In the EHEC O157:H7 EDL933 strain, the proQ gene is expressed from the minus strand and located between z2879 (encoding a hypothetical protein) and prc (encoding a periplasmic protease) ( Supplementary Figure S1A ). To investigate the cellular role of proQ in EHEC O157:H7, a proQ deletion strain (ΔproQ) was constructed. Given that the promoter region of prc starts within the proQ coding sequence (Kerr et al., 2014), only the first 190 nucleotides of proQ gene (the region upstream the prc promoter start site) were removed to prevent the disruption of prc promoter. Consistent with this design, the level of prc transcript remained unchanged in the ΔproQ strain compared to wild type (WT) ( Supplementary Figure S2 ), suggesting that the partial deletion of the proQ sequence did not affect prc promoter activity. Moreover, ΔproQ exhibited comparable growth to WT in LB medium (pH 7.2) at 37°C ( Supplementary Figure S1B ), indicating a dispensable role of proQ in EHEC O157:H7 fitness under normal conditions.

EHEC O157:H7 must withstand the harsh conditions of the stomach and small intestine before adhering to intestinal epithelial cells to initiate colonization in the large intestine (Connolly et al., 2015). To test the role of proQ in these environments, we challenged WT, ΔproQ, and complemented (proQ-c) strains with low pH (2.5), 1% bile salts, and 25 μM human defensin-5 (HD-5), simulating stomach and small intestine conditions. In all cases, deletion of proQ significantly impaired bacterial survival, and this defect was rescued by expressing proQ from a low-copy plasmid under the control of its native promoter ( Figures 1A–C ). These results highlight the importance of proQ expression for survival in stomach- and small intestine-like environments before entering the large intestine. Moreover, adherence of EHEC O157:H7 to follicle-associated epithelium in the intestine brings the bacteria into contact with macrophages, where they encounter oxidative stress (Etienne-Mesmin et al., 2011). To determine if proQ also contributes to oxidative stress resistance of EHEC O157:H7, we exposed the WT, ΔproQ and proQ-c strains to 5 mM H₂O₂ for 30 minutes. We observed that survival of ΔproQ was significantly reduced compared to WT and proQ-c strains after H₂O₂ treatment ( Figure 1D ). Altogether, our data suggest that proQ expression is required for EHEC O157:H7 to survive in a range of adverse environment in host gastrointestinal tract, and to establish a successful infection eventually.

To further assess the role of ProQ in the virulence of EHEC O157:H7, WT, ΔproQ, and proQ-c strains were grown to mid-log phase (OD_600nm = 0.6), incubated with Caco-2 cells, and their adherence abilities were compared. We observed that the ΔproQ strain exhibited significantly higher adherence to Caco-2 cells compared to WT ( Figure 1E ). Notably, prc is also required for robust translocation of EspF, one of the LEE effectors, which could potentially contribute to the virulence phenotype of the ΔproQ strain. However, the expression of this gene was not affected by proQ deletion ( Supplementary Figure S2 ). This suggests that the increased adherence observed in the ΔproQ strain is primarily due to the loss of proQ expression rather than indirect effects from a disrupted prc expression. Interestingly, we also found that the expression of the proQ gene was significantly lower in attached EHEC O157:H7 cells compared to non-attached cells ( Figure 1F ), indicating that proQ is downregulated following cell adherence. Together, these findings suggest that proQ acts as a negative regulator of EHEC O157:H7 adherence to host cells.

Given that the LEE locus is one of the most critical virulence factors for EHEC O157:H7 colonization, we next investigated whether ProQ affects its expression. Using qRT-PCR, we compared the expression of representative genes from the seven LEE operons [ler (LEE1), escT and escC (LEE2), escN (LEE3), espB (LEE4), eae and tir (LEE5)] across different proQ genetic backgrounds. As shown in Figure 2A , the expression of all tested LEE genes was significantly upregulated in the ΔproQ strain and returned to WT levels upon proQ complementation. Consistently, EspB secretion increased markedly in the absence of proQ and was restored by proQ complementation ( Supplementary Figure S3 ), indicating that proQ negatively regulates T3SS-mediated secretion of LEE-encoded effectors. Notably, the upregulation of the master regulator ler in the ΔproQ strain suggests that ProQ may indirectly affect other LEE genes through ler. To further dissect how ProQ regulates ler expression, we fused the ler promoter to a lux reporter gene and measured its activity in WT and ΔproQ strains. We observed an enhanced ler promoter activity in the ΔproQ background ( Figure 2B ), suggesting a transcriptional regulation of ler by ProQ. Given that ProQ is also known to function post-transcriptionally, we examined whether ProQ affects ler mRNA stability. RNA decay assays revealed increased ler transcript stability in the ΔproQ strain compared to WT and proQ complementation strains ( Table 1 ; Figure 2C ), indicating that ProQ regulates ler at both transcriptional and post-transcriptional levels. We then conducted an RNA co-immunoprecipitation (RIP) assay to determine whether ProQ directly binds ler mRNA in vivo. Using a chromosomally 3xFLAG-tagged ProQ strain, we immunoprecipitated associated RNAs and assessed ler enrichment by qRT-PCR. Interestingly, no enrichment of ler mRNA was detected in the ProQ-3xFLAG sample ( Figure 2D ), suggesting that ProQ does not directly bind ler mRNA and the regulation of proQ on ler is likely mediated by other factors.

The LEE expression was known to be modulated by multiple transcriptional regulators under various stresses (Woodward et al., 2019; Wale et al., 2021), providing potential indirect pathways for ProQ-mediated regulation on ler. Although little is known about how proQ affects gene expression in EHEC O157:H7, RNA-seq analysis has identified hundreds of genes whose expression is altered by proQ deletion or overexpression in the non-pathogenic E. coli K-12 strain (Melamed et al., 2020). Among these, three genes—ihfA, hns, and adhE—encode transcriptional regulators that are conserved in EHEC and known to modulate LEE expression through different mechanisms (Friedberg et al., 1999; Beckham et al., 2014; Levine et al., 2014). We then used qRT-PCR to assess whether these genes are affected by ProQ in EHEC O157:H7. Our results showed that deletion of proQ led to nearly a 2-fold decrease in hns expression and a 3-fold decrease in ihfA expression, both of which were restored by proQ complementation ( Figure 3A ). In contrast, adhE expression remained unchanged upon proQ deletion or complementation ( Figure 3A ). To further investigate the underlying mechanism, we examined ihfA and hns mRNA stability in the presence and absence of proQ expression. RNA decay assays revealed that both transcripts showed shorter half-lives in the ΔproQ strain compared to WT and proQ complementation ( Table 1 ; Figures 3B, C ). However, their promoter activities showed no differences between the two strains ( Figure 3D ), suggesting that ProQ only regulates ihfA and hns post-transcriptionally, which likely contributes to the observed changes in LEE expression and bacterial adherence in the proQ deletion strain.

As an RNA binding protein, ProQ has been demonstrated to function as a chaperone to affect the regulation or stability of its sRNA targets (Smirnov et al., 2017; Melamed et al., 2020; Ghandour et al., 2025). We next investigated whether ProQ impacts sRNAs involved in LEE regulation in EHEC O157:H7. Using qRT-PCR, we analyzed the expression of five sRNAs, EvrS, DicF, MavR, GlmY and GlmZ, that were shown to influence LEE expression (Gruber and Sperandio, 2014; Sudo et al., 2018; Melson and Kendall, 2019; Sauder and Kendall, 2021; Han et al., 2024). Deletion of proQ markedly decreased the level of GlmY and GlmZ, whereas the level of the other sRNAs remained largely unchanged ( Figure 4A ). RNA decay assays showed that transcripts of GlmY and GlmZ were less stable in the ΔproQ strain ( Table 1 ; Figures 4B, C ). In contrast, complementation of proQ in the deletion strain restored their stability ( Table 1 ; Figures 4B, C ), suggesting that the ProQ protein protects these sRNAs from degradation. Consistent with these findings, deletion of proQ significantly altered the expression of several GlmY/GlmZ-regulated genes. Specifically, upon the deletion of proQ, the tryptophan synthesis gene tnaA was upregulated, while the acid resistance gene gadA and the non-LEE effector gene nleA were downregulated ( Figure 4D ). Expression of the curli formation regulatory gene, csgD, was also repressed when proQ was deleted ( Figure 4D ). To further investigate whether ProQ is involved in biofilm formation, crystal violet staining assays were performed. As shown in Figure 4E , the abolishment of proQ expression severely impaired biofilm formation, indicating that ProQ is a positive regulator of biofilm formation in EHEC O157:H7, which is consistent with the decreased csgD expression in ΔproQ. Altogether, these results show that ProQ stabilizes GlmY and GlmZ and further regulates various downstream genes and pathways controlled by these sRNAs.

Besides LEE, flagellar genes also play critical roles in EHEC O157:H7 pathogenesis (Sun et al., 2022a). To test the importance of proQ on flagella production, we first compared the bacterial motility between WT and ΔproQ. As shown in Figure 5A , the ΔproQ strain exhibited severely impaired motility, with an approximately tenfold reduction in swimming radius after 10 hours on LB agar compared to WT. This defect was rescued by proQ complementation, suggesting that ProQ positively regulates motility of EHEC O157:H7.

The biosynthesis and assembly of a flagellum in E. coli is controlled by genes in three hierarchical classes (Chaban et al., 2015). The Class I gene flhDC encodes the master regulator, which activates Class II genes responsible for structural and assembly components of the hook-basal body, as well as the regulatory proteins FliA and FlgM. Class III genes (e.g., fliC, and motAB-cheAW) encode distal structural elements of the flagellum and proteins essential for flagellar rotation and chemotaxis. To determine which class of flagellar genes is involved in ProQ-mediated regulation of motility in EHEC O157:H7, we assessed the expression of flhD, fliA, flgM, and fliC following proQ deletion. Among these, only expression of fliC showed a significant reduction in the ΔproQ strain, while the expression of others remained unchanged ( Figure 5B ), suggesting that ProQ specifically modulates motility by regulating fliC. To further explore how proQ regulates fliC, we examined the effect of proQ deletion on transcription and stability of fliC. As indicated in Figure 5C , the fliC promoter activity was unaffected by proQ deletion. However, RNA decay assays revealed a remarkable decrease in fliC mRNA stability in the ΔproQ strain compared to WT and proQ complementation strains ( Table 1 ; Figure 5D ). These data strongly indicate that ProQ positively regulates the expression of fliC and bacterial motility of EHEC O157:H7 at the post-transcriptional level.

In addition to regulating virulence genes, ProQ has also been implicated in antibiotic persistence in several pathogenic bacteria (Rizvanovic et al., 2022; Cianciulli Sesso et al., 2024). For instance, in Salmonella, ProQ contributes to the formation of a subpopulation of growth-arrested cells capable of surviving high concentrations of various antibiotics (Rizvanovic et al., 2022). These findings prompted us to investigate whether EHEC O157:H7 also exhibits antibiotic persistence and whether ProQ plays a role in this process. We exposed the WT and ΔproQ strains to several antibiotics commonly used for EHEC treatment—including ampicillin, kanamycin, and tetracycline—at concentrations far above the MICs that are usually used to test bacterial antibiotic persistence (Balaban et al., 2019; Rizvanovic et al., 2022). These antibiotics have been shown to reduce EHEC infection without inducing Shiga toxin release or causing intestinal damage, making them promising options for effective treatment (Mühlen et al., 2020). After 1 hour of antibiotic exposure at 50× MICs, most EHEC O157:H7 cells were efficiently killed ( Figure 6A ). However, a biphasic pattern with a plateau of surviving persister was observed after 2 hours of tetracycline treatment and after 4 hours of ampicillin or kanamycin treatment ( Figure 6A ). To assess the role of ProQ in persister formation, we challenged the ΔproQ strain with tetracycline at 50× MIC for 6 hours and compared its survival to that of the WT and proQ-complemented strains. As shown in Figure 6B , deletion of proQ significantly reduced the persister level compared to the WT, and complementation restored persistence to the WT level. Notably, MIC testing confirmed that differences in survival were not due to altered antibiotic resistance ( Table 2 ). These results demonstrate that ProQ is essential for persister formation in EHEC O157:H7 under antibiotic stress.

It is worth noting that ProQ-dependent persister formation in Salmonella is associated with the activation of metabolically costly processes, such as virulence gene expression (Rizvanovic et al., 2022). In line with this, we found that deletion of fliC significantly reduced persister level after 6 hours of tetracycline treatment ( Figure 6C ), whereas deletion of ler had little effect. This suggests that flagellar synthesis, but not T3SS synthesis, contributes to antibiotic persistence in EHEC O157:H7. Furthermore, deletion of proQ did not further decrease persister levels in the ΔfliC background, whereas a similar effect was not observed in the Δler strain ( Figure 6C ). Collectively, these findings indicate that ProQ promotes antibiotic persistence in EHEC O157:H7 primarily through regulation of fliC, rather than ler.

While we have demonstrated the regulatory roles of ProQ in several genes and sRNAs important for EHEC O157:H7 virulence, it remained unclear whether and how ProQ directly binds these RNAs. To investigate this, we first performed RNA co-immunoprecipitation (co-IP) experiments and assessed the enrichment of hns, ihfA, and fliC mRNAs, as well as the sRNAs GlmY and GlmZ. All tested RNAs were enriched in the ProQ-coIP samples ( Figures 7A , 8A ), indicating that ProQ binds these RNAs in vivo. To further confirm direct binding, we conducted filter binding assays using purified ProQ protein and radiolabeled hns, ihfA, and fliC mRNAs. As shown in Figures 7B–D , ProQ bound directly to all three mRNAs. In contrast, no binding was detected for ler ( Figure 7B ). Importantly, hairpin structures were found in the 3’ UTRs of hns, ihfA, and fliC ( Figure 7E ). Given that ProQ typically recognizes RNA substrates through structured elements at the 3’ end, we generated mutated versions of these mRNAs (hns/ihfA/fliC mut) in which the hairpin structures were disrupted ( Figure 7E ). Binding assays revealed that these mutations severely diminished or abolished ProQ binding ( Figures 7B–D ), demonstrating that ProQ directly interacts with hns, ihfA, and fliC through these specific structures. These findings are consistent with previous studies showing that ProQ preferentially binds RNAs with distinct secondary structures (Holmqvist et al., 2018; Mamonska et al., 2025).

We also confirmed the in vitro binding between ProQ protein with GlmY and GlmZ ( Figures 8B, C ), with similar binding affinities observed for these two sRNAs. Considering that GlmY and GlmZ were previously reported to contain three prominent hairpin structures (Urban and Vogel, 2008), we tested variants of GlmY and GlmZ carrying mutations that disrupted each individual hairpin to identify which hairpins are critical for ProQ binding ( Figure 8D ). Interestingly, disruption of the 5’ hairpin did not significantly affect ProQ binding affinity for both sRNAs, whereas mutations in the middle and 3’ hairpins substantially impaired ProQ binding ( Figures 8B, C ). These results suggest that the central and 3’ hairpins are essential for the interaction between ProQ and GlmY/GlmZ.

ProQ is a conserved protein among many bacterial pathogens, suggesting that it mediates similar post-transcriptional regulatory mechanisms across different species. To further test this, we examined the influence of proQ deletion on LEE and fliC expression in EHEC O145:H28 (a second prevalent EHEC species) and EPEC O55:H7 (the ancestor of EHEC O157H7). qRT-PCR analysis revealed that deletion of proQ orthologous genes significantly elevated expression of ler, espB and tir, but reduced level of fliC gene in both strains ( Figure 9 ). Similar regulatory effects of proQ were also previously reported in another EPEC serotype O127:H6 (Mihaita et al., 2025). These results suggest that the regulatory roles of ProQ on LEE and fliC gene expression is conserved in a range of EHEC and EPEC strains.

The bacterial strains (see a complete list of strains in Supplementary Table S1 ) were grown in standard Luria-Bertani (LB) media with shaking at 200 rpm at 37°C. When required, antibiotics were added (chloramphenicol at 35 μg/mL or kanamycin at 50 μg/mL). Caco-2 cells were purchased from ATCC and maintained at 37°C with 5% CO₂ in Dulbecco’s modified Eagle medium (DMEM; ThermoFisher) supplemented with 10% heat-inactivated fetal calf serum (GE Healthcare) and 10,000 U/mL of Penicillin-Streptomycin (ThermoFisher) and G418 (Sigma) at 500 μg/mL.

The isogenic mutants lacking proQ in EHEC or EPEC strains were created using the λ red recombination technique, using pKD4 to create the deletion PCR products and pCP20 to remove the antibiotic cassette (Datsenko and Wanner, 2000). The mutant was confirmed by colony PCR and Sanger sequencing. The chromosomally tagged ProQ strain was constructed by replacing the stop codon with a 3×FLAG sequence using the same strategy as for gene deletion. The complementation plasmid expressing proQ was engineered by cloning the coding sequence of proQ and its native promoter (100 nt upstream of the transcriptional start site) between the XbaI/SphI restriction sites on the low copy plasmid pACYC184, or the SacII/BamHI sites on pWSK29. The constructed plasmids were transformed into the ΔproQ to construct the complementation strains. To construct the plasmid for ProQ and EspB purification, the coding sequence of proQ or espB was amplified by PCR and cloned in between the SacII/BamHI restriction sites on the pET28a plasmid. The recombined plasmids were transformed into the BL21 (DE3) strain. The transcriptional reporter fusions were constructed by inserting sequences corresponding to -150 – +10 nt relative to the transcriptional start site of each targeted gene (hns, ihfA, and fliC) between into XhoI/BamHI digested pMS402 plasmid. For ler-lux construction, we used the sequence -300 to +10 bp with respect to the transcription start site of the proximal promoter.

All plasmids and primers used in this study are listed in Supplementary Table S2 and Supplementary Table S3 , respectively.

Total RNA was extracted from bacterial cultures growing at OD_600nm of 0.6 in LB media. Briefly, 10 OD of bacterial cell pellets were collected and resuspended in 1 mL TRIzol Reagent (ThermoFisher). Cells were mechanically lysed by beadbeating on ice (using 5 cycles of 10 s on and 20 s off). Supernatant was collected by centrifugation for 10 min at 8,000g, 4°C. 300 μL of phenol: chloroform:isoamyl alcohol (25:24:1, pH4.5, Roth) was then added, mixed with the cell lysates, and incubated for 3 min at room temperature, followed by centrifugation for 15 min at 13,000g, 4°C. The upper aqueous phase was removed and transferred to a new tube for another round of extraction using 300 μL chloroform (Sigma-Aldrich). RNA was mixed with 2.5 volume of ice-cold ethanol and 0.3 M sodium acetate, followed by the incubation at -20°C for overnight. RNA pellets were collected by centrifugation for 10 min at 13,000g, 4°C, washed twice with ice-cold 75% ethanol and resuspended in 50 μL of nuclease-free water (ThermoFisher). RNA was then treated with DNase I (ThermoFisher) (0.25 U per 1 µg of RNA) for 45 min at 37°C and further purified using the same strategy as described above. The integrity of purified RNA was evaluated by electrophoresis. The concentration of RNA samples was determined by Nanodrop (ThermoFisher).

RT-qPCR was performed using the Luna one step RT-qPCR kit (New England Biolabs) and the Applied Biosystems ViiA 7 instruments with QuantStudio Real-Time software v1.3 (Applied Biosystems) according to the manufacturer’s instructions. For qRT-PCR analysis of sRNAs with short length (DicF, GlmY and GlmZ), primers carrying overhang sequences were used, following a protocol published before (Sharbati-Tehrani et al., 2008). Data were normalized to the expression levels of the 16S rRNA gene (rrsH). Fold changes in expression were determined using the 2^-ΔΔCT method (Schmittgen and Livak, 2008). Three biological replicates were analyzed for each experiment. Primer sequences used for qRT-PCR are summarized in Supplementary Table S2 .

Bacterial adherence assay was performed as described previously (Liu et al., 2020). Briefly, bacteria were grown in LB media with an OD_600nm of 0.6, collected by centrifugation at 4,000g for 3 min. The cells were washed with and resuspended in DMEM medium and added to Caco-2 epithelial cells at a multiplicity of infection (MOI) of 100. At 3 h post-infection, the Caco-2 cells were washed three times with sterile 1× PBS to remove non-attached bacteria and lysed with 1 mL of 0.1% Triton X-100 for 5 min. The cell lysates were serially diluted with sterile 1× PBS and plated on LB agar plates. After incubation at 37°C overnight, the bacterial colonies were counted. The adhesion efficiency of different strains was compared by calculating the number of adhered bacteria per mL.

The plasmids carrying transcriptional fusions were transformed into the WT or ΔproQ strains. The resulting strains were then grown at 37°C in LB media to an OD_600nm of 0.6. Cells were washed with and resuspended in sterile 1× PBS. 200 µl of each diluted cell culture was added to 96-well plates (Corning), and the lux signal intensity was recorded using the Infinite M200 PRO plate reader (Tecan). The OD_600nm was also measured in parallel for normalization.

Bacterial cultures were grown to an OD_600nm of 0.6, and 250 µg/ml rifampicin was added. Cells were incubated at room temperature and 4 OD samples were collected at different time intervals post-treatment. Each sample was immediately mixed with 1/5 volume of 95% ethanol and 5% phenol and frozen in liquid nitrogen. Total RNA extraction and qRT-PCR analysis were carried out as described above.

ELISA assays were performed to evaluate the secretion of EspB in EHEC O157:H7 as described previously (Carlson-Banning and Sperandio, 2016). Bacterial cultures were grown in LB medium at 37°C to an OD_600nm of 0.6. Cell growth was then quenched with STOP solution (0.92 M sodium azide and 100 µl of Sigma protease inhibitor cocktail in sterile 1× PBS). The reaction mixtures were diluted 1:2 with sterile 1× PBS and incubated in Dynatech Laboratories Microtiter ELISA plates. Wells were blocked with 5% milk in PBST (1× PBS with 0.1% Tween 20), and samples were washed with PBST before incubation with an anti-EspB primary antibody (1:1,000; ANTIBODIES-ONLINE), followed by a goat anti-rabbit HRP-conjugated secondary antibody (1:1,000; ThermoFisher). Plates were developed using a 3,3’,5,5’-Tetramethylbenzidine substrate (Sigma-Aldrich) and the reaction was stopped with 2N HCl. Absorbance was measured at 450 nm using a Tecan plate reader. Absolute EspB concentrations were determined from standard curves generated by titrating purified EspB protein of known concentration, buffer-exchanged into 1× PBS. Two independent experiments were performed, each with three biological replicates.

To purify ProQ-associated RNAs in vivo, RNA co-immunoprecipitation (RNA co-IP) assays were performed as previously described (Smirnov et al., 2016). An EHEC O157:H7 strain carrying chromosomally 3xFLAG-tagged ProQ were grown at 37°C in LB media to an OD_600nm of 0.6. 50 OD of bacteria were harvested by centrifugation and lysed as described in the “RNA Isolation” section. The cell lysates were split into two: one half incubated with 40 µl of monoclonal anti-FLAG M2 antibody (Sigma-Aldrich) for 30 min at 4°C (as the co-IP sample), while the other was left untreated with antibody and served as the mock co-IP control. 75 µl of prewashed Protein A Sepharose (Sigma-Aldrich) was then added and incubated for another 30 min at 4°C. Afterwards, Sepharose were washed extensively with the lysis buffer and resuspended in the lysis buffer. An equal volume of phenol:chloroform:isoamyl alcohol (25:24:1, pH4.5, Roth) was added for RNA exaction. RNA samples were further purified through two rounds of chloroform extractions and precipitated with ice-cold ethanol and 0.3 M sodium acetate.

To assess if one RNA is co-purified with ProQ-3xFLAG (indicating a direct binding), the RNA samples obtained from RNA co-IP experiments were subjected to qRT-PCR using specific primers targeting genes of interest. The enrichment of target RNAs in the co-IP or mock-coIP experiments was calculated as the ratio of abundance between the FLAG-tagged strains (FLAG) and the untagged control strain (WT). Each experiment was performed in three biological replicates.

The ProQ purification was carried out using a HisTrap crude column (GE Healthcare) on a FPLC system (AKTA pure) as described before (Stein et al., 2020). RNA was transcribed using the MEGAshortscript T7 Transcription Kit (ThermoFisher) following the manufacturer’s instructions. The DNA templates used in in vitro transcription reactions (sequences shown in Supplementary Table S3 ) were synthesized by IDT. P³² labelling and purification of the tested RNAs were performed as described before (Han et al., 2022). Filter binding assay was performed by incubating purified ProQ proteins at various concentrations (0, 5, 12.5, 25, 50, 100, and 200 nM for mRNA binding, and 0, 1.25, 2.5, 5, 12.5, 25, and 50 nM for sRNA binding) with different radiolabeled RNA targets, followed by the protocol described before (Rio, 2012). Each filter binding assay was conducted in two (for GlmY/GlmZ and ler) or three (for hns, ihfA and fliC) replicates.

To test if proQ affects the growth of EHEC O157:H7 under normal growth condition, bacterial overnight cultures were diluted into fresh LB media to achieve an OD_600nm of 0.03. The cultures were then grown at 37°C with shaking at 200 rpm. The OD_600nm values was measured every hour for 14 h using a spectrophotometer (Tecan). Each strain was analyzed in triplicate and the average OD_600nm values were used to plot growth curves. To test the effect of proQ on resistance of EHEC O157:H7 under acidic condition, pre-cultures of EHEC O157:H7 WT, proQ deletion and proQ complementation strains grown in LB media at 37°C (at an OD_600nm of 0.6) were diluted 1:100 in fresh LB with a pH adjusted to 2.5 with HCl. The cells were then incubated at 37°C with shaking at 200 rpm for 2 h. Resistance to bile salts, HD-5 and oxidative stress was determined by incubating bacterial cells grown in LB at an OD_600nm of 0.6 with 1% bile salts or 25 mM HD-5 at 37°C for 1 h, or with 5 mM H₂O₂ for 30 min at 4°C in the dark, respectively. Final bacterial cultures were serially diluted and plated onto LB plates for overnight incubation to determine viability. The survival rate of the cells was determined by comparing the viable counts of the challenged cells to those of the initial inoculum. For all these assays, two independent experiments were carried out with three biological replicates each.

Strains of WT, proQ deletion and proQ complementation were cultured at 37°C in LB media to an OD_600nm of 0.6. 1 µl of cell cultures were spotted on the center of freshly made 0.3% soft agar plates. After incubation for 10 h at 37°C, diameter of the swimming zone for each strain was measured. All strains were tested in biological triplicate.

Overnight cultures were diluted to an OD_600nm of 0.01 in fresh LB media in flat-bottom 96-well microtiter plates (Corning) and grown at 30°C for 48 h under static condition to enable biofilm formation. Wells were washed and cells stained with 0.1% crystal violet (Millipore Sigma) for 15 min. Crystal violet was then removed from the wells, followed by extensive washing with distill water. The remaining dye in the wells was dissolved in 200 μl of ethanol-acetone (4:1) solution and quantified by measuring the OD_590nm values. All strains were tested in biological triplicates, and the experiments were independently repeated twice.

RNA structures were predicted by RNAfold (http://rna.tbi.univie.ac.at/cgi-bin/RNAWebSuite/RNAfold.cgi) (Garcia-Martin et al., 2013) and ViennaRNA (http://rna.tbi.univie.ac.at/forna/) (Gruber et al., 2015), and drawn using R2DT (https://r2dt.bio/) (McCann et al., 2025) and RNACanvas (https://rna2drawer.app/) (Johnson and Simon, 2023).

WT or deletion strains were grown in LB media at 37°C until an OD_600nm reached 0.6. Tetracycline (75 μg/mL), kanamycin (15 μg/mL), or ampicillin (250 μg/mL) were then added. After 1, 2, 4, 6, or 8 h, cells were washed three times and diluted in sterile 1× PBS and serially diluted for plating on agar plates. CFU counts of antibiotic treated and untreated cells were determined after overnight incubation at 37°C. The experiments were performed in six biological replicates.

Unless otherwise indicated, all data are expressed as the mean ± standard deviation (SD) from three independent biological replicates. Statistical analyses were assessed using two-tailed Student’s t-test, one-way ANOVA or two-way ANOVA analysis, or two-sided Mann-Whitney rank-sum test. Differences were considered statistically significant when P ≤ 0.05 (^*), while P values of ≤ 0.01 and ≤ 0.001 were considered highly and extremely significant (^** and ^***), respectively.