HOMEOSTASIS OF THE GRAM-NEGATIVE CELL ENVELOPE

Abstract

The Gram-negative bacterial cell envelope is a complex structure and its homeostasis is essential for bacterial survival. Envelope stress responses (ESRs) are signal transduction pathways that monitor the fidelity of envelope assembly during normal growth and also detect and repair envelope damage caused by external assaults, including immune factors, protein toxins, and antibiotics. In this review, we focus on three best-studied ESRs and discuss the mechanisms by which ESRs detect various perturbations of envelope assembly and integrity and regulate envelope remodeling to promote bacterial survival. We will highlight the complex relationship of ESRs with envelope biogenesis pathways and discuss some of the challenges in this field on the road to mapping the global regulatory network of envelope homeostasis.

Introduction

The Gram-negative bacterial (GNB) cell envelope is an essential structure. It defines the boundary of a living cell with its environment, dictates cell shape, and is the site of many essential processes such as selective nutrient uptake, respiration, and secretion of metabolites and macromolecules into the milieu. To monitor this extracytoplasmic compartment, bacteria rely on specialized pathways called envelope stress responses (ESRs). When envelope integrity or biogenesis is compromised, this information is transduced to the cytoplasmic transcriptional machinery by ESRs to induce an adaptive gene expression response. ESRs play two functions. The first function is housekeeping. ESRs ensure the fidelity of envelope biogenesis by regulating the flux of envelope precursors, preventing their mistargeting, and mitigating adverse effects in the case of malfunction. The second function is a stress response to external assaults. The cell envelope is a constant target for biological warfare. Envelope-targeting agents include host immune system factors (e.g., cationic antimicrobial peptides (CAMPs), and lysozyme), bacterial toxins, and toxic metabolites. Many of these metabolites are clinically used as antibiotics (e.g., β-lactams and polymyxins) or are chemical inhibitors of envelope biogenesis pathways. ESRs confer a high level of intrinsic resistance to these agents by minimizing envelope damage and, in some cases, directly preventing penetration of these agents or inhibiting their activities.

Despite substantial advances in understanding the molecular architecture and mechanisms of individual ESRs, homeostatic control of the envelope as a whole is not understood. Here, we will focus on the three best-studied ESRs, called Rcs, Cpx, and σE, in the model GNB Escherichia coli (Figure 1), and discuss the complex relationship of ESRs with envelope biogenesis pathways and each other.

Figure 1. Molecular architecture and mechanism of ESRs.

(for the most recent review see [21], only the latest references are included in the legend). ESR components (orange) are shown in the context of GNB structure and biogenesis pathways (Box 1). OMPs are shown in solid dark blue, while lipoproteins are shown with pattern fill. Stress (red stars) can directly or indirectly (through inhibition of a biogenesis pathway) activate ESRs. Solid arrows represent signaling events that are relatively well-established and dashed lines represent proposed events that not yet fully understood.

(A) Rcs (regulator of capsule synthesis). The OM sensory lipoprotein RcsF and the IM negative regulator IgaA are central to regulation of RcsCDB phosphorelay, in which RcsC is the hybrid histidine kinase, RcsD the phosphotransfer and RcsB, the cytoplasmic transcriptional response regulator. RcsB works as a homodimer or as a heterodimer with RcsA. In the absence of stress, IgaA keeps the system inactive by repressing the phosphotransfer protein RcsD [16]. In response to stress or when it accumulates at the IM, RcsF interacts with the periplasmic domain of IgaA, releasing the inhibition of phosphorelay. At the OM, the Bam complex assembles RcsF in a complex with OMPs, resulting in partial surface exposure. This topology allows RcsF to monitor the integrity of the LPS layer. While a growing body of evidence suggests that RcsF is not a sensor of Bam complex activity [34], it remains unknown whether periplasmic retention of RcsF increases signaling.

(B) Cpx (conjugative pilus expression). CpxAR is a classical two-component system with positive (NlpE) and negative (CpxP) regulators. The OM lipoprotein NlpE is only required to sense a subset of cues, such as lipoprotein biogenesis defects. When export of NlpE is compromised, it interacts directly with the periplasmic domain of CpxA histidine kinase, leading to phosphorylation of the transcriptional response regulator CpxR [18,19]. CpxA can also be activated by PG defects or IM-mistargeted OMPs; however, the underlying mechanism remains unknow. Periplasmic CpxP is both an effector and negative regulator of CpxA and provides negative feedback, helping to fine-tune Cpx output. While some misfolded proteins help to titrate CpxP and promote signaling, attenuation of CpxP is not generally required for CpxA activation.

(C) σE. σE is the extracytoplasmic function (ECF) alternative σ factor. In the absence of stress, the transmembrane anti-σ factor RseA inhibits σE. The signaling pathway relies on regulated proteolysis of RseA by DegS, RseP, and ClpXP proteases. PDZ domains of DegS recognize the YxF motif of periplasmic uOMPs, initiating a proteolytic cascade. Defects in biosynthesis and export of LPS synergistically induce σE. While the precise underlying mechanism is unknown, it involves releasing RseB inhibition of DegS-dependent RseA cleavage.

Structure and assembly of the GNB cell envelope

The GNB cell envelope comprises a cytoplasmic or inner membrane (IM) and outer membrane (OM) (Fig. 1). Both membranes occlude an aqueous space called the periplasm, which contains the peptidoglycan (PG) cell wall. PG is a mesh-like structure that serves as an exoskeleton, dictating cell shape and protecting against turgor pressure.

The IM is a typical fluid-mosaic biological membrane consisting of phospholipids (PLs) and integral α-helical proteins. The OM has an unusual architecture because it is incredibly rich in β-barrel OM proteins (OMPs) and asymmetric with PLs at in the inner leaflet, and a unique to GNB glycolipid called lipopolysaccharide (LPS) at the outer leaflet (Fig. 1) [1]. LPS is negatively charged, and bridging with cations, such as Mg²⁺ and Ca²⁺, promotes the establishment of strong lateral interactions between LPS molecules. Consequently, the OM outer leaflet is not fluid, and mobility of OMPs is greatly restricted. LPS also provides rigidity to the OM, and the OM functions as a load-bearing structure, similar to PG [2,3]. LPS makes the OM bilayer impermeable to polar and non-polar compounds. Selective permeability is achieved by OMPs, which often function as non-specific pores or selective channels mediating nutrient uptake. OMPs are extremely abundant and occupy the majority of the OM surface area [4], and contribute substantially to overall OM stability. In contrast to OMPs, OM lipoproteins are mainly peripheral and tethered to the OM by a covalent lipid moiety. In E. coli, the most abundant lipoprotein Lpp covalently attaches the OM to PG [5], whereas other GNB can utilize OMPs for this function [6,7]. OM crosslinking to PG provides envelope further mechanical stability [5].

Although each envelope component has a dedicated biogenesis pathway, known as Lol (localization of lipoproteins), Bam (β-barrel assembly machinery), and Lpt (LPS transport) (Box 1 and Fig. 1), these pathways are tightly intertwined. For example, the biosynthesis pathways of glycopolymers (LPS and PG) and lipids (LPS and PLs) are interconnected at the level of metabolic precursors [8,9]. PLs serve as donors of acyl chains for lipoprotein modifications [10]. The Lol pathway transports Bam and Lpt complex lipoprotein components, and the Bam complex also assembles the LPS translocon OMP LptD [10,11]. This interconnection and interdependency have several important implications for envelope homeostasis. First, disruption of one pathway indirectly affects others, leading to pleiotropic phenotypes. Consequently, it is often challenging to identify a precise molecular signal that induces a specific ESR. Second, the same conditions directly or indirectly induce multiple ESRs, making it difficult to assess the contribution of a particular ESR to the observed cellular adaptive response. Finally, any ESR-induced modification of an envelope component or biogenesis pathway requires compensatory adaptation of other components, leading to hierarchical envelope remodeling. Although this area is not well-studied, we will discuss some examples of ESR interactions that may regulate this process.

Box 1. OM biogenesis pathways.

Lipoprotein biogenesis [10].

Lipoproteins have a specialized signal sequence (SS) with a conserved Cys residue. After Sec-dependent translocation across the IM, lipoproteins undergo complex processing that involves Cys modification with acyl chains derived from IM PLs and SS cleavage, after which Cys becomes the first amino acid in a mature protein. Mature OM lipoproteins are recognized by the Lol pathway (Fig. 1). LolCDE mediate energy-dependent extraction of lipoproteins from the IM and transfer them to the periplasmic chaperone LolA. LolA binds to lipid moieties and protects lipoproteins from aggregation. The OM acceptor lipoprotein LolB insertion incoming lipoprotein substrate into the OM inner leaflet. There is genetic evidence for LolAB-independent trafficking to the OM, but the components have not been identified.

OMP biogenesis [11].

After Sec-dependent translocation and SS cleavage, OMPs are released into the periplasm, where they bind to chaperones that prevent them aggregating and deliver them to the Bam complex (Fig. 1). This complex comprises five proteins, among which essential BamA, an OMP itself, folds and inserts OMPs substrates into the OM. The remaining components (BamB–E) are lipoproteins with regulatory functions. BamD and at least two of the three remaining components are essential for growth.

LPS biogenesis [1].

Lipid A and the conserved core of LPS is synthesized at the cytoplasmic side of the IM. After flipping to the outer leaflet of the IM by MsbA, side modifications, and a variable O-antigen ligation, the LPS molecule is transferred to the Lpt pathway, which forms a continuous transenvelope bridge (Fig. 1). LptBCFG extracts LPS molecules and pushes them through the LptA bridge to the OMP/lipoprotein subcomplex LptD/E, which inserts LPS directly into the OM outer leaflet.

PG biogenesis [50].

PG consists of strands of β-(1,4)linked N-acetylglucosamine (NAG) and N-acetylmuramic acid (NAM), which are covalently crosslinked by small peptides attached to NAM. The PG precursor molecule lipid II, a membrane-tethered NAG-NAM-pentapeptide, is synthesized on the cytoplasmic side of the IM and flipped to the periplasmic side by MurJ. The NAG-NAM-pentapeptide is incorporated into PG by multiprotein complexes containing enzymes with glycosyl transferase and transpeptidase activities.

Cellular response to lipoprotein biogenesis defects

Rcs and Cpx are strongly induced upon lipoprotein biogenesis inhibition (Fig. 1A and B). Rcs was implicated in lipoprotein biogenesis early on because the OM sensory lipoprotein RcsF is constitutively active when it mislocalizes to the IM or periplasm [12–14], where it interacts with IgaA, thereby releasing the inhibition of RcsCDB phosphorelay [15,16] (Fig. 1B). However, recent research provides strong evidence that this Rcs activation is not a physiological response but simply a side effect of RcsF mislocalization. Counterintuitively to the expected protective role of ESRs, Rcs activation is toxic and causes cell death under these conditions. Indeed, Rcs activation is one reason for the essentiality of the lipoprotein export components LolA and LolB[17] and the PL biosynthesis enzyme PgsA [13]. Rcs promotes expression of various lipoprotein genes, which likely places an additional burden on the already compromised lipoprotein export pathway. One such highly expressed Rcs-dependent lipoprotein, OsmB, is a source of Rcs toxicity upon LolB depletion [17].

Cpx is a true ESR dedicated to monitoring lipoprotein biogenesis, a function that requires NlpE [17–19]. When export of NlpE to the OM is compromised, it accumulates at the IM, stimulating CpxA for signaling (Fig. 1A) [18,19]. Unlike Rcs, Cpx induces a protective response against chemical and genetic inhibitors of lipoprotein biogenesis [17,19]. Cpx mitigates the consequences of abnormal lipoprotein localization at the IM, although it remains unknown which specific regulon genes are responsible for this function. This Rcs/Cpx case study demonstrates that generation of a protective adaptive response in addition to ESR activation is required to link an ESR to a physiological cellular response.

σE is not directly linked to sensing lipoprotein biogenesis defects. However, one of its effectors, small RNA MicL, downregulates expression of Lpp, the most abundant lipoprotein in E. coli (estimated 750,000 molecules per cell) [20]. This likely changes substrate flux through the Lol pathway, helping to prioritize export of lipoproteins essential for OM assembly, including Bam and Lpt components.

Cellular response to OMP biogenesis defects

Various defects in OMP biogenesis induce σE, Cpx, and Rcs. σE is the central and best understood ESR controlling OMP biogenesis [21]. When OMP assembly at the Bam complex is compromised, unfolded OMPs (uOMPs) accumulate in the periplasm, which exposes C-terminal sequences harboring the conserved YxF motif. YxF peptide binds to the periplasmic PDZ domain of the protease DegS, initiating RseA proteolysis (Fig. 1C). Once RseA is degraded, σE directs RNA polymerase to transcribe its effector genes in order to minimize uOMP accumulation by multiple mechanisms, such as decreasing OMP expression, proteolytically removing uOMPs in the periplasm, and promoting OMP folding at the OM. σE is the only ESR that is essential during normal growth of E. coli because of its housekeeping function in coordinating the rate of OMP synthesis in the cytoplasm with the capacity for OMP assembly at the OM [22]. Increasing the basal activity or induction kinetics of σE allows cells to survive otherwise lethal perturbations of the OMP assembly pathway [23,24], while decreasing σE activity using DegS and RseP inhibitors synergizes with Bam complex inhibition and membrane-targeting antibiotics [22,25,26].

Prolonged uOMP accumulation in the periplasm increases the likelihood of OMP folding into the IM. Tethering of OMPs to the IM induces Cpx (Fig. 1A). Cpx prevents the toxicity of mistargeted OMPs by downregulating expression of the σE-dependent periplasmic chaperone Skp, thereby preventing OMP folding into the IM [27]. This Skp downregulation highlights the complex relationship between Cpx and σE. On the one hand, they are partially redundant in combating envelope protein mistargeting and co-regulate important effectors, such as the periplasmic protease DegP [28]. On the other hand, they display a negative interaction because in addition to Skp, Cpx downregulates expression of σE itself [29]. This indicates that Cpx can inhibit or attenuate the σE response, at least under some conditions.

The relationship between OMP assembly and Rcs is less straightforward and not yet fully understood. Like the Lol pathway, the Bam complex is required for proper RcsF localization because it assembles RcsF with partner OMPs, leading to a partially surface-exposed topology [15,30–32] (Fig. 1A). Rcs is induced in some partial loss-of-function bam mutants; therefore, it was proposed that RcsF also monitors Bam complex function during OMP assembly by monitoring its own export to the cell surface [15]. According to this model, under normal growth conditions, BamA prevents RcsF from signaling by assembling it with OMPs. The recent structure of the RcsF/BamA complex demonstrated occlusion of the RcsF signaling C-terminal domain (CTD) [33]. When BamA activity is compromised, RcsF accumulates in the periplasmic orientation, promoting downstream signaling.

Interpreting the phenotypes of the bam mutants is difficult. The Bam complex is required for assembly of the LPS LptD/E translocon and therefore bam mutations broadly affect OM/LPS biogenesis. Hence, Rcs induction in bam mutants may be indirect owing to LPS stress, a well-recognized Rcs-inducing cue (see below). Indeed, stabilizing Mg²⁺-mediated LPS cross-bridges prevents Rcs induction in bam mutants as well as upon the chemical inhibition of BamA, supporting the primary function of RcsF in sensing LPS stress (see below) and not Bam complex activity [34]. Rcs is activated only by those bam mutations that don’t affect RcsF/OMP assembly. All bam mutations that prevent RcsF/OMP assembly inhibit the ability of RcsF to signal despite the presence of the OM/LPS defects, demonstrating the importance of proper RcsF localization for its sensory function [30,34].

Cellular response to LPS biogenesis defects

LPS biogenesis is tightly regulated because under- and over-production of LPS is detrimental to cells. At the core of this regulation is not an ESR but proteolysis of LpxC, the first committed enzyme in lipid A biosynthesis by the essential YejM/YciM/FtsH protease complex (reviewed in [35]). When LPS is underproduced, PLs flip to the outer leaflet, increasing OM permeability. Bacteria remove PLs from the outer leaflet by several mechanisms, including their degradation by the phospholipase PldA [36]. PldA generates fatty acids that are reimported into the cytoplasm and act as signaling molecules to ultimately inhibit LpxC degradation, resulting in increased LPS production [37]. On the other hand, when LPS is overproduced or the Lpt pathway is compromised, LPS backs up at the IM and binds to YejM (PbgA) [38], which stimulates LpxC degradation, thereby reducing the level of LPS and preventing its toxic accumulation at the IM [38–40]. This regulated proteolysis of LpxC shows interesting parallels to σE in terms of the way in which it coordinates LPS biosynthesis at the cytoplasmic side of the IM with the LPS assembly needs and capabilities at the OM.

Disruption of LPS packing by direct binding of CAMPs (e.g., polymyxin B (PMB)) or genetic alteration of the charge and structure of LPS is a strong Rcs-inducing signal [30,41–43]. Unlike other envelope biogenesis-targeting antibiotics, including cell wall, lipoprotein, and Bam complex inhibitors, CAMPs cause rapid onset of Rcs independent of de novo protein synthesis, suggesting that LPS stress is a direct signal [30,42]. Proper cell surface localization of the RcsF N-terminal domain (NTD) is required for LPS sensing because mutants defective in RcsF/OMP assembly are unresponsive to PMB treatment [30,34]. The cationic charge of the RcsF NTD is also required, suggesting there are electrostatic interactions between LPS and RcsF; however, the precise mechanism underlying RcsF induction is unknown [30]. The RcsF/OMP complex does not disassemble, and downstream signaling is likely induced by conformational changes of this complex, allowing the signaling CTD of RcsF to interact with the periplasmic domain of IgaA (Fig. 1B).

Colanic acid (M-antigen) is an Rcs effector that contributes to bacterial survival upon PMB treatment [44]. Unlike other capsular polysaccharides, colanic acid can be secreted or covalently linked to LPS, generating so-called M_LPS and helping to stabilize the LPS layer [45]. Rcs also regulates other capsule polysaccharides. Owing to their negative charge, they can act as potential molecular decoys for CAMPs. Rcs also shares a partially overlapping regulon with PhoPQ, a two-component system induced under low Mg²⁺ conditions, and they synergistically promote intrinsic CAMP resistance [46–48].

Genetic alterations of LPS induce σE synergistically with OMP defects, partly because LPS is required for efficient OMP folding [4], but also due to generation of an independent signal via RseB, alleviating inhibition of DegS-mediated cleavage of RseA [49] (Fig. 1C). The precise mechanism and function of this regulation has not been fully established.

Cellular responses to PG cell wall defects

PG biogenesis, regulation, and repair in response to various conditions were recently reviewed [50]. Here, we will focus on the connection of PG to ESRs. Cpx and Rcs have emerged as two major ESRs monitoring PG integrity because they are induced by and confer resistance to various PG-targeting inhibitors, including β-lactams and A22 [15,51,52]. The mechanism by which PG defects induce Cpx and Rcs remains elusive because no component of these pathways directly interacts with PG.

β-Lactams inhibit penicillin-binding proteins (PBPs) that crosslink PG strands, leading to a toxic cycle of PG synthesis and degradation [53]. The Cpx response to β-lactams is best-studied at the level of its regulon members, LdtD and Slt [54]. LdtD is a transpeptidase that forms non-canonical PG crosslinks, and increasing expression of LdtD and the stress alarmone ppGpp confers broad β-lactam resistance [55]. While Cpx activity is required for intrinsic β-lactam resistance, its constitutive overactivation causes LdtD-dependent toxicity [51], suggesting that LdtD activity is tightly regulated and is typically restricted to specific conditions. Further support for the stress response function of LdtD comes from the recent discovery that LdtD-mediated PG remodeling can overcome otherwise lethal defects in LPS transport [56], uncovering a surprising connection between the PG and LPS biogenesis pathways.

Slt degrades non-crosslinked PG strands, facilitating PG turnover [53]. Recent characterization of a clinical cpxA* allele suggests that PG degradation products that accumulate upon PBP inhibition and subsequent Slt hydrolysis are molecular signals for Cpx, implicating Slt as a central player in the Cpx-mediated PG monitoring and remodeling pathway [57].

Rcs promotes lysozyme resistance by regulating expression of lysozyme inhibitors [58]. Rcs is required for intrinsic β-lactam resistance and L-form survival[52,59], but the mechanism remains unknown. Unlike Cpx, Rcs activation does not affect the rate of PG synthesis or turnover [60]. Rcs likely promotes resistance by mitigating the downstream consequences of PG defects, rather than preventing them [60]. It is possible that this Rcs function is related to its role in the OM/LPS homeostasis as the OM/LPS structural integrity is critical for L-form survival [3,59]. The Rcs response to β-lactams is also significantly delayed in comparison with that to CAMPs, suggesting that Rcs induction upon PG stress is indirect [42]. An example is Cpx-dependent Rcs activation in response to deletion of a specific subset of PBPs [61]. One possible explanation for this is that envelope modification by Cpx generates a Rcs-inducing cue(s). While the molecular mechanism is unknown, this study provides the first evidence that ESRs can be organized into complex hierarchical signaling cascades and function in concert to control envelope homeostasis [61].

Concluding remarks

Envelope biogenesis pathways and ESRs are tightly linked. Perturbations of these pathways often not only generate an ESR-inducing signal, but affect the localization and function of ESR components. While it is difficult to differentiate the true physiological events from the unintended consequences of protein mislocalization, the key lies in the outcome: the function of ESRs is to protect the envelope from damage and promote bacterial survival under stress conditions. Further elucidation of the global regulatory network underlying envelope homeostasis will depend heavily on the ability to determine both direct molecular cues for ESR activation and the functions of ESR effector genes, and how they contribute to envelope remodeling and stress resistance.

Highlights.

• A network of biogenesis pathways builds the Gram-negative bacterial cell envelope.

• ESRs monitor envelope integrity and biogenesis during normal growth and stress.

• Connections among biogenesis pathways make it difficult to study individual ESRs.