241
Journal of Virological Methods, 39 (1992) 241-258 0 1992 Elsevier Science Publishers B.V. / All rights reserved / 0166-0934/92/%05.00 VIRMET 01383
Recombinant parvovirus B19 capsids as a new substrate for detection of B 19-specific IgG and IgM antibodies by an enzyme-linked immunosorbent assay Marcel
M.M. Salimans,
Mario J.A.W.M. van Bussel, Caroline and Willy J.M. Spaan
S. Brown
Department of Virology, University Hospital Leia’en and Faculty of Medicine. Leiden (The Netherlands) (Accepted
16 April 1992)
Summary A new enzyme-linked immunosorbent assay for the detection of B19-specific IgG and IgM antibodies was established using B19 capsids synthesized in a baculovirus expression system. These 1319 capsids, consisting of either coat protein VP2 alone or of both VP1 and VP2, have been shown to be similar to native virus in size and appearance. The results obtained for the detection of B 19-specific antibodies showed good correlations with a radioimmunoassay which uses native B19 virus and an immunofluorescence assay based on insect cells expressing coat protein VPl. The course of the antibody response could be followed by determining the titers of sequential serum samples taken after a recent B19 infection. Both types of recombinant capsids form an excellent source of antigen for the detection of both B19 IgG and IgM antibodies and are a very promising substitute for native virus. Human parvovirus B19; Recombinant capture IgM ELISA
viral antigen; Diagnosis; ELISA; Antibody
Correspondence to: M.M. Salimans, Department of Virology, University Hospital L&den, P.O. Box 320, 2300 AH L&den, The Netherlands.
248
Introduction Human parvovirus B19, since its discovery in 1975 (Cossart et al., 1975), has been associated with several defined clinical syndromes. The virus is the causal agent of erythema infectiosum, the fifth disease of childhood (Anderson et al., 1984) and some cases of acute arthritis and arthralgia, especially in adult women (Reid et al., 1985; White et al., 1985). A B19 infection in patients with chronic hemolytic anemia can lead to an aplastic crisis (Serjeant et al., 1981) and the virus has been associated with serious complications during pregnancy such as abortion, intrauterine fetal death (Hall, 1985; Mortimer et al., 1985), hydrops fetalis (Anand et al., 1987; Maeda et al., 1988; CDC, 1989) and with chronic anemia in the immunodeficient patient (Kurtzmann et al., 1988). The development of diagnostic tests for the detection of B19-specific IgG and IgM antibodies (Anderson et al., 1982; Anderson et al., 1986; Schwartz et al., 1988) has been hindered by the limited amount of viral antigen available. The virus has not yet been propagated in a stable cell-line, although low-level propagation of virus in bone marrow explants (Ozawa et al., 1986) and fetal liver cells (Yaegashi et al., 1989; Brown et al., 1991) has been achieved. Isolation of the virus from serum relies on the random screening of blood donors since symptoms occur when a patient is no longer in the viraemic phase. To overcome this limited supply of antigen, attention has been focussed on recombinant DNA techniques for the production of antigen. Recent publications have described the successful expression of B19 proteins in several systems: Escherichia coli (Sisk et al., 1987; Morinet et al., 1989), Chinese Hamster Ovary (CHO) cells (Kajigaya et al., 1989) and insect cells, using a baculovirus expression system (Brown et al., 1990a; C.S. Brown et al., 1991; Kajigaya et al., 1991). The first attempts to develop diagnostic tests based on these renewable sources of B19 antigens have been described previously (Brown et al., 1990b; Schwartz et al., 1991). The baculovirus expression system was chosen for the production of viral antigens due to the high levels of expression achieved for foreign proteins. Large amounts of the B19 structural proteins have been produced in this system and insect cells expressing VP1 show a strong and characteristic signal in the immunofluorescence assay when reacted with BlPspecific IgG and IgM antibodies (Brown et al., 1990b). More recently, it has been shown for a number of viruses, such as poliovirus (Urakawa et al., 1989), bluetongue virus (French et al., 1990) and rotavirus (Labbe et al., 1990), that virus-like particles can be synthesized from baculovirus recombinants in insect cells. These particles mimic the native virus and are non-infectious. We and others have lately shown that when VP2 alone or both VP1 and VP2 are simultaneously expressed in the baculovirus system, empty B19 capsids are formed (Brown et al., 1991; Kajigaya et al., 1991). Due to their similarity to native B19 and their antigenicity, these capsids promised to be an excellent candidate for the development of a more convenient serological test. In this paper we describe the use of empty B19 capsids for the detection of
249
BlPspecific IgG antibody and the labeling and use of these capsids for specific IgM antibody detection in an enzyme-linked immunosorbent assay (ELISA).
Materials and Methods Clinical specimens
Groups of sera tested: (i) serum specimens previously tested in the radioimmunoassay (RIA) with native B19 virus at the Public Health Laboratory Service in London, kindly supplied by Dr. B.J. Cohen; (ii) sera from patients with a clinical indication for a parvovirus B19 infection, previously tested in the immunofluorescence assay (IFA) in our clinical diagnostic laboratory; (iii) random blood donor sera from the Blood Bank of the University Hospital Leiden, The Netherlands; (iv) sera positive for rheumatoid factor (RF), as determined by latex agglutination and; (v) sera positive for rubella-specific IgM antibodies. Expression of B19 proteins in insect cells
The cloning of the genes for the B19 coat proteins VP1 and VP2 for expression in insect cells has been described previously in detail (Brown et al., 1990a; Brown et al., 1991). Briefly, the VP1 gene was expressed from the polyhedrin promoter in a recombinant baculovirus produced upon recombination between wildtype baculovirus and the plasmid vector pAcYM 1 (Matsuura et al., 1987), containing the VP1 DNA sequence behind the polyhedrin promoter, in Spodoptera frugiperda monolayers. The same procedure was followed for the expression of VP2 (Brown et al., 1990a). Double recombinant virus, expressing both VP1 and VP2, was produced by recombination between the VP2 recombinant virus, expressing the VP2 gene from the polyhedrin promoter, and the plasmid vector pAcAS3 (Vlak et al., 1988), which contains the VP1 DNA sequence behind the p10 promoter. Purification of B19 protein capsids
The isolation of B19 capsids from insect cells infected with the VP2 recombinant and double recombinant baculovirus was essentially the same as described previously (Brown et al., 1991), except for the final purification step. Instead of a sucrose gradient (Brown et al., 1991), the particles were purified by centrifugation in an isopycnic CsCl-gradient (28%, wt/wt), performed with a Beckmann SW55 TI rotor at 100000 x g for 24 h at 5°C. An opaque protein band was clearly detectable in the gradient, which contained the capsids. The bands were extracted, dialysed overnight against PBS and stored at -80°C. The purity of the capsid preparations was checked by running 1 lug amounts on SDS-PAGE and staining (Fast Green or Coomassie brilliant blue) the gel: 90-
95% purity was obtained. Labeling of the B19 capsids For the IgM-ELISA the B19 capsids were biotinylated as follows: the particles were dialysed for 2.5 h at 4°C against 0.1 M NaHC03 (pH 8.7) and diluted to a final concentration of 0.1-1.0 mg/ml. Subsequently, biotinyl-l\rhydroxysuccinimid (BNHS) was added during constant gentle mixing to a final concentration of 0.1 mg per mg protein. BNHS was dissolved, as a stock solution, in dimethylsulfoxide to a concentration of 2.5 mg/ml. After a 3 h incubation at room temperature, the biotinylated particles were dialysed overnight against PBS at 4°C and stored at -80°C. ELISA for Bl9-specific
IgG antibodies
To test for the presence of parvovirus B19-specific IgG antibodies in patient sera, microtiter plates (polysorb F96-3, Nunc) were coated with 25 ng per well of B19 capsids (the optimal amount was determined by dilution experiments) overnight at 4°C in PBS. After washing four times with PBS containing 0.05% Tween-20 (washing-buffer), 100 ,ul of diluted serum was added and incubated for 1 h at 37°C. As a standard procedure, all sera were diluted 1:lOO in dilutionbuffer (2% fetal calf serum, 0.05% Tween-20 and 0.1 mg/ml merthiolate in PBS). The wells were washed four times with washing buffer and 100 ~1 of a peroxidase-labeled rabbit-anti-human IgG (F(ab)2, Dakopatts, Denmark) was added at a 1:4000 dilution (determined by dilution experiments) in dilution buffer and incubated for 1 h at 37°C. After washing four times with washing buffer the plates were incubated with 0-phenylenediamine.2HCl (OPD, Abbott, USA) for 30 min in the dark at room temperature and the reaction was stopped with 4 N HzS04. The optical density (A) was measured at 492 nm. The ‘cut off value for the ELISA was determined by the mean A-value of 50 IFA-negative sera (A = 0.20-0.25) to which 2-times the standard deviation was added. The A-value of the cut off was usually between 0.30-0.35 for both VP2 and VPl-VP2 particles. Titers of the sera tested were determined, using a calibration curve, as previously described by Wielaard et al. (1985). Titer 100 was considered to be the cut-off value. During our ELISA experiments, it became clear that the choice of microtiter plates could influence the results. In our hands, the polysorb-plates (Nunc) gave the best results in the IgG- as well as the IgM-ELISA. The background was low and positive sera gave high values when compared to other plates. ELISA for Bl9-speciJic
IgM antibodies
The presence of B19-specific IgM antibodies was determined by means of an anti-p capture assay. Microtiter plates were coated with a 1:lOOO dilution of rabbit-anti-human IgM (p-chain, Dakopatts, Denmark) in PBS overnight at
251
4°C. After washing (see IgG procedure) sera were added at a dilution of 1:lOO in dilution buffer (see IgG procedure) and incubated for 3 h at 37°C or overnight at 4°C. The plates were washed and 50 ng per well of biotin-labeled B19 capsids were added and incubated for 2 h at 37°C. After washing, bound particles were detected by incubation with horse radish peroxidase-labeled avidine (Dakopatts, Denmark) at a 1:25 000 dilution in dilution buffer for 1 h at 37°C. Finally, the wash-steps and staining reaction was the same as for the IgG-procedure. The cut off and the calculation of titers were essentially the same as described for the IgG-ELISA. The A-value of the cut-off serum was approx. 0.25-0.30 (mean value of 50 negative sera was 0.18-0.23) for both VP2 and VPl-VP2 particles and a titer of 75 was taken as the cut-off value. Immunofluorescence assay (IFA) The immunofluorescence assay was performed on insect cells that had been infected with the recombinant baculovirus expressing VPl. The procedure for the detection of B19-specilic IgG and IgM antibodies was essentially the same as described before (Brown et al., 1990b). Briefly, glass slides, onto which insect cells expressing VP1 had been fixed, were incubated with sera and bound antibody was detected with FITC-conjugated goat-anti-human IgG or FITCconjugated goat-anti-human IgM (Institute Pasteur, France). To avoid false-positive IgM results due to the presence of RF and to prevent competition by IgG antibodies, sera to be tested for B19 IgM antibodies were pretreated with a goat-anti-human-IgG preparation (GULLSORB, Gull Lab., USA).
Results and Discussion Due to the limited supply of native B19 virus, an ELISA for the detection of IgG and IgM antibodies was developed, based on recombinant capsids produced in insect cells. The capsids, formed from VP2 expressed alone or VP1 and VP2 expressed simultaneously in insect cells, could be simply and rapidly purified and high yields of antigen were obtained (Brown et al., 1991). These capsids have been shown to resemble native virus, which is an important factor if conformational epitopes are torbe recognized. That conformational epitopes play an important role in the immune response against B19 was demonstrated by the fact that 13 sera, that reacted with insect cells expressing VP2 in an IFA, did not react with the same recombinant VP2 in a western blot (Brown et al., 1990a). Also, 17 sera that reacted with both types of capsids in the ELISA did not react with VP2 in a western blot (results not shown). This demonstrates that the capsids possess conformational determinants that are recognized by antibodies in patient sera, whereas these antibodies do not recognize nonconformational epitopes in a Western blot assay. The application of recombinant B19 capsids in an EIA has recently been demonstrated by
252
Kajigaya
and coworkers
(1991) who used lysates of insect cells expressing VP2.
IgG-ELBA Thirty serum samples, tested previously by RIA based on native B19 virus (Cohen et al., 1983), were tested for the presence of B19-specific IgG in the ELISA with both types of capsids. A comparison of the results is shown in Fig. 1 demonstrating an optimal correlation between the ELISA titers and the RIA values throughout the whole range. In most cases, high ELISA titers corresponded with high RIA values. In the equivocal range of the RIA (l-3 a.u.) some of the sera were weakly positive in the ELISA while most of them were negative. All negative sera in the RIA were also negative in the ELISA. Titers obtained in the ELISA compare very well for both VP2 and VPl-VP2 particles, although the VPl-VP2 capsids gave somewhat higher titers. The capsids that incorporate VP1 may more closely resemble native B19 and thus react better with the sera. However, since no serum samples have been found so far that only react with VPl-VP2 particles, the contribution of the incorporated VP1 seems to be limited. Both types of capsids have a similar structure to the native virus and gave, on the whole, comparable results. A second comparison was made between the ELISA, based on the VP2 capsids, and the IFA, based on recombinant VP1 protein (Brown et al., 1990b), for 99 serum samples. These sera, from patients with a clinical indication for a B19 infection, were tested in the IFA at dilutions 150, 1:500, 15000 and scored positive or negative. The results (Table 1) show that the same 66 sera were positive and the same 28 negative in both assays, indicating an excellent correlation. Comparable results were obtained with VPl-VP2 particles. When compared to the IFA the ELISA showed a higher sensitivity and specificity in the IgGtest. The sensitivity was 94.3% and the specificity 96.5% when VP2 particles were used and 92.9%, respectively, 100% when VPl-VP2 particles were used. The prevalence of B19-specific IgG antibodies in the general population was analyzed by testing a random selection of blood donor sera with both types of protein particles. 255 sera were tested and the results (Table 2) show that 79% TABLE 1 Comparison ELISA result
Pos. Pos. Neg. Neg.
of IFA and ELISA for detection of Bl9-specific IFA result
Pos. Neg. Pos. Neg.
IgG antibodies
in 99 sera
No. of specimens ELISA VP2 particles
VP1 -VP2 particles
66
65 0
:a 28
‘Weakly positive in the IFA at a 1:50 dilution. b4 of these sera were weakly positive in the IFA at a 1:50 dilution.
b
253 TABLE 2 Prevalence of B19-specific IgG antibodies ELISA titer
in blood donor sera (n = 255)
No. of sera tested in ELISA
< 100 (=neg.) 100-1000 1oOck1oooo > 10000
VP2 particles
VP1 -VP2 particles
1;:
54 106 63 32
15 18
(n = 201) sera reacted positively when VP2 and VPl-VP2 particles were used. The distribution of titer values shows that 46% (for VP2) and 47% (for VPlVP2) of these positive sera had titers > 1000. These results compare very well with previous findings in this age group (Cohen et al., 1988; Brown et al., 1990b). IgM-ELISA
To avoid false-positive reactivity due to the presence of RF a p-capture assay was developed. The use of biotin-labeled B19 capsids meant that bound antibodies could be detected without the use of B19-specific monoclonal antibodies. Thirty sera, with known values for B19-specific IgM in the RIA, were tested. The comparison is shown in Fig. 2 and, as for the IgG test, an optimal correlation was seen between the two assays: high and low values in the RIA correspond with high and low titers in the ELISA, with both types of particles. All negative sera in the RIA were also negative in the ELISA’s. A
0
10
20
30
40
50
60
70
80
90
100 ,100
R.I.A.(a.u.)
Fig. 1. Comparison of ELISA and RIA for detection of B19 specific IgG antibodies. Sera were tested in an ELISA with VP2 (+) and VPl-VP2 particles (0). ELISA titers were calculated in relation to a cut off and a calibration serum (Wielaard et al., 1985). The continuous line indicates the cut-off value (titer 100). RIA values are given in arbitrary units (a.u.): c 1, negative result; l-3, equivocal range; > 3, positive result. The dotted line indicates the upper limit of the equivocal range (a.u., 3).
254
0
10
20
30
40
50
60
70
80
90
100 ,100
R.1.A.Ca.u.)
Fig. 2. Comparison of ELISA and RIA for detection of B19 specific IgM antibodies. Sera were tested for IgM antibodies with biotinylated B19 VP2 particles (+) and VPI-VP2 particles (0). ELISA titers were calculated as described in Fig. 1 (Wielaard et al., 1985). The continuous line indicates the cut off-titer (titer 75). RIA values are given in a.u. The dotted line shows the upper limit of the equivocal range.
comparison was also made between the ELISA and the IFA for 101 sera from patients with a clinical indication for a B19 infection, using 1:lO and 150 dilutions in the IFA (Table 3). The same 39 sera were positive and the same 49 sera negative in both tests when VP2 particles (ELISA) were used while 41 sera were positive and 47 negative with VPl-VP2 particles. The sensitivity of the ELISA was slightly lower than that of the IFA. However, 6 of the 10 sera that were positive in the IFA and negative in the ELISA showed a very weak reaction in the IFA: a faint signal was only detectable at a 1:lO dilution, which is considered to be an extremely low titer. The sera that were positive by ELISA and negative by IFA always showed very low values, slightly above the cut-off value, perhaps indicating an equivocal range. When these six equivocal sera (in the IFA) were not included, the sensitivity was 90.7% and specificity 94.2% when VP2 particles were used, and 95.4%, respectively, 90.4% when VPl-VP2 particles were used. Thus, the correlation between the two assays was comparable to that found for the IgG results. TABLE 3 Comparison ELISA result
Pos. Pos. Neg. Neg.
of IFA and ELISA for detection IFA result
Pos. Neg. Pos. Neg.
of Bl9-specific
IgM antibodies
in 101 sera
No. of specimens ELISA VP2 particles
VPI-VP2 particles
39 3 loa 49
41
a6 of these sera were weakly positive in the IFA at a 1:lO dilution.
;a 47
To test the specificity of the IgM-ELISA, 14 sera known to contain RF were examined. None of these sera was reactive in the B19 IgM-ELISA with both types of particles, indicating that a possible interference by RF was prevented. To exclude nonspecific reactivity in the B19 IgM-ELISA of sera of patients with other acute viral infections, such as rubella (Kurtz et al., 1985), we tested 19 rubella IgM positive sera. None of these sera was reactive in the B19 IgMELISA, although one of these sera gave an A-value near the cut-off, using both types of particles. Antibody
response after infection
The antibody response after a B19 infection was examined. Sera taken at two or three time intervals from patients proven to be recently infected with B19 were tested in the ELISA for IgM and IgG antibodies with VP2 and VPl-VP2 capsids. The results obtained with VPZcapsids are shown in Fig. 3. Vpl-VP2 capsids gave similar values (results not shown). The first serum of each series tested had high IgM titers, decreasing in time in the subsequent sera of these series, whereas IgG titers increased in these sera. This pattern is typical for the part of the curve after a primary B19 infection when the IgM response is decreasing (Anderson et al., 1986). Due to the difficulty of obtaining sera directly after an infection, increasing IgM titers characteristic of the first part of the response curve could not be demonstrated. Approximately eight to ten weeks after infection, the sera became negative in the IgM test while IgG titers increased and elevated levels persisted. These results agree with former findings
1
2 10
2 721 weeks
24 after
28 onset
25
2 6
of illness
Fig. 3. Course of the antibody response after a recent B19 infection in six patients. Sera were tested in an ELISA using VPZ-capsids and titers were calculated in relation to two reference sera with known titers: a cut-off serum and a calibration serum.
256
(Anderson et al., 1986), indicating the reliability of the tests described. The ELISA described here is a sensitive and specific test for the detection of B19 IgG and IgM antibodies. Since the capsids mimic the native virus they react well with antibodies specific for conformational determinants, which may prove to be more important than linear epitopes. Other tests that have recently been described for B19 antibodies include an ELISA based on cultured B19 virus (Yaegashi et al., 1989) which has the limitation of small yields of virus produced in fetal liver cells that are difficult to culture. This is in contrast to the large amounts of capsids produced in insect cells. Approximately 500 pg capsids of high purity can be obtained from lo8 infected cells which is sufficient to coat 150 microtiter plates for IgG or 75 microtiter plates for IgM testing. Two ELISAs, based on linear peptides have also been recently described. In one case, the peptide represented the VPl-specific part of the capsid protein and was only shown to be effective for the detection of B19-specific IgG (Schwartz et al., 1991). The second test utilizes a peptide representing a short sequence of the VP2 protein (Fridell et al., 1989) which would not be expected to react with antibodies recognizing conformational epitopes. The ELISA described here thus offers solutions to a number of problems. The recombinant capsids resemble native virus, have a high purity and can be produced easily in large amounts. Acknowledgements
This work was supported (in part) by the ‘Praeventiefonds’ (project: 281517). We thank A.M.W. van Elsacker-Niele, P.E. Vermeer-de-Bondt and H.T. Weiland for providing serum specimens and helpful discussion. References Anand, A., Gray, ES., Brown, T., Clewley, J.P. and Cohen, B.J. (1987) Human parvovirus infection in pregnancy and hydrops fetalis. N. Engl. J. Med. 316, 183-186. Anderson, L.J., Tsou, C., Parker, R.A., Chorba, T.L., Wulff, H., Tattersall, P. and Mortimer, P.P. (1986) Detection of antibodies and antigens of human parvovirus B19 by enzyme-linked immunosorbent assay. J. Clin. Microbial. 24, 522-526. Anderson, M.J., Davis, L.R., Jones, S.E. and Pattison, J.R. (1982) The development and use of an antibody capture immunoassay for specific IgM to a human parvovirus-like agent. J. Hyg. Camb. 88, 309324. Anderson, M.J., Lewis, E., Kidd, J.M., Hall, S.M. and Cohen, B.J. (1984) An outbreak of erythema infectiosum associated with human parvovirus infection. J. Hyg. 93, 85-93. Brown, C.S., Salimans, M.M.M., Noteborn, M.H.M and Weiland, H.T. (1990a) Antigenic parvovirus B19 coat proteins VP1 and VP2 produced in large quantities in a baculovirus expression system. Virus Res. 15, 197-212. Brown, C.S., van Bussel, M.J.A.W.M., Wassenaar, A.L.M., van Elsacker-Niele, A.M.W., Weiland, H.T. and Salimans, M.M.M. (1990b) An immunofluorescence assay for the detection of parvovirus B19 IgG and IgM antibodies based on recombinant viral antigen. J. Virol. Methods 29. 5362.
257 Brown, C.S., van Lent, J.W.M., Vlak, J.M. and Spaan, W.J.M. (1991) Assembly of empty capsids by using baculovirus recombinants expressing human parvovirus B19 structural proteins. J. Virol. 65, 2702-2706. Brown, K.E., Mori, J., Cohen, B.J. and Field, A.M. (1991) In vitro propagation of parvovirus B19 in primary foetal liver culture. J. Gen. Virol. 72, 741-745. Centers for Disease Control. (1989) Risks associated with human parvovirus B19 infection. Morbid. Mortal. Weekly Rep. 38, 81-97. Cohen, B.J., Mortimer, P.P. and Pereira, M.S. (1983) Diagnostic assays with monoclonal antibodies for the human serum parvovirus-like virus (SPLV). J. Hyg. 91, 113-130. Cohen, B.J. and Buckley, M.S. (1988) The prevalence of antibody to human parvovirus B19 in England and Wales. J. Med. Microbial. 25, 151-153. Cossart, Y.E., Cant, B., Field, A.M. and Widdows, D. (1975) Parvovirus-like particles in human sera. Lancet i, 72-13. French, T.J. and Roy, P. (1990) Synthesis of bluetongue virus (BTV) corelike particles by a recombinant baculovirus expressing the two major structural core proteins of BTV. J. Virol. 64, 1530-1536. Fridell, E., Trojnar, J. and Warren, B. (1989) A new peptide for human parvovirus B19 antibody detection. Stand. J. Inf. Dis. 21, 597-603. Hall, S. (1985) Infection with parvovirus during pregnancy. Br. Med. J. 290, 713-714. Kajigaya, S., Shimada, T., Jujita, S. and Young, N.S. (1989) A genetically engineered cell line that produces empty capsids of B19 (human) parvovirus. Proc. Natl. Acad. Sci. USA 86,7601-7605. Kajigaya, S., Fujii, H., Field, A., Anderson, S., Rosenfeld, S., Anderson, L.J., Shimada, T. and Young, N.S. (1991) Self-assembled B19 parvovirus capsids, produced in a baculovirus system, are antigenically and immunogenically similar to native virions. Proc. Natl. Acad. Sci. USA 88, 46464650. Kurtz, J.B. and Anderson, M.J. (1985) Cross-reaction in rubella and parvovirus specific IgM tests. Lancet ii, 1356. Kurtzmann, G.J., Cohen, B., Meyers, P., Amunullah, A. and Young, N.S. (1988) Persistent B19 parvovirus infection as a cause of severe chronic anemia in children with acute lymphocytic leukemia. Lancet ii, 1159-l 162. Labbe, M., Charpilienne, A., Crawford, SE., Estes, M.K. and Cohen, J. (1991) Expression of rotavirus VP2 produces empty corelike particles. J. Virol. 65, 29462952. Maeda, H., Shimokawa, H., Satoh, S., Nakano, H. and Nunoue, T. (1988) Non-immunologic hydrops fetalis resulting from intrauterine human parvovirus B19 infection: report of two cases. Obstet. Gynecol. 9, 482-485. Matsuura, Y., Possee, R.D., Overton, H.A. and Bishop, D.H.L. (1987) Baculovirus expression vectors: the requirements for high level expression of proteins, including glycoproteins. J. Gen. Virol. 68, 1233-1250. Morinet, F., D’Auriol, L., Tratschin, J.D. and Galibert, F. (1989) Expression of the human parvovirus B19 protein fused to protein A in Escherichiu coli: recognition by IgM and IgG antibodies in human sera. J. Gen. Virol. 70, 3091-3097. Mortimer, P.P., Cohen, B.J., Buckley, M.M., Cradock-Watson, J.E., Ridehalgh, M.K.S., Burkhardt, F. and Schilt, U. (1985) Human parvovirus and the fetus. Lancet ii, 1012. Ozawa, K., Kurtzman, G. and Young, N. (1986) Replication of the B19 parvovirus in human bone marrow cell culture. Science 233, 883-886. Reid, D.M., Reid, T.M.S., Brown, T., Rennie, J.A.N. and Eastmond, C.J. (1985) Human parvovirus-associated arthritis: a clinical and laboratory description. Lancet i, 422-425. Schwarz, T.F., Roggendorf, M. and Deinhardt, F. (1988) Human parvovirus B19: ELISA and immunoblot assays. J. Virol. Methods 20, 155-168. Schwarz, T.F., Modrow, S., HottentrPger, B., Hoflacher, B., Jiiger, G., Schartl, W., Sumazaki, R., Wolf, H., Middeldorp, J., Roggendorf, M. and Deinhardt, F. (1991) New oligopeptide immunoglobulin G test for human parvovirus B19 antibodies. J. Clin. Microbial. 29, 431435. Serjeant, G.R., Mason, K., Topley, J.M., Serjeant, B.E., Pattison, J.R., Jones, SE. and Mohamed, R. (1981) Outbreak of aplastic crisis in sickle cell aneamia associated with parvovirus-like agent.
258 Lancet ii, 595-598. Sisk, W.P. and Berman, M.L. (1987) Expression of human parvovirus Bl9 structural protein in Escherichia coli and detection of antiviral antibodies in human sera. Biotechnology 5, 1077-1080. Urakawa, T., Ferguson, M., Minor, P.D., Cooper, J., Sullivan, M., Almond, J.W. and Bishop, D.H.L. (1989) Synthesis of immunogenic, but not-infectious, poliovirus particles in insect cells by a baculovirus expression vector. J. Gen. Virol. 70, 1453-1463. Vlak, J.M., Klinkenberg, F.A., Zaal, K.J.M., Usmany, M., Klinge-Roode, E.C., Geertvliet, J.B.F., Roosien, L. and van Lent, J.W.M. (1988) Functional studies on the pl0 gene of Autographa californica nuclear polyhedrosis virus using a recombinant expressing a plO+galactosidase fusion gene. J. Gen. Virol. 69, 765-776. White, D.G., Woolf, A.D., Mortimer, P.P., Cohen, B.J., Blake, D.R. and Bacon, P.A. (1985) Human parvovirus arthropathy. Lancet i, 419421. Wielaard, F., Denissen, A., van Elleswijk-v.d.Berg, J. and van Gemert, G. (1985) Clinical validation of an antibody-capture anti-rubella IgM-ELISA. J. Virol. Methods 10, 349-354. Yaegashi, N., Shiraishi, H., Takeshita, T., Nakamura, M., Yajima, S. and Sugamura, K. (1989) Propagation of human parvovirus Bl9 in primary culture of erythroid lineage cells derived from fetal liver. J. Virol. 63, 2422-2426. Yaegashi, N., Shiraishi, H., Tada, K., Yajima, A. and Sugamura, K. (1989) Enzyme-linked immunosorbent assay for IgG and IgM antibodies against human parvovirus Bl9: use of monoclonal antibodies and viral antigen propagated in vitro. J. Virol. Methods 26, 171-182.
Journal of Virological Methods, 39 (1992) 241-258 0 1992 Elsevier Science Publishers B.V. / All rights reserved / 0166-0934/92/%05.00 VIRMET 01383
Recombinant parvovirus B19 capsids as a new substrate for detection of B 19-specific IgG and IgM antibodies by an enzyme-linked immunosorbent assay Marcel
M.M. Salimans,
Mario J.A.W.M. van Bussel, Caroline and Willy J.M. Spaan
S. Brown
Department of Virology, University Hospital Leia’en and Faculty of Medicine. Leiden (The Netherlands) (Accepted
16 April 1992)
Summary A new enzyme-linked immunosorbent assay for the detection of B19-specific IgG and IgM antibodies was established using B19 capsids synthesized in a baculovirus expression system. These 1319 capsids, consisting of either coat protein VP2 alone or of both VP1 and VP2, have been shown to be similar to native virus in size and appearance. The results obtained for the detection of B 19-specific antibodies showed good correlations with a radioimmunoassay which uses native B19 virus and an immunofluorescence assay based on insect cells expressing coat protein VPl. The course of the antibody response could be followed by determining the titers of sequential serum samples taken after a recent B19 infection. Both types of recombinant capsids form an excellent source of antigen for the detection of both B19 IgG and IgM antibodies and are a very promising substitute for native virus. Human parvovirus B19; Recombinant capture IgM ELISA
viral antigen; Diagnosis; ELISA; Antibody
Correspondence to: M.M. Salimans, Department of Virology, University Hospital L&den, P.O. Box 320, 2300 AH L&den, The Netherlands.
248
Introduction Human parvovirus B19, since its discovery in 1975 (Cossart et al., 1975), has been associated with several defined clinical syndromes. The virus is the causal agent of erythema infectiosum, the fifth disease of childhood (Anderson et al., 1984) and some cases of acute arthritis and arthralgia, especially in adult women (Reid et al., 1985; White et al., 1985). A B19 infection in patients with chronic hemolytic anemia can lead to an aplastic crisis (Serjeant et al., 1981) and the virus has been associated with serious complications during pregnancy such as abortion, intrauterine fetal death (Hall, 1985; Mortimer et al., 1985), hydrops fetalis (Anand et al., 1987; Maeda et al., 1988; CDC, 1989) and with chronic anemia in the immunodeficient patient (Kurtzmann et al., 1988). The development of diagnostic tests for the detection of B19-specific IgG and IgM antibodies (Anderson et al., 1982; Anderson et al., 1986; Schwartz et al., 1988) has been hindered by the limited amount of viral antigen available. The virus has not yet been propagated in a stable cell-line, although low-level propagation of virus in bone marrow explants (Ozawa et al., 1986) and fetal liver cells (Yaegashi et al., 1989; Brown et al., 1991) has been achieved. Isolation of the virus from serum relies on the random screening of blood donors since symptoms occur when a patient is no longer in the viraemic phase. To overcome this limited supply of antigen, attention has been focussed on recombinant DNA techniques for the production of antigen. Recent publications have described the successful expression of B19 proteins in several systems: Escherichia coli (Sisk et al., 1987; Morinet et al., 1989), Chinese Hamster Ovary (CHO) cells (Kajigaya et al., 1989) and insect cells, using a baculovirus expression system (Brown et al., 1990a; C.S. Brown et al., 1991; Kajigaya et al., 1991). The first attempts to develop diagnostic tests based on these renewable sources of B19 antigens have been described previously (Brown et al., 1990b; Schwartz et al., 1991). The baculovirus expression system was chosen for the production of viral antigens due to the high levels of expression achieved for foreign proteins. Large amounts of the B19 structural proteins have been produced in this system and insect cells expressing VP1 show a strong and characteristic signal in the immunofluorescence assay when reacted with BlPspecific IgG and IgM antibodies (Brown et al., 1990b). More recently, it has been shown for a number of viruses, such as poliovirus (Urakawa et al., 1989), bluetongue virus (French et al., 1990) and rotavirus (Labbe et al., 1990), that virus-like particles can be synthesized from baculovirus recombinants in insect cells. These particles mimic the native virus and are non-infectious. We and others have lately shown that when VP2 alone or both VP1 and VP2 are simultaneously expressed in the baculovirus system, empty B19 capsids are formed (Brown et al., 1991; Kajigaya et al., 1991). Due to their similarity to native B19 and their antigenicity, these capsids promised to be an excellent candidate for the development of a more convenient serological test. In this paper we describe the use of empty B19 capsids for the detection of
249
BlPspecific IgG antibody and the labeling and use of these capsids for specific IgM antibody detection in an enzyme-linked immunosorbent assay (ELISA).
Materials and Methods Clinical specimens
Groups of sera tested: (i) serum specimens previously tested in the radioimmunoassay (RIA) with native B19 virus at the Public Health Laboratory Service in London, kindly supplied by Dr. B.J. Cohen; (ii) sera from patients with a clinical indication for a parvovirus B19 infection, previously tested in the immunofluorescence assay (IFA) in our clinical diagnostic laboratory; (iii) random blood donor sera from the Blood Bank of the University Hospital Leiden, The Netherlands; (iv) sera positive for rheumatoid factor (RF), as determined by latex agglutination and; (v) sera positive for rubella-specific IgM antibodies. Expression of B19 proteins in insect cells
The cloning of the genes for the B19 coat proteins VP1 and VP2 for expression in insect cells has been described previously in detail (Brown et al., 1990a; Brown et al., 1991). Briefly, the VP1 gene was expressed from the polyhedrin promoter in a recombinant baculovirus produced upon recombination between wildtype baculovirus and the plasmid vector pAcYM 1 (Matsuura et al., 1987), containing the VP1 DNA sequence behind the polyhedrin promoter, in Spodoptera frugiperda monolayers. The same procedure was followed for the expression of VP2 (Brown et al., 1990a). Double recombinant virus, expressing both VP1 and VP2, was produced by recombination between the VP2 recombinant virus, expressing the VP2 gene from the polyhedrin promoter, and the plasmid vector pAcAS3 (Vlak et al., 1988), which contains the VP1 DNA sequence behind the p10 promoter. Purification of B19 protein capsids
The isolation of B19 capsids from insect cells infected with the VP2 recombinant and double recombinant baculovirus was essentially the same as described previously (Brown et al., 1991), except for the final purification step. Instead of a sucrose gradient (Brown et al., 1991), the particles were purified by centrifugation in an isopycnic CsCl-gradient (28%, wt/wt), performed with a Beckmann SW55 TI rotor at 100000 x g for 24 h at 5°C. An opaque protein band was clearly detectable in the gradient, which contained the capsids. The bands were extracted, dialysed overnight against PBS and stored at -80°C. The purity of the capsid preparations was checked by running 1 lug amounts on SDS-PAGE and staining (Fast Green or Coomassie brilliant blue) the gel: 90-
95% purity was obtained. Labeling of the B19 capsids For the IgM-ELISA the B19 capsids were biotinylated as follows: the particles were dialysed for 2.5 h at 4°C against 0.1 M NaHC03 (pH 8.7) and diluted to a final concentration of 0.1-1.0 mg/ml. Subsequently, biotinyl-l\rhydroxysuccinimid (BNHS) was added during constant gentle mixing to a final concentration of 0.1 mg per mg protein. BNHS was dissolved, as a stock solution, in dimethylsulfoxide to a concentration of 2.5 mg/ml. After a 3 h incubation at room temperature, the biotinylated particles were dialysed overnight against PBS at 4°C and stored at -80°C. ELISA for Bl9-specific
IgG antibodies
To test for the presence of parvovirus B19-specific IgG antibodies in patient sera, microtiter plates (polysorb F96-3, Nunc) were coated with 25 ng per well of B19 capsids (the optimal amount was determined by dilution experiments) overnight at 4°C in PBS. After washing four times with PBS containing 0.05% Tween-20 (washing-buffer), 100 ,ul of diluted serum was added and incubated for 1 h at 37°C. As a standard procedure, all sera were diluted 1:lOO in dilutionbuffer (2% fetal calf serum, 0.05% Tween-20 and 0.1 mg/ml merthiolate in PBS). The wells were washed four times with washing buffer and 100 ~1 of a peroxidase-labeled rabbit-anti-human IgG (F(ab)2, Dakopatts, Denmark) was added at a 1:4000 dilution (determined by dilution experiments) in dilution buffer and incubated for 1 h at 37°C. After washing four times with washing buffer the plates were incubated with 0-phenylenediamine.2HCl (OPD, Abbott, USA) for 30 min in the dark at room temperature and the reaction was stopped with 4 N HzS04. The optical density (A) was measured at 492 nm. The ‘cut off value for the ELISA was determined by the mean A-value of 50 IFA-negative sera (A = 0.20-0.25) to which 2-times the standard deviation was added. The A-value of the cut off was usually between 0.30-0.35 for both VP2 and VPl-VP2 particles. Titers of the sera tested were determined, using a calibration curve, as previously described by Wielaard et al. (1985). Titer 100 was considered to be the cut-off value. During our ELISA experiments, it became clear that the choice of microtiter plates could influence the results. In our hands, the polysorb-plates (Nunc) gave the best results in the IgG- as well as the IgM-ELISA. The background was low and positive sera gave high values when compared to other plates. ELISA for Bl9-speciJic
IgM antibodies
The presence of B19-specific IgM antibodies was determined by means of an anti-p capture assay. Microtiter plates were coated with a 1:lOOO dilution of rabbit-anti-human IgM (p-chain, Dakopatts, Denmark) in PBS overnight at
251
4°C. After washing (see IgG procedure) sera were added at a dilution of 1:lOO in dilution buffer (see IgG procedure) and incubated for 3 h at 37°C or overnight at 4°C. The plates were washed and 50 ng per well of biotin-labeled B19 capsids were added and incubated for 2 h at 37°C. After washing, bound particles were detected by incubation with horse radish peroxidase-labeled avidine (Dakopatts, Denmark) at a 1:25 000 dilution in dilution buffer for 1 h at 37°C. Finally, the wash-steps and staining reaction was the same as for the IgG-procedure. The cut off and the calculation of titers were essentially the same as described for the IgG-ELISA. The A-value of the cut-off serum was approx. 0.25-0.30 (mean value of 50 negative sera was 0.18-0.23) for both VP2 and VPl-VP2 particles and a titer of 75 was taken as the cut-off value. Immunofluorescence assay (IFA) The immunofluorescence assay was performed on insect cells that had been infected with the recombinant baculovirus expressing VPl. The procedure for the detection of B19-specilic IgG and IgM antibodies was essentially the same as described before (Brown et al., 1990b). Briefly, glass slides, onto which insect cells expressing VP1 had been fixed, were incubated with sera and bound antibody was detected with FITC-conjugated goat-anti-human IgG or FITCconjugated goat-anti-human IgM (Institute Pasteur, France). To avoid false-positive IgM results due to the presence of RF and to prevent competition by IgG antibodies, sera to be tested for B19 IgM antibodies were pretreated with a goat-anti-human-IgG preparation (GULLSORB, Gull Lab., USA).
Results and Discussion Due to the limited supply of native B19 virus, an ELISA for the detection of IgG and IgM antibodies was developed, based on recombinant capsids produced in insect cells. The capsids, formed from VP2 expressed alone or VP1 and VP2 expressed simultaneously in insect cells, could be simply and rapidly purified and high yields of antigen were obtained (Brown et al., 1991). These capsids have been shown to resemble native virus, which is an important factor if conformational epitopes are torbe recognized. That conformational epitopes play an important role in the immune response against B19 was demonstrated by the fact that 13 sera, that reacted with insect cells expressing VP2 in an IFA, did not react with the same recombinant VP2 in a western blot (Brown et al., 1990a). Also, 17 sera that reacted with both types of capsids in the ELISA did not react with VP2 in a western blot (results not shown). This demonstrates that the capsids possess conformational determinants that are recognized by antibodies in patient sera, whereas these antibodies do not recognize nonconformational epitopes in a Western blot assay. The application of recombinant B19 capsids in an EIA has recently been demonstrated by
252
Kajigaya
and coworkers
(1991) who used lysates of insect cells expressing VP2.
IgG-ELBA Thirty serum samples, tested previously by RIA based on native B19 virus (Cohen et al., 1983), were tested for the presence of B19-specific IgG in the ELISA with both types of capsids. A comparison of the results is shown in Fig. 1 demonstrating an optimal correlation between the ELISA titers and the RIA values throughout the whole range. In most cases, high ELISA titers corresponded with high RIA values. In the equivocal range of the RIA (l-3 a.u.) some of the sera were weakly positive in the ELISA while most of them were negative. All negative sera in the RIA were also negative in the ELISA. Titers obtained in the ELISA compare very well for both VP2 and VPl-VP2 particles, although the VPl-VP2 capsids gave somewhat higher titers. The capsids that incorporate VP1 may more closely resemble native B19 and thus react better with the sera. However, since no serum samples have been found so far that only react with VPl-VP2 particles, the contribution of the incorporated VP1 seems to be limited. Both types of capsids have a similar structure to the native virus and gave, on the whole, comparable results. A second comparison was made between the ELISA, based on the VP2 capsids, and the IFA, based on recombinant VP1 protein (Brown et al., 1990b), for 99 serum samples. These sera, from patients with a clinical indication for a B19 infection, were tested in the IFA at dilutions 150, 1:500, 15000 and scored positive or negative. The results (Table 1) show that the same 66 sera were positive and the same 28 negative in both assays, indicating an excellent correlation. Comparable results were obtained with VPl-VP2 particles. When compared to the IFA the ELISA showed a higher sensitivity and specificity in the IgGtest. The sensitivity was 94.3% and the specificity 96.5% when VP2 particles were used and 92.9%, respectively, 100% when VPl-VP2 particles were used. The prevalence of B19-specific IgG antibodies in the general population was analyzed by testing a random selection of blood donor sera with both types of protein particles. 255 sera were tested and the results (Table 2) show that 79% TABLE 1 Comparison ELISA result
Pos. Pos. Neg. Neg.
of IFA and ELISA for detection of Bl9-specific IFA result
Pos. Neg. Pos. Neg.
IgG antibodies
in 99 sera
No. of specimens ELISA VP2 particles
VP1 -VP2 particles
66
65 0
:a 28
‘Weakly positive in the IFA at a 1:50 dilution. b4 of these sera were weakly positive in the IFA at a 1:50 dilution.
b
253 TABLE 2 Prevalence of B19-specific IgG antibodies ELISA titer
in blood donor sera (n = 255)
No. of sera tested in ELISA
< 100 (=neg.) 100-1000 1oOck1oooo > 10000
VP2 particles
VP1 -VP2 particles
1;:
54 106 63 32
15 18
(n = 201) sera reacted positively when VP2 and VPl-VP2 particles were used. The distribution of titer values shows that 46% (for VP2) and 47% (for VPlVP2) of these positive sera had titers > 1000. These results compare very well with previous findings in this age group (Cohen et al., 1988; Brown et al., 1990b). IgM-ELISA
To avoid false-positive reactivity due to the presence of RF a p-capture assay was developed. The use of biotin-labeled B19 capsids meant that bound antibodies could be detected without the use of B19-specific monoclonal antibodies. Thirty sera, with known values for B19-specific IgM in the RIA, were tested. The comparison is shown in Fig. 2 and, as for the IgG test, an optimal correlation was seen between the two assays: high and low values in the RIA correspond with high and low titers in the ELISA, with both types of particles. All negative sera in the RIA were also negative in the ELISA’s. A
0
10
20
30
40
50
60
70
80
90
100 ,100
R.I.A.(a.u.)
Fig. 1. Comparison of ELISA and RIA for detection of B19 specific IgG antibodies. Sera were tested in an ELISA with VP2 (+) and VPl-VP2 particles (0). ELISA titers were calculated in relation to a cut off and a calibration serum (Wielaard et al., 1985). The continuous line indicates the cut-off value (titer 100). RIA values are given in arbitrary units (a.u.): c 1, negative result; l-3, equivocal range; > 3, positive result. The dotted line indicates the upper limit of the equivocal range (a.u., 3).
254
0
10
20
30
40
50
60
70
80
90
100 ,100
R.1.A.Ca.u.)
Fig. 2. Comparison of ELISA and RIA for detection of B19 specific IgM antibodies. Sera were tested for IgM antibodies with biotinylated B19 VP2 particles (+) and VPI-VP2 particles (0). ELISA titers were calculated as described in Fig. 1 (Wielaard et al., 1985). The continuous line indicates the cut off-titer (titer 75). RIA values are given in a.u. The dotted line shows the upper limit of the equivocal range.
comparison was also made between the ELISA and the IFA for 101 sera from patients with a clinical indication for a B19 infection, using 1:lO and 150 dilutions in the IFA (Table 3). The same 39 sera were positive and the same 49 sera negative in both tests when VP2 particles (ELISA) were used while 41 sera were positive and 47 negative with VPl-VP2 particles. The sensitivity of the ELISA was slightly lower than that of the IFA. However, 6 of the 10 sera that were positive in the IFA and negative in the ELISA showed a very weak reaction in the IFA: a faint signal was only detectable at a 1:lO dilution, which is considered to be an extremely low titer. The sera that were positive by ELISA and negative by IFA always showed very low values, slightly above the cut-off value, perhaps indicating an equivocal range. When these six equivocal sera (in the IFA) were not included, the sensitivity was 90.7% and specificity 94.2% when VP2 particles were used, and 95.4%, respectively, 90.4% when VPl-VP2 particles were used. Thus, the correlation between the two assays was comparable to that found for the IgG results. TABLE 3 Comparison ELISA result
Pos. Pos. Neg. Neg.
of IFA and ELISA for detection IFA result
Pos. Neg. Pos. Neg.
of Bl9-specific
IgM antibodies
in 101 sera
No. of specimens ELISA VP2 particles
VPI-VP2 particles
39 3 loa 49
41
a6 of these sera were weakly positive in the IFA at a 1:lO dilution.
;a 47
To test the specificity of the IgM-ELISA, 14 sera known to contain RF were examined. None of these sera was reactive in the B19 IgM-ELISA with both types of particles, indicating that a possible interference by RF was prevented. To exclude nonspecific reactivity in the B19 IgM-ELISA of sera of patients with other acute viral infections, such as rubella (Kurtz et al., 1985), we tested 19 rubella IgM positive sera. None of these sera was reactive in the B19 IgMELISA, although one of these sera gave an A-value near the cut-off, using both types of particles. Antibody
response after infection
The antibody response after a B19 infection was examined. Sera taken at two or three time intervals from patients proven to be recently infected with B19 were tested in the ELISA for IgM and IgG antibodies with VP2 and VPl-VP2 capsids. The results obtained with VPZcapsids are shown in Fig. 3. Vpl-VP2 capsids gave similar values (results not shown). The first serum of each series tested had high IgM titers, decreasing in time in the subsequent sera of these series, whereas IgG titers increased in these sera. This pattern is typical for the part of the curve after a primary B19 infection when the IgM response is decreasing (Anderson et al., 1986). Due to the difficulty of obtaining sera directly after an infection, increasing IgM titers characteristic of the first part of the response curve could not be demonstrated. Approximately eight to ten weeks after infection, the sera became negative in the IgM test while IgG titers increased and elevated levels persisted. These results agree with former findings
1
2 10
2 721 weeks
24 after
28 onset
25
2 6
of illness
Fig. 3. Course of the antibody response after a recent B19 infection in six patients. Sera were tested in an ELISA using VPZ-capsids and titers were calculated in relation to two reference sera with known titers: a cut-off serum and a calibration serum.
256
(Anderson et al., 1986), indicating the reliability of the tests described. The ELISA described here is a sensitive and specific test for the detection of B19 IgG and IgM antibodies. Since the capsids mimic the native virus they react well with antibodies specific for conformational determinants, which may prove to be more important than linear epitopes. Other tests that have recently been described for B19 antibodies include an ELISA based on cultured B19 virus (Yaegashi et al., 1989) which has the limitation of small yields of virus produced in fetal liver cells that are difficult to culture. This is in contrast to the large amounts of capsids produced in insect cells. Approximately 500 pg capsids of high purity can be obtained from lo8 infected cells which is sufficient to coat 150 microtiter plates for IgG or 75 microtiter plates for IgM testing. Two ELISAs, based on linear peptides have also been recently described. In one case, the peptide represented the VPl-specific part of the capsid protein and was only shown to be effective for the detection of B19-specific IgG (Schwartz et al., 1991). The second test utilizes a peptide representing a short sequence of the VP2 protein (Fridell et al., 1989) which would not be expected to react with antibodies recognizing conformational epitopes. The ELISA described here thus offers solutions to a number of problems. The recombinant capsids resemble native virus, have a high purity and can be produced easily in large amounts. Acknowledgements
This work was supported (in part) by the ‘Praeventiefonds’ (project: 281517). We thank A.M.W. van Elsacker-Niele, P.E. Vermeer-de-Bondt and H.T. Weiland for providing serum specimens and helpful discussion. References Anand, A., Gray, ES., Brown, T., Clewley, J.P. and Cohen, B.J. (1987) Human parvovirus infection in pregnancy and hydrops fetalis. N. Engl. J. Med. 316, 183-186. Anderson, L.J., Tsou, C., Parker, R.A., Chorba, T.L., Wulff, H., Tattersall, P. and Mortimer, P.P. (1986) Detection of antibodies and antigens of human parvovirus B19 by enzyme-linked immunosorbent assay. J. Clin. Microbial. 24, 522-526. Anderson, M.J., Davis, L.R., Jones, S.E. and Pattison, J.R. (1982) The development and use of an antibody capture immunoassay for specific IgM to a human parvovirus-like agent. J. Hyg. Camb. 88, 309324. Anderson, M.J., Lewis, E., Kidd, J.M., Hall, S.M. and Cohen, B.J. (1984) An outbreak of erythema infectiosum associated with human parvovirus infection. J. Hyg. 93, 85-93. Brown, C.S., Salimans, M.M.M., Noteborn, M.H.M and Weiland, H.T. (1990a) Antigenic parvovirus B19 coat proteins VP1 and VP2 produced in large quantities in a baculovirus expression system. Virus Res. 15, 197-212. Brown, C.S., van Bussel, M.J.A.W.M., Wassenaar, A.L.M., van Elsacker-Niele, A.M.W., Weiland, H.T. and Salimans, M.M.M. (1990b) An immunofluorescence assay for the detection of parvovirus B19 IgG and IgM antibodies based on recombinant viral antigen. J. Virol. Methods 29. 5362.
257 Brown, C.S., van Lent, J.W.M., Vlak, J.M. and Spaan, W.J.M. (1991) Assembly of empty capsids by using baculovirus recombinants expressing human parvovirus B19 structural proteins. J. Virol. 65, 2702-2706. Brown, K.E., Mori, J., Cohen, B.J. and Field, A.M. (1991) In vitro propagation of parvovirus B19 in primary foetal liver culture. J. Gen. Virol. 72, 741-745. Centers for Disease Control. (1989) Risks associated with human parvovirus B19 infection. Morbid. Mortal. Weekly Rep. 38, 81-97. Cohen, B.J., Mortimer, P.P. and Pereira, M.S. (1983) Diagnostic assays with monoclonal antibodies for the human serum parvovirus-like virus (SPLV). J. Hyg. 91, 113-130. Cohen, B.J. and Buckley, M.S. (1988) The prevalence of antibody to human parvovirus B19 in England and Wales. J. Med. Microbial. 25, 151-153. Cossart, Y.E., Cant, B., Field, A.M. and Widdows, D. (1975) Parvovirus-like particles in human sera. Lancet i, 72-13. French, T.J. and Roy, P. (1990) Synthesis of bluetongue virus (BTV) corelike particles by a recombinant baculovirus expressing the two major structural core proteins of BTV. J. Virol. 64, 1530-1536. Fridell, E., Trojnar, J. and Warren, B. (1989) A new peptide for human parvovirus B19 antibody detection. Stand. J. Inf. Dis. 21, 597-603. Hall, S. (1985) Infection with parvovirus during pregnancy. Br. Med. J. 290, 713-714. Kajigaya, S., Shimada, T., Jujita, S. and Young, N.S. (1989) A genetically engineered cell line that produces empty capsids of B19 (human) parvovirus. Proc. Natl. Acad. Sci. USA 86,7601-7605. Kajigaya, S., Fujii, H., Field, A., Anderson, S., Rosenfeld, S., Anderson, L.J., Shimada, T. and Young, N.S. (1991) Self-assembled B19 parvovirus capsids, produced in a baculovirus system, are antigenically and immunogenically similar to native virions. Proc. Natl. Acad. Sci. USA 88, 46464650. Kurtz, J.B. and Anderson, M.J. (1985) Cross-reaction in rubella and parvovirus specific IgM tests. Lancet ii, 1356. Kurtzmann, G.J., Cohen, B., Meyers, P., Amunullah, A. and Young, N.S. (1988) Persistent B19 parvovirus infection as a cause of severe chronic anemia in children with acute lymphocytic leukemia. Lancet ii, 1159-l 162. Labbe, M., Charpilienne, A., Crawford, SE., Estes, M.K. and Cohen, J. (1991) Expression of rotavirus VP2 produces empty corelike particles. J. Virol. 65, 29462952. Maeda, H., Shimokawa, H., Satoh, S., Nakano, H. and Nunoue, T. (1988) Non-immunologic hydrops fetalis resulting from intrauterine human parvovirus B19 infection: report of two cases. Obstet. Gynecol. 9, 482-485. Matsuura, Y., Possee, R.D., Overton, H.A. and Bishop, D.H.L. (1987) Baculovirus expression vectors: the requirements for high level expression of proteins, including glycoproteins. J. Gen. Virol. 68, 1233-1250. Morinet, F., D’Auriol, L., Tratschin, J.D. and Galibert, F. (1989) Expression of the human parvovirus B19 protein fused to protein A in Escherichiu coli: recognition by IgM and IgG antibodies in human sera. J. Gen. Virol. 70, 3091-3097. Mortimer, P.P., Cohen, B.J., Buckley, M.M., Cradock-Watson, J.E., Ridehalgh, M.K.S., Burkhardt, F. and Schilt, U. (1985) Human parvovirus and the fetus. Lancet ii, 1012. Ozawa, K., Kurtzman, G. and Young, N. (1986) Replication of the B19 parvovirus in human bone marrow cell culture. Science 233, 883-886. Reid, D.M., Reid, T.M.S., Brown, T., Rennie, J.A.N. and Eastmond, C.J. (1985) Human parvovirus-associated arthritis: a clinical and laboratory description. Lancet i, 422-425. Schwarz, T.F., Roggendorf, M. and Deinhardt, F. (1988) Human parvovirus B19: ELISA and immunoblot assays. J. Virol. Methods 20, 155-168. Schwarz, T.F., Modrow, S., HottentrPger, B., Hoflacher, B., Jiiger, G., Schartl, W., Sumazaki, R., Wolf, H., Middeldorp, J., Roggendorf, M. and Deinhardt, F. (1991) New oligopeptide immunoglobulin G test for human parvovirus B19 antibodies. J. Clin. Microbial. 29, 431435. Serjeant, G.R., Mason, K., Topley, J.M., Serjeant, B.E., Pattison, J.R., Jones, SE. and Mohamed, R. (1981) Outbreak of aplastic crisis in sickle cell aneamia associated with parvovirus-like agent.
258 Lancet ii, 595-598. Sisk, W.P. and Berman, M.L. (1987) Expression of human parvovirus Bl9 structural protein in Escherichia coli and detection of antiviral antibodies in human sera. Biotechnology 5, 1077-1080. Urakawa, T., Ferguson, M., Minor, P.D., Cooper, J., Sullivan, M., Almond, J.W. and Bishop, D.H.L. (1989) Synthesis of immunogenic, but not-infectious, poliovirus particles in insect cells by a baculovirus expression vector. J. Gen. Virol. 70, 1453-1463. Vlak, J.M., Klinkenberg, F.A., Zaal, K.J.M., Usmany, M., Klinge-Roode, E.C., Geertvliet, J.B.F., Roosien, L. and van Lent, J.W.M. (1988) Functional studies on the pl0 gene of Autographa californica nuclear polyhedrosis virus using a recombinant expressing a plO+galactosidase fusion gene. J. Gen. Virol. 69, 765-776. White, D.G., Woolf, A.D., Mortimer, P.P., Cohen, B.J., Blake, D.R. and Bacon, P.A. (1985) Human parvovirus arthropathy. Lancet i, 419421. Wielaard, F., Denissen, A., van Elleswijk-v.d.Berg, J. and van Gemert, G. (1985) Clinical validation of an antibody-capture anti-rubella IgM-ELISA. J. Virol. Methods 10, 349-354. Yaegashi, N., Shiraishi, H., Takeshita, T., Nakamura, M., Yajima, S. and Sugamura, K. (1989) Propagation of human parvovirus Bl9 in primary culture of erythroid lineage cells derived from fetal liver. J. Virol. 63, 2422-2426. Yaegashi, N., Shiraishi, H., Tada, K., Yajima, A. and Sugamura, K. (1989) Enzyme-linked immunosorbent assay for IgG and IgM antibodies against human parvovirus Bl9: use of monoclonal antibodies and viral antigen propagated in vitro. J. Virol. Methods 26, 171-182.
