Journal of Autoimmunity
(1992) 5,351-361
Use of a Molecularly Cloned Human SS.B Antigen to Detect Anti43S.B Antibodies
Robert A. Owe-Young ,* Sheila Horn,* John P. Edmondst and Allan D. Sturgesst *Dept.
of Rheumatology,
The St George Hospital,
University
TDepartment
of Rheumatology,
of New South Wales, Australia
(Received 24 May 1991 and accepted 30 December 1991) The aim of this study was to examine the utility of diagnostic assays based on recombinant SS.B/La (rSS.B). Using this antigen, we have developedan ELISA and an immunoblot and compared these recombinant antigenbased assays with traditional thymus extract-based counterimmunoelectrophoresis (CIEP). Using the recombinant ELISA, the incidence of anti-SS.B in 184 normal blood donors was 2.2% (four sera). These four sera were all low titre, i.e., 3-5 SD above the mean. Of 38 sera positive for anti-SS.B by CIBP, 37 were positive in both recombinant assays (97.4% concordance). Anti-SS.B titre in CBIP correlated strongly with results of the rSS.B-based ELISA, but the ELISA was 3,000-fold more sensitive. In an analysis of 152 autoimmune sera containing anti-DNA, anti-RNP, anticentromere, anti-SS.A/Ro or anti-cardiollpln, all of which were negative for anti-SS.B/La by CIEP, the recombinant assays detected 17 new antiSS.B positives. These positive results were found only in sera which had previously been characterised by CIEP as anti-SS.A/Ro positive. AntiSS.B/La antibodies detected by recombinant SS.B assays were found to be highly predictive of primary Sjiigren’s syndrome. Our results show that rSS.B can have an important role in the design of sensitive and specific assays for anti-SS.B. The diagnostic significance of anti-SS.B/La as a guide to primary Sjdgren’s syndrome is not diminished by the increased sensitivity of recombinant SS.B assays.
Introduction Identification of a specific antinuclear antibody (ANA) can yield valuable diagnostic information, e.g., anti-topoisomerase 1 (Scl-70) is highly indicative of scleroderma
Correspondence to: Dr A. D. Sturgess, Dept. of Rheumatology, Kogarah, NSW 2217, Australia.
St George Hospital, Belgrave St,
351 0896-841
l/92/030351
+ 11$03.00/0
0 1992 Academic Press Limited
352 Robert A. Owe-Young et al. [l], as is anti-Sm for systemic lupus erythematosus (SLE) [2]. Recent reports indicate that anti-SS.B (also known as anti-La) is a powerful predictor of primary Sjogren’s syndrome [3]. Indirect immunofluorescence cannot be used to characterise accurately anti-SS.B since a number of other ANAs produce an identical immunofluorescent pattern. The traditional methods of characterising ANAs such as SS.B. RNP, SS.A and Scl-70 are either slow and insensitive (e.g., Ouchterlony gel diffusion or counterimmunoelectrophoresis (CIEP)) or are technically exacting (e.g. immunoblotting using mammalian cell extracts) .and are therefore not easily applicable to a routine laboratory situation. Recently, molecularly cloned autoantigens have become available and can be used as substrates for the specific detection and characterisation of ANAs [4-71. Using recombinant human SS.B antigen (rSS.B), we have developed an ELISA and an immunoblot to detect anti-SS.B. We have investigated the prevalence of anti-SS.B in normal and disease sera and compared recombinant antigen-based assays with traditional CIEP. Methods Counterimmunoelectrophoresis This was performed according to the method of Kurata and Tan [8], using as antigen sources a lyophilised acetone extract of rabbit thymus (Pel Freeze, AR, USA), or fresh human spleen obtained from non-diseased trauma victims. Precipitin lines were identified by testing unknowns against in-house reference sera standardised against CDC reference sera. Expression and purification of recombinant SS.B antigen A cDNA for SS.B was cloned from a human B cell library as described [9]. The clone encodes the carboxy-terminal 355 amino acids of human SS.B lacking 53 amino acids at the amino terminus. At least two epitopes for anti-SS.B binding exist on this clone [9]. The rSS.B was produced as a fusion protein with 8-galactosidase in E. coli, and the fusion protein was purified by differential solubility [lo]. The recovery of recombinant fusion protein was assayed using the bicinchoninic acid (BCA) protocol (Pierce Chemical Co., Rockford, IL, USA). Fusion protein purity was assessed by sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE) performed in a BioRad Mini Protean electrophoresis cell according to the method of Laemmli [ll]. ELISA
design
For the ELISA various antigen and second antibody dilutions were tested in a checkerboard protocol to find optimal conditions. A local reference serum was verified against a CDC reference serum (Center for Disease Control, Atlanta, CA, USA 30333; ANA human reference serum no.2, cat. no. IS2073, lot no. 82-0008). To reduce background due to the presence of anti-E. coliantibodies, sera were absorbed at 1: 10 for at least 1 h at 4°C with lysate from a control culture of E. coli expressing
Cloned SS.B antigen assays
353
P-galactosidase. Sera which produced discrepant results on the ELISA and immunoblot were absorbed for a second time to reduce further anti-B coli background. The wells of Nunc Maxi-Sorp plates (Denmark) were coated overnight at 4°C with 2 pg/well recombinant protein in 50 ~1 bicarbonate buffer (0.04 M Na,CO,, 0.06 M NaHCO,, pH 9.6). The remaining steps of the assay were carried out at room temperature. Residual protein-reactive sites were blocked for 1 h with 50 ul/well 10”” adult bovine serum in phosphate-buffered saline (blocking buffer). After blocking, the wells were washed three times with 0.05% Tween 20 in PBS before the addition of 50 l.d serum at 1: 100 dilution in blocking buffer for 1 h. The wells were washed again and 50 ~1 of a dilution of goat anti-human IgG (H + L)-peroxidase conjugate (KPL, Gaithersberg, MD, USA; cat. no. 041006) in blocking buffer was added for 90 mins. After a further six washes, TMB substrate (KPL) was used to develop the plate at 37°C; the absorbance of the wells containing the local reference serum was monitored at 655 nm until the optical density was 0.430. This was found to be the absorbance at which the addition of 1 M phosphoric acid to stop further colour development would give the local reference serum an OD of approximately 1.0 at the subsequent 450 nm reading wavelength. The CDC standard, using this control, produced an OD of 1.385. Absorbances were read in a BioRad model 3550 plate reader, which automatically subtracted the OD of ‘no antigen’ control wells from the OD of test wells. Positive sera were defined as those having absorbances greater than 3 standard deviations (SD) above the mean of 184 blood bank donors. Sera with optical densities from 3 to 5 SD above the mean were termed low titre, 5 to 10 SD above the mean were termed moderate titre, and over 10 SD above the mean were termed high titre. Immunoblotting Sera which were positive in the rSS.B ELISA were also tested in immunoblots using the rSS.B as the antigen. This was done to verify that positive results in the ELISA were due to reactivity to rSS.B and not to any contaminating E. coli proteins. Nitrocellulose strips on which rSS.B had been blotted [12] were incubated for 1 h with 5”, non-fat skim milk in PBS (‘Blotto’) to block residual protein-reactive sites. Test and control sera were added to the strips at I:200 dilution for 1 h. After three 5 min washes with Blotto, a 1:7,500 dilution of goat anti-human IgG-alkaline phosphatase conjugate (Promega) was added for 1 h. After a further three washes, antibody binding was detected with NBT/BCIP substrate. Inhibition of ELISA
results
As an additional test of the validity of ELISA results, sera from patients who were anti-SS.B positive were absorbed extensively with rSS.B-producing E. coli and compared with serum aliquots absorbed with control E. coli culture expressing 8galactosidase alone. Optical densities of sera absorbed with control and rSS.Bproducing E. coli were analysed using Student’s paired t-test. Specificity
of rSS.B
assays
The specificity and sensitivity of the rSS.B-based ELISA and the immtmoblot was evaluated by testing sera from patients with a variety of autoantibodies previously
354 Robert A. Owe-Young et al. classified by routine diagnostic methods (CIEP, Farr assay, immunofluorescence on Hep2 cells). These included 31 anti-DNA and 30 anti-RNP sera from SLE patients, 27 anti-centromere sera from limited scleroderma patients, four anti-cardiolipin sera from primary antiphospholipid syndrome patients, and 50 anti-SSA positive/antiSS.B negative sera from 27 SLE patients, 18 primary Sjogren’s patients and 15 patients in whom no diagnosis was known. Thirty eight anti-SSB positive/antiSS.A positive sera were also tested. Twenty two were from patients with primary Sjogren’s syndrome, and 16 were from patients with SLE or from patients in whom no diagnosis was known.
Correlation
of ELISA
with CIEP
Spearman’s rank correlation coefficient was used to assess the relationship between the level of SS.B antibodies as measured by optical density in the rSS.B ELISA and the titre of SS.B antibodies as measured by CIEP. Significance was accepted at the 0.05 level. Results Expression and purification of recombinant protein The BCA protein assay showed the yield of rSS.B protein to be 0.3 mg/mL bacterial culture. SDS-PAGE on 7.5% gels revealed the presence of a band corresponding to the molecular weight of the rSS.B+galactosidase fusion protein at about 160 kDa which was not present in preparations from bacteria expressing 8-galactosidase alone. The rSS.B was aliquoted and stored at -3O”C, as repeated freeze-thawing abrogated its activity.
ELZSA The normal blood donor sera (n = 184) had a mean absorbance of 0.065 + 0.050 (1 SD); there were four low titre positives, giving an incidence of 2.2%. The OD cutoff for positivity was established at 0.215 (3 SD above mean). Inter-assay coefficient of variation (CV) was 13.8% and intra-assay CV was 4.3%.
ELISA
specificity and immunoblotting
Anti-SS.B sera Thirty seven of 38 sera (97.4%) which were positive for anti-SS.B by CIEP were positive in the ELISA, with ODs ranging from 0.550 to 2.082. All 37 ELISApositive sera were confirmed by immunoblot (Table 1, Figure 1). Other autoimmune sera Of 152 sera, 36 were positive when initially tested in the ELISA. Only 30 of these were available for confirmatory testing by immunoblot. Seventeen sera were confirmed. Thus, 13 sera produced discrepant results in the two rSS.B assays. These 13 sera were retested in the ELISA after a second round of absorption to reduce further
Cloned SS.B antigen assays
355
1. Results of rSS.B-based assays. Comparison between recombinant SS.B-based assays (ELISA and immunoblot) with serological groups defined by traditional assays. When sera were treated with only a single round of absorption with a lysate of control E. coli, there were 26 false positives, which were not confirmed by immunoblot. A second absorption, to deplete anti-E. coli antibodies, removed these false positives. Conjirmed positives are only found in anti-SS.A and anti-SS. B positive groups
Table
Serological
group
Number
DNA RNP ACA aC1 SS.A+/SS.BSS.A+/SS.B+
ELISAfve
31 30 27 4 50 38
Western
0 0 0 0 17 37
blot + ve 0 0 0 0 17 37
-I
n=3B @
2
: 6
8 n=31 0.5 -
0 0
n=30 Q
Q 0
8
DNA
RNP
0
0
0 6
0
n-4
0 Cut-off
0
p
0
0
ACA
aCL
SS.A+/SS.B-
SS.A+/SS.B+
CIEP group
Figure 1. OD measurements in rSS.B-based ELISA. Cut-off for positivity was set at the mean + 3 SD of blood donors. Sera were treated with a single round of absorption with an anti-E. coli lysate to reduce anti-E. coli antibodies. The positive results in the anti-DNA, anti-RNP, aCL and anti-centromere groups were not present after a second absorption procedure. 184
the anti-E. coli background. After this treatment, none of the 13 were positive in the ELISA. No confirmed positives were found in the anti-BNP, anti-DNA, anticentromere or anti-cardiolipin groups. Only in the anti-SS.A sera group were confirmed anti-SS.B positives found (17 out of 50,34%). Of the anti-SS.A group, 13 of the 27 primary Sjogren’s patients were positive, compared with four of 33 patients with other diseases, P< 0.005.
356 Robert A. Owe-Young et al.
-40 t
-6O-
-took I
ControlE. co/i
SS.B-E.
cob
Figure 2. Changes in rSS.B ELISA OD after absorption of sera with rSS.B. These sera were all k3.B positive in CIEP, and moderate or high titre in the rSS.B ELISA. The changes in OD are represented as percentage changes. Most sera showed at least a 50”/6 fall in OD after specific absorption.
Inhibition of ELISA results by rSS.B Figure 2 shows the OD results in the ELISA after absorption with control E. coli or with rSS.B-producing E. coli. After absorption of the CIEP-positive sera with SS.Bproducing E. coli, most of the anti-SS.B binding activity was deleted in all except one serum. Similarly, when ELISA-positive, CIEP-negative sera were treated this way (Figure 3), the ODs after incubation with rSS.B-producing E. coliwere significantly different from the ODs after absorption with control E. coli (P
(1992) 5,351-361
Use of a Molecularly Cloned Human SS.B Antigen to Detect Anti43S.B Antibodies
Robert A. Owe-Young ,* Sheila Horn,* John P. Edmondst and Allan D. Sturgesst *Dept.
of Rheumatology,
The St George Hospital,
University
TDepartment
of Rheumatology,
of New South Wales, Australia
(Received 24 May 1991 and accepted 30 December 1991) The aim of this study was to examine the utility of diagnostic assays based on recombinant SS.B/La (rSS.B). Using this antigen, we have developedan ELISA and an immunoblot and compared these recombinant antigenbased assays with traditional thymus extract-based counterimmunoelectrophoresis (CIEP). Using the recombinant ELISA, the incidence of anti-SS.B in 184 normal blood donors was 2.2% (four sera). These four sera were all low titre, i.e., 3-5 SD above the mean. Of 38 sera positive for anti-SS.B by CIBP, 37 were positive in both recombinant assays (97.4% concordance). Anti-SS.B titre in CBIP correlated strongly with results of the rSS.B-based ELISA, but the ELISA was 3,000-fold more sensitive. In an analysis of 152 autoimmune sera containing anti-DNA, anti-RNP, anticentromere, anti-SS.A/Ro or anti-cardiollpln, all of which were negative for anti-SS.B/La by CIEP, the recombinant assays detected 17 new antiSS.B positives. These positive results were found only in sera which had previously been characterised by CIEP as anti-SS.A/Ro positive. AntiSS.B/La antibodies detected by recombinant SS.B assays were found to be highly predictive of primary Sjiigren’s syndrome. Our results show that rSS.B can have an important role in the design of sensitive and specific assays for anti-SS.B. The diagnostic significance of anti-SS.B/La as a guide to primary Sjdgren’s syndrome is not diminished by the increased sensitivity of recombinant SS.B assays.
Introduction Identification of a specific antinuclear antibody (ANA) can yield valuable diagnostic information, e.g., anti-topoisomerase 1 (Scl-70) is highly indicative of scleroderma
Correspondence to: Dr A. D. Sturgess, Dept. of Rheumatology, Kogarah, NSW 2217, Australia.
St George Hospital, Belgrave St,
351 0896-841
l/92/030351
+ 11$03.00/0
0 1992 Academic Press Limited
352 Robert A. Owe-Young et al. [l], as is anti-Sm for systemic lupus erythematosus (SLE) [2]. Recent reports indicate that anti-SS.B (also known as anti-La) is a powerful predictor of primary Sjogren’s syndrome [3]. Indirect immunofluorescence cannot be used to characterise accurately anti-SS.B since a number of other ANAs produce an identical immunofluorescent pattern. The traditional methods of characterising ANAs such as SS.B. RNP, SS.A and Scl-70 are either slow and insensitive (e.g., Ouchterlony gel diffusion or counterimmunoelectrophoresis (CIEP)) or are technically exacting (e.g. immunoblotting using mammalian cell extracts) .and are therefore not easily applicable to a routine laboratory situation. Recently, molecularly cloned autoantigens have become available and can be used as substrates for the specific detection and characterisation of ANAs [4-71. Using recombinant human SS.B antigen (rSS.B), we have developed an ELISA and an immunoblot to detect anti-SS.B. We have investigated the prevalence of anti-SS.B in normal and disease sera and compared recombinant antigen-based assays with traditional CIEP. Methods Counterimmunoelectrophoresis This was performed according to the method of Kurata and Tan [8], using as antigen sources a lyophilised acetone extract of rabbit thymus (Pel Freeze, AR, USA), or fresh human spleen obtained from non-diseased trauma victims. Precipitin lines were identified by testing unknowns against in-house reference sera standardised against CDC reference sera. Expression and purification of recombinant SS.B antigen A cDNA for SS.B was cloned from a human B cell library as described [9]. The clone encodes the carboxy-terminal 355 amino acids of human SS.B lacking 53 amino acids at the amino terminus. At least two epitopes for anti-SS.B binding exist on this clone [9]. The rSS.B was produced as a fusion protein with 8-galactosidase in E. coli, and the fusion protein was purified by differential solubility [lo]. The recovery of recombinant fusion protein was assayed using the bicinchoninic acid (BCA) protocol (Pierce Chemical Co., Rockford, IL, USA). Fusion protein purity was assessed by sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE) performed in a BioRad Mini Protean electrophoresis cell according to the method of Laemmli [ll]. ELISA
design
For the ELISA various antigen and second antibody dilutions were tested in a checkerboard protocol to find optimal conditions. A local reference serum was verified against a CDC reference serum (Center for Disease Control, Atlanta, CA, USA 30333; ANA human reference serum no.2, cat. no. IS2073, lot no. 82-0008). To reduce background due to the presence of anti-E. coliantibodies, sera were absorbed at 1: 10 for at least 1 h at 4°C with lysate from a control culture of E. coli expressing
Cloned SS.B antigen assays
353
P-galactosidase. Sera which produced discrepant results on the ELISA and immunoblot were absorbed for a second time to reduce further anti-B coli background. The wells of Nunc Maxi-Sorp plates (Denmark) were coated overnight at 4°C with 2 pg/well recombinant protein in 50 ~1 bicarbonate buffer (0.04 M Na,CO,, 0.06 M NaHCO,, pH 9.6). The remaining steps of the assay were carried out at room temperature. Residual protein-reactive sites were blocked for 1 h with 50 ul/well 10”” adult bovine serum in phosphate-buffered saline (blocking buffer). After blocking, the wells were washed three times with 0.05% Tween 20 in PBS before the addition of 50 l.d serum at 1: 100 dilution in blocking buffer for 1 h. The wells were washed again and 50 ~1 of a dilution of goat anti-human IgG (H + L)-peroxidase conjugate (KPL, Gaithersberg, MD, USA; cat. no. 041006) in blocking buffer was added for 90 mins. After a further six washes, TMB substrate (KPL) was used to develop the plate at 37°C; the absorbance of the wells containing the local reference serum was monitored at 655 nm until the optical density was 0.430. This was found to be the absorbance at which the addition of 1 M phosphoric acid to stop further colour development would give the local reference serum an OD of approximately 1.0 at the subsequent 450 nm reading wavelength. The CDC standard, using this control, produced an OD of 1.385. Absorbances were read in a BioRad model 3550 plate reader, which automatically subtracted the OD of ‘no antigen’ control wells from the OD of test wells. Positive sera were defined as those having absorbances greater than 3 standard deviations (SD) above the mean of 184 blood bank donors. Sera with optical densities from 3 to 5 SD above the mean were termed low titre, 5 to 10 SD above the mean were termed moderate titre, and over 10 SD above the mean were termed high titre. Immunoblotting Sera which were positive in the rSS.B ELISA were also tested in immunoblots using the rSS.B as the antigen. This was done to verify that positive results in the ELISA were due to reactivity to rSS.B and not to any contaminating E. coli proteins. Nitrocellulose strips on which rSS.B had been blotted [12] were incubated for 1 h with 5”, non-fat skim milk in PBS (‘Blotto’) to block residual protein-reactive sites. Test and control sera were added to the strips at I:200 dilution for 1 h. After three 5 min washes with Blotto, a 1:7,500 dilution of goat anti-human IgG-alkaline phosphatase conjugate (Promega) was added for 1 h. After a further three washes, antibody binding was detected with NBT/BCIP substrate. Inhibition of ELISA
results
As an additional test of the validity of ELISA results, sera from patients who were anti-SS.B positive were absorbed extensively with rSS.B-producing E. coli and compared with serum aliquots absorbed with control E. coli culture expressing 8galactosidase alone. Optical densities of sera absorbed with control and rSS.Bproducing E. coli were analysed using Student’s paired t-test. Specificity
of rSS.B
assays
The specificity and sensitivity of the rSS.B-based ELISA and the immtmoblot was evaluated by testing sera from patients with a variety of autoantibodies previously
354 Robert A. Owe-Young et al. classified by routine diagnostic methods (CIEP, Farr assay, immunofluorescence on Hep2 cells). These included 31 anti-DNA and 30 anti-RNP sera from SLE patients, 27 anti-centromere sera from limited scleroderma patients, four anti-cardiolipin sera from primary antiphospholipid syndrome patients, and 50 anti-SSA positive/antiSS.B negative sera from 27 SLE patients, 18 primary Sjogren’s patients and 15 patients in whom no diagnosis was known. Thirty eight anti-SSB positive/antiSS.A positive sera were also tested. Twenty two were from patients with primary Sjogren’s syndrome, and 16 were from patients with SLE or from patients in whom no diagnosis was known.
Correlation
of ELISA
with CIEP
Spearman’s rank correlation coefficient was used to assess the relationship between the level of SS.B antibodies as measured by optical density in the rSS.B ELISA and the titre of SS.B antibodies as measured by CIEP. Significance was accepted at the 0.05 level. Results Expression and purification of recombinant protein The BCA protein assay showed the yield of rSS.B protein to be 0.3 mg/mL bacterial culture. SDS-PAGE on 7.5% gels revealed the presence of a band corresponding to the molecular weight of the rSS.B+galactosidase fusion protein at about 160 kDa which was not present in preparations from bacteria expressing 8-galactosidase alone. The rSS.B was aliquoted and stored at -3O”C, as repeated freeze-thawing abrogated its activity.
ELZSA The normal blood donor sera (n = 184) had a mean absorbance of 0.065 + 0.050 (1 SD); there were four low titre positives, giving an incidence of 2.2%. The OD cutoff for positivity was established at 0.215 (3 SD above mean). Inter-assay coefficient of variation (CV) was 13.8% and intra-assay CV was 4.3%.
ELISA
specificity and immunoblotting
Anti-SS.B sera Thirty seven of 38 sera (97.4%) which were positive for anti-SS.B by CIEP were positive in the ELISA, with ODs ranging from 0.550 to 2.082. All 37 ELISApositive sera were confirmed by immunoblot (Table 1, Figure 1). Other autoimmune sera Of 152 sera, 36 were positive when initially tested in the ELISA. Only 30 of these were available for confirmatory testing by immunoblot. Seventeen sera were confirmed. Thus, 13 sera produced discrepant results in the two rSS.B assays. These 13 sera were retested in the ELISA after a second round of absorption to reduce further
Cloned SS.B antigen assays
355
1. Results of rSS.B-based assays. Comparison between recombinant SS.B-based assays (ELISA and immunoblot) with serological groups defined by traditional assays. When sera were treated with only a single round of absorption with a lysate of control E. coli, there were 26 false positives, which were not confirmed by immunoblot. A second absorption, to deplete anti-E. coli antibodies, removed these false positives. Conjirmed positives are only found in anti-SS.A and anti-SS. B positive groups
Table
Serological
group
Number
DNA RNP ACA aC1 SS.A+/SS.BSS.A+/SS.B+
ELISAfve
31 30 27 4 50 38
Western
0 0 0 0 17 37
blot + ve 0 0 0 0 17 37
-I
n=3B @
2
: 6
8 n=31 0.5 -
0 0
n=30 Q
Q 0
8
DNA
RNP
0
0
0 6
0
n-4
0 Cut-off
0
p
0
0
ACA
aCL
SS.A+/SS.B-
SS.A+/SS.B+
CIEP group
Figure 1. OD measurements in rSS.B-based ELISA. Cut-off for positivity was set at the mean + 3 SD of blood donors. Sera were treated with a single round of absorption with an anti-E. coli lysate to reduce anti-E. coli antibodies. The positive results in the anti-DNA, anti-RNP, aCL and anti-centromere groups were not present after a second absorption procedure. 184
the anti-E. coli background. After this treatment, none of the 13 were positive in the ELISA. No confirmed positives were found in the anti-BNP, anti-DNA, anticentromere or anti-cardiolipin groups. Only in the anti-SS.A sera group were confirmed anti-SS.B positives found (17 out of 50,34%). Of the anti-SS.A group, 13 of the 27 primary Sjogren’s patients were positive, compared with four of 33 patients with other diseases, P< 0.005.
356 Robert A. Owe-Young et al.
-40 t
-6O-
-took I
ControlE. co/i
SS.B-E.
cob
Figure 2. Changes in rSS.B ELISA OD after absorption of sera with rSS.B. These sera were all k3.B positive in CIEP, and moderate or high titre in the rSS.B ELISA. The changes in OD are represented as percentage changes. Most sera showed at least a 50”/6 fall in OD after specific absorption.
Inhibition of ELISA results by rSS.B Figure 2 shows the OD results in the ELISA after absorption with control E. coli or with rSS.B-producing E. coli. After absorption of the CIEP-positive sera with SS.Bproducing E. coli, most of the anti-SS.B binding activity was deleted in all except one serum. Similarly, when ELISA-positive, CIEP-negative sera were treated this way (Figure 3), the ODs after incubation with rSS.B-producing E. coliwere significantly different from the ODs after absorption with control E. coli (P
