Viral and host factors in influencing ADE in dengue

DENV has a positive-sense RNA genome that encodes for seven nonstructural proteins and three structural glycoproteins (the capsid shell, envelope (E) and premembrane (prM) proteins) that are responsible for virus attachment, entry, and maturation [6]. Although all serotypes are closely related, a significant degree of sequence diversity exists [4]. As a result, a subset of IgGs that target viral proteins from one serotype may cross-react with those from other serotypes and exhibit poor neutralising function. For instance, the noninfectious immature virions have a “spiky” appearance, because of higher surface exposure of immunodominant epitopes (pr fragment of the prM proteins and fusion loop in the E proteins) [7]. These epitopes are recognised by low levels of cross-reactive IgGs, promoting entry into FcγR-bearing cells that lead to infectious virus maturation and immune suppression that favours viral replication [4,8,9]. Additionally, decreased neutralising antibody concentrations due to waning immunity may also predispose to ADE. Decay in maternal anti-DENV antibody titres, estimated at <1:20 (determined by serial dilution of plasma via ELISA technique), was associated with increased likelihood of severe dengue during an infant’s primary DENV infection [10]. In another paediatric cohort, high anti-DENV antibody levels, >1:320 titres, protected against symptomatic dengue, whereas individuals with anti-DENV antibody titres between 1:21 and 1:80 were more susceptible to severe dengue [2,3]. A mouse model of dengue was used to delineate ADE pathways, and in pups born to DENV-1 immune dams, older pups (>3 weeks old) were more likely to succumb to vascular leakage and death compared to their younger counterparts (2 weeks old) when challenged with DENV-2. In the same study, sera from older pups displayed lower neutralising activities, and their sera demonstrated ADE-mediated increased viremia in vitro [11]. Similarly, in a nonhuman primate model, passive transfer of monoclonal anti-DENV IgG 1A5 at lower concentration (0.22 to 0.67mg/kg) demonstrated increased viremia titre when challenged with DENV-4 [12]. Overall, these observations reinforce the hypothesis that waning immunity may contribute to ADE-mediated outcomes (Fig 1A).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Schematic representation of IgG antibodies and the interaction with DENV.

(A) Anti-DENV IgGs can bind to DENV antigens including PrM and E proteins, which promotes interaction with FcγRs, such as FcγRI (CD64), FcγRII (CD32), and FcγRIII (CD16), expressed by phagocytes. When these antibodies are insufficient to neutralise the DENV, this may then result in enhanced immature virion uptake into the phagocytes (extrinsic pathway). In the phagocytes, the suppression of a pro-inflammatory response and induction of a Th2-type immune response can further enhance viral replication (intrinsic pathway), and subsequently cause excessive cytokine production. Together, this process is termed as antibody-dependent enhancement. On the other hand, in the absence of IgG, DENV is taken into the phagocytes less efficiently through canonical receptor-mediated endocytosis and is unlikely to contribute to ADE. When neutralising anti-DENV IgGs are present at high levels, DENV is fully neutralised and severe pathology can be prevented. (B) IgG is often depicted as a “Y-shaped” molecule consisting of two identical light chains and two identical heavy chains. The antigen-binding site recognises specific antigens, while the Fc region can engage with FcγR expressed by immune cells, leading to activation of effector mechanisms. A posttranslational modification in the Fc region results in the catalysation of a fucose moiety at the asparagine 297 site via FUT8. Afucosylated IgGs lack the fucose structure and have increased binding affinity towards FcγRIII. So far, only afucosylated IgG1s have been reported in dengue patients. The stimuli that trigger IgG1 afucosylation in dengue are not well understood, but DENV surface antigens and microRNAs may play a role in the process by modulating FUT8 expression. ADE, antibody-dependent enhancement; DENV, dengue virus; E, envelope; FcγR, Fcγ receptor; FUT8, fucosyltransferase 8; IgG, immunoglobulin G; PrM, premembrane.

https://doi.org/10.1371/journal.ppat.1011223.g001

Given that FcγRs mediate DENV immune complex uptake, their phenotypic and genetic expression on immune cells may also determine the risk of ADE. FcγRs are surface receptors on lymphoid and myeloid cells, with a unique distribution on each cell type. They can be classified as activating or inhibitory receptors, depending on their intracellular domains. In vitro studies reported that FcγRIIa (CD32a) and FcγRI (CD64) mediate ADE in primary monocytes [13], while FcγRIIa and FcγRIIIa (CD16a) were shown to promote ADE in U937 promonocytic cell line [14]. In contrast, interaction between DENV immune complexes and FcγRIIb, an inhibitory FcγR, prevents phagocytosis and reduces the likelihood ADE in monocytes [15]. In populational studies, genetic polymorphism in FcγRIIa results in a single amino acid alteration (arginine instead of histidine at position 131) and is protective against dengue haemorrhagic fever. This is presumably due to increased binding affinity of their FcγRIIa for IgGs, which subsequently accelerates DENV elimination [16,17]. However, due to limited studies, the association between genetic polymorphism in FcγR genes and the risk of ADE is still not well understood. Lastly, immune cell populations and subpopulations are known to express different levels and types of FcγR, which can change during an infection. For example, FcγRIIa expression can be down-regulated in classical monocytes and up-regulated in intermediate and nonclassical monocytes in dengue [18,19]. Whether or not these changes influence the risk of ADE remains to be determined.

Antigen-specific IgG1s afucosylation and its regulation

In humans, there are four different subclasses of IgGs (IgG1, IgG2, IgG3, and IgG4). Most of the circulating IgGs undergo fucosylation (approximately 94% in adults), a posttranslation modification in the IgG–Fc region where a fucose moiety is attached to the N-linked glycan in the Fc region (Fig 1B) [20]. This creates a steric hindrance and reduces their Fc-FcγRIII binding affinity. In contrast, afucosylated IgGs, which lack the fucose molecule, circulate at low levels under homeostatic condition and have high affinity for FcγRIII. In various infectious diseases, increased levels of antigen-specific afucosylated IgG1s have been reported, and these correlated with different disease severities. For example, afucosylated IgG1s appear to mediate severe disease in enveloped virus infections (COVID-19 and dengue) by promoting immune-mediated pathologies [14,21–23]. While in HIV-1 infected individuals, the presence of antigen-specific afucosylated IgG1 was associated with better control of viral load, suggesting a protective mechanism [24].

The mechanisms and stimuli involved in inducing antigen-specific IgG1 afucosylation in dengue are unknown, but these are likely to be regulated by factors such as exposure to specific DENV antigens, cytokine activities and activation of innate immune signalling receptors. It was proposed that B cell recognition of cell membrane–bound antigens, rather than soluble antigens, in enveloped viral infections, may be one of the stimuli that triggers IgG1 afucosylation. Accordingly, afucosylated antibodies were shown to be specific against E protein of DENV, spike protein of SARS-CoV-2, and envelope glycoprotein of HIV, all of which are antigens associated with host cell membrane [21,23,24]. A recent study hypothesised that the recognition of host cell membrane, in addition to the foreign antigen by B cell, provides a “self” signal to the B cell and may trigger signalling pathways that result in afucosylation; this remains to be proven in future studies [25]. Fucosyltransferase 8 (FUT8) is the enzyme involved in IgG fucosylation. The knockout of FUT8 gene in cell lines led to increased abundance of afucosylated monoclonal antibodies, while the expression of FUT8 gene by plasmablast was reported to inversely correlate with IgG1–Fc afucosylation [26,27]. MicroRNAs (miR) may play a role in the regulation of FUT8 gene, where miR-26a, miR-455-3p, miR-449a, and miR-34a can negatively regulate FUT8 expression in cancer cells (Fig 1B) [28,29]. Both miR-449a and miR-34a were reported to regulate immune responses in DENV infection, but their expression levels in dengue patients and how they may affect IgG1 afucosylation through FUT8 expression in B cells remain elusive [30].

Afucosylated IgG1s in dengue

While having preexisting anti-DENV antibodies during heterotopic secondary infection predisposes to the risk of severe dengue, only a minority (<5%) progresses to severe outcomes [2,3]. A plausible explanation is that ADE is only mediated by antibodies of particular specificity or type, which leads to different levels of susceptibility in patients. For instance, afucosylated IgG1s have increased binding affinity to FcγRIIIa and FcγRIIIb; this may trigger higher uptake of DENV immune complexes and cause extensive inflammation, which are commonly associated with ADE [14,23]. Afucosylated IgG1s from dengue patients recognise the E proteins, and these antibody levels correlated with increased disease severity [21,22]. Furthermore, in early disease phase, higher afucosylated IgG1 abundance was observed in secondary DENV infection compared to primary infection, and these antibodies were predictive of severe dengue [21]. Having ≥10% of afucosylated anti-E IgG1 was reported to be a significant risk factor for thrombocytopenia and blocking of FcγRIIIa protected against platelet reduction in mice [22]. Similarly, the presence of ≥10% of afucosylated maternal anti-E IgG1s predicted symptomatic primary dengue in infants [14].

Potential pathways triggered by afucosylated IgG1s in dengue

The downstream pathways triggered by afucosylated IgG1s in dengue and their roles in ADE are not well understood, but these are likely to be dependent on the type of immune cells that they interact with (Fig 2). For example, natural killer (NK) cells highly express FcγRIIIa, and increased binding affinity of afucosylated IgG1s may enhance antibody-dependent cellular cytotoxicity (ADCC). In dengue, ADCC has been correlated with reduced ADE in vitro, and ADCC promoted IFN-γ expression that was associated with protection against symptomatic dengue in humans [31,32] (Fig 2A). Additionally, compared to fucosylated IgGs, afucosylated antibodies can enhance degranulation and killing of tumour cells by NK cells, as well as promoting quicker detachment of the NK cells from target cells, enabling them to attack more cells [33,34]. Together, it is tempting to speculate that patients who can produce high levels of afucosylated IgGs in early primary DENV infection may be able clear the infection more effectively.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Proposed effector pathways in different immune cells following stimulation with afucosylated IgG1–DENV immune complexes.

(A) NK cells highly express FcγRIIIa (CD16a) and may recognise afucosylated IgG1–DENV immune complex. This may then promote antibody-dependent cell-mediated cytotoxicity, leading to direct killing of DENV-infected cells and DENV, contributing to protection from symptomatic dengue. (B) Afucosylated IgG1s may promote extrinsic ADE in phagocytes, as the absence of fucose molecule greatly enhances the binding of afucosylated IgG1–DENV immune complex to FcγRIIIa expressed by monocytes/macrophages and dendritic cells. This may lead to immature virion uptake, maturation, and replication, leading to increased viremia and inflammation. Neutrophils express FcγRIIIb, and DENV-activated neutrophil can release potent antimicrobial molecules that are associated with severe dengue. However, whether interaction between afucosylated IgG1–DENV immune complex and FcγRIIIb results in ADE and/or release of potent antimicrobial is unknown. Further investigations are needed to provide evidence on afucosylated lgG1s involvement in the proposed pathways above. ADE, antibody-dependent enhancement; DENV, dengue virus; IgG, immunoglobulin G; NK, natural killer.

https://doi.org/10.1371/journal.ppat.1011223.g002

In contrast, interaction between afucosylated IgG1s and FcγRIIIa on monocytes/macrophages/dendritic cells may enhance immature virion uptake, thus promoting ADE-mediated viral replication; this is known as the extrinsic ADE pathway (Fig 2B) [9]. Using in vitro monocytic cell line, afucosylated DENV immune complexes engaged FcγRIIIa to increase immunoreceptor tyrosine-based afucosylated IgG motif (ITAM) signalling that enhanced infection [35]. In vivo studies revealed that afucosylated IgG1–FcγRIIIa interacts with splenic macrophages that promoted viral replication and anti-afucosylated IgG1 nanobodies protected against ADE pathogenicity in mice [36,37]. In SARS-CoV-2 infection, the uptake of afucosylated immune complexes by mononuclear cells was shown to trigger pro-inflammatory cytokine production such as TNF, IL-1β, and IL-6 [23]. It is unclear how afucosylated IgG1 immune complexes can affect cytokine production in DENV infection, but the aforementioned pro-inflammatory mediators have been associated with ADE pathology in dengue [9]. On the contrary, DENV immune complexes have also been demonstrated to suppress pro-inflammatory cytokine production in THP-1 monocytic cell line and to subsequently promote viral replication; this is also known as the intrinsic ADE pathway [38]. FcγRIIIb is highly expressed on neutrophils, and activated neutrophils have been associated with severe dengue [5]. Whether afucosylated IgG1s interact with FcγRIIIb to promote ADE or degranulation in neutrophils is unknown. These observations highlight the complexity in delineating the precise ADE-mediated pathology, but it is likely that fucosylated and afucosylated antibodies may induce different inflammatory responses during DENV infection, and further studies to investigate the roles of these antibodies in ADE are warranted.

Conclusions

In dengue, afucosylated antibodies appear to be correlated with severe outcomes, but the mechanisms by which they contribute to ADE in dengue are unclear. Further studies on the roles of afucosylated IgGs in dengue, and whether their expression differ between naturally exposed and vaccine-induced individuals, are needed to improve our understanding on ADE-mediated pathology in dengue patients.