Outer membrane protein A and OprF: versatile roles in Gram-negative bacterial infections


  • Subramanian Krishnan,

    1.  Division of Infectious Diseases, Department of Pediatrics, The Saban Research Institute, Children’s Hospital Los Angeles, CA, USA
    Search for more papers by this author
  • Nemani V. Prasadarao

    1.  Division of Infectious Diseases, Department of Pediatrics, The Saban Research Institute, Children’s Hospital Los Angeles, CA, USA
    2.  Department of Surgery, The Saban Research Institute, Children’s Hospital Los Angeles, CA, USA
    3.  Keck School of Medicine, University of Southern California, Los Angeles, CA, USA
    Search for more papers by this author

N. V. Prasadarao, Keck School of Medicine, University of Southern California, Los Angeles, CA 90027, USA
Fax: 323-361-2867
Tel: 323-361-5465
E-mail: pnemani@chla.usc.edu


Outer membrane protein A (OmpA) is an abundant protein of Escherichia coli and other enterobacteria and has a multitude of functions. Although the structural features and porin function of OmpA have been well studied, its role in the pathogenesis of various bacterial infections has emerged only during the last decade. The four extracellular loops of OmpA interact with a variety of host tissues for adhesion to and invasion of the cell and for evasion of host-defense mechanisms when inside the cell. This review describes how various regions present in the extracellular loops of OmpA contribute to the pathogenesis of neonatal meningitis induced by E. coli K1 and to many other functions. In addition, the function of OmpA-like proteins, such as OprF of Pseudomonas aeruginosa, is discussed.


blood–brain barrier


C4b-binding protein


complement control protein module 3


cystic fibrosis


central nervous system


complement receptor 3


dendritic cell


human brain microvascular endothelial cell






inducible nitric oxide synthase


logarithmic phase




major histocompatibility complex


necrotizing enterocolitis


nuclear localization signal


outer membrane protein A


nuclear factor-kappaB


polymorphonuclear leukocyte


reactive oxygen species


transforming growth factor-β


tumor necrosis factor-α


toll-like receptor


ulcerative colitis


Bacterial infections contribute to significant mortality throughout the world. Globally important diseases, such as tuberculosis, kill millions of people every year. Other infections such as pneumonia, which can be caused by Streptococcus and Pseudomonas, and food-borne illnesses caused by Shigella, Salmonella and Campylobacter, together with other illnesses such as tetanus, typhoid, diphtheria, syphilis, leprosy and meningitis contribute to high morbidity rates. In the USA, 71% of urinary tract infections, 65% of pneumonia episodes, 34% of surgical-site infections and 24% of bloodstream infections are caused by Gram-negative bacilli. The rates of nosocomial infections caused by Gram-negative bacteria are even higher in many parts of the world. Treatment of these infections has become problematic due to the emergence of antibiotic-resistant strains and therefore understanding the pathogenesis of various diseases caused by these bacteria would yield useful information for the development of new preventative or therapeutic strategies. Bacterial pathogens have several surface structures (including pili, fimbriae, outer membrane proteins and various secretion systems) that have the potential to interact with host tissues to mediate attachment and invasion. Gram-negative bacteria contain several classes of outer membrane proteins that form monomeric or trimeric barrels to facilitate the transport of nutrients into the cell.

Outer membrane protein A (OmpA) is a multifunctional major outer membrane protein of Escherichia coli and other Enterobacteriaceae. OmpA has been highly conserved among Gram-negative bacteria throughout evolution. In addition to its role as a structural protein, OmpA serves as a receptor for several bacteriocins, namely colicin U, colicin L and bacteriocin 28b. It also serves as a receptor for several bacteriophages, such as K3, Ox2 and M1. Furthermore, the interaction of OmpA with the F-plasmid-encoded outer membrane protein TraN is required for conjugal mating. The current understanding of the structure and assembly of OmpA is reviewed as a part of this minireview series by Reusch RN and elsewhere [1,2]. OmpA is a β-barrel protein that contains four extracellular loops, and a comparison of sequence similarities in the extracellular portions of OmpA from various Gram-negative bacteria is shown in Fig. 1. Recently, OmpA has been shown to exist as two different allelic forms: ompA1 and ompA2. These allelic forms have specific differences in their amino acids, especially in the amino acids of loop 2 and loop 3 [3]. Here, the function of OmpA in the pathogenesis of various diseases, caused by different bacteria, is reviewed.

Figure 1.

 Sequence homology of OmpA from various Gram-negative bacteria. Representative pathogens for which OmpA has been shown to play a role in virulence were subjected to protein sequence alignment using the online ClustalW2 alignment tool (http://www.ebi.ac.uk/ClustalW2). The four loops of OmpA are shown within boxes and the differences in amino acid sequences are highlighted in yellow. The cleavage site of the mature protein is indicated by an arrow.

Pathogenesis of E. coli K1-induced neonatal meningitis: all-round action of OmpA

Bacterial meningitis in the neonatal population is a serious central nervous system (CNS) infection distinguished by severe inflammation of the meninges and the subarachnoid space. The mortality rates associated with this disease range from 20% to 30% in infected infants and reach 100% if untreated. Although antibiotic treatment reduces the mortality rate, 50% of survivors suffer neurological sequelae such as mental retardation, cortical blindness and hearing impairment. E. coli K1 is one of the prominent pathogens that cause meningitis in neonates. The bacterium initially colonizes nasopharyngeal or gastrointestinal sites and then penetrates into the circulation in which it multiplies to reach high-grade bacteremia. Subsequently, E. coli K1 binds to and invades the brain microvascular endothelial cells, which line the blood?brain barrier (BBB). During these steps, the bacteria must evade the host-defense mechanisms, and E. coli K1 has developed a variety of survival strategies in this respect.

Role of E. coli K1 OmpA in evasion of complement attack

The first report of a role for OmpA in virulence was demonstrated in a neonatal rat model of E. coli K1 meningitis. E. coli K1 lacking OmpA expression was inefficient in inducing bacteremia in neonatal rats and embryonated chick embryo [4]. Survival in serum is an important initial step to allow E. coli K1 to reach a threshold level of bacteremia and for subsequent crossing of the BBB. Several bacterial pathogens have developed strategies to evade complement-mediated killing by interfering with complement activation [5]. Our studies have demonstrated that the ability of E. coli K1 to survive in serum is caused by the binding of OmpA to C4b-binding protein (C4BP), which is the predominant serum inhibitor of C3b activation via the classical pathway. C4BP bound to E. coli K1 acts a cofactor in Factor I-mediated cleavage of C4b to C4d and increases the dissociation of the classical pathway C3 convertase. The N-terminal extracellular loops of OmpA interact with complement control protein module 3 (CCP3), which is found in the alpha chain of C4BP, to prevent complement-mediated killing [6]. Of note, when synthetic peptides representing sequences of CCP3 were pre-incubated with E. coli K1, a reduced serum-survival rate of the bacteria was observed. Logarithmic phase (LP) cultures of OmpA+E. coli had a higher survival rate in serum than post-LP cultures. Increased binding of LP OmpA+E. coli to C4BP is responsible for this resistance to serum killing owing to a concomitant decrease in the formation of C3b and other downstream proteins required for bacteriolysis [7].

Recent studies of loops 1, 2 and 4 of OmpA demonstrated that mutation to alanine of three amino acids at a time, in different positions of each of loops 1, 2 and 4, resulted in resistance of E. coli K1 to serum bactericidal activity as these mutants were unable to bind C4BP efficiently (Fig. 2). Of note, mutation of three residues in loop 4 increased the virulence of E. coli in the newborn mouse model of meningitis. Loop 4 mutant E. coli K1 caused high-level bacteremia in a shorter time-period than did the wild-type E. coli K1, indicating that this mutant survived efficiently during the initial stages of infection. In addition, the loop 4 mutant also induced the production of higher levels of tumor necrosis factor-α (TNF-α), interleukin (IL)-1β, IL-6 and IL-12, and recruited more microglia, B cells, macrophages and granulocytes compared with E. coli K1 to the sites of infection. Therefore, it is possible that the binding of a serum protein that initiates serum killing depends on the loop 4 sequence during the initial stages of infection. Once the bacteria survive in serum, evasion of immunocyte-mediated killing is the next step for survival.

Figure 2.

 Interaction of C4BP with Escherichia coli K1 which expresses mutant OmpA. Two-dimensional structure of OmpA depicting regions of three or four amino acids mutated to alanine (A). E. coli K1 expressing either wild-type or mutated OmpA was incubated with C4BP for 30 min, washed and the bound proteins were evaluated by flow cytometry (B). (Previously published in J. Biol. Chem. DOI: 10.1074/jbc. M110.178236.)

Role of OmpA in the interaction of E. coli with immune cells

Neutrophils or polymorphonuclear leukocytes (PMNs) provide the first line of defense against invading microorganisms. Binding of microorganisms to the PMN surface-receptors generates signals that stimulate the PMNs to kill the microbial intruder, which requires delivery of the cellular bactericidal arsenal to the prey [8]. Killing may be achieved through production of reactive oxygen species (ROS) generated by NADPH oxidase [9]. However, pathogenic microbes develop strategies to avoid killing by PMNs, by suppressing ROS generation [10]. The OmpA of E. coli K1 is shown to interact with gp96 (a heat shock protein expressed on the surface of neutrophils) for entry and survival in the cells. The interaction of OmpA with gp96 increases the expression of toll-like receptor (TLR)2 and suppresses the expression of TLR4 and complement receptor 3 in neutrophils. In addition, OmpA+E. coli down-regulates the expression of gp91phox, Rac1 and Rac2 (the components of NADPH oxidase), thereby avoiding killing by PMNs. Mutation of three amino acids in the extracellular loops of OmpA and their interaction with PMNs revealed that two regions in loops 1 and 2 are important for interaction. The interaction of loop 1 and loop 2 mutants with PMNs significantly increased the expression of TLR4 and CR3, thereby increasing the phagocytic ability of PMNs. Of note, depletion of neutrophils or gp96 in newborn mice renders them resistant to E. coli K1 infection, indicating that PMNs are initial safe havens for E. coli to persist and multiply.

The expression of OmpA also influences the entry into and survival of E. coli K1 in macrophages. Opsonization with antibody or complement was not required for bacterial entry into macrophages, indicating that the interaction of OmpA with a specific receptor is critical for entry of bacteria into macrophages. Invasion of macrophages was also dependent on microfilament and microtubule condensation [11]. Although after infection with certain bacteria macrophages undergo apoptosis to limit the spread of the pathogen, many bacterial pathogens have developed anti-apoptotic mechanisms to extend the life of these cells for their own benefit. Similarly, E. coli K1 prevents macrophage apoptosis by inducing expression of the anti-apoptotic protein BclXL. Surprisingly, OmpA+E. coli K1-infected macrophages were resistant to staurosporine-induced apoptosis, indicating that the bacterium takes over the control of macrophage function [12]. An intriguing aspect of the interaction of E. coli K1 with macrophages is that OmpA was shown to bind the alpha chain of Fcγ receptor I (CD64a) of macrophages for entry and survival (Fig. 3A,B). The interaction of OmpA with CD64a also prevented the association of the γ-chain to CD64a and induced a tyrosine-phosphorylation pattern distinct from Ig2a-induced phosphorylation. Binding of OmpA to CD64a enhanced the expression of additional CD64 and TLR2 molecules on the surface, whereas it suppressed the expression of TLR4 and CR3. Interestingly, owing to accelerated clearance of the bacteria as a result of increased CR3 expression, CD64−/− mice were not susceptible to infection (Fig. 3C) [13]. OmpA-mediated interaction with macrophages suppressed the expression of TNF-α, IL-1β and other pro-inflammatory cytokines by preventing the phosphorylation of IκB, which in turn was responsible for the suppression of nuclear factor-kappaB (NF-κB) activation. Macrophages infected with OmpA+E. coli were unresponsive to lipopolysaccharide (LPS)-induced NF-κB activation and pro-inflammatory secretion [14].

Figure 3.

 Interaction of OmpA with CD64 is critical for survival of Escherichia coli K1 in macrophages. (A) RAW 264.7 macrophages were infected with OmpA+E. coli or OmpAE. coli for various periods of time, washed, fixed, and extracellular and intracellular bacteria were stained with antibodies to S-fimbriae. (B) RAW 264.7 macrophages were transfected with control short hairpin RNA (shRNA), FcγRI or CR3 shRNA and the cells were used for bacterial binding and invasion assays. Both the binding to and entry of E. coli were significantly reduced in FcγRI macrophages. (C) FcγRI−/− infant mice at day 3 were infected with 103 colony-forming units (CFU) of E. coli K1. After 72 h, the brains were collected, paraffin embedded and tissue sections were stained with hematoxylin and eosin. Wild-type mice showed gliosis, neuronal apoptosis and neutrophil infiltration, whereas FcγRI−/− newborn mice showed no such pathology. (Previously published in PLoS Pathogens, DOI: 10.1371/journal.ppat. 1001203.)

Dendritic cells (DCs) are vital for the initiation and the modulation of specific immune responses and are the most potent antigen-presenting cells. Immature DCs sample the external environment and capture various antigens, including whole bacteria, in the periphery and the submucosa. After ingestion of bacterial pathogens, DCs migrate to secondary lymphoid tissue where they present processed antigen to stimulate antigen-specific T cells. The activation of DCs relies on the up-regulation of costimulatory molecules and abundant surface expression of major histocompatibility complex (MHC) class II molecules, resulting in so-called mature DCs, which are potent stimulators of naive T cells. E. coli K1 has been shown to invade and survive in DCs, for which OmpA expression, again, is critical. OmpA+E. coli promotes its survival in DCs by suppressing DC maturation markers such as CD40, HLA-DR and CD86 [15]. OmpA+E. coli also induces the interaction of CD47 and thrombospondin-1, which has a negative regulatory effect on markers required for DC maturation [16]. Previous studies have demonstrated that recombinant OmpA from Klebsiella pneumoniae interacts with TLR2. However, this interaction induced the maturation of DCs and the production of IL-12, which is a distinctly different response compared with the interaction of OmpA from E. coli.

Interaction of E. coli K1 OmpA with brain endothelial cells for invasion

After reaching a certain threshold level of bacteremia, E. coli K1 interacts with the BBB to invade and enter the CNS. Our studies demonstrate that expression of OmpA is crucial for the invasion of human brain microvascular endothelial cells (HBMECs) by E. coli K1, which is an in vitro culture model for the BBB. Studies using OmpAE. coli K1 showed that these bacteria invade HBMECs at a rate 25- to 50-fold lower than OmpA+E. coli K1. Of note, antibodies to OmpA, which recognizes the N-terminal regions of OmpA, significantly prevent invasion of E. coli K1. In agreement with the role of N-terminal regions of OmpA, two synthetic peptides that represent the sequence of loops 1 and 2 of OmpA significantly inhibited bacterial invasion of HBMECs [17]. Interestingly, one of these peptides had sequence homology to the loop 2 of OmpA2, encoded by ompA allele 2, found commonly in pathogenic bacteria. In addition, E. coli K1 OmpA was found to interact specifically with GlcNAc1 and 4GlcNAc (chitobiose) moieties present on the surface of HBMEC. The chito-oligomers also prevented the onset of meningitis in a rat model of hematogenous meningitis, further substantiating the role of these sugars in E. coli K1 pathogenesis [18]. Computer modeling studies demonstrated that the extracellular loops are highly mobile, which upon interaction with chitobiose moieties initially stabilizes the three-dimensional structure of OmpA for subsequent interaction with the peptide backbone of the receptor [19] (Fig. 4). Modeling studies also revealed that the chitobiose binds in two different grooves formed by loops 1 and 2 and by loop 4 [20].

Figure 4.

 Simulation snapshots of the interaction of OmpA with GlcNAc1 and 4GlcNAc moieties. Wild-type OmpA binds to two GlcNAc1, 4GlcNAc moieties (chitobiose): one at the tips of loops 1 and 2, and the second near the membrane region at the base of loops 2 and 4. The simulation studies showed that the OmpA structure stabilizes by 10 ns with more favorable free energy. In contrast, mutation of three amino acids in loop 2 (amino acids 61-64) to alanine renders the chitobiose unable to bind and exhibited less favorable energy. WT, wild type. (Previously published in J. Biol. Chem. DOI: 10.1074/jbc.M110.122804.)

The receptor for OmpA on HBMECs was identified as Ecgp96, a homolog of gp96 belonging to the Hsp90 heat shock protein family [21,22]. Interaction of OmpA with Ecgp96 induces a cascade of cellular signals in HBMECs that aid in the successful invasion of the bacterium [23–33]. It was demonstrated that the purified N-terminal portion of OmpA directly bound HBMECs, while a mutant peptide lacking loop regions could not [34]. A recent study employed E. coli K1 OmpA mutants that completely lacked the loop regions and showed that loops 1 and 2 were critical for Ecgp96 binding [35]. However, complete deletion of the loop regions might severely impair OmpA folding and localization in the outer membrane. A more logical model, where selected amino acids in the loop regions were mutated to alanine, was recently used in two different studies. One study showed that mutating specific residues in loops 1, 2 and 4 to alanine significantly prevented bacterial binding to and invasion of HBMECs [20]. Although the lack of OmpA expression in E. coli K1 modulates the expression of type 1 fimbriae, some of the loop mutations did not affect the expression of other surface structures, including type 1 fimbriae, suggesting that the interaction of OmpA with Ecgp96 is truly required for invasion of HBMECs. In addition, we have recently demonstrated that Ecgp96 interacts with TLR2 more efficiently upon infection with OmpA+E. coli, rather than with OmpAE. coli, and that the C-terminal portions of both of these molecules are important for inducing signaling events in HBMECs for invasion (S. Krishnan, S. Chen, M. Arditi and N.V. Prasadarao, unpublished data).

OmpA was also shown to be important for the invasion of astrocytes by E. coli K1. OmpA expression induced actin rearrangements in astrocytes. Intracerebral injection of E. coli K1 preincubated with recombinant OmpA in C57BL/6 mice had a protective effect while mice infected with bacteria alone died within 36 h. Preincubation with recombinant OmpA also prevented astrocyte activation and neutrophil infiltration in the brain [36].

Pathogenesis of meningitis and necrotizing enterocolitis by Cronobacter sakazakii: twisting the signal by OmpA

Cronobacter sakazakii, previously known as Enterobacter sakazakii, is an emerging neonatal pathogen that causes necrotizing enterocolitis (NEC) and meningitis. NEC is an inflammatory intestinal disorder that affects 2–5% of all premature infants and is associated with high mortality rates. The OmpA of C. sakazakii was required for bacterial invasion of the INT407 intestinal cell line. Microtubule and microfilament condensation was also shown to be important for the invasion process [37]. Another porin, OmpX, along with OmpA was later shown to promote basolateral invasion of C. sakazakii in epithelial cells [38]. OmpA of C. sakazakii was also important for invasion of HBMECs but the invasion process required microtubule condensation only and not microfilaments [39]. In agreement with the requirement for OmpA expression, only OmpA+C. sakazakii induced meningitis in a newborn mouse model of meningitis and was accompanied by neutrophil infiltration, gliosis and hemorrhage (Fig. 5A). C. sakazakii has also been shown to induce NEC in a newborn rat model under hypoxia conditions and in a mouse model. The binding of C. sakazakii to enterocytes induced apoptosis of the cells by increasing the production of IL-6. Preceding apoptosis, C. sakazakii-infected enterocytes produced significant quantities of inducible nitric oxide, which in turn is responsible for the disruption of tight junctions (Fig. 5B). Feeding newborn rats with Lactobacillus bulgaricus prior to infection with C. sakazakii completely prevented the onset of NEC. Subsequent studies have shown that OmpA expression in C. sakazakii is critical for the binding of the bacterium to enterocytes, both in vitro and in animal models.

Figure 5.

 OmpA+Cronobacter sakazakii induces NEC in newborn mice by disrupting the tight junctions. (A) Newborn mice were infected with OmpA+C. sakazakii or OmpAC. sakazakii and intestines were processed for hematoxylin and eosin staining. The arrow indicates the disruption of the villi structure. (B) Confluent monolayers of Caco-2 cells were infected with OmpA+C. sakazakii or OmpAC. sakazakii for 4 h, washed, fixed and stained with antibodies to ZO-1 followed by secondary antibody coupled to Cy3. Stained cells were imaged using a confocal microscope LSM710. OmpA CS, OmpAC. sakazakii; OmpA+ CS; OmpA+C. sakazakii. (Previously published by The American Association of Immunologists, Inc. in J. Immunology, DOI: 10.4049/jimmunol.1100108.)

C. sakazakii OmpA binds DC-SIGN (DC-specific ICAM nonintegrin) for survival in DCs but is not required for DC invasion. OmpA+C. sakazakii also induced the anti-inflammatory cytokines transforming growth factor-β (TGF-β) and IL-10 to prevent DC maturation. OmpA+C. sakazakii could not induce DC maturation and the infected DCs could not present antigens to T cells, despite pretreatment with LPS [40]. Intestinal epithelial cell monolayers pretreated with supernatant from an OmpA+C. sakazakii/DC culture markedly enhanced membrane permeability and enterocyte apoptosis, whereas OmpAC. sakazakii/DC culture supernatant had no effect. Analysis of supernatant from an OmpA+C. sakazakii/DC co-culture revealed production of significantly higher levels of TGF-β compared with the levels produced by infection with OmpAC. sakazakii. TGF-β levels were elevated in the intestinal tissue of mice infected with OmpA+C. sakazakii. Co-cultures of Caco-2 cells, and DCs in a ‘double-layer’ model (Caco-2 cells on the top of the filter and DCs at the bottom in transwell inserts) followed by infection with OmpA+C. sakazakii significantly enhanced monolayer leakage by increasing TGF-β production. These studies thus revealed a correlation between DCs and intestinal epithelial cells, which was mediated by TGF-β.

Elevated levels of inducible nitric oxide synthase (iNOS) were also observed in the double-layer infection model, and suppression of iNOS expression inhibited the C. sakazakii-induced Caco-2 cell monolayer permeability, even in the presence of DCs or OmpA+C. sakazakii/DC supernatant. Blocking TGF-β activity with a neutralizing antibody suppressed iNOS production and prevented apoptosis and monolayer leakage. Depletion of DCs in newborn mice protected against C. sakazakii-induced NEC, while adoptive transfer of DCs rendered the animals susceptible to infection. Therefore, the interaction of C. sakazakii with DCs in the intestine increases intestinal epithelium damage and the onset of NEC as a result of increased production of TGF-β. In agreement with the role of DCs, infection of newborn mice at day 3 with C. sakazakii induced the recruitment of significantly greater number of DCs to the lamina propria in the intestine without altering the recruitment of other immune cells during the pathogenesis of NEC. However, OmpA+C. sakazakii survived only in intestine and attached to enterocytes, indicating that interaction of C. sakazakii with DCs is a critical step in the pathogenesis.

Pseudomonas aeruginosa pathogenesis: OprF, the ortholog of OmpA

Pseudomonas aeruginosa is a ubiquitous Gram-negative bacterium that causes ventilator-associated pneumonia, chronic lung infections in patients with cystic fibrosis (CF), skin and soft-tissue infections in burn victims, and bacteremia and sepsis in cancer patients. Flagella, pili and LPS initially help P. aeruginosa to bind to the cell-surface glycolipid asialo-GM1 during lung damage. Subsequently, P. aeruginosa uses a type III secretion system to inject virulence factors into the cytoplasm of eukaryotic cells. So far, four effector proteins have been identified in P. aeruginosa: ExoU, ExoS, ExoT and Exo Y.

OprF of P. aeruginosa is widely considered as an ortholog of OmpA with significant amino acid similarity in their C-terminal domains [41]. OprF is a general porin that allows the nonspecific diffusion of ionic particles and small polar nutrients. Recent advances in OprF porin biology have been reviewed in this minireview series [42]. Being described as an anchoring protein and expressed on the cell surface, the role of OmpA in host–pathogen interactions has recently been getting more attention. OprF has been shown to be responsible for the adhesion to pulmonary epithelial cells, glial cells and Caco-2 cells. OprF also contributes to host pathogenicity in two metazoan models, namely the plant Cichorium intybus and the nematode Caenorhabditis elegans. The H103 strain of P. aeruginosa induced necrosis in the leaves of C. intybus 8 days after inoculation in the middle vein of its leaves while an OprF mutant of the same strain could only induce limited necrosis. P. aeruginosa was also able to kill C. elegans, which required ingestion of the bacteria by the worm and subsequent proliferation in the worm gut.

Surprisingly, OprF is not expressed in tandem with other adherence factors during colonization of CF lung. Pili and flagella expression are completely down-regulated during this stage. OprF is essential for microaerobic growth of P. aeruginosa [43]. The binding of OprF to interferon-γ (IFN-γ) has been shown to up-regulate another adhesin, LecA, through quorum sensing. Activation of IFN-γ also increased expression of pyocyanin, which is another quorum-sensing-related virulence product. Activation of LecA and pyocyanin leads to the disruption of epithelial cell function [44]. OprF expression is also imperative for the formation of anerobic biofilms [45]. OprF mutants are unable to adhere to animal cells and lack the ability to secrete ExoT and ExoS toxins via the type III secretion system. An OprF mutant was deficient in the production of signal molecules N-(3-oxododecanoyl)-l-homoserine lactone and N-butanoyl-l-homoserine lactone, both of which regulate the timing and production of pyocyanin, elastase, lectin PA-1L and exotoxin A. OprF is considered as a sensor and modulates quorum sensing to enhance bacterial virulence [46]. A recent study demonstrated the efficacy of two different vector types (Adc7 and Ad5 nonhuman primate vectors) cloned with the oprF gene and their effect on anti-P. aeruginosa systemic and lung immunity in mice. Administration of AdC7OprF to the respiratory tract directly resulted in an increase of OprF-specific IgG and IgA in lung epithelial lining fluid (ELF), OprF-specific IFN-γ in lung T cells and OprF-specific IgG in serum compared with immunization with Ad5OprF. Furthermore, challenge with a lethal dose of P. aeruginosa to these animals increased the survival rates [47].

Acinetobacter baumannii: AbOmpA

Acinetobacter baumannii is an emerging nosocomial pathogen that causes very severe to fatal pneumonia. The bacterium requires OmpA (AbOmpA) and invades epithelial cells in a zipper-like mechanism by a process that requires both microtubule and microfilament condensation. The presence of OmpA in A. baumannii induced severe lung pathology in a murine model of infection and bacterial burden in the blood, whereas the OmpA strain did not [48]. AbOmpA was further shown to localize to mitochondria and induce apoptosis in epithelial cells [49]. Transient expression of AbOmpA in epithelial cells led to the localization of the protein in the nucleus and required a novel monopartite nuclear localization signal (NLS). Mutation of NLS led to cytoplasmic localization of AbOmpA [50]. AbOmpA also induced apoptosis of DCs by targeting mitochondria and inducing the production of ROS [51].

Enterohemorrhagic E. coli: EHEC OmpA

Enterohemorrhagic E. coli (EHEC) strains are responsible for frequent food-borne and water-borne outbreaks of disease, causing nonbloody diarrhea to bloody discharge in humans. A number of diarrheagenic intestinal pathogens, including enteropathogenic E. coli, EHEC and Citrobacter rodentium induce a characteristic pattern of adherence upon binding to epithelial cells. An isogenic OmpA mutant of the noninvasive enteric pathogen EHEC showed reduced adherence on cultured HeLa cells, and anti-OmpA serum prevented EHEC binding to HeLa cells [52]. Further studies on EHEC adherence on alfalfa sprouts showed that OmpA was important for binding [53]. EHEC OmpA also induced murine DCs to secrete IL-1, IL-10 and IL-2. However, EHEC OmpA could induce DC migration across polarized epithelium [54].

Pasteurella multocida: PmOmpA

Pasteurella multocida is a small, Gram-negative bacterium, frequently found as a commensal in the upper respiratory tracts of many animals, especially cats and dogs. Humans are often infected with P. multocida as a result of an animal bite, scratch or lick. P. multocida OmpA (PmOmpA) binds extracellular matrix proteins such as heparin and fibronectin [55]. However, immunization of mice with recombinant PmOmpA did not protect the animal against bacterial challenge in spite of inducing potent serum IgG responses [56].

Klebsiella pneumonia: KpOmpA

Klebsiella pneumoniae is a capsulated Gram-negative pathogen that causes community-acquired and nosocomial pneumonia. K. pneumoniae OmpA (KpOmpA) binds and activates DCs and macrophages and triggers cytokine production and DC maturation [57]. However, in airway epithelial cells KpOmpA modulates the inflammatory responses and is responsible for immune evasion. A mutant lacking OmpA triggers a higher level of cytokine responses and is attenuated in a pneumonia animal model [58]. OmpA mutants activate IL-8 induction via NF-κB-dependent and p38- and p44/42-dependent pathways. In addition, OmpA mutants of K. pneumonia engage TLR2 and TLR4 to activate NF-κB. These strategies utilized by KpOmpA may facilitate pathogen survival in the hostile environment of the lung. The ompA gene showed a protective effect against K. pneumoniae when administered as a DNA vaccine [59].

Salmonella typhimurium: SalOmpA

The genus Salmonella contains over 2000 serotypes and is one of the most important pathogens in the family Enterobacteriaceae, which cause diarrhea, fever and abdominal cramps up to 72 h postinfection. Salmonella typhimurium is one of the common salmonella serovars that cause salmonellosis. S. typhimurium OmpA (SalOmpA), although it does not play a role in bacterial adherence or invasion, has been shown to be a potential vaccine candidate in various studies. The immune response during typhoid fever is predominantly directed against OmpA [60–62]. S. typhimurium OmpA also induces DC maturation via TLR4, p38 and extracellular signal related kinases 1 and 2 (ERK1/2) activation and further enhances T helper 1 (Th1) cell polarization [63]. SalOmpA also induces DC maturation and profound activation of MHC II molecules. It also induces a CD4- and CD8-specific response when pulsed along with E7 and PADRE peptides, and shows a promising, long-term protective anti-tumor effect in mice [64].

Other bacteria and OmpA

Bacteroides vulgatus strains derived from patients with ulcerative colitis (UC) revealed the presence of two OmpA variants. Tissue adherence of UC-derived strains was stronger than of non UC-derived strains. Adherence of DH5-α transformed with OmpA of UC-derived strains was also better than of non-UC strains [65]. Reimerella anatipestifer OmpA has also been recently shown to be important for binding and invasion of Vero cells in vitro. The bacterium is a major disease-causing agent in farm ducks [66].


OmpA has been the subject of many studies carried out to unravel its structural features and porin activity. Its role in the pathogenesis of various bacterial infections is gaining importance. Despite its conservation throughout evolution among pathogenic and nonpathogenic bacteria, OmpA interacts with specific receptors for initiating pathogenesis in some Gram-negative bacterial infections. Computer modeling studies have indicated that the extracellular loops of OmpA are highly mobile, and thus a variety of three-dimensional structures have been proposed. In reality, the membrane architecture of the spatial arrangement of OmpA may depend on the presence of other molecules in that particular pathogen, thereby enhancing its virulence. An OmpA homolog, OprF, has emerged as a virulence factor for P. aeruginosa through its binding to eukaryotic cells and/or the production of other virulence factors. In addition, OprF is a target for IFN-γ to act as a sensor for quorum sensing, leading to activation of virulence factors when in contact with the host. However, owing to the emergence of antibiotic-resistant strains, the treatment options for reducing the severity of Gram-negative bacterial infections will be limited in the future. Therefore, a greater understanding of the pathogenesis mediated by outer membrane proteins, such as OmpA and OprF, may lead to the development of novel antivirulence drug targets.


The work presented in this review was supported by NIH grants AI40567, AI73115, HD41525 and American Heart Association Grants-in Aid (to NVP) and Research Career Development Fellowships by Children’s Hospital Los Angeles.