protection from complement-mediated cell lysis - CiteSeerX

Cloning and expression of the complementreceptor glycoprotein C from Herpesvirussimiae (herpes B virus): protection fromcomplement-mediated cell lysis

Hartwig P. Huemer,1,2 Christian Wechselberger,23 Alice M. Bennett,4

Dietrich Falke3 and Lesley Harrington4

Correspondence

Hartwig Huemer

[email protected]

1Institute for Hygiene and Social Medicine, University of Innsbruck, Fritz-Pregl-Str. 3, A-6020Innsbruck, Austria

2Institute of Molecular Biology, Austrian Academy of Sciences, Salzburg, Austria

3Institute of Virology, University of Mainz, Germany

4Department of Microbiology, CBD Porton Down, Salisbury, UK

Received 4 November 2002

Accepted 17 January 2003

Simian herpes B virus (SHBV) is the herpes simplex virus (HSV) homologue for the species

Macaca. Unlike in its natural host, and unlike other animal herpesviruses, SHBV causes high

mortality in accidentally infected humans. SHBV-infected cells, like those infected with HSV-1 and

equine herpesvirus types 1 and 4, express complement C3 receptor activity. To study

immunoregulatory functions involved in susceptibility/resistance against interspecies transmission,

the SHBV glycoprotein C (gCSHBV) gene (encoding 467 aa) was isolated. Sequence analysis

revealed amino acid identity with gC proteins from HSV-2 (46?9%), HSV-1 (44?5%) and

pseudorabies virus (21?2%). Highly conserved cysteine residues were also noted. Similar to

gCHSV-2, gCSHBV is less glycosylated than gCHSV-1, resulting in a molecular mass of 65 kDa if

expressed in replication-deficient vaccinia virus Ankara. Stable transfectants expressing full-length

gCSHBV on the cell surface induced C3 receptor activity and were substantially protected from

complement-mediated lysis; no protection was observed with control constructs. This suggests

that expression of the gC homologues on infected cell surfaces might also contribute to the survival

of infected cells in addition to decreased virion inactivation. Interestingly, soluble gCSHBV isolated

from protein-free culture supernatants did not interfere with the binding of the alternative

complement pathway activator properdin to C3b, which is similar to our findings with gCHSV-2 and

could be attributed to major differences in the amino-terminal portion of the protein with extended

deletions in both gCSHBV and gCHSV-2. Binding of recombinant gCSHBV to polysulphates was

observed. This, together with the heparin-sensitivity of the gCSHBV–C3 interaction on the infected

cell surface, suggests a role in adherence to heparan sulphate, similar to the gC proteins of

other herpesviruses.

INTRODUCTION

As a rare exception among animal members of the familyHerpesviridae, which do not usually infect humans, severedisease leading to high mortality in humans is caused bythe primate simian herpes B virus (SHBV, Cercopithecineherpesvirus type 1). Accidental infection of humans has

been described on several occasions with SHBV, originatingfrom old world monkeys of the genus Macaca.

In contrast to the situation in its natural host, where diseasesymptoms resemble those of human herpes simplex virus(HSV), severe encephalitis associated with high mortalityhas been observed in humans (Davenport et al., 1994).

Interestingly, human HSV can also act as a ‘killer virus’ iftransferred to certain new world monkey species (Huemeret al., 2002), indicating that the pathogenicity and virul-ence of herpesviruses are difficult to predict in differenthost environments.

3Present address: Upper Austrian Research GmbH, Center forBiomedical Nanotechnology, Linz, Austria.

The DNA sequence of the gC gene of Cercopithecine herpesvirus type 1reported in this paper has been deposited (30 October 1998) in EMBLunder nucleotide accession no. SHE012474 and in GenBank underno. AJ012474.

0001-8949 G 2003 SGM Printed in Great Britain 1091

Journal of General Virology (2003), 84, 1091–1100 DOI 10.1099/vir.0.18949-0

Immunocompetent humans are not usually endangered byalphaherpesviruses from other species, a phenomenon thathas been also observed in animals, where, in most cases,interspecies transmission seems to be rather restricted(Engels et al., 1992).

In contrast, a broad spectrum of cells from different speciesis susceptible to infection with alphaherpesviruses in vitro,which is most likely due to attachment to the heparansulphate ‘receptors’ found on many cell types. Thus, it seemsunlikely that a restriction at the cellular level is solelyresponsible for the observed host specificity of herpesvirusesin vivo and a key topic is to investigate possible mechanismsfor this widely observed ‘natural immunity’ (reviewed bySkinner et al., 2001). Identification of virus and host factorsthat cause susceptibility or resistance to herpesviruses fromother species may represent a possible way forward forfuture prophylactic or therapeutic measures.

In the early phase of virus infection, neither a specifichumoral nor a cellular immune response is available. Thefate of an infectious agent depends primarily on its abilityto avoid the mechanisms of the so-called ‘nonspecific’immune response. One important antibody-independentsystem, acting without precedent priming with antigen, isthe alternative complement pathway, a phylogenetically oldsystem that represents a first barrier against invadingmicroorganisms (Cooper, 1991). Its importance for com-plex viruses arises from the finding that several membersof the Herpesviridae and also the Poxviridae express pro-teins on the virion and infected cell surface; these proteinsare able to interact with proteins of the complement cascadeat different levels (Ohmann et al., 1988; Isaacs et al., 1992;Huemer et al., 1992b, 1993a; Rother et al., 1994).

Thus, HSV glycoprotein C (gCHSV-1/2) of both serotypeshas been shown to bind to the central complementcomponent C3 and thereby interfere with the antibody-independent alternative complement pathway (Fries et al.,1986; McNearney et al., 1987). gCHSV-1 was found to bevery similar to known human complement regulatoryproteins, thus competing for binding with the alternativepathway activator properdin or regulator protein factor H,which was not observed using purified gCHSV-2 (Huemeret al., 1993b). Absence of gCHSV-1 or pseudorabies virus(PRV) gIII (gCPRV) on virions has been shown to causeincreased virus lysis via the alternative complement path-way and interestingly, species variations in the efficiencyof lysis using different complement sources have beenobserved (Hidaka et al., 1991; Ikeda et al., 2000; Maeda et al.,2002). This is in accordance with the finding of differencesin the binding to complement C3 of herpesvirus gC pro-teins of different species (Huemer et al., 1993a).

Expression of gC homologues is highly conserved and thegC proteins of different HSV serotypes are found in clinicalisolates, with the exception of isolates originating fromimmunologically ‘privileged’ locations (Hidaka et al., 1990).Moreover, the finding of an attenuated phenotype of other

species isolates with deletions or diminished expression ofthe gC homologues (Mettenleiter et al., 1988; Liang et al.,1992; Moffat et al., 1998) also suggests that these proteinsperform vital functions in vivo.

Furthermore, the gC proteins have been shown to contri-bute to binding to cell surface heparan sulphate, which isalso true for other herpesvirus glycoproteins such as gB(reviewed by Sawitzky et al., 1990; Liang et al., 1992).However, it should be noted that binding to charged surfacemolecules, like heparan or chondroitin sulphate, appears tobe a general mechanism of virus adsorption in nature, whichis also found in many other viruses from different families.

gC binding to heparan sulphate is not essential for HSV-2attachment (Gerber et al., 1995) and studies using gC-deficient HSV-1 mutants suggest that gC plays little role inthe adsorption process (Griffiths et al., 1998). Although thisis still a controversial topic (Laquerre et al., 1998), there isgrowing evidence that binding to heparan sulphate maynot represent the predominant biological function of thegC homologues in vivo, at least in some of the strainstested. Additionally, the C3-binding sites on gCHSV-1 andthe binding sites for heparin, dextran sulphate and otherpolysulphates seem to be related, as these substances blockthe C3 receptor function of gCHSV-1 (Huemer et al., 1992a).

In this study, we have investigated the complementreceptor of SHBV, and cloned and sequenced gCSHBV.Expression of the gene in eukaryotic expression systemsfacilitates characterization of the antigenic structure andfurthers studies of possible immune regulatory properties.The constructs have also been used for diagnostic serologytesting. Serodiagnosis of SHBV is difficult due to extensivecross-reactivity with HSV specimens (Falke, 1964; Hilliardet al., 1989; Norcott & Brown, 1993). In addition, vacciniavirus constructs using the established modified vaccinevirus strain Ankara (MVA) have the potential for use inimmunization studies.

METHODS

Cells and viruses. The macaque isolate from SHBV used in thisstudy has been described (Falke, 1961). HSV-1 strain Ang andequine herpesvirus type 1 (EHV-1) strain Piber were used for com-parison. Virus stocks of SHBV were produced in Vero cells (ATCC,#CCL-81). For expression experiments, the rabbit kidney RK13 cellline (ATCC, #CCL-37) was used. Chicken-adapted MVA was kindlyprovided by A. Mayr, LMU-Munich, Germany. Wild-type virus andrecombinants were raised in primary chicken fibroblasts (CHF).

DNA purification, identification of fragments and cloningprocedures. As isolation of the full-length gCSHBV gene by PCRamplification using HSV-specific primers was not successful, andbecause amplification of long SHBV fragments proved to be difficultdue to the high G/C content of the SHBV genome, a conventionalcloning strategy was chosen. Viral DNA was isolated from SHBV-infected roller bottles of Vero cells, which were lysed in 1 % laurylsarcosine and digested with pronase E. Viral DNA was separatedfrom genomic DNA by density centrifugation in CsCl gradients.Fractions containing viral DNA were precipitated with isopropanol

1092 Journal of General Virology 84

H. P. Huemer and others

and digested with rare-cutting restriction enzymes. Restriction digestswere separated on 1 % agarose, blotted onto nylon membranes andhybridized with gCHSV-1- and gCEHV-1-specific probes under condi-tions of low stringency (50˚C, 0?5 % SSC wash). As probes, restric-tion fragments were excised from plasmids containing the codingsequences of HSV-1 strain KOS (Huemer et al., 1989) and EHV-1strain Piber (Huemer et al., 1995) and labelled by random primingusing Klenow polymerase and digoxigenin-labelled UTP. Blotswere developed using the anti-digoxigenin system from BoehringerMannheim. Cross-hybridizing SHBV fragments were isolated, sub-cloned into cloning plasmids and subjected to sequence analysis.For comparison, the gC gene of HSV-1 strain Ang was isolated in asimilar manner.

Sequence analysis. Subcloning of several subfragments was neces-sary as the high G/C content of SHBV prevented long sequencingruns due to the formation of secondary structures. Constructedplasmids were analysed by automated sequencing in an ABI PrismDNA sequencer and a MWG LiCor using both dye labelling anddye primer sequencing techniques. Homology alignments with gCsequences obtained from EMBL were performed using the DNAStarprogram.

Construction of expression plasmids and expression ofrecombinant proteins. After sequence analysis, a 1?4 kb BamHI–SacI fragment was excised and religated to a 570 bp NcoI–BamHIfragment producing the complete gCSHBV-encoding sequence. Thewhole coding sequence following a NcoI site containing the startcodon was inserted into the expression vector pCR3.1 (Invitrogen)under the control of the cytomegalovirus immediate-early promoter.

A construct lacking the putative transmembrane sequence wasgenerated subsequently by PCR amplification of the 59-terminalsequence, including the PstI site at position 1420 using kinase-treatedforward (59-GATGGAGTTCGGGAGCGGCGA-39) and reverse (59-GACCCCGTGGGCCAGCAGGTGACCCACCATCACCATCACCAT-TGA-39) primers. The PCR fragment was cloned unidirectionally intopCR3.1 using a single-sided dephosphorylated vector (Invitrogen).The fragment was excised with PstI/EcoRI and inserted into thecorresponding sites of pCR3.1SHBV, thus replacing the hydrophobicputative transmembrane sequences following aa 418 with six histidineresidues and a stop codon.

For comparison, similar expression plasmids were constructed,inserting the full-length gC genes from EHV-1 (Huemer et al., 1995)and Marek’s disease virus (MDV, kindly provided by K. Osterrrieder,IMB/BFAV, Insel Riems, Germany) into pCR3.1.

Plasmids were transfected into Vero and RK13 cells using Lipofection(Qiagen) and the expression of gCSHBV was monitored by immuno-fluorescence. Additionally, stable cell lines were generated by electro-poration of RK13 cells using a Biorad GenePulser (0?4 cm cuvettes,350 V, 500 mF). Recombinant cell lines were selected by outgrowth inmedium containing 500 mg geneticin ml21 (G418). These stable cellclones were finally moved to protein-free culture conditions, whichenabled isolation of the secreted recombinant protein by filtration.Among the protein-free media tested, Cytoferr CHO medium (PAALaboratories) supported growth of our transfected RK13 cells. Super-natants were concentrated to 100-fold by Amicon filtration (Millipore)using membranes with a 10 kDa cut-off pore. Filtrates were used as asource of recombinant protein.

Construction of vaccinia virus recombinants. As insertion intothe thymidine kinase gene leads to loss of replicative activity ofMVA (Scheiflinger et al., 1996), insertion into the haemagglutinin(HA) gene of vaccinia virus was performed using recombinationplasmid pHA11k-gpt, described recently by Huemer et al. (2000a).Full-length and truncated gCSHBV and the homologous sequences

from HSV-1 and EHV-1 were ligated into the recombination vectorpHA11k-gpt and transfected into CHF cells infected earlier withMVA. Recombinant viruses expressing gpt-resistance were selectedin medium containing xanthine, hypoxanthine and mycophenolicacid and isolated by several rounds of plaque purification usingsemi-solid agar overlays.

Antisera and immunoassays. A polyclonal anti-SHBV antiserumhas been raised in rabbits (Falke, 1964). Sera against cynomolgus-and rhesus-derived isolates of SHBV have been kindly provided byM. Slomka (CPHL, Colindale, UK). Macaque sera were obtainedfrom C. Coulibaly (Paul-Ehrlich-Institute, Germany). Production ofantisera and monoclonal antibodies against gCHSV-1 has beendescribed previously (Huemer et al., 1989). Monoclonal and poly-clonal antibodies against gC homologues from PRV, bovine herpes-virus type 1 (BHV-1) and EHV-1 have been used as describedearlier (Huemer et al., 1992b, 1993a, 1995). Antisera were tested byimmunofluoresence using 12-well microscopic slides coated withacetone/methanol-fixed SHBV-infected Vero or RK13 cells stablytransfected with gCSHBV or control constructs. Infected cell lysates,transfected cells and recombinant gCSHBV protein were reacted withpolyclonal anti-SHBV rabbit sera for ELISA and for immunofluore-sence assays, according to the methods described for gCHSV-1

(Huemer et al., 1989, 1992a).

Radioimmunoprecipitation of vaccinia virus lysates was performedby infecting 35S-labelled RK13 cells with recombinant gCSHBV-MVAconstructs for 2 days. Cells were lysed with 0?5 % NP-40 and 0?5 %deoxycholate and supernatants were obtained by low-speed centrifuga-tion. Lysates were then reacted with the indicated mono- or poly-clonal antibodies and precipitated by formalin-fixed Staphylococcusaureus protein A (Sigma). Samples were washed in lysis buffer andanalysed by SDS-PAGE. MVA constructs containing gCHSV-1, gCEHV-1

and the parental MVA were used for comparison.

Complement factors, C3-binding assays and polyanion bind-ing. The purification of complement component C3 from humanserum and other species has been described elsewhere (Huemer et al.,1992b, 1993a, b). Macaque C3 was purified accordingly from cyno-molgus monkey serum by ion exchange chromatography overMono-Q and subsequent molecular size fractionation on CL4b–Sepharose (Pharmacia). Haemolytic activity of the purified C3protein was tested by lysis of rabbit erythrocytes using C3-deficienthuman serum (Quidel).

Binding of C3 to surface-expressed gCSHBV was tested by FACS analy-sis of recombinant RK13 cells expressing full-length gCSHBV. Stabletransformants producing the truncated gCSHBV lacking the trans-membrane domain and the parental RK13 cell line were used ascontrols. Cells were incubated with complement component iC3(100 mg ml21 in PBS) on ice; iC3 represents haemolytically inactiveC3 with the internal thioester hydrolysed, leading to C3b/iC3b-likeproperties of the molecule, including binding to different humancomplement receptors (Cole et al., 1985). This was confirmed bybinding of the samples used to different complement receptor-carryinghuman cell lines, such as Raji and U937. Cells were washed with PBSand reacted with FITC-labelled antisera directed against C3d or C3c(Dako). After a final washing step, cells were fixed with formaldehydeand the intensity of fluorescence was determined using a FACScanIII(Beckton-Dickinson).

Binding of gCSHBV from virus-infected cell lysates was performedaccording to the immunoprecipitation described for gCPRV (Huemeret al., 1992b). In brief, purified C3 was coupled to CNBr–Sepharoseand incubated with membrane extracts of SHBV- or HSV-1-infectedVero cells, which were labelled previously with [14C]glucosamine.

http://vir.sgmjournals.org 1093

Herpes B virus gC protects against complement

Sepharose coupled with ovalbumin or purified mouse immuno-globulin served as negative controls. Washed precipitates were runon 10 % SDS-polyacrylamide gels and exposed to radiography.

Initial testing for binding of C3 to SHBV-infected Vero cells wasperformed by rosetting using sheep red blood cells (SRBC) coated withhuman complement component C3b. This has been found previouslyto be useful for detection of complement receptors expressed by HSV-1,EHV-1, EHV-4 and a guinea pig herpesvirus, but not for HSV-2,BHV-1 or PRV (Huemer et al., 1992a, 1993a, 1995). Inhibition ofSRBC binding of SHBV-infected cells was performed using heparinsulphate, dextran sulphate and other polysulphates under conditionsidentical to those described for HSV-1 (Huemer et al., 1992a).

Binding of recombinant gCSHBV from protein-free cell culturesuperatants to the polysulphates dextran sulphate and chondroitinsulphate was tested in ELISA. Thus, different concentrations of thesubstance were covalently coupled to CovaLink microtitre plates(Nunc). Plates were then blocked with 100 mM Tris/glycine followedby 1 % ovalbumin in PBS. Concentrated cell culture supernatants ofstably transfected cells secreting gCSHBV were used as well as super-natant from control cells concentrated 100 times by membranefiltration. Bound gCSHBV was detected by polyclonal anti-SHBVserum and peroxidase-labelled secondary antibody. Plates werewashed with PBS containing 0?1 % Tween-20, developed withABTS [azino-bis(3-ethylbenzthiazoline-6-sulfonic acid)] and read ata wavelength of 405 nm.

Properdin- and factor B-binding assays were performed as describedpreviously for gCHSV-1/2 (Huemer et al., 1993b). C3b-coated plateswere preincubated with 2-fold serial dilutions of gCSHBV, startingwith 5 mg ml21, and binding of purified properdin was tested usinga specific monoclonal antibody (reagents from Quidel). Additionally,binding of serum-derived factor B to plate-bound C3b was testedusing a polyclonal anti-factor B antiserum raised in goats (AtlanticAntibodies). Plates coated with BSA only and wells without viralglycoprotein served as background controls.

Complement lysis assay of transfected RK13 cells. Humanand porcine sera were obtained by clotting blood samples at 4˚C.Guinea pig serum, used for diagnostic complement fixation assays,was obtained from a commercial source. Haemolytic activity of serawas tested using antibody-coated SRBC, verifying high complementtitres of the preparations.

RK13 cells were lysed by porcine serum but not by human and guineapig sera, as could be detected by chromium-release assays. ProliferatingRK13 cells were trypsinized, washed and 56106 cells labelled at 37˚Cfor 1 h with 100 mCi (3?7 MBq) of sodium chromate containing 51Cr.Cells were washed to remove excess chromium, seeded into 96-wellplates and after a further 1 h of incubation, serial dilutions of serumwere added to the culture medium. The specific release of chromiumwas counted after 4–5 h by subtracting the background release ofuntreated cells. Percentage of cell lysis was compared to a maximumrelease value obtained by cell lysis using 2 % Triton X-100. Heatinactivation of sera or addition of EDTA inhibited complement-mediated lysis, completely reducing levels to background.

RESULTS

Expression of complement receptor activity onSHBV-infected cells sensitive to polyanions

SHBV-infected Vero cells showed binding of complementcomponent C3 coupled to SRBC and CNBr–Sepharose.Comparable to gCHSV-1 (Huemer et al., 1992a), this binding

was inhibited by heparin or dextran sulphate (Fig. 1). SRBC,lacking C3 only, showed no binding to SHBV-infectedcells; binding was also inhibited by anti-C3 antibodies (datanot shown).

Precipitations of SHBV-infected Vero cells labelled with[14C]glucosamine revealed a C3-binding protein of about65 kDa precipitated by human and macaque C3–Sepharose,whereas the fully glycosylated form of gCHSV-1 appeared at120 kDa. Furthermore, no binding was observed of gCHSV-1

and gCSHBV to the porcine complement component as wellas to the control–Sepharose (Fig. 1).

Isolation and sequencing of the gCSHBV gene

SHBV DNA hybridized to gCHSV-1 but not to gCEHV-1-specific DNA probes under conditions of low stringency(data not shown).

The gCSHBV-encoding sequence was contained within twoBamHI fragments of 960 bp and 3 kb. An open readingframe of 1401 bp encoding 467 aa was found and in contrastto the HSV proteins, two additional cysteine residues wereidentified in the leader sequence, in addition to the eightcysteines in the coding sequences at highly conservedpositions (Fig. 2).

There are six potential N-glycosylation sites in gCSHBV andcomparison of the protein sequence with the HSVhomologues revealed major structural differences in theamino-terminal domain. Both gCHSV-2 and gCSHBV showdeletions in the amino-terminal domain when comparedwith gCHSV-1, with the extended SHBV deletions preceding

Fig. 1. C3 receptor function in SHBV-infected Vero cells andinhibition by heparin and related polyanions. (a) Binding of[14C]glucosamine-labelled SHBV and HSV-1 lysates to C3–Sepharose. The binding of SHBV- and HSV-1-infected celllysates labelled with [14C]glucosamine to Sepharose-coupled,purified C3 from the indicated species. Lanes: 1, human C3; 2,porcine C3; 3, cynomolgus C3; 4, IgG–Sepharose. (b) Bindingof SHBV-infected cells to complement C3 on SRBC (upperpanel); binding can be inhibited by the addition of heparin(10 mg ml21) or related polyanions such as dextran sulphate(lower panel).



the deletion of the HSV-2 protein. Despite these differ-ences, the overall amino acid identity was about 44?5 % forgCHSV-1 and 46?9 % for gCHSV-2. The highly conservedcentral region of the molecule suggests a comparablestructure, which is also reflected by C3 binding.

Comparison with the homologous proteins of humanvaricella-zoster virus (VZV), PRV, BHV-1, EHV-1 andMDV revealed an overall amino acid identity of 21?2 % forPRV, 20?3 % for BHV-1, 20?1 % for EHV-1, 16?9 % forMDV and only 16?3 % for VZV.

Expression of gCSHBV in recombinant vacciniavirus

Cells infected with MVA constructs containing thegCSHBV-encoding sequence express the protein in the cellmembrane, as detected by immunofluorescence with anti-SHBV-specific antiserum. No cross-reactivity was observedusing anti-gCEHV-1 and anti-gCBHV-1 or gCPRV-specificpolyclonal antisera or different anti-gCHSV-1 monoclonalantibodies (data not shown).

Expression of gCSHBV in recombinant MVA was comparedwith the homologous proteins from EHV-1 and HSV-1 inFig. 3. Immunoprecipitation revealed a protein with amolecular mass of about 65 kDa, which indicates that theSHBV protein, similar to gCHSV-2, appears to be muchless glycosylated than the HSV-1 homologue, although,underglycosylated gCHSV-1 precursor precipitated also. Thestronger precipitations of gCHSV-1 and gCEHV-1 do notindicate a stronger rate of expression but reflect the highbinding capacity of monoclonal antibodies as comparedto polyclonal antisera directed against all SHBV proteins,with only a fraction of precipitated immunoglobulinrepresenting anti-gCSHBV antibodies. No precipitationswere observed with control MVA, thereby demonstratingthe specificity of the antibodies used.

Fig. 2. Comparison of the gCSHBV protein sequence withgCHSV-1 and gCHSV-2. Conserved amino acids are boxed andcysteine residues are highlighted. Gaps are included to maxi-mize alignment.

Fig. 3. Production of gCSHBV in MVA and RK13 transfectants.(a) Expression of gCSHBV in recombinant vaccinia virus. Immuno-precipitations of 35S-labelled supernatants of RK13 cells infectedwith MVA constructs and the parent MVA virus are shown.Lysates were precipitated with anti-HSV-1 gC and anti-EHV-1 gC monoclonal antibodies and anti-SHBV polyclonal rabbitserum. Relative molecular masses in kDa are shown on theright. (b) Secretion of recombinant gCSHBV in supernatants oftransfected cells. gCSHBV produced by constructs lacking thetransmembrane region was detected by Western blotting(lane 1), whereas only trace amounts could be detected inthe supernatant of stable RK13 transfectants expressing themembrane-associated protein (lane 2). In lane 3, gCEHV-1 isused as a control, which, due to glycosylation, shows up asa diffuse band. The amido black-stained blot of the proteinmarker is depicted on the right with bars indicating the relativemolecular mass in kDa.



Detection of C3 receptor activity on transfectedcells expressing surface-bound gCSHBV

Construction of stable cell lines expressing full-length and truncated gCSHBV. Soluble gCSHBV wasdetected by Western blotting in the supernatant of cellslacking the transmembrane anchor sequence as a proteinof approximately 65 kDa, whereas no secretion wasdetected in cells expressing the full-length protein (Fig. 3).

Expression of recombinant proteins was detected on thesurface and within transfected cells by immunofluorescenceusing a polyclonal anti-SHBV antiserum. Whereas the full-length protein is expressed homogeneously on the cellsurface, constructs lacking the gC transmembrane domainshow a protein staining mainly in the perinuclear (pre-sumably Golgi) region (Fig. 4).

Stable RK13 cell lines were obtained expressing full-lengthand truncated gCSHBV as well as clones expressing gCMDV.Interestingly, no clones showing sufficient expression ofgCHSV-1 were obtained, presumably due to toxicity of theHSV-1 protein in eukaryotic cells; this presents a furtherdifference between the simian and the human herpesvirusgC protein.

Cells expressing gCSHBV on the cell surface were comparedwith gCSHBV constructs lacking the membrane anchorsequence in addition to the parental RK13 cell line forbinding of complement C3 by FACS analysis. As depicted inFig. 5, binding of human C3 was observed only to surface-bound gCSHBV but not to parental RK13 cells or gCSHBV

constructs lacking the membrane anchor.

Inhibition of complement-mediated cell lysis bysurface-expressed gCSHBV

Protection from alternative complement pathway lysis ofthe stable RK13 cell lines expressing surface-bound gCconstructs was studied in chromium-release assays. Asshown in Fig. 6, complement lysis of untransfected RK13cells occurred in a concentration-dependent manner up toserum dilutions of 1 : 80.

Of interest in RK13 cells expressing gCSHBV on the cellmembrane was that complement lysis was diminishedsubstantially at different serum concentrations, with some-thing similar to a plateau or saturation point being reachedaround dilutions of 1 : 20–1 : 40. Higher serum concentra-tions exceeded the complement regulatory capacity of themembrane protein, although a significant reduction wasalso observed at lower dilutions.

As controls, the untransfected RK13 ancestor cell line andstable transformants expressing gCMDV on the cell mem-brane were used, with the latter also showing no protec-tive effect against complement-mediated cell lysis. The

(a)

(b)

Fig. 4. gCSHBV expression on the cell surface and secretion bytransfected RK13 cells. RK13 cells transfected with full-lengthgCSHBV (left panels) and a truncated version lacking the trans-membrane (TM) region of the gC gene (right panels) weretested for recombinant protein expression by immunofluore-scence using a SHBV-specific antiserum (a). UntransfectedRK13 cells served as the negative control (b). Note the peri-nuclear Golgi-type staining of secreted gCSHBV in comparisonwith the surface staining of the membrane-bound protein.

Fig. 5. Binding of C3 to cell lines expressing surface-boundgCSHBV. Binding of complement factor C3 was determined byFACS analysis. Filled curves represent cells incubated withcomplement component C3. Open curves indicate antibodycontrols without addition of C3. Stable transfectants and con-trols are shown as follows: S+, RK13 cells expressinggCSHBV, including the transmembrane region; S2, RK13 cellsexpressing truncated gCSHBV without the transmembraneregion; RK13, negative control (untransfected ancestor RK13cell line); Raji, positive control (complement receptor-positiveBurkitt’s lymphoma tumour B-cell line).



percentage of lysis of the gCSHBV RK13 transfectantsseemed to be largely dependent on the amount of gCSHBV

protein expressed on the cell surface, as variations in lysisamong different stable cell clones correlated with the rate ofmembrane expression of gCSHBV, as detected by immuno-fluorescence, and no inhibition was found with the trun-cated gCSHBV RK13 clone produced subsequently (datanot shown).

Binding of recombinant gCSHBV to polysulphatesand failure to block C3b binding of alternativepathway complement activator properdin

As depicted in Fig. 7, secreted gCSHBV bound to dextransulphate coupled covalently to polystyrol surfaces in a dose-dependent manner. No binding was observed to microtitreplates coated with ovalbumin or BSA. Weak binding wasalso observed to sodiumhyaluronate, which is a chargedsubstance found in connective tissue.

Soluble gCSHBV did not inhibit the binding of purifiedalternative complement pathway activator properdin tosurface-bound C3b, which is similar to gCHSV-2. The dif-ferent concentrations of the viral protein represented bythe data-points (Fig. 7, lanes 1–3) did not lead to a reduc-tion in properdin binding to C3b (Fig. 7,$), but, interest-ingly, even led to slightly enhanced levels of properdinbinding as compared to the data-point without viral pro-tein added (Fig. 7, lane 4). BSA-coated wells were used asa comparison; no binding of properdin was observed, asdetected by monoclonal anti-properdin antibody (Fig. 7,#).

Binding of factor B from human serum to our C3b-coatedplates was rather weak compared to the strong binding of

the purified properdin and also the use of a polyvalentantiserum led to considerable background. Furthermore,gCSHBV did not have any visible influence on this inter-action (Fig. 7, .).

DISCUSSION

Infection of humans with SHBV occurs after animal bitesamong animal care workers but infected organ or cellculture material is also hazardous for laboratory staff andthe disease can be transferred to close contacts (Davenportet al., 1994). Serological testing for SHBV is rather difficultdue to extensive cross reactivity with human HSV (Falke,1964; Hilliard et al., 1989; Norcott & Brown, 1993). Therecombinant constructs and proteins produced in thisstudy may consequently provide a useful tool for serologicaltesting of infected animals and human populations thatmay be at risk.

Fig. 6. Expression of gCSHBV on the cell surface protects fromcomplement-mediated cell lysis. Stably transfected RK13 cellsshowing high expression of gCSHBV on the cell membrane ($)were compared with untransfected RK13 cells (#) and cellsexpressing gCMDV on the cell surface (m). Means ±SE oftriplicate experiments are shown. Lysis of cells by the indicatedserum dilutions was measured by release of chromium andexpressed as the percentage of total cell lysis achieved withdetergent. Heat inactivation of serum completely inhibitedcomplement-mediated lysis of RK13 cells (e).

Fig. 7. Secreted gCSHBV binds to polysulphates but does notinhibit properdin binding to C3b. (a) Binding of soluble gCSHBV

to polysulphates. Microtitre plates were coated with the indi-cated substances by covalent binding. Serial 2-fold dilutions ofrecombinant gCSHBV were added in lanes 1–3 and bindingdetected in ELISA. Lane 4 indicates the background controlwith no gC added. (b) No inhibition of alternative complementpathway activator properdin by gCSHBV. Serial 2-fold dilutionsof soluble gCSHBV were added to C3b-coated plates in lanes1–3. Lane 4 served as positive reference with no gC added.Binding of purified properdin to C3b- ($) and BSA-coatedplates (#) was monitored by monoclonal antibody. Binding ofcomplement factor B from human serum to plate-bound C3b(m) and BSA (n) was examined for comparison.



The sequencing of the gCSHBV gene revealed a close rela-tionship with the HSV homologues, which was expectedfrom their comparable genome organization (Harringtonet al., 1992). Regions with high amino acid identities andconserved cysteine residues indicate a similar structure,which is also reflected in the functional properties of theprotein. Regarding the differences in the amino-terminaldomain of gCSHBV, it should be noted that there are alsomajor differences within the two human HSV subtypes inthis region, with the deletion of 27 aa in the HSV-2 protein.

These differences in the amino terminus of the gC pro-teins possibly explain the different functional behaviour ofgCHSV-1 and gCHSV-2 upon competition with complementregulatory proteins, including properdin, factor H or com-plement receptors CR1 and CR2, with the failure of gCHSV-2

to inhibit factor H or alternative pathway activator pro-perdin (Huemer et al., 1993b). Therefore, it was notsurprising that the recombinant gCSHBV was also unableto block binding of properdin but may even partiallystabilize the properdin–C3b interaction. Whether thisobservation has any significance for the function ofgCSHBV and gCHSV-2 remains questionable, as these proteinsmight possibly exert their function only in their physio-logical (i.e. membrane-associated) conformation, which isalso true for other membrane receptors.

Regarding the binding to surface-bound C3, SHBV, des-pite the HSV-2-like amino-terminal deletions, seems tobe functionally similar to HSV-1, as gCHSV-2 does notmediate binding of C3-coated SRBC to HSV-2-infectedcells. Whether this reflects influences such as differingdensities or clustering of receptors, differences in lateralmobility and/or other membrane/receptor peculiarities ofthe behaviour of gCHSV-2 remain unclear. A suggestedinhibitory influence of other HSV-2 glycoproteins in thecell membrane seems rather unlikely, as there were nodetectable interferences of HSV-2 glycoproteins usingintertypic recombinant strains expressing gCHSV-1 (Huemeret al., 1992a).

The sensitivity of the gCSHBV–C3 interaction to heparin anddextran sulphate and the binding of gCSHBV to surface-bound polysulphates is also analogous to the situationfound with gCHSV-1 (Huemer et al., 1992a), suggesting acomparable heparan sulphate ‘receptor’ activity in SHBV.However, studies investigating adherence of gC-null HSVstrains from both subtypes suggested that mediatingadherence to heparan sulphate may not be the predomi-nant function of gC (Gerber et al., 1995; Griffiths et al.,1998). Of interest, heparin also interacts with a variety ofproteins in the clotting system, the immune system, growthfactors, adherence molecules and receptors, including dif-ferent endogenous complement regulatory proteins (Edenset al., 1993).

From the sequence data, it appears that SHBV seems to bemore closely related to HSV-2. Furthermore, PRV wasfound as the most closely related animal herpesvirus, which

is in accordance with the neurotropic characteristics ofboth viruses. One of the characteristics of PRV is its broadhost-range, which is unusual for the Alphaherpesvirinae,and it would be interesting to know more about the tro-pism of SHBV in this respect.

The interference of gCSHBV with cell lysis by porcinecomplement was somewhat surprising, as there was nobinding to purified porcine C3, resisting the vigorouswashing procedure in immunoprecipitation. Whether thisindicates differences in affinity or the numbers of bindingsites, etc., has to be clarified. Failure of direct binding onthe surface does not necessarily correspond to a lack offunction, like the situation with known human comple-ment regulators such as decay-accelerating factor (DAF).DAF is a cell surface glycoprotein that regulates complementactivity by accelerating the decay of C3/C5 convertases.Interacting directly with membrane-bound C3b/C4b, itprevents uptake of C2 and factor B but is not a C3b receptordetectable by the rosetting technique (Medof et al., 1984).

Furthermore, gCHSV-2 has been shown to interfere withthe alternative complement pathway, although no directbinding of C3 was observed on the cell surface (McNearneyet al., 1987). Similar observations have been made mostrecently with PRV, which has been shown to be protectedagainst porcine serum by gCPRV (Maeda et al., 2002),although binding to porcine C3 was only found with thepurified protein and not by binding of complement-coatedSRBC to PRV-infected cells (Huemer et al., 1992b, 1993a).This difficult scenario is complicated further by the factthat herpesviruses, as well as other enveloped viruses, areable to incorporate host cell-derived complement regulatoryproteins into their membrane (Spear et al., 1995). This hasbeen shown to play an important role in variations ofcomplement-mediated lysis of PRV grown in differenthost cells (Maeda et al., 2002) and highlights further thecomplex differences among human and animal comple-ment regulatory proteins, as well as possible influencesof conformational requirements or varying affinities fordifferent C3 species of the virally encoded complementreceptor proteins.

Purified gCHSV-1 also has a reduced binding capacity forC3 derived from other species and binds very poorly toporcine C3 (Huemer et al., 1993a). This observation hasbeen confirmed most recently by plasmon surface resonanceanalysis (Biacore), whereas human alternative complementpathway regulator factor H was able to bind to the porcinecomplement C3 in this set-up (M. Holmberg, Helsinki,personal communication).

Certain species-specific selectivity is observed amonghuman complement regulators (Horstmann et al., 1985;Yu et al., 1997) and marked differences in the efficiency ofcomplement-mediated herpesvirus lysis using serum ofdifferent species as a complement source have beendescribed (Hidaka et al., 1991; Maeda et al., 2002). Thefindings that HSV was lysed far more efficiently by serum of



rodents as compared to human serum than earlier reportshave suggested appears to play a crucial role in animalstudies using herpesvirus vectors for anti-tumour therapy(Ikeda et al., 2000). This suggests a close evolutionaryrelationship of microbial-encoded complement escapemechanisms and the ‘unspecific’ host defence and high-lights once more the importance of using appropriateanimal hosts as models of herpesvirus infections.

Nevertheless, this class of molecules, acting as viruspathogenicity factors due to their conservation amongherpesviruses, represents a good candidate for furthervaccination studies, which are under way also with gCantigens from other species (Huemer et al., 2000b).

ACKNOWLEDGEMENTS

We thank Dr Y. Li (Oswell Laboratory, University of Birmingham, UK)for help with automated sequencing and Drs M. Slomka (CPHL,Collindale, UK) and C. Coulibaly (Paul Ehrlich Institute, Langen,Germany) for donation of antisera. The authors also appreciatethe technical help provided by Bridget Rickard, Janet Drake andSteven Hibbs (all from the Chemical and Biological DefenceEstablishment, Porton Down, UK). The work was supported bythe Austrian Academy of Sciences programme for the advancementof research and technology (APART), the Jubilaumsfonds of theAustrian National Bank and the Biotechnology Programme of theEuropean Commission.

REFERENCES

Cole, J. L., Housley, G. A., Jr, Dykman, T. R., MacDermott, R. P. &Atkinson, J. P. (1985). Identification of an additional class of C3-binding membrane proteins of human peripheral blood leukocytesand cell lines. Proc Natl Acad Sci U S A 82, 859–863.

Cooper, N. R. (1991). Complement evasion strategies of micro-organisms. Immunol Today 12, 327–331.

Davenport, D. S., Johnson, D. R., Holmes, G. P., Jewett, D. A., Ross,S. C. & Hilliard, J. K. (1994). Diagnosis and management of human Bvirus (Herpesvirus simiae) infections in Michigan. Clin Infect Dis 19,33–41.

Edens, R. E., Linhardt, R. J. & Weiler, J. M. (1993). Heparin is not justan anticoagulant anymore: six and one-half decades of studies on theability of heparin to regulate complement activity. In ComplementToday (Complement Profiles), pp. 96–120. Edited by J. M. Cruse &R. E. Lewis, Jr. Basel: Karger.

Engels, M., Palatini, M., Metzler, A. E., Probst, U., Kihm, U. &Ackermann, M. (1992). Interactions of bovine and caprineherpesviruses with the natural and the foreign hosts. Vet Microbiol33, 69–78.

Falke, D. (1961). Isolation of two variants with different cytopathicproperties from a strain of herpes B virus. Virology 14, 492–495.

Falke, D. (1964). Serologic relationship between B virus and herpessimplex virus using complement fixation assay. Z Hyg 150, 185–193(in German).

Fries, L. F., Friedman, H. M., Cohen, G. H., Eisenberg, R. J.,Hammer, C. H. & Frank, M. M. (1986). Glycoprotein C of herpessimplex virus 1 is an inhibitor of the complement cascade. J Immunol137, 1636–1641.

Gerber, S. I., Belval, B. J. & Herold, B. C. (1995). Differences in therole of glycoprotein C of HSV-1 and HSV-2 in viral binding maycontribute to serotype differences in cell tropism. Virology 214,29–39.

Griffiths, A., Renfrey, S. & Minson, T. (1998). Glycoprotein C-deficient mutants of two strains of herpes simplex virus type 1exhibit unaltered adsorption characteristics on polarized or non-polarized cells. J Gen Virol 79, 807–812.

Harrington, L., Wall, L. V. & Kelly, D. C. (1992). Molecular cloningand physical mapping of the genome of simian herpes B virus andcomparison of genome organization with that of herpes simplexvirus. J Gen Virol 73, 1217–1226.

Hidaka, Y., Sakuma, S., Kumano, Y., Minagawa, H. & Mori, R.(1990). Characterization of glycoprotein C-negative mutants ofherpes simplex virus type 1 isolated from a patient with keratitis.Arch Virol 113, 195–207.

Hidaka, Y., Sakai, Y., Toh, Y. & Mori, R. (1991). Glycoprotein C ofherpes simplex virus type 1 is essential for the virus to evadeantibody-independent complement-mediated virus inactivation andlysis of virus-infected cells. J Gen Virol 72, 915–921.

Hilliard, J. K., Black, D. & Eberle, R. (1989). Simian alphaherpes-viruses and their relation to the human herpes simplex virus. ArchVirol 109, 83–102.

Horstmann, R. D., Pangburn, M. K. & Muller-Eberhard, H. J. (1985).Species specificity of recognition by the alternative pathway ofcomplement. J Immunol 134, 1101–1104.

Huemer, H. P., Broker, M., Larcher, C., Lambris, J. D. & Dierich,M. P. (1989). The central segment of herpes simplex virus type 1glycoprotein C (gC) is not involved in C3b binding: demonstrationby using monoclonal antibodies and recombinant gC expressed inEscherichia coli. J Gen Virol 70, 1571–1578.

Huemer, H. P., Larcher, C., Dierich, M. P. & Falke, D. (1992a).Factors influencing the interaction of herpes simplex virusglycoprotein C with the third component of complement. ArchVirol 127, 291–303.

Huemer, H. P., Larcher, C. & Coe, N. E. (1992b). Pseudorabiesglycoprotein III derived from virions and infected cells binds to thethird component of complement. Virus Res 23, 271–280.

Huemer, H. P., Larcher, C., Van Drunen Littel-van den Hurk, S.& Babiuk, L. A. (1993a). Species selective interaction ofAlphaherpesvirinae with the ‘unspecific’ immune system of thehost. Arch Virol 130, 353–364.

Huemer, H. P., Wang, Y., Garred, P., Koistinen, V. & Oppermann, S.(1993b). Herpes simplex virus glycoprotein C: molecular mimicry ofcomplement regulatory proteins by a viral protein. Immunology 79,639–647.

Huemer, H. P., Nowotny, N., Crabb, B. S., Meyer, H. & Hubert, P. H.(1995). gp13 (EHV-gC): a complement receptor induced by equineherpesviruses. Virus Res 37, 113–126.

Huemer, H. P., Strobl, B., Shida, H. & Czerny, C. P. (2000a).Induction of recombinant gene expression in stably transfected celllines using attenuated vaccinia virus MVA expressing T7 RNA poly-merase with a nuclear localisation signal. J Virol Methods 85, 1–10.

Huemer, H. P., Strobl, B. & Nowotny, N. (2000b). Use of apathogenicvaccinia virus MVA expressing EHV-1 gC as basis of a combinedrecombinant MVA/DNA vaccination scheme. Vaccine 18, 1320–1326.

Huemer, H. P., Larcher, C., Czedik-Eysenberg, T., Nowotny, N. &Reifinger, M. (2002). Fatal infection of a pet monkey with humanherpesvirus. Emerg Infect Dis 8, 639–642.

Ikeda, K., Wakimoto, H., Ichikawa, T., Jhung, S., Hochberg, F. H.,Louis, D. N. & Chiocca, E. A. (2000). Complement depletionfacilitates the infection of multiple brain tumors by an intravascular,



replication-conditional herpes simplex virus mutant. J Virol 74,4765–4775.

Isaacs, S. N., Kotwal, J. & Moss, B. (1992). Vaccinia virus complement-control protein prevents antibody-dependent complement-enhancedneutralization of infectivity and contributes to virulence. Proc NatlAcad Sci U S A 89, 628–632.

Laquerre, S., Argnani, R., Anderson, D. B., Zucchini, S., Manservigi,R. & Glorioso, J. C. (1998). Heparan sulfate proteoglycan binding byherpes simplex virus type 1 glycoproteins B and C, which differ intheir contributions to virus attachment, penetration, and cell-to-cellspread. J Virol 72, 6119–6130.

Liang, X., Babiuk, L. A. & Zamb, T. J. (1992). An in vivo study of aglycoprotein gIII-negative bovine herpesvirus 1 (BHV-1) mutantexpressing b-galactosidase: evaluation of the role of gIII in virusinfectivity and its use as a vector for mucosal immunization. Virology189, 629–639.

Maeda, K., Hayashi, S., Tanioka, Y., Matsumoto, Y. & Otsuka, H.(2002). Pseudorabies virus (PRV) is protected from complementattack by cellular factors and glycoprotein C (gC). Virus Res 84, 79–87.

McNearney, T. A., Odell, C., Holers, V. M., Spear, P. G. & Atkinson,J. P. (1987). Herpes simplex virus glycoproteins gC-1 and gC-2bind to the third component of complement and provide protec-tion against complement-mediated neutralization of viral infectivity.J Exp Med 166, 1525–1535.

Medof, M. E., Kinoshita, T. & Nussenzweig, V. (1984). Inhibition ofcomplement activation on the surface of cells after incorporation ofdecay-accelerating factor (DAF) into their membranes. J Exp Med160, 1558–1578.

Mettenleiter, T. C., Schreurs, C., Zuckermann, F., Ben-Porat, T. &Kaplan, A. S. (1988). Role of glycoprotein gIII of pseudorabies virusin virulence. J Virol 62, 2712–2717.

Moffat, J. F., Zerboni, L., Kinchington, P. R., Grose, C., Kaneshima,H. & Arvin, A. M. (1998). Attenuation of the vaccine Oka strain of

varicella-zoster virus and role of glycoprotein C in alphaherpesvirusvirulence demonstrated in the SCID-hu mouse. J Virol 72, 965–974.

Norcott, J. P. & Brown, D. W. G. (1993). Competitive radio-immunoassay to detect antibodies to herpes B virus and SA8 virus.J Clin Microbiol 31, 931–935.

Ohmann, H. B. & Babiuk, L. A. (1988). Induction of receptors forcomplement and immunoglobulins by herpesviruses of variousspecies. Virus Res 9, 335–342.

Rother, R. P., Rollins, S. A., Fodor, W. L., Albrecht, J. C., Setter, E.,Fleckenstein, B. & Squinto, S. P. (1994). Inhibition of complement-mediated cytolysis by the terminal complement inhibitor ofherpesvirus saimiri. J Virol 68, 730–737.

Sawitzky, D., Hampl, H. & Habermehl, K. O. (1990). Comparison ofheparin-sensitive attachment of pseudorabies virus (PRV) and herpessimplex virus type 1 and identification of heparin-binding PRVglycoproteins. J Gen Virol 71, 1221–1225.

Scheiflinger, F., Falkner, F. G. & Dorner, F. (1996). Evaluation of thethymidine kinase (tk) locus as an insertion site in the highlyattenuated vaccinia MVA strain. Arch Virol 141, 663–669.

Skinner, G. R. B., Ahmad, A. & Davies, J. A. (2001). The infrequencyof transmission of herpesviruses between humans and animals:postulation of an unrecognised protective host mechanism. CompImmunol Microbiol Infect Dis 24, 255–269.

Spear, G. T., Lurain, N. S., Parker, C. J., Ghassemi, M., Payne, G. H.& Saifuddin, M. (1995). Host cell-derived complement controlproteins CD55 and CD59 are incorporated into the virions of twounrelated enveloped viruses. Human T cell leukemia/lymphomavirus type I (HTLV-I) and human cytomegalovirus (HCMV).J Immunol 155, 4376–4381.

Yu, J., Dong, S., Rushmere, N. K., Morgan, B. P., Abagyan, R. &Tomlinson, S. (1997). Mapping the regions of the complementinhibitor CD59 responsible for its species selective activity.Biochemistry 36, 9423–9428.



Documents

protection from complement-mediated cell lysis - CiteSeerX