Abstract
Two-component systems and phosphorelays play central roles in the ability of bacteria to rapidly respond to changing environments. In E. coli and related enterobacteria, the complex Rcs phosphorelay is a critical player in changing bacterial behavior in response to antimicrobial peptides, beta-lactam antibiotics, and other challenges to the cell surface. The Rcs system is unusual in that IgaA, an inner membrane protein, is essential due to its negative regulation of the RcsC/RcsD/RcsB phosphorelay. While it has previously been shown that IgaA transduces signals from the outer membrane lipoprotein RcsF, how it interacts with the phosphorelay was unknown. Here we use in vivo interaction assays and genetic dissection of the critical proteins to demonstrate that IgaA interacts with the phosphorelay protein RcsD, and that this interaction is necessary for regulation. Interactions in periplasmic domains of these two proteins anchor repression of signaling. However, the signaling response depends on a weaker interaction between cytoplasmic loop 1 of IgaA and a truncated PAS domain in RcsD. A point mutation in the PAS domain increases interactions between the two proteins and is sufficient to abolish induction of this phosphorelay. RcsC, the histidine kinase that initiates phosphotransfer through the phosphorelay, appears to be indirectly regulated by IgaA via the contacts with RcsD. Unlike RcsD, and unlike many other histidine kinases, the periplasmic domain of RcsC is not necessary for the response to inducing signals. The multiple contacts between IgaA and RcsD form a poised sensing system, preventing over-activation of this apparently toxic phosphorelay but allowing it to be rapidly and quantitatively responsive to signals.
Author Summary The Rcs phosphorelay plays a central role in allowing enterobacteria to sense and respond to antibiotics, host-produced antimicrobials, and interactions with surfaces. A unique negative regulator, IgaA, keeps signaling from this pathway under control when it is not needed, but how it controls the phosphorelay has been unclear. We define a set of critical interactions between IgaA and the phosphotransfer protein RcsD. A periplasmic contact between IgaA and RcsD provides a necessary inhibition of Rcs signaling, modulated further by regulated interactions in the cytoplasmic domains of each protein. This multipartite interaction provides a sensitive regulatory switch.
Introduction
Bacteria must constantly monitor their cell wall and envelope integrity to withstand environmental insult. Osmotic stress, redox stress and envelope disruption demand that the bacterium remodel its exterior to provide protection, often in the form of capsular polysaccharide. Enterobacterales use the Rcs phosphorelay to integrate complex signals from the outer membrane and periplasm, changing gene regulation in response to stress [1, 2]. The Rcs phosphorelay is a complex signal transduction pathway, involving an outer membrane lipoprotein (RcsF) and three inner membrane proteins (IgaA, RcsC and RcsD), leading to changes in the phosphorylation of the transcriptional regulator (RcsB). The Rcs phosphorelay regulates production of virulence-associated capsules as well as motility and the expression of many stress-related genes.
Signaling through the pathway is complex, and not fully understood. Briefly, outer membrane stress such as cationic polypeptides or cell wall stresses such as beta-lactams cause RcsF to change its interaction with IgaA (originally identified in Salmonella and named for intracellular growth attenuation; the E. coli version of this gene, yrfF, is referred to here as IgaA). The activated RcsF/IgaA interaction allows the hybrid histidine kinase RcsC to auto-phosphorylate and then pass phosphate to phosphorelay protein RcsD, a process studied here, which passes it to response regulator RcsB (Figure 1A). Over-signaling through the phosphorelay leads to cell death, possibly because of the global nature of the RcsB regulon. IgaA is essential because of its role as a gating/braking mechanism for the phosphorelay. Deletion of IgaA is only possible in cells containing mutations in RcsC, RcsD, or RcsB. For this reason, the poorly understood IgaA mechanism of action is of key interest. Multiple studies have focused on the interaction of RcsF with IgaA when cell wall stress is detected [3–7], but the downstream action of IgaA is less well understood [8]. In this work we define RcsD as the direct binding partner of IgaA and define the regions in RcsD that are critical for interaction with IgaA. Production of RcsD variants that are deficient in IgaA binding cause massive over-signaling, mucoidy, and often cell death, consistent with phenotypes seen upon loss of IgaA itself. These results contrast with previous assumptions that IgaA was likely to directly interact with and regulate the histidine kinase RcsC, because RcsC initiates the phosphorelay.
Results
A sensitive and flexible assay for the Rcs phosphorelay
We have revisited the Rcs signaling pathway using a newly developed sensitive in vivo fluorescent reporter assay. RprA is a small RNA that is a sensitive, specific target of Rcs regulation. RprA levels are nearly undetectable in the absence of RcsB, its direct transcriptional activator, and increase in cells in which the Rcs system is activated, for instance by constitutively active rcsC mutations [9]. A rprA transcriptional fusion to mCherry allows continuous detection of a wide range of Rcs activation levels, using a plate reader. This growth and fluorescence assay can be viewed as a function of fluorescence over average OD, showing a large, early change in slope when a given strain is induced to respond to an Rcs stimulus (Fig S1A-C). Bar graphs of fluorescence at OD600 0.4. Figure 1B shows the increase in signal when wild-type cells are exposed to a small molecule stimulator of Rcs signaling, the non-toxic cationic polymyxin B nonapeptide (PMBN). PMBN stimulus is a useful indicator of pathway status; normal signaling yields a measured level of response to PMBN, while pathway disruptions (modification or deletion of pathway components) cause dampened or loss of PMBN response. Cells deleted for rcsB lose all signal, including the low basal level seen in the absence of PMBN (Fig 1B, S1A-C Fig). The absence of RcsF lowers overall signal (compare ΔrcsF::chl, - PMBN, to WT, - PMBN); this decrease has been reported before [3, 10–12], but is particularly clear with this assay. Lack of RcsF also greatly dampens response to an outer membrane stress like PMBN (Fig 1B). It is known that Rcs signaling can be induced in the absence of RcsF, so the small activation of the rcsF::chl strain in the presence of PMBN is possibly stimulating Rcs in this (still unexplored) manner [10, 13]; (Majdalani et al, unpublished).
The hybrid histidine kinase RcsC and the phosphorelay protein RcsD play both positive and negative roles in regulation of RcsB activity. Loss of RcsC or RcsD blocks the response to PMBN, but also lead to significantly higher levels of PrprA-mCherry in the absence of normal inducing signal (Fig 1B). This is consistent with previous work, in which expression of an PrprA- lac reporter was increased upon deletion of rcsC or rcsD [9, 11]. Deletion of rcsC is thought to cause the loss of ability to de-phosphorylate RcsB that has acquired phosphate from other sources [14–16]. In our assay conditions, the rcsC deletion strain produces a signal that is 3-4 fold above WT.
The rcsD deletion shown, rcsD541, increased PrprA-mCherry expression in a manner similar to the rcsC deletion. rcsD is encoded upstream of rcsB, with the major promoters for rcsB inside the rcsD coding region [17]. This affects the way rcsD deletion alleles can be constructed. In addition, in both Salmonella and E. coli, some transcripts from the rcsD promoter may continue through to rcsB, though apparently at a much lower level [17, 18]. Four different rcsD alleles were examined, each designed to leave the rcsB promoters intact (S1D Fig). These include rcsD carrying an H842A mutation in the phosphotransfer domain active site, as well as rcsD containing stop codons after codon 841 (rcsD841*). The overall amount of PrprA-mCherry expression seems to differ modestly between rcsD541 and rcsD543, even though the deleted regions in both alleles share the same boundaries. rcsD541 also gave higher PrprA-mCherry signal than the point mutants (S1D Fig). When checked by western blot with a polyclonal RcsD antibody, it is evident that rcsD541, rcsD543 and rcsD841* are all devoid of detectable RcsD (S1E Fig). As previously seen [11], rcsD541 and rcsD543 had no significant effect on RcsB levels, nor did rcsD H842A and rcsD841* (S1E Fig, right panel).
The somewhat increased level of PrprA-mCherry common to all rcsD and rcsC strains is likely in part due to phosphorylation of RcsB by the small molecule acetyl phosphate (AcP), in the absence of the dephosphorylation carried out by RcsD and RcsC [15, 19]. The influence of AcP can be demonstrated in an ackA deletion strain that builds up large amounts of AcP [15] (S1F Fig). While WT cells showed a modest increase in signal in the ackA mutant (compare black and gray bars), all tested rcsC and rcsD mutations produced high levels of PrprA-mCherry I the ackA mutant, consistent with failure to dephosphorylate RcsB (S1F Fig). The increase in signal is fully dependent upon RcsB (last bar in graph, S1F).
IgaA and RcsD interact directly
We began interrogating how IgaA might interfere with Rcs signaling by examining the interactions of IgaA with downstream members of the phosphorelay, using the bacterial adenylate cyclase two hybrid assay (BACTH). In this assay, cells produce beta-galactosidase when Bordetella adenylate cyclase fragments (T18 and T25, Cya) are reconstituted by fusion to interacting proteins in a cyclase-defective host [20, 21]. IgaA interacted robustly with RcsD in two orientations (IgaA-T18/RcsD-T25 and IgaA-T25/RcsD-T18), expressing beta-galactosidase activity approximately 20-fold greater than either fusion paired with an empty cognate vector, the standard background control (Fig 2A; S2A Fig). This interaction occurred irrespective of the chromosomal presence or absence of other Rcs members (S2B,C Fig). However, no significant interaction was detected between IgaA and RcsC in parallel assays (Fig 2A, S2B,D Fig).
The IgaA, RcsD, and RcsC fused to Cya fragments were all tested for their ability to function in the Rcs phosphorelay and were found to be functional (see Materials and Methods, S2E Fig); blots for the proteins showed the expected bands (S2F Fig). Therefore, lack of interaction by RcsC in the bacterial two-hybrid assay cannot be attributed to significant misfolding or lack of protein. These results suggest that IgaA interacts with RcsD but does not interact with RcsC in this assay.
Regions in RcsD necessary and sufficient for interaction with IgaA were defined in the bacterial two-hybrid assay (Fig 2B). Using C-terminally truncated RcsD constructs, we found that IgaA bound just as well to RcsDN-683 (RcsD without the ABL and Hpt domains) as it did to wild type. It also bound to RcsDN-522 (no HATPase, ABL or Hpt domains), but less strongly than to wild type or RcsDN-683. The interactions were unaffected by the presence of RcsD in the chromosome (S2G Fig). Strikingly, RcsDN-462 also bound IgaA almost as well as full length RcsD (Fig. 2B). This region is predicted to contain an incomplete Per-Arndt-Sim (PAS) domain (shown on the graphic as “PAS” to emphasize that it is not a complete PAS domain), which has been associated with signal detection in sensor histidine kinases [22]. A further truncation, removing most of the cytoplasmic regions upstream of the HATPase domain (RcsDN-383), did not produce a measurable IgaA interaction, suggesting a critical role for at least some of the cytoplasmic domain. A fully cytoplasmic RcsD construct (RcsD326-C) that began directly after the membrane-bound portion and included the rest of the RcsD C-terminus also failed to show any detectable interaction with IgaA. Finally, a construct in which the periplasmic region was deleted (RcsDΔ45-304) still interacts with IgaA, although only at a level of about half that seen with the WT construct (Fig 2B).
These results are most consistent with RcsD interactions with IgaA both within the trans-membrane/periplasmic portion of RcsD and within the initial cytoplasmic regions (bounded perhaps by residue 462), with neither interaction sufficient for a full signal in this two-hybrid assay. Intriguingly, the constructs that gave positive interactions (RcsDN-522, RcsDN-683 and RcsDN-462, but not RcsDN-383 and RcsD326-C) also caused mucoidy, a reflection of activation of the Rcs phosphorelay, in the cloning strain (Stellar E. coli, Clontech), which is wild-type for all genes of the Rcs phosphorelay. This was further examined, using the PrprA-mCherry assay.
Titration of IgaA by overexpression of truncated RcsD
If IgaA repression of Rcs signaling depends on the direct interaction of RcsD with IgaA, suggested by the BACTH results, overproduction of the regions of RcsD sufficient for this interaction may titrate IgaA away from the chromosomally-encoded RcsD, allowing unregulated signaling through the Rcs phosphorelay. The same RcsD fragments studied in the BACTH experiments were cloned without tags into a pBAD24 plasmid, under the control of the arabinose-inducible pBAD promoter. In this plasmid in the absence of arabinose, RcsD is expressed levels sufficient to complement the modest increase in signal seen in an rcsD mutant (Fig 3A, right panel; compare RcsD to V); the protein is significantly overexpressed in the presence of arabinose. These plasmids were assayed in both rcsD+ and rcsD541 strains containing the PrprA-mCherry reporter fusion. We would expect the WT strain to be active for the reporter when the Rcs phosphorelay is activated. Indeed, in the rcsD+ host, overproduction after arabinose induction of RcsD fragments capable of interacting with IgaA (RcsDN-683, RcsDN-525 and RcsDN-462) activated the PrprA-mCherry fusion, while RcsDN-383, which was negative for interaction with IgaA (Fig 2B), did not (Fig 3A, S3A Fig). Significant cellular growth arrest and lysis occurred when RcsDN-683 or RcsDN-525 were overproduced (in the presence of arabinose), making them difficult to compare quantitatively to non-lysed cells. In S3A Fig, their activation at lower ODs (before lysis) is shown, and fluorescence graphed as a function of the OD throughout growth is shown.
The ability of fragments to activate in an rcsD+ host correlates well with ability to interact with IgaA in the bacterial two-hybrid assay, consistent with the model that activation by overproduction is a result of binding to and titrating IgaA, freeing the wild-type RcsD to signal. These results reinforce the conclusion from the BACTH assays that at least two regions of RcsD interact with IgaA. Both regions are necessary for efficient titration, one region presumably between 1-383 (encompasses trans-membrane and periplasmic region), present in all the titrating plasmids, but not sufficient for titration, and a second region between 383 and 462 (including the incomplete PAS-like domain).
Note that overexpressing full-length RcsD did not induce mucoidy or signaling. Since the full-length protein has all regions that should bind and titrate IgaA, this result suggests the possibility that the C-terminal domains of RcsD, missing in all the activating/titrating truncations, might play a negative role that blocks or is epistatic to the titration seen with the truncated plasmids. This is consistent with the report that the ABL domain can bind and inhibit phosphorylation of RcsB [23], and is confirmed below.
The RcsD Hpt domain is necessary for transmitting a signal from the RcsC response regulator domain to RcsB [24]. Therefore, we expected that plasmids expressing truncated RcsD constructs that lack the Hpt domain would be completely unable to activate the phosphorelay in a strain mutant for rcsD. In the rcsD541 mutant allele background, the basal level of expression is above that in a WT host (Fig 1B, Fig 3A, compare V to WT V); a plasmid expressing the intact RcsD reduces PrprA-mCherry expression to levels comparable to the WT strain, consistent with complementation of the rcsD541 mutant (Fig. 3A). The activating fragment RcsDN-462 did not induce significant PrprA-mCherry activity in this host, consistent with expectation, since it does not contain the Hpt domain (Fig 3A).
Unexpectedly, cells expressing somewhat longer RcsD fragments, truncations RcsDN-522 and RcsDN-683, both missing the Hpt domain, were able to significantly increase signal when induced in the rcsD541 host (S3A Fig). However, these same plasmids caused lysis in WT and rcsD541 cells (S3A Fig). These constructs contain regions of RcsD that are not well understood, including an ancestral histidine kinase structure between residues 462 and 683. To further investigate the basis for this unexpected signaling, the same plasmids were tested in three additional rcsD mutants (S3B Fig). In a strain with a chromosomal mutation in the phosphotransfer active site, RcsD H842A,or in a strain containing a stop codon at residue 842, right before the active site (rcsD841*), the RcsDN-522 and RcsDN-683 truncations did not raise the level of PrprA-mCherry, suggesting that the nature of the chromosomal rcsD mutation is contributing to the effect caused by overproduction of the fragments (S3B Fig). An intact Hpt domain in the chromosomal copy of rcsD appears to be necessary to allow this modest activation, possibly suggesting that the rcsD541 mutant may express a low level of the Hpt domain. This signaling is fully dependent upon RcsB (S3C Fig). Why some of our truncated constructs but not others act in this way is not further investigated here.
Signaling by RcsD alleles with reduced capacity to interact with IgaA
Plasmids carrying C-terminal portions of RcsD (and thus the Hpt domain) were constructed and tested in both the WT and rcsD541 strain. Expression of the ABL-Hpt domains (RcsD686-C), the Hpt domain alone (RcsD792-C), the cytoplasmic portion of RcsD (RcsD326-C), or an rcsD allele deleted for the periplasmic region (rcsD Δ45-304) did not induce signaling in wild type cells (Fig 3B, left panel). However, each of these plasmids led to significantly increased signaling in the rcsD541 strain, even in the absence of arabinose (Fig 3B). The cytoplasmic portion of RcsD (RcsD326-C), which had no effect on signal in rcsD+ cells, led to a 12-fold increase in signal in rcsD541 over the vector control, approximately a 24-fold increase over the normal wild type level of signal (Fig. 3B), grew poorly (6 hour OD600 of 0.12, used in Fig 3B), and colonies containing this plasmid became mucoid in the absence of arabinose. Less pronounced induction of signaling was seen in cells expressing the ABL-Hpt domain or the Hpt domain alone in the absence of arabinose. These results are fully consistent with the idea that IgaA represses signaling via interactions with regions in the N-terminus of RcsD; expression of derivatives of RcsD that retain the Hpt domain (and thus can transfer signal from RcsC to RcsB) but that have lost or compromised the IgaA interaction region are now able to activate in the absence of an inducing signal such as PMBN.
We confirmed that this high level of signal is dependent upon RcsC; in an rcsCD double mutant, expression of RcsD326-C or RcsD Hpt (rcsD792-C) does not give rise to signal above the value obtained by adding full length RcsD back to an rcsCD double mutant (S3D Fig). The properties of these fragments demonstrate clearly that phosphotransfer from RcsC to RcsB requires, minimally, the Hpt domain of RcsD, and that, in the absence of the IgaA/RcsD inhibitory interaction, RcsC promotes high, constitutive signaling. Finally, because this signaling is not seen in an rcsD+ strain, signaling by these fragments of RcsD is recessive to the full-length protein.
As noted above, it has been reported that overproduction of the RcsD ABL domain can bind and inhibit RcsB [23]. We further investigated this with overproduction of RcsD regions in our plasmids, testing their potential for induction by PMBN in cells carrying a wild-type chromosomal rcsD copy (S3E Fig). If RcsB (or any other step in the pathway) is inhibited, we would expect to block PMBN induction.
Strikingly, while the RcsD plasmid does not affect PMBN-dependent induction in the absence of arabinose (low levels of RcsD), it fully inhibits it when arabinose is present (high levels of RcsD). The plasmid containing the ABL-Hpt domain (Rcs686-C) has a very similar profile, as predicted if the ABL domain is necessary for inhibition. RcsD Δ45-304, deleted for the periplasmic region, also acts similarly. The Hpt domain itself (RcsD792-C) was somewhat less effective in blocking PMBN signaling, consistent with a role for the ABL domain in inhibition. When these plasmids were expressed in a rcsD541 host (S3F Fig), RcsD was still able to respond to PMBN without arabinose induction, but not with arabinose. High levels of the ABL-Hpt domain also inhibited the constitutive signaling otherwise seen with low levels of this RcsD fragment, confirming that the inhibition is independent of and downstream of the IgaA/RcsD interaction, not present for this piece of RcsD.
Two of the plasmids gave results that were more difficult to interpret. Expression of a full-length RcsD mutant for the active site histidine (RcsD H842A) had higher activity than expected in an rcsD+ host but not in the rcsD541 host (S3E and S3F Fig). Because this higher activity was seen with or without arabinose, the results suggest that the wild-type RcsD (from the chromosome) and RcsDH842A (from the plasmid) interact, increasing signaling through RcsD. It is also possible that even this low level of RcsDH842A allows some titration of IgaA. The other plasmid that gave unexpected results was RcsD326-C. In a wild-type host, this construct acts much like the vector control (low activity without PMBN, increase with PMBN), as if it lacks the inhibitory activity of some of the other constructs. While this might suggest that it is not made in significant amounts, it is clearly able to stimulate activity, independent of PMBN, in an rcsD541 host (S3F Fig). It seems likely therefore that for this construct, interaction with RcsD+ in the WT host interferes with the ability of the RcsD326-C both to signal (low activity – or + arabinose in rcsD+ host) and to inhibit (vector-like activity with PMBN, rather than the inhibitory activity seen with ABL-Hpt, containing RcsD686-C). These results would suggest that the interaction with the chromosomal RcsD requires sequences upstream of the ABL domain, quite likely the defective HisKA and/or HATPase domains. HisKA domains have been shown to contain dimer interfaces in histidine kinases, and it seems quite possible that the mutant HisKA domain here participates in RcsD/RcsD interactions [25].
These results all suggest that in the absence of arabinose, the pBAD-RcsD constructs express levels of RcsD comparable to the chromosomal level, able to complement and signal but lacking the inhibitory activities seen only with high levels of RcsD achieved after induction with arabinose. To further confirm the behavior of the truncated RcsD proteins, selected alleles were introduced into the bacterial chromosome in place of the native rcsD gene. In these strains, RcsD should be expressed from the native promoter, at the native level. These alleles could generally be introduced into an rcsB deletion, where signaling is off, but some alleles were difficult to isolate or were clearly unstable in rcsB+ cells. rcsD326-C and rcsD Δ48-304 could be introduced into the rcsB+ strain, but cells became quite mucoid and had significant PMBN-independent signaling; RcsD ABL-Hpt was better tolerated (S3G Fig). These results parallel the observations with the RcsD plasmids in an rcsD541 strain (Fig 3B; S3F Fig).
Mutant strains were further tested for the ability to support deletion of igaA, by P1 transduction from a bioH::kanR igaA::chlR donor (EAW66, containing a bioH::kan mutant closely linked to igaA::chlR, in a strain containing rcsD541), selecting for kanamycin resistance and then testing kanamycin resistant colonies for linkage of the igaA::chlR (S3H Fig). In a recipient defective for rcsB, rcsC, or rcsD, the linkage of the bioH::kan and igaA::chlR markers is >70%; in a strain WT for the Rcs phosphorelay, linkage was <1 chlR/100 kanR. Mutations in rcsD were introduced into the chromosome, and then used as recipients for P1 transduction (S3H Fig). Strains carrying the rcsD541 and rcsD841* mutations tolerated loss of igaA well, as expected for strains null for rcsD (S3H Fig).
rcsD326-C, carrying all of the cytoplasmic regions of RcsD, did not tolerate loss of IgaA (S3H Fig). Although chlR colonies were isolated, those colonies had unstable phenotypes; restreaking yielded colonies that were not as mucoid or fluorescent as the parent strain, strongly suggesting that the igaA deletions were only surviving when mutations or deletions of components of the Rcs phosphorelay genes were also present. These results are fully consistent with the behavior of rcsD326-C on plasmids (Fig 3B), and consistent with the model that there are critical regulatory contacts between IgaA and RcsD not only in the periplasm but in the cytoplasmic domains as well. Thus, while rcsD326-C was negative in the bacterial two-hybrid interaction with IgaA (Fig 2B), the continued dependence on IgaA for viability is consistent with it retaining a critical contact with IgaA. As expected, deleting rcsB as well (EAW54, S3H Fig) allowed introduction of the igaA::chlR mutation.
Not all rcsD alleles could be introduced into the chromosome. rcsD Δ45-304, which retains a strong interaction with IgaA (Fig 2B) but interferes with cell growth when expressed from a plasmid in rcsD541, even without induction (Fig 3B), was lethal, and the chromosomal version of this mutant could not be constructed without accumulating secondary loss-of-function mutations in rcsD or rcsB. A chromosomal mutant derivative with a slightly longer periplasmic domain deletion, rcsD Δ48-304, could be constructed, but was mucoid, constitutively active (S3G Fig) and did not tolerate introduction of the igaA deletion (EAW106, S3H), consistent with a critical RcsD-IgaA contact that participates in repression beyond the periplasmic region. This allele may be modestly defective for phosphorelay function, and therefore is better tolerated than the rcsD Δ45-304 allele.
RcsD686-C (ABL-Hpt), which we expect to lack all regions involved in IgaA interaction, led to a lower level of signaling (S3G Fig), is non-mucoid, and, as expected, tolerates loss of igaA (EAW108, S3H Fig). We would suggest, based on its phenotypes, that this construct is not fully active for passing signal from RcsC to RcsB.
These results lead to the following conclusions: 1) the RcsD periplasmic region is essential for full interaction with IgaA and, most strikingly, for full inhibition by IgaA, but is not sufficient for binding or titration of IgaA. This is most consistent with a direct interaction of the RcsD periplasmic loop and IgaA. The precise role of the trans-membrane (TM) regions flanking the periplasmic loop have not yet been explored. 2) An additional region or regions of interaction exist, in the cytoplasmic PAS-like domain of RcsD; this interaction is not sufficient to allow IgaA-dependent repression, but presumably improves binding in the presence of the TM/periplasmic region (thus allowing binding to and titration of IgaA) and contributes significantly to repression by IgaA. 3) Constructs with the Hpt domain but lacking the periplasmic loop (or lacking both the periplasmic region and TM helices) of RcsD are capable of Rcs induction-independent signaling, presumably because they are blind to IgaA repression. 4) The Hpt domain on its own, or the full cytoplasmic domain, is recessive to RcsD+, and is thus not able to constitutively signal in the presence of functional RcsD.
Critical residues in the cytoplasmic PAS-like domain of RcsD
Alanine scanning mutagenesis of individual conserved residues in the PAS domain region of RcsD was carried out in the pBAD-RcsD plasmid. Plasmids were initially screened for level of fluorescence in an rcsD541 mutant strain, in the absence of arabinose. In this assay, functional RcsD expressed from the plasmid reduces fluorescence by complementing the rcsD541 allele, nulls would be expected to not affect fluorescence, and mutations that were capable of passing phosphate from RcsC to RcsB but were less sensitive to IgaA repression might be expected to have higher fluorescence (see right-hand panel in Fig. 3B, for example). Unexpectedly, many of the plasmids that appeared to give strong signals and were thus thought to possibly have become blind to IgaA instead had acquired stop codons. These mutants were not further investigated here. Thus, we instead focused on alanine mutants that retained function, measured by the ability to complement the rcsD541 mutant, reducing the elevated signal found in this mutant to the lower level found in rcsD+ strains (compare lane 1 and lane 3, S4A Fig). Among the six mutants screened, one was striking in that it was unresponsive to PMBN induction, suggesting that it somehow locked RcsD into an “off” configuration. This mutant allele, RcsD T411A, was further analyzed.
The rcsD T411A mutation was introduced into the chromosome and tested for its response to PMBN (Fig 4A). The mutant had a lower basal level of Rcs signaling and, as was seen with the plasmid-borne copy, this mutant had a very muted response to PMBN. A22, an inhibitor of MreB, and mecillinam have also been reported to induce the Rcs System [3, 26], and we confirmed that induction with our reporter (Fig 4A, WT). The T411A mutant also failed to respond to A22 or mecillinam (Fig 4A, red bars).
We can imagine two general ways in which T411A might block induction. It might affect the ability of signal to move through the phosphorelay, possibly locking RcsD in the “phosphatase” confirmation. In this case, we might expect it to be indifferent to the presence of IgaA. In the alternative model, rcsDT411A is locked off because it no longer releases the interaction with IgaA when signal is received; if so, it will still be sensitive to loss of IgaA. To test whether the RcsD T411A was causing a “locked” state in which the protein could no longer pass signal to RcsB, we tried to delete igaA in the chromosome. The strain did not tolerate IgaA deletion (EAW121, S3H Fig), which suggested that RcsD T411A abrogates activation by increased or changed interaction with IgaA.
We next turned to IgaA to begin to identify the regions likely to interact with RcsD. Based on our observations with the RcsD truncations, we would predict at least two regions of interaction betweenRcsD and IgaA: between the RcsD periplasmic region (critical for IgaA regulation) and the periplasmic region of IgaA, as well as additional important interactions between the cytoplasmic PAS-like region of RcsD (and possibly other regions) and cytoplasmic domains of IgaA.
The periplasmic domain of IgaA (see Fig 4B) has previously been found to interact with RcsF [8]. Here, we find that deletion of the periplasmic domain (IgaA Δ384-649) fully abrogates the interaction of RcsD and IgaA; T411A modestly restores this interaction, consistent with the mutant leading to increased interaction within the cytoplasmic regions (Fig 4B and S4B Fig). Deletion of either cytoplasmic loop 1 or cytoplasmic loop 2 of IgaA had essentially no effect on the interaction with wild-type RcsD, suggesting that the primary interactions that drive the bacterial two hybrid signal are between the periplasmic regions. Periplasmic point mutation L643P is a stable protein (S4C Fig) that caused a partial loss of function mutant in igaA [27]; this mutation led to loss of interaction of IgaA and RcsD (S4D Fig). However, other alleles at this position (L643A) or nearby did not disrupt interaction or activity (S4D and S4E Fig), suggesting that L643 is not itself a critical residue but that L643P may disrupt IgaA folding or localization. The specific regions within the IgaA periplasmic domain that contact RcsD remains for future analysis.
Chromosomal deletions of either one of the igaA cytoplasmic loops (Δ36-181, cyt1; Δ263-330, cyt2) or of the periplasmic loop (Δ384-649, peri) (see Fig 4B) were constructed in cells mutant for rcsD and carrying the PrprA-mCherry reporter. A complete igaA deletion was used for comparison. Introduction of the RcsD plasmid, even in the presence of glucose (low levels of RcsD expressed) was poorly tolerated in all the igaA deletions, with secondary mutations arising at a rapid rate (see inset, Fig 4C). Therefore, assays in liquid were considered untrustworthy, and the phenotypes of the primary transformants were evaluated on agar plates (Fig 4C).
Transformation of the RcsD plasmid into cells carrying a deletion of igaA, the igaA periplasmic domain (igaA Δ384-649) or the second cytoplasmic domain (igaA Δ263-330) gave rise to highly mucoid growth, consistent with lack of IgaA function. Introduction of the RcsD+ plasmid into cells deleted for the first cytoplasmic domain (igaA Δ36-181) gave less mucoid growth, although the PrprA-mCherry reporter was well expressed compared to WT and rcsD541 (Fig 4C left panel), consistent with an important role for this domain of IgaA as well. The plasmid expressing RcsDT411A rather than RcsD+ was introduced into these strains. This mutation was unable to decrease the signal in the full deletion of igaA or the deletion of the periplasmic domain. However, it reduced mucoidy and allowed more robust colony growth in cells carrying the second cytoplasmic domain deletion (Fig 4C). This result is most consistent with RcsDT411A improving the interaction with IgaA cytoplasmic loop 1 and therefore abrogating induction (Fig 4A).
Analysis of RcsC domains and involvement in signaling
From previous work, it is clear that RcsC plays an essential role in signaling in the Rcs phosphorelay, as the source of phosphorylation [16]. However, as shown above, full length RcsC did not interact with IgaA in the bacterial two hybrid assay. The plasmids expressing T18 and T25 fusions to full-length RcsC interfered with cell growth, and also did not interact with RcsD, and thus while our data strongly supports the interaction of IgaA with RcsD, we are cautious in interpreting this negative result with RcsC and IgaA. A construct expressing only the cytoplasmic domains of RcsC (rcsC326-C) interacts well with RcsD as well as with the cytoplasmic regions of RcsD, although a bit less strongly (S5A Fig). It interacts as well with a version of RcsD missing residues beyond aa 683 (RcsDN-683), but not at all with RcsDN-525, suggesting that the region between aa 525 and 683 of RcsD, including its inactive HATPase domain is essential for this interaction (S5A Fig). Given that phosphate flows from the C-terminal REC domain of RcsC to the RcsD Hpt domain, we would predict a further, possibly transient, interaction between the RcsC REC domain and the RcsD Hpt domain.
Unlike RcsD, RcsC constructs in pBAD24 were often cytotoxic, causing massive cell lysis without any detectable increase in PrprA-mCherry signal above background, and slow growth even in rich defined glucose media, where the pBAD promoter should be only modestly active. To avoid this overproduction toxicity, deletions and substitutions of interest were introduced into the chromosomal copy of rcsC and tested for response to induction by PMBN. RcsC carrying a mutation in the kinase active site (H479A) had low activity and was not responsive to PMBN, as expected (Fig 5A). Note that the activity in this mutant is more like the wild-type strain without PMBN than like the deletion of rcsC (Fig 5A), in support of experiments reported by Clarke et al that the active site His is not necessary for phosphatase activity [16].
Unexpectedly, a deletion of the periplasmic portion of RcsC, leaving the TM helices (RcsCΔ48-314), had a lower basal level of signal than WT, yet responded strongly to PMBN and A22 (Fig 5 A, B). This allele still requires RcsF for PMBN signal detection, which suggests that this signal comes from RcsF through IgaA to RcsD to RcsC (Fig 5A). Cells carrying the rcsCΔ48-314 mutation tolerate IgaA deletion, although they become mucoid and unstable (EAW70, S3H Fig); deletion of rcsC or rcsCH479A was unaffected by loss of igaA (S3H Fig). We suggest that this allele has a modestly decreased ability to signal, which in other experiments allows cells to support deletion of igaA. Overall, this result strongly suggests that IgaA regulation of the phosphorelay and signal transduction via RcsF are not acting through the periplasmic loop of RcsC.
Although the periplasmic region is not necessary for RcsC function, it would appear that membrane association is. Cells carrying a deletion of the membrane spanning portion (RcsC326-C) acted in a similar manner to an rcsC deletion, with a constitutive level of reporter expression and no response to PMBN (Fig 5A). Consistent with a loss of function for the rcsC326-C allele, the deletion of igaA could be introduced into this strain, and cells remained non-mucoid (EAW56, S3H Fig). A chimeric construct in which the MalF TM and periplasmic region replaced the rcsC periplasmic region restored the ability of the cell to respond to PMBN (S5B Fig), albeit with a higher basal level of signaling in the absence of PMBN. Finally, a series of periplasmic deletions with different linker lengths all responded to PMBN to some extent, although constructs with shorter turns had somewhat lower basal levels (S5B Fig).
Although these results demonstrate that the RcsC periplasmic region is not necessary for sensing the OM stress signal elicited by PMBN, it seemed possible that other inducing stresses, such as mecillinam or A22, might act in a way that was dependent upon the periplasmic region of RcsC. This was tested in our system and demonstrated that the RcsCΔ48-314 mutation still showed induction in response to all three stimuli (Fig 5B). These additional Rcs stimuli elicited an increase in PrprA-mCherry that had less dynamic range and more cell death than PMBN. At the published concentrations, A22 (5 µg/mL; Sigma) and Mecillinam (0.3 µg/mL; Sigma) could induce the WT strain and the RcsC Δ48-314 strain, while RcsDT411A failed to respond (Fig 4A). Therefore, for A22 and Mecillinam, as for PMBN, the rcsC periplasmic region is not required for sensing and responding to signal.
DISCUSSION
The results reported here provide a new view of how IgaA transduces inducing signals within the complex Rcs phosphorelay (Fig 1A). IgaA, a multipass membrane protein, is a strong negative regulator of Rcs. As previously described, signals, such as PMBN, believed to disrupt LPS interactions, or A22 and Mecillinam, peptidoglycan disruptors, change the nature of the RcsF/IgaA interaction. This leads to a change, presumably a decrease, in IgaA’s interaction with the downstream phosphorelay. We find that the point of interaction of IgaA is with the phosphotransfer protein RcsD, rather than with the RcsC histidine kinase. In fact, while RcsC function requires membrane association, the TM and periplasmic sequences of RcsC are not required (Fig. 5, S5B Fig). The change in the IgaA-RcsD interaction frees RcsC-generated phosphate to flow from RcsC to RcsD, and from there to RcsB, activating signaling downstream of RcsB. Deletion and mutation analysis of RcsD identified multiple regions important for IgaA-dependent regulation, separate from the regions critical for phosphorelay signal flow from RcsC to RcsD. These observations help to explain why RcsD includes not only an Hpt domain but also trans-membrane and signaling domains. We suggest that the use of RcsD as the direct target of IgaA has allowed the development of a poised signaling complex, without impinging on structures necessary for histidine kinase activity. In addition, this branched signaling pathway allows the possibility of other signals directly regulating RcsC activity.
Multipartite interactions of RcsD and IgaA regulate signaling: anchors and switches
Our analysis of the regions of RcsD and IgaA necessary for interaction and regulation suggest multiple points of contact between these proteins, each with different roles in regulation. The first contact is in the periplasmic loops of these two proteins. The periplasmic domain of RcsD (amino acids 45-304) is necessary but not sufficient for repression by IgaA; the periplasmic region of IgaA (aa 384-649) is similarly essential for IgaA function and drives the interaction of IgaA and RcsD in a bacterial two-hybrid assay (Fig 2B, Fig 4B). We assume that these periplasmic domains directly contact each other (Fig 6). Because others have demonstrated that overproduction of IgaA missing the periplasmic domain can support repression and allow depletion of wild-type IgaA [8], we suggest that this strong periplasmic interaction provides an anchor for interaction with RcsD, but likely not the region in which signal is sensed. Thus, the periplasmic contact can be bypassed by overproduction, but is critical for repression at normal levels of the interacting proteins.
The cytoplasmic PAS-like domain of RcsD is also necessary for regulation, interaction in the bacterial two-hybrid assay and interaction as judged by titration of IgaA (Fig 2B, Fig 3A). It seems likely that this cytoplasmic domain contacts one or both cytoplasmic loops of IgaA. Both cytoplasmic loop 1 and cytoplasmic loop 2 of IgaA are necessary for RcsD to function properly (Fig 4C), in agreement with previous work [8]. Because the bacterial two-hybrid interactions are primarily driven by the periplasmic domains (Fig 4B), the contacts of RcsD with the IgaA cytoplasmic loops are likely to be weaker. We do not currently have any direct evidence that cytoplasmic loop 2 is contacting RcsD, but certainly deletion of this loop, like deletion of the periplasmic region, abrogates repression (Fig 4C). We suggest that the interaction of cytoplasmic loop 1 and RcsD, in the region around T411, constitutes the regulatory switch for this system. Deletion of loop 1 is the least detrimental in terms of bacterial growth and signaling (Fig 4C), suggesting that the contacts outside cytoplasmic loop 1 are sufficient for enough IgaA repression of RcsD to support viability. Our model suggests that the additional repressive interaction in loop 1 is normally lost upon Rcs stimulus (in the presence of PMBN, for instance), and that the anchor contacts in the periplasm and with IgaA loop 2 ensure that signaling is never so high that the cell dies. In the T411A mutant, this stimulus-sensitive contact becomes stronger, so that the system becomes uninducible (Fig 4A). This can be seen in the bacterial two-hybrid assay as some restored interaction in the absence of the IgaA periplasmic region (Fig 4B).
In work by Collet and coworkers, overproduction of cytoplasmic loop 1 [8] was, by itself, capable of repressing an Rcs reporter to a similar extent to that seen with both cytoplasmic regions, further supporting a critical role for this region of Rcs.
RcsD is an unusual phosphorelay protein
Phosphate flow in complex phosphorelays such as Rcs is from His (kinase domain) to Asp (response regulator domain of RcsC) to His (RcsD phosphotransfer protein) to Asp (RcsB response regulator domain). RcsD, is a large inner membrane protein with many additional domains; its domain organization suggests that duplication of an ancestral protein may have given rise to RcsC and RcsD. Our results suggest critical roles for these additional regions of RcsD.
Consistent with its role as an anchor for IgaA, alignments suggest significant regions of conservation within the periplasmic domain of RcsD, apparently more so than the similarly sized RcsC periplasmic domain, which we show here is not critical for signaling (Fig 5A, S6 Fig). There is significant conservation as well in the truncated PAS domain, but less conservation in the inactive HATPase domain than in the active parallel RcsC domains. Future work will be necessary to identify the periplasmic interaction points of RcsD with IgaA and to understand whether the RcsD ATPase domain plays any critical role in regulation.
Alternative signaling pathways remain to be understood
The complexity of the Rcs phosphorelay provides opportunities for signals to regulate RcsB activity independently of the RcsF-IgaA-RcsD interaction network. Some transcription factors interact directly with RcsB, independent of its phosphorylation, to make heterodimers that regulate specific sets of genes (reviewed in [1]). In addition, there is evidence for activation of RcsB-dependent genes, dependent upon RcsC and RcsD, but independent of RcsF. The two cases in which this has been reported involve overproduction of the DjlA DnaJ-like chaperone and mutation in the periplasmic disulfide bond formation protein, DsbA, possibly suggesting that alterations in protein folding may be the inducing event [1, 16, 28].
One other unexplored aspect of our work is the possible expression of low levels of the C-terminal domains of RcsD, to produce a short phosphotransfer protein that would not be subject to IgaA regulation. For instance, the modest activity of RcsDN-522 and RcsDN-683 in the rcsD541 host (S3A Fig) was unexpected. Because this same increase was not seen when the host contained other rcsD alleles (rcsD841*, for instance), we suggest it may be due to low level expression of a C-terminal fragment of the chromosomal RcsD protein able to transfer phosphate from RcsC to RcsB. In other experiments, we found that unplanned stop codons were found in some plasmids expressing rcsD alanine scan mutants. These plasmids, rather than acting like nulls, had activity significantly above that of a null strain, again suggesting that they might be expressing a C-terminal fragment of RcsD. Whether this is ever made under wild-type physiologically relevant conditions remains to be determined but would provide the possibility of an IgaA-resistant signaling pathway.
Overall, while a critical step in the best understood signaling pathway is clarified here, there is still much to learn about Rcs, other modes of signaling to the phosphorelay, and exactly how the IgaA/RcsD interactions modulate phosphate movement from RcsC through RcsD to RcsB. Given the range of genes regulated by RcsB, and the importance of these genes for bacterial behavior, the options for multiple ways for the system to be regulated may not be surprising.
Materials and Methods
Bacterial growth conditions and strain construction
Cells were grown in LB with appropriate antibiotics (ampicillin 100µg/mL, kanamycin 30-50µg/mL, chloramphenicol 10µg/mL for the cat cassette in cat sacB strains and 25µg/mL for others (chlR), tetracycline 25µg/mL, gentamicin 10µg/mL, zeocin 50µg/mL); 1% glucose was added in some cases to reduce basal level expression of PBAD and PLac promoters. For fluorescence/growth assays, strains were grown in MOPS minimal glucose or minimal glycerol (Teknova). Strains were constructed via recombineering and/or P1 transduction with selectable markers, as outlined in S1 Table. Strains, plasmids,oligonucleotides and gBlocks used in this study are listed in S1-S3 Tables. Oligonucleotides and gBlocks were from IDT DNA, Coralville, IA.
For recombineering, cells carrying the chromosomal mini-λ Red system or the plasmid-borne Red system (pSIM27) were grown in LB, without or with Tetracycline respectively, at 32°C to an OD600 of ∼0.4-06. At mid-log, cultures were transferred to a water bath at 42°C to induce expression of the λ-Red system for 15 minutes and then immediately chilled in an ice-water slurry for 10 minutes prior to washing in sterile ice-cold water to make electrocompetent cells. 100 ng of ss oligo DNA or dsDNA (PCR product or gBlock) were used in the electroporation; 1 ml of LB or SOC was added for recovery before plating on selective plates [29]. Truncations and point mutations were introduced in place of the wild-type chromosomal copies of genes, leaving no marker or scar, unless otherwise indicated, by first inserting the counter-selectable ara-kan-kid cassette from CAI_91 and simultaneously deleting the gene of interest, and then replacing it with the desired allele, provided either as a PCR product or a gBlock. This cassette, a gift of C. Ranquet (BGene Genetics, Grenoble), expresses the Kid toxin under the control of an arabinose-inducible promoter. Cells carrying the ara-kan-kid counter-selectable marker cassette were grown with added 1% glucose in the media to repress. Counter-selection for removal of the ara-kan-kid cassette was done on LB-1% arabinose plates. All plasmid and chromosomal mutations were confirmed by sequencing using flanking primers.
DNA and strain manipulation and mutagenesis
Polymerase chain reactions were performed using Pfu Ultra II polymerase (Agilent) or Clontech Hifi polymerase (Takara). Primers used in this study are listed in S3 Table. PCR products were purified using column purification (Qiagen) according to the manufacturer’s instructions. Gibson assemblies were performed using the Clontech In-Fusion HD Cloning Kit (Takara) and transformed into either Clontech Stellar Cells or NEB Turbo cells containing LacIq.
Alanine-scanning mutagenesis was carried out by SGI-DNA (San Diego, CA), using their BioXP system, on pBAD24-RcsD (pEAW11). We ordered single mutants targeted to conserved residues within the cytoplasmic region of RcsD, from residue 326-683. Plasmids were first transformed into Stellar E. coli (Clontech), extracted and retransformed into EAW19, screening for fluorescence on minimal glucose-ampicillin agar plates, in comparison to cells carrying pBAD-RcsD+ or the empty pBAD vector. Out of 35 mutants screened, ten had fluorescence levels comparable to the pBAD-RcsD+ control; six were further studied (S4A Fig). Another 17 had higher fluorescence than either the pBAD vector or the pBAD-RcsD+ control, but sequencing of these isolates showed that they had all contained, in addition to the designed mutation, additional unexpected stop codon mutations and were not further studied here.
Bacterial Adenylate Cyclase Two-Hybrid Assay
In the bacterial adenylate cyclase two hybrid assay (BACTH), an adenylate cyclase mutant strain is used to assay for beta-galactosidase activity engendered when the T18 and T25 portions of adenylate cyclase are reconstituted, allowing cAMP/CRP to activate the lac operon. On their own, T18 and T25 will not form adenylate cyclase efficiently unless they are fused to two interacting proteins [21]. Tags were C-terminal to avoid interference with protein insertion into the membrane.
The RcsD and RcsC fusion proteins were tested for determine if they were functional and thus presumably membrane-localized. Plasmids expressing RcsD-T25 and RcsC-T25 were introduced into strains containing deletions for those two genes; after transformation, the cells were transduced with P1 grown on a strain NM357, containing igaA::chlR, selecting for chloramphenicol resistance. In a strain deleted for rcsD or rcsC, the igaA deletion can be introduced by P1 transduction. However, the fusion plasmids blocked the ability of cells to be transduced with igaA::chlR, consistent with them complementing their respective deletions (S2E Fig).
igaA co-transduction frequencies
bioH/igaA co-transduction frequencies were used to determine which strains could support loss of IgaA. bioH, at 3544844 nt, is linked to igaA (position 3526469). The bioH::kanR mutant from the Keio collection [30] was introduced by P1 transduction into an rcsD541 igaA::chlR mutant (EAW17), selecting for kanamycin resistance and retention of chloramphenicol resistance (igaA::chlR), to create EAW66. Because rcsD is inactive in this strain, it can tolerate loss of igaA. P1 transduction from this donor to recipient strains was carried out, selecting for Kanamycin Resistance and then screening 50-100 colonies for linkage to igaA::chlR. In rcsB, rcsC or rcsD null recipients, the co-transduction frequency was 78%. In a wild-type strain, the linkage dropped to zero, consistent with the known lethality of an igaA deletion [3, 31] (S3H Fig).
Fluorescence assays
Fluorescence assays for Rcs activation were performed in 96 well plates in a Tecan Spark 10m spectrophotometer. These strains carried a transcriptional fusion of mCherry, at the ara locus, to the promoter for sRNA RprA, as a reporter for Rcs pathway activation, referred to here as PrprA-mCherry. Fluorescence of cells was measured in MOPS glucose minimal media (Teknova) unless otherwise stated. The pBAD24 plasmid was used for overexpression of RcsD fragments in strains expressing araE constitutively to ensure homogenous arabinose uptake [32]. For cells expressing proteins from pBAD, overnight cultures in MOPS minimal glucose were washed with MOPS minimal glycerol to eliminate residual glucose, then diluted into fresh MOPS minimal glycerol media (.05% glucose, .5% glycerol) with 0.02% arabinose or 0.2% glucose as an uninduced control. Polymyxin B nonapeptide (PMBN; Sigma), a non-toxic polymyxin derivative, was used at 20 ug/mL to produce Rcs induction. To check for Rcs induction by other known compounds, A22, an MreB inhibitor was used at 5ug/mL, and mecillinam was used at 0.3ug/mL.
Each strain/condition combination was performed in technical triplicate on the plate, with biological replicates performed on different days. Optical density and mCherry fluorescence were monitored every fifteen minutes for seven hours. At the end of six hours, measurements of fluorescence at equivalent OD600 values (0.4 +/- 0.03 after starting at OD600 .03-.05) were converted to bar graphs of fold change of fluorescence with respect to the wild type strain. Some strains arrested growth early and never achieved 0.4 OD600, and the OD600 at 6 hours for those are noted on the graph. Six hours marks the time when the wild type strain begins to transition to stationary phase, and ODs become less interpretable due to cell aggregation in the well bottom.
Supporting Information
Supplemental Tables
S1 Table: Strains used in this study.
S2 Table: Plasmids used in this study.
S3 Table: Primers used in this study.
Acknowledgments
We thank S. Buchanan and members of her laboratory for discussion throughout this work, as well as members of the LMB. Erin Wall was supported during much of this work by a PRAT Fi2 fellowship GM123943 from NIGMS. This research was supported in part by the Intramural Research Program of the NIH, National Cancer Institute, Center for Cancer Research.