The Complement Receptor C5aR2: A Powerful Modulator of Innate and Adaptive Immunity

The complement system is an evolutionarily ancient component of the innate immune system that is fundamental for first-line defense against foreign organism invasion, infection, and host tissue damage (1, 2). One important component of the complement system is C5a, which is liberated following complement activation–mediated proteolytic cleavage of factor C5. Being a well-recognized chemoattractant and anaphylatoxin, C5a exerts broad and powerful immunomodulatory roles (3). Complement activation is tightly controlled through several complement regulatory proteins, and failure of such control can lead to excessive or unresolved C5a production, which has been implicated in a plethora of acute and chronic diseases (4). As such, both C5a and its receptors have been attractive targets for fundamental research and preclinical drug development. There are two identified C5a receptors, termed C5a receptor 1 (C5aR1), also referred to as CD88, and C5a receptor 2 (C5aR2), also referred to as C5L2 or GPR77. Although the C5a–C5aR1 interaction is generally acknowledged as having proinflammatory and disease-inducing responses, the role of C5aR2 remains hotly debated, partly because of a dearth of molecular and pharmacological tools. Through this review, we aim to highlight the knowns and unknowns present in the C5aR2 field as well as provide a timely update on the exciting recent breakthroughs concerning both the molecular mechanisms and pharmacological tools of this receptor.

Human C5a is a 74-aa glycoprotein with a helix bundle structure (2). The first described primary receptor for C5a, C5aR1, is a canonical G protein–coupled receptor (GPCR) expressed on a broad range of cell types, including all cells of myeloid origin (such as neutrophils, subpopulations of monocytes, and macrophages), some lymphocytes, and many nonmyeloid cells, including epithelial cells (3). From a signaling perspective, C5aR1 predominantly couples to the pertussis toxin–sensitive Gα_i2 proteins in most cell types and the pertussis toxin–insensitive G protein Gα₁₆ in hematopoietic cells and monocytes (3, 5), and its activation alters cAMP/PKA, ERK1/2, p38 MAPK, calcium mobilization, and β-arrestin–mediated signaling responses (2, 3, 6, 7). C5a possesses pleiotropic functions both within and outside of the immune system that are highly cell type and context dependent (8). From an immune perspective, C5a acts on immune cells to enhance adhesion molecule expression, chemotaxis, degranulation, phagocytosis, and oxidative burst and to modulate cytokine production and secretion (9). Given the highly potent proinflammatory nature of C5a, uncontrolled activation of the C5a–C5aR1 axis has been associated with a myriad of acute and chronic inflammatory diseases (3, 4, 10–14). With C5aR1 being a highly promising therapeutic target, a large number of ligands have been developed over the years (15). A detailed review covering C5aR1 inhibitors can be found elsewhere (16, 17).

The second C5a receptor, C5aR2, was initially reported by Ohno et al. (18) in 2000, in which the novel human gene coding for this orphan receptor was cloned in primary human immature dendritic cells generated from venous blood monocyte culture. Since its discovery, a number of studies have been devoted to interrogate the structure, expression, ligand binding, and signaling of C5aR2.

Located on chromosome 19, the C5aR2 gene appears to be closely related to the C5a, C3a, and the fMLF subfamily of chemoattractant receptors, based on similarity scores (2, 18, 19). C5aR2 possesses one potential N-linked glycosylation site in the N-terminal domain and seven serine and threonine residues serving as potential phosphorylation sites for GPCR kinases (18, 20) (Fig. 1). However, unlike C5aR1, C5aR2 contains the PKC phosphorylation motif in the C-terminal region rather than the third intracellular loop (2, 18). Another distinctive feature of C5aR2 is the replacement of arginine by a leucine residue in the DRY motif, a sequence located in the intracellular side of the third transmembrane domain that allows for heterotrimeric G protein coupling and activation in Class A GPCRs (21).

Based on C5aR2 mRNA studies in human cells and tissues, C5aR2 has approximate concomitant expression with C5aR1, but at consistently lower levels (2, 22). In humans, abundant C5aR2 expression is found in tissues, such as bone marrow, spleen, and lung, and in immune cells, including most myeloid cells and selected T cell subsets (21–24). Recently, significant progress was made in understanding the expression of murine C5aR2 with the creation and use of a novel floxed tandem dye Tomato–C5aR2 knock-in mouse (25), demonstrating human-comparable patterns. This included strong C5aR2 expression in the brain, bone marrow, and airways, and in all myeloid cells (with variable degrees of tissue-specific differences), B cells, and NK cells (25). Surprisingly, both naive and activated T cells failed to display any positive staining for C5aR2, contradicting the two prior studies (26, 27). This discrepancy may be because of the comparatively weaker C5aR2 expression on these cells, which masked its detection, and truly distinct expression patterns between strains/species.

Not surprisingly the cellular localization of C5aR2 is also highly cell type specific (2). Whereas C5aR2 displays a predominant intracellular expression in primary human neutrophils, monocytes, monocyte-derived macrophages, and NK cells (22, 24, 28), the receptor is largely located on the cell surface in human CD4⁺ T cells, LAD2 human mast cell line, and C5aR2-transfected human embryonic kidney–293–C5aR2 cells (23, 24, 26, 29).

Unlike C5aR1, C5aR2 is commonly recognized as being incapable of G protein coupling, a conclusion derived from the absence of the highly conserved DRY and NPXXY motifs in the intracellular end of the third and seventh transmembrane helices, respectively, in C5aR2 (Fig. 1). Indeed, both of these motifs are indispensable for ligand-induced G protein recognition/coupling in Class A GPCRs (2, 21, 30, 31). However, the failure to restore G protein–mediated calcium signaling upon substituting the two motifs in C5aR2 with the corresponding sequences in C5aR1 (D131LCD131RF; N287PMLF-N287PMLY) suggests the presence of multiple repressive mechanisms in C5aR2 (30). The lack of any detectable G protein coupling and conventional signaling, together with the ability to undergo ligand-induced phosphorylation and constitutive internalization, contributed to the early notion of C5aR2 being a decoy receptor for C5a (21, 30). This concept suggests that C5aR2 binds C5a and desarginated C5a (C5a-desArg) and internalizes them for intracellular degradation, thereby providing a negative regulatory mechanism to remove excess C5a from the circulation (2, 21, 30).

The model of an inactive C5aR2 decoy receptor, however, has been challenged by more recent studies. For example, despite the inability of C5aR2 to couple to G proteins, multiple studies have now indicated a consistent signaling role for this receptor through β-arrestins (7, 22, 29, 32). A clear colocalization of C5aR2 and β-arrestin 1 after C5a stimulation was demonstrated using immunocytochemistry experiments and coimmunoprecipitation in human polymorphonuclear neutrophils (22). Time-dependent β-arrestin 2 translocation and colocalization with C5aR2 was also observed in transfected cell lines (29, 32, 33). This colocalization was also validated using the highly sensitive and quantitative molecular technique bioluminescence resonance energy transfer (BRET) to detect C5aR2 and β-arrestin 2 interaction, demonstrating a high level of constitutive C5aR2–β-arrestin 2 complexes (7).

In addition to the ability to recruit β-arrestins, C5aR2 has also been found to physically interact with C5aR1 to form heteromers (22, 24, 34). This C5aR1–C5aR2 heterodimerization was suggested to be indispensable for optimal C5a-mediated AP2 recruitment and C5aR1 internalization (35). Based on these findings, instead of being a decoy receptor as previously proposed, C5aR2 appears to function as an important regulator of C5aR1 signaling. A possible working model of C5aR2 has been proposed: stimulating cells with a high concentration of C5a triggers C5aR1–C5aR2 heterodimerization and β-arrestin recruitment to C5aR2, which in turn facilitates C5aR1 internalization and downregulates C5aR1-mediated ERK signaling (7, 24, 36). The degree of constitutive C5aR2–β-arrestin interaction may be responsible for regulating the pool of C5aR2 available for this inhibitory activity (7). In addition, C5aR2 likely signals through a β-arrestin–scaffolded kinase cascade (37). A detailed picture of C5aR2 signaling and its functional consequences however remains to be elucidated.

The endogenous C5aR2 ligands include C5a and its C5a-desArg form (38–41). Another highly contentious endogenous ligand proposed for C5aR2 is the acylation-stimulating protein, a desarginated form of C3a capable of increasing triglyceride synthesis and glucose uptake in adipocytes (42). However, the activity of desarginated C3a/ASP on C5aR2 is currently debated, for no positive data were observed in later, more specific binding studies performed (41, 43).

To date, only two functionally selective C5aR2 ligands have been identified (43). In a study by Croker et al. (43), BRET-based C5aR2-mediated β-arrestin 2 recruitment assays were employed to screen a panel of linear peptides, resulting in the discovery of two ligands, P32 (Ac-RHYPYWR-OH, EC₅₀ = 5.2 μM) and P59 (Ac-LIRLWR-OH, EC₅₀ = 6.4 μM). Both compounds displayed selective, partial agonistic activity at C5aR2, without any detectable signaling activity at C3aR or C5aR1 (44). Functionally, in human monocyte-derived macrophages, P32 and P59 caused C5aR2-depdendent downregulation of both C5a-induced ERK signaling and LPS-induced proinflammatory cytokine release. In vivo, P32, when administered i.v., significantly inhibited C5a-triggered neutrophil mobilization (43). Despite limitations, such as low potencies and linear peptide structures, as the first C5aR2-selective agonists, P32 and P59 have since served as important tools for deciphering the roles of C5aR2 in physiology and pathophysiology (26, 27, 43). Several other nonselective ligands have also been reported, including A8^D71−73 (45), and staphylococcal bicomponent pore-forming toxins (46), as summarized in Table I. Notably, to date, no C5aR2-selective inhibitors (i.e., antagonists) have been identified.

In the early years following its discovery, C5aR2 was generally regarded as a decoy receptor because of its higher binding affinity for the metabolized form of C5a, C5a-desArg, slow on- and off-rates for ligand binding, inability to couple to G proteins, and constitutive ligand-independent internalization (21, 30). The decoy model proposed that C5aR2 competed with C5aR1 for binding to C5a and C5a-desArg and internalized these ligands for intracellular degradation, thereby reducing the concentrations of circulating C5a and C5a-desArg (3). Conflicting evidence was later shown when the blockade of C5aR2 using an Ab failed to impact on either ligand uptake or C5aR endocytosis in human blood neutrophils (22). This latter finding, however, can possibly be attributed to the low C5a concentrations (2 nM) used in the study. Subsequent investigations using confocal microscopy and BRET assays suggest that C5a-induced C5aR1–C5aR2 heterodimerization, internalization, and β-arrestin signaling as more plausible mechanisms for C5aR2-mediated downregulation of C5a responses (7, 22, 24).

An undifferentiated view of C5aR2 as a simple negative regulator of C5aR1 responses is also at odds with other studies showing C5aR2 deficiency or blockade is beneficial in disease (35, 47–51) or in which the combined inhibition of both C5a receptors was required for an improved outcome (52, 53). Curiously, both C5a- and C3a-stimulated ERK1/2 and Akt phosphorylation events were significantly impaired in murine C5aR2-deficient bone marrow–derived macrophages, which suggests an indispensable role of C5aR2 toward both C5aR1 and C3aR functions (37). A plausible hypothesis is that the heterodimerization between C5aR2 and C5aR1 and potentially with C3aR is important for the delivery and recycling of C5aR1/C3aR after its synthesis/internalization. This delivery role of C5aR2 could partly be the reason behind the high degree of constitutive interaction between C5aR1 and C5aR2 in unstimulated cells (24). So far, however, there has been little investigation on the potential impact of C5aR2 deficiency on the basal trafficking and expression of C5aR1, and, furthermore, the potential interplay between C3aR and C5aR2 remains largely uncharted.

Both the complement system and various pattern recognition receptors provide critical frontline defense against external and internal insults, which thus necessitates the presence of an intricate cross-talk between complement and pattern recognition receptors to mount an optimized immune response (54). One popular area of research has been on the potential influence of C5aR2 toward the activities of TLRs, which thus alter the release of inflammatory mediators. One example is the nuclear protein high mobility group box 1 (HMGB1), a key late inflammatory mediator in sepsis; its in vivo production requires C5a (49, 55). Interestingly, the LPS-induced in vivo production of HMGB1, although unaffected by C5aR1 deficiency, was diminished by C5aR2 deficiency. Correspondingly, macrophages that only expressed C5aR2, without C5aR1, showed robust HMGB1 production upon C5a stimulation associated with MEK1/2, JNK1/2, and PI3K/Akt activation (49).

C5a was further found to have modulatory effects, either synergistic or downregulatory, on LPS-induced TNF-α and IL-6 release from murine neutrophils and macrophages, which were both diminished in C5ar2^−/− cells (37). Additionally, C5a enhanced LPS-induced release of G-CSF in murine peritoneal macrophages, and this effect was diminished in the absence of either C5aR1 or C5aR2, indicating that both C5a receptors were required for this immunomodulatory effect of C5a on G-CSF secretion (53). Furthermore, a study by Croker et al. (24) reported the potentiation effect of C5a on LPS-induced secretion of anti-inflammatory IL-10 by human monocyte-derived macrophages. Surprisingly, in both of the above findings, the potentiation effect was markedly reduced for C5a-desArg despite its similar binding affinities and signaling potencies relative to C5a (24, 53). This discrepancy in immunomodulatory activity was associated with the inability of C5a-desArg to trigger C5aR1–C5aR2 heterodimerization, a potential key mechanism mediating C5a-mediated G-CSF and IL-10 upregulation.

Further insights into C5aR2 interaction with TLR4 were identified through the use of the selective C5aR2 agonists P32 and P59. Both compounds are partial agonists for C5aR2-mediated β-arrestin 2 recruitment, without displaying any observable effect on C5aR1–C5aR2 heterodimerization (43). Interestingly, both ligands potently inhibited LPS-induced IL-6 release from primary human macrophages but failed to demonstrate any significant effect on the production of TNF-α or IL-10, which could be attributed to the additional requirement of C5aR1–C5aR2 heterodimerization for IL-10 modulation (43). The involvement of C5aR2 in the downregulation of IL-6 release was consistent with previous findings by Chen et al. (37) in mouse primary macrophages, where C5a suppressed LPS-induced IL-6 and TNF-α release, which was diminished in C5ar2^−/− cells. Another theory of how C5aR2 modulates TLR activity comes from experimental evidence using human PBMCs (56). Preactivation of TLRs induced hypersensitivity to C5a-driven cytokine production, which was demonstrated to occur through TLR4-induced downregulation of C5aR2 expression. The above studies demonstrate the presence of a direct C5aR2–TLR cross-talk that is independent of C5aR1.

Other than TLRs, potential modulatory activities of C5aR2 have also been suggested toward the NLRP3 inflammasome system. Specifically, in mouse peritoneal macrophages, C5a was found to dampen NLRP3-mediated IL-1β release following LPS and ATP treatment, which was surprisingly only marginally affected by C5aR1 deficiency, implying that the downregulation was primarily dependent on C5aR2 (57). Aside from the limited number of studies on C5aR2 and TLR4/NLRP3 cross-talk, the potential immunomodulation of C5aR2 on other TLRs and pattern recognition receptors remains to be explored. Notably, C5a has been documented to alter signaling of multiple TLRs, including TLR1/2 and TLR6/7-mediated cytokine release in human dendritic cells (58), which could be influenced by C5aR2. Another interesting study conducted in human monocytes reported a TLR activation-induced cell hypersensitivity to C5a associated with a dampened C5aR2 expression (56). It is hoped that the emergence of selective C5aR2 agonists will allow for exciting new findings addressing potential novel C5aR2–pattern recognition receptor cross-talk.

Apart from the potential modulatory effects on innate immune receptors, C5aR2 is increasingly recognized for its functional roles in adaptive immunity, particularly in regulating T cell development and polarization. In a house dust mite–induced experimental allergic asthma model, C5aR2-deficient mice were protected from developing airway hypersensitivity and IgE allergic responses, which was accompanied by an augmentation in pulmonary IL-17A (59), a proinflammatory cytokine secreted by Th17 cells (60). Further experiments supported the robust effect of C5aR2 deficiency on enhancing the release of IL-23 and IFN-γ from myeloid dendritic cells. IL-23 stimulates the generation of Th17 cells, whereas IFN-γ serves to promote and stabilize Th1 polarization, both of which are indispensable in the pathogenesis of autoimmune diseases (61). It is possible that C5aR2 holds important regulatory functions at the dendritic cell/CD4⁺ T cell interface to suppress the induction, maintenance, and differentiation of Th17 and Th1 cells and, therefore, steer immune reactions toward Th2-mediated allergic responses (59).

Aligning with these findings, a study by Arbore et al. (26) highlighted the negative regulatory effect exerted by surface-expressed C5aR2 on intracellular C5a–C5aR1 signaling and NLRP3 inflammasome assembly in activated human CD4⁺ T cells. Specifically, mitochondrial-expressed C5aR1 is engaged by intracellular C5a upon TCR activation and triggers the generation of reactive oxygen species that, in turn, nucleate the assembly of a canonical NLRP3 inflammasome. This proposed functional “complosome” (i.e., intracellular complement/inflammasome system) provides intrinsic IL-1β that is required to drive sustained IFN-γ production and Th1 differentiation in an autocrine fashion. Whereas human resting and activated CD4⁺ T cells express the C5aR2 extra- and intracellularly, agonism of cell surface C5aR2 via P32 allowed for full control over the intracellular C5a–C5aR1 axis and curtailed IL-1β secretion and Th1 induction. Given the study relied on pharmacological blockade of surface-expressed C5aR2 in CD4⁺ T cells, the functional role of intracellular C5aR2, which constitutes the major proportion of the receptor pool, remains to be explored.

Regulatory T cells (Treg) is another T cell subtype critical for maintaining self-tolerance and keeping immune responses in check (61). A recent study highlighted the potential role of C5aR2 in inducing Treg generation, supported by the upregulation of C5aR2 gene expression during Treg generation in a GFP knock-in C5ar2 reporter mouse (27). Accordingly, overexpression of C5aR2 in naive CD4⁺ T cells enhanced their polarization into induced Tregs. Considering the important role played by Tregs in developing long-term tolerance after organ transplant (61), the effect of C5aR2 deficiency was further studied in a mouse model of Treg-dependent cardiac allograft survival, in which accelerated graft rejection was observed following overexpression of C5aR2 and reduced in the absence of C5aR2 (27). To explain these results, the authors proposed the regulatory effects of C5aR2 to be through its scavenging effect on locally produced C5a, which in consequence reduced C5a binding to the C5aR1, an interaction that was inhibitory to induced Treg development. A direct cross-talk between the two C5a receptors could not be concluded from the study (27).

Several studies have now highlighted the potential role of C5aR2 in suppressing the induction, maintenance, and differentiation of Th17 and Th1 cells and supporting the development of Tregs (26, 27). Nevertheless, the actions of C5aR2 appeared to be achieved through the indirect inhibitory effects of the receptor on C5aR1 activity rather than a direct effect of C5aR2-signaling itself. Whether C5aR2 independently affects T lymphocyte development, differentiation, or activation remains to be answered, and should be an area of future research significance.

Although C5aR1 is highly regarded as proinflammatory and pathogenic in a multitude of inflammatory diseases (3, 4), the nature of C5aR2 appears to be much more nuanced and multifaceted (62). Owing to the lack of C5aR2-selective ligands, the majority of earlier C5aR2 studies were conducted in germline C5aR2-knockout animal models. Consistent with the role of C5aR2 as a negative regulator and balancer of C5aR1 surface expression (7, 22), the receptor was reported to be protective in a range of disease models. However, other studies have reported C5aR2 deletion to be protective in disease, thus leading to C5aR2 being considered a “controversial” C5a receptor. A summary highlighting selected studies identifying protective or pathogenic roles for C5aR2 is provided in Table II. One caveat of the current study design paradigm in the C5aR2 research field is the overreliance on knockout animal models. A germline deficiency in C5aR2 may have many downstream effects, potentially affecting the trafficking, recycling, and expression of C5aR1 and/or C3aR, which may partially contribute or wholly account for many of the reported disease outcomes (2).

The potential roles C5aR2 plays in pathophysiology are in its infancy and remain largely unresolved. As highlighted in this review, the functions of C5aR2 are highly dynamic and incorporate multiple systems beyond the canonical complement and C5a receptor pathways. As such, several factors could contribute to the results observed in prior studies. For example, C5aR2 has a clear modulatory effect on C5aR1 expression and activity, so the effect of C5aR2 is dependent on the role of C5aR1 in a particular disease condition and stage. The compositions of cell types involved in inflammatory responses also vary greatly between diseases, as does the expression, localization and function of the two C5a receptors among these cells. Indeed, although human CD4⁺ T cells show low or no C5aR1 but high C5aR2 cell surface expression, this expression pattern is reversed in myeloid cell populations (22, 24, 26, 63). Furthermore, it is well recognized that a major component of the regulatory machinery controlling GPCR activity lies in the trafficking of proteins that move GPCRs between cellular compartments (64, 65), suggesting C5aR1/2 signaling roles may vary significantly between cell type and their activation states.

A plausible model of C5aR2 regulation could be as follows: when expressed at the cell surface, C5aR2 predominantly serves as a downregulator of C5aR1 functions. The receptors sequester excessive extracellular C5a (e.g., during acute inflammation), signal through β-arrestins, and additionally form heterodimers with C5aR1 to facilitate their internalization, thereby helping to prevent an overactivation of C5aR1 activity (Fig. 2A). Surface-expressed C5aR2 also responds to C5a in a C5aR1-independent manner, as exemplified by the presence of C5aR2–TLR4 cross-talk, which could either synergize with or dampen the C5aR1 effect (Fig. 2B). In contrast, during the resting state or under chronic inflammation, in which extracellular C5a levels are low or remain largely unchanged, the intracellular function of C5aR2 dominates. Unlike their surface counterparts, intracellular C5aR2 functions in recycling and delivering C5aR1 back to the cell surface, serving to maintain the pool of surface C5aR1 for a prompt physiological response during an inflammatory insult (Fig. 2C). In the first scenario, deficiency in surface C5aR2 can lead to an overactivation of C5a-mediated acute inflammation, whereas in the second scenario, C5aR2 deficiency can impair C5aR1 responses, which may indeed be protective under certain pathological conditions. This model could thus explain the conflicting findings by multiple laboratories showing both protective and pathogenic functions of C5aR2 in disease models. Nevertheless, the relationship between C5aR2 localization and C5aR1 function remains largely unexplored, especially in the murine system, where the majority of disease models are based. The main contributing factor of this inadequacy is the lack of a functional tools (including drugs and Abs) that could allow for selective targeting of murine C5aR2. Interesting progress is expected in the coming years with the recent development of a floxed tandem dye Tomato–C5aR2 knock-in mouse (25) that will allow for cell tracing and conditional deletion, along with the development of the selective C5aR2 ligands P32 and P59, which have demonstrated functional effects in mice (25, 43).

From a therapeutic perspective, C5aR2 has demonstrated multifaceted modulatory roles in diverse diseases. Whereas C5aR2 activity appears to be protective and beneficial under some conditions, the receptor may also synergize with C5aR1 to exacerbate disease outcomes. On the one hand, pharmacological activation of C5aR2 could be beneficial toward chronic inflammation and autoimmune diseases by promoting intrinsic regulatory mechanisms that keep C5a–C5aR1 in check and enhancing self-tolerance. On the other hand, under acute conditions such as sepsis, in which overwhelming C5a–C5aR1 activities dominate, C5aR2 blockade may be a method to arrest the recycling and re-expression of C5aR1. Another key aspect is that therapeutically activating C5aR2 may provide an effective means to inhibit the intracellular C5a–C5aR1 system in T lymphocytes that are at present largely out of the reach of the conventional C5aR1 inhibitors (26). Given the dual nature of C5aR2 activity across both innate and adaptive immunity, partial C5aR2 agonists (such as P32 and P59) may provide unique clinical advantages by moderately upregulating C5aR2 functions without excessively altering immune pathways.

With an increased understanding of the complexity of the C5a–C5a receptor system, the most recently identified C5a receptor, C5aR2, is accumulating attention for its unique role in dampening C5a signaling, modulating C5aR1 activity, and more recently, independent cross-talk with other pattern recognition receptors and intracellular inflammasomes. Rather than treating C5aR2 as a single entity with controversial functions, we should begin to consider the receptor as a multifaceted modulator affecting multiple systems and cell types. However, several challenges remain that impede the further advancement in the C5aR2 field, including a dearth of selective ligands and an overreliance on genetic knockout animal models and mammalian nonimmune cell lines. To address these issues, more potent C5aR2-selective ligands, both agonists and antagonists, that can enable specific targeting of C5aR2 without necessarily a direct interference with C5aR1 function will need to be developed. Additionally, there is a genuine need for the development and implementation of better in vitro assays that can be exploited to decipher the complex molecular mechanisms underlying C5aR2 function and signaling, and how they may interact with other immune pathways.

T.M.W. is an inventor on patents describing C5aR2 ligands. The other authors have no financial conflicts of interest.