Understanding Fc function for rational vaccine design against pathogens

ABSTRACT

Antibodies represent the primary correlate of immunity following most clinically approved vaccines. However, their mechanisms of action vary from pathogen to pathogen, ranging from neutralization, to opsonophagocytosis, to cytotoxicity. Antibody functions are regulated both by antigen specificity (Fab domain) and by the interaction of their Fc domain with distinct types of Fc receptors (FcRs) present in immune cells. Increasing evidence highlights the critical nature of Fc:FcR interactions in controlling pathogen spread and limiting the disease state. Moreover, variation in Fc-receptor engagement during the course of infection has been demonstrated across a range of pathogens, and this can be further influenced by prior exposure(s)/immunizations, age, pregnancy, and underlying health conditions. Fc:FcR functional variation occurs at the level of antibody isotype and subclass selection as well as post-translational modification of antibodies that shape Fc:FcR-interactions. These factors collectively support a model whereby the immune system actively harnesses and directs Fc:FcR interactions to fight disease. By defining the precise humoral mechanisms that control infections, as well as understanding how these functions can be actively tuned, it may be possible to open new paths for improving existing or novel vaccines.

INTRODUCTION

Vaccine campaigns to date have resulted in a significant reduction of morbidity and mortality across the ages, estimated to prevent approximately 2–3 million deaths annually (1, 2). However, new vaccines are still urgently needed to combat widely circulating agents such as Mycobacterium tuberculosis, human immunodeficiency virus (HIV), Plasmodium, SARS-related coronaviruses, etc. In addition, new vaccination strategies are needed to address disease prevalence in vulnerable populations that respond poorly to current vaccine strategies such as the elderly, neonates, immunocompromised, etc. Thus, while existing vaccine platforms have successfully revolutionized the battle against many pathogens to date, a deeper appreciation of immune mechanisms of protection is needed for next-generation formulations and platforms.

Induction of antibodies represents the primary correlate of protection against infection/disease for most vaccines (3). Emerging data clearly illustrate that vaccines drive protection via mechanistically distinct antibody functions including (i) neutralization, (ii) opsonophagocytosis, (iii) cytokine secretion, (iv) inflammatory cell degranulation, and (v) complement deposition and several others. For example, neutralizing antibodies are crucial for protection against measles (4, 5), opsonophagocytic antibodies are key to defense against Streptococcus pneumoniae (6), and antibodies able to drive cytotoxic destruction of infected cells are linked to immunity against influenza (7, 8). It is important to note that these functions frequently synergize with each other to confer protection, and analysis of one function alone is likely not sufficient to account for the entirety of “protective immunity.” Recently developed sophisticated functional immune assays that investigate antibody mechanism of action represent a more pivotal readout of immunity and marker of true clinical efficacy. This review aims to review antibody mechanisms of protection across pathogens, explore the mechanisms by which these antibody functions are driven selectively by the immune system, and discuss novel strategies to harness these functions for future vaccine development.

MECHANISMS OF ANTIBODY-MEDIATED PROTECTION

Antibodies recognize pathogens through their antigen-binding fragments (Fab). Each human immunoglobulin is composed of two Fab domains and a constant crystallizable fragment (Fc fragment) domain. The Fab fragments are composed of a heterodimer of a light chain (LC) and a heavy chain (HC). The Fc-domain is composed of HC only and linked to the Fab via a hinge region (Fig. 1A) (9). The Fc domain interacts with effector cells via their Fc-receptors (FcR) and/or complement receptors (Fig. 1B). The Fc domain binds to specialized cell-surfaced and secreted receptors, including FcR, complement, and other non-canonical receptors found on immune cells (10). Thus, the Fab provides antigen specificity, tethering the antibody to the pathogen or pathogen-associated peptides, whereas the Fc links the adaptive (Fab) and innate (FcR) immune systems.

Fig 1.

IgG structure and Fc-mediated effector mechanisms and antibody glycosylation. (A) An IgG consists of two fragments, an antigen-binding fragment (Fab) and a crystallizable fragment (Fc) connected by a flexible hinge region. The Fab region includes variable and constant domains both on the heavy (VH and VL) and light chains (CL and CH1). The Fc fragment contains complementary CH2 and CH3 domains with an N-linked glycan attached to a conserved glycosylation site at Asn297 of the CH2 (simplified picture). (B) Antibodies recognize and bind to pathogens via the Fab fragment. The Fc-region, then, interacts with Fc-receptors (FcRs) present in various types of immune cells. (C) Antibody subclasses and isotypes bind to FcRs with different affinities and are differentially expressed across various cell types. (D) The structure of the biantennary N-linked glycan, attached to asparagine 297 of the IgG heavy chain constant domain (CH2), sits between the arms of the Fc and regulates antibody function by altering the structure of the Fc. Changes in the shape and flexibility of the Fc alter the affinity of the Fc for Fc receptors. The core glycan comprises of two consecutive N-acetylglucosamine (GlcNAc) molecules, followed by a mannose and two additional mannose antennae, each with a single GlcNAc attached. In addition, four sugars can be added in variable combinations. Two galactoses can be added to the antennary GlcNAcs (galactosylation), followed by sialic acid that can be added to each of the galactose residues (sialylation). In addition, a single GlcNAc can be added to the core mannose, creating a bisecting arm (bisection). Finally, fucose can be added to the first GlcNAc (fucosylation).

The Fc-domain is a multifaceted linker, providing specific instructions to the innate immune system on how it should process the target to which the antibody binds. FcRs recognize the Fc regions of IgG antibodies and include Fc-gamma receptor I (FcγRI, also known as CD64), FcγRII (also known as CD32), and FcγRIII (also known as CD16). These will be referred to as FcγR throughout this review. Additional FcRs exist for other antibody isotypes including FcϵRI:IgE, FcαRI/CD89:IgA, FcμR:IgM, and Fcα/μR:IgA/IgM (11). The wide array of FcRs provides a critical mechanism for antibodies of distinct isotypes to interact with and activate signaling cascades of numerous cells of the immune system upon pathogen recognition.

IgG can also bind to non-canonical receptors such as the neonatal Fc receptor (FcRn). FcRn binds the Fc-domain of IgG with high affinity at low pH, allowing for efficient capture and release of antibodies across cell barriers. This mechanism also plays an essential role in serum IgG homeostasis. In addition to their role in trafficking IgG1, FcRn is expressed on multiple immune cells, such as dendritic cells (DCs), macrophages, monocytes, neutrophils, and B cells, to facilitate the uptake and degradation of IgG-opsonized pathogens. For example, the association of hemaglutinin (HA) head-specific antibody-viral complexes with FcRn prevents viral ribonucleoprotein transport to the nucleus, preventing virus replication (12). In addition to FcRs, complement receptors (CRs) expressed by antigen-presenting cells (APCs) are essential for mediating effector function. Complement proteins C3b and C4b, deposited on immune complexes following C1q engagement, can bind to the transmembrane CR1 found on innate immune cells such as neutrophils, monocytes, and dendritic cells. The binding of the C3b-immune complex to membrane CR1 on antigen-presenting cells induces opsonization and antigen presentation in the context of major histocompatibility complex (MHC) molecules to enhance T-cell immunity (13). In addition, C3b, as well as other derivatives iC3b, C3dg, and C3d, bind to the B-cell receptors CR2 and CR1 and play a key role in enhancing B-cell immunity by increasing their entrance and survival in germinal centers (GCs) as well as stimulating memory B-cell development (14, 15).

FcRs are expressed in diverse combinations on different cell types, which leads to unique response patterns. The majority of innate effector cells (monocytes, macrophages, DCs, basophils, and mast cells) co-express both activating FcRs (FcγRI, FcγRIIa, and FcγRIIIa) and inhibitory FcR (FcγRIIb). This allows for the balancing of Ig-mediated cellular activation (16). FcγRI binds to IgG much more potently than other receptors. FcγRII and FcγRIII are characterized by a low affinity for IgG (with K_D for human IgG1 = 10⁻⁵–10⁻⁷M) and are also able to engage multimeric immune complexes (IgG-antigen complexes) through high avidity interactions (17).

Antibody isotype/subclass selection and glycosylation collectively shape the binding affinity of antibodies to FcRs and complement. Following initial antigen recognition, either infection or immunization, affinity maturation of the Fab and class switch recombination (CSR) of the Fc lead to the production of higher affinity IgGs of different subclass/isotypes (18, 19). The process leads to the directed combination of particular antigen specificities with unique functions via the maintenance or replacement of the Ig heavy chain (HC) with one of 10 distinct constant regions [e.g., Cμ (IgM), Cγ (IgG1-4), Cα (IgA1 or 2), or Cε (IgE)] each with a unique affinity for Fc-receptors and complement (20). Humans possess four IgG subclasses (IgG1, IgG2, IgG3, and IgG4) with different functionality and affinity for FcγRs (21). IgG subclasses demonstrate different affinities for FcγRs, with IgG3 having the highest affinity for most FcγRs, followed by IgG1, IgG4, and IgG2 (Fig. 1C) (22).

In addition to isotype and subclass selection, changes in Fc-glycosylation at conserved sites on the antibody-heavy chain profoundly alter antibody interactions with FcRs (Fig. 1D). For example, when the N-glycan is enzymatically cleaved from IgG, the antibody’s ability to bind to FcγRs is abrogated (23). Given the importance of IgG Fc-glycosylation on shaping FcγR binding, specific Fc-glycan structures have been linked to distinct Fc-effector functions. Along these lines, a loss of fucosylation (afucosylation) improves IgG affinity for FcγRIIIa (24). In addition, increased sialylation has been linked to anti-inflammatory functions (25).

Shifts in glycosylation patterns have been documented in different pathological and physiological conditions such as between autoimmune and infectious diseases, as well as age, vaccination, and pregnancy. Specifically, reproducible shifts toward enhanced agalactosylated antibodies emerge under chronic inflammatory conditions and a rapid loss of galactose on disease-specific antibodies precedes the onset of autoimmune flares (26). After vaccination, antigen-specific antibody populations are decorated with different Fc-glycans (27, 28). Moreover, these glycans shift in a pathogen-specific manner (29). Similarly, glycan signatures can shift depending on a vaccine adjuvant, even if the vaccine is targeted to the same antigen (29). These data collectively suggest that antibody glycosylation is a highly regulated process (30). Intracellularly, a complex cascade of glycosyltransferases and glycosidases acts to shape the Fc-glycan in a sequential and highly organized manner in the Golgi apparatus. Analysis of the sialyltransferase St6Gal1 and the fucosyltransferase Fut8 expression in plasmablasts and memory B cells at different time points following trivalent influenza virus vaccination has demonstrated significant shifts in enzyme expression, likely to be key to shaping the antibody glycans (31). Along these lines, alterations in Fc glycoforms on HA-specific IgGs have been directly linked to glycosyltransferase expression profiles (30).

The Fab represents an evolutionary marvel, permitted to evolve genetically, via somatic hypermutation, to explore a nearly infinite number of sequence-based solutions to develop ultra-specific binding to an antigen. This process of evolution and integrated perpetual selection within the germinal center (GC) of the lymph node enables the selection of antibodies to mount a multipronged targeting program to a pathogen at distinct regions of interest. The ability of B cells to create this diversified repertoire of binding antibodies is owed to their somatic hypermutation (SHM) through mechanisms such as activation-induced cytidine deaminase (AID). Cytidine deamination converts the nucleotide to uracil, which can result in (1) C → T mutation during replication as uracil will be read as thymine by DNA polymerase; (2) activation of the double-strand break repair pathway and translesion DNA synthesis, resulting in mutations at the C or neighboring nucleotides; and (3) non-canonical mismatch repair pathway activation in which the highly mutagenic DNA polymerase η is recruited to the G:U mismatch (32–35). This entire mutagenic process occurs largely within the variable (v) and diversity (d) gene segments of the immunoglobulin loci within B cells (36–38). The hotspots of these mutations within the v and d gene segments correspond to the complementarity-determining regions (CDR). These CDRs, located on the heavy chain of the immunoglobulin, are responsible for the direct contact of the Fab with the antigen (39–41). Therefore this deliberate, gene-segment localized mutagenic process can give rise to a bevy of unique antibody-producing B cells. Immunoglobulins produced by B cells that exhibit low affinity to the target antigen undergo apoptosis whereas B cells producing antibodies of strong antigen binding potential are expanded (42). Via this process, the immune system is progressively populated by waves of antibodies that evolve to bind more rapidly, more specifically, and with higher affinity to target antigens of interest with a diversity of functions.

A notable class of antibodies are those that specifically recognize regions of pathogens used to mediate penetrance. These “neutralizing antibodies” selectively displace or interfere with pathogen-antigen binding with stunning precision. For example, neutralizing antibodies to HIV and Dengue virus must bind to specific minimal, often quaternary epitopes, on virion glycoproteins to drive neutralization (43, 44). Similarly, antibodies able to block and displace bacterial toxins, such as those produced by Corynebacterium diphtheriae and Bacillus anthracis, are also key to vaccine-mediated protection (45, 46). Increasing evidence, however, has illustrated that blocking a toxin or pathogen from accessing entry receptors cannot completely ablate disease. Data increasingly point to the important role of Fc-mediated recruitment of innate immune effector functions in the clearance of toxins and pathogens. For some neutralizing antibodies against viruses, such as HIV, Fc-mediated functions are also essential for protection in vivo. In this case, protective activity can be lost if antibody-mediated interactions with FcRs are abrogated (47). Therefore, there exists an intimate collaboration between the Fab and the Fc for protection.

While neutralization is critical for protection against several infections, neutralization is not necessary against all infections (48, 49). For example, in the case of influenza, some antibodies to the stem of the hemagglutinin (HA) antigen, involved in binding and infection, do not neutralize the virus in vitro. However, these antibodies can block infection in vivo (50, 51). Interestingly, the protective activity of these antibodies depends on the ability of the antibodies to interact with FcRs and appears to depend on the selective recruitment of NK cells to exert antibody-dependent cellular cytotoxicity (ADCC) (52). Moreover, population-level studies have also demonstrated increased levels of antibody-dependent complement deposition (ADCD), and antibody-dependent cellular phagocytosis (ADCP) by myeloid-lineage cells and neutrophils in cohorts of individuals that exhibit a reduced likelihood of infection (49, 53).

Antibody-dependent cellular phagocytosis

ADCP is the humoral immune function whereby antibody-opsonized pathogens or infected cells bind to FcRs or complement on the surface of phagocytes to drive immune complex uptake (Fig. 2A). Professional phagocytes include monocytes, macrophages, neutrophils, DCs, or eosinophils, all of which express different combinations of FcRs, including FcRs for all antibody isotypes. However, FcR expression varies on these cells in a tissue-dependent manner, where mucosal macrophages and DCs express high levels of FcαRI (CD89), not found on peripheral macrophages and DCs, enabling these cells to respond to antibody-opsonized material in a tissue-specific manner.

Fig 2.

The impact of different routes of vaccination and formulation on Fc-mediated antibody effector functions. (A) ADCP, ADCC, and antibody-dependent complement activation are the main Fc-receptor-mediated antibody effector functions leading to the destruction and clearance of the pathogen and therefore protection against infection/disease. Binding of the pathogen-specific antibodies to the surface of the pathogen leads to the formation of immune complexes. The interaction of the Fc-domain of antibodies from immune complexes and Fc-receptors present on multiple types of innate immune cells mediates the downstream effector functions. (B) Different routes of immunization, such as intravenous (IV), intradermal (ID), intramuscular (IM), and mucosal, enable to target distinct types of APCs leading to the production of particular cytokines further shaping the antibody type and their Fc-effector functions as well as proper choice of a delivery system and adjuvant initiate distinct early immune response at the site of vaccine administration and therefore influence the Fc-subclass/Fc-glycosylation selection (C).

The early steps of ADCP depend on the number and combination of FcRs engaged (relative to the ligand and antibody density on the target antigen). The binding of activating classical FcRs leads to the phosphorylation of immunoreceptor tyrosine-based activation motifs (ITAM) (on activating FcRs) and initiation of signaling cascades (54). Given the higher affinity of IgG1 and IgG3 for FcγR binding, immune complexes composed of these antibodies lead to rapid triggering and robust phagocytic clearance of these antibody complexes. In addition, IgA and IgM have also been implicated in rapid and robust immune complex uptake, triggered by both isotype-specific Fc-receptors and complement receptors.

Depending on the immune cell triggered, different inflammatory responses may evolve following phagocytosis. In human blood, neutrophils are a critical first line of defense against pathogens. Due to their unique FcR repertoire, including FcγRIIa (CD32a), FcγRIIIb (CD16b), FcγRI (CD64), and FcαRI (CD89), these innate immune cells represent highly specialized gatekeepers able to rapidly respond to antibody-complexes, drive rapid degradation and clearance, and support antigen cross-presentation (55). However, depending on the FcRs triggered, distinct inflammatory consequences can evolve. For example, IgA/FcαRI binding can lead to rapid neutrophil extracellular trap formation, or NETosis killing (56). Conversely, IgG activation of neutrophils can lead to degranulation and cytokine secretion. DCs, on the other hand, pick up antibody complexes, and depending on the antibody/FcR binding profiles, can lead to lysosomal proteolytic degradation and/or antigen presentation through MHCI and MHCII. This simultaneously allows for a response to existing pathogens and a priming of new memory responses (57). This DC switch is regulated by antibody-binding to the neonatal Fc-receptor, FcRn, which drives uptake in a pH-dependent manner.

Beyond FcRs, immune complex uptake may also occur following complement deposition on immune complexes. In addition to their role in creating pores in membranes, the complement components deposited on immune complexes interact with complement receptors (CR) found in phagocytic immune cells. For example, the first component of the classical complement cascade, C1q, interacts with the C1q-receptor (C1qr) found in myeloid cells (58). In addition, following the activation of C3 and deposition of C3b, iC3b (after cleavage of C3b), and C3d on immune complexes, complement receptors (CRs) can mediate phagocytosis. These receptors include CRIg, the multifunctional CR1, CR2, and the β2-integrin members CR3 and CR4. Surface expression of these complement receptors differs, with CR1 and CR2 being expressed on myeloid, neutrophils B cells, and macrophages, CR3 and CR4 being present on myeloid, neutrophils, activated lymphocytes, and natural killer cells, and CRIg being present on tissue-specific macrophages (13, 28, 59–68). Complement-mediated phagocytosis is particularly essential during the acute immune response to infection, where low-affinity IgMs efficiently recruit C1q and drive rapid and effective phagocytic clearance. In addition, this IgM/C1q phagocytic axis also plays a critical response in protection against T-cell-independent antigens. Moreover, because T cells also express C3aR, C5aR1, and C5aR2, adaptive immune cells may also be activated by antibody complexes, resulting in broader immune activation (69–72). Given the extensive expression of CRs, these receptors likely collaborate with FcRs to drive clearance, intracellular routing, and immune activation/signaling to nearby cells.

ADCP has been implicated in the protection against bacteria (73), parasites (61), fungi (74), and viruses (75). Opsonophagocytosis plays an important protective role against Streptococcus pneumoniae, Staphylococcus aureus, Pseudomonas aeruginosa, and Klebsiella pneumoniae. Surface-exposed proteins and cell wall components are targets for antibodies able to interact with FcγRs and CRs, leading to rapid bacterial uptake and destruction. For example, meningococcal-specific polysaccharide-specific IgG2 that interacts with FcγRIIa leads to increased killing via neutrophils (73, 76). Similarly, neutrophil-mediated opsonophagocytic killing plays a critical role in immunity against fungal Candida albicans infection (77), shown recently to be critical in the defense against multidrug-resistant fungal pathogens (74).

Recent evidence points to a critical role for opsonophagocytosis in antiviral and anti-parasitic immunity. For example, vaccine-induced antibody-mediated phagocytosis has been linked to protection against the simian immunodeficiency virus (SIV), the evolutionary precursor of HIV (78). Interestingly, using the same DNA/Adenoviral-based SIV vaccine, monocyte recruiting IgG phagocytic responses were induced by intra-muscular vaccination, whereas IgA-driven neutrophil phagocytic responses were induced by intranasal immunization, both resulting in ~50% protection from infection. Given the presence of higher levels of phagocytic antibodies among human spontaneous controllers of HIV infection (79) and our emerging appreciation for the role of homozygosity for the low-affinity FcγRIIa allele among HIV rapid progressors (defined by a low CD4⁺ T-cell count) (80), these data highlight the role of opsinophagocytosis in HIV control. Moreover, the data point to opportunities to drive distinct phagocytic responses, via monocytes or neutrophils, by vaccination route.

Phagocytic responses by neutrophils were linked with long-term vaccine-mediated protection against Ebolavirus (EBOV) in non-human primates (81). Interestingly, phagocytic responses correlated with survival post-virus challenge were against the truncated EBOV glycoprotein (GP), termed the soluble GP (or sGP) (81–83). These functionally leveraged antibodies were more clear correlates of protection than neutralizing antibodies themselves. The vaccine platform was a live-attenuated vesicular stomatitis virus that encodes for EBOV GP, administered intramuscularly (84). This again hints at vaccination platform-based and/or immunization sites as key determinants of antibody functions (see section Prospects for Fc programming).

The COVID-19 pandemic provided unprecedented insight into how ADCP functions as a novel pathogen. Humoral immunity was low in the human population prior to 2019, allowing for the generation of a novel, polyfunctional antibody response through vaccination and/or recovery. The primary target of SARS-CoV-2-directed antibodies is the spike glycoprotein. ADCP of the spike is mediated through antibodies recognizing the viral protein at multiple subdomains (85–90), whereas neutralizing antibodies are much more specifically targeted to the receptor binding domain. The presence of ADCP-leveraged antibodies is linked to positive clinical outcomes and protection against antigenically drifted SARS-CoV-2 variants (85, 86, 91).

Similarly, vaccine-induced phagocytic antibodies against malaria have been implicated in the protection against parasitic infection (92). Recent immune correlates analysis of the most advanced malaria vaccine candidate, RTS’S (93, 94), have pointed to a critical role for phagocytosis as a key mechanism of protection following vaccination (92). While decades of data had clearly linked protection from infection following vaccination to vaccine-induced titers, functional characterization of these antibodies pointed to an enrichment of phagocytic functions in individuals able to resist the Plasmodium challenge (93).

Given the abundance of phagocytic cells in the skin, blood, and liver, it remains unclear where this protective activity may occur on a pathogen-by-pathogen basis. For example, the presence of spike-reactive IgG1 in the lungs was a correlate of protection for COVID-19 but compartment-specific functions of ADCP have not yet been delineated for COVID-19. This is also true for Plasmodium, which has a multi-compartment life cycle. Future studies into how ADCP is regionally activated are warranted.

Antibody-dependent cellular cytotoxicity

ADCC represents the mechanism by which innate effector cells (lacking specificity for an antigen) recognize antibody-opsonized pathogens/antigens, or entire infected cells, and trigger the cytotoxic destruction of that antibody-opsonized target (Fig. 2A). Cytotoxicity is driven by the release of cytotoxic granules containing perforins and granzymes from cytotoxic effector cells. ADCC is best described for IgG-mediated natural killer cell (NK) activation. Specifically, IgG1 and IgG3 appear to drive ADCC most effectively due to their affinity for FcγRIIIa (CD16), a dominant FcγR found in NK cells. However, the engagement of FcγRI (CD64) and FcγRII (CD32) has also been implicated in macrophage-, neutrophil-, or eosinophil-mediated degranulation and cytotoxicity (16). Furthermore, emerging data point to a role for IgA-FcαR and IgE-FcεR as additional mediators of ADCC by neutrophils, particularly in the setting of cancer treatments (95, 96). ADCC-like activities can also occur in cytolytic granule-producing neutrophils, monocytes, macrophages, basophils, eosinophils, NK T cells, αβ, and γδ CD8+ T cells, via Fc-mediated activation (97). The distinct granule contents found in each of these cell types can result in a remarkable breadth of distinct inflammatory and cytotoxic outcomes.

ADCC has been implicated in protection against multiple viruses and some intracellular bacteria and parasites. ADCC can act both indirectly to eliminate the pathogen within infected cells as well as directly on particular pathogens. For example, ADCC activity has been linked to disease control following HIV, influenza virus, Ebolavirus, or respiratory syncytial virus (RSV) challenge(s). Several studies have noted enhanced ADCC-inducing antibodies among HIV spontaneous controllers compared to individuals progressing to AIDS (98). Similarly, both animal and human studies have demonstrated a critical role for ADCC in the control of influenza, where survival and recovery following H7N9 infection were preferentially observed in the presence of anti-HA antibodies able to activate primary NK cells (99). Conversely, non-survivors of both HIV and influenza infections generate lower levels of FcγR binding antibodies and reduced ADCC (99). Along the same lines, reduced NK cell counts have been observed in hospitalized children with RSV lower respiratory infection (100). Moreover, NK cell-activating monoclonal antibodies have been shown to provide enhanced protection against Ebolavirus, particularly for antibodies targeting the base of GP (101). For COVID-19, the kinetics of FcγRIIIa-binding antibodies were linked to severe disease survival (91). Finally, as mentioned above, cross-influenza protective antibodies that target the HA-stem, which is more conserved across influenza viruses, utilize ADCC-like activities for protection (102). Collectively, the ability to recruit NK cells appears key to protection against a variety of viruses.

Interestingly, ADCC has also been implicated in protection against intracellular bacteria as well as some parasites. Because cytotoxic granules must fuse with the target membrane and release granzymes into the cell to drive destruction, canonical ADCC activity is not compatible with all pathogen membranes. For example, NK cells may have a limited capacity to eliminate highly coated bacteria, but emerging data point to a critical role for NK cells in the destruction of intracellular bacteria once inside monocytes/macrophages. Specifically, both elevated levels of FcR-expressing NK cells (103) and enhanced ADCC-inducing antibodies have been observed in controlled Mycobacterium tuberculosis (Mtb) infection compared to individuals with active, uncontrolled disease (104). Likewise, NK cells have been implicated in the control of parasite infections such as malaria and trypanosomiasis. For Plasmodium, NK cells have been implicated in killing the parasite during the blood stage by attacking infected red blood cells (105) and eliminating the liver stage of the infection via the killing of infected hepatocytes (106). Similarly, ADCC has been implicated in the direct elimination of Trypanosoma cruzi (107), mechanistically pointing to FcRγIIIa-dependent ADCC activity as a key mechanism for the direct elimination of these pathogens.

Antibody-dependent complement activation

In addition to the directed role of innate immune cells, antibodies can also recruit the complement cascade to destroy pathogens. Complement is a complex of soluble blood proteins that can be deposited to mark antibody complexes for destruction. Complement is an evolutionarily conserved process that represents one of the first lines of innate immune defense against many bacteria, yeasts, and viruses. Activation of the complement cascade can lead to direct or indirect destruction of pathogens through (i) the formation of a self-organized membrane attack complex (MAC) on the surface of the pathogen membrane, (ii) directed pathogen uptake, or (iii) the activation of the innate immune system by inflammatory mediators (Fig. 2A).

IgM, followed by IgG3 and IgG1, display the highest potential to activate the classical complement pathway. This pathway is triggered by IgM deposition on a target antigen or the formation of IgG hexamers, either of which allows C1q to bind (63). This initial binding of C1q to surface-bound IgM or IgG results in the autoactivation of C1r and cleavage of C1s. Active C1s cleave C4 and C2, which collectively form the C3 convertase, C4bC2a. The C3 convertase cleaves C3, the major complement protein in serum. Upon cleavage, C3a (an anaphylatoxin) is released, and C3b (the opsonizing agent and ligand for complement receptors) deposits on the surface of the antibody target (14, 108). Complement activation can drive phagocytic uptake or can continue to catalyze the deposition of complement components, leading to the formation of the MAC. The formation of the MAC is blocked on the surface of most human cells by MAC-disassembling mechanisms aimed at preventing non-specific cellular destruction (108). Most pathogens lack these factors, providing an opportunity for the complement to act in pathogen destruction. Interestingly, many enveloped viruses, including HIV, human cytomegalovirus, hepatitis C, and vaccinia virus (62, 109, 110), steal human complement decay activating factor (DAF/CD55) on their viral membranes, allowing them to subvert complement-mediated destruction. However, because C1q aggregates antibodies, complement can enhance viral neutralization by blocking and clustering viral receptors (111) as well as direct viruses to opsonophagocytic uptake.

Complement is essential in the control of multiple bacterial infections, as observed in individuals with complement deficiencies, who experience higher rates of recurrent encapsulated bacterial infections such as Streptococcus pneumoniae, Haemophilus influenzae, and Neisseria meningitidis, especially in early childhood (112). Moreover, complement is key to protection against enteric pathogens including Salmonella typhimurium and Shigella (113) as individuals with high levels of SBA-inducing antibodies exhibited reduced disease.

In the context of viral pathogens, a study conducted in complement-deficient mice (a classical pathway C1q^−/−, a lectin pathway C4^−/−, an alternative pathway factor D (fD)^−/−, and factor B (fB)^−/−) demonstrate that all complement activation pathways are essential for control of severe West Nile Virus disease (114). Similarly, complement-mediated viral lysis was observed against HIV in patients with acute HIV infection (115). Complement also plays a key role in vaccinia virus (VACV) neutralization, highlighting the collaboration between the Fab and the Fc in antibody-mediated protection against VACV (protection confirmed in mouse models of VACV infection) (116). Thus, both direct complement-mediated killing and indirect complement-mediated phagocytic clearance of pathogens are key to protection against pathogens.

PROSPECTS FOR FC PROGRAMMING

Given our mounting appreciation for the role of diverse humoral immune functions in the control and elimination of multiple pathogens, defining novel mechanisms to control and tune both Fab-evolution and Fc-effector function may offer novel opportunities to strategically design more effective vaccines. Distinct delivery platforms and adjuvants have been shown to elicit unique Fc-effector function signatures, even when directed toward the same antigen (117–119). Several approaches exist to tune early inflammatory cues to shape and direct the vaccine-induced humoral immune response (Fig. 2B).

Route of vaccination

The route of vaccination and the initial types of innate immune cells triggered can dramatically alter the quality of the humoral immune response. Specifically, different innate immune APCs exist within different tissues, programmed to drive tissue-appropriate immunity (120, 121). Given that APCs are loaded with diverse sets of pattern recognition receptors expressed in a tissue-specific manner, APCs can rapidly respond to new antigens. This local and APC-specific activation drives downstream inflammatory responses that shape the adaptive immune response (Fig. 2B). For example, vaccines delivered mucosally target mucosal resident APCs poised to elicit mucosal-centric immunity, including the induction of higher levels of IgA (122). Intramuscular (IM) vaccination results in the recruitment of DCs that drive a systemic-centric immune response, largely populated by IgG. The skin is highly enriched in multiple subsets of professional APCs including Langerhans cells, dermal dendritic cells (dDCs), and macrophages, and thus can respond aggressively to novel antigens (123). Along these lines, multiple studies have posited that intradermal delivery of influenza vaccines results in improved immunogenicity and more durable immune responses than the typical IM route (124–127). Also, intravenous vaccination approaches against malaria (128) and Mycobacterium tuberculosis (129) showed strong protection related to the induction of enhanced tissue-resident immunity within the liver and lung, respectively.

The IM route of vaccination can stimulate highly functional IgG responses. This is true for a variety of platforms including mRNA (87, 130, 131), live-attenuated (132), inactivated (133), and component-based vaccines (134). These functions include the FcγR-leveraged functions such as ADCP (both myeloid-lineage and neutrophil), ADNK, and ADCD. Interestingly, stimulation of Fc-effector functions through IM vaccination appears to also be bolstered through localized inflammation induced by adjuvants (134, 135). Intranasal (IN) route of vaccination has also been shown to stimulate a durable humoral response, including antibody effector functions. In a study comparing IM and IN vaccination, ADCP by neutrophils was significantly higher in IN-delivered vaccinations, while ADCP driven by monocytes was similar between IM and IN vaccination groups (136). Thus, the route of delivery can result in unique antibody programming and can elicit more tissue/compartment targeted protection.

Adjuvants

While the route can shape the localization of the immune response, additional inflammatory signals delivered at the time of immunization have been shown to be key to shaping both the Fab and Fc-of the vaccine-induced immune response. Critically, the co-delivery of particular adjuvants, such as synthetic TLR ligands (e.g., TLR4 ligand—monophosphoryl lipid A (MPLA) and TLR7 ligand—Imiquimod) have been demonstrated to drive enhanced affinity maturation and epitope spreading in influenza vaccines in rhesus macaques (137). However, adjuvants have also been shown to drive distinct cytokine and chemokine expression profiles, shaping both T-helper quality and antibody subclass/isotype selection (Fig. 2B and C). For example, the use of alum, the squalene-based oil-in-water emulsion MF59, and other adjuvants in mice drive enhanced functional IgG2a selection compared to standard largely IgG1 selection (138). Poly I:C elicits a highly balanced IgG1, IgG2b, and IgG2c response compared to alum or Addavax (139), pointing to the potential rational selection of adjuvants that give skewed or directed subclass selection in mice. Moreover, a study in non-human primates immunized with HIV envelope glycoprotein gp140 in different adjuvant formulations demonstrated adjuvant-induced differences in antibody glycan profiles, which further correlated with Fc-mediated effector functions (29). Specifically, formulation of the antigen with alum/TLR7 adjuvant resulted in the production of antibodies that strongly bound to the rhesus FcγRs: Rh.RcγRIIa, Rh.FcγRIIb, and Rh.FcγRIIIa and exhibiting high ADCD and ADCP activity (29). Similarly, humans immunized with H5 Influenza HA adjuvanted with MF59 demonstrated higher levels of phagocytic-linked IgG1 and IgG3 via enhanced binding to FcγRIIa, but no enhancement in ADCC linked to poor FcγRIIIa binding (140). The increased IgG3 and phagocytic signal in the absence of ADCC were postulated to occur via altered Fc-glycosylation, pointing to skewed adjuvant B cell programming (140).

More recently, protein-based COVID-19 vaccines and boosters (prefusion spike) have been shown to elicit highly cross-reactive IgG when administered with the AS03 adjuvant (141–143). Moreover, the presence of AS03 in the immunization formulation was required for functional antibody responses. Notably, FcγR-binding antibodies were robustly induced when the prefusion spike was administered with AS03 compared to a formulation without AS03; in fact, a lower dose prefusion spike + AS03 elicited higher FcγR-binding than a higher dose prefusion spike without AS03 (1 µg prefusion spike + AS03 vs 3 µg prefusion spike −AS03). This was also true for non-neutralizing functions such as ADCD, ADCP, ADNP, and ADNKA where the lower dose vaccine with AS03 outperformed the higher dose vaccine without AS03. These non-human primates immunized with AS03 formulations showed strong protection against the SARS-CoV-2 challenge, and antibodies generated could be passively transferred and confer protection (135). While dissecting the protection afforded exclusively by Fcγ- and/or Fcα-binding antibodies can be challenging in studies like these, it is clear that non-neutralizing antibodies whose functions are leveraged through their Fc domain play a substantial role in disease mitigation. The presence of adjuvants appears highly correlated with this.

Adjuvants appear to selectively skew antibody effector function both through selective isotype/subclass selection and altered Fc-glycosylation. Defining the overall impact of all clinically approved adjuvants may offer an opportunity to customize vaccine and adjuvant combinations based on the desired target immune profile. In addition, understanding what Fc-effector functions remain poorly programmable via the currently available adjuvants may point to novel opportunities for the design of novel classes of adjuvants that add value to vaccine development pipelines.

Delivery systems

Finally, beyond the location of administration and adjuvant technologies, accumulating data suggest that the size, stoichiometry, shape, and arrangement of vaccine antigens can also shape the quality of the immune response (144–147). While single antigens can be less immunogenic than whole pathogens, diverse carrier particles including flexibly shaped and sized biodegradable nanoparticles [such as polylactic acid (PLA) or chitosan] (148, 149), protein nanoparticles (150, 151), inorganic nanocarriers (e.g., gold, carbon, and silica) (152–155) or liposomes (156), can mimic pathogens by triggering specific combinations of receptors, arming the immune system in a highly effective manner (Fig. 2B). Moreover, it has been demonstrated that the controlled release of antigen from particles promotes affinity maturation through increased exposure and perpetual selection of B cells (157).

Distinct delivery systems have been linked with different Fc-mediated functions. Recent studies on COVID-19 vaccinations have given us previously unrealized details on this. For example, mRNA-based vaccines can elicit strong FcγR-binding antibody levels and non-neutralizing functions (87), which have been correlated with reduced disease severity (91). Importantly, while antigenic drift can decrease neutralizing antibody recognition (158, 159), FcγR-binding antibodies and non-neutralizing functions can persist in mRNA-vaccinated individuals. Interestingly, NK-signaling antibodies generated by mRNA vaccines showed a stronger subdomain preference than an inactivated-vaccine-generated response (130, 133). This could be due to how the antigen is presented, with mRNA-based COVID-19 vaccines encoding for a stabilized, prefusion spike, whereas an inactivated COVID-19 vaccine can contain spike in numerous conformations (160–163).

Adenovirus-based delivery systems have also demonstrated an ability to induce functional antibodies that were mostly ambivalent to genetic drift of the antigen, particularly in regards to complement deposition and NK-signaling (164). In addition, recombinant viral vectors can leverage viral carrier danger signals at the time of vaccination aimed at shaping the humoral immune responses. Viral vectors such as adenovirus-based platforms have been successfully employed for COVID-19 vaccines. However, population-based immunity toward the adenovirus vector itself needs to be taken into account when utilizing these platforms because neutralization/inactivation of the vector will decrease antigen production and processing. While emerging data clearly illustrate the highly distinct Fc-effector profiles leveraged by distinct viral vector strategies (165), the specific antibody profiles elicited by each vaccine vector remain incompletely understood. It is tempting to speculate that, like adjuvants, viral vectors could provide strategic information on how immunity is trained to most effectively drive a desired target immune profile.

Discreet mechanistic insight into how various delivery platforms/modalities shape Fc-effector function remains largely unexplored. While it is clear that platforms, adjuvants, and dose schedule impact how Fc functions are leveraged, future studies detailing how this process occurs are desperately needed, especially as the link between Fc functions and clinical outcomes becomes clearer.

CONCLUSION

While vaccines have revolutionized our ability to limit many previously highly lethal infectious diseases, as well as recently emerged pathogens, current vaccine design strategies and platforms can clearly be improved. This is of particular importance for pathogens with intrinsically high immune escape potential and those with multi-tissue life cycles. With our growing appreciation for the role of both the Fab and the Fc in mechanistic control against pathogens, next-generation vaccine strategies should leverage both ends of the antibody to provide a comprehensive protection strategy.

The COVID-19 pandemic presented itself at a critical moment in shifting this paradigm. The COVID-19 vaccines were developed, manufactured, and distributed within a stunning timeframe. As the pandemic progressed, genetic drift within the SARS-CoV-2 spike, particularly the receptor binding domain, allowed the virus to evade neutralizing antibodies (158, 159, 166–173). However, protection against disease did not see a corresponding drop, signifying that immune correlates of protection beyond neutralization existed. Fc-mediated functions of antibodies are key determinants of COVID-19 disease severity (85, 86, 91, 174–176). Fully dissecting how various platforms, antigen designs, adjuvants, delivery modalities, and routes of administration affect these is critical. Therefore, to accomplish this goal, next-generation vaccine design will require a holistic understanding of the mechanisms of immune protection, taking into account all immune effector functions and the full potential of the humoral immune response. This will require a commitment to integrate and compare immune profiles across geographic, age, sex, gender, etc. demographics, and identifying signatures that predict outcomes. These commitments will not only help to reduce the disease burden of currently circulating pathogens but also allow for the rapid deployment of highly effective vaccines for emerging infectious diseases.