solid phase anti-igm elisa for detection of rubella specific igm antibodies

Acta path. microbiol. scand. Sect. B, 89: 123-128. 1981

SOLID PHASE ANTI-IgM ELISA FOR DETECTION OF RUBELLA SPECIFIC IgM ANTIBODIES

MOGENS VEJTORP

Rubella Department, Statens Seruminstitut, Copenhagen. Denmark

Vejtorp, M. Solid phase anti-IgM ELISA for detection of rubella specific IgM antibodies. Acta path. rnicrobiol. scand. Sect. B. 89: 123-1 28. 198 I.

IgM in the test sample was bound to anti-lgM on a solid phase consisting of a polystyrene microtest plate and rubella specific IgM antibody subsequently detected by incubation with crude rubella virus antigen and anti-rubella conjugate. High levels of rubella specific IgG, which decreased the sensitivity of an indirect ELISA for rubella IgM, did not interfere in the anti-IgM ELISA, whereas non-rubella IgM interfered by a competition for the anti-IgM on the solid phase. False-positive results owing to IgM rheumatoid factor could be prevented by the addition of aggregated IgG 0.1 mg/ml to the serum diluent. The median duration of the antibody response detected by the solid phase anti-IgM assay was 60 days (range 24-1 40) in tests of serial serum samples from I 7 patients with rubella. The assay appears suitable for routine diagnosis of postnatal rubella and will probably be particularly valuable for diagnosis of prenatal rubella.

Key words: Enzyme-immunoassay; rheumatoid factor.

M. Vejtorp. Rubella Department, Statens Seruminstitut, Amager Boulevard 80. DK-2300 Copenhagen S. Denmark.

Received 24.xi.80 Accepted 19.xii.80 The solid phase anti-IgM technique was introdu-

ced by Flehmig for a radioimmunoassay (8) and later by Duermeyer & van der Veen for an enzyme- linked immunosorbent assay (ELISA) ( 5 , 6 ) for detection of IgM class antibodies to hepatitis A. The technique is also used for demonstration of IgM class antibodies to Sendai virus (71, Mycoplusmu pneumoniue (16) and to hepatitis B virus core antigen (9). The indirect ELISA for detection of rubella IgM antibodies utilizes rubella virus antigen physically bound to polystyrene in microtitration plates and an anti-human IgM alkaline phosphatase conjugate (20). The method is simple and specific, but the interference of IgM rheumatoid factor (RF) is a draw-back ( I 9). A further complication is the competitive inhibition by rubella specific IgG which was shown in preliminary experiments. In order to overcome these problems the anti-IgM ELISA was established for demonstration of rubella IgM antibodies. The assay is described together with the results of tests of 200 blood donors' sera and of 2 2 3

serial serum samples collected from 17 patients during a 10-year period after clinical and serologi- cally confirmed rubella infection. Results of studies on the interference of rubella specific IgG, non- rubella IgM and of IgM RF in the assay are reported.

MATERIALS AND METHODS

Serum Specimens A total of 223 sera from I7 patients with rubella were

collected during a period from 38 days before to I0 years after the rash. The clinical diagnosis of rubella was confirmed by the haemagglutination-inhibition test as previously described (20). Included in the study were also sera from 200 blood donors and 5 sera containing high levels of IgM RF. Immunoglobulin fractions of sera from 7 patients with recent or past rubella infections were tested after separation by rate zonal ultracentrifugation (22).

Rubella Virus Antigen The rubella virus antigen was prepared in monolayer

cultures of BHK cells essentially as previously described

I23

(18). However, the outgrowth medium contained 10% newborn calf serum and the maintenance medium was Glasgow minimum essential medium with glutamine 2 mmol/l. NaHCO3 0.027 mmol/l and 10% tryptose phosphate broth. The supernatant from the infected cells was concentrated by ultrafiltration and purified by rate zonal ultracentrifugation as described by Meurman et al. (1977). Sucrose was removed from the fractions containing virus antigen by gelfiltration through Sepha- dex G 25 M in a prepacked disposable PD 10 column (Pharmacia, code no. 17-0850-01), which was equilibra- ted with TRIS-buffer (15) and used according to the instructions of the manufacturer.

Anti-rubella Antibody Conjugate The IgG antibodies were isolated from a I0 ml pool of

human sera with high concentrations of rubella IgG by (NH&S04 precipitation and ion exchange chromato- graphy through DEAE-Sephadex A 50 (Pharmacia) ( 1 1). The eluate concentrated to 4 ml by (NH&S04 precipitation ( 1 1) was applied to the top of a column 2.5 x 90 cm containing Sephadex G 200 (Pharmacia). The buffer consisted of 0.01 mol/l TRIS pH 7.2 and 0.15 mol/l sodium chloride; the flow rate was 18 ml/h, and 4.5 ml fractions were collected. The protein concentration of the eluate was monitored by spectrophotometry at 280 nm (LKB Uvicord 111). The fractions eluted in the IgG peak were concentrated by vacuum dialysis in collodium bags to a final volume of 1.58 ml containing 34.5 g/ l protein as determined by refractometry. IgM was not detected in the concentrate by single radial immunodiffusion ( I 4). DAKO immunoglobulins A/S kindly conjugated 7 mg of the purified IgG to 28 mg horse radish peroxidase by a two step glutaraldehyde procedure (2). The conjugate in a volume of 2 ml was diluted I to 4 in phosphate buffered saline-Tween (PBS-Tween) (23) with Merthiolate" 0.2 g / l and bovine serum albumin 5 g / l (PBS-BSA) and stored at 4 OC.

Solid Phase Anti-lgM ELlSA The solid phase was 96 wells flat bottom polystyrene

microtest plates (NUNC Cat. no. 2-69787). The wells were coated with rabbit anti-human IgM (DAKO code no. 10-09 1 ) diluted 1 / I000 in a carbonate-bicarbonate buffer (231, by the addition of 200 pl to each well, storage overnight at 4 OC and washing four times with PBS-Tween on an automatic plate washer (Dynawasher AM 7 I ). The serum samples were diluted I /200 in PBS- BSA and 200 pl added to each of four wells. After incubation for 3 h on a water bath at 37 OC, the plates were washed. Crude rubella virus antigen, i.e. the supernatant of the infected BHK cells (6 parts) was mixed with normal serum albumin (human) 25% (Cutter Laboratories, order no. 684-7 1) (3 parts) and phosphate buffered saline pH 7.4 ( I part) and 100 pl added to two of the four wells. After overnight incubation at room temperature the wells were washed and 100 pl conjugate added after dilution 1 / 2000 in PBS-Tween (6 parts) with normal serum albumin (human) (4 parts). In experiments with addition of aggregated IgG to serum diluent, the conjugate was diluted I / 1600. The plates were incubated for 3 h at room temperature and washed. A 200 pI

volume of a solution of I0 mg orthophenylene diamine in 20 ml citrate buffer (citric acid, H f l 34.7 mmol/l, Na2HP04, 2 H2O 66.7 mmol/l, pH 5) and 7 .5 pl H202 was added to each well. After incubation for 35 min at room temperature the reaction was stopped by the addition of 50 pl H2SO4 3 mol/l. The results were read at 492 nm on a Titertek Multiscan" photometer and adjusted according to the mean of two measurements of a positive reference serum in order to reduce day-to-day variations. An E-value was calculated as the difference between the means of two tests with and two without the virus antigen.

Indirect ELISAs The indirect ELISAs for rubella IgG and IgM

antibodies were performed essentially as previously described ( 1 8, 20). However, the antigen was purified on sucrose density gradient and the control antigen was omitted in the rubella IgG ELISA. IgM RF was determined by an indirect ELISA as previously described (2 I ) .

Aggregated IgG IgG was isolated from a pool of human sera without

rubella antibody and aggregated by heating to 63 O C for 15 min (4).

Dithiothreitol Reduction IgM antibodies were in one experiment reduced before

the assay by dilution of the serum to 1 / I 0 in dithiothreitol 0.01 mol/l (17) and incubation for I h at 37 OC and overnight at 4 OC.

Total IgM Concentrations

single radial immunodiffusion ( 1 4). The concentration of total IgM was measured by

RESULTS

Rubella IgM in Blood Donors' Sera The results of tests of 200 blood donors' sera by

the solid phase anti-IgM ELISA appear from Fig. 1 . The median and mean E-values were 0.00, the standard deviation 0.022, and the values of 199 sera were below 0.08. The median absorbance in tests of the 200 sera without the virus antigen was 0.16 (range 0.08-0.33). An E-value of one serum of 0.17 was reduced to 0.05 after incubation of the sample diluted 1 / 100 in an equal volume of purified rubella virus antigen for 1 h at 37 OC and overnight at 4 O C

and the E-value was diminished to 0.04 after reduction of the IgM antibodies by dithiothreitol. Rubella specific IgM was not detected in this sample by the indirect ELISA, and the concentration of IgM RF was approximately 2 IU/ml.

Rubella IgM in Sera from Patients with Rubella The results of the solid phase anti-IgM assay of

149 serial serum samples obtained from 17 patients

I24

NO OF SERA

2 5

20

15

10

5

L i -0.04 -0.02 L 0.04 0.06 0.08 0.10 0.12 0.14 0.16 0.18

E -VALUE

Fig. I . Sera from 200 blood donors tested by the solid phase anti-IgM ELISA for rubella IgM.

with rubella between 38 days before and 200 days after the rash are shown on Fig. 2. Using a cut-off value of 0.10 between antibody negative and positive samples, the median duration of the rubella IgM antibody response was 60 days (range 24- 140). The E-values of 74 sera collected between 200 days and 10 years after the rash were all below 0.10. The coefficient of variation was 10% in 30 assays on 13 different days of a serum with a mean E-value of 0.28 (range 0.24-0.36).

E- VALUE

Interference of Rubella IgG and Non-rubella lgA4 A serum fraction with a low concentration of

rubella IgM was added to PBS-Tween and to 5 sera diluted 1/200 in PBS-Tween and tested in the indirect ELISA for rubella IgM. The admixture of intermediate or high levels of rubella IgC resulted in a decrease of the rubella IgM E-values and a falsely negative test (Table I). The addition of such sera to a fraction containing rubella IgM also resulted in a reduction of the E-values measured by the solid

0.2-

0. . . c :* 0. -c_

1 I , . , ~ ~ r , . , . . , , , , , , , , , , , , . .

0 50 100 150 200 DAYS AFTER RASH

Fig. 2. Rubella IgM antibodies determined by the solid phase anti-IgM ELISA of 149 sera from 17 patients with rubella. The continuous lines indicate upper and lower range of the E-values.

125

TABLE I . Indirect ELISA of a Sucrose Density Gradient Fraction Containing Rubella IgM after Addition of

Rubella IgG Negative and Positive Sera

Rubella IgG E-value of added serum Rubella IgM E-value

(buffer) 0.0 0.2 0.5 1 .o 1.1

0.12 0.15 0.1 1 0.07 0.03 0.05

Interference of IgM RF Five sera containing between 180 and 540 IU/ml

IgM RF were tested by the indirect and the solid phase anti-IgM ELISA for rubella IgM. The absorbances in the solid phase anti-IgM assays without virus antigen were 0.3 1-0.6 I , but decreased to 0.1 1 4 . 1 6 after addition of aggregated IgG 0. I mg/ml to the serum diluent.

The false-positive results of the solid phase anti- IgM ELISA (Table41 were prevented by the addition of aggregated IgG, whereas the results of tests of sera from two patients with recent rubella were unchanged.

phase anti-IgM technique (Table 2). However, the E-values were not correlated to the rubella IgG concentrations of the added Sera (p>o.Io), Spear- man’s rank correlation), whereas there was a

inverse correlation to the total I ~ M

TABLE 4. Indirect and Solid Phase Anti-IgM ELISA with and without Addition of Aggregated I& 0 . 1 mgfml to Serum Diluent of 5 Sera (n0.s 1-51 Containing High Levels of IgM RF and o f 2 Sera (N0.s 6-7) from Patients

with Recent Rubella concentrations (P<0.02).

TABLE 2. Solid Phase Anti-IgM ELISA of a Sucrose Density Gradient Fraction Containing Rubella IgM after

Addition of Rubella IgG Negative and Positive Sera

Rubella IgC E-value Rubella IgM Total IgM of added serum E-value concentration (mg/ml)

(buffer) 0.42 0.05 0.0 0.16 0.82 0.3 0.15 0.82 0.5 0.1 I I .54 0.8 0.22 0.54 I .o 0.24 0.52

The addition of IgG fractions containing different levels og rubella specific IgG to an IgM fraction resulted in only minor changes of the IgM E-values (Table 3).

TABLE 3. Solid Phase Anti-IgM ELISA of a Sucrose Density Gradient Fraction Containing Rubella IgM after Addition of Rubella Antibody Negative and Positive IgG

Fraction

Rubella IgG E-value of added fraction Rubella IgM E-value

(buffer) 0.0 0.3 0.6 0.8 1 . 1

0.50 0.54 0.55 0.5 I 0.55 0.57

Solid phase anti-IgM ELISA (E-value)

serum diluent

Indirect ELISA

(E-value) without IgG with IgG no.

1 0.56 0.23 0.00 2 0.37 0.19 0.00 3 0.33 0.36 0.01 4 0.22 0.33 0.02 5 0.19 0.14 0.04 6 0.80 0.49 0.5 1 7 0.14 0.30 0.28

DISCUSSION

The competition between specific IgG and IgM antibodies for anwen on the solid phase, which was demonstrated in an indirect ELISA for rubella IgM (Table I), is also observed in the indirect immunofluorescence technique (31, but not in an analogous radioimmunoassay (1 2). The IgG/IgM competition may reduce the duration of the IgM antibody response detected by indirect assays, and may particularly decrease the sensitivity for diagnosis of congenital rubella, as sera from prenatally infected infants in addition to small concentrations of rubella IgM may contain high levels of transplacentally transferred maternal IgG (3). The IgG antibodies can be removed by adsorption to protein-A Sepharose (13) or to anti-yFc (101, in a separation procedure before the assay. This may, however, be impractical for large-scale use.

126

The first step in the solid phase anti IgM ELISA is a binding of the IgM class antibodies in the sample to anti-IgM on the solid phase, whereas the IgG antibodies are removed during the ensuing washing procedure and the interference of these antibodies thereby avoided (Table 3).

A competition was detected in the solid phase anti-IgM ELISA between rubella specific IgM and non-rubella specific IgM for anti-IgM on the solid phase. ,This interference reduced the sensitivity in tests of sera containing high levels of non-rubella IgM (Table 2). However, specific IgM probably constitutes a large proportion of the total IgM in sera from patients with post-natal infections and particularly in sera from prenatally infected infants ( 1 , 16).

The interference of IgM RF in the anti-IgM assay (Table41 is probably caused by a binding of this IgM class anti-IgG to anti-IgM on the solid phase and a subsequent attachment of the human IgG conjugate to free binding sites on RF. Rubella IgG in the serum sample may also be attached to RF and afterwards bind virus antigen and conjugate. Aggregated human IgG added to the serum diluent as described by Gerlich & Luer (1979) probably saturated the free binding sites on RF and thus prevented the false positive results.

The cut-off E-value between the rubella IgM antibody negative and positive sera was set at 0.10, which corresponded approximately to the mean plus 5 standard deviations in tests of 200 blood donors sera (Fig. 1 ). The applicability of this limit has to be evaluated further. However, in the present material a higher threshold would decrease the sensitivity and cause an occasionally very short duration of the antibody response. A lower limit, on the other hand, would in some instances increase the duration of the antibody response for more than 200 days (Fig. 2) and thus reduce the diagnostic value of the test.

The cut-off value was exceeded by one of 200 blood donors' sera. The positive IgM result has, however, presumably been caused by a low concentration of rubella specific IgM, as the E-value of this serum decreased below the threshold after absorption by purified rubella virus antigen and after reduction by dithiothreitol. The median duration of the IgM antibody response in samples from patients with rubella was 60 days (range 24- 1401, and thus slightly longer than the persistence for 50 days (range 17-1 96) detected by the indirect ELISA of the same sera (20).

Binding of the conjugate directly to anti-IgM on the solid phase probably explains the relatively high >>background(( values in tests of sera without rubella IgM and in tests without the virus antigen. This was

compensated for by a subtraction of the absorbances measured in tests without virus antigen from that obtained in parallel tests with the rubella virus antigen. Improvement in the quality of anti-IgM antibody and conjugate may reduce the ))back- groundcc absorbance.

The major advantages of the new solid phase anti-IgM ELISA for detection of rubella IgM is the lack of interference of specific IgG antibodies and the easy method for prevention of the interference of IgM RF. The test, which is suitable as a routine method, was slightly more sensitive than the indirect ELISA for diagnosis of postnatal rubella and will probably be particularly sensitive for the diagnosis of prenatal rubella.

ADDENDUM

Upon completion of the study following modifica- tions have been found economical. Anti-IgM for coating of the solid phase can be used diluted to 1 / 1 2 500 in phosphate buffered saline, pH 7.4 and albumin in the diluent for conjugate may be replaced by 2 .5% (v/v) human serum without rubella IgG.

Agnete Ingild, DAKO Immunoglobulins A /S prepared the conjugates for the solid phase anti-IgM ELISA. Bendr Mama, Deparrmenr of Biophysics, Srarens Seruminstirur, performed the rate zonal ultracentrifugation, the gel filtrations and the determination of IgM concentration. The excellent technical assistance of Miss Karin Hansen and Miss Maria Borgman is gratefully acknowledged. E. Fanee. general practitioner, collected the serial serum samples from his patients with rubella.

1 .

2.

3.

4.

5.

REFERENCES

Aljbrd Jr . , C . A , : Rubella. In: Remingron, J . S. & Klein, J . 0. (Eds.): Infectious Diseases of the fetus and newborn infant. W. B. Saunders Company, Philadelphia 1976, p. 86. Avrameas. S. & Ternynck, T.: Peroxidase labeled antibody and Fab conjugates with enhanced intracel- Mar penetration. Immunochemistry 8: 1175-1 179, 1971. Cradock- Watson, J. E., Ridehalgh, K . S . & Chant- ler, s.: Specific immunoglobulins in infants with the congenital rubella syndrome. J. Hyg. (Camb.) 76: 109-123, 1976. Dissanayake, S., Hay, F. C . & Roitr, I . M.: The binding constants of IgM rheumatoid factors and their univalent fragments for native and aggregated human IgC. Immunology 32: 309-31 8, 1977. Duermeyer. W. & van der Veen, J.: Specific detection of IgM-antibodies by ELISA, applied in hepatitis A. Lancet 2, 684-685, 1978.

127

6 . Duermeyer, W . , Wielaard, F. & van der Veen, J . : A new principle for the detection of specific IgM antibodies applied in an ELISA for hepatitis A. J. Med. Virol. 4: 25-32, 1979.

7. E d 3 H. C. J. , Gerlich, W . & Koszinowski, I / . H . : Detection of antibodies to Sendai virus by enzyme- linked immunosorbent assay (ELISA). J. Immunol. Methods 28: 163-1 76, 1979.

8. Flehmig, B.: Laboratoriumsdiagnose der Hepatitis A-Infection. Bundesgesundhbl. 17: 277-283, 1978.

9 . Gerlich. W . H . & Liier, W . : Selective detection of IgM-antibody against core antigen of the hepatitis B virus by a modified enzyme immuno assay. J. Med. Virol. 4: 227-238, 1979.

10. Cispen, R . , Nagel, J . , Brand-Saathof; B . & de Graaf; S.: Immunofluorescence test for IgM rubella antibodies in whole serum after absorption with anti-yFc. Clin. exp. Immunol. 22: 431-437, 1975.

1 1 . Harboe, N . & Ingild, A , : Immunization, isolation of immunoglobulins, estimation of antibody titre. In: Axelsen, N . H.. Krell. J . & Weeke, 8. (Eds.): A manual of quantitative immunoelectrophoresis. Blackwell Scientific Publications, Oxford 1973. pp.

12. Kangro, H . 0.. Pattison, J . R . & Heath, R . B. : The detection of rubella-specific IgM antibodies by radioimmunoassay. Br. J. exp. Path. 59: 577-583, 1978.

13. Leinikki, P. O., Shekarchi, I . , Dorsett, P . & Sever. J . L.: Determination of virus-specific IgM antibodies by using ELISA: elimination of false-positive results with protein h-Sepharose absorption and subsequent IgM antibody assay. J. Lab. Clin. Med. 92:

14. Mancini, G. . Carbonara. A . 0. & Heremans. J . F.: Immunochemical quantitation of antigens by single

I6 1-1 64.

849-857, 1978.

radial immunodiffusion. Immunochemistry 2: 235- 254, 1965.

IS. Meurman, 0. H., Vl!ianen# M. K . & Granfors, K.: Solid-phase radioimmunoassay of rubella virus immunoglobulin M antibodies: Comparison with sucrose density gradient centrifugation test. .I. Clin. Microbiol. 5: 257-262, 1977.

16. Price, P. C.: Direct radioimmunoassay for the detection of IgM antibodies against Mycoplasma pneumoniae. J. Immunol. Methods 32: 261-273, 1980.

17. Schrohenloher. R . E.: Copper-catalysed reoxidation of human monoclonal IgM. Scand. J. Immunol. 8:

18. Vejtorp, M.: Enzyme-linked immunosorbent assay for determination of rubella IgG antibodies. Acta path. microbiol. scand. Sect. B, 86: 387-392, 1978.

19. Vejtorp, M.: The interference of IgM rheumatoid factor in enzyme-linked immunosorbent assays of rubella IgM and IgG antibodies. J. Virol. Methods

20. Vejtorp, M . , Fanee, E. & Leerhoy, J.: Diagnosis of postnatal rubella by the enzyme-linked immunosorbent assay for rubella IgM and IgG antibodies. Acta path. rnicrobiol. scand. Sect. B, 87: 155-1 60, 1979.

21. Vejtorp, M., Heier-Madsen, M. & Halberg, P.: Enzyme-linked immunosorbent assay for determination of IgM rheumatoid factor. Scand. J. Rheuma-

22. Vesikari. T. & Vaheri, A . : Rubella: a method for rapid diagnosis of a recent infection by demonstration of the IgM antibodies. Brit. Med. J. 1: 22 1-223, 1968.

23. Voller, A . . Bidwell, D. E. & Bartlett. A . : The enzyme linked immunosorbent assay (ELISA). Dynatech Europe, Guernsey, G. B., 1979.

443-449, 1978.

I: 1-9, 1980.

tol. 8: 65-70, 1979.

128

solid phase anti-igm elisa for detection of rubella specific igm antibodies

Documents