Filetype PDF | Posted on 14 Sep 2022 | 3 years ago
Authentication
digital resources
184x Filetype PDF File size 0.65 MB Source: www.ncbi.nlm.nih.gov
Clin. exp. Immunol. (1978) 31, 50-54.
Rubella serology by solid-phase radioimmunoassay:
its potential for screening programmes
CHIEKOSUGISHITA,* SIOBHANO'SHEA,JENNIFERM.BEST&J. E. BANATVALADepartmentof
Virology, St Thomas' Hospital and Medical School, London
(Received 18 May 1977)
SUMMARY
Sera from 269 adult females who had experienced naturally acquired or vaccine-induced infection
by rubella virus, including immune persons challenged intranasally with rubella vaccine (RA27/3)
as well as sera from 100 patients attending antenatal clinics, were tested for rubella antibodies by
the conventional haemagglutination inhibition tests (HAI), as well as a newly developed solid-
phase radioimmunoassay (RIA) for rubella immunoglobulin G (IgG) antibodies. Following both
naturally acquired and vaccine-induced infection, titres by RIA were approximately ten-fold
higher than by HAI. The RIA test was particularly useful in assessing the true immune status of
those with apparently low levels ofHAI antibody and has the added advantage that pre-treatment
of sera to remove inhibitors of haemagglutination and red cell agglutinins is unnecessary. The
RIA test has potential for the large-scale screening programmes which need to be carried out if
the Department of Health and Social Security recommendation, that women attending antenatal
and family planning clinics be screened for rubella antibodies, is to be effectively met.
INTRODUCTION
In the U.K., it is current policy to offer rubella vaccination to 11 to 14-year-old girls without prior
antibody screening (Department of Health and Social Security, 1970) and to women of child-bearing
age who, because of their occupation, are at particular risk of exposure to rubella, provided serological
studies show that they lack immunity (Department ofHealth and Social Security, 1972). More recently,
it has been recommended that women attending antenatal clinics should also be screened for rubella
antibodies and those found seronegative should be offered rubella vaccine in the immediate post-partum
period (Department ofHealth and Social Security, 1976). However, there is as yet no available data on a
nationwide basis on the number ofadults screened or the number vaccinated.
For screening purposes, rubella antibodies are generally detected by HAI tests. Although the HAI
test is a very reliable diagnostic test, for screening purposes it has limitations due to the difficulty of
removing lipoprotein inhibitors of haemagglutination present in variable amounts in all sera (Schmidt
&Lennette, 1970; Haukenes, Haram & Solberg, 1977; Pattison, 1977). Furthermore, red cell agglutinins
may sometimes be difficult to remove and interfere with the reading of tests, particularly at low serum
dilutions (Schmidt & Lennette, 1970; Mortimer, 1976). Because it is difficult to differentiate between
antibody and inhibitor at low serum dilutions (Banatvala & Best, 1973; Haukenes & Blom, 1975;
Mortimer, 1976), many laboratories screen sera at a dilution of 1: 16 or 1:20 and recommend vaccination
for those lacking antibody at this level, although a proportion ofthese patients may be immune.
This paper describes a solid-phase RIA technique, which is not influenced by the presence of serum
inhibitors, for detecting rubella IgG antibodies. Since it is more reliable than the HAI test for detecting
low levels of antibody and is a technique that can readily be automated, it may provide a suitable test
for large-scale screening programmes.
* Present address: Department ofMaternal and Child Health, School ofHealth Sciences, Faculty ofMedicine, University
of Tokyo, Hongo, Bunkyo-ku, Tokyo 113, Japan.
Correspondence: Professor J.E. Banatvala, Department of Virology, St Thomas' Hospital and Medical School, London
SE1 7EH. 0099-9104/78/0100-0050302.00 (C 1978 Blackwell Scientific Publications
50
RIA
Rubella IgG antibody by 51
MATERIALS AND METHODS
Sera tested by RIA. Seronegative samples. 134 sera shown to be negative by HAI (< 1:4) were obtained from women who
were nurses, medical students, physiotherapists or laboratory workers at St Thomas' Hospital, most of whom were later
vaccinated against rubella. In addition, serial serum samples from nine volunteers who failed to develop a satisfactory
HAI response following vaccination (< 1:8) were also included.
Seropositive samples. (a) 126 sera were obtained from 107 women who had experienced naturally acquired infection or
who had been vaccinated subcutaneously (s.c.) with one of four rubella vaccines (Cendehill, HPV77.DE-5, RA27/3 or
To-336). 100 of these sera had HAI titres of > 1:16 and twenty-six had titres of 1:4 or 1:8.
(b) Sera were also obtained from seven volunteers with either naturally acquired or vaccine-induced immunity whose
HAI titres ranged from < 1:4 to 1:64 and RIA titres from 1:80 to 1:2500. They were challenged with intranasally (i.n.)
administered RA27/3, using a ten-fold greater dose (104.69 TCID 50) than is usually given in trials with RA27/3 i.n. Sera
were obtained before challenge and 10, 30 and 90 days afterwards.
Antenatal screening programme. Sera from 100 unselected patients attending antenatal clinics at St Thomas' Hospital were
tested by RIA. The sera, which had already been screened by HAI at 1: 16 dilution, were tested by RIA at 1:20.
Haemagglutination inhibition test. A modification of the method described by Cooper et al. (1969) was used. Briefly, sera
were pre-treated with heparin and manganous chloride, the residual manganous chloride being precipitated with 0 025 M
sodium carbonate and absorbed with 1-day-old chick erythrocytes. 4 to 8 units ofrubella HA antigen were used in the test.
Radioimmunoassay reagents. Antigens. Rubella (Judith strain) and control antigens were prepared by previously described
methods (Al-Nakib, Best & Banatvala, 1975), but were not treated with Tween 80 and diethyl ether before use. The rubella
antigen used had an HA titre of 1:640-1: 1280.
125I-labelled anti-human IgG (125I-labelled AHG). A gamma globulin-rich fraction ofsheep antihuman IgG (Wellcome
Reagents) was labelled with 1251 (Al-Nakib etal., 1975) and diluted 1:10 in 50% inactivated foetal calfserum (FCS) (Flow
Laboratories Ltd.) in 0 01 M Tris-HCl buffer, pH 7-2, containing 0-1% sodium azide and stored at 4VC. Free iodine was
removed every 1 to 2 weeks by dialysis against 0 04 M sodium phosphate buffer, pH 7-4. Precipitation with 10% trichloracetic
acid indicated that at least 90% of the radioactivity was incorporated in the protein fraction. Specific activities of the 1251_
labelled AHG preparations ranged from 40-90 pCi/,pg ofprotein. The optimal dilution ofeach preparation of 125I-labelled
AHGwas determined by titration against a positive control serum. The dilution producing the maximum specific binding
was used.
Diluent. All reagents except the antigen were diluted in Tris-buffer (0-002 M Tris, 0 3 M NaCl, 0 003 M EDTA), pH
7-8, containing 10% inactivated FCS.
The radioimmunoassay technique. A modification of the technique described by Rosenthal, Hayashi & Notkins (1972)
for detecting vaccinia and Herpes simplex antibodies was used. 50pl volumes ofcontrol and rubella antigens, diluted 1:4 in
Tris buffer containing 0 01 M calcium chloride and 0 01 M magnesium chloride, were dispensed in polyvinyl microplates
(Titertek/Linbro, Flow Laboratories Ltd.) and allowed to dry for 3 hr under a current ofwarm air. 75 p1 volumes of filler
(50% inactivated FCS in Tris buffer) were then added to all wells and dried overnight at room temperature. Two- or four-
fold dilutions of each serum were distributed in 50 pl volumes on both control and rubella antigens in triplicate. Plates
were incubated for 1 hr at 370C in a humidified box and then washed ten times in tap water. A 50 pl volume of 125I-labelled
AHG at its optimal dilution was added to each well; plates were then incubated for 1 hr at 370C and washed as before.
Finally, plates were cut up and the individual wells counted for 100 sec in a Nuclear Enterprises 8312 gamma counter.
Theend-pointtitre was taken as the highest serum dilution to give a binding ratio (BR) >0 5 above the BR for the negative
serum, where the:
BR = mean counts for a serum dilution on rubella antigen
mean counts for a serum dilution on control antigen
The following five controls in triplicate were included in each test: a serum-free control, a negative serum at the lowest
serum dilution to be included in the test, a positive convalescent rubella serum at 1:500 and 1:5000, 1:5000 being the end-
point dilution for this serum, and empty wells for background counts. A BR ofapproximately 1 0 was obtained for serum-
free and negative serum controls. When the positive control serum at 1:5000 failed to give a BR > 1V5, the activity of the
labelled antiglobulin was considered to have deteriorated and a fresh label prepared. Labelled antiglobulins could be used
for 3 to 4 weeks after preparation.
RESULTS
There was a good correlation between the HAI and RIA tests (r = 0 899), although the RIA test was
considerably more sensitive (Fig. 1). Thus, the geometric mean titre (GMT) of the hundred sera with
HAI titres of > 1: 16 was 1:720 by RIA and 1:78 by HAI. However, when comparing individual sera,
antibody titres ranged from five to eighty times greater by RIA. Of 134 sera which were seronegative
by HAI, eight were found to be seropositive by RIA (Table 1). Six of the twenty-six sera with low
52 Chieko Sugishita et al.
1024
512
256 .. . .
128
I/
2 64 * /
a)
cc_
32
16 -0
80 160 320 640 1280 2560 5120 10240
Reciprocal RIA fitres
FIG. 1. Relationship between HAI and RIA titres of 100 serum specimens with HAI titres > 1: 16. Straight
line obtained using linear regression analysis.
HAI titres (1:4 or 1: 8) were seronegative by RIA, while sixteen (62%) of these sera by RIA had titres
of >1:80.
Anadditional test ofthe sensitivity ofRIA was obtained by determining antibody levels in the sera of
nine volunteers who apparently failed to develop a satisfactory HAI response following vaccination
(Table 2). Cases 1 to 4 were seronegative by both RIA and HAI, and case 5 produced very low levels of
antibodies by RIA 8 weeks after vaccination, but developed a satisfactory response when revaccinated.
Case 6 was probably already immune when vaccinated. Her HAI titres were tested on several occasions
but were difficult to interpret, titres being either <1:4 or 1:8. By RIA the titre was 1:80 both
pre- and post-vaccination. On repeated testing, undetectable or low antibody titres were obtained in
cases 7 to 9 after vaccination, but results of RIA tests indicated that these volunteers had developed a
TABLE 1. Comparison ofHAI and RIA antibody titres of 260 sera
Reciprocal RIA titres*
Reciprocal 80 160 320 640 1280 2560 5120 10240 Total
HAI titres < 20 20 40
< 4 126 3 4 1 134
4 2 1 1 4
8 4 4 2 4 2 5 1 22
16 4 2 5 11
32 2 7 3 5 4 21
64 2 12 4 4 22
128 12 12 2 1 27
256 2 5 3 4 14
512 1 1 1 3
1024 1 1 2
Total 132 3 4 10 8 39 18 29 8 8 1 260
* Results as number of sera in each class.
RIA 53
Rubella IgG antibody by
TABLE 2. HAI and RIA titres following primary vaccination and revaccination ofnine vaccinees who, when tested by HAI,
had apparently failed to seroconvert after primary vaccination
Pre-vaccination titre Post-vaccination titre Titre following revaccination
Case No. HAI RIA HAI RIA HAI RIA
1 < 1:4 < 1:10 < 1:4 < 1:10 1:32 1:2500
2 < 1:4 < 1:10 < 1:4 < 1:10 >1:16 n.t.
3 < 1:4 < 1:10 < 1:4 < 1:10 1:64 >1:640
4 < 1:4 < 1:10 < 1:4 < 1:10 1:64 n.t.
5 < 1:4 < 1:10 < 1:4 1:20 1:32 1:500
6 < 1:4(1:8)* 1:80 1:8(< 1:4) 1:80
7 < 1:4 < 1:10 < 1:4 (1:16) 1:80 1:16 1:160
8 < 1:4 < 1:10 1:8 (< 1:4) 1:160 1:16 1:320
9 < 1:4 < 1:10 < 1:8 (1:8) (1:64) 1:80 1:64 n.t.
n.t. = Not tested.
* Result ofrepeated HAI test in parentheses.
satisfactory immune response. On revaccination, no significant change in the HAI or RIA antibody
titres was observed when the sera were tested in parallel.
Table 3 compares HAI and RIA titres in naturally acquired and vaccine-induced infections. Although
similar trends were apparent, the GMTs show that following naturally acquired infection the results
obtained by RIA were approximately ten-fold higher and, following vaccination, five-fold higher than
by HAI. However, in both groups individual RIA titres were occasionally up to forty times greater than
HAI titres.
TABLE 3 Comparison ofantibody titres as tested by HAI and RIA following natural infection and 1 year after vaccination
with four different vaccines
HAI RIA
No.
Infection or vaccination tested Range Median GMT Range Median GMT
Natural infection
Onset 1-5 years before test 5 1:64-1:1024 1:256 1:294-1 1:1280-1:5120 1:1280 1:2228-6
Onset 10yearsbeforetest 10 1:32-1:256 1:128 1:90-5 1:320-1:5120 1:1280 1:970-0
Vaccine
Cendehill 15 1:16-1:128 1:64 1:61-1 1:80-1:1280 1:320 1:385 0
HPV77.DE-5 15 1:32-1:1024 1:128 1:161-3 1:320-1:2560 1:640 1:769-9
RA27/3 (s.c.) 15 1:32-1:256 1:128 1:92-6 1:80-1:1280 1:320 1:557-1
To-336 15 1:32-1:512 1:128 1:127-9 1:80-1:5120 1:320 1:735-2
Total vaccinees 60 1:16-1:1024 1:128 1:103-9 1:80-1:5120 1:320 1:590-2
Ofseven seropositive volunteers challenged with RA27/3 i.n., a significant rise (four-fold or greater)
in antibody titre was detected in three by the RIA but in only one ofthese by the HAI.
Results of screening 100 unselected sera from patients attending antenatal clinics confirmed the
greater sensitivity ofthe RIA test. Thus, ofnineteen sera which were seronegative by HAI, only fourteen
were seronegative by RIA. One serum sample was seropositive by HAI but seronegative by RIA.
Reproducible results were consistently obtained with the RIA test, since the same antibody titre
(i.e. 1: 5000) was obtained for the positive control serum when tested with seven different preparations
of 125I-labelled AHG and five batches ofantigen over a period of 14 months. Identical titres were also
obtained when eleven other sera, with titres ranging from 1: 80-1:5120, were re-tested 7 months later
using different preparations of 125I-labelled AHG and antigen.
no reviews yet
Please Login to review.
