CsrA Positively and Directly Regulates the Expression of the pdu, pocR, and eut Genes Required for the Luminal Replication of Salmonella Typhimurium

ABSTRACT Enteric pathogens, such as Salmonella, have evolved to thrive in the inflamed gut. Genes located within the Salmonella pathogenicity island 1 (SPI-1) mediate the invasion of cells from the intestinal epithelium and the induction of an intestinal inflammatory response. Alternative electron acceptors become available in the inflamed gut and are utilized by Salmonella for luminal replication through the metabolism of propanediol and ethanolamine, using the enzymes encoded by the pdu and eut genes. The RNA-binding protein CsrA inhibits the expression of HilD, which is the central transcriptional regulator of the SPI-1 genes. Previous studies suggest that CsrA also regulates the expression of the pdu and eut genes, but the mechanism for this regulation is unknown. In this work, we show that CsrA positively regulates the pdu genes by binding to the pocR and pduA transcripts as well as the eut genes by binding to the eutS transcript. Furthermore, our results show that the SirA-CsrB/CsrC-CsrA regulatory cascade controls the expression of the pdu and eut genes mediated by PocR or EutR, which are the positive AraC-like transcriptional regulators for the pdu and eut genes, respectively. By oppositely regulating the expression of genes for invasion and for luminal replication, the SirA-CsrB/CsrC-CsrA regulatory cascade could be involved in the generation of two Salmonella populations that cooperate for intestinal colonization and transmission. Our study provides new insight into the regulatory mechanisms that govern Salmonella virulence. IMPORTANCE The regulatory mechanisms that control the expression of virulence genes are essential for bacteria to infect hosts. Salmonella has developed diverse regulatory mechanisms to colonize the host gut. For instance, the SirA-CsrB/CsrC-CsrA regulatory cascade controls the expression of the SPI-1 genes, which are required for this bacterium to invade intestinal epithelium cells and for the induction of an intestinal inflammatory response. In this study, we determine the mechanisms by which the SirA-CsrB/CsrC-CsrA regulatory cascade controls the expression of the pdu and eut genes, which are necessary for the replication of Salmonella in the intestinal lumen. Thus, our data, together with the results of previous reports, indicate that the SirA-CsrB/CsrC-CsrA regulatory cascade has an important role in the intestinal colonization by Salmonella.

in the presence of propanediol and ethanolamine, respectively, which are nonfermentable carbon compounds that are metabolized in the lumen of the inflamed intestine via tetrathionate respiration (4)(5)(6)(7). In laboratory conditions, the SPI-1 genes are expressed when Salmonella is grown in lysogeny broth (LB) (under SPI-1-inducing conditions) (8,9). The pdu and eut genes are expressed in LB at low levels, and their expression is activated in the presence of propanediol or ethanolamine, respectively (10,11).
A myriad of regulators control the expression of the SPI-1 genes, most of which act on HilD, which is an AraC-like transcriptional regulator that is encoded within SPI-1 that directly or indirectly activates the expression of the SPI-1 genes and other related genes that are located outside of SPI-1 (2,3,12). Among the regulators controlling HilD are the BarA/SirA two-component system (TCS) and the Csr system (12,13). The BarA/ SirA and Csr systems are present in numerous bacteria, in which they control a wide variety of cellular processes by acting in a regulatory cascade (14,15). The BarA/SirA TCS activates the transcription of the csrB and csrC genes encoding the CsrB and CsrC (CsrB/C) small RNAs (sRNAs), which bind to the RNA-binding protein CsrA (16,17). CsrA binds to sequences containing conserved GGA motifs, which are generally located within the loops of hairpin structures in RNAs (18,19). CsrB and CsrC contain multiple CsrA-binding sites, and they therefore antagonize CsrA activity on target transcripts (20,21). The most common CsrA-mediated regulatory mechanism involves CsrA binding to multiple sites in 5' RNA leader regions, one of which overlaps the Shine-Dalgarno (SD) sequence, thereby repressing translational initiation, which often leads to the degradation of mRNAs (15,(22)(23)(24)(25)(26)(27). However, CsrA can also activate the expression of some genes either by binding to leader sequences in mRNAs and thereby preventing the formation of secondary structures that sequester the SD sequence or by protecting mRNAs from attack by RNases and thereby stabilizing transcripts (15,28,29). CsrA binds to the hilD mRNA and blocks its translation, while the BarA/SirA TCS activates the expression of the CsrB/C sRNAs that antagonize the effect of CsrA, thereby favoring the expression of hilD (13,30).
The expression of the pdu and eut genes is positively controlled by the AraC-like transcriptional regulators PocR and EutR, respectively (10,11,31,32). Additionally, global expression studies implicate the BarA/SirA and Csr systems in the regulation of the expression of the pdu and eut genes (33,34). However, the mechanism for this regulation is unknown. In this study, we show that CsrA activates the expression of the pdu and eut genes by directly binding to the pocR, pduA, and eutS transcripts, while SirA-CsrB/C reduces the expression of these genes by counteracting the effect of CsrA.
Our results indicate that the regulation of the pdu and eut genes by the SirA-CsrB/C-CsrA cascade requires the presence of the transcriptional regulators PocR and EutR, respectively.

RESULTS
SirA, CsrB/C, and CsrA regulate pdu and eut expression. To examine the regulation of the pdu and eut genes by SirA, CsrB/C, and CsrA, we first constructed plasmids with lacZ translational fusions carrying the full-length intergenic regulatory region upstream and the first codons of the pduA or eutS genes (the pduA and eutS are the first genes of the pdu and eut operons, respectively) ( Fig. 1). Then, the expression of the generated pduA'-'lacZ and eutS'-'lacZ translational fusions was quantified in the WT S. Typhimurium SL1344 strain and its isogenic DsirA and DcsrB DcsrC mutants, as well as in the WT strain overexpressing CsrA and in the DsirA mutant overexpressing SirA or CsrB. We did not analyze the effect of the DcsrA mutant, as this strain presented a severe growth defect (13,35). The bacterial strains were grown in LB at 37°C, which are conditions under which SirA, CsrB/C, and CsrA control the expression of the SPI-1 virulence genes (SPI-1-inducing conditions) (8,9). The expression of both the pduA'-'lacZ and eutS'-'lacZ fusions was increased in the DsirA and DcsrB DcsrC mutants as well as in the WT strain overexpressing CsrA from the plasmid pK3-CsrA, compared with the WT strain either with or without an empty vector (Fig. 2). Conversely, the expression of both fusions was decreased in the DsirA mutant overexpressing SirA or CsrB from the plasmids pK3-SirA and pK3-CsrB, respectively, with respect to the DsirA mutant with or without an empty vector (Fig. 2). These results indicate that SirA and CsrB/C negatively control the expression of the pdu and eut genes, whereas CsrA positively controls their expression.
The regulation of pdu and eut by SirA-CsrB/C-CsrA requires the presence of PocR or EutR. The expression of the pdu and eut genes is positively controlled by the transcriptional regulators PocR and EutR, respectively (10,11,31,32). Therefore, we asked whether the regulation of pdu and eut by the SirA-CsrB/C-CsrA cascade involves PocR or EutR. To test this possibility, we examined expression of the pduA'-'lacZ and eutS'-'lacZ fusions in the DsirA DpocR, DcsrB DcsrC DpocR, DsirA DeutR, and DcsrB DcsrC eutR mutant strains as well as in strains overexpressing CsrA (pK3-CsrA) in the presence or absence of PocR or EutR. The increased expression of pduA'-'lacZ and eutS'-'lacZ caused by the absence of SirA or CsrB/C or by the overexpression of CsrA (Fig. 2) was lost in the absence of the respective regulator PocR or EutR; the expression of the pduA'-'lacZ and eutS'-'lacZ fusions was barely detectable in the absence of PocR and EutR, respectively (Fig. 3). These results demonstrate that regulation of the pdu and eut genes by the SirA-CsrB/C-CsrA cascade requires the presence of PocR and EutR, respectively.
SirA, CsrB/C, and CsrA regulate PocR and EutR expression. Our results suggest that SirA-CsrB/C-CsrA regulates the pdu and eut genes through PocR and EutR. Therefore, we sought to determine whether SirA-CsrB/C-CsrA controls the expression of PocR and EutR.
The pocR gene is located near the pdu genes, and it is expressed as a single gene operon (Fig. 1A). We constructed a pocR'-'lacZ translational fusion carrying the fulllength intergenic region upstream and the first codons of pocR (Fig. 1A), and we analyzed the expression of this translational fusion in the different S. Typhimurium strains that were assessed in this study. The expression of pocR'-'lacZ increased in the absence of SirA or CsrB/C and by the overexpression of CsrA from pK3-CsrA (Fig. 4A). Conversely, the expression of this fusion decreased by the expression of SirA or CsrB from pK3-SirA and pK3-CsrB, respectively (Fig. 4A). Consistent with these results, the production of CsrA from pK3-CsrA increased the chromosomal expression of 3xFLAGtagged PocR (PocR-FLAG) by 3.4-fold in the WT strain (Fig. 4B). These results show that the SirA-CsrB/C-CsrA cascade regulates the expression of pocR. To determine whether the presence of PocR is required for this regulation, we next examined the expression of the pocR'-'lacZ fusion in the DsirA DpocR and DcsrB DcsrC DpocR mutant strains as well as in strains overexpressing CsrA from pK3-CsrA in the WT and DpocR genetic backgrounds. The absence of SirA or CsrB/C as well as the overexpression of CsrA increased the expression of pocR'-'lacZ in the DpocR mutants (Fig. 5A), indicating that PocR is not required for the regulation of the pocR gene by the SirA-CsrB/C-CsrA cascade.
The pocR'-'lacZ fusion was expressed at similar levels in the WT strain and its isogenic DpocR mutant when both were carrying an empty vector (Fig. 5A), indicating that PocR does not autoregulate its own expression. Consistent with this finding, electrophoretic mobility shift assays (EMSAs) showed MBP-PocR binding to the pduA promoter region but not to the pocR or eutR promoters ( Fig. 5B and C). In agreement with these findings, previous studies indicated that PocR activates the expression of the pdu genes but does not regulate itself (10,36). It should be noted that PocR binding to the pduA promoter had not been determined previously.
eutR is the last gene of the eut operon, and it is transcribed primarily from the promoter located upstream of eutS; however, an additional promoter upstream of eutR has been reported ( Fig. 1B) (11). Our results show that the SirA-CsrB/C-CsrA cascade controls the expression of eutS (Fig. 2B), implying that SirA-CsrB/C-CsrA would also regulate the expression of eutR. Consistent with this prediction, the production of CsrA from pK3-CsrA increased the chromosomal expression of 3xFLAG-tagged EutR (EutR-FLAG) by 2.1-fold in the WT strain (Fig. 4D). To determine whether the SirA-CsrB/C-CsrA cascade also regulates the expression of a translational fusion driven by promoter P2 immediately upstream of eutR (Fig. 1B), we constructed and analyzed a P2-eutR'-'lacZ translational fusion. The absence of SirA or CsrB/C did not affect the expression of the P2-eutR'-'lacZ fusion (Fig. 4C). Intriguingly, the production of CsrA from pK3-CsrA significantly reduced the expression of the P2-eutR'-'lacZ fusion in the WT strain (Fig. 4C), which was an opposite effect to that observed for CsrA on eutS and EutR-FLAG (Fig. 2B, 3B, and 4D) as well as on pduA, pocR, and PocR-FLAG ( Fig. 2A, 3A, 4A, 4B, and 5A).
Together, these results indicate that the SirA-CsrB/C-CsrA cascade regulates the expression of pocR by acting on the transcript generated by the promoter upstream of this gene and that it regulates the expression of eutR by acting on the transcript generated by the promoter upstream of eutS.
SirA and CsrB/C regulate pdu and eut expression in the presence of inducer molecules. The expression of the pdu and eut genes is activated to high levels in the presence of propanediol and ethanolamine, respectively; vitamin B 12 has also been shown to act as an inducer molecule for the expression of the eut genes (10,11,37). We found that SirA and CsrB/C also control the expression of the pdu and eut genes in the presence of inducer molecules (propanediol 1 vitamin B 12 or ethanolamine 1 vitamin B 12 ), but this is to a lower extent than that observed under SPI-1-inducing conditions (lacking inducers) (Fig. 6). CsrA binds to the pocR, pduA, and eutS transcripts. To determine the direct targets of CsrA for the regulation analyzed in this study, we performed EMSAs using purified CsrA-H6 and the 59-end-labeled RNA of pocR, pduA, and eutS. A band with lower mobility was detected with a CsrA-H6 concentration between 1 and 8 nM for pocR RNA, between 8 and 63 nM for pduA RNA, and between 2 and 16 nM for eutS RNA (Fig. 7), indicating that CsrA binds to these three transcripts. A nonlinear least-squares analysis of these data yielded apparent K d values of 7.4, 31, and 5.7 nM for pocR, pduA, and eutS RNAs, respectively (Fig. 7), indicating that CsrA has a higher affinity for pocR and eutS than for pduA. Additionally, the specificity of the CsrA interaction with pocR, pduA, and eutS RNAs was evaluated by competition experiments with specific (pocR, pduA, and eutS) and nonspecific (phoB) unlabeled RNA competitors. Unlabeled pocR, pduA, or eutS RNAs were effective competitors, whereas phoB RNA was not (Fig. 7), indicating that these interactions are specific. Collectively, our results show that CsrA positively regulates the expression of the pdu genes by binding to the pocR and pduA transcripts, and it also positively regulates the expression of the eut genes by binding to the eutS transcript.

DISCUSSION
In the mouse intestine as well as under SPI-1-inducing in vitro conditions, Salmonella differentiates into two subpopulations that are genetically identical but phenotypically distinct: one subpopulation expresses SPI-1 genes (SPI-1 ON ), and the other does not (SPI-1 OFF ) FIG 4 The SirA-CsrB/C-CsrA cascade regulates the expression of PocR and EutR. The b-galactosidase activity of the pocR9-9lacZ (A) and P2-eutR9-9lacZ (C) translational fusions was quantified in the indicated strains. The data represent the average and the standard deviation of three independent experiments done in duplicate. The P values were calculated using one-way ANOVAs with Tukey's post hoc tests (***, P , 0.001). Western blot analysis of PocR-FLAG (B) and EutR-FLAG (D) expression in the WT strain carrying a chromosomal 3xFLAG-tagged pocR or eutR gene, respectively, with the indicated plasmids. Monoclonal anti-FLAG antibodies were used to detect the FLAG-tagged proteins. As a loading control, the expression of GroEL was also detected using polyclonal anti-GroEL antibodies. The blots were performed three times in independent experiments. Representative images of the blots are shown. The fold change for the expression of FLAG-tagged proteins were calculated as the ratio of AU (arbitrary units) normalized with GroEL, using the ImageJ software package. The P values were obtained by using unpaired Student's t tests (*, P , 0.05; **, P , 0.01). Western blots and b-galactosidase assays were performed with samples taken from bacterial cultures that were grown overnight in LB at 37°C. (38)(39)(40)(41)(42). Different studies indicate that Salmonella colonizes the gut via a division of labor between these two subpopulations: the SPI-1 ON is able to invade cells from the intestinal epithelium and trigger an inflammatory response that provides a particular niche where the SPI-1 OFF replicates in the intestinal lumen, thereby displacing the microbiota (39)(40)(41)(43)(44)(45).  b-galactosidase assays were performed with samples taken from bacterial cultures that were grown overnight in LB at 37°C. The data represent the average and the standard deviation of three independent experiments done in duplicate. The P values were calculated using one-way ANOVAs with Tukey's post hoc tests (***, P , 0.001). Nonradioactive EMSAs using purified MBP-PocR and the DNA fragments contained in the pduA9-9lacZ (B) and pocR9-9lacZ (C) fusions. As a negative control, the DNA fragment contained in the eutR9-9lacZ fusion was included in each DNA-binding reaction. The immunodetection assays using anti-MBP monoclonal antibodies (the images below the EMSAs) show that the MBP-PocR protein forms overly large complexes with (pduA) or without (pocR and eutR) bound DNA, and these remained near the wells of the gels.
Alternative electron acceptors, such as tetrathionate (5), which is utilized by Salmonella for propanediol and ethanolamine metabolism (4)(5)(6)(7)46), are generated in the inflamed gut. The pdu and eut genes encode the enzymes necessary for the use of propanediol and ethanolamine as carbon sources (32,47). The BarA/SirA-CsrB/C-CsrA regulatory cascade controls the expression of the SPI-1 genes under SPI-1-inducing conditions (13). BarA/SirA-CsrB/C induces the expression of these genes by antagonizing the translational repression exerted by CsrA on the hilD transcript (13,30), which codes for the central positive regulator of SPI-1 (48)(49)(50). Recently, we reported that SirA is required for the generation of the SPI-1 ON subpopulation (51). Results from the present study, together with the results of previous reports (33,34), show that the BarA/SirA-CsrB/C-CsrA cascade also controls the expression of the pdu and eut genes under SPI-1-inducing conditions, but, interestingly, this occurs in a manner that is opposite to that of the control exerted by this regulatory cascade on the SPI-1 genes (Fig. 8).
Our data indicate that SirA-CsrB/C negatively controls the pdu and eut genes by counteracting the direct positive regulation of CsrA on the pocR-pduA and eutS transcripts, respectively. We found that CsrA binds to the pocR and eutS RNAs with a similar affinity to that reported for its interaction with hilD RNA (13) and that the affinity of CsrA for pduA RNA was approximately 5-fold lower. CsrA primarily represses the translation of target mRNAs (13,15,(22)(23)(24)(25)(26)(27). However, positive regulation by CsrA on some transcripts has been described. For instance, CsrA positively controls the expression of the master regulator of flagellar genes, namely, FlhD 4 C 2 , by binding to the flhDC mRNA and thereby protecting it from degradation by RNase E (28). Furthermore, CsrA activates the expression of YmdA, which is a protein of unknown function, by binding to the ymdA mRNA and thereby exposing the SD sequence for translation initiation (29). The precise mechanism(s) by which CsrA activates the expression of pocR, pduA, and eutS remains to be determined.
Vitamin B 12 is necessary for the metabolism of propanediol and ethanolamine because it acts as a cofactor for the propanediol dehydratase enzyme that is encoded by the pdu genes as well as for the ethanolamine ammonia-lyase enzyme that is encoded by the eut genes (31,32,47,52,53). Salmonella anaerobically synthesizes vitamin B 12 with the enzymes encoded in the cbi-cob operon (36,54,55). In addition, the reduction of tetrathionate during the metabolism of propanediol and ethanolamine requires tetrathionate reductase, thiosulphate reductase, and sulfite reductase enzymes, which are encoded by the ttr, phs, and asr operons, respectively (4,56). Interestingly, our unpublished results and previous reports support that the BarA/SirA-CsrB/C-CsrA regulatory cascade controls the expression of the cbi-cob, phs, and asr genes in a similar way to that observed for the pdu and eut genes (33,34). Based on the results described above, it is tempting to speculate that the BarA/ SirA-CsrB/C-CsrA regulatory cascade could be involved in the generation of the SPI-1 ON and SPI-1 OFF subpopulations, which is a matter of our current investigation. Molecules or cues that are present in the intestine, acting through the BarA/SirA TCS and/or other regulatory pathways, could mediate whether the expression of SPI-1 or that of the pdu/eut/cbi-cob genes is induced. For instance, the short-chain fatty acids (SCFAs) acetate and formate act through the BarA/SirA TCS to activate SPI-1 gene expression (57)(58)(59). Other SCFAs or long-chain fatty acids (LCFAs), such as propionate, butyrate, oleate, myristate, and palmitate, repress the expression of the SPI-1 genes (57,(60)(61)(62)(63). It is necessary to know how all of these molecules or signals acting on SPI-1 affect the expression of the pdu/eut/cbi-cob genes. This will help to integrate the possible pathways that oppositely control the expression of the SPI-1 and the pdu/eut/cbi-cob genes, which could be involved in the generation of the SPI-1 ON and SPI-1 OFF subpopulations. It should be noted that propanediol, which is the inducer molecule for pdu expression, indirectly represses SPI-1 via propionate production from propanediol metabolism (64). Our study adds an additional layer to the complex regulatory network that controls Salmonella virulence.

MATERIALS AND METHODS
Bacterial strains and growth conditions. The bacterial strains used in this study are listed in Table 1. The cultures for the b-galactosidase and Western blot assays were grown in test tubes containing 5 mL of lysogeny broth (LB)-Miller (1% tryptone, 0.5% yeast agar, and 1% NaCl [pH 7.5]), that were incubated for 16 h at 37°C with shaking at 200 rpm. When necessary, the culture medium was supplemented with streptomycin (100 mg/mL), ampicillin (200 mg/mL), or kanamycin (30 mg/mL).
Construction of plasmids. The plasmids and primers used in this study are listed in Tables 1 and 2, respectively. To construct the peutS-lacZ, ppocR-lacZ, and peutR-lacZ plasmids, the regulatory regions of eutS, pocR, and eutR were amplified by PCR using the primer pairs eutS-FwEcoRI/R2eutS-BamHI, pocR-FwEcoRI/pocR-RvBamHI and F2eutR-EcoRI/R2eutR-BamHI, respectively. The PCR products were digested with EcoRI and BamHI and were then cloned into the pRS414 vector (67) digested with the same enzymes. To construct the pMAL-PocR plasmid expressing the MBP-PocR fusion protein, the pocR structural gene was amplified by PCR using the primers F2PocR-MBP and R2PocR-MBP. The PCR product was digested with BamHI and PstI and was cloned into vector pMAL-c2x digested with the same enzymes. Chromosomal DNA from the WT S. Typhimurium SL1344 strain was used as the DNA template in all of the PCRs. All of the plasmids were sequenced and then transformed via electroporation into S. Typhimurium genetic backgrounds, as specified.
b-galactosidase assay. The protein quantification and b-galactosidase activity measurements were performed as previously described (68,69), with the following modifications. Samples of cells were  The rate of each reaction was obtained by recording the change in absorbance at 405 nm every 15 s for 5 min by using an ELx808 scanning microplate reader and the Gen5 software package (BioTek). The activities were obtained via interpolation with a standard curve (0 to 5,400 U) that was previously stored in the Gen5 software. The protein concentration of each cell extract was obtained by using a bicinchoninic acid (BCA) protein assay (Pierce). Bovine serum albumin was used as the protein standard. The enzyme activity and protein concentration values were used to calculate the b-galactosidase specific activity (U/mg).
Purification of maltose binding protein (MBP)-PocR. E. coli BL21/DE3 carrying the pMAL-PocR was grown in two flasks containing 100 mL of LB with 0.2% glucose at 37°C in a shaken water bath. At an optical density of 0.6, the expression of MBP-PocR was induced by the addition of 1 mM isopropyl-b-D-thiogalactopyranoside (IPTG). Then, the cultures were incubated overnight at 18°C. Bacterial cells were collected by centrifugation at 8,000 rpm at 4°C. The pellet was washed once with ice-cold column buffer (200 mM Tris-HCl [pH 7.5], 200 mM NaCl, 1 mM EDTA, and 10 mM b-mercaptoethanol) and resuspended in 30 mL of the same buffer. The bacterial suspension was sonicated in a Soniprep 150 sonicator (Sonics and Materials, Inc.). Bacterial debris was separated by centrifugation at 4°C, and the soluble extract was loaded three times into an amylose column (New England Biolabs) equilibrated with column buffer. The column was then washed with 13 volumes of column buffer. MBP-PocR was eluted with column buffer containing 10 mM maltose (Sigma). The fractions were analyzed by using SDS-12% polyacrylamide gels, and those containing purified MBP-PocR were concentrated in 1 mL of dialysis buffer (  The sequences corresponding to the template plasmids pKD4 or pSUB11 are underlined. c RE, restriction enzyme for which a site was generated in the primer. of purified MBP-PocR in 20 mL of binding buffer containing 10 mM Tris (pH 8.0), 50 mM KCl, 1 mM DTT, 0.5 mM EDTA, 5% glycerol, and 10 mg/mL bovine serum albumin (BSA). The binding reaction mixtures were incubated at room temperature for 20 min, mixed with 2 mL of 5Â DNA-loading dye (0.25% bromophenol blue, 0.25% xylene cyanol FF, 30% glycerol, and 50Â Gel Red [Biotum]), and then analyzed on 6% nondenaturing Tris-borate-EDTA (TBE)-buffered acrylamide gels in 0.5Â TBE buffer at 85 V and room temperature. After electrophoresis, the DNA fragments were visualized by UV light illumination (Bio-Rad Molecular Imager, Gel Doc TM, XR1 Imaging System, USA). The MBP-PocR complex was detected by Western blotting. RNA EMSAs. The electrophoretic mobility shift assays (EMSAs) were performed using published procedures (70). His-tagged CsrA (CsrA-H6) was purified as previously described (71). Note that the sequences of CsrA from Salmonella and E. coli are identical. RNA was synthesized in vitro by using a RNAMaxx Transcription Kit (Agilent Technologies). The PCR fragments that were used as the templates in the transcription reactions contained a T7 promoter and pocR, pduA, and eutS sequences that extended from 291 to 13, 244 to 138, and 247 to 148, relative to the start of the translation, respectively. Gel-purified RNA was 59-end-labeled with [g -32 P]-ATP (7,000 Ci/mmol). RNA suspended in TE buffer was heated to 90°C for 1 min, and this was followed by slow cooling to room temperature. The binding reaction mixtures (10 mL) contained 10 mM Tris-HCl (pH 7.5), 10 mM MgCl 2 , 100 mM KCl, 200 ng/mL yeast RNA, 0.2 mg/mL BSA, 7.5% glycerol, 20 mM DTT, 0.1 mg/mL xylene cyanol, 0.2 nM (eutS and pduA) or 0.1 nM (pocR) RNA, and various concentrations of purified CsrA-H6. The competition assay mixtures also contained unlabeled competitor RNA. The reaction mixtures were incubated for 30 min at 37°C to allow for CsrA-RNA complex formation. The samples were then fractionated through native 10% polyacrylamide gels using 0.5Â TBE buffer. Radioactive RNA bands were visualized with a Typhoon 9410 phosphorimager (GE Healthcare) and quantified using the ImageQuant 5.2 software package. The apparent equilibrium binding constants (K d ) of the CsrA-RNA interaction were calculated as previously described (72).
Statistical analysis. Statistical significance was analyzed by using the Prism 8 program, version 8.01 (GraphPad Software, San Diego, CA), using either one-way analyses of variance (ANOVA) with Tukey's post hoc tests or unpaired Student's t tests. A P value of ,0.05 was considered to be indicative of a statistically significant result.