIG 4

The MxiE box in icaR and icaT is well conserved. Alignment of the promoter, the 5’UTR, and the 5′ end of the coding sequence for (A) icaR. (B) icaT. The prefix of each sequence indicates the Shigella subgroups or the E. coli phylogroups to which each corresponding strain belongs. The region matching the MxiE box, and the beginning of the coding sequence are indicated. The consensus sequence represented in larger font at the bottom of each alignment indicated the strong conservation of the promoter and 5’ untranslated region of these genes.

FIG 5

The coproduction of MxiE and IpgC activates the expression of icaR and icaT in E. coli. Quantification of the expression, as the number of copies of transcript per microliter, of icaR, icaT, and recA by ddPCR in −/−, ipgC⁺, mxiE⁺ or ipgC⁺ mxiE⁺ strains obtained by transformation with the relevant plasmids. (A) BS176, a plasmid cured derivative of S. flexneri 5a str. M90T. (B) E. coli K-12 str. MG1655, a representative of phylogroup A. (C) E. coli O157:H7 str. ATCC 43888, a representative of phylogroup E. The mean values and standard deviations based on three biological replicates are represented. The statistical significance of these data was tested with a one-way ANOVA and a Bonferroni correction for pairwise post hoc tests with a 95% confidence interval; *, P < 0.05; **, P < 0.01; ***, P < 0.001; NS, not significant. (D to F) Detection of the production of MxiE and IpgC and of IcaR from its endogenous locus by immunoblotting the TCL of strains described in (A to C). These results are representative of three independent experiments.

FIG 6

The coproduction of EivF and YgeG does not activate the expression of icaR and icaT in E. coli. Quantification of the expression, as the number of copies of transcript per microliter, of icaR, icaT, and recA by ddPCR in −/−, ygeG⁺, eivF⁺ or ygeG⁺ eivF⁺ strains obtained by transformation with the relevant plasmids. (A) E. coli K-12 str. MG1655. (B) E. coli O157:H7 str. ATCC 43888. The mean values and standard deviations based on three biological replicates are represented. The statistical significance of these data was tested with a one-way ANOVA and a Bonferroni correction for pairwise post hoc tests with a 95% confidence interval. There were no statistically significant variations in the data within any given group. (C to D) Detection of the production of YgeG and EivF by immunoblotting the TCL of strains described in (A and B). These results are representative of three independent experiments.