USEFUL IN DIAGNOSTIC METHOD

Background of the Invention

Technical Field

This invention relates to diagnosis and treatment of Epstein-Barr virus (EBV) infection.

Background Art Epstein-Barr virus is a member of the herpesvirus family, which infects most humans. Most persons are infected with the virus in early childhood and then carry the virus for life. The majority of primary infections in young children are asymptomatic. In adolescents and young adults, infectious mononucleosis (IM) is the most common manifestation of primary infection with EBV (G.H. Roberts, Diagnostics and Clinical Testing 27;16-21. 1989) . EBV has also been implicated in rheumatoid arthritis, chronic fatigue syndrome, Burkitt's lymphoma, nasopharyngeal carcinoma, aplastic anemia, aggressive lymphoproliferative disorders, and rupture of the spleen. EBV is known to be an opportunistic pathogen, often developing in immunosuppressed subjects, such as for example patients with allografts, or persons having acquired immune deficiency syndrome ("AIDS") .

Diagnosis of IM is often based on the presentation of characteristic symptoms including primarily fever, pharyngitis, and cervical lymphadenopathy. However, IM may be complicated by splenomegaly, hepatitis, and myalgia, as well as headache and further nonspecific symptoms (S.H. Cheeseman, Semin. Hematol. 25:261-268, 1988) . Moreover, other pathogens, mainly cytomegalovirus but also adenovirus, rubella virus, human

immunodeficiency virus (HIV), and Toxoplasma gondii , can cause a similar syndrome. At present, testing for heterophil antibodies is frequently used to relate IM to EBV infection. Such test results are not necessarily conclusive for EBV, because heterophil antibodies can be induced by viruses other than EBV and because the incidence of heterophil antibodies among patients with IM is, at most, 90% (A.S. Evans, et al., J.Infect.Pis. 132:546-553, 1975). In particular, young children often lack a heterophil antibody response (C.V. Sumaya and Y. Ench, Pediatrics 75:1011-1019. 1985).

Serological diagnostic methods have progressed from testing for the presence of non-specific heterophil antibodies to EBV specific serodiagnosis ( . Henle, et al., Hum. Pathol. 5:551-565. 1974). To date, this is performed predominantly by indirect immunofluorescence (IF) or immunoperoxidase (IP) staining of EBV-infected cells. Standard assays make use of fixed cells derived from selected EBV-infected cell cultures which specifically express EBV-associated antigens such as the viral capsid antigens (VCAs) , the set of early antigens (EAs, which include EA diffuse and EA restricted) , and EBV nuclear antigens (EBNAs) . These antigens correspond to the late lytic phase, the early lytic phase, and the phase of latency of EBV infections, respectively. Qualitative and quantitative determinations of immunoglobulin IgM and/or IgG antibodies against individual antigens or combinations of these antigens then allows discrimination among primary, past, and reactivated EBV infections. However, IF assays are time- consuming and are not suitable for large-scale testing. Because of variability of antigen-producing cells as well as subjective reading of results, these tests are also difficult to standardize. Moreover, IgM-IF tests may suffer from non-specific reactions when rheumatoid factors are present in the serum (G. Henle, et al., Clin. Exp. Immunol. 36:415-422, 1979).

Various attempts have been made to facilitate specific serodiagnosis of EBV infections by developing microtiter plate assays based either on native antigens purified from EBV-infected cells (G.R. Pearson, J. Virol. Methods 1:97-104, 1988) or on synthetic peptides selected empirically or by computer analysis from the a ino acid sequences of presumptively relevant EBV- associated antigens. (R.I. Fox, et al. J. Clin. Lab. Anal. 1:140-145, 1987; J.E. Geltofsky, et al. J. Clin. Lab. Anal. 1:153-162, 1987; J.M. Middledorp and P. Herbrink, J. Virol. Methods 21:133-146. 1988).

One of the best indicators of active EBV infection is the presence of antibody to viral capsid antigens, structural proteins necessary for replication of the virus (G.R. Pearson and J. Luka, Characterization of the virus-Determined Antigens. In The Epstein-Barr Virus: Recent Advances. Epstein and Achong, Eds. Wiley: New York, pp. 47-73, 1986). Viral capsid antigens are present in every cell infected with EBV. VCA-IgG is the first antibody to appear at the onset of the primary EBV- infection, and it peaks within four to six weeks. VCA-IgG gradually declines during convalescence but remains detectable for life (G.H. Roberts, Diagnostics and Clinical Testing 22:16-21, 1989). Almost all patients are positive for VCA-IgM by the third week following onset and can be expected to become VCA-IgM negative 4 to 8 weeks later. Detection of VCA-IgM antibodies remains the most sensitive procedure for mononucleosis because an absence of VCA-IgM from serum in which VCA-IgG is detected excludes a diagnosis of primary EBV infection (ibid.).

Two major polypeptides associated with the VCA complex have been identified. These are a glycosylated protein of 125,000 daltons size, designated gpl25, and a non-glycosylated protein of 160,000 daltons size, designated pl60. Studies have shown that gpl25 is the most dominant antigen in the VCA complex and therefore the strongest immunogen (G.R. Pearson et al . (1986),

supra) . gpl25 has already been found to be most useful for diagnosing a recent or past EBV infection. 100% of sera tested from individuals having confirmed infectious mononucleosis have been positive for IgM antibodies to this glycoprotein. gpl25 has also served as a reliable substrate for measuring IgA antibodies to VCA (G.R. Pearson et al . (1986), supra) , which have been found to be a useful marker for nasopharyngeal carcinoma (G.H. Roberts (1989) , supra) . Conventionally, gpl25 has been isolated and purified from EBV-infected cells. The cell culture system is very expensive and time consuming, however, and yields only small quantities of the protein. Further, suitable culture systems for bulk production of gpl25 are lacking, and use of gpl25 for routine diagnosis has therefore been unsupported.

M. Gong et al . (1987), Jour. Virol. 61:499-508. describes expression of an EBV open reading frame BALF4 in E. coll . Recently, four diagnostically useful recombinant EBV proteins, representative of the viral capsid antigen (pl50) , the diffuse early antigen (p54) , the major DNA- binding protein (pl38) , and the EBV nuclear antigen (p72) , were used to set up individual enzyme-linked immunosorbent assays (ELISAs) for detection of immunoglobulin IgM and IgG antibodies (M. Gorgievski- Hrisoho, et al. J. Clin. Microbiol. 28:2305-2311 _f 1990). In direct comparison with results obtained by standard IF or IP assays, the recombinant EBV ELISAs were shown to provide a means for specific and sensitive serodiagnosis of infectious mononucleosis (IM) .

Summary of the Invention

Disclosure of Invention

In one aspect, in general, the invention features pure recombinant EBV-gpl25 protein having sufficient selectivity and sensitivity to be useful in an

immunoassay for EBV VCA antibodies, and methods for producing recombinant gpl25 protein in sufficient amount and purity to enable its use in assays such as enzyme- linked immunosorbent assays (ELISA) for the qualitative and quantitative detection of EBV VCA-IgG and VCA-IgM antibodies, which would facilitate the diagnosis of primary and past EBV infection. A recombinant gpl25 protein according to the invention "has sufficient selectivity and sensitivity to be useful in an immunoassay for EBV VCA antibodies" if, when used in an immunoassay of serum for EBV VCA antibodies, the results indicate substantially the same pattern of seropositives and seronegatives as are indicated in a similar immunoassay using an EBV gpl25 viral capsid antigen conventionally isolated and purified from EBV-transformed cells.

The terms "recombinant gpl25 protein" and, alternatively, "recombinant EBV-gpl25 protein", as used herein, mean and include the full-length recombinant gpl25 polypeptide, whether or not partially or fully glycosylated; and mean and include any recombinant polypeptide fragment of the gpl25 polypeptide, whether or not partially or fully glycosylated, that is immunoreactive with VCA antibodies with which the EBV gpl25 viral capsid antigen is immunoreactive. Thus, a recombinant gpl25 protein according to the invention can be a polypeptide, glycosylated or not, having a length less than that of the full-length gpl25 viral capsid antigen; and, for example, I have shown that a purified recombinant gpl25 protein according to the invention, made by expression in a Jaculovirus system of a DNA sequence corresponding to that portion of the EBV DNA sequence between a start codon at 159,322 and a stop codon at base 156,755, can be as effective as EBV gpl25 protein isolated and purified from EBV-infected cells in an ELISA for VCA antibodies.

Preferred recombinant gpl25 protein is made by expression of the EBV viral capsid protein-encoding DNA

sequence in a baculovirus expression system, as the resulting expression product can have substantially better specificity and sensitivity than product obtained by expression of the same EBV viral capsid protein- encoding DNA in, for example, E. coli or yeast.

Moreover, the recombinant gpl25 protein according to the invention can be expressed in high yield in a baculovirus system.

Recombinant gpl25 protein according to the invention can be obtained in greater quantities at lower cost, and can be purified to higher degrees of purity, than obtainable for gpl25 obtained by conventional purification from EBV-infected cell culture.

Preferably, the method for making recombinant EBV- gpl25 protein in high amounts and purity includes steps of cloning and expressing the EBV-gpl25 protein in a baculovirus expression system, and isolating and purifying the resulting protein by metal affinity chromatography. In another general aspect, the invention features recombinant gpl25 protein made by such a method.

In another general aspect, the invention features a serologic method for diagnosing present or past primary EBV infection in a subject by using recombinant gpl25 protein in an immunoassay to detect gpl25 protein-binding antibody in serum from the subject; and apparatus for carrying out the method. The apparatus preferably includes the recombinant gpl25 protein immobilized on a solid support and blocked for nonspecific antibody; and is provided with means for contacting the blocked immobilized recombinant gpl25 protein with the sample (such as serum) to be tested; and is provided with a secondary antibody conjugated with IgG and/or IgM having a detectable label; and is provided with means for detecting the label.

Description of Preferred Embodiments

Best Mode of Carrying out the Invention Purified EBV-gpl25 Protein. Purified recombinant gpl25 protein can be made in high quantity according to the invention generally by inserting a DNA fragment containing a gpl25-encoding sequence into an expression vector that is capable of expression in a suitable host, transforming host cells with the vector, and culturing the cells. Recombinant gpl25 can then be purified from lysates of the cultured cells. A preferred protocol follows.

Expression of EBV-gpl25 Protein in insect cells. The complete (172,282 base pairs) nucleotide sequence of Epstein-Barr virus has been established (R. Baer, et al. Nature 310:207-211. 1984). The gpl25 protein has been mapped to a BamHI A DNA fragment. A plasmid DNA, designated as pTZ19-110, encoding gpl25 was obtained from Dr. Frederick Wang (Harvard Medical School, Boston, MA) . In this recombinant plasmid, a 3.5 kb BamHi-.EcoRI fragment encoding gpl25 was cloned in a plasmid vector pTZ19R (Pharmacia, Piscata Way, NJ) . This EBV sequence represents nucleotides 159,469 to 155,969 on the EBV genome. From the DNA sequence, it was found that the protein start codon ATG (methionine) is at nucleotide

159,322 and the stop codon is at base 156,755. The size of the protein was predicted from the DNA sequence as 95,640 daltons. Because gpl25 is a glycosylated protein, the size of the protein isolated from EBV-infected cells was determined to be 125,000 daltons.

Preferably the gpl25 protein is expressed in a baculovirus system.

For expression of the protein in baculovirus, the plasmid pTZ19-110 was digested with restriction enzyme Nhel , at a site located 33 bp downstream from ATG of gpl25 coding sequence. The coding sequence was then released from the vector by digesting with restriction enzyme .EcoRI. This Nnel-JJcoRI fragment coding gpl25 was

isolated from low melting 0.7% agarose gel using spindBind DNA extraction unit (FMC, Rockland, ME) . In order to obtain the correct reading frame in baculovirus transfer vector pBlueBacHis C (Invitrogen, San Diego, CA) , the DNA fragment was cloned into the Nhel-EcoRI site of E. coli vector pTrcHis C (Invitrogen, CA) . Then the Ncol-Hindlll coding gpl25 was subcloned at the corresponding site of pBlueBacHis C. Since this vector is derived from E. coli pUC vector, all the manipulations were done in E. coli Top 10 strain (Invitrogen,, CA) .

The pBlueBacHis C vector is designed for efficient protein expression and purification from recombinant baculovirus clones in insect cells. High levels of expression of DNA sequences cloned into the pBlueBacHis C vector is made possible by the presence of the polyhedrin promoter. The polyhedrin protein is the major structural component of the baculovirus occlusion bodies and accounts for more than 50% of the total "stainable" protein of infected Spodoptera frugiperda on SDS- polyacrylamide gels. This vector contains the natural polyhedrin leader sequence followed by a sequence which codes for (in 5' to 3' direction from N-terminal to C- ter inal) and ATG translation initiation codon, a tract of six histidine residues that function as a metal binding domain in the translated protein, a transcript stabilizing sequence from gene 10 of phage T7, and an enterokinase cleavage recognition sequence. A multiple cloning region positioned downstream of this sequence allows insertion of the foreign gene in the correct reading frame relative to the initiation codon. This vector also allows co-expression of 3-galactosidase upon transfection enabling rapid identification of the recombinant plaques in the presence of the substrates X- gal or Blue-gal. Sequences which flank the polyhedrin gene in the wild-type baculovirus genome are positioned 5' to 3' to the expression cassettes on the pBlueBacHis C transfer vector. Following co-transfection with the wild-type

viral DNA, homologous recombination between these sequences results in a recombinant virus with the gene of interest expressed under the control of the viral polyhedrin enhancer/promoter elements. The BaculoGold linearized baculovirus DNA (Pharmingen, San Diego, CA) was used for co-transfection of pBlueBacHis C. This DNA contains a lethal deletion, and thus the transfected virus DNA cannot make infectious virus particles in insect cells unless the deletion is complemented by co- transfected polyhedrin-based pBlueBacHis C. Transfection of insect cell Sf9 and production and purification of recombinant virus was done according to the method described in the Invitrogen manual.

Expression of a gpl25-encoding DNA fragment in pBluBacHis C was confirmed by western blot analysis.

About l ml insect cells infected three days earlier with the recombinant virus, and 1 ml of a EBV-transformed human B cell line (Raji cell line) were pelleted and dissolved in 100 μl of Laemmli buffer. The sample was boiled for 2 minutes and then loaded on a 7.5% SDS- polyacrylamide gel and electrophoresed overnight at 70 volts. The protein was transferred electrophoretically to a nitrocellulose membrane. Nonspecific binding of the protein was blocked by treating the membrane with 5% nonfat dry milk in TBST (10 mM Tris, pH 8.0), l mM EDTA, 0.05 % Tween-20) . Monoclonal antibody to gpl25 was obtained from DuPont Company (Cat # NEA 9243) and added to the membrane-bound protein and incubated for 1 hour, washed 3 times with TBST, and then incubated with anti- mouse IgG-alkaline phosphatase conjugate for 30 minutes. The membrane was washed again 3 times with TBST, and the color was developed with NBT (nitro blue tetrazolium) and BCIP (5-bromo-4-chloro-3-indolyl phosphate) . A single protein band that reacted with the monoclonal antibody in the extract of recombinant virus transfected cell line Sf9 migrate close to that of the monoclonal antibody- reacting band of the EBV-gpl25 viral capsid protein from the EBV-transformed Raji cell line. This result confirms

expression of gpl25 protein using the baculovirus system in the insect cells.

Purification of Recombinant gpl25 Protein.

Recombinant gpl25, expressed in transfected insect culture Sf9 for example as described above, can be purified from the culture by using affinity column chromatography methods, for example.

The vector pBluBacHis C contains DNA sequence below the translation start codon which produces 6 histidine peptide at the N-terminal region of the expressed protein. This region has the affinity to bind to Ni ² charged ProBond resin (Invitrogen, CA) and thus allows one-step purification. About 50 ml of insect cell Sf9 was grown at a density of 2 x 10° cells/ml in a 100 ml spinner flask. Cells were infected at the multiplicity of infection of 6 and grown at 27 °C for 3 days. The cells were pelleted by centrifugation and suspended in 20 mM sodium phosphate and 500 mM NaCl and lysed by sonification and centrifuged to remove cell debris. The recombinant protein was purified from the supernatant through ProBond resin by the method described by Invitrogen. The N-terminal fusion peptide was removed from gpl25 protein by enterokinase cleavage and subsequent purification through the ProBond resin. This method gave about 1 mg of protein from 50 ml of culture, and the purity of the protein was confirmed on 7.5 % SDS polyacrylamide gel.

Industrial Applicability ELISA Using Recombinant gp!25 Protein.

The following ELISA demonstrates by way of example the use of recombinant gpl25 protein according to the invention in an assay for EBV serological parameters in sera from patients having confirmed EBV infection. About 100 μl of a recombinant gpl25 protein solution (1 μg/ml of the purified recombinant gpl25 protein in PBS buffer, pH 7.5) made as described above was added to Maxiabsorb Nunc microwell strips and incubated overnight

at 4° C. Each plate was washed 2 times with 200 μl of TBST buffer (10 mM Tris pH 8.0, 1 mM EDTA, 150 mM NaCl and 0.05% Tween-20) at room temperature for 5 minutes. Each plate was blocked overnight at 4 °C with 300 μl of 0.1% nonfat dry milk in TBST buffer. A 100 μl aliquot of diluted test serum (1:21 dilution in TBST buffer) was added and incubated for 1 hour at 37 °C, and then the wells were washed 3 times for 5 minutes at 37 °C with TBST. A 100 μl aliquot of alkaline phosphatase- conjugated anti-human IgM and/or IgG was added and incubated for 30 minutes at 37° C. The wells were washed 3 times with TBST for 5 minutes at 37 °C, and then 100 μl of alkaline phosphatase substrate (prepared by adding 5 mg p-nitrophenylphosphate and 1 ml of 5x diethanolamine buffer (Kierkegaard and Perry, Gaithersburg, MD) to 4 ml of distilled water) was added and incubated at 37 °C for 30 minutes. Optical density was measured at 405 n .

The results are shown in the Table. The Table shows a comparison of an ELISA method ["TTI"], using recombinant gpl25 protein according to the invention, with the commercially available Gull's Immunofluorescence and ELISA kit, using gpl25 isolated and purified from EBV-infected cells. The results for the Gull's kit and for the method of the invention are similar. Owing to the small quantity of protein obtainable by isolation from EBV-infected cells, and owing to the expensive cell culture and isolation method used in so obtaining the material, the Gull's kit is very expensive. The method of the invention provides for producing large quantities of protein relatively inexpensively, providing for a kit that can be produced in quantity and very cost effectively. Conventional methods, using for example immunofluorescence, are more time-consuming and labor- intensive, and yield less reliable determinations, owing to substantial cross-reactivity.

Table

Comparison of EBV serological parameters determined by standard immunofluorescence techniques (IF) and by Gull and TTI recombinant ELISAs in sera from patients with confirmed recent or acute EBV infection.

Patient # IF ELISA

Gull TTI

VCA-IgG VCA-IgM

IgG IgM gG IgM

IM

1 + + 1.432 0.631 1.550 0.792 2 + + 1.341 0.827 1.466 0.927

3 + + 1.263 0.498 1.629 0.619

4 + + 1.172 1.020 1.429 1.495

5 + + 1.265 0.396 1.779 0.500

6 + + 1.486 0.440 1.783 0.540 7 + + 1.192 0.344 1.232 0.430

8 + + 1.622 1.058 1.841 1.120

9 + + 1.634 0.766 1.764 0.786 N

1 0.068 0.075 0.112 0.115 2 + - 1.885 0.047 1.958 0.049

3 + - 1.351 0.110 1.526 0.090

4 - - 0.083 0.052 0.076 0.036

5 + - 1.455 0.048 1.581 0.052

6 - - 0.094 0.039 0.086 0.028 7 + - 1.646 0.051 1.773 0.041

8 + - 1.383 0.072 1.456 0.037

9 + - 1.788 0.061 1.858 0.056 10 - - 0.096 0.045 0.082 0.078

Other Embodiments

Other embodiments are within the following claims. For example, the recombinant gpl25 protein can be made using an expression system other than the baculovirus system described above. Expression in a baculovirus system is much preferred, according to the invention, over expression in, for example, E. coli or yeast or mammalian cells, because the baculovirus system provides for production of high-purity recombinant gpl25 protein in higher yield. Moreover, the configuration (folding, glycosylation) of the recombinant gpl25 protein made in a baculovirus system is sufficiently like that of the EBV gpl25 viral capsid antigen from EBV-transformed human cells to provide for a high degree of specificity and selectivity in an immunoassay system. Recombinant gpl25 protein expressed in E. coli or in yeast apparently lacks some conformational characteristics that are required for high specificity and selectivity. Moreover, yields of recombinant gpl25 protein are lower in yeast expression systems. Mammalian expression systems are difficult and costly to maintain, and purification of recombinant gpl25 expressed in mammalian cells is costly and time- consuming. I have found that recombinant gpl25 expressed in

E . coli , for example, is unsatisfactory in an immunoassay for EBV VCA antigens, as it is not sufficiently selective or sensitive.

For expression of the protein in E. coli , plasmid pTZ19-110 was modified as follows. A sequence of about 183 bp was deleted by digesting plasmid pTZ19-110 with restriction enzyme Nhel , at a site located 33 bp downstream from ATG, and BaiαRI , at a site located 150 bp upstream from ATG. Both sites were blunt-ended with T DNA polymerase and ligated, and the resulting circular plasmid was transformed into E. coli NM 522. This manipulation resulted in formation of a new BamHI site. The plasmid was then double digested with BamHI and EcoRI

and the fragment encoding gpl25 was cloned into the same translational reading frame of the expression vector pTrHisC (obtained from Invitrogen, San Diego, CA) . This vector is designed for high level expression of recombinant protein in E. coli under the control of the inducible tac (trp-lac) promoter. This vector also contains the translation start codon, the polyhistidine metal binding domain, and enterokinase cleavage sites. Expression of a gpl25-encoding DNA fragment in pTrcHisC was confirmed by Western blot analysis. In this procedure, transformants were grown in LB to ODβooO.δ at 30 °C, and then IPTG was added to final concentration of 1 mM to induce expression of gpl25. After induction the cells were grown for 3 hours at 30 °C. About 200 μl of the culture was spun down and resuspended in SDS buffer. The sample was boiled for 2 minutes and then loaded on a 7.5% SDS-polyacrylamide gel and electrophoresed overnight at 70 volts. The proteins were transferred electrophoretically to a nitrocellulose membrane. Nonspecific binding of the protein was blocked by treating the membrane with 5 % nonfat dry milk in TBST (10 mM Tris pH 8.0, 1 mM EDTA, 0.05 % Tween-20) . Sera containing VCA-IgM, previously confirmed by IF test, were added to the membrane-bound protein and incubated for 1 hour, washed 3 times with TBST, and then incubated with anti-human IgM-alkaline phosphatase conjugate for 30 minutes. The sera were washed again 3 times with TBST, and the color was developed with NBT (nitro blue tetrazolium) and 5-bromo-4-chloro-3-indolyl phosphate. A single protein band having a molecular weight of approximately 95 kd reacted with the VCA-IgM sera, suggesting that gpl25 is produced as a 95 kd protein in E. coli , but the size of the gpl25 molecule isolated from EBV-infected cells was 125 kd. This size difference is most likely due to a lack of protein glycosylation in E. coli .

Recombinant 95 kd gpl25 protein, expressed in culture in E. coli for example as described above, can be

purified from the culture for example by using affinity column chromatography methods. By way of example, recombinant 95 kd gpl25 was purified from E. coli transformants made as described above, according to the following protocol.

About 1.2 liter of LB broth + glucose and ampicillin was inoculated with 12 ml of an overnight culture of cells containing the fusion plasmid. The cells were grown for 3 hours at 30 °C, and then IPTG to 1 mM final concentration was added to induce expression. The cells were incubated for a further 3 hours at 30 °C, and then were harvested and suspended in lysis buffer (10 mM sodium phosphate pH 7.0, 30 mM NAC1, 0.25 % Tween-20, 10 mM /..-mercaptoethanol and 1 mM PMSF (phenylmethylsulfonyl fluoride) . The cells were lysed by sonication and centrifuged.

The recombinant gpl25 protein expressed in E. coli was found to be insoluble, and was present in the pellet obtained from the crude fraction obtained as described above. The pellet was described in a denaturing buffer containing 8 M urea, 20 mM sodium phosphate pH 7.8, and 0.5 M NaCl. This was then passed through a ProBond metal affinity chromatography resin column (Invitrogen, San Diego, CA) . The column was washed and the protein was eluted using procedures set out in the Invitrogen manual.

As noted above, the recombinant gpl25 protein expressed in E. coli is unacceptable for use in an immunoassay for EBV VCA antibody because it is not sufficiently selective and sensitive to yield consistent and reliable results. Its poor performance in such an immunoassay may owe in part to the fact that it is denatured during processing, so that it no longer has a configuration presenting the appropriate antigen binding sites.

There are apparently post-translation modifications that result in differences between the recombinant gpl25 protein expressed in E. coli and that expressed in the

baculovirus system. Recombinant gpl25 protein expressed in E. coli as described above has a molecular size about 94 kd, corresponding to the molecular size that would be predicted for a non-glycosylated polypeptide product of expression of the DNA sequence. Recombinant gpl25 protein made according to the invention in a baculovirus system has, as noted above, much better specificity and sensitivity than that made in E. coli ; this may be owing to the substantial glycosylation in the Jaculovirus- expressed protein (which has a molecular size comparable to that of the EBV gpl25 viral capsid antigen as isolated from EBV infected cells) . There may also be differences in folding between the E. coli-expressed gpl25 protein and the Jbaσulovirus-expressed gpl25 protein, that may affect the conformation of antibody-binding sites and compromise the specificity or the sensitivity of antibody-binding at the sites.

As I have demonstrated, recombinant gpl25 protein resulting from expression according to the invention of fractions less than the whole length of the EBV gpl25 viral capsid antigen-encoding DNA can display antibody- binding specificities and sensitivities comparable to those of EBV gpl25 viral capsid antigen isolated from EBV-infected cells. Fragments smaller than that described in the examples are within the claims, and can be found as a matter of routine, for example by the following protocol. First, the amino acid sequence can be examined for hydrophilic (and hydrophobic) regions that are more (or less) likely to appear on the surface of the immunologically active molecule, and therefore are more (or less) likely to make up part of an antibody- binding domain. Then sequences that appear less likely to make up part of a binding domain can be excised using standard techniques, and the resulting candidate fragment or fragments can be expressed, purified, and isolated and then used in an assay generally as described above to determine their VCA-Ig-binding specificities and sensitivities.

Alternatively, candidate polypeptide fragments can be constructed synthetically according to well- established and automated techniques, and then used in an assay generally as described above to determine their VCA-Ig-binding specificities and sensitivities. As will be appreciated, however, to the extent that post- translational modification of the recombinant gpl25 protein (or of any fragment of it) can have an effect on the specificity and selectivity of the resulting molecule, as suggested above, synthetically-generated polypeptide fragments may not be as effective in an immunoassay for EBV VCA antibodies as the recombinant gpl25 protein (or recombinant fragments of it) as expressed in, for example, the baculovirus system, as detailed above.