Relevance of Antibody-Dependent Enhancement in COVID-19

Abstract

Antibody-dependent enhancement (ADE) is a well-established mechanism of pathology in several viral diseases, but its relevance in COVID-19 is not yet recognized. Although several studies in humans have shown an association between antibody responses and disease severity, long term studies addressing the presence of antibodies before infection and their neutralization capacity are needed to establish ADE. Mechanistic studies have determined that the entry of SARS-CoV-2 into host cells can be mediated by immune complexes through Fcγ receptors or by favoring ACE2 conformation. However, the impact on viral replication is not clear. There is evidence for enhancing effects of immune complexes on Fcγ receptor-mediated effector mechanisms and cytokine secretion after modulation of cell signaling in immune cells, specially by antibodies with altered glycosylation, which points to ADE that can contribute to COVID-19 pathology. However, more studies are needed to determine the impact of antibodies both in naturally infected and vaccinated subjects, which can lead to their use as a prognostic marker and increase vaccine safety.

1. Introduction

Antibody-dependent enhancement (ADE) is a mechanism that contributes to pathology in viral infections through a complex interaction between host cells and immune complexes formed by viral antigens and antibodies produced in response to the infection. Ideally, the anti-viral response produces neutralizing antibodies that block viral infection of host cells and leads to clearance of immune complexes without inflammation [1]. In ADE, non-neutralizing antibodies bind to viral antigens promoting viral entry into host cells and/or cell signaling that alters immune cell function, thus contributing to pathology [2,3].

ADE is a well-established contributor in severe dengue that develops after secondary infection with a different dengue virus (DENV) type than the primary infection, where antibodies specific to the first type are non-neutralizing for the new infecting DENV [4]. The facilitation of viral entry by antibodies leads to enhancement of viral replication and viral load, key contributors to the appearance of severe pathology [5,6]. This mechanism has been termed “extrinsic ADE”. Other immune phenomena mediated by altered cell signaling induced by antibody-antigen complexes have been classified as “intrinsic ADE” and result in inhibition of anti-viral defenses or excessive inflammation, both of which can also promote pathology [2,7,8]. In the case of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection, ADE has been recognized, but since it is a new virus, its relevance has not been proved yet [9].

SARS-CoV-2 is the etiological agent of coronavirus disease of 2019 (COVID-19), an illness with a poorly understood pathophysiology. However, excessive inflammation is one of the main components of severe presentations [10]. In these cases, SARS-CoV-2 infection of type II pneumocytes and other cell types expressing the angiotensin converting enzyme 2 (ACE2) receptor leads to a massive secretion of proinflammatory mediators known as “cytokine storm”, which induces recruitment and activation of granulocytes, macrophages and other inflammatory cells [10,11]. The resulting hyperinflammation leads to severe lung injury clinically manifested as acute respiratory distress syndrome (ARDS), which can affect several other organ systems [12,13]. These events happen in only a small subset of infected individuals and the factors that contribute to their appearance remain obscure. Thus, it is imperative to establish the importance of any phenomena that may lead to increased inflammation in the context of SARS-CoV-2 infection, among them ADE. For this reason, the objective of this review is to summarize the latest findings that support a role for ADE in SARS-CoV-2 infection. Three aspects will be discussed: (1) evidence of the impact of the antibody response on COVID-19 pathology both in humans and animal models; (2) immune complex-mediated mechanisms that favor SARS-CoV-2 entry into host cells; and (3) immune complex-mediated mechanisms that generate signals that modulate inflammation.

2. Relationship Between the Antibody Response and COVID-19 Pathology

ADE occurs when antibodies do not neutralize the virus, either because of low concentration, low affinity for the antigen or both [2,8]. However, establishing the relevance of ADE for a particular disease is very challenging because of several variables that influence its occurrence. First, a virus has several antigens, each with numerous epitopes, and therefore the polyclonal antibody response to the virus is a mixture of antibodies with different affinities and specificities, some capable of neutralization and others not. Second, antigenic variation by viruses determines changes in antibody binding, which in some cases leads to sub-neutralizing activity. Third, antibody concentration wanes over time, hampering the neutralizing capacity of the response. For these reasons, to determine if ADE contributes to increased severity in COVID-19, it is necessary not only to establish a temporal relationship between antibody production and severe presentations but also prove that the antibody response has lost its neutralizing capacity because it faces a variation of an antigen, its concentration has diminished or both. In the case of DENV infection, proof of the involvement of ADE in severe disease has been obtained with long-term studies that included big cohorts [14,15]. In contrast, convincing evidence for the relevance of ADE in SARS-CoV-2 infection is yet to be obtained, although several studies both in humans and animal models of the disease exist and are summarized in the following section and in Table 1.

2.1. Antibody Response and COVID-19 Severity in Human Studies

A temporal relation between the start of the antibody response and the development of severe disease in COVID-19 may be an indication that ADE is a contributing factor. Therefore, it is important to determine if antibody production precedes severe pathology analyzing human samples. Several reports that study the relationship between the severity of COVID-19 and antibody levels are summarized in Table 1. To determine COVID-19 severity, most of these studies use the World Health Organization criteria or a similar method based on respiratory pathology and hospitalization requirements. The timing of the samples varies widely (range from day 0 of infection to a year post-infection), but several find a higher level of anti-SARS-CoV-2 IgG antibodies in patients with more severe pathology [16,17,18,19,20,21,22,23]. For example, Hashem et al. [16] investigated the characteristics and dynamics of SARS-CoV-2-specific antibody responses in 87 hospitalized COVID-19 patients, analyzing serum samples spanning 70 days after the start of symptoms. Anti-spike and anti-nucleocapsid (N) IgM and IgG were measured in 240 serum samples of patients with disease classified as mild, moderate or severe. Antibody concentrations had a positive correlation with the severity of disease, as well as the need for ICU and the occurrence of death. The most significant correlation found was that of anti-N antibodies, which were also shown to be most abundant during the first 15 days after the start of symptoms [16]. Siles Alvarado et al. [24] also stratified patients in four severity levels from asymptomatic to critically ill and found that anti-N, anti-RBD and anti-spike antibody titers all correlated to more severity in acute disease (1 month after testing positive for SARS-CoV-2 infection), with the correlation persisting for up to a year with anti-N antibodies [24]. Zhao [23] found a similar dynamic in a group of 173 patients, with anti-SARS-CoV-2 IgG being produced from 15 days post-infection and having significantly higher (approximately four-fold) concentrations at day 39 in critically ill patients in relation to those with mild symptoms [23]. Morgan et al. studied a cohort of 938 patients for up to a year after infection and found higher anti-N and anti-RBD titers in those who needed hospitalization [18]. Finally, de Oliveira et al. found that hospitalized patients showed higher levels of IgG at 90 days than mild cases and that these levels persisted up to 180 days after onset [25].

The conclusion that can be drawn from this group of studies is that antibody production precedes severe presentations in some patients, but no description for mechanisms for ADE are provided. In contrast, a study by Park et al. [26] established that severity of pneumonia, quantified by X ray scores, preceded high antibody titers, indicating that disease severity may be the cause of a stronger antibody response [26]. Additional evidence for the role of ADE has been provided by a meta-analysis by Gan et al. [27], which established that patients that had immunity either by natural infection or vaccination were protected from severe disease. This analysis included 11 studies, but data did not discriminate between SARS-CoV-2 variants and included only studies completed prior to the emergence of the omicron strain [27]. Thus, epidemiological studies specifically designed to address the role of ADE in SARS-CoV-2 infection are needed.

Table 1. Studies that analyze antibody response and COVID-19 severity in humans.

Country	Year	Subjects	Time of Sampling	Main Finding	Reference
Saudi Arabia	2020	87	4 to 70 days	Anti-N Abs correlated to severe disease; nAbs to ICU and death	[16]
China	2020	173	5 to 40 days	Total Abs two weeks after onset associated with disease severity	[23]
Estonia	2022	123	3, 6, 12 months	Severe disease had higher anti-RBD titers after 3 and 6 months	[17]
Canada	2025	938	1, 6 and 12 months	Anti-RBD, anti-N and nAbs higher in hospitalized at one month	[18]
Thailand	2024	110	14 days to 1 year	Anti-RBD levels associated with pneumonia at 14 days	[19]
China	2020	32	0 to 28 days	Higher anti-spike on days 14 and 21	[28]
China	2020	87	4 to 41 days	Anti-RBD higher in severe and moderate cases on days 16 and 25	[20]
China	2021	63	0 to 51 days	Anti-RBD, anti-N and nAbs higher in severe cases; correlated with IL1-β	[22]
Turkey	2023	208	0 to 15 days	Anti-spike and nAbs higher in severe and moderate cases	[21]
México	2022	111	5 to 50 days	Low probability of survival correlated to low neutralization	[29]

Abs: antibodies; nAbs: neutralizing antibodies; ICU: intensive care unit; N: nucleocapsid; RBD: receptor binding domain of spike protein.

Table 1. Studies that analyze antibody response and COVID-19 severity in humans.

Country	Year	Subjects	Time of Sampling	Main Finding	Reference
Saudi Arabia	2020	87	4 to 70 days	Anti-N Abs correlated to severe disease; nAbs to ICU and death	[16]
China	2020	173	5 to 40 days	Total Abs two weeks after onset associated with disease severity	[23]
Estonia	2022	123	3, 6, 12 months	Severe disease had higher anti-RBD titers after 3 and 6 months	[17]
Canada	2025	938	1, 6 and 12 months	Anti-RBD, anti-N and nAbs higher in hospitalized at one month	[18]
Thailand	2024	110	14 days to 1 year	Anti-RBD levels associated with pneumonia at 14 days	[19]
China	2020	32	0 to 28 days	Higher anti-spike on days 14 and 21	[28]
China	2020	87	4 to 41 days	Anti-RBD higher in severe and moderate cases on days 16 and 25	[20]
China	2021	63	0 to 51 days	Anti-RBD, anti-N and nAbs higher in severe cases; correlated with IL1-β	[22]
Turkey	2023	208	0 to 15 days	Anti-spike and nAbs higher in severe and moderate cases	[21]
México	2022	111	5 to 50 days	Low probability of survival correlated to low neutralization	[29]

Abs: antibodies; nAbs: neutralizing antibodies; ICU: intensive care unit; N: nucleocapsid; RBD: receptor binding domain of spike protein.

In regard to the relationship between neutralizing capacity and COVID-19 severity, four studies established that the presence of neutralizing antibodies (nAbs) was correlated to more severity [16,18,21,22], while one concluded that a low degree of neutralization was associated with a fatal outcome [29]. However, the interpretation of these studies is problematic because neutralization was defined by assays that evaluate blocking of the interaction of ACE2 expressed in cell lines or attached to the bottom of an assay plate with recombinant or pseudovirus-expressed spike from one SARS-CoV-2 strain. In some cases, the spike was from the Wuhan Hu-1 original strain [16,18], while one study used a strain isolated from one of the study subjects [21] and other studies do not provide this information [22,29]. Thus, the neutralization capacity of a particular sample is linked to the assay strain, which is not necessarily the one that infected the patients who donated the samples. Interestingly, the study that found low neutralization associated with low survival was the only one that analyzed neutralization degrees in patients infected with different variants. This study, which reported low neutralization in patients infected with delta and omicron variants, used a commercial neutralization assay that does not identify the origin of the RBD sequence. However, the authors concluded that mutations present in the RBD from these variants were the cause of low neutralization, which implies that the assay used a sequence from the original SARS-CoV-2 isolate [29]. This would mean that the antibodies present in these patients probably had a high neutralization capacity for their infecting variants and thus neutralization would be associated with a poorer outcome. Together with the other cited reports, this points to neutralization predisposing to pathology, a finding that would contradict the principle behind ADE. However, the design of these studies does not allow us to establish if the antibodies involved neutralizing the infecting variant or if they were present before infection, important factors for supporting ADE. Thus, prospective studies that establish the neutralization capacity of pre-existing antibodies and then follow individuals that are infected by a variant to which neutralization is low would be needed to establish ADE.

2.2. Antibody Response and SARS-CoV-2 Infection in Animal Models

Animal models of SARS-CoV-2 infection provide a means to test antibodies that are possible therapeutics and thus are concerned with ruling out ADE. Accordingly, Reed et al. found that in a ferret model, the polyclonal anti-SARS-CoV-2 antibody SAB185 did not mediate any clinical worsening of infection or pulmonary pathology in relation to controls [30]. Likewise, Li et al. tested a panel of anti-RBD monoclonal antibodies in a mouse SARS-CoV-2 infection model, evaluating their capacity to either neutralize the virus or enhance disease. They found that these antibodies did not increase viral load, cytokine secretion or lung inflammation in vivo, including one (called DH1041) that showed enhancement of viral entry to cells in vitro [31]. In contrast, one study provides evidence that the antibodies can contribute to pathology in SARS-CoV-2 infection, thus supporting a role for ADE. Matveev et al. evaluated the effect of a highly neutralizing anti-RBD monoclonal antibody called RS2 in a hamster model. Administration of RS2 after SARS-CoV-2 infection resulted in a higher viral load in nasal fluids in relation to hamsters that received a control antibody, suggesting ADE. This contrasted to the effect of the antibody given before infection, which resulted in lower viral load, suggesting neutralization. Importantly, they related this finding with human disease by studying the existence of antibodies directed to the same epitope recognized by RS2 in COVID-19 patients and found that severe disease was associated with higher levels of these antibodies, thus supporting the claim that antibodies that can mediate ADE in vivo are present in severely ill patients [32].

3. Antibody-Dependent Enhancement of SARS-CoV-2 Entry into Host Cells

The main step in the viral life cycle influenced by ADE is entry into host cells. Normally, this process depends on cellular receptors that interact specifically with viral molecules. However, when virus-specific antibodies are present in the moment of infection, the need for interaction with the viral receptor may be bypassed because receptors for the Fc portion of antibodies bound to the virus can mediate entry. Therefore, studying this event is paramount for the understanding of ADE.

The cellular receptor that mediates SARS-CoV-2 entry into host cells is ACE2, which is highly expressed in the surface of type II pneumocytes and other cell types form the respiratory epithelium [33,34]. This receptor binds the receptor binding domain (RBD) of the spike protein located in SARS-CoV-2 envelope. Virus binding induces spike protein cleavage by the enzyme PMPRSS2, which induces fusion of the viral envelope and the cell membrane, allowing viral entry [35]. This process can be inhibited by neutralizing antibodies that bind to the spike protein and block its interaction with ACE2, an effector mechanism that is the hallmark of a protective immune response. Since the generation of antibodies takes several days after the primary infection, neutralization is relevant in secondary infections and is the goal of vaccination strategies. However, when antibodies are non-neutralizing, which may happen when infection is via a different virus variant or antibody concentrations wane through time, the opposite effect may ensue, facilitation of viral entry because of antibody binding. This may be mediated by Fc receptors (FcRs), particularly the ones that recognize IgG (FcγRs), the main neutralizing antibody isotype; but in the case of SARS-CoV-2, antibody binding to the virus may also facilitate the entry through ACE2 by inducing conformational changes. Both these mechanisms will be reviewed in the following sections.

3.1. FcγR-Mediated SARS-CoV-2-Antibody Complex Entry into Host Cells

There is evidence that SARS-CoV-2-specific antibodies can increase its entry into cells through FcγRs [36,37,38,39,40]. Most studies addressing this phenomenon use monoclonal antibodies obtained from naturally infected patients or vaccinated individuals [36,37,38,39] because these types of antibodies recognize only one epitope from the target antigen, resulting in better assay reproducibility. Likewise, most researchers express the spike protein in a pseudovirus that expresses a luciferase reporter which allows quantification of cell entry [36,37,40]. Finally, infected cells are commonly cell lines that can be transfected for FcγR and/or ACE2 expression [36,37,39,40]. Thus, most evidence has resulted from highly artificial systems, although their value lies in the ability to quantify the magnitude of entry and determine the required components.

To determine if SARS-CoV-2 entry can be mediated through FcγRs, Zhou et al. [37] obtained a panel of 48 RBD-specific monoclonal IgG antibodies from a patient infected with SARS-CoV-2 whose sera showed a high degree of neutralization. For the entry assays they used Raji cells, a B lymphoblast cell line that expresses FcγRII. They found that 11 of these antibodies induced viral entry instead of neutralization, both for pseudoviruses and SARS-CoV-2 particles. When the latter were used, no viral replication was observed after viral entry. An analysis of the epitopes recognized by the entry-enhancing antibodies revealed that they bind to a different RBD epitope than neutralizing antibodies. They also determined that entry was FcγR-mediated because a mutation in FcγRII resulted in loss of the effect [37]. In a similar study, Wang et al. [36] tested the capacity of an anti-spike neutralizing monoclonal antibody derived from a convalescent COVID-19 patient to mediate SARS-CoV-2 entry to several cell lines transfected with different FcγRs. They found that immune complexes formed by this antibody and spike-expressing pseudovirus can infect cells that express FcγRIIA, but only at sub-neutralizing concentrations [36]. In another study, Paciello et al. [39] tested the capacity of a panel of 479 monoclonal antibodies obtained from naturally infected and vaccinated subjects to mediate entry of spike-coated pearls into the FcγR-expressing THP-1 monocytic cell line. They found most of the antibodies tested could mediate virus entry, being more efficient those that recognized the N-terminal domain (NTD) of the spike protein, a region different to RBD and thus not implicated in neutralization [39].

The studies described above employed cell lines, but the capacity of antibodies to enhance viral entry has also been shown in primary human cells. For example, Severa et al. [38] tested monoclonal antibodies obtained from convalescent patients and found two capable of mediating SARS-CoV-2 entry into peripheral blood mononuclear cells. In these assays, they showed that no viral replication occurred in these cells and that entry was FcγR-dependent [38]. In a similar fashion, Maemura et al. [40] designed assays with primary macrophages using the polyclonal antibodies contained in plasma from 15 convalescent patients and found that pseudovirus entry was enhanced by antibodies. Their study was complemented by transfecting the BHK cell line with different receptors, which allowed them to support the claim that viral entry mediated by these plasmas was dependent on FcγRIIA and FcγRIIIA [40].

Taken together, these studies allow the conclusion that spike-specific IgG can mediate SARS-CoV-2 entry into FcγR-expressing cells. There is evidence that points to a lack of neutralization capacity as a factor that results in entry enhancement. Additionally, most FcγR-expressing cell types are not permissive for viral replication once SARS-CoV-2 enters, making the possibility of ADE less likely. However, viral proteins have been shown to interfere with both RNA processing [41] and protein trafficking pathways [42], mechanisms that may contribute to inhibit the secretion of anti-viral factors in FcγR-expressing cells after SARS-CoV-2 entry.

3.2. Antibody-Dependent Enhancement of SARS-CoV-2 Entry into ACE2 Expressing Cells Independent of FcγRs

ACE2-mediated entry of SARS-CoV-2 can also be enhanced by antibody binding to the virus without participation of FcγRs through an alteration of spike conformation [43,44]. In this protein, an interaction between the NTD and RBD regions results in a “pre-fusion” conformation that must transition to an “up” position after ACE2 binding for viral entry to take place [45]. Two studies address the effect of antibody binding SARS-CoV-2 in this conformational change and the consequent effect on viral entry. First, Liu et al. [43] tested monoclonal antibodies generated from patients and found that some of the RBD-specific ones could neutralize the virus by blocking ACE2 binding, while others that recognize the NTD region paradoxically enhance virus binding to ACE2 and facilitate entry into host cells. This model, which employed pseudovirus and the HEK293T cell line transfected with ACE2, allowed the researchers to observe a four-fold increase in viral replication after entry. Additionally, assays that evaluated epitope accessibility by using antibodies for different spike epitopes revealed that anti-NTD antibodies that enhance virus entry also favor the “up” position of the spike protein [43]. The other study by Connor et al. [44] used a panel of 1213 monoclonal antibodies derived from 8 patients infected with SARS-CoV-2 and found that 72 of them enhanced entry, most (68%) specifically to the NTD region. The assays used G293B cells transfected with ACE2 that did not express FcRs, indicating that virus entry was FcγR-independent. When the assay used epithelial cell lines that did not express ACE, the effect of antibody binding was not observed [44]. These two studies indicate that the ACE2-mediated entry of SARS-CoV-2 into host cells can also be enhanced by antibodies without FcγR participation, which impacts the quantity of virus that enters natural host cells in the respiratory system capable of sustaining viral replication. Interestingly, Liu et al. also evaluated the amount of entry-enhancing antibodies specific to NTD in the plasma of different groups of COVID-19 patients and found more abundance in severe clinical presentations [43].

3.3. Impact of Antibodies on Virus Fusion

Another important aspect that needs evaluation is the impact of antibodies on the fusion of the SARS-CoV-2 envelope and the host cell membrane, a crucial step in the viral life cycle that follows the binding of spike to ACE2. With this aim, Kibria et al. [46] employed a chemiluminescence-based assay to test the capacity of monoclonal antibodies derived from a convalescent COVID-19 patient to mediate membrane fusion between one set of HEK293T cells transfected with FcγRI and another transfected with spike. They found four anti-RBD neutralizing antibodies capable of augmenting cell fusion and observed the formation of cellular aggregates within 15 min of antibody binding to the proteins expressed in the cells, which was not observed with control antibodies. They also tested sera form convalescent patients and polyclonal IgG from vaccinated individuals, finding the same fusion-enhancing capacity, albeit at a lower magnitude than that induced by the monoclonal antibodies. Using sophisticated methods for the generation of chimeras of antibodies and receptors, they determined that ACE2 was not required for fusion in this system, since cell signaling through FcγRs or a membrane immunoglobulin was sufficient [46]. This study shows that membrane fusion is another vital step in cell entry that can be enhanced by anti-SARS-CoV-2 antibodies.

3.4. Enhancement of SARS-CoV-2 Entry by Antibodies Specific for Other Coronaviruses

Severe acute respiratory syndrome coronavirus (SARS-CoV) and Middle East respiratory syndrome coronavirus (MERS-CoV) are two viruses that belong to the same genera as SARS-CoV-2, Betacoronavirus, and can cause severe respiratory disease in humans. FcγR-mediated entry of both viruses complexed to specific antibodies has been shown, bypassing the need for binding to the specific viral receptor (ACE2 for SARS-CoV and dipeptidyl peptidase 4 for MERS-CoV) [47,48]. Since these viruses share the same antigenic structure as SARS-CoV-2, it is important to investigate whether ADE can arise when antibodies generated against SARS-CoV or MERS-CoV cross-react with SARS-CoV-2 antigens. This is probable because, as described for DENV, antibodies that can neutralize the original antibody-inducing virus (SARS-CoV or MERS-CoV) may bind SARS-CoV-2 antigens with lower affinity and not be able to neutralize virus entry but rather result in ADE. With this hypothesis, Thomas et al. [49] designed assays with FcγRII-transfected BHK cells and SARS-CoV-2 spike-expressing pseudoviruses to test whether the polyclonal antibodies contained in serum from 16 patients infected with MERS-CoV could mediate viral entry. They found that 9 of the 16 sera were effective at enhancing cell entry and showed low neutralizing activity against SARS-CoV-2. None of these sera could mediate entry of pseudovirus expressing spike from MERS-CoV, indicating that ADE was due to a cross-reaction with the SARS-CoV-2 spike protein [49]. Thus, a previous infection with another Betacoronavirus may predispose to ADE.

4. Pathologic Mechanisms in COVID-19 Associated with Immune Complex-Mediated Cell Signaling

In addition to viral entry into host cells, immune complexes formed by SARS-CoV-2 and antibodies can ignite cell signaling cascades that modify immune cell function and modulate inflammation. The main event involved in this process is cross-linking of FcγRs, which can lead to diverse effects that can impact COVID-19 pathophysiology, evidence of which will be summarized below.

4.1. Enhancement of FcγR-Mediated Effector Mechanisms

The normal function of FcγR activation by immune complexes in viral infections is the generation of diverse effector mechanisms that contribute to clear the virus. These include antibody-dependent cellular phagocytosis (ADCP), antibody-dependent cellular cytotoxicity (ADCC) and antibody-dependent complement deposit (ADCD), which are executed by myeloid cells (monocytes, macrophages and dendritic cells), neutrophils and natural killer (NK) cells [50,51]. In the context of SARS-CoV-2 infection, they may be ineffective in viral defense and instead contribute to pathology by damaging tissues and increasing inflammation, thus becoming a component of ADE. Therefore, the finding by Paciello et al. [39] of significant enhancement of ADCP of SARS-CoV-2 spike-coated pearls by monocytic THP-1 cells is relevant. This study also evaluated ADCD by measuring the deposit of complement factor C3 mediated by antibodies obtained from COVID-19 patients on the surface of Expi293F cells that expressed spike. When they analyzed the neutralizing capacity of the antibodies involved, they found an inverse relation to ADCP and ADCD: anti-RBD antibodies that were highly neutralizing were poor at mediating FcγR-dependent effector mechanisms, while anti-NTD antibodies that were poorly neutralizing enhanced both ADCP and ADCD [39].

The activation of the classical complement pathway by immunocomplexes formed between polyclonal antibodies contained in plasma of convalescent COVID-19 patients and SARS-CoV-2 antigens was evaluated by Jarlhelt et al. [52] with an ELISA assay that detected C3 and C4 deposition and membrane attack complex formation. The results indicated that these immune complexes were able to activate the complement system and, importantly, showed a correlation of this activity with COVID-19 clinical presentation, being significantly higher in plasma from severely affected patients [52]. Zedan et al. [53] studied ADCC using plasma from SARS-CoV-2 infected subjects or vaccinated individuals in assays where FcγRIIIA-expressing Jurkat cells directed their cytotoxicity towards immobilized complexes formed between the antibodies contained in the plasmas and spike protein. They determined that both infection and vaccination induced ADCC-mediating spike IgG, with the highest activity shown by plasma from people who received mRNA vaccines. Additionally, they observed that ADCC was more efficient with plasma from symptomatic patients in relation to infected asymptomatic subjects and showed lower levels in patients infected with the omicron variant of SARS-CoV-2 [53]. Fernandez-Soto et al. [54] also evaluated ADCC in the context of COVID-19, testing antibodies specific for SARS-CoV-2 M protein obtained from patients with different degrees of severity. The ADCC assays revealed that immune complexes formed with antibodies from COVID-19 patients induced more NK cell degranulation than those from healthy controls, and those from critically ill patients also showed a higher activity than those from ambulatory patients [54].

The conclusion derived from this group of studies is that anti-SARS-CoV-2 antibodies enhance ADCP, ADCD and ADCC by binding FcγRs, mechanisms that may contribute to pathology.

4.2. Cytokine Secretion Induced by Immune Complexes

The role of excessive proinflammatory cytokine secretion in severe COVID-19 has been recognized [55]. However, the cause of this phenomenon is not clear, and it is likely that multiple factors that vary from patient to patient contribute. In this context, different manners in which the regulation of leukocytes that constitute the primary source of these cytokines may be altered should be analyzed, including signaling induced after FcγR-crosslinking by immune complexes.

To explore this aspect, Severa et al. [38] obtained anti-spike neutralizing monoclonal antibodies (nAbs) from COVID-19 convalescent patients and tested their capacity to induce cytokine secretion from plasmacytoid dendritic cells (pDCs) after immune complex formation. They found that complexes formed SARS-CoV-2 and nAbs induced significantly higher interferon-γ (IFN-γ), interleukin (IL)-6, tumor necrosis factor-α (TNF-α) and IL-8 secretion than the virus alone, which was associated with the generation of a pDC inflammatory phenotype. Furthermore, CD14⁺ monocytes were isolated from the peripheral blood of these patients and after stimulation with the immune complexes they converted to a CD86-expressing inflammatory phenotype that secreted high amounts of IL-6, TNF-α and IL-8 [38]. Jarlhelt et al. [52] also evaluated cytokine secretion by stimulating blood monocytes purified from healthy donors and THP-1 monocytic cells with immune complexes formed with polyclonal antibodies from convalescent COVID-19 patients and recombinant RBD. They found TNF-α secretion was significantly enhanced in both cell types in relation to controls, in a magnitude comparable to immobilized lipopolysaccharide, a well-established inflammatory stimulus. However, immune complexes did not induce IL-6 or IL-1β secretion [52].

CCL3 (also known as MIP-1α) is a relevant inflammatory chemokine that was evaluated by Fernández-Soto et al. [54] in NK cell cultures stimulated with immune complexes formed between SARS-CoV-2 and polyclonal anti-protein M antibodies purified from plasma of COVID-19 patients. The results indicated that the chemokine was more abundant in cultures stimulated with the immune complex, and that antibodies from critically ill patients had the most notable effect [54]. Likewise, Larsen et al. [56] evaluated IL-6 and C reactive protein (an inflammatory marker) secretion induced by complexes formed with anti-spike and anti-M antibodies obtained from SARS-CoV-2-infected individuals. They used macrophages differentiated in vitro from peripheral blood monocytes. These cells were high IL-6 secretors after stimulation with immune complexes and a ligand for toll-like receptor (TLR)-3, a receptor for viral RNA. They also determined that both IL-6 and C reactive protein were more abundant when antibodies from severe COVID-19 presentations were used and when tested early in the disease process (approximately 20 days post-infection) in relation to later time points [56]. One final study regarding cytokine secretion by Hoepel et al. [57] compared the effects of anti-spike antibody-positive and negative sera obtained from COVID-19 patients that required admission to the intensive care unit with antibodies from sera of non-infected individuals. They found that the inflammatory response induced in primary macrophages by anti-spike IgG-positive sera from COVID-19 patients was higher than that from anti-spike-negative patients or non-infected donors [57].

Taken together, the studies that assess cytokine secretion induction by immune complexes in COVID-19 indicate that this mechanism is induced and thus can contribute significantly to the inflammatory milieu in this pathology. However, it is worth noting that the study by Maemura et al. did not find cytokine secretion after stimulation of primary macrophages with immune complexes formed with antibodies from convalescent COVID-19 patients [40]. This could be explained by the fact that this study did not recruit patients with severe COVID-19, a factor that was associated with higher cytokine secretion induction as described above for other studies.

4.3. Impact of Antibody Glycosylation in Induction of Inflammation in COVID-19

N-glycosylation of the IgG1 heavy chain, specifically in the asparagine residue in position 297, is an important factor in the activation of leukocytes by immune complexes [58]. Different carbohydrates can glycosylate this amino acid residue, which confers different FcγR-activating properties. It has been shown that the absence of fucose (afucosylated IgG1) increases affinity for FcγRIIIA, leading to excessive activation and enhancement of inflammatory effector mechanisms [58]. For this reason, several researchers have addressed the degree and effect of IgG glycosylation in the context of SARS-CoV-2 infection.

Siekman et al. [59] analyzed IgG1 glycosylation profiles in COVID-19 patients and found that anti-spike IgG1 had lower fucosylation than total IgG1 and that it occurred more in patients that required hospitalization in relation to those that did not [59]. Larsen et al. [56] obtained similar results, finding significantly lower anti-spike IgG fucosylation in COVID-19 patients with ARDS who required admission to the intensive care unit compared with plasma donors who had mild disease or were asymptomatic, suggesting that low IgG1 fucosylation may be associated to hyperinflammation in COVID-19 [56].

Chakraborty et al. [60] also found higher production of afucosylated IgG1 antibodies in severe COVID-19 and thus aimed to explore the effect of immune complexes formed with these antibodies on cytokine secretion after stimulation of FcγRIIIA. They determined the affinity of antibodies purified from severe COVID-19 patients for FcγRIIIA and found that IgG with more than 20% of afucosylation had an affinity as much as three times higher than fucosylated IgG. In vitro stimulation of monocytes obtained from healthy donors showed that these antibodies with higher affinity for FcγRIIIA induced a higher secretion of IL-6 and TNF-α [60]. To expand these findings, this group studied IgG fucosylation in different stages of COVID-19 and found afucosylated antibodies in early stages of the disease in patients who later progressed to severe pathology. They tested plasma from patients with afucosylated anti-spike IgG in humanized mice that expressed human FcγRIIIA, revealing that pulmonary infiltration by inflammatory cells (indicated by neutrophil and monocyte count in alveolar fluid) was higher than that induced by plasma with fucosylated IgG. Furthermore, they found higher levels of TNF-α, IL-6 and the chemokines CXCL1, CCL3 and CCL4 [61]. Hoepel et al. [57] also found a correlation between IgG1 afucosylation and cytokine secretion in polarized M2 macrophages. They used anti-spike antibodies obtained from critically ill COVID-19 patients to form immune complexes with the spike protein and stimulate the macrophages, finding upregulation of IL-8, IL-6, IL-1β and TNF-α secretion. In addition, they tested a monoclonal anti-spike antibody obtained from a patient with and without fucosylation, confirming the association of lack of fucose and cytokine secretion [57].

The cited studies indicate that low fucosylation of anti-SARS-CoV-2 IgG is a potential contributing factor to hyperinflammation in COVID-19. It is therefore relevant to evaluate the level of IgG fucosylation in vaccinated individuals. The study by Chakraborty et al. found high fucosylation in subjects vaccinated with the BNT162b2 mRNA vaccine [61]. Additionally, Van Coillie et al. [62] measured fucosylation in subjects that received this vaccine, dividing them in two groups: those naïve for SARS-CoV-2 and those who were seropositive before vaccination. They found transitory afucosylation in the first group and high levels of afucosylation in the latter, but immune complexes from neither group were able to induce inflammatory cytokine secretion from macrophages, indicating that the amount of afucosylated IgG was not sufficient to activate FcγRs [62]. These results suggest that the BNT162b2 vaccine is not associated with the risk of excessive inflammation due to low antibody fucosylation.

5. Relevance of ADE in Vaccination Strategies

The relevance of ADE in vaccination is paramount because vaccinated individuals are healthy and thus the possibility of enhancing the pathology that antibodies are meant to protect from must be eliminated. In the previous sections, several studies that show ADE have included antibodies from samples of vaccinated individuals [39,46,53,62]. In addition, Shimizu et al. [63] tested the capacity of sera from vaccinated individuals to mediate ADE by culturing FcR-expressing myeloid cells with the SARS-CoV-2 omicron strain complexed with antibodies from individuals who had received two doses of the Moderna mRNA vaccine. They reported a ten-fold increase in viral replication in relation to pre-vaccination sera, highlighting the potential for ADE. Since antibodies generated by vaccination and natural infection have the same structure and function, this finding is not surprising. It is worth noting that the neutralization capacity of the antibodies tested in this study for the omicron variant was low [63], raising the possibility that vaccination may enhance disease caused by SARS-CoV-2 variants that have an antigenic difference from the vaccine protein. However, large clinical studies that address specifically the impact of ADE in vaccinated individuals are not yet available, although they are warranted because of the need to increase vaccine safety.

6. Conclusions

Although the current clinical and epidemiological studies do not allow us to conclude that ADE is a relevant phenomenon in SARS-CoV-2 infection, several in vitro studies support mechanisms for both intrinsic and extrinsic ADE, as depicted in Figure 1.

The body of evidence for ADE in COVID-19 infection summarized in this review warrants its consideration as an important pathology-inducing mechanism. However, there is a need to define several key elements to establish its relevance. First, long-term studies aimed at defining the effect of antibodies formed prior to infection, whether through a primary infection or vaccination, as well as their neutralizing capacity for the SARS-CoV-2 variant involved, should be designed. Second, efforts should be made to determine if the effect of immune complexes on viral entry to host cells results in higher viral loads in vivo, either by inducing replication in FcγR-expressing cells or enhancing the normal mechanism mediated by ACE2. Finally, studies pursuing the application of afucosylated IgG as a prognostic marker may be beneficial, since numerous studies establish their occurrence with disease severity.