Flagellin: a unique microbe-associated molecular pattern and a multi-faceted immunomodulator

Mammals peacefully co-exist with vast numbers of different bacteria most of the time. Whether pathogenic, nonpathogenic or commensal to a mammalian host, bacteria produce microbe-associated molecular patterns (MAMPs) or pathogen-associated molecular patterns (PAMPs) that are recognized by multiple classes of germline-encoded pattern-recognition receptors (PRRs). Flagellin, a major structural protein of bacterial flagella, is composed of four domains (D0, D1, D2 and D3), of which the D0 and D1 domains are the most highly conserved. As a MAMP derived from bacteria, flagellin is dually recognized by Toll-like receptor 5 (TLR5) in the extracellular space and by NLR neuronal apoptosis inhibitory protein 5 (NAIP5) and NAIP6 in the cytosol.^{1, 2} TLR5 detects flagellin by binding specifically to its highly conserved domain D1, whereas NAIP5 or NAIP6 detects intracellular flagellin by specific binding to its C-terminus via 35 amino acids in the D0 domain. Flagellin-activated TLR5 triggers proinflammatory genes, including IL-6, TNF-α, and KC, whereas flagellin-activated NAIP5/6 promotes the assembly of the NLRC4 inflammasome, activation of caspase 1, secretion of IL-1β/IL-18 and pyroptosis of infected cells.¹

In the past century, our understanding of the interplay between host innate immunity and microbes is mostly derived from the study of pathogenic bacteria and their related diseases. In the recent decade, our view on the interplay is largely broadening to include questions about the symbiosis of bacteria in our gut and its influence on our health.³ According to our current understanding, commensals are even welcomed by the host because they compete for niches in the extracellular milieu and help to protect against pathogens. It is a curious fact that the immune system can distinguish commensals from pathogens. However, little information is currently known about how the immune system discriminates between a pathogen and a commensal. For example, both pathogenic Salmonella typhi and commensal Escherichia coli may produce equivalent amounts of chemically identical LPS, which are detected by TLR4. However, the immune system responds significantly to the pathogen and not to the commensal. Regarding flagellin, its vast diversity in size and composition derived from different bacteria strains and species together with its dual checks by TLR5 and NAIP5/NLRC4 pathways may provide an opportunity to answer the question of distinction. We comparatively analyzed two recombinant flagellins derived from S. typhi and the well-known nonpathogenic bacteria strain E. coli K12 strain MG1655 and tested their ability to activate the TLR5 or NAIP5/NLRC4 pathways. Similar activation of the TLR5 pathway can be detected by either pathogenic or nonpathogenic flagellin. However, nonpathogenic E. coli flagellin induces minimal activation of NAIP5/NLRC4 and caspase 1, whereas Salmonella Typhi flagellin elicits robust caspase 1, increased IL-1β/IL-18 levels and cell death (see Figure 1). This finding suggests that macrophages might be able to discriminate commensal and invading pathogenic bacteria by differential recognition of flagellins and initiate innate immune responses only to pathogenic bacteria.⁴ This study offers a possible molecular explanation for the findings of Franchi et al.,⁵ which demonstrated that intestinal mononuclear phagocytes (iMP) can actively promote host defense against pathogenic but not commensal bacteria in the intestine by differential production of mature IL-1β through NLRC4 inflammasome activation.⁵ In this regard, MAMP is a more suitable term than PAMP in discussing the interplay between host innate immunity and microbes, including both pathogens and commensals. It is conceivable that the immune system needs to set a threshold for its response to MAMPs derived from commensal or pathogenic microbes. Otherwise, some diseases, such as ulcerative colitis and Crohn's disease, may occur, which are hypothesized as problems in distinguishing commensals from pathogens or overreacting to commensals.⁶ However, the specific reasons underlying the distinction of MAMPs derived from different bacteria remain elusive and warrant further intensive investigation.

Figure 1.

Diagram depicting how to recognize flagellins and thereby trigger immune responses to bacteria. As dual detectors, TLR5 and NAIP5/6 recognize flagellins and detect the infected bacteria outside and inside the cells, respectively. Once it passes through the mucosa and is captured by macrophages, pathogen-derived flagellin (PF) tightly binds NAIP5/6, resulting in increased levels of IL-1β/IL-18 and macrophage pyroptosis. In addition, commensal-derived flagellin (CF) exhibits low affinity to NAIP5/6, thus slightly triggering the assembly of the NLRC4 inflammasome in macrophages. The cytokine and macrophage profile may determine the differential recognition for commensal and pathogenic bacteria (right panel). Epithelial cells recognize PF or CF via TLR5 and secrete certain soluble factors, modulating DC. The modulated DCs regulate the B cell response in a T-cell-dependent or -independent manner to enhance the flagellin-specific antibody response. In addition, NLRC4 activation in macrophages downregulates TLR5-mediated antibody immune responses (left panel).

As an evolutionarily conserved and vital element for mobile bacteria regardless of pathogenic, nonpathogenic or commensal characteristics, flagellin can be recognized by TLR5 on the cell surface and/or detected by cytosolic receptor NAIP5 or NAIP6 inside cell. The recognition and response profile of flagellin by the innate immune system is possibly location dependent (extracellular or intracellular), type dependent (amino acid sequence of flagellin) and quantity dependent (amount of flagellin molecule). The dual sensor strategy implies that the immune system can recognize flagellated pathogens and initiate responses based not only on whether MAMPs are presented but also on where those MAMPs are presented.⁷ The stronger and weaker response strategy implies that the immune system can also recognize flagellated bacteria and initiate responses based on the bacterial genome, which determines the relationship between host and microbe.⁴ Whether a threshold value is defined by the immune system for recognition and response to flagellin, either for TLR5 or NAIP5, remains unclear. In this respect, we may conceive different combinations of immune responses to flagellin: only TLR5 activated, only NAIP5 activated, both TLR5 and NAIP5 activated, stronger TLR5 activated and stronger NAIP5 activated. Thus, a complex cross-talk between the dual sensor TLR5 and NAIP5, and the differential output must exist in the immune recognition of and response to distinct flagellins. However, the dual activities of flagellin make this molecule a very good tool for the investigation of possible cross-talk between TLR5 and NAIP5. Using this tool, we assessed the effects of flagellin and its mutants lacking the ability to activate TLR5 and NAIP5/NLRC4 alone or in combination on the adaptive immune responses against flagellin. Activation of NLRC4 in macrophage downregulates TLR5-mediated antibody immune responses (see Figure 1).⁸ The finding revealed an interaction between NLRC4 activation and TLR5 activation by flagellin, which affects the output of specific immune responses. Using recombinant variant flagellins, we also found that NAIP5 and NAIP6 differentially recognize flagellins (unpublished data). However, how the cross-talk and interaction between NAIP5/6 and TLR5 signaling on specific immune responses affect bacterial infections remains to be elucidated.

The dual activities of flagellin in triggering the TLR5 and NAIPs/NLRC4 inflammasome pathway also suggest that this molecule to be a unique multi-faceted immunomodulator. Recent studies have demonstrated that recombinant flagellin or its derivate exhibit potential as a potent immunotherapy agent for treatment of acute radiation syndrome, viral infection and even tumors. A pharmacologically optimized flagellin derivative named entolimod has demonstrated preclinical potential in the immunotherapy of several cancers and has been approved for Phase I clinical trial.^{9, 10} The immune modulation activity of flagellin can also be harnessed to enhance the efficacy of vaccines by eliciting a stronger immune response to the immunogen. In this regard, flagellin is a promising novel vaccine adjuvant. Of particular interest, flagellin is an effective mucosal vaccine adjuvant when administered via a mucosal route, such as the intranasal (i.n.) route, which can prominently increase a protective IgA response.¹¹ We developed a series of recombinant flagellins based on different combinations between its dual activities for mucosal adjuvant, which can enhance the antigen-specific IgA antibody response by priming nasal epithelial cells to modulate airway DCs and B cells.¹² Using a fusion strategy, we developed an anti-caries mucosal vaccine by either direct fusion or replacement of the highly variable region of flagellin with caries vaccine antigen, which can promote a protective salivary IgA antibody response and inhibit dental caries prophylactically and therapeutically in rat caries model (unpublished data).^{13, 14}

As a unique microbe-associated molecular pattern, does flagellin possess the ability to activate some other unknown pathway in addition to the TLR5 and NAIPs/NLRC4 inflammasome pathway? If yes, considerably more complex interactions and cross-talks among these activities of flagellin might exist. Thus, more potential underlying physiological and pathological mechanisms might be explored to further understand the relationship between host and pathogen vs host and commensal.