Journal of Virological Methods, 34 (1991) 37-43 0 1991 Elsevier Science Publishers B.V. / All rights reserved / 0166-0934/91/$03.50 kDONIS 0166093491004120

VIRMET 01207

31

Kinetics of rubella-specific IgM antibody in postnatal rubella infection

response

Patricia Cordoba, Silvia Nates, James Mahony and Marta Zapata Immunology Laboratory, Virology Institute “Dr. Jose M. Vanella” Medicine Faculty,

Cordoba National University, Cordoba, Argentina

(Accepted 8 April 1991)

The serological diagnosis of primary postnatal rubella infection is based on detection of rubella-virus-specific IgM antibody or a four-fold rise in rubella- specific IgG antibody. Although there are several different methods of enzyme immunoassays that are commercially available, the cost benefit evaluation makes them impractical for use in developing countries. For this reason, we have standardized the measurement of rubella IgM antibody by HA1 following serum fractionation by ion-exchange chromatography. The sera samples obtained from pregnant women infected with rubella virus at different times during gestation were fractionated and tested by HAI. Seven out of nine sera collected within the first two days after onset of rash showed detectable levels of rubella IgM antibody. All 57 sera collected between 3 and 30 days after the onset of rash contained rubella IgM antibody. After 30 days, only 1 of 5, or 20%, of sera contained IgM antibody. The HA1 testing method was rapid and specific and the cost was not prohibitive. HAI-IgM testing could be used to diagnose primary rubella infections in developing countries where expensive EIAs are unaffordable.

Rubella IgM; Kinetic; Primary infection

Correspondence to: Patricia Cordoba, Immunology Laboratory, Virology Institute “Dr. Jose M. Vanella” Medicine Faculty, Ciudad Universitaria - Estafeta 32, 5000 Cordoba, Argentina.

38

Introduction

Serological diagnosis of primary postnatal rubella is based on the detection of rubella-specific IgM antibody or the presence of a four-fold rise in IgG titre for paired acute- and convalescent-phase sera. Measurement of rubella-specific IgM antibody obviates the need to collect a convalescent-phase serum which is often received in the laboratory too late to assist with the diagnosis.

Rubella virus reinfection can be distinguished from primary infection by the presence of very low levels of rubella-virus-specific IgM antibody in sera, which is only detected by highly sensitive methods. Rubella-virus-specific IgM antibody can be measured following serum fractionation by physicochemical methods such as sucrose gradient centrifugation or by measuring IgM antibody in fractionated sera (Zapata and Marquez, 1985). The latter method is more rapid but requires the removal of rheumatoid factor and IgG-class antibody, which may cause false-positive results.

Serum IgM fractionation can be done by sucrose-density-gradient centrifugation (Best et al., 1969), gel filtration (Morgan-Capner et al., 1980) or ion-exchange chromatography (Nates et al., 1987). Rubella-specific IgM antibodies can be measured by testing the IgM immunoglobulin fraction by haemagglutination inhibition (HAI) or an immunofluorescent test.

In an effort to develop an inexpensive serological technique for rapid diagnosis of rubella infection in developing countries, we used a rubella virus HAI test following serum fractionation by ion-exchange chromatography (Nates et al., 1987). We report that rubella IgM antibody can be measured using this method and that it appears as early as three days after the onset of rash and persists for up to 35 days after the onset of symptoms.

Materials and Methods

Specimens

164 sera received in the laboratory for rubella diagnosis and stored at 4°C were used in the study. 71 sera were obtained from pregnant women at various times during the course of infection with rubella virus. All had clinical histories compatible with rubella virus infection and had a four-fold antibody rise as diagnosed by HAI. 93 sera obtained from healthy women who were seropositive for rubella antibody were used as controls. They had no recent exanthematic illness, nor contact with rubella patients.

HAI

The haemagglutination inhibition test used routinely in our laboratory was carried out according to Palmer et al. (1970). Rubella haemagglutinin antigen was obtained from Vero cells infected with the Gilschrist rubella virus strain by

treatment with Tween 80 and ether (Norrby, 1962). Serum samples were treated with MnCl*-heparin to eliminate non-specific inhibitors (Plotkin et al., 1968) and the agglutinins were adsorbed with a 50% suspension of baby chicken erythrocytes in modified Alsever solution. The indicator system was a 0.25% suspension of one-day-old chick erythrocytes in HSAG diluent (1% Hepes saline, 3% bovine albumin and 2.5% gelatin). Titres < l/8 were considered negative for rubella antibody.

Separation of serum IgM and IgG was carried out by ion-exchange chromatography according to the method described by Johnsson and Libby (1980). This method was adapted by Nates et al. (1987) for detection of IgM antibodies to rubella virus. 250 ,ul of serum were fractionated in small columns containing quaternary aminoethyl (QAE) and Sephadex A-50, using 0.0077 M EDTA, 0.073 M acetic acid, pH 7, as buffer wash solution. The IgM fractions were obtained by elution with 0.26 M acetic acid, 0.34 M NaCl, pH 4.2. Specific antibodies to rubella virus were assayed by HAI. 70% of the total IgM was recovered.

Results and Discussion

Analyses of sera obtained from rubella patients and those from healthy pregnant women are shown in Table 1. The sera are grouped according to time since the onset of rash for 0-2,3-5,6-lO, ll-15,16-20,21-25,26-30 and 31-35 days. Results are expressed as the percentage of sera positive for rubella IgM antibody for each group. During the first two days after the onset of exanthem levels of IgM antibodies were detected in 66.6% of the serum samples; between days 3 and 30 it was detected in 100% of the sera assayed. In contrast, only 1 of 5, or 20%, of sera collected 30 days after the onset of rash had rubella IgM antibody. All control sera were negative for rubella IgM antibody.

The distribution of rubella antibody titres in total sera (squares) and IgM

TABLE 1

Rubella specific IgM (percentage of positives) measured by HAI after ion-exchange chromato- graphy in 164 sera from 71 primary rubella infections and 93 sera from immune pregnant women

Days after onset Number of sera IgM fraction Positive of rash

Positive (> l/8) Negative W)

O-2 3-5 5-10

11-15 16-20 21-25 26-30 31-35 Total Control sera

7

:: 11 9

2 1

65 0

2 66.6 - - :: _ 100 - 100 - 100 4 100

20

94 0

40

t

ml El

t o&t davs after onset

Fig. 1. Distribution of rubella HA1 antlboay Utres in rotal serum (squares) and IgM fractions (circles) obtained by ion-exchange chromatography. Sera were collected from pregnant women with confirmed

primary rubella, at various days after onset of the rash.

fractions (circles) is shown in Fig. 1. The number of sera with various HA1 titres is indicated by the number of symbols inside each square or circle. Rubella-specific IgM antibody was first detected two days after the onset of rash. HAT antibody titres of total unfractionated sera are indicated by squares and were always greater than IgM antibody titres. From days 3 to 22, the mean IgM antibody titre was l/32 and for unfractionated serum l/256. Between days 22 and 35, the IgM titres decreased rapidly while in total serum it remained high. These results indicate that rubella-specific IgM antibody was present in 57 of 57 sera collected between 3 and 30 days after the onset of rash as measured by HA1 following serum fractionation by ion-exchange chromatography and that this test procedure can be used within this time period to diagnose successfully all rubella infections.

The remaining problems of rubella diagnosis during pregnancy in developing countries today are those of acute rubella at various stages of pregnancy, rubella reinfection and congenital rubella. The availability of serological tests to measure rubella-specific IgM antibody will help to ensure the correct diagnosis.

We detected rubella-virus-specific IgM antibodies by HA1 after ion- exchange chromatography. The method evaluated in this study is 100% accurate for the diagnosis of primary rubella between 3 and 30 days post- exanthem since 57 of 57 sera tested contained rubella-specific IgM antibody. In contrast, sera collected during the first two days or after 30 days after the onset of rash contained specific IgM antibody in only 66 and 20% of samples respectively. All of 93 sera collected from healthy pregnant women, randomly

41

selected, independent of their HA1 antibody titres in total serum, were all negative for rubella IgM antibody.

The persistence of rubella-specific IgM antibody varies with the methodol- ogy used. Others have shown that immunofluorescence or HA1 following sucrose gradient centrifugation detects IgM antibody for up to four to five weeks post exanthem (Hornsleth et al., 1975). Antibody capture enzyme immunoassays detected IgM antibody up to eight weeks after onset of rash, while indirect enzyme immunoassays detected IgM antibody over shorter time intervals (Enders and Knotek, 1985).

Other investigators have used commercial tests, (Enzygnost, Behring; Rubazyme, Abbott; Rubenz M, Northumbria/Biosigma), to follow rubella- specific IgM antibody response after infection. These were compared with the conventional HAI test following sucrose density gradient ultracentrifugation. The results obtained by all methods showed that during the first three days after onset of rash, low concentrations of rubella IgM-specific antibodies were detected with between 27 and 97% of sera positive for rubella IgM antibody. Between days 4 and 15, high IgM antibody concentrations were found with nearly all the methods used, while between days 16 and 35 high concentrations were still found in 83-100% of sera tested. From days 36 to 70, the percentage of serum samples with detectable rubella IgM antibody declined and they could be detected between days 71 and 180 in only 9-76% of specimens. After day 180, detection of IgM antibodies was rarely found. This is not surprising since the ELISA for rubella IgM antibody has been reported to be more sensitive than the HA1 test (Bonfanti et al., 1985; Chernesky et al., 1984). Linde (1985) using the Rubazyme M test found that sera drawn from zero to three days after the onset of the rash contained low amounts of rubella virus specific. IgM antibody. From days 4 to 14 all specimens were IgM antibody positive and at day 60 the ELISA absorbance for rubella virus-specific IgM had fallen considerably. Even when other workers have used different ELISA test, including anti-IgM capture ELISA (Enders, 1985; Echevarria et al., 1985), which were reported to be more sensitive than HAI, only the anti-IgM capture ELISA Rubenz M detected rubella-specific antibodies in 100% of the sera tested between days 4 and 35 after onset of exanthem.

Our data correlate well with the ELISA results, but the slope of the anti-IgM capture ELISA is lower than that allowed with the ion-exchange fractionation technique, indicating that the capture ELISA test is more sensitive and can detect rubella IgM antibody longer after the onset of symptoms.

Our method requires the separation of IgM and IgG providing an IgM fraction “free” of IgG. The low-concentration IgG detection in IgM fractions (0.5 mg %) may correspond to IgG4, that has affinity to the QAE-Sephadex A- 50. The analysis of subclass distribution of rubella virus-specific immunoglo- bulins G (Linde, 1985) showed that IgG4 is not present in primary rubella infections, and so cannot give false-positive results. The IgM fraction could be contaminated by IgA antibody but the studies of Grangeot-Keros et al. (1988) and Ponzi et al. (1985) indicated that IgA antibodies follow the same time

42

course as IgM antibody following acute rubella, suggesting that rubella IgA antibody could be a useful index of recent infection.

The HA1 test was the first widely used technique for measuring rubella antibody and is the standard against which all other rubella immunity screening and diagnostic tests performed in Argentina are measured. The HA1 is time consuming, highly technical and requires pretreatment of serum. Nevertheless, HA1 serology has correlated with both immunity and acute infection in the case of seroconversion or four-fold rise in titres. The presence of rubella HA1 antibody correlates with immunity and protection from primary infection, and no other test has been shown to correlate with the immune status.

As previously indicated, the techniques that require previous separation of IgM from IgG are now being replaced by various enzyme immunoassays, including capture IgM assays that are commercially available. These tests, however, are too expensive for developing countries and therefore not used widely. The HA1 ion-exchange fractionation procedure described in this paper is inexpensive, easy to perform and has the advantage of HA1 standardization, making it an ideal test for diagnosing primary rubella infection in developing countries.

Acknowledgement

We thank Angelica Hoeller for valuable technical assistance.

References

Best, J.M., Banatvala, J.E. and Watson, D. (1969) Serum IgM and IgG responses in postnatally acquired rubella. Lancet ii, 65568.

Bbnfanti, C., Meurinan, 0. and Halonen, P. (1985) Detection of specific immunoglobulin M antibody to rubella virus by use of an enzyme-labeled antigen. J. Clin. Microbial. 21, 963-968.

Chernesky, M.A., Wyman, L., Mahony, J., Castriciano, S., Safford, I.W. and Metzel, P.S. (1984) Clinical evaluation of the sensitivity and specificity of a commercial enzyme immunoassay for detection of rubella virus specific immunoglobulin M. J. Clin. Microbial. 20, 40@404.

Echevarria, J.M., Sainz, C., De Ory, F. and Najera, R. (1985) Evaluation of commercial methods of enzyme immunoassay (EIA) for the measurement of rubella-specific IgM. J. Virol. Methods 11, 177-187.

Enders, G. (1985) Serological test combinations for safe detection of rubella infections. Rev. Infect. Dis. Vol. 7, Suppl. 1, 113-121.

Enders, G. and Knotek, F. (1985) Comparison of the performance and reproducibility of various serological methods and diagnostic kits for the detection of rubella antibodies. J. Virol. Methods 11, l-14.

Grangeot-Keros, L., Pillot, J., Daffos, F. and Forestier, F. (1988) Prenatal and postnatal production of IgM and IgA antibodies to rubella virus studied by antibody capture immunoassay. J. Infect. Dis. 158, No. 1, 138-143.

Hornsleth, A., Leerhoy, J., Grauballe, P. and Spangaard, H. (1975) Persistence of rubella virus- specific immunoglobulin A antibodies: investigation of successive serum samples with lowered immunoglobulin G concentration. Infect. Immun. 11, 804-808.

43

Johnson, R.B., Jr. and Libby, R. (1980) Separation of immunoglobulin M (IgM) essentially free of IgG from serum for use in systems requiring assay of IgM-type antibodies without interference from rheumatoid factor. J. Clin. Microbial. 12, 451454.

Linde, G.A. (1985) Subclass distribution of rubella virus-specific immunoglobulin G. J. Clin. Microbial. 21, 117-121.

Morgan-Capner, P., Davies, E. and Paltison, J.R. (1980) Rubella-specific IgM detection using Sephacryl S-300 gel filtration. J. Clin. Pathol. 32, 1082-1085.

Nates, S., Ferreira Centeno, A., Marquez, A. and Zapata, M. (1987) Utilization de un metodo rapid0 para determinar infection primaria por virus rubeola. Rev. Arg. Microbial. 19, 153-159.

Norrby, E. (1962) Haemagglutination by measles virus 4 - a simple procedure for production of high potency antigen for haemagglutination-inhibition (III) test. Proc. Sot. Exp. Biol. Med. 111, 814818.

Palmer, D.F., Herrmann, K.L., Lincoln, R.E., Hearn, M.V. and Fuller, J.M. (Eds) (1970) A procedural guide to the performance of the standardized rubella haemagglutination-inhibition test. Center for Disease Control Immunity, series No. 2, Atlanta, GA, U.S.A.

Plotkin, S.A., Bechtel, D.J. and Sedwick, W.D. (1968) A simple method for removal of rubella haemagglutination inhibitors from serum adaptable to finger-tip-blood. Am. J. Epidemiol. 88, 301-304.

Ponzi, A.N., Merlino, C., Angeretti, A. and Penna, R. (1985) Virus-specific polymeric immunoglobulin A antibodies in serum from patients with rubella, measles, varicella and herpes zoster virus infections. J. Clin. Microbial. 22, 505-509.

Zapata, M. and Marquez, A. (1985) Metodos immunologicos de detection de antigen0 e IgM especitica en el diagnostic0 virologico rapido. Adelantos en Microbiologia y Enfermedades Infecciosas 4, 84-108.