Selective Detection of Autoimmune Antibodies to Singleand Double-Stranded DNA by Enzyme Immunoassay

Transcription

1 142 Annals of Clinical & Laboratory Science, vol. 33, no. 2, 2003 Selective Detection of Autoimmune Antibodies to Singleand Double-Stranded DNA by Enzyme Immunoassay Gregory S. Makowski 1 and Melinda L. Ramsby 2 Departments of Laboratory Medicine 1 and Medicine, 2 University of Connecticut School of Medicine, Farmington, Connecticut Abstract. Presence in serum of anti-double-stranded deoxyribonucleic acid antibodies (anti-dsdna) is one of the diagnostic criteria for systemic lupus erythematosus. Anti-single stranded DNA antibodies (anti-ssdna) also occur, but their clinical significance is unclear. Use of enzyme immunoassay (EIA) kits that are specific for anti-dsdna is desirable but problematic, since preparation of dsdna of high integrity is difficult (ie, regions of ssdna may persist). This study evaluated an EIA kit that uses low Mw plasmid/ bacteriophage DNA as a nucleic acid source. Anti-dsDNA results were compared to results obtained by an anti-ssdna EIA and an immunofluorescent assay (IFA) that uses Crithidia luciliae (specific for anti-dsdna). Consecutive serum samples (n=139, 88% female) submitted to the clinical laboratory for anti-dsdna analysis were evaluated. EIA precision was determined at three levels. Intra-assay precision [mean ± SD (CV)] for anti-dsdna: 36 ± 3.5 IU/ml (9.8%); 98 ± 4.4 (4.5%); and 245 ± 5.8 (2.4%); and anti-ssdna: 40 ± 1.4 U/ml (3.5%); 190 ± 4.8 (2.5%); and 283 ± 10.3 (3.7%) (n=8). Inter-assay precision for antidsdna: 36 ± 6.3 IU/ml (17.5%); 90 ± 5.9 (6.6%); and 207 ± 20.9 (10.1%); and anti-ssdna: 48 ± 8.2 U/ ml (17.0%); 193 ± 12.7 (6.6%); and 263 ± 21.5 (8.2%) (n=8). Linearity was assessed with (a) high dsdna/high ssdna and (b) low dsdna/high ssdna samples (no high dsdna/low ssdna samples were identified). Linearity (>200 IU/ml) was found for both sample types with a correlation coefficient (r) of Anti-dsDNA immunoreactivity was not apparent with the low dsdna/high ssdna sample. More patients were positive for anti-ssdna (54%), compared to anti-dsdna (17%). IFA confirmation (Crithidia) indicated a relative sensitivity and specificity of 94.1% and 93.4% for the anti-dsdna EIA. IFA positivity correlated with increased anti-dsdna level: 20% (30-60 IU/ml); 70% ( IU/ml); and 89% (>200 IU/ml). Of specimens that were anti-ssdna positive and anti-dsdna negative (n=51), only one was IFA positive. When IFA was compared to an anti-dsdna EIA kit that used high Mw calf thymus DNA, lower relative specificity (88.2%) and sensitivity (72.6%) was obtained. Anti-ssDNA was found in many false positive specimens (87%). (received 22 October 2002; accepted 30 November 2002) Keywords: lupus erythematosis, autoimmunity; anti-dna antibodies, enzyme immunoassay Introduction The presence in serum of circulating autoimmune antibodies to native double-stranded deoxyribonucleic acid (anti-dsdna) [1] is one of the American Address correspondence to Gregory S. Makowski, Ph.D., Department of Laboratory Medicine, University of Connecticut Health Center, MC-2235, 263 Farmington Ave., Farmington, CT , USA; tel ; fax ; makowski@nso1.uchc.edu. The current address of Melinda L Ramsby, M.D., Ph.D., is: Farmington Valley Arthritis and Rheumatology, LLC, 54 West Avon Road, Avon, CT 06001, USA. College of Rheumatology s diagnostic criteria for systemic lupus erythematosus (SLE) [2]. Assay methods for detecting anti-dsdna include radioimmunoprecipitation (Farr method) [3], immunofluorescence assay (IFA) using Crithidia luciliae [4], and various enzyme immunoassays (EIAs) [5-8]. Despite their convenience, the EIAs have failed to supplant more labor intensive methods such as IFA [9]. The EIAs that were first developed for antidsdna used microtiter plate wells coated with immobilized DNA from calf thymus as the antibody binding substrate. However, preparation of intact /03/0200/0140 $ by the Association of Clinical Scientists, Inc.

2 Enzyme immunoassays for dsdna and ssdna 143 high molecular weight (Mw) dsdna was difficult since single-stranded regions may persist or be artifactually introduced [10-12]. The non-double stranded regions of DNA contribute to immunoreactivity with other antibodies, such as anti-single stranded DNA antibodies (anti-ssdna) [12]. Although anti-ssdna antibodies also occur (with and without anti-dsdna), their clinical significance in SLE is unclear [13-15]. Efforts to enhance analytical performance in the detection of anti-dsdna have included improved methods for preparation of intact DNA [8,12], the introduction of reference antiserum [16], and the use of an international standardized material [17]. Despite these innovations, inconsistent performance of commercial EIA kits for autoimmune antibodies, especially anti-dsdna, has persisted [18-21]. In this study, we evaluated the analytical performance of an anti-dsdna EIA kit that uses a unique combination of plasmid and bacteriophage lambda DNA as a low Mw nucleic acid source. Results obtained with this kit were compared with those obtained with (a) an EIA kit specific for antissdna, (b) an indirect immunofluorescence assay (IFA) using Crithidia luciliae (highly specific for antidsdna), and (c) an EIA kit that uses immobilized high Mw DNA (less specific for anti-dsdna). Material and Methods Patient specimens. A total of 139 consecutive serum specimens submitted to the clinical laboratory for routine anti-dsdna antibody testing were evaluated. The samples were collected in serum separator tubes (Becton-Dickinson) and allowed to clot for 30 min at room temperature. After centrifugation, serum samples were aliquoted and stored frozen at -80 C. More than 90% of the specimens were from patients in the rheumatology clinic. The patients were predominantly female (n = 124, age [mean ± SD] 40 ± 17 yr), versus male (n = 15, age 46 ± 14 yr). The Institutional Review Board of the University of Connecticut Health Center approved this study. Enzyme immunoassay. Enzyme immunoassay (EIA) kits for detection of anti-dsdna antibodies and anti-ssdna antibodies were obtained from Helix Diagnostics (West Sacramento, CA). Serum samples were diluted 1:100 (v/v) in assay diluent and assays were done according to the manufacturer s directions. Absorbance measurements (450 nm) were performed with a Thermomax microplate reader (Molecular Devices, Inc., Menlo Park, CA). The manufacturer s recommended cut-off values for anti-dsdna antibody positivity were used: <30 IU/ ml (negative); IU/ml (low positive); IU/ml (positive); and >200 IU/ml (strong positive). The results were compared to those obtained by an EIA kit containing immobilized high Mw calf thymus DNA (kit #2), obtained from the Binding Site, Inc. (San Diego, CA). Immunofluorescence assay. An immunofluorescence assay (IFA) for anti-dsdna was performed by a kit that used Crithidia luciliae (Binding Site, Inc.). The IFA assay was performed according to the manufacturer s directions; IFA positivity was determined at a screening dilution of 1:10 (v/v). Statistical analysis. Results were expressed as the mean ± SD. Statistical tests included Student s t- test and Spearman s rank correlation coefficient (r); p <0.05 was considered significant. Results Precision. Assay performance characteristics were determined at 3 decision levels for the anti-dsdna EIA (kit #1) using patient specimens (Table 1). Similar analytical performance was obtained for the anti-ssdna EIA kit. In general, good reproducibility was obtained; the CVs for intra-assay precision were generally <10% and were only slightly greater for inter-assay studies. Similar precision results were obtained with EIA kit #2, which used wells coated with high Mw dsdna (data not shown). Linearity. Linearity was assessed for the anti-dsdna (kit #1) and anti-ssdna antibody EIAs using serially diluted patient samples containing (a) high dsdna/ high ssdna (sample 1) and (b) low dsdna/high ssdna antibodies (sample 2) (Fig. 1). Samples with high dsdna/low ssdna were not identified.

3 144 Annals of Clinical & Laboratory Science Table 1. Measurements of intra-assay (n = 8) and inter-assay (n = 8) precision of enzyme immunoassays (EIA) for dsdna and ss DNA at 3 decision levels, using kits from Helix Diagnostics. The data are reported as IU/ml (mean ± SD; CV). Assay Serum Intra-assay Inter-assay specimen precision precision dsdna Level 1 36±3.5 (9.8%) 36±6.3 (17.5%) Level 2 98±4.4 (4.5%) 90±5.9 (6.6%) Level 3 245±5.8 (2.4%) 207±20.9 (10.1%) ssdna Level 1 40±1.4 (3.5%) 48±8.2 (17.0%) Level 2 190±4.8 (2.5%) 193±12.7 (6.6%) Level 3 283±10.3 (3.7%) 263±21.5 (8.2%) Table 2. Comparisons of enzyme immunoassays (EIA) for dsdna and ssdna in 139 serum specimens using kits from Helix Diagnostics. The number of serum specimens that tested positive for dsdna by the Crithidia immunofluorescence assay (IFA) are listed in parentheses. ssdna (U/ml) ds DNA (IU/ml) by EIA by EIA < >200 <30 64 (0) 0 (0) 0 (0) 0 (0) (1) 3 (0) 1 (0) 0 (0) (0) 2 (1) 5 (3) 0 (0) > (0) 0 (0) 4 (4) 9 (8) Totals: 115 (1) 5 (1) 10 (7) 9 (8) Positive by Crithidia IFA: 1% 20% 70% 89% Table 3. Comparisons of the results of enzyme immunoassays (EIA ) for dsdna and ss DNA with the results of Crithidia immunofluorescence assay (IFA) for dsdna. dsdna (low Mw DNA) EIA (kit #1) pos neg Crithidia IFA pos 16 1 neg dsdna (high Mw DNA) EIA (kit #2) Crithidia IFA pos 15 2 neg ssdna EIA Crithidia IFA pos 1 16 neg Excellent linearity (>200 IU/ml) was found for samples 1 and 2 by both EIAs, with correlation coefficients (r) from Virtually no immunoreactivity of the low dsdna/high ssdna sample (sample 2) was detected with the antidsdna EIA kit (kit #1) at any of the dilutions. Correlation studies. Anti-dsDNA antibody levels obtained with kit #1 showed weakly positive correlation (r = 0.493) with anti-ssdna antibody concentration (Fig. 2A). There were a large number of highly positive anti-ssdna specimens that contained little anti-dsdna antibody. When antidsdna results >30 IU/ml were compared with the anti-ssdna antibody results, the correlation increased substantially (r = 0.741) (Fig. 2B). EIA comparison with IFA. A substantial number of patient specimens submitted for anti-dsdna antibody determination were positive for antissdna (54%, 79/139), versus anti-dsdna (17%, 16/139), using the 30 IU/ml cut-off (Table 2). No specimens with high dsdna and low ssdna antibodies were detected. Results by the Crithidia IFA correlated positively with increased anti-dsdna antibody concentration (Table 2). Of the 115 specimens with an anti-dsdna antibody <30 IU/ ml (ie, negative), only one was positive by the IFA using Crithidia. Levels of anti-ssdna antibody did not correlate with the IFA results. Of 51 anti-ssdna antibody samples that were >30 U/ml, only one specimen was tested positive by IFA. IFA positivity rate in sera from females (16/124, 12.9%) was approximately double that in sera from males (1/ 15, 6.6%). Because of the small number of male patients, the statistical significance of this finding is unclear. However, the overall distribution of antidsdna antibody levels was comparable in males vs females, respectively: <30 IU/ml to IU/ml range (86.7% vs 86.3%); to >200 IU/ml range (13.3% vs 13.7%). Positivity rates. Only one of the 17 positive specimens by the Crithidia IFA was falsely negative by the anti-dsdna antibody EIA (kit #1) (Table 3). Of 8 false-positive anti-dsdna results, 7 were from specimens with anti-dsdna levels in the low

4 Enzyme immunoassays for dsdna and ssdna 145 Fig. 1. Studies of the linearity of enzyme immunoassays (EIAs) of dsdna and ssdna. Linearity was assessed by serial dilutions of 2 patient samples containing high dsdna/high ssdna (sample 1) and low dsdna/high ssdna (sample 2). The samples were analyzed by dsdna (A,C) and ssdna (B,D) EIAs. Relative dilutions (1.0 = no dilution) and correlation coefficients (r = 1.000, perfect correlation) are shown. Table 4. The relative sensitivity, specificity, and efficiency of enzyme EIAs for dsdna and ssdna, compared to the results obtained by the Crithidia immunofluorescence assay (IFA). Parameter dsdna EIA (low Mw, kit #1) dsdna EIA (high MW, kit #2) ss DNA Relative sensitivitiy 94.1% 88.2% 5.8% Relative specificity 93.4% 72.6% 46.4% Relative efficiency 93.5% 69.8% 46.8%

5 146 Annals of Clinical & Laboratory Science Fig. 2. Correlation charts for the results of EIAs for dsdna and ssdna in the entire group of 139 serum samples (panel A) and for samples above the 30 IU/ml cutoff for positivity (panel B). Correlation coefficients (r) are indicated (r = 1.000, perfect correlation). Fig. 3. Correlation charts for the results of EIAs for dsdna using high Mw DNA (kit #2) versus low Mw DNA (kit #1; panel A) and versus the results of EIA for ssdna (panel B). Correlation coefficients (r) are indicated (r = 1.000, perfect correlation). positive to positive range. Similar false-negative rates were obtained using the high Mw DNA anti-dsdna antibody kit (kit #2), ie, only 2 false-negative results were obtained. However, the number of falsepositive specimens increased substantially, ie, to 31, when kit #2 was used. As expected, the results of the anti-ssdna antibody test did not correlate with those of the Crithidia IFA. Relative sensitivity and specificity. The highest relative sensitivity and specificity were found using the anti-dsdna kit containing immobilized low Mw DNA (kit #1) (Table 4). Lower values for relative sensitivity and specificity were obtained with the high Mw DNA kit (kit #2). The relative efficiency of kit #2 (<70%) was substantially lower than kit #1 (>90%). In general, results obtained

6 Enzyme immunoassays for dsdna and ssdna 147 with the anti-ssdna antibody EIA kit showed poor relative sensitivity and specificity in comparison to the Crithidia IFA method. Correlation between anti-dsdna EIA methods. Overall, results obtained with the two anti-dsdna antibody kits (kit #1 and kit #) correlated well with each other (r = 0.834) (Fig. 3A). Due to apparent immunoreactivity with anti-ssdna antibodies, the results obtained with kit #2 correlated substantially better with the anti-ssdna antibody kit (Fig. 3B), compared to the anti-dsdna kit that contained low Mw DNA (kit #1) (see Fig. 2). Discussion Detection of anti-dsdna antibodies is important for the diagnosis of systemic lupus erythematosus [2,13] and may be of use for monitoring disease activity, including exacerbations [13,15,21-25]. Numerous EIA kits are commercially available, but the lack of consistent assay performance among manufacturers [19,20] has sustained the continued use of traditional methods of anti-dsdna antibody detection such as IFA, although IFA is semiquantitative at best [9]. More quantitative methods (eg, EIA) may be compromised by the integrity of DNA immobilized in the microtiter wells. The reactivities of anti-dsdna antibodies (specific for SLE) and anti-ssdna antibodies (nonspecific for SLE) appear to be especially problematic for these EIAs, due to the apparent technical complexity of preparing intact double-stranded DNA, especially that of high Mw from calf thymus [10-12]. This study evaluated the analytical performance of an anti-dsdna EIA kit that uses a combination of plasmid and bacteriophage lambda DNA as a low Mw nucleic acid source (kit #1). EIA results were compared to those by an indirect immunofluorescence assay (IFA) using Crithidia luciliae, a highly specific assay for anti-dsdna antibodies [3,4]. The analytical sensitivity and specificity were compared to those obtained with a anti-dsdna EIA kit that uses immobilized high Mw DNA (kit #2). Discrepancies in analytical performance between kit #1 and kit #2 were examined by testing for the presence of anti-ssdna antibodies. Detection of anti-dsdna antibodies by the EIA composed of immobilized low Mw DNA (kit #1) showed high sensitivity and specificity compared to the Crithidia IFA. Virtually no immunoreactivity was detected in EIA kit #1 using samples containing high levels of anti-ssdna antibodies (crossreactivity <2% at ssdna 600 U/mL). In contrast, lower relative sensitivity and specificity were obtained with an EIA that uses immobilized high Mw DNA obtained from calf thymus (kit #2). Discrepancies in analytical performance obtained using kit #2 were attributable to a higher false positive rate from specimens containing only anti-ssdna antibodies. The high relative sensitivity and specificity for antidsdna antibodies in kit #1 resulted in excellent relative efficiency (>90%) in comparison to the Crithidia IFA method. This study demonstrates that the use of an antidsdna EIA kit containing immobilized low Mw DNA gives excellent relative sensitivity and specificity in comparison to the immunofluorescence antibody assay using Crithidia luciliae. Decreased assay performance with an EIA kit containing immobilized high Mw DNA was most likely attributable to immunoreactivity with anti-ssdna antibodies, although this could not be determined unambiguously. Although use of low Mw DNA in microtiter wells improves the specific detection of anti-dsdna, it is no guarantee of high specificity [19]. Because of this uncertainty, it has been recommended that anti-dsdna kits be evaluated for non-anti-ssdna antibody binding [26]. The Arthritis Foundation/Centers for Disease Control and Prevention has a reference antiserum for antidsdna (CDC1), but this preparation also contains anti-ssdna antibodies [19]. An appropriate reference antiserum containing only anti-ssdna antibodies is unavailable from this source. However, samples containing only anti-ssdna can readily be obtained as described in this study. Such samples may be aliquoted and stored frozen as reference materials to validate assay methods. As found in this study, serum containing only anti-dsdna antibodies is very rare [15] and therefore unlikely to serve as an appropriate reference material.

7 148 Annals of Clinical & Laboratory Science Acknowledgements The authors have no conflicts of interest in respect to this investigation. References 1. Tan EM, Schur PH, Carr RI, Kunkel HG. Deoxyribonucleic acid (DNA) and antibodies to DNA in serum of patients with systemic lupus erythematosus. J Clin Invest 1966;45: Tan EM, Cohen AS, Fries JF, Masi AT, McShane DJ, Rothfield NF, Schaller JG, Talal N, Winchester RJ. The 1982 revised criteria for the classification of systemic lupus erythematosus. Arth Rheum 1982;25: Pincus T, Schur PH, Rose JA, Decker JL, Talal N. Measurement of serum DNA-binding activity in systemic lupus erythematosus. New Engl J Med 1969;281: Aarden LA, De Groot ER, Feltkamp TEW. Immunology of DNA. III. Crithidia luciliae, a simple substrate for the determination of anti-dsdna with the immunofluorescence technique. Ann NY Acad Sci 1975;254: Eaton RB, Schnneider G, Schur PH. Enzyme immunoassay for antibodies to native DNA. Specificity and quality of antibodies. Arth Rheum 1983;26: Klotz JL, Minami RM, Teplitz RL. An enzyme-linked immunosorbent assay for antibodies to native and denatured DNA. J Immunol Meth 1979;29: Miller TE, Lahita RG, Zarro VJ, MacWilliam J, Koffler D. Clinical significance of anti-double stranded DNA antibodies detected by a solid phase enzyme immunoassay. Arth Rheum 1981;24: Emlen W, Jarusiripipat P, Burdick G. A new ELISA for detection of double-stranded DNA antibodies. J Immunol Meth 1990;132: S2 Immunology (Special), Participant Summary Report, Surveys 2002, College of American Pathologists, Northfield, IL. 10. Engvall E. Determination of antibodies to DNA by ELISA. Lancet 1976;2: Pyeritz RE, Schlegel RA, Thomas CA Jr. Hydrodynamic shear breakage of DNA may produce single-stranded terminals. Biochim Biophys Acta 1972;272: Locker JD, Medof ME, Bennett RM, Sukhupunyaraksa S. Characterization of DNA used to assay for anti-dna antibodies; determination of the specificities on anti-dna antibodies in SLE and nonsle rheumatic disease states. J Immunol 1977;118: Kavanaugh A, Tomar R, Reveille J, Solomon DH, Homburger HA. Guidelines for clinical use of the antinuclear antibody test and tests for specific autoantibodies to nuclear antigens. Arch Pathol Lab Med 2000;124: Hahn BH. Antibodies to DNA. N Engl J Med 1998;338: Pisetsky DS. Anti-DNA antibodies in systemic lupus erythematosus. Rheum Dis Clin North Amer 1992;18: Tan EM, Fritzler MJ, McDougal JS, McDuffie FC, Nakamura RM, Reichlin M, Reimer CB, Sharp GC, Schur PH, Wilson MR, Winchester RJ. Reference sera for antinuclear antibodies. I. Antibodies to native DNA, Sm, nuclear RNP, and SS-B/La. Arth Rheum 1982;25: Feltkamp TEW, Kirkwood TBL, Maini RN, Aarden LA. The first international standard for antibodies to double stranded DNA. Ann Rheum Dis 1988;47: Smolen JS, Steiner G, Tan EM. Standards of care: the value and importance of standardization. Arth Rheum 1997;40: Tan EM, Smolen JS, McDougal JS, Butcher BT, Conn D, Dawkins R, Fritzler MJ, Gordon T, Hardin JA, Kalden JR, Lahita RG, Maini RN, Rothfield NF, Smeenk R, Takasaki Y, van Venroij WJ, Wiik A, Wilson M, Koziol JA. A critical evaluation of enzyme immunoassays for detection of antinuclear antibodies of defined specificities. I. Precision, sensitivity, and specificity. Arth Rheum 1999;42: Tan EM, Smolen JS, McDougal JS, Fritzler MJ, Gordon T, Hardin JA, Kalden JR, Lahita RG, Maini RN, Reeves WH, Rothfield NF, Takasaki Y, Wiik A, Wilson M, Koziol JA. A critical evaluation of enzyme immunoassays for detection of antinuclear antibodies of defined specificities. II. Potential for quantitation of antibody content. J Rheumatol 2002;29: Kavanaugh A. The utility of immunologic laboratory tests in patients with rheumatic diseases. Arth Rheum 2001;44: ter Borg EJ, Horst G, Hummel EJ, Limburg PC, Kallenberg CGM. Measurement of increases in antidouble-stranded DNA antibody levels as a predictor of disease exacerbation in systemic lupus erythematosus. A long-term, prospective study. Arth Rheum 1989;33: Esdaile JM, Abrahamowicz M, Joseph L, MacKenzie T, Li Y, Danoff D. Laboratory tests as predictors of disease exacerbations in systemic lupus erythematosus. Why some tests fail. Arth Rheum 1996;39: Okamura M, Kanayama Y, Amastu K, Negoro N, Kohda S, Takeda T, Inoue T. Significance of enzyme linked immunosorbent assay (ELISA) for antibodies to double stranded and single stranded DNA in patients with lupus nephritis: correlation with severity of renal histology. Ann Rheum Dis 1993;52: Ho A, Magder LS, Barr SG, Petri M. Decreases in antidouble-stranded DNA levels are associated with concurrent flares in patients with systemic lupus erythematosus. Arth Rheum 2001;44: Burlingame RW, Rubin RL. Enzyme-linked immunosorbent assays for diagnostically important antinuclear antibodies. In: Manual of Clinical Laboratory Immunology, 6th ed (Rose NR, Hamilton RG, Detrick B, Eds), ASM Press, Washington, DC, 2002; pp