An immunofluorescence assay for the detection of parvovirus B19 IgG and IgM antibodies based on recombinant viral antigen

Journal of ~rofogica~Methods,29 (1990) 53-62 Elsevier

53

VIRMET 01037

An immunofluorescence assay for the detection of parvovirus B19 IgG and IgM antibodies based on recombinant viral antigen Caroline S. Brown, Mario J.A.W.M. van Bussel, Alfred L.M. Wassenaar, Anne-Marie W, van Elsacker-Niele, Harro T. Weiland and Marcel M.M. Salimans Depa~ent

of Virology, Facufty ofMedicine, Universes Hospital Leiden, The ~e#her~a~ (Accepted

27 March 1990)

Summary An indirect immunofluorescence assay for serum IgG and IgM antibodies to human parvovirus B19 was established using recombinant B19 viral antigen, the capsid protein VPl, which had been produced in a baculovirus expression system. This protein gives a strong and characteristic signal in the immunofluorescence assay, making it a suitable candidate for this test system. The test results showed a good correlation with results obtained with a solid-phase capture ra~oi~unoassay (Cohen et al., 1983). 76% of sera from a random selection of blood donors were positive for B19 IgG which agrees with previous findings. The course of the IgM and IgG antibody response to B19 infection could be followed with the immunofluorescence assay by determining the titers of series of sera taken after a recent B19 infection. Investigation of 24 sera containing rubella-specific IgM showed no cross-reactivity with the recombinant B19 VP1 used in this test system. The test described here has the advantage of being based on a renewable source of antigen and will be further evaluated for routine diagnostic use in comparison with radioimmunoassay. Human p~ovi~s B19; ~~ornbin~t assay; B 19 diagnosis

viral antigen; Indirect i~unofluo~scence

Correspondence to: Caroline S. Brown, Department of Virology, P.0. Box 320, 2300 AH Leiden, The Netherlands. 0166~0934/90/$03.50 0 1990 Elsevier Science Publishers B.V. (Biomedical Division)

54

Introduction Since its discovery in 1975 by Cossart et al., the clinical relevance of human parvovirus B19 has been established. It is the causal agent of erythema i~ectiosum (EI), the fifth disease of childhood, while B 19 infection in adults, especially women, can result in arthralgia or arthritis. Complications related to the virus’s propensity to infect and lyse red-cell progenitors can be severe or even life threatening, Examples include aplastic crisis in patients with sickle cell anaemia (Serjeant et al., 1981) and other chronic haemolytic anaemias, and prolonged anaemia in immunocompromised patients (Young, 1988; Weiland et al., 1989). B 19 infection in pregnant women has been indicated as a cause of spontaneous abortion, hydrops fetalis and intrauterine fetal death (Anderson et al., 1988; Maeda et al., 1988; van Elsacker-Niele et al., 1989) and the extent of its involvement in such complications is still the subject of a number of studies. There are few established tests for B19 IgG and IgM antibodies. Two of these, solid phase radioimmunoassay (RIA) developed by Cohen et al. (1983) and an ELISA developed by Schwarz et al. (1988) based on the antibody capture principle, require whole virus. This is not readily available for two reasons: firstly, it has not been possible to grow parvovirus B19 in a stable cell-line, although low-level propagation of virus has been demonstrated in erythroid progenitor cells from human bone marrow (Ozawa et al., 1986; Srivastava and Lu, 1988) and fetal liver (Yaegashi et al., 1989a). Secondly, isolation of virus from patient serum is hampered by the fact that symptoms occur when the patient is no longer in the viraemic phase (Anderson et al., 1985). This has led to a random screening of blood donors to obtain virus, of whom only about one in 40000 is viraemic at any time. In view of these difficulties and in order to obtain a renewable source of B19 viral antigens, recombinant DNA techniques have been employed. The two coatproteins, VP1 and VP2 (84 kDa and 58 kDa respectively) were expressed in large amounts in a baculovirus system in insect cells (Brown et al., 1990). Preliminary experiments with insect cells expressing VP1 showed that this protein gave a very distinct signal in indirect immunofluorescence. This makes it a suitable candidate for the development of an immunofluorescence assay (IFA) for the detection of B 19specific IgG and IgM antibodies in the sera of infected individuals. The development of this test and its applications are described.

Materials and Methods

Sera from patients with a clinical indication for a parvovirus B19 infection that had been tested for the presence of B19 antibodies in the RIA by Dr B.J. Cohen at the PHLS, Colindale, London, were tested in the IFA. Series of sera from seven patients were also tested for which the presence of B 19-specific antibodies had been con~~ed for one serum sample of each series in the RIA. Donor sera (89) were

55

PARVOVIRUS B19

2444 I

1

NSP 436

2449

4787 I

VP2 3125

1 47387

Fig. 1. Parvovirus I319 genome showing coding regions for the major protein species (non-structural protein [NSP], VP1 and VP2).

obtained from the Blood Bank of the University Hospital, Leiden, The Netherlands. Twelve sera positive for rheumatoid factors (RF) and 24 sera positive for rubella IgM were also used. Preparu~ion of reco~~~~~

Bl9 VP1

In order to produce parvovirus B19 VP1 coat protein in large quantities, B19 DNA coding for this protein was cloned into the transfer vector pAcYM1 (Matsuura et al., 1987) for expression in a baculovirus expression system. The cloning procedure. has recently been described in detail (Brown et al., 1990). Briefly, B19 virus DNA was isolated from the serum of a viraemic patient. The 2.5 Kb HindIII/.ScaI fragment, representing the coding region of VP1 (see Fig. l), was cloned into the BamHI site of pAcYM1 behind the strong polyhedrin promoter. The resulting construct was co-transfected with Autographica californica nuclear poiyhedrosis virus (AcNPV) DNA onto.monolayers of Spodopderafwgiperda (Sf) insect cells, during which process the re~ombin~t baculovirus AcB19VPlL was produced by homologous ~combination. After purification, this recombinant baculovirus was shown to contain the complete coding region for B 19 VP1 by Southern blot analysis. It expressed high levels of the 84 kDa B19 VP1 protein as shown by direct analysis on SDS-polyacrylamide gels and Western blot analysis with sera containing parvovirus B19-specific antibodies and with monoclonal antibodies 3A2 and 3B2 (donated by Dr J. Middeldorp, Organon). AcBlgVPlL-infected cells expressing VP1 were subsequently used for the development of an IFA. Indirect immuno$uorescence assay Monolayers of Sf cells infected with the ~~ornbin~t ba~ulovi~s at an MO1 of 1 were resuspended in TC-100 medium (G~CO~~)

AcBlOVPlL 4 days post

56

infection (p.i.) at which stage about half of the cells are expressing detectable amounts of VPl. The infected cells were mixed with uninfected cells at a ratio of 1: 1, so that about 25% of the cells should give a positive reaction upon testing with a serum containing B19-specific IgM or IgG. The cells were spotted onto glass slides at a concentration of 5000 cells/well in a 5 1.11volume (l&well slides) or 10 000 cells/well in a 10 ~1 volume (lo-well slides), and dried with an electric fan. All subsequent drying steps were done with a fan. The cells were fixed in methanol (15 min at 4°C followed by 30 min at -70°C) and slides were either stored at -80°C or used directly in the IFA, Directly before use, slides were thawed, dried briefly, washed for 10 min in distilled water and dried again. For all incubations, the minimum volume for incubation was 5 ~1 per well for l&well slides and 10 ~1 for IO-well slides.

Detection of B19 IgG antibodies A 30 min blocking step was carried out at 37°C in a moist chamber with veronalbuffered saline (VBS, pH 7.5) containing 10% guinea pig serum (GPS) to cover aspecific sites. The slides were dipped in phosphate-buffered saline (PBS) then in distilled water and dried. Sera to be tested for the presence of B19-specific IgG were diluted (as described in Results and the figure legends) in VBS containing 10% GPS. Incubation was performed for 1 h at 37°C in a moist chamber. Slides were dipped in PBS, washed a further 10 min in PBS at room temperature (RT), dipped in distilled water and dried. Bound antibody was detected with a 1:40 dilution of FITC-conjugated goat-antihuman IgG (Kallestad) in VBS, containing 10% GPS and a 1:5OOdilution of Evans Blue (Hoffman-La Roche). Incubation was for 30 min at 37°C in a moist chamber. The slides were dipped in PBS, washed four times in PBS for 5 min at RT, dipped in water and dried. Slides were mounted with Tris-glycerol buffer, pH 9.9 (1 g of T~s~ydroxymethyl]methyl~ine in 100 ml of 10% PBS and 80% glycerol) and examined with a fluorescence microscope.

Detection of BIB IgiV antibodies Blocking was not carried out before incubation with sera to be tested for the presence of B19-IgM. Instead, sera were pretreated with a goat-anti-human IgG preparation (GullSORB, Gull Laboratories). This removes any competing specific IgG, which can be present since the IgM and IgG response overlap (Anderson et al., 1985), and avoids false-positive IgM results due to the presence of RF. Sera were diluted 1: 10 in GullSORB, incubated 1.5min at RT and cen~fuged for 30 min at 3~ rpm at 4°C to pellet any formed complexes. The supematant was removed and diluted in VBS containing 10% GPS, as described in Results and the figure legends, and incubated for 2 h at 37°C in a moist chamber. The following steps in the procedure were essentially the same as for IgG except that FITC-conjugated goat-anti-human IgM was used for the antibody detection step.

Fig. 2. Immunofluorescence analysis of recombinant B 19 VP1 expressed in insect cells. Panel A shlows the I-action with a serum containing Bl9-specific IgG, panel B shows the reaction with B19-spe cific The arTOW W and panel C shows the reaction with a serum con~ning no B19-specific ~ti~dies. in panel C indicates an insect ceil expressing the VP1 protein.

58

Results

immu~o~uoresce~t signal seen with B19 IgG and IgM Fig. 2A shows the immunofluorescence seen when human serum containing 319~specific IgG antibodies is incubated with insect cells expressing 319 VP1 protein and detected with ~C-la~ll~ conjugate for human IgG. Fig. 2B shows the reaction with human serum containing Bl9-IgM upon detection with FITClabelled conjugate for human IgM. The characteristic signal seen with VP1 is due to the fo~ation of aggregates by this protein in the insect cells. The signal with IgG is stronger than that seen with IgM although the characteristic shapes formed in the IgM reaction can easily be distinguished from background fluorescence. Fig. 2C shows the signal seen when insect cells expressing VP1 are reacted with sera containing no B19-specific IgG or IgM antibodies. Cells expressing VP1 can be distinguished from cells not containing recombinant protein by their morphology but there is no specific fluorescence.

Evaluation of the IFA for detective of ZgG ~~t~~odies Thirty serum samples that had been tested in the RIA (Cohen et al., 1983) for the presence of Bl9-specific IgG were tested in a blind experiment to determine if sera scoring positive in the RIA also scored positive in the IFA. Seventeen of the sera tested were positive in the RIA and in the IFA while 13 were negative in both assays. This 100% agreement between the results with the RIA and the IFA was found at 1:SO and 1:500 dilutions of sera, The IFA titers of 19 sera with known RIA values for Bl94gG (ranging from <:I arbitrary units [a.u.] to >lOO a.u.) were determined by carrying out two-fold dilution series starting at a 1:32 dilution. The titer of the serum was the last dilution at which the serum still scored positive. The results are shown in Table 1 where a good correlation can be seen between the IFA titres and the RIA values throughout the range of the table. The titrations were performed in duplo (Table 1 shows the results for one titration assay) and for some sera the titre varied by one dilution step although all positive sera remained positive and all negative sera remained negative. Upon testing twenty sera negative for B 19 antibodies (as defined by the RIA) in the IFA for IgG at dilutions ranging from 1: 10 to 1: 100, it was found that in some sera, a dilution lower than 1:30 gave a high degree of background fluorescence. Therefore, a start dilution of greater than 1:30 was used when testing for IgG, A random group of blood donors were screened for the presence of B19-specific IgG. Of the 89 donors tested, 68 (76%) were positive at a 1:50 dilution. At a 1500 dilution, 62 were positive (70%) and at 1~5000, 29 were positive (33%).

Evaluution of the IFA for the detection ofIgh4 antibodies All sera to be tested for B19 IgM in the IFA were pretreated as described in Materials and Methods. After pretreatment, I319-specific IgG antibodies could no longer be detected in a I:10 dilution of sera which were previously positive

59

TABLE 1 Comparison of RIA and IFA for detection of B19 antibodies IgG RIA (a.u.)

IFA (the)

100 100 91

8 192 >I6 384 >16 384

IgM RIA (a.u.) >lOO ::

IFA (titre) 1280 1280 1280

z; 40 25 15.5 11 8.4 7.6 4.2 l.50

8 192 8 192 8 192 4 096 4 096 1024 512 512 ~32 512

;: 16 16 8.3 5.3 :4

2560 1280 640 320 40
4:1 3.4i.k

1:8 1.2
128 ~32 <32 ~32 ~32

2.3;‘; 1.k
:: 10
a.u. = arbitrary units;
for Bl9-IgG, The pretreatment also reduced aspecific background fluorescence as demonstrated by testing 20 negative sera for B 19-IgM at 1: 10 to 1: 100 dilutions. At a 1: 10 dilution it was possible to distinguish specific from background fluorescence. Twelve sera that were known to contain RF were tested in the IFA for IgM without pretreatment and all 12 scored positive at a 1: 10 and 1:40 dilution. After pretreatment, all 12 became negative for B 19-IgM at these dilutions, indicating that possible B19-IgG and RF involvement could effectively be abolished. In order to examine the possibility that rubella-specific IgM antibodies crossreact with the recombinant B19 VP1 antigen, 24 sera from patients with a recent rubella infection were tested for B19-IgM at 1: 10 and 1:50 dilutions. Only one of the sera was weakly positive at a 1: 10 dilution. Nineteen sera with known values for B 19-specific IgM in the RIA (ranging from ~1 a.u. to >lOO au.) were titrated in the IFA and the same correlation was found as in the IgG titration series, namely that a high RIA value corresponded to a high titre in the IPA. The results are shown in Table 1. The IFA titrations were performed in duplo (Table 1 shows the results for one titration) and, as was seen in the IgG titrations, some sera gave two different titres. The difference was never more than one dilution step and positive and negative sera remained so. When two values are shown for the RIA, the serum sample was tested twice.

10000

1

1000

100

10 Cl0

5

0

Weeks -1

-+--

10 after

15

onset

of

__A__ 2 __+-

5 -B-

6 -)_-

20

25

symptoms 3

+

4

7

Fig. 3. Course of B19 IgM response in seven patients. The IgM titres of the sera (3 samples for patients 1, 2, 4 and 5 and 2 samples for patients 3,6 and 7) were determined in the TFA by performing two-fold dilution series with 1:lO as start dilution.

Antibody response in the course of a B19 infection Sera taken at two or three time intervals from seven patients, which had a proven B 19 infection, were tested in the IFA for the presence of specific IgG and IgM antibodies. The titres obtained for IgM are shown in Fig. 3. The first serum sample showed a high IgM titre in all cases, ranging from 320-1920, while the second and third samples all showed a decrease in the IgM titre. In five patients (1, 2, 4, 6 and 7) where the last serum sample was taken more than 10 weeks after onset of symptoms, B 19-IgM was no longer detectable (
Discussion Current tests for anti-B19 antibody, such as RIA (Cohen et al., 1983) and ELISA (Brown et al., 1989) have had to rely on a source of B19 virus obtained by

61

screening blood donors. Recently, an ELISA has been described using viral antigen propagated in vitro in a primary culture of fetal liver erythroid cells (Yaegashi et al., 1989b). Although the in vitro propagated viral antigen compared favourably in sensitivity with in vivo B19 virus obtained from patient serum, the in vitro preparation contained lOOO-fold less virus per ml compared with virus obtained from patient serum. Also, the source of fetal liver cells is reliant on donors. The system described here utilises an expression system in which large amounts of recombinant B 19 antigen are produced in insect cells that are relatively easy to grow. With cells harvested from a 80 cm2 culture flask (containing 20 million cells), approximately 4000 IFA tests can be carried out. Furthermore, the viral antigen is presented in the insect cells in aggregate form which gives a very characteristic signal in the IFA when a reaction is carried out with a serum containing B19specific IgG or IgM antibodies. In order to evaluate the immunofluorescence assay described here, results were compared with the RIA, a well established test for B19 antibodies. First, it was established that it was possible to distinguish between sera positive for B 19-specific IgG and sera negative for B19-IgG. It was then shown that, of a random selection of 89 blood donor sera, 76% were positive for B19-IgG in the IFA which compares with previous findings in the blood donor age-group range (Cohen and Buckley, 1988). Titrations of sera in the IFA with a range of RIA values for B19-IgG and IgM showed a good correlation between the two test systems, as shown in Table 1. The titres obtained with the IFA were shown to be reproduceable by carrying out the ti~a~ons in duplo. In what is considered to be the equivocal range in the RIA, between about l-3 au., sera tested either positive or negative in the IFA. The diagnostic value of these findings is still to be determined. By determining the IFA titres of seven series of sera from patients that had had a recent B19 infection, it was possible to follow the course of the IgM antibody response which is typical for the second part of the response curve when the IgM response is decreasing (Anderson et al., 1985). This is shown in Fig. 3 where the IgM titre was high shortly after infection and decreased in subsequent sera. In the cases where a serum sample was available more than 10 weeks after onset of illness, the IgM titre was negative. These data show that the IFA-IgM test is a reliable instrument for the detection of recent B19 infections. The exclusion of the possibility of cross-reaction between rubella-specific IgM and the B19 antigen was impo~~t as both viruses may cause similar symptoms upon infection. The problem of cross-reactivity has been discussed by Kurtz and Anderson (1985) where it was shown that sera positive for rubella-specific IgM gave equivocal results when tested for B19-specific IgM and vice versa. The test described here is sensitive and can be made available on a large scale for routine diagnostic use. Furthermore, the recombinant antigen used in this system is being evaluated for use in other test systems, such as the ELISA, and for the production of monoclonal antibodies. The availability of such tools will facilitate extensive epidemiological studies resulting in a more complete picture of infections with this virus.

62

Acknowledgements We thank G.W. van Gemert for practical guidance and critical assessment of this manuscript and P.E. Veneer-de-Bondt for the discussion of clinical matters. Also, B.A.C. Dijkmans for the RF-positive sera.

References Anderson, L.J. and Hurwitz, E.S. (1988) Human parvovirus B19 and pregnancy. Clin. Perinatol. 15, 273-286. Anderson, M.J., Higgins, P.G., Davis, L.R., Williams, J.S., Jones, SE., Kidd, I.M., Pattison, J.R. and Tyrrell, A.J. (1985) Experimental parvovirus infection in humans. J. Infect. Dis. 152, 257-265. Brown, C.S., %&mans, M.M.M., Noteborn, M.H.M. and Weiland, H.T. (1990) Antigenic parvovirus B19 coat proteins VP1 and VP2 produced in large quantities in a baculovirus expression system. Virus Res. 15, 197-212. Brown, K.E., Buckley, M.M., Cohen, B.J. and Samuel, D. (1989) An amplified ELISA for the detection of parvovirus B19 IgM using monoclonal antibody to FITC. J. Viral. Methods 26, 189-198. Cohen, B.J., Mortimer, P.P. and Pereira, MS. (1983) Diagnostic assays with mon~lonal antibodies for the human serum parvovirus-like virus (SPLV). J. Hyg. 91, 113-130. Cohen, B.J. and Buckley, M.M. (1988) The prevalence of antibody to human parvovirus BlP in England and Wales. J. Med. Microbial. 25, 151-153. Cossart, Y.E., Field, A.M., Cant, B. and Widdows, D. (1975) Parvovirus-like particles in human sera. Lancet I, 72. Kurtz, J.B. and Anderson, M.J. (1985) Cross-reactions in rubella and parvovirns specific IgM tests. Lancet II, 1356. Maeda, H., Shimokawa, H., Satoh, S., Nakano, H. and Nunoue, T. (1988) Non-immunologic hydrops fetalis resulting from intrauterine human parvovirus BlP 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. Viral. 68, 1233-f 250. Ozawa, K., Kurtzman, G. and Young, N. (1986) Replication of the B19 parvovi~s in human bone marrow cell cultures. Science 233, 883-886. Schwarz, T.F., Roggendorf, M. and Deinhardt, F. (1988) Human parvovirus Bl9: ELISA and immunoblot assays. J. Virol. Methods 20, 155-168. Serjeant, G.R., Mason, K., Topley, J.M., Serjeant, B.E., Pattison, J.R., Jones, SE. and Mohamed, R. (1981) Outbreak of aplastic crises in sickle ceil anaemia associated with parvovirus-like agent. Lancet II, 595-598. Srivastava, A. and Lu, L. (1988) Replication of Bl9 parvovirus in highly enriched hematopoietic progenitor cells from normal human bone marrow. J. Viral. 62, 3059-3063. van Elsacker-Niele, A.M.W., Salimans, M.M.M., Weiland, H.T.. Vermey-Keers, C., Anderson, M.J. and Versteeg, J. (1989) Fetal pathology in human parvovirus BIB infection. Br. J. Obstet. Gynaecol. 96, 768-775. Weiland H.T., Salimans, M.M.M., Fibbe, WE., Kluin, P.M. and Cohen, B.J. (1989) prolonged parvovirus B19 infection with severe anemia in a bone manow transplant recipient. Br. I. Haematol. 71, 300. Yaegashi, N., Shiraishi, H., Takeshita, T., Nakamura, M., Yajima, S. and Sugamum, K. (1989a) Propagation of human parvovirus B19 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. (1989b) Enzyme-linked immunosorbent assay for IgG and IgM antibodies against human parvovirus B19: use of monoelonal antibodies and viral antigen propagated in vitro. J. Viral. Methods 26, 171-182. Young, N. (1988) Hematologic and hematopoietic consequences of B19 parvovirus infection. Semin. Hematol. 2.5(2), 159-172.