- Email: [email protected]
Evaluation of an enzyme-linked immunosorbent assay for quantitation of antibodies to Junin virus in human sera
Journal of Virological Methods, Elsevier
19 (1988) 299-306
299
JVM 00702
Evaluation of an enzyme-linked immunosorbent assay for quantitation of antibodies to Junin virus in human sera S. Garcia
France,
A.M.
Ambrosio,
M.R.
Feuillade
and J.I. Maiztegui
lnstituto National de Estudios sobre Virosis Hemorrcigicas, Pergamino, Argentina (Accepted
7 January
1988)
Summary An enzyme-linked immunosorbent assay (ELBA) was evaluated for the quantitation of anti-Junin virus (JV) antibodies, in 83 selected cases of Argentine haemorrhagic fever (AHF). Serum samples were studied in two groups to facilitate comparative analysis; the first group was ELISA with indirect immunofluorescence (IF) test, in the second ELISA with plaque reduction neutralization test (PRNT). From the results obtained by using ELBA and IF on the same serum samples, a clear tendency of ELISA to demonstrate seroconversion for JV earlier and at higher frequency than IF test was noted. Simultaneous titration of specific antibodies by ELISA and PRNT tests rendered significantly correlated titers (r=0.81), both methods being equivalently specific (100%). The demonstration of specific antibodies by ELISA in two cases that were undetected by the PRNT test resulted in a higher sensitivity index for ELISA than for PRNT (100% vs 97%). It is concluded that ELISA could efficiently replace IF and PRNT tests for the diagnosis of AHF. ELISA;
Junin
virus;
Argentine
haemorrhagic
fever diagnosis
Introduction Argentine Junin virus Correspondence tina.
haemorrhagic fever (AHF) is an acute systemic disease caused by (JV), a member of the Arenavirus group. The disease courses with to: Susana
Garcia
France,
BioSidus
S.A.,
Larrea
926, (1117) Buenos
Aires,
Argen-
haematological, cardiovascular, neurological and renal involvement, and the mortality rates can reach 25% (Maiztegui, 1975). The transfusion of immune plasma has proven to be a very effective therapeutic approach, reducing the mortality rates to less than 3% - when administered within the first eight days of illness (Maiztegui et al., 1979; Enria et al., 1984). Etiological diagnosis is usually established by means of viral isolation and/or serological conversion. The demonstration of specific antibodies in the convalescence period has a double relevance since it allows confirmation of the diagnosis of the disease, and selection of immune plasma donors for therapeutic purposes. Complement fixation (CF) (Sever, 1962) and indirect immunofluorescence (IF) (Peters et al., 1973) tests are currently used to demonstrate seroconversion for JV. The plaque reduction neutralization test (PRNT) (Webb et al., 1969) was always considered to be the most reliable assay, rendering quantitative, specific results, even when its performance is rather slow and cumbersome. In recent years, the enzyme-linked immunosorbent assay (ELISA) developed by Engvall and Perlmann (1971) has proved to be one of the most suitable assay systems for routine laboratory practice in hospitals and for seroepidemiological surveillance (Voller et al., 1976). Several previous works (Ivanov et al., 1981; Niklasson et al., 1984; Kurstak et al., 1986) have reported the use of ELISA systems for the detection of Arenavirus antigens and specific antibodies. Viewing these tests as, a means for simplification in the serology required for the diagnosis of AHF, as well as for the selection of AHF immune plasma donors, the present work was carried out to standardize a direct ELISA for the detection of antibodies to JV, and, preliminarily, to explore the feasibility of replacing IF and PRNT tests by this ELISA method.
Materials
and Methods
Test sera Sera from 91 individuals were selected two groups of samples, as follows:
for comparative
purposes,
and studied
as
Group 1. Paired sera from 22 serologically diagnosed cases of AHF were included for preliminary comparison of results using ELISA and IF tests. Serum obtained on admission of each patient was used as the first sample. These were paired with serum samples obtained 30 days after the clinical onset in 8 cases, and 60 days after clinical onset in the remaining cases. Samples from five other individuals were used as negative controls for this group; three of them seroconverted for LCMV, and the remaining two were patients diagnosed as having non-AHF diseases. Group 2. Serum samples from 59 immune plasma donors, who had AHF between 90 days and 18 years earlier, were selected to compare the results obtained by ELISA and PRNT tests. As controls for this group, serum samples from five
301
individuals admitted with febrile diseases were included. These sera were negative when assayed by PRNT, but they had particular characteristics: (1) from one of these patients, JV was isolated by co-cultivation of blood mononuclear cells during the acute period of the disease: (2) seroconversion for JV by CF test was shown from another in the early convalescence, and (3) the remaining three sera were from patients with non-AHF diagnosis. A pool of normal human sera and one commercial normal IgG were used as negative controls for ELISA. A sample of human anti-JV serum with a high titer (15120) of neutralizing antibodies was included as a positive control in every test. The validity of this control was determined by assaying this positive serum simultaneously with purified IgG from human immune plasma in ELISA and PRNT tests.
Indirect immunofluorescence
(IF) test
Serum samples were tested on acetone-fixed JV-infected Vero cell spots and LCMV-infected L-cells, following already described procedures (Peters et al., 1973). Rabbit anti-human IgG, IgM, and IgA fluorescein conjugate (Dako F200) was used to detect specific reactions, thus enabling a direct comparison with the ELISA results.
Plaque reduction neutralization test (PRNT) Assays were conducted pressed as the maximum JV-infected mouse brain
according to Webb and Johnson (1969). Titers were exserum dilution that inhibited 80% plaque formation of homogenate on Vero cells under agarose.
ELBA test BHK - 21 cell monolayers Antigen.
grown in T1.50 flasks (150 cm2 of growth surface) were infected with the XJCl, strain of JV (MOI=O.l). Five days later, supernatants were harvested, and the purification of the viral content was performed following procedures described elsewhere (Objeski et al., 1976). Protein determinations were done according to Lowry et al. (1951).
Solid phase.
Polystyrene used; however, a suitable (Nunc l-67000).
removable well strips (Dynatech, 011010630-1) were alternative was found to be 96-well tissue culture plates
Anti-human IgG alkaline phosphatase conjugate (Sigma, A31.50) was Conjugate. used in every assay. The ELISA test was performed by the simple sandwich method. 100 ~1 of purified virus, containing 20 p,g/ml of protein in 0.5 M sodium borate buffer (SBB), pH 8.3, were placed at the bottom of each microtiter well. The plate was incubated at 37°C for 3 h or left overnight at 4°C. After two washes with 0.1 M sodium phosphate buffer (pH 7.4), plus 0.05% Tween-20 (PBS-T), the wells were filled with 100 l~,l of 1% bovine serum albumin in SBB. The plate was incubated for 1
302
h at 37”C, the wells were washed three times, and 100 ~1 of two-fold diluted test serum were added. After incubation at 37°C for 1 h, the wells were washed three times with PBS-T. 100 j.~lof conjugate were placed in each well, and the incubation of the plate was run for 1 h at 37°C or overnight at 4°C. After three washes, 200 ~1 of freshly prepared solution of p-nitrophenyl phosphate disodium (1 ml/ml) in 10% diethanolamine buffer containing 0.01% MgCl, (pH 9.8) were added to each well. The plate was incubated for 20-30 min at room temperature. The reaction was interrupted by adding 50 ~1 of 3 N NaOH. Positive reactions were revealed by the yellow coloration in the wells. Objective reading of the results was performed by using a spectrophotometer to measure the absorbance (OD) of the content of each well at 405 nm. Results were considered positive when the ratio of sample OD to control OD was 2 1.8. The sensitivity and specificity of ELISA and PRNT tests were calculated with the following formulas (Lilienfeld and Lilienfeld, 1980). Percent sensitivity =
TABLE Results Patients
true positives + false negatives
xl00
1 of JV antibody
detection
IF antibodies 30
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22
true positives
64* 64 8 16 64 64 Neg Neg Neg Neg Neg Neg Neg Neg Neg Ncg Neg Neg Neg Neg Neg Neg
*Titers are expressed
by means tested 60
64 64 Neg Neg Neg Neg Neg Neg Neg Neg Neg Neg Neg Neg Ncg Neg as the reciprocal
of ELISA
on day
and IF tests. ELISA
90
30
antibodies
tested 60
1280 1280 160 5120 320 5120 320 640 8 8 16 8 16 16 Neg Neg Neg Neg Neg Neg Neg Neg of the dilution.
160 320 1280 320 640 640 160 160 640 320 320 1280 80 640
on day
303
Percent
specificity
=
true negatives true negatives
+ false positives
x 100
and terms were defined: True positives:
JV isolation and/or serologic conversion from patients whose clinical and epidemiological data were indicative of AHF.
False positives:
Seroconversion from patients with negative viral isolation sence of clinical and epidemiological grounds.
True negatives:
Patients with negative and no seroconversion
False negatives:
viral isolation, no clinical evidence for JV by any available test.
in ab-
of AHF,
No seroconversion from patients with clear clinical and epidemiological data of AHF, and/or positive JV isolation.
Results The comparative results of testing paired serum samples from 22 cases of AHF by ELISA and IF tests are shown in Table 1. Patients 1.5-22 were diagnosed as AHF by PRNT on days 60-90 after clinical onset. The serum samples from the three patients diagnosed as cases of LCM rendered negative results by ELISA tests, as well as the two remaining non-AHF cases; these five individuals are not included in Table 1. The frequency of anti-J\/’ antibody detection by ELBA on serum samples from a second group of 64 individuals is compared with that obtained by PRNT in Table 2. Antibody titers of the 59 samples that were positive by both tests resulted in a high correlation coefficient (r=0.81), as it is depicted in Fig. 1. The two serum samples that rendered positive by ELISA test (titer=l:20) and negative by PRNT, were considered true positives, since they were from patients with virological and/or serological data supporting an AHF diagnosis. JV was isolated by co-cultivation of blood mononuclear cells from one of them, and a transient seroconversion by CF test had been obtained from the second in the early convalescence. TABLE
2
Comparative
frequencies
of antibodies
to JV detected
by ELISA
Total
+ -
tests. Total
ELISA
PNRT
and PRNT
+
_
59 2
0 3
59 5
61
3
64
304
10240.-
.
5120..
. ..
..
2 II w
l
.
60..
.
.
40,.
.
l .*
l .* l **
...
l
...
.:: ...
..
.. . .
...
.H l
.. l.
l
.
l*
r= 0.81 P
20..
l
l.
I 20
40
80
NEUTRALIZATION
160
320 (Reciprocal
640
1280 serum
2560
5120
10240
dilution)
Fig. 1. Correlation between titers of antibodies to JV assessed by ELBA and PRNT tests.
Repeated testing of the same serum samples by ELISA rendered highly reproducible titers, differing in no more than one dilution factor between assays. Sensitivity indices, calculated by using the values in Table 2 and the formulae described in Methods, resulted in 97% (59161) for PRNT versus 100% (61161) for ELISA. As for specificity, indices of 100% were obtained for both tests, since three serum samples were true negatives, and no false positives were found.
Discussion The results of applying the ELISA system described here indicate that this is a highly sensitive method to detect antibodies to JV. Serologic diagnosis of AHF is usually performed with samples obtained from each patient on admission, and 30, 60 and 90 days after the clinical onset. Data from our laboratory show that IF test allows to diagnose 90% of the cases of AHF (55% with the sample from day 30, and 35% with the 60 or 90 days sample), considering as 100% the number of seroconversions obtained by PRNT. The results shown in Table 1 demonstrate that the ELBA assayed would detect seroconversion for JV earlier than IF and, furthermore, it can render an AHF diagnosis of cases that would be unconfirmed if IF were the only available test. Data shown in Table 2 clearly indicate that ELISA is a more sensitive test than
305 PRNT, revealing seroconversion for JV in two cases of AHF that were undetected by PRNT. Comparison of the titers of anti-JV antibodies obtained by PRNT and IF with those assessed by ELISA leads to clearly different results. Titers of anti-JV antibodies obtained by PRNT and ELISA were highly correlated (r=0.81), whereas, titers obtained by ELISA show a consistent tendency to be higher than those determined by IF test (Table 1). Although this implies that the ELISA assay is determining antibodies directed toward JV envelope glycoprotein and that the IF test is mainly detecting nucleocapsid antibodies, this remains to be experimentally verified. The results presented strongly suggest that the ELISA system assayed is an adequate approach to circumvent some limitations of the methods currently used to detect JV antibodies. The high specificity and sensitivity of this ELISA test, in addition to the simplicity of its performance, make this method suitable for AHF diagnosis, seroepidemiological surveys, and also for the evaluation of specific antibodies content in the plasma units that are used in the treatment of the patients with AHF.
Acknowledgements We thank Dr. Delia Enria for helpful discussion of this manuscript and Dr. Z. Martinez Segovia for allowing us to use the Virus Laboratory facilities at the Instituto National de Microbiologia ‘Dr. Carlos G. Malbran’, Buenos Aires and Mrs. M. Denegri for secretarial assistance. References Engvall, E. and Perlmann, F. (1971) Immunochemistry 8, 871. Enria, D., Briggiler, A., Fermindez, N., Levis, S. and Maiztegui, J. (1984) Lancet ii, 255. Ivanov, A., Bashkirtsev, V.N. and Tkachenko, E.A. (1981) Arch. Virol. 67, 71. Kurstak, E., Tijssen, P., Kurstak, C. and Morisset, R, (1986) Bull. W.H.O. 64, 465. Lilienfeld, A. and Lilienfeld, D. (1980) Foundations of Epidemiology, Oxford University Press, ford. Lowry, O.H., Rosebrough, N.J., Fahr, A.L. and Randall, R.J. (1951) J. Biol. Chem. 193, 265. Maiztegui, J. (1975) Bull. W.H.O. 52, 567. Maiztegui, J., Fernandez, N. and de Damilano, A.J. (1979) Lancet ii, 1216. Niklasson, B.S., Jahrling, P.B. and Peters, C.J. (1984) J. Clin. Microbial. 20, 239. Objeski, J., Bishop, D.H.L., Murphy, F. and Palmer, E. (1976) J. Virol. 19, 985. Peters, C.J., Webb, P.A. and Johnson, K.M. (1973) Proc. Sot. Exp. Biol. Med. 142, 526. Sever, J.L. (1962) J. Immunol. 88, 320. Voller, A., Bidwell, D. and Bartlett, A. (1976) Bull. W.H.O. 53, 55. Webb, P., Johnson, K. and Mackensie, R. (1969) Proc. Sot. Exp. Biol. Med. 130, 1013.
(IX-
19 (1988) 299-306
299
JVM 00702
Evaluation of an enzyme-linked immunosorbent assay for quantitation of antibodies to Junin virus in human sera S. Garcia
France,
A.M.
Ambrosio,
M.R.
Feuillade
and J.I. Maiztegui
lnstituto National de Estudios sobre Virosis Hemorrcigicas, Pergamino, Argentina (Accepted
7 January
1988)
Summary An enzyme-linked immunosorbent assay (ELBA) was evaluated for the quantitation of anti-Junin virus (JV) antibodies, in 83 selected cases of Argentine haemorrhagic fever (AHF). Serum samples were studied in two groups to facilitate comparative analysis; the first group was ELISA with indirect immunofluorescence (IF) test, in the second ELISA with plaque reduction neutralization test (PRNT). From the results obtained by using ELBA and IF on the same serum samples, a clear tendency of ELISA to demonstrate seroconversion for JV earlier and at higher frequency than IF test was noted. Simultaneous titration of specific antibodies by ELISA and PRNT tests rendered significantly correlated titers (r=0.81), both methods being equivalently specific (100%). The demonstration of specific antibodies by ELISA in two cases that were undetected by the PRNT test resulted in a higher sensitivity index for ELISA than for PRNT (100% vs 97%). It is concluded that ELISA could efficiently replace IF and PRNT tests for the diagnosis of AHF. ELISA;
Junin
virus;
Argentine
haemorrhagic
fever diagnosis
Introduction Argentine Junin virus Correspondence tina.
haemorrhagic fever (AHF) is an acute systemic disease caused by (JV), a member of the Arenavirus group. The disease courses with to: Susana
Garcia
France,
BioSidus
S.A.,
Larrea
926, (1117) Buenos
Aires,
Argen-
haematological, cardiovascular, neurological and renal involvement, and the mortality rates can reach 25% (Maiztegui, 1975). The transfusion of immune plasma has proven to be a very effective therapeutic approach, reducing the mortality rates to less than 3% - when administered within the first eight days of illness (Maiztegui et al., 1979; Enria et al., 1984). Etiological diagnosis is usually established by means of viral isolation and/or serological conversion. The demonstration of specific antibodies in the convalescence period has a double relevance since it allows confirmation of the diagnosis of the disease, and selection of immune plasma donors for therapeutic purposes. Complement fixation (CF) (Sever, 1962) and indirect immunofluorescence (IF) (Peters et al., 1973) tests are currently used to demonstrate seroconversion for JV. The plaque reduction neutralization test (PRNT) (Webb et al., 1969) was always considered to be the most reliable assay, rendering quantitative, specific results, even when its performance is rather slow and cumbersome. In recent years, the enzyme-linked immunosorbent assay (ELISA) developed by Engvall and Perlmann (1971) has proved to be one of the most suitable assay systems for routine laboratory practice in hospitals and for seroepidemiological surveillance (Voller et al., 1976). Several previous works (Ivanov et al., 1981; Niklasson et al., 1984; Kurstak et al., 1986) have reported the use of ELISA systems for the detection of Arenavirus antigens and specific antibodies. Viewing these tests as, a means for simplification in the serology required for the diagnosis of AHF, as well as for the selection of AHF immune plasma donors, the present work was carried out to standardize a direct ELISA for the detection of antibodies to JV, and, preliminarily, to explore the feasibility of replacing IF and PRNT tests by this ELISA method.
Materials
and Methods
Test sera Sera from 91 individuals were selected two groups of samples, as follows:
for comparative
purposes,
and studied
as
Group 1. Paired sera from 22 serologically diagnosed cases of AHF were included for preliminary comparison of results using ELISA and IF tests. Serum obtained on admission of each patient was used as the first sample. These were paired with serum samples obtained 30 days after the clinical onset in 8 cases, and 60 days after clinical onset in the remaining cases. Samples from five other individuals were used as negative controls for this group; three of them seroconverted for LCMV, and the remaining two were patients diagnosed as having non-AHF diseases. Group 2. Serum samples from 59 immune plasma donors, who had AHF between 90 days and 18 years earlier, were selected to compare the results obtained by ELISA and PRNT tests. As controls for this group, serum samples from five
301
individuals admitted with febrile diseases were included. These sera were negative when assayed by PRNT, but they had particular characteristics: (1) from one of these patients, JV was isolated by co-cultivation of blood mononuclear cells during the acute period of the disease: (2) seroconversion for JV by CF test was shown from another in the early convalescence, and (3) the remaining three sera were from patients with non-AHF diagnosis. A pool of normal human sera and one commercial normal IgG were used as negative controls for ELISA. A sample of human anti-JV serum with a high titer (15120) of neutralizing antibodies was included as a positive control in every test. The validity of this control was determined by assaying this positive serum simultaneously with purified IgG from human immune plasma in ELISA and PRNT tests.
Indirect immunofluorescence
(IF) test
Serum samples were tested on acetone-fixed JV-infected Vero cell spots and LCMV-infected L-cells, following already described procedures (Peters et al., 1973). Rabbit anti-human IgG, IgM, and IgA fluorescein conjugate (Dako F200) was used to detect specific reactions, thus enabling a direct comparison with the ELISA results.
Plaque reduction neutralization test (PRNT) Assays were conducted pressed as the maximum JV-infected mouse brain
according to Webb and Johnson (1969). Titers were exserum dilution that inhibited 80% plaque formation of homogenate on Vero cells under agarose.
ELBA test BHK - 21 cell monolayers Antigen.
grown in T1.50 flasks (150 cm2 of growth surface) were infected with the XJCl, strain of JV (MOI=O.l). Five days later, supernatants were harvested, and the purification of the viral content was performed following procedures described elsewhere (Objeski et al., 1976). Protein determinations were done according to Lowry et al. (1951).
Solid phase.
Polystyrene used; however, a suitable (Nunc l-67000).
removable well strips (Dynatech, 011010630-1) were alternative was found to be 96-well tissue culture plates
Anti-human IgG alkaline phosphatase conjugate (Sigma, A31.50) was Conjugate. used in every assay. The ELISA test was performed by the simple sandwich method. 100 ~1 of purified virus, containing 20 p,g/ml of protein in 0.5 M sodium borate buffer (SBB), pH 8.3, were placed at the bottom of each microtiter well. The plate was incubated at 37°C for 3 h or left overnight at 4°C. After two washes with 0.1 M sodium phosphate buffer (pH 7.4), plus 0.05% Tween-20 (PBS-T), the wells were filled with 100 l~,l of 1% bovine serum albumin in SBB. The plate was incubated for 1
302
h at 37”C, the wells were washed three times, and 100 ~1 of two-fold diluted test serum were added. After incubation at 37°C for 1 h, the wells were washed three times with PBS-T. 100 j.~lof conjugate were placed in each well, and the incubation of the plate was run for 1 h at 37°C or overnight at 4°C. After three washes, 200 ~1 of freshly prepared solution of p-nitrophenyl phosphate disodium (1 ml/ml) in 10% diethanolamine buffer containing 0.01% MgCl, (pH 9.8) were added to each well. The plate was incubated for 20-30 min at room temperature. The reaction was interrupted by adding 50 ~1 of 3 N NaOH. Positive reactions were revealed by the yellow coloration in the wells. Objective reading of the results was performed by using a spectrophotometer to measure the absorbance (OD) of the content of each well at 405 nm. Results were considered positive when the ratio of sample OD to control OD was 2 1.8. The sensitivity and specificity of ELISA and PRNT tests were calculated with the following formulas (Lilienfeld and Lilienfeld, 1980). Percent sensitivity =
TABLE Results Patients
true positives + false negatives
xl00
1 of JV antibody
detection
IF antibodies 30
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22
true positives
64* 64 8 16 64 64 Neg Neg Neg Neg Neg Neg Neg Neg Neg Ncg Neg Neg Neg Neg Neg Neg
*Titers are expressed
by means tested 60
64 64 Neg Neg Neg Neg Neg Neg Neg Neg Neg Neg Neg Neg Ncg Neg as the reciprocal
of ELISA
on day
and IF tests. ELISA
90
30
antibodies
tested 60
1280 1280 160 5120 320 5120 320 640 8 8 16 8 16 16 Neg Neg Neg Neg Neg Neg Neg Neg of the dilution.
160 320 1280 320 640 640 160 160 640 320 320 1280 80 640
on day
303
Percent
specificity
=
true negatives true negatives
+ false positives
x 100
and terms were defined: True positives:
JV isolation and/or serologic conversion from patients whose clinical and epidemiological data were indicative of AHF.
False positives:
Seroconversion from patients with negative viral isolation sence of clinical and epidemiological grounds.
True negatives:
Patients with negative and no seroconversion
False negatives:
viral isolation, no clinical evidence for JV by any available test.
in ab-
of AHF,
No seroconversion from patients with clear clinical and epidemiological data of AHF, and/or positive JV isolation.
Results The comparative results of testing paired serum samples from 22 cases of AHF by ELISA and IF tests are shown in Table 1. Patients 1.5-22 were diagnosed as AHF by PRNT on days 60-90 after clinical onset. The serum samples from the three patients diagnosed as cases of LCM rendered negative results by ELISA tests, as well as the two remaining non-AHF cases; these five individuals are not included in Table 1. The frequency of anti-J\/’ antibody detection by ELBA on serum samples from a second group of 64 individuals is compared with that obtained by PRNT in Table 2. Antibody titers of the 59 samples that were positive by both tests resulted in a high correlation coefficient (r=0.81), as it is depicted in Fig. 1. The two serum samples that rendered positive by ELISA test (titer=l:20) and negative by PRNT, were considered true positives, since they were from patients with virological and/or serological data supporting an AHF diagnosis. JV was isolated by co-cultivation of blood mononuclear cells from one of them, and a transient seroconversion by CF test had been obtained from the second in the early convalescence. TABLE
2
Comparative
frequencies
of antibodies
to JV detected
by ELISA
Total
+ -
tests. Total
ELISA
PNRT
and PRNT
+
_
59 2
0 3
59 5
61
3
64
304
10240.-
.
5120..
. ..
..
2 II w
l
.
60..
.
.
40,.
.
l .*
l .* l **
...
l
...
.:: ...
..
.. . .
...
.H l
.. l.
l
.
l*
r= 0.81 P
20..
l
l.
I 20
40
80
NEUTRALIZATION
160
320 (Reciprocal
640
1280 serum
2560
5120
10240
dilution)
Fig. 1. Correlation between titers of antibodies to JV assessed by ELBA and PRNT tests.
Repeated testing of the same serum samples by ELISA rendered highly reproducible titers, differing in no more than one dilution factor between assays. Sensitivity indices, calculated by using the values in Table 2 and the formulae described in Methods, resulted in 97% (59161) for PRNT versus 100% (61161) for ELISA. As for specificity, indices of 100% were obtained for both tests, since three serum samples were true negatives, and no false positives were found.
Discussion The results of applying the ELISA system described here indicate that this is a highly sensitive method to detect antibodies to JV. Serologic diagnosis of AHF is usually performed with samples obtained from each patient on admission, and 30, 60 and 90 days after the clinical onset. Data from our laboratory show that IF test allows to diagnose 90% of the cases of AHF (55% with the sample from day 30, and 35% with the 60 or 90 days sample), considering as 100% the number of seroconversions obtained by PRNT. The results shown in Table 1 demonstrate that the ELBA assayed would detect seroconversion for JV earlier than IF and, furthermore, it can render an AHF diagnosis of cases that would be unconfirmed if IF were the only available test. Data shown in Table 2 clearly indicate that ELISA is a more sensitive test than
305 PRNT, revealing seroconversion for JV in two cases of AHF that were undetected by PRNT. Comparison of the titers of anti-JV antibodies obtained by PRNT and IF with those assessed by ELISA leads to clearly different results. Titers of anti-JV antibodies obtained by PRNT and ELISA were highly correlated (r=0.81), whereas, titers obtained by ELISA show a consistent tendency to be higher than those determined by IF test (Table 1). Although this implies that the ELISA assay is determining antibodies directed toward JV envelope glycoprotein and that the IF test is mainly detecting nucleocapsid antibodies, this remains to be experimentally verified. The results presented strongly suggest that the ELISA system assayed is an adequate approach to circumvent some limitations of the methods currently used to detect JV antibodies. The high specificity and sensitivity of this ELISA test, in addition to the simplicity of its performance, make this method suitable for AHF diagnosis, seroepidemiological surveys, and also for the evaluation of specific antibodies content in the plasma units that are used in the treatment of the patients with AHF.
Acknowledgements We thank Dr. Delia Enria for helpful discussion of this manuscript and Dr. Z. Martinez Segovia for allowing us to use the Virus Laboratory facilities at the Instituto National de Microbiologia ‘Dr. Carlos G. Malbran’, Buenos Aires and Mrs. M. Denegri for secretarial assistance. References Engvall, E. and Perlmann, F. (1971) Immunochemistry 8, 871. Enria, D., Briggiler, A., Fermindez, N., Levis, S. and Maiztegui, J. (1984) Lancet ii, 255. Ivanov, A., Bashkirtsev, V.N. and Tkachenko, E.A. (1981) Arch. Virol. 67, 71. Kurstak, E., Tijssen, P., Kurstak, C. and Morisset, R, (1986) Bull. W.H.O. 64, 465. Lilienfeld, A. and Lilienfeld, D. (1980) Foundations of Epidemiology, Oxford University Press, ford. Lowry, O.H., Rosebrough, N.J., Fahr, A.L. and Randall, R.J. (1951) J. Biol. Chem. 193, 265. Maiztegui, J. (1975) Bull. W.H.O. 52, 567. Maiztegui, J., Fernandez, N. and de Damilano, A.J. (1979) Lancet ii, 1216. Niklasson, B.S., Jahrling, P.B. and Peters, C.J. (1984) J. Clin. Microbial. 20, 239. Objeski, J., Bishop, D.H.L., Murphy, F. and Palmer, E. (1976) J. Virol. 19, 985. Peters, C.J., Webb, P.A. and Johnson, K.M. (1973) Proc. Sot. Exp. Biol. Med. 142, 526. Sever, J.L. (1962) J. Immunol. 88, 320. Voller, A., Bidwell, D. and Bartlett, A. (1976) Bull. W.H.O. 53, 55. Webb, P., Johnson, K. and Mackensie, R. (1969) Proc. Sot. Exp. Biol. Med. 130, 1013.
(IX-