Microbial Modulation of Host Immunity with the Small Molecule Phosphorylcholine

Abstract

All microorganisms dependent on persistence in a host for survival rely on either hiding from or modulating host responses to infection. The small molecule phosphorylcholine, or choline phosphate (ChoP), is used for both of these purposes by a wide array of bacterial and parasitic microbes. While the mechanisms underlying ChoP acquisition and expression are diverse, a unifying theme is the use of ChoP to reduce the immune response to infection, creating an advantage for ChoP-expressing microorganisms. In this minireview, we discuss several benefits of ChoP expression during infection as well as how the immune system fights back against ChoP-expressing pathogens.

INTRODUCTION

Phosphorylcholine [(CH₃)₃N⁺CH₂CH₂PO₄⁻] is a small (0.18-kDa) zwitterionic molecule expressed by a number of microbes across taxonomic kingdoms which infect humans and other eukaryotic hosts (1–4). Phosphorylcholine will be referred to here as ChoP, for choline phosphate. Most microbes that express ChoP acquire choline from their host. While choline is not required for growth in most prokaryotes, it is an essential nutrient in eukaryotes. Furthermore, choline is readily available to microbes during infection, as ChoP is a component of the most abundant eukaryotic membrane phospholipid, phosphatidylcholine. The turnover of phosphatidylcholine results in the release of glycerophospholipids containing choline (5). While microbes can use choline as a nutrient source (6) or as an osmoprotectant in the form of glycine betaine (7, 8), many also modify proteins or glycoconjugates with ChoP. ChoP may be either attached to the surface of the microbe or secreted on modified products.

The advantages of ChoP modification have been explored in both bacterial and parasitic systems. For extracellular bacteria that colonize the respiratory tract, where ChoP-expressing microbes are particularly common, ChoP is always attached to the bacterial surface (1, 9, 10). Surface expression of ChoP can affect epithelial cell adhesion and immune recognition. The intracellular bacterium Legionella pneumophila, in contrast, modifies a host protein with ChoP, aiding bacterial survival within host cells (11). In filarial nematodes, a secreted product that contains ChoP moderates the host inflammatory response, creating an environment amenable to parasite persistence (12). Despite the vast differences between these systems, a common theme is the use of ChoP to modulate the host response in order to support microbial survival.

The identification of ChoP on pathogenic microbes has led to the investigation of vaccine approaches to aid the recognition and clearance of ChoP-expressing pathogens. However, these efforts have been limited by the ability of ChoP to stimulate protective immunity, which is affected by the molecule to which it is attached as well as its accessibility within that molecule. The prevalence of ChoP expression, both in microbes and their host, demonstrates its importance in the constant interplay between these organisms. In this minireview, we discuss how bacteria and parasites use ChoP expression to promote survival in their respective hosts. Figure 1 summarizes the most-well-characterized pathways of ChoP-dependent host modulation during infection. The links between these different mechanisms may provide new avenues of investigation into the role of ChoP expression and how it can be targeted therapeutically.

Fig 1.

Schematic of the mechanisms used by microbes to modulate host immune responses with phosphorylcholine. In bacteria such as S. pneumoniae, the attachment of ChoP to the bacterial surface allows for adhesion to the eukaryotic platelet-activating factor receptor (PAFr). This is an example of adhesion through molecular mimicry, in this case, mimicry of the natural PAFr-binding partner PAF, which also contains ChoP. In H. influenzae, ChoP expression reduces binding of bactericidal antibody. In L. pneumophila, the secreted protein AnkX attaches phosphorylcholine to host Rab1 GTPase in order to modify its activity. The secretion of ES-62 in nematodes results in ChoP-dependent inhibition of proinflammatory Th1 immunity.

MICROBIAL ACQUISITION AND EXPRESSION OF ChoP

Choline is an essential nutrient in humans, as it is required for the synthesis of several molecules, including cell membrane lipids, methionine, platelet-activating factor receptor, and the neurotransmitter acetylcholine (13). Choline turnover occurs continuously in eukaryotic cells during the synthesis and recycling of the major membrane lipid phosphatidylcholine (5). As a consequence, many microbes have adapted strategies to take advantage of host-accessible choline for use as a nutrient source, osmoprotectant, and mechanism for the evasion of host immune responses. The phosphorylated choline molecule ChoP is associated with a wide range of microorganisms. The two most well-characterized groups include bacterial species that colonize the upper respiratory tract and parasitic nematodes that secrete ChoP-modified products (9, 10, 14, 15). Importantly, ChoP is associated with commensal as well as pathogenic microbes. A summary of the microorganisms that express ChoP-modified molecules can be found in Table 1.

Table 1.

Microbes that use ChoP for structural modification^a

Organism group and taxonomic classification	ChoP-modified structure	Reference(s)
Gram-positive bacteria
Streptococcus pneumoniae	TA, LTA	1
Streptococcus oralis	TA	134
Streptococcus mitis	TA	134
Gram-negative bacteria
Haemophilus influenzae	LPS	2
Haemophilus haemolyticus	LPS	135
Pseudomonas aeruginosa	Elongation factor Tu	10, 27
Neisseria meningitidis	Pilus	10
Neisseria gonorrhoeae	Pilus	10
Neisseria lactima	LPS	54
Neisseria subflava	LPS	54
Neisseria flavescens	LPS	54
Histophilus somni	LPS	136
Aggregatibacter actinomycetemcomitans	LPS	14
Acinetobacter baumannii	Porin D	55
Pasteurella multocida	LPS	137
Legionella pneumophila	Host Rab1 GTPase	11
Proteus mirabilis	LPS	138
Treponema pallidum	Membrane lipid	132
Morganella morganii	O antigen	139
Mollicutes
Mycoplasma fermentans	Membrane proteins	3
Nematodes
Caenorhabditis elegans	ASP-6, glycolipids	40, 140
Acanthocheilonema viteae	ES-62	4
Brugia pahangi	Secreted proteins	141
Brugia malayi	Secreted proteins	142
Dictyocaulus viviparus	GP300	59
Ascaris suum	Glycolipids	143
Ascaris lumbricoides	Glycolipids	144
Haemonchus contortus	GP300	59
Cooperia oncophora	GP300	59
Onchocerca volvulus	Glycolipids, glycoproteins	145, 146
Onchocerca gibsoni	Glycolipids	146
Trichinella spiralis	Tsp, glycolipids	147, 148
Nippostrongylus brasiliensis	C substance	149
Wuchereria bancrofti	PC-Ag	150
Protozoa
Plasmodium falciparum	Surface, secreted proteins	58
Eimeria bovis	Polypeptides	151

^aThe table does not include microorganisms that contain ChoP-modified phospholipids or those for which the ChoP-modified structure has not been elucidated. Ag, antigen.

Choline is required for growth in some, but not all, bacterial species. The human respiratory pathogen Streptococcus pneumoniae, which cannot synthesize choline, is dependent on choline for growth (16). While S. pneumoniae can grow without choline in the presence of a choline structural analog, bacteria under these conditions form extended chains and are unable to autolyze or undergo transformation (17). When environmental choline is available, it is transported into the cell, transformed to ChoP, and incorporated into the cell wall teichoic acid (TA) or lipoteichoic acid (LTA) through the genes in the lic operon (18). The lic operon was first identified in another respiratory tract bacterial pathogen, Haemophilus influenzae, and is also present in commensal Neisseria strains (19, 20). The lic operon allows for molecular thievery, whereby microbes utilize a host resource to their own advantage. Parasites, in contrast, use the Kennedy pathway for phosphatidylcholine synthesis, similar to most eukaryotes (21–23). Some bacterial species can also synthesize phosphatidylcholine, either through the Kennedy pathway or by methylation of phosphatidylethanolamine, which is an alternative method for phosphatidylcholine synthesis also used by eukaryotes (24–26). Pathogenic Neisseria strains contain enzymes that are homologous to those used in the methylation pathway to synthesize ChoP (26).

Most microbes with ChoP-modified molecules have evolved mechanisms to vary ChoP expression and its location. In Pseudomonas aeruginosa, for example, ChoP is attached to the elongation factor Tu in a temperature-dependent manner (27). For S. pneumoniae, which attaches ChoP to the TA (cell wall) and LTA (cell membrane), the number of ChoP groups per cell wall repeating unit and the amount of TA itself are both variable (28, 29). Also, S. pneumoniae contains several choline-binding proteins which can attach to the ChoP-modified cell wall (30). While the choline-binding proteins of S. pneumoniae are diverse, from the cell wall hydrolytic enzyme LytA to the abundant surface molecule PspA, all have the potential to obscure antibody recognition of the ChoP epitope when bound to the pneumococcal cell wall (31, 32). In an S. pneumoniae PspA mutant, for example, there is increased binding of anti-ChoP antibodies as well as serum proteins that recognize ChoP (33).

The location of ChoP attachment itself is variable between different Neisseria species. In the pathogenic species Neisseria meningitidis and Neisseria gonorrhoeae, ChoP is covalently linked to serine residues on the bacterial pilus, while in commensal species, including Neisseria lactima and Neisseria subflava, it is found on the lipopolysaccharide (LPS) (20, 34). Pilus modification is a rare example of the direct attachment of ChoP to a protein residue as opposed to linkage through glycans, as for the majority of cases. ChoP attachment to the pilus in N. gonorrhoeae is in competition with attachment of phosphoethanolamine, and the “winner” is determined by the expression level of the protein PilV (34). In N. meningitidis, pilin modification is variable due to the presence of an unstable homopolymeric repeat in the ChoP transferase gene pptA (35). The on-off switching in expression due to the repeating sequence in pptA is an example of phase variation. ChoP attachment to the LPS in commensal Neisseria species and to the LPS in H. influenzae is also determined by phase variation.

Phase variation in commensal Neisseria spp. and H. influenzae is due to the presence of a track of tetranucleotide repeats in the first gene of the lic locus (19, 36). Phase variation in these bacteria occurs through slipped-strand mispairing, resulting in stochastic variation in ChoP attachment (37). Rather than controlling expression in response to an external stimulus, phase variation allows for rapid on-off switching of ChoP expression within the population. This strategy for regulation, or, rather, the absence of it, allows for the rapid selection of advantageous traits within a heterogeneous population exposed to new environments. In addition, there are multiple alleles of the ChoP transferase gene licD in H. influenzae strains. The licD allele carried by a given strain determines the location of attachment to the LPS (38). Finally, some strains of H. influenzae contain a duplication of the lic locus, resulting in the attachment of two ChoPs per LPS molecule (39).

Parasitic expression of the levels of several ChoP-modified molecules is also variable, albeit through different mechanisms than those found in bacterial species. While there is a range of different molecules modified by ChoP in Caenorhabditis elegans, stage-specific ChoP modification of glycosphingolipids occurs in this as well as other nematodes (40). For example, in the filarial nematode Acanthocheilonema viteae, ChoP is attached to the secretory product ES-62. Secretion of ES-62 is stage specific, restricted to the latest larval stage and adult worms, and occurs through posttranslational control of ES-62 production (41).

The acquisition and control of ChoP expression by so many diverse microorganisms allude to the importance of this small molecule. This is especially true during host infection, in which choline is readily available. Even the human host itself can undergo stage-specific modification of proteins with ChoP, such as during pregnancy. It has been shown that placental polypeptides are modified with ChoP, which has been proposed to help shield the fetus from immune recognition (42). In the following sections, the advantages and disadvantages of microbial ChoP expression in different host environments are highlighted.

ChoP AFFECTS HOST RECOGNITION

ChoP is found on the surface of several bacterial species, both commensal and pathogenic, that colonize the upper respiratory tract. Despite the constant phase variation of ChoP, ChoP-expressing phase variants of H. influenzae are the dominant population isolated from humans and from colonized mice (43–45). In a 6-day colonization study in healthy human adults, inoculation with mixed-phase-variant populations resulted in selection for ChoP-expressing phase variant-dominant populations following colonization (J. J. Poole, E. D. Foster, K. Chaloner, J. R. Hunt, M. P. Jennings, T. Bair, K. Knudtson, R. S. Munson, P. L. Winokur, E. S. Christensen, and M. A. Apicella, submitted for publication). While ChoP is important during colonization, it also plays a role during disease. For example, ChoP-expressing H. influenzae strains are more strongly associated with the development of otitis media in chinchillas (46). Correspondingly, ChoP-expressing H. influenzae strains are correlated with increased persistence in children with otitis media (47). H. influenzae strains expressing ChoP also have delayed clearance from the lungs of infected mice (48). ChoP expression is advantageous for other bacteria as well. S. pneumoniae strains without ChoP-modified TA are unable to colonize the upper respiratory tract in mice and are less virulent in a murine sepsis model (49). ChoP expression in Histophilus somni has also been shown to increase colonization of its bovine host (50).

One of the advantages of ChoP expression for survival in the respiratory tract is the ability of ChoP to increase bacterial adhesion to the epithelial cell surface. ChoP is part of the recognition domain of the host protein platelet-activating factor (PAF), which binds to host platelet-activating factor receptor (PAFr) on epithelial cells. Bacterial species, including H. influenzae, S. pneumoniae, Aggregatibacter actinomycetemcomitans, H. somni, P. aeruginosa, and commensal Neisseria strains, exhibit this version of molecular mimicry, allowing binding to PAFr in vitro (51–54). PAFr binding also increases cell invasion for several bacterial species, including S. pneumoniae, P. aeruginosa, A. actinomycetemcomitans, and Acinetobacter baumannii (27, 51, 52, 55). Invasion of epithelial cells allows bacteria to avoid extracellular immune responses, although intracellular survival is dependent on evasion of antibacterial intracellular immunity. Invasion of epithelial cells may also aid bacterial access to the bloodstream, exposing bacteria to yet another host environment. In support of a role for PAFr during bacterial infection, PAFr knockout mice are less susceptible to S. pneumoniae-induced pneumonia (56). S. pneumoniae also exhibits ChoP-dependent binding to the epithelial receptor asialo-GM1 (57). Although the mechanism for this binding remains unclear, this suggests that S. pneumoniae adheres to epithelial cells by binding multiple host receptors through ChoP. A role for ChoP in adhesion has also been proposed for some parasitic species. In the parasite Plasmodium falciparum, a Var family protein involved in parasite adhesion is modified with ChoP, although whether binding is dependent on ChoP has not been investigated (58). In contrast, a ChoP-modified glycoprotein in Dictyocaulus viviparus may inhibit, rather than aid, attachment to the host (59). While several bacterial species have increased adhesion to host cells through ChoP expression, this is not the only example of how ChoP contributes to microbial survival. Also, the in vivo impact of PAFr binding remains unclear for some species. For example, the role of PAFr is not always ChoP dependent, as PAFr has also been shown to be important in infection models with bacteria that do not express ChoP (60–62). In addition, there are examples of ChoP-expressing bacterial infections, such as colonization with H. influenzae, that are not affected by the absence of PAFr (63).

ChoP expression by microbes can also affect survival in the presence of host immune factors, including antibody, complement, and antimicrobial peptides. It was recently shown that ChoP expression in H. influenzae reduces antibody binding to the bacterial surface, resulting in increased survival in the presence of complement (64). ChoP attachment to the LPS in H. influenzae alters the physical properties of the outer membrane, resulting in decreased membrane accessibility, reduced membrane permeability, and an altered membrane melting temperature (64). ChoP expression in H. influenzae also increases resistance to antimicrobial peptides such as the human cathelicidin LL-37 (65). ChoP-expressing Pasteurella multocida strains are also more resistant to cathelicidins produced by their natural host, chickens (66). Increased resistance to antimicrobial peptides has been observed for other phosphorylated amine molecules, including phosphoethanolamine (PEtn) attached to lipopolysaccharide (LPS). When PEtn is expressed on lipid A or the LPS outer core, bacterial species, including Salmonella enterica and Citrobacter rodentium, have increased resistance to the antimicrobial peptide polymyxin B (67, 68). These effects may extend beyond LPS modification, as similar effects on polymyxin B sensitivity were also observed for PEtn attachment to flagella in Campylobacter jejuni (69). It is possible that phospho-amine-associated modifications serve to increase membrane stability in a number of bacterial strains, and the impact of these modifications on host recognition has only recently begun to be explored. ChoP-expressing H. influenzae strains are also associated with biofilms in vitro (70), and in a chinchilla model of otitis media, biofilm persistence was correlated with ChoP expression (71). While a cause-and-effect relationship between biofilms and ChoP expression in H. influenzae has not been established, these data suggest that ChoP-expressing phase variants have an advantage during biofilm establishment in the development of otitis media.

ChoP MODULATION OF THE HOST RESPONSE

In addition to altering host recognition, ChoP is used by microbes to modify host responses during infection. Recently, a mechanism for ChoP-dependent host modification was identified in the intracellular bacterium Legionella pneumophila. In this pathogen, a secreted protein reversibly phosphorylcholinates the host GTPase Rab1 (72). The bacterial secreted protein, AnkX, is a ChoP transferase, and ChoP modification of Rab1 results in host cell defects in endosome formation (11). It has been proposed that Legionella uses AnkX and the dephosphorylcholinase Lem3 as an alternative to the GTP/GDP exchange system used in human cells (73). It is possible that other intracellular bacteria use ChoP to modify the activity of host proteins, although whether this phenomenon is limited to the decoration of host GTPases is unclear. Alteration of endosome processing creates an advantage for L. pneumophila during intracellular survival, much as the increased epithelial cell adherence and reduced immune recognition aid extracellular bacteria during host colonization.

Microbes also use ChoP to modify the host immune response. Interestingly, all examples to date of ChoP-mediated immune modulation occur in parasites. However, there may be common links between the role of ChoP in parasites and its role in bacteria that have not yet been explored. The most thoroughly studied example of a ChoP-modified immunomodulatory molecule is ES-62, a secretory product of parasitic nematodes. Several of the immunomodulatory properties of ES-62 are dependent on the presence of ChoP. In A. viteae, ES-62 induces a Th2 immune response while reducing Th1 immunity. As part of the induction of a Th2-skewed immune response, macrophage and dendritic cell (DC) exposure to ES-62 reduces the production of interleukin 12 (IL-12) and tumor necrosis factor alpha (TNF-α) in response to classic inflammatory signals such as LPS (74). Also, exposure to ES-62 results in the maturation of dendritic cells that increase IL-4 and reduce gamma interferon (IFN-γ) cytokine production in T cells (75). ES-62 exposure additionally desensitizes B cell activation and proliferation through the traditional PI3K and Ras mitogen-activated protein (MAP) kinase signaling pathways (76, 77). Finally, ES-62 can induce a Th2-skewed antibody response, resulting in production of increased IgG1 and decreased IgG2a (78). The filarial nematode Brugia malayi also produces ChoP-modified antigens which suppress B and T cell signaling in a ChoP-dependent manner (79). The suppression of Th1 inflammation allows for parasite persistence within the host.

The phenotypes described above have also been replicated using the ChoP-conjugated OVA peptide. ChoP-OVA exposure alone results in the induction of Th2 immunity, demonstrating that ChoP is essential for the immune modulation observed during ES-62 exposure (80). Indeed, even the synthetic oxidized 1-palmitoyl-2-arachidonoyl-sn-glycero-3-phosphorylcholine, or oxPAPC, molecule, which contains ChoP, has been shown to bind TLR4 and induce IL-8 secretion (81). Interestingly, PAFr has been found on the surface of macrophages as well as B and T cells, raising the possibility that ChoP-mediated immune modulation occurs through PAFr binding (82, 83). However, while PAFr is functional on macrophages, it is not on B and T cells, suggesting the presence of additional immune activators responsible for ChoP-dependent effects on the immune system (84).

While the exposure or secretion of ChoP-modified molecules creates an advantage for microbes, the immune modulation that results also has the potential to affect immune homeostasis in the host. For example, the anti-inflammatory effect of ES-62 exposure also reduces arthritis in a ChoP-dependent manner (85). Additionally, the importance of ChoP-mediated effects on the immune response to concurrent or subsequent microbial infections remains to be explored.

HOST RESPONSES TO ChoP

Microbial expression of ChoP-modified molecules provides several potential advantages during host infection. However, the ability to vary ChoP expression in most microbes suggests that ChoP-modified molecules may be advantageous only in certain host environments. The acute-phase reactant protein C-reactive protein (CRP) recognizes ChoP and can initiate classical-pathway, complement-mediated killing of H. influenzae and increases phagocytosis of S. pneumoniae (43, 86). Even the position of ChoP on the LPS of H. influenzae affects the level of CRP-mediated bacterial killing (38). A role for CRP has also been demonstrated in vivo, as transgenic mice expressing human CRP have increased survival and reduced bacteremia following infection with S. pneumoniae, and the majority of these effects are dependent on complement (87, 88). Also, CRP binding to ChoP-modified pilin increases uptake of N. meningitidis by phagocytes (89). It was recently shown that in S. pneumoniae, CRP binding is inhibited by the surface protein PspA, which is anchored to ChoP (33). Inhibition of CRP binding also reduces complement deposition on the bacterial surface. CRP levels are initially low both in the blood and at the mucosal surface, although these increase dramatically during inflammation (90). When CRP is readily available, such as in rat or human serum during infection, ChoP-expressing H. influenzae strains are more sensitive to complement-mediated killing than bacteria not expressing ChoP (2, 91). In contrast, ChoP-expressing bacteria have the advantage in mouse and rabbit models of bloodstream infection, in which CRP levels remain low (92, 93). The pentraxin serum amyloid protein (SAP) can also bind ChoP, although no bactericidal effect has been demonstrated (94). Instead, SAP binding inhibits classical-pathway complement activity (95). The difference in the impact of CRP versus SAP binding highlights the concept that ChoP expression may be advantageous only in certain host conditions.

The second major host factor involved in ChoP recognition is the expression of anti-ChoP antibodies. Anti-ChoP antibodies are B1-lineage, natural antibodies. B1 antibodies are part of the repertoire of IgM synthesized prior to microbial exposure and have lower affinity than antibodies induced following infection (96). The major idiotype of anti-ChoP antibodies is T15, originally identified in mice (97, 98). While ChoP modification of host molecules, including the major cell membrane lipid phosphatidylcholine, is very common, it has been shown that T15 antibodies do not recognize the ChoP epitope in intact human membrane lipids (99, 100). Anti-ChoP antibodies do, however, recognize ChoP on intact or digested microbes, and it has been shown that there are detectable levels of circulating ChoP-containing antigens following infection with ChoP-expressing parasites in the bloodstream (101).

While anti-ChoP antibodies play a role in the recognition of microbes with ChoP-modified surfaces (102), they also are important during clearance of apoptotic cells and may be beneficial during atherosclerosis (103). As a case in point, anti-ChoP antibodies that bind to dental plaque bacteria cross-react with oxidized low-density lipoprotein (oxLDL), which contains ChoP (104). Reduced anti-ChoP IgM levels in patients with acute cardiovascular disease correspond with increased risk of a new cardiovascular event (105, 106). Also, higher anti-ChoP IgM levels correlate with decreased symptomology for systemic lupus erythematosus (SLE) patients (107). These effects, however, are not limited to anti-ChoP antibodies. CRP, known to play an important role in killing ChoP-expressing microbes, binds to oxLDL as well as apoptotic cells in a ChoP-dependent manner (108). However, ChoP-dependent recognition of LDL may not always be beneficial. A study investigating the effect of anti-ChoP antibodies on atherosclerosis found that opsonization of minimally modified LDL increased inflammation generated by DCs and NK cells (109). These studies suggest that infection with ChoP-expressing microbes can impact noninfectious-disease processes in the host.

Evasion of host recognition is critical for microbial survival during infection. However, microbes themselves are susceptible to infection, as exemplified by phage predation of bacteria. It has been shown that choline-modified TA in S. pneumoniae increases susceptibility to certain phages (110). Several pneumococcal phages contain proteins homologous to the cholin-binding proteins of S. pneumoniae, and choline-binding interactions have been demonstrated for some of these (111, 112). As LPS molecules can serve as phage receptors (113, 114), it is possible that ChoP attachment to the LPS affects phage susceptibility in other bacteria as well. In this way, ChoP expression may also impact intermicrobial competition in addition to the ongoing competition between microbe and host.

ChoP-BASED VACCINES

Vaccines that successfully stimulate a protective anti-ChoP response have the potential to be used for immunization against a wide range of ChoP-expressing microbes. Even though B1 anti-ChoP antibodies are low affinity, they provide some protection, as B1-deficient Xid mice (115) are more susceptible to S. pneumoniae infection (31). Beyond preexisting anti-ChoP B1 antibodies, stimulation of anti-ChoP antibody production occurs during exposure to ChoP-expressing microbes. Carriage of S. pneumoniae in young children is associated with increased levels of anti-ChoP antibodies (116). It remains unclear, however, whether these antibodies are able to protect children from additional infections with S. pneumoniae or other ChoP-expressing microbes.

While there is no vaccine currently in use that targets ChoP directly, there is one example of a licensed vaccine that induces a response against a protein involved in choline acquisition. One of the vaccines for S. pneumoniae contains 10 pneumococcal polysaccharide serotypes conjugated to protein D. Protein D is a glycerophodiester phosphodiesterase from H. influenzae that scavenges host choline (117). In children, the protein D-containing vaccine has 35% efficacy against otitis media caused by H. influenzae (118). Passive transfer of sera from children immunized with this vaccine protects chinchillas against H. influenzae otitis media with similar efficacy (119). The protein D-containing conjugate vaccine induces antibodies that inhibit the glpQ activity of protein D (120), suggesting that its efficacy is related to its ability to repress choline scavenging in vivo. In recent work, it was shown that abrogation of protein D activity results in reduced ChoP expression and decreased epithelial cell adherence and fitness in vivo (121). The success of the protein D-containing conjugate vaccine against H. influenzae infection, while limited, demonstrates the potential for targeting ChoP expression to protect against other ChoP-expressing microbes.

More direct approaches to target ChoP, such as inducing anti-ChoP antibodies, may have a greater impact on stimulating immunity to ChoP-expressing microbes. While the majority of ChoP-specific antibodies are IgM natural antibodies (122), human ChoP-specific IgG is protective against H. influenzae and S. pneumoniae in mouse models of infection (102). Patients with specific antibody defects, such as IgA-deficient patients, also have elevated levels of anti-ChoP IgG (123). Several vaccine formulations have demonstrated promise for the induction of ChoP-specific responses in animal models of infection with S. pneumoniae (124), N. meningitidis (125), and H. influenzae (126). In S. pneumoniae, the level of protection with anti-ChoP IgG is dependent on capsule type and the effectiveness of its shielding (127). ChoP vaccine formulations often involve conjugation of ChoP to a protein carrier, such as keyhole limpet hemocyanin (128, 129). ChoP is an example of a T cell-independent antigen, as with polysaccharides. Conjugation to a protein carrier allows for a repertoire shift in the anti-ChoP idiotype, resulting in the involvement of T cells and the development of higher-affinity antibodies (130). The induction of anti-ChoP antibodies may also provide protection against other ChoP-expressing bacteria, including A. actinomycetemcomitans and Treponema pallidum (131, 132). Interestingly, vaccination with S. pneumoniae induces T15 IgM that binds oxLDL, resulting in reduced atherosclerosis (133).

Targeting ChoP surface molecules is an attractive option as a strategy to immunize against multiple pathogens simultaneously. However, this would also affect ChoP-expressing commensal bacteria, and depletion of normal flora due to vaccination is a potential concern. Also, the variable expression of ChoP in many microbes suggests that even in the presence of protective anti-ChoP antibodies, phase variation or covering the ChoP epitope would allow infection to continue. Given the diversity in ChoP expression, additional targets may be required for vaccines against ChoP to have the most significant impact.

CONCLUDING REMARKS

The use of choline beyond a nutrient source has proven to be a useful strategy for microbes as a way to hide from (molecular mimicry) or alter host immune responses. The role of ChoP-modified molecules varies by microbe, although there may be undiscovered functional links. It is unclear, for example, whether ChoP-modified glycans or pilin in bacteria can modulate the immune response like ES-62 does during nematode infection. ChoP modulation of host immune responses may be dependent on the context in which it is presented. Additional studies should give insight into the development of ChoP-targeted vaccines and immune therapies using ChoP-modified molecules. Likewise, surface expression of ChoP on bacteria as demonstrated for H. influenzae may impact survival in the presence of antibody, complement, and antimicrobial peptides. ChoP itself may not be the only molecule capable of these and other effects, as other phospho-amine molecule modifications of the LPS increase resistance to antimicrobial peptides. Finally, there may be additional examples of related microbial products that can modify the activity of host proteins, as highlighted by the discovery of the ChoP transferase of L. pneumophila. Regardless of the interplay between different systems, the ability to manipulate ChoP has a marked effect on host recognition and response to microbes during infection.

There has been extensive interest in stimulating immune responses targeting ChoP-expressing microbes. However, the effects of anti-ChoP immunity also impact noninfectious-disease conditions, including arthritis, SLE, and atherosclerosis. There is clearly a dual role for anti-ChoP immune responses, which target ChoP-expressing microbes but also respond to host molecules with an exposed ChoP epitope, such as oxLDL. Even in the context of an effective anti-ChoP immune response, the seemingly universal capacity of microbes to vary either ChoP expression or accessibility limits the effectiveness of anti-ChoP immunity in eradicating infections. This minireview highlights ChoP as a key player in the constant battle between host and microbe during infection. The contribution of this dynamic molecule to microbial survival occurs through several different mechanisms, and the interplay between these and the host response to ChoP-expressing microbes is an important area for ongoing research.