Identification of hot spots in the variola virus complement inhibitor (SPICE) for human complement regulation

Abstract

Variola virus, the causative agent of smallpox, encodes a soluble complement regulator named SPICE. Previously, SPICE has been shown to be much more potent in inactivating human complement than the vaccinia virus complement control protein (VCP), although they differ only in 11 amino acid residues. In the present study, we have expressed SPICE, VCP, and mutants of VCP by substituting each or more of the 11 non-variant VCP residues with the corresponding residue of SPICE to identify hot spots that impart functional advantage to SPICE over VCP. Our data indicate that (i) SPICE is approximately 90-fold more potent than VCP in inactivating human C3b, and the residues Y98, Y103, K108 and K120 are predominantly responsible for its enhanced activity; (ii) SPICE is 5.4-fold more potent in inactivating human C4b, and residues Y98, Y103, K108, K120 and L193 mainly dictate this increase; (iii) the classical pathway decay-accelerating activity of activity is only twofold higher than that of VCP, and the 11 mutations in SPICE do not significantly affect this activity; (iv) SPICE possesses significantly greater binding ability to human C3b compared to VCP, although its binding to human C4b is lower than that of VCP; (v) residue N144 is largely responsible for the increased binding of SPICE to human C3b; and (vi) the human specificity of SPICE is dictated primarily by residues Y98, Y103, K108, and K120 since these are enough to formulate VCP as potent as SPICE. Together, these results suggest that principally 4 of the 11 residues that differ between SPICE and VCP partake in its enhanced function against human complement.

FIG. 1.

Space-filling model of SPICE and sequence comparison of SPICE and VCP depicting the 11 mutations in SPICE compared to VCP, and SDS-PAGE analysis of purified recombinant SPICE, VCP, and various substitution mutants of VCP. (A) Front and back faces of SPICE model built by utilizing crystal structure of VCP (Ig40) (36) as the template using the program SWISS-MODEL (12, 40, 47). The 11 residues that differ between SPICE and VCP are labeled and depicted in red. (B) Sequence alignment of VCP and SPICE identifying CCP domains and linkers. Residues highlighted in red and indicated by asterisks represent those that differ between SPICE and VCP. These residues were mutated in VCP to generate SPICE, 11 single-amino-acid substitution mutants, and tetra- and pentamutants. (C) SDS-PAGE analysis of purified proteins. Purified VCP, SPICE, single-point, and tetra- and pentamutants of VCP were analyzed by SDS-12% PAGE under reducing conditions and stained with Coomassie blue. Molecular weight (MW) was determined by SDS-PAGE: VCP, 29,000; SPICE, 27,000; Q77H, 29,000; H98Y, 28,500; S103Y, 29,000; E108K, 28,500; E120K, 27,500; S131L, 28,000; E144N, 28,000; D178N, 29,000; S193L, 29,000; K214T, 29,000; K236Q, 28,000; tetramutant, 27,500; pentamutant, 27,500. Molecular mass is expressed in kilodaltons in the figure.

FIG. 2.

Time course analysis of factor I cofactor activity of SPICE, VCP, and the single-amino-acid substitution mutants of VCP for human complement protein C3b. Cofactor activity was determined by incubating human C3b with VCP, SPICE, or the mutants in the presence or absence of human factor I. The reactions were stopped after the indicated time periods by adding sample buffer containing dithiothreitol, and C3b cleavages were visualized by subjecting the samples to SDS-PAGE analysis on a 9% gel. During C3b cleavage, the α′-chain was cleaved into N-terminal 68-kDa and C-terminal 46-kDa fragments. The C-terminal fragment is subsequently cleaved into a 43-kDa fragment. Generation of these fragments indicates the inactivation of C3b. The amount of α′-chain remaining was assessed by measuring its intensity by densitometric analysis and is represented graphically in the lower panel.

FIG. 3.

Time course analysis of factor I cofactor activity of SPICE, VCP, and the single-amino-acid substitution mutants of VCP for human complement protein C4b. Cofactor activity was determined by incubating human C4b with VCP, SPICE, or the mutants in the presence or absence of human factor I. The reactions were stopped after the indicated time periods by adding sample buffer containing dithiothreitol, and C4b cleavages were visualized by subjecting the samples to SDS-PAGE analysis on an 11% gel. During C4b cleavage, the α′-chain is cleaved into N-terminal 27-kDa, C-terminal 16-kDa, and central C4d fragments; generation of these fragments indicates the inactivation of C4b. The amount of α′-chain left was assessed by measuring its intensity by densitometric analysis and is represented graphically in the lower panel.

FIG. 4.

Analysis of CP C3-convertase decay-accelerating activity of SPICE, VCP, and the single-amino-acid substitution mutants of VCP. The CP C3 convertase C4b,2a was assembled on EA by sequentially incubating the cells with human C1, C4 and C2 (Calbiochem). The C3 convertase was allowed to decay by incubating EA-C4b,2a with various concentrations of VCP, SPICE, or the mutants for 5 min at 22°C, and the remaining enzyme activity was assessed by measuring cell lysis after incubation of the cells with guinea pig sera diluted 1:100 in DGVB containing 40 mM EDTA for 30 min at 37°C. The data obtained were normalized by setting lysis that occurred in the absence of an inhibitor (VCP or SPICE or the mutants) as 100% lysis.

FIG. 5.

SPR analysis of SPICE, VCP, and the single-amino-acid substitution mutants of VCP. (A) Binding of SPICE, VCP, and the mutants to human C3b. (B) Binding of SPICE, VCP, and the mutants to human C4b. SPICE, VCP, and the mutants were oriented onto the NTA sensor chip, and C3b (1 μM) or C4b (250 nM) was flown over the chip to measure binding.

FIG. 6.

Time course analysis of factor I cofactor activity of SPICE and the tetra- and pentamutants of VCP for human complement proteins C3b and C4b. The cofactor activity was determined by incubating human C3b or C4b with SPICE, tetramutant (H98Y, S103Y, E108K, and E120K), or pentamutant (H98Y, S103Y, E108K, E120K, and S193L) of VCP in the presence or absence of human factor I. The reactions were stopped after the indicated time periods by adding sample buffer containing dithiothreitol. C3b and C4b cleavages were visualized by subjecting the samples to SDS-PAGE analysis on 9 and 11% gels, respectively. The amount of α′-chain remaining was assessed by measuring its intensity by densitometric analysis and is represented graphically. The upper panels show the cleavage of C3b by tetramutant and pentamutant, and the lower panels depict the cleavage of C4b.

FIG. 7.

SPR analysis of SPICE and the tetra- and pentamutants of VCP. (A) Binding of SPICE, tetramutant (H98Y, S103Y, E108K, and E120K), and pentamutant (H98Y, S103Y, E108K, E120K, and S193L) to human C3b. (B) Binding of SPICE, tetramutant, and pentamutant to human C4b. SPICE and the mutants were oriented onto the NTA sensor chip, and C3b (1 μM) or C4b (250 nM) was flown over the chip to measure binding.

FIG. 8.

Inhibition of human and dog complement by SPICE, VCP, and the tetra- and pentamutants of VCP. Inhibition of human (A) and dog (B) complement by SPICE, VCP, or tetramutant (H98Y, S103Y, E108K, and E120K), and pentamutant (H98Y, S103Y, E108K, E120K, and S193L) of VCP was studied my measuring their effect on the alternative pathway-mediated lysis of rabbit erythrocytes (E_R).