Extracytoplasmic function or ECF sigma factors are the most abundant class of alternative sigma factors in bacteria. Members of the rpoE subclass of ECF sigma factors are implicated in sensing stress in the cell envelope of Gram-negative bacteria and are required for virulence in many pathogens. The best-studied member of this family is rpoE from Escherichia coli, encoding the σE protein. σE has been well studied for its role in combating extracytoplasmic stress, and the members of its regulon have been largely defined. σE is required for viability of E. coli, yet none of the studies to date explain why σE is essential in seemingly unstressed cells. In this work we investigate the essential role of σE in E. coli by analyzing the phenotypes associated with loss of σE activity and isolating suppressors that allow cells to live in the absence of σE. We demonstrate that when σE is inhibited, cell envelope stress increases and envelope integrity is lost. Many cells lyse and some develop blebs containing cytoplasmic material along their sides. To better understand the connection between transcription by σE and cell envelope integrity, we identified two multicopy suppressors of the essentiality of σE, ptsN and yhbW. yhbW is a gene of unknown function, while ptsN is a member of the σE regulon. Overexpression of ptsN lowers the basal level of multiple envelope stress responses, but not that of a cytoplasmic stress response. Our results are consistent with a model in which overexpression of ptsN reduces stress in the cell envelope, thereby promoting survival in the absence of σE.
Citation: Hayden JD, Ades SE (2008) The Extracytoplasmic Stress Factor, σE, Is Required to Maintain Cell Envelope Integrity in Escherichia coli. PLoS ONE 3(2): e1573. doi:10.1371/journal.pone.0001573
Academic Editor: Steve Sandler, University of Massachusetts, United States of America
Received: December 18, 2007; Accepted: January 13, 2008; Published: February 6, 2008
Copyright: © 2008 Hayden, Ades. 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.
Funding: This work was supported by grant MCB-0347302 from the National Science Foundation. The funding agency did not play any role in this study or in preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
All organisms have stress responses that allow them to sense and respond to damaging conditions by altering gene expression. An additional level of complexity is introduced when the inducing signal is sensed on one side of a membrane and that information must be communicated across the membrane for a response to be generated. In Gram-negative bacteria this intercompartmental signaling is required to maintain the cell envelope, which consists of the inner and outer membranes, periplasmic space, and peptidoglycan layer .
The cell envelope is a complex, dynamic compartment that is crucial for the survival of the cell. It is not a static structure and can be remodeled in response to environmental conditions. The chemical environment of the cell envelope is distinct from that of the cytoplasm. The envelope lacks ATP, is oxidizing, and can be subject to fluctuations in ionic strength due to passive diffusion of small molecules through outer membrane porins . As such, Gram-negative bacteria possess stress responses that are uniquely targeted to the cell envelope. These stress responses include the CpxAR (Cpx), BaeRS (Bae), and Rcs phosphorelays, the response governed by the alternative sigma factor σE, and the phage shock (PSP) response –. Each of these responses is activated following perturbation of particular components of the envelope.
Although stress responses are important for reacting to damaging conditions, many stress proteins also play important roles in basic cellular physiology. This is particularly true for the σE-dependent response in E. coli, as the rpoE gene, which encodes σE, is essential for viability . Despite the wealth of information about the role of σE in response to cell envelope stress and the identification of the σE regulon, the essential role of σE is still unclear , . One suppressor of ΔrpoE lethality, a deletion of the ydcQ gene, has been identified . However, the function of this gene is not well understood and it is not clear how it suppresses the essentiality of rpoE.
σE is activated by stresses that interfere with the folding of outer membrane proteins (OMPs) such as heat shock, overexpression of OMP genes, and mutations in genes encoding chaperones required for OMP folding –. In unstressed cells, σE activity is low because σE is sequestered at the inner membrane by the antisigma factor RseA (Fig. 1) , . RseA is a single-pass inner membrane protein that binds to σE and prevents σE from interacting with RNA polymerase –. During envelope stress, RseA is degraded in response to unfolded porins by the sequential action of two inner membrane proteases, DegS and RseP, followed by the cytoplasmic protease ClpXP (Fig. 1) –. A periplasmic protein, RseB, binds to the periplasmic domain of RseA and enhances inhibition by RseA, protecting it from proteolysis (Fig. 1) , , , . Continual degradation of RseA is required to provide the cell with sufficient free σE to support viability, and deletion of either degS or rseP is toxic due to the stabilization of RseA and consequent sequestration of σE , , . σE can also be activated independently of the RseA-dependent stress-signaling pathway by the cytoplasmic alarmone ppGpp, whose levels change in response to nutrient availability .
Figure 1. The σE-dependent extracytoplasmic stress response.
σE is held at the membrane by the antisigma factor RseA. RseB binds to the periplasmic domain of RseA and protects RseA from proteolysis. Unfolded OMPs activate the protease DegS, which cleaves the periplasmic domain of RseA. The partially degraded RseA is now a substrate for RseP. RseP cleaves RseA further, releasing the cytoplasmic domain of RseA bound to σE. This remaining domain of RseA is degraded by ClpXP thereby freeing σE to interact with RNA polymerase and direct transcription of its regulon. The major classes of known genes in the σE regulon are indicated. The alarmone ppGpp and protein DksA can activate σE-dependent transcription once σE is released from RseA, but are not shown for clarity. (IM inner membrane, OM outer membrane).doi:10.1371/journal.pone.0001573.g001
As expected from its role in the stress response, the σE regulon includes genes encoding periplasmic foldases, proteases, and chaperones that aid in OMP folding. In addition, σE transcribes an array of biosynthetic enzymes that are involved in phospholipid, fatty acid, LPS, and membrane-derived oligosaccharide synthesis and transport, and a number of other cell envelope proteins including lipoproteins, inner membrane proteins, and envelope proteins of unknown function , , . The σE regulon also includes several genes that encode major components of complexes responsible for ushering LPS and OMPs across the periplasm and properly assembling them in the outer membrane , . Recently σE has been shown to transcribe several small RNAs that decrease the expression of OMPs, providing a mechanism to decrease the flux of OMPs to the envelope during times of stress –. In addition to the envelope-associated proteins, σE-dependent promoters are found upstream of genes involved in cytoplasmic processes such as transcription, translation, DNA replication, and DNA/RNA modification , , .
The σE regulon contains a number of essential genes, and the simplest explanation for why σE is essential is that it is required to transcribe one or more of these genes. Alternatively, loss of expression of multiple non-essential genes dependent on σE for expression could be lethal. However, many genes in the σE regulon are also transcribed by σ70. Chromatin immoprecipitation followed by genomic microarray (ChIP-chip) experiments examining the location of σ70 on the E. coli chromosome suggest that approximately 40% of σE-regulated promoters overlap with σ70 promoters . This number is likely to be an underestimation of the extent of dual regulation, since it does not include genes that also have σ70 promoters farther away from the σE promoter . As such, σE may be essential because genes in its regulon that are also transcribed by σ70 are misregulated with lethal effect in the absence of σE.
In this work we took two approaches to better understand the essential role of σE. The first approach was to characterize the phenotypes of cells when σE activity was inhibited to gain insights into what cellular functions were impacted. We demonstrate that σE is required to maintain the integrity of the cell envelope; in the absence of σE the cells developed blebs along their sides and lysed. The second approach was to isolate multicopy suppressors that allow cells to live in the absence of σE activity. Two multicopy suppressors, ptsN and yhbW, were isolated that suppress both the lethality due to inhibition of σE activity and the essentiality of the rpoE gene. ptsN is a member of the σE regulon, while yhbW is not, and neither gene has previously been associated with cell envelope functions. Overexpression of ptsN lowered the basal activity of several envelope-sensing pathways, implying that it may suppress ΔrpoE lethality by reducing stress in the cell envelope.
Inhibition of σE activity is toxic and increases envelope stress
To gain a better understanding of the role of a particular gene in cellular physiology, it is often informative to examine the phenotypes of a deletion mutant lacking the gene of interest. Because σE is encoded by an essential gene, it is not possible to examine the phenotype of a ΔrpoE strain. Therefore we examined the effects of loss of σE activity by inducing the overexpression of its inhibitors, rseA and rseB, encoded under the IPTG-inducible trc promoter on the pTrc99a plasmid, pRseAB. This method efficiently inhibits σE (Fig. 2A) by preventing its association with RNA polymerase –. Overexpression of rseA and rseB and the consequent sequestration of σE were lethal for the bacterium. In liquid culture, the cfu/ml began to decrease after 2 hours (~2.5 generations) following the addition of IPTG and the optical density stopped increasing and began to decrease slightly within 3 hours (~3.5 generation) (Fig. 2B and ). On solid media, the plating efficiency in the presence of IPTG was reduced by three orders of magnitude compared with non-inducing conditions (Table 1). To ensure that the observed phenotypes were due to inhibition of σE and not overproduction of RseA and RseB, we introduced a point mutation into the rseA gene on the plasmid encoding rseA and rseB, changing the aspartate residue at position 11 of RseA to histidine. This mutation abrogates the antisigma factor activity of RseA (Fig. 2A and ). When overproduced along with RseB, the RseAD11H protein reached similar steady-state levels as the wild-type protein, was properly localized to the inner membrane, did not inhibit σE activity, did not induce lysis, and did not reduce the plating efficiency (Fig. 2, Table 1, and data not shown).
Figure 2. Inhibition of σE activity caused cell death.
Strains SEA007 (squares, pRseAB), SEA008 (circles, pTrc99a vector control), and SEA4044 (triangles, pRseAD11H) were grown in LB at 30°C with shaking in a gyrotary waterbath. IPTG was added at OD600 ~0.1 to induce overexpression of rseA and rseB, indicated by the arrow. (A) σE was inhibited by overexpression of rseA and rseB, but not rseAD11H and rseB, as shown by β-galactosidase activity measured from the σE-dependent rpoHP3-lacZ reporter. Miller units (open symbols, right axis) and OD600 (closed symbols, left axis) are plotted at each time point. (B) Overexpression of rseA and rseB, but not rseAD11H and rseB, caused a reduction in the colony forming units and optical density. OD600 (closed symbols, left axis) and cfu/ml (open symbols, right axis) were measured throughout the growth curve. A representative experiment is shown in both panels.doi:10.1371/journal.pone.0001573.g002
Table 1. Plating efficiencies following inhibition of σEdoi:10.1371/journal.pone.0001573.t001
Because σE is known to respond to envelope stress and direct the synthesis of many components of the cell envelope, it is likely that this compartment will be negatively impacted and envelope stress will increase when σE activity is blocked. We measured activation of reporter genes for the Cpx, Bae, and Rcs envelope stress responses (cpxP-lacZ activated by CpxR, spy-lacZ activated by BaeR and CpxR, rprA-lacZ activated by RcsB, Fig. 3). β-galactosidase activity from the envelope stress reporters increased by 2.5 hours following inhibition of σE (Fig. 4). Since activation of these stress responses could be due to a general increase in stress associated with cell death, we also monitored activation of the cytoplasmic stress response mediated by σ32 using the σ32-regulated htpG-lacZ reporter fusion. σ32 activity increased relatively little (~2-fold) following inhibition of σE compared with the larger increases seen for the envelope stress responses (3.5-fold for Cpx, 8-fold for Rcs, and 12-fold for the Bae/Cpx-dependent reporters). These results suggest that loss of σE activity does increase stress in the cell envelope, but does not increase cytoplasmic stress to the same extent.
Figure 3. Diagram of stress regulators and their respective regulon members used as reporters in this work.
All reporter constructs are present as lacZ fusions integrated as a single-copy in the chromosome. The rpoHP3-lacZ fusion is regulated by σE. The degP-lacZ fusion is regulated by σE and CpxR. The cpxP-lacZ fusion is regulated by CpxR. The spy-lacZ fusion is regulated by CpxR and BaeR. The rprA-lacZ fusion is regulated by RcsB. The htpG-lacZ fusion is regulated by σ32. σE, Cpx, Bae, and Rcs all monitor envelope stress, while σ32 monitors cytoplasmic stress.doi:10.1371/journal.pone.0001573.g003
Figure 4. Envelope stress responses are activated following inhibition of σE.
β-galactosidase activity was measured from the htpG-lacZ (σ32, SEA4183), cpxP-lacZ (Cpx, SEA4177), rprA-lacZ (Rcs, SEA4190), and spy-lacZ (Cpx and Bae, SEA4187) reporters. The fold change in Miller units (M.U.) with and without rseA and rseB overexpression (M.U. with IPTG/M.U. no ITPG) is shown for samples taken at 0 (white), 1 (grey), 2 (blue), 2.5 (red), and 3 (black) hours after addition of IPTG.doi:10.1371/journal.pone.0001573.g004
Loss of σE activity does not cause major changes in the composition of membranes
Since inhibition of σE increases envelope stress and σE can transcribe genes that are localized to the inner and outer membranes, we next asked if the envelope membranes were altered. Samples were taken from cultures 2.5 hours following overexpression of rseA and rseB. This time point was chosen for analysis because the cell envelope stress responses were activated, the cfu/ml had started to decrease, and the morphological defects described below were evident. Cell lysates were fractionated by centrifugation through discontinuous sucrose density gradients and several assays were used to detect fractions containing inner and outer membrane components. In each of these assays the results from cultures in which σE activity was inhibited were similar to those from control cultures. No alterations in the overall protein content of the fractions were detected on Coomassie-stained SDS polyacrylamide gels (Fig. 5A). The inner and outer membrane fractions were well separated, as determined by western blots probed with antibodies to the inner membrane protein, FtsH, and the outer membrane protein, FepA (Fig. 5C). Fractions containing inner membrane were also identified using a β-NADH oxidase activity assay (Fig. 5B). Thus, loss of σE activity does not result in gross defects in the steady-state protein composition or density of the inner and outer membranes.
Figure 5. Fractionation of strain SEA007 lysates with and without inhibition of σE on sucrose density gradients.
At 2.5 hours following inhibition of σE, cells were harvested and lysates separated by discontinuous sucrose density gradient centrifugation. Lysates from comparable numbers of cells (as determined by OD600) were loaded on the gradients. (A) Coomassie-stained 12% SDS-polyacrylamide gels of gradient fractions from the control culture (top) and culture in which σE activity was inhibited (bottom). M denotes molecular weight markers. (B) β-NADH oxidase activity (closed symbols) and specific gravity (open symbols) of the fractions in (A), inhibition of σE (circles) and control (squares). (C) Western blots using polyclonal antibodies to detect FtsH (inner membrane) and FepA (outer membrane) in fractions from (A), control (top panels) and inhibition of σE (bottom panels). L denotes whole-cell lysates before fractionation.doi:10.1371/journal.pone.0001573.g005
Morphological changes following inhibition of σE
We next examined the morphology of cells following inhibition of σE. By 2 hours after addition of IPTG to overexpress rseA and rseB (the time that the cfu/ml began to drop), two phenotypes were evident by phase contrast microscopy. Ghosted cells and cells with blebs began to appear in the culture. The blebs formed primarily along the lateral wall of the cells and were found less frequently at either the poles or the septum. Usually only one bleb formed per cell. By 2.5 hours after addition of IPTG, approximately 20% of cells in the culture had blebs (Fig. 6) and the number of ghosted cells increased along with the amount of cellular debris, indicative of lysis. In a typical experiment, approximately 100 cells were viewed at each time point. The location of the blebs and the timing of their appearance with respect to loss of viability were very reproducible, suggesting that the phenotypes were related. By 3–4 hours after addition of IPTG some cells had additional blebs, the number of ghosted cells and the amount of cell debris increased, and many cells clumped together. Similar phenotypes were observed when σE was inhibited by proteolytic stabilization of RseA through depletion of the DegS or RseP proteases (data not shown) and were not found following overexpression of rseAD11H and rseB from pRseAD11HB, providing further evidence that the phenotypes were due to inhibition of σE and not overexpression of rseA and rseB. In addition, the same phenotypes were seen when the cells were grown in glucose or glycerol minimal media supplemented with amino acids indicating that the phenotypes were not a function of growth rate or medium composition.
Figure 6. Cell envelope defects resulting from inhibition of σE.
Cultures of SEA007 and SEA008 were grown as in Fig. 2. Samples were taken approximately 2.5 hrs. after induction of rseA and rseB overexpression. (A–D) Images of live cells are shown using differential interference microscopy (DIC, column 1) and fluorescence microscopy following addition of FM4-64 to visualize membranes (red, column 2), DAPI to visualize DNA (blue, column 3), and expression of YFP to visualize the cytoplasm (green, column 4). The three fluorescent micrographs are overlaid in column 5. (A) Images of the SEA008 control strain in which σE was not inhibited are shown. (B–D) Images of SEA007 following σE inhibition, reveal blebs that contain YFP and stain with DAPI (B and C) and that contain YFP but do not stain with DAPI (D). In (D), the arrow marks a bleb lacking DAPI staining and the arrowhead marks a lysed cell that retained DAPI staining, but lost YFP. Scale bars are 2 µm. Over 1,000 cells were examined by fluorescence microscopy and typical micrographs are represented here. Scanning electron micrographs (E) and transmission electron micrographs (F) of SEA007 following σE inhibition. Scale bars are 1 µm. No blebs were seen on cells in control cultures.doi:10.1371/journal.pone.0001573.g006
To further probe the morphology of cells following inhibition of σE, fluorescent probes were used to detect cytoplasm, nucleic acids, and membranes in live cells. We visualized the cytoplasm by expressing the soluble fluorescent protein, YFP, from an expression construct integrated in the chromosome. Membranes and nucleic acids were visualized respectively with the FM4-64 and DAPI fluorescent dyes, which were added directly to the cultures. Neither the dyes nor cytoplasmic expression of YFP altered the phenotypes associated with inhibition of σE. All of the cells and blebs were outlined with FM4-64, and both the cells and the blebs contained YFP within the FM4-64 boundary, with the exception of the ghosted cells (Fig. 6A–D). The ghosted cells were outlined with FM4-64, but no longer contained YFP (Fig. 6D, indicated by the arrowhead). YFP is not proteolytically labile, therefore the lack of a YFP signal in the ghosted cells is likely to be due to release of cytoplasmic contents upon lysis. All of the cells had DAPI staining, and both the compaction of the nucleoids and the number of nucleoids per cell did not differ significantly from control cells, suggesting that loss of σE activity does not cause major defects in chromosome replication or segregation (Fig. 6A–D). Approximately 50% of the cells with blebs contained DAPI staining within the bleb, in addition to YFP (Fig. 6B,C). We hypothesize that some blebs do not contain DAPI staining (Fig. 6D, indicated by the arrow), either because the nucleoid was simply not drawn into the blebs, or there were not enough nucleic acids present to be visualized with DAPI.
Electron microscopy was employed to determine if there were any additional changes in cell morphology not detectable with light microscopy. Scanning electron micrographs revealed no significant alterations other than cells with blebs and ghosted cells (Fig. 6E). The surface of the cells with blebs and ghosted cells were similar to those of cells from control cultures without overexpression of rseA and rseB. Using transmission electron microscopy (TEM) we obtained several images containing sections through a bleb. In these images, the inner and outer membranes could be clearly traced around the bleb, and cytoplasmic material was present in the bleb (Fig. 6F). No evidence of outer membrane vesiculation, which has been associated with alterations in σE activity , , or other gross morphological defects were observed in the TEM or SEM images.
Isolation of multicopy suppressors of the requirement of σE activity for viability
The phenotypic studies demonstrate that when σE activity is depleted, cell envelope integrity is compromised, resulting in lysis and bleb formation. However, these phenotypes are not readily explained by loss of expression of any gene(s) in the σE regulon (see Discussion). To gain additional information about why σE is essential, we turned to genetics and isolated multicopy suppressors that allowed E. coli to live in the absence of σE activity. Cells containing pRseAB (strain SEA007) were transformed with pools of plasmids from the ASKA ORF library, in which each E. coli ORF was cloned under the control of an IPTG-inducible promoter . Transformants were plated in the presence of IPTG to induce overexpression of rseA and rseB and the gene encoded on the ASKA plasmid. Of approximately 10,000 transformants, 44 putative suppressors were recovered.
Two classes of genes can be isolated with this selection, those encoding genes that restore σE activity and those that no longer require σE for growth. We distinguished between these classes using the chromosomally encoded σE-dependent lacZ reporter gene (rpoHP3-lacZ ) in strain SEA007 and detected σE activity by plating in the presence of Xgal and IPTG. Of the 44 clones isolated in the selection, 26 formed blue colonies, indicating that overexpression of the gene on the plasmid restored σE activity, and 18 formed white colonies, indicating that σE activity was not restored and therefore was no longer required for growth. This latter class contained plasmids that were potential suppressors of the requirement of σE for viability. For each of these 18 plasmids, we verified that suppression was linked to the plasmid, the plating efficiencies when rseA and rseB were overexpressed were no longer reduced, and the morphological defects were significantly reduced.
The selection identified plasmids that permit growth in the presence of the overexpression of rseA and rseB. However, these strains could have low levels of σE activity that contribute to viability. We therefore determined whether the rpoE gene could be deleted when the potential suppressor genes were overexpressed using a cotransduction assay. A nadB::Tn10 allele immediately upstream of a rpoE::kan deletion was introduced by P1 transduction into wild-type strains containing each of the 18 plasmids, and nadB::Tn10 transductants were selected. Because the genes are tightly linked, the rpoE::kan allele will be cotransduced unless rpoE is essential (Table 2). The rpoE deletion was cotransduced in strains carrying 2 of the 18 plasmids (Table 2). For the remainder of the strains, the rpoE deletion was not cotransduced, indicating that the rpoE gene was still required for viability. The two multicopy suppressors of rpoE essentiality were yhbW and ptsN (Tables 1 and 2). yhbW encodes a putative luciferase-like monooxygenase of unknown function. ptsN encodes EIIANtr, a protein related to the enzyme IIA components of PTS carbohydrate transport systems . However, EIIANtr is not known to be associated with any known transporters in E. coli and the Ntr phosphoryl relay is instead thought to play a role in signaling –. ptsN is a member of the rpoE regulon  and we focused on it for further studies.
Table 2. Cotransduction of nadB::Tn10 and rpoE::kandoi:10.1371/journal.pone.0001573.t002
Phosphorylation of EIIANtr is required for suppression
Since EIIANtr is part of a phosphoryl relay and the phosphorylation state of EIIANtr is important for its activity , , we determined whether phosphorylation of EIIANtr was important for suppression. EIIANtr is phosphorylated at His73 and possibly His120 . We constructed two variants, ptsNH73A and ptsNH73A/H120A, and tested them for their ability to suppress the requirement of σE for viability in the cotransduction assay. The rpoE::kan allele could not be cotransduced with nadB::Tn10 (Table 2) suggesting that phosphorylation of EIIANtr is required for survival in the absence of σE. Overexpression of both EIIANtr variants could be detected on Coomassie-stained protein gels indicating that lack of suppression was not owed to instability of the variant proteins (data not shown).
The ptsP and ptsO genes encode EINtr and NPr, respectively, and form a phosphoryl relay that transfers a phosphate from phosphoenol pyruvate (PEP) to EINtr to NPr to EIIANtr , . Since phosphorylation of EIIANtr was required for suppression, we asked whether the other components of the phosphoryl relay were also required by determining whether overexpression of ptsN could still suppress the essentiality of rpoE in strains lacking ptsP or ptsO. These genes were not required for suppression by overexpressed ptsN; nadB::Tn10 rpoE::kan cotransductants were obtained in both strains (Table 2). Additionally, the rpoE gene could not be deleted in ΔptsN, ΔptsO, or ΔptsP strains without overexpression of ptsN. Since the results with the point mutations in ptsN indicate that phosphorylation of EIIANtr is required for suppression, it is likely that paralogous proteins from other phosphoryl relays might compensate for the loss of the Ntr components and phosphorylate EIIANtr in their absence.
PtsN has been shown to bind to and inhibit potassium uptake by the potassium transporter TrkA . K+ is the major cation in the E. coli cytoplasm and participates in a variety of processes related to known functions of σE including adaptation to osmotic stress and maintenance of turgor pressure –. In addition, K+ binds and regulates a number of intracellular enzymes including RNA polymerase (as potassium glutamate) . Therefore, if overexpression of ptsN suppresses the requirement of σE for viability by inhibiting TrkA, rpoE should not be essential in a ΔtrkA strain. However, when a ΔtrkA strain was used as the recipient in the cotransduction assay, only nadB::Tn10 transductants were obtained indicating that ΔtrkA is not a suppressor (Table 2). Furthermore, the trkA gene was not required for suppression by ptsN, as cotransductants were readily obtained in a ΔtrkA strain when ptsN was overexpressed (Table 2).
Overexpression of ptsN lowers envelope stress
The basal levels of several envelope stress responses are lower in a ΔydcQ strain, the other characterized suppressor ΔrpoE lethality . In addition, σE-dependent rpoHP3-lacZ activity is low in a strain containing an unmapped ΔrpoE suppressor . Consistent with these results, overexpression of ptsN lowered activity of the σE-dependent rpoHP3-lacZ reporter in a wild-type rpoE+ strain (Fig. 7 and 8A). To determine whether this effect was specific to σE, or if ptsN overexpression affected other envelope stress responses, we measured the effects of ptsN overexpression on reporters for the following envelope stress responses (Fig. 3): Cpx (cpxP-lacZ reporter), σE and Cpx (degP-lacZ reporter), Bae and Cpx (spy-lacZ reporter), and Rcs (rprA-lacZ reporter). A strain carrying a reporter for the σ32-dependent cytoplasmic heat shock response (htpG-lacZ) was included to assay whether ptsN also lowered cytoplasmic stress (Fig. 3). Overexpression of ptsN significantly lowered expression of the lacZ fusions that are regulated by envelope stress response factors, but had no effect on the σ32-dependent cytoplasmic stress reporter (Fig. 7). The latter result indicates that overexpression of ptsN does not lower β-galactosidase activity per se, nor does it have a dampening effect on overall gene expression.
Figure 7. Overexpression of ptsN reduces cell envelope stress.
β-galactosidase activity from the rpoHP3-lacZ (σE, SEA4254), degP-lacZ (σE and Cpx, SEA4181), cpxP-lacZ (Cpx, SEA4179), rprA-lacZ (Rcs, SEA4199), spy-lacZ (Cpx and Bae, SEA4189), and htpG-lacZ (σ32, SEA4185) fusions was measured in overnight cultures, with and without overexpression of ptsN. Activity of each lacZ fusion was normalized to that of cultures with no ptsN overexpression and set to 100%. Average values and standard deviations from a minimum of 3 experiments are shown.doi:10.1371/journal.pone.0001573.g007
Figure 8. ptsN requires known regulators to reduce stress sensed by the σE, Rcs, and Cpx envelope stress pathways.
In each panel the β-galactosidase activity was measured with and without overexpression of ptsN in a WT strain and in a strain lacking known regulators of the respective stress responses. Measurements are expressed as the percentage of reporter gene activity in the WT strain without ptsN overexpression and this value is noted above the columns. Relative reporter gene activities are shown for (A) the rpoHP3-lacZ fusion in WT and ΔrseA strains (SEA4254 and SEA4228), (B) the rprA-lacZ fusion in WT and ΔrcsB strains (SEA4199 and SEA4203), and (C) the cpxP-lacZ fusion in WT and cpxR::Ωspec strains (SEA 4179 and SEA4251). The inset in C shows the cpxR::Ωspec strain with and without ptsN overexpression. The basal level of expression is >250-fold lower in cpxR::Ωspec cells than in WT cells and is not easily seen on the main plot.doi:10.1371/journal.pone.0001573.g008
ptsN overexpression could reduce the basal levels of envelope stress that activate the stress signaling pathways or trigger a cytoplasmic signal that downregulates the effectors of the signaling pathways independently of cell envelope stress. If the latter idea is correct, then ptsN should lower reporter activity independently of the envelope signal transduction pathways. In a strain lacking the antisigma factor RseA, σE activity is high and unresponsive to known regulatory signals that originate in the cell envelope, but is still responsive to levels of the cytoplasmic signaling factor, ppGpp . Overexpression of ptsN did not lower activity of the rpoHP3-lacZ reporter in a ΔrseA strain, suggesting that signal transduction through RseA is required for ptsN to lower σE activity (Fig. 8A).
If ptsN overexpression does in fact lower the basal level of envelope stress, then the decrease in reporter gene expression should be dependent on the transcriptional activators associated with the envelope stress pathways. Interestingly, in the ΔydcQ strain the decrease in reporter gene expression of the Cpx-regulated cpxR-lacZ reporter was not dependent on the response regulator CpxR . We measured the effects of overexpression of ptsN on the Cpx-dependent cpxP-lacZ and Rcs-dependent rprA-lacZ reporter fusions in strains lacking their respective transcriptional regulators, CpxR and RcsB. In a WT strain, overproduction of ptsN lowered expression of the rprA-lacZ reporter 3.2-fold compared with 1.8-fold in the ΔrcsB strain (Fig. 8B) and lowered expression of the cpxP-lacZ reporter 4.3-fold compared with 2.2-fold in the cpxR::Ωspec strain (Fig. 8C). It has been reported previously that the cpxP-lacZ reporter is not expressed in a ΔcpxR variant of the E. coli strain MC4100 , . However, we were able to reproducibly measure β-galactosidase activity from this reporter in a cpxR::Ωspec variant of strain MG1655, although expression was quite low (~250-fold lower than in a WT background). These results indicate that decreased activity of the reporters due to ptsN overexpression is at least partially dependent on cpxR and rcsB, suggesting that EIIANtr lowers envelope stress sensed by both pathways.
Despite extensive knowledge about the regulatory pathways that control σE activity, the stresses that activate these pathways, and the genes that σE transcribes, the essential role of σE in the cell has remained elusive. In this work we demonstrate that σE is required for maintenance of cell envelope integrity. When σE activity was inhibited, envelope stress sensed by several stress responses increased, indicating that the cell envelope was impacted. Many cells lysed and some developed blebs containing cytoplasmic material along their sides. Interestingly, these phenotypes were not accompanied by major alterations in the composition of the envelope membranes. We identified two multicopy suppressors, ptsN and yhbW, that allowed cells to live in the absence of the rpoE gene. ptsN is a member of the σE regulon  and lowers the basal level of activity of several envelope stress responses by an unknown mechanism when overexpressed.
How does loss of σE activity lead to bleb formation and lysis?
The blebs that form following inhibition of σE resemble bulges formed when E. coli are treated with penicillin-like antibiotics that inhibit peptidoglycan synthesis, or when an inactive variant of the peptidoglycan synthase PBP 1B is overproduced , . These treatments are also lytic, and lysis is thought to be triggered by holes formed in the peptidoglycan layer by murein hydrolases acting in the absence of peptidoglycan synthesis , . The similarity in the phenotypes suggests that inhibition of σE leads to weakening of and/or defects in the peptidoglycan layer. These defects potentially have two major outcomes. Either a hole forms in the peptidoglycan that is large enough to accommodate extrusion of the inner membrane and cytoplasm, evidenced by YFP signal in all of the blebs, or the cell lyses, evidenced by loss of YFP signal in the ghosted cells. The observation that blebs form in only 20% of the cells is likely to be due to the inherent stochastic nature of the process. Many cells may lyse before a bleb can form. Further support for the model that inhibition of σE leads to defects in the peptidoglycan layer comes from preliminary results with a strain lacking four enzymes involved in peptidoglycan synthesis, PBPs 4, 5, 6, and 7, which results in irregular peptidoglycan synthesis . This strain develops blebs and lyses more quickly following inhibition of σE than the isogenic WT strain (data not shown). If the peptidoglycan is affected, we should be able to reduce cell death by growing the cells in high osmolarity medium. Unfortunately, this experiment was not technically feasible. High osmolarity induced the σE response such that σE was no longer inhibited by RseA, even when RseA was overexpressed (data not shown).
Even though these phenotypes are consistent with the model that loss of σE activity leads to alterations in the peptidoglycan layer resulting in bleb formation and lysis, we cannot eliminate the possibility that bleb formation is caused by a different event, such as aberrant cell division. Future experiments examining the integrity of the peptidoglycan layer in more detail and the localization of cell division proteins following inhibition of σE should help distinguish between these possibilities.
Relationship between the phenotypes and the known functions of σE regulon members
Because the only reported function of σE is to direct transcription, altered regulation of one or more of the genes in the σE regulon should ultimately lead to bleb formation and loss of viability when σE activity is blocked. However, given the known functions of σE regulon members, we cannot point to one specific gene or set of genes in the regulon that is obviously associated with these phenotypes. Depletion phenotypes have been characterized in detail for five essential σE-regulated genes (yaeT, yfiO, imp, lptA and lptB), which encode components of complexes that assemble LPS or OMPs in the outer membrane , –. Depletion of these genes causes striking defects in the cell envelope, but not those observed following inhibition of σE , –.
The blebbing phenotype suggests that peptidoglycan or cell division is affected by loss of σE. Although σE does not transcribe any known murein hydrolases or biosynthetic enzymes directly involved in peptidoglycan synthesis, it does transcribe bacA, which encodes an undecaprenyl pyrophosphate phosphatase , . BacA hydrolyzes a phosphate moiety from undecaprenyl pyrophosphate to generate undecaprenyl phosphate, which is a lipid carrier required for synthesis of peptidoglycan and other cell wall polymers . However, E. coli possesses at least three additional undecaprenyl pyrophosphate phosphatase enzymes and cells lacking bacA alone have no obvious morphological phenotypes . A σE-dependent promoter is also found upstream of the cell division protein FtsZ , but this promoter is one of multiple promoters that contribute to the complex regulation of ftsZ expression , making it unlikely that loss of σE activity would have a large effect under non-stressed growth conditions.
Since no direct connection can yet be made convincingly between σE regulon members and the observed phenotypes, we think it likely that the phenotypes are caused by indirect effects resulting from loss of or altered regulation of σE-dependent genes. For example, σE transcribes genes required for synthesis of the lipid A component of LPS, lpxA, lpxB, and lpxD . LpxA catalyzes the first step in lipid A synthesis and uses the same substrate, UDP-N-acetyl-D-glucosamine, as is used in the first dedicated step of peptidoglycan biosynthesis . Changes in lpxA and lpxD expression in the absence of σE could alter the flux of substrates through these pathways, affecting peptidoglycan synthesis and the integrity of the peptidoglycan layer.
Suppressors of ΔrpoE lethality
In this work two genes were identified whose overexpression suppresses both the loss of viability following inhibition of σE and the requirement of the rpoE gene for viability. Of the three known suppressors of ΔrpoE lethality (ΔydcQ  and overexpression of ptsN and yhbW), the only gene with any known roles in the cell is ptsN. ptsN is a member of the σE regulon, which suggests that it plays a role in cell envelope processes . Several functions have been associated with the nitrogen PTS system, of which ptsN is a member, including balancing nitrogen and carbon metabolism, suppression of a temperature-sensitive allele of the essential GTPase Era by a ΔptsN mutant, and inhibition of the potassium transporter TrkA , , . This latter role of ptsN held the most promise for providing insights into the mechanism of suppression because K+ is important for osmotic regulation and can regulate RNA polymerase, all activities associated with σE , , . However, we demonstrated that trkA is not required for suppression by ptsN, nor is ΔtrkA a suppressor, indicating that EIIANtr has additional functions, yet to be discovered, that are responsible for suppression of ΔrpoE lethality. As discussed below, our results indicate that this novel activity likely involves regulation of cell envelope stress responses.
Lowered envelope stress responses, a characteristic of σE suppressors
The basal level of activity of multiple envelope stress responses is reduced in the presence of all σE suppressors for which data is available, overexpression of ptsN, ΔydcQ  and an uncharacterized suppressor , suggesting that lower activity of envelope stress responses is an important part of the suppression mechanism. Preliminary results indicate that the basal levels of cell envelope stress responses are also lower following overexpression of yhbW, the other multicopy suppressor that we isolated. In light of these results, one model for suppression is that the suppressors strengthen the cell envelope such that loss of σE activity and the resultant impact on the cell envelope are no longer lethal. If this model is correct, then the suppressors should lower envelope stress sensed by the signal transduction pathways that monitor the cell envelope. We found that the effects of ptsN overexpression on reporter gene expression for the Rcs and Cpx pathways were at least partially dependent on their respective stress regulators. In addition, σE activity was not altered by ptsN overexpression in a strain lacking rseA, the only known envelope sensor for the pathway. Although it is possible that EIIANtr was unable to overcome the high level of activation of σE in a ΔrseA strain background, we think it unlikely based on observations with another regulator of σE, which does not act through RseA. σE activity can be both activated and reduced in a ΔrseA strain in response to the cytoplasmic regulator ppGpp , suggesting that the effects of ptsN overexpression should be discernible. In contrast, the reduction in activity of the Cpx reporter gene examined in the ΔydcQ strain was not dependent on cpxR . However, epistasis experiments were only reported for the one reporter, and it is not known whether the activity of other cell envelope stress responses were reduced in the ΔydcQ background independently of their respective signal transduction pathways .
Why is σE essential in E. coli?
The functions of core members of the σE regulon are conserved in a number of bacterial species . However, σE is not essential in many of these bacteria including closely related species such as Salmonella typhimurium . This observation led, in part, to a model proposing that σE is not essential because it is required to transcribe a particular gene or set of genes . Instead, σE is essential because the bacteria overreact to its absence and, through an unknown mechanism that requires the ydcQ gene product, activate an unknown cell death pathway that kills the cells . Our data are consistent with this model. However, we suggest that even though the functions of core regulon members are conserved in bacterial species in which σE is not essential , significant differences can exist in the properties of the cell envelopes of different species and even strains within a species that may determine whether the bacteria can survive in the absence of σE. For example, the majority of the work on σE, including that presented in this paper, has been performed in K12 strains of E. coli, which lack the O-antigen of LPS . This variation could significantly alter the properties of the bacterial cell envelope, making the bacteria more sensitive to disruptions in cell envelope integrity so that σE would be essential. Therefore, we propose that σE is indeed essential because it is required to transcribe genes in its regulon. In the absence of σE these genes are not properly regulated, leading to disruption of cell envelope integrity and lysis. Future experiments delving in greater depth into the phenotypes of cells lacking σE, the roles of σE regulon members of unknown function, and the mechanism of suppression should help to distinguish between these models and provide new insights into the role of this important cell envelope stress response.
Materials and Methods
Media, strains, and plasmids
Strains are derivatives of MG1655, unless otherwise noted, and are listed in Table 3. Cultures were grown in Luria Bertani (LB) broth at 30°C with shaking in a gyrotary water bath. Antibiotics were used at the following concentrations: ampicillin (amp) 100 µg/ml, kanamycin (kan) 30 µg/ml, tetracyline (tet) 10 µg/ml, spectinomycin (spec) 50 µg/ml, and chloramphenicol (cam) 20 µg/ml. Altered alleles and the stress response reporters were transferred into the appropriate recipient strains by transduction with P1vir according to standard methods , with the exception that sodium citrate was not used because it can be toxic to cells with compromised membranes. The rseAD11H mutation was made in the pRseAB plasmid and the ptsN H73A and H73A/H12A mutations in the pPtsN plasmid using the QuikChange site-directed mutagenesis kit (Stratagene) and verified by sequencing. The ptsN::kan, ptsO::kan, ptsP::kan, and trkA::kan alleles were obtained from the Keio collection  and moved by P1 transduction into SEA001. The kan allele was then removed with FLP recombinase by the method of Datsenko and Wanner  to create strains SEA4131, SEA4143, SEA4144, and SEA4230.
Table 3. Strains and Plasmidsdoi:10.1371/journal.pone.0001573.t003
β-galactosidase assays were performed largely as described . Overnight cultures were diluted to OD600 of 0.02 and grown with aeration at 30°C. rseA and rseB overexpression was induced by the addition of IPTG to 1 mM when cultures reached an OD600 of 0.1–0.2 for experiments shown in Figs. 2 and 4. In Figs. 7 and 8, β-galactosidase activity was determined from cultures grown for 12–14 hours with and without 1mM IPTG. Experiments with exponential phase cultures yielded similar results. A minimum of three independent experiments were performed for each strain and condition.
Cells of SEA007 and SEA008 to be used for electron microscopy were grown in LB to an OD600 of 0.1–0.2, at which point IPTG was added to induce rseA and rseB overexpression in SEA007. Cells were harvested 2.5 hours later by centrifugation. Samples for scanning electron microscopy (SEM) were subjected to a primary fixation step in 6.25% glutaraldehyde in Sorenson's Buffer for 1.5 hours at room temperature then washed in 0.1 M cacodylate buffer. A secondary fixation step was performed by treating the samples with 2% osmium tetroxide in 0.1 M cacodylate buffer, followed by washing as before and dehydration in ethanol. Samples were dried, sputter coated and imaged on a JEOL JSM 5400 scanning electron microscope with a PGT Prism light element detector.
Samples for transmission electron microscopy (TEM) were prepared by cryo-fixation and freeze substitution. The samples were washed in 0.1 M cacodylate buffer and incubated in 18% glycerol (v/v) in 0.1 M cacodylate for 20 minutes at room temperature for cryo-protection. The sample was pelleted by centrifugation and 10 µl of the pellet was blotted onto filter paper and frozen in liquid nitrogen. Freeze substitution was carried out in 2% osmium tetroxide, 0.1% uranyl acetate in acetone at −90°C for 3 days. Samples were slowly warmed at −60°C for 14 hours, −30°C for 14 hours, and then 0°C for 1 hour. Samples were transferred to acetone, brought to room temperature, and then infiltrated with resin. Ultrathin sections were obtained using a Reichart-Jung Ultracut E microtome with a diamond knife. Sections were mounted on hexagonal mesh or open slot copper grids and stained with 2% uranyl acetate in 50% ethanol for 5 minutes followed by Reynold's lead stain for 2 minutes and washed with water. After drying, the grids were visualized on a JEOL JEM 1200 EXII transmission electron microscope. Images were taken with a digital TEITZ camera.
For fluorescence microscopy, location of the cytoplasm was visualized in strains SEA4287 and SEA4288, which express soluble YFP throughout the cytoplasm, but not the periplasm. Overnight cultures grown in LB/Amp were subcultured into media containing 1 ng/ml anhydrous tetracycline, which induces expression of YFP. IPTG was added when cells reached an OD600 of 0.1–0.2 and samples were taken for microscopy between 1.5 and 3.5 hours later. FM4-64 (Molecular Probes T-3166) and 4,6-diaminidino-2-phenylindole, DAPI, were added directly to samples at respective final concentrations of 5 µg/ml and 1 µg/mL. Cells were mounted on freshly prepared polylysine coated coverslips and visualized using an Olympus BX-61 epi-fluorescent microscope with a 100× phase oil objective (NA of 1.3), using an Olympus 41028 filter cube for YFP, Olympus 31000v2 filter cube for DAPI, and an Olympus 41001 filter cube for FM4-64. Images were captured with a Hamamatsu Orca-ER camera using Slidebook 4.1 software (Compix, Inc.). Deconvolution was performed using the AutoQuant software package (Media Cybernetics). Generally, less than 10 iterations were necessary to deconvolve images.
Sucrose density gradient fractionation
Cultures of SEA007 were grown to an OD600 of 0.1 at which point IPTG was added to 1 mM to one culture to induce rseA and rseB overexpression. After 2.5 hours of growth, cells were harvested by centrifugation, resuspended in 10 mM HEPES pH 7.8, and lysed by French press. Discontinuous sucrose density gradients were prepared by layering 860 µl of 50%, 45%, 40%, 35%, and 30% sucrose solutions over a 200 µl 55% sucrose cushion. Lysates from equal numbers of cells for treated and control cultures were loaded onto the gradients, based on the OD600 of the cultures at the time of sampling. After centrifugation at 50,000 rpm for 15 hours, 150 µl fractions were collected from the top of the gradients. The specific gravity of each fraction was determined by measuring the refractive index. β-NADH oxidase activity of each fraction was determined as previously reported . Total protein content was analyzed on 12% SDS-polyacrylamide gels with Coomassie blue staining. Western blotting was performed with polyclonal antibodies to the outer membrane protein FepA (gift of Kathleen Postle) and inner membrane protein FtsH (gift of Christophe Herman) and detected with horseradish peroxidase-conjugated donkey anti-rabbit IgG (GE Healthcare) using the ECL Plus kit (GE Healthcare).
A genetic selection was carried out to identify genes whose overexpression allowed cells to live when σE activity was inhibited. Strain SEA007 (carrying pRseAB) was transformed by electroporation with 5 pools of plasmids from the ASKA plasmid library , and the cells were plated on LB agar with 1 mM IPTG, Amp (to select for pRseAB), and Cam (to select for the library plasmids). The ASKA plasmid library (without the GFP tag) used in these studies consists of plasmids containing E. coli ORF's cloned under the control of an IPTG-inducible promoter . Colonies that grew on IPTG were streaked again on LB/IPTG/Amp/Cam to confirm growth. Plasmids were then isolated, and library plasmids purified by transformation into DH5α with selection on LB/Cam. The selected ASKA plasmids were then transformed back into SEA007 to verify that they conferred growth on IPTG. Cells were also grown on plates containing Xgal to determine whether σE activity was restored (blue colonies) or was still inhibited (white colonies). The selected genes encoded on the ASKA plasmids were identified by sequencing.
Determination of the essentiality of rpoE
The nadB::Tn10 rpoE::kan linked marker construct used in cotransduction assays was made in a strain that contains an unidentified suppressor of ΔrpoE (sup, CAG43113). The rpoE gene was deleted first, using the method of Datsenko and Wanner  to generate strain SEA4041. The nadB::Tn10 insertion from SEA2000 was then moved into the rpoE::kan strain by P1 transduction, selecting for resistance to Kan and Tet to generate strain SEA4114. Transductants were verified by PCR to ensure that the kan allele was present and the rpoE gene was absent. The essentiality of rpoE was determined by assessing the cotransduction frequency of nadB::Tn10 and rpoE::kan in recipient strains. The markers are tightly linked, so nadB::Tn10 should be transduced independently only in strains in which rpoE is essential. Plasmids that suppressed lethality due to overexpression of rseA and rseB without restoring σE activity were transformed into SEA001. The resulting strains were transduced with a P1 lysate from SEA4114. Colonies were first plated on LB/Tet/IPTG to select for nadB::Tn10 and to overexpress the putative suppressor gene on the plasmid. Transductants were then streaked on LB/Kan/IPTG to determine if rpoE had been deleted. PCR was used to verify the presence of the nadB::Tn10 and rpoE::kan alleles and the absence of the rpoE gene. Deletion of the rpoE gene was also verified by western blotting with polyclonal antibodies specific to σE. In each cotransduction experiment, SEA001 and CAG43113 were used as negative and positive controls, respectively.
We thank Susan Gottesman, Tracy Raivio, Carol Gross, Mark Goulian, and Hirotada Mori for sharing strains; Mary Laubacher for making strains; and Kathleen Postle, Marta Manning, and Charles Bulathsinghala for assistance with sucrose gradients and the FepA antibody. All fluorescence microscopy was done at the Cytometry Facility, University Park (Huck Institutes of the Life Sciences, Penn State University) and the facility is funded, in part, under a grant with the Pennsylvania Department of Health using Tobacco Settlement Funds. The Department specifically disclaims responsibility for any analyses, interpretations, or conclusions. All electron microscopy was done at the Electron Microscopy Facility at the Huck Institutes of the Life Sciences, Penn State University.
Conceived and designed the experiments: SA JH. Performed the experiments: JH. Analyzed the data: SA JH. Wrote the paper: SA JH.
- 1. Kellenberger E (1996) Structure and Function at the Subcellular Level. In: Neidhardt FC, Curtiss R III, Ingraham JL, Lin ECC, Low KB, Magasanik B, Reznikoff WS, Riley M, Schaechter M, Umbarger HE, editors. Escherichia coli and Salmonella: Cellular and Molecular Biology. 2nd ed. Washington, DC: ASM Press. pp. 17–28.
- 2. Nikaido H (1994) Porins and specific diffusion channels in bacterial outer membranes. J Biol Chem 269: 3905–3908.
- 3. Majdalani N, Gottesman S (2005) The Rcs phosphorelay: a complex signal transduction system. Annu Rev Microbiol 59: 379–405.
- 4. Raivio TL (2005) Envelope stress responses and Gram-negative bacterial pathogenesis. Mol Microbiol 56: 1119–1128.
- 5. Rowley G, Spector M, Kormanec J, Roberts M (2006) Pushing the envelope: extracytoplasmic stress responses in bacterial pathogens. Nat Rev Microbiol 4: 383–394.
- 6. Ruiz N, Silhavy TJ (2005) Sensing external stress: watchdogs of the Escherichia coli cell envelope. Curr Opin Microbiol 8: 122–126.
- 7. De Las Penas A, Connolly L, Gross CA (1997) SigmaE is an essential sigma factor in Escherichia coli. J Bacteriol 179: 6862–6864.
- 8. Alba BM, Gross CA (2004) Regulation of the Escherichia coli sigma-dependent envelope stress response. Mol Microbiol 52: 613–619.
- 9. Rhodius VA, Suh WC, Nonaka G, West J, Gross CA (2006) Conserved and variable functions of the σE stress response in related genomes. PLoS Biol 4: e2.
- 10. Button JE, Silhavy TJ, Ruiz N (2007) A suppressor of cell death caused by the loss of σE downregulates extracytoplasmic stress responses and outer membrane vesicle production in Escherichia coli. J Bacteriol 189: 1523–1530.
- 11. Mecsas J, Welch R, Erickson JW, Gross CA (1995) Identification and characterization of an outer membrane protein, OmpX, in Escherichia coli that is homologous to a family of outer membrane proteins including Ail of Yersinia enterocolitica. J Bacteriol 177: 799–804.
- 12. Missiakas D, Betton JM, Raina S (1996) New components of protein folding in extracytoplasmic compartments of Escherichia coli SurA, FkpA and Skp/OmpH. Mol Microbiol 21: 871–884.
- 13. Rouviere PE, De Las Penas A, Mecsas J, Lu CZ, Rudd KE, et al. (1995) rpoE, the gene encoding the second heat-shock sigma factor, σE, in Escherichia coli. Embo J 14: 1032–1042.
- 14. Rouviere PE, Gross CA (1996) SurA, a periplasmic protein with peptidyl-prolyl isomerase activity, participates in the assembly of outer membrane porins. Genes Dev 10: 3170–3182.
- 15. De Las Penas A, Connolly L, Gross CA (1997) The σE-mediated response to extracytoplasmic stress in Escherichia coli is transduced by RseA and RseB, two negative regulators of σE. Mol Microbiol 24: 373–385.
- 16. Missiakas D, Mayer MP, Lemaire M, Georgopoulos C, Raina S (1997) Modulation of the Escherichia coli σE (RpoE) heat-shock transcription-factor activity by the RseA, RseB and RseC proteins. Mol Microbiol 24: 355–371.
- 17. Campbell EA, Tupy JL, Gruber TM, Wang S, Sharp MM, et al. (2003) Crystal structure of Escherichia coli σE with the cytoplasmic domain of its anti-sigma RseA. Mol Cell 11: 1067–1078.
- 18. Ades SE, Connolly LE, Alba BM, Gross CA (1999) The Escherichia coli σE-dependent extracytoplasmic stress response is controlled by the regulated proteolysis of an anti-sigma factor. Genes Dev 13: 2449–2461.
- 19. Alba BM, Leeds JA, Onufryk C, Lu CZ, Gross CA (2002) DegS and YaeL participate sequentially in the cleavage of RseA to activate the σE-dependent extracytoplasmic stress response. Genes Dev 16: 2156–2168.
- 20. Alba BM, Zhong HJ, Pelayo JC, Gross CA (2001) degS (hhoB) is an essential Escherichia coli gene whose indispensable function is to provide σE activity. Mol Microbiol 40: 1323–1333.
- 21. Flynn JM, Levchenko I, Sauer RT, Baker TA (2004) Modulating substrate choice: the SspB adaptor delivers a regulator of the extracytoplasmic-stress response to the AAA+ protease ClpXP for degradation. Genes Dev 18: 2292–2301.
- 22. Kanehara K, Ito K, Akiyama Y (2002) YaeL (EcfE) activates the σE pathway of stress response through a site-2 cleavage of anti-σE, RseA. Genes Dev 16: 2147–2155.
- 23. Grigorova IL, Chaba R, Zhong HJ, Alba BM, Rhodius V, et al. (2004) Fine-tuning of the Escherichia coli σE envelope stress response relies on multiple mechanisms to inhibit signal-independent proteolysis of the transmembrane anti-sigma factor, RseA. Genes Dev 18: 2686–2697.
- 24. Costanzo A, Ades SE (2006) Growth phase-dependent regulation of the extracytoplasmic stress factor, σE, by guanosine 3′,5′-bispyrophosphate (ppGpp). J Bacteriol 188: 4627–4634.
- 25. Dartigalongue C, Missiakas D, Raina S (2001) Characterization of the Escherichia coli σE regulon. J Biol Chem 276: 20866–20875.
- 26. Rezuchova B, Miticka H, Homerova D, Roberts M, Kormanec J (2003) New members of the Escherichia coli σE regulon identified by a two-plasmid system. FEMS Microbiol Lett 225: 1–7.
- 27. Ruiz N, Kahne D, Silhavy TJ (2006) Advances in understanding bacterial outer-membrane biogenesis. Nat Rev Microbiol 4: 57–66.
- 28. Sperandeo P, Cescutti R, Villa R, Di Benedetto C, Candia D, et al. (2007) Characterization of lptA and lptB, two essential genes implicated in lipopolysaccharide transport to the outer membrane of Escherichia coli. J Bacteriol 189: 244–253.
- 29. Johansen J, Rasmussen AA, Overgaard M, Valentin-Hansen P (2006) Conserved small non-coding RNAs that belong to the σE regulon: role in down-regulation of outer membrane proteins. J Mol Biol 364: 1–8.
- 30. Thompson KM, Rhodius VA, Gottesman S (2007) σE regulates and is regulated by a small RNA in Escherichia coli. J Bacteriol 189: 4243–4256.
- 31. Udekwu KI, Wagner EG (2007) σE controls biogenesis of the antisense RNA MicA. Nucleic Acids Res 35: 1279–1288.
- 32. Wade JT, Roa DC, Grainger DC, Hurd D, Busby SJ, et al. (2006) Extensive functional overlap between sigma factors in Escherichia coli. Nat Struct Mol Biol 13: 806–814.
- 33. McBroom AJ, Johnson AP, Vemulapalli S, Kuehn MJ (2006) Outer membrane vesicle production by Escherichia coli is independent of membrane instability. J Bacteriol 188: 5385–5392.
- 34. McBroom AJ, Kuehn MJ (2007) Release of outer membrane vesicles by Gram-negative bacteria is a novel envelope stress response. Mol Microbiol 63: 545–558.
- 35. Kitagawa M, Ara T, Arifuzzaman M, Ioka-Nakamichi T, Inamoto E, Toyonaga H, Mori H (2005) Complete set of ORF clones of Escherichia coli ASKA library (a complete set of E. coli K-12 ORF archive): unique resources for biological research. DNA Res 12: 291–299.
- 36. Mecsas J, Rouviere PE, Erickson JW, Donohue TJ, Gross CA (1993) The activity of σE, an Escherichia coli heat-inducible sigma-factor, is modulated by expression of outer membrane proteins. Genes Dev 7: 2618–2628.
- 37. Powell BS, Court DL, Inada T, Nakamura Y, Michotey V, et al. (1995) Novel proteins of the phosphotransferase system encoded within the rpoN operon of Escherichia coli. Enzyme IIANtr affects growth on organic nitrogen and the conditional lethality of an erats mutant. J Biol Chem 270: 4822–4839.
- 38. Commichau FM, Forchhammer K, Stulke J (2006) Regulatory links between carbon and nitrogen metabolism. Curr Opin Microbiol 9: 167–172.
- 39. Rabus R, Reizer J, Paulsen I, Saier MH Jr (1999) Enzyme I(Ntr) from Escherichia coli. A novel enzyme of the phosphoenolpyruvate-dependent phosphotransferase system exhibiting strict specificity for its phosphoryl acceptor, NPr. J Biol Chem 274: 26185–26191.
- 40. Lee CR, Cho SH, Yoon MJ, Peterkofsky A, Seok YJ (2007) Escherichia coli enzyme IIANtr regulates the K+ transporter TrkA. Proc Natl Acad Sci USA 104: 4124–4129.
- 41. Lee CR, Koo BM, Cho SH, Kim YJ, Yoon MJ, et al. (2005) Requirement of the dephospho-form of enzyme IIANtr for derepression of Escherichia coli K-12 ilvBN expression. Mol Microbiol 58: 334–344.
- 42. Bianchi AA, Baneyx F (1999) Hyperosmotic shock induces the σ32 and σE stress regulons of Escherichia coli. Mol Microbiol 34: 1029–1038.
- 43. Wood JM (1999) Osmosensing by bacteria: signals and membrane-based sensors. Microbiol Mol Biol Rev 63: 230–262.
- 44. Csonka LN, Epstein W (1996) Osmoregulation. In: Neidhardt FC, Curtiss R III, Ingraham JL, Lin ECC, Low KB, Magasanik B, Reznikoff WS, Riley M, Schaechter M, Umbarger HE, editors. Escherichia coli and Salmonella: Cellular and Molecular Biology. Washington DC: ASM Press. pp. 1210–1223.
- 45. Gralla JD, Vargas DR (2006) Potassium glutamate as a transcriptional inhibitor during bacterial osmoregulation. EMBO J 25: 1515–1521.
- 46. Walsh NP, Alba BM, Bose B, Gross CA, Sauer RT (2003) OMP peptide signals initiate the envelope-stress response by activating DegS protease via relief of inhibition mediated by its PDZ domain. Cell 113: 61–71.
- 47. Danese PN, Silhavy TJ (1998) CpxP, a stress-combative member of the Cpx regulon. J Bacteriol 180: 831–839.
- 48. Meisel U, Holtje JV, Vollmer W (2003) Overproduction of inactive variants of the murein synthase PBP1B causes lysis in Escherichia coli. J Bacteriol 185: 5342–5348.
- 49. Spratt BG (1975) Distinct penicillin binding proteins involved in the division, elongation, and shape of Escherichia coli K12. Proc Natl Acad Sci USA 72: 2999–3003.
- 50. Denome SA, Elf PK, Henderson TA, Nelson DE, Young KD (1999) Escherichia coli mutants lacking all possible combinations of eight penicillin binding proteins: viability, characteristics, and implications for peptidoglycan synthesis. J Bacteriol 181: 3981–3993.
- 51. Braun M, Silhavy TJ (2002) Imp/OstA is required for cell envelope biogenesis in Escherichia coli. Mol Microbiol 45: 1289–1302.
- 52. Charlson ES, Werner JN, Misra R (2006) Differential effects of yfgL mutation on Escherichia coli outer membrane proteins and lipopolysaccharide. J Bacteriol 188: 7186–7194.
- 53. Doerrler WT, Raetz CR (2005) Loss of outer membrane proteins without inhibition of lipid export in an Escherichia coli YaeT mutant. J Biol Chem 280: 27679–27687.
- 54. Malinverni JC, Werner J, Kim S, Sklar JG, Kahne D, et al. (2006) YfiO stabilizes the YaeT complex and is essential for outer membrane protein assembly in Escherichia coli. Mol Microbiol 61: 151–164.
- 55. Wu T, McCandlish AC, Gronenberg LS, Chng SS, Silhavy TJ, et al. (2006) Identification of a protein complex that assembles lipopolysaccharide in the outer membrane of Escherichia coli. Proc Natl Acad Sci USA 103: 11754–11759.
- 56. El Ghachi M, Derbise A, Bouhss A, Mengin-Lecreulx D (2005) Identification of multiple genes encoding membrane proteins with undecaprenyl pyrophosphate phosphatase (UppP) activity in Escherichia coli. J Biol Chem 280: 18689–18695.
- 57. Flardh K, Garrido T, Vicente M (1997) Contribution of individual promoters in the ddlB-ftsZ region to the transcription of the essential cell-division gene ftsZ in Escherichia coli. Mol Microbiol 24: 927–936.
- 58. Anderson MS, Raetz CR (1987) Biosynthesis of lipid A precursors in Escherichia coli. A cytoplasmic acyltransferase that converts UDP-N-acetylglucosamine to UDP-3-O-(R-3-hydroxymyristoyl)-N-acetylglucosamine. J Biol Chem 262: 5159–5169.
- 59. Humphreys S, Stevenson A, Bacon A, Weinhardt AB, Roberts M (1999) The alternative sigma factor, σE, is critically important for the virulence of Salmonella typhimurium. Infect Immun 67: 1560–1568.
- 60. Liu D, Reeves PR (1994) Escherichia coli K12 regains its O antigen. Microbiology 140: 49–57.
- 61. Miller JH (1972) Experiments in Molecular Genetics. Cold Spring Harbor, NY: Cold Spring Harbor Laboratories.
- 62. Baba T, Ara T, Hasegawa M, Takai Y, Okumura Y, et al. (2006) Construction of Escherichia coli K-12 in-frame, single-gene knockout mutants: the Keio collection. Mol Syst Biol 2: 2006 0008.
- 63. Datsenko KA, Wanner BL (2000) One-step inactivation of chromosomal genes in Escherichia coli K-12 using PCR products. Proc Natl Acad Sci USA 97: 6640–6645.
- 64. Osborn MJ, Gander JE, Parisi E, Carson J (1972) Mechinism of assembly of the outer membrane of Salmonella typhimurium. Isolation and characterization of cytoplasmic and outer membrane. J Biol Chem 247: 3962–3972.
- 65. Nonaka G, Blankschien M, Herman C, Gross CA, Rhodius VA (2006) Regulon and promoter analysis of the E. coli heat-shock factor, σ32, reveals a multifaceted cellular response to heat stress. Genes Dev 20: 1776–1789.
- 66. Majdalani N, Hernandez D, Gottesman S (2002) Regulation and mode of action of the second small RNA activator of RpoS translation, RprA. Mol Microbiol 46: 813–826.
- 67. Ades SE, Grigorova IL, Gross CA (2003) Regulation of the alternative sigma factor σE during initiation, adaptation, and shutoff of the extracytoplasmic heat shock response in Escherichia coli. J Bacteriol 185: 2512–2519.
- 68. Danese PN, Snyder WB, Cosma CL, Davis LJ, Silhavy TJ (1995) The Cpx two-component signal transduction pathway of Escherichia coli regulates transcription of the gene specifying the stress-inducible periplasmic protease, DegP. Genes Dev 9: 387–398.
- 69. Raivio TL, Laird MW, Joly JC, Silhavy TJ (2000) Tethering of CpxP to the inner membrane prevents spheroplast induction of the cpx envelope stress response. Mol Microbiol 37: 1186–1197.