Abstract
Microbial membranes are the first line of defense against environmental stress as well as the site of many metabolic processes. As such, maintenance of membrane integrity and homeostasis is key to cell survival. Altering membrane protein and lipid composition is an important strategy for maintaining membrane integrity in response to many environmental stresses. There are now numerous examples of small RNA (sRNA)-mediated regulation of membrane protein production, but less is known about how sRNAs regulate the types and relative proportions of different fatty acids in the membrane. The only sRNA known to regulate membrane fatty acid composition is RydC, which stabilizes cfa mRNA, encoding cyclopropane fatty acid (CFA) synthase, resulting in increased production of the synthase and higher levels of cyclopropane fatty acids (CFAs) in the cell membrane. Here, we report that three additional sRNAs, ArrS, CpxQ, and GadF also alter cfa translation and thus the amount of CFAs in cell membranes. RydC, ArrS, and GadF bind at sites that overlap a known RNase E cleavage site in the cfa mRNA 5′-untranslated region (UTR), resulting in increased cfa mRNA stability and translation. In contrast, CpxQ binds to a different site in the cfa mRNA 5′-UTR, and acts to reduce cfa translation. The physiological role of CFAs in membrane lipids is poorly understood, but CFAs have been shown to promote bacterial resistance to acid stress. We show that cfa translation increases in an sRNA-dependent manner when cells are subjected to mild acid stress. RydC is necessary for an increase in cfa translation at pH 5.0 and a rydC mutant is more sensitive to acid shock than the rydC+ strain. Alteration of membrane lipid composition is a key mechanism for bacterial responses to many environmental stresses, including acid stress. This work suggests an important role for sRNAs in these responses through their regulation of cfa mRNA.
Introduction
Bacteria modify the biophysical properties of their membranes to adapt to changing environmental conditions, such as pH, temperature, and pressure fluctuations (1). Membrane properties can be altered by changing the type or abundance of proteins embedded in the membrane or by modifying the relative proportions of different types of phospholipids. Membrane fluidity is a crucial biophysical property as it affects membrane associated functions such as permeability to solutes, solute transport, and protein- protein interactions. Because the length and saturation of the acyl chains in the phospholipids determines the fluidity of the membrane, bacteria can adjust the ratio of saturated to unsaturated fatty acids (UFA) to adapt to their environment. For example, bacteria can increase resistance to toxic compounds, such as antimicrobial peptides produced by the immune system, by increasing the production of saturated fatty acids, which will form a tightly packed and less fluid membrane compared to a membrane rich in unsaturated fatty acids.
While de novo production of new fatty acids is an important adaptation mechanism, sometimes bacteria are subjected to abrupt environmental changes and require rapid modification of the fatty acids already incorporated in the membrane. One post-synthetic modification is the conversion of a pre-exisiting UFA to a cyclopropane fatty acid (CFA) by the enzyme cyclopropane fatty acid synthase (encoded by cfa). CFAs occur in the phospholipids of many species of bacteria and have been linked to pathogenesis (2-⇓4). CFAs are formed by the addition of a methylene group, derived from S-adenosyl methionine, across the double bond of an UFA incorporated in a phospholipid (3, 5). Although widely studied, the physiological function of CFAs in phospholipids remains unclear. However, the conversion of UFAs to CFAs is energetically expensive, as it requires three ATP molecules to generate one S-adenosyl methionine (6). Given this energetic cost, it is expected that CFA formation, under some conditions, is advantageous. In fact, inactivation of a cfa homologue that introduces a cyclopropane ring in the major cell envelope component (α-mycolates) of Mycobacterium tuberculosis leaves the bacteria unable to establish a persistent infection (4). One hypothesis is that the formation of CFAs may reduce membrane fluidity and permeability, thereby preventing undesirable molecules from entering the cell (3). Chang and Cronan identified the first phenotype for CFAs, which is protection from acid shock (7, 8). Escherichia coli cfa mutants were unable to survive a rapid shift from pH 7 to pH 3. CFAs in membranes have been shown to protect bacteria from acid shock because the conversion of UFA to CFA decreases permeability of the membrane to protons (9). CFAs are thought to be important for pathogenesis because an acid shock such as rapid shift from pH 7 to pH 3 occurs as an ingested pathogen goes through the stomach (2). In fact, it has been noted that pathogenic E. coli strains contain more CFAs and are more resistant to acid than nonpathogenic E. coli strains (8).
Production of CFAs is regulated in part by control of cfa transcription. In E. coli and Salmonella, the cfa gene has two promoters (10, 11). The distal promoter is σ70-dependent and yields a longer cfa transcript with a 210-nt 5′-untranslated region (UTR). The proximal promoter is controlled by the general stress response σ factor (σs) encoded by rpoS and produces a shorter cfa transcript with a 33-nt 5′-UTR. The σ70-dependent promoter is functional throughout growth, whereas transcription from the σs promoter occurs only during stationary phase. CFAs are produced during early stationary phase when typically all phospholipid UFAs are converted to CFAs (10).
Recently, cfa mRNA was also shown to be post-transcriptionally controlled by the sRNA RydC (12). RydC is 64-nucleotides (nt) long and binds to the RNA chaperone Hfq (13). The sequence of RydC is conserved in many enteric species, particularly at the 5′ end, which contains nucleotides important for base pairing with cfa and other known target mRNAs (12). To date, RydC is known to regulate three mRNA targets, cfa (12), yejABEF (14), and csgD (15) mRNAs, though the conditions and signals stimulating RydC production and the physiological role of RydC in bacterial physiology remain unknown. RydC interaction with cfa mRNA stabilizes cfa mRNA, and in turn increases production of cyclopropane fatty acid (CFA) synthase (12). RydC specifically base pairs in the 5′-UTR of the longer isoform of cfa mRNA and prevents RNase E-dependent decay of the mRNA. In Salmonella, RydC overexpression increased the abundance of CFA synthase about 11-fold and increased levels of CFAs in membrane lipids about 4-fold but the physiological importance of this has yet to be elucidated (12).
A recent study on global mapping of sRNA-RNA interactions using in vivo crosslinking and sequencing identified 3 additional sRNAs (ArrS, GadF and CpxQ) that interact with cfa mRNA (16). Interestingly, these sRNAs were previously linked to cell envelope stress or acid response. The sRNA CpxQ is part of the Cpx two-component system that responds to cell envelope stress (17). CpxQ is processed from the 3′ UTR of cpxP mRNA, which encodes a chaperone of the Cpx system (17). Both the Cpx two-component system and the processing of CpxQ are induced when cell membrane potential is dissipated. Together with Hfq, CpxQ was shown to reduce translation of nhaB (a sodium-proton antiporter), agp (glucose-1-phosphatase), fimA (the major fimbrial subunit of type 1 pilus), and skp (a periplasmic chaperone) (17, 18). ArrS is an antisense sRNA that is encoded upstream of gadE, encoding the major acid resistance transcription factor, and is complementary to the 5′-UTR of the longest gadE transcript (19). ArrS expression is induced by low pH via σs and GadE. Overexpression of arrS leads to increased acid survival, compared to WT, in a GadE- dependent manner (20). GadF was identified as an sRNA located in the 3′-UTR of gadE mRNA (16). GadF has been detected as an sRNA via northern blot and is putatively processed from the gadE transcript by an RNase. To date, GadF has not been extensively studied.
Here, we sought to elucidate the mechanisms of regulation of cfa mRNA by these four sRNAs, RydC, ArrS, GadF, and CpxQ, and to gain insight into the physiological importance of sRNA-mediated regulation of membrane CFA composition in the model organism E. coli. Our results indicate that RydC, ArrS and GadF activate cfa translational fusions to varying degrees, but all three require a binding site in the 5′-untranslated region. RydC, ArrS and GadF all regulate cfa translation in a positive manner, and all three bind at overlapping sites that correspond to a previously identified RNase E recognition site. Our results are consistent with the idea that all three activate cfa by a similar mechanism that was reported previously for RydC, involving protection of cfa mRNA from RNase E-dependent turnover. In contrast, CpxQ represses cfa translation by a mechanism requiring sequences upstream of the RNase E cleavage site. While the physiological role for regulation of cfa by each sRNA is still unclear, our data implicate RydC-mediated activation of cfa in protection from acid stress.
Results
Four sRNAs regulate CFA synthase mRNA in an isoform-specific manner
Previous studies have characterized two transcription start sites for the cfa gene, encoding cyclopropane fatty acid (CFA) synthase in E. coli (10). Transcription from the distal promoter is σ70-dependent and yields a longer cfa transcript with a 212-nt 5′-UTR (Fig. 1A) that is subject to post-transcriptional regulation by sRNAs. The sRNA RydC binds to a site in the 5′-UTR and stabilizes cfa mRNA by inhibiting RNase E-dependent cleavage (12). A global study of sRNA-RNA interactions recently identified three additional sRNAs that bind to the 5′-UTR of the long cfa mRNA isoform: ArrS, GadF, and CpxQ (16). The base pairing interactions between the latter three sRNAs and cfa mRNA have not yet been characterized. To verify that these sRNAs alter cfa translation in vivo, we constructed two translational cfa‘-’lacZ fusions under the control of the arabinose-inducible PBAD promoter. One fusion has the full 212-nt 5′-UTR (called PBAD-cfa‘-’lacZ-Long, Fig. 1B), which contains the RydC binding site (12), while the other has only the 34-nt 5′-UTR (called PBAD-cfa’-’lacZ-Short, Fig. 1C). Base pairing predictions from IntaRNA (22) suggested that ArrS, GadF and CpxQ would base pair in the same region of cfa mRNA as RydC (Fig. 1A, “RydC BS”), so we hypothesized that these sRNAs would alter the translation of the PBAD-cfa’-’lacZ-Long but not the PBAD-cfa’-’lacZ-Short. When the individual sRNAs were ectopically expressed from a plasmid, all sRNAs affected PBAD-cfa’-’lacZ-Long activity (Fig. 1B) and had no effect on PBAD-cfa’-’lacZ-Short activity (Fig. 1C). RydC, ArrS, and GadF activated the long fusion 18-, 10-, and 1.3-fold, respectively, while CpxQ repressed the long fusion ~3-fold. These results indicate that cfa mRNA can be post-transcriptionally activated or repressed by multiple sRNAs.
Post-transcriptional regulation of cfa mRNA by sRNAs changes membrane lipid composition
CFA synthase forms CFAs by transferring a methylene group from S-adenosyl methionine to the double bond of an unsaturated fatty acid (UFA) of a mature phospholipid that is already incorporated in the membrane. Specifically, palmitoleic acid (C16:1) is converted into methylene-hexadecanoic acid (C17CFA) and vaccenic acid (C18:1) is converted into methylene-octadecanoic acid (C19CFA). To test how post-transcriptional regulation of cfa by sRNAs alters the proportion of CFAs in membrane fatty acids, we conducted gas chromatography on membrane lipids isolated from strains where each sRNA was ectopically expressed (Figure 2, Table S1). The FA composition of all strains was similar except for CFA and UFA content (Table S1). In strains carrying PIac-cfa, there were reduced levels of 16:1 and 18:1 UFAs and ~6-fold increase in levels of C17CFA, compared to the strain carrying the vector control. Similarly, ArrS-producing cells had ~5-fold higher levels of C17CFA and reduced 16:1 and 18:1 UFAs compared to vector control. RydC-producing cells likewise had reduced levels of 16:1 and 18:1 UFAs and ~8-fold increased levels of C17CFA compared to the control strain. GadF production resulted in a slight increase in C17CFA (1.5-fold higher than vector control), while CpxQ-producing strains had modestly reduced C17CFA (~2-fold reduced compared to vector control). These data show that regulation of cfa mRNA translation by each sRNA is correlated with changes in membrane CFA content, implying that the regulation by sRNAs could contribute to meaningful changes in cell membrane structure and function.
Role of sRNAs in surviving acid shock
Because CFAs have been implicated in resistance to acidic pH (7, 8), we next investigated whether sRNAs promote survival after acid shock. Survival of wild-type, Δcfa, ΔrydC, ΔarrS, ΔgadF and ΔcpxQ strains and complemented mutants was measured after rapid shift of cultures from pH 7 to pH 3 as described in Materials and Methods. Survival was determined by the ratio of CFUs recovered after acid shock compared to CFUs before acid shock (Fig. 3). Compared to the wild-type strain, the ΔrydC, and Δcfa strains had lower survival rates after the acid shock. The Δcfa mutant had the lowest survival rate of all the strains tested (Δcfa 1.0 ± 0.3 compared to WT 4.3 ± 0.5). The ΔrydC strain had a survival rate intermediate between that of Δcfa and wild-type (ΔrydC, 2.5 ± 0.4) suggesting that RydC might be important for activating cfa or other targets to promote resistance to acid shock. The ΔarrS, ΔgadF and ΔcpxQ strains showed the same survival rates as wild-type, implying that these three sRNAs do not play an important role in regulating cfa or other targets to promote acid shock resistance under these conditions.
Complementation of the cfa, rydC and arrS mutations by expression of the corresponding gene from a plasmid resulted in enhanced survival of acid shock compared to the wild-type strain. Enhanced acid resistance for ArrS-producing strains was observed previously and was attributed to ArrS positively regulating gadE expression (20). Our data suggest that ArrS-dependent activation of cfa might also contribute to enhanced acid resistance of ArrS-producing cells. The complemented gadF and cpxQ strains were not different from wild-type (Fig. 3). We have shown so far that ArrS, GadF, and RydC all activate cfa translation and increase the CFA content in cell membranes but only deletion of rydC renders otherwise wild-type cells more susceptible to acid shock. These results suggest that at least RydC-mediated activation of cfa translation promotes cell survival during an acid shock.
Small RNA-dependent regulation of cfa at acidic pH
To further investigate sRNA-dependent regulation during acid stress, we examined sRNA-dependent regulation of cfa translational fusions at neutral and acidic pH. Cells carrying PBAD-cfa’-’lacZ-Long (Fig. 4A), PBAD-cfa’-’lacZ-Short (Fig. 4B), or a fusion containing a deletion of the 5′ region of the cfa 5′-UTR containing the sRNA binding sites (PBAD-cfa’-’lacZ-ΔsRNABS, Fig. 4C) were grown in TB medium pH 7.0 to early exponential phase then subcultured into TB medium at either pH 7 or at pH 5. Samples were harvested 120 minutes later and assayed for β-galactosidase activity. For PBAD-cfa’-’lacZ-Long, activity was higher at pH 5 compared to pH 7 (Fig. 4A). Activity of PBAD-cfa’-’lacZ-Short was similar at both pH 5 and pH 7 (Fig. 4B). Likewise, PBAD-cfa’-’lacZ-ΔsRNABS had similar activity at pH 5 and pH 7 (Fig. 4C). These observations indicate that activation of cfa mRNA in response to acidic pH occurs post-transcriptionally, and the loss of regulation of a fusion lacking the sRNA binding sites suggested that one or more sRNAs could be responsible for regulation under these conditions.
To further probe the effects of sRNA regulation of cfa mRNA at acidic pH, the experiment was performed in wild-type background and strains with mutations in each sRNA (Fig. 5). Activity of PBAD-cfa’-‘lacZ-Short was not affected by changes in pH or by deletion of any of the sRNAs (Fig. S1). For PBAD-cfa’-‘lacZ-Long, wild-type, ΔarrS, and ΔgadF strains all had similar activities at pH 7 and pH 5, with clearly increased activity at pH 5 (Fig. 5A). The ΔcpxQ mutant had higher cfa’-’lacZ activity at pH 7 compared to wild-type (Fig. 5A). Since ectopic expression of cpxQ repressed cfa translation (Fig. 1B), the increased levels of cfa’-’lacZ activity in the ΔcpxQ mutant suggests that at neutral pH, CpxQ is produced at sufficient levels to repress cfa. The ΔcpxQ mutation had no effect on PBAD-cfa’-’lacZ-Long at pH 5, where levels were similar to wild-type. The ΔrydC mutant had lower cfa’-’lacZ activity at both pH 7 and pH 5 compared to wild-type, suggesting that RydC may be produced at sufficient levels at both neutral and acidic pH to have an activating effect on cfa translation under both conditions. We have shown that ArrS, GadF, and RydC all activate cfa translation and increase the CFA content in cell membranes but only deletion of rydC affects PBAD-cfa’-’lacZ-Long activity at pH 5.0. This observation fits with our previous result that deletion of rydC renders otherwise wild-type (cfa+) cells more susceptible to acid shock (Fig. 3).
To further investigate the interplay of the sRNAs on cfa mRNA translation at different pH values, we deleted rydC in combination with every other sRNA in the PBAD-cfa’-’lacZ-Long strain (Fig. 5B) and PBAD-cfa’-’lacZ-Short (Fig. S1B) and measured β-galactosidase activity of the reporter fusions at pH 7 and pH 5. Deletions of each sRNA in the ΔrydC background had no effect on PBAD-cfa’-’lacZ-Short activity (Fig. S1B). Deletion of arrS or gadF in the ΔrydC background had no effect on PBAD-cfa’-’lacZ-Long activity compared to the ΔrydC parent (Fig. 5B), suggesting that neither ArrS nor GadF play a role in regulation of cfa translation under these conditions.
Compared with the Δrydc parent, the ΔcpxQ ΔrydC strain had slightly higher PBAD-cfa’-‘lacZ-Long activity at pH 7 (Fig. 5B). In contrast, at pH 5, the Δrydc parent and ΔcpxQ ΔrydC strain had similar levels of PBAD-cfa’-’lacZ-Long activity (Fig. 5B). Altogether, the data are consistent with the hypothesis that RydC exerts a positive effect on cfa mRNA at both neutral and acidic pH under our growth conditions. In contrast, CpxQ has a mild repressive effect on cfa mRNA only at neutral pH.
RpoS is not responsible for observed differences in cfa mRNA regulation
The σs-dependent cfa promoter was previously implicated in increased CFA levels when cells were grown at pH 5 compared to pH 7 (7). PBAD-cfa’-’lacZ-Long contains region encompassing the σs dependent promoter, so it was possible that RpoS could be regulating cfa transcription in the context of the PBAD-cfa’-’lacZ-Long during mild acid stress. To determine if RpoS impacts regulation of cfa under our conditions, we measured PBAD-cfa’-’lacZ-Long activity at pH 5 and 7 in a Δrpos background (Fig. 6A). Deletion of rpoS had no effect on PBAD-cfa’-‘lacZ-Long activity at pH 5 or 7 and we observed the same increase in activity in response to pH 5 in both wild-type and Δrpos backgrounds, indicating that RpoS does not impact the observed activity of the PBAD-cfa’-‘lacZ-Long fusion.
We showed that ectopic production of each sRNA, RydC, ArrS, GadF, and CpxQ can regulate PBAD-cfa’-‘lacZ-Long (Fig. 1B). While the experiment above suggests that pH-dependent regulatory effects are not mediated by RpoS, we wanted to further test whether sRNAs could be acting indirectly on cfa via post-transcriptional regulation of rpoS. We ectopically expressed CpxQ, RydC, ArrS, or GadF in wild-type and Δrpos PBAD-cfa’-‘lacZ-Long strains and did not observe any differences in sRNA-dependent regulation between wild-type and ΔrpoS (Fig. 6B), indicating that the sRNA-dependent regulation is not mediated indirectly through RpoS. These data further support the idea that the regulation of cfa by these sRNAs is direct.
Activating and repressing sRNAs bind distinct sites on cfa mRNA
RydC, ArrS, GadF, and CpxQ all regulated the cfa translational fusion containing the cfa 212-nt 5′-UTR but not the cfa translational fusion containing only the 34-nt 5′-UTR suggesting that each sRNA base pairs within the −212 to −34 (relative to the start of translation) region of cfa mRNA. The mechanism of RydC regulation of cfa mRNA was shown to be stabilization via pairing with an RNAse E recognition site to prevent cleavage (12). To determine the mechanism of regulation by ArrS, GadF, and CpxQ, we used previously published data paired with computational methods to predict each sRNA binding site (Fig. 7A). To genetically test these base pairing predictions, we first constructed a translational fusion with a 133-nt deletion from the beginning of the 212-nt 5′-UTR, which contains predicted binding sites for all sRNAs (PBAD- cfa’-‘lacZ-ΔsRNABS, Fig. 7B), including the RNase E recognition site that RydC protects. Of note, the PBAD-cfa’-‘lacZ-ΔsRNABS fusion had a higher basal level of activity compared to PBAD- cfa’-‘lacZ-Long (Fig. 7C compared to Fig. 1B), presumably because the PBAD-cfa’-‘lacZ-ΔsRNABS fusion lacks the known RNase E cleavage site at position 109-nt (12). When each individual sRNA was ectopically expressed from a plasmid, none of the sRNAs affected PBAD- cfa’-‘lacZ-ΔsRNABS activity (Fig. 7C), suggesting each sRNA binds at a site upstream of −79.
To test base-pairing predictions with each sRNA more specifically, we did a mutational analysis. RydC, GadF, and ArrS are all predicted to bind the same region of cfa mRNA (Fig. 2A, Fig. 3A-C). To investigate the importance of these interactions, we introduced three point mutations in each sRNA that were predicted to disrupt base pairing to PBAD-cfa’-‘lacZ-Long (C7G, G8C, A9U for arrS1 (Fig. 8A), C26G, G28C, A29U for gadF1 (Fig. 8B), C5G, G6C, A7U for rydC6 (Fig. 8C)). These mutated sRNA variants, arrS1, gadF1 and rydC6, could no longer activate PBAD-cfa’-‘lacZ-Long (Fig. 8D). We made the compensatory mutations (U103A, C102G, G101C) to create PBAD-cfa’-‘lacZ-LongAGC, to restore putative base pairing interactions with the mutant sRNAs (Fig. 8A-C, E). Wild-type ArrS, GadF, and RydC could not activate the mutant PBAD-cfa’-‘lacZ -LongAGC fusion (Fig. 8E), consistent with the idea that these residues in the cfa 5′-UTR are critical for interactions with all three sRNAs. In contrast, all three mutants: RydC6, GadF1, and ArrS1, activated the PBAD-cfa’-‘lacZ-LongAGC fusion (Fig. 8E), with fold-activation restored to levels similar to the wild-type mRNA-sRNA pairs (Figs. 1B, 8D).
To further characterize CpxQ-mediated regulation of cfa mRNA, we deleted the first 22-nt of the 212-nt 5′-UTR (−212 to −190, called PBAD-cfa’-‘lacZ-ACpxQBS, Fig. 9A) which includes the putative CpxQ binding site (Fig. 7A). This fusion was insensitive to CpxQ, suggesting that CpxQ-dependent repression requires a site within the −212 to −190 region of cfa 5′-UTR. Interestingly, PBAD-cfa’-‘lacZ-ΔCpxQBS had lower basal activity compared to PBAD-cfa’-‘lacZ-Long, suggesting that the −212 to −190 region of the 5′ UTR cfa mRNA may play a role in cfa mRNA structure and stability. Based on these data, we hypothesize that ArrS and GadF regulate cfa by the same mechanism as RydC, stabilizing the cfa mRNA transcript by impairing RNAse E-mediated decay (12). CpxQ-mediated repression must occur via a different mechanism that has yet to be determined.
Discussion
Small RNAs (sRNAs) are important post-transcriptional regulators that play key roles in the response to environmental stress. We are interested in how sRNAs can contribute to membrane modification in response to stressors, particularly how sRNAs can lead to altered phospholipid composition. In the present work, we investigate the roles of multiple sRNAs in regulating the translation of cfa mRNA, encoding a key enzyme used to alter lipid composition of bacterial membranes. We show that the cfa gene can be post-transcriptionally regulated by four sRNAs, all of which act on the same long isoform of cfa mRNA that results from transcription from the constitutive σ70-dependent promoter. The sRNAs RydC, ArrS, and GadF activate cfa translation resulting in increased membrane CFA levels, while the sRNA CpxQ represses cfa translation leading to reduced membrane CFA levels. Our data suggest that one of the physiological functions of sRNA-mediated regulation of cfa mRNA is protection of cells from acid stress. Loss of rydC made cfa+ cells less resistant to acid shock (Fig. 3), which was consistent with the observation that loss of RydC reduced cfa translation at both neutral and acid pH (Fig. 5A,B). Acid stress phenotypes were also apparent for strains ectopically expressing ArrS (Fig. 3), which may be due to both ArrS-mediated activation of cfa and gadE.
Previous studies determined that the RydC-cfa mRNA base pairing prevents RNase E-mediated decay and stabilizes the mRNA to allow increased translation (12). We hypothesize, because ArrS and GadF base pair with the same region of cfa mRNA as RydC, that ArrS and GadF would regulate cfa mRNA stability via a similar mechanism. However, CpxQ must be repressing cfa translation by a different mechanism. CpxQ has been shown to repress other mRNA targets using one of two conserved seed regions (17). One of these seed regions matches the region of CpxQ predicted to base pair with cfa mRNA. CpxQ represses its other targets by base pairing near the Shine-Dalgarno sequence to prevent ribosome binding and directly inhibit translation initiation, or by base pairing within the coding region and stimulating mRNA decay by RNase E (17). Neither of these mechanisms is analogous to regulation of cfa by CpxQ. We observed that when we deleted the putative CpxQ binding site from the PBAD-cfa’-‘lacZ-Long fusion, the basal level of activity of the fusion decreased. This may indicate that the CpxQ base pairing site is required for cfa mRNA stability, leading to increased turnover of cfa mRNA when the site is deleted. We propose that CpxQ base pairing with cfa mRNA may also stimulate mRNA turnover. The known RNase E recognition site protected by RydC far downstream of the putative CpxQ binding site (Fig. 7A). It is possible that CpxQ binding upstream promotes a structural rearrangement that makes that site more sensitive to RNase E. Future work will analyze the full structure of the cfa 5′-UTR in the presence and absence of CpxQ, and determine whether CpxQ-mediated regulation of cfa depends on RNase E.
Regulation of cfa mRNA by multiple sRNAs suggests that under some condition(s) these sRNAs influence membrane lipid composition and membrane integrity to promote fitness or stress resistance. Towards elucidating the physiological roles of RydC, ArrS, GadF, and CpxQ in the regulation of membrane composition, we investigated the role of each sRNA in acid stress, a condition where CFA synthase is known to play a role (7, 8). We determined that cfa translation was higher when cells were exposed to pH 5 compared to pH 7. When rydC was deleted, cfa translation could not reach maximum wild-type levels at pH 5 and Δrydc mutants were more susceptible to acid shock compared to wild-type, giving us the first phenotype associated with RydC. However, it is not currently known if activation of cfa at acidic pH is the major physiological role of RydC, as the conditions and signals stimulating RydC production are not yet known.
While ectopic expression of arrS increased cfa translation and promoted higher levels of acid resistance, deletion of arrS did not affect cfa translation or acid shock survival, indicating that ArrS may not play a role acid stress under the conditions tested. ArrS has been characterized as an activator of the transcriptional activator GadE and ectopic expression of ArrS increases acid resistance though its activation of gadE but a deletion phenotype has not been reported (19, 20). We propose that ArrS may activate cfa under another condition where CFAs in the membrane contribute to resistance, such as heat or pressure. Currently, the only know stimulus of ArrS transcription is acid (19).
Expression of GadF only slightly increased cfa translation and a gadF deletion did not affect cfa translation or acid shock survival, indicating that GadF also does not play a role in acid stress under the conditions tested. GadF is relatively uncharacterized but was shown to activate acrB 12-fold compared to the reported 2-fold activation of cfa (16). AcrB is an efflux pump that uses proton motive force to pump compounds, such as antibiotics, organic solvents, and detergents, out of the cell (23). GadF regulation of cfa mRNA may be relevant under these conditions, as CFA membrane content has been shown to promote resistance to organic solvents (24). Understanding the regulation of GadF itself would provide insight into the importance of its regulation of cfa mRNA, however, it is not currently known how gadE mRNA (from which GadF is derived) is processed or in response to what stimuli.
CpxQ is the only sRNA of the four we have characterized that represses cfa translation. Deletion of cpxQ resulted in higher cfa translation at pH 7 compared to wild-type, but did not affect cfa translation at pH 5. Responding to acid stress is likely not the relevant physiological role for CpxQ, since deletion of cpxQ did not affect acid shock survival. However, our observation that CpxQ represses cfa translation at neutral pH combined with what is known about CpxQ may provide some insight. CpxQ is part of the Cpx two-component system that responds to cell envelope stress (17, 18). Both the Cpx two-component system and the processing of CpxQ are induced when cell membrane potential is dissipated (17). CpxQ was shown to reduce translation of four inner membrane proteins including nhaB, a sodium-proton antiporter (17). When the cytoplasm becomes basic, NhaB can pump in 3H+ in exchange for 2Na+ out. However, if NhaB pumps in too many H+, the inner membrane potential is dissipated and CpxQ is induced to repress nhaB translation (17). We predict that CpxQ would also repress cfa translation under this condition because CFAs in the inner membrane are known the decrease membrane permeability to H+ and in a situation where H+ concentration is too high in the cytoplasm, H+ could passively cross the inner membrane to get out of the cytoplasm. Determining how and under what conditions CpxQ differentially regulates its regulon would elucidate the link between inner membrane stress and acid stress.
The UFA to CFA conversion by CFA synthase is energetically expensive, so it is not surprising that CFA synthase production is tightly controlled at the transcriptional level and at the post-transcriptional level by multiple sRNAs. Regulation of cfa mRNA by multiple sRNAs may have many advantages. There are a few examples of multiple sRNAs acting on the same target. Post-transcriptional control of the important transcription factor RpoS (σs) by multiple sRNAs to mediate responses to diverse conditions is well documented (25). The rpoS gene is post-transcriptionally regulated by at least four sRNAs in response to different conditions. sRNA regulation of rpoS highlights the potential importance of regulation of the same mRNA target by multiple sRNAs: synthesis of the sRNAs are induced by different stresses which allows RpoS in turn to properly respond to these various stress signals. The stimulus that causes synthesis of ArrS, CpxQ, GadF, and RydC would give more insight into how and why cfa mRNA is regulated by multiple sRNAs. It is known that transcription of arrS is induced by acid stress and that CpxQ is processed in response to cell membrane potential dissipation, but how that relates to their regulation of cfa is not understood. The regulation of rydC itself has not been elucidated, and it is not currently known how gadE is processed or in response to what stimuli.
Regulation of cfa mRNA by multiple sRNAs may provide an additional mechanism to deal with acid stress beyond that of RpoS and the σS-dependent promoter of cfa. The σ70-dependent promoter, which yields the longer cfa transcript, is functional throughout growth, whereas transcription from the σs promoter occurs only during stationary phase (10). Thus, the cfa transcript from the σ70-dependent promoter would potentially be in the cell available to be stabilized by RydC if the pH becomes acidic. Consistent with this hypothesis, we found that transcription from the σ70 dependent promoter does not increase upon exposure to mild acid (data not shown) but that an sRNA produced under those conditions can stabilize the already present transcripts.
Previous work has extensively shown that sRNAs play an integral role in changing membrane protein composition in response to changing environments but how sRNAs influence membrane fatty acid composition is not well studied. Here, we build on the first example of an sRNA altering fatty acid composition by showing regulation of cfa by four sRNAs. Currently, the sRNAs studied here are the only examples of sRNAs altering the fatty acid composition of membranes, which can consequently alter the stability and permeability of the membrane.
Materials and Methods
Strain and plasmid construction
Strains and plasmids used in this study are listed in Table S2. All strains used in this study are derivatives of E. coli K-12 DJ624 (D. Jin, National Cancer Institute). Oligonucleotide primers used in this study are listed in Table S3. Integrated DNA Technologies synthesized the primers. ΔrpoS and Δcfa mutations were made via P1 vir transduction from the Keio collection (26). All other chromosomal mutations were made using λ Red recombination (27, 28) and marked alleles were moved between strains by P1 vir transduction (29). Kanamycin markers were removed using pcp20 (27).
PBAD- cfa’-‘lacZ -Long, PBAD- cfa’-‘lacZ -Short, PBAD- cfa’-‘lacZ -ΔsRNABS, and PBAD- cfa’-‘lacZ -ΔCpxQBS translational lacZ reporter fusions under the control of the PBAD promoter were constructed by PCR amplifying the cfa 5′UTR fragment of interest using primers containing 5′ homologies to PBAD and lacZ (Table S3). PBAD- cfa’-‘lacZ -Long and PBAD- cfa’-‘lacZ -Short were made by Alisa King (King et al. 2018). PBAD- cfa’-‘lacZ -LongAGC was made using Gibson Assembly (30). pBR322 was amplified using primers with 5′ homologies to the cfa 5′UTR containing the three point mutations. Gibson reaction was performed using NEBuilder HiFi-DNA Assembly Master Mix (New England Biolabs) according to the manufacture’s protocol. The Gibson product was transformed into XL10 competent cells and the plasmid was purified. The mutated cfa 5′UTR fragment of interest was PCR amplified from this plasmid using primers containing 5′ homologies to PBAD and lacZ (Table S3). PCR products were recombined into PM1205 using λ Red homologous recombination and counter-selection against sacB as previously described (31).
Plasmids containing CpxQ, ArrS, RydC, GadF, Cfa under the control of the PLlacO promoter were constructed by PCR amplifying each gene from E. coli MG1655 chromosomal DNA using oligonucleotides containing HindIII and BamHI restriction sites (Table S3). PCR products and vector pBRCS12 (32) were digested with HindIII and BamHI (New England Biolabs) restriction endonucleases. Digestion products were ligated using DNA Ligase (New England Biolabs) to produce the plasmids.
PLlaco-arrSI, PLlaco-gadF1, and PLlaco-rydC6 were created using QuikChange mutagenesis (Agilent Technologies) using primers in Table S3.
Media and growth conditions
Bacteria were cultured in Luria-Bertani (LB) broth medium or on LB agar plates at 37°C, unless stated otherwise. When necessary, media were supplemented with antibiotics at following concentrations: 100 μg ml-1 ampicillin (Amp) or 25 μg ml-1 kanamycin (Kan). Isopropyl β-D-1-thiogalactopyranoside (IPTG) was used at 0.1 mM for induction of expression from PLlacO-1 promoter.
β-Galactosidase assays
Ectopic expression of sRNAs
Bacterial strains harboring translational lacZ reporter fusions carrying an empty, Plac- arrS, Plac- cpxQ, Plac- gadF, or Plac- rydC plasmid were cultured overnight in Terrific Broth (TB) with Amp and 0.002% L-arabinose then subcultured 1:100 to fresh TB medium containing Amp and 0.002% L-arabinose. Strains were grown at 37°C with shaking to early exponential phase then 0.1 mM IPTG was added to induce sRNA expression. Samples were harvested 60 minutes later and β-galactosidase assays were then performed as previously described (29).
Acid
Bacterial strains harboring translational lacZ reporter fusions were cultured overnight in TB medium pH 7.0 with 0.002% L-arabinose then subcultured 1:100 to fresh TB medium pH 7.0 containing 0.002% L-arabinose. Strains were grown at 37°C with shaking for 2 hours then strains were subcultured 1:100 again into TB medium with 0.002% L-arabinose at pH 7.0 or pH 5.0. Samples were harvested 120 minutes later and β-galactosidase assays were then performed as previously described (29).
Analysis of fatty acids
Cells were cultured overnight in LB containing Amp and subcultured 1:100 to fresh LB containing Amp. Cells were grown to an OD600 of 0.1 then plasmids were induced with 0.1M IPTG for one hour. Cells were then spun down and the cell pellet was washed twice with MilliQ water. Fatty acids were extracted according to a previously described protocol, method 3.1 (33). Pentadecanoic acid methyl ester was used as an internal standard. The bacterial sample was spiked with 500 μg of the internal standard, pentadecanoic acid, and then the sample was transesterified with 2 mL of 0.5M sodium methoxide for 1 min at room temperature. Fatty acid methyl esters (FAMEs) were then extracted using 2 mL of hexane. The solution was centrifugation at 2000 rpm for 5 min and the organic upper phase was removed and dried under nitrogen. FAMEs were treated with trimethylsilyldiazomethane in methanol to ensure complete methylation. FAMEs were identified on an Agilent 6890N GC with a 5973 MS and a Zebron WAX column (30m × 0.25mm × 0.25pm).
Acid Shock Assay
Strains were grown overnight in LB with Amp then subcultured 1:100 into LB with Amp at pH 7.0 and grown at 37°C with shaking until OD600= 0.1. Plasmids were induced with 0.1mM IPTG. When strains reached an OD600 of 0.2, cultures were diluted 10x into LB pH 3.0 and incubated at 37°C with no shaking for 60 minutes. Survival was determined by the ratio of CFUs on the LB plates after acid shock to CFUs on the LB plate before acid shock.