Fig 1.
T3SS activation is only detected at the edge of Y. enterocolitica microcolonies.
a) Schematic depiction of the T3SS injectisome (adapted from [52]); right, common nomenclature names of main T3SS components [53,54]. b) Confocal microscopy image of a Y. enterocolitica ΔsctW PyopE-sfGFP-ssrA microcolony section at 37°C. T3SS activity, which results in a strong upregulation of the yopE promoter and cellular fluorescence, is only detected at the edge of the microcolony. Scale bar, 50 µm, n = 3.
Fig 2.
Assembly and activity of the T3SS are significantly reduced at higher bacterial densities.
a) Top, graphical depiction of secretion assay; bottom, effectors secreted by the T3SS at increasing ODin (left to right: 0.1, 0.3, 0.7, 1.0, 1.5). Left, assignment of effectors; right, molecular weight (MW) standards. b) T3SS reporter assay (PyopE-sfGFP-SsrA) [47,56] at increasing ODin values as in a). Intensities are adjusted for OD600 and displayed relative to the highest OD-adjusted intensity value, growth curves of cultures used in these experiments are displayed in S5 Fig. c) Quantification of EGFP-SctQ foci per bacterium, corresponding to assembled injectisomes, visualized by fluorescence microscopy at increasing ODin as in b). Number of measurements and foci counts listed in S1 Table. Sample micrographs depicted at bottom, scale bar, 2 µm. d) Single-cell T3SS reporter assay (PyopE-sfGFP-SsrA) of Yersinia cells to assess the heterogeneity of T3SS activation at the indicated low and intermediate activation levels (ODin=0.1, 0.3, as indicated). Measurements were taken 4h and 5h post-secretion induction, intensities are displayed as in b). Each dot represents a single cell, black bars represent the average fluorescence intensity. n = 3 for all panels. Gel image in a) shows representative result; for c, n cells = 300; for d, n cells = 100, statistical analysis was performed using a non-parametric t-test, ****, p < 0.0001, whiskers denote standard deviation.
Table 1.
Density-dependent regulation of expression of T3SS components.
Fig 3.
The density-dependent downregulation of the T3SS is specific and reversible.
a) Volcano plot showing differences in the level of Y. enterocolitica proteins between cultures grown under secreting conditions at ODin 1.5 and 0.1. All proteins with at least three detected peptides under both conditions are displayed; T3SS-related proteins are marked by larger red dots. b) T3SS reporter assay (PyopE-sfGFP-ssrA) in strains with ODin=0.1 or 1.5 and concentration or dilution at the time of the temperature shift to 37°C, as indicated. Measurements were performed 150 min after dilution/concentration. c) T3SS reporter assay (PyopE-sfGFP-ssrA) in strains with ODin values as indicated. Medium exchange at the time of the temperature shift to 37°C as depicted; C indicates constant incubation without medium change. Measurements were performed 150 min after medium exchange. n = 3 biological replicates for all panels. In panels b) and c), n cells = 150, each spot represents a single measurement, whiskers denote standard deviation. Statistical analysis was performed using a non-parametric t-test; ****, p < 0.0001.
Fig 4.
Quorum sensing and the alternative sigma factor RpoS do not significantly contribute to the density-dependent downregulation of the T3SS.
a) Secretion assay of wild-type (WT) and quorum sensing mutant strains at different ODin as used in Fig 1 (left to right, ODin=0.1, 0.3, 0.7, 1.0, 1.5). b) Secretion assay of the indicated strains at ODin 0.1 and 1.5, as displayed. For in trans complementation of the rpoS mutant from pBAD::RpoS, RpoS expression was induced by the addition of 0.2% arabinose at the time of the shift to 37°C. n = 3, gel images show representative results.
Fig 5.
Additional expression of the transcriptional regulator VirF at higher densities restores T3SS secretion.
a) Secretion assay showing effector secretion at ODin=0.1 (left) and 1.5 (right). Lanes 3-6, increasing expression of VirF in trans (arabinose concentrations of 0.001, 0.003, 0.008, 0.03%, respectively). b) Confocal microscopy image of a Y. enterocolitica ΔsctW PyopE-sfGFP-ssrA microcolony additionally expressing VirF (induced by 0.2% arabinose) at 37°C. T3SS activity, visualized by the sfGFP signal, is detected throughout the colony. Scale bar, 50 µm; n = 3 for all panels, images display representative results.
Fig 6.
Increased abundance of the regulatory RNAs CsrBC at high bacterial density and posttranscriptional control of the virF transcript.
a) Organization of the sctG-virF operon, bold lines depict coding regions [18,69,73], red arrows indicate the position of RT-PCR primers. b) Protein levels of SctG and VirF at ODin 1.5 vs. 0.1 (displayed as log2 fold change), as measured by quantitative proteomics. c) Fluorescence microscopy of SctG-sfGFP (top) and the VirF-dependent T3SS component mCherry-SctL, both expressed from their native genetic background, at ODin = 0.1 (left) or 1.5 (right). n = 3, micrographs display representative results. d) Transcript levels of sctG and virF at ODin 1.5 vs. 0.1, as measured by reverse transcriptase quantitative PCR (see panel A). e) Schematic depiction of the influence of the CsrABC system on VirF production [26,55,74]. The regulatory RNAs CsrB and CsrC sequester CsrA, which in the closely related Y. pseudotuberculosis system (S13 Fig) prevents it from binding and stabilizing the virF transcript. f) Transcript levels of CsrB and CsrC at ODin 1.5 vs. 0.1, as measured by reverse transcriptase quantitative PCR. In b), d) and f), points indicate individual biological replicates, whiskers denote standard deviation. Statistical analysis was performed using a non-parametric t-test, **, p < 0.01.
Fig 7.
Host-cell adhesion is significantly downregulated at higher cell densities.
a) Protein levels of YadA at ODin 1.5 vs. 0.1, as measured by quantitative proteomics. Points indicate individual biological replicates, whiskers denote standard deviation. n = 3. b) Attachment of wild-type (WT) or ΔyadA bacteria to HeLa cells at the conditions indicated (ODin=0.1, 1.5) (displayed as fraction of attached bacteria). Points indicate individual biological replicates, whiskers denote standard deviation. Statistical analysis was performed using a non-parametric t-test **, p < 0.01, ns, p > 0.05.
Fig 8.
Model of the influence of local higher cell density on T3SS secretion and on host-cell attachment.
a) At lower cell densities (left), Yersinia relies on the adhesion protein YadA (depicted in grey on the bacterial surface) and effector secretion by the T3SS (depicted in red on the bacterial surface) to colonize the host by fighting the host immune system. In a microcolony, where higher cell densities are perceived (right), Yersinia specifically downregulates T3SS secretion and YadA adhesion to efficiently replicate and ultimately spread and disseminate. b) Both adhesion and secretion are regulated by the transcription factor VirF. Upon host entry, sensed as 37°C shift, the virF transcript is stabilized thanks to the global mRNA regulator CsrA. The efficient binding of CsrA to virF mRNA and resulting expression of T3SS components are ensured by the low basal levels of the two CsrA inhibitory RNAs CsrBC [26,55,74]. At higher local cell densities, increased CsrC levels sequester CsrA, which causes a higher turnover of the virF transcript, leading to a decreased expression of T3SS components and YadA. The resulting decrease in adhesion and T3SS activity promotes Yersinia replication and dissemination at later stages of infection.