Identification of RNases implicated in type III secretion of Y. pseudotuberculosis

In our attempt to identify RNases that influence virulence traits of Y. pseudotuberculosis, in particular the expression of the Ysc-T3SS/Yop machinery, we constructed in-frame deletions in genes of identified putative RNases (rnb, RNase II; rnc, RNase III, rnd, RNase D; rng, RNase G; rnhA, RNase HI; rph, RNase PH; yoaB, L-PSP(YoaB); tcdF, TdcF; tcdF-1, TdcF-1; tcdF-2, TdcF-2; and pnp PNPase), except RNase E as its loss is lethal [20]. The chosen RNases differ in their target specificity, cleavage mechanism, and role in bacterial physiology [7,10], and previous global expression analysis by RNA-seq showed that these RNases are all controlled in response to virulence-relevant parameters such as temperature and/or nutrient availability [36,37]. As an uncontrolled/induced expression of the Ysc-T3SS/Yop machinery was previously shown to result in a growth reduction [38–40], we first characterized the growth behavior of the RNase mutants. Most RNase mutants did not influence bacterial growth (S1 Fig). Only the PNPase and RNase III-deficient strains (YP139 (Δpnp), YP356 (Δrnc)) exhibited a growth delay at 25°C and 37°C, and complementation by the respective RNase gene could restore the phenotype of the wildtype (Fig 1A). In contrast to the Δpnp mutant, the growth defect of the rnc mutant was much stronger at 37°C (Fig 1A). A similar growth arrest has been observed upon a constitutive expression of the Ysc-T3SS of Yersinia [38–40]. We therefore also tested growth at 37°C under Ca²⁺-limiting conditions, as a substitute for host cell contact, to induce Ysc-T3SS/Yop expression and Yop secretion. In agreement with previous results [39, 40, 38], induction of Yop secretion is accompanied by a strong growth reduction of the wildtype which was more severe in both RNase-deficient mutants (Fig 1A).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Loss of RNases PNPase and RNase III affects growth and the formation of T3S-like structures of Y. pseudotuberculosis.

(A) Overnight cultures of Y. pseudotuberculosis YPIII and the isogenic RNase mutants YP139 (Δpnp) and YP356 (Δrnc) harboring the empty vector (pV) or the complementation plasmids (prnc⁺ or ppnp⁺) were diluted to an OD₆₀₀ of 0.2 in LB and growth at 25°C, 37°C, and 37°C/-Ca²⁺ was followed by measurement of OD₆₀₀. Data represent the mean ± SD from experiments done in triplicates. (B) Scanning (left panel) and transmission (right panel) microscopy of Y. pseudotuberculosis strains YPIII (wt), YP139 (Δpnp), and YP356 (Δrnc) grown at 25°C, 37°C and 37°C/-Ca²⁺. Flagella are indicated by grey and T3S injectisome-like structures by white arrows, respectively. Bars indicate 200 nm.

https://doi.org/10.1371/journal.ppat.1011965.g001

This prompted us to select the Δpnp and Δrnc mutant for further analysis by electron microscopy (Fig 1B). In contrast to the wildtype and the Δpnp mutant, additional surface-exposed structures resembling needles of T3S injectisomes [41] were detected on the Δrnc mutant at 37°C with scanning and negative staining electron microscopy, which are also seen with the wildtype under secretion conditions (37°C/-Ca²⁺) (Fig 1B). This indicated that the production of the Ysc-T3SS/Yop machinery is induced at 37°C without Ca²⁺ depletion in the Δrnc mutant, conditions under which the production and assembly of the apparatus are normally repressed [34,42].

Analysis of the influence of RNase III and PNPase on Yop secretion and expression

To test and characterize the influence of the PNPase and RNase III on the Ysc-T3SS/Yop machinery, Y. pseudotuberculosis wildtype strain YPIII and the isogenic Δpnp and Δrnc mutants were cultivated at 25°C (i.e. non-expression conditions), 37°C (i.e. T3SS/Yop-non-secretion conditions) and at 37°C/Ca²⁺-limitation which mimics host cell contact and triggers expression and activity of the Ysc-T3SS/Yop machinery (i.e. T3SS/Yop-secretion conditions) [34,42]. A mutant strain deficient in the T3SS component YscS was used as a negative control as the encoded protein is essential for Yop secretion [43]. Yop secretion by the wildtype was only detectable at 37°C/-Ca²⁺ (T3SS-secretion conditions), but not at 25°C and 37°C (Fig 2A). In contrast, slightly higher amounts of Yops were already secreted by the Δpnp mutant under non-secretion conditions (37°C), and a significantly higher amount of secreted Yop proteins was detectable under identical conditions in the absence of RNase III (Fig 2A). This did not result from bacterial cell lysis (S2A Fig) and the derepressed effect of the Δrnc mutant could be complemented by an rnc-positive plasmid (S2B Fig). The observed Yop secretion pattern is in full agreement with the detection of numerous T3S injectisome-like structures on the surfaces of the bacteria (Fig 1B). This might also explain the growth reduction observed at 37°C for the Δrnc mutant (Fig 1A), the phenotype which is characteristic for Yop secretion conditions (37°C/-Ca²⁺) [38–40].

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Influence of PNPase and RNase III on Yop protein secretion.

(A) Y. pseudotuberculosis strains YPIII (wt), YP139 (Δpnp), and YP356 (Δrnc) were grown at 25°C, 37°C, and 37°C/-Ca²⁺; the secreted proteins in the supernatant of the cultures were precipitated with TCA and separated on SDS gels (left panels). The Yop secretion-deficient mutant YP101 (ΔyscS) was used as negative control. Secreted Yop proteins were quantified using ImageJ. Data represent the mean ± SD from three independent biological replicates relative to the amounts of Yops secreted by the wildtype at 25°C (right panel). Significant differences were determined using the one-way Anova test and are indicated by asterisks (*P <0.05, **P = 0.01). c: indicates an unidentified secreted protein used to control the efficiency of protein precipitation (B) Secretion of a plasmid-encoded YopE-beta-lactamase (BlaM) fusion protein (illustrated in the upper panel) expressed in Y. pseudotuberculosis strains YPIII (wt), and YP356 (Δrnc) grown at 37°C and 37°C/-Ca²⁺ for 4 h was determined. Data represent the mean ± SD from three independent biological replicates (lower panel). Significant differences were determined using the Student’s t-test and are indicated by asterisks (****P<0.0001).

https://doi.org/10.1371/journal.ppat.1011965.g002

To confirm the influence of RNase III on the secretion of a specific Yop protein, we quantified the secretion of an introduced, plasmid-encoded YopE-β-lactamase fusion protein (YopE-BlaM) by nitrocefin assays after a shift from 25°C to 37°C +/- Ca²⁺ (Fig 2B). In agreement with our previous results, secretion of the YopE-BlaM fusion protein was significantly increased in the Δrnc mutant at 37°C, but not in the wildtype, whereas similar high amounts of secreted YopE-BlaM were detectable with both wildtype and Δrnc mutant under secretion conditions (37°C/- Ca²⁺).

To test, whether the synthesis of the Ysc-T3SS components and/or the Yops, or solely the secretion of Yop proteins is affected, Yop protein levels in whole cell extracts of Y. pseudotuberculosis YPIII (wt), and the Δpnp and Δrnc RNase mutant strains grown under Yop secretion (37°C/-Ca²⁺) or non-secretion conditions (37°C) were analyzed by Western blotting using an anti-all Yop antibody (Fig 3). We found that higher amounts of the Yops are synthesized in the Δpnp mutant at 37°C/non-secretion conditions. This is in agreement with a previous study, reporting that certain T3SS and yop genes and equivalent proteins (e.g. YpkA, YopD, YscC) were about 1.5–2.3 fold higher expressed in the absence of PNPase [18]. The overall increase of the amounts of secreted Yops under 37°C/non-secretion conditions is even higher and also significant in the Δrnc mutant when compared to wildtype (Fig 3B). This supported the secretion data (Fig 2), demonstrating that RNAse III has a stronger influence on the T3SS/yop machinery than PNPase. The overall Yop protein levels of both mutants are almost identical under secretion and non-secretion conditions and comparable to the Yop levels of the wildtype under secretion conditions (37°C/-Ca²⁺); except for YopE. This effector seems to be considerably reduced in the absence of RNase III compared to the wildtype (Fig 3B and 3C). Notably, also the synthesis of the YopE-BlaM reporter is strongly increased at 37°C in the Δrnc strain based on the β-lactamase activity (Fig 2B), similar to the synthesis of other Yops (Fig 3C). In contrast, the YopE protein was barely detectable in the same strain (Fig 3C). This indicated that not yopE expression but rather post-transcriptional steps of YopE synthesis/stability are affected in the absence of RNase III.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. Yop synthesis is altered in PNPase- and RNase III-deficient strains.

Y. pseudotuberculosis strains YPIII (wt), YP139 (Δpnp), and YP356 (Δrnc) without (A,B) or with a plasmid-encoded YopE-beta-lactamase (BlaM) fusion protein (C) were grown at 37°C (+) and 37°C/-Ca²⁺ (-); whole cell extracts were prepared and separated on SDS gels (left panels). Synthesized Yop proteins in the YPIII (wt) (A-C), YP139 (Δpnp) (A), YP356 (Δrnc) (B), and (C) YP356 (Δrnc) pIVO13 (yopE-blaM) were detected by Western blotting using a multi-Yop antiserum (left panel) and were quantified by ImageJ (right panel). An antiserum against H-NS was used for loading control. Data represent the mean ± SD from three independent biological replicates and relative amounts are documented with respect to the wildtype grown at 37°C. Significant differences were determined using Student’s t-test and indicated by asterisks (*P <0.05; ****P<0.0001).

https://doi.org/10.1371/journal.ppat.1011965.g003

To elucidate how the loss of PNPase and RNase III triggers the synthesis of the Ysc-T3SS/Yop components under non-inducing conditions, we next addressed the impact of the pnp and rnc knock-out mutations on known Ysc-T3SS/Yop control mechanisms and regulators.

PNPase and RNase III influence on Ysc-T3SS/Yops does not occur through an increase of the pYV copy number

Synthesis of the Ysc-T3SS/Yop injectisome components is controlled on different levels, including the regulation of the copy number of the virulence plasmid pYV (also named pIB1) in response to temperature and host cell contact/secretion [44]. As the copy number of plasmids is subjected to complex regulation involving RNases [45], we employed qPCR to determine whether a knockout of the rnc or the pnp gene influences the pYV copy number. However, similar to the wildtype, we found only a subtle increase of the copy number from 1 to 2 copies in the mutants upon a shift from 25°C to 37°C, while an increase to 3–4 copies of pYV was observed under T3SS/Yop-secretion conditions (37°C/-Ca²⁺) (S3 Fig). Accordingly, increased Yop expression and synthesis at 37°C (Fig 3) cannot be explained by an increase in the pYV copy number and must be based on the deregulation of other control factors.

RNase III and PNPase mainly influence Yop protein expression through the control of LcrF

Expression of the yop and the Ysc-T3S component genes is mainly activated by the AraC-type transcriptional activator LcrF in response to temperature and host cell contact [19,31,33]. Therefore, we tested whether the absence of RNase III and PNPase influenced LcrF synthesis. We found that the amount of the LcrF protein was significantly increased at 37°C/non-inducing conditions in both RNase mutant strains (Fig 4A–4C). The overall influence of the loss of RNase III was more pronounced compared to the loss of PNPase (Fig 4A and 4B), and could be fully complemented by the rnc-encoding plasmid (Fig 4C). Interestingly, high LcrF levels in the Δrnc strain at 37°C (no secretion) and 37°C/-Ca²⁺ (secretion) were comparable, demonstrating that LcrF synthesis is independent of the secretion control process in the absence of RNase III (Fig 4B and 4C).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 4. Influence of RNase III and PNPase on the synthesis of LcrF.

Y. pseudotuberculosis strains YPIII (wt), YP139 (Δpnp) (A), and YP356 (Δrnc) (B) were grown at 37°C (+) and 37°C/-Ca²⁺ (-). The secretion-deficient mutant YP101 (ΔyscS) and the LcrF-deficient mutant strain YP179 (ΔlcrF) were used as a negative control. Whole-cell extracts were prepared and separated by SDS-PAGE. The transcriptional activator LcrF was detected by Western blotting using an LcrF-specific polyclonal antiserum. An antiserum against H-NS was used for loading control (left panel). Data were quantified by ImageJ and represent the mean ± SD from three independent biological replicates (right panel). Significant differences were determined using Student’s t-test and indicated by asterisks (*P <0.05; ***P<0.001). (C) Y. pseudotuberculosis strains YPIII (wt), YP139 (Δrnc), and YP356 (Δrnc) pIVO20 (prnc⁺) were grown at 37°C (+) and 37°C/-Ca²⁺ (-); the LcrF-deficient mutant strain YP179 was used as a negative control. Whole-cell extracts were prepared and separated on SDS gels. Synthesized LcrF was detected by Western blotting using a polyclonal LcrF antiserum. An antiserum against RNA polymerase (RNAP) was used for loading control.

https://doi.org/10.1371/journal.ppat.1011965.g004

RNase III and PNPase influence lcrF transcript levels

To unravel the molecular mechanism by which the RNases control LcrF levels, we first investigated the lcrF transcript level in the Δrnc and Δpnp mutant compared to wildtype at 37°C. In agreement with previous studies [19,33], we observed that the lcrF transcript is generally very unstable in the wildtype as visualized by a diffuse pattern of different-length transcripts hybridizing with the lcrF probe on Northern blots (Fig 5A). Although this general pattern did not change in the absence of both RNases a significantly higher amount of lcrF transcripts was observed in the absence of RNase III at 37°C (Fig 5A), and this effect could be fully complemented by a plasmid-encoded copy of the rnc gene (Fig 5B). Loss of PNPase caused a smaller upregulation of lcrF mRNAs under these conditions, whereas no influence of the RNases was observed under secretion conditions at 37°C/-Ca²⁺ (Fig 5A).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 5. Influence of RNase III and PNPase on lcrF transcript levels.

(A) Y. pseudotuberculosis strains YPIII (wt), YP139 (Δpnp), YP356 (Δrnc), and YP375 (Δrnc, Δpnp) were grown at 37°C and 37°C/-Ca²⁺. Total RNA of the samples was prepared and the lcrF transcript was detected by Northern blotting; 16S rRNA was used as loading control. (B) Y. pseudotuberculosis strains YPIII (wt), YP139 (Δpnp), YP356 (Δrnc), and YP356 (Δrnc) pIVO20 (prnc⁺) and the ΔlcrF strain (YP179) were grown at 37°C. Total RNA of the samples was prepared and the lcrF transcript was detected by Northern blotting (left panel) and quantified by ImageJ (right panel). Data represent the mean ± SD from three independent biological replicates. Significant differences were determined using Student’s t-test and indicated by asterisks (*P<0.05). (C) RNA-Seq coverage tracks of the yscW-lcrF locus in Y. pseudotuberculosis YPIII (wt) and YP356 (Δrnc) grown at 37°C. RNA-Seq reads were mapped to the pYV virulence plasmid (NC_006153.2) and analyzed using a DESeq2 pipeline. The mapping visualizes the nucleotide score of the combined biological replicates for the two strains with an equal score for both. (D) RNA-Seq coverage tracks of the yscW-lcrF locus of the YPIII (wt) and YP356 (Δrnc) were adjusted to nucleotide scores of 900 and 150, respectively to compare the read pattern of both strains.

https://doi.org/10.1371/journal.ppat.1011965.g005

In a parallel experiment, we also performed a comparative transcription profile analysis of the wildtype and the Δrnc mutant grown at 37°C by RNA-sequencing (for more detail see also below). In agreement with the results of the Northern blots, we could detect a massive increase of reads mapping to lcrF in the RNA-seq data set (Fig 5C and S4 Dataset), whereby the overall read patterns covering the lcrF gene were comparable (see enlargement of the wildtype transcript profile in Fig 5D).

Upregulation of lcrF transcript levels in the Δrnc and Δpnp mutant could have occurred by the direct influence of the RNases on the stability of the lcrF mRNA or by an indirect effect on the activation of lcrF transcription. We first tested whether the overall increase of lcrF transcript levels in the absence of RNase III and PNPase is indirect and caused by an increase of lcrF transcription. For this purpose, we first used two reporter constructs, which both harbor the promoter region of the yscW-lcrF operon from -311 to +8 fused to phoA (transcriptional fusion, pMV53) and from -304 to +281 (including the entire 5’ untranslated region) relative to the transcriptional start site fused to lacZ (translational fusion, pKB35) to compare the activity of the promoter at 37°C. However, no increase in promoter P_yscW-lcrF activity was detectable with both reporters in the Δrnc and the Δpnp mutant, respectively (S4 Fig). This demonstrated that an influence of the RNases on the initiation of lcrF transcription can be excluded.

To investigate whether the absence of RNase III and PNPase affects the stability of the lcrF transcript, we determined the effects of RNase III on lcrF mRNA decay and measured the half-life of the transcript in the wildtype, the Δrnc, Δpnp, and Δpnp/Δrnc mutant by Northern blot analysis (Fig 6A and 6B). The half-life of the lcrF mRNA in the Δrnc mutant was two-fold higher than in the wildtype. Loss of PNPase also increased lcrF stability although to a lesser extent and not significantly, and the effect of both RNases on lcrF transcript stabilization seems to be additive (Fig 6B).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 6. Influence of RNase III and PNPase on the stability of the lcrF transcript.

Y. pseudotuberculosis strains YPIII (wt), YP356 (Δrnc), YP139 (Δpnp), and YP375 (Δrnc Δpnp) were grown at 37°C. Rifampicin was added to the cultures to stop transcription. Samples were taken before (0 min) and 1, 3 and 7.5 min after rifampicin addition. Total RNA of the samples was extracted and subjected to Northern blotting, using an lcrF-specific probe. 16S and 23S rRNA served as loading controls. A representative Northern blot for each strain is shown from three independent biological replicates. (B) The three independent biological replicates were used for the quantification of lcrF transcript levels with ImageJ software. 16S rRNA was used as a control for quantification. Relative transcript levels were plotted in Graphpad Prism 8 to determine the lcrF mRNA decay rate. Data represent the mean ± SD from three independent biological replicates. Significant differences were determined at 7.5 min using Student’s t-test and indicated by asterisks (*P <0.05).

https://doi.org/10.1371/journal.ppat.1011965.g006

Next, we investigated whether the RNases act directly on the lcrF mRNA, but according to our analysis, it is very unlikely that the lcrF transcript is a direct target of RNase III. Firstly, we did not identify potential RNase III cleavage sites (e.g. ~22 nt double-stranded RNA segments including base-pairing preferences such as 5’ GAAAnnAAAG/CUUUnnUUUG 3’) within the yscW-lcrF transcript, which were identified in direct targets in previous surveys [46–48]. Only one longer double-stranded RNA segment without a potential RNase III cleavage sequence was identified in the 5’ untranslated region (5’-UTR) of the yscW-lcrF transcript. However, its presence in the 5’-UTR of the P_yscW-lcrF-lacZ reporter fusion (pKB35) did not result in an expression change in the Δrnc mutant compared to wildtype (S4B Fig). Secondly, no smaller-size degradation products of the lcrF transcript were detectable in the mRNA pool of the wildtype compared to the Δrnc strain in gels with a high resolution (S5 Fig). Thirdly, although we detected a massive increase of reads mapping to the lcrF mRNA in our RNA-seq analysis when visualized with the same nucleotide score height, no obvious differences in the RNA read patterns covering the lcrF gene region of the wildtype and the Δrnc mutant were detectable (Fig 5C and 5D). Moreover, no typical endoribonuclease-mediated processing events generating additional sharp/trimmed read ends in the wildtype were identified (Fig 5D) as described for direct mRNA targets [49,50]. These results indicated that RNase III is unlikely to target the lcrF mRNA directly at 37°C, but rather affects post-transcriptional regulators that have an overall influence on lcrF transcript levels, such as the effector protein YopD and the carbon storage regulatory system CsrABC [19].

Mutual influence of RNase III, PNPase, and the translocon protein YopD

Besides its role as a structural component of the translocation pore of the T3SS for Yop delivery [51,52], YopD also acts as a negative post-transcriptional regulator that represses the synthesis of the master regulator LcrF and components of the Ysc-T3SS/Yop system under non-secretion conditions [19,53,54]. To test whether PNPase and RNase III affect YopD synthesis, we compared intracellular YopD levels between wildtype and RNase mutants.

The amount of YopD was strongly increased in the absence of RNase III at 37°C with a similar level observed in the wildtype at 37°C/-Ca²⁺ (Fig 7A). However, YopD upregulation does not explain the observed phenotypes of the Δrnc mutant, as the presence of YopD was shown to reduce and not increase lcrF mRNA levels, and Ysc-T3SS/Yop synthesis [19,53,54]. On the contrary, YopD seems to be upregulated in the absence of RNase III as part of the entire Ysc-T3SS/Yop secretion complex due to higher LcrF levels (Fig 3B).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 7. Influence of RNase III on the synthesis of YopD.

Y. pseudotuberculosis strains YPIII (wt), YP91 (ΔyopD), YP179 (ΔlcrF), YP139 (Δpnp), YP218 (ΔyopD/Δpnp), YP372 (ΔyopD/Δrnc), and YP375 (Δpnp/Δrnc) were grown at 37°C and 37°C/-Ca²⁺. Whole-cell extracts were prepared and separated by SDS-PAGE. Synthesized YopD (A) and LcrF (B) were detected by Western blotting using a YopD-detecting polyclonal antiserum. An antiserum against H-NS was used for loading control. (C) Y. pseudotuberculosis strains YPIII (wt), YP139 (Δpnp), YP356 (Δrnc) and YP375 (Δpnp/Δrnc) were grown at 37°C. The pnp (left) and the rnc (right) transcripts were detected by Northern blotting using specific probes. 16S RNA served as loading controls. (D) Scheme illustrating the control of LcrF and Ysc-T3SS-Yop synthesis by PNPase and RNase III.

https://doi.org/10.1371/journal.ppat.1011965.g007

Moreover, the amount of the transcriptional activator LcrF is similarly increased in the Δrnc and ΔyopD mutant compared to wildtype, but the overall LcrF level is much higher in the Δrnc/ΔyopD double compared to the single mutants (Fig 7B), indicating an additive effect. This shows that RNase III regulates ysc-T3SS/yop gene expression not via but rather together with YopD.

Loss of PNPase resulted also in the upregulation of YopD, but its influence was much less pronounced (Fig 7A). A comparison of the ΔyopD and the Δpnp/ΔyopD mutant revealed no difference in LcrF levels, whereby the overall repressive effect of YopD was generally stronger than that of PNPase (Fig 7B). Together, this suggested that PNPase could act upstream of YopD.

Surprisingly, the production of LcrF and YopD did not further increase in the Δpnp/Δrnc double mutant as expected from the single mutation. In contrast, the Δrnc/Δpnp double mutant contained less YopD and LcrF than the Δrnc single mutant (Fig 7A and 7B). This indicated an interplay of the RNases in which the loss of PNPase seems to counteract the upregulation of LcrF promoted by the loss of RNase III. It is known that RNase III cleaves within the 5’ end of the pnp transcript and initiates its rapid degradation in E. coli; as a result, PNPase mRNA levels are significantly increased in the E. coli Δrnc mutant strain [55,56]. Significantly higher amounts of the pnp transcripts were also identified in the Δrnc mutant of Y. pseudotuberculosis grown at 37°C by the RNA-seq analysis (log₂ = 2.2 (T1) 1.9 (T2); S4 Dataset) and by Northern blotting (Fig 7C), indicating a similar process in this pathogen. This suggests an interactive, regulatory pattern illustrated in Fig 7D, in which RNase III strongly represses LcrF production via a YopD-independent pathway, which is partially compensated by a second pathway controlling PNPase and YopD.

Impact of RNase III on the Csr system

As the mechanism of how RNase III controls lcrF mRNA levels independently of YopD remained unclear, we tested whether RNase III influences the amount of CsrA. CsrA is another post-transcriptional regulator known to increase the stability of the lcrF transcript in response to host cell contact/secretion [19]. CsrA belongs to the carbon storage regulator (Csr) system together with the two regulatory RNAs CsrB and CsrC which sequester multiple CsrA dimers to eliminate their function (Fig 8A, right panel). Loss of RNase III did not influence the overall amount of CsrA (Fig 8A). However, expression of the counteracting CsrB and CsrC sRNAs was significantly reduced in the absence of RNase III (Fig 8B). This, in turn, increases the amount of unbound, active CsrA, which is known to strongly reduce the degradation of the lcrF mRNA [19].

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 8. Influence RNase III on the carbon storage system components CsrA, CsrB, and CsrC.

(A) (Left) Y. pseudotuberculosis strains YPIII (wt) and YP356 (Δrnc) were grown at 37°C in the presence (+) and absence of Ca²⁺ (-). Whole-cell extracts were prepared and separated by SDS-PAGE. Synthesized CsrA protein was detected by Western blotting using a polyclonal antiserum against CsrA. An antiserum against H-NS was used for loading control. (Right) Scheme of the interaction of the carbon storage regulatory systems components: the RNA binding protein CsrA and the CsrA-sequestering regulatory RNAs CsrB and CsrC. (B) (Left) Y. pseudotuberculosis strains YPIII (wt), YP356 (Δrnc), and the complemented strains were grown at 37°C in the presence of Ca²⁺ (non-secretion conditions). Total RNA of the samples was prepared and CsrB and CsrC were detected by Northern blotting. (Right) Amounts of the CsrB and CsrC RNA were quantified by ImageJ and represent the mean ± SD from three independent biological replicates. Significant differences were determined using Student’s t-test and indicated by asterisks (**P<0.01). (C) Plasmid pKB34 encoding the yscWlcrF-lacZ translational fusion was transformed into Y. pseudotuberculosis strains YPIII (wt) pAKH85 (pV, empty vector), YP356 (Δrnc) pAKH85 (pV, empty vector) and the complementing strains YPIII (wt) pIVO20 (prnc⁺) and YP356 pIVO20 (prnc⁺). The transformants were grown at 37°C and beta-galactosidase activity was determined, respectively. Data represent the mean ± SD from three independent biological replicates. Significant differences were determined using Student’s t-test and indicated by asterisks (*P <0.05).

https://doi.org/10.1371/journal.ppat.1011965.g008

The molecular mechanism of how CsrA stabilizes the lcrF mRNA is currently unknown [19]. However, as CsrA was found to enhance translation of the lcrF transcript [19], and a higher ribosome occupancy upon upregulation of translation was shown to protect mRNAs from RNase-mediated degradation [57,58], it seemed possible that increased availability of active CsrA in the Δrnc mutant enhances translation and thus lcrF mRNA stability. Based on this assumption, we tested the influence of RNase III on the expression of a translational reporter fusion in which the lacZ gene is fused to codon 25 of the lcrF gene, harboring the intergenic RBS and the entire promoter region of the yscW-lcrF operon up to position -304 according to the transcriptional start site (Fig 8C). In contrast to the transcriptional fusions (S4 Fig), the translational reporter revealed a significant increase in the synthesis of the LcrF-LacZ fusion protein in the absence of RNase III (Fig 8C).

RNase III influence on the translational machinery

The increase of active CsrA could explain the Δrnc phenotype as the CsrB and CsrC levels are reduced to approximately 50% in the RNase III-deficient strain (Fig 8B). However, additional identified expression changes of components of the translation machinery may also have an influence. Our RNA-Seq analysis comparing the transcription profiles of the wildtype and the Δrnc mutant revealed that the transcript levels of 11 ribosomal proteins (r-proteins), as well as several translation initiation factors and rRNA/tRNA-modifying enzymes, were significantly altered with a log₂ fold change of +/- ≤1 (S1 Table). Among them are (i) the r-protein transcripts of rpsJ (S10, log₂ = 3.059) and rplC (L3, 2.784) encoded by the S10 operon involved in rRNA and r-protein regulation [59]; (ii) the translational initiation factor IF-3 (InfC) (log₂ of 1.02), an essential bona fide factor that promotes 30S initial complex formation, accurate tRNA selection, and fidelity of bacterial translation initiation [60,61], (iii) the 16S rRNA methyltransferase RsmG/GidB with the ability to enhance the loading of r-proteins and 30S ribosome assembly in the presence of the 5’-leader [62], and (iv) the 16S rRNA-processing protein RimM assisting in the assembly of the 30S subunit, and promoting translation efficiency [63,64]. The higher abundance of these translation factors in the Δrnc mutant could also increase (or assist CsrA-mediated activation of) lcrF translation and thus lcrF mRNA stability.

RNase III-mediated global reprogramming of the Y. pseudotuberculosis transcriptome in response to virulence-relevant conditions

Based on our results, we assumed that loss of RNase III has a global influence on the gene expression pattern of the Yersinia chromosome and the virulence plasmid. To obtain more information about the overall influence of RNase III, we performed transcriptomic profiling by strand-specific RNA sequencing followed by a differential expression analysis (DESeq) by comparing sequence reads from strand-specific libraries of Y. pseudotuberculosis wildtype and the Δrnc mutant strain. For this purpose, the bacteria were grown at 25°C (T0) (environmental, non-T3SS expression conditions), and then shifted to 37°C (host, T3SS non-secretion), and 37°C/-Ca²⁺ (host, T3SS secretion conditions) for 1 h (T1) and 4 h (T2). 1 h was chosen as this is the earliest time point at which an LcrF/Yop upregulation can be detected, and 4 h was chosen for maximal T3SS induction indicated by the strong expression of the activator LcrF, and synthesis of the Yops (S6 Fig). From each library, between 2–3 million cDNA reads were generated and mapped to the YPIII genome sequence (NC_010465) and the pYV virulence plasmid (NC_006153) (S1 Dataset). The global gene expression profiles of the different samples were distinct, and profiles of the three biological triplicates clustered together (S7 Fig).

The detailed analysis of the transcriptome revealed that approximately >99% of the reads of the Y. pseudotuberculosis YPIII (wt) and the Δrnc mutant grown at 25°C mapped to the chromosome (4,7 Mb, 4,250 genes). The read distribution was only slightly changed in the wildtype when the bacteria were shifted to 37°C. Under these conditions, more reads were found to map to the virulence plasmid pYV (99 genes, 66 kb) at T1 and T2. However, this still accounts for less than 3% of the total reads (Fig 9A and S2 Dataset). This changed significantly under secretion conditions (induced by the depletion of Ca²⁺). Approximately 56% (T2) of the reads were assigned to the pYV, and only 44% (T2) were mapped to the chromosome. This indicated a major shift of the transcriptome from chromosomally- to virulence plasmid-encoded genes. Most strikingly, a similar shift was observed in the Δrnc mutant grown at 37°C under non-secretion conditions (+Ca²⁺) (Fig 9A and S3 Dataset). About 47% (T2) of the entire transcripts originated from the virulence plasmid, which resembled the situation in the wildtype under secretion conditions (-Ca²⁺). This is particularly striking because the virulence plasmid represents only 1.5% of the entire Y. pseudotuberculosis genome.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 9. Global influence of RNase III on the Y. pseudotuberculosis transcriptome.

Y. pseudotuberculosis strains YPIII (wt), and YP356 (Δrnc) were grown at 25°C, and at 37°C in the presence and absence of Ca²⁺ for 1 h (T1) or 4 h (T2). Total RNA of the different strains was prepared and transcriptional profiling by strand-specific RNA sequencing followed by a differential expression analysis (DESeq) by comparing sequencing reads of wildtype and the Δrnc mutant strain was performed. (A) RNA-Seq read the distribution on the chromosome and the virulence plasmid (pYV). RNA-Seq reads which uniquely mapped to the chromosome or pYV under the different tested conditions are shown. (B) Influence of RNase III on global pathways at T1 and T2 at 37°C. Differentially expressed genes between wildtype and the Δrnc mutant with a log2FC cut-off of -2/+2 and P-value ≤ 0.05 were used for the analysis of global biological pathways based on the categories established according to the KEGG database. (C) Quantitative analysis of differentially expressed genes between wildtype and the Δrnc mutant. Bar plots show the absolute number of differentially expressed genes (left panel) after Ca²⁺ depletion after 1 h (T1) or 4 h (T2) of the wildtype (wt) or the Δrnc mutant, or (right panel) between the wildtype (wt) and the Δrnc mutant at 25°C (T0), before the shift to 37°C for 1 h (T1) or 4 h (T2) in the presence or absence of Ca²⁺.

https://doi.org/10.1371/journal.ppat.1011965.g009

To identify which global pathways are particularly affected by the absence of RNase III at 37°C under non-secretion conditions, we summarized the genes up- or downregulated (log2-fold change (log2FC) ≥ +/- 2, p-value ≤ 0.05) in groups based on their association with the pathway categories found in the KEGG Orthology database. (Fig 9B and S4 Dataset). We found that mostly pYV-encoded genes and several regulatory RNAs are upregulated in the wildtype and the Δrnc mutant 1 h (T1) after the shift from non-secretion (37°C) to secretion conditions (37°C/-Ca²⁺) (Fig 9C, left panel and S2 and S3 Datasets). This pattern changed after 4 h (T2) when an additional upregulation of chromosomally encoded transcripts mostly for metabolic enzymes and transporters was observed in the wildtype (likely to compensate for the energetic burden caused by T3SS/Yop synthesis), but not in the Δrnc strain (Fig 9C, left panel and S2 and S3 Datasets).

Our results further showed that loss of RNase III at 25°C (T0: 33 genes) was rather moderate, whereas a shift to 37°C had a significant impact on the transcripts of numerous genes compared to wildtype (37°C, T1: 194 genes, T2: 252 genes) (S4 Dataset and Fig 9C, right panel). This indicated that the influence of RNase III is strongly temperature-dependent. At 37°C, many more metabolic gene transcripts were detected in the wildtype (S4 Dataset and Fig 9C, right panel). This may explain the generally lower fitness of the RNase III-deficient strain at this growth condition (Fig 1A).

Notably, the abundance of almost all pYV-encoded Ysc-T3SS/Yop-encoding transcripts, including the transcripts of the yscW-lcrF operon was strongly increased in the Δrnc mutant at 37°C compared to wildtype at 37°C (S2 Table and S2 Dataset and Fig 9B, right panel), but it was very similar to the wildtype at 37°C/-Ca²⁺ (S2 Table and S2 Dataset and Fig 9C, right panel). Based on this analysis, we hypothesized that secretion-relevant genes are already induced under non-secretion conditions at 37°C in the absence of RNase III. To prove this assumption, we compared transcriptome changes of the wildtype upon the shift from non-secretion to secretion conditions (YPIII, 37°C/+Ca²⁺ vs. 37°C/-Ca²⁺) (S2 Dataset) with changes of the wildtype and the Δrnc mutant at 37°C under non-secretion conditions (YPIII, 37°C/+Ca²⁺ vs. Δrnc, 37°C/+Ca²⁺) (S4 Dataset). Notably, of the 28 pYV-encoded transcripts which were found in a higher abundance in the Δrnc mutant at 37°C at T2, 27 were also increased under secretion conditions in the wildtype to almost the same magnitude. This indicated that lcrF and T3SS/yop expression upon host cell contact could be triggered by a reduction of functional RNase III during the infection.

We further found that besides pYV-encoded factors, only a few other chromosomally-encoded transcripts were induced in the Δrnc mutant at 37°C at T1; among them are several translation-relevant factors, e.g. the ribosome-association factor Rai/YfiA (log2FC = 1.5), the ribosomal protein S2/RpsB (log2FC = 1.3), and the translation initiation factor IF3 (log2FC = 1). These components of the translational machinery could also promote/assist in the global reprogramming of the Yersinia transcriptome in the absence of RNase III.

RNase III-mediated influence on virulence-relevant Yop-injection and phagocytosis

Next, we analyzed whether global reprogramming in the absence of RNase III under non-secretion conditions (37°C), which includes the upregulation of all pYV-encoded secretion-relevant genes, improves the defense against the attack of phagocytic cells. To do so, we first infected activated human THP-1 macrophages with the wildtype and the Δrnc mutant harboring the plasmid-encoded YopE-β-lactamase fusion protein (YopE-BlaM). Translocation assays demonstrated that the loss of RNase III resulted in a strong increase in YopE-BlaM injection into THP-1 macrophages. As shown in Fig 10A and 10B, the majority of THP-1 cells treated with the Δrnc mutant turned blue as a measure of translocation. Upon infection with the wildtype, considerably fewer THP-1 macrophages showed a blue fluorescence, and most of them exhibited a lower blue fluorescence intensity (Fig 10A and 10B). As expected, macrophages treated with the strains harboring the control plasmid remained green fluorescent. A similar YopE-BlaM translocation pattern was observed when the translocation was analyzed in human epithelial cells (S8 Fig). This indicated that the Yop translocation efficiency is considerably increased in the absence of RNase III.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 10. Influence of RNase III on Yop translocation into macrophages and phagocytosis.

(A) Translocation of a plasmid pIVO13-encoded YopE-beta-lactamase (BlaM) fusion protein expressed in Y. pseudotuberculosis strains YPIII (wt), and YP356 (Δrnc) into human THP-1 macrophages was determined by a green to blue fluorescence shift. Bacteria transformed with the empty plasmid (pV) were used as negative control. White bar: 50 μm. (B) Relative translocation of the YopE-BlaM construct determined by the ratio of blue to green fluorescent cells is illustrated. Data represent the mean ± SD from three independent biological replicates relative to the YopE-BlaM translocation determined for the wildtype defined as 1.0. Significant differences were determined using the Student’s t-test and are indicated by asterisks (*** P<0.001). (C) Y. pseudotuberculosis strains YPIII (wt) and YP356 (Δrnc) were added to human THP-1 macrophages in an MOI of 30, and after 1 h intracellular amounts of the bacteria were determined as described in Material and Methods. Data represent the average ± SD from three independent biological replicates relative to the phagocytosis of the wildtype defined as 1.0. Significant differences were determined using the Student’s t-test and are indicated by asterisks (**** P<0.0001).

https://doi.org/10.1371/journal.ppat.1011965.g010

As Yop injection has repeatedly been shown to prevent the phagocytosis of Yersinia [28,29], we tested whether the loss of RNase III affects the phagocytic activity of macrophages. For this, wildtype and Δrnc mutant bacteria were grown at 37°C and added to activated THP-1 macrophages, and the number of intracellular bacteria was quantified. We found that the number of phagocytosed bacteria was significantly reduced with the Δrnc mutant compared to wildtype (Fig 10C). This strongly suggested that global reprogramming initiated by a loss/downregulation of RNase III is accompanied by reduced growth, but a higher capacity to successfully defend against the attack of macrophages.

Cell culture, media, and growth conditions

Overnight cultures of E. coli were routinely grown in LB at 37°C, Yersinia strains were grown at 25°C or 37°C in LB (Luria Bertani) broth. For the analysis of Yop and LcrF synthesis and secretion under secretion and non-secretion conditions, overnight cultures of the Yersinia strains grown at 25°C were diluted in LB to an OD₆₀₀ of 0.2, grown for 2 h at 25°C and then either further grown at 25°C, or shifted to 37°C in the presence or absence of the Ca²⁺ chelators 20 mM NaOx and 20 mM MgCl₂ (37°C/-Ca²⁺). If necessary, antibiotics were added at the following concentrations: carbenicillin 100 μg ml^-1, chloramphenicol 30 μg ml^-1, kanamycin 50 μg ml^-1, rifampicin 1 mg/ml and Triclosan 20 μg ml^-1. Human macrophages (THP-1) and human epithelial cells (HEp-2) were cultured in RPMI 1640 medium (sigma) containing 10% (vol/vol) fetal calf serum (FCS) and Eagle’s Minimum Essential Medium (EMEM) media + 10% FCS, respectively. All cultures were incubated at 37°C with 5% CO₂.

Strain and plasmid constructions

All DNA manipulations, restriction digestions, ligations, and transformations were performed using standard genetic and molecular techniques [86,87]. Plasmid DNA was isolated using QIAprep Spin Miniprep Kit (Qiagen) or Nucleospin Plasmid Kit (Macherey-Nagel). DNA-modifying enzymes and restriction enzymes were purchased from New England Biolabs. The oligonucleotides used for amplification by PCR, and sequencing were purchased from Metabion and Eurofins. PCRs were done in a 50 μl mix for 30 cycles using Phusion High-Fidelity DNA polymerase (New England Biolabs). Purification of PCR products was routinely performed using the QIAquick PCR Purification Kit (Qiagen) or the NucleoSpin Gel and PCR Clean-up Kit (Macherey-Nagel). All constructed plasmids were sequenced by Microsynth AG (Balgach, Switzerland).

Strains and plasmids used in this study are listed in S3 Table and primers for plasmid generation are listed in S4 Table. Plasmid pIVO13 was constructed by the insertion of a PCR fragment harboring the regulatory region of the encoding region of yopE. The fragment was amplified from total DNA of Y. pseudotuberculosis strain YPIII using primer set VIII193/VIII194 cloned into the BamHI/XhoI sites of plasmid pPCH1, replacing the full-length coding sequence of yopE. The resulting plasmid pIVO13 was sequenced using primers I984 and V824. Plasmids encoding the rnc and pnp genes for complementation analyses were constructed by amplification of DNA segments including the respective native promoter identified by Nuss et al. [37], the coding sequence and the 3’-UTR of the RNase gene using primer pairs VIII365/VIII387 for rnc and VIII636/VIII/637 for pnp (S4 Table). The generated PCR fragments were cloned into plasmid pAKH85 generating pIVO20 and pIVO21, and their sequence was confirmed using primer VIII385 and VIII388. Plasmid pTT15 was constructed by ligation of the tet promoter and the MCS with primers VI527 and II143 into the BamHI and NotI sites of pFU86. The yscW-lcrF promoter region was amplified with primers IX729 and IX730 and ligated into the BamHI and KpnI sites of pTT15, generating pMV53.

Construction of the Y. pseudotuberculosis deletion mutants

The Y. pseudotuberculosis strains used in this study are derived from wild-type strain YPIII. All deletion mutants were generated via homologous recombination. The mutant strains YP356, YP372, and YP375 in which the RNase III gene rnc was deleted were constructed in a three-step PCR. First, the Yersinia genomic DNA was used as a template to amplify 500-bp regions flanking the rnc gene. The upstream fragment was amplified with primer pairs (VII882/VII881) of which the reverse primer contained an additional 20 nt at the 5´-end which were homologous to the upstream region of the rnc gene. The downstream fragment was amplified with primer pairs (VII879/VII880) of which the forward primer contained additional 20 nt at the 5´-end which were homologous to the downstream region of rnc. Subsequently, a PCR reaction was performed with the forward primer of the upstream fragment and the reverse primer of the downstream fragment using the upstream and downstream PCR products as templates. The PCR fragment was digested with SacI and was ligated into the SacI site of the suicide plasmid pAKH3 generating plasmid pIVO11. The resulting plasmid was sequenced using primers III981/III982. Next, the plasmid was transformed into S17-1λpir and was transferred into Y. pseudotuberculosis YPIII, YP91 (ΔyopD), and YP139 (Δpnp) via conjugation. Chromosomal integration of the plasmid was selected by plating on LB supplemented with kanamycin. Single colonies were tested for the correct integration of the plasmid via colony PCR using primer pair VII366/VII367. Mutants were subsequently grown on LB agar plates containing 10% sucrose. Single colonies were tested for growth on carbenicillin and sequenced using primer pair VII366/VII367.

RNA isolation and Northern blotting

Bacterial cultures were grown under the desired conditions. The bacteria were pelleted by centrifugation for 1 min at 16000 x g. The pellets were resuspended in 0.2 volume parts of stop solution (5% water-saturated phenol, 95% ethanol) and immediately snap-frozen in liquid nitrogen. After thawing on ice, bacteria were pelleted for 1 min at 16,000 x g at 4°C. The pellet was resuspended in lysozyme solution (50 mg lysozyme/ml TE-buffer) and incubated for at least 5 min at RT. The total RNA of lysed bacteria was isolated with the ’SV Total RNA Isolation System’ (Promega, USA) according to the manufacturer’s instructions. RNA concentration and quality were determined by measurement of A₂₆₀ and A₂₈₀.

For the analysis of lcrF transcripts total cellular RNA (5–30 μg) was separated on MOPS agarose gels (1.2%) or 8% urea acrylamide (RNA <250 nt), and transferred in 10 x SSC buffer (1.5 M NaCl, 0.15 M sodium citrate pH 7.0) for 1.5 h at a pressure of 5 cm Hg to a positively charged nylon membrane by vacuum blotting or by semi-dry blotting (250 mA, 2.5 h) and UV cross-linked. Prehybridization, hybridization to DIG-labeled lcrF PCR probes, and membrane washing were conducted using the DIG luminescent Detection kit (Roche) according to the manufacturer’s instructions. The lcrF transcripts of interest were detected with a DIG-labeled PCR fragment (DIG-PCR nucleotide mix, Roche) amplified with primer pairs for the individually detected according to the manufacturer’s instructions. (S4 Table). For the detection of the CsrB and CsrC RNAs and the 5S and 16S control rRNA, fluorophore-coupled DNA oligonucleotide primers purchased from Integrated DNA technologies were used (S4 Table). The primers were incubated with the membrane in a final concentration of 15 nM for CsrB and CsrC and 5 nM for 5 S rRNA in church moderate buffer (10 mg/ml BSA, 500 nM Na₂HPO₄, 0.34% (w/v) H₃PO₄, 15% (v/v) formamide, 1 mM EDTA, 7% SDS, adjusted to pH 7.2) for 6 h. The membrane was washed with 2 x SSC and 0,2% SDS for 5 min and the RNAs were detected at 700 nm (CsrB, CsrC) or 800 nm (5 S rRNA) with the Odyssey Fc Imaging system (Li-COR).

RNA isolation for RNA-seq

To yield intact high-quality RNA for RNA-seq analyses, hot phenol RNA extraction was performed. The bacteria were harvested by centrifugation at 4,000 rpm and the bacterial pellets were snap-frozen in liquid nitrogen. Bacterial pellets were resuspended in 1/8 volume of RNA resuspension buffer (0.3 M saccharose, 0.01 M NaOAc) on ice. An equal volume of RNA lysis buffer (2% SDS, 0.01 M NaOAc pH 4.5) was added, mixed, and incubated at 65°C for 90 sec. An equal volume of pre-heated (65°C) saturated aqua-phenol was added, mixed with the lysate, incubated for 3 min at 65°C, and then immediately frozen in liquid nitrogen. Subsequently, the samples were centrifuged at 16,000 x g for 10 min at room temperature and the aqueous layer was extracted and the procedure was repeated three times. For the final extraction of phenol, an equal volume of chloroform/isoamyl alcohol (24:1) was added, mixed, and centrifuged for 10 min at 16,000 x g at room temperature. The RNA in the aqueous layer was precipitated by adding 1/10 volume of 3 M sodium acetate (pH 4.5) and 2.5 volumes of ice-cold ethanol at -20°C. The RNA was recovered by centrifugation at 16,000 x g for 30 min at 4°C, the pellets were washed at 4°C with ice-cold 70% (v/v) ethanol two times and the RNA pellets were air-dried and resuspended in nuclease-free water.

RNA stability assay

Y. pseudotuberculosis strains YPIII, YP139 (Δpnp), YP356 (Δrnc), and YP375 (Δpnp, Δrnc) were grown at 37°C. Subsequently, rifampicin was added in a final concentration of 1 mg/ml to block transcription, and samples were taken in intervals before (0), and 1, 3, and 7.5 min after the addition of rifampicin. To finally analyze the decay rate of the RNA transcripts, Northern blots were performed with an lcrF-specific probe as described [19], and the detected relative amount of RNA in each sample was determined using the Odyssey Fc Imaging system and software (LiCor).

Quantitative PCR (qPCR) analysis of the copy number of pYV

The copy number of the Y. pseudotuberculosis virulence plasmid pYV was determined using quantitative PCR (qPCR). Total DNA of YPIII or its derivatives, grown at the indicated conditions, was isolated by the addition of chloroform:isoamylalcohol (24:1). The sample was mixed, centrifuged at 16,000 x g and the supernatant was mixed with four volumes of 100% ethanol. The precipitated DNA was pelleted by centrifugation (16,000 x g) and the pellet was resuspended in nuclease-free water. For the qPCR, three primer pairs for chromosomal genes (glnA (VII922/VII923), rpoB (VII924/VII925), YPK_3178 (VII926/VII927) and two primer pairs for plasmid genes (yscM (VII918/VII919), repA (VII920/VII921)) were chosen adapted from Wang et al. [44]. 1 ng/μl of total DNA was amplified using the primer sets in SensiFAST SYBR No-ROX (Bioline). qPCR was realized in technical duplicates in the Rotor-Gene Q real-time PCR cycler (Qiagen). Two reactions with nuclease-free water without DNA were used as controls. A melting curve was employed to ensure specificity. The plasmid copy number of the strains was calculated as described by Wang et al. [44].

RNA sequencing (RNA-seq)

Y. pseudotuberculosis YPIII and the isogenic rnc mutant YP356 were grown in LB medium to exponential phase (OD₆₀₀ 0.5). The cultures were split and grown at 25°C (T0), 37°C, or 37°C in the presence of 20 mM Na₂C₂O₄ and 20 mM MgCl₂ in triplicates. After 1 h (T1) and 4 h (T2), total bacterial RNA was isolated by a hot phenol extraction protocol (see above), DNA was digested using the TURBO DNase (Ambion), purified with phenol:chlorophorm:isopropanol, and the quality was assessed using the Agilent RNA 6000 Nano Kit on the Agilent 2100 Bioanalyzer (Agilent Technologies) or Qubit (Thermo Scientific). Samples with an RIN of >8 were considered for RNA-Seq sample preparation. From the total RNA the rRNA was depleted using Ribo-Zero (Illumina). ERCC spike-ins (external RNA controls consortia Mix 1 and Mix 2) were added before rRNA depletion as a means to control for sequencing and platform performance.

Further processing included fragmentation by sonication to a median size of 200 nt, and cDNA library preparation (NEBNext Ultra II Directional RNA Library Prep Kit—Bacteria (New England Biolabs) according to the manufacturer’s instructions. Multiple oligos for Illumina sequencing (#E6440L) were obtained from New England Biolabs and used as indices that allow uniqueness in both directions. Illumina sequencing was performed with the NovaSeq 6000 sequencing system (S1 flow cell, Illumina) with 100 cycles and paired-end reads (PE50, 2 x 50 bp reads). The fluorescent images were processed into sequences and transformed to FastQ format using the Genome Analyzer Pipeline Analysis software 1.8.2 (Illumina). The sequence output was controlled for general quality features, sequencing adapter clipping, and demultiplexing using the fastq-mcf and fastq-multx tools of ea-utils [88].

Read mapping, bioinformatics and statistics

Quality of the sequencing output was analyzed using FastQC (Babraham Bioinformatics). All sequenced libraries were mapped to the YPIII genome (NC_010465) and the pYV plasmid (accessions NC_006153) using Bowtie2 (version 2.1.0) with default parameters [89]. After read mapping, SAMtools was employed to filter the resulting bam files for uniquely mapped reads (both strands), which were the basis for downstream analyses. To detect ORFs and trans-encoded sRNAs that are differentially expressed in YPIII and the rnc mutant YP356, we used DESeq2 [90] for all differential expression (DE) analyses following the default analysis steps described in the package’s vignette. For each comparison, HTSeq in union count mode was used to generate raw read counts required by DESeq as the basis for DE analysis.

Data access

The high-throughput read data is deposited at the Gene Expression Omnibus (GEO) with accession no.: GSE (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE249386). The comparative transcriptome analyses are given in S2–S4 Datasets.

Gel electrophoresis, preparation of cell extracts, and Western blotting

For the detection of proteins of interest, Y. pseudotuberculosis cultures were grown under indicated growth conditions. Cell extracts of equal amounts of bacteria were prepared and separated on 12–15% SDS polyacrylamide gels [86]. The gels were either stained with Coomassie Blue or the samples were transferred onto a PVDF membrane via electro-blotting and probed with the respective antibody as described [91]. For visualization of the proteins mentioned below, generated polyclonal (Davids Biotechnology) or purchased monoclonal antibodies were used (anti-LcrF 1:1,000, gift of G. Plano; anti-CsrA 1:8000 anti-rabbit, anti-β-lactamase 1:10,000 anti-mouse, Abcam; anti-YopD 1:6,000 and anti all Yops 1:12,000 gift of A. Forsberg).

Yop effector secretion assay

The Yop secretion assay was performed as described [92]. Bacteria were grown overnight at 25°C in LB medium, diluted 1/50 in fresh LB medium, and grown at 25°C for 2 h and then shifted to 37°C in the presence or absence of 20 mM NaOx and 20 mM MgCl₂. Proteins in the filtered medium supernatant were precipitated with TCA and harvested by centrifugation. Precipitated proteins were washed with 95% acetone with 0.5% SDS and resuspended in equal amounts of sample buffer. Subsequently, the proteins were separated on 15% SDS polyacrylamide gels and visualized by Coomassie brilliant blue staining or Western blotting with polyclonal antibodies directed against secreted Yop proteins of Y. pseudotuberculosis.

Nitrocefin-based and fluorescent-based Yop effector secretion assay

A Nitrocefin-based secretion assay was used to quantify Yop effector secretion of Y. pseudotuberculosis strains. For this purpose, Y. pseudotuberculosis strains harboring plasmid pIVO13 encoding a yopE-blaM fusion were grown under indicated conditions and adjusted to an OD₆₀₀ of 0.8. Bacterial cells were pelleted and 95 μl of the supernatant were mixed with 5 μl of a 2 mmol Nitrocefin diluted in PBS and incubated for 1 h at room temperature. The absorption at 486 nm (A₄₈₆) and 390 nm (A₃₉₀) was measured at specific time points using the Varioskan Flash (Thermo Scientific). The specific β-lactamase activity was calculated as ratio between A₄₈₆/A₃₉₀ (sample) / A₄₈₆/A₃₉₀ (medium).

A fluorescent-based translocation assay was employed to quantify Yop effector translocation by the Y. pseudotuberculosis strains. For this purpose, a Live-BLAzer-FRET B/G Loading Kit with CCF4-AM (Invitrogen) was used. Human macrophages (THP-1) were differentiated with 50 ng/ml of phorbol 12-myristate 13-acetate (PMA) for 3 days in μ-slide 8 well (ibidi) at a seeding density of 5x10⁴ cells/well. Subsequently, culture medium was refreshed and cells were further incubated for two additional days. Human epithelial cells (HEp-2) were seeded in μ-slide 8 well (ibidi) at a concentration of approximately 3x10⁴ cells/well and allowed to attach overnight. The next day, cells were washed and infected with a multiplicity of infection (MOI) of 30 of Y. pseudotuberculosis WT and RNase mutant strains harboring plasmid pIVO13 encoding a yopE-blaM fusion or the empty plasmid as negative control. After incubation for 1 h, cells were washed with phosphate-buffered saline (PBS) and stained with loading dye containing 50 μg/ml gentamicin according to the manufacturer’s protocol. After staining for 1 h, YopE translocation was visualized by fluorescence microscopy using a compact fluorescence microscope (BZ-9000, Keyence).

Analysis of the Yersinia phagocytosis by human macrophages

In preparation for the phagocytosis assay, THP-1 cells were seeded at a cell density of 6x10⁴ cells/well in a 48-well plate and stimulated with 50 ng/ml PMA. Cells were replaced with fresh medium before the addition of bacteria at MOI of 30. One hour post-infection, the cells were washed extensively with PBS and added medium containing gentamicin at a concentration of 50 μg/ml to eliminate the remaining extracellular bacteria. After incubation for an additional hour, macrophages were washed and the total number of intracellular bacteria was determined by lysing the cells with 0.1% Triton X-100. Subsequently, bacteria were plated on LB agar and colonies were counted. The phagocytosis rate was calculated based on the initial bacterial inoculum.

β-galactosidase assays

Y. pseudotuberculosis strains, carrying lacZ reporter fusions of interest were grown under indicated growth conditions in LB medium. The β-galactosidase activity of the lacZ fusion constructs was measured in permeabilized cells as described previously [87]. The activities were calculated as follows: β-galactosidase activity OD₄₁₅ · 6,75 · OD₆₀₀^-1 ·t (min)^-1 · Vol (ml)^-1.

Cell viability assay

Cell viability of bacterial cultures was measured using the BacTiter-Glo Microbial Cell Viability Assay (Promega) by detection of present ATP. Bacterial cultures were cultivated under indicated conditions and the OD₆₀₀ adjusted to 0.8 in the growth medium. The culture was mixed 1:1 with the BacTiter-Glo substrate and luminescence was measured using the Varioskan Flash (Thermo Scientific). Cell viability was determined in duplicates by calculating the ratio of luminescence of the sample and medium.

Scanning and transmission electron microscopy

In order to identify T3SS needles on the bacterial surface, negative staining for transmission electron microscopy and field emission scanning electron microscopy was used. Bacterial overnight cultures were diluted 1:50 in LB and grown at 25°C, 37°C, and 37°C/-Ca²⁺ for 4 h. Three independent cultures were pooled and cultures were mixed with cooled glutaraldehyde (2%) and formaldehyde (5%). For transmission electron microscopy thin carbon support films were prepared by sublimation of carbon on freshly cleaved mica. Using 300 mesh copper grids, the samples were negatively stained with 2% (w/v) aqueous uranyl acetate, according to the method of Valentine et al. [93], and examined in a transmission electron microscope (TEM910, Zeiss, Germany) at an acceleration voltage of 80 kV at calibrated magnifications. Images were recorded digitally with a Slow-Scan CCD-Camera (ProScan, 1024 x 1024, Scheuring, Germany) applying the ITEM-Software (Olympus Soft Imaging Solutions, Münster, Germany).

For field emission scanning electron microscopy glass coverslips were coated with a poly-L-lysine solution (Sigma, Munich, Germany), 30 μl of the samples were added, left for 10 min, washed gently with TE buffer (TRIS EDTA 0,01 M, pH 6.9) and then fixed in 2% glutaraldehyde in cacodylate buffer. After washing with TE buffer dehydration was carried out in a graded series of acetone (10, 30, 50, 70, 90, 100%) on ice for 15 min for each step. Samples were then critical-point dried with liquid CO₂ (CPD 300, Leica, Wetzlar) and covered with a gold film by sputter coating (SCD 040, Balzers Union, Liechtenstein). For examination in a field emission scanning electron microscope (Merlin, Zeiss, Germany), an Everhart Thornley SE-detector was used with the inlens SE-detector in a 25:75 ratio at an acceleration voltage of 5 kV. Images were recorded applying the SmartSEM software version 6.06.