The results presented in this study reveal a number of important insights. Murine macrophages and dendritic cells responded to GlyAg stimulation through the production of NO. More importantly, the NO-driven oxidation of engulfed GlyAg led directly to the release of protons and ultimately endosomal acidification. The pH-sensitive probe LysoSensor Green showed that cells capable of producing NO have a strong correlation between GlyAg oxidation and vesicular acidification independent of resident proton pump activity, whereas NO-deficient cells failed to show acidified endosomal compartments. This relationship was detectable in both BafA-treated and untreated cells. In addition, the induction of NO alone via P3C-mediated TLR2 stimulation failed to generate acidic vesicles, demonstrating that GlyAg oxidation by NO was the key to the generation of acidic vesicular microenvironments.
The physiologic significance of these fundamental observations was assessed under three circumstances: conventional protein antigen processing, TLR9 signaling, and GlyAg processing. First, OVA processing was strongly inhibited by BafA treatment, yet was largely restored in WT but not NO-deficient cells after GlyAg incubation. These experiments conclusively demonstrate that GlyAg-induced acidification directly supports conventional protein Ag processing through activation of resident proteases. Second, CpG DNA stimulation of TLR9 resulting in TNFα production was eliminated with BafA treatment, but the TNFα response was restored to ~50% of that of BafA negative control cells when the BafA-treated cells were also incubated with GlyAg. These results show that GlyAg-induced acidification is sufficient to promote TLR9 signaling. Finally, in vitro oxidation of GlyAg revealed the release of protons as a reaction product, which lowered the pH of the local environment and inhibited further GlyAg cleavage. These data suggest a novel feedback mechanism of GlyAg oxidation via the release of protons into the endosomal compartment that protects the GlyAg epitope from complete degradation. Our findings point to a novel model in which capsular carbohydrate oxidation leads to vesicular acidification and subsequently promotes adaptive immune responses through the promotion of conventional Ag processing and negative regulation of GlyAg processing as well as innate immune responses against microbial DNA through TLR9 recognition and signaling.
The results described in this study are highly compatible with previous findings, as these classical immune mechanisms are known to be highly sensitive to pH. Most recently, TLR9-mediated TNFα production in response to microbial DNA was shown to be dependent on vesicular pH-sensitive proteolysis of the luminal TLR9 ectodomain (14
). Similarly, conventional protein Ag processing is triggered by acidic pH because of the proteolytic enzymes, such as the cathepsin family, which relies on low pH for activation (30
). As a consequence, the observation that NO-mediated oxidation of endocytosed carbohydrate and the resulting low pH leads TLR9 signaling and OVA processing is in harmony with these mechanisms through the common link of acid-dependent proteolysis. These findings also provide an explanation for why GlyAg fragments <~4 kDa have never been reported within endosomal compartments (2
), because oxidation generally favors neutral or alkaline pH (33
). This is a key point because fragments <4 kDa fail to trigger a T cell response (12
), likely owing to their failure to associate with MHC II (11
), thus providing specific functionality to the self-limiting nature of GlyAg oxidation in APCs.
Many microbes, including commensal organisms of the gut microbiota, are encapsulated by carbohydrates. Upon phagocytosis, these microbes are lysed through the action of oxidant molecules that cleave the surface polysaccharides into fragments. This results in the spilling of microbial contents, including nucleic acids and antigenic proteins, into the lumen of the endosome. This endosomal milieu is rich in potential immune targets, including GlyAgs—molecules that are recognized by PRR molecules such as TLR9 and microbial proteins that can be processed by proteases for MHC II loading and T cell activation. GlyAg molecules are known ligands of pattern recognition receptors (e.g., NOD, TLR) that initiate signaling cascades, which trigger the innate microbial clearance pathways and can shape the adaptive immune response (4
). In addition, GlyAgs expressed by commensal organisms have been implicated in directing the maintenance of immune homeostasis within the gut. The B. fragilis
GlyAg PSA was shown to normalize the TH
2 imbalance found in gnotobiotic mice (34
), whereas CD4+
T cells recognizing GlyAg via MHC II presentation (3
) can downregulate inflammation in the IL-10−/−
model of inflammatory bowel disease (35
). It is clear from these studies that the importance of carbohydrates within the immune response is becoming more widely recognized as a critical portion of the overall immune response to microbes. Our findings show that the pathways surrounding GlyAg oxidation, protein processing and TLR9 signaling mechanistically intersect in the endosomal compartment because of pH, thus providing a novel view of the events immediately following oxidative death of encapsulated microbes via oxidation within phagocytes.
Although our findings are focused on NO-mediated oxidation, there is also strong corroborating data in the literature pointing to the general importance of oxidation and pH in the immune response. In agreement with our findings, the NADPH oxidase complex that produces superoxide is effective only at near-neutral pH, because a more neutral environment has been found to be more conducive to oxidation-mediated killing of pathogens (36
). Interestingly, phagocytes isolated from chronic granulomatous disease (CGD) model mice, a primary immunodeficiency in which a genetic defect in the NADPH oxidase results in diminished reactive oxygen species production, have been shown to contain more acidic phagosomes than seen in normal WT cells and are thus unable to kill ingested pathogens (37
) despite the presence of normal to high concentrations of NO (40
). Given our findings, it is not surprising that patients with CGD are particularly susceptible to infection by heavily encapsulated microbes (42
), including Staphylococcus aureus
that carry T cell activating GlyAgs in their capsule (44
) and Aspergillus
species. Our model suggests that the lack of carbohydrate oxidation and cleavage at acidic pH in CGD directly contributes to decreased microbial killing and leads to a failure to provide MHC II and other PRR molecules with the agonists required for appropriate innate and adaptive immune responses.
The importance of achieving and maintaining a proper pH gradient along the endocytic pathway is also highlighted by findings that some pathogenic bacterial and fungal species can modulate vesicular acidification by various mechanisms as a means to circumvent detection and killing by immune cells. For example, Mycobacterium tuberculosis
inhibits acidification and persists in phagosomes by preventing phagosomal maturation and the acquisition of the V-type adenosine triphosphatase (45
). The pathogenic fungus Histoplasm capsulatum
also inhibits acidification by preventing acquisition of the proton pump (48
). If Helicobacter pylori
crosses the gastric epithelium and undergoes phagocytosis, it can prevent vesicular acidification by expressing urease, which breaks urea down and buffers the vesicles carrying the bacteria to neutrality (46
). Other bacteria including Salmonella
, Legionella pneumophilia
, and Chlamydia
can alter vesicular pH by blocking vesicular fusion and maturation (50
). In all of these cases, it is conceivable that administration of GlyAg could induce microbicidal acidification within the endocytic pathway independent of the V-type adenosine triphosphatase and endosomal maturation, thereby assisting in the eradication of the resident pathogen.
We have found that NO-mediated oxidation of endocytosed carbohydrates leads to the release of protons that acidify the local endosomal environment, resulting in GlyAg epitope protection, conventional Ag retention and processing, and TLR9 signaling. Our model therefore provides important new insights into the induction of TLR9-dependent innate as well as protein and GlyAg-mediated adaptive immune responses to microbial products from heavily encapsulated pathogens. These insights may also point to a novel antibiotic approach for some intracellular pathogens.