Mark T. Xiao1 · Calder R. Ellsworth· Xuebin Qin,: 3 February 2024
Abstract
The complement system, a key component of innate immunity, provides the first line of defense against bacterial infection; however, the COVID-19 pandemic has revealed that it may also engender severe complications in the context of viral respiratory disease. Here, we review the mechanisms of complement activation and regulation and explore their roles in both protecting against infection and exacerbating disease. We discuss emerging evidence related to complement-targeted therapeutics in COVID-19 and compare the role of the complement in other respiratory viral diseases like influenza and respiratory syncytial virus. We review recent mechanistic studies and animal models that can be used for further investigation. Novel knockout studies are proposed to better understand the nuances of the activation of the complement system in respiratory viral diseases.
Introduction
The complement system, a key component of innate immunity, provides the first line of defense against microbial invasion. It also participates in the pathogenesis of many chronic, non-infectious human disorders, including autoimmune diseases, hyper acute graft rejection, paroxysmal nocturnal hemoglobinuria (PNH), atypical hemolytic uremic syndrome (aHUS), and atherosclerosis [1, 2]. Since its first identification in 1895 [3], our understanding of the complement system has led to the development of successful complement-targeted therapeutics used to treat diseases ranging from PNH/aHUS to the SARS-CoV-2 infection (COVID-19)]. It is well established that bacterial invasion activates the complement system, forming activated bioproducts and the terminal membrane attack complex (MAC), a terminal complement activation product, thereby culminating in lysis and pathogen clearance [5]. However, it is less clear how viral infections such as SARS-CoV-2 activate the complement system. Like in SARS-CoV-2, the complement system appears to be maladaptive in respiratory syncytial virus (RSV) infection, which can similarly cause acute respiratory distress syndrome (ARDS), a life-threatening lung injury that allows fluid to leak into the lungs [6, 7]. However, in other respiratory viral infections that can cause ARDS, such as influenza A & B, the role of the complement system is complicated. C3 may be protective while C3a and C5a signaling and MAC may be detrimental to the host [8, 9]. The cellular and molecular mechanisms by which the components of the complement system contribute to the pathogenesis of these respiratory viral infections remain unclear. We will review (1) complement activation and regulation, (2) clinical evidence of the detrimental role of the complement in COVID-19, (3) beneficial effect of complement-related therapeutics on COVID-19, 4) the pathogenic roles of the complement in other respiratory viral infections including influenza and RSV, and 5) experimental approaches to dissect the complement-mediated mechanisms during acute respiratory viral infection.
Complement activation and regulation The complement system is an essential component of the innate immune system. It has been evolutionarily preserved for hundreds of millions of years and comprises roughly 30 membrane-bound and soluble proteins (Fig. 1) [10, 11]. The complement system is activated by three distinct pathways that occur both on pathogenic surfaces and in plasma (Fig. 1) [1, 12]. The classical pathway is primarily triggered by antigen-bound antibodies. Specifically, and the Fc regions of these activated antibodies (primarily IgM and IgG) bind to C1q, which initiates the complement classical pathway [13–15]. Human IgG3 and IgG1 bind and activate C1 readily, whereas IgG2 does so poorly, and IgG4 exhibits no activity [14–16]. The alternative pathway can be stimulated by attachment of C3b, a cleavage product of complement component 3 (C3), to foreign particles and damaged tissue, or by spontaneous cleavage of C3. The mannose-binding lectin (MBL) pathway is initiated when the plasma MBL protein complexes with MBL-associated serine protease-1 and -2 (MASP-1/2), which in turn binds to microbial surface oligosaccharides and acetylated residues (Fig. 1) [17]. The three extracellular pathways (classical, alternative, and MBL) coalesce via the activation of C3 convertase and the cleavage of C3 to C3a and C3b [1, 12, 18]. This engenders a cascade of cleavage and activation events, including the cleavage of C5 to C5a and C5b via a C3b complex that culminates in the formation of the membrane attack complex (MAC; Fig. 1) [1, 12, 18]. MAC formation begins with the sequential recruitment of C6, C7, and C8 to C5b. The MAC embeds itself in the phospholipid bilayer, and C8 induces the polymerization of C9 molecules to form a pore-like structure. The MAC is a macromolecular pore capable of inserting itself into cell membranes and lysing foreign pathogens, and heterologous cells. Under certain pathological conditions such as PNH, the loss of complement regulators on erythrocytes leads to lytic MAC formation which can cause hemolysis [12, 19–21]. Additionally, the formation of MAC at sublytic concentrations in a cell membrane of nuclear cells such as monocytes and endothelial cells can stimulate signaling cascades [22–30] that lead to the activation of monocytes and mediate infammation on blood vessel without the lysis of the cells [2, 31–34]. In addition to complement activation on cell surfaces and in serum, recent evidence points towards intracellular complement activation via local production, endocytosis, and phagocytosis [35, 36]. Classical Pathway Lectin Pathway Alternative Pathway Antibody binds antigens Lectins bind sugars Pathogens or damaged tissue C1r/C1s + C1q + Activated antibody Pathogen Mannose + MLB/FCN + MASP C4, C2 C4b2b (C3 Convertase) C4a, C2a, C4d C3 C4b2b3b (C5 Convertase) Pathogen LPS C3b C3 Fasctor B Factor D C3bBb (C3 Convertase) C3 C3bBb3b (C5 Convertase) C5b C5, C3 C5 C3a C3a C6, 7, 8, 9 C5a C3a C3a C5a Anaphylatoxins & MAC Regulators CD46 FI CR1 CD55 CD59 FI FH CD55 C1INH Therapeutics Eculizumab Tesidolumab Ravulizumab Vilobelimab AMY-101 Berinert Conestat Alfa Narsoplimab C5b-9 (MAC) FP C3 and C5 cleavage has been observed in various immune cells, including lymphocytes, monocytes, and neutrophils, and may also have a signaling function [37].
To prevent excessive complement-mediated damage, several plasma and membrane-bound protein regulators have evolved to attenuate and restrict the complement system at different stages of activation (Fig. 1) [1, 12, 18]. Plasma factor I (fI) controls the production of active C3b by cleaving C3b into inactive iC3b and C3d [1, 12, 18] (Fig. 1).
Factor H, also soluble, regulates the alternative pathway by accelerating the decay of C3 convertase and is a cofactor for factor-I-mediated inactivation of C3b [38]. The soluble C1 inhibitor regulates the classical pathway upstream [1, 12, 18] (Fig. 1). Membrane-bound protein regulators consist of complement receptor 1 (CR1), membrane cofactor protein (MCP) CD46, decay-accelerated factor (DAF) CD55, and CD59 (Fig. 1). These regulators are expressed on host cell membranes and protect the host from complement attack by inhibiting upstream complement convertases, deactivating complement products, and restricting the formation of MAC by inhibiting complement pathway activation at varying levels of the cascade [1, 12, 18]. For example, CD59 inhibits the formation of the MAC by directly binding to C8 and C9 and preventing the polymerization of C9, while CD55 accelerates the decay of C3 and C5 convertases [1, 11, 12, 18]. These regulators mediate a delicate balance between adaptive and toxic immune responses.
Proper regulation of the complement cascade is essential for a healthy immune response. Over- or under activation is associated with disease pathologies; for example, although PNH and aHUS are distinct in terms of clinical manifestation, they both feature a mechanistic over activation of the complement system via various malfunctioning complement regulators including CD55, CD59, Factor H, and Factor I [39, 40]. Figure 2 details examples of SARSCoV-2-mediates complement activation and tissue damage, using various images available online under creative commons use licenses [41–45]. On the other hand, some tumors may develop resistance to complement-dependent cytotoxic (CDC) chemotherapy via the overexpression of CD59 (and consequently, the inhibition of MAC) [46, 47]. CD59 also plays a large role in virulence: Increased incorporation of CD59 into viral envelopes can protect against antibody dependent complement-mediated lysis in HIV-1, cytomegalovirus, and herpes, among other pathogenic viruses [48–53].
Respiratory viruses such as SARS-CoV-2, influenza, and RSV all show evidence of heightened activation of the complement system upon infection, but the consequences of this C3b ACE2 Bb C3b Bb ACE2 mediated endocytosis/viral entry C1q S and N protein binding. Classical pathway acvaon Platelet Acvaon
Endothelial damage (Thrombosis, MAC, complement deposition) Hemolysis Relapse of PNH and TMA Recruitment of monocytes and neutrophils ARDS
Tissue damage (MAC formation, inflammation, necrosis) Complement byproducts in ssue/BALF (MAC, C5a, C3c) Acute kidney injury
Complement byproducts in renal/glomerular arteries (MAC, C3) Hemolysis Relapse of aHUS Organ damage/dysfunction Complement deposition (MAC) Inhibition of C3bBb convertase disassociation N Protein S Protein
Fig. 2 SARS-CoV-2 mediates complement activation and tissue damage and causes the relapse of complement dysregulation-related diseases: SARS-CoV-2 cellular invasion is mediated by ACE2 receptor binding. Beyond viral mediated cell lysis, SARS-CoV-2 can cause damage to tissues in the lungs, endothelia, kidneys, and other organs via the direct activation (and the inhibition of regulation) of the complement system. SARS-CoV-2 can also mediate the relapse of complement dysregulation-mediated diseases such as atypical hemolytic uremic syndrome (aHUS), Complement-mediated thrombotic microangiopathy (TMA) and paroxysmal Nocturnal Hemoglobinuria (PNH) (highlighted in red). BALF bronchiolar Lavage Fluid, ARDS Acute Respiratory Distress Syndrome. S Protein Spike Protein, MAC Membrane Attack Complex, N Protein Nucleocapsid Protein. Images used were imported from online sources under creative commons use
licenses or generated in Microsoft PowerPoint. 94 Page 4 of 18 M. T. Xiao et al. activation vary dramatically (as reviewed below). The mechanisms underlying these distinct host responses to specific infections are unclear and warrant further investigation.
Increased complement activation is associated with severe COVID‑19
COVID-19-related morbidity and mortality are increasingly thought to be related to excessive immune response [54]. While the pathogenesis of severe COVID-19 varies, key manifestations include platelet activation, thrombosis, endothelial dysfunction, immune activation, and cytokine storm [55–58]. Proteomic analysis from tissue samples obtained early in the pandemic identified complement over activation as one of the strongest indicators of COVID-19 infection. For instance, hospitalized COVID-19 patients have significantly higher levels of circulating sC5b-9 than similarly sick influenza patients and C5a and alternative pathway markers are associated with increased COVID-19 severity [59]. In addition to the alternative pathway, which is predominant earlier in the course of diseases, clinical evidence points to complement activation across the MBL and classical pathways as well [60].
While evidence increasingly points to COVID-19 being a systemic disease [61, 62], pathogenesis of acute COVID-19 infection begins in and most severely impacts the lungs via respiratory failure and inflammation. Early clinical observations indicate heightened complement deposition in various tissue, including C3 and C5b-9 on lung endothelium and C5a in bronchoalveolar lavage fluid (BALF) [63, 64]. Heightened complement activity may also play a critical role in thromboinfammation in severe COVID-19 patients via the platelet/neutrophil extracellular traps (NETs)/thrombin axis [56] (Fig. 2). Coagulation and inflammation drive ARDS and contribute to the severity of the disease, and ARDS has long been associated with complement activation [65–67].
Some evidence suggests that complement activity may contribute to COVID-19 severity in a dose-dependent manner, with higher concentrations of sC5b-9 (solubilized MAC), C5a, and C3 byproducts in particular being associated with respiratory failure [57, 61, 63, 68–70] (Fig. 2).
The kidneys follow as the second most commonly affected organ system in COVID-19 patients, with an estimated 25% of patients hospitalized due to COVID-19 developing acute kidney injury [71, 72] (Fig. 2). Again, biopsies reveal that these patients had similar/elevated levels of complement activation products in renal and glomerular arteries relative to those with other complement-related kidney pathologies [73]. In fact, immunofluorescence staining of samples from nine patients with fatal COVID-19-related acute respiratory failure revealed that C5b-9 deposition patterns in the kidneys were remarkably similar to those in the lungs [64](Fig. 2).
The similarity in COVID-19-related complement pathologies across organs suggests systemic complement over activation.
The complement system has been implicated in other COVID19-mediated organ damage as well. For instance, deposits of various complement proteins have been observed in the livers, hearts, and systemic and cutaneous vasculature of infected patients [62, 64, 74–76] (Fig. 2). Together, extensive clinical evidence indicates that upon SARS-CoV2 infection, all three complement pathways can be activated and participate in the pathogenesis of severe COVID-19. SARS-CoV-2. COVID‑19 induces the relapse of complement‑related human diseases
As discussed previously, PNH and aHUS both feature a similar pathological under-regulation of the complement system via the altered expression or function of various complement cascade proteins (Fig. 2). Patients with preexisting irregularities are particularly sensitive to any level of complement over activation. In one report, a 28-year-old female with previously diagnosed aHUS due to a heterozygous missense variant of the C3 gene suffered a relapse brought on by COVID-19 infection [77]. Other patients have been in similar situations, with COVID-19 and preexisting complement dysfunctions such as nonsense Factor H, CD46, and PIG-A mutations, and gain of function C3 mutations contributing to acute thrombotic microangiopathy (TMA) [78–81]. Utilization of eculizumab, a clinically used complement therapeutics to block C5a and MAC formation to suppressing complement activity in those patients showed immediate beneficent effects [77–81]. PNH and aHUS can be triggered by complement activation engendered by S protein antigens, and asymptomatic COVID-19 has also been reported to trigger multisystem inflammatory syndrome (MIS-C) in children (MIS-C), a rare, life-threatening complication of COVID-19 [82]. Decrease in complement activation is closely associated with rapid improvement of mMIS-C after intravenous immunoglobulin treatment [82].
These cases further highlight the sensitivity of the complement system and how COVID-19 can disrupt the delicate balance between regulation and activation, thereby accelerating complement dysregulation-related diseases (Fig. 2).
However, our understanding of the complement pathway activation on SARS-CoV-2 infection is limited and warrants further investigation. Therefore, in the following section, we will review the current understanding of complement pathway activation in COVID-19 and promising areas of investigation for future studies.
Three pathways of complement activation in COVID‑19
The general pathophysiology of acute severe COVID-19 consists of two successive phases: an initial viral invasion
Emerging role of complement in COVID-19 and other respiratory virus diseases Page 5 of 18 94 followed by an aberrant immunopathological response [7].
Viral invasion is facilitated by binding of the viral spike (S) protein to angiotensin-converting enzyme 2 (ACE2) cellular receptors [83] (Fig. 2). ACE2 is expressed in many tissue types, particularly in the lungs, heart, kidneys, brain, and endothelia [83]. This may account for the tropism in infection previously discussed, and the wide variety of corresponding symptoms that may occur. Tissue damage can stem from not only productive viral infection but also complement activation due to adjected infected tissue (particularly in the case of microvascular endothelial cells) [84].
Classical pathway activation in COVID‑19
The role of the classical pathway in severe COVID-19 is less investigated relative to the lectin and alternative pathways.
While the classical pathway is associated with the adaptive immune response, it is notably not entirely antibody-dependent. For example, the classical pathway components C1q can bind directly to apoptotic blebs to initiate the complement cascade in bacterial infections [85, 86]. This seems tobe consistent for SARS-CoV-2 as well, as anti-virus IgGs,
S1, and N proteins bind to C1q and gC1qR [61, 70]. Additionally, complement activity assays using normal and C1qdepleted human serum have suggested viral protein and C1q binding in a dose-dependent manner [70]. Direct pathogen binding and complement reactive protein are both plausible factors for the activation of C1. Recent evidence has also highlighted a potential protective role of complement components C1q and C4b-bp. C1q produced by alveolar type II cells and local macrophages, independent of further complement activation, are known to modulate influenza A virus infection and replication, and ACE2 expressing cells infected with SARS-CoV-2 treated with C1q and C4b-bp were found to have reduced levels of inflammatory cytokines [87, 88]. This suggests that local, non-activating classical pathway activity may be beneficial to host defense. A deeper investigation using classical pathway protein knockouts is necessary to better understand how the classical pathway impacts the pathogenesis of severe COVID-19.
Lectin pathway in COVID‑19
The lectin pathway is clearly implicated in the pathogenesis of severe COVID-19. The concentration of lectin pathway s pecifc products, such as the MASP-1/C1-INH complex, have been correlated with COVID-19 severity [89]. Binding assays have shown that the SARS-CoV-2 N protein can interact with MASP-2, and complement deposition assays have linked N protein levels to activated C3 in a calcium dependent manner [90] (Fig. 2). However, a separate cell study found that varying N protein treatment levels had no impact on cell viability, suggesting the possibility of sub-lytic complement activation via the lectin pathway [91]. MASP has also been implicated in animal studies of COVID-19. In one study, transgenic human ACE2-K18 mice susceptible to SARS-CoV-2 had improved symptomatic and phenotypic outcomes when treated with rescue HG4, an anti-MASP-2 antibody [92]. On the other hand, the role of the MBL protein is less clear. In a recent study from 2023, certain MBL polymorphisms (and lower functional MBL levels in general) were associated with a higher inflammatory response and disease severity [93]. This may be because MBL can suppress TNF-α and IFN-γ in natural killer cells and attenuate the acute immune response.
Other MBL polymorphisms showed no correlation with any disease outcomes [89, 93]. Once again, the complement system’s activation-independent functions seem to play an adaptive role, as opposed to downstream complement activation efects. The net efects of complement components are not well understood; further investigation via single knockout animal models is warranted.
Alternative pathway in COVID‑19
The alternative pathway is continuously active at low levels and can amplify the response initiated by any pathway, providing a rapid response to pathogens. However, the rapid amplification of the complement cascade can be detrimental in certain disease contexts; the alternative pathway is disproportionately implicated in complement dysregulation-related diseases such as aHUS and PNH, among others [39, 94].
Evidence highlights that the alternative pathway is similarly implicated in COVID-19 complications; while complement products were generally upregulated among all patients, alternative pathway-specific biomarkers (such as increased byproducts including Ba/Bb and iC3b) and decreased alternative pathway regulator (properdin) in admission samples were correlated with increased severity and risk of mortality [60, 95]. Alternative pathway-specific SNPs, including factor B SNPs, have similarly been correlated with severity [96]. Clinical samples have highlighted a surge of factor D during the early phase severe COVID-19; in the human cell and macaque COVID-19 models, anti-factor D treatment mitigated complement activation and protected endothelial cells post-viral infection [97]. Enhanced C3 cleavage and downstream release of C3a and C5a may contribute to the severe COVID-19 phenotype by promoting myeloid cell infiltration and propagating inflammation.
These emerging clinical studies clearly indicate that SARS-CoV-2 infection activates three complement pathways. Interestingly, a recent study shows that SARSCoV-2–encoded ORF8 protein may contribute to the decay of C3-convertase; this finding may help explain how the virus escapes elimination via the complement system and further highlights the gap between local and systemic 94 Page 6 of 18 M. T. Xiao et al. complement responses [98]. SARS-CoV-2 may develop other mechanisms to inhibit the host complement pathway activation for escaping the complement attack, which requires further investigation. The relevance of local versus systemic complement response has been documented in bacterial pneumonia via a conditional complement knockout mouse model. Although the liver is primarily responsible for complement biosynthesis, complement production by lung epithelia relevant [99]. Using a tamoxifen-inducible Cre-Lox system, researchers selectively ablated the gene encoding C3 from lung epithelia and hepatic cells. Lung C3 knockout mice experienced greater acute lung injury caused by Pseudomonas aeruginosa infection than controls while maintaining near-normal levels of circulating C3. Conversely, mice deficient in liver-generated C3 had similar levels of the injury as wild-type controls [99]. Single-cell transcriptome analysis of the lungs of patients with COVID-19 has revealed that C3 is upregulated in various lung epithelial cells, including AT1, AT2, and club cells [100]. However, the distinction between local and systemic complement production and how they contribute to the hyper-immune activation seen in severe COVID-19 pathogenesis is unclear and also warrants further investigation. Regardless of the source, complement activation contributes to the overexpression of pro-inflammatory cytokines, including IL-6 and TNF-α, and worsening inflammation, organ damage, and general disease severity [56, 63, 101]. Beneficial effects of complement therapeutics in COVID‑19 patients
In this section, we will discuss the evidence for complement-related COVID-19 therapies, and how they may deepen our understanding of the pathogenesis of the disease. Specifically, these therapies highlight how introducing external complement therapeutics at diferent stages of the activation cascade can help attenuate the hyper-inflammation engendered by COVID-19. For instance, downstream protein C5 has been a popular target due to preexisting complement therapies. In particular, the anti-C5 antibody eculizumab has shown positive effects in severe COVID-19 patients across various institutions (Fig. 1).
Eculizumab, when used in conjunction with standard treatment in severe COVID-19 patients, improves survival rates and reduces comorbidities like hemolysis [102, 103].
While eculizumab may engender a drop in various inflammatory markers and reactive complement protein levels [77, 81, 103, 104], the associated drop in complement activity may in turn be associated with a greater incidence of infectious complications such as ventilator-associated pneumonia [103]. While eculizumab is currently the most extensively investigated complement inhibitor, other antiC5 antibodies have also shown early success [105–107].
Tesidolumab, which binds to a separate C5 epitope, and ravulizumab, which shares the same epitope on C5 with eculizumab but has longer plasma stability, have also been investigated in preliminary single-arm studies of severeCOVID-19 patients (Fig. 1).
Vilobelimab, an anti-C5a antibody, has also shown efcacy as a treatment for acute COVID-19 (Fig. 1). Vilobelimab, when used in conjunction with standard treatment, have shown improved survival outcomes [108]. Unlike eculizumab, vilobelimab can attenuate the anaphylatoxic and hyper-infla mmatory effects of complement activation without paralyzing terminal MAC formation nor upstream complement activation; this may be beneficial for patients with comorbidities or who are at risk for coinfection.
In contrast, upstream complement inhibition has also been investigated. Anti-C3 therapy via AMY-101, a compstatin analog, has been promising. Data from an FDA
Phase 2 randomized controlled trial of severe COVID19 patients indicated that AMY-101 was associated with reduced markers for inflammation and immunothrombosis without inducing further complications [109] (Fig. 1).
Initial trials have shown that AMY-101 is comparable to eculizumab in reducing lung infammation [110]. Berinert, a C1 esterase inhibitor (iC1e), has also been investigated as a potential therapeutic. iC1e is a human-derived protein used to treat hereditary angioedema primarily via the inhibition of the kinin-kallikrein system, but it also inhibits C1s and downstream classical pathway activity
(Fig. 1). Conestat alfa, a recombinant rabbit-generated iC1e, has also undergone preliminary investigation. While preliminary investigation has produced mixed results, this approach can better isolate the effects of the classical pathway-specific response in severe COVID-19 [111, 112]. Narsoplimab, an anti-MASP-2 antibody, is a similarly upstream complement inhibitor that blocks lectin pathway activation. Narsoplimab treatment was associated with reduced circulating endothelial cell counts (a measure for endothelial damage), reactive complement proteins, inflammatory markers (IL-6, Il-8), and improved patient outcomes [58]. All these trial studies further shed light on the important role of the complement pathway activation and complement activation bioproducts such as C3a, and C5a and MAC participate in the pathogenesis of severe COVID-19. However, the cellular and molecular mechanism underlying complement-accelerated severe
COVID-19 remains unclear and requires further experimental investigation. Emerging role of complement in COVID-19 and other respiratory virus diseases Page 7 of 18 94 Animal models for the investigation of the role of the complement system in coronaviruses
Animal models for the investigation of MERS‑CoV and SARS‑CoV Complement activation not only occurs in COVID-19 patients but also in patients who are infected with Severe Acute Respiratory Distress Syndrome and Middle East
Respiratory Syndrome (SARS-CoV and MERS-CoV), two similar viruses from the coronavirus family that can also result in acute lung injury and ARDS [90]. Increased complement activation has been suggested to play a critical role in the pathogeneses of severe SARS and MERS [113, 114]. Like SARS-CoV2’s N protein, the N proteins of SARS-CoV, MERS-CoV were also found to bind MASP-2, increasing complement activation by potentiating MBL-dependent MASP-2 activation, and the deposition of MASP-2, C4b,activated C3 and MAC [90]. While the three diseases are similar, SARS-CoV and MERS-CoV have distinctly higher reported fatality rates (9.5% and 34.4% respectively) [115, 116]. Additionally, MERS-CoV targets the DDP-4 receptor rather than ACE2 and is associated with a greater incidence of renal failure [117, 118].
Experimentally, SARS-CoV has been investigated using C3 knockout murine models. While there are no SARSCoV-2 knockout studies published, investigators infected knockout and control (C56BL/6J) mice with a mouse adapted SARS-CoV virus passaged 15 times [119]. They found that C3−/− mice were most protected from SARS-CoV, while path-specifc knockouts showed weak protection [119]
(Table 1). This supports a multi-pronged approach toward COVID-related complement over activation. C3−/− mice did not display any weight loss following infection, and respiratory function was improved compared to controls. Staining and fow cytometry of lung tissue revealed much lower inflammatory monocyte and neutrophil populations [119].
Analysis of serum cytokines revealed that while infected control mice showed dramatic spikes in levels of IL-5, IL-6, CXCL1, and G-CSF on day two, their concentrations remained relatively constant over seven days of infection for C3−/− mice, and there was no difference in viral load in the lungs of C3−/− compared to control mice [119]. Further studies should clarify whether this phenomenon is present in COVID-19 as well and determine which complement associated responses to SARS are essential for protection.
The complement system’s role has also been investigated in mouse models of the MERS-CoV [120]. hDPP4 transgenic mice were infected with MERS-CoV and administered an anti-C5aR antibody or sham treatment (Table 1).
Histopathological analysis revealed antibody treatment was associated with alleviated lung damage, though there were no differences in overall survival. In contrast to SARS-CoV, anti-complement treatment was associated with lower viral load in the lungs and brain, but it was unclear whether these inconsistencies were driven by differences in the pathogen or the mouse model used in the experiments [120]. Inhibition of the C5a-C5aR axis also reduced splenic damage (measured by caspase-3 and TUNEL staining) and increased T-cell
regeneration (measured by PCNA) in red pulp (parenchymatous tissue of the spleen that consists of loose plates or cords infiltrated with red blood cells) [120]. Viral replication did not occur in the spleen (as hDPP4 expression, essential for viral uptake, was low), indicating that the damage induced by the complement system is at least partially systemic in nature. Although there are notable differences between MERS and SARS infection, over activation of the complement seems to be a common factor. While these pathogens are distinct from SARS-CoV-2, understanding the similarities between the three will improve our understanding of the complement system and better equip us to deal with future coronavirus outbreaks.
Animal models for COVID‑19
In addition to mice, Syrian golden hamsters and rhesus macaques also show promise as infection models for SARSCoV-2. Like mice, these animals show upregulated markers of complement activation following acute COVID-19 infection [121, 122]. Golden hamsters and rhesus macaques typically recover from SARS-CoV-2, so they may serve as potential alternative animals for non-lethal models [123].
Hamsters and nonhuman primates share a structural similarity in their ACE2 receptor that closely resembles that of humans and can be directly infected with the human virus strain, which could more closely replicate human COVID-19 infection and inform the design of antiviral antibodies. Complement factor D antibody reduces endothelial dysfunction, cytokine, and coagulation in the primate COVID-19 model, which highlights the importance of alternative pathway activation in COVID-19 as discussed above [97] (Table 1). Nonhuman primate models most closely mimic human disease phenotypes, so they are particularly useful for developing new therapeutics and experimentally evaluating the mechanisms of COVID-19 and long COVID-19, however, unlike mice, few tools are available to manipulate hamsters. And rhesus macaques are costly and can be logistically difficult to manage. Mice remain the most suitable models for a sequential investigation of complement component knockouts and illuminating the nuances of complement function.
SARS-CoV-2 does not recognize the mouse ACE2 receptor, and as a result, mice are not susceptible to severe infection by ordinary means. Current mouse models utilize different strategies; for example, K18-hACE2 (K18) transgenic mice express the human ACE2 receptor under the keratin 18 promoter to induce expression in epithelia. This facilitates 94 Page 8 of 18 M. T. Xiao et al. Table 1 Animal Models used Virus Complement associated harm or beneft Author and Year Complement factors investigated Model details Key findings Coronaviruses SARS-CoV Harmful Lisa Gralinski (2018) [119] C3 C3−/− and C57BL/6J control mice were infected with a mouse adapted SARS-CoV strain and compared against each other C3 deficient mice had significantly reduced respiratory disease even though viral load was unchanged. Complement-deficient mice have reduced neutrophilia in their lungs and reduced systemic inflammation MERS-CoV Harmful Yuting Jiang (2018) [120] C5a/C5aR Mice were genetically modified to express the hDDG4 receptor to facilitate MERS-CoV virus infection and replication. MERS induced increased C5a and C5b-9 complement activation products in sera and lung tissues. anti-C5aR antibody treatment led to decreased viral replication in lung tissues, and increased proliferation in the spleen SARS-CoV2 Harmful Youssif M Ali (2018) [92] Lectin pathway K18-hACE2 transgenic mice Administration of HG4, an anti- MASP-1 antibody significantly reduced the lung injury score including alveolar inflammatory cell infiltration, alveolar edema and alveolar hemorrhage. SARS-CoV2 Harmful Eri Kawakami (2023) [97] Alternative pathway SARS-CoV2-infected rhesus macaques. Anti-complement factor D monoclonal antibody mitigated abnormal complement activation, protected endothelial cells, and curtailed the innate immune response post-viral exposure RSV Harmful Fernando Polack (2002) [159] C3 C3−/−, B-cell deficient, and BALB/c control mice immunized with a formalin inactivated RSV and subsequently were challenged by RSV. C3 deficient mice had lower airway hyper responsiveness, but similar lung inflammation as their WT counterparts. The complement sys- tem plays an important role in the immune complex mediated damage from RSV Harmful Monali Bera (2011) [6] C3a/C3aR C3aR1−/− mice were infected with a human RSV strain and compared against BALB/c control mice. C3aR deficient RSV-infected mice had similar airway sensitivity as healthy controls. Bone marrow transplantation from C3aR −/− donors engendered a rescue pheno- type in WT mice who were infected with RSV
Emerging role of complement in COVID-19 and other respiratory virus diseases Page 9 of 18 94 Table 1 (continued) Virus Complement associated harm or benefit Author and Year Complement factors investigated Model details Key findings No effect or protective Alexander Bukreyev (2012) [160] C3, C5 (Cobra venom) WT BALB/c mice were treated with either cobra venom factor to deplete complement levels, or a saline control. The complement system does not reduce RSV viral titers in the lungs of RSV naïve animals. However, it plays a role in antibody-mediated complement cytotoxicity. Influenza A Beneficial Manfred Kopf (2002) [8] C3, Cr1, Cr2 C3−/−, Cr2−/−, and control C57BL/6 mice were infected with influenza (PR8) and then compared. C3 plays an important role during influenza infection via the induction of T-cell lymphocyte priming and migration. Beneficial Kevin O’Brien (2009) [164] C3, C5a, iC3b C3−/− and wt C57BL/6 mice were infected with influenza (PR8) and compared C3 deficiency was associated with delayed viral clearance, lower morbidity, as measured by body weight, and lower evidence of histological damage. No elevated C3 levels in mice infected with PR8 or CA09 Harmful Shihui Sun (2013) [164] C3aR, C5a and H5N1-infected mice treated with C3aR antagonist, anti-C5a anti- body, or cobra venom factor treatment of H5N1-infected mice with a C3aR antagonist, anti-C5a antibody or with cobra venom factor significantly reduced acute lung injury Harmful Nianping Song (2018) [162] C5a-C5aR1, C3, MAC C5aR1−/− and control BALB/c mice were infected with a lethal dose of influenza(PR8) and then compared. C5aR1 deficiency was associated with less lung damage and infammatory markers, although overall survival and body weight were unchanged Harmful M. Paula Longhi (2007) [165] MAC mCd59a−/− mice were infected with influenza virus, strain E61- 13-H17 CD59a, previously defned as a complement regulator, modulates both the innate and adaptive immune response to influenza virus by both complement-dependent and -independent mechanisms 94 Page 10 of 18 M. T. Xiao et al.
viral entry in the airway, simulating how infection begins in, humans. SARS-CoV-2-infected K18 mice recapitulate various COVID-19 phenotypes seen in humans, ranging from moderate to severe disease and even death [124–126]. The meta-analysis of single-cell data across five model species showed that the K18-hACE2 mouse model, followed by the hamster model, most closely resembled human COVID-19 lung pathology [127]. The severity of the disease in this model is dependent on the infectious dose of the SARSCoV-2 virus [128]. As discussed above, the inhibition of the lectin pathway of complement activation with anti-MASP-2 antibody (HD4) reduces ARDS severity in SRAS-CoV2- infected K18-hACE2 mice [92] (Table 1). This model has been widely used to evaluate the efficacy of potential therapeutics and vaccines; however, it is rarely used for mechanistic study, as the level and pattern of hACE2 expression in K18 mice differ from that of humans. While hACE2 expression in lung epithelia is similar in effect across both species, mouse models may not accurately reflect its level of expression across different organs in humans. For instance, hACE2 expression in mouse brain may be elevated compared to human brain tissue [129].
Other researchers have developed a non-lethal mouse adapted (MA) strain via a serial passage in BALB/c mice lungs, selecting virulent strains. After 10 passages, one strain (MA10) was identified that showed high virulence in standard laboratory mice without decreased fitness in human cells [130]. Importantly, this model imitated the age-dependent severity we see in humans, with older BALB/c micehaving more severe lung damage, weight loss, and mortality.
While this strain had high pathogenicity in BALB/c mice, it was mild in C57BL/6 mice. The highest dose tested (105 PFU) engendered a transient ~ 10% loss in weight, but no mortality.
In 2021, a more lethal mouse-adapted strain was identifed after 30 passages (MA30) through 8–10-week-old BALB/c mice. Among other differences, the MA30 strain showed spike protein mutations, including Q498R, Q493R,and K417M, all of which are suspected to enhance mACE2 binding [131]. Mutations in the S protein receptor-binding domain may limit the viability of this approach for antibody development, as this region is a primary target for neutralizing antibodies. A 5*104
PFU dose of the MA30 strain was highly lethal in young (6–10 weeks) BALB/c and middle aged (6–9 months) C57BL/6 mice, and non-lethal in young C57BL/6 mice [131]. The range in severity makes this a suitable strain for C57BL/6 mice [131]. Viral load was present but attenuated in other organs relative to the lungs. In K18-hACE2 mice, viral load in many organs is comparable to that of the lungs, including the heart and brain [129].
This may be because mACE2 expression more closely replicates ACE2 expression in human tissue than the transgene inserted under the K18 promoter. The C57BL/6 mouse strain is widely used for investigating the pathogenesis ofhuman disease in mice. C57BL/6 mice can be infected by the MA30 strain directly and have been crossed with a number of transgenic and knockout mice to introduce molecularly engineered genes in the C57BL/6 background. As such, MA30 can be easily used to dissect the pathogenesis of severe COVID-19 and potentially long COVID-19 as well. Dissecting the role of complement and complement activation in COVID‑19 using complement knockout mice.
While the MA COVID-19 strains and hACE2-K18 mice are useful for pathogenic studies, substantial species differences in complement and complement-regulatory proteins exist between humans and mice and must be considered in the application of a mouse model for human diseases. In mice, there are two CD55 (or DAF) proteins (one GPI-linked and the other expressed as a transmembrane protein [132, 133]), two CD59 GPI-linked proteins (mCd59a [134] and mCd59b [135]), and Crry, a structural and functional analog of human CD46 and CD55 that restricts three pathway convertases [136]. In contrast, humans express only one CD55
and one CD59. As Crry is expressed only in mice and rats, it is an appropriate and relevant inhibitor for studying C3 blockade in rodents. Of note, Crry is a structural and functional ortholog of human CR1, a soluble form of which has been shown to be safe in humans. mCd59a is the primary source of CD59 anti-MAC activity in mice under physiological conditions [137, 138]. mCd59b is expressed lower level on erythrocytes and a higher level on testis as compared with mCd59a [139, 138, 140]. Also, a recent study found that non–GPI-anchored intracellular isoforms of human CD59 and mouse CD59b are implicated in normal insulin secretion [141].
To our knowledge, there are no single-gene knockout studies of the complement system for SARS-CoV-2. Using pathway-specific, anaphylatoxin, MAC, and other immune system-related knockout models with the MA30 strain infected B6 mice or hACE2-K18 model can provide valuable insight into the specific functions, components, and regulators that contribute to complement-mediated pathogenesis.
For instance, to define how the complement system is activated, three complement pathways specific knockout mice including the mice deficiency in C1q [142], MBL [143], Factor B [144] and Factor D [145], key components for the classical, lectin, and alternative pathway activations can be used for conducting in vivo studies respectively. C3, C4 and C5 knockout mice together with Crry, DAF, C3aR and/or C5aR1 and C5aR2 knockout can be used for defining the role of complement in general and C3a-C3aR and C5aC5aR signaling in the pathogenesis of severe COVID-19 and long COVID-19 [136, 146–151]. C5, C7 and C9 and mCd59 knockouts can be used to define the role of MAC in Emerging role of complement in COVID-19 and other respiratory virus diseases Page 11 of 18 94 severe COVID-19 [152–156]. In addition to cell lysis, MAC formation has a sub-lethal role in activating the inflammasome [157]. A study comparing caspase-1 (a complement independent pro inflammatory enzyme) knockout mice and wild mice would illuminate which pro inflammatory mechanisms are most relevant in severe COVID-19. Conditional C3 or C5 knockout mice are useful to dissect the role of local complement activation and production in the COVID-19 pathogenesis [158]. A systemic investigation of complement knockouts together with specific pathway therapeutic inhibitors as described above may unveil non-canonical functions of the complement system and improve our understanding of COVID-19, other respiratory viral diseases as we will discuss below and SARS for future pandemics.
The role of the complement system in other respiratory viral diseases
The role of complement in the pathogenesis of other respiratory viral infections, including RSV and influenza, has been investigated. The net impact of the role of the complement seems to vary between different disease contexts; summarizing its impact across similar diseases may provide a beneficial context for the case of COVID-19.
The detrimental role of the complement in RSV infection
Like in COVID-19, complement hyper activation also plays a prominent role in the context of RSV. RSV is a common viral infection that primarily affects the respiratory tract, particularly in young children. RSV is a member of the Paramyxoviridae family, with a (-)ssRNA genome. Much like COVID-19, healthy individuals respond to RSV by developing mild cold-like symptoms, while infants, older adults, and immunocompromised individuals may develop more severe respiratory illness including bronchiolitis and pneumonia. In one mouse model simulating enhanced RSV disease, a particularly severe manifestation of RSV, researchers found symptoms were abrogated in C3-defcient mice [159]
(Table 1). Prior to infection, these mice had been administered an RSV vaccine which had previously been associated with increased morbidity in human children. While histopathology revealed similar levels of lung infammation, C3-defcient mice had much lower measured airway hyper responsiveness compared to their wild-type counterparts [159]. High airway hyper responsiveness persists in C5-defcient mice, highlighting the effects of C3a in particular [159]. This hypothesis was further corroborated by a separate study investigating the inhibition of the C3a-C3aR axis in RSV-infected mice [6] (Table 1). They found that C3aR1-defcient RSV-infected mice had similar levels of, airway resistance to healthy (sham-infected) controls. These mice had lower inflammation and faster viral clearance than wild-type infected mice. C3aR-defcient mice transplanted with wild-type bone marrow exhibited wild-type airway hyper responsiveness, while wild-type mouse recipients with C3aR−/− donors exhibited a rescue phenotype, highlighting the importance of bone marrow-derived immune cells in complement over activation. Molecular profiling revealed heightened production of Th17 cytokines as a potential mechanism for acute RSV-associated pathophysiology via complement C3a and tachykinins [6].
In another study, BALB/c mice were administered cobra venom factor to deplete complement components by activating the complement system [160]. Viral titers from isolated lungs revealed similar viral expression between complement-deficient and normal mice; however, the complement system may be more adaptive in repeat infections of RSV, since the complement system appears to assist in the antibody-mediated restriction of RSV replication [160] (Table 1). This suggests that the complement system does not seem to prevent RSV replication in infection-naïve mice.
Additionally, RSV can escape immune surveillance via the incorporation of complement regulators into viral particles.
One in vitro study revealed that CD55 KO (but not CD46 KO) cell lines had greater affinity to C3 deposition and had 2–3 times higher percentages of opsonized viral particles compared to wild-type cells [161]. Viral camouflage via host complement regulators combined with a direct contribution to harmful pathologies like airway hyper responsiveness reduces the efficacy of complement activity in RSV.
This may serve as a parallel for COVID-19. Indeed, some evidence suggests that gC1qR may play a role both in camouflaging COVID-19 viral particles and acting as a platform for complement activation and C3a and C5a production [70].
The role of complement in influenza infection
While cytokine storm and complement over activation are issues in influenza A [162, 163], evidence suggests that the complement plays a more complicated role in corresponding immune response as compared to coronavirus and RSV infections. Influenza, colloquially referred to as the fu, is, caused by viruses of the Orthomyxoviridae family. Various studies have highlighted the complement system’s ability to contribute to defense against influenza and control infection. C3-defcient mice show delayed viral clearance and increased viral loads in the lung. Fewer CD4+ and CD8+T-cells were detected in the BALF of these mice, indicating that C3 plays an important role in promoting specific immunity during influenza infection via the induction of T-cell lymphocyte priming and migration [8] (Table 1).
Another study of C3 knockout mice infected with avian influenza (H5N1) produced similar results. C3 deficiency was again associated with delayed viral clearance, along with higher morbidity, as measured by decreased body 94 Page 12 of 18 M. T. Xiao et al. weight, and greater histological damage [164] (Table 1). These experimental results indicate that C3 has a protective role in influenza infection.
In contrast, Sun et al. reported that treatment of H5N1- infected mice with a C3aR antagonist, anti-C5a antibody or with cobra venom factor significantly reduced acute lung injury [163]. Consistently, using C5aR1 deficient mice and the mice treated with an anti-C5aR1 antibody, Song et al. also show that C5a-C5aR1 signaling activation contributes to the development of influenza-induced acute lung injury [162] (Table 1). The removal of regulators like CD59 (CD59a-/-) in mice infected with influenza is associated with increased lung inflammation [165] (Table 1). However, treatment of mice with soluble complement receptor 1 reduced levels of lung-infiltrating neutrophils but not CD4+T-cells, suggesting that canonical complement regulators may be associated with non-terminal activation and complement independent pathways [165]. These results indicate that excessive complement activation accelerates H5N1-induced acute lung injury [163]. A better understanding of complement camouflage and the mechanistic role of complement activity in different pathologies may prove vital in managing future respiratory viral diseases.
Other respiratory diseases
Among other respiratory diseases, rhinoviruses, a diverse group of Picornaviridae responsible for the common cold, are known to inhibit intracellular complement activation (via the C3 signaling mechanism) through the expression of a viral cytosolic 3C protease [166, 167]. As nonenveloped viruses, intracellular complement signaling is particularly important. Interestingly, cell studies have revealed that intracellular C3 activation is basal during infection with RSV, an enveloped virus. Cell studies have also revealed that rupintrivir, a 3C protease inhibitor, can restore intracellular complement sensing and normalize complement activation.
Conclusions and future directions
The complement system plays diverse roles in the pathogenesis of respiratory viral infections. It contributes to host defense via the antibody-mediated complement-dependent lysis of virions and infected cells, helps neutralize viral particles, and primes the adaptive immune system by inducing T-cell lymphocytes; however, it also exacerbates morbidity and mortality via the hyper inflammatory effects of complement components, such as the anaphylatoxins C3a and C5aand the formation of MAC. While these maladaptive effects are particularly severe in COVID-19 and RSV, the role of the complement system in influenza A is complicated (namely, C3 may be protective while C3a-C3aR, C5a-C5aR may be harmful to the host). The mechanisms responsible for this difference remain unclear. Our current understanding leaves us with three broad questions:
1. What mediates the complement system’s deleterious effects in coronaviruses and RSV compared to its impact on influenza infection?
2. What are the mechanisms by which SARS-CoV-2 andRSV viral particles evade the complement system?
3. What are the relative roles of the activation of the three complement pathways in COVID-19 and other respiratory viral diseases?
Addressing these questions requires a comparison study such as fu vs COVID-19 infection in animal models [168]together with complement knockout mouse models as discussed above and a proteomic approach. Clinical observations also fail to differentiate between sub-lytic and lethal MAC formation. While C5 inhibition via eculizumab has been successful in treating these patients, the mechanism of damage must be clarified before better-targeted therapeutics can be designed. While clinical trials have shown some success modulating severe COVID-19 with monoclonal antibodies used to treat similar complement-related diseases, the lack of a fundamental understanding of the complement system precludes the development of more efficacious treatments. The following research questions will reveal the mechanistic roles of complement components in the pathogenesis of viral respiratory diseases:
1. What are the cellular mechanisms underlying complement-accelerated COVID-19? Do complement-mediated platelet activation, thrombosis, endothelial cell dysfunction, and/or myeloid cell activation contribute to COVID-19?
2. How do the anaphylactic, opsonizing, and terminal arms of the complement contribute to COVID-19 pathogenesis?
3. How does local complement production contribute to disease pathology compared to liver-derived components?
4. How does intracellular complement activation functionally differ from activation on the membrane/in plasma? Does this vary by cell type or disease context?
5. How do C3a, C5a, and sublytic MAC formation contribute to disease?
Addressing these questions requires further experimental investigation via animal models together with other approaches such as cell ablation, single-cell and special transcriptomics, and proteomics analysis [168–171]. Complement-related loss-of-function and conditional knockout Emerging role of complement in COVID-19 and other respiratory virus diseases Page 13 of 18 94models across the complement cascade, including organ based (lung/liver) complement deficiencies, should be extensively studied and compared against pre-existing models. A
thorough investigation of complement activation and regulation may provide further targets for therapy and build a strong foundation to defend against future pandemics.