Bacterial Subversion of Host Innate Immune Pathways

See allHide authors and affiliations

Science  10 May 2013:
Vol. 340, Issue 6133, pp. 697-701
DOI: 10.1126/science.1235771

Defense and Counter-Defense

Provided a pathogen can enter the body and survive coughing and spluttering, peristalsis, and mucus, the first active responses the host evokes to an invading organism will be at the level of the first cell encountered, well before classical cellular immunity and antibody responses are initiated. Randow et al. (p. 701) review the range of intracellular defenses against incoming pathogens and describe how compartmental boundaries within the cell provide multiple levels at which pathogens can be thwarted in their attempts to subjugate the cell to do their bidding. Baxt et al. (p. 697) review the range of evasion tactics that bacterial pathogens can summon to counter host repulsion and establish a niche in which to replicate and ensure onward transmission.


The pathogenesis of infection is a continuously evolving battle between the human host and the infecting microbe. The past decade has brought a burst of insights into the molecular mechanisms of innate immune responses to bacterial pathogens. In parallel, multiple specific mechanisms by which microorganisms subvert these host responses have been uncovered. This Review highlights recently characterized mechanisms by which bacterial pathogens avoid killing by innate host responses, including autophagy pathways and a proinflammatory cytokine transcriptional response, and by the manipulation of vesicular trafficking to avoid the toxicity of lysosomal enzymes.

In recent years, cellular mechanisms involved in membrane trafficking and the elimination of intracellular particles—including pathogens, membrane fragments, and protein aggregates—have been elucidated in increasing molecular detail. Simultaneously, a variety of cellular mechanisms for sensing infecting pathogens have been discovered (1), including recognition of pathogen-associated molecular patterns (PAMPs) and host-derived danger signals induced by microbes. These mechanisms variously result in engulfment and elimination of the pathogen via autophagy, proinflammatory cytokine production, lytic cell death via pyroptosis, and induction of inflammatory responses in the host organism.

These innate immune mechanisms work in parallel with one another and in a coordinated fashion to eliminate bacterial threats. Each provides specialized protection for specific subcellular compartments (1). Which mechanism is most effective or appropriate to defend against a particular pathogen depends largely on the subcellular compartment occupied by the pathogen (Fig. 1).

Fig. 1 Innate immune responses commonly encountered by pathogenic bacteria.

The specific immune responses faced by bacteria during infection are dictated in part by the subcellular compartment that they occupy. Extracellular bacteria are subject to phagocytosis by professional phagocytes and to complement-mediated lysis. Intravacuolar and intracytoplasmic bacteria are subject to engulfment by autophagy and killing via lysosomal fusion with the membrane-bound compartment. Detection of bacterial PAMPs by extracellular or intracellular receptors activates signaling cascades that lead to a proinflammatory transcriptional response. Detection of bacterial PAMPs in the cytosol also triggers activation of inflammasomes, which activate proinflammatory cytokines. Steps in these pathways known to be inhibited by bacteria are marked with a red X. Blue ellipses and circles, rod-shaped and coccal bacteria, respectively; orange hexagons, complement; green diamonds, degradative enzymes within a lysosome; yellow star, inflammasome; pink “cups” on cell surface, Toll-like receptors (extracellular PRRs); white arrow, proinflammatory cytokine transcriptional response.

In parallel with the explosion of insights into fundamental innate immune processes, specific processes by which pathogenic microorganisms subvert these innate immune pathways have been discovered. Prototypical bacterial mechanisms for altering host innate immunity are the specialized secretion systems (particularly types III and IV and, in some instances, type VI) of Gram-negative bacteria, which enable infecting bacteria to inject effector proteins directly into the cytoplasm of host cells. The effector proteins directly modify the function of host factors engaged in innate immune signaling, cytoskeletal dynamics, membrane trafficking, phosphoinositide lipid metabolism, host cell signaling, ubiquitin modification pathways, transcription, and protein modification, among others. In addition, the host innate immune response has evolved the capability to recognize these secretion systems.

In this Review, we examine examples from a range of mechanisms by which bacterial pathogens avoid killing by host-cell innate responses. In turn, the discovery of these mechanisms of pathogen survival has contributed to our understanding of the underlying protective host pathways.

Evading Autophagy

One of the earliest defense responses encountered by bacterial pathogens within minutes or hours of entry into a host cell is autophagy, a process that engulfs and delivers intracellular bacteria for lysosomal degradation. Several pathogens have evolved mechanisms to evade the autophagic response, but we have mechanistic insight into only a few. Inhibition of autophagy is mediated by secreted bacterial proteins that block various steps of the process.

Evasion of Autophagic Recognition

We know that cytoplasmic bacterial pathogens circumvent recognition by the autophagy machinery. For example, in Shigella, evasion is by direct inhibition of autophagic proteins, whereas Listeria monocytogenes recruits host cell proteins that mask the pathogen from recognition (2, 3). Both of these pathogens induce their uptake into mammalian cells: Shigella predominantly into epithelial cells of the colon and L. monocytogenes predominantly into macrophages.

Shortly after entry, Shigella escapes the uptake vacuole and, mediated by the surface protein IcsA, polymerizes host actin at the bacterial surface into a propulsive tail that facilitates bacterial spread. In addition to recruiting host actin polymerization machinery, IcsA is recognized by the host’s autophagy protein Atg5 (2). Shigella counters this attack by means of a protein secreted by the organism’s type III secretion system, IcsB, which binds IcsA to block Atg5 recognition, thereby protecting the bacterium from autophagy. The Burkholderia pseudomallei type III–secreted effector protein BopA is a homolog of IcsB and is likewise required for evasion of autophagy, although whether a similar mechanism operates is unknown (4).

Like Shigella spp., L. monocytogenes escapes the uptake vacuole and spreads by polymerization of an actin tail. Actin-tail formation by L. monocytogenes is also mediated by a surface protein, ActA, which is functionally distinct from Shigella IcsA. L. monocytogenes uses two mechanisms for disguise, both requiring host proteins. First, ActA recruits the host actin polymerization machinery, consisting of Arp2/3, actin, and Ena/VASP proteins, which polymerize and elongate actin tails at the bacterial surface. In the absence of ActA, significantly more L. monocytogenes are engulfed in autophagosomes (3). The combination of recruitment of host actin-associated proteins and the formation of a propulsive actin tail allows L. monocytogenes to evade autophagic recognition. A second L. monocytogenes protein, internalin K (InlK), has a nonredundant role in avoidance of autophagy, such that if InlK and ActA are lacking, more bacteria are targeted to autophagy than if one or the other is missing (5). Recruitment of the host’s major vault protein to the bacterial surface by InlK correlates with escape from autophagosomes, albeit by an unknown mechanism (5).

Inhibition of Autophagy

After pathogen recognition, the autophagy initiation complex assembles. This complex comprises combinations of Unc-51–like kinase, Beclin-1 (Atg6/Bcl-2–interacting protein 1), phosphatidylinositol 3-kinase, Vps15, Vps34, and Atg14. During assembly a double-membrane structure (phagophore) forms that elongates into a mature autophagic vacuole that encloses the desired target for destruction.

The membranes of the autophagosome become decorated with the lipidated phosphatidylethanolamine (PE)–conjugated form of LC3 (LC3-II). This maturation process appears to be disrupted by several pathogens, resulting in a decrease of LC3-II relative to uninfected cells (6, 7). For example, L. pneumophila creates an endoplasmic reticulum (ER)–derived vacuolar niche, the Legionella-containing vacuole (LCV), which is protected from innate immune responses of the host. RavZ, a L. pneumophila effector protein, secreted by the specialized type IV secretion system, protects intracellular organisms from autophagosome maturation by irreversibly deconjugating PE from LC3 (7). Such deconjugation involves proteolytic cleavage of the lipid moiety one residue amino-terminal to the point of conjugation, which prevents reconjugation (7). However, in the absence of RavZ, many LCVs survive, indicating that other mechanisms of autophagy evasion function during infection.

Modulation of Autophagosomal Maturation

Bacterial pathogens replicating within vacuoles in host cells can still be recognized and targeted for autophagy by the host. Consequently, several bacterial pathogens have evolved mechanisms to modulate autophagosomal maturation, apparently by inhibition of autophagosomal-lysosomal fusion. Mechanistic insight into these processes is largely lacking, but preliminary studies with several pathogens suggest that secreted bacterial effectors are involved (Fig. 2). For example, the Gram-negative obligate intracellular pathogen Anaplasma phagocytophilum lives within an autophagosome-like compartment that consists of a double lipid bilayer decorated with several autophagosome markers. The Ats-1 protein is secreted by the organism’s type IV secretion system and appears to induce autophagy through binding the initiation complex component Beclin-1 (8). Despite robust induction of autophagy, vacuoles containing bacteria lack lysosomal markers, indicating that autophagosomal-lysosomal fusion is inhibited.

Fig. 2 Manipulation of autophagy by intracellular bacteria.

Autophagy pathways and sites at which bacterial effector proteins interfere with these pathways are depicted for intravacuolar (top) and intracytosolic (bottom) pathogens. Multiple bacterial effector proteins from a wide range of bacteria have been identified as interacting with these pathways (2, 3, 510, 3943, 44). Bacterial strains are given, and the relevant effector proteins are indicated in parentheses. Blue arrow, activation of cellular process; black arrows, progression in autophagy pathways; red lines, inhibition of cellular process; green circles, foci of LC3-II; compartment containing blue diamonds, lysosome.

Do Bacterial Pathogens Modify the Overall State of Autophagy?

In addition to removing invading pathogens, autophagy pathways recognize other factors that accompany infection. Infection-specific intracytoplasmic aggregates, or aggresome-like structures, are of particular interest. Aggresomes are ubiquitinated by host ligases, which enables them to be recognized by the autophagy pathway. This leads to activation of autophagy in the cell, thereby increasing the overall level of autophagy and increasing the likelihood of pathogen destruction. Ubiquitinated intracytoplasmic aggregates have been observed to accumulate around Salmonella-containing vacuoles (SCVs), but this pathogen circumvents any harm to itself by means of a deubiquitinase, SseL, secreted by the organism’s type III secretion system (9). Similar aggregates form during infection of macrophages by L. pneumophila, and here expression of the type IV Dot/Icm specialized secretion system blocks their ubiquitination (10). The genomes of several pathogenic Escherichia coli encode proteins with considerable sequence similarity to SseL, including conservation of an active cysteine residue, leading us to speculate that this mechanism of autophagy suppression may be widespread among bacterial pathogens and raising the possibility that damping the overall level of autophagy may also be of benefit to extracellular pathogens.

Interfering with Induction of the Proinflammatory Transcriptional Response

In response to microbial infection, a proinflammatory transcriptional response is activated that triggers recruitment of phagocytic cells and other components of the immune response to the site of infection. Activation is mediated by mitogen-activated protein kinase (MAPK) cascades and translocation of nuclear factor κB (NF-κB) into the nucleus, leading to increased transcription of a panel of immune molecules, including the proinflammatory cytokines interleukin-1β (IL-1β) and IL-18. The efferent activators of these pathways are the pattern-recognition receptors (PRRs), which recognize microbial PAMPs in the extracellular, vacuolar, and cytoplasmic compartments. Effector proteins from bacterial pathogens intercept NF-κB activation and MAPK kinase (MAPKK) signaling in numerous ways (Fig. 3), in several cases via previously undescribed enzymatic or molecular activities.

Fig. 3 Manipulation of the proinflammatory transcriptional response by bacterial effector proteins.

(A) Pathways of MAPK and NF-κB activation and sites at which bacterial effector proteins interfere with these pathways are depicted. MAP kinase signaling is triggered by detection of microbial PAMPs by PRRs. Signaling leads to transcriptional activation of proinflammatory responses, including transcription of the cytokines IL-1β and IL-18. Bacterial effector proteins that have been identified as interacting with these pathways are indicated (1115, 18, 19, 4550). Arrows indicate activation of cellular process; lines denote inhibition of cellular process. MAPKKK, MAPKK kinase; Ub, ubiquitin; LF, lethal factor, P, phosphate. (B) Detection of bacterial PAMPs in the cytosol by NLRs activates the inflammasome pathway. Inflammasome activation leads to processing of caspase-1 to its active form. In turn, active caspase-1 processes the cytokines IL-1β and IL-18 to their active forms. Yersinia spp. YopM interferes with bacterium-induced activation of inflammasomes by binding to mature caspase-1 (22).

The Shigella spp. IpaH family of type III secreted effector proteins function as E3 ligases, one of which modulates NF-κB activity (11). Similarly, Cif (a type III secreted effector of pathogenic E. coli) and Cif homolog in Burkholderia pseudomallei (CHBP) block degradation of the inhibitor of NF-κB (IκBα) by deaminating a glutamine residue of the ubiquitin-like molecule (Ubl) NEDD8 (neural precursor cell expressed, developmentally down-regulated 8) and, to a lesser extent, ubiquitin itself (12).

OspF, a type-III secreted effector protein of Shigella spp., is a phosphothreonine lyase that specifically inactivates MAPKs by irreversibly removing phosphate groups (13), reducing the influx of acute inflammatory cells into Shigella-infected tissues in mice (14, 15). Similarly, the Salmonella spp. homolog of OspF, SpvC, is a phosphothreonine lyase that inactivates phospho-Erk and inhibits inflammation in the intestines of mice (16, 17).

Yersinia spp. translocate into cells the acetyltransferase YopJ (YopP in Y. enterocolitica), which inactivates multiple cellular kinases by acetylation of serine and threonine residues (18, 19). YopJ inactivation of MAPKK and inhibitor of NF-κB kinase subunit beta (IKKβ) induces apoptosis, release of activated caspase-1, and secretion of IL-1β. Inactivation of kinases downstream of the intracellular nucleotide-binding oligomerization domain (NOD)–like receptor (NLR)–2 has a similar effect and, in mice, causes intestinal barrier dysfunction (1821).

Yet, Yersinia spp. also block activation of proinflammatory cytokines. The inflammasome is a pro–caspase-1 containing macromolecular platform containing NLR family proteins in response to PAMPs. Activation results in cleavage of pro–caspase-1 to caspase-1, which processes pro–IL-1β and pro–IL-18 into their mature and active forms (Fig. 3), leading to both release of the proinflammatory cytokines and pyroptotic cell death. The Yersinia type III secreted effector protein YopM binds and sequesters mature caspase-1, preventing substrate binding (22). A four-residue sequence in an exposed loop of YopM, similar to the binding loops in caspase-1 substrates and regulators, is required for YopM-mediated inhibition of caspase-1 (22), indicating that inhibition of caspase-1 occurs by substrate mimicry. Thus, Yersinia spp. both block and activate proinflammatory signals. Precisely how these seemingly paradoxical effects of Yersinia effector proteins are orchestrated during infection in vivo remains unclear.

Multiple bacterial factors trigger inflammasome activation (23), and inhibition of caspase-1 activation or reduction in release of IL-1β has been described for many pathogens, although the mechanistic details have been elucidated in only a few instances.

Manipulating Vesicular Trafficking

As well as evading autophagy and intracellular inflammatory processes, survival of bacterial pathogens also requires modification of normal cell functions that contribute to host defense, such as host phagocytic, transport, and secretory pathways. Pathogens enhance their survival by avoiding phagocytosis or subsequent fusion of bacteria-containing vacuoles with lysosomes. For example, it has long been known that Neisseria gonorrhoeae inactivates complement using the host’s own complement-inhibitory proteins, thwarting complement-mediated opsonophagocytosis and killing (Fig. 1). Pathogenic E. coli, on the other hand, use the type III secreted effector EspF to avoid antibody-independent direct uptake by professional phagocytic cells. A second E. coli type III secreted effector, EspJ, inhibits macrophage opsonophagocytosis generally (24).

Modification of Intravacuolar Niches

Another common theme in the promotion of pathogen survival is the exploitation of host guanosine triphosphatase (GTPase) signaling to co-opt host cell vesicle trafficking (Fig. 4). These small regulatory switch molecules cycle between an inactive guanosine diphosphate (GDP)–bound form and an active GTP-bound form with the help of guanine nucleotide exchange factors and GTPase-activating proteins.

Fig. 4 Host GTPase signaling and vesicular trafficking by pathogenic bacteria.

Cellular vesicle transport and secretory pathways contribute to host defense by containing pathogens in vacuoles. In phagocytic cells, lysosome fusion with the phagosome leads to killing of intraphagosomal bacteria by lysosomal enzymes. In many cell types, bacterial manipulation of vesicular trafficking promotes bacterial survival. This occurs largely by the manipulation of GTPase signaling by bacterial effector proteins. The examples of bacterial manipulation of host secretory and transport pathways discussed in the text are shown here (2532, 35, 36), and steps at which secreted bacterial effector proteins intersect host vesicular trafficking are indicated. CRV, Coxiella-replicative vacuole; MPR, mannose-6-phosphate receptor; PI3P, phosphatidylinositol 3-phosphate.

The S. Typhimurium type III secreted effector SopB, a phosphoinositide phosphatase, is required for the generation and persistence of phosphatidylinositol 3-phosphate on SCVs and for SCV maturation (25). SopB’s phosphatase activity alters the electrostatic membrane surface charge of the SCV, affecting the recruitment of Rab GTPases to the SCV and, ultimately, inhibiting SCV fusion with the lysosome (26).

Additionally, the S. Typhimurium type III secreted effector, SifA, promotes the formation of Salmonella-induced filaments (SIFs), membranous tubules that extend from the SCV to the plasma membrane. SIFs and other types of tubules induced during Salmonella infection are thought to be required for vacuole stability. SifA also sequesters the small GTPase Rab9 and other host proteins to subvert retrograde trafficking of mannose 6-phosphate receptors, thereby decreasing lysosomal activity in infected cells to favor intracellular growth (27). Thus, S. Typhimurium can modify Rab GTPase activity to both exploit vesicular trafficking and maintain the integrity of its own vacuolar niche.

During L. pneumophila infection, bacterial induction of LCV formation is critical to its protection from innate immune responses. The L. pneumophila protein RalF, which is secreted by the Dot/Icm type IV secretion system, recruits and activates host GTPase adenosine diphosphate ribosylation factor 1 to LCVs for LCV biogenesis (28, 29). L. pneumophila also recruits and activates the cellular GTPase Rab1 to the LCV via the type IV secreted effector protein SidM (DrrA) that, along with a second L. pneumophila effector (LidA), recruits ER-derived vesicles to the LCV (30). SidM also functions as a GDP dissociation inhibitor displacement factor (31). The type IV secreted effector, LepB, deactivates Rab1 once LCV maturation and fusion with the host ER is complete. Thus, L. pneumophila orchestrates the recruitment and activation of Rab1, as well as its timely deactivation and removal, for the maintenance of its intracellular niche.

Although less well understood, Coxiella burnetii provides perhaps the ultimate example of vacuolar niche development as a mechanism of immune evasion in that it survives in a lysosome-like environment. The Coxiella-replicative vacuole is an acidic autophagolysosome-like vacuole generated by fusion and interaction with the phagocytic, endocytic, and secretory pathways (32). As for other intracellular pathogens, the mechanisms of this distinct adaptation require secretion of effector proteins by a specialized secretion system (33, 34).

Modulating Vesicular Trafficking

The type III secreted effector, EspG, of pathogenic E. coli displays GTPase-activating activity for Rab1 that disrupts ER-to-Golgi trafficking and general secretory pathways (35, 36). Interestingly, the EspG homolog, VirA, of Shigella spp., displays similar activity to disrupt ER-to-Golgi trafficking and suppress host autophagy (35). Whether EspG alters host autophagy during E. coli infection or whether VirA affects the general secretory pathway during Shigella infection is not yet clear.


The pathogenesis of infection is a constantly evolving battle between the host, which needs to restrict an infecting microorganism, and the pathogen, which needs to replicate and survive for transmission to other hosts. Although the past decade has brought numerous insights into the molecular mechanisms involved in this exchange, the understanding of bacterial resistance to the host response lags behind insights into the responses themselves.

An important reason for this is that most mechanisms of bacterial interference with host innate immune responses have been worked out in cell culture. In vitro systems oversimplify innate responses due to lack of normal cytokine context, absence of complex tissue and cell types, and use of immortalized cell lines that do not reproduce responses seen in primary cells. Moreover, it is becoming increasingly clear that each host response pathway likely regulates others: For example, autophagy negatively regulates activation of inflammasomes (37, 38), and vesicular trafficking is linked to autophagy (35). Moving forward, studies in intact animals are critically important, as they will allow analysis of the host response under conditions that provide for the interplay of multiple innate immune pathways.

References and Notes

  1. Acknowledgments: M.B.G. holds a patent with Avant Immunotherapeutics for an iron-regulated bacterial promoter.
View Abstract

Navigate This Article