The classical pathway of complement is crucial to the immune system, but it also contributes to inflammatory diseases when dysregulated. Binding of the C1 complex to ligands activates the pathway by inducing autoactivation of associated C1r, after which C1r activates C1s. C1s cleaves complement component C4 and then C2 to cause full activation of the system. The interaction between C1s and C4 involves active site and exosite-mediated events, but the molecular details are unknown. In this study, we identified four positively charged amino acids on the serine protease domain that appear to form a catalytic exosite that is required for efficient cleavage of C4. These residues are coincidentally involved in coordinating a sulfate ion in the crystal structure of the protease. Together with other evidence, this pointed to the involvement of sulfate ions in the interaction with the C4 substrate, and we showed that the protease interacts with a peptide from C4 containing three sulfotyrosine residues. We present a molecular model for the interaction between C1s and C4 that provides support for the above data and poses questions for future research into this aspect of complement activation.

Complement, an essential system for both innate and adaptive immunity, responds to the presence of a foreign pathogen within the vertebrate host. As well as directly affecting lysis of a pathogen, the complement cascade functions in inflammation and phagocytosis by acting as an opsonin, enhancing the migration of phagocytic cells to the infected area (1, 2), as well as initiating adaptive immune responses and regulating T and B cells (3, 4). The complement system generates a proteolytic cascade that initiates via one of three pathways of activation (57): classical, lectin, or alternative. Complement pathways have been associated with both unnecessary activation, causing inflammation in host tissues (8), and deficiencies, which contribute to autoimmunity and chronic infections (9, 10).

The classical pathway is initiated by the recognition molecule C1q, which is part of the 790-kDa C1 complex that assists in clearing infection and plays a role in immune tolerance and xenograft transplantation rejection (11, 12). The C1q molecule classically recognizes Ab–Ag complexes, but it also reacts with structurally different self and nonself targets, including C-reactive protein, bacterial porins, apoptotic cells, extracellular matrix proteins (13), polysaccharides, and prion–protein β-amyloid fibrils (1416). C1q has an associated Ca2+-dependent tetramer composed of the serine proteases (SPs) C1s-C1r-C1r-C1s (17). Binding of C1 to a target ligand leads to autoactivation of C1r, after which the activated enzyme cleaves proenzyme C1s at an Arg-Ile scissile bond within the SP domain to form a two-chain, active enzyme (7, 18). The highly specific C1s binds and cleaves its substrates, C4 and C2, which subsequently form the C3 convertase (C4bC2a) (8).

C1s consists of six domains, of which the CUB1-EGF-CUB2 domains assist in binding to the C1 complex via Ca2+ binding sites (19), whereas the C-terminal CCP1-CCP2-SP fragment functions in the catalysis of C4 and C2. The exact mechanism by which C1s interacts with C4 has not been elucidated, although evidence suggests that the CCP domains contain an additional binding site (exosite) for C4, whereas the SP domain executes both the recognition and cleavage of C2 (2022).

In this study, we show that the SP domain of C1s contains an exosite(s) that makes a major contribution to the binding and catalysis of C4. The residues making up an exosite on the C1s SP that contributes to the catalysis of C4 cleavage were identified. In parallel, evidence for the likely binding site on C4 for the C1s exosite was accumulated, and we constructed a molecular model to visualize the potential interaction and guide further work in this regard.

The cDNA for C1s was synthesized by GenScript (Piscataway, NJ). For all recombinant constructs, the pET-17b expression vector was digested with NheI and EcoRI restriction endonucleases, and PCR products with identical sticky ends were ligated into the plasmid. Preceding the C1s sequences, the forward primer contained the codons for the three amino acids (Ala-Ser-Met) of the T7-Tag sequence, which increase the efficiency of recombinant protein expression. C1s fragments were amplified using the following primers: CCP1 5′ (GCGGCTAGCATGAAGGGCTGGAAACTTCGCTATCAT), CCP2 5′ (GCGCGCTAGCCCTGTGGACTGTGG), SP 5′ (GCGGCTAGCCCAGTCTGTGGAGTC), CCP2 3′ (GCGGAATTCTTACTGTTTTTCTTCAAAGGGTTCTCTGGG), and SP 3′ (CGCGAATTCTTAGTCCTCACGGGGGGT). The proteins that were finally produced constituted Lys281 to Asp688 for the CCP1-CCP2-SP form and Pro423 to Asp688 for the SP form. Mutagenesis of the C1s fragments was carried out using the QuikChange Site-Directed Mutagenesis Kit, following the manufacturer’s methods and using splice-overlap PCR.

The expression plasmids were transformed into an Escherichia coli BL21Star(DE3) pLysS host strain, and the transformants were selected on Luria–Bertani medium plates containing ampicillin (100 μg/ml) and chloramphenicol (34 μg/ml). E. coli cultures were grown in 1 l 2YT at 37°C until an OD600 ∼ 0.7 was reached, at which point a final concentration of 1 mM isopropyl-d-thiogalactoside was added. Following induction, the culture was centrifuged (27,000 × g, 20 min, 4°C), the cells were collected in 30 ml 50 mM Tris-HCl, 20 mM EDTA (pH 7.4), and then frozen at -80°C. The cells were thawed and sonicated on ice six times for 30 s, with ≥10 s between sonication cycles. The inclusion bodies were collected by centrifugation (27,000 × g, 20 min, 4°C), and the supernatant was discarded. The pellet was washed three times with 10 ml 50 mM Tris-HCl, 20 mM EDTA (pH 7.4) and frozen at −80°C in between each wash. The inclusion bodies of the CCP1-CCP2-SP and SP constructs were solubilized in 8 M urea, 0.1 M Tris-HCl, 100 mM DTT (pH 8.3) at room temperature (RT) for 3 h. All denatured proteins were centrifuged at 27,000 × g for 10 min at 4°C. The solubilized proteins were drip-diluted at RT overnight into the refolding buffers (CCP1-CCP2-SP: 50 mM Tris-HCl, 3 mM reduced glutathione, 1 mM oxidized glutathione, 5 mM EDTA, and 0.5 M arginine [pH 9]; SP: 50 mM Tris-HCl, 3 mM reduced glutathione, 1 mM oxidized glutathione, 5 mM EDTA, and 0.5 M arginine [pH 10]).

Following an overnight refolding procedure, the renatured protein solutions were dialyzed against 50 mM Tris-HCl (pH 9), and renatured proteins were purified on a 5-ml Q-Sepharose-Fast Flow column (GE Healthcare, Piscataway, NJ). The column was equilibrated with 50 mM Tris-HCl (pH 9) for six column volumes (CVs) before loading. The column was washed with 6 CVs 50 mM Tris-HCl (pH 9), and protein was eluted with a linear NaCl gradient from 0 to 400 mM over 35 ml at 1 ml/min. Fractions were analyzed by 12.5% reducing SDS-PAGE. The recombinant proteins were further purified using a Superdex 75 16/60 column (GE Healthcare) in a buffer of 50 mM Tris-HCl, 145 mM NaCl (pH 7.4), aliquoted, snap-frozen, and maintained at −80°C. The enzymes were activated prior to use by incubating them overnight at RT with C1r purified from human plasma (Complement Technologies, Tyler, TX) immobilized on an NHS-Sepharose (GE Healthcare) matrix. Following overnight incubation, the activated enzymes were eluted using 50 mM Tris-HCl, 145 mM NaCl (pH 7.4).

C1s CCP1-CCP2-SP and SP concentrations were based on OD280 values using the calculated absorption coefficients (ε1%, 1 cm) of 18.7 and 17.8, respectively. Active concentrations of plasma-derived C1s (Complement Technologies), CCP1-CCP2-SP, and SP were determined by titrating the enzymes against C1 inhibitor (23) (Complement Technologies).

Assays were carried out in fluorescence assay buffer [0.05 M Tris-HCl, 0.15 M NaCl, 0.2% (w/v) PEG 8000, 0.02% (w/v) NaN3 (pH 7.4)] at 37°C using final substrate concentrations ranging from 0.5 to 20 μM. The C4 P4-P4′ fluorescence quenched substrate [2Abz-GLQRALEI-Lys(Dnp)-NH2] (GL Biochem, Shanghai, China) was solubilized in 10% (v/v) N,N-dimethylformamide. The final concentration of the C1s protease fragments in the assays was 400 nM. The rate of increase of fluorescence for the assays was measured on a BMG Technologies FLUOstar Galaxy fluorescent plate reader (BMG Labtech, Offenburg, Germany) using excitation and emission wavelengths of 320 and 420 nm, respectively. The initial reaction rate was estimated at a single concentration of enzyme from duplicate measurements over a range of substrate concentrations. To determine steady-state reaction constants (Vmax [maximum velocity], K0.5 [half saturation constant], and h [Hill coefficient]), the experimental results were fitted, using GraphPad Prism Version 5.0 software (GraphPad Software, San Diego, CA), to an equation describing positive cooperativity (V = Vmax[S]h/[S]h + [K0.5]h), which defines the relationship between reaction rate (V) and substrate concentration ([S]) when more than one binding site is present (24). The catalytic efficiency (kcat) values were calculated as described previously (24).

C1s protease fragments were incubated with H-Glu-Gly-Arg-chloromethylketone (EGRck) (10 μM; Sigma, Sydney, Australia) for 4 h at 37°C to prepare inhibited fragments for ELISA tests. To check for complete inhibition, the activity of 5 μg/ml the preparation was tested against Leu-Gly-Arg-NHMec (Bachem, Bubendorf, Switzerland) substrate (50 μM) at 37°C using a FLUOstar Optima Plate Reader with excitation and emission wavelengths of 370 and 460 nm, respectively.

For the determination of EC50 values, C1s fragments were diluted to final concentrations of 1–1500 nM in assay buffer (20 mM sodium phosphate, 150 mM NaCl, 5 mM EDTA [pH 7.4]). C4 (Complement Technologies) was diluted to 1 μM in assay buffer. The proteases and substrate were incubated separately at 37°C for 5 min and combined, and the cleavage reaction was allowed to proceed at 37°C for 1 h.

For time-course analyses, C1s single/cluster mutants were diluted to 1 nM, whereas the substrate C4 was diluted to 1 μM in assay buffer. The proteases and substrate were incubated separately at 37°C for 5 min and combined, and the cleavage reaction was allowed to proceed at 37°C for 0 to 120 min.

To determine IC50 values for C4-derived peptides, all proteins and peptides were tested at a final molar ratio of 1:1:1 (protease/peptide/substrate). The sulfated tyrosine-containing peptide (Ac-[Nle]-EANED[sY]ED[sY]E[sY]DELPAKDDPD-NH2) was synthesized via fluorenylmethyl (Fmoc)-strategy solid-phase peptide synthesis. The sulfated tyrosine residues were introduced via coupling of the synthetic building block Fmoc-Tyr[OSO3CH2C(CH3)3], which was prepared and introduced to the growing peptide chain using the method described previously (25, 26). The phosphorylated (Ac-[Nle]-EANED[pY]ED[pY]E[pY]DELPAKDDPD-NH2) and unmodified (Ac-[Nle]-EANEDYEDYEYDELPAKDDPD-NH2) C4 peptides were synthesized by GL Biochem using Fmoc-strategy solid-phase peptide synthesis. The modified and unmodified peptides were used at final concentrations of 5–500 μM. C1s CCP1-CCP2-SP and C1s SP were used at final concentrations of 25 and 370 nM, respectively. Protease fragments were incubated with individual peptide concentrations for 1 h at RT in assay buffer. C4 was diluted in assay buffer, incubated at 37°C for 5 min, and added to the protease/peptide mixture; the cleavage assay took place for 1 h at 37°C. The effect of the C4-derived peptides on the time course of C4 cleavage by C1s fragments was measured in a similar manner, using the same protease concentrations, whereas peptides were used at a final concentration of 500 μM. The cleavage reaction was allowed to proceed at 37°C for 0 to 120 min.

All reactions were stopped by the addition of reducing SDS-PAGE loading buffer, and samples were incubated at 90°C for 5 min, loaded onto 12.5% SDS-PAGE, and electrophoresed. Gels were stained with Coomassie blue R-250 stain and destained. The cleavage of C4 α band was analyzed using the γ band as a loading control. A Typhoon Trio (488-, 532-, and 632-nm lasers) was used for densitometry analysis using IQTL ImageQuant software (1D Gel Analysis) (GE Healthcare).

EC50 values were derived by analyzing the data points by nonlinear regression using the following equation: Y = Ymin + (Ymax − Ymin)/{1 + 10[(LogEC50 − X)*h)]}, where Ymin is the minimum Y value, Ymax is the maximum Y value, and h is the Hill slope.

All incubation steps were carried out using 50-μl solutions for 1 h at RT, except where stated otherwise. Every step was followed by four washes with 50 mM Tris-HCl, 150 mM NaCl, 2 mM CaCl2, 0.1% (v/v) Tween (pH 7.5). Microtiter plates (Maxisorp; Nunc) were coated overnight at 4°C with C1s mutants (10 μg/ml) diluted in 75 mM Na2CO3 (pH 9.6). The wells were blocked for 2 h with 230 μl 1% (w/v) BSA in PBS (8 mM sodium phosphate, 2 mM potassium phosphate, 0.14 M NaCl, 0.01 M potassium chloride [pH 7.4]) (blocking solution). Dilutions of human C4, diluted in 50 mM HEPES, 100 mM NaCl, 2 mM CaCl2 (pH 7.4), were added to the plates and incubated for 2 h. After washing, plates were incubated with specific rabbit polyclonal Abs against C4c (Dako) diluted 1:4000 in blocking solution. HRP-conjugated secondary Abs against rabbit IgG (Dako) were diluted 1:2000 in blocking solution and allowed to bind. Bound enzyme was quantified using the 3,3′5,5′-tetramethylbenzidine/H2O2 colorimetric assay (Sigma), which was quenched with 2 M H2SO4, and A450 values were measured. The A450 values were corrected for the background absorbance of substrate alone. Data points were analyzed using GraphPad Prism Version 5.0, using a single-site binding equation: Y = Bmax*X/(KD + X). This equation describes the binding of ligand to a receptor that follows the law of mass action. Bmax represents the maximal binding, and KD is the concentration of ligand required to reach one-half maximal binding.

Unless stated otherwise, all incubations were carried out at RT in 50 μl solution, and each step was followed by washing, as described above for the ELISA binding assay. Aggregated IgG was diluted to 2.5 μg/ml in 75 mM Na2CO3 (pH 9.6) and coated onto MaxiSorp microtiter plates (Nunc) overnight at 4°C. The wells were blocked for 2 h with 200 μl 1% (w/v) BSA in PBS (blocking solution). Various concentrations of C1s constructs, diluted in Gelatin veronal buffer [2.5 mM Veronal buffer (pH 7.3) 150 mM NaCl, 0.1% (w/v) gelatin, 1 mM MgCl2], were preincubated with 0.3% human serum for 15 min. C1s–serum mixtures were added to the plates and incubated for 20 min at 37°C, followed by 1 h of incubation with specific rabbit polyclonal Abs against C4c (Dako), diluted 1:4000 in blocking solution. HRP-conjugated secondary Abs against rabbit (Dako) were diluted 1:2000 in blocking solution and allowed to bind for 30 min. Bound enzyme was quantified using the 1,2-phenylenediamine dihydrochloride/H2O2 colorimetric assay (Dako), which was quenched with 0.5 M H2SO4, after which A490 values were determined.

ELISAs were carried out as described above, unless otherwise indicated. C1s CCP1-CCP2-SP S632A (10 μg/ml) was coated overnight. Following blocking, wells were incubated with either 7 or 70 nM C4 substrate, and the plate was washed four times and incubated with various concentrations of human IgG anti-sulfotyrosine, or the rabbit anti-C4c Ab described above, for 2 h at RT. Abs were diluted in blocking buffer, and concentrations of human IgG anti-sulfotyrosine were selected from a titration curve, showing low, medium, and high efficiency of binding to C4. Following four wash steps, goat anti-human IgG/HRP secondary Ab was incubated with wells containing human IgG anti-sulfotyrosine for 1 h at RT; wells containing the rabbit anti-C4c Ab were incubated as described previously.

Phosphorylated C4 peptide was synthesized (GL Biochem) with a triglycine linker to a biotinylated C terminus (Ac-[Nle]-EANED[pY]ED[pY]E[pY]DELPAKDDPD-GGGK[Biotin]). The peptide (1 mg) was solubilized in 10% (v/v) N,N-dimethylformamide, and the remaining 90% was solubilized in column-binding buffer (20 mM sodium phosphate, 0.15 M NaCl [pH 7.5]). The HiTrap Streptavidin column (1 ml) was equilibrated with 10 CV column binding buffer, and the peptide was loaded onto the column at 0.1 ml/min. Both ends of the column were capped and incubated for 1 h at 4°C, after which the column was washed with 10 CV column-binding buffer. Approximately 0.3–1 mg purified protein was diluted in 5 ml buffer A (50 mM Tris-HCl [pH 7.4]) and then bound to the HiTrap Streptavidin column at 0.2 ml/min. The column was washed with 6 CV buffer A and then eluted with a linear NaCl gradient to buffer B (50 mM Tris-HCl, 1 M NaCl [pH 7.4]) over 7 ml at 0.2 ml/min. The A280 (AU) and conductivity (mS/cm) values were plotted using GraphPad Prism Version 5.0.

A model of the C1s–C4 complex was created to investigate the hypothesis that the sulfotyrosine residues of C4 interact with C1s. This model is based on the published C1s structure (1ELV), and we used the published C3 (2A73) structure as the template structure for C4, building in two missing loops: the 720 loop, containing the scissile bond, and the 1350 loop, which is the large insert containing the three sulfotyrosine residues. The C1s was docked onto the 720 loop using the extended substrate interaction mode observed in crystal structures of serpins with thrombin (2729). Docking of C1s on the 720 loop results in the alignment of N-terminal “stock” of the 1350 loop, containing the sulfotyrosine residues, with the region where a sulfate binds to the C1s structure. A small amount of structural regularization was undertaken to alleviate steric clashes. All molecular modeling and docking were carried out using the programs PyMol and Coot.

Pure C1s domain fragments (Fig. 1) were obtained following refolding and purification at yields varying from 0.3 to 2.3 mg/ml per 1 l of E. coli culture, depending on the fragment being produced. All C1s domain mutants were purified to homogeneity using a two-step chromatography protocol, as indicated by SDS-PAGE analysis (Fig. 1) and reacted to Abs raised in chickens against CCP and SP domains (data not shown). C1s cannot autoactivate; therefore, we used C1r immobilized on an NHS-Sepharose column to activate the SP-containing fragments. All activated fragments were N-terminally sequenced to ensure that they had been correctly cleaved by C1r (data not shown). The activated fragments were able to react entirely with C1 inhibitor at a 1:1 ratio, indicating that they were 100% active (data not shown).

FIGURE 1.

SDS-PAGE analysis of purified recombinant C1s fragments. Schematic representation of C1s purified from human plasma or expressed as recombinant proteins. Plasma C1s is composed of the six domains: CUB1, EGF, CUB2, and a catalytic portion composed of the CCP1-CCP2-SP domains. Arrow indicates the Arg-Ile bond cleavage point for activation of the SP, held together by a disulfide bond indicated by the line in the SP. Asterisks label the active site S632A mutation. The numbers correspond to the wells in the SDS-PAGE (lower panel). Reduced samples were electrophoresed on 12.5% SDS-PAGE. Lane 1, Plasma-derived C1s; lane 2, CCP1-CCP2-SP zymogen; lane 3, activated CCP1-CCP2-SP; lane 4, activated CCP1-CCP2-SP S632A; lane 5, SP zymogen; lane 6, SP activated; lane 7, SP S632A zymogen; lane 8, activated SP S632A; lane 9, cluster Gln mutant of C1s CCP1-CCP2-SP [K575Q, R576Q, R581Q, K583Q]; lane 10, cluster Ala mutant of C1s CCP1-CCP2-SP [K575A, R576A, R581A, K583A]. The latter two lanes are representative of all mutant enzyme forms.

FIGURE 1.

SDS-PAGE analysis of purified recombinant C1s fragments. Schematic representation of C1s purified from human plasma or expressed as recombinant proteins. Plasma C1s is composed of the six domains: CUB1, EGF, CUB2, and a catalytic portion composed of the CCP1-CCP2-SP domains. Arrow indicates the Arg-Ile bond cleavage point for activation of the SP, held together by a disulfide bond indicated by the line in the SP. Asterisks label the active site S632A mutation. The numbers correspond to the wells in the SDS-PAGE (lower panel). Reduced samples were electrophoresed on 12.5% SDS-PAGE. Lane 1, Plasma-derived C1s; lane 2, CCP1-CCP2-SP zymogen; lane 3, activated CCP1-CCP2-SP; lane 4, activated CCP1-CCP2-SP S632A; lane 5, SP zymogen; lane 6, SP activated; lane 7, SP S632A zymogen; lane 8, activated SP S632A; lane 9, cluster Gln mutant of C1s CCP1-CCP2-SP [K575Q, R576Q, R581Q, K583Q]; lane 10, cluster Ala mutant of C1s CCP1-CCP2-SP [K575A, R576A, R581A, K583A]. The latter two lanes are representative of all mutant enzyme forms.

Close modal

The kinetic parameters for cleavage of the fluorescent quenched C4 P4-P4′ (2Abz-GLQRALEI-Lys(Dnp)-NH2) peptide substrate by the recombinant C1s fragments were compared with each other and plasma-derived C1s (Supplemental Table I). Overall, the kinetic values were similar for all forms of the enzyme, indicating that there was little significant alteration to the active site upon deletion of the N-terminal domains of C1s. These results are analogous to previous studies (20). The results also indicate that all recombinant forms of the enzyme were efficiently activated by treatment with C1r.

The effect of each C1s domain mutant on C4 deposition via the classical pathway was tested. The classical pathway was activated using wells coated with aggregated IgG. The human serum used in the assay was preincubated in the absence or presence of increasing concentrations of the C1s domain mutants. The effect on C4b deposition is illustrated in Fig. 2, which demonstrates that the SP S632A (C1s numbering, S195A chymotrypsin numbering) form of C1s had the highest overall impact on C4b deposition, with complete inhibition achieved at 1 μM. The CCP1-CCP2-SP S632A form behaved similarly, with ∼80% inhibition at a concentration of 1 μM.

FIGURE 2.

Recombinant C1s fragments inhibit C4b deposition. Aggregated human IgG was immobilized on the wells of a microtiter plate and allowed to activate 0.3% human serum containing various concentrations of C1s CCP1-CCP2-SP S632A (▪), C1s SP S632A (♦), or BSA (●) as a negative control (preincubated for 15 min at RT). After 20 min of incubation at 37°C, the plates were washed, and the deposited C4b was detected with specific polyclonal Abs. The absorbance obtained in the absence of C1s fragments was defined as 100%. The average of three independent experiments performed in duplicate is presented, with bars indicating SD.

FIGURE 2.

Recombinant C1s fragments inhibit C4b deposition. Aggregated human IgG was immobilized on the wells of a microtiter plate and allowed to activate 0.3% human serum containing various concentrations of C1s CCP1-CCP2-SP S632A (▪), C1s SP S632A (♦), or BSA (●) as a negative control (preincubated for 15 min at RT). After 20 min of incubation at 37°C, the plates were washed, and the deposited C4b was detected with specific polyclonal Abs. The absorbance obtained in the absence of C1s fragments was defined as 100%. The average of three independent experiments performed in duplicate is presented, with bars indicating SD.

Close modal

The affinity of each of the C1s forms for C4 was determined using an ELISA in which the enzyme fragments were coated to the plate. It was not possible to measure binding with C4 coated to the plate, possibly because of alterations to the substrate protein that precluded binding by the enzyme. Data for the binding of C4 to immobilized C1s fragments fitted well to a single-site binding equation to allow determination of KD values (Supplemental Fig. 1). Plasma-derived C1s and the CCP1-CCP2-SP forms of C1s exhibited similarly tight binding to C4, with KD values of 0.81 and 0.73 nM, respectively (Table I). The SP domain retained high affinity for C4, albeit with a 7-fold higher KD value than C1s CCP1-CCP2-SP. The active site S632A mutant of the CCP1-CCP2-SP form showed an ∼7-fold increased KD value compared with the active form of the enzyme, whereas the value for the S632A mutant of the SP form was essentially similar to the active SP form. Because it is possible that the KD values for active forms of the enzymes could be heavily influenced by C4b deposition following cleavage of C4 by these enzymes, we also determined the values for enzymes inhibited by the covalent inhibitor AEBSF. The KD values obtained for such inhibited forms of the enzymes were in the range of those found for the active forms (data not shown), suggesting that C4b deposition by the active enzymes did not account for the KD values observed in these analyses.

Table I.
Equilibrium binding constants and EC50 values for binding and cleavage of C4 by activated forms of C1s
C1s FormKD (nM)KD (nM) + EGRckaEC50 (nM)
Plasma-derived 0.81 ± 0.18 — — 
CCP1-CCP2-SP 0.73 ± 0.1 2.15 ± 0.4 
CCP1-CCP2-SP S632A 4.93 ± 1.10 — — 
SP 4.95 ± 0.8 25.9 ± 6.7 141 
SP S632A 6.59 ± 0.99 — — 
C1s FormKD (nM)KD (nM) + EGRckaEC50 (nM)
Plasma-derived 0.81 ± 0.18 — — 
CCP1-CCP2-SP 0.73 ± 0.1 2.15 ± 0.4 
CCP1-CCP2-SP S632A 4.93 ± 1.10 — — 
SP 4.95 ± 0.8 25.9 ± 6.7 141 
SP S632A 6.59 ± 0.99 — — 
a

KD value was determined in the presence of 10 μM EGRck.

—, Not determined.

The SP domain appeared to retain a significant binding affinity for C4, but it was not clear what role the active site (beyond the catalytic S632 residue) played in the binding process. The CCP1-CCP2-SP and SP fragments were inhibited with EGRck, an irreversible inhibitor targeted to the active site. EGRck would be expected to occupy the P3–P1 subsites of C1s, which were shown to be vital for interaction with cleavage sites in peptide and most likely protein substrates (30). The active site-inhibited CCP1-CCP2-SP fragment showed a 3-fold increased KD value in comparison with the active form (Table I), suggesting that the active site played a small role in the binding of the enzyme to C4. The C1s SP domain still exhibited binding to C4, despite the addition of the active site inhibitor, with a 5-fold increase in the KD value for the EGRck-treated form in comparison with the active SP domain. These results suggest that the active site has a role in the interaction between C1s and C4, but clearly an additional binding site(s) or exosite(s) on the SP domain is playing a major role.

Previous reports demonstrated that the SP domain is unable to cleave C4 without the assistance of the CCP1-CCP2 domains (20); however, the results from both the C4-deposition assays and ELISA reported in this article showed significant interactions between this domain and C4. Therefore, the cleavage of native C4 by C1s CCP1-CCP2-SP and SP fragments was investigated. Various concentrations of each protease were incubated with C4 (1 μM) for 1 h at 37°C. Reactions were stopped with reducing SDS-PAGE loading buffer and subjected to electrophoresis using SDS-PAGE (Supplemental Fig. 2). The cleavage of C4 was calculated by quantifying the disappearance of the α-chain of the protein using densitometry; these values were plotted against enzyme concentration to determine EC50 values of 1 and 141 nM for the CCP1-CCP2-SP and SP fragments, respectively (Table I). These results suggest that the SP domain of C1s is able to bind to C4 sufficiently to facilitate cleavage, albeit at a 140-fold lower efficiency than the CCP1-CCP2-SP form. It is likely that the results obtained showing that the SP domain retains the ability to cleave C4 are due to the higher concentrations of the SP domain used in the assays compared with the previous study (20). Therefore, we investigated whether the SP domain of C1s contained an exosite likely to form interactions with C4.

An electrostatic potential contour map of C1s CCP2-SP (PDB: 1ELV) showed that the majority of the protease surface was negatively charged, with only a few positively charged patches (Fig. 3). One positively charged patch at 90° to the active site binds a putative sulfate ion within the x-ray crystal structure of C1s (17). Four positively charged amino acids from within this area in the SP domain surround the sulfate ion. These residues were chosen for mutagenesis, where conservative mutations were made, altering Lys or Arg residues to Gln, either singularly or in a cluster (Fig. 3).

FIGURE 3.

C1s CCP2-SP (PDB: 1ELV) electrostatic map, with sulfate ions shown as yellow spheres. The residues of C1s mutated to Gln residues are indicated. Red represents predicted negatively charged areas, whereas blue indicates predicted positive charge.

FIGURE 3.

C1s CCP2-SP (PDB: 1ELV) electrostatic map, with sulfate ions shown as yellow spheres. The residues of C1s mutated to Gln residues are indicated. Red represents predicted negatively charged areas, whereas blue indicates predicted positive charge.

Close modal

The first aim was to investigate the kinetics of cleavage of synthetic peptide substrates representing the cleavage sequences in C4 (P4-P4′) (Supplemental Table I). Most single and cluster mutants showed modest changes in the individual kcat and K0.5 values; however, generally these changes compensated for each other, resulting in little overall variation in their kcat/K0.5 values compared with wild-type C1s. This suggests that the mutations had minor effects on the active site of C1s but little overall effect on the catalytic ability of the enzyme.

Interestingly, the mutations to the C1s SP had only little or no effect on the KD values for the interaction of the enzyme catalytic fragment with C4 (Table II). The results indicate that the identified cluster of positively charged amino acids may not constitute a binding exosite for C4.

Table II.
Equilibrium binding constants and EC50 values for binding and cleavage of C4 by mutants of C1s
C1s MutantKD (nM)EC50 (nM)
CCP1-CCP2-SP 0.84 ± 0.05 
CCP1-CCP2-SP K575Q 3.22 ± 0.22 2.4 
CCP1-CCP2-SP R576Q 2.07 ± 0.20 16.7 
CCP1-CCP2-SP R581Q 3.51 ± 0.26 39.1 
CCP1-CCP2-SP K583Q 2.38 ± 0.09 31.9 
CCP1-CCP2-SP K575Q, R576Q, R581Q, K583Q 1.06 ± 0.07 — 
CCP1-CCP2-SP K575A, R576A, R581A, K583A 1.42 ± 0.13 — 
C1s MutantKD (nM)EC50 (nM)
CCP1-CCP2-SP 0.84 ± 0.05 
CCP1-CCP2-SP K575Q 3.22 ± 0.22 2.4 
CCP1-CCP2-SP R576Q 2.07 ± 0.20 16.7 
CCP1-CCP2-SP R581Q 3.51 ± 0.26 39.1 
CCP1-CCP2-SP K583Q 2.38 ± 0.09 31.9 
CCP1-CCP2-SP K575Q, R576Q, R581Q, K583Q 1.06 ± 0.07 — 
CCP1-CCP2-SP K575A, R576A, R581A, K583A 1.42 ± 0.13 — 

—, Not determined.

The cleavage of C4 was initially examined by determining EC50 values for cleavage of the substrate. This analysis demonstrated that three single mutants, R576Q, R581Q, and K583Q, were significantly altered in their efficiency of cleavage of the protein substrate (Table II). The mutant R576Q had an EC50 value that was 17-fold higher than the wild-type enzyme, whereas the R581Q and K583Q enzymes had even more significantly increased EC50 values (32–39-fold higher). Mutant K575Q showed little change and had an EC50 value that was increased 2-fold relative to wild-type.

The effect of the single mutations on the time course of cleavage of C4 was examined. The time course analysis revealed that the K575Q mutant cleaves C4 at a slower rate than does the wild-type enzyme, but it has a similar end point. The remaining three mutants (R576Q, R581Q, and K583Q) all displayed considerably slower C4 catalysis, achieving ∼40% C4 cleavage after 120 min, compared with the wild-type enzyme (Fig. 4A). Mutants of the entire cluster of positively charged residues were then made, with the residues changed to either Gln or Ala. Preliminary determination of the EC50 values demonstrated that large concentrations of the cluster mutant enzymes were required for 50% cleavage of substrates (data not shown); therefore, the effects on the catalysis of C4 were tested using a time course analysis. Mutants were incubated with a standard concentration of C4 (1 μM), and the reaction was stopped at time points ranging from 0 to 120 min. As seen in Fig. 4B, the Gln cluster mutant only cleaved C4 to ∼25% of the extent of wild-type C1s, whereas the Ala cluster mutant had <5% cleavage of the substrate compared with the wild-type enzyme, even after 16 h.

FIGURE 4.

Time course analysis of C4 cleavage by single and quadruple mutants of C1s. Cleavage of C4 (1 μM) was carried out at 37°C, and the final enzyme concentration used in each case was 1 nM. (A) Single mutants: K575Q, R576Q, R581Q, K583Q. Reactions were stopped at 0, 1, 2, 5, 15, 30, 60, or 120 min. (B) Gln (K575Q, R576Q, R581Q, K583Q) and Ala (K575A, R576A, R581A, K583A) quadruple mutants: reactions were stopped at 0, 0.5, 1, 2, 4, 8, or 16 h. In both cases, the loss of the α-chain was quantified using densitometry, and the γ-chain was used as a loading control. Results are representative of three replicate experiments.

FIGURE 4.

Time course analysis of C4 cleavage by single and quadruple mutants of C1s. Cleavage of C4 (1 μM) was carried out at 37°C, and the final enzyme concentration used in each case was 1 nM. (A) Single mutants: K575Q, R576Q, R581Q, K583Q. Reactions were stopped at 0, 1, 2, 5, 15, 30, 60, or 120 min. (B) Gln (K575Q, R576Q, R581Q, K583Q) and Ala (K575A, R576A, R581A, K583A) quadruple mutants: reactions were stopped at 0, 0.5, 1, 2, 4, 8, or 16 h. In both cases, the loss of the α-chain was quantified using densitometry, and the γ-chain was used as a loading control. Results are representative of three replicate experiments.

Close modal

Because we had established that the residues on the surface of the SP domain of C1s making contact with a sulfate ion in the crystals used for determination of the enzyme’s structure were important for catalysis of the C4 substrate, we next wished to investigate whether sulfotyrosine residues on the surface of C4 were an important point of contact for C1s in its interaction with the substrate protein. C4 contains three sulfotyrosine residues on the α-chain contained within a 20-aa stretch of mostly negatively charged residues. To provide initial evidence that the area of interest on C4 may serve as a binding site for C1s, an experiment was conducted to investigate whether anti-sulfotyrosine Abs could still bind to C4 in the presence of C1s. This was performed using two concentrations of C4 and three concentrations of the Ab to overcome any possible equilibrium-binding constraints. The anti-sulfotyrosine Ab concentration used was derived from titration curves, yielding concentrations showing low-, medium-, and high-efficiency binding to C4. Binding of the anti-sulfotyrosine Ab to C4 alone was used as a control to reveal 100% binding at each concentration of the Ab. Additionally, the binding of C4 to C1s was demonstrated using an anti-C4 Ab, which showed that C4 was equivalently present in these cases and should have been available for binding to the anti-sulfotyrosine Ab if the sulfotyrosine residues were available for interactions with the Ab. The results indicated that C1s strongly inhibited the binding of the anti-sulfotyrosine Ab, suggesting that the sulfotyrosine-containing area on C4 is indeed involved in the interaction with C1s (Fig. 5).

FIGURE 5.

Effect of C1s on binding of anti-sulfotyrosine Abs to C4. C1s inhibition of binding of human anti-sulfotyrosine IgG (1–50 nM) to C4 at 7 nM (black bars) or 70 nM (diagonal black line bars). Values were normalized using the binding of anti-sulfotyrosine IgG to C4 alone at 7 nM (gray bars) or 70 nM (white bars) as 100%. The presence of C4 binding to C1s at 7 nM (horizontal lined bar) or 70 nM (cross-hatched bar) was confirmed in each case by measuring the binding to an anti-C4 Ab. The ability of anti-sulfotyrosine Ab to bind to C1s was also measured (stippled white bar). Results are representative of three replicate experiments. ***p < 0.0001, two-way ANOVA.

FIGURE 5.

Effect of C1s on binding of anti-sulfotyrosine Abs to C4. C1s inhibition of binding of human anti-sulfotyrosine IgG (1–50 nM) to C4 at 7 nM (black bars) or 70 nM (diagonal black line bars). Values were normalized using the binding of anti-sulfotyrosine IgG to C4 alone at 7 nM (gray bars) or 70 nM (white bars) as 100%. The presence of C4 binding to C1s at 7 nM (horizontal lined bar) or 70 nM (cross-hatched bar) was confirmed in each case by measuring the binding to an anti-C4 Ab. The ability of anti-sulfotyrosine Ab to bind to C1s was also measured (stippled white bar). Results are representative of three replicate experiments. ***p < 0.0001, two-way ANOVA.

Close modal

Three C4 peptides were synthesized with three sulfotyrosine (Ac-[Nle]-EANED[sY]ED[sY]E[sY]DELPAKDDPD-NH2), three phosphorylated tyrosine (Ac-[Nle]-EANED[pY]ED[pY]E[pY]DELPAKDDPD-NH2), and three underivatized tyrosine (Ac-[Nle]-EANEDYEDYEYDELPAKDDPD-NH2) residues to assess the effect of the peptide alone, as well as the modifications of the tyrosine residues by sulfate or phosphate groups. The relatively difficult synthesis of the sulfotyrosine-containing peptide resulted in a low yield of the material. It was shown previously that the substitution of sulfate with phosphate groups on tyrosine residues had minimal effects on the binding interaction of certain proteins and ligands (31); therefore, the effects of the phosphate modification on the interaction were also investigated. The replacement of the sulfate groups with phosphate greatly improved the yield and stability of the peptide, which was beneficial in later experiments.

Initial experiments demonstrated that a concentration of 500 μM of the C4 peptides showed the most significant impact on C4 cleavage for both C1s domain fragments; therefore, this concentration was chosen to investigate their effects on the time course of C4 cleavage (data not shown). Proteases were incubated either alone or with C4 peptide and then allowed to cleave C4 over a 2-h time course. Preincubation of sulfated peptide with both C1s fragments reduced their C4 cleavage, but the most significant results were once again seen for C1s SP (Fig. 6A, 6B). Typically, the percentage of C4 cleavage was half of that for the enzyme in the absence of the peptide at each time point when the protease was preincubated with the sulfated C4 peptide. Preincubation of the enzymes with the phosphorylated C4 peptide resulted in small reductions in C4 cleavage, and both C1s fragments had cleaved the same percentage of C4 by the end of the assay in the presence or absence of the phosphorylated peptide (Fig. 6C, 6D). Finally, preincubation with the nonsulfated C4 peptide had minimal effects on the cleavage of C4 by the C1s SP fragment, and it decreased the C4 cleavage activity of C1s CCP1-CCP2-SP by ∼20% (Fig. 6E, 6F). This latter effect may be due to relatively nonspecific binding of the peptide to the CCP domains of the protease.

FIGURE 6.

Analysis of the effect of C4-derived peptides on C4 cleavage by C1s fragments. C1s CCP1-CCP2-SP and SP were incubated with 500 μM of sulfated (●) (A, B), phosphorylated (▪) (C, D), or nonsulfated (▴) (E, F) C4 peptide and allowed to cleave C4 (1 μM) at 37°C over the indicated time range. (A) Sulfated peptide and CCP1-CCP2-SP. (B) Sulfated peptide and SP. (C) Phosphorylated peptide and CCP1-CCP2-SP. (D) Phosphorylated peptide and SP. (E) Nonsulfated peptide and CCP1-CCP2-SP. (F) Nonsulfated peptide and SP. Protease alone (--*--) was used as a positive control. In all experiments, densitometry was calculated from the loss of the α-chain, using the γ-chain as a control. Results are representative of two replicate experiments.

FIGURE 6.

Analysis of the effect of C4-derived peptides on C4 cleavage by C1s fragments. C1s CCP1-CCP2-SP and SP were incubated with 500 μM of sulfated (●) (A, B), phosphorylated (▪) (C, D), or nonsulfated (▴) (E, F) C4 peptide and allowed to cleave C4 (1 μM) at 37°C over the indicated time range. (A) Sulfated peptide and CCP1-CCP2-SP. (B) Sulfated peptide and SP. (C) Phosphorylated peptide and CCP1-CCP2-SP. (D) Phosphorylated peptide and SP. (E) Nonsulfated peptide and CCP1-CCP2-SP. (F) Nonsulfated peptide and SP. Protease alone (--*--) was used as a positive control. In all experiments, densitometry was calculated from the loss of the α-chain, using the γ-chain as a control. Results are representative of two replicate experiments.

Close modal

The binding efficiency of each C1s fragment for the C4 peptide was then investigated. Ideally, a sulfated C4 peptide would have been preferred for such assays because it had the strongest effects on C4 cleavage by C1s enzyme fragments, but such assays were difficult because of the very low amounts of peptide available. The results above show that the phosphorylated C4 peptide was able to interact with the proteases enough to inhibit their ability to cleave C4; therefore, a biotinylated phosphorylated C4 peptide was synthesized. A triglycine linker was attached to the C-terminal end of the peptide, followed by a biotin molecule. The biotin-labeled peptide was bound to a streptavidin column, and proteins, in turn, bound to the peptide were eluted using a linear NaCl gradient (0–1 M); the proteins with high binding efficiency for the C4 peptide eluted from the column at higher NaCl concentrations. Activated C1s SP once again had the highest affinity for the peptide (Fig. 7), followed closely by activated C1s CCP1-CCP2-SP. The four C1s CCP1-CCP2-SP single mutants of interest (K575Q, R576Q, R581Q, K583Q) and both the quadruple Ala and Gln mutants demonstrated no binding to the C4 peptide. The interaction between the exosite on the SP of C1s and the sulfated stretch of amino acids on C4 was modeled (Fig. 8), indicating that the exosite on the SP could engage the loop containing the sulfotyrosine residues in C4 simultaneously to the engagement of the active site of the enzyme with the cleavage site on C4.

FIGURE 7.

Binding of C1s to a phosphotyrosine-containing C4 peptide. The indicated forms of C1s were applied to an affinity column with an attached peptide of the sequence from C4 containing phosphorylated tyrosine residues. The column was washed and then eluted with NaCl, as indicated by the gray dotted line showing the increase in conductivity as the salt concentration increased. The peaks for elution of the C1s forms are indicated, with all C1s mutants eluting during the washing phase. Results are representative of two replicate experiments.

FIGURE 7.

Binding of C1s to a phosphotyrosine-containing C4 peptide. The indicated forms of C1s were applied to an affinity column with an attached peptide of the sequence from C4 containing phosphorylated tyrosine residues. The column was washed and then eluted with NaCl, as indicated by the gray dotted line showing the increase in conductivity as the salt concentration increased. The peaks for elution of the C1s forms are indicated, with all C1s mutants eluting during the washing phase. Results are representative of two replicate experiments.

Close modal
FIGURE 8.

Model for interaction between C1s and C4. C1s CCP2-SP (PDB: 1ELV) surface map in dock-model with C3 (PDB: 2A73) (pink) refined with C4 sulfotyrosine-containing loop 1414NED-KVV1459. Inset, Modeled interaction sites between C1s and C4. C4 tyrosine residues (green sticks) 1420 and 1422 interacting with C1s sulfate-binding pocket (K575, R576, R581, and K583). Contact points: C4 tyrosine residues (green sticks) bound simultaneously with C4 cleavage peptide (hot pink sticks) in active site.

FIGURE 8.

Model for interaction between C1s and C4. C1s CCP2-SP (PDB: 1ELV) surface map in dock-model with C3 (PDB: 2A73) (pink) refined with C4 sulfotyrosine-containing loop 1414NED-KVV1459. Inset, Modeled interaction sites between C1s and C4. C4 tyrosine residues (green sticks) 1420 and 1422 interacting with C1s sulfate-binding pocket (K575, R576, R581, and K583). Contact points: C4 tyrosine residues (green sticks) bound simultaneously with C4 cleavage peptide (hot pink sticks) in active site.

Close modal

The cleavage of C4 by the C1s protease of the C1 complex is a key point in the activation of the classical pathway of the complement system. Because the unregulated activation of complement underlies the pathogenesis of many inflammatory diseases (8), understanding the mechanisms of its initiation and regulation is vital to potentially develop therapeutic agents targeting the system. The modular nature of C1s has given rise to some debate as to the molecular mechanism by which it interacts with C4 and cleaves it to form C4b (2022). In this article, we provided considerable insight into the interaction between the enzyme and its primary substrate.

Previous data in the field indicated a key role for the CCP domains of C1s in the cleavage reaction for C4, and some inferences have been made about the binding of the enzyme to the substrate based on kinetic constants for the catalytic reaction (20, 22). Such inferences assume a linear relationship between the Km values for the catalytic reaction and equilibrium dissociation constants for the enzyme–substrate reaction that rarely exist. In this study, we probed the specific role of the SP domain of the catalytic region of C1s using a combination of binding and catalytic reactions. The data from our experiments are in agreement with previous literature in showing that the CCP domains of the enzyme play a role in binding and catalysis by the enzyme (2022). However, our studies showed that the SP domain of the enzyme played a greater role in binding to C4 than previously thought and, in particular, they showed that region(s) on the SP outside of the active site of the enzyme are important for binding C4, consistent with the C1s SP housing an exosite(s) for C4. This is somewhat different from the situation with the MASP-2 protease of the lectin pathway, which also has C4 as its primary substrate. For MASP-2, it appears that the CCP2 domain of the protease is the primary location of an exosite for C4 (32).

In our efforts to identify likely regions of the C1s SP that might constitute the exosite for C4, we noted that the protein contains sulfate ions in the form of three sulfotyrosine residues incorporated into the α-chain of C4 in close proximity to each other among a stretch of negatively charged residues 1412EANEDYEDYEYDELPAKDDPD1432 (sulfated tyrosines are underlined). This pattern of amino acids is widely accepted to be a protein–protein interaction motif (26, 3335). Interestingly, nonsulfated C4 was shown to have 50% less hemolytic activity in complement activation assays, and a 10-fold higher concentration of C1s is required for cleavage of C4 (36). We also noted that the crystallized CCP2-SP enzyme had a sulfate ion bound to its surface in the midst of a cluster of positively charged residues: K575, R576, R581, and K583 (Fig. 3). The presence and location of the sulfate ions in the crystal structure of C1s and the potential for sulfate groups to be involved in the interaction between C1s and C4 assisted in selecting positively charged residues of C1s for mutagenesis to elucidate their role in interactions with C4.

The interaction site on C4 for C1s was investigated using a range of techniques and representative sequences from the substrate. The kinetic values for cleavage of the synthetic substrates based on the cleavage sequences in C4 were determined for all mutants. Mutation of the four residues surrounding the sulfate ion on the C1s SP domain had little effect on the kinetics of cleavage of peptide substrates by the mutant enzymes, indicating that these residues were not involved in direct interactions between the enzyme and the cleavage sequences of the substrates. However, mutagenesis of the four residues resulted in enzymes that were significantly altered with respect to their cleavage of the protein substrate C4 compared with wild-type C1s. The K575Q mutant appeared to be the least affected of the mutants, showing a moderate change in EC50 value and rate of cleavage of C4 in a time course assay. The other three mutants, R576Q, R581Q, and K583Q, showed markedly increased EC50 values for cleavage of C4 in comparison with wild-type C1s. All of these mutants were also substantially affected in time course analyses of C4 cleavage. It is interesting to note that these are the three residues that are in close vicinity to the trilobed sulfate ion, with K575 slightly further away. The data and characteristics fit with the four residues making up an exosite for the C4 substrate. Therefore, these residues were targeted to create two quadruple mutants, altering the Lys or Arg residues to either Gln or Ala residues. Cleavage of C4 was significantly reduced for both the Gln and Ala cluster mutants, confirming that the four charged residues together most likely form an exosite that mediates efficient cleavage of C4. The single mutants or quadruple mutants of C1s showed little alteration in their equilibrium dissociation constants for C4. This may indicate that there are still yet-to-be-discovered binding residues on the C1s SP for C4 or that further investigation of the individual rate constants of the overall reaction is required to better understand the kinetic mechanism of C4 binding. In any event, it is clear that the cluster of positively charged residues on the C1s SP at the very least constitutes an exosite required for efficient catalysis of the substrate by C1s. This implies that the cluster of positive charges ideally positions the protease on the substrate for efficient cleavage.

As noted previously, the major potential interaction site for a positively charged region on C1s would be the highly negatively charged region on C4 containing three sulfated tyrosine residues. The inability of the anti-sulfotyrosine Ab to bind C4 in the presence of C1s, as well as the ability of peptides with the sequence of the sulfotyrosine-containing stretch of C4 to interfere with cleavage of the protein by C1s, provided strong indications that this region of C4 is important for interactions with C1s, particularly when coupled with previous literature indicating the importance of sulfotyrosines for interaction with the enzyme (35). The ability of the CCP1-CCP2-SP and SP forms of C1s to bind to the immobilized phospho-tyrosine peptide provided further evidence that a site on the SP was able to interact with the peptide, whereas the inability of any of the mutants of C1s to bind under physiological salt conditions strongly suggests that the identified positively charged exosite on the C1s SP is the site of interaction of this C4 sequence with C1s. A model of C4 containing the large loop on which the sulfotyrosine residues are located was developed, and this was docked to C1s without any prior assumptions. Interestingly, the model shows only the SP domain interacting with the substrate C4. The structure demonstrates that the middle and C-terminal sulfotyrosine residues (Tyr1420, Tyr1422) interact with the C1s binding pocket containing the K575, R576, R581, and K583 residues identified to be important for our study (Fig. 8). Importantly, binding of the sulfotyrosine loop with the exosite of C1s occurred simultaneously with binding of the C4 cleavage sequence AGLQRALEIL into the active site.

The data presented in this article provide the most detailed elucidation of the molecular details of the mechanism of interaction between C1s and C4. We identified a critical role for exosite(s) on the SP domain of the enzyme in the binding and catalysis of C4 and defined the location of one such positively charged exosite on the C1s SP domain necessary for efficient cleavage of C4 by C1s. We also provided strong evidence to suggest that this exosite of C1s binds to a stretch of negatively charged residues on C4 containing three sulfotyrosine residues and provided preliminary validation for this interaction using molecular modeling. The studies provide a firm basis to guide a full elucidation of the molecular details underpinning the kinetic mechanism by which C1s interacts with C4. It is possible that the identified exosite also plays a role in the binding of highly sulfated heparin by C1s, an interaction that is vital for full regulation of the protease by the serpin, C1 inhibitor (37, 38). Because unregulated complement activation underlies many inflammatory diseases, the data revealed in this study provide a basis for the design of therapeutic molecules for the targeted prevention of activation of complement via the classical pathway.

We thank Dr. Andrew Bradbury of Los Alamos Laboratory, U.S. Department of Energy (Los Alamos, NM) for the gift of the human anti-sulfotyrosine Ab.

This work was supported by a National Health and Medical Research Council of Australia Program Grant (490900 to R.N.P.) and the Swedish Research Council (K2009-68X-14928-06-3 to A.B.).

The online version of this article contains supplemental material.

Abbreviations used in this article:

CV

column volume

EGRck

H-Glu-Gly-Arg-chloromethylketone

Fmoc

fluorenylmethyl

kcat

catalytic efficiency

RT

room temperature

SP

serine protease.

1
Frank
M. M.
,
Fries
L. F.
.
1991
.
The role of complement in inflammation and phagocytosis.
Immunol. Today
12
:
322
326
.
2
Vegh
Z.
,
Kew
R. R.
,
Gruber
B. L.
,
Ghebrehiwet
B.
.
2006
.
Chemotaxis of human monocyte-derived dendritic cells to complement component C1q is mediated by the receptors gC1qR and cC1qR.
Mol. Immunol.
43
:
1402
1407
.
3
Stäger
S.
,
Alexander
J.
,
Kirby
A. C.
,
Botto
M.
,
Rooijen
N. V.
,
Smith
D. F.
,
Brombacher
F.
,
Kaye
P. M.
.
2003
.
Natural antibodies and complement are endogenous adjuvants for vaccine-induced CD8+ T-cell responses.
Nat. Med.
9
:
1287
1292
.
4
Pekkarinen
P. T.
,
Vaali
K.
,
Junnikkala
S.
,
Rossi
L. H.
,
Tuovinen
H.
,
Meri
S.
,
Vaarala
O.
,
Arstila
T. P.
.
2011
.
A functional complement system is required for normal T helper cell differentiation.
Immunobiology
216
:
737
743
.
5
Götze
O.
,
Müller-Eberhard
H. J.
.
1971
.
The c3-activator system: an alternate pathway of complement activation.
J. Exp. Med.
134
:
90
108
.
6
Ikeda
K.
,
Sannoh
T.
,
Kawasaki
N.
,
Kawasaki
T.
,
Yamashina
I.
.
1987
.
Serum lectin with known structure activates complement through the classical pathway.
J. Biol. Chem.
262
:
7451
7454
.
7
Sim
R. B.
,
Tsiftsoglou
S. A.
.
2004
.
Proteases of the complement system.
Biochem. Soc. Trans.
32
:
21
27
.
8
Sørensen
R.
,
Thiel
S.
,
Jensenius
J. C.
.
2005
.
Mannan-binding-lectin-associated serine proteases, characteristics and disease associations.
Springer Semin. Immunopathol.
27
:
299
319
.
9
Lewis
M. J.
,
Botto
M.
.
2006
.
Complement deficiencies in humans and animals: links to autoimmunity.
Autoimmunity
39
:
367
378
.
10
Degn
S. E.
,
Jensenius
J. C.
,
Thiel
S.
.
2011
.
Disease-causing mutations in genes of the complement system.
Am. J. Hum. Genet.
88
:
689
705
.
11
Botto
M.
,
Walport
M. J.
.
2002
.
C1q, autoimmunity and apoptosis.
Immunobiology
205
:
395
406
.
12
Dalmasso
A. P.
1992
.
The complement system in xenotransplantation.
Immunopharmacology
24
:
149
160
.
13
Sjöberg
A.
,
Önnerfjord
P.
,
Mörgelin
M.
,
Heinegård
D.
,
Blom
A. M.
.
2005
.
The extracellular matrix and inflammation: fibromodulin activates the classical pathway of complement by directly binding C1q.
J. Biol. Chem.
280
:
32301
32308
.
14
Ghai
R.
,
Waters
P.
,
Roumenina
L. T.
,
Gadjeva
M.
,
Kojouharova
M. S.
,
Reid
K. B.
,
Sim
R. B.
,
Kishore
U.
.
2007
.
C1q and its growing family.
Immunobiology
212
:
253
266
.
15
Kishore
U.
,
Ghai
R.
,
Greenhough
T. J.
,
Shrive
A. K.
,
Bonifati
D. M.
,
Gadjeva
M. G.
,
Waters
P.
,
Kojouharova
M. S.
,
Chakraborty
T.
,
Agrawal
A.
.
2004
.
Structural and functional anatomy of the globular domain of complement protein C1q.
Immunol. Lett.
95
:
113
128
.
16
Tissot
B.
,
Montdargent
B.
,
Chevolot
L.
,
Varenne
A.
,
Descroix
S.
,
Gareil
P.
,
Daniel
R.
.
2003
.
Interaction of fucoidan with the proteins of the complement classical pathway.
Biochim. Biophys. Acta
1651
:
5
16
.
17
Gaboriaud
C.
,
Thielens
N. M.
,
Gregory
L. A.
,
Rossi
V.
,
Fontecilla-Camps
J. C.
,
Arlaud
G. J.
.
2004
.
Structure and activation of the C1 complex of complement: unraveling the puzzle.
Trends Immunol.
25
:
368
373
.
18
Gaboriaud
C.
,
Rossi
V.
,
Bally
I.
,
Arlaud
G. J.
,
Fontecilla-Camps
J. C.
.
2000
.
Crystal structure of the catalytic domain of human complement c1s: a serine protease with a handle.
EMBO J.
19
:
1755
1765
.
19
Busby
T. F.
,
Ingham
K. C.
.
1990
.
NH2-terminal calcium-binding domain of human complement C1s- mediates the interaction of C1r- with C1q.
Biochemistry
29
:
4613
4618
.
20
Rossi
V.
,
Bally
I.
,
Thielens
N. M.
,
Esser
A. F.
,
Arlaud
G. J.
.
1998
.
Baculovirus-mediated expression of truncated modular fragments from the catalytic region of human complement serine protease C1s. Evidence for the involvement of both complement control protein modules in the recognition of the C4 protein substrate.
J. Biol. Chem.
273
:
1232
1239
.
21
Rossi
V.
,
Cseh
S.
,
Bally
I.
,
Thielens
N. M.
,
Jensenius
J. C.
,
Arlaud
G. J.
.
2001
.
Substrate specificities of recombinant mannan-binding lectin-associated serine proteases-1 and -2.
J. Biol. Chem.
276
:
40880
40887
.
22
Bally
I.
,
Rossi
V.
,
Thielens
N. M.
,
Gaboriaud
C.
,
Arlaud
G. J.
.
2005
.
Functional role of the linker between the complement control protein modules of complement protease C1s.
J. Immunol.
175
:
4536
4542
.
23
Beinrohr
L.
,
Murray-Rust
T. A.
,
Dyksterhuis
L.
,
Závodszky
P.
,
Gál
P.
,
Pike
R. N.
,
Wijeyewickrema
L. C.
.
2011
.
Serpins and the complement system.
Methods Enzymol.
499
:
55
75
.
24
O’Brien
G.
,
Quinsey
N. S.
,
Whisstock
J. C.
,
Pike
R. N.
.
2003
.
Importance of the prime subsites of the C1s protease of the classical complement pathway for recognition of substrates.
Biochemistry
42
:
14939
14945
.
25
Simpson
L. S.
,
Widlanski
T. S.
.
2006
.
A comprehensive approach to the synthesis of sulfate esters.
J. Am. Chem. Soc.
128
:
1605
1610
.
26
Simpson
L. S.
,
Zhu
J. Z.
,
Widlanski
T. S.
,
Stone
M. J.
.
2009
.
Regulation of chemokine recognition by site-specific tyrosine sulfation of receptor peptides.
Chem. Biol.
16
:
153
161
.
27
Li
W.
,
Adams
T. E.
,
Nangalia
J.
,
Esmon
C. T.
,
Huntington
J. A.
.
2008
.
Molecular basis of thrombin recognition by protein C inhibitor revealed by the 1.6-A structure of the heparin-bridged complex.
Proc. Natl. Acad. Sci. USA
105
:
4661
4666
.
28
Johnson
D. J.
,
Langdown
J.
,
Li
W.
,
Luis
S. A.
,
Baglin
T. P.
,
Huntington
J. A.
.
2006
.
Crystal structure of monomeric native antithrombin reveals a novel reactive centre loop conformation.
J. Biol. Chem.
218
:
35478
35489
.
29
Li
W.
,
Johnson
D. J.
,
Esmon
C. T.
,
Huntington
J. A.
.
2004
.
Structure of the antithrombin-thrombin-heparin ternary complex reveals the antithrombotic mechanism of heparin.
Nat. Struct. Mol. Biol.
11
:
857
862
.
30
Kerr
F. K.
,
O’Brien
G.
,
Quinsey
N. S.
,
Whisstock
J. C.
,
Boyd
S.
,
de la Banda
M. G.
,
Kaiserman
D.
,
Matthews
A. Y.
,
Bird
P. I.
,
Pike
R. N.
.
2005
.
Elucidation of the substrate specificity of the C1s protease of the classical complement pathway.
J. Biol. Chem.
280
:
39510
39514
.
31
Hofsteenge
J.
,
Stone
S. R.
,
Donella-Deana
A.
,
Pinna
L. A.
.
1990
.
The effect of substituting phosphotyrosine for sulphotyrosine on the activity of hirudin.
Eur. J. Biochem.
188
:
55
59
.
32
Duncan
R. C.
,
Bergström
F.
,
Coetzer
T. H.
,
Blom
A. M.
,
Wijeyewickrema
L. C.
,
Pike
R. N.
.
2012
.
Multiple domains of MASP-2, an initiating complement protease, are required for interaction with its substrate C4.
Mol. Immunol.
49
:
593
600
.
33
Dong
J. F.
,
Li
C. Q.
,
López
J. A.
.
1994
.
Tyrosine sulfation of the glycoprotein Ib-IX complex: identification of sulfated residues and effect on ligand binding.
Biochemistry
33
:
13946
13953
.
34
Farzan
M.
,
Mirzabekov
T.
,
Kolchinsky
P.
,
Wyatt
R.
,
Cayabyab
M.
,
Gerard
N. P.
,
Gerard
C.
,
Sodroski
J.
,
Choe
H.
.
1999
.
Tyrosine sulfation of the amino terminus of CCR5 facilitates HIV-1 entry.
Cell
96
:
667
676
.
35
Feng
H.
,
Sun
Z.
,
Farzan
M. R.
,
Feng
P.
.
2010
.
Sulfotyrosines of the Kaposi’s sarcoma-associated herpesvirus G protein-coupled receptor promote tumorigenesis through autocrine activation.
J. Virol.
84
:
3351
3361
.
36
Hortin
G. L.
,
Farries
T. C.
,
Graham
J. P.
,
Atkinson
J. P.
.
1989
.
Sulfation of tyrosine residues increases activity of the fourth component of complement.
Proc. Natl. Acad. Sci. USA
86
:
1338
1342
.
37
Murray-Rust
T. A.
,
Kerr
F. K.
,
Thomas
A. R.
,
Wu
T.
,
Yongqing
T.
,
Ong
P. C.
,
Quinsey
N. S.
,
Whisstock
J. C.
,
Wagenaar-Bos
I. C.
,
Freeman
C.
,
Pike
R. N.
.
2009
.
Modulation of the proteolytic activity of the complement protease C1s by polyanions: implications for polyanion-mediated acceleration of interaction between C1s and SERPING1.
Biochem. J.
422
:
295
303
.
38
Rossi
V.
,
Bally
I.
,
Ancelet
S.
,
Xu
Y.
,
Frémeaux-Bacchi
V.
,
Vivès
R. R.
,
Sadir
R.
,
Thielens
N.
,
Arlaud
G. J.
.
2010
.
Functional characterization of the recombinant human C1 inhibitor serpin domain: insights into heparin binding.
J. Immunol.
184
:
4982
4989
.

The authors have no financial conflicts of interest.