The classical pathway is antibody-dependent and initiated by C1 complex, when binding of C1q to antibody IgM or IgG complexed with antigens (Duncan and Winter, 1988; Gaboriaud et al., 2014; Pednekar et al., 2016). Such binding events cause C1q conformational changes, which lead to the activation of C1r and C1s. Consequently, activated C1s splits C4 and C2, which then form C3 convertase (C4b2a) (Law and Dodds, 1997).

In addition to antigen-antibody immune complexes, C1q can bind directly with certain proteins on the surface of pathogens like bacterial cell walls, or polyanionic structures such as the lipoteichoic acid (Prellner, 1981). The binding of phosphatidylserine of the apoptotic cells, and other agents such as pentraxins (C-reactive protein and pentraxin 3) to C1q also initiates the classical pathway (McGrath et al., 2006; Biro et al., 2007).

The lectin pathway is activated by microbial sugars instead of immune complexes. It is stimulated when MBL, ficolins (M-ficolin, H-ficolin, and L-ficolin), or collectins (collectin-10 and collectin-11) bind to mannose residues on the surface of the pathogen (Bohlson et al., 2007; Garred et al., 2016). Once bound with glycans, the MASPs become activated from inactive zymogens. A conformational change occurs in MASP-1, following which MASP-2 is cleaved (Garred et al., 2016). Activated MASP-2 then cleaves C4 and C2 to form C3 convertase (C4b2a) (Heja et al., 2012), which continues the cascade as the classical pathway until the formation of the MAC on the membrane of the pathogen. As the lectin pathway can be activated in the absence of antibody, it plays an important role in early infection even before the antibody is generated.

The alternative pathway is activated by the spontaneous hydrolysis of C3 to from C3(H₂O). Physiologically, the alternative pathway maintains constitutive activation at low-levels in plasma by a hydrolytic process known as tick-over. This spontaneous hydrolysis of a labile thioester bond converts C3 to a bioactive form C3(H₂O) (Merle et al., 2015a). This results in a dramatic change in structure that exposes a binding site for FB (Merle et al., 2015a). FB binds to both C3b and C3(H₂O), which can be cleaved into Ba and Bb by a plasma protease, FD. At this point, C3bBb and C3(H₂O)Bb are formed as the C3 convertase of the alternative pathway. As C3bBb and C3(H₂O)Bb are unstable, properdin (factor P, FP) binds with them to stabilize the alternative pathway C3 convertase (Kemper and Hourcade, 2008).

All three pathways converge at the formation of C3 convertase, which promotes the cleavage of C3 into C3a and C3b. When a large quantity of C3b is generated, C5 convertase is formed by the binding of C3b to C4b2a or C3bBb (C4b2a3b or C3bBbC3b), which marks the activation of the terminal pathway (Schmidt et al., 2016). C5 is cleaved into the potent anaphylatoxin C5a and the bigger split fragment C5b by C5 convertases (Rawal and Pangburn, 2001). Large quantities of C3a and a limited quantity of C5a are released into the fluid phase, and then act on endothelial cells and mast cells (Dias Da Silva and Lepow, 1967; Kohl, 2001; Schraufstatter et al., 2002) as anaphylatoxins because of their ability to induce a shock-like syndrome similar to an allergic reaction (Coulthard and Woodruff, 2015; Hawksworth et al., 2017). The cascade of reactions leading to the assembly of the MAC on the membrane of the pathogen is initiated by C5b. This cascade starts with the binding of C5b to the monomer C6 to form the C5b6 complex, followed by further binding to the monomer C7 to expose the hydrophobic site on C7. The conformationally altered C7 is inserted into the membrane to bind C8 onto itself. Subsequently, C8β binds C5b67, and C8α-γ is inserted into the membrane, leading to the polymerization and insertion of 10–16 molecules of C9 into the membrane to form a pore (Parker and Sodetz, 2002). This 10-nm aperture is formed by the MAC, which disrupts the integrity of the lipid bilayer and allows the transfer of solutes and water across the membrane, resulting in osmotic lysis of the pathogen (Ricklin et al., 2010).

To inhibit the initiation stage, C1-INH, a plasma serine protease inhibitor, binds to the active enzymes C1r/s through covalent bond and dissociates it from C1q. C1-INH also binds through covalent bond to MASP-1 and MASP-2 and makes them dissociated from MBL. This leads to limit activation of the classical pathway and lectin pathway (Jiang et al., 2001).

Negative regulatory proteins that are also present in the plasma and/or on the cell membrane inhibit C3 and C5 convertases (Zipfel and Skerka, 2009). The C4BP has a high binding affinity to C4b and the C4b binding leads to the displacement of C2a from C3 convertase (C4b2a) (Schmidt et al., 2016). FI, an active serine protease in plasma, prevents the formation of C3 convertase and inactivates C4b2a and C3bBb through binding to C3b and C4b with the help of the cofactors FH, C4BP, MCP, and CR1. C3b is cleaved by FI and MCP first into membrane-bound iC3b and C3f, after which C3f is released into the fluid phase. The iC3b is then further cleaved by FI and CR1 into C3dg bound on the surface, and C3c released into the fluid phase. In the alternative pathway, C3bBb is permanently inactivated. Similarly, C4b is inactivated by FI and cleaved into C4c and C4d, thereby inhibiting the formation of C4b2a.

Conformational change in C3b exposes an extended surface in C3b that allows complement regulators FH, DAF, MCP, and CR1 to bind, resulting in the accelerated decay of the alternative pathway C3-convertase and inactivation of C3b (Alcorlo et al., 2015). FH is capable of competing with FB to bind with C3b (Zipfel and Skerka, 2009). DAF binds to C4b and dissociates C2a and Bb from C3 convertases (Fujita et al., 1987). It also displaces Bb from C3bBb to inhibit the formation of C3bBb or C3bBbC3b (Dho et al., 2018). In addition, CRIg can bind with remnant fragments of C3b (C3b, iC3b, and C3c) on the cell membrane to inhibit the alternative pathway (Zipfel and Skerka, 2009).

At the stage of MAC formation, several inhibitory proteins prevent the insertion of the MAC into the lipid bilayer. The S-protein, also known as vitronectin, binds with C5b67, C5b-8, and C5b-9 to inhibit MAC insertion. The SP40/40 not only binds directly to the C5b-9 complex but serves as a cofactor of the S-protein. Both C8BP and CD59 inhibit the binding of C9 to C5b-8, and C9 polymerization (Zipfel and Skerka, 2009).

Thus, regulation of complement activation at various stages of all three pathways spontaneously maintains homeostasis. Otherwise, any breaks in this balance could lead to autoimmune syndromes, such as atypical hemolytic uremic syndrome, with clinical features of damaged red blood cells and platelets, and kidney inflammation due to malfunction of the complement system (Yoshida et al., 2018).

When complement is activated, it contributes to the defense against pathogens, which is not only evolved as part of the innate immune system but remains a link to the adaptive immune system. Complement destroys pathogens directly via the MAC. Once complement activation has been initiated, all three pathways lead to the formation of C3 convertase, which initiates a cascade of enzymatic reactions that eventually lead to the formation of the MAC on the surface of pathogens, resulting in osmotic lysis of invading bacteria (especially Gram-negative bacteria), viruses, and parasites (Merle et al., 2015b).

Moreover, complement fragment C5a recruits neutrophils, monocytes, and macrophages to clear pathogens through a process known as opsonization (Ricklin et al., 2010; Martin and Blom, 2016). The C3b and C4b bind with CR1 expressed on phagocytes and erythrocytes to enhance IgG-mediated phagocytosis or clear soluble immune complexes (Freeley et al., 2016). In addition, iC3b can be specifically recognized by CR3 and CR4 on monocytes and neutrophils (Wang et al., 2015; Lubbers et al., 2017). This causes the phagocytes to release toxic reactive oxygen compounds and microbicidal components, such as lysozyme and proteases that kill pathogens (Martin and Blom, 2016).

Besides the participation of complement in innate immune defense, some of the original complement proteins, or intermediate components generated during activation, have many other immunity-enhancing functions. In addition to initiating the classical activation of complement, C1q has many unconventional functions. It is noteworthy that apart from its role as part of the complement system, C1q recognizes its receptors expressed on phagocytes, such as neutrophils and macrophages, and stimulates these phagocytes to release reactive oxygen species, such as H₂O₂ and superoxide, to attack pathogens. Furthermore, C1q plays a role in modulating dendritic cell (DC) maturation (Son et al., 2012), B cell tolerogenic capacity, and IgM or IgG production (Young et al., 1991). Moreover, C1q is reportedly involved in the clearance of apoptotic cells via CD91, which is a multi-protein-binding scavenger receptor complex containing CRT (Ogden et al., 2001).

Anaphylatoxins C3a and C5a stimulate DCs to express the major histocompatibility complex II (MHC-II) and CD86 via C3a and C5a receptors, thereby promoting their maturation as APCs (Peng et al., 2008, 2009). Locally produced C3a and C5a also interact with their receptors on T lymphocytes involved in maintaining T cell proliferation and differentiation (Strainic et al., 2008). The interaction between C3d and complement receptor CR2 on B cells during antigen-induced activation promotes B cell activation and facilitates the transformation of naïve B cells into antibody-producing effector and memory B cells (van den Elsen and Isenman, 2011; Donius et al., 2013). These alternative functions of complement components bridge the innate immune response with the adaptive immune response to coordinate the elimination of invading pathogens. The functions of the complement system are shown in Figure 1.

The C1-INH is a soluble regulator of complement activation that negatively regulates the classical and lectin pathways by inhibiting C1r, C1s, MASP-1, and MASP-2, the activating proteases of the complement cascade. The intracellular protozoa, Plasmodium falciparum, which causes the deadliest malaria in humans (Rich et al., 2009), could recruit and utilize C1-INH to inhibit complement activation (Mejia et al., 2016). Merozoites, the invasive form of blood-stage malarial parasites, actively recruit C1-INH to their surface when exposed to human serum. A member of the merozoite surface protein 3 family, PfMSP3.1, worked as a direct interactive partner to bind with C1-INH, to inhibit C1s, MASP1, and MASP2 (Kennedy and Wijeyewickrema, 2017).

Host complement FH, a CRP, is the main soluble inhibitor of the alternative pathway against the formation of C3-convertase. It achieves this inhibition by supporting the conversion of C3b to iC3b (Zipfel and Skerka, 2009). Echinococcus granulosus, a cestode that causes cystic echinococcosis in humans, expresses a protein that sequestrates FH on the hydatid cyst wall to inhibit complement C3b deposition (Diaz et al., 1997; Breijo et al., 2008). The binding molecule remained unknown until further research showed that it was InsP6, a major component of the acellular laminated layer (LL) of the hydatid cyst wall, that was bound with FH to inhibit the alternative pathway (Irigoin et al., 2008). Mosquito midgut epithelial cells also express two proteins (40 and 100 kDa) as receptors that captured FH to inhibit the deposition of C3b and impair activation of the alternative complement pathway (Khattab et al., 2015).

Decay-accelerating factor (DAF; CD55), another human complement regulatory factor, causes the accelerated decay of C3 and C5 convertases by associating with C4b and C3b deposited on the cell membrane (Fujita et al., 1987; Heine et al., 2003). Schistosoma mansoni is a blood fluke that causes intestinal schistosomiasis. When incubated with normal human erythrocytes, but not with DAF-deficient erythrocytes, S. mansoni became resistant to complement lysis in vitro (Horta and Ramalho-Pinto, 1991). Further study showed that S. mansoni acquired DAF from host erythrocytes via the expression of a GPI anchor on the surface of the worm (Ramalho-Pinto et al., 1992). The ability of the trypsin-treated S. mansoni worm to acquire DAF was reduced (Ramalho-Pinto et al., 1992). Treatment with GPI-specific phospholipase D (GPI-PLD) facilitated the binding of DAF to the surface of the schistosomula (Carvalho et al., 1994).

As a membrane-bound inhibitor of the cytolytic MAC, CD59 reduces C9 polymerization on the cell surface by binding to C8α and C9 (Venneker and Asghar, 1992). The N-linked glycosylation of CD59 is related to its complement-inhibitory activity (Ninomiya et al., 1992). P. falciparum is able to acquire the intrinsic host factor, CD59, to restrict complement attack on the infected erythrocyte (Wiesner et al., 1997). Furthermore, P. falciparum expressed mannosyltransferase (PfPIG-M), which is involved in GPI synthesis, and thereafter increased the levels of the GPI-anchored protein, CD59, on the cells, indicating that the GPI anchor is involved in the capture of CD59 on the surface of P. falciparum-infected erythrocytes (Kim and Hong, 2007). Trichomonas vaginalis, an anaerobic flagellated protozoan parasite, also acquired CD59 from different host cells, including red blood cells, during infection, to protect the parasite from being lysed by host complement (Ibanez-Escribano et al., 2015).

To avoid complement-mediated lysis, some parasites express a variety of CRPs on their surfaces. As more parasite genomes have been deciphered, parasitic helminths have been shown to express many mammalian-like receptors for host growth factors, cytokines, or hormones to regulate the growth, development, signaling, and reproduction of the parasite (Hu et al., 2003). More results suggest that parasites modulate host anti-parasite immune responses by expressing host immunological inhibitors and receptors of some immunological components (Ramirez-Toloza and Ferreira, 2017). The antigenic similarity between host- and parasite-expressed proteins may mask the host’s immune system to recognize invading parasites, and thus protect the parasite from elimination (Abu-Shakra et al., 1999).

Parasites express orthologs of host complement components or regulators to modulate or inhibit the functions of the host complement. S. japonicum, a blood fluke that causes Asian schistosomiasis, expresses protein (Schistosome CRIT) that share similarities with host complement C2 receptor inhibitor trispanning (CRIT) (Inal, 2005). This expression is indicative of their involvement in host complement activation or regulation. Schistosome CRIT is located on the surface tegument of the Schistosoma parasite and enables it to bind C2 via its extracellular domain. It subsequently inhibits the binding of C2 to C4b, to interfere with the formation of C3 convertase (C4b2a). The CRIT is an example of molecular mimicry, as it reportedly binds C2 with a domain that is homologous to one region of human C4b. Both the classical and lectin complement pathways are interrupted when C2 is hijacked (Cestari Idos et al., 2009). The C2 binding site of schistosome CRIT is located at an 11-amino acid sequence at the C-terminus of the first extracellular domain, which is involved in the inhibition of the classical complement pathway and reduction of immune complex-mediated inflammation (Inal et al., 2003). Trypanosoma cruzi, an intracellular protozoan parasite that causes Chagas disease, also expresses CRIT on the surface of trypomastigotes to inhibit C2-associated complement activation (Ramirez-Toloza and Ferreira, 2017).

In addition to the C2 receptor inhibitor, adult schistosomes express the C3 receptor on their surface tegument. During complement activation, C3 binds to the worm’s surface through the C3 receptor and stimulates the replacement of the outer tegument, which is shed during complement attack (Silva et al., 1993). Through this expression of the C3 receptor and shedding of the C3–C3 receptor, parasites are able to consume C3 from serum, and consequently become non-activators of the alternative pathway (Marikovsky et al., 1986; Rasmussen and Kemp, 1987).

In addition to its recruitment of host DAF on its surface to avoid host complement lysis, the T. cruzi trypomastigote also expresses DAF (T-DAF) on the surface of its virulent forms to inhibit complement activation by blocking C3, similar to mammalian DAF (Joiner et al., 1988; Kipnis et al., 1988; Tambourgi et al., 1993). Further studies have demonstrated that T. cruzi expressed a 160 kDa (GP160) complement regulatory glycoprotein on the surface of trypomastigotes (Norris et al., 1989). The gp160 gene was verified to share significant DNA sequence homologous with the human DAF gene (Norris et al., 1991). GP160 can inhibit the formation of the alternative and classical C3 convertase as it is a member of the C3/C4 binding family of complement regulators. This prevents the activation and amplification of the complement cascade on the parasite’s surface (Norris and Schrimpf, 1994; Norris et al., 1997).

An earlier study described a schistosome complement inhibitor, a 94-kD protein of S. mansoni (SCIP-1), expressed on the surface of S. mansoni larvae and adults, which was found to be functionally and antigenically related to human CD59. It binds to human C8 and C9, and inhibits the assembly of C5b-9 (Parizade et al., 1994). In addition, other CD59 homologs have been identified in the schistosome genome displaying the consensus CCXXXCN sequence at the C terminus (Wilson and Coulson, 2009) and in the membrane fraction of the live schistosome tegument (Castro-Borges et al., 2011). CD59 homologs (FhCD59-1,2,3) have also been found on the surface tegument of the trematode, Fasciola hepatica, a liver fluke (Shi et al., 2014). FhCD59-2 showed a phylogenetic relationship with SmCD59-2 on the surface tegument of S. mansoni (Shi et al., 2014). However, analogs of mammalian cell-expressed recombinant schistosome CD59 showed no inhibition of complement activity in vitro, which possibly differs from the action of native proteins expressed in the tegument (Farias et al., 2013). And both the location and potential function of the CD59-like proteins in F. hepatica even require further biochemical analyses to elucidate.

In addition to their expression of parasite-encoded regulators, which mimic host complement regulators, to inhibit complement activation, parasites also express or secrete a variety of proteins that directly bind to some complement components to inhibit their activation by targeting various stages (Figure 2).