Dissecting the Regions of Virion-Associated Kaposi's Sarcoma-Associated Herpesvirus Complement Control Protein Required for Complement Regulation and Cell Binding

Abstract

Complement, which bridges innate and adaptive immune responses as well as humoral and cell-mediated immunity, is antiviral. Kaposi's sarcoma-associated herpesvirus (KSHV) encodes a lytic cycle protein called KSHV complement control protein (KCP) that inhibits activation of the complement cascade. It does so by regulating C3 convertases, accelerating their decay, and acting as a cofactor for factor I degradation of C4b and C3b, two components of the C3 and C5 convertases. These complement regulatory activities require the short consensus repeat (SCR) motifs, of which KCP has four (SCRs 1 to 4). We found that in addition to KCP being expressed on the surfaces of experimentally infected endothelial cells, it is associated with the envelope of purified KSHV virions, potentially protecting them from complement-mediated immunity. Furthermore, recombinant KCP binds heparin, an analogue of the known KSHV cell attachment receptor heparan sulfate, facilitating infection. Treating virus with an anti-KCP monoclonal antibody (MAb), BSF8, inhibited KSHV infection of cells by 35%. Epitope mapping of MAb BSF8 revealed that it binds within SCR domains 1 and 2, also the region of the protein involved in heparin binding. This MAb strongly inhibited classical C3 convertase decay acceleration by KCP and cofactor activity for C4b cleavage but not C3b cleavage. Our data suggest similar topological requirements for cell binding by KSHV, heparin binding, and regulation of C4b-containing C3 convertases but not for factor I-mediated cleavage of C3b. Importantly, they suggest KCP confers at least two functions on the virion: cell binding with concomitant infection and immune evasion.

The complement system is an important component of the innate immune response and links innate and adaptive immunity. It is comprised of a number of serum proenzymes and proteins that interact in an amplification cascade (reviewed in reference 54). Complement can be activated by specific antibody-pathogen binding (classical pathway), surfaces of pathogens (alternative pathway), or through recognition of foreign microbial carbohydrate configurations (lectin pathway). Central to this process is the cleavage of proenzymes to enable the formation of the C3 and C5 convertase enzymatic complexes, with the release of smaller chemoattractant and anaphylatoxin fragments. The covalent attachment of C4b and C3b to pathogen and infected cell surfaces also enhances recognition by phagocytes and increases the humoral response to those pathogens (12, 56). In order to protect host cells from autologous complement attack, a number of soluble and membrane-bound complement regulators have evolved to limit inflammation to the infected site.

An important group of complement regulators is encoded in the regulators of complement activation (RCA) gene cluster at chromosome 1 (locus 1q32). All of these proteins, including membrane cofactor protein (CD46), complement receptor 1 (CR1; CD35), decay-accelerating factor (DAF; CD55), factor H (FH), and C4b binding protein, contain 4 to 35 short consensus repeat (SCR) domains and share significant homology, as well as complement inhibition mechanisms (28). SCR domains consist of approximately 60 amino acids and contain conserved proline and hydrophobic residues organized into a compact hydrophobic core surrounded by short β-strands (41). This structure is stabilized by four invariant cysteine residues that form two pairs of disulfide bonds in a 1—3 and 2—4 manner.

RCA proteins regulate C3 and C5 convertases by two mechanisms (reviewed in reference 11). First, they may accelerate the decay of convertase enzyme complexes through dissociation of C2a from C4b (the components of the classical and lectin activation pathway C3 convertases) or dissociation of factor Bb from C3b (which form the alternative activation pathway C3 convertase). Secondly, they may act as cofactors to induce or enhance the cleavage of either C3b or C4b to inactive fragments incapable of forming convertases, through recruitment of the serine protease, factor I (FI). However, the precise mechanism of action for these events, including binding specificity, location of binding on C3b or C4b, and requirement for conformational changes in proteins, is still under investigation.

Since complement does not require previous antigen exposure to be fully effective and accelerates and enhances the adaptive immune responses to pathogens, it represents a potentially important antiviral immune response. Evidence in support of this contention is provided by the diversity of complement evasion strategies adopted by viruses that we are now beginning to understand. For example, not only does human cytomegalovirus upregulate the expression of host RCA complement regulators on infected cells, progeny viruses acquire these cell surface-expressed host regulators upon egress (48, 50, 51). Human immunodeficiency virus, human T-cell leukemia virus type 1, and vaccinia virus also incorporate host regulators into their envelope during egress (46-48). However, to date only two virus families, Herpesviridae and Poxviridae, have been found to encode their own complement inhibitors (reviewed in references 11 and 21).

The effects of complement on viral infection include the following: (i) lysis of infected cells and enveloped virus via formation of a pore or “membrane attack complex” in the cell or viral surface; (ii) coating of infected cells and virions with component C3b to enhance phagocytosis or block viral infection; and (iii) production of potent anaphylatoxins, which exert a variety of effects on the immune system, including the recruitment of inflammatory cells to the site of infection.

Kaposi's sarcoma-associated herpesvirus (KSHV) (16) is the etiologic agent of Kaposi's sarcoma and perhaps primary effusion lymphoma (PEL) and multicentric Castleman's disease (see reference 18). The orf4 gene of this virus encodes the lytic KSHV complement control protein (KCP), which consists of four N-terminal SCR domains, a dicysteine motif, a long serine (S)- and threonine (T)-rich domain, and a C-terminal hydrophobic element sufficient to act as a transmembrane anchor (49, 52). The full protein is 550 amino acids in length, including a 19-amino-acid signal peptide, but PEL cells naturally infected with KSHV also express two isoforms of KCP in which either the S/T region or the S/T region plus the dicysteine motif is removed through alternative splicing of the orf4 transcript. All three isoforms retain the four SCR domains and the transmembrane region, and all can regulate complement by accelerating the decay of the classical pathway C3 enzyme complex and by acting as a cofactor for factor I-mediated inactivation of C3b and C4b (49). There is no evidence for an abundant soluble form of KCP, unlike, for example, that encoded by the related gammaherpesvirus herpesvirus saimiri (4).

In the present study, we hypothesized that for maximal protection of KSHV against complement-mediated elimination, evolutionary pressure would yield KCP expression on the surfaces of cells infected de novo and on the surfaces of extracellular KSHV virions. Indeed, KCP was found on experimentally infected endothelial cells. Importantly, we demonstrate KCP is also present on the virion itself, where it may function as a ligand for virion binding to cells via its affinity with other glycosaminoglycans, such as heparan sulfate (HS), a known receptor for KSHV cell attachment (3, 8). Treating KSHV with an anti-KCP monoclonal antibody (MAb), BSF8, inhibited infection of cells by 35%. Epitope-mapping analyses revealed that this MAb binds to a region of KCP within SCR domains 1 and 2, also the region of the protein involved in heparin binding. This MAb strongly inhibited classical C3 convertase decay acceleration by KCP and factor I cofactor degradation of C4b but not C3b. Thus, our data suggest KCP is at least bifunctional, conferring upon the virion the capability to bind cells and evade complement.

MATERIALS AND METHODS

Cell culture and KSHV purification.

The immortalized human endothelial cell lines tDMVEC (36), HUVEC, and HMVEC were maintained in complete endothelial cell growth medium (EGM-2; Biowhittaker). CHO cells, obtained from the European Collection of Animal Cell Cultures (Salisbury, United Kingdom), were maintained in Dulbecco's modified Eagle's medium, supplemented with 10% fetal bovine serum, 1% L-glutamine, and 1% nonessential amino acids. CHO cells engineered to express KCP (52) were grown with hygromycin selection (400 μg ml⁻¹). The serum supplement for media was heat inactivated. KSHV was prepared from the naturally infected PEL cell lines BCBL-1c containing rKSHV.152, a recombinant virus expressing green fluorescent protein (GFP) (53), and JSC-1 (14) as described previously (35). For routine culture, BCBL1-c cell lines were grown in the presence of G-418 (250 μg ml⁻¹) to maintain the recombinant virus. PEL cell lines, SLK (25), and HEK293 cells were maintained in RPMI medium supplemented with 10% fetal bovine serum and 1% L-glutamine. Heat-inactivated KSHV was prepared by incubating virus at 60°C for 40 to 60 min.

Antibody production.

The monoclonal anti-KCP antibody BSF8 was generated in house. Briefly, BALB/c mice were immunized via the intraperitoneal route with a mixture of KCP-Fc (Fig. 1) and Freund's incomplete adjuvant. After four immunizations, spleens were removed and hybridoma cells were created with SP2/0 fusion partners via standard techniques. Positive splenocyte-myeloma fused cells were selected by flow cytometry. Anti-KCP MAb-secreting cells were identified by MAb binding to CHO cells expressing full-length KCP (52). The specificity of the anti-KCP MAb was mapped by Western blotting and enzyme-linked immunosorbent assay (ELISA) (see below) with recombinant KCP proteins (see Fig. 1). Mapping studies were also performed by flow cytometry using glycophosphotidylinositol (GPI)-anchored SCR deletion constructs expressed on CHO cells, where the Fc fusion region was replaced by the GPI sequence from human DAF (52). BSF8 was of the murine immunoglobulin G₁ (IgG₁) isotype as determined by an Isostrip typing kit (Boehringer Mannheim). Large quantities of monoclonal antibody were made by the ascites production method (purchased as a service from Eurogentec), purified by high-pressure liquid chromatography using a HiTrap protein G-Sepharose column (Amersham Biosciences, United Kingdom) according to the manufacturer's instructions, and standardized to 4-mg/ml stocks. The rabbit polyclonal antibody p0107 was raised against bacterially expressed and refolded KCP SCRs 1 and 2 (SCRs 1-2) domains (33), while rabbit polyclonal BSORF4 was raised against soluble KCP SCR 1 to SCR 4 (SCRs 1-4) domains enzymatically released from KCP-Fc (see below).

FIG. 1.

Schematic of the recombinant forms of KCP. The wild-type and SCR deletion and replacement recombinant forms of KCP constructed and studied in the present investigation are represented.

Detection of KCP on surfaces of experimentally infected cells.

Endothelial cells were plated 48 h prior to infection and inoculated with KSHV for 48 h. Immunofluorescence assay (IFA) was performed on cells detached with cell dissociation buffer (Sigma-Aldrich, United Kingdom) and applied to eight-well multitest glass slides (ICN). Cells were fixed in ice-cold acetone-methanol (1:1) for 5 min. KCP staining was detected with p0107 polyclonal anti-KCP antibody and fluorescein isothiocyanate (FITC)-conjugated antirabbit antibody (Sigma-Aldrich, United Kingdom). Cell nuclei were stained with propidium iodide (1 μg ml⁻¹), and cells were visualized with a laser-scanning microscope (Zeiss LSM).

Measuring KSHV binding to cells by flow cytometry.

Concentrated KSHV was incubated with HEK293 cells at 4°C. Unbound virus was removed by washing the cells twice (phosphate-buffered saline [PBS]-2% fetal calf serum-0.05% sodium azide). Bound virus was detected with anti-K8.1 primary antibody (a gift from B. Chandran), followed by staining with a FITC-conjugated secondary antibody (Sigma-Aldrich, United Kingdom). Cells were fixed in 1% formaldehyde and analyzed by flow cytometry (FACScalibur; Becton-Dickinson, United Kingdom). Staining with antibody to KSHV latency-associated nuclear antigen (LANA) (Novacastra, United Kingdom) served as a negative control. KCP on the surfaces of virions was detected by replacing the anti-K8.1 primary antibody with the polyclonal KCP antibody BSORF4.

Immune electron microscopy.

KSHV was applied to Parlodion-coated 200-mesh nickel grids (Agar Scientific, United Kingdom). KCP was detected by staining with monoclonal or polyclonal anti-KCP antibody. The grids were washed in PBS and bound antibody detected using antimouse or antirabbit antibody conjugated to 10-nm gold (British Biocell, United Kingdom). Preparations were negatively stained with phosphotungstic acid, pH 7 (Agar Scientific, United Kingdom), and viewed using a Jeol 100s electron microscope (80 kV; magnification, ×40,000).

Purification of recombinant KCP (KCP-Fc).

The smallest isoform of KCP (KCP Short; see Fig. 1), retaining all four SCR domains, was cloned into the pDR2ΔEF1α expression vector (a gift from Anegon, INSERM U437, Nantes, France) (17) such that the transmembrane region was replaced with part of the human IgG₁ gene as described previously (52). Constructs were transfected into CHO cells and stable clones selected in hygromycin B (Roche, United Kingdom). Cell supernatant containing KCP-Fc was stored at 4°C until protein purification. Contaminating proteins were precipitated at 40% saturation of (NH₄)₂SO₄ and removed by centrifugation. KCP was precipitated with 60% saturation of (NH₄)₂SO₄ and centrifugation (5,000 × g, 25 min). Following precipitation, the KCP-Fc pellet was redissolved in sterile PBS and purified with protein A-Sepharose (44). Purified protein was dialyzed against PBS using 12- to 14-kDa-cutoff dialysis tubing (size 9; Medicell International Ltd., United Kingdom) and concentrated using 50-kDa-cutoff filters (Millipore, United Kingdom). To obtain the Fc moiety and monomeric forms of KCP, the recombinant dimeric KCP forms were cleaved from the recombinant protein with factor Xa (Boehringer Mannheim UK Ltd.) according to the manufacturer's instructions, since a specific enzyme recognition site had been engineered between KCP and the Fc fusion protein.

Determining the contribution of KCP to KSHV infection.

HEK293 cells, plated in either 35-mm tissue culture plates (Nalgene Nunc International) or eight-well chamber slides (Nalgene Nunc International), were infected with rKSHV.152. Cells were inoculated for 2 to 4 h at 37°C and washed, and fresh medium was added. Infected cells were quantified by determining the number of GFP-expressing cells, either by fluorescence microscopy (Zeiss LSM) or flow cytometry (FACSCalibur; Becton Dickinson), 24 and 48 h postinfection. To determine the effect of antibodies to KCP upon infection, virus was first incubated with antibody at 37°C for 30 to 60 min prior to inoculation. To analyze the effect of recombinant KCP on infection, a protein of known concentration was incubated with cells at 4°C for 30 to 60 min prior to infection.

Generating KCP SCR-Fc fusion protein deletion mutants.

The following procedure generated KCP with deleted SCR 1, 1 plus 2, 1 plus 2 plus 3, 4, or 3 plus 4 (KCP Δ1, KCP Δ12, KCP Δ123, KCP Δ4, and KCP Δ34; see Fig. 1). Primers were designed for the 5′ and 3′ ends of each of the four SCR domains (see Table S1 in the supplemental material). Using different combinations of these primers and cDNA for wild-type KCP as the template, PCR products were generated in which single or multiple SCR domains were deleted. For recombinant proteins missing C-terminal SCR domains, PCR products were purified, digested with XbaI (this recognition sequence was added to the 5′ PCR primer) and NotI (the recognition sequence was added to the 3′ PCR primer), and cloned into an XbaI-NotI-digested pTorsten eukaryotic expression vector. This vector is an in-house hybrid vector that was constructed by combining the eukaryotic expression vector pDR2ΔEF1α with the human IgG₁ Fc-encoding region from the pIgPlus vector (R&D Systems, United Kingdom). It allows the in-frame addition of the hinge and Fc regions of human IgG₁ to the C termini of recombinant proteins. For the N-terminal SCR deletions, the CD33 signal sequence (from the vector sigpIgplus; R&D Systems, United Kingdom) was added in-frame to the 5′ end of the cDNA to enable expression of recombinant protein. PCR primers designed to add the CD33 signal sequence contained the restriction enzyme SpeI so that the recombinant cDNA encoded XbaI-CD33 signal-SpeI-SCR domains-NotI-BamHI-Fc region. The integrity of all recombinant cDNA was confirmed by sequencing.

Internal SCR deletions.

Deletion of SCR 2 and SCR 3 was achieved by using Δ12 KCP-Fc or Δ123 KCP-Fc sequence as a template and replacing the CD33 signal sequence with the KCP signal sequence and SCR 1 (to generate Δ2) or signal sequence and KCP SCR domains 1 and 2 (to generate Δ3), respectively. For these cDNA exchanges, CD33 signal sequence was removed by digestion with XbaI and SpeI, and PCR product containing KCP SCR 1 or SCR 1 and 2 was inserted in its place using the same restriction enzymes. The integrity of all recombinant cDNA was confirmed by sequencing (ABI sequencing systems).

Generation of KCP SCR-Fc fusion protein swap mutants.

Mutant forms of KCP were generated in which SCR domain 1, 2, 3, or 4 from human complement receptor 2 (CR2; CD21; a gift from K. Marchbank, Cardiff University) was exchanged for the corresponding SCR in KCP (see Fig. 1). Primers were designed for the 5′ and 3′ ends of each SCR domain in CR2. Domain swaps 1 and 4 were generated by digesting Δ1 and Δ4 pDR2-EF1 with XbaI/SpeI or NotI/BamHI, respectively. CR2 SCR domains 1 and 4 were generated by PCR, purified, and digested with XbaI/SpeI or NotI/BamHI, respectively. CR2 SCR domain 1 or 4 was then cloned into Δ1 KCP-Fc or Δ4 KCP-Fc to make KCP SCR domain swaps 1 and 4 (Swap1 and Swap4).

In order to make KCP Swap2 and Swap3, PCR fragments containing KCP SCR 1 or 1 plus 2 and CR2 SCR 2 or SCR 3 were generated and purified. The KCP SCR fragments were engineered to contain an EcoRI site at the 3′ end and the CR2 SCR fragments to begin with EcoRI sites at the 5′ end. These fragments were digested, ligated, and used as a PCR template to generate XbaI-KCP-EcoRI-CR2-SpeI products. These products were then used to replace the CD33 signal in Δ12 KCP-Fc or Δ123 KCP-Fc by using XbaI and SpeI enzymes and ligation to create KCP Swap2 and Swap3, respectively. For these internal SCR domain swaps, the restriction enzyme sites were removed by site-directed mutagenesis to return the entire inter-SCR domain sequence (including that from the CR2 sequence) to that of KCP using the primers described (see Table S2 in the supplemental material) and the QuikChange XL site-directed mutagenesis kit (Stratagene, United Kingdom) according to the manufacturer's instructions. The resultant SCR swap mutants contained only residues consistent with KCP sequence in the inter-SCR hinge regions, since previous modeling experiments for KCP suggested that these regions were often involved in the functional domains of KCP (33).

Measuring antibody binding to KCP by ELISA.

Monoclonal antibody epitopes were determined by ELISA. Ninety-six-well ELISA plates were coated with 0.5 or 1 μg/well of wild-type or recombinant forms of KCP-Fc, blocked with 1% bovine serum albumin in PBS or 3% fish gelatin in 50 mM Tris, 150 mM NaCl, pH 8.0, incubated with a dilution range of antibody (300 to 0.5 ng/ml), and washed, and bound antibody was detected with horseradish peroxidase-conjugated goat antimouse immunoglobulin secondary antibody (Bio-Rad Laboratories, United Kingdom). Plates were developed using o-phenylenediamine dihydrochloride substrate (Dako Corporation, United Kingdom) and read in a microplate reader using the 490-nm filter. Recombinant forms of KCP investigated included those with the SCR domain deleted and exchanged, as well as those containing one to three mutated amino acid residues (33).

Complement assays.

C1 (23), C2 (20), C4 (5), FH (10), and FI (19) were purified from plasma as described previously. C3, C3b, and C4b were purchased from Advanced Research Technologies. C1 and C2 were functionally pure, since they were devoid of other complement factors. FH, FI, C4, C3, and C3b were at least 95% pure, as determined by Coomassie staining of proteins separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). All purified proteins were stored in aliquots at −80°C. Protein concentrations were determined by measuring absorbance at 280 nm or from amino acid analysis following 24 h of hydrolysis in 6 M HCl. C3b and C4b were labeled with ¹²⁵I using the chloramine T method (24), and the specific activity was determined to be 0.4 to 0.5 MBq/μg of protein.

Factor I cofactor activity assay.

Wild-type or mutant forms of KCP-Fc (0.3 μM) were incubated with C3b (75 μg/ml) or C4b (25 μg/ml), FI (60 nM), trace amounts of ¹²⁵I-labeled C3b or C4b, and buffer (50 mM TRIS-HCl and 150 mM NaCl, pH 7.4) to a final volume of 50 μl at 37°C for 2 h. The assay was stopped by addition of SDS loading buffer containing the reducing agent dithiothreitol and proteins separated on a 10 to 15% SDS-PAGE gradient gel. Radiolabeled proteins were then visualized by autoradiography, and bands corresponding to C3b and C4b cleavage fragments were quantified by densitometry.

Inhibiting factor I cofactor activity with monoclonal antibodies.

Wild-type KCP-Fc (0.3 μM) was preincubated for 1 h with 40 μg, 4 μg, or 0.4 μg of MAb antibody BSF8 (representing molar ratios of 10:1, 1:1, and 1:10 relative to the number of copies of KCP) at room temperature. After the incubation period, this mixture was used in the C3b and C4b degradation assays described above.

Decay-accelerating activity for the classical pathway C3 convertase.

Sheep erythrocytes were washed twice with DGVB⁺⁺ (2.5 mM veronal buffer [pH 7.3], 72 mM NaCl, 140 mM glucose, 0.1% gelatin, 1 mM MgCl₂, and 0.15 mM CaCl₂), suspended at a concentration of 10⁹ cells/ml, and incubated for 20 min at 37°C with an equal volume of amboceptor to make EA cells (diluted 1:2,000 in DGVB⁺⁺; Boehring Diagnostics, Germany). EAC1 cells were made by washing the EA cells twice with ice-cold DGVB⁺⁺ and resuspending at 10⁹ cells/ml, and then C1 was added to 10¹⁰ cells drop-wise to a final concentration of 5 μg/ml and the mixture incubated with agitation for 20 min at 30°C. EAC1 cells were washed twice with ice-cold buffer and incubated with agitation for 20 min at 30°C with 1 μg/ml of C4. The resultant EAC14 cells were incubated in DGVB⁺⁺ containing C2 (5 μg/ml) for 5 min to allow formation of C3-convertase. The cells were then placed on ice for 1 min, centrifuged, and resuspended in prewarmed (30°C) DGVB⁺⁺. An equal volume of these EAC142 cells was added to a range of prediluted recombinant forms of KCP-Fc (in 50 mM Tris, pH 7.35) and allowed to incubate at 30°C with constant shaking for 5 min. One-hundred-microliter aliquots of each sample were removed and added to 100 μl of guinea pig serum diluted 1:50 in 40 mM EDTA-GVB (2.5 mM veronal buffer [pH 7.3], 72 mM NaCl, 15 mM EDTA), and resultant erythrocyte lysis was determined following incubation at 37°C for 60 min. The amount of released hemoglobin was directly proportional to the residual C3-convertase activity remaining on the EA142 cells and was measured at 405 nm in the supernatant after removal of the unlysed cells by centrifugation (1,000 × g, 4 min). Soluble recombinant DAF-Fc (49) was used as a positive control in these experiments, and all inhibitors were compared to the amount of lysis observed in the absence of added inhibitors.

Inhibiting decay-accelerating activity with MAb antibodies.

Wild-type KCP-Fc (0.3 μM in 100 μl) was preincubated for 1 h with 40 μg, 8 μg, 4 μg, 0.4 μg, or 0.04 μg BSF8 at room temperature. The amount of wild-type KCP-Fc used in these assays was determined empirically to inhibit 75% of the erythrocyte lysis in the above assay. After the incubation period, these mixtures were used to set up decay acceleration assays (as described above).

RESULTS

KCP is expressed on experimentally infected endothelial cells.

KCP is expressed on PEL cells naturally infected with KSHV when lytic replication is induced by chemical treatment (52). To determine if this protein is expressed in cells upon their de novo infection with KSHV, three endothelial cell lines were inoculated with virus concentrated from JSC-1 cells. IFA revealed KCP expression on the surfaces of these cells after 48 h and 7 days postinfection (data not shown). In these experiments, at least 50% of the inoculated cells were infected, as determined by IFA quantification of cells staining for LANA (data not shown). However, no more than 2% of the infected (LANA-positive) cells demonstrated KCP expression, commensurate with our own data that KCP is a lytic-cycle viral protein in PEL cells (52) and those of others documenting orf4 as a lytic-cycle gene in PEL cells (27, 43) and endothelial cells (40). The endothelial cells in which KCP expression was detected exhibited degenerate morphology, consistent with the cytopathic effect induced by lytic virus replication. Attempts to increase the extent of lytic replication, and thus increase the number of cells expressing KCP, by treatment with sodium butyrate or phorbol esters were unsuccessful (data not shown) and confirm results in previous studies (6).

KCP expression on the virion surface.

To determine if KCP is also present on KSHV virions, a virus-cell binding assay was established in which permissive HEK293 cells were inoculated with KSHV and viral proteins on the cell surface were quantified by flow cytometry after staining with appropriate antibodies. Thus, KSHV was incubated with HEK293 cells at 4°C and unbound virus removed by washing. Confirmation that virions had bound was provided by staining with antibody to K8.1, a viral glycoprotein known to be expressed on the virion surface (Fig. 2A, upper panel) (15, 58). KCP was also detected in this assay after staining with specific antibodies (Fig. 2A, middle panel), suggesting the presence of KCP on the surface of KSHV. As expected, staining of the cells with antibody to non-virion-associated LANA failed to generate a substantial signal (Fig. 2A, lower panel).

FIG. 2.

Detection of KCP on the surface of KSHV. (A) KCP is detectable on the surfaces of cells inoculated with KSHV. KSHV was incubated with HEK293 cells at 4°C. Unbound virus was removed by washing, and any remaining virus was detected by flow cytometry with antibody to either glycoprotein K8.1 (positive control, upper panel) or KCP (middle panel). Antibody to non-virion-associated LANA (lower panel) served as the negative control. FITC-conjugated secondary antibody detected primary antibody binding, and the fluorescence intensities of cells were measured (FL1-H). Cells incubated in the absence or presence of KSHV are represented. (B) KCP detection on virions by immune electron microscopy. KSHV was applied to nickel grids and stained with anti-KCP antibody, followed by gold-conjugated secondary antibody and finally negative staining. KCP is localized to the surfaces of KSHV virions (from JSC-1 cells) stained with anti-KCP MAb (left panel) or polyclonal (right panel) antibody. Gold is specifically associated with virions and not other material (labeled with an asterisk; left panel). No association of gold particles was observed with virions in the absence of primary antibody (data not shown). Each scale bar represents 200 nm.

To obviate the possibility that the virus-cell binding assay was detecting nonspecific binding of the anti-KCP antibody, for example, by membranous debris embedded with viral proteins, immune electron microscopy studies were performed on virions. KSHV was concentrated from PEL cell supernatants and stained with either polyclonal (BSORF4) or MAb (BSF8) anti-KCP antibody (52). KCP localized strongly to both enveloped virions (Fig. 2B) and light particles but not to nonenveloped capsids (data not shown). Labeling was specific to the viral surface, since there was negligible evidence of gold-conjugated secondary antibodies interacting with other protein aggregates within the sample (Fig. 2B, left panel). KSHV virions of both JSC-1 (Fig. 2B) and KSHV.152 (data not shown) demonstrated comparable staining patterns with both MAb and polyclonal anti-KCP antibodies.

Recombinant KCP binds to heparin.

HS is a known receptor for KSHV cell attachment (3), interacting with K8.1 (8, 55) and glycoprotein B (1). We determined whether KCP might bind polyanions like heparin, a glycosaminoglycan closely related to HS, and HS itself, facilitating KSHV infection. Indeed, heparin binding by viral complement regulatory proteins is not unprecedented (22, 39) and has been shown for KCP (34). Heparin binding proteins are often also able to bind HS (31), and KCP is able to interact with HS-expressing CHO cells but not with CHO cells lacking expression of glycosaminoglycans (34). We confirmed our previous observations that KCP-Fc (Fig. 1) bound specifically to heparin in a heparin-agarose column, eluting at 0.2 to 0.3 M NaCl; Fc alone was applied to the column and did not bind (data not shown). Our K64Q/K65Q/K88Q KCP mutant lacking complement regulatory activity (33) demonstrated reduced binding compared to the wild-type protein, with most of the mutant protein being removed by washing the column prior to NaCl elution and residual bound mutant protein eluting in 0.1 to 0.2 M NaCl (data not shown).

Purified KCP-Fc and anti-KCP antibodies suppress infection of 293 cells.

Since KCP is present on the virion and binds heparin, we determined whether it contributes to cell infection by KSHV. HEK293 cells were inoculated with rKSHV.152 to enable quantification of infection. Treatment with either soluble recombinant KCP-Fc or antibodies to KCP reduced infection of these cells in a dose-dependent manner (Fig. 3). Thus, soluble KCP-Fc inhibited infection, by >50%, at a concentration of 500 μg ml⁻¹ but not 100 μg ml⁻¹. Both MAb (BSF8) and polyclonal antibody (p0107) inhibited infection by at least 35% and 20%, respectively. This inhibition of infection, with either recombinant protein or antibodies, was statistically significant compared to the positive control of no inhibitory protein added (P ≤ 0.05; two-sample t test). The block to infection with heparin in our assay is not as extensive as that shown by others (3, 8), but it is still statistically significant. Presumably methodology with our assay, such as cell type differences and our quantification of recombinant KSHV infection by flow cytometric measurement of GFP expression, as opposed to quantification by IFA, accounts for this quantitative difference.

FIG. 3.

Recombinant KCP and anti-KCP antibodies can reduce KSHV infection of HEK293 cells. Cells were infected with rKSHV.152 (53) in the presence of either recombinant KCP-Fc or anti-KCP antibody. The effect on infection levels was determined by flow cytometry, measuring the mean fluorescence intensities (of cells 48 h postinfection). The figure shows the percentages of infection compared to the positive control (no antibody or recombinant protein added) from a single experiment performed in duplicate but is representative of at least four experiments. BSF8 is an anti-KCP monoclonal antibody; p0107 is a purified polyclonal antibody raised against KCP SCR domains 1 and 2. Buffer alone had no effect on infection (data not shown). Heat-inactivated virus served as a negative control for infection. Statistically significant inhibition of infection compared to the positive control is indicated (*, P < 0.05, two-sample t test). Error bars represent ±1 standard error of the mean.

Functional inhibition of KCP by MAb BSF8.

KCP regulates complement by dissociating the classical and lectin C3 convertases (comprised of the proteins C4b and C2a) with relative ease (49) but not the alternative complement C3 convertases (formed from C3b and Bb). KCP can also regulate complement by acting as a cofactor for the factor I-mediated degradation of the C4b component of the C3 convertases and of the C3b component of the C5 and alternative pathway C3 convertases (49). An overview of these pathways, including flow diagrams, was published recently (11).

In order to determine if the domains of KCP responsible for virus-mediated cell binding were shared with those of complement regulation, KCP function-blocking studies were performed with MAb BSF8, the antibody that suppressed KSHV infection by up to 35% (Fig. 3). Functional studies revealed that BSF8 completely blocked the ability of KCP-Fc to dissociate the classical and lectin pathway C3 convertases at a molar ratio of 2:1 (Fig. 4A).

FIG. 4.

Inhibition of KCP complement regulatory activity by preincubation of wild-type KCP-Fc with MAb BSF8. (A) Inhibition of classical C3 convertase decay acceleration. The data from two representative experiments (1 and 2) are shown where the antibodies are added at increasing molar ratios relative to the concentration of KCP-Fc. BSF8 completely inhibited KCP activity in this assay at a molar ratio of antibody:KCP-Fc of at least 2 (▴), whereas control antibody () did not. Assays were performed in triplicate and repeated at least twice. Error bars represent standard deviations. (B) Inhibition of C4b degradation cofactor assay. KCP-Fc was incubated with 40, 4, or 0.4 μg/ml of antibody representing molar ratios of 10:1, 1:1, and 1:10 prior to incubation with radiolabeled C4b and factor I. Factor I cofactor activity was measured as the appearance of the C4d fragment visualized by autoradiography following separation by SDS-PAGE under reducing conditions (upper panel). The relative inhibition of KCP activity by antibodies was determined by quantifying the intensity of the C4d band by densitometry (lower panel) compared to the activity of KCP-Fc in the absence of antibody (not shown). One hundred percent C4d is the amount of cleavage product obtained in the absence of antibody. C4b is composed of three polypeptides, α (93 kDa), β (75 kDa), and γ (33 kDa), labeled on the upper panel. When C4 is activated, a short fragment (C4a) is cleaved from the α chain, yielding C4b, and this α chain can then be degraded by factor I to C4d. (C) Inhibition of C3b degradation cofactor assay. KCP-Fc was incubated with 40, 10, or 0.4 μg/ml of antibody representing molar ratios of 10:1, 1:1, and 1:10 prior to incubation with radiolabeled C3b and factor I. Factor I cofactor activity was measured by the appearance of three iC3b fragments (68, 46, and 43 kDa) visualized by autoradiography following their separation by SDS-PAGE under reducing conditions (upper panel). The relative inhibition of KCP activity by antibodies was determined by quantifying the intensity of the 68-kDa band of the iC3b product of C3b cleavage (lower panel), compared to the activity of KCP-Fc in the absence of antibody (not shown). One hundred percent iC3b is the amount of cleavage product obtained in the absence of antibody. C3b is composed of two polypeptides, α (119 kDa) and β (75 kDa), labeled on the upper panel. When C3 is activated, a short fragment (C3a) is cleaved from the α chain, yielding C3b, and this α chain can then be degraded by factor I to iC3b.

KCP can also regulate complement by acting as a factor I cofactor for the degradation of C4b and C3b. At a molar ratio of 10:1 (antibody to KCP-Fc), BSF8 decreased C4b cleavage by 75% (Fig. 4B) but reduced KCP-induced cleavage of C3b by only 25% (Fig. 4C). These data were determined by densitometric quantification of C4d (the degradation product of C4b cleavage by factor I; Fig. 4B, lower panel) and iC3b (the 68-kDa product of C3b cleavage by factor I; Fig. 4C, lower panel).

Mapping the region of KCP to which MAb BSF8 binds.

Since MAb BSF8 suppressed KSHV infection of HEK293 cells (Fig. 3), KCP dissociation of the classical C3 convertase (Fig. 4A), as well as C4b (Fig. 4B), but not C3b (Fig. 4C) cleavage by factor I, the data suggested that cell binding and certain complement regulatory activities can be topologically dissociated. We therefore mapped the region of KCP to which BSF8 binds with domain deletion and insertion mutants of KCP.

Recombinant forms of KCP-Fc were generated in which individual or multiple SCR domains were deleted or in which individual SCR domains were exchanged for the equivalent SCR from CR2 (Fig. 1). Previously, we have shown that KSHV expresses three isoforms of KCP, all of which are equivalent in complement-regulating function (49, 52). Essentially, these isoforms vary in the amount of the KCP serine and threonine-rich stalk that they carry between the region encompassing the four SCR domains and the transmembrane domain (Fig. 1). Therefore, all of the recombinant proteins of the present study were based on the shortest form of KCP (KCP Short; Fig. 1). Furthermore, the inter-SCR spacing was standardized to four amino acid residues between cysteines of adjacent SCR domains, since this distance is consistent with that found in wild-type KCP.

The apparent molecular masses of the different recombinant forms of KCP with SCR deletions are shown in Fig. 5A and B, while those of recombinant KCP-Fc with SCR domains swapped with those of CR2 are shown in Fig. 5C. The 2- to 3-kDa variation between single SCR domain deletion and swap forms of KCP-Fc can be accounted for by minor variations in presumptive N-linked glycosylation of KCP and CR2, as predicted using the NetNGlyc 1.0 server website (www.cbs.dtu.dk/services/NetNGlyc/). However, constructs involving the deletion or exchange of SCR 4 also lack adjacent 48 amino acid residues of the KCP stalk (Fig. 1), which were predicted to contain a further two N glycosylation sites, accounting for the larger changes in apparent molecular mass relative to other exchanged/deleted domains. The recombinant KCP forms missing SCR 1, SCR 4, or SCRs 3 and 4 were also expressed on the cell surfaces of CHO cells by exchanging the recombinant Fc fusion tag for the GPI anchor signal from human DAF to enable epitope mapping by flow cytometry.

FIG. 5.

Visual analyses of recombinant KCP proteins. Western blot analysis (A and C) or silver staining (B) of recombinant KCP-Fc proteins missing multiple (A) or single (B) SCR domains or recombinant forms where single SCR domains have been exchanged with CR2 (KCP Swap1-Fc to Swap4-Fc). Numbering indicates which SCR domains have been deleted, as identified by the symbol Δ (see Fig. 1). All proteins were separated by SDS-PAGE under reducing conditions.

Initial dot blot analyses with deletion and swap forms of KCP suggested that BSF8 bound to this protein within the SCR 1-2 domains (Table 1). The ability of BSF8 to bind CHO cells expressing the entire KCP cDNA or recombinant GPI-anchored forms of SCRs 2-4, SCRs 1-3, or SCRs 1 and 2 was then quantified by flow cytometry (Fig. 6A). Smaller GPI-anchored recombinant forms have higher levels of expression on the cell surface, and BSF8 binding to all but one construct was saturated at 1 mg/ml (Fig. 6A). BSF8 bound less well to KCP missing SCR 1, indicating partial loss of the binding epitope. This reduced binding was confirmed by ELISA analyses (Fig. 6B). The extent of binding of BSF8 to KCP-Fc constructs with SCR exchanges for SCR 3 or 4 (Fig. 6B) or deleted for SCR 3 or 4 (data not shown) was not affected. Furthermore, removing KCP SCR 2, either through exchange with CR2 SCR 2 (Fig. 6B) or deletion (not shown), abrogated BSF8 binding. In the ELISA mapping of BSF8, we also utilized recombinant forms of KCP-Fc that we previously generated to contain two or three amino acid substitutions (Fig. 6C) (33). Binding of BSF8 was identical to that of nonmutated KCP-Fc for all point mutants except those exchanging either arginine residues at positions 20, 33, and 35 or lysine residues at 64, 65, and 88 to glutamine residues. The first set of amino acid mutations is located in SCR 1, while the second set is located in SCR 2 and the hinge region between SCRs 1 and 2. However, all six of these residues are predicted to contribute to a continuous positive groove on one face of the KCP molecule (33), consistent with them also participating in a single MAb conformational epitope.

TABLE 1.

Summary of dot blot analysis for mapping MAb BSF8 binding to KCP^a

KCP fusion protein	Binding of BSF8
Wild-type KCP-Fc	+
KCP(Δ1)-Fc	−
KCP(Δ12)-Fc	−
KCP(Δ123)-Fc	−
KCP(Δ4)-Fc	+
KCP(Δ34)-Fc	+
KCP(Δ14)-Fc	−
KCP(Δ2)-Fc	−
KCP(Δ3)-Fc	+
Swap1-Fc	−
Swap2-Fc	−
Swap3-Fc	+
Swap4-Fc	+

^aResults are summarized from a minimum of three repeated analyses. Five micrograms of each purified recombinant protein was dried on a nitrocellulose membrane prior to blocking, incubating with monoclonal antibody detecting bound antibody with horseradish peroxidase-conjugated goat antimouse immunoglobulin. Binding to wild-type KCP-Fc provided the positive control, and lack of binding to an irrelevant Fc fusion protein served as the negative control. Positive results represented a strong signal; negative results represent negligible signal. See Fig. 1 and the text for an explanation of swap mutants.

FIG. 6.

Mapping the region of KCP to which MAb BSF8 binds. (A) Mapping by flow cytometry. Mean cellular fluorescence from BSF8 binding to CHO cells expressing full-length wild-type KCP cDNA (░⃞) or recombinant GPI-anchored forms lacking SCR 1 (•), SCR 4 (▪), or SCRs 3 and 4 (▴) are shown relative to input concentration of antibody. All assays were performed in triplicate. (B) Mapping by ELISA. ELISA plates were coated with full-length KCP (░⃞) or recombinant forms where SCR 1 was deleted (▪) or KCP SCR domains were swapped for equivalent CR2 domains: SCR 2 (•), SCR 3 ( ), or SCR 4 (▴). Binding of antibody relative to input concentration of antibody is shown. All assays were performed in duplicate. (C) Mapping by ELISA with point mutants of KCP. Same assay as shown in B, except that plates were coated with full-length KCP (░⃞) or point mutant forms of KCP, R20Q/R33Q/R35Q, D21A/D23A/E28A, K64Q/K65Q/K88Q, M113A/M120A, K131Q/K133Q/H135Q, R136Q/K138Q, H158A/H171A/H213A, E99Q/E152Q/D155Q, F195A/F207A/F209A, from a previous study. These assays were performed in duplicate.

Thus, MAb BSF8, which is capable of inhibiting KSHV infection of HEK293 cells by up to 35%, binds to the SCR 1 and 2 domains of KCP, perhaps within the region between these domains and defined by the K64Q/K65Q/K88Q mutant.

Complement regulation by recombinant forms of KCP.

BSF8, the anti-KCP MAb, which binds to KCP SCR domains 1 and 2 (Fig. 6), inhibited KSHV infection and suppressed the complement-regulating function of KCP: it completely blocked the ability of KCP-Fc to dissociate the classical and lectin pathway C3 convertases (Fig. 4A) and decreased C4b cleavage by 75% (Fig. 4B). To verify the involvement of those structural domains imposing complement regulation by KCP, suggested from the function-blocking studies with BSF8, functional studies of the recombinant KCP forms were performed.

Recombinant KCP forms missing SCR 1, 2, or 3, either by deletion or SCR exchange with CR2, completely lost their ability to regulate the classical and lectin complement activation pathways as assessed by lack of acceleration of the decay of the C3 convertase components, C4b and C2a (Fig. 7). On the other hand, removal of SCR 4, by deletion or exchange, had no effect on decay acceleration of this C3 convertase (Fig. 7).

FIG. 7.

Comparison of acceleration of the classical C3 convertase decay by wild-type and KCP-Fc recombinant proteins. Purified complement components were used to deposit the C4b2a C3 convertase on the surface of sheep erythrocytes, which are incubated with putative inhibitors before allowing the complement attack to proceed to completion and lysis. Relative abilities of recombinant KCP proteins to decrease target cell lysis are shown compared to that of wild-type KCP. (A) KCP mutants with deleted SCR domains. (B) KCP mutants with SCR domains exchanged with CR2. SCR domains have been deleted or swapped for SCR 1 (▪), SCR 2 (▴), SCR 3 (•), or SCR 4 (⧫); wild-type KCP as an Fc fusion protein, ░⃞. All assays were performed in triplicate, and the analysis was repeated at least twice. Error bars represent standard deviations. One hundred percent lysis refers to lysis without the addition of inhibitor.

The ability of the recombinant KCP forms to mediate factor I cleavage of C4b and C3b was also investigated. Only SCR 4 was found to be dispensable for cofactor activity in the cleavage of C4b to C4d (Fig. 8). However, none of the single SCR-deleted or -exchanged recombinant forms of KCP could mediate the factor I cleavage of C3b (data not shown), which was found for KCP-Fc containing all four SCR domains (49). SCR 1 was indispensable for the decay-accelerating activity (Fig. 7) and degradation of C3b or C4b cleavage (Fig. 8; also data not shown). We also found that removal of SCR 1 abrogated the ability of KCP to bind heparin (data not shown), indicating that the heparin binding site is distal to the complement component binding region. The deleted or swapped recombinant KCP forms determined SCR 4 as being required only for degradation of C3b by factor I.

FIG. 8.

C4b degradation cofactor assay for wild-type KCP-Fc compared to recombinant KCP proteins. Radiolabeled C4b was incubated with test cofactors and factor I for 2 h before being separated by SDS-PAGE under reducing conditions and visualized by autoradiography. Top panel, C4b degradation when single SCR domains were deleted from KCP (Δ). Bottom panel, C4b degradation when SCR domains from CR2 were substituted for those of KCP. C4b degradation was represented by the appearance of the C4d fragment of the alpha chain but is not affected by the variation in lane loading intensity, since it is a qualitative, not quantitative, assay. No degradation of C4b was seen in the absence of factor I (fI) with full-length KCP (KCP, no fI). The bands labeled α, β, and γ are described in the legend to Fig. 4B.

DISCUSSION

KCP is expressed during KSHV lytic cycle replication on the surface of naturally infected PEL cells (52) and the complement-regulating function is localized to the four SCR domains (49). Specifically, regulation of the classical pathway of complement, through decay acceleration and factor I cleavage of C4b, requires a cluster of positively charged amino acids in the SCR 1 to SCR 2 domain, as well as positively and negatively charged regions of SCRs 2 and 3. The regulation of the alternative pathway overlaps with classical pathway regulatory sites but requires more C-terminal residues in SCRs 3 and 4 (33).

In the present study, we found KCP to be expressed on the surfaces of virions (Fig. 2) and de novo-infected endothelial cells (data not shown), potentially helping to protect them from complement-mediated lysis. Moreover, anti-KCP antibody blocking studies revealed that MAb BSF8 repeatedly reduced infection of HEK293 cells by ca. 35% (Fig. 3), suggesting that the virion location of KCP contributes to KSHV infection.

Consistent with our hypothesis that KCP facilitates KSHV infection, KCP binds heparin in vitro, and our K64Q/K65Q/K88Q KCP mutant has substantially reduced heparin-binding activity. In this mutant, the region of positively charged lysine residues between SCR domains 1 and 2 was substituted with uncharged glutamine residues. Heparin and HS are glycosaminoglycans: polysaccharides that are highly negatively charged due to the presence of sulfate groups on the carbohydrate backbone. Ionic interactions are therefore important in mediating protein binding to glycosaminoglycans. The reduced binding of the KCP mutant to heparin indicates that this positively charged region of KCP is involved in the interaction with heparin, and by extrapolation HS (31), via electrostatic bonds. Indeed we have found previously that the K64Q/K65Q/K88Q mutant does not interact with CHO cells in contrast to wild type KCP (34). Consistent with these findings, MAb BSF8 did not bind either the R20Q/R33Q/R35Q KCP mutant or the K64Q/K65Q/K88Q mutant, both of which are deficient in heparin binding ability (34). Our data implicate KCP in binding to cells via polyanions like HS and indicate that the region of KCP linking SCR domains 1 and 2 is involved in this interaction. Others have speculated that the heparin-binding region of KCP maps to the first SCR domain but presented no data in this regard (38). Therefore, KCP is one of a growing list of KSHV virion-associated proteins to play a role in virus-cell interactions; K8.1 and gB also bind HS (1, 3, 8, 55), and gB additionally binds a KSHV cell entry receptor, integrin α3β1 (2).

Previous mass spectrometry (MS) determination of the protein content of KSHV virions by two groups did not identify KCP in virions (7, 57). We attribute this discrepancy with our own data either to the increased sensitivity of electron microscopic evaluation over that of MS or to the loss of KCP from the virion surface during the purification procedures employed in these studies. In support of our present findings, two groups have independently identified the KCP homologue of rhesus rhadinovirus by MS on the surface of this nonhuman primate virus (42; Scott Wong, personal communication).

During our MAb BSF8 epitope mapping studies, we determined the location of complement binding and regulation within the SCR domains. We found that only the three N-terminal SCR domains were required for KCP classical pathway decay-accelerating activity, since KCP Δ4-Fc and KCP Swap4-Fc had activity equivalent to that of wild-type KCP-Fc (Fig. 7). Furthermore, KCP domains SCRs 1-3 were sufficient to retain the cofactor activity for factor I-mediated degradation of C4b (Fig. 8). Decay-accelerating activity (Fig. 4A) and factor I-mediated degradation of C4b (Fig. 4B) were inhibited by MAb BSF8, but poor inhibition of factor I-mediated degradation of C3b (Fig. 4C) by BSF8 suggests that different structural aspects of SCR domains 1 to 3 participate in these activities.

SCR domain deletion/swapping and function-blocking antibodies have previously been used to map the complement binding and regulatory regions of human RCA gene cluster proteins. Complement receptor 1 (CD35) and factor H, which each consist of 20 to 37 SCR domains, were found to contain multiple C3b and/or C4b binding sites that varied in their decay-accelerating and cofactor activities (reviewed in references 30 and 59). C4b-binding protein has eight SCR domains, but C4b binding and complement regulation required only the three N-terminal SCR domains (9). Decay-accelerating factor (CD55) and membrane cofactor protein (CD46) are similar to KCP in that they have only four SCR domains. CD55 lacks cofactor activity, and SCRs 2-3 were required for decay acceleration of the classical C3 convertase, while SCRs 2-4 were required for decay of the alternative C3 convertase (13). Interestingly, the best fit sequence identity between KCP and CD55 (as determined by two-sequence BLAST at www.ncbi.nlm.nih.gov/BLAST/) aligns KCP SCRs 1-3 to CD55 SCRs 2-4 (41% identity, 55% similarity), consistent with the dispensable nature of CD55 SCR 1 and KCP SCR 4 for decay-accelerating activity (reference 13 and Fig. 8, respectively). However, unlike CD55, KCP does not have significant decay-accelerating activity for the alternative C3 convertase. CD46 also lacks decay-accelerating activity, and SCRs 3 and 4 were required to bind C3b, while SCRs 2 to 4 were required for C4b and C3b cofactor activities (13).

Of the viral complement regulators composed of SCR domains, investigation of the structural requirements has been performed only for KCP (33, 38) and the vaccinia virus complement control protein (VCP). Rosengard et al. (45) utilized recombinant chimeric molecules composed of SCR domains from CR2 (CD21) and VCP. Isaacs et al. (26) mapped the functional domains of VCP by using SCR-specific function-blocking monoclonal antibodies, and Mullick et al. identified complement regulatory domains with deletion mutants (37). The complement binding and complement-regulating regions of the herpes simplex virus type 1 glycoprotein C have also been investigated, both in vitro and in vivo (29, 32). However, this protein does not contain SCR domains.

In conclusion, KCP is located on the virion surface, where it participates in virus binding to cells. This binding occurs via a region of the protein also important for classical complement regulation, as determined by inhibition of both activities by MAb BSF8 (i.e., cell infection and accelerated decay of C3 convertase). Our epitope mapping of the BSF8 recognition site and competitive inhibition studies show that cell binding is likely mediated through a heparan sulfate binding region, delineated by a cluster of six positively charged amino acids (R20, R33, R35, K64, K65, and K88) spanning SCR domains 1 and 2. While SCR 1 removal abrogated heparin binding and C4b-regulating function, it does not abrogate C4b binding (data not shown), indicating that the role of SCR 1 in complement regulation is not necessarily via complement component binding. KCP contains only four SCR domains, yet it combines the regulatory mechanisms of both CD55 and CD46 within the same relative size. Studies to address further the role of KCP on the surfaces of virions in the context of KSHV biology are in progress with recombinant viruses.