Natural variation of a sensor kinase controlling a conserved stress response pathway in Escherichia coli

Manuela Roggiani; Srujana S. Yadavalli; Mark Goulian

doi:10.1371/journal.pgen.1007101

Abstract

Previous studies have shown that exponentially growing Escherichia coli can detect mild acidity (~pH 5.5) and, in response, synthesize enzymes that protect against severe acid shock. This adaptation is controlled by the EvgS/EvgA phosphorelay, a signal transduction system present in virtually every E. coli isolate whose genome has been sequenced. Here we show that, despite this high level of conservation, the EvgS/EvgA system displays a surprising natural variation in pH-sensing capacity, with some strains entirely non-responsive to low pH stimulus. In most cases that we have tested, however, activation of the EvgA regulon still confers acid resistance. From analyzing selected E. coli isolates, we find that the natural variation results from polymorphisms in the sensor kinase EvgS. We further show that this variation affects the pH response of a second kinase, PhoQ, which senses pH differently from the closely related PhoQ in Salmonella enterica. The within-species diversification described here suggests EvgS likely responds to additional input signals that may be correlated with acid stress. In addition, this work highlights the fact that even for highly conserved sensor kinases, the activities identified from a subset of isolates may not necessarily generalize to other members of the same bacterial species.

Author summary

Bacteria employ a class of proteins, sensor kinases, to sense environmental cues and initiate cellular responses through phosphorylation of partner response regulator proteins. Individual kinases are generally assumed to have the same sensory activity across members of a bacterial species. In this work, we report an unexpected counterexample in which the well-established capacity of the kinase EvgS to sense mild acidity is limited to a subset of Escherichia coli isolates. Despite this natural variation, EvgS activation still confers resistance to acid stress in strains that have lost EvgS pH-sensing activity. Thus, most E. coli share a conserved output of the Evg system but do not require identical sensory functions. This work highlights the potential for significant functional divergence of a sensor kinase within a species and also indicates that there are additional input signals for the highly conserved EvgS protein.

Citation: Roggiani M, Yadavalli SS, Goulian M (2017) Natural variation of a sensor kinase controlling a conserved stress response pathway in Escherichia coli. PLoS Genet 13(11): e1007101. https://doi.org/10.1371/journal.pgen.1007101

Editor: Patrick H. Viollier, University of Geneva Medical School, SWITZERLAND

Received: May 24, 2017; Accepted: November 3, 2017; Published: November 15, 2017

Copyright: © 2017 Roggiani et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: All relevant data are within the paper and its Supporting Information files.

Funding: This work was supported by National Institutes of Health, grant GM080279. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Competing interests: The authors have declared that no competing interests exist.

Introduction

The species Escherichia coli comprises a remarkably diverse collection of bacteria, reflecting their capacity to colonize and manipulate disparate in vivo and ex vivo niches. Most of the well-documented phenotypic differences between E. coli strains are associated with genes found in only a subset of isolates [1]. Genes that are ubiquitous or nearly ubiquitous across the species, on the other hand, are generally assumed to have the same function in each cell type. However, polymorphisms in these core genes could have significant effects on the activities of the proteins that they encode and contribute to natural variation across the species. Furthermore, highly conserved networks of interacting proteins can be perturbed by components that are not encoded in all strains. Thus, the properties of a conserved regulatory circuit may depend on the E. coli isolate and be quite different from the properties established in the well-studied laboratory strain, E. coli K-12. Here we report an unexpected example of such natural variation in the EvgS/EvgA phosphorelay, a two-component system that has been identified in virtually all E. coli isolates.

The E. coli EvgS/EvgA phosphorelay is at the top of a pathway associated with acid and drug resistance [2–10]. Studies of this system in E. coli K-12 indicate that the sensor kinase EvgS is stimulated by mild acidity (pH 5.5–5.7), possibly via the protein’s periplasmic domain [9–11], resulting in EvgS autophosphorylation and subsequent phosphoryl transfer to the response regulator EvgA [12]. Phosphorylated EvgA regulates transcription of a number of genes, including the safAydeO operon, which is a node for two branches of the Evg network (Fig 1A) [6]. The transcription factor YdeO is a key component of the glutamate-dependent acid resistance network AR2 [2, 7, 8, 13] (for comprehensive reviews on E. coli acid resistance (AR) systems see [14–16]) that upregulates the activator GadE, leading to increased expression of AR2 effector genes. SafA encodes a small membrane protein that activates the sensor kinase PhoQ, thereby connecting the EvgS/EvgA and PhoQ/PhoP two-component signaling systems (Fig 1A) [17, 18]. PhoQ is stimulated by conditions of low divalent cations (Mg⁺⁺ and Ca⁺⁺) and antimicrobial peptides [19, 20]. In Salmonella, PhoQ is also stimulated directly by low pH [21], but in E. coli pH stimulation of PhoQ is indirect via SafA [8]. PhoQ controls the phosphorylation state of the response regulator PhoP, which in turn regulates transcription of a large regulon that includes genes associated with acid resistance. In exponential phase cells, PhoP contributes to AR2 by indirectly elevating RpoS levels [22], which contributes to the expression of the central regulator GadE as well as downstream effectors (Fig 1A) [22]. Thus, EvgS is believed to contribute to acid resistance via both the SafA and YdeO branches of the pathway.

Download:

Fig 1. The Evg acid resistance network.

A: Simplified diagram of the regulatory pathway leading to acid resistance in exponential phase-minimal medium cultures of E. coli. Thick arrows represent genes, with black denoting regulators, and hatched arrows indicate genes whose promoters were used as reporters in this study. Gray rectangles represent two-component system proteins, black ovals and diamonds represent other types of regulators. For clarity, only representative AR effector-encoding genes are shown, and genes of the PhoP regulon are omitted. B: Locus in MG1655 containing safA and ydeO (depicted as dark grey arrows) that is missing in E. coli MP1. Black vertical lines represent the boundaries of the region missing in MP1. Figure not drawn to scale.

https://doi.org/10.1371/journal.pgen.1007101.g001

The evgAevgS operon is found in almost all of the fully-sequenced E. coli genomes currently in the NCBI database. In contrast with this almost universal conservation among E. coli, close orthologs of evgA and evgS have not been identified in other bacterial species, including other species within the Escherichia genus. In addition, the safAydeO operon has a similarly high level of conservation in E. coli and, like evgAevgS, has not been identified in other species.

Recently, we noticed that safAydeO is missing in E. coli MP1, a mouse commensal isolate [23], suggesting key links in the Evg network may be severed in this strain. This observation was the starting point for the work presented here. While establishing the effects of these missing genes, we determined that EvgS is unresponsive to pH in MP1, as well as in many other E. coli isolates, despite the high conservation of the Evg network. We also show that the divergence is due to natural variation in the EvgS sequence. In addition, we find that low pH activation of the PhoQ/PhoP system is similarly variable across E. coli.

Results

E. coli MP1 lacks two branches of the EvgS/EvgA pathway

E.coli strain MP1 lacks a 13 kb segment of DNA containing the genes safA and ydeO that mediate two branches of the EvgS/EvgA AR2 pathway (Fig 1A and 1B). To explore the physiological effects of this disruption on the Evg network, we monitored the transcription of a PhoQ/PhoP-regulated promoter (P_mgrB) [24], [25] and a GadE-regulated promoter (P_hdeA) [6, 26] using fluorescent protein fusions. The reporter constructs were integrated in the chromosome at ectopic sites, leaving the native loci undisturbed.

The EvgS/EvgA two-component system is activated by mild acidity (pH 5.5–5.7) in glucose minimal medium [7, 8, 10]. In addition, the signaling cascade can be initiated with a constitutively active EvgS variant, EvgS1, that has the amino acid substitution F577S [4, 6, 22]. Both mild acidity and the presence of the EvgS1 allele activated mgrB and hdeA transcription in the standard laboratory strain E. coli K-12 (MG1655). However, neither condition had an effect on transcription of these genes in E. coli MP1 (Fig 2A). Transcription from the mgrB promoter in MP1 was not stimulated over a range of acid pH values (5.1–7) and was similar to the behavior of MG1655 ΔsafA (S1A Fig). To rule out the possibility that the PhoQ/PhoP system itself was compromised in MP1, we verified that mgrB transcription is activated by low Mg²⁺ in a similar fashion in both MG1655 and MP1 (S1B Fig). These results are consistent with the observation that safA and ydeO are absent in the MP1 genome and that no other proteins in MP1 perform equivalent functions.

Download:

Fig 2. Two branches of the EvgS/EvgA pathway are interrupted, and acid resistance in exponential phase cultures is severely impaired in MP1.

A: Activation of fluorescent reporters. Strains MMR175, MMR180, MP138, MP139 (top panel, P_hdeA-yfp reporter) and strains TIM63, MMR170, MP131, and MP140 (bottom panel, P_mgrB-yfp reporter) were grown in minimal medium at pH 5.7 (induced wild-type strains) and pH 7 (evgS1 strains and non-induced wild-type strains) to OD₆₀₀ ~0.2. Cultures of the P_mgrB-yfp reporter strains contained 10 mM Mg⁺⁺ so that PhoQ activation was fairly low in the absence of pH induction. Fluorescence was quantified as described in Materials and methods. Values represent the average fluorescence from two representative experiments and error bars represent the range. B: Survival after acid challenge at pH 2.5 for various times. Strains MG1655 and MP1 were cultured in minimal medium at pH 5.7 and pH 7 (MG1655 only) to OD₆₀₀ ~0.2. Cultures were shocked in LB at pH 2.5, and aliquots were withdrawn at the indicated times. Values are the average percent survival from two representative experiments and error bars represent the range.

https://doi.org/10.1371/journal.pgen.1007101.g002

The absence of two branches of the Evg network suggests that MP1 may be less acid resistant than other E. coli isolates. We therefore grew MP1 to exponential phase in minimal medium at pH 5.7, to induce the Evg system, and then shocked the cells at pH 2.5 in rich medium, as described previously [13]. After one hour, MP1 survival was over 3 orders of magnitude lower than the survival of MG1655 (Fig 2B). This difference emerged quite rapidly, with MP1 showing a 40-fold reduction compared to MG1655 by 5 minutes following the transition to pH 2.5. The acid shock sensitivity of MP1 in these growth conditions was comparable to that of MG1655 cultures in exponential phase at neutral pH, a condition for which the EvgS/EvgA two-component system is inactive (Fig 2B) [2, 7, 8, 13, 27]. In contrast, MP1 and MG1655 in stationary phase withstand acid challenge equally well, regardless of pre-exposure to mildly acidic pH (S2A Fig). These observations confirm that inducible exponential phase acid resistance is impaired in MP1, consistent with the absence of safAydeO.

Restoring safAydeO in MP1 is not sufficient for low pH activation of GadE and PhoQ

To determine whether the safAydeO operon from MG1655 (Fig 1B) would restore the interrupted EvgS-PhoQ and EvgS-GadE pathways in MP1, we introduced this segment of DNA into MP1 on a single copy plasmid (psafAydeO_MG1655), see Materials and methods. Comparison of the acid resistance of MP1 carrying either the empty vector or psafAydeO_MG1655 revealed that restoration of safAydeO did not increase survival of MP1 (Fig 3A). However, psafAydeO does complement a ydeO deletion in MG1655 and fully restores acid resistance in this strain. In addition, transduction of a segment of DNA that includes the full 13 kb region missing in MP1 (Fig 1B) failed to rescue acid resistance (S2B Fig).

Download:

Fig 3. EvgS does not respond to mild pH induction in MP1.

A: Resistance to acid challenge. Strains MG1655 or its ΔydeO derivative (MMR239), and MP1, with the empty vector pSMART or plasmid psafAydeO_MG1655 (pMR76) were cultured in minimal medium at pH 5.7 to OD₆₀₀ ~0.2, and shocked for an hour at pH 2.5 in LB as described in Materials and methods. Percent survival values are the average of three independent experiments and error bars represent the standard deviations of the means. B: Activation of fluorescent reporters. Strains MP142 and MP143 (top panel, P_mgrB-yfp reporter) and strains MP138 and MP139 (bottom panel, P_hdeA-yfp reporter), each carrying the empty vector pSMART or its derivative psafAydeO_MG1655 (pMR76) were cultured in minimal medium at pH 5.7 (induced wild-type strains) and pH 7 (evgS1 strains and non-induced wild type strains) to OD₆₀₀ ~0.2 and fluorescence was quantified as described in Materials and methods. Fluorescence values are the average of two representative experiments and error bars represent the range. C: As in (B), but with strains MMR173, MMR179, MP136, and MP137 (P_emrK-yfp reporter).

https://doi.org/10.1371/journal.pgen.1007101.g003

We also found that MP1/ psafAydeO_MG1655 did not activate transcription of the PhoQ/PhoP and GadE reporters P_mgrB-yfp and P_hdeA-yfp, respectively, in response to low pH (Fig 3B). In contrast, the constitutively active EvgS_MG1655 variant EvgS1 activated expression of both reporters when psafAydeO_MG1655 was present. These results indicate that the SafA-PhoQ and YdeO-GadE interactions were successfully restored by psafAydeO_MG1655 in MP1 and that components upstream of safA and ydeO involved in pH sensing are divergent between MP1 and MG1655.

It is also noteworthy that in MG1655, neither safA nor phoQP deletions affected acid resistance, even after prolonged exposure to low pH, in contrast with the behavior of a ΔevgAS strain (S2C Fig). These results indicate that for exponential phase cultures, the SafA-PhoQ-PhoP branch of the Evg network does not provide increased protection to acid shock following induction at pH 5.7.

The MP1 EvgS/EvgA system is not induced by low pH

Based on the above results, we hypothesized that low pH might not function as an input signal for the EvgS/EvgA phosphorelay in MP1. We therefore measured pH induction of the emrK promoter, which is directly regulated by EvgA [4] (Fig 1A). We found that transcription was induced in MG1655, as expected, but not in MP1 (Fig 3C, S3 Fig). In contrast, the constitutively active EvgS1 mutant was able to induce the emrK reporter in MP1. The failure of low pH to activate EvgS in MP1 could be due to differences between the evgAS operon in MG1655 and in MP1, or due to an upstream factor required for acid-sensing that is missing or non-functional in MP1. To explore these possibilities, we compared the pH induction of P_emrK-yfp in MP1 ΔevgAS and MG1655 ΔevgAS transformed with single-copy plasmids expressing the evgAS operon from one or the other strain (pevgAS_MG1655 or pevgAS_MP1). We found that pevgAS_MG1655 restores pH induction of P_emrK-yfp in both MG1655 ΔevgAS and MP1 ΔevgAS (Fig 4A). In contrast, pevgAS_MP1 shows minimal pH induction in either strain, although pevgAS_MP1 in MG1655 does show a small amount of induction, suggesting that there may be some factors outside of the evgAevgS operon that contribute to pH sensing. Overall, however, the above results indicate that the primary differences in pH response for the EvgS/EvgA systems in MG1655 and MP1 are due to differences in the sensor kinase and/or response regulator proteins themselves. Furthermore, since the EvgA amino acid sequences from MG1655 and MP1 are identical whereas the EvgS sequences differ at 39 residues, the different pH response in the two strains is likely due to polymorphisms in EvgS. This conclusion is further supported by the fact that a plasmid expressing a hybrid operon consisting of evgA_MP1evgS_MG1655 restores pH induction of P_emrK-yfp in MG1655 ΔevgAS (S4 Fig).

Download:

Fig 4. EvgAS_MG1655 is sufficient to restore response to mild acidity in MP1.

A: Activity of the fluorescent reporter P_emrK-yfp. Strains with wild type evgAS (MMR182 and MP146) or ΔevgAS (MMR191 and MP162), and with the empty vector (pSMART) or plasmids pevgAS_MG1655 (pMR78) or pevgAS_MP1 (pMR84) were cultured in minimal medium at pH 5.7 or pH 7 to OD₆₀₀ ~0.2, and fluorescence was quantified as described in Materials and methods. Values are the average fluorescence from two representative experiments and error bars represent the range. B: Resistance to acid challenge. Wild type or ΔevgAS strains derivatives of MG1655 (MMR182 and MMR191) and of MP1 (MP146 and MP162), and with the empty vector pSMART or plasmids pevgAS_MG1655 (pMR78), psafAydeO_MG1655 (pMR76), or pevgAS_MG1655-safAydeO_MG1655 (pMR128) were cultured in minimal medium at pH 5.7 to OD₆₀₀~0.2, and shocked for an hour at pH 2.5 as described in Materials and methods. Percent survival values are the average of two representative experiments and error bars represent the range.

https://doi.org/10.1371/journal.pgen.1007101.g004

The above results identify two properties of MP1 that potentially affect acid resistance: the chromosomal deletion containing safAydeO (Fig 1B) and differences in EvgS between MP1 and MG1655. To determine whether these factors account for the absence of inducible exponential phase acid resistance in MP1, we tested the survival of an MP1 derivative containing one or both of these loci from MG1655 following acid shock. Incorporation of both loci from MG1655 into MP1 (MP1 ΔevgAS / pevgAS_MG1655 -safAydeO_MG1655) rescued the inducible acid resistance phenotype by two orders of magnitude compared to MP1 with either plasmid psafAydeO_MG1655 or plasmid pevgAS_MG1655 (Fig 4B). These results indicate that in addition to the absence of the chromosomal segment containing safAydeO, sensitivity to acid shock in MP1 results from the inability of EvgS_MP1 to respond to stimulation by mild acidity. We also note that the amino acid substitution F577S, which renders the EvgS_MG1655 allele constitutively active (EvgS1), causes the same effect in EvgS_MP1 (S5 Fig). This finding supports the hypothesis that although EvgS_MP1 is expressed and functional, the protein cannot sense pH change.

The 39 residues in EvgS that differ between MG1655 and MP1 are distributed throughout the protein (S6 Fig). In an attempt to determine if a subset of these residues that are localized to a particular domain account for the pH insensitivity of EvgS_MP1, we tested the activity of several EvgS hybrids containing swapped regions of EvgS_MP1 and EvgS_MG1655 (S4A Fig). These constructs were expressed from a single copy plasmid in a MG1655 ΔevgS strain, and activation of the EvgA-dependent reporter P_emrK-yfp in response to acid stimulation was assessed. We found that all of the hybrids showed a strong pH-response (S4B Fig), indicating that pH insensitivity of EvgS_MP1 cannot be ascribed to a single domain of EvgS_MP1.

Natural variation in pH sensing by the EvgS/EvgA phosphorelay among E. coli

Strains MG1655 and MP1 belong to different phylogenetic groups: A and B2, respectively [23]. We therefore wondered whether the properties noted above are unique to MP1 or are shared by other E. coli isolates. We considered eight representative strains (Table 1), which include commensals of group A and B2 (HS and Nissle respectively), intestinal pathogens (H10407, EDL933, and E2348/69), extra-intestinal pathogens (CFT073 and UTI89), and an “atypical” E. coli isolate classified in Clade I and of enterotoxigenic pathotype (TW10509). EvgS amino acid sequences from these strains have varying degrees of divergence from the MG1655 sequence (Table 1, S6 Fig), and a tree based on these sequences clusters according to each strain’s phylogenetic group (Fig 5A). A similar analysis that includes 285 EvgS sequences from fully sequenced E. coli genomes indicates that clustering according to the phylogenetic group is a general characteristic (S7 Fig).

Download:

Fig 5. Analysis of EvgS and its sensitivity to mild acidity in selected E. coli isolates.

A: Dendrogram based on EvgS sequence alignments was constructed as described in Materials and methods. Gray shaded boxes indicate phylogenetic groups. B: Activity of the EvgA fluorescent reporter P_yfdX-yfp. All of the strains carry a single copy plasmid with the EvgA reporter P_yfdX-yfp_MG1655 (pMR86). The strains were cultured in minimal medium at pH 5.7 or pH 7 to OD₆₀₀ ~0.2, and fluorescence was quantified as described in Materials and methods. Values are the average fluorescence from two experiments and error bars represent the range. All of the pH 7 data bars, and the pH 5.7 data bars for some strains, are too small to be visible in the figure. C: Resistance to acid challenge. Strains were cultured in minimal medium at pH 5.7 to OD₆₀₀ ~0.2, and shocked for an hour in LB at pH 2.5 as described in Materials and methods. Percent survival values are the average of two experiments and error bars represent the range. D: Plot of fold induction of the fluorescent reporter P_yfdX-yfp from panel B versus percent survival after exposure to pH 2.5, from panel C. The E. coli strains are described in detail in Table 1. Dashed lines represent the limit of detection of the assay.

https://doi.org/10.1371/journal.pgen.1007101.g005

Download:

Table 1. Characteristics of isolates used in this study.

https://doi.org/10.1371/journal.pgen.1007101.t001

Among the ten isolates, pairwise EvgS divergences are as large as 5.43%, and no two strains share 100% EvgS amino acid sequence identity (S8 Fig). This natural variation in EvgS is not a property of proteins encoded in neighboring genes: the EvgA amino acid sequence is identical across all ten strains, and YfdE, which is encoded by a gene just downstream of evgS, shows only a moderate level of divergence (S8 Fig). Additionally, the variation is not a general property of hybrid sensor kinases: two other hybrid kinases in E. coli, ArcB and BarA, are highly conserved among the ten isolates considered in this study (S8 Fig).

Based on the substantial sequence variation in EvgS, we hypothesized that the Evg systems in different E. coli strains would show varying responsiveness to acid pH that would be correlated with the degree of divergence from EvgS_MG1655. To test this hypothesis, we assessed the pH induction of EvgS/EvgA in each strain using a single-copy plasmid containing a transcriptional fusion of yfp to the yfdX promoter (P_yfdX-yfp), which is directly activated by phosphorylated EvgA [6]. We found that for group B2 strains, EvgS is not responsive to acid pH, with the exception of the EPEC strain E2348/69 (Fig 5B). The EvgS sequence from this isolate has the fewest substitutions (relative to MG1655) within the group B2 strains that we tested (Table 1). For strains outside the B2 clade, reporter expression varied from a 60-fold induction for HS, whose EvgS sequence is the closest to that of MG1655, to 8-fold induction in isolate TW10509.

We also compared the survival from acid shock for the various isolates following growth at pH 5.7 (Fig 5C). The pattern across the E. coli strains shows some correlation with EvgS sequence relatedness. Isolates in group A that were tested have comparably high resistance to acid challenge whereas the group B2 isolates Nissle, UTI89, and MP1 are quite sensitive. The association is imperfect however, as the two B2 isolates E2348/69 and CFT073 had significantly higher survival, albeit both were still more susceptible to acid shock than the group A strains (Fig 5C). Interestingly, CFT073 is very sensitive to slightly harsher acid shock (pH 2.25), as shown in S9A Fig. There is also a correlation between the extent of activation of the EvgS/EvgA system from growth at pH 5.7 (as assessed with a yfdX transcriptional reporter) and resistance to acid shock across the strains (Fig 5D). CFT073 is an exception to this trend for acid shock at pH 2.5, suggesting that this isolate has an exponential phase acid resistance pathway that does not require activation of the Evg system for this stress. However this strain clusters with the other closely related Group B2 isolates when shocked at pH 2.25 or lower pH (S9B Fig).

Above we showed that low pH fails to directly activate PhoQ/PhoP in MP1 (Fig 2A). Since we found that EvgS is not stimulated by mild acidity for three other group B2 isolates that we tested (Nissle 1917, CFT073, UTI89), we wished to determine if the PhoQ/PhoP system in these strains could be activated by low pH. To test this, we used a GFP reporter plasmid containing the PhoP-activated phoPphoQ promoter [25] and measured fluorescence of exponential phase cultures at pH 5.7 relative to pH 7. We found that like MP1, the three additional B2 strains that we tested showed no induction of the PhoP reporter (Fig 6), consistent with the conclusion that EvgS in these isolates is unresponsive to pH and that E. coli PhoQ is not directly stimulated by mild acidity (S1A Fig) [8].

Download:

Fig 6. The PhoQ/PhoP system in group B2 isolates is not activated by mild acidity.

Strains (in the same order as in the figure) MG1655, TIM136, MP1, MP200, Nissle 1917, CFT073, UTI89, HS containing a reporter plasmid with a PhoP-regulated promoter (P_phoPQ-gfp) were cultured in minimal medium at pH 7 and pH 5.7, with 10 mM MgSO₄, to OD₆₀₀ ~0.2. Fluorescence of the reporter was measured as described in Materials and methods, and reported as fold induction from pH 7 to pH 5.7. Values are the average from two representative independent experiments. Error bars represent the range.

https://doi.org/10.1371/journal.pgen.1007101.g006

evgAS from MG1655 confers resistance to acid shock in acid-sensitive isolates

The strains Nissle 1917 and UTI89 are similar to MP1 in their sensitivity to acid shock and inability to activate the Evg system in response to low pH (Fig 5). However, unlike MP1, these strains (as well as all of the other E. coli isolates that we tested) have an intact safAydeO locus. We therefore tested whether EvgS from MG1655 is sufficient to restore acid resistance in these strains by transforming each with the plasmid pevgAS_MG1655. For both Nissle 1917/ pevgAS_MG1655 and UTI89/ pevgAS_MG1655, growth in mild acidity induced expression of the P_yfdX-yfp reporter (Fig 7A). Likewise, in these strains EvgS_MG1655 can activate the GadE-dependent reporter P_hdeA-yfp (YdeO-dependent pathway) and the PhoQ/PhoP-dependent reporter P_mgrB-yfp (SafA-dependent pathway) (S10 Fig), although the extent of activation is lower than that of MG1655. In addition, with the pevgAS_MG1655 plasmid, both Nissle 1917 and UTI89 were as resistant to acid shock as MG1655 (Fig 7B). Thus, the pathway leading from mild pH induction to exponential phase acid resistance for strains UTI89 and Nissle 1917can be rescued with the evgAevgS operon from MG1655.

Download:

Fig 7. EvgAS_MG1655 is sufficient to restore response to mild acidity in E. coli Nissle 1917 and E. coli UTI89.

A: Activity of the fluorescent reporter P_yfdX-yfp. Strains MMR178 (MG1655 P_yfdX-yfp), MMR227 (Nissle 1917 P_yfdX-yfp), and MMR228 (UTI89 P_yfdX-yfp), carrying the empty vector pSMART or pevgAS_MG1655 (pMR78) were cultured in minimal medium at pH 5.7 or pH 7 to OD₆₀₀~0.2, and fluorescence was quantified. Values are the average fluorescence from two representative experiments and error bars represent the range. B: Resistance to acid challenge. The same cultures as in (A) were shocked for an hour in LB at pH 2.5 as described in Materials and methods. Percent survival values are the average of two representative experiments and error bars represent the range. Dashed line represents the limit of detection of the assay.

https://doi.org/10.1371/journal.pgen.1007101.g007

Natural variation in evgS sequence is consistent with weak purifying selection

Since EvgA is identical across all of the E. coli strains used in this study, the above results suggest that the strain variability in pH response is due to differences in EvgS. We therefore analyzed the nucleotide sequence evolution of evgS. To test functional conservation across evgS orthologs, we calculated the dN/dS ratio, which is a measure of selection on protein-coding sequences [28, 29]. For any given gene, dN/dS is defined as the ratio of the average number of nucleotide substitutions per non-synonymous site (dN) to the average number of substitutions per synonymous site (dS). A dN/dS ratio that is less than one indicates purifying or stabilizing selection, dN/dS equal to one indicates neutral selection, and dN/dS greater than one indicates positive or adaptive selection.

To estimate dN/dS, evgS sequences were aligned using TranslatorX [30] and analyzed with the Synonymous Non-synonymous Analysis Program (SNAP v2.1.1) [31]. The dN/dS ratio calculated for the ten E. coli isolates is less than one (0.1), indicating that evgS is under purifying selection and, overall, is functionally conserved (Table 2, S4 Table).

Download:

Table 2. Sequence evolution of evgS based on dN/dS analysis.

https://doi.org/10.1371/journal.pgen.1007101.t002

To assess whether different clades experience varying selective pressures, we also calculated dN/dS for evgS within phylogenetic groups A and B2 separately. The group A isolates have a dN/dS ratio of 0.32, which is significantly higher than that of the group B2 isolates as well as the overall ratio for all ten organisms (~0.1). In addition, the Z-test for neutral selection (MEGA 4.0, [32]) gives a p-value = 0.2 for group A evgS. We therefore cannot reject the null hypothesis that dN = dS, suggesting that evgS from group A is under weaker purifying selection compared to that of B2. We observed a similar trend when we analyzed only the periplasmic region of evgS (1–537, 87–234) (Table 2, S4 Table).

We extended our dN/dS analysis to the two genes that flank evgS in the E. coli genome (evgA and yfdE) and also two other hybrid sensor kinases (barA and arcB) (Table 2, S4 Table). In contrast with evgS, these four genes appear to be under strict purifying selection, with similar values for dN/dS for all ten isolates together and for groups A and B2 separately.

To check if the dN/dS statistics estimated here are consistent across other E. coli, we calculated the dN/dS values for evgS as well as control genes for all complete E. coli genomes in the NCBI database. The resulting values closely match those obtained for the ten isolates (Tables 2, S4 and S5 Tables).

The EvgS/EvgA system in E. coli is homologous to the virulence-associated BvgS/BvgA system in Bordetella pertussis [33] and KvgS/KvgA system in Klebsiella pneumoniae [34]. We therefore tested whether these orthologs showed similar levels of variation as that of EvgS/EvgA. We repeated the dN/dS analyses for the bvgAbvgS and kvgAkvgS genes using the available complete sequences from B. pertussis and K. pneumoniae, respectively (S5 Table). For bvgAbvgS, the sequences from different isolates are highly invariant and showed very few non-synonymous or synonymous changes, if at all. This result is consistent with previous observations [35], and suggests that bvgAbvgS may either be recently acquired by B. pertussis and/or that the selective pressures to maintain the sequences of bvgAbvgS are strong, which is not surprising given the significant role of BvgS/BvgA system in virulence regulation.

In contrast with BvgS, the variation for K. pneumoniae kvgS is similar to that of E. coli evgS (S5 Table). However, whereas E. coli evgA shows very little variation relative to evgS, the variation of K. pneumoniae kvgA is comparable to that of kvgS, suggesting weaker selection for kvgA relative to evgA. In addition, the flanking gene yfdX in K. pneumoniae has a high level of variation as well, and the variations of all three genes kvgA, kvgS, yfdX are substantially higher than those of K. pneumoniae arcB and barA.

Discussion

It is well established that the EvgS/EvgA system in the laboratory strain E. coli K-12 is stimulated by moderately acidic pH and that this response enables exponentially growing cells to resist severe acid shock [7, 8]. As shown here, this behavior extends to additional E. coli isolates—namely those in the same phylogenetic group as the K-12 strain, group A, as well as isolates from several additional groups (Fig 5). However, for other E. coli isolates, the Evg system is much less responsive, or not responsive at all, to acid pH. Furthermore, this behavior correlates with a decreased ability to survive extreme acid shock following growth in mild acidity (Fig 5). For the three strains with the largest difference in acid pH response relative to that of E. coli K-12—MP1, Nissle 1917, UTI89—we found that the primary differences in pH-sensing capacity are intrinsic to the EvgS/EvgA system itself, rather than arising from accessory proteins or other factors that vary between cells. For two of these strains (Nissle 1917 and UTI89), transformation with a K-12 EvgS/EvgA system led to acid resistance that was comparable to K-12. Thus, the suite of genes required to confer protection from acid stress is intact in these strains and is under control of the Evg system, despite the fact that the native system cannot be activated by mild acidity. In contrast, for the mouse commensal strain MP1, activation of EvgS/EvgA was not sufficient to provide protection from acid shock due to the absence of a chromosomal segment encoding the EvgA-regulated transcription factor YdeO and the connector protein SafA. Restoration of this chromosomal region from K-12, in combination with the K-12 EvgS/EvgA proteins, led to exponential phase acid stress resistance comparable to that of K-12. The absence of the safAydeO region in MP1 indicates that the Evg network has been significantly reduced, since YdeO is a key component of this network due to its control of GadE, which in turn regulates many downstream genes involved in acid resistance [6–8, 13, 26, 36]. In addition, since SafA functions as a connector between the Evg and PhoQ/PhoP systems, its absence disconnects the two pathways, abrogating activation of the PhoQ/PhoP system by signals that stimulate EvgS. However, we find that this branch of the Evg pathway does not play an important role in exponential phase acid resistance, at least for the acid stress assays that we employed.

Our results indicate that the diversity in pH response of the Evg system stems from natural variation in the EvgS sensor kinase. Furthermore, a dN/dS analysis indicates that evgS is under purifying selection, but that the selection is significantly weaker than that of the hybrid sensor kinase genes arcB and barA. This observation is consistent with the fact that the Evg system is broadly conserved in E. coli but that the ability of pH to function as an input signal is highly variable across isolates.

The pH-sensing mechanism of EvgS (for those EvgS variants that have this capacity) is not known. An analysis of EvgS mutants in MG1655 or related K-12 strains has implicated the periplasmic and cytoplasmic domains as playing a role in the pH response [9–11], although it remains to be established whether the sensing is direct or requires additional cellular components. Our own analysis of hybrid molecules in which domains of EvgS were substituted by analogous portions of the pH insensitive EvgS_MP1 suggests that differences across multiple domains account for the absence of pH sensing in MP1. These observations are consistent with a recently proposed model of EvgS pH sensing in which pH modulates the strength of EvgS dimerization mediated by interactions in the periplasmic, transmembrane, and cytoplasmic domains [11]. In addition, from the EvgS amino acid sequence alignments (S6 Fig), we were unable to identify specific residues that likely account for the differences in pH-responsiveness of EvgS natural variants. For example, EvgS_E2348/69, which is stimulated by low pH and is in group B2, differs from the four group B2 EvgS proteins that are not stimulated by pH at residues S382 (in the periplasmic domain), N859 (in the histidine kinase domain), and A1191 (at the C-terminus). However, strain E2348/69 is an exception since these residues are conserved across all of the nine other strains in this study. These observations indicate that multiple independent polymorphisms account for variability in pH sensing by EvgS.

The diversification of the Evg two-component system within E. coli suggests that pH is not the only input signal for the EvgS sensor kinase, and that the primary selective pressure for maintaining this signaling system may be associated with some other (unknown) stimulus. The fact that for some strains with a pH-nonresponsive EvgS the EvgA regulon still confers acid resistance further suggests the unknown input signal is strongly correlated with conditions of acid stress. The emergence of these EvgS variants may reflect selection for different dose-response behaviors with respect to other signals. However, since some isolates such as MP1 have evolved a network that disconnects acid resistance effectors from the Evg regulon (through loss of safAydeO), it seems likely that Evg input signals are not always correlated with acid stress.

The presence of the EvgS/EvgA system in the vast majority of E. coli isolates indicates there is a significant selective pressure to maintain this system, which makes the natural variation and genetic flexibility of EvgS all the more striking. This behavior stands in stark contrast with the common assumption that sensor kinases and response regulators of conserved two-component systems behave uniformly across a species (or even across closely related genera). Similar diversification may be present in other core signaling systems and poses a challenge for extrapolating from well-studied members of a bacterial species.

Materials and methods

Growth media and conditions

Liquid cultures were grown at 37°C in minimal A medium [37] supplemented with 0.2% glucose and 0.1% casamino acids, or in minimal A medium buffered at pH 5.7 with 100 mM 2-(N-morpholino)ethanesulfonic acid (MES, Sigma-Aldrich) and HCl. Minimal medium cultures of strain UTI89, which is auxotrophic for nicotinamide, and its derivatives were supplemented with 5 μg/ml nicotinamide. Bacterial cultures that were used to prepare electro-competent cells were grown in SOB with the appropriate antibiotic, when necessary. Cultures for preparing P1_vir lysates and for transductions were grown in LB (Miller) broth. LB-agar plates were used to grow cultures for CFU (colony forming unit) counts. To select for antibiotic resistance and to maintain plasmids, antibiotics were added to culture media to the following concentrations: ampicillin 50 μg/ml; for MG1655 and its derivatives—kanamycin 25 μg/ml, chloramphenicol 25 μg/ml or 12.5 μg/ml for single copy plasmids (pSMART derivatives); for all other strains—chloramphenicol 12 μg/ml or 6 μg/ml for single copy plasmids (pSMART derivatives); kanamycin 50 μg/ml for Nissle 1917 and its derivatives, 35 μg/ml for all other strains.

Strains and plasmids

Strains and Plasmids used in this study are described in S1 and S2 Tables, respectively. Primers used in this study are listed in S3 Table. Transformations of plasmids and linear DNA for chromosomal integration were performed by electroporation. Antibiotic cassettes flanked by FRT sites were removed, when necessary, with the plasmid pCP20 as described in [38]. Transductions were conducted with phage P1_vir [37]. For details on strain and plasmid constructions, see S1 Methods.

Fluorescence quantification

Fluorescence was quantified by microscopy. Cultures were inoculated from single colonies on LB Agar plates and grown in minimal medium aerobically at 37°C overnight to saturation, then diluted 1:1000 into fresh medium that was at pH 7 or pH 5.7 (as indicated). To test PhoQ/PhoP activation via stimulation of EvgA and the connector SafA and to avoid stimulation by low magnesium, overnight cultures were diluted in minimal medium with 10 mM MgSO₄. Conversely, to activate PhoQ/PhoP with low magnesium, overnight cultures were diluted in minimal medium at pH 7 and containing 1 μM MgSO₄. Cultures were grown at 37°C to an optical density at 600 nm (OD₆₀₀) of 0.2–0.3, then rapidly cooled in an ice-water slurry and kept on ice for at least 1 hour. Fluorescence microscopy and data analysis were performed as previously described [39]. For each data set, the fluorescence of at least one hundred cells was recorded. The mean fluorescence from each data set was background subtracted, and the average between replicas was calculated.

Acid resistance assay

Cultures were grown as described above for fluorescence microscopy to OD₆₀₀ of 0.2. Cultures were concentrated to one sixth of the volume by centrifugation. Because of the low density of the cultures, this step was necessary to lower the limit of detection of viable cells following acid challenge. Fifty microliters of the concentrated cultures (approximately 1-5x10⁷ cells) were transferred to 1 ml sterile phosphate-buffered saline (PBS 137mM NaCl, 2.7mM KCl, 10mM Na₂HPO₄, 2mM KH₂PO₄), pH 7.4, and 50 μl were transferred to 1 ml of LB broth, pH 2.5 (acidified with HCl), that was pre-warmed to 37°C. The bacterial suspensions in acidified LB broth were incubated at 37°C for one hour, or the indicated times. Both LB and PBS cell suspensions were serially diluted in PBS, and aliquots were immediately plated in triplicate. After incubation overnight, CFUs were counted. The percentage survival was calculated as the number of CFUs/ml of acid shocked cultures divided by the number of CFUs/ml of the cultures diluted in PBS. For S9A Fig, the experiments were conducted in the same way, except the cultures were exposed to LB at pH 2.25, 2.0, or 1.75 as indicated.

Sequence alignments and phylogenetic analysis

Full-length EvgS, EvgA, YfdE, BarA, and ArcA sequences were extracted from complete E. coli genomes in NCBI using tblastn (http://blast.ncbi.nlm.nih.gov) using the corresponding sequences from E. coli K-12. Sequence alignments of S6 and S8 Figs were obtained with Clustal Omega, [40, 41] on the EMBL-EBI server. EvgS sequences in Fig 5A and S7 Fig were aligned and neighbor-joining trees were constructed using Muscle [42] on the EMBL-EBI server with the default parameters and displayed using Figtree (http://tree.bio.ed.ac.uk/software/figtree/) with midpoint rooting. For S7 Fig, the phylogenetic groups for the corresponding E. coli isolates from which these EvgS sequences were derived were determined as described in [43].

dN/dS analysis to assess nucleotide divergence

Nucleotide sequences for the genes of interest were obtained from the NCBI database. Multiple sequence alignments were generated in a codon-delimited format on the TranslatorX server [30] using MUSCLE alignment software [42]. To obtain gene-specific dN/dS estimates, we utilized the Synonymous Non-synonymous Analysis Program (SNAP v2.1.1) implementing Nei and Gojobori’s method [28] and its statistic output tool [31] available on the HIV sequence database website (www.hiv.lanl.gov, Korber, 2000). The codon-aligned nucleotide sequence alignments (.aln or fasta file) were provided as input into the SNAP tool to compute dN/dS ratios. Alignments were imported into MEGA software v4.0 [32] to perform a Z-test of neutral selection, for a null hypothesis (dN = dS) using modified Nei-Gojobori (Jukes-Cantor) method.

Supporting information

S1 Methods. Construction of strains and plasmids used in this study.

https://doi.org/10.1371/journal.pgen.1007101.s001

(PDF)

S1 Table. List of bacterial strains used in this study.

https://doi.org/10.1371/journal.pgen.1007101.s002

(PDF)

S2 Table. List of plasmids used in this study.

https://doi.org/10.1371/journal.pgen.1007101.s003

(PDF)

S3 Table. Primers used in this study.

https://doi.org/10.1371/journal.pgen.1007101.s004

(PDF)

S4 Table. Sequence evolution of evgS based on dN/dS analysis.

https://doi.org/10.1371/journal.pgen.1007101.s005

(PDF)

S5 Table. dN/dS analysis for evgS (Escherichia coli), bvgS (Bordetella pertussis), kvgS (Klebsiella pneumoniae) and related genes based on currently available complete genomes at NCBI.

https://doi.org/10.1371/journal.pgen.1007101.s006

(PDF)

S1 Fig. Activation of the PhoQ/PhoP system from different stimuli.

A: The PhoQ/PhoP system is not activated by mild acidity over a range of pH values. Transcriptional reporter derivatives of MG1655 (TIM92), MG1655 ΔsafA (SAM74), and MP1 (MP131) were cultured to OD₆₀₀ ~0.2 in minimal medium containing 10 mM MgSO₄ and buffered with 100 mM MES, at pH 5.1, 5.3, 5.5, 5.7, 5.9, 6.1, 6.3, and 7. Fluorescence was determined as described in Materials and methods. Fluorescence values are the average from two independent experiments. Error bars represent the range. B: mgrB is activated in MP1 by low Mg⁺⁺ similarly to E. coli K-12. Strains TIM63 (MG1655 P_mgrB-yfp) and MP131 (MP1 P_mgrB-yfp) were cultured to OD₆₀₀ ~0.2 in minimal medium at pH 7 with either 10 mM or 1 μm MgSO₄. Fluorescence was determined as described in Materials and methods. Fluorescence values are the average from two independent experiments. Error bars represent the range.

https://doi.org/10.1371/journal.pgen.1007101.s007

(PDF)

S2 Fig. Resistance to acid shock.

A: resistance of strains MG1655 and MP1 in different growth stages. Strains were cultured in minimal medium at pH 7 to stationary phase (16 hours) and at pH 5.7 to exponential phase (OD₆₀₀ ~0.2). Cultures were shocked for an hour at pH 2.5 as described in Materials and methods. Values are the average percent survival from two representative experiments and error bars represent the range. B: Resistance of MP1*, a strain derived from MP1 by transducing from MG1655 a region encompassing the 13 kb segment that is absent in MP1. Strains MG1655, MP1 and MP1* (MP144) were cultured at pH 5.7 to exponential phase (OD₆₀₀ ~0.2), and acid resistance was assayed as in A. Values are the average percent survival from two representative experiments and error bars represent the range. C: Strains MG1655, MMR241 (MG1655 ΔevgAS), TIM96 (MG1655 ΔphoQP), and SAM74 (MG1655 ΔsafA) were cultured and assayed for acid resistance as in B, except samples were withdrawn at the indicated times. Values are the average percent survival from two representative experiments and error bars represent the range.

https://doi.org/10.1371/journal.pgen.1007101.s008

(PDF)

S3 Fig. EvgS_MP1 is not responsive to mild acidity in the pH range effective for EvgS_MG1655.

Strains MMR182 (MG1655 P_emrK-yfp) and MP146 (MP1 P_emrK-yfp) were cultured in minimal medium at pH 7 or in minimal medium buffered with 100 mM MES at pH 5.1, 5.3, 5.5, 5.7, 5.9, 6.1, 6.3, and 7. Cultures were harvested at OD₆₀₀~0.2, and fluorescence of the reporter was measured as described in Materials and methods. Fluorescence values are the average from two representative independent experiments. Error bars represent the range.

https://doi.org/10.1371/journal.pgen.1007101.s009

(PDF)

S4 Fig. EvgS domain swap between MG1655 and MP1.

A: Diagram illustrating the plasmids and hybrid EvgS proteins utilized in this assay. Predicted transmembrane domains (TM), and residues they span, are indicated. (*) indicates a transmembrane domain that is predicted in some databases (www.uniprot.org) but is inconsistent with a homology model based on the ortholog BvgS [11]. Residues delimiting the swapped EvgS regions are numbered at the bottom and indicated by the dotted lines. B: Activity of the EvgA-dependent reporter P_emrK-yfp. The hybrid names correspond to those indicated in panel A. Strains with either wild type evgAS (MMR182) carrying the empty vector (pSMART), or ΔevgAS (MMR191) with the empty vector, or one of the plasmids pMR78, pMR117, pMR84, pMR80, pMR92, pMR82, pMR83 (in this order in the figure) were cultured in minimal medium at pH 7 and pH 5.7 to OD₆₀₀~0.2. Fluorescence was quantified by microscopy as described in Materials and methods. Values are the average fluorescence from two independent experiments and error bars represent the range.

https://doi.org/10.1371/journal.pgen.1007101.s010

(PDF)

S5 Fig. The single residue change that makes EvgS_MG1655 constitutively active has a similar effect on EvgS_MP1.

Derivatives of MG1655 and MP1 with the P_emrK-yfp reporter and with WT evgS (MMR182 and MP146) or with the the reporter and the F577S amino acid substitution associated with the evgS1 allele (MMR183 and MP145) were cultured in minimal medium at pH 5.7 (induced wild-type strains) and pH 7 (F577SEvgS) strains and non-induced wild-type strains) to OD₆₀₀~0.2 and fluorescence of the reporter was measured as described in Materials and methods. Fluorescence values are the average from two representative independent experiments. Error bars represent the range.

https://doi.org/10.1371/journal.pgen.1007101.s011

(PDF)

S6 Fig. Alignment of EvgS from ten E. coli isolates.

EvgS protein sequences from the indicated ten E. coli isolates were aligned as described in Materials and methods. (*) indicates a predicted transmembrane domain (www.uniProt.org) that is inconsistent with a homology model based on the ortholog BvgS [11]. The predicted PAS domain is shaded gray, and the green highlights indicate residues identified in previous studies as being involved in EvgS activity [4, 9–11].

https://doi.org/10.1371/journal.pgen.1007101.s012

(PDF)

S7 Fig. EvgS phylogeny in E. coli isolates. Dendrogram based on EvgS sequences from 285 E. coli genomes.

The colors correspond to phylogenetic groups: red–group A, orange–group C, green–group B1, light blue–group D, dark blue–group E, purple–group B2. The phylogenetic groups are also appended to the genome identifiers in the figure. Full-length EvgS sequences were extracted from complete E. coli genomes in NCBI using tblastn (http://blast.ncbi.nlm.nih.gov) with the E. coli K-12 EvgS sequence, and the large number of redundant hits from the genomes of E. coli K-12 derivatives were removed, resulting in 285 EvgS sequences. The tree construction is described in Materials and methods.

https://doi.org/10.1371/journal.pgen.1007101.s013

(PDF)

S8 Fig. Protein divergence matrices of ten E. coli isolates.

The indicated protein sequences from ten E. coli isolates were aligned as described in Materials and methods. Matrix entries indicate the percentage of amino acids that differ for the corresponding pairs of strains. In addition to EvgS and EvgA, matrices are shown for YfdE, a protein encoded by a gene adjacent to evgS, and for two other hybrid histidine kinases, BarA and ArcB. All five matrices use the indicated color range scale.

https://doi.org/10.1371/journal.pgen.1007101.s014

(PDF)

S9 Fig. Sensitivity of selected E. coli isolates to acid challenge in a range of pH values.

A: indicated strains were cultured in minimal medium at pH 5.7 to OD₆₀₀ ~0.2, and shocked for an hour in LB at pH 2.25 or pH 2.0, or pH 1.75 as described in Materials and methods. Percent survival values are the average of two representative experiments and error bars represent the range. As a reference, the figure includes the values of challenge at pH 2.5 from Fig 5C. B: Correlations between fold induction of the fluorescent reporter P_yfdX-yfp from Fig 5B versus percent survival after exposure to pH 2.25, pH 2.0, and pH 1.75 from panel A. The E. coli strains are described in detail in Table 1. Dashed lines represent the limit of detection of the assay.

https://doi.org/10.1371/journal.pgen.1007101.s015

(PDF)

S10 Fig. EvgAS_MG1655 rescue of the SafA and YdeO pathways in E. coli Nissle 1917 and E. coli UTI89.

Derivatives of MG1655, Nissle1917 and UTI89 with the P_hdeA-yfp reporter (Panel A, MMR175, MMR254 and MMR255) or with the P_mgrB-yfp reporter (Panel B, TIM95, MMR272 and MMR273), and carrying the empty vector pSMART or its derivative pevgAS_MG1655 (pMR78), were cultured in minimal medium at pH 7 and pH 5.7 to OD₆₀₀~0.2. Fluorescence of the reporter was measured as described in Materials and methods. Fluorescence values are the average from two representative independent experiments. Error bars represent the range.

https://doi.org/10.1371/journal.pgen.1007101.s016

(PDF)

Acknowledgments

We thank Joshua Plotkin and Premal Shah for helpful discussions concerning the dN/dS analysis. We also thank the following researchers for providing strains used in this study: Jun Zhu for H10407; Ilan Rosenshine for TUV93-0 and E2348/69; Jeffrey Weiser for CFT073; Maria Hadjifrangiskou for UTI89.

References

1. Leimbach A, Hacker J, Dobrindt U. E. coli as an all-rounder: the thin line between commensalism and pathogenicity. Curr Top Microbiol Immunol. 2013;358:3–32. pmid:23340801
- View Article
- PubMed/NCBI
- Google Scholar
2. Masuda N, Church GM. Escherichia coli gene expression responsive to levels of the response regulator EvgA. J Bacteriol. 2002 Nov;184(22):6225–34. pmid:12399493
- View Article
- PubMed/NCBI
- Google Scholar
3. Nishino K, Yamaguchi A. Overexpression of the response regulator evgA of the two-component signal transduction system modulates multidrug resistance conferred by multidrug resistance transporters. J Bacteriol. 2001 Feb;183(4):1455–8. pmid:11157960
- View Article
- PubMed/NCBI
- Google Scholar
4. Kato A, Ohnishi H, Yamamoto K, Furuta E, Tanabe H, Utsumi R. Transcription of emrKY is regulated by the EvgA-EvgS two-component system in Escherichia coli K-12. Biosci Biotechnol Biochem. 2000 Jun;64(6):1203–9. pmid:10923791
- View Article
- PubMed/NCBI
- Google Scholar
5. Eguchi Y, Oshima T, Mori H, Aono R, Yamamoto K, Ishihama A, et al. Transcriptional regulation of drug efflux genes by EvgAS, a two-component system in Escherichia coli. Microbiology. 2003 Oct;149(Pt 10):2819–28. pmid:14523115
- View Article
- PubMed/NCBI
- Google Scholar
6. Itou J, Eguchi Y, Utsumi R. Molecular mechanism of transcriptional cascade initiated by the EvgS/EvgA system in Escherichia coli K-12. Biosci Biotechnol Biochem. 2009 Apr 23;73(4):870–8. pmid:19352034
- View Article
- PubMed/NCBI
- Google Scholar
7. Ma Z, Masuda N, Foster JW. Characterization of EvgAS-YdeO-GadE branched regulatory circuit governing glutamate-dependent acid resistance in Escherichia coli. J Bacteriol. 2004 Nov;186(21):7378–89. pmid:15489450
- View Article
- PubMed/NCBI
- Google Scholar
8. Burton NA, Johnson MD, Antczak P, Robinson A, Lund PA. Novel aspects of the acid response network of E. coli K-12 are revealed by a study of transcriptional dynamics. J Mol Biol. 2010 Sep 3;401(5):726–42. pmid:20603130
- View Article
- PubMed/NCBI
- Google Scholar
9. Johnson MD, Bell J, Clarke K, Chandler R, Pathak P, Xia Y, et al. Characterization of mutations in the PAS domain of the EvgS sensor kinase selected by laboratory evolution for acid resistance in Escherichia coli. Mol Microbiol. 2014 Sep;93(5):911–27. pmid:24995530
- View Article
- PubMed/NCBI
- Google Scholar
10. Eguchi Y, Utsumi R. Alkali metals in addition to acidic pH activate the EvgS histidine kinase sensor in Escherichia coli. J Bacteriol. 2014 Sep;196(17):3140–9. pmid:24957621
- View Article
- PubMed/NCBI
- Google Scholar
11. Sen H, Aggarwal N, Ishionwu C, Hussain N, Parmar C, Jamshad M, et al. Structural and Functional Analysis of the Escherichia coli Acid-Sensing Histidine Kinase EvgS. J Bacteriol. 2017 Sep 15;199(18).
- View Article
- Google Scholar
12. Perraud AL, Kimmel B, Weiss V, Gross R. Specificity of the BvgAS and EvgAS phosphorelay is mediated by the C-terminal HPt domains of the sensor proteins. Mol Microbiol. 1998 Mar;27(5):875–87. pmid:9535079
- View Article
- PubMed/NCBI
- Google Scholar
13. Masuda N, Church GM. Regulatory network of acid resistance genes in Escherichia coli. Mol Microbiol. 2003 May;48(3):699–712. pmid:12694615
- View Article
- PubMed/NCBI
- Google Scholar
14. Foster JW. Escherichia coli acid resistance: tales of an amateur acidophile. Nat Rev Microbiol. 2004 Nov;2(11):898–907. pmid:15494746
- View Article
- PubMed/NCBI
- Google Scholar
15. De Biase D, Lund PA. The Escherichia coli Acid Stress Response and Its Significance for Pathogenesis. Adv Appl Microbiol. 2015;92:49–88. pmid:26003933
- View Article
- PubMed/NCBI
- Google Scholar
16. Kanjee U, Houry WA. Mechanisms of acid resistance in Escherichia coli. Annu Rev Microbiol. 2013;67:65–81. pmid:23701194
- View Article
- PubMed/NCBI
- Google Scholar
17. Eguchi Y, Itou J, Yamane M, Demizu R, Yamato F, Okada A, et al. B1500, a small membrane protein, connects the two-component systems EvgS/EvgA and PhoQ/PhoP in Escherichia coli. Proc Natl Acad Sci U S A. 2007 Nov 20;104(47):18712–7. pmid:17998538
- View Article
- PubMed/NCBI
- Google Scholar
18. Eguchi Y, Ishii E, Yamane M, Utsumi R. The connector SafA interacts with the multi-sensing domain of PhoQ in Escherichia coli. Mol Microbiol. 2012 Jul;85(2):299–313. pmid:22651704
- View Article
- PubMed/NCBI
- Google Scholar
19. Garcia Vescovi E, Soncini FC, Groisman EA. Mg2+ as an extracellular signal: environmental regulation of Salmonella virulence. Cell. 1996 Jan 12;84(1):165–74. pmid:8548821
- View Article
- PubMed/NCBI
- Google Scholar
20. Bader MW, Sanowar S, Daley ME, Schneider AR, Cho U, Xu W, et al. Recognition of antimicrobial peptides by a bacterial sensor kinase. Cell. 2005 Aug 12;122(3):461–72. pmid:16096064
- View Article
- PubMed/NCBI
- Google Scholar
21. Prost LR, Daley ME, Le Sage V, Bader MW, Le Moual H, Klevit RE, et al. Activation of the bacterial sensor kinase PhoQ by acidic pH. Mol Cell. 2007 Apr 27;26(2):165–74. pmid:17466620
- View Article
- PubMed/NCBI
- Google Scholar
22. Eguchi Y, Ishii E, Hata K, Utsumi R. Regulation of acid resistance by connectors of two-component signal transduction systems in Escherichia coli. J Bacteriol. 2011 Mar;193(5):1222–8. pmid:21193607
- View Article
- PubMed/NCBI
- Google Scholar
23. Lasaro M, Liu Z, Bishar R, Kelly K, Chattopadhyay S, Paul S, et al. Escherichia coli isolate for studying colonization of the mouse intestine and its application to two-component signaling knockouts. J Bacteriol. 2014 May;196(9):1723–32. pmid:24563035
- View Article
- PubMed/NCBI
- Google Scholar
24. Miyashiro T, Goulian M. Stimulus-dependent differential regulation in the Escherichia coli PhoQ PhoP system. Proc Natl Acad Sci U S A. 2007 Oct 9;104(41):16305–10. pmid:17909183
- View Article
- PubMed/NCBI
- Google Scholar
25. Kato A, Tanabe H, Utsumi R. Molecular characterization of the PhoP-PhoQ two-component system in Escherichia coli K-12: identification of extracellular Mg2+-responsive promoters. J Bacteriol. 1999 Sep;181(17):5516–20. pmid:10464230
- View Article
- PubMed/NCBI
- Google Scholar
26. Hommais F, Krin E, Coppee JY, Lacroix C, Yeramian E, Danchin A, et al. GadE (YhiE): a novel activator involved in the response to acid environment in Escherichia coli. Microbiology. 2004 Jan;150(Pt 1):61–72. pmid:14702398
- View Article
- PubMed/NCBI
- Google Scholar
27. Johnson MD, Burton NA, Gutierrez B, Painter K, Lund PA. RcsB is required for inducible acid resistance in Escherichia coli and acts at gadE-dependent and -independent promoters. J Bacteriol. 2011 Jul;193(14):3653–6. pmid:21571995
- View Article
- PubMed/NCBI
- Google Scholar
28. Nei M, Gojobori T. Simple methods for estimating the numbers of synonymous and nonsynonymous nucleotide substitutions. Mol Biol Evol. 1986 Sep;3(5):418–26. pmid:3444411
- View Article
- PubMed/NCBI
- Google Scholar
29. Hurst LD. The Ka/Ks ratio: diagnosing the form of sequence evolution. Trends Genet. 2002 Sep;18(9):486. pmid:12175810
- View Article
- PubMed/NCBI
- Google Scholar
30. Abascal F, Zardoya R, Telford MJ. TranslatorX: multiple alignment of nucleotide sequences guided by amino acid translations. Nucleic Acids Res. 2010 Jul;38(Web Server issue):W7–13. pmid:20435676
- View Article
- PubMed/NCBI
- Google Scholar
31. Ota T, Nei M. Divergent evolution and evolution by the birth-and-death process in the immunoglobulin VH gene family. Mol Biol Evol. 1994 May;11(3):469–82. pmid:8015440
- View Article
- PubMed/NCBI
- Google Scholar
32. Tamura K, Dudley J, Nei M, Kumar S. MEGA4: Molecular Evolutionary Genetics Analysis (MEGA) software version 4.0. Mol Biol Evol. 2007 Aug;24(8):1596–9. pmid:17488738
- View Article
- PubMed/NCBI
- Google Scholar
33. Arico B, Scarlato V, Monack DM, Falkow S, Rappuoli R. Structural and genetic analysis of the bvg locus in Bordetella species. Mol Microbiol. 1991 Oct;5(10):2481–91. pmid:1791760
- View Article
- PubMed/NCBI
- Google Scholar
34. Lai YC, Lin GT, Yang SL, Chang HY, Peng HL. Identification and characterization of KvgAS, a two-component system in Klebsiella pneumoniae CG43. FEMS Microbiol Lett. 2003 Jan 21;218(1):121–6. pmid:12583907
- View Article
- PubMed/NCBI
- Google Scholar
35. Herrou J, Debrie AS, Willery E, Renauld-Mongenie G, Locht C, Mooi F, et al. Molecular evolution of the two-component system BvgAS involved in virulence regulation in Bordetella. PLoS One. 2009 Sep 14;4(9):e6996. pmid:19750014
- View Article
- PubMed/NCBI
- Google Scholar
36. Ma Z, Gong S, Richard H, Tucker DL, Conway T, Foster JW. GadE (YhiE) activates glutamate decarboxylase-dependent acid resistance in Escherichia coli K-12. Mol Microbiol. 2003 Sep;49(5):1309–20. pmid:12940989
- View Article
- PubMed/NCBI
- Google Scholar
37. Miller JH. A short course in bacterial genetics: a laboratory manual and handbook for Escherichia coli and related bacteria. Plainview, N.Y.: Cold Spring Harbor Laboratory Press; 1992.
38. Datsenko KA, Wanner BL. One-step inactivation of chromosomal genes in Escherichia coli K-12 using PCR products. Proc Natl Acad Sci U S A. 2000 Jun 6;97(12):6640–5. pmid:10829079
- View Article
- PubMed/NCBI
- Google Scholar
39. Roggiani M, Goulian M. Oxygen-Dependent Cell-to-Cell Variability in the Output of the Escherichia coli Tor Phosphorelay. J Bacteriol. 2015 Jun 15;197(12):1976–87. pmid:25825431
- View Article
- PubMed/NCBI
- Google Scholar
40. Sievers F, Wilm A, Dineen D, Gibson TJ, Karplus K, Li W, et al. Fast, scalable generation of high-quality protein multiple sequence alignments using Clustal Omega. Mol Syst Biol. 2011 Oct 11;7:539. pmid:21988835
- View Article
- PubMed/NCBI
- Google Scholar
41. Goujon M, McWilliam H, Li W, Valentin F, Squizzato S, Paern J, et al. A new bioinformatics analysis tools framework at EMBL-EBI. Nucleic Acids Res. 2010 Jul;38(Web Server issue):W695–9. pmid:20439314
- View Article
- PubMed/NCBI
- Google Scholar
42. Edgar RC. MUSCLE: multiple sequence alignment with high accuracy and high throughput. Nucleic Acids Res. 2004;32(5):1792–7. pmid:15034147
- View Article
- PubMed/NCBI
- Google Scholar
43. Clermont O, Christenson JK, Denamur E, Gordon DM. The Clermont Escherichia coli phylo-typing method revisited: improvement of specificity and detection of new phylo-groups. Environ Microbiol Rep. 2013 Feb;5(1):58–65. pmid:23757131
- View Article
- PubMed/NCBI
- Google Scholar

[ref1] 1. Leimbach A, Hacker J, Dobrindt U. E. coli as an all-rounder: the thin line between commensalism and pathogenicity. Curr Top Microbiol Immunol. 2013;358:3–32. pmid:23340801
View Article
PubMed/NCBI
Google Scholar

[2] View Article

[3] PubMed/NCBI

[4] Google Scholar

[ref2] 2. Masuda N, Church GM. Escherichia coli gene expression responsive to levels of the response regulator EvgA. J Bacteriol. 2002 Nov;184(22):6225–34. pmid:12399493
View Article
PubMed/NCBI
Google Scholar

[6] View Article

[7] PubMed/NCBI

[8] Google Scholar

[ref3] 3. Nishino K, Yamaguchi A. Overexpression of the response regulator evgA of the two-component signal transduction system modulates multidrug resistance conferred by multidrug resistance transporters. J Bacteriol. 2001 Feb;183(4):1455–8. pmid:11157960
View Article
PubMed/NCBI
Google Scholar

[10] View Article

[11] PubMed/NCBI

[12] Google Scholar

[ref4] 4. Kato A, Ohnishi H, Yamamoto K, Furuta E, Tanabe H, Utsumi R. Transcription of emrKY is regulated by the EvgA-EvgS two-component system in Escherichia coli K-12. Biosci Biotechnol Biochem. 2000 Jun;64(6):1203–9. pmid:10923791
View Article
PubMed/NCBI
Google Scholar

[14] View Article

[15] PubMed/NCBI

[16] Google Scholar

[ref5] 5. Eguchi Y, Oshima T, Mori H, Aono R, Yamamoto K, Ishihama A, et al. Transcriptional regulation of drug efflux genes by EvgAS, a two-component system in Escherichia coli. Microbiology. 2003 Oct;149(Pt 10):2819–28. pmid:14523115
View Article
PubMed/NCBI
Google Scholar

[18] View Article

[19] PubMed/NCBI

[20] Google Scholar

[ref6] 6. Itou J, Eguchi Y, Utsumi R. Molecular mechanism of transcriptional cascade initiated by the EvgS/EvgA system in Escherichia coli K-12. Biosci Biotechnol Biochem. 2009 Apr 23;73(4):870–8. pmid:19352034
View Article
PubMed/NCBI
Google Scholar

[22] View Article

[23] PubMed/NCBI

[24] Google Scholar

[ref7] 7. Ma Z, Masuda N, Foster JW. Characterization of EvgAS-YdeO-GadE branched regulatory circuit governing glutamate-dependent acid resistance in Escherichia coli. J Bacteriol. 2004 Nov;186(21):7378–89. pmid:15489450
View Article
PubMed/NCBI
Google Scholar

[26] View Article

[27] PubMed/NCBI

[28] Google Scholar

[ref8] 8. Burton NA, Johnson MD, Antczak P, Robinson A, Lund PA. Novel aspects of the acid response network of E. coli K-12 are revealed by a study of transcriptional dynamics. J Mol Biol. 2010 Sep 3;401(5):726–42. pmid:20603130
View Article
PubMed/NCBI
Google Scholar

[30] View Article

[31] PubMed/NCBI

[32] Google Scholar

[ref9] 9. Johnson MD, Bell J, Clarke K, Chandler R, Pathak P, Xia Y, et al. Characterization of mutations in the PAS domain of the EvgS sensor kinase selected by laboratory evolution for acid resistance in Escherichia coli. Mol Microbiol. 2014 Sep;93(5):911–27. pmid:24995530
View Article
PubMed/NCBI
Google Scholar

[34] View Article

[35] PubMed/NCBI

[36] Google Scholar

[ref10] 10. Eguchi Y, Utsumi R. Alkali metals in addition to acidic pH activate the EvgS histidine kinase sensor in Escherichia coli. J Bacteriol. 2014 Sep;196(17):3140–9. pmid:24957621
View Article
PubMed/NCBI
Google Scholar

[38] View Article

[39] PubMed/NCBI

[40] Google Scholar

[ref11] 11. Sen H, Aggarwal N, Ishionwu C, Hussain N, Parmar C, Jamshad M, et al. Structural and Functional Analysis of the Escherichia coli Acid-Sensing Histidine Kinase EvgS. J Bacteriol. 2017 Sep 15;199(18).
View Article
Google Scholar

[42] View Article

[43] Google Scholar

[ref12] 12. Perraud AL, Kimmel B, Weiss V, Gross R. Specificity of the BvgAS and EvgAS phosphorelay is mediated by the C-terminal HPt domains of the sensor proteins. Mol Microbiol. 1998 Mar;27(5):875–87. pmid:9535079
View Article
PubMed/NCBI
Google Scholar

[45] View Article

[46] PubMed/NCBI

[47] Google Scholar

[ref13] 13. Masuda N, Church GM. Regulatory network of acid resistance genes in Escherichia coli. Mol Microbiol. 2003 May;48(3):699–712. pmid:12694615
View Article
PubMed/NCBI
Google Scholar

[49] View Article

[50] PubMed/NCBI

[51] Google Scholar

[ref14] 14. Foster JW. Escherichia coli acid resistance: tales of an amateur acidophile. Nat Rev Microbiol. 2004 Nov;2(11):898–907. pmid:15494746
View Article
PubMed/NCBI
Google Scholar

[53] View Article

[54] PubMed/NCBI

[55] Google Scholar

[ref15] 15. De Biase D, Lund PA. The Escherichia coli Acid Stress Response and Its Significance for Pathogenesis. Adv Appl Microbiol. 2015;92:49–88. pmid:26003933
View Article
PubMed/NCBI
Google Scholar

[57] View Article

[58] PubMed/NCBI

[59] Google Scholar

[ref16] 16. Kanjee U, Houry WA. Mechanisms of acid resistance in Escherichia coli. Annu Rev Microbiol. 2013;67:65–81. pmid:23701194
View Article
PubMed/NCBI
Google Scholar

[61] View Article

[62] PubMed/NCBI

[63] Google Scholar

[ref17] 17. Eguchi Y, Itou J, Yamane M, Demizu R, Yamato F, Okada A, et al. B1500, a small membrane protein, connects the two-component systems EvgS/EvgA and PhoQ/PhoP in Escherichia coli. Proc Natl Acad Sci U S A. 2007 Nov 20;104(47):18712–7. pmid:17998538
View Article
PubMed/NCBI
Google Scholar

[65] View Article

[66] PubMed/NCBI

[67] Google Scholar

[ref18] 18. Eguchi Y, Ishii E, Yamane M, Utsumi R. The connector SafA interacts with the multi-sensing domain of PhoQ in Escherichia coli. Mol Microbiol. 2012 Jul;85(2):299–313. pmid:22651704
View Article
PubMed/NCBI
Google Scholar

[69] View Article

[70] PubMed/NCBI

[71] Google Scholar

[ref19] 19. Garcia Vescovi E, Soncini FC, Groisman EA. Mg2+ as an extracellular signal: environmental regulation of Salmonella virulence. Cell. 1996 Jan 12;84(1):165–74. pmid:8548821
View Article
PubMed/NCBI
Google Scholar

[73] View Article

[74] PubMed/NCBI

[75] Google Scholar

[ref20] 20. Bader MW, Sanowar S, Daley ME, Schneider AR, Cho U, Xu W, et al. Recognition of antimicrobial peptides by a bacterial sensor kinase. Cell. 2005 Aug 12;122(3):461–72. pmid:16096064
View Article
PubMed/NCBI
Google Scholar

[77] View Article

[78] PubMed/NCBI

[79] Google Scholar

[ref21] 21. Prost LR, Daley ME, Le Sage V, Bader MW, Le Moual H, Klevit RE, et al. Activation of the bacterial sensor kinase PhoQ by acidic pH. Mol Cell. 2007 Apr 27;26(2):165–74. pmid:17466620
View Article
PubMed/NCBI
Google Scholar

[81] View Article

[82] PubMed/NCBI

[83] Google Scholar

[ref22] 22. Eguchi Y, Ishii E, Hata K, Utsumi R. Regulation of acid resistance by connectors of two-component signal transduction systems in Escherichia coli. J Bacteriol. 2011 Mar;193(5):1222–8. pmid:21193607
View Article
PubMed/NCBI
Google Scholar

[85] View Article

[86] PubMed/NCBI

[87] Google Scholar

[ref23] 23. Lasaro M, Liu Z, Bishar R, Kelly K, Chattopadhyay S, Paul S, et al. Escherichia coli isolate for studying colonization of the mouse intestine and its application to two-component signaling knockouts. J Bacteriol. 2014 May;196(9):1723–32. pmid:24563035
View Article
PubMed/NCBI
Google Scholar

[89] View Article

[90] PubMed/NCBI

[91] Google Scholar

[ref24] 24. Miyashiro T, Goulian M. Stimulus-dependent differential regulation in the Escherichia coli PhoQ PhoP system. Proc Natl Acad Sci U S A. 2007 Oct 9;104(41):16305–10. pmid:17909183
View Article
PubMed/NCBI
Google Scholar

[93] View Article

[94] PubMed/NCBI

[95] Google Scholar

[ref25] 25. Kato A, Tanabe H, Utsumi R. Molecular characterization of the PhoP-PhoQ two-component system in Escherichia coli K-12: identification of extracellular Mg2+-responsive promoters. J Bacteriol. 1999 Sep;181(17):5516–20. pmid:10464230
View Article
PubMed/NCBI
Google Scholar

[97] View Article

[98] PubMed/NCBI

[99] Google Scholar

[ref26] 26. Hommais F, Krin E, Coppee JY, Lacroix C, Yeramian E, Danchin A, et al. GadE (YhiE): a novel activator involved in the response to acid environment in Escherichia coli. Microbiology. 2004 Jan;150(Pt 1):61–72. pmid:14702398
View Article
PubMed/NCBI
Google Scholar

[101] View Article

[102] PubMed/NCBI

[103] Google Scholar

[ref27] 27. Johnson MD, Burton NA, Gutierrez B, Painter K, Lund PA. RcsB is required for inducible acid resistance in Escherichia coli and acts at gadE-dependent and -independent promoters. J Bacteriol. 2011 Jul;193(14):3653–6. pmid:21571995
View Article
PubMed/NCBI
Google Scholar

[105] View Article

[106] PubMed/NCBI

[107] Google Scholar

[ref28] 28. Nei M, Gojobori T. Simple methods for estimating the numbers of synonymous and nonsynonymous nucleotide substitutions. Mol Biol Evol. 1986 Sep;3(5):418–26. pmid:3444411
View Article
PubMed/NCBI
Google Scholar

[109] View Article

[110] PubMed/NCBI

[111] Google Scholar

[ref29] 29. Hurst LD. The Ka/Ks ratio: diagnosing the form of sequence evolution. Trends Genet. 2002 Sep;18(9):486. pmid:12175810
View Article
PubMed/NCBI
Google Scholar

[113] View Article

[114] PubMed/NCBI

[115] Google Scholar

[ref30] 30. Abascal F, Zardoya R, Telford MJ. TranslatorX: multiple alignment of nucleotide sequences guided by amino acid translations. Nucleic Acids Res. 2010 Jul;38(Web Server issue):W7–13. pmid:20435676
View Article
PubMed/NCBI
Google Scholar

[117] View Article

[118] PubMed/NCBI

[119] Google Scholar

[ref31] 31. Ota T, Nei M. Divergent evolution and evolution by the birth-and-death process in the immunoglobulin VH gene family. Mol Biol Evol. 1994 May;11(3):469–82. pmid:8015440
View Article
PubMed/NCBI
Google Scholar

[121] View Article

[122] PubMed/NCBI

[123] Google Scholar

[ref32] 32. Tamura K, Dudley J, Nei M, Kumar S. MEGA4: Molecular Evolutionary Genetics Analysis (MEGA) software version 4.0. Mol Biol Evol. 2007 Aug;24(8):1596–9. pmid:17488738
View Article
PubMed/NCBI
Google Scholar

[125] View Article

[126] PubMed/NCBI

[127] Google Scholar

[ref33] 33. Arico B, Scarlato V, Monack DM, Falkow S, Rappuoli R. Structural and genetic analysis of the bvg locus in Bordetella species. Mol Microbiol. 1991 Oct;5(10):2481–91. pmid:1791760
View Article
PubMed/NCBI
Google Scholar

[129] View Article

[130] PubMed/NCBI

[131] Google Scholar

[ref34] 34. Lai YC, Lin GT, Yang SL, Chang HY, Peng HL. Identification and characterization of KvgAS, a two-component system in Klebsiella pneumoniae CG43. FEMS Microbiol Lett. 2003 Jan 21;218(1):121–6. pmid:12583907
View Article
PubMed/NCBI
Google Scholar

[133] View Article

[134] PubMed/NCBI

[135] Google Scholar

[ref35] 35. Herrou J, Debrie AS, Willery E, Renauld-Mongenie G, Locht C, Mooi F, et al. Molecular evolution of the two-component system BvgAS involved in virulence regulation in Bordetella. PLoS One. 2009 Sep 14;4(9):e6996. pmid:19750014
View Article
PubMed/NCBI
Google Scholar

[137] View Article

[138] PubMed/NCBI

[139] Google Scholar

[ref36] 36. Ma Z, Gong S, Richard H, Tucker DL, Conway T, Foster JW. GadE (YhiE) activates glutamate decarboxylase-dependent acid resistance in Escherichia coli K-12. Mol Microbiol. 2003 Sep;49(5):1309–20. pmid:12940989
View Article
PubMed/NCBI
Google Scholar

[141] View Article

[142] PubMed/NCBI

[143] Google Scholar

[ref37] 37. Miller JH. A short course in bacterial genetics: a laboratory manual and handbook for Escherichia coli and related bacteria. Plainview, N.Y.: Cold Spring Harbor Laboratory Press; 1992.

[ref38] 38. Datsenko KA, Wanner BL. One-step inactivation of chromosomal genes in Escherichia coli K-12 using PCR products. Proc Natl Acad Sci U S A. 2000 Jun 6;97(12):6640–5. pmid:10829079
View Article
PubMed/NCBI
Google Scholar

[146] View Article

[147] PubMed/NCBI

[148] Google Scholar

[ref39] 39. Roggiani M, Goulian M. Oxygen-Dependent Cell-to-Cell Variability in the Output of the Escherichia coli Tor Phosphorelay. J Bacteriol. 2015 Jun 15;197(12):1976–87. pmid:25825431
View Article
PubMed/NCBI
Google Scholar

[150] View Article

[151] PubMed/NCBI

[152] Google Scholar

[ref40] 40. Sievers F, Wilm A, Dineen D, Gibson TJ, Karplus K, Li W, et al. Fast, scalable generation of high-quality protein multiple sequence alignments using Clustal Omega. Mol Syst Biol. 2011 Oct 11;7:539. pmid:21988835
View Article
PubMed/NCBI
Google Scholar

[154] View Article

[155] PubMed/NCBI

[156] Google Scholar

[ref41] 41. Goujon M, McWilliam H, Li W, Valentin F, Squizzato S, Paern J, et al. A new bioinformatics analysis tools framework at EMBL-EBI. Nucleic Acids Res. 2010 Jul;38(Web Server issue):W695–9. pmid:20439314
View Article
PubMed/NCBI
Google Scholar

[158] View Article

[159] PubMed/NCBI

[160] Google Scholar

[ref42] 42. Edgar RC. MUSCLE: multiple sequence alignment with high accuracy and high throughput. Nucleic Acids Res. 2004;32(5):1792–7. pmid:15034147
View Article
PubMed/NCBI
Google Scholar

[162] View Article

[163] PubMed/NCBI

[164] Google Scholar

[ref43] 43. Clermont O, Christenson JK, Denamur E, Gordon DM. The Clermont Escherichia coli phylo-typing method revisited: improvement of specificity and detection of new phylo-groups. Environ Microbiol Rep. 2013 Feb;5(1):58–65. pmid:23757131
View Article
PubMed/NCBI
Google Scholar

[166] View Article

[167] PubMed/NCBI

[168] Google Scholar

Figures

Abstract

Author summary

Introduction

Results

E. coli MP1 lacks two branches of the EvgS/EvgA pathway

Restoring safAydeO in MP1 is not sufficient for low pH activation of GadE and PhoQ

The MP1 EvgS/EvgA system is not induced by low pH

Natural variation in pH sensing by the EvgS/EvgA phosphorelay among E. coli

evgAS from MG1655 confers resistance to acid shock in acid-sensitive isolates

Natural variation in evgS sequence is consistent with weak purifying selection

Discussion

Materials and methods

Growth media and conditions

Strains and plasmids

Fluorescence quantification

Acid resistance assay

Sequence alignments and phylogenetic analysis

dN/dS analysis to assess nucleotide divergence

Supporting information

S1 Methods. Construction of strains and plasmids used in this study.

S1 Table. List of bacterial strains used in this study.

S2 Table. List of plasmids used in this study.

S3 Table. Primers used in this study.

S4 Table. Sequence evolution of evgS based on dN/dS analysis.

S5 Table. dN/dS analysis for evgS (Escherichia coli), bvgS (Bordetella pertussis), kvgS (Klebsiella pneumoniae) and related genes based on currently available complete genomes at NCBI.

S1 Fig. Activation of the PhoQ/PhoP system from different stimuli.

S2 Fig. Resistance to acid shock.

S3 Fig. EvgSMP1 is not responsive to mild acidity in the pH range effective for EvgSMG1655.

S4 Fig. EvgS domain swap between MG1655 and MP1.

S5 Fig. The single residue change that makes EvgSMG1655 constitutively active has a similar effect on EvgSMP1.

S6 Fig. Alignment of EvgS from ten E. coli isolates.

S7 Fig. EvgS phylogeny in E. coli isolates. Dendrogram based on EvgS sequences from 285 E. coli genomes.

S8 Fig. Protein divergence matrices of ten E. coli isolates.

S9 Fig. Sensitivity of selected E. coli isolates to acid challenge in a range of pH values.

S10 Fig. EvgASMG1655 rescue of the SafA and YdeO pathways in E. coli Nissle 1917 and E. coli UTI89.

Acknowledgments

References

S3 Fig. EvgS_MP1 is not responsive to mild acidity in the pH range effective for EvgS_MG1655.

S5 Fig. The single residue change that makes EvgS_MG1655 constitutively active has a similar effect on EvgS_MP1.

S10 Fig. EvgAS_MG1655 rescue of the SafA and YdeO pathways in E. coli Nissle 1917 and E. coli UTI89.