Balanced cell division is secured by two different regulatory sites in OxyS RNA

Abstract

The hydrogen peroxide-induced small RNA OxyS has been proposed to originate from the 3′ UTR of a peroxide mRNA. Unexpectedly, phylogenetic OxyS targetome predictions indicate that most OxyS targets belong to the category of “cell cycle,” including cell division and cell elongation. Previously, we reported that Escherichia coli OxyS inhibits cell division by repressing expression of the essential transcription termination factor nusG, thereby leading to the expression of the KilR protein, which interferes with the function of the major cell division protein, FtsZ. By interfering with cell division, OxyS brings about cell-cycle arrest, thus allowing DNA damage repair. Cell division and cell elongation are opposing functions to the extent that inhibition of cell division requires a parallel inhibition of cell elongation for the cells to survive. In this study, we report that in addition to cell division, OxyS inhibits mepS, which encodes an essential peptidoglycan endopeptidase that is responsible for cell elongation. Our study indicates that cell-cycle arrest and balancing between cell division and cell elongation are important and conserved functions of the oxidative stress-induced sRNA OxyS.

INTRODUCTION

The origin of small RNAs (sRNAs) and how they evolved to be part of bacterial regulatory networks is a fascinating question. A recent study focusing on the evolution of sRNAs has identified protein-coding genes as a potential source of sRNAs (Krieger et al. 2022). This study proposes, for example, that the peroxide-induced OxyS sRNA evolved from a 3′-UTR fragment of a peroxidase mRNA. The oxidative stress-induced sRNA of Escherichia coli is one of the most well-studied sRNAs. More than two decades of research on OxyS have shown that it is involved in format metabolism, regulation of stationary phase, and resistance to cephalothin (Zhang et al. 1998; Argaman and Altuvia 2000; Cho and Kim 2018). Moreover, we found that this sRNA plays a role in protecting cells against the damaging effects of mutagens such as hydrogen peroxide and alkylating agents (Altuvia et al. 1997), while overexpression of OxyS has been found to be toxic. Studying the molecular mechanism mediating the protective function of OxyS and/or the toxic phenotype, we showed that the antimutator function of OxyS is actually the other side of the coin of its cytotoxic phenotype (Barshishat et al. 2018). By repressing the expression of the essential transcription termination factor nusG, OxyS enables read-through transcription into a cryptic prophage encoding kilR. The KilR protein interferes with the function of the major cell division protein FtsZ, thus imposing growth arrest. This transient growth inhibition facilitates DNA damage repair, enabling cellular recovery, thereby increasing viability following stress. Namely, by indirectly inhibiting cell division, OxyS buys the cell more time to properly repair its DNA. Once the oxidative stress has been resolved, OxyS is no longer expressed and the cell returns to its normal life cycle (Barshishat et al. 2018).

Here, we show that OxyS sRNA represses the expression of mepS encoding peptidoglycan (PG) endopeptidase in addition to the transcription termination factor nusG.

MepS promotes cell elongation by making space for the insertion of new PG material (Singh et al. 2015; Truong et al. 2020). PG is an essential cross-linked macromolecule that confers bacterial cell shape and rigidity. In order to maintain the integrity of the PG during its expansion, cross-link cleavage by hydrolases such as MepS must be tightly coupled with cross-link formation catalyzed by synthases. As cell division and cell elongation are contrasting functions, mutants impaired in cell division are sensitive to increased activity of PG endopeptidases (Truong et al. 2020). Thus, it makes perfect sense that inhibition of cell division by OxyS is accompanied by repression of mepS expression.

RESULTS AND DISCUSSION

Base-pairing between OxyS loop A and mepS mRNA inhibits the expression of mepS

Searching for OxyS targets using CopraRNA (Wright et al. 2013), we noticed mepS (murein endopeptidase, formerly Spr). mepS encoding an essential PG endopeptidase influences bacterial morphogenesis by incorporating new murein (Singh et al. 2015; Truong et al. 2020). As OxyS repression of nusG results in inhibition of cell division (Barshishat et al. 2018), and mutants impaired in cell division are sensitive to increased activity of PG endopeptidases (Truong et al. 2020), we wondered whether OxyS controls the expression of mepS, in addition to nusG.

Based on the CopraRNA program, 11 nt at the 5′-end of OxyS (part of loop A) were predicted to base-pair with the ribosome-binding site of mepS mRNA (Fig. 1A,B). To examine the effect of OxyS on mepS expression, we constructed a mepS-lacZ translation fusion, Plac-oxyS wild-type, and two OxyS mutants, Plac-oxyS Δ15–33 and Plac-oxyS C18G G30C, in which the binding site to mepS was deleted or disrupted, respectively (Fig. 1A,B). β-Galactosidase assays detected a 4.6-fold decrease in mepS-lacZ expression in the presence of wild-type OxyS, while OxyS mutants (Δ15–33 and C18G G30C) failed to repress mepS expression (Fig. 1C).

FIGURE 1.

OxyS represses mepS expression. (A) Nucleotides 50–84 in OxyS (marked in green) are involved in nusG binding, while nucleotides 25–33 (purple) bind mepS. In OxyS, Δ15–33 mutant nts 15–33 were deleted. In OxyS C18G G30C, the mutation G30C complements (-18)G in mepS, while C18G restores the disrupted OxyS structure. (B) Predicted base-pairing between mepS (red) and OxyS (purple) RNAs. Indicated are the initiation codon (bold) and the Shine–Dalgarno sequence of mepS (underlined). The position of mepS and OxyS complementary mutations is indicated. (C,D) Cultures (MG1655 mal::lacI^q ΔlacZ::Tn10) carrying mepS-lacZ wild-type and mepS C(-18)G-lacZ (pSC101*) translational fusions and Plac-OxyS wild-type and mutants were treated with IPTG (1 mM) at OD₆₀₀ of 0.1–0.2. β-Galactosidase activity was measured 60 min after treatment. Results are displayed as mean of three to six biological experiments ± standard deviation. (E) To detect the effect of OxyS on MepS protein levels, the sequential peptide affinity (SPA) tag was inserted into the MG1655 (mal::lacI^q) chromosome adjacent to the carboxy-terminal amino acid of MepS. At OD₆₀₀ of 0.5, the cultures were exposed to 1 mM H₂O₂ for the indicated time points, whereas cultures carrying OxyS plasmids were treated with 1 mM IPTG for 30 min to induce OxyS expression. Plasmid-encoded (see Fig. 2E, wild-type) and H₂O₂-induced (Fig. 1E, northern) OxyS RNA levels were detected using end-labeled OxyS-specific primer (3708). Loading control (tm RNA) was detected using primer 1912.

Inhibition of nusG by OxyS involves nucleotides 50–84 in OxyS. As inhibition of mepS by OxyS involves nucleotides 25–35, repression of mepS by the Plac-oxyS C76U C77U mutant that is unable to repress nusG but carries an intact mepS recognition domain was similar to that of wild-type OxyS. Repression of mepS-lacZ carrying C(-18)G, the complementary mutation to OxyS C18G G30C, was restored when base-pairing was restored (Fig. 1D).

To detect the effect of OxyS on MepS protein levels, the SPA tag (Zeghouf et al. 2004) was inserted into the MG1655 (mal::lacI^q) chromosome adjacent to the carboxy-terminal amino acid of MepS. At OD₆₀₀ of 0.5, the cultures were exposed to 1 mM H₂O₂ for 5 and 15 min to induce OxyS transcription from the chromosome, whereas cultures carrying OxyS plasmids were treated with 1 mM IPTG for 30 min to induce plasmid-encoded OxyS. Figure 1E (left panel) shows that upon exposure to H₂O₂, MepS-SPA levels decrease with time, while OxyS levels increase. Similarly, MepS-SPA levels decreased dramatically in the presence of wild-type OxyS and OxyS C76U C77U, while OxyS Δ15–33 and OxyS C18G G30C, in which the binding site to mepS was deleted or disrupted, had no effect on MepS protein levels (Fig. 1E, right panel; for plasmid-encoded OxyS RNA levels, see the legends to Figs. 1E and 2E).

FIGURE 2.

Short mepS transcript is responsive to OxyS regulation. (A) mepS is transcribed from two promoters P2 and P1 producing long and short mRNAs. To inactivate the downstream promoter P1, the sites −10 and −35 (underlined) were mutated as shown beneath the sequence. The initiation codon (bold) and the Shine–Dalgarno sequence of mepS (underlined, red) are indicated. The sequence in mepS that can bind OxyS is in red. (B,C) OxyS affects the expression of short mepS mRNA (lacZ assays). Cultures (MG1655 mal::lacI^q ΔlacZ::Tn10) carrying (ΔP2)P1-mepS-lacZ and P2(ΔP1)-mepS-lacZ (pSC101*) translational fusions and Plac-oxyS wild-type and mutants were treated with IPTG (1 mM) at OD₆₀₀ of 0.1–0.2. β-Galactosidase activity was measured 60 min after treatment. Results are displayed as mean of four to five biological experiments ± standard deviation. (D) OxyS-mepS RNA interaction in vivo. Cultures (MG1655 mal::lacI^q) carrying OxyS plasmids and chromosomally encoded wild-type full-length, short [(ΔP2)P1], and long [P2(ΔP1)] mepS transcripts were treated with IPTG (1 mM) at OD₆₀₀ of 0.2–0.3 for 30 min. The cDNA products generated by primer extension using 30 µg of total RNA and end-labeled mepS-specific primer 3412 were analyzed in 6% acrylamide 8 M urea-sequencing gel alongside with pUC19-MspI labeled marker. (RI) Relative intensity. Band intensities were determined by the ImageLab program. mepS mRNA levels were normalized to tm RNA loading control. mepS/Plac was used as a 100% reference. (E) In vivo RNA samples (10 µg) as in D were separated using 6% urea-polyacrylamide gels (northern blot). The membranes were probed with end-labeled OxyS (3708) and tm RNA (1912) specific primers. tm RNA serves as a loading control. (F) OxyS-mepS RNA binding in vitro. In vitro synthesized (0.05 pmol) short (139 nt) and long (248 nt) mepS mRNAs incubated with and without (5 pmol) synthesized OxyS RNAs (173 nt wild-type and 158 nt Δ15–33) at 25°C for 15 min. Primer extension was carried out for 7.5 min at 37°C using end-labeled mepS-specific primer (3740). The products were analyzed in 6% acrylamide 8 M urea-sequencing gel alongside with sequencing reactions.

Short mepS mRNA is more responsive to OxyS regulation

Genome-wide studies identified two promoters upstream of mepS (Fig. 2A; Mendoza-Vargas et al. 2009; Chung et al. 2013). The upstream promoter (P2) is more active in the exponential phase whereas the downstream promoter (P1) is stronger in the stationary phase (Thomason et al. 2015). Folding of the two mRNA species produced by the two promoters showed that the site to which OxyS binds in the long mepS mRNA molecule (G146–G156) is predicted to be sequestered, while the same site in the short molecule (G37–G47) seems more accessible (Supplemental Fig. S1). We examined the effect of OxyS on the short mepS transcript using a (ΔP2)P1-mepS-lacZ fusion in which the P2 was deleted leaving the downstream promoter intact and on the long transcript using P2(ΔP1)-mepS-lacZ in which the −10 and −35 sites of P1 were mutated, leaving the upstream promoter intact (Fig. 2A). β-Galactosidase assays detected an eightfold reduction in mepS-lacZ produced by the short mRNA due to OxyS (Fig. 2B), while repression by OxyS of mepS-lacZ produced by the long transcript resulted in a threefold reduction only (Fig. 2C), indicating that mepS expression produced by the short transcript is more readily regulated by OxyS. Furthermore, mepS-lacZ produced by the long transcript is less accessible to the translation machinery, as indicated by the 5.6-fold decrease in β-galactosidase basal activity levels.

In vivo RNA analysis of chromosomally encoded short [(ΔP2)P1] and long [P2(ΔP1)] mepS mRNA constructed by the scarless mutation methodology showed that wild-type OxyS expression results in approximately four- and twofold decrease in mepS short and long mRNA levels, respectively. Both OxyS mutants with disrupted complementarity to mepS have no effect on mepS RNA levels, while expression of OxyS C76U C77U that is unable to repress nusG but capable of repressing mepS reduces mepS transcript levels, similar to OxyS wild-type (Fig. 2D). As the plasmid-encoded wild-type OxyS levels were similar to the levels of OxyS mutants, this further indicated that repression of mepS is due to mepS-OxyS interaction (Fig. 2E).

Given that stable RNA hybrids may block the elongation by reverse transcriptase, we examined the interaction of in vitro-synthesized OxyS with short and long mepS RNAs, using primer extension assays. These assays showed that the interaction of OxyS and OxyS C76U C77U with short mepS mRNA results in a termination signal that maps to the site of complementarity (Fig. 2F). No termination signals were detected when OxyS Δ15–33 was incubated with short mRNA or upon incubation of wild-type OxyS with the long mepS RNA (Fig. 2F).

Simultaneous inhibition of cell division and cell elongation by OxyS promotes cell recovery

Given that cell division and cell elongation are contrasting functions, mutants impaired in cell division are sensitive to increased activity of PG endopeptidases (Truong et al. 2020). Here we show that OxyS represses the expression of mepS, which is responsible for cell elongation, in addition to cell division inhibition. The effect of mepS repression on growth was examined using an OxyS mutant in which the complementary sequence to mepS was deleted. Cells carrying Plac-oxyS Δ15–33 were more inhibited than cells expressing wild-type OxyS (Fig. 3A), suggesting that decreasing mepS expression is advantageous when cell division is inhibited. As shown previously, the growth of cells expressing the Plac-oxyS C76U C77U mutant that is unable to repress nusG was similar to that of the control (Barshishat et al. 2018).

FIGURE 3.

Simultaneous inhibition of cell division and cell elongation by OxyS promotes cell recovery. (A) E. coli (MG1655 mal::lacI^q) carrying Plac-oxyS wild-type and mutants were treated with 1 mM IPTG at dilution, growth (OD600) was measured as indicated. (B) Cultures (MG1655 mal::lacI^q) carrying control and OxyS plasmids, and a chromosomally encoded wild-type short (ΔP2)P1 mepS transcript that was constructed using the scarless mutations methodology, were treated with IPTG (1 mM) at dilution; CFU was determined 1, 2, and 4 h after dilution. (C) Cultures (MG1655 mal::lacI^q) carrying PBAD-kilR and OxyS plasmids were treated with arabinose (0.2%) and IPTG (1 mM) at dilution; CFU was determined as indicated in the figure. Results are displayed as mean of five biological experiments ± standard deviation. (D) Cultures of wild-type OxyS and two chromosomally encoded OxyS mutants (Δ15–33 and C76U C77U) were treated with 0.2 mM of H₂O₂ for 30 min to induce OxyS expression and then exposed to 10 mM H₂O₂ to induce mutations. CFU and rifampicin resistant cells were monitored after 18 h of growth. Results are displayed as mean of six biological experiments ± standard deviation.

Similarly, cell survival assay of strains carrying chromosomally encoded short mepS mRNA and plasmids encoding wild-type OxyS and OxyS Δ15–33 revealed that while cells carrying the control plasmid continued to grow and divide, cells expressing OxyS stopped dividing and cells expressing the OxyS Δ15–33 showed a decrease in the number of CFU, indicating that stopping cell division without stopping cell elongation is highly detrimental (Fig. 3B).

By repressing the expression of the essential transcription termination factor nusG, OxyS enables read-through transcription into a cryptic prophage encoding kilR. The KilR protein interferes with the function of the major cell division protein FtsZ, thus imposing growth arrest (Barshishat et al. 2018). To characterize the effect on cell survival of OxyS regulation of mepS, separately from nusG, we expressed KilR in trans from an inducible promoter (PBAD-kilR). Expression of both plasmid-encoded and chromosomally encoded kilR resulted in a dramatic decrease in cell survival in the absence of mepS repression (compare lanes 3 and 6 in OxyS Δ15–33, Fig. 3C), whereas mepS repression by the OxyS mutant that is unable to induce the chromosomally encoded kilR, rescues cells from the toxic effects of plasmid-encoded kilR (compare lanes 3 and 9), further confirming the importance of OxyS regulation of both functions together for cells to survive. In trans expression of KilR (PBAD-kilR) in a ΔkilR mutant strain resulted in 3.8-fold reduction in cell survival (Supplemental Fig. S2), similar to the reduction displayed by cells carrying both PBAD-kilR and OxyS that is unable to produce KilR (OxyS C76U C77U), further indicating that KilR expression is more detrimental without the mitigating effects mediated by the OxyS 5′ part (Fig. 3C).

Previously, we showed that by inhibiting cell division, OxyS buys the cells more time to properly repair its DNA (Barshishat et al. 2018). We tested the relative effect of cell division versus cell elongation on DNA damage repair by monitoring the mutation rate in cultures carrying wild-type OxyS and two chromosomally encoded OxyS mutants (Δ15–33 and C76U C77U), constructed using the scarless mutations methodology. The cultures were first treated with 0.2 mM of H₂O₂ for 30 min to induce OxyS expression and then exposed to 10 mM H₂O₂ to induce mutations. Inhibition of cell division only, while cell elongation is active (OxyS Δ15–33), results in a slight increase (1.2-fold) in mutation frequency compared to wild-type OxyS which inhibits both functions. In contrast, inhibiting cell elongation while cell division is active (OxyS C76U C77U) results in a more pronounced increase (1.7-fold) in the rate of mutations (Fig. 3D).

Simultaneous inhibition of cell division and cell elongation by OxyS is phylogenetically conserved

OxyS is a conserved sRNA in the Enterobacteriaceae family which evolved from the 3′ UTR of a peroxidase gene (Krieger et al. 2022). To find out what is the most common function of OxyS and to learn whether it involves linking between cell division and cell elongation, we conducted a phylogenetic prediction of OxyS targetomes. A total of 1340 OxyS homologs were detected with the GLASSgo tool in the NCBI nt database (Lott et al. 2018). A phylogenetically weighted sequence logo based on an alignment of all detected homologs shows that OxyS 5′ part is rather variable, while the internal region and the Rho-independent terminator are rather conserved (Supplemental Fig. S3). We selected 146 representative homologs that cover the whole phylogenetic distribution of OxyS for an evolutionary target conservation analysis based on CopraRNA (Supplemental Fig. S4; Wright et al. 2013). For each organism, we selected 14 partners for the comparative CopraRNA prediction and performed a gene set enrichment analysis on the top 200 predicted targets (Supplemental Fig. S5; Supplemental Table S1). Strikingly, the most frequent category was “cell cycle” that was enriched in 121 of the 146 organisms containing cell cycle and cell division targets including mraY (77/146), murG (72/146), nagZ (70/146), tolB (65/146), zapC (68/146), and ftsZ (54/146). Other categories such as membrane organization, PG metabolic process and extracellular polysaccharide biosynthetic process were also frequently enriched, indicating a possible role of OxyS in cell elongation and membrane growth. Widely conserved predicted targets from these categories are, for example, bamA (124/146), bamD (60/146), or mepS (97/146).

To test the validity of these predictions, we chose zapC, ftsZ, murG, minD, bamD, and mepS of Salmonella for further investigation. Among genes involved in cell division, we found zapC and ftsZ-lacZ fusions to be affected by OxyS (Supplemental Fig. S6). Intriguingly, ftsZ-lacZ regulation by Salmonella OxyS is mediated by the same region that is involved in the regulation of nusG by E. coli OxyS. Thereby, wild-type OxyS of E. coli represses the expression of Salmonella ftsZ-lacZ, and the OxyS mutant that fails to inhibit nusG (C76U C77U) also fails to inhibit ftsZ-lacZ (Supplemental Fig. S6). This result also rules out the possibility that OxyS represses ftsZ expression via the second more canonical site located upstream of the AUG (−2 to −8). Although zapC-lacZ is regulated by the 5′ domain of the site involved in nusG regulation (position 52–66) (Barshishat et al. 2018), E. coli OxyS likely fails to regulate Salmonella zapC because the sequence of the interaction site differs between Salmonella and E. coli, which may indicate independent target evolution. The mepS target and the interaction site are conserved between Salmonella and E. coli. Thus, E. coli OxyS represses Salmonella mepS, and OxyS mutants that fail to inhibit mepS of E. coli also fail to repress mepS of Salmonella (Supplemental Fig. S6).

Fluorescence microscopy imaging of cell morphology

We visualized the effect of the separation between cell division and cell elongation on their morphology by fluorescence microscopy. Surprisingly, we found that inhibition of cell division only, while cell elongation is active (OxyS Δ15–33), results in a unique morphology where cells cluster together and adhere to each other (Fig. 4). Given that the morphology of E. coli cells expressing KilR, which is known to interfere with cell division, was similar to that of cells expressing the OxyS Δ15–33 mutant indicates that “clumping” results from an imbalance between division and elongation. E. coli cells expressing wild-type OxyS displayed elongated filaments, whereas cells expressing the mutant OxyS C76U; C77U that is unable to produce KilR formed short rod-shaped cells very similar to control cells (Fig. 4). The morphology of Salmonella expressing wild-type OxyS appears to be different from the morphology detected in E. coli, with comparable less and shorter cell filaments. Possibly because the effect of E. coli OxyS on cell division is amplified by the indirect interference with the function of FtsZ, mediated through NusG and KilR, while in Salmonella in which kilR is absent, OxyS decreases ftsZ expression directly conceivably by base-paring (Supplemental Fig. S7).

FIGURE 4.

Fluorescence microscopy images of E. coli expressing OxyS or KilR. Cultures of E. coli mal::lacI^q carrying Plac, Plac-OxyS, Plac-OxyS Δ15–33, and Plac-OxyS C76U C77U and Plac-kilR were grown for 3 h in the presence of IPTG from dilution. Scale bar, 100 µm.

In conclusion, in this study we show that E. coli OxyS regulates two different target genes associated with cell division using two separate sites: One site interferes with the function of ftsZ by inhibiting the expression of nusG, whereas the second site inhibits the expression of mepS to prevent cell elongation. By inhibiting both cell division and cell elongation, OxyS rescues cells arrested in the division from the destructive consequences of increased activity of PG endopeptidases and buys the cell more time to properly repair its DNA.

A phylogenetic evolutionary target conservation analysis of OxyS homologs to identify OxyS’ most common functions revealed that the majority of OxyS targets belong to the category of “cell cycle” (121/146), followed by “peptidoglycan metabolism” (82/146). Notably, while the involvement of OxyS in cell-cycle arrest appears to be broadly conserved, in many of the cases it is achieved by regulating different sets of genes. In Salmonella, OxyS affects the cell cycle by regulating at least three cell-cycle-related genes, ftsZ, zapC, and mepS, while in E. coli cell-cycle arrest is achieved through the regulation of nusG and mepS. Intriguingly, ftsZ regulation by Salmonella OxyS is mediated by roughly the same region that is involved in the regulation of nusG by E. coli OxyS. The regulation of nusG is likely a more recent evolutionary adaptation in the genus Escherichia. While the nusG UTR is largely conserved between all 146 investigated organisms, only Escherichia-specific changes in the rather unconserved loop b in OxyS enabled its regulation. In contrast, the mepS-OxyS interaction site is conserved between Salmonella and E. coli.

Hundreds of bacterial sRNAs with diverse metabolic functions have been identified so far. However, the number of sRNAs known to influence cell division, either directly or indirectly, is still scarce. One example is the E. coli prophage-encoded DicF sRNA that was discovered more than two decades ago as an inhibitor of cell division (Faubladier et al. 1990). DicF prevents the assembly of the septal ring structure by ftsZ by repressing the translation of ftsZ mRNA (Balasubramanian et al. 2016). The sRNA StsR of Rhodobacter sphaeroides is induced upon stress conditions and during the stationary phase by alternative sigma factors. Expression of StsR provides a regulatory link between cell division and environmental cues (Grützner et al. 2021). EcpR1 is an example of an sRNA that modulates the cell cycle. High-level expression of the sRNA EcpR1 in the plant-symbiotic Sinorhizobium meliloti has resulted in cell elongation (Robledo et al. 2015).

Exploring the RIL-Seq data for possible association between sRNAs and genes involved in cell division predicts several putative interactions between CyaR sRNA and genes of cell division, of which none were confirmed (Melamed et al. 2016). Unlike the above, a search for RprA sRNA partners led to the discovery that translation of cpoB mRNA encoding a cell division coordinator was reduced by 50% in the presence of RprA (Lalaouna et al. 2018).

Most sRNAs control the expression of multiple targets with complementary functions that together compose a mini-regulon. For example, the sRNA RyhB is specifically transcribed under iron-depleted conditions and regulates the expression of a large set of target mRNAs, many of which encode iron-containing and iron-storage proteins and are down-regulated by RyhB (Masse and Gottesman 2002; Massé et al. 2005). Likewise, GcvB sRNA controls the expression of a large number of functionally related mRNAs whose encoded proteins belong to the class of ABC transporters involved in amino acid metabolism including uptake and biosynthesis (Urbanowski et al. 2000; Sharma et al. 2007; Pulvermacher et al. 2009a,b,c; Sharma et al. 2011). In response to glucose-phosphate stress, SgrS sRNA down-regulates the expression of sugar transporters and up-regulates phosphatase yigL that dephosphorylates the accumulated sugars to facilitate their export (Vanderpool and Gottesman 2004; Rice and Vanderpool 2011; Papenfort et al. 2013; Sun and Vanderpool 2013). McaS, a midstationary phase sRNA regulated by carbon source availability, represses csgD, a key player in biofilm formation, and activates the expression of flhD and pgaA, affecting motility and biofilm, respectively, thus creating a mini-regulon that disfavors biofilm formation and promotes motility (Jørgensen et al. 2012; Thomason et al. 2012). Slightly different from the above, OxyS controls the expression of two linked functions where changes in one function require concomitant changes in the other in order for the cells to survive.

Finally, cell morphology images detected by fluorescence microscopy are striking. Cells in which cell division is inhibited while cell elongation is active display a unique fireball-like morphology. Truong and coworkers showed that filamentous ΔftsP cells overproducing MepS display half the number of normal Z-rings relative to wild-type cells carrying a control plasmid, thus interfere with the assembly of mature divisomes (Truong et al. 2020). The phase-contrast images of mutant strains defective for division with in trans overproduction of MepS are somewhat reminiscent of the images detected with overproduction of KilR or OxyS Δ15–33.

In total, 18 of the top 100 CopraRNA targets are predicted to interact via the OxyS 15–35 nt region in E. coli (Supplemental Table S2). Of those especially mltC (rank 5) is of interest. MltC is a membrane-bound lytic murein transglycosylase and mltC mutants grow as short chains of cells probably due to a defect in cell separation (Heidrich et al. 2002). The predicted mltC/OxyS interaction site ends ∼100 nt upstream of the start codon and could in theory comply with a positive regulation by OxyS. Whether these “fireballs” are formed as a result of an unfavorably high expression of MepS in the context of cell division inhibition or because of OxyS regulation of other targets, and what mechanism leads to their formation remains to be seen.

MATERIALS AND METHODS

Bacterial growth conditions

Escherichia coli and Salmonella cultures were grown at 37°C (200 rpm) in LB medium (pH 6.8). Ampicillin (100 µg/mL), tetracycline (10 µg/mL), chloramphenicol (20 µg/mL), and kanamycin (40 µg/mL) were added where appropriate. PlacO promoter was induced with isopropyl β-d-thiogalactoside (IPTG; 1 mM), and PBAD promoter was induced with 0.2% arabinose as indicated. Strains, plasmids and primers used in this study are listed in Supplemental Tables S3–S5.

Strain construction

To construct ΔoxySF::kan, the chromosomal region flanked by genome coordinates 4158287 and 4158395 (GenBank entry NC_000913.3) was replaced by the kan gene from pKD4 plasmid, using primers 3651 and 3652 as described (Datsenko and Wanner 2000). oxyS gene disruption was examined by PCR using flanking primers 2026 and 2027. To construct Salmonella (SL1344 ΔhisG::lacIq:Cm), a PCR fragment carrying PlacI-lacIq between AatII and XhoI and downstream from cm cassette was amplified using primers 2156 and 2157. The fragment was inserted into hisG46 mutant of Salmonella. Chromosomal scarless point mutations in mepS and oxyS were carried out as described (Li et al. 2013). Briefly, the tetA-sacB cassette from the XTL634 strain chromosome was amplified using the primer pairs 3425–3420 (oxyS Δ15–33), 3640–3641 (oxyS CC76,77UU), 3722–3723 [mepS (ΔP2)], and 3728–3729 [mepS (ΔP1)]. The PCR products were inserted into MG1655 mal::lacIq carrying pKD46 plasmid generating insertions of tetA-sacB in the selected genes. Next, the products of Gibson reactions carried out on two PCR fragments generated using primer pairs 3427–3428 and 3429–3430 to construct oxyS Δ15–33; 3427–3642 and 3643–3430 to construct oxyS CC76,77UU; 3726–3724 and 3725–3727 to construct mepS (ΔP2) as well as 3730–3726 and 3731–3727 to construct mepS (ΔP1) were used to transform the various tetA-sacB strains. Colonies sensitive to tetracycline were selected on fusaric acid-containing plates. The newly chromosomally generated scarless mutations were verified by sequencing. MepS-SPA fusion in the chromosome was constructed using primers (3751 and 3752) designed to amplify the SPA tag together with the kanamycin resistance cassette from plasmid pJL148 (Zeghouf et al. 2004). The PCR products were gel purified and then transformed into DY378 cells that were grown at 30°C to OD₆₀₀ of 0.5 then transferred to 42°C for 15 min. Insertions were confirmed by PCR using primers 3753 and 3691. The products were sequenced using primer 2227. The fusions were transferred into MG1655 mal::lacIq by P1 transduction.

Plasmid construction

To construct Plac-oxyS Δ15–33 and Plac-oxyS C18G G30C, whole plasmid PCR was carried out using Plac-oxyS plasmid as template and primers 3419–3420 and 3745–3746, respectively. To construct mepS-lacZ translation fusion pBOG552 (Supplemental Table S4; Hershko-Shalev et al. 2016), mepS fragments were PCR amplified using primers 3411–3412 and chromosomal wild-type and mepS mutants (ΔP1, ΔP2) as templates. The PCR products were then cloned into the EcoRI and BamHI sites of pBOG552. To construct mepS-C(-18)G-lacZ fusion, the EcoRI and BamHI fragment was inserted into pGEM3 plasmid and subjected to whole plasmid PCR using primers 3754–3755 of which 3754 carries C(-18)G. The mutated fragment was then inserted into pBOG552. To construct PBAD-kilR, kilR sequence was amplified from MG1655 chromosomal DNA using primers 3922 and 2701 and cloned into the PstI and HindIII restriction sites of pEF21 plasmid.

To construct Salmonella mepS and zapC-lacZ translation fusions, PCR fragments amplified using primers 3940–3941 and 3942–3943, respectively, were cloned into the EcoRI and BamHI sites of pBOG552. To construct PlacO-ftsZ-lacZ translation fusion, the PlacO PCR fragment amplified using primers 3319–3320 and pBR PlacO as template was cloned into EcoRI and KpnI. Salmonella ftsZ fragment was PCR amplified using primers 3952–3953 and cloned into the KpnI and BamHI sites.

β-Galactosidase assays

Overnight cultures of MG1655 mal::lacI^q, ΔlacZ::Tn10, carrying mepS-lacZ translational fusion Plac and Plac-oxyS plasmids (Supplemental Table S4), were diluted 1/100 in 10 mL LB medium containing ampicillin and kanamycin and grown to OD₆₀₀ of 0.1–0.2. Thereafter, the cultures were treated with IPTG (1 mM) to induce transcription of OxyS. β-Galactosidase activity was measured 60 min after IPTG induction. Cultures of Salmonella (SL1344 ΔhisG::lacI^q:cm) carrying lacZ fusion plasmids were induced by IPTG (1 mM) at dilution. Samples were taken 150 min after dilution.

Western of SPA-tagged mepS

Overnight cultures of SPA-tagged strain were diluted 1/100 and grown shaking (200 rpm) at 37°C in LB. H₂O₂ (1 mM) was added at OD₆₀₀ of 0.5 as indicated in the figure. Expression of OxyS was induced by 1 mM IPTG, 1 h after dilution for 30 min. Pellets were resuspended in 1× Laemmli sample buffer, heated at 95°C for 5 min and analyzed on SDS–PAGE (10%). The proteins were transferred to a nitrocellulose membrane (Bio-Rad) after blocking with BSA and skim milk as described in Basu and Altuvia (2021) and probed with FLAG M2 monoclonal antibody (Sigma-Aldrich) according to the manufacturer's protocol. The proteins were visualized using the secondary antibody Anti-Mouse IgG-HRP using Clarity Max Western ECL Substrate (BioRAD).

Northern analysis

RNA samples (10 µg) isolated from strains as indicated were denatured for 10 min at 70°C in 98% formamide loading buffer, separated on 6% acrylamide 8 M urea gels and transferred to Zeta Probe GT membranes (Bio-Rad laboratories) by electroblotting. To detect OxyS, the membrane was hybridized with end-labeled OxyS primer (3708) in modified CHURCH buffer (1 mM EDTA, pH 8.0, 0.5 M NaHPO₄, pH 7.2, and 5% SDS) for 2 h at 45°C and washed as previously described (Ben-Zvi et al. 2019). Tm RNA (10Sa) was used as a loading control (1912).

In vitro RNA synthesis

DNA templates for RNA synthesis: short mepS (139 nt) was generated using primers 3720–3721; long mepS (248 nt) was generated using 3719–3721; OxyS (109 nt) was generated using primers 2238–2027; OxyS Δ15–33 (94 nt) was generated using primers 2238–2027. The RNAs were synthesized in 50 µL reactions containing T7 RNA polymerase (25 units; New England Biolabs), 40 mM Tris–HCl (pH 7.9), 6 mM MgCl₂, 10 mM dithiothreitol (DTT), 20 units RNase inhibitor (CHIMERx), 500 µM of each NTP, and 200 ng of purified PCR templates carrying the sequence of the T7 RNA polymerase promoter. Synthesis was allowed to proceed for 2 h at 37°C followed by 10 min at 70°C. To remove the DNA template, 4 U of turbo DNase I (Ambion) was added (37°C, 30 min), followed by phenol/chloroform extraction and ethanol precipitation in the presence of 0.3 M ammonium acetate.

Primer extension

In vivo

Total RNA (30 µg) extracted using Tri reagent (Sigma) from strains as indicated was incubated with end-labeled mepS-specific primer (3412) at 70°C for 5 min, followed by 10 min in ice. The reactions were subjected to primer extension at 42°C for 45 min using 1 unit of MMLV-RT (Promega) and 0.5 mM of dNTPs. Extension products were analyzed on 6% acrylamide 8 M urea-sequencing gels next to pUC19-MspI labeled marker.

In vitro

Annealing mixtures containing in DEPC-treated water, 0.05 pmol of in vitro-synthesized mepS RNA, without or with 5 pmol of in vitro-synthesized OxyS RNAs (wild-type and mutants) were incubated for 15 min at RT. The mixtures were incubated for another 10 min at RT with 0.6 pmol of end-labeled mepS-specific primer (3740) in 20 mM Tris–HCl, 10 mM magnesium acetate, 0.1 M NH₄Cl, 0.5 mM EDTA, 2.5 mM β-mercaptoethanol, and 0.5 mM each dNTP upon which reverse transcriptase (Promega; 40 units) was added. cDNA synthesis was allowed to proceed for 7.5 min at 37°C. The extension products were separated on 6% acrylamide 8 M urea-sequencing gels alongside sequencing reactions.

Survival assays

Overnight cultures of MG1655 mal::lacI^q grown from fresh transformation plates were diluted 1/100 in LB supplemented with the appropriate antibiotics (20 mL in 125 mL flasks) and grown at 37°C (200 rpm). IPTG (1 mM) to induce OxyS and arabinose to induce KilR (0.2%) were added at the time of dilution. Samples to estimate CFU were taken at the indicated time points.

Mutagenesis assays

To estimate the number of rifampicin resistant mutants, ON cultures of wild-type and chromosomally encoded oxyS mutants were diluted 1/100 in LB and grown shaking at 37°C (200 rpm) to OD₆₀₀ of 0.1–0.2. Thereafter, cultures treated by 0.2 mM H₂O₂ (30 min, 200 rpm) were exposed to 10 mM H₂O₂ for 20 h (200 rpm). To determine frequencies of mutagenesis, aliquots were taken after 24 h and plated on LB plates containing 100 µg/mL of rifampicin. The numbers of Rif^r mutants were normalized to the numbers of viable cells at the 24-h time point.

Fluorescence microscopy

Escherichia coli overnight cultures carrying plasmids were subcultured in fresh LB medium supplemented with 1 mM IPTG and Amp. Cultures were grown at 37°C for 3 h before imaging. Salmonella cultures carrying Salmonella OxyS (Plac St-oxyS) were grown for 20 h in the presence of IPTG. For imaging, cultures were centrifuged and suspended in 20 µL of FM4-64 stain at a final concentration of 100 mg/mL. Samples were adhered using poly-l-lysine coated slides and photographed using Eclipse Ti2 microscope (Nikon), equipped with Prime BSI camera (Photometrics, Roper Scientific). System control and image processing were performed using NIS-Elements AR Analysis (version 5.30.05; Nikon).

Phylogenetically weighted sequence logo

The information content in the multiple sequence alignment (MSA) was used to judge the conservation of sRNAs and sRNA promoters. One thousand three hundred and forty OxyS homologs, including 150 nt upstream of the start of sRNA were aligned with MAFFT (Katoh and Standley 2013) (maxiterate 1000, localpair) and the resulting MSA was transformed into a count table storing the appearance of the four nucleotides at each alignment position. Gaps were not counted. To account for the nonuniform organism coverage for different phylogenetic groups, the CopraRNA weighting procedure based on 16S rDNA was used. Instead of counting any appearance of a character by 1, it was counted by the weight of the respective organism. The weighted count table was visualized with WebLogo3 (Crooks et al. 2004).

Phylogenetic tree

A 16S rDNA alignment was done with MAFFT (Katoh and Standley 2013) (retree 2, maxiterate 0), the distance matrix (F81 distance correction) and a maximum likelihood tree was calculated with the phangorn R-package (Schliep 2011). The Yersinia pseudotuberculosis YPIII (NC_010465) 16S rDNA was used as outgroup for rooting.

Comparative target prediction

For each of the 146 organisms, 14 organisms from their local phylogenetic neighborhood were selected based on the 16s rDNA-based distance matrix (Supplemental Fig. S2). For each set of 15 organisms, a target prediction with CopraRNA was done (Wright et al. 2013; Georg et al. 2020) using the Turner99 energy model.

Interaction site conservation and clustering

The homologous sRNAs and 5′ UTRs were aligned using MAFFT (Katoh and Standley 2013) (maxiterate 1000, localpair). IntaRNA 2.0 (Mann et al. 2017) was used to compute position-wise spot probabilities how likely a combination of positions is covered by an interaction for each sRNA/UTR pair. The resulting probability matrix was mapped to the sRNA and UTR alignments were weighted by the organism-specific CopraRNA weight summed up field-wise to get a combined interaction probability landscape. Peak areas, representing potential conserved interaction sites, were identified by the R “contourLines” function. The optimal and the first suboptimal IntaRNA predicted sRNA/UTR interactions for each organism were assigned to the identified peaks. Interactions assigned to the same peak share the same interaction regions in sRNA and UTR, but they might differ on sequence level or in the actually interacting nucleotide pairs. In the next step, all interactions assigned to the same peak are clustered based on their sequence identity and the positions of the interaction pairs. This step relies on a good local alignment of the subsequences covered by the peak area. Thus, a subalignment of the sRNA and UTR sequences is cut from the global alignment based on the peak coordinates and the resulting sequences are realigned with DIALIGN-TX (Subramanian et al. 2008).

Gene set enrichment analysis

Gene ontology annotations were extracted from UniProt (The UniProt Consortium 2023). For organisms not represented in UniProt, the E. coli annotations were used for homologous genes. The top 200 predicted targets for each organism were subjected to a gene set enrichment analysis with the topGO R-package (Alexa et al. 2006) using the Fisher test statistic. All terms with a P-value ≤0.2 were pooled for the conservation analysis. The number of terms with high semantic similarity was reduced with the GOSemSim R-package (https://academic.oup.com/bioinformatics/article/26/7/976/213143). For terms with a semantic similarity of ≥0.4 after the “wang” method and an overlap of enriched genes ≥50, only the more relevant term was kept. Relevance was scored based on the geometric mean of the information content and the normalized frequency of the term in the investigated organisms.

SUPPLEMENTAL MATERIAL

Supplemental material is available for this article.

MEET THE FIRST AUTHOR

Maya Elgrably-Weiss.

Meet the First Author(s) is an editorial feature within RNA, in which the first author(s) of research-based papers in each issue have the opportunity to introduce themselves and their work to readers of RNA and the RNA research community. Maya Elgrably-Weiss is the first author of this paper, “Balanced cell division is secured by two different regulatory sites in OxyS RNA.” Maya did her PhD with Professor Shoshy Altuvia at the Hebrew University Jerusalem, Israel. Her research focused on the identification of genes that are involved in the Salmonella typhimurium response to oxidative stress. In the following years, Maya continued to investigate the functions and the mechanisms of action of sRNA in Salmonella.

What are the major results described in your paper and how do they impact this branch of the field?

The most significant result in our paper is the finding that the two opposing physiological pathways, cell elongation and cell division, are regulated by a single sRNA molecule, OxyS. This regulation occurs simultaneously by the repression of two different genes through two separate regions in OxyS. Moreover, the balance obtained by this regulation is essential for the survival of the bacteria in the presence of hydrogen peroxide. In the OxyS Δ15–33 mutant, in which only cell division is prevented, this balance is violated, and the cells adopt a “fireball”-like morphology.

What led you to study RNA or this aspect of RNA science?

From the beginning, as an undergraduate student in the laboratory of Professor Shoshy Altuvia, I was fascinated by the idea that sRNA molecules enable a quick response to various stress conditions. To my mind, the possibility of studying both sRNA's physiological responses and their molecular mechanisms makes this field of research a fascinating one.

During the course of these experiments, were there any surprising results or particular difficulties that altered your thinking and subsequent focus?

The images of “fireball”-like morphology were a total surprise and unexpected. We expected the cells to elongate but never thought we would see a shape like the one we saw. It is still an intriguing puzzle.

Are there specific individuals or groups who have influenced your philosophy or approach to science?

In my academic journey, I encountered exceptional researchers from whom I learned a great deal. Particularly with my mentor, Professor Shoshy Altuvia, I experienced numerous scientific revelations, marked by enthusiasm, persistence, and a profound love for science. Collaborating with Dr. Jens Georg and his PhD student, Fayyaz Hussain, as well as with the PhD student, Bushra Shraiteh, who helped me with the microscopy, has been both enriching and inspiring. I carry a prayer for a more global collaboration in science as well as in general.

What are your subsequent near- or long-term career plans?

In the future, I plan to continue my research on sRNAs alongside teaching science in high school. I believe that by making science more approachable for young people and teaching them the love of science and the enthusiasm for it while inoculating analytical scientific thinking, we can lead to a better generation of future researchers.