Functional Analysis of Two Divergent C4 Isotypes in the Classical and Lectin Pathways of Complement Activation in the Common Carp (Cyprinus carpio)

Abstract

In the evolution of the complement system, a major humoral innate immune factor, the existence of multiple isotypes of the complement components is considered as a key strategy to enhance innate immune defense. Complement C4 is also diversified in a wide range of vertebrate species including teleost fish, possibly supporting the robust activation mechanism of the complement. To better understand the functional diversity of C4 isotypes in the teleost complement system, two C4 isotypes, C4-1 and C4-2, sharing only 32% amino acid sequence identity, were examined for binding specificities towards model target molecules representing microbe antigens and towards Gram-positive and -negative bacteria. The results suggest that C4-1 and C4-2 behave similarly in binding to the tested targets, despite the predicted difference in binding specificity based on the thioester catalytic site. The participation of C4-1 in the classical and lectin pathways of complement activation was also explored using pathway-specific activating enzyme complexes, C1r/s and MBL-MASP2. As a result, C4-1 can be activated in both the classical and the lectin pathways, at higher efficiency in the classical pathway. Taken together, the present results imply that both C4-1 and C4-2 isotypes are fully functional in the complement activation cascades, probably playing comparable roles in innate immunity.

1. Introduction

The immune system is a complex array working as the host defense and homeostatic system. The evolution of intricate host defense system has allowed mammals and other organisms to protect themselves against various intruders, including viruses, bacteria, fungi, and foreign bodies. As one of the oldest existing parts of the immune system, the complement system is present in evolution long before the development of adaptive immunity, in which the mammalian system is believed to be the most highly developed and understood [1,2,3]. The complement system of mammals consists of more than 30 proteins, present either as soluble proteins in the blood (plasma proteins) or as membrane-associated proteins, playing pivotal roles in pathogen clearance by direct killing and opsonization and in the promotion of inflammatory and adaptive immune responses [1,4]. Three distinct routes of complement activation have been identified: the classical, lectin, or alternative pathways. The lectin and alternative pathways are antibody-independent activation cascades and are believed to have the most ancient origins. As for the classical pathway, it is triggered primarily by antibodies complexed with their antigens and is regarded as the newest in animal phylogeny [5].

In classical pathway activation, the initial component C1, comprising subcomponents C1q, C1r, and C1s as a Ca²⁺-dependent complex, recognizes the Fc region of immunoglobulins using C1q and becomes an active form with proteolytically active C1r and C1s [6]. The active C1s cleaves complement components of C4 and C2, generating their active fragments of C4b and C2b, respectively. C4b can bind covalently to a target surface through its intramolecular thioester site, whereas C2b with serine protease activity can interact with C4b to form a complex, C4b2b, which now cleaves C3, the central complement component, to generate C3a and C3b fragments with potent physiological activities such as opsonization and anaphylatoxic leukocyte activation [1]. This C3 activation mechanism is well conserved in the lectin pathway as well, in which the C1 complex is replaced by a Ca²⁺-dependent complex of mannose-binding lectin (MBL) and proteases designated as MBL-associated serine proteases (MASPs). Three homologous MASPs (MASP1 to MASP3) have been identified in this complex, and MASP2 is responsible for the proteolytic activation of C4 and C2 [7].

Bony fish is one of the most primitive vertebrates that possess the three activation pathways of the complement system, thereby attracting much interest of immunologists from an evolutionary point of view [8]. One of the striking features of the bony fish complement system is the presence of highly diversified isotypes of many complement components, such as C3, C4, C7, and factor B in teleost (the largest group of the ray-finned fish class) [3,8,9]. Based on amino acid sequence diversity found in the complement isotypes, their possible functional diversity has been inferred [3,9] but is poorly supported by experimental evidence at the protein and cellular levels.

One of the key components shared by the lectin and the classical pathways is C4, which is a ~200 kDa glycoprotein composed of three disulfide-linked polypeptides, α-, β-, and γ-chains (94, 72, and 30 kDa, respectively, in human C4 and 95, 66, and 35 kDa, respectively, in carp C4) [10,11]. Located in its α-chain is the thioester site, which allows for the covalent binding of C4b to target molecules upon proteolytic activation [12]. In most mammals, including humans, two C4 isotypes have been characterized as C4A and C4B [13]. They share a 99.4% amino acid identity but differ in target specificity of the covalent binding and their physiological roles [12,14]. The key difference in the primary structures lies in the catalytic residues, located at the ~100 residue C-terminal of the thioester site, where C4B has a His residue (His-type) and C4A has an Asp (non-His-type). Due to the catalytic His, C4B can bind to a hydroxy group via the thioester site by making an ester bond, whilst C4A lacks the catalyzed binding and shows specificity towards the amino group, resulting in amide formation [12]. Due to the extremely high sequence similarity, C4A and C4B have been interpreted to be a recent gene duplication product in some mammalian species [5,13]. In addition, His-type C4 has been believed to be an authentic lineage conserved in most species analyzed to date, whilst non-His-type C4 is an option [13,14]. Our discovery of two C4 isotypes from teleost, C4-1 (non-His-type) and C4-2 (His-type), sharing only 32% identity, however, suggested their possible ancient evolutionary origin at least in vertebrates [11]. Moreover, the cloning of C4 from other teleost [15] and cartilaginous fish [16] and a recent genome-wide survey of diversified C4 isotypes revealed a much wider phylogenetic distribution of the non-His-type C4 isotype than expected [17]. Even some animal species of bony fish, reptiles, and birds have been shown to possess only non-His-type C4, suggesting that non-His-type C4 plays a significant role in complement-mediated immunity [17]. In the teleost group, functional studies of C4 protein have been reported only for rainbow trout [18] and the common carp [11]. The two forms of rainbow trout C4 studied were both His-type and were suggested to be trout-specific gene-duplication products [18]. The functional diversity of His-type and non-His-type C4 isotypes in non-mammalian vertebrates is still yet to be clarified.

In the present study, we aimed to elucidate the functional difference in carp C4-1 and C4-2 isotypes with regard to their target binding specificity and participation in the activation cascades of the classical and lectin pathways in order to gain insights into the functional significance of the gene duplication of the complement C4.

2. Materials and Methods

2.1. Carp Serum

Adult common carp (Cyprinus carpio) were obtained from a local fish farm in Fukuoka, Japan. Blood was collected from the caudal vessel and left to clot at 4 °C for 2 h. Carp serum was obtained via centrifugation at 3000 rpm for 5 min at 4 °C and aliquoted in 15 mL tubes and kept at −80 °C until use [19].

2.2. Buffers

The buffer formulas were as follows: PBS, 10 mM sodium phosphate buffer (pH 7.4) containing 0.9% NaCl; T-PBS, PBS containing 0.05% Tween 20; GGVB²⁺, 2.5 mM sodium barbital buffer (pH 7.5) containing 2.5% glucose, 72 mM NaCl, 0.1% gelatin, 0.15 mM CaCl₂, and 1 mM MgCl₂; EDTA-GGVB, 2.5 mM sodium barbital buffer (pH 7.5) containing 2.5% glucose, 72 mM NaCl, 0.1% gelatin, and 10 mM EDTA; TBS²⁺, 20 mM Tris-HCl (pH 7.5) containing 200 mM NaCl, 0.3 mM CaCl₂, and 2 mM MgCl₂.

2.3. Chromatography Columns

A Superdex 200 Increase column (1 × 30 cm) and HiTrap NHS-activated column (1 mL) were obtained from Cytiva (Tokyo, Japan). GlcNAc-agarose was purchased from Sigma-Aldrich (Tokyo, Japan) and packed in a PD10 column (Cytiva). Anti-carp C1q-HiTrap was prepared by conjugating 10 mg of anti-C1q A-chain rabbit IgG with a HiTrap NHS-activated column (1 mL), following the manufacturer’s instructions. Immobilized peptide columns were prepared using Toyopearl AF-Epoxy-650 (Tosoh Bioscience, Tokyo, Japan).

2.4. Preparation of Anti-Carp C1q-A Chain, Anti-Carp C4-1 Netrin Domain, and Anti-Carp C4-2 Netrin Domain

These antibodies were prepared as reported in the previous papers [20] and purified on HiTrap Protein G HP (Cytiva) according to the manufacturer’s instructions.

2.5. Preparation of Anti-Peptide Antibody Directed to α-Chain of C4-1 and C4-2 of Carp

Synthetic peptides, CYNYEGDEDQKDEPM (representing from ¹³⁸²Tyr to ¹³⁹⁵Met of carp C4-1 isotype; GenBank accession no. BAB03284) and CSYVQQTYNYYEDYE (representing from ¹³⁹⁰Ser to ¹⁴⁰³Glu of carp C4-2 isotype; GenBank accession no. BAB03285), were coupled to keyhole limpet hemocyanin (KLH) (Fuji Film Wako Pure Chemical, Osaka, Japan) through the thiol group of the N-terminal Cys using the m-maleimidobenzyol-N-hydroxysuccinimide ester as a cross-linker. Rabbit antisera were raised by immunization with the peptide-KLH conjugates emulsified in Freund’s complete adjuvant in SCRUM Inc (Tokyo, Japan). Anti-peptide antibodies were purified from the antisera using immobilized peptide columns. Specific antibodies were eluted from the column with 0.1 M glycine-HCl buffer (pH 2.5) and immediately neutralized with 1/10 volume of 1 M Tris-HCl (pH 9.0).

2.6. Preparation of Target Cells

Staphylococcus aureus (NCIMB8625 strain) and Aeromonas hydrophila (KAH8501 strain) were cultured using Soybean Casein Digest (SCD) plates and SCD broth (Fuji Film Wako Pure Chemical, Osaka, Japan) at 37 °C and 25 °C, respectively, and killed with 0.5% formalin or by UV-irradiation, as reported elsewhere [21].

2.7. Preparation of Carp Antisera against Bacteria

Carp (two individuals with an average body weight of about 100 g) were immunized via the intraperitoneal injection of formalin-killed A. hydrophila and S. aureus (2 × 10⁹ cells/mL-PBS, 100 µL/fish) eight times at weekly intervals. The bacteria concentration was determined by using Burker–Turk’s counting chamber. The antiserum raised towards each bacteria species was heat-inactivated at 50 °C for 20 min. Anti-A. hydrophila and anti-S. aureus showed agglutination titers of 256 and 1024, respectively, against each antigen bacteria.

2.8. Purification of Carp MBL-MASP2 Complex

The MBL-MASP2 complex was isolated from 20 mL of carp serum, following the published method [22], using GlcNAc-agarose affinity chromatography.

2.9. Purification of Carp C1r/s Complex

A published method reported for human and goldfish C1r/s [23,24] was closely followed with some modifications. Briefly, carp serum (10 mL) supplemented with 0.5 mM PMSF and 2 mM Pefabloc was made to 6% (w/v) with polyethylene glycol 4000 (Nacalai Tesque, Kyoto, Japan), and the precipitated fraction was dissolved in TBS²⁺ and passed through anti-carp C1q-HiTrap column (1 mL) equilibrated with TBS²⁺. After thoroughly washing the column with TBS²⁺, the C1r/s complex was dissociated from the C1 complex trapped on the column via elution with 10 mM EDTA in TBS.

2.10. ELISA for C4-Binding Assay

Wells of ELISA plate (Nunc-Immuno Plate) were coated with 5 µg/mL yeast mannan (Sigma), lipopolysaccharide (LPS) from E. coli O111:B4 (Sigma), poly-L-lysine (Sigma), or carp IgM, by closely following the reported procedure [21]. After blocking with 0.5% BSA-0.2% gelatin-PBS, 100 µL of normal carp serum diluted (1/20) with GGVB2+ or EDTA-GGVB was added to each well and incubated at room temperature for 30 min. After washing with 0.05% Tween 20-PBS, the plate was incubated with anti-C4-1-NTR or with anti-C4-2-NTR (1/1000 dilution), followed by peroxidase-conjugated second antibody and ABTS substrate. Absorbance at 405 nm was measured with the ImmunoMini NJ-2300 plate reader (Thermo Fischer, Tokyo, Japan). The results were expressed as means and SDs from triplicate measurements.

2.11. Flow Cytometric Assay of C4-Binding on Bacterial Target

The bacteria suspension (2 × 10⁹ cells/mL) was incubated with the same volume of 1/20 diluted anti-bacteria carp antiserum at room temperature for 30 min and then washed with PBS. The sensitized bacteria and non-sensitized control bacteria were incubated with 1/20 diluted normal carp serum at room temperature for 30 min and then washed with PBS. Activated C4 fragments (presumably C4b) bound to the bacteria were detected by incubation with anti-C4-1-NTR (20 µg/mL) or anti-C4-2-NTR (5 µg/mL), then with FITC-conjugated anti-rabbit IgG secondary antibody, followed by flow cytometry using Beckman Coulter Epics XL, essentially as described elsewhere [21].

2.12. SDS-PAGE and Western Blotting Analyses

SDS-PAGE and Western Blotting were carried out using a 7.5% separating gel and 4% stacking gel using Laemmli’s buffer system [25]. Samples were heat-treated (100 °C, 2 min) in 0.125 M Tris-HCl (pH 6.8) containing 2% SDS, 0.001% bromophenol blue, and 10% glycerol for non-reducing conditions. For the reduction of the sample, 5% 2-mercaptoethanol was included in the sample buffer. The separated proteins were transferred to PVDF membranes (Cytiva) in the transfer buffer (25 mM Tris-HCl, 192 mM glycine, and 10% (v/v) methanol, pH 8.3). The Western blotting was carried out using the primary antibody: 1/1000 diluted anti-C4-1 α peptide antibody, and the secondary antibody: 1/5000 diluted anti-rabbit IgG (H+L chains)-HRP (Medical Biology Laboratory Co., Tokyo, Japan). Color development of the membrane was achieved by using EzWest Blue TMB substrate (ATTO, Tokyo, Japan).

2.13. Dot-Blotting

The peptide-KLH conjugate solution and KLH alone dissolved in PBS (1 µL each) were dot-blotted onto PVDF membranes at protein doses of 1, 10, 100, and 1000 ng. After blocking with 3% (w/v) skim milk-PBS, the membranes were incubated with 1/1000 diluted affinity-purified anti-C4-1 α or anti-C4-2 α peptide antibody, prepared as above, and then with 1/2000 diluted HRPO-conjugated secondary antibody. The antigen dots were visualized with 4-chloro-1-naphthol substrate as previously described [21].

2.14. C4-Cleavage Assay

Carp serum (2.5 µL) was incubated with 5 µg of carp C1r/s complex or MBL-MASP2 complex in 10 µL of TBS at 25 °C for 2 h and analyzed via SDS-PAGE under non-reducing conditions and Western blotting with anti-C4-1 α, as described above.

3. Results

3.1. Specificity of Anti-Carp C4 Isotype Antibodies

The reactivity of anti-peptide antibodies directed to the C4-1 α-chain and C4-2 α-chain of carp was examined by dot-blotting against antigen peptides and Western blotting against carp serum proteins fractionated via gel-filtration. As shown in Figure 1, the affinity-purified anti-C4-1α and anti-C4-2α detected respective antigen peptides with no cross-reactivity. In addition, these antibodies gave only a slight signal with KLH carrier protein at the highest concentration (1 µg).

In Figure 2, anti-C4-1α detected a 190 kDa band in the fractions corresponding to a molecular mass of about 200 kDa in the gel-filtration of carp serum protein, suggesting that the antibody is highly specific to carp C4-1. On the other hand, anti-C4-2α did not give a positive band against any fraction.

Antibodies against the NTR domain of carp C4-1 and carp C4-2 were also tested for their reactivity via Western blotting using the C4-rich fraction from the gel-filtration of carp serum described above. Under reducing conditions, anti-C4-1-NTR and anti-C4-2-NTR detected single bands of 34 kDa and 35 kDa, respectively, corresponding to their γ-chain, which harbors the NTR domain (Figure 3). On the other hand, these antibodies gave no specific antigen band in Western blotting after SDS-PAGE under non-reducing conditions.

3.2. Binding Spectra of Carp C4 Isotypes against Model Targets

The deposition of C4 isotypes to LPS, mannan, poly-L-lysine, and carp IgM was assayed by ELISA, using anti-C4-1-NTR and anti-C4-2-NTR. As shown in Figure 4, C4-1 and C4-2 showed similar binding spectra against the model target molecules, while poly-L-lysine gave poor binding of both C4 isotypes. These results infer that the actual binding specificities of C4-1 and C4-2 to natural targets’ activation cannot be specified by the binding specificity through the thioester site towards the hydroxy group (accelerated by the catalytic His in C4-2) or amino group (non-catalyzed in C4-1).

3.3. Binding Spectra of Carp C4 Isotypes against Bacteria

Complement activation in normal carp serum was triggered by Gram-negative bacteria, A. hydrophila, and by Gram-positive bacteria, S. aureus, in the presence or absence of a specific carp antibody for sensitization. Sensitized bacteria can trigger the classical pathway, whilst non-sensitized bacteria are expected to trigger the lectin pathway, which was ensured by the addition of anti-carp C1q to block the classical pathway activation. Carp serum supplemented with 10 mM EDTA was also employed as a negative control that prohibits any pathway activation.

As shown in Figure 5A, non-sensitized A. hydrophila led to a low-level binding of C4-1 and C4-2, where C4-2 showed slightly more binding than C4-1. Sensitization with antibodies remarkably enhanced the deposition of both C4-1 and C4-2 on this target surface. Similarly, non-sensitized S. aureus yielded no binding of C4-1 and a little binding of C4-2; the deposition of both isotypes was increased to some extent by sensitization with antibodies (Figure 5B). These results suggest that the activation efficiency of C4-1 isotype is lower than that of C4-2 in the lectin pathway, and that both isotypes can be activated equally in the classical pathway.

3.4. Proteolytic Activation of C4-1 by C1r/s and MBL-MASP2 Complexes

The involvement of the C4-1 isotype in both the classical and lectin pathways was further examined using the initial C4-activating enzyme complexes, the C1r/s complex and MBL-MASP2 complex, respectively, purified from carp serum. As shown in Figure 6, the incubation of normal carp serum with purified C1r/s generated a 180 kDa band, probably representing a C4b fragment, from the 190 kDa band of intact C4-1. The MBL-MASP2 complex also yielded the 180 kDa band with a partially remaining 190 kDa band of intact C4-1. These results indicate that C4-1 can be proteolytically activated both in the classical and lectin pathways, while its activation efficiency would be higher in the classical pathway.

4. Discussion

In the present study, two different approaches were used to prepare specific antibodies against carp C4-1 and C4-2 isotypes. One was to raise polyclonal antibodies using a recombinant NTR domain harbored in the C4 γ-chain. As expected from their low amino acid sequence identity (30%) over the NTR domain between the two isotypes, the anti-C4-1-NTR and anti-C4-2-NTR were proven to be specific to C4-1 and C4-2, respectively, with no detectable cross-reactivity, and recognized native C4-1 and C4-2 proteins as well as their γ-chain in the Western blotting of reduced polypeptides [20]. These antibodies worked successfully in the ELISA and flow cytometry to determine the binding specificity, but unfortunately failed to detect the C4 isotypes in the Western blotting under non-reducing conditions, possibly due to differences in the tertiary structure of carp C4 between non-reduced and reduced forms. We therefore attempted to prepare a second set of antibodies that recognize the α-chain or whole C4 polypeptides, enabling the monitoring of the decrease in the molecular mass of the α-chain or whole C4 molecule via Western blotting. To this aim, the anti-peptide antibodies were made specific to 14 residues in the C-terminal region of the α-chain of the two isotypes. Despite the high reactivity and specificity towards the respective immunogen peptides, only the anti-peptide antibody directed to C4-1 successfully recognized its polypeptide band in Western blotting under the non-reducing conditions, but not the α-chain band under the reducing conditions. Thus, we could not examine the fragmentation of the C4-2 isotype due to the poor reactivity/specificity of the prepared anti-peptide antibody.

The functional characterization of human C4A and C4B as well as a line of evidence from in vitro mutagenesis study on the catalytic site of human C4 have clarified that the binding specificity of the thioester site of C4 is primarily determined by the catalytic residue, namely His or non-His [10,11,12]. His-type human C4B shows better binding to microbial targets covered by carbohydrate cell walls rich in hydroxy group, whereas human C4A (non-His-type) preferentially binds to proteinous target such as soluble immune complexes due to the relative abundance of the amino group [12,26]. Based on the experimental evidence, carp C4-1 (non-His-type) and C4-2 (His-type) probably show the same selectivity of target binding through the thioester site as those of human C4A and C4B.

The binding spectra of carp C4-1 and C4-2 with model target molecules do not agree with the binding specificities expected from the above presumptions. Namely, C4-1 (non-His-type), which cannot bind to hydroxy groups, actually bound to LPS and mannan, which are poor in hydroxy groups. In addition, both C4-1 and C4-2 showed indistinguishable binding spectra against the target molecule tested. It is also intriguing to note that C4-1 and C4-2 showed little to no binding to poly-L-lysine, rich in amino groups and with no hydroxy group, in contrast to the similar assay results on carp C3 isotypes, in which a non-His-type C3-S isotype showed a detectable level of binding to poly-L-lysine [18]. This discordance may suggest that poly-L-lysine is a poor activator of the lectin or classical pathway of carp complement but could activate the alternative pathway, which activates C3 without the participation of C4 and C2 components [1].

Since it seems unlikely that non-His-type C4-1, of which thioester should have a binding specificity towards amino groups, could bind directly to the hydroxy groups of LPS and mannan, the apparent binding of C4-1 to these hydoxy-group-rich targets might reflect its indirect binding mediated by other serum protein, presumably immunoglobulins, MBL, and C1, deposited on these targets [1,3].

The binding preference of carp C4-1 and C4-2 towards natural bacterial targets was also explored for a better understanding of their roles in bio-defense, employing Gram-negative fish pathogenic bacteria, A. hydrophila, and Gram-positive model bacteria, S. aureus. The results suggest that the binding of C4-1 is slightly weaker than that of C4-2 when they are activated in the lectin pathway. When the classical pathway is activated by sensitizing the target bacteria with antibodies, C4-1 binding becomes comparable to C4-2 binding. These results imply the possibility that C4-1 does not participate in the lectin pathway activation or that the activation efficiency of C4-1 is lower than that of C4-2 in the lectin pathway.

To examine these hypotheses, the pathway-specific activation of C4-1 was explored using the C1r/s complex, specific activating proteases of the classical pathway, and the MBL-MASP2 complex, in which MASP2 is responsible for the proteolytic activation of C4 and C2 in the lectin pathway. Purification methods of these activators from carp serum have been established in our laboratory and successfully applied to the present study. In the current study, C4-1 could be fragmented into C4b, the activated form of C4, by both the C1r/s and MBL-MASP2 complexes, indicating that C4-1 can participate in the lectin and classical pathways. It seems interesting that intact C4-1 remained after incubation with the MBL-MASP2 complex. This may suggest the relatively inefficient activation of C4-1 in the lectin pathway compared to that in the classical pathway, apparently in accordance with the results from the binding assay to bacteria targets. In this context, it is interesting that the lectin pathway may play an important role in the innate defense of the grass carp against A. hydrophila, involving the elevation of the gcMASP1 transcript, although this gcMASP1 seems most likely an ortholog of MASP3 of other animals [27]. Functional analyses at the protein level would provide insights into the defensive contribution of the lectin pathway and C4 isotypes against bacterial infection in aquaculture.

A recent homology-modeling study on grass carp C4 and MASP2 reported the possible formation of a C4-MASP2 complex, suggesting their tight functional linkage [28]. In future studies, this hypothesis could be verified at the protein level by utilizing a purified MBL-MASP2 complex and an anti-C4-1-NTR/anti-C4-2-NTR antibody for immunoprecipitation in the common carp, a close relative of the grass carp.

5. Conclusions

The present study revealed that both C4-1 and C4-2 isotypes, despite their divergent primary structure over an entire open-reading frame, are fully functional with similar binding spectra against natural targets.