Production of Herpes B Virus Recombinant Glycoprot
http://www.100md.com
微生物临床杂志 2005年第2期
Viral Immunology Center, Georgia State University, Atlanta, Georgia
ABSTRACT
B virus (cercopithecine herpesvirus 1) is the only deadly alphaherpesvirus that is zoonotically transmissible from macaques to humans. The detection of humoral immune responses is the method of choice for the rapid identification of B virus-infected animals. We evaluated the diagnostic potential of recombinant B virus glycoproteins for the detection of immunoglobulin G (IgG) antibodies in monkey and human sera. Glycoproteins B, C, and E and secreted (sgG) and membrane-associated (mgG) segments of glycoprotein G (gG) were expressed in the baculovirus expression system, while gD was expressed in CHO cells. We developed recombinant protein-based IgG enzyme-linked immunosorbent assays (ELISAs) and compared their diagnostic efficacies by using B virus antibody-negative (n = 40) and -positive (n = 75) macaque sera identified by a whole antigen-based ELISA and Western blotting. The diagnostic sensitivities of the gB-, gC-, gD-, and mgG-ELISAs were 100, 97.3, 88.0, and 80.0%, respectively. The specificities of the gB-, gC-, and gD-ELISAs and of the mgG-ELISA were 100 and 97.5%, respectively. In contrast, the sensitivities and specificities of sgG- and gE-ELISAs were low, suggesting that sgG and gE are less effective diagnostic antigens. Sera from nonmacaque monkeys cross-reacted with gB, gC, and gD, and only baboon sera reacted weakly with mgG. Human herpes simplex virus type 1 (HSV-1)- and HSV-2-positive sera pools reacted with gB and gD, whereas sera from B virus-infected individuals reacted with all four antigens. These data indicate that gB, gC, gD, and mgG have a high diagnostic potential for B virus serodiagnosis in macaques, whereas mgG may be a valuable antigen for discrimination between antibodies induced by B virus and those induced by other, closely related alphaherpesviruses, including HSV-1 and -2.
INTRODUCTION
Human infection with B virus (also called cercopithecine herpesvirus 1, monkey B virus, and herpes B virus) is the most feared occupational hazard among individuals working with macaque monkeys, since fatality is often the outcome of infection, which proceeds in the absence of effective antiviral therapy (25, 56). The use of macaques in research has been steadily growing over the last decade and is expected to rise quickly in the near future due to the increasing demands for these animals for use in HIV/AIDS investigations, vaccine trials, drug testing, and research into bioterrorism agents. As macaque usage increases, frequencies of human exposures to B virus are increasing as well. Rapid and accurate diagnostic tests are urgently needed to aid in the early identification of clinical cases, which is essential for a timely initiation of antiviral therapies in zoonotically infected humans. In addition to human diagnostics, enhanced assays are required for monitoring specific-pathogen-free (SPF) macaque colonies established by the National Institutes of Health for the breeding of B virus-free animals (55), as these animals often demonstrate only very low levels of antibody.
Unfortunately, a direct diagnosis of infection by virus detection (cell culture or PCR) is impossible in most cases, since, similar to other alphaherpesviruses, B virus establishes a lifelong latency in sensory ganglia of macaques and seldom reactivates (9, 53, 58). Current diagnoses of B virus infections in humans and monkeys rely mainly on the detection of serum antibodies to B virus proteins. Indirect enzyme-linked immunosorbent assays (ELISA) and other rapid serological tests based on a solubilized, B virus-infected cell antigen have been developed and used for the identification of infected animals (8, 21, 28, 39), with subsequent confirmation by Western blotting to identify specific targets that are immunoreactive with serum antibodies (54). Serodiagnoses of zoonotic infections are performed by Western blotting, which is a time-consuming technique that requires visual interpretations of complex patterns. This method is not specific enough to unequivocally identify B virus infections in herpes simplex virus (HSV)-positive humans because antibodies to HSV type 1 (HSV-1) and HSV-2 are highly cross-reactive with B virus proteins (12, 23), making differentiation a complex task. Most importantly, currently used serological assays utilize B virus-infected cell lysates as an antigen, and these can only be produced in a maximum containment laboratory (biosafety level 4), which limits the number of facilities that are capable of providing antigen. Antigens may also suffer from lot-to-lot variation, compromising outcome measures based on assays using these antigens.
Recombinant-based serological assays have been developed for the diagnosis of many viral infections, including human cytomegalovirus (11), hepatitis C virus (27), hepatitis E virus (46), human papillomavirus (49), Ebola virus (45), and many others. Several recombinant glycoprotein G (gG)-based immunoassays for HSV type-specific serodiagnosis are commercially available (19, 44). However, the use of recombinant antigens for B virus serodiagnosis has not been widely investigated. Recently, recombinant gD was shown to be useful for B virus serodiagnosis by dot blot and Western blot assays, but the performance of this antigen in ELISAs was not studied (51). In an earlier study, we produced a fusion protein containing a single B virus-specific immunodominant epitope of gD and demonstrated its efficacy for the identification of B virus infections by using an indirect ELISA (43). The serodiagnosis of infections, however, cannot be based exclusively upon the presence of antibodies to a single epitope of a pathogen due to variations in individual responses to a selected epitope. Moreover, the existence of cross-reacting antibodies against similar epitopes in other proteins may result in a false-positive diagnosis. Ideally, the identification of an appropriate cocktail of recombinant proteins would provide the opportunity to detect the broad range of differentially induced antibodies during the course of infection, including different antibody subclasses and isotypes.
Individual polypeptide targets of antibody responses following HSV infections have been studied by several investigators. Most immune sera from HSV-infected individuals react with 12 to 20 viral polypeptides in immunoblots. The envelope glycoproteins gB, gD, gC, gE, and gG, the major capsid protein VP5, and the nucleocapsid complex p40 have been identified as primary targets of IgG antibody responses in patients with HSV-1 and/or HSV-2 infections (2, 5, 13, 14, 38). The B virus's homology to HSV-1 (79.9% amino acid identity for gB, 57% identity for gD, 49.9% identity for gC, 46% identity for gE, and 29.2% identity for gG [42]) suggests that the specificity of B virus-induced antibody responses is similar to that of responses induced by HSV types 1 and 2. Ideally, these glycoproteins may be primary candidates for evaluation as recombinant diagnostic reagents.
We report for the first time the production and purification of recombinant B virus antigens, specifically gB, gC, gE, the secreted segment of gG (sgG), and the membrane-associated segment of gG (mgG), by use of a baculovirus expression system and an investigation of immunogenic and antigenic properties of these proteins, along with recombinant gD produced in CHO cells. Diagnostic efficacies of indirect ELISAs using each recombinant protein as a coating antigen (Rec-ELISAs) were determined with a large panel of positive (n = 75) and negative (n = 40) serum samples from rhesus and cynomolgus macaque monkeys representing a spectrum of antibody levels. Reactivities of sera from different monkey species and of human sera containing HSV-1, HSV-2, and B virus-specific antibodies with recombinant B virus antigens were also studied to identify antigens that are suitable for a B virus differential serodiagnosis for both natural and inadvertent foreign hosts.
MATERIALS AND METHODS
Viruses, cells, and media. All cells, vectors, and supplies were purchased from Invitrogen (Carlsbad, Calif.) unless otherwise noted. Plasmids were propagated in Escherichia coli Top10 cells. Recombinant baculoviruses were produced and titrated in Spodoptera frugiperda (Sf-9) cells. Sf-9 cells were maintained at 27°C in serum-free Sf-900 II SFM insect cell culture medium with antibiotics.
Construction of baculovirus donor plasmid pFBMV5H. The expression cassette of the Bac-to-Bac donor plasmid pFastBacHTc was modified by the addition of a honeybee melittin secretion sequence (HBM) and the V5 epitope coding sequence. The HBM sequence was amplified by PCR from another transfer vector, pMelBac, by use of the forward primer Mel-RsrII and the reverse primer RecBac778 (Table 1). The amplified fragment was cloned into a pCR-XL-TOPO plasmid and then excised with the RsrII and EcoRI restriction enzymes from the resulting recombinant plasmid. A fragment encoding V5 epitope and His tags was excised from a pcDNA3.1A vector with the EcoRI and PmeI restriction enzymes. The pFastBacHTc plasmid was digested with HindIII, its ends were blunted with the Klenow fragment, and then the plasmid was digested with RsrII. Both fragments were ligated into the linearized pFastBacHTc plasmid to produce a new donor plasmid, pFBMV5H. The introduced sequences were confirmed to be correct and in frame by sequencing of the pFBMV5H DNA by use of a BigDye Terminator sequencing kit (ABI PRISM).
Protein analyses. Analyses of protein structures were performed with the DNASTAR Protean program (version 4.0.3), the signal peptide prediction program SignalP v.2.0.b2 (http://www.cbs.dtu.dk/services/SignalP-2.0/), and the transmembrane prediction program TMHMM v2.0 (http://www.cbs.dtu.dk/services/TMHMM/).
Preparation of recombinant baculoviruses. The DNA segments encoding the extracellular domains of the B virus glycoproteins B, C, E, and G were amplified from the genomic DNA of B virus strain E2490 by PCRs with the primers listed in Table 1. PCR fragments were cloned into the pCR-TOPO-XL vector. Recombinant plasmids were sequenced to verify the identities of the inserts. Cloned fragments were then excised with the HindIII and EcoRI restriction enzymes and cloned into the corresponding sites of the pFBMV5H donor plasmid to obtain gB, gC, gE, sgG, and mgG donor plasmids.
The transposition of expression cassettes into a baculovirus shuttle vector (bacmid) was performed by the transformation of E. coli DH10Bac competent cells with the constructed donor plasmids. Successful transpositions were verified by PCR amplification of the prepared bacmid DNAs with gene-specific forward primers and the M13/pUC reverse primer. Recombinant viruses were produced by transfecting monolayers of Sf-9 cells grown in six-well cell culture plates with bacmid DNAs by use of the Cellfectin reagent. At 72 h postinfection, virus-containing medium from the transfected cells was harvested and titrated by a plaque assay on Sf-9 cell monolayers. Five individual plaques were picked from each transfection for further analysis. High-titer stocks of recombinant viruses were then produced in Sf-9 cells. All procedures were performed in accordance with the manual for the Bac-to-Bac baculovirus expression system.
Optimization of recombinant protein expression. Suspension cultures of Sf-9 cells (30 ml each) grown in 100-ml spinner flasks were infected with each recombinant virus at multiplicities of infection (MOIs) of 1, 4, and 7. Supernatant samples were collected at 0, 24, 48, 72, and 96 h postinfection, mixed 1:1 with 2x sodium dodecyl sulfate (SDS) buffer, boiled for 5 min, fractionated by SDS-10% polyacrylamide gel electrophoresis (SDS-10% PAGE), and then analyzed by immunoblotting with an anti-V5 monoclonal antibody (MAb) (1:2,000 dilution) (Sigma, St. Louis, Mo.). The largest quantities of nondegraded proteins were produced at 72 h postinfection at MOIs of 4 and 7.
Production and purification of recombinant proteins. Sf-9 suspension cultures were grown in 250-ml spinner flasks to a density of 1.6 x 106 cells/ml and then were infected with the recombinant viruses at an MOI of 4. Media were harvested at 72 h postinfection and centrifuged at 15,000 x g for 10 min at 4°C. Supernatants were collected and either stored at –80°C or used immediately for purification. The collected supernatants were concentrated and dialyzed against column buffer (40 mM sodium phosphate [pH 7.8], 0.3 M NaCl) by tangential flow filtration with a Labscale TFF system containing a Pellicon XL device (10-kDa cutoff) (Millipore, Bedford, Mass.). The concentrated supernatant was then loaded onto a 1-ml HiTrap chelating column (Amersham Biosciences, Piscataway, N.J.) charged with Ni2+ ions and equilibrated with column buffer. After washing of the column with 10 ml of column buffer supplemented with 20 mM imidazole, recombinant proteins were eluted with column buffer containing 250 mM imidazole. Fractions were analyzed by immunoblotting with the anti-V5 MAb. Those containing recombinant protein were pooled and dialyzed against phosphate-buffered saline (PBS), and aliquots were stored at –80°C. Recombinant B virus gD was produced in suspension cultures of Chinese hamster ovary cells that were stably transfected with the gD expression construct (L. Perelygina, H. Zurkuhlen, I. Patrusheva, N. Patrushev, and J. K. Hilliard, Abstr. 25th Int. Herpesvirus Workshop, abstr. 5.50, 2000).
Western blot analysis. Lysates of uninfected Vero cells and of Vero cells infected with B virus lab strain E2490 were prepared as described previously (28). Proteins were fractionated by SDS-10% PAGE and then either stained in a gel with Coomassie blue (GelCode Blue staining reagent; Pierce, Rockford, Ill.) or transferred onto a nitrocellulose membrane. The membrane was blocked with BLOTTO (5% skim milk, 1% normal goat serum, and 0.05% Tween 20 in PBS) for 1 h and then incubated for 1 h with primary antibodies diluted in BLOTTO. After being washed three times for 10 min in PBS-0.05% Tween 20, the membrane was incubated for 1 h with peroxidase-conjugated anti-rabbit, anti-mouse, or anti-human immunoglobulin G (IgG) (1:40,000 dilution in BLOTTO) (Pierce). The membrane was washed again, and bands were visualized on Kodak X-Omat film after detection with ECL Western blot detection reagents (Amersham Biosciences).
Mass spectrometry. Recombinant protein bands were excised from an SDS-10% PAGE gel and subjected to in-gel digestion by modified trypsin (Promega, Madison, Wis.) as suggested by the manufacturer. Matrix-assisted laser desorption ionization-time of flight mass spectrometry of the resulting digests was performed on a Voyager-DE PRO MALDI-TOF workstation (Applied Biosystems, Framingham, Mass.) by the Advanced Biotechnology Core Facility at Georgia State University. For protein identification, the measured masses of the tryptic peptides were compared with published databases by use of the MS-FIT module of Protein Prospector software (http://prospector.ucsf.edu/ucsfhtml4.0/msfit.htm).
Production of rabbit antisera. Two hundred fifty micrograms of each purified glycoprotein was diluted in 0.5 ml of PBS, mixed with an equal volume of Freund's complete adjuvant (Sigma), and injected subcutaneously into multiple sites of New Zealand White female rabbits. Rabbits were given subcutaneous booster injections with 250 μg of each protein in Freund's incomplete adjuvant (Sigma) every 4 weeks for 6 months. Preimmune sera were collected at the time of the first immunization. Test and production bleeds were performed on the 14th day after each injection. The end-point Western blot (WB) titer of each serum was determined by immunoblotting of membrane strips containing fractionated B virus antigen with twofold serial dilutions of rabbit antisera. The titer was determined as the reciprocal of the lowest serum dilution at which no reactivity with B virus proteins was detected.
Monkey and human sera. All sera were kindly provided by the National B Virus Resource Center (Atlanta, Ga.). B virus antibody-negative and -positive serum pools from macaques and HSV-negative, HSV-1-positive, and HSV-2-positive serum pools from humans were prepared as described previously (43). Antibody-positive serum pools from baboons, langurs, sooty mangabeys, and African green monkeys infected with herpesvirus papio 2, herpes virus langur, mangabey herpesvirus, and simian agent 8, respectively, were described and characterized previously (29). B virus antibody-positive (n = 75) or -negative (n = 40) rhesus and cynomolgus serum samples were randomly selected for assessments of macaque serum reactivities with recombinant proteins. All sera were serotyped by the B virus diagnostic laboratory by use of the previously described whole B virus antigen-based ELISA and Western blot assays (28, 32). ELISA titers of positive sera ranged from 600 to 230,000.
Rec-ELISA. The optimal concentrations of antigens, sera, and conjugate were determined by checkerboard titration. Purified recombinant proteins were diluted to a concentration of 1 μg/ml (gB) or 2 μg/ml (gC, gD, gG, and gE) in 20 mM Tris-HCl, pH 8. Wells of Maxisorp immunoplates (Nalge Nunc International, Rochester, N.Y.) were coated with antigens (100 μl/well) for 1 h on a plate shaker at room temperature and then overnight at 4°C. The plates were washed with 20 mM Tris-HCl, pH 8, to remove unbound antigens and were blocked with 150 μl of ELISA-BLOTTO (borate-buffered saline containing 2.5% [each] nonfat dry milk and liquid gelatin)/well. Blocking and subsequent incubations with sera and conjugate were done for 1 h on a plate shaker at room temperature. Three washes with borate-buffered saline-0.05% Tween 20 were performed after each incubation step by use of an automatic plate washer. Sera were diluted 1:50 in ELISA-BLOTTO, and 50 μl was added to each well. Alkaline phosphatase-conjugated goat anti-human IgG Fc fragment (Sigma) was diluted 1:4,000 (for tests with monkey sera) or 1:6,000 (for tests with human sera) in ELISA-BLOTTO, and 50 μl was added to each well. A substrate solution was prepared by dissolving p-nitrophenyl phosphate (pNPP) tablets (Sigma) in 1 M diethanolamine buffer containing 0.5 mM MgCl2, pH 9.8 (1 mg of pNPP/ml), and 200 μl of a freshly prepared mixture was added to each well. After a 25-min incubation at room temperature, the reaction was stopped by adding 50 μl of 3 N NaOH per well, and the optical densities (ODs) were read on an automatic microplate reader at 405 nm. B virus antibody-positive and -negative rhesus monkey serum pools were included in each plate to serve as positive and negative controls and for use in calculations of positive-to-negative (P/N) ratios. P/N values were determined as follows: the OD value of a test or positive control serum reacted with an antigen was divided by the OD value of the negative control serum reacted with the same antigen. A test was considered valid if the P/N value of the positive control serum was >2.2.
RESULTS
The Bac-to-Bac baculovirus expression system allows the rapid and efficient generation of multiple recombinant baculoviruses. However, the expression cassette of the commercial donor plasmids from the Bac-to-Bac system does not contain specific DNA sequences that allow a recombinant fusion protein to be secreted into media and to be easily detected. We constructed a new donor plasmid, pFBMV5H, from pFastBacHTc by adding the HBM secretion signal to achieve a high level of protein secretion and the V5 epitope to enhance the detection of recombinant proteins. A diagram of the expression cassette of pFBMV5H is shown in Fig. 1A.
The locations of the signal peptides and transmembrane domains in B virus glycoproteins (Fig. 1B) were predicted with the SignalP and TMHMM programs. The location of the cleavage site in the B virus gG precursor was predicted based on data available for the homologous HSV-2 protein (50). DNA fragments encoding the entire extracellular domains of gB, gC, gE, sgG, and mgG were amplified by PCR from the genomic DNA of B virus E2490 and then cloned into the BamHI and EcoRI sites of the donor plasmid pFBMV5H. The sequences of the cloned fragments were identical to the genomic sequences of the corresponding B virus genes (42), with the exception of the mgG fragment, as there was a 132-bp deletion (positions 1096 to 1227 in the gG gene) in the constructed mgG donor plasmid. This deleted DNA fragment, which encodes a PAPTTT amino acid repeat, is present in the gG sequence of the E2490 strain but is not found in any of the other sequenced gG genes of B virus clinical isolates (40). Since the deletion construct encodes an mgG protein that actually mimics those produced during natural infections, we decided to use this construct for production of the mgG recombinant baculovirus. Five recombinant baculoviruses were subsequently constructed as described in Materials and Methods and were designated bac-gB37-719, bac-gC27-425, bac-gE29-408, bac-gG22-310, and bac-gG311-613.
To produce large amounts of recombinant proteins, we infected Sf-9 suspension cultures with the recombinant viruses at an MOI of 4 and harvested the media at 72 h postinfection. Recombinant proteins were affinity purified from the supernatants of infected insect cells by nickel chromatography. Western blotting with a V5 MAb identified the column fractions containing recombinant proteins. The peak column fractions were then pooled. After measuring the protein concentrations in the pooled fractions, we estimated the following yields of purified proteins: 6 mg of gB, 10 mg of gC and mgG, and 1 mg of gE and sgG per 1 liter of insect cell culture medium. Substantial quantities of recombinant gE and sgG were retained in the infected cells, as determined by immunoblotting of infected cell pellets and supernatants with the anti-V5 MAb (data not shown), resulting in a low level of secretion and low yields of purified gE and sgG.
All recombinant proteins migrated in an SDS gel as broad bands with larger molecular masses than predicted, most probably due to glycosylation (Fig. 2A; Table 2), since multiple N-glycosylation signals were found in all proteins (Fig. 1B). Similarly, cotranslational N-glycosylation of baculovirus-expressed HSV-1 glycoproteins resulted in a decrease in their electrophoretic mobilities in denaturing polyacrylamide gels (15, 16, 47). The anti-V5 MAb reacted strongly with all of the recombinant proteins (Fig. 2B), but a macaque B virus-positive serum pool recognized only recombinant gB, gC, and mgG (Fig. 2C). Interestingly, mgG and sgG migrated as multiple bands, and each reacted with the anti-V5 MAb and B virus-positive macaque sera. To verify the identities of the produced proteins, we cut the protein bands from an SDS-PAGE gel, digested them with trypsin, and analyzed the resulting fragments by mass spectrometry. Comparisons of the tryptic profiles of the recombinant proteins with published databases confirmed their identities as B virus glycoproteins gB, gC, gE, and gG.
To verify the immunogenicity of the recombinant proteins and to further validate their authenticity, we used each of the five glycoproteins to immunize rabbits to induce monospecific, polyclonal antibodies. With the exception of sgG (WB titer, <800), each glycoprotein induced a high-titer antibody in rabbits (WB titer, <32,000). Western blot analyses revealed that each immune serum reacted with specific protein bands from the B virus-infected cell lysates (Fig. 3), whereas preimmune sera did not (data not shown). Rabbit anti-gB, -gC, and -gE sera reacted with B virus polypeptides of 120, 72, and 80 kDa, respectively. Sera directed against the different gG fragments recognized multiple protein bands: the 85-kDa band was recognized by sgG antisera, the 160-, 48-, and 20-kDa polypeptides were recognized by mgG antisera, and the 115-kDa band was recognized by both types of antisera. The 48- and 20-kDa bands were weak and could be seen only after long exposures of immunoblots to X-ray films. The polypeptides recognized by the produced monospecific antisera were in size agreement with the previously reported molecular masses of selected B virus glycoproteins (7, 41, 48). These data validate that the antigenic and immunogenic properties of the prepared recombinant proteins are similar to those of viral glycoproteins produced during B virus infection.
The performance of each baculovirus-produced recombinant, i.e., gB, gC, gE, sgG, and mgG, in an ELISA was subsequently evaluated. Recombinant gD purified from the medium of stably transfected CHO cells was also included in the evaluation. To determine the optimal coating concentration for microtiter plates (producing the highest OD and P/N values), we performed a checkerboard titration with positive and negative serum pools from macaques. The optimal amount of recombinant antigen per well was 50 ng for gB and 100 ng for all other proteins.
The results of Rec-ELISA showed that the B virus antibody-positive serum pool reacted strongly with gB, gC, gD, and mgG, weakly with gE, and not at all with sgG (Fig. 4), confirming the Western blot results for these proteins. To further assess the value of sgG and gE for serodiagnosis, we studied the reactivities of individual negative (n = 10) and positive (n = 10) macaque sera with these proteins at a 1:50 dilution. For sgG-ELISA, positive samples reacted weakly with sgG (OD values varied from 0.16 to 0.83), suggesting that the secreted portion of B virus gG did not stimulate the production of a high level of antibodies during natural B virus infection and, therefore, that sgG has limited value for serodiagnosis. For gE-ELISA, both positive and negative macaque sera reacted with gE, producing OD values in the ranges of 0.3 to 1.3 and 0.2 to 0.66, respectively. Moreover, an HSV and B virus antibody-negative human pool reacted very strongly with gE (OD value of 2.9). This gE property renders the produced recombinant gE antigen unsuitable for diagnostic purposes, at least in this form. However, more specific epitopes of gE may be useful immunodiagnostic reagents.
Extended evaluations of the gB, gC, gD, and mgG antigens were performed with large panels of individual macaque sera. B virus antibody-positive (n = 75) and -negative (n = 40) serum panels included approximately equal numbers of samples from rhesus and cynomolgus macaques, since these macaque species are used most often for biomedical research. To determine the extents of IgG antibody reactivities with gB, gC, gD, and mgG, we used each recombinant antigen separately to coat individual microplate wells and to react with each serum from the selected panels. The highest average immunoreactivity of the positive serum panel (mean OD value ± 1 standard deviation [SD]) occurred with gB (2.23 ± 0.69), followed closely by gC (1.58 ± 1.00), mgG (1.05 ± 1.06), and gD (0.67 ± 0.48) (Fig. 5, solid horizontal lines). Cutoff values for positive results were determined from Rec-ELISA results by use of the negative serum panel. The reactivities of negative sera were low with gC, gD, and mgG (OD values of 0.06 to 0.2) but were high with gB (OD values of 0.4 to 0.5) due to binding of the conjugate to this antigen. Later, we identified a different lot of conjugate without cross-reactions with gB, and we currently use it to perform Rec-ELISAs. The OD cutoff values were calculated separately for each antigen as the mean of the results from 40 negative sera plus 3 SD (Fig. 5, dashed lines). The P/N values of negative sera for each antigen ranged from 0.3 to 3.7. The P/N cutoff value of 2.2 was determined as the mean of the P/N values of negative sera plus 3 SD.
According to the established cutoffs, all B virus-positive samples (n = 75) were positive for gB antibodies, 73 were positive for gC antibodies, 66 were positive for gD antibodies, and 60 were positive for mgG antibodies, resulting in diagnostic sensitivities of 100, 97.3, 88.0, and 80.0% for gB-, gC, gD-, and mgG-ELISAs, respectively, compared to the reference assays (BV-ELISA and Western blotting). Antibodies to all four glycoproteins were found in 76% of positive sera. Rec-ELISA results for negative samples revealed one discordant result between mgG-ELISA and the reference assays. Therefore, the relative diagnostic specificity of the gB-, gC-, and gD-ELISAs was 100%, and the relative specificity of the mgG-ELISA was 97.5%. Notably, the majority of mgG antibody-negative samples (11 of 15) were from cynomolgus macaques, and consequently, the diagnostic sensitivities of mgG-ELISAs for cynomolgus sera (n = 38) and rhesus sera (n = 37) differed substantially, at 71.1 and 89.2%, respectively. These data indicate that all four recombinant antigens have excellent diagnostic potential for macaque testing. The mgG antigen, however, might be well suited only for the serodiagnosis of infections with rhesus strains of B virus, as further studies are needed to assess the usefulness of this antigen for cynomolgus serum testing.
The serodiagnosis of B virus infection in foreign hosts, e.g., humans or nonmacaque, Old World monkeys, is often complicated by the presence of antibodies that are cross-reactive with species-specific alphaherpesviruses. To find out whether the recombinant proteins have immunodiagnostic value for identifying B virus infections in foreign hosts, we determined the extent to which antibodies induced by closely related simplex alphaherpesviruses in foreign hosts reacted with the B virus recombinant antigens. Nonmacaque monkey serum pools were prepared from samples containing high antibody titers against species-specific alphaherpesviruses, and thus Rec-ELISAs were performed with sera diluted 1:500 for the gB antigen and 1:100 for the other three recombinant antigens. All human sera were tested at a dilution of 1:50. Rec-ELISA data for each positive serum were normalized to OD values obtained with a negative serum pool from the same species and then expressed as P/N values (Table 3). The threshold value was set at 2.2. As expected, gB antibodies from all tested sera were the most cross-reactive, producing the highest OD values and P/N ratios. All serum pools from nonmacaque monkeys cross-reacted with the gD and gC antigens, but only baboon sera showed reactivities with mgG that were above the threshold. Individual serum samples from B virus-infected humans (both HSV negative and HSV-1 positive) reacted with all B virus recombinants, whereas pooled sera from HSV-positive and B virus-negative patients recognized only the gB and gD antigens. Interestingly, mgG appears to have the best diagnostic value for the identification of B virus zoonoses, since HSV-positive sera as well as sera from nonmacaque monkeys infected with herpes simplex viruses did not react with this protein in Rec-ELISAs. Furthermore, HSV antibody-positive serum pools did not react detectably with gC, suggesting that gC may be an additional useful antigen for human serodiagnosis.
DISCUSSION
The identification of herpesvirus infections is generally based upon indirect analyses of antibodies because the viruses elude detection. Interpretations of antibody reactivity can be complicated due to the plethora of cross-reactive interactions among antibodies and closely related viral antigens. These challenges can be reduced by the use of virus-specific proteins for the detection of antibodies. In this study, we presented a comparative analysis of the diagnostic capacities of recombinant B virus glycoproteins, i.e., gB, gC, gD, gE, sgG, and mgG, versus that of a whole B virus antigen with respect to the diagnostic efficacy of each for the detection of B virus-specific antibodies. The selection of these glycoproteins was based upon extensive Western blot analyses of tens of thousands of macaque serum samples, representing many time points after B virus infection, to identify virus protein targets of the host immune system as part of the National B Virus Resource Center (www.gsu.edu/bvirus).
In the first part of this study, we reported the cloning of B virus gB, gC, gE, sgG, and mgG into baculovirus by use of the modified donor vector pFBMV5H. Production and affinity purification of each recombinant glycoprotein resulted in lower relative yields of gE and sgG (1.0 mg/liter) than of gB, gC, and mgG (6 to 10 mg/liter). The data indicated that gE and sgG are secreted into the medium at reduced levels, although the insect-specific secretion signal was identical for each recombinant. These glycoproteins appeared to remain bound to the cell, as shown by a Western blot analysis of cells. Since the correct folding of expressed proteins in the endoplasmic reticulum is known to be essential for the effective excretion of secretory and membrane proteins (10, 22, 57), one possible explanation for the relatively low yields of gE and sgG may be that these truncated forms were incorrectly processed in insect cells and, as a result, were misfolded and retained in the endoplasmic reticulum. In similar studies by other investigators using HSV-2, notably less protein was secreted by insect cells infected with sgG-2- and full-size recombinant gG-2-expressing baculoviruses than with an mgG-2-expressing baculovirus (26). The production of purified gE and sgG in our studies, however, resulted in yields sufficient to perform 10,000 tests with 100 ng of each protein per diagnostic sample.
We next demonstrated by Western blotting and ELISA that the baculovirus-produced gB, gC, and mgG glycoproteins are recognized by sera from naturally infected hosts, establishing the essential retention of antigenic epitopes found in the authentic viral glycoprotein counterparts. The consistent absence of detectable antibody reactivities with recombinant gE and sgG in Western blots suggested either a lack of gE and/or sgG authenticity, resulting in a failure of naturally induced antibodies to recognize these recombinant proteins, or that linear epitopes of these viral proteins are not targeted strongly by host humoral immunity during natural infection. Since the authenticity of the recombinants was verified by mass spectrometry, we suggest that humoral responses to B virus infections may not include reactivities to linear (continuous) epitopes in gE and sgG. In accordance with our data, a lack of antibodies against gE and sgG linear epitopes was also demonstrated with sera from HSV-infected patients (30, 34).
Recombinant viral glycoproteins were used to produce rabbit antibodies that were then verified to be immunoreactive with each authentic viral glycoprotein present in B virus-infected cell lysates. Each recombinant protein elicited strong antibody responses in rabbits, with the exception of sgG. For ELISAs, the sgG recombinant protein was weakly reactive with sera from naturally infected macaques, suggesting that sgG is simply not capable of inducing a strong B-cell response, nor is its authentic counterpart from B virus. Similar observations have been made for HSV-2-infected humans with sgG-2, which induces only a weak antibody response late in infection relative to membrane-associated gG (17).
The next critical phase of analysis of the B virus recombinant reagents for use in immunodiagnosis established their value for the detection of IgG responses induced in infected macaques as well as humans. Four of the six recombinant reagents, when used in ELISA, proved to be both specific and sensitive relative to a whole B virus antigen-based ELISA. Antibody detection by ELISA was most sensitive when we used recombinant gB to evaluate defined panels of rhesus and cynomolgus macaque sera. Thus, gB appears to be the most promising candidate for immunodiagnostics. However, gB is a highly conserved protein, and therefore homologous proteins of other macaque herpesviruses, e.g., simian varicella virus (SVV), may have the capability to induce antibodies that are cross-reactive with gB. Several SVV strains have been isolated, including macaque strains (18), but SVV antibody prevalence in macaque populations has not been reported to date. A sequence comparison of the B virus and SVV gB proteins (GenBank accession no. AAA58703) revealed a 48% amino acid identity between their ectodomains, suggesting that SVV gB polyclonal antibody responses may include antibodies that cross-react with B virus gB, although early limited studies in our lab showed no cross-reactivity between B virus and SVV for sera from naturally infected macaques. The ability of several human herpesviruses, such as HSV-1, HSV-2, varicella-zoster virus, cytomegalovirus, and Epstein-Barr virus, to induce cross-reactive gB polyclonal antibodies, however, has been reported repeatedly in the literature (4, 6, 31). Future studies will be performed to investigate the cross-reactivity between SVV and B virus, but even if they are cross-reactive, the use of appropriate combinations of B virus recombinant proteins, e.g., gC, gD, and mgG, determined the specificity of induced antibodies with excellent sensitivities (97.3, 88.0, and 80.0%, respectively) for nearly 100% of the samples tested for this study. Most importantly, these reagents eliminate the need for and the danger associated with the production of large-scale lots of B virus for immunodiagnostics.
We were very interested in whether recombinant B virus gG would prove to be a valuable immunodiagnostic reagent in view of its usefulness for the differentiation of HSV types 1 and 2. Of two produced gG recombinants, only mgG performed satisfactorily in Rec-ELISA, demonstrating 80% relative diagnostic sensitivity and 97.5% specificity with our test sera. The secreted N-terminal fragment of gG appeared to have little diagnostic value, as low serum reactivity was observed with the sgG antigen even when readily detectable mgG-specific antibodies were present. A recently developed sgG-2-based ELISA for HSV-2 antibody detection similarly demonstrated a reduced sensitivity relative to ELISA with mgG-2 or whole gG-2 (17, 34). We demonstrated in this study that the diagnostic sensitivity of the mgG-based ELISA is lower than that of the gB-, gC-, or gD-based ELISA. Further work will be required to understand these observations. With similar observations from HSV studies, investigators have speculated several possibilities that encompass aspects of both virus and host correlates of infection, e.g., late kinetics of gG antibody development, infections with gG null mutants, or a quick waning of gG antibodies (3, 20, 35, 38). Future studies are planned to determine the value of specific gG epitopes for the differentiation of B virus-induced antibodies in different species of macaques and in humans.
The ability of these reagents to differentiate antibodies induced against human herpesviruses is important to our evaluations because of the large number of potential exposures to B virus among members of the biomedical community working with macaque monkeys. Many exposures have included individuals with a positive history of HSV antibodies. For this reason, we were particularly interested in mgG- and gC-based ELISAs, as immunodominant, type-specific epitopes are clustered in the C-terminal region of the closely related HSV-2 gG (33, 36) and because type-specific gC responses have also been noted (1). The experiments described in this article demonstrate that HSV antibody pools do not react in the mgG- or gC-based ELISA for the detection of B virus, whereas zoonotically B virus-infected humans do react unambiguously. Importantly, one serum was obtained from a symptomatic patient on day 30 after the onset of disease (24), whereas another, convalescent-phase serum was obtained from the person at >10 years postinfection. Although only two samples were analyzed, our results suggest that B virus infection in humans elicits type-specific gG and gC antibody responses. However, the performance of gC and mgG antigens in ELISA must be further investigated by use of a large panel of individual human sera before we apply these recombinant antigens to B virus serodiagnosis for humans.
An added value of these specific immunodiagnostic reagents is the fact that selected recombinants can be used to detect B virus-specific antibodies in nonmacaque species that are coinfected with closely related simplex alphaherpesviruses. Antibodies induced by these viruses from sooty mangabeys, langurs, baboons, and African green monkeys reacted either poorly or not at all with mgG in ELISAs. Cross-species infections in nonhuman primates can thus be identified with confidence, which is essential for the prevention of B virus outbreaks in primate colonies housing different monkey species in close proximities (37, 52).
In summary, we showed with this work that the recombinant B virus antigens gB, gC, gD, and mgG are valuable diagnostic reagents for the identification of B virus infections in macaques. A mixture of the recombinant antigens is likely to provide the highest sensitivity and specificity in B virus diagnostic tests. In addition to being relatively economical to produce and easy to subject to quality control and standardization, recombinant proteins are safe to use in laboratories that are not equipped with biocontainment facilities. Because the early identification of B virus infections in animals within a primate colony, and especially in newly acquired animals, is important for preserving the integrity of an SPF colony and for preventing additional fatal zoonotic infections, these highly standardized recombinant reagents should prove to be invaluable for the design of assays that can be performed readily in a wide variety of clinical laboratory settings.
ACKNOWLEDGMENTS
We thank the personnel of the National B Virus Resource Laboratory for providing monkey and human serum samples. We are grateful to Deborah Walthall for performing mass spectrometry analysis.
This study was supported by Public Health Service grants R01 RR03162 and P40 RR05062 from the NIH National Center for Research Resources and by the generous and continued support of the Georgia Research Alliance.
Present address: Division of Cardiology, Department of Medicine, Emory University, Atlanta, GA 30322.
REFERENCES
Ackermann, G., F. Ackermann, H. J. Eggers, U. Wieland, and J. E. Kuhn. 1998. Mapping of linear antigenic determinants on glycoprotein C of herpes simplex virus type 1 and type 2 recognized by human serum immunoglobulin G antibodies. J. Med. Virol. 55:281-287.
Ashley, R., J. Benedetti, and L. Corey. 1985. Humoral immune response to HSV-1 and HSV-2 viral proteins in patients with primary genital herpes. J. Med. Virol. 17:153-166.
Ashley, R. L., J. Militoni, F. Lee, A. Nahmias, and L. Corey. 1988. Comparison of Western blot (immunoblot) and glycoprotein G-specific immunodot enzyme assay for detecting antibodies to herpes simplex virus types 1 and 2 in human sera. J. Clin. Microbiol. 26:662-667.
Balachandran, N., D. E. Oba, and L. M. Hutt-Fletcher. 1987. Antigenic cross-reactions among herpes simplex virus types 1 and 2, Epstein-Barr virus, and cytomegalovirus. J. Virol. 61:1125-1135.
Bernstein, D. I., Y. J. Bryson, and M. A. Lovett. 1985. Antibody response to type-common and type-unique epitopes of herpes simplex virus polypeptides. J. Med. Virol. 15:251-263.
Bernstein, D. I., L. M. Frenkel, Y. J. Bryson, and M. G. Myers. 1990. Antibody response to herpes simplex virus glycoproteins gB and gD. J. Med. Virol. 30:45-49.
Blewett, E. L., D. Black, and R. Eberle. 1996. Characterization of virus-specific and cross-reactive monoclonal antibodies to herpesvirus simiae (B virus). J. Gen. Virol. 77:2787-2793.
Blewett, E. L., J. T. Saliki, and R. Eberle. 1999. Development of a competitive ELISA for detection of primates infected with monkey B virus (herpesvirus simiae). J. Virol. Methods 77:59-67.
Boulter, E. A. 1975. The isolation of monkey B virus (herpesvirus simiae) from the trigeminal ganglia of a healthy seropositive rhesus monkey. J. Biol. Stand. 3:279-280.
Braakman, I., and E. van Anken. 2000. Folding of viral envelope glycoproteins in the endoplasmic reticulum. Traffic 1:533-539.
Daiminger, A., U. Bader, and G. Enders. 1998. An enzyme linked immunoassay using recombinant antigens for differentiation of primary from secondary or past CMV infections in pregnancy. J. Clin. Virol. 11:93-102.
Eberle, R., D. Black, and J. K. Hilliard. 1989. Relatedness of glycoproteins expressed on the surface of simian herpes-virus virions and infected cells to specific HSV glycoproteins. Arch. Virol. 109:233-252.
Eberle, R., S. W. Mou, and J. A. Zaia. 1985. The immune response to herpes simplex virus: comparison of the specificity and relative titers of serum antibodies directed against viral polypeptides following primary herpes simplex virus type 1 infections. J. Med. Virol. 16:147-162.
Eberle, R., S. W. Mou, and J. A. Zaia. 1984. Polypeptide specificity of the early antibody response following primary and recurrent genital herpes simplex virus type 2 infections. J. Gen. Virol. 65:1839-1843.
Ghiasi, H., R. Kaiwar, A. B. Nesburn, and S. L. Wechsler. 1992. Baculovirus expressed herpes simplex virus type 1 glycoprotein C protects mice from lethal HSV-1 infection. Antivir. Res. 18:291-302.
Ghiasi, H., R. Kaiwar, A. B. Nesburn, and S. L. Wechsler. 1992. Expression of herpes simplex virus type 1 glycoprotein B in insect cells. Initial analysis of its biochemical and immunological properties. Virus Res. 22:25-39.
Gorander, S., B. Svennerholm, and J. A. Liljeqvist. 2003. Secreted portion of glycoprotein G of herpes simplex virus type 2 is a novel antigen for type-discriminating serology. J. Clin. Microbiol. 41:3681-3686.
Gray, W. L., and N. J. Gusick. 1996. Viral isolates derived from simian varicella epizootics are genetically related but are distinct from other primate herpesviruses. Virology 224:161-166.
Groen, J., B. Hersmus, H. G. Niesters, W. Roest, G. van Dijk, W. van der Meijden, and A. D. Osterhaus. 1999. Evaluation of a fully automated glycoprotein G-2 based assay for the detection of HSV-2 specific IgG antibodies in serum and plasma. J. Clin. Virol. 12:193-200.
Hashido, M., F. K. Lee, S. Inouye, and T. Kawana. 1997. Detection of herpes simplex virus type-specific antibodies by an enzyme-linked immunosorbent assay based on glycoprotein G. J. Med. Virol. 53:319-323.
Heberling, R. L., and S. S. Kalter. 1987. A dot-immunobinding assay on nitrocellulose with psoralen inactivated herpesvirus simiae (B virus). Lab. Anim. Sci. 37:304-308.
Helenius, A., T. Marquardt, and I. Braakman. 1992. The endoplasmic reticulum as a protein-folding compartment. Trends Cell Biol. 2:227-231.
Hilliard, J. K., D. Black, and R. Eberle. 1989. Simian alphaherpesviruses and their relation to the human herpes simplex viruses. Arch. Virol. 109:83-102.
Holmes, G. P., J. K. Hilliard, K. C. Klontz, A. H. Rupert, C. M. Schindler, E. Parrish, D. G. Griffin, G. S. Ward, N. D. Bernstein, T. W. Bean, et al. 1990. B virus (herpesvirus simiae) infection in humans: epidemiologic investigation of a cluster. Ann. Intern. Med. 112:833-839.
Huff, J. L., and P. A. Barry. 2003. B-virus (cercopithecine herpesvirus 1) infection in humans and macaques: potential for zoonotic disease. Emerg. Infect. Dis. 9:246-250.
Ikoma, M., J. A. Liljeqvist, J. Groen, K. L. Glazenburg, T. H. The, and S. Welling-Wester. 2002. Use of a fragment of glycoprotein G-2 produced in the baculovirus expression system for detecting herpes simplex virus type 2-specific antibodies. J. Clin. Microbiol. 40:2526-2532.
Kalamvoki, M., V. Miriagou, A. Hadziyannis, U. Georgopoulou, A. Varaklioti, S. Hadziyannis, and P. Mavromara. 2002. Expression of immunoreactive forms of the hepatitis C NS5A protein in E. coli and their use for diagnostic assays. Arch. Virol. 147:1733-1745.
Katz, D., J. K. Hilliard, R. Eberle, and S. L. Lipper. 1986. ELISA for detection of group-common and virus-specific antibodies in human and simian sera induced by herpes simplex and related simian viruses. J. Virol. Methods 14:99-109.
Katz, D., W. Shi, P. W. Krug, R. Henkel, H. McClure, and J. K. Hilliard. 2002. Antibody cross-reactivity of alphaherpesviruses as mirrored in naturally infected primates. Arch. Virol. 147:929-941.
Kuhn, J. E., G. Dunkler, K. Munk, and R. W. Braun. 1987. Analysis of the IgM and IgG antibody response against herpes simplex virus type 1 (HSV-1) structural and nonstructural proteins. J. Med. Virol. 23:135-150.
Kuhn, J. E., K. Klaffke, K. Munk, and R. W. Braun. 1990. HSV-1 gB and VZV gp-II crossreactive antibodies in human sera. Arch. Virol. 112:203-213.
Lamiel, J. M., J. A. Ward, and J. K. Hilliard. 1993. Detection of type-specific herpesvirus antibodies by neural network classification of Western blot densitometer scans, p. 1731-1736. In Proceedings of the IEEE International Conference on Neural Networks. Institute of Electrical and Electronics Engineers, Inc., Piscataway, N.J.
Levi, M., U. Ruden, and B. Wahren. 1996. Peptide sequences of glycoprotein G-2 discriminate between herpes simplex virus type 2 (HSV-2) and HSV-1 antibodies. Clin. Diagn. Lab. Immunol. 3:265-269.
Liljeqvist, J. A., E. Trybala, J. Hoebeke, B. Svennerholm, and T. Bergstrom. 2002. Monoclonal antibodies and human sera directed to the secreted glycoprotein G of herpes simplex virus type 2 recognize type-specific antigenic determinants. J. Gen. Virol. 83:157-165.
Liljeqvist, J. A., B. Svennerholm, and T. Bergstrom. 1999. Herpes simplex virus type 2 glycoprotein G-negative clinical isolates are generated by single frameshift mutations. J. Virol. 73:9796-9802.
Liljeqvist, J. A., E. Trybala, B. Svennerholm, S. Jeansson, E. Sjogren-Jansson, and T. Bergstrom. 1998. Localization of type-specific epitopes of herpes simplex virus type 2 glycoprotein G recognized by human and mouse antibodies. J. Gen. Virol. 79:1215-1224.
Loomis, M. R., T. O'Neill, M. Bush, and R. J. Montali. 1981. Fatal herpesvirus infection in patas monkeys and a black and white colobus monkey. J. Am. Vet. Med. Assoc. 179:1236-1239.
Lopez, C., A. Arvin, and R. Ashley. 1993. Immunity to herpesvirus infections in humans, p. 397-245. In B. Roizman, R. J. Whitley, and C. Lopez (ed.), The human herpesviruses. Raven Press, Ltd., New York, N.Y.
Norcott, J. P., and D. W. Brown. 1993. Competitive radioimmunoassay to detect antibodies to herpes B virus and SA8 virus. J. Clin. Microbiol. 31:931-935.
Ohsawa, K., D. H. Black, H. Sato, and R. Eberle. 2002. Sequence and genetic arrangement of the U(S) region of the monkey B virus (cercopithecine herpesvirus 1) genome and comparison with the U(S) regions of other primate herpesviruses. J. Virol. 76:1516-1520.
Perelygina, L., I. Patrusheva, H. Zurkuhlen, and J. K. Hilliard. 2002. Characterization of B virus glycoprotein antibodies induced by DNA immunization. Arch. Virol. 147:2057-2073.
Perelygina, L., L. Zhu, H. Zurkuhlen, R. Mills, M. Borodovsky, and J. K. Hilliard. 2003. Complete sequence and comparative analysis of the genome of herpes B virus (cercopithecine herpesvirus 1) from a rhesus monkey. J. Virol. 77:6167-6177.
Perelygina, L., H. Zurkuhlen, I. Patrusheva, and J. K. Hilliard. 2002. Identification of a herpes B virus-specific glycoprotein D immunodominant epitope recognized by natural and foreign hosts. J. Infect. Dis. 186:453-461.
Prince, H. E., C. E. Ernst, and W. R. Hogrefe. 2000. Evaluation of an enzyme immunoassay system for measuring herpes simplex virus (HSV) type 1-specific and HSV type 2-specific IgG antibodies. J. Clin. Lab. Anal. 14:13-16.
Saijo, M., M. Niikura, S. Morikawa, T. G. Ksiazek, R. F. Meyer, C. J. Peters, and I. Kurane. 2001. Enzyme-linked immunosorbent assays for detection of antibodies to Ebola and Marburg viruses using recombinant nucleoproteins. J. Clin. Microbiol. 39:1-7.
Sehgal, D., P. S. Malik, and S. Jameel. 2003. Purification and diagnostic utility of a recombinant hepatitis E virus capsid protein expressed in insect larvae. Protein Expr. Purif. 27:27-34.
Sisk, W. P., J. D. Bradley, R. J. Leipold, A. M. Stoltzfus, M. Ponce de Leon, M. Hilf, C. Peng, G. H. Cohen, and R. J. Eisenberg. 1994. High-level expression and purification of secreted forms of herpes simplex virus type 1 glycoprotein gD synthesized by baculovirus-infected insect cells. J. Virol. 68:766-775.
Slomka, M. J., L. Harrington, C. Arnold, J. P. Norcott, and D. W. Brown. 1995. Complete nucleotide sequence of the herpesvirus simiae glycoprotein G gene and its expression as an immunogenic fusion protein in bacteria. J. Gen. Virol. 76:2161-2168.
Stacey, S. N., J. S. Bartholomew, A. Ghosh, P. L. Stern, M. Mackett, and J. R. Arrand. 1992. Expression of human papillomavirus type 16 E6 protein by recombinant baculovirus and use for detection of anti-E6 antibodies in human sera. J. Gen. Virol. 73:2337-2345.
Su, H. K., J. D. Fetherston, M. E. Smith, and R. J. Courtney. 1993. Orientation of the cleavage site of the herpes simplex virus glycoprotein G-2. J. Virol. 67:2954-2959.
Tanabayashi, K., R. Mukai, and A. Yamada. 2001. Detection of B virus antibody in monkey sera using glycoprotein D expressed in mammalian cells. J. Clin. Microbiol. 39:3025-3030.
Thompson, S. A., J. K. Hilliard, D. Kittel, S. Lipper, W. E. Giddens, Jr., D. H. Black, and R. Eberle. 2000. Retrospective analysis of an outbreak of B virus infection in a colony of DeBrazza's monkeys (Cercopithecus neglectus). Comp. Med. 50:649-657.
Vizoso, A. D. 1975. Recovery of herpes simiae (B virus) from both primary and latent infections in rhesus monkeys. Br. J. Exp. Pathol. 56:485-488.
Ward, J. A., and J. K. Hilliard. 1994. B virus-specific pathogen-free (SPF) breeding colonies of macaques: issues, surveillance, and results in 1992. Lab. Anim. Sci. 44:222-228.
Ward, J. A., and J. K. Hilliard. 2002. Herpes B virus-specific pathogen-free breeding colonies of macaques: serologic test results and the B-virus status of the macaque. Contemp. Top. Lab. Anim. Sci. 41:36-41.
Whitley, R. J., and J. K. Hilliard. 2001. Cercopithecine herpesvirus (B virus), p. 2835-2848. In D. M. Knipe et al. (ed.), Fields virology. Lippincott Williams & Wilkins, Philadelphia, Pa.
Zhang, J. X., I. Braakman, K. E. Matlack, and A. Helenius. 1997. Quality control in the secretory pathway: the role of calreticulin, calnexin and BiP in the retention of glycoproteins with C-terminal truncations. Mol. Biol. Cell 8:1943-1954.
Zwartouw, H. T., and E. A. Boulter. 1984. Excretion of B virus in monkeys and evidence of genital infection. Lab. Anim. 18:65-70.(Ludmila Perelygina, Irina)
ABSTRACT
B virus (cercopithecine herpesvirus 1) is the only deadly alphaherpesvirus that is zoonotically transmissible from macaques to humans. The detection of humoral immune responses is the method of choice for the rapid identification of B virus-infected animals. We evaluated the diagnostic potential of recombinant B virus glycoproteins for the detection of immunoglobulin G (IgG) antibodies in monkey and human sera. Glycoproteins B, C, and E and secreted (sgG) and membrane-associated (mgG) segments of glycoprotein G (gG) were expressed in the baculovirus expression system, while gD was expressed in CHO cells. We developed recombinant protein-based IgG enzyme-linked immunosorbent assays (ELISAs) and compared their diagnostic efficacies by using B virus antibody-negative (n = 40) and -positive (n = 75) macaque sera identified by a whole antigen-based ELISA and Western blotting. The diagnostic sensitivities of the gB-, gC-, gD-, and mgG-ELISAs were 100, 97.3, 88.0, and 80.0%, respectively. The specificities of the gB-, gC-, and gD-ELISAs and of the mgG-ELISA were 100 and 97.5%, respectively. In contrast, the sensitivities and specificities of sgG- and gE-ELISAs were low, suggesting that sgG and gE are less effective diagnostic antigens. Sera from nonmacaque monkeys cross-reacted with gB, gC, and gD, and only baboon sera reacted weakly with mgG. Human herpes simplex virus type 1 (HSV-1)- and HSV-2-positive sera pools reacted with gB and gD, whereas sera from B virus-infected individuals reacted with all four antigens. These data indicate that gB, gC, gD, and mgG have a high diagnostic potential for B virus serodiagnosis in macaques, whereas mgG may be a valuable antigen for discrimination between antibodies induced by B virus and those induced by other, closely related alphaherpesviruses, including HSV-1 and -2.
INTRODUCTION
Human infection with B virus (also called cercopithecine herpesvirus 1, monkey B virus, and herpes B virus) is the most feared occupational hazard among individuals working with macaque monkeys, since fatality is often the outcome of infection, which proceeds in the absence of effective antiviral therapy (25, 56). The use of macaques in research has been steadily growing over the last decade and is expected to rise quickly in the near future due to the increasing demands for these animals for use in HIV/AIDS investigations, vaccine trials, drug testing, and research into bioterrorism agents. As macaque usage increases, frequencies of human exposures to B virus are increasing as well. Rapid and accurate diagnostic tests are urgently needed to aid in the early identification of clinical cases, which is essential for a timely initiation of antiviral therapies in zoonotically infected humans. In addition to human diagnostics, enhanced assays are required for monitoring specific-pathogen-free (SPF) macaque colonies established by the National Institutes of Health for the breeding of B virus-free animals (55), as these animals often demonstrate only very low levels of antibody.
Unfortunately, a direct diagnosis of infection by virus detection (cell culture or PCR) is impossible in most cases, since, similar to other alphaherpesviruses, B virus establishes a lifelong latency in sensory ganglia of macaques and seldom reactivates (9, 53, 58). Current diagnoses of B virus infections in humans and monkeys rely mainly on the detection of serum antibodies to B virus proteins. Indirect enzyme-linked immunosorbent assays (ELISA) and other rapid serological tests based on a solubilized, B virus-infected cell antigen have been developed and used for the identification of infected animals (8, 21, 28, 39), with subsequent confirmation by Western blotting to identify specific targets that are immunoreactive with serum antibodies (54). Serodiagnoses of zoonotic infections are performed by Western blotting, which is a time-consuming technique that requires visual interpretations of complex patterns. This method is not specific enough to unequivocally identify B virus infections in herpes simplex virus (HSV)-positive humans because antibodies to HSV type 1 (HSV-1) and HSV-2 are highly cross-reactive with B virus proteins (12, 23), making differentiation a complex task. Most importantly, currently used serological assays utilize B virus-infected cell lysates as an antigen, and these can only be produced in a maximum containment laboratory (biosafety level 4), which limits the number of facilities that are capable of providing antigen. Antigens may also suffer from lot-to-lot variation, compromising outcome measures based on assays using these antigens.
Recombinant-based serological assays have been developed for the diagnosis of many viral infections, including human cytomegalovirus (11), hepatitis C virus (27), hepatitis E virus (46), human papillomavirus (49), Ebola virus (45), and many others. Several recombinant glycoprotein G (gG)-based immunoassays for HSV type-specific serodiagnosis are commercially available (19, 44). However, the use of recombinant antigens for B virus serodiagnosis has not been widely investigated. Recently, recombinant gD was shown to be useful for B virus serodiagnosis by dot blot and Western blot assays, but the performance of this antigen in ELISAs was not studied (51). In an earlier study, we produced a fusion protein containing a single B virus-specific immunodominant epitope of gD and demonstrated its efficacy for the identification of B virus infections by using an indirect ELISA (43). The serodiagnosis of infections, however, cannot be based exclusively upon the presence of antibodies to a single epitope of a pathogen due to variations in individual responses to a selected epitope. Moreover, the existence of cross-reacting antibodies against similar epitopes in other proteins may result in a false-positive diagnosis. Ideally, the identification of an appropriate cocktail of recombinant proteins would provide the opportunity to detect the broad range of differentially induced antibodies during the course of infection, including different antibody subclasses and isotypes.
Individual polypeptide targets of antibody responses following HSV infections have been studied by several investigators. Most immune sera from HSV-infected individuals react with 12 to 20 viral polypeptides in immunoblots. The envelope glycoproteins gB, gD, gC, gE, and gG, the major capsid protein VP5, and the nucleocapsid complex p40 have been identified as primary targets of IgG antibody responses in patients with HSV-1 and/or HSV-2 infections (2, 5, 13, 14, 38). The B virus's homology to HSV-1 (79.9% amino acid identity for gB, 57% identity for gD, 49.9% identity for gC, 46% identity for gE, and 29.2% identity for gG [42]) suggests that the specificity of B virus-induced antibody responses is similar to that of responses induced by HSV types 1 and 2. Ideally, these glycoproteins may be primary candidates for evaluation as recombinant diagnostic reagents.
We report for the first time the production and purification of recombinant B virus antigens, specifically gB, gC, gE, the secreted segment of gG (sgG), and the membrane-associated segment of gG (mgG), by use of a baculovirus expression system and an investigation of immunogenic and antigenic properties of these proteins, along with recombinant gD produced in CHO cells. Diagnostic efficacies of indirect ELISAs using each recombinant protein as a coating antigen (Rec-ELISAs) were determined with a large panel of positive (n = 75) and negative (n = 40) serum samples from rhesus and cynomolgus macaque monkeys representing a spectrum of antibody levels. Reactivities of sera from different monkey species and of human sera containing HSV-1, HSV-2, and B virus-specific antibodies with recombinant B virus antigens were also studied to identify antigens that are suitable for a B virus differential serodiagnosis for both natural and inadvertent foreign hosts.
MATERIALS AND METHODS
Viruses, cells, and media. All cells, vectors, and supplies were purchased from Invitrogen (Carlsbad, Calif.) unless otherwise noted. Plasmids were propagated in Escherichia coli Top10 cells. Recombinant baculoviruses were produced and titrated in Spodoptera frugiperda (Sf-9) cells. Sf-9 cells were maintained at 27°C in serum-free Sf-900 II SFM insect cell culture medium with antibiotics.
Construction of baculovirus donor plasmid pFBMV5H. The expression cassette of the Bac-to-Bac donor plasmid pFastBacHTc was modified by the addition of a honeybee melittin secretion sequence (HBM) and the V5 epitope coding sequence. The HBM sequence was amplified by PCR from another transfer vector, pMelBac, by use of the forward primer Mel-RsrII and the reverse primer RecBac778 (Table 1). The amplified fragment was cloned into a pCR-XL-TOPO plasmid and then excised with the RsrII and EcoRI restriction enzymes from the resulting recombinant plasmid. A fragment encoding V5 epitope and His tags was excised from a pcDNA3.1A vector with the EcoRI and PmeI restriction enzymes. The pFastBacHTc plasmid was digested with HindIII, its ends were blunted with the Klenow fragment, and then the plasmid was digested with RsrII. Both fragments were ligated into the linearized pFastBacHTc plasmid to produce a new donor plasmid, pFBMV5H. The introduced sequences were confirmed to be correct and in frame by sequencing of the pFBMV5H DNA by use of a BigDye Terminator sequencing kit (ABI PRISM).
Protein analyses. Analyses of protein structures were performed with the DNASTAR Protean program (version 4.0.3), the signal peptide prediction program SignalP v.2.0.b2 (http://www.cbs.dtu.dk/services/SignalP-2.0/), and the transmembrane prediction program TMHMM v2.0 (http://www.cbs.dtu.dk/services/TMHMM/).
Preparation of recombinant baculoviruses. The DNA segments encoding the extracellular domains of the B virus glycoproteins B, C, E, and G were amplified from the genomic DNA of B virus strain E2490 by PCRs with the primers listed in Table 1. PCR fragments were cloned into the pCR-TOPO-XL vector. Recombinant plasmids were sequenced to verify the identities of the inserts. Cloned fragments were then excised with the HindIII and EcoRI restriction enzymes and cloned into the corresponding sites of the pFBMV5H donor plasmid to obtain gB, gC, gE, sgG, and mgG donor plasmids.
The transposition of expression cassettes into a baculovirus shuttle vector (bacmid) was performed by the transformation of E. coli DH10Bac competent cells with the constructed donor plasmids. Successful transpositions were verified by PCR amplification of the prepared bacmid DNAs with gene-specific forward primers and the M13/pUC reverse primer. Recombinant viruses were produced by transfecting monolayers of Sf-9 cells grown in six-well cell culture plates with bacmid DNAs by use of the Cellfectin reagent. At 72 h postinfection, virus-containing medium from the transfected cells was harvested and titrated by a plaque assay on Sf-9 cell monolayers. Five individual plaques were picked from each transfection for further analysis. High-titer stocks of recombinant viruses were then produced in Sf-9 cells. All procedures were performed in accordance with the manual for the Bac-to-Bac baculovirus expression system.
Optimization of recombinant protein expression. Suspension cultures of Sf-9 cells (30 ml each) grown in 100-ml spinner flasks were infected with each recombinant virus at multiplicities of infection (MOIs) of 1, 4, and 7. Supernatant samples were collected at 0, 24, 48, 72, and 96 h postinfection, mixed 1:1 with 2x sodium dodecyl sulfate (SDS) buffer, boiled for 5 min, fractionated by SDS-10% polyacrylamide gel electrophoresis (SDS-10% PAGE), and then analyzed by immunoblotting with an anti-V5 monoclonal antibody (MAb) (1:2,000 dilution) (Sigma, St. Louis, Mo.). The largest quantities of nondegraded proteins were produced at 72 h postinfection at MOIs of 4 and 7.
Production and purification of recombinant proteins. Sf-9 suspension cultures were grown in 250-ml spinner flasks to a density of 1.6 x 106 cells/ml and then were infected with the recombinant viruses at an MOI of 4. Media were harvested at 72 h postinfection and centrifuged at 15,000 x g for 10 min at 4°C. Supernatants were collected and either stored at –80°C or used immediately for purification. The collected supernatants were concentrated and dialyzed against column buffer (40 mM sodium phosphate [pH 7.8], 0.3 M NaCl) by tangential flow filtration with a Labscale TFF system containing a Pellicon XL device (10-kDa cutoff) (Millipore, Bedford, Mass.). The concentrated supernatant was then loaded onto a 1-ml HiTrap chelating column (Amersham Biosciences, Piscataway, N.J.) charged with Ni2+ ions and equilibrated with column buffer. After washing of the column with 10 ml of column buffer supplemented with 20 mM imidazole, recombinant proteins were eluted with column buffer containing 250 mM imidazole. Fractions were analyzed by immunoblotting with the anti-V5 MAb. Those containing recombinant protein were pooled and dialyzed against phosphate-buffered saline (PBS), and aliquots were stored at –80°C. Recombinant B virus gD was produced in suspension cultures of Chinese hamster ovary cells that were stably transfected with the gD expression construct (L. Perelygina, H. Zurkuhlen, I. Patrusheva, N. Patrushev, and J. K. Hilliard, Abstr. 25th Int. Herpesvirus Workshop, abstr. 5.50, 2000).
Western blot analysis. Lysates of uninfected Vero cells and of Vero cells infected with B virus lab strain E2490 were prepared as described previously (28). Proteins were fractionated by SDS-10% PAGE and then either stained in a gel with Coomassie blue (GelCode Blue staining reagent; Pierce, Rockford, Ill.) or transferred onto a nitrocellulose membrane. The membrane was blocked with BLOTTO (5% skim milk, 1% normal goat serum, and 0.05% Tween 20 in PBS) for 1 h and then incubated for 1 h with primary antibodies diluted in BLOTTO. After being washed three times for 10 min in PBS-0.05% Tween 20, the membrane was incubated for 1 h with peroxidase-conjugated anti-rabbit, anti-mouse, or anti-human immunoglobulin G (IgG) (1:40,000 dilution in BLOTTO) (Pierce). The membrane was washed again, and bands were visualized on Kodak X-Omat film after detection with ECL Western blot detection reagents (Amersham Biosciences).
Mass spectrometry. Recombinant protein bands were excised from an SDS-10% PAGE gel and subjected to in-gel digestion by modified trypsin (Promega, Madison, Wis.) as suggested by the manufacturer. Matrix-assisted laser desorption ionization-time of flight mass spectrometry of the resulting digests was performed on a Voyager-DE PRO MALDI-TOF workstation (Applied Biosystems, Framingham, Mass.) by the Advanced Biotechnology Core Facility at Georgia State University. For protein identification, the measured masses of the tryptic peptides were compared with published databases by use of the MS-FIT module of Protein Prospector software (http://prospector.ucsf.edu/ucsfhtml4.0/msfit.htm).
Production of rabbit antisera. Two hundred fifty micrograms of each purified glycoprotein was diluted in 0.5 ml of PBS, mixed with an equal volume of Freund's complete adjuvant (Sigma), and injected subcutaneously into multiple sites of New Zealand White female rabbits. Rabbits were given subcutaneous booster injections with 250 μg of each protein in Freund's incomplete adjuvant (Sigma) every 4 weeks for 6 months. Preimmune sera were collected at the time of the first immunization. Test and production bleeds were performed on the 14th day after each injection. The end-point Western blot (WB) titer of each serum was determined by immunoblotting of membrane strips containing fractionated B virus antigen with twofold serial dilutions of rabbit antisera. The titer was determined as the reciprocal of the lowest serum dilution at which no reactivity with B virus proteins was detected.
Monkey and human sera. All sera were kindly provided by the National B Virus Resource Center (Atlanta, Ga.). B virus antibody-negative and -positive serum pools from macaques and HSV-negative, HSV-1-positive, and HSV-2-positive serum pools from humans were prepared as described previously (43). Antibody-positive serum pools from baboons, langurs, sooty mangabeys, and African green monkeys infected with herpesvirus papio 2, herpes virus langur, mangabey herpesvirus, and simian agent 8, respectively, were described and characterized previously (29). B virus antibody-positive (n = 75) or -negative (n = 40) rhesus and cynomolgus serum samples were randomly selected for assessments of macaque serum reactivities with recombinant proteins. All sera were serotyped by the B virus diagnostic laboratory by use of the previously described whole B virus antigen-based ELISA and Western blot assays (28, 32). ELISA titers of positive sera ranged from 600 to 230,000.
Rec-ELISA. The optimal concentrations of antigens, sera, and conjugate were determined by checkerboard titration. Purified recombinant proteins were diluted to a concentration of 1 μg/ml (gB) or 2 μg/ml (gC, gD, gG, and gE) in 20 mM Tris-HCl, pH 8. Wells of Maxisorp immunoplates (Nalge Nunc International, Rochester, N.Y.) were coated with antigens (100 μl/well) for 1 h on a plate shaker at room temperature and then overnight at 4°C. The plates were washed with 20 mM Tris-HCl, pH 8, to remove unbound antigens and were blocked with 150 μl of ELISA-BLOTTO (borate-buffered saline containing 2.5% [each] nonfat dry milk and liquid gelatin)/well. Blocking and subsequent incubations with sera and conjugate were done for 1 h on a plate shaker at room temperature. Three washes with borate-buffered saline-0.05% Tween 20 were performed after each incubation step by use of an automatic plate washer. Sera were diluted 1:50 in ELISA-BLOTTO, and 50 μl was added to each well. Alkaline phosphatase-conjugated goat anti-human IgG Fc fragment (Sigma) was diluted 1:4,000 (for tests with monkey sera) or 1:6,000 (for tests with human sera) in ELISA-BLOTTO, and 50 μl was added to each well. A substrate solution was prepared by dissolving p-nitrophenyl phosphate (pNPP) tablets (Sigma) in 1 M diethanolamine buffer containing 0.5 mM MgCl2, pH 9.8 (1 mg of pNPP/ml), and 200 μl of a freshly prepared mixture was added to each well. After a 25-min incubation at room temperature, the reaction was stopped by adding 50 μl of 3 N NaOH per well, and the optical densities (ODs) were read on an automatic microplate reader at 405 nm. B virus antibody-positive and -negative rhesus monkey serum pools were included in each plate to serve as positive and negative controls and for use in calculations of positive-to-negative (P/N) ratios. P/N values were determined as follows: the OD value of a test or positive control serum reacted with an antigen was divided by the OD value of the negative control serum reacted with the same antigen. A test was considered valid if the P/N value of the positive control serum was >2.2.
RESULTS
The Bac-to-Bac baculovirus expression system allows the rapid and efficient generation of multiple recombinant baculoviruses. However, the expression cassette of the commercial donor plasmids from the Bac-to-Bac system does not contain specific DNA sequences that allow a recombinant fusion protein to be secreted into media and to be easily detected. We constructed a new donor plasmid, pFBMV5H, from pFastBacHTc by adding the HBM secretion signal to achieve a high level of protein secretion and the V5 epitope to enhance the detection of recombinant proteins. A diagram of the expression cassette of pFBMV5H is shown in Fig. 1A.
The locations of the signal peptides and transmembrane domains in B virus glycoproteins (Fig. 1B) were predicted with the SignalP and TMHMM programs. The location of the cleavage site in the B virus gG precursor was predicted based on data available for the homologous HSV-2 protein (50). DNA fragments encoding the entire extracellular domains of gB, gC, gE, sgG, and mgG were amplified by PCR from the genomic DNA of B virus E2490 and then cloned into the BamHI and EcoRI sites of the donor plasmid pFBMV5H. The sequences of the cloned fragments were identical to the genomic sequences of the corresponding B virus genes (42), with the exception of the mgG fragment, as there was a 132-bp deletion (positions 1096 to 1227 in the gG gene) in the constructed mgG donor plasmid. This deleted DNA fragment, which encodes a PAPTTT amino acid repeat, is present in the gG sequence of the E2490 strain but is not found in any of the other sequenced gG genes of B virus clinical isolates (40). Since the deletion construct encodes an mgG protein that actually mimics those produced during natural infections, we decided to use this construct for production of the mgG recombinant baculovirus. Five recombinant baculoviruses were subsequently constructed as described in Materials and Methods and were designated bac-gB37-719, bac-gC27-425, bac-gE29-408, bac-gG22-310, and bac-gG311-613.
To produce large amounts of recombinant proteins, we infected Sf-9 suspension cultures with the recombinant viruses at an MOI of 4 and harvested the media at 72 h postinfection. Recombinant proteins were affinity purified from the supernatants of infected insect cells by nickel chromatography. Western blotting with a V5 MAb identified the column fractions containing recombinant proteins. The peak column fractions were then pooled. After measuring the protein concentrations in the pooled fractions, we estimated the following yields of purified proteins: 6 mg of gB, 10 mg of gC and mgG, and 1 mg of gE and sgG per 1 liter of insect cell culture medium. Substantial quantities of recombinant gE and sgG were retained in the infected cells, as determined by immunoblotting of infected cell pellets and supernatants with the anti-V5 MAb (data not shown), resulting in a low level of secretion and low yields of purified gE and sgG.
All recombinant proteins migrated in an SDS gel as broad bands with larger molecular masses than predicted, most probably due to glycosylation (Fig. 2A; Table 2), since multiple N-glycosylation signals were found in all proteins (Fig. 1B). Similarly, cotranslational N-glycosylation of baculovirus-expressed HSV-1 glycoproteins resulted in a decrease in their electrophoretic mobilities in denaturing polyacrylamide gels (15, 16, 47). The anti-V5 MAb reacted strongly with all of the recombinant proteins (Fig. 2B), but a macaque B virus-positive serum pool recognized only recombinant gB, gC, and mgG (Fig. 2C). Interestingly, mgG and sgG migrated as multiple bands, and each reacted with the anti-V5 MAb and B virus-positive macaque sera. To verify the identities of the produced proteins, we cut the protein bands from an SDS-PAGE gel, digested them with trypsin, and analyzed the resulting fragments by mass spectrometry. Comparisons of the tryptic profiles of the recombinant proteins with published databases confirmed their identities as B virus glycoproteins gB, gC, gE, and gG.
To verify the immunogenicity of the recombinant proteins and to further validate their authenticity, we used each of the five glycoproteins to immunize rabbits to induce monospecific, polyclonal antibodies. With the exception of sgG (WB titer, <800), each glycoprotein induced a high-titer antibody in rabbits (WB titer, <32,000). Western blot analyses revealed that each immune serum reacted with specific protein bands from the B virus-infected cell lysates (Fig. 3), whereas preimmune sera did not (data not shown). Rabbit anti-gB, -gC, and -gE sera reacted with B virus polypeptides of 120, 72, and 80 kDa, respectively. Sera directed against the different gG fragments recognized multiple protein bands: the 85-kDa band was recognized by sgG antisera, the 160-, 48-, and 20-kDa polypeptides were recognized by mgG antisera, and the 115-kDa band was recognized by both types of antisera. The 48- and 20-kDa bands were weak and could be seen only after long exposures of immunoblots to X-ray films. The polypeptides recognized by the produced monospecific antisera were in size agreement with the previously reported molecular masses of selected B virus glycoproteins (7, 41, 48). These data validate that the antigenic and immunogenic properties of the prepared recombinant proteins are similar to those of viral glycoproteins produced during B virus infection.
The performance of each baculovirus-produced recombinant, i.e., gB, gC, gE, sgG, and mgG, in an ELISA was subsequently evaluated. Recombinant gD purified from the medium of stably transfected CHO cells was also included in the evaluation. To determine the optimal coating concentration for microtiter plates (producing the highest OD and P/N values), we performed a checkerboard titration with positive and negative serum pools from macaques. The optimal amount of recombinant antigen per well was 50 ng for gB and 100 ng for all other proteins.
The results of Rec-ELISA showed that the B virus antibody-positive serum pool reacted strongly with gB, gC, gD, and mgG, weakly with gE, and not at all with sgG (Fig. 4), confirming the Western blot results for these proteins. To further assess the value of sgG and gE for serodiagnosis, we studied the reactivities of individual negative (n = 10) and positive (n = 10) macaque sera with these proteins at a 1:50 dilution. For sgG-ELISA, positive samples reacted weakly with sgG (OD values varied from 0.16 to 0.83), suggesting that the secreted portion of B virus gG did not stimulate the production of a high level of antibodies during natural B virus infection and, therefore, that sgG has limited value for serodiagnosis. For gE-ELISA, both positive and negative macaque sera reacted with gE, producing OD values in the ranges of 0.3 to 1.3 and 0.2 to 0.66, respectively. Moreover, an HSV and B virus antibody-negative human pool reacted very strongly with gE (OD value of 2.9). This gE property renders the produced recombinant gE antigen unsuitable for diagnostic purposes, at least in this form. However, more specific epitopes of gE may be useful immunodiagnostic reagents.
Extended evaluations of the gB, gC, gD, and mgG antigens were performed with large panels of individual macaque sera. B virus antibody-positive (n = 75) and -negative (n = 40) serum panels included approximately equal numbers of samples from rhesus and cynomolgus macaques, since these macaque species are used most often for biomedical research. To determine the extents of IgG antibody reactivities with gB, gC, gD, and mgG, we used each recombinant antigen separately to coat individual microplate wells and to react with each serum from the selected panels. The highest average immunoreactivity of the positive serum panel (mean OD value ± 1 standard deviation [SD]) occurred with gB (2.23 ± 0.69), followed closely by gC (1.58 ± 1.00), mgG (1.05 ± 1.06), and gD (0.67 ± 0.48) (Fig. 5, solid horizontal lines). Cutoff values for positive results were determined from Rec-ELISA results by use of the negative serum panel. The reactivities of negative sera were low with gC, gD, and mgG (OD values of 0.06 to 0.2) but were high with gB (OD values of 0.4 to 0.5) due to binding of the conjugate to this antigen. Later, we identified a different lot of conjugate without cross-reactions with gB, and we currently use it to perform Rec-ELISAs. The OD cutoff values were calculated separately for each antigen as the mean of the results from 40 negative sera plus 3 SD (Fig. 5, dashed lines). The P/N values of negative sera for each antigen ranged from 0.3 to 3.7. The P/N cutoff value of 2.2 was determined as the mean of the P/N values of negative sera plus 3 SD.
According to the established cutoffs, all B virus-positive samples (n = 75) were positive for gB antibodies, 73 were positive for gC antibodies, 66 were positive for gD antibodies, and 60 were positive for mgG antibodies, resulting in diagnostic sensitivities of 100, 97.3, 88.0, and 80.0% for gB-, gC, gD-, and mgG-ELISAs, respectively, compared to the reference assays (BV-ELISA and Western blotting). Antibodies to all four glycoproteins were found in 76% of positive sera. Rec-ELISA results for negative samples revealed one discordant result between mgG-ELISA and the reference assays. Therefore, the relative diagnostic specificity of the gB-, gC-, and gD-ELISAs was 100%, and the relative specificity of the mgG-ELISA was 97.5%. Notably, the majority of mgG antibody-negative samples (11 of 15) were from cynomolgus macaques, and consequently, the diagnostic sensitivities of mgG-ELISAs for cynomolgus sera (n = 38) and rhesus sera (n = 37) differed substantially, at 71.1 and 89.2%, respectively. These data indicate that all four recombinant antigens have excellent diagnostic potential for macaque testing. The mgG antigen, however, might be well suited only for the serodiagnosis of infections with rhesus strains of B virus, as further studies are needed to assess the usefulness of this antigen for cynomolgus serum testing.
The serodiagnosis of B virus infection in foreign hosts, e.g., humans or nonmacaque, Old World monkeys, is often complicated by the presence of antibodies that are cross-reactive with species-specific alphaherpesviruses. To find out whether the recombinant proteins have immunodiagnostic value for identifying B virus infections in foreign hosts, we determined the extent to which antibodies induced by closely related simplex alphaherpesviruses in foreign hosts reacted with the B virus recombinant antigens. Nonmacaque monkey serum pools were prepared from samples containing high antibody titers against species-specific alphaherpesviruses, and thus Rec-ELISAs were performed with sera diluted 1:500 for the gB antigen and 1:100 for the other three recombinant antigens. All human sera were tested at a dilution of 1:50. Rec-ELISA data for each positive serum were normalized to OD values obtained with a negative serum pool from the same species and then expressed as P/N values (Table 3). The threshold value was set at 2.2. As expected, gB antibodies from all tested sera were the most cross-reactive, producing the highest OD values and P/N ratios. All serum pools from nonmacaque monkeys cross-reacted with the gD and gC antigens, but only baboon sera showed reactivities with mgG that were above the threshold. Individual serum samples from B virus-infected humans (both HSV negative and HSV-1 positive) reacted with all B virus recombinants, whereas pooled sera from HSV-positive and B virus-negative patients recognized only the gB and gD antigens. Interestingly, mgG appears to have the best diagnostic value for the identification of B virus zoonoses, since HSV-positive sera as well as sera from nonmacaque monkeys infected with herpes simplex viruses did not react with this protein in Rec-ELISAs. Furthermore, HSV antibody-positive serum pools did not react detectably with gC, suggesting that gC may be an additional useful antigen for human serodiagnosis.
DISCUSSION
The identification of herpesvirus infections is generally based upon indirect analyses of antibodies because the viruses elude detection. Interpretations of antibody reactivity can be complicated due to the plethora of cross-reactive interactions among antibodies and closely related viral antigens. These challenges can be reduced by the use of virus-specific proteins for the detection of antibodies. In this study, we presented a comparative analysis of the diagnostic capacities of recombinant B virus glycoproteins, i.e., gB, gC, gD, gE, sgG, and mgG, versus that of a whole B virus antigen with respect to the diagnostic efficacy of each for the detection of B virus-specific antibodies. The selection of these glycoproteins was based upon extensive Western blot analyses of tens of thousands of macaque serum samples, representing many time points after B virus infection, to identify virus protein targets of the host immune system as part of the National B Virus Resource Center (www.gsu.edu/bvirus).
In the first part of this study, we reported the cloning of B virus gB, gC, gE, sgG, and mgG into baculovirus by use of the modified donor vector pFBMV5H. Production and affinity purification of each recombinant glycoprotein resulted in lower relative yields of gE and sgG (1.0 mg/liter) than of gB, gC, and mgG (6 to 10 mg/liter). The data indicated that gE and sgG are secreted into the medium at reduced levels, although the insect-specific secretion signal was identical for each recombinant. These glycoproteins appeared to remain bound to the cell, as shown by a Western blot analysis of cells. Since the correct folding of expressed proteins in the endoplasmic reticulum is known to be essential for the effective excretion of secretory and membrane proteins (10, 22, 57), one possible explanation for the relatively low yields of gE and sgG may be that these truncated forms were incorrectly processed in insect cells and, as a result, were misfolded and retained in the endoplasmic reticulum. In similar studies by other investigators using HSV-2, notably less protein was secreted by insect cells infected with sgG-2- and full-size recombinant gG-2-expressing baculoviruses than with an mgG-2-expressing baculovirus (26). The production of purified gE and sgG in our studies, however, resulted in yields sufficient to perform 10,000 tests with 100 ng of each protein per diagnostic sample.
We next demonstrated by Western blotting and ELISA that the baculovirus-produced gB, gC, and mgG glycoproteins are recognized by sera from naturally infected hosts, establishing the essential retention of antigenic epitopes found in the authentic viral glycoprotein counterparts. The consistent absence of detectable antibody reactivities with recombinant gE and sgG in Western blots suggested either a lack of gE and/or sgG authenticity, resulting in a failure of naturally induced antibodies to recognize these recombinant proteins, or that linear epitopes of these viral proteins are not targeted strongly by host humoral immunity during natural infection. Since the authenticity of the recombinants was verified by mass spectrometry, we suggest that humoral responses to B virus infections may not include reactivities to linear (continuous) epitopes in gE and sgG. In accordance with our data, a lack of antibodies against gE and sgG linear epitopes was also demonstrated with sera from HSV-infected patients (30, 34).
Recombinant viral glycoproteins were used to produce rabbit antibodies that were then verified to be immunoreactive with each authentic viral glycoprotein present in B virus-infected cell lysates. Each recombinant protein elicited strong antibody responses in rabbits, with the exception of sgG. For ELISAs, the sgG recombinant protein was weakly reactive with sera from naturally infected macaques, suggesting that sgG is simply not capable of inducing a strong B-cell response, nor is its authentic counterpart from B virus. Similar observations have been made for HSV-2-infected humans with sgG-2, which induces only a weak antibody response late in infection relative to membrane-associated gG (17).
The next critical phase of analysis of the B virus recombinant reagents for use in immunodiagnosis established their value for the detection of IgG responses induced in infected macaques as well as humans. Four of the six recombinant reagents, when used in ELISA, proved to be both specific and sensitive relative to a whole B virus antigen-based ELISA. Antibody detection by ELISA was most sensitive when we used recombinant gB to evaluate defined panels of rhesus and cynomolgus macaque sera. Thus, gB appears to be the most promising candidate for immunodiagnostics. However, gB is a highly conserved protein, and therefore homologous proteins of other macaque herpesviruses, e.g., simian varicella virus (SVV), may have the capability to induce antibodies that are cross-reactive with gB. Several SVV strains have been isolated, including macaque strains (18), but SVV antibody prevalence in macaque populations has not been reported to date. A sequence comparison of the B virus and SVV gB proteins (GenBank accession no. AAA58703) revealed a 48% amino acid identity between their ectodomains, suggesting that SVV gB polyclonal antibody responses may include antibodies that cross-react with B virus gB, although early limited studies in our lab showed no cross-reactivity between B virus and SVV for sera from naturally infected macaques. The ability of several human herpesviruses, such as HSV-1, HSV-2, varicella-zoster virus, cytomegalovirus, and Epstein-Barr virus, to induce cross-reactive gB polyclonal antibodies, however, has been reported repeatedly in the literature (4, 6, 31). Future studies will be performed to investigate the cross-reactivity between SVV and B virus, but even if they are cross-reactive, the use of appropriate combinations of B virus recombinant proteins, e.g., gC, gD, and mgG, determined the specificity of induced antibodies with excellent sensitivities (97.3, 88.0, and 80.0%, respectively) for nearly 100% of the samples tested for this study. Most importantly, these reagents eliminate the need for and the danger associated with the production of large-scale lots of B virus for immunodiagnostics.
We were very interested in whether recombinant B virus gG would prove to be a valuable immunodiagnostic reagent in view of its usefulness for the differentiation of HSV types 1 and 2. Of two produced gG recombinants, only mgG performed satisfactorily in Rec-ELISA, demonstrating 80% relative diagnostic sensitivity and 97.5% specificity with our test sera. The secreted N-terminal fragment of gG appeared to have little diagnostic value, as low serum reactivity was observed with the sgG antigen even when readily detectable mgG-specific antibodies were present. A recently developed sgG-2-based ELISA for HSV-2 antibody detection similarly demonstrated a reduced sensitivity relative to ELISA with mgG-2 or whole gG-2 (17, 34). We demonstrated in this study that the diagnostic sensitivity of the mgG-based ELISA is lower than that of the gB-, gC-, or gD-based ELISA. Further work will be required to understand these observations. With similar observations from HSV studies, investigators have speculated several possibilities that encompass aspects of both virus and host correlates of infection, e.g., late kinetics of gG antibody development, infections with gG null mutants, or a quick waning of gG antibodies (3, 20, 35, 38). Future studies are planned to determine the value of specific gG epitopes for the differentiation of B virus-induced antibodies in different species of macaques and in humans.
The ability of these reagents to differentiate antibodies induced against human herpesviruses is important to our evaluations because of the large number of potential exposures to B virus among members of the biomedical community working with macaque monkeys. Many exposures have included individuals with a positive history of HSV antibodies. For this reason, we were particularly interested in mgG- and gC-based ELISAs, as immunodominant, type-specific epitopes are clustered in the C-terminal region of the closely related HSV-2 gG (33, 36) and because type-specific gC responses have also been noted (1). The experiments described in this article demonstrate that HSV antibody pools do not react in the mgG- or gC-based ELISA for the detection of B virus, whereas zoonotically B virus-infected humans do react unambiguously. Importantly, one serum was obtained from a symptomatic patient on day 30 after the onset of disease (24), whereas another, convalescent-phase serum was obtained from the person at >10 years postinfection. Although only two samples were analyzed, our results suggest that B virus infection in humans elicits type-specific gG and gC antibody responses. However, the performance of gC and mgG antigens in ELISA must be further investigated by use of a large panel of individual human sera before we apply these recombinant antigens to B virus serodiagnosis for humans.
An added value of these specific immunodiagnostic reagents is the fact that selected recombinants can be used to detect B virus-specific antibodies in nonmacaque species that are coinfected with closely related simplex alphaherpesviruses. Antibodies induced by these viruses from sooty mangabeys, langurs, baboons, and African green monkeys reacted either poorly or not at all with mgG in ELISAs. Cross-species infections in nonhuman primates can thus be identified with confidence, which is essential for the prevention of B virus outbreaks in primate colonies housing different monkey species in close proximities (37, 52).
In summary, we showed with this work that the recombinant B virus antigens gB, gC, gD, and mgG are valuable diagnostic reagents for the identification of B virus infections in macaques. A mixture of the recombinant antigens is likely to provide the highest sensitivity and specificity in B virus diagnostic tests. In addition to being relatively economical to produce and easy to subject to quality control and standardization, recombinant proteins are safe to use in laboratories that are not equipped with biocontainment facilities. Because the early identification of B virus infections in animals within a primate colony, and especially in newly acquired animals, is important for preserving the integrity of an SPF colony and for preventing additional fatal zoonotic infections, these highly standardized recombinant reagents should prove to be invaluable for the design of assays that can be performed readily in a wide variety of clinical laboratory settings.
ACKNOWLEDGMENTS
We thank the personnel of the National B Virus Resource Laboratory for providing monkey and human serum samples. We are grateful to Deborah Walthall for performing mass spectrometry analysis.
This study was supported by Public Health Service grants R01 RR03162 and P40 RR05062 from the NIH National Center for Research Resources and by the generous and continued support of the Georgia Research Alliance.
Present address: Division of Cardiology, Department of Medicine, Emory University, Atlanta, GA 30322.
REFERENCES
Ackermann, G., F. Ackermann, H. J. Eggers, U. Wieland, and J. E. Kuhn. 1998. Mapping of linear antigenic determinants on glycoprotein C of herpes simplex virus type 1 and type 2 recognized by human serum immunoglobulin G antibodies. J. Med. Virol. 55:281-287.
Ashley, R., J. Benedetti, and L. Corey. 1985. Humoral immune response to HSV-1 and HSV-2 viral proteins in patients with primary genital herpes. J. Med. Virol. 17:153-166.
Ashley, R. L., J. Militoni, F. Lee, A. Nahmias, and L. Corey. 1988. Comparison of Western blot (immunoblot) and glycoprotein G-specific immunodot enzyme assay for detecting antibodies to herpes simplex virus types 1 and 2 in human sera. J. Clin. Microbiol. 26:662-667.
Balachandran, N., D. E. Oba, and L. M. Hutt-Fletcher. 1987. Antigenic cross-reactions among herpes simplex virus types 1 and 2, Epstein-Barr virus, and cytomegalovirus. J. Virol. 61:1125-1135.
Bernstein, D. I., Y. J. Bryson, and M. A. Lovett. 1985. Antibody response to type-common and type-unique epitopes of herpes simplex virus polypeptides. J. Med. Virol. 15:251-263.
Bernstein, D. I., L. M. Frenkel, Y. J. Bryson, and M. G. Myers. 1990. Antibody response to herpes simplex virus glycoproteins gB and gD. J. Med. Virol. 30:45-49.
Blewett, E. L., D. Black, and R. Eberle. 1996. Characterization of virus-specific and cross-reactive monoclonal antibodies to herpesvirus simiae (B virus). J. Gen. Virol. 77:2787-2793.
Blewett, E. L., J. T. Saliki, and R. Eberle. 1999. Development of a competitive ELISA for detection of primates infected with monkey B virus (herpesvirus simiae). J. Virol. Methods 77:59-67.
Boulter, E. A. 1975. The isolation of monkey B virus (herpesvirus simiae) from the trigeminal ganglia of a healthy seropositive rhesus monkey. J. Biol. Stand. 3:279-280.
Braakman, I., and E. van Anken. 2000. Folding of viral envelope glycoproteins in the endoplasmic reticulum. Traffic 1:533-539.
Daiminger, A., U. Bader, and G. Enders. 1998. An enzyme linked immunoassay using recombinant antigens for differentiation of primary from secondary or past CMV infections in pregnancy. J. Clin. Virol. 11:93-102.
Eberle, R., D. Black, and J. K. Hilliard. 1989. Relatedness of glycoproteins expressed on the surface of simian herpes-virus virions and infected cells to specific HSV glycoproteins. Arch. Virol. 109:233-252.
Eberle, R., S. W. Mou, and J. A. Zaia. 1985. The immune response to herpes simplex virus: comparison of the specificity and relative titers of serum antibodies directed against viral polypeptides following primary herpes simplex virus type 1 infections. J. Med. Virol. 16:147-162.
Eberle, R., S. W. Mou, and J. A. Zaia. 1984. Polypeptide specificity of the early antibody response following primary and recurrent genital herpes simplex virus type 2 infections. J. Gen. Virol. 65:1839-1843.
Ghiasi, H., R. Kaiwar, A. B. Nesburn, and S. L. Wechsler. 1992. Baculovirus expressed herpes simplex virus type 1 glycoprotein C protects mice from lethal HSV-1 infection. Antivir. Res. 18:291-302.
Ghiasi, H., R. Kaiwar, A. B. Nesburn, and S. L. Wechsler. 1992. Expression of herpes simplex virus type 1 glycoprotein B in insect cells. Initial analysis of its biochemical and immunological properties. Virus Res. 22:25-39.
Gorander, S., B. Svennerholm, and J. A. Liljeqvist. 2003. Secreted portion of glycoprotein G of herpes simplex virus type 2 is a novel antigen for type-discriminating serology. J. Clin. Microbiol. 41:3681-3686.
Gray, W. L., and N. J. Gusick. 1996. Viral isolates derived from simian varicella epizootics are genetically related but are distinct from other primate herpesviruses. Virology 224:161-166.
Groen, J., B. Hersmus, H. G. Niesters, W. Roest, G. van Dijk, W. van der Meijden, and A. D. Osterhaus. 1999. Evaluation of a fully automated glycoprotein G-2 based assay for the detection of HSV-2 specific IgG antibodies in serum and plasma. J. Clin. Virol. 12:193-200.
Hashido, M., F. K. Lee, S. Inouye, and T. Kawana. 1997. Detection of herpes simplex virus type-specific antibodies by an enzyme-linked immunosorbent assay based on glycoprotein G. J. Med. Virol. 53:319-323.
Heberling, R. L., and S. S. Kalter. 1987. A dot-immunobinding assay on nitrocellulose with psoralen inactivated herpesvirus simiae (B virus). Lab. Anim. Sci. 37:304-308.
Helenius, A., T. Marquardt, and I. Braakman. 1992. The endoplasmic reticulum as a protein-folding compartment. Trends Cell Biol. 2:227-231.
Hilliard, J. K., D. Black, and R. Eberle. 1989. Simian alphaherpesviruses and their relation to the human herpes simplex viruses. Arch. Virol. 109:83-102.
Holmes, G. P., J. K. Hilliard, K. C. Klontz, A. H. Rupert, C. M. Schindler, E. Parrish, D. G. Griffin, G. S. Ward, N. D. Bernstein, T. W. Bean, et al. 1990. B virus (herpesvirus simiae) infection in humans: epidemiologic investigation of a cluster. Ann. Intern. Med. 112:833-839.
Huff, J. L., and P. A. Barry. 2003. B-virus (cercopithecine herpesvirus 1) infection in humans and macaques: potential for zoonotic disease. Emerg. Infect. Dis. 9:246-250.
Ikoma, M., J. A. Liljeqvist, J. Groen, K. L. Glazenburg, T. H. The, and S. Welling-Wester. 2002. Use of a fragment of glycoprotein G-2 produced in the baculovirus expression system for detecting herpes simplex virus type 2-specific antibodies. J. Clin. Microbiol. 40:2526-2532.
Kalamvoki, M., V. Miriagou, A. Hadziyannis, U. Georgopoulou, A. Varaklioti, S. Hadziyannis, and P. Mavromara. 2002. Expression of immunoreactive forms of the hepatitis C NS5A protein in E. coli and their use for diagnostic assays. Arch. Virol. 147:1733-1745.
Katz, D., J. K. Hilliard, R. Eberle, and S. L. Lipper. 1986. ELISA for detection of group-common and virus-specific antibodies in human and simian sera induced by herpes simplex and related simian viruses. J. Virol. Methods 14:99-109.
Katz, D., W. Shi, P. W. Krug, R. Henkel, H. McClure, and J. K. Hilliard. 2002. Antibody cross-reactivity of alphaherpesviruses as mirrored in naturally infected primates. Arch. Virol. 147:929-941.
Kuhn, J. E., G. Dunkler, K. Munk, and R. W. Braun. 1987. Analysis of the IgM and IgG antibody response against herpes simplex virus type 1 (HSV-1) structural and nonstructural proteins. J. Med. Virol. 23:135-150.
Kuhn, J. E., K. Klaffke, K. Munk, and R. W. Braun. 1990. HSV-1 gB and VZV gp-II crossreactive antibodies in human sera. Arch. Virol. 112:203-213.
Lamiel, J. M., J. A. Ward, and J. K. Hilliard. 1993. Detection of type-specific herpesvirus antibodies by neural network classification of Western blot densitometer scans, p. 1731-1736. In Proceedings of the IEEE International Conference on Neural Networks. Institute of Electrical and Electronics Engineers, Inc., Piscataway, N.J.
Levi, M., U. Ruden, and B. Wahren. 1996. Peptide sequences of glycoprotein G-2 discriminate between herpes simplex virus type 2 (HSV-2) and HSV-1 antibodies. Clin. Diagn. Lab. Immunol. 3:265-269.
Liljeqvist, J. A., E. Trybala, J. Hoebeke, B. Svennerholm, and T. Bergstrom. 2002. Monoclonal antibodies and human sera directed to the secreted glycoprotein G of herpes simplex virus type 2 recognize type-specific antigenic determinants. J. Gen. Virol. 83:157-165.
Liljeqvist, J. A., B. Svennerholm, and T. Bergstrom. 1999. Herpes simplex virus type 2 glycoprotein G-negative clinical isolates are generated by single frameshift mutations. J. Virol. 73:9796-9802.
Liljeqvist, J. A., E. Trybala, B. Svennerholm, S. Jeansson, E. Sjogren-Jansson, and T. Bergstrom. 1998. Localization of type-specific epitopes of herpes simplex virus type 2 glycoprotein G recognized by human and mouse antibodies. J. Gen. Virol. 79:1215-1224.
Loomis, M. R., T. O'Neill, M. Bush, and R. J. Montali. 1981. Fatal herpesvirus infection in patas monkeys and a black and white colobus monkey. J. Am. Vet. Med. Assoc. 179:1236-1239.
Lopez, C., A. Arvin, and R. Ashley. 1993. Immunity to herpesvirus infections in humans, p. 397-245. In B. Roizman, R. J. Whitley, and C. Lopez (ed.), The human herpesviruses. Raven Press, Ltd., New York, N.Y.
Norcott, J. P., and D. W. Brown. 1993. Competitive radioimmunoassay to detect antibodies to herpes B virus and SA8 virus. J. Clin. Microbiol. 31:931-935.
Ohsawa, K., D. H. Black, H. Sato, and R. Eberle. 2002. Sequence and genetic arrangement of the U(S) region of the monkey B virus (cercopithecine herpesvirus 1) genome and comparison with the U(S) regions of other primate herpesviruses. J. Virol. 76:1516-1520.
Perelygina, L., I. Patrusheva, H. Zurkuhlen, and J. K. Hilliard. 2002. Characterization of B virus glycoprotein antibodies induced by DNA immunization. Arch. Virol. 147:2057-2073.
Perelygina, L., L. Zhu, H. Zurkuhlen, R. Mills, M. Borodovsky, and J. K. Hilliard. 2003. Complete sequence and comparative analysis of the genome of herpes B virus (cercopithecine herpesvirus 1) from a rhesus monkey. J. Virol. 77:6167-6177.
Perelygina, L., H. Zurkuhlen, I. Patrusheva, and J. K. Hilliard. 2002. Identification of a herpes B virus-specific glycoprotein D immunodominant epitope recognized by natural and foreign hosts. J. Infect. Dis. 186:453-461.
Prince, H. E., C. E. Ernst, and W. R. Hogrefe. 2000. Evaluation of an enzyme immunoassay system for measuring herpes simplex virus (HSV) type 1-specific and HSV type 2-specific IgG antibodies. J. Clin. Lab. Anal. 14:13-16.
Saijo, M., M. Niikura, S. Morikawa, T. G. Ksiazek, R. F. Meyer, C. J. Peters, and I. Kurane. 2001. Enzyme-linked immunosorbent assays for detection of antibodies to Ebola and Marburg viruses using recombinant nucleoproteins. J. Clin. Microbiol. 39:1-7.
Sehgal, D., P. S. Malik, and S. Jameel. 2003. Purification and diagnostic utility of a recombinant hepatitis E virus capsid protein expressed in insect larvae. Protein Expr. Purif. 27:27-34.
Sisk, W. P., J. D. Bradley, R. J. Leipold, A. M. Stoltzfus, M. Ponce de Leon, M. Hilf, C. Peng, G. H. Cohen, and R. J. Eisenberg. 1994. High-level expression and purification of secreted forms of herpes simplex virus type 1 glycoprotein gD synthesized by baculovirus-infected insect cells. J. Virol. 68:766-775.
Slomka, M. J., L. Harrington, C. Arnold, J. P. Norcott, and D. W. Brown. 1995. Complete nucleotide sequence of the herpesvirus simiae glycoprotein G gene and its expression as an immunogenic fusion protein in bacteria. J. Gen. Virol. 76:2161-2168.
Stacey, S. N., J. S. Bartholomew, A. Ghosh, P. L. Stern, M. Mackett, and J. R. Arrand. 1992. Expression of human papillomavirus type 16 E6 protein by recombinant baculovirus and use for detection of anti-E6 antibodies in human sera. J. Gen. Virol. 73:2337-2345.
Su, H. K., J. D. Fetherston, M. E. Smith, and R. J. Courtney. 1993. Orientation of the cleavage site of the herpes simplex virus glycoprotein G-2. J. Virol. 67:2954-2959.
Tanabayashi, K., R. Mukai, and A. Yamada. 2001. Detection of B virus antibody in monkey sera using glycoprotein D expressed in mammalian cells. J. Clin. Microbiol. 39:3025-3030.
Thompson, S. A., J. K. Hilliard, D. Kittel, S. Lipper, W. E. Giddens, Jr., D. H. Black, and R. Eberle. 2000. Retrospective analysis of an outbreak of B virus infection in a colony of DeBrazza's monkeys (Cercopithecus neglectus). Comp. Med. 50:649-657.
Vizoso, A. D. 1975. Recovery of herpes simiae (B virus) from both primary and latent infections in rhesus monkeys. Br. J. Exp. Pathol. 56:485-488.
Ward, J. A., and J. K. Hilliard. 1994. B virus-specific pathogen-free (SPF) breeding colonies of macaques: issues, surveillance, and results in 1992. Lab. Anim. Sci. 44:222-228.
Ward, J. A., and J. K. Hilliard. 2002. Herpes B virus-specific pathogen-free breeding colonies of macaques: serologic test results and the B-virus status of the macaque. Contemp. Top. Lab. Anim. Sci. 41:36-41.
Whitley, R. J., and J. K. Hilliard. 2001. Cercopithecine herpesvirus (B virus), p. 2835-2848. In D. M. Knipe et al. (ed.), Fields virology. Lippincott Williams & Wilkins, Philadelphia, Pa.
Zhang, J. X., I. Braakman, K. E. Matlack, and A. Helenius. 1997. Quality control in the secretory pathway: the role of calreticulin, calnexin and BiP in the retention of glycoproteins with C-terminal truncations. Mol. Biol. Cell 8:1943-1954.
Zwartouw, H. T., and E. A. Boulter. 1984. Excretion of B virus in monkeys and evidence of genital infection. Lab. Anim. 18:65-70.(Ludmila Perelygina, Irina)