- Email: [email protected]
Confirmation of specificity by neutralisation in immunoradiometric assay for hepatitis B surface antigen
Journal of Viroiogikalklethods, @ Elsevier/North-Holland
CONFIRMATION
1 (1980)
Biomedical
113-
113
116
Press
OF SPECIFICITY
BY NEUTRALISATION
IN IMMUNORADIO-
METRIC ASSAY FOR HEPATITIS B SURFACE ANTIGEN
C.H. CAMERON Department (Accepted
and MOYA BRIGGS
ofMicrobiology, 9 January,
The Middlesex Hospital Medical School, London
WI, England
1980)
Two types of neutralisation tion of positive results obtained
or blocking test using human anti-HB, are described by immunoradiometric screening for HBsAg.
for the confirma-
INTRODUCTION Since the earliest days of HB,Ag detection by immunoradiometric assay (RIA), it has been apparent that confirmation of the specificity of a positive reaction, even if repeated
and found consistently positive, is essential before the sample can be reported as HB,Agpositive. Incorrect positive reports give rise not only to false information in epidemiological studies, but also to a variety of unnecessary difficulties for persons wrongfully designated as hepatitis B positive. In the case of high-titre HB,Ag-positive specimens found by RIA confirmation can be provided by the use of other less sensitive tests. In the case of low-titre specimens, detectable only by RIA, confirmation must be made using an RIA method. We describe 2-stage and 3-stage neutralisation or blocking tests for use in different circumstances. These techniques have been for the most part employed with our own RIA system, but have also been used with various commercial RIA kits, and are in principle applicable to any similar solid-phase sandwich immunoradiometric MATERIALS
assay.
AND METHODS
RIA reagents These were described in outline
by Heathcote
et al. (1974).
A more detailed descrip-
tion is in preparation. Polystyrene tubes were coated with globulins from horses immunised with purified HB,Ag, the globulins having been absorbed against normal human serum and plasma proteins coupled to agarose beads. Labelled antibody was made from a similar globulin preparation from immunised rabbits by absorbing specific anti-HB, on to purified HB,Ag coupled to agarose beads, the antibody repurified.
was then eluted, iodinated
with “‘1, and finally
114
RIA test for HBfig Antibody-coated Overnight
tubes were inoculated
incubation
at room temperature
with 100 ~1 volumes of test or control serum. was followed by washing and incubation
for
1.5-2 h at room temperature with 100 ~1 volumes of labelled antibody, and finally by washing and counting for radioactivity. An alternative, more rapid, procedure used two incubation stages of 1.5 h each at 45°C. Neutralising serum The human antiserum used had a high titre of anti-HB, (200 I.U./ml) and was from a donor with no history of having received blood transfusions or blood products. The negative control serum was made from a pool of human sera, all found negative by RIA for HB,Ag, anti-HB, and anti-HB, . These sera were used at a dilution ing 0 .l % sodium azide.
of 1 : 2 in 0.02 M Tris/HCl buffer, pH 7.6, contain-
2-Stage neutralisation test Pairs of coated tubes were inoculated
with 20 ~1 volumes of neutrahsing
serum, followed by 100 ~1 volumes of the test serum. The procedure
or control
then followed that
of the standard test. 3-Stage neutralisation test Pairs of coated tubes were inoculated with 100 ~1 volumes of test serum and given the usual first incubation. The tubes were then washed and inoculated with 120 ~1 volumes of neutralising or control serum. After the additional incubation stage of 1 h at room temperature the tubes were washed and the labelled antibody added, as in the final stage of the standard
test. The radioactivity
remaining
after washing in the neutralised
and
control tubes was compared. RESULTS
The 2-stage and 3-stage neutralisation tests were performed on a series of dilutions of an ad antigen-positive serum and an ay antigen-positive serum, both of high titre, using the human neutralising antiserum (Table 1). The composite data are shown graphically (Fig. 1) to indicate approximately the degree of neutralisation to be expected with this antiserum in the presence of varying amounts of HB,Ag in either the 2-stage or the 3-stage procedure. DISCUSSION It will
amounts
be seen that in the 2-stage test the antiserum in unable to neutralise large of HB,Ag, whereas in the 3-stage test the amount of antigen remaining in the
115
TABLE
1
Results
of 2- and 3-stage
concentrations
neutralisation
of HBsAg subtypes
tests using a human
Control/neutralised
HB,&
anti-HB,
serum (100 I.U./ml)
and different
ad and ay ratio
(ng/ml) HBsAg subtype
ad
HB,Ag
subtype
ay -
2-stage
3-stage
test
test
2-stage
test
3-stage
250,000
1
3.2
1
1.9
25,000
1
3.7
1
2.1
2,500
2
4.8
1.2
2.4
250
12
7.7
6.2
4.5
8.3
11.4
5.6
64
29
11
16.7
5.9
32
35
115
28
16
42
11.5
28.6
10.5
128
175
test
8.3
8
56
12.7
24.4
10.6
4
21
15.4
30
10
2
17
11.8
15.2
7.7
1
7
8.3
9.1
5.3
Control N&d RATIO
100,000
10,000
1,000
“g/ml
Fig. 1. Neutrahsation in the 2- and 3-stage
of different neutralisation
100 HEsAg
concentrations tests.
105
21
of HBsAg
by a human
anti-HB,
serum
(100 I.U./ml)
116
tube after the first wash is never too great to be significantly In the presence neutralisation
of lesser amounts
ratios because the antibody
antibody
to antigen,
presence
of even smaller amounts
neutralised
by the antiserum.
of antigen (e.g. 20 ng/ml) the 2-stage test gives high is blocking
but also the initial binding
not only the binding
of antigen
of labelled
to the coated tube. In the
(e.g. 2 ng/ml) the neutralisation
ratios in both tests
become lower because the unneutralised counts continue to fall, towards the end of the titration, after the neutralised counts have already reached their lowest, base-line values. We have always found non-specific reactions to give counts comparable only to those given by lower-titre true positive specimens. The failure to neutralise a strongly positive reaction should immediately arouse the suspicion that the positive reaction is specific but the neutralisation inadequate. The need for a common sense approach in such circumstances is well illustrated by Nath et al. (1979). The use of very high-titre neutralising sera from hyperimmunised necessarily a satisfactory alternative to lower-titre human sera. Human
animals antisera
is not of un-
usually high-titre from multi-transfused subjects, such as haemophiliacs, we also regard as potentially unsatisfactory. The 2-stage test is convenient in that, having no extra incubation stage, it is easily included in a batch of screen tests. For this reason it may be preferable to dilute a specimen known to be otherwise too strong, or to alter the relative proportions of test serum to neutralising and control serum, in order to use the 2-stage test. The 3-stage test is useful for confirming specimens of unknown strength, and for testing samples of plasma or blood products which might possibly coagulate when mixed with serum. Initially we used six-tube neutralisation tests, with immune and non-immune rabbit, horse and human sera, to assess the reliability of the confirmatory tests. A true HB,Agpositive gave a pattern of reactions showing significant inhibition by all the immune sera but not by the non-immune sera. In the case of one commercial RIA which employed human “‘1 anti-HB, we found that animal antisera could only be used to confirm weak HB,Ag-positive
sera after absorption
against human plasma proteins.
False positives of various types gave a variety of reaction patterns. For example, the ‘anti-rabbit’ type of false positive (in a labelled rabbit antibody RIA system) gave increased counts in all tubes except the immune and non-immune rabbit tubes. There is, however, a possibility of error if an unfortunate combination of single rabbit sera is used, and this may perhaps apply to other species; it is possible for a particular immune serum to inhibit a false-positive reaction which is not inhibited by the particular non-immune serum employed. This combination of sera would, if used alone, give an erroneous apparent confirmation
of specificity.
The human antiserum we have used has never neutralised a positive reaction that has not also been neutralised by rabbit and horse anti-HB,. For this reason we believe that the use of human anti-HB, alone can give satisfactory confirmation. REFERENCES
Heathcote, J., C.H. Cameron and D.S. Dane, 1974, Lancet 1, 71. 19, 321. N., A. Armbruster and F.R. Ellis, 1979, Transfusion
Natk
CONFIRMATION
1 (1980)
Biomedical
113-
113
116
Press
OF SPECIFICITY
BY NEUTRALISATION
IN IMMUNORADIO-
METRIC ASSAY FOR HEPATITIS B SURFACE ANTIGEN
C.H. CAMERON Department (Accepted
and MOYA BRIGGS
ofMicrobiology, 9 January,
The Middlesex Hospital Medical School, London
WI, England
1980)
Two types of neutralisation tion of positive results obtained
or blocking test using human anti-HB, are described by immunoradiometric screening for HBsAg.
for the confirma-
INTRODUCTION Since the earliest days of HB,Ag detection by immunoradiometric assay (RIA), it has been apparent that confirmation of the specificity of a positive reaction, even if repeated
and found consistently positive, is essential before the sample can be reported as HB,Agpositive. Incorrect positive reports give rise not only to false information in epidemiological studies, but also to a variety of unnecessary difficulties for persons wrongfully designated as hepatitis B positive. In the case of high-titre HB,Ag-positive specimens found by RIA confirmation can be provided by the use of other less sensitive tests. In the case of low-titre specimens, detectable only by RIA, confirmation must be made using an RIA method. We describe 2-stage and 3-stage neutralisation or blocking tests for use in different circumstances. These techniques have been for the most part employed with our own RIA system, but have also been used with various commercial RIA kits, and are in principle applicable to any similar solid-phase sandwich immunoradiometric MATERIALS
assay.
AND METHODS
RIA reagents These were described in outline
by Heathcote
et al. (1974).
A more detailed descrip-
tion is in preparation. Polystyrene tubes were coated with globulins from horses immunised with purified HB,Ag, the globulins having been absorbed against normal human serum and plasma proteins coupled to agarose beads. Labelled antibody was made from a similar globulin preparation from immunised rabbits by absorbing specific anti-HB, on to purified HB,Ag coupled to agarose beads, the antibody repurified.
was then eluted, iodinated
with “‘1, and finally
114
RIA test for HBfig Antibody-coated Overnight
tubes were inoculated
incubation
at room temperature
with 100 ~1 volumes of test or control serum. was followed by washing and incubation
for
1.5-2 h at room temperature with 100 ~1 volumes of labelled antibody, and finally by washing and counting for radioactivity. An alternative, more rapid, procedure used two incubation stages of 1.5 h each at 45°C. Neutralising serum The human antiserum used had a high titre of anti-HB, (200 I.U./ml) and was from a donor with no history of having received blood transfusions or blood products. The negative control serum was made from a pool of human sera, all found negative by RIA for HB,Ag, anti-HB, and anti-HB, . These sera were used at a dilution ing 0 .l % sodium azide.
of 1 : 2 in 0.02 M Tris/HCl buffer, pH 7.6, contain-
2-Stage neutralisation test Pairs of coated tubes were inoculated
with 20 ~1 volumes of neutrahsing
serum, followed by 100 ~1 volumes of the test serum. The procedure
or control
then followed that
of the standard test. 3-Stage neutralisation test Pairs of coated tubes were inoculated with 100 ~1 volumes of test serum and given the usual first incubation. The tubes were then washed and inoculated with 120 ~1 volumes of neutralising or control serum. After the additional incubation stage of 1 h at room temperature the tubes were washed and the labelled antibody added, as in the final stage of the standard
test. The radioactivity
remaining
after washing in the neutralised
and
control tubes was compared. RESULTS
The 2-stage and 3-stage neutralisation tests were performed on a series of dilutions of an ad antigen-positive serum and an ay antigen-positive serum, both of high titre, using the human neutralising antiserum (Table 1). The composite data are shown graphically (Fig. 1) to indicate approximately the degree of neutralisation to be expected with this antiserum in the presence of varying amounts of HB,Ag in either the 2-stage or the 3-stage procedure. DISCUSSION It will
amounts
be seen that in the 2-stage test the antiserum in unable to neutralise large of HB,Ag, whereas in the 3-stage test the amount of antigen remaining in the
115
TABLE
1
Results
of 2- and 3-stage
concentrations
neutralisation
of HBsAg subtypes
tests using a human
Control/neutralised
HB,&
anti-HB,
serum (100 I.U./ml)
and different
ad and ay ratio
(ng/ml) HBsAg subtype
ad
HB,Ag
subtype
ay -
2-stage
3-stage
test
test
2-stage
test
3-stage
250,000
1
3.2
1
1.9
25,000
1
3.7
1
2.1
2,500
2
4.8
1.2
2.4
250
12
7.7
6.2
4.5
8.3
11.4
5.6
64
29
11
16.7
5.9
32
35
115
28
16
42
11.5
28.6
10.5
128
175
test
8.3
8
56
12.7
24.4
10.6
4
21
15.4
30
10
2
17
11.8
15.2
7.7
1
7
8.3
9.1
5.3
Control N&d RATIO
100,000
10,000
1,000
“g/ml
Fig. 1. Neutrahsation in the 2- and 3-stage
of different neutralisation
100 HEsAg
concentrations tests.
105
21
of HBsAg
by a human
anti-HB,
serum
(100 I.U./ml)
116
tube after the first wash is never too great to be significantly In the presence neutralisation
of lesser amounts
ratios because the antibody
antibody
to antigen,
presence
of even smaller amounts
neutralised
by the antiserum.
of antigen (e.g. 20 ng/ml) the 2-stage test gives high is blocking
but also the initial binding
not only the binding
of antigen
of labelled
to the coated tube. In the
(e.g. 2 ng/ml) the neutralisation
ratios in both tests
become lower because the unneutralised counts continue to fall, towards the end of the titration, after the neutralised counts have already reached their lowest, base-line values. We have always found non-specific reactions to give counts comparable only to those given by lower-titre true positive specimens. The failure to neutralise a strongly positive reaction should immediately arouse the suspicion that the positive reaction is specific but the neutralisation inadequate. The need for a common sense approach in such circumstances is well illustrated by Nath et al. (1979). The use of very high-titre neutralising sera from hyperimmunised necessarily a satisfactory alternative to lower-titre human sera. Human
animals antisera
is not of un-
usually high-titre from multi-transfused subjects, such as haemophiliacs, we also regard as potentially unsatisfactory. The 2-stage test is convenient in that, having no extra incubation stage, it is easily included in a batch of screen tests. For this reason it may be preferable to dilute a specimen known to be otherwise too strong, or to alter the relative proportions of test serum to neutralising and control serum, in order to use the 2-stage test. The 3-stage test is useful for confirming specimens of unknown strength, and for testing samples of plasma or blood products which might possibly coagulate when mixed with serum. Initially we used six-tube neutralisation tests, with immune and non-immune rabbit, horse and human sera, to assess the reliability of the confirmatory tests. A true HB,Agpositive gave a pattern of reactions showing significant inhibition by all the immune sera but not by the non-immune sera. In the case of one commercial RIA which employed human “‘1 anti-HB, we found that animal antisera could only be used to confirm weak HB,Ag-positive
sera after absorption
against human plasma proteins.
False positives of various types gave a variety of reaction patterns. For example, the ‘anti-rabbit’ type of false positive (in a labelled rabbit antibody RIA system) gave increased counts in all tubes except the immune and non-immune rabbit tubes. There is, however, a possibility of error if an unfortunate combination of single rabbit sera is used, and this may perhaps apply to other species; it is possible for a particular immune serum to inhibit a false-positive reaction which is not inhibited by the particular non-immune serum employed. This combination of sera would, if used alone, give an erroneous apparent confirmation
of specificity.
The human antiserum we have used has never neutralised a positive reaction that has not also been neutralised by rabbit and horse anti-HB,. For this reason we believe that the use of human anti-HB, alone can give satisfactory confirmation. REFERENCES
Heathcote, J., C.H. Cameron and D.S. Dane, 1974, Lancet 1, 71. 19, 321. N., A. Armbruster and F.R. Ellis, 1979, Transfusion
Natk