The detection and identification of antibodies to saline extractable nuclear antigens by counterimmunoelectrophoresis.код для вставкиСкачать
1026 THE DETECTION AND IDENTIFICATION OF ANTIBODIES TO SALINE EXTRACTABLE NUCLEAR ANTIGENS BY COUNTERIMMUNOELECTROPHORESIS HAROLD D. KEISER and JOSHUA WEINSTEIN Sera containing antibodies to the saline extractable nuclear antigens RNP, Sm, and Ha can be identified with counterimmunoelectrophoresis (CIE) by matching their precipitin reactions to those of reference sera. The CIE reaction of Sm is unusual in that two precipitin lines are formed, one of which shows partial immunologic identity with the reaction to RNP. Sera were identified by CIE which reacted with additional acidic antigens in rabbit thymus extract, some of which had the same enzyme sensitivity profiles as RNP and Ha but were immunologically not identical and could not be detected by hemagglutination with red cells coated with calf thymus nuclear extract. Although the incidence of reactivity to RNP, Sm, and Ha was similar to previous reports, review of the clinical manifestations of 128 patients with systemic lupus erythematosus (SLE) did not provide statistically significant confirmation of previously reported correlations between the presence and specificity of reactivity and particular patterns of disease expression. With due attention paid to technical artifacts, CIE is a sensitive and specific method for identifying sera reactive with acidic nuclear antigens that has practical advantages over alternative methods. From the Department of Medicine, Albert Einstein College of Medicine, Bronx, New York. Supported by research awards from the S.L.E. Foundation of America and Research Career Development Award AM00024 of the National Institutes of Health (H.D.K.). Harold D. Keiser, MD: Associate Professor of Medicine; Joshua Weinstein, MD: Assistant Professor of Medicine, Albert Einstein College of Medicine. Address reprint requests to Harold D. Keiser, MD, Department of Medicine, Albert Einstein College of Medicine, 1300 Morris Park Avenue, Bronx, New York 10461. Submitted for publication September 21, 1979; accepted in revised form February 19, 1980. Arthritis and Rheumatism, Vol. 23, No. 9 (September 1980) Many of the antinuclear antibodies in the sera of patients with connective tissue diseases react with components of mammalian cell nuclei which are solubilized by normal saline solutions. These antibodies are of considerable interest because they comprise as much as 20% of total serum immunoglobulin in some patients (l), and because antibodies to specific saline-extractable nuclear antigens (ENA) have been associated with disease subgroups with characteristic clinical features and prognoses (2-4). Thus, the presence of antibodies to the nuclear antigen RNP is said to be the hallmark of a group of patients with overlapping features of systemic lupus erythematosus (SLE) and another rheumatic disease who have a low incidence of renal abnormality and a good prognosis (5,6). Antibodies to the nuclear antigen Sm seem to be specific for patients with SLE (3,4,7), and antibodies to the nuclear antigen Ha-also termed SS-B or La ( 8 b a r e reported to occur with high frequency in patients with idiopathic or SLE-associated Sjogren’s syndrome who may be at greater risk of disease progression or of developing pseudolymphoma (9). The methods most commonly employed in the detection and identification of antibodies to ENA, agarose double immunodiffusion and hemagglutination, are not entirely satisfactory. Immunodiffusion is a relatively insensitive technique, and it is usually necessary to test multiple serum dilutions to be certain of including the proper ratio for antigen-antibody precipitation. Hemagglutination is a highly sensitive technique, but Ha, and possibly other ENA, cannot be detected by hemagglutination (lo), and it may not be possible to distinguish among the multiple ribonuclease-sensitive or insensitive ENA which could be responsible for a hemagglutination reaction. The studies reported here indicate that counter- ANTIBODIES TO SALINE ENA immunoelectrophoresis (CIE), as described by Kurata and Tan (1 l), may be a more practical method for detecting antibodies to ENA. With proper placement of wells and recognition of anomolous or nonimmunologic reactions, sera containing antibodies to RNP, Sm, Ha, and other ENA, singly or in combination, can be readily and accurately identified by CIE. MATERIALS AND METHODS Sera. The serum samples examined for antibodies to ENA were originally submitted to this laboratory for serum hemolytic complement and anti-native DNA antibody level determinations. Only sera of the patients known to have antinuclear antibodies were tested for antibodies to ENA, and when multiple samples were available from a patient, the sample examined was chosen from a period of active disease and/or depressed serum complement. Sera were stored at -20°C and routinely heated at 56OC for 1 hour prior to testing, although the testing of a dozen serum samples in parallel suggested that heat inactivation may not be necessary. Reference sera for RNP, Sm, and Ha were provided by Dr. Halstead R. Holman, and a reference serum for Ro was provided by Dr. Peter Maddison. Counterimmunoelectrophoresis. CIE was carried out essentially as described by Kurata and Tan (ll), the major modification being the arrangement of the wells. The gel cutter consisted of two parallel rows of punches 4 mm in outside diameter separated by 5 mm from outer edge to outer edge, with the punches in each vertical row approximately 0.5 mm apart. The row of wells on the left (cathodal) side was filled with antigen solution and the row on the right side (anodal) with sera. Electrophoresis was performed in a Durrum type apparatus in 0.05M barbital buffer, pH 8.6, with a constant current of 4 mamp per slide and was stopped when the orange G marker dye had migrated approximately 15 mm. This took 1 to 1% hours, which is somewhat longer than reported by Kurata and Tan (11). Slides were read after overnight incubation in a moist chamber at room temperature. Washing the slides for several days with cold 5% sodium citrate or normal saline occasionally improved the definition of precipitin 1027 lines, but subsequent drying and staining with Coomassie blue were not helpful. Hemagglutination. Passive hemagglutination with sheep red cells sensitized with saline extract of calf thymus nuclei was performed by Dr. Peter Barland (7). Nuclear antigens. A commercial preparation of lyophilized rabbit thymus extract (Pel-Freeze Biologicals, Rogers, Arkansas) was used as the source of saline-extractable nuclear antigens. Extraction and enzyme treatment were performed as described by Kurata and Tan (1 1); digestion with trypsin (DCC-treated, type XI, Sigma Chemical Co., St. Louis, Missouri), not described by these authors, was performed at a 1 :25 ratio of enzyme to substrate with incubation in a 1 : 1 volume mixture for 30 minutes at 37°C. Extracts and digests were stored in aliquots at -20°C for as long as 4 months with no apparent loss of activity. A 0.35M NaCl extract of purified rat liver chmmatin was provided by Dr. Irving Listowsky. Clinical data. A detailed review was performed of the clinical features of the 106 of 149 patients with ENA-positive sera and of the 41 SLE patients with ENA-negative sera whose medical records were available for analysis. The diagnosis of SLE was established on the basis of the presence of four provisional American Rheumatism Association criteria (12) or by the presence of three criteria and one of the following: 1) a positive LE band test (i.e. dermal-epidermal deposits of immunoglobulins and complement components in clinically uninvolved skin); 2) elevated levels of antibodies to native DNA by the Millipore filter assay (13) with Iz5I-labeled filtered DNA (14); or 3) classic electron microscopic and immunofluorescence findings on renal biopsy. The diagnoses of dermato-polymyositis and scleroderma were made on the basis of the presence of the usual clinical features (15). Patients who met the diagnostic criteria for SLE and in addition had biopsy, electromyographic, or enzyme evidence of myositis or who had clinical, biopsy, or radiographic evidence of scleroderma were considered to have overlapping SLE-polymyositis or SLE-scleroderma respectively. Statistics. The statistical significance of the differences in incidence of selected clinical parameters between specific serologically defined groups and all other SLE patients studied was determined by the 2 test. Figure 1. Counterimmunoelectrophoresiswith rabbit thymus extract (TE) and ribonuclease (RNAse), deoxyribonuclease (DNAse), and trypsin (Try) digests of thymus extract in the sets of wells on the left (cathodal) and standard reference sera for RNP (a), Sm (b), and Ha (c) in the sets of wells on the right (anodal). KEISER AND WEINSTEIN 1028 b o lo TE TE 0 1 0 !O 0 13 0 U Figure 2. Counterimmunoelectrophoresiswith thymus extract or its digests in the anodal sets of wells as denoted in Figure 1 and patient sera in the cathodal wells. Sera are denoted by their specificity as determined by enzymatic sensitivity and immunologic matching with reference sera, with the individual patient’s initials as subscript (e.g., RNPG, denotes serum with RNP reactivity obtained from patient GP). RESULTS RNP, Ha, Sm. Sera known to contain antinuclear antibodies were reacted by CIE with a saline extract of rabbit thymus and with extract treated with ribonuclease, deoxyribonuclease, or trypsin. The CIE results using standard sera for RNP, Sm, and Ha obtained from Dr. H. R. Holman are shown in Figure 1. Antibodies to RNP form a single sharp precipitin line which is weak or absent after digestion of the thymus extract with ribonuclease or trypsin. Antibodies to Ha also form a single precipitin line, but this reaction is not affected by treatment of the thymus extract with any of the enzymes used. By contrast, antibodies to Sm usually produce both a sharp precipitin line and, adjacent to it on the side nearer to the antigen well, a broad fuzzy band or line. The sharper precipitin line of Sm is weak or absent with ribonuclease-treated thymus extract, but it persists and even appears enhanced after trypsin digestion of the thymus extract. Sera reactive with RNP, Ha, or Sm can be definitively identified by CIE on the basis of the relationship between their precipitin bands and those of standard sera run in adjacent sets of wells. Thus, it can be appreciated that serum in the middle well in Figure 2A con- tains antibodies to both RNP and Ha, inasmuch as the precipitin line nearer to the antigen well merges smoothly with the line formed by a standard anti-RNP serum, and the precipitin line nearer to the serum well merges smoothly with the line formed by a standard anti-Ha serum. Crossing-over or wide divergence of precipitin lines, as with those formed by the reactions to Ha and RNP or Sm in adjacent sets of wells (Figure 2B, top and bottom), are indicative of immunologic nonidentity. The immunologic relationship of RNP and Sm as shown by CIE is complex (Figure 2B, middle). Both of the precipitin lines of Sm cross over the precipitin line of RNP, but the RNP precipitin line crosses over only the less distinct Sm precipitin line nearer to the antigen well and appears to merge with the sharper Sm precipitin line. The resultant spurring of the sharper Sm precipitin line over the RNP precipitin line indicates partial immunologic identity. Sera that contain antibodies to both Sm and RNP produce two precipitin lines upon reaction with thymus extract (Figure 2C). A standardized RNP precipitin line formed by the reaction in an adjacent set of wells merges with the precipitin line nearer to the antigen well (Figure 2D); both precipitin lines appear to merge with the sharper precipitin ANTIBODIES TO SALINE ENA 1029 Figure 3. Countetimmunoelectrophoresis with thymus extract or its digests in the anodal sets of wells and patient sera in the cathodal wells. Notations are the same as in Figures 1 and 2. line formed by the reaction to Sm in an adjacent set of wells (Figure 2D). Additional acidic nuclear antigens. Over the course of these studies, two or more sera were identified that reacted by CIE with four additional components of the thymus extract: 1) RNP-2, a ribonuclease and trypsin sensitive antigen which is immunologically distinct from RNP (Figures 3A and B); 2) P-2, which is resistant to digestion with ribonuclease, deoxyribonuclease, or trypsin but is immunologically distinct from Ha (Figure 3C and D); 3) “DNA,” a reaction sensitive to digestion with deoxyribonuclease and detected thus far only in sera with high titers of antibodies to native DNA (Figure 3E); and 4) P-3, which, unlike the other antigens described, is sensitive to trypsin but resistant to ribonuclease (Figure 3F). Several additional antigens have been detected in the thymus extract which have thus far been found to react with only a single serum. The incidence of reactivity to the rabbit thymus extract and to each of its recognized antigens, as determined by CIE, is shown in Table 1. Of interest is the frequency with which antibodies to two ENA were identified by CIE in the same sera (Table 1, bottom). Thirty percent of sera with anti-RNP and 49 percent of sera with anti-Sm also reacted with another ENA, the most common combination being reactivity to both RNP and Sm. With the exception of the sera reactive with DNA and P-3, which were not tested, each of the antigens found by CIE with rabbit thymus extract were also found with a 0.35M NaCl extract of purified rat liver chromatin. These antigens were found in the precipitate formed by the thymus extract in 2% TCA, indicating that they are components of the poorly characterized LMG fraction of chromatin (16). A serum reactive with Table 1. Incidence of ENA reactivities in ANA positive sera ENA reactivity RNP* smt Ha RNP-2 “DNA P-2 P-3 Unmatched Number reactive sera 87 37 22 10 6 4 2 12 9’0 ENA positive sera (total = 149) 58 25 15 7 4 3 1 8 % ANA positive sera (total = 357) 24 10 6 3 2 1 <I 3 * RNP + another ENA:26 (Sm, 12; Ha, 6; RNP-2, 6; unmatched, 2). t Sm + another ENA:20 (RNP, 12; Ha, 2; RNP-2, 2; P-2, 1; unmatched, 3). KEISER AND WEINSTEIN 1030 Figure 4. Counterimmunoelectrophoresiswith empty anodal wells (a) or with thymus extract and its digests in the anodal wells @-d) and patient sera in the cathodal wells. Sera non-reactive with thymus extract are denoted by ENA-,with individual sera identified by the patient’s initials in the subscript. The well labeled Hacs-SP (b, bottom) was deliberately overfilled to allow the serum to spill over the well edge. Other notations are the same as in previous figures. the cytoplasmic antigen Rc-equivalent to SS-A (8)gave no CIE reaction with the rabbit thymus extract. Problems of CIE interpretation. In interpreting the results of CIE, care must be taken to avoid certain common artifacts. Arcs surrounding the serum well on the side nearer the antigen are frequently noted and do not represent an immunologic reaction inasmuch as they can be found after electrophoresisin the absence of thymus extract (Figure 4A). Overfilling of a serum well with consequent spillage into the wells above or below can produce a spurious precipitin line near those wells, as shown in Figure 4B. A large, ill-defined precipitin reaction between serum and trypsin-digested thymus extract may occur even with sera that do not react with the untreated thymus extract, especially after prolonged incubation or washing of the CIE plate (Figure 4C). Finally, the finding of an indistinct precipitin band, delayed in its appearance relative to most of the other precipitin lines, is not a distinguishing characteristic of Sm, as has been maintained previously (1 1). Similar blurred and delayed reactions occur with other thymus antigens and individual sera, and different Sm sera produce reactions which vary markedly in intensity and in the extent to which they are distinct from the sharp precipitin line of Sm (Figure 4D). Comparison with hemagglutination. Sera reactive with each of the antigenic components of rabbit thymus extract identified above, with the exception of anti-P-3, were tested for anti-ENA antibodies by the hemagglutination technique with calf thymus extract-sensitized red blood cells. As indicated in Table 2, only the sera reactive with RNP or Sm were identified by hemagglutination and, with the limited number of successive dilutions used, sera with antibodies to both RNP and Sm, as determined by CIE, were identified by hemagglutination as reacting only with Sm. Table 2. Comparison of ENA determination by CIE and hemagglutination Patient CIE result Ha titer GP SA MM MS AMF BW JO RNP Sm Ha P-2 “DNA” RNP-2 RNP + Sm 1 :65,500 RNP 1:65,500 Sm 0 0 0 0 1:65,000Sm ANTIBODIES TO SALINE ENA 103 1 Table 3. Diagnoses of ENA-positive patients reviewed* ENA reactivity SLE (75) RNP, total RNP only Sm, total Sm only Ha RNP-2 P-2 P-3 “DNA’ Unmatched 44 29 23 9 12 7 3 1 3 I SLE/ PM (6) SLE/ PSS (6) 3 2 3 2 0 0 0 0 1 0 5 1 3 0 PM (5) PSS (4) Misc. (10) 4 3 0 0 3 3 0 0 0 0 0 0 0 7 2 2 I I 0 0 1 0 4 0 1 1 0 0 0 0 1 0 0 0 1 1 * SLE = systemic lupus erythematosus; PM = polymyositis; PSS = progressive systemic sclerosis; Misc. = miscellaneous. Clinical correlations. Sufficient clinical information was available to substantiate the diagnosis in 106 of the patients with antibodies to ENA. Seventy-five of these patients had SLE and an additional 12 had SLE with features of either polymyositis or scleroderma, 5 patients had polymyositis, 4 had scleroderma, 3 had undefinable connective tissue disorders, and 1 patient each had tuberculosis, lymphoma, chronic active hepatitis, essential mixed cryoglobulinemia, idiopathic Sjogren’s syndrome, rheumatoid arthritis, and Takayasu’s arteritis. As may be seen in Table 3, antibodies to RNP were the most common and the least diagnostically specific. Antibodies to Sm were found in 29 SLE patients, in 2 patients with probable rheumatoid arthritis, and in a patient with tuberculosis, antinuclear antibodies which were possibly drug-induced, and no lupus-like manifestations. In view of the relatively small number of positive sera, the diagnostic specificity of antibodies to ENA other than RNP and Sm is unclear. A detailed analysis was performed of the clinical, laboratory, and renal biopsy findings in the 87 SLE patients with and the 41 SLE patients without antibodies to ENA whose medical records were sufficiently complete. As shown in Tables 4, 5, and 6, the various serologically defined groups were quite similar with respect to the incidence of the major clinical and laboratory features of SLE. Mortality was significantly increased in our SLE patients with antibodies to Ha, and fewer patients with anti-Sm had high levels of antibodies to native DNA. Clinical features overlapping with other connective tissue disorders, i.e. Raynaud‘s phenomenon, scleroderma-like skin changes, or myositis, were more common in patients with anti-RNP, but the difference in incidence did not’ reach the level of statistical significance. The increased incidence of overlap features was significant only for the subgroup of patients not listed in Table 4 with antibodies to both RNP and Sm (P < 0.05). Finally, there was a tendency toward more benign pathologic findings on light microscopic examination of renal biopsy specimens from patients with anti-RNP, but the sample size was too small for the statistical test to be valid. Furthermore, the incidence of clinically apparent renal abnormality and the incidence of subendothelial electron-dense deposits in biopsy specimens were not significantly different from those found in SLE patients without anti-RNP. DISCUSSION The studies reported here indicate that the identification of antibodies to specific ENA, present alone or in combination, can be conveniently carried out with CIE by first establishing the enzyme sensitivity of the reaction and then matching the reaction with those formed by standard antisera to the limited number of antigens with the same enzyme sensitivity profile. With due attention paid to the artifacts created by overfilled wells and non-immunologic reactions (Figure 4), CIE has distinct advantages over the other available meth- Table 4. Demographic and clinical features in patients with systemic lupus erythematosus correlated with ENA reactivity Demographic features Clinical features* ENA reactivity No. patients Age, average (range) Years duration, average (range) Deaths Negative RNP, total RNPonly Sm, total Sm only Ha RNP-2 41 52 32 29 11 12 8 38(13-60) 41 (18-78) 40(18-65) 41 (16-76) 39(21-54) 38(16-62) 41 (28-78) 9.0(2-23) 8.7(2-24) 9.3 (2-24) 7.8(2-21) 8.6 (2-20) 6.0(1-13) 11.3(420) 4(10) 5(10) 2 ( 6) 3(10) 0 4(33)j 0 * Number of patients; percentages shown in parentheses. jP < 0.05. CNS Dermal Raynaud’s Sicca involvement vasculitis phenomenon syndrome 16(39) 18(35) 12(38) 13(45) 4(36) 6(50) 3 (38) lO(24) 21 (40) lO(31) 13(45) 6 (55) 8(67) 6(75) I 1 (27) 25 (48) 14(44) 13(45) 4(36) 4(33) 3(38) 4(10) 4 ( 8) 0 1 ( 3) 0 1 ( 8) l(13) Myositis 4(10) 3 ( 6) 2(6) 3(10) 2(18) 0 0 Scleroderma 0 5(10) 1(3) 3(10) 0 0 l(13) KEISER AND WEINSTEIN 1032 Table 5. Laboratory findings in patients with SLE correlated with ENA reactivity Laboratory findings; ENA reactivity No. patients CH50 580 (nl > 150) AntinDNA Platelets 40,000 Negative RNP, total RNP only Sm, total Sm only Ha RNP-2 41 52 32 29 11 12 8 22(54) 22(42) 14(44) 14(48) 3(29) 6(50) 5 (63) 28(68) 26(50) 16(50) 11 (38)t 2(18) 5 (42) 6(75) 15(37) 21 (40) I 1 (34) 12(41) 2(18) 7(58) 3 (33) Coombs Proteinuria (>350 mg/24 hr) Azotemia (Cr > 1.5 mg/dl) 11 (27) 24(46) 13 (41) 8(28) 2(18) 7(58) 4(50) 16(39) 21 (40) 13(41) 12(41) 3(27) 6(50) 3 (38) lO(24) 14(27) S(25) 7(24) 0 6(50) 2(25) + * Numbers of patients; percentages shown in parentheses. t P < 0.05. ods for the routine detection and identification of antibodies to ENA. Anionic antigens, such as the major ENA, can be detected with greater sensitivity by CIE than by Ouchterlony agarose gel immunodiffusion, because in CIE antigen and antibody do not diffuse randomly but are directed toward each other by the electrical field. With appropriately spaced wells, the immunologic relationships among various antigens can be determined just as in Ouchterlony immunodiffusion. Hemagglutination is a.more sensitive technique than CIE and is more easily quantitated. However, our results indicate that antibodies to antigens other than RNP and Sm which are detectable in rabbit thymus extract and in a purified extract of rat liver chromatin by CIE are not detected by hemagglutination with red cells sensitized with calf thymus nuclear extract (Table 2). Furthermore, the hemagglutination assay as conventionally performed may fail to detect the presence of antibodies to RNP in sera which also contain antibodies to Sm. This can occur because in the hemagglutination assay, antibodies to RNP and Sm are distinguished by comparing the result obtained from red cells coated with ribonuclease-treated antigen to that obtained from red cells coated with untreated antigen; similar hemagglutination titers indicate the presence of anti-Sm, and a much lower titer with ribonuclease-treated antigen indicates the presence of anti-RNP. If serum dilutions are arbitrarily limited to 1:65,000 or 1:100,000, as is generally the case, rather than continued to the several million-fold dilution often necessary to reach a negative end point (9,the drop in hemagglutination titer with ribonuclease-treated antigen indicative of the presence of anti-RNP will not be evident should the serum also contain antibodies to Sm in titers above the arbitrarily set level. Our findings with respect to the enzyme sensitivities of major ENA are not entirely in accord with previous reports. We found Ha to be resistant to treatment with trypsin, even when using a standard anti-Ha serum provided to us by investigators who described Ha as trypsin-sensitive (10). Different conditions of digestion or our use of a chymotrypsin inhibitor-treated trypsin preparation may be responsible for this discrepancy. Sm has not previously been noted to be ribonuclease-sensi- Table 6. Renal biopsy findings in patients with SLE correlated with ENA reactivity Renal biopsy findings* Electron microscopic examination Light microscopic examination ENA reactivity No. patients biopsied Negative RNP, total RNP only Sm, total Sm only Ha RNP-2 20 20 11 13 2 8 3 WNLor mesangial 3 (15) 16 (30) 2(18) 4 (31) 0 (13) 2 (67) Membranous + Membranous proliferative 2 (10) 4 (20) 2 (18) 3 (23) 2 (100) 2 (25) 0 3 (15) 4 (20) 2 (18) 2(15) 0 2(25) 1 (33) * Numbers of patients; percentages shown in parentheses. Proliferative 12(60) 6(30) 5 (45) 4(31) 3 (38) 0 SubendoNo. thelial performed deposits 13 15 8 10 2 6 2 7(54) 4(27) 3 (38) 4(40) 1(50) 3 (50) 0 ANTIBODIES TO SALINE ENA tive, but this may not have been appreciated, especially in studies in which hemagglutination or complement fixation is used, because nuclear extracts seem to contain at least two components with Sm reactivity, one sensitive to the action of ribonuclease and one resistant. Findings consistent with the existence of two Sm components, and with the partial identity of one of these components to RNP, have been observed previously, but these findings have been interpreted in different ways. Kurata and Tan found as we have that the reaction of Sm and anti-Sm produced two precipitin lines on CIE, but they attributed only the late-appearing diffuse line to Sm (1 1). Dorsch and coworkers, using Ouchterlony immunodiffusion, identified sera reactive only with Sm on the basis of their producing a single precipitin line upon diffusion against nuclear extract which persisted when ribonuclease-digested extract was used and which showed a reaction of complete identity with a standard anti-Sm serum. However, they noted that this Sm-anti-Sm reaction spurred over the reaction with nuclear extract of sera containing anti-RNP, indicating that there was a relationship of partial immunologic identity between the two systems. Dorsch et a1 concluded that anti-Sm sera must also contain antibodies to RNP (17). Mattioli and Reichlin had previously reported the same immunodiffusion findings, but they offered a different explanation for the spurring of Sm-anti-Sm over RNP-anti-RNP. On the basis of complement fixation studies with absorbed sera, Mattioli and Reichlin concluded that RNP and Sm must exist in nuclear extracts as components of molecular complexes rather than as individual molecular species, that Sm occurs both alone and in combination with RNP, and that RNP occurs only in a complex with Sm (18). The complex nature and relatedness of Sm and RNP postulated by Mattioli and Reichh are supported by the recent report of Lerner and Steitz that the immune precipitates formed by 32Pand 3sS-methioninelabeled nuclear extracts and antibodies to RNP and Sm contain the same seven polypeptides associated with several overlapping species of small nuclear RNA (19). We believe that our findings on the reaction of Sm and anti-Sm and the relationship between Sm and RNP represent a confirmation by CIE of the conclusions reached by Mattioli and Reichlin (18). During electrophoresis of nuclear extract and sera reactive with Sm, antibodies to Sm meet and precipitate with the RNP-Sm molecular complex to form the sharp, rapidly developing, ribonuclease-sensitive precipitin line. Many anti-Sm sera are of high titer (1 1) such that there is a considerable excess of antibody present which may, in 1033 effect, diffuse through this precipitin line to meet and precipitate with the molecular complexes containing Sm only. This second ribonuclease-resistant precipitin line of Sm is delayed and diffuse in its appearance because it arises from subsequent diffusion rather than from the initial electrophoresis. Thus, with many anti-Sm sera, the conditions of CIE may permit the two molecular species of Sm to produce separate precipitin reactions even though only one is evident under the conditions of Ouchterlony immunodiffusion. Rabbit or calf thymus extracts, as well as extracts of purified mammalian chromatin, clearly contain multiple antigens reactive with patients’ sera in addition to RNP, Sm, and Ha. Since different ENA may have similar temperature and enzyme sensitivities, and since these parameters may not be constant from laboratory to laboratory, the only reliable way to determine the specificity of an ENA-reactive serum is by establishing its immunologic relationship to a standard serum (8). Thus, it is likely that some or all of the new antigens identified in this study are identical to nuclear or cytoplasmic antigens previously described by other investigators (20-23), but this can be sorted out only by reacting standard sera together against thymus extract by immunodiffusion or CIE. Additional antigens not present in thymus extracts, some with specificity for different rheumatic diseases, have been found in cytoplasmic extracts, in the nuclei of proliferating cells, and in nuclear extracts from a human B lymphocyte cell line (8,24). One such antigen, Ro (SS-A), could not be detected in rabbit thymus extract by CIE. Use of a Wil, extract in addition to thymus extract in testing patients’ sera by CIE could extend the advantages of this method to the detecting and identification of other acidic antigens with different clinical specificities. The incidence in SLE patients of antibodies to RNP, Sm, and Ha found in this study (41%, 23%, and 9% respectively) is virtually identical to that reported by Kurata and Tan (1 l), who also used CIE, and is substantially higher than that reported in studies with Ouchterlony immunodiffusion or hemagglutination (3,9,17,25,26). Such differences in incidence are to be expected on the basis of the differences in sensitivity and specificity of these methods as discussed above. Our findings confirm previous reports that antibodies to Sm occurred almost exclusively in patients with SLE, whereas the ENA reactivity of sera from patients with diseases other than SLE was most often due to antibodies to RNP (3,4,7). However, a review of the clinical, laboratory, and renal biopsy manifestations of our SLE patients did not clearly support previously suggested as- KEISER AND WEINSTEIN 1034 sociations of ENA specificity and specific patterns of disease expression (2-4). In particular, clinical-serologic correlations such as the increased incidence of overlap features, the low incidence of serious renal disease and the better prognosis associated with antibodies to RNP (2-4), the increased incidence of vasculitis and the decreased incidence of serious central nervous system and renal disease reported in patients with anti-Sm (27,28), and the high frequency of antibodies to Ha reported in SLE patients with Sjogren’s syndrome (9) were either not present in our SLE patients or were not statistically significant. The clinical-serologic correlations in our patients which were statistically significant, the increased mortality rate in patients with antibodies to Ha, and the increased incidence of overlap features in patients with antibodies to both Sm and RNP require independent confirmation because of the relatively small number of patients involved and the possibililty of unperceived selection bias. In view of recent reports of instances of variation in titers of antibodies to ENA with time and level of disease activity (9,23,29), it may be that these contradictory results are due to the inappropriateness of studies of single serum samples. Longitudinal studies correlating antibody specificity and titers with disease activity may be required to determine the clinical significance, if any, of the various ENA antibodies. ACKNOWLEDGMENTS We are grateful to Drs. H. R. Holman, P. Maddison, and I. Listowsky for generously providing reference sera and purified chromatin extract, to the physicians on the Einstein faculty who provided access to their clinical records, to Dr. D. M. Marcus for reviewing the manuscript, to Bernice Samuels for technical assistance, to Dorothy Vasco for secretarial support, and to Dr. S. Wassertheil-Smolar and Susan Slagel for assistance with the statistical analyses. REFERENCES 1. Maddison PJ, Reichlin M: Quantitation of precipitating antibodies to certain soluble nuclear antigens in SLE: their contribution to hypergammaglobulinemia. Arthritis Rheum 20:819-824, 1977 2. Sharp GC, Irvin WS, LaRoque RL, Velez C, Daly V, Kaiser AD, Holman HR: Association of auto-antibodies to different nuclear antigens with clinical patterns of rheumatic disease and responsiveness to therapy. J Clin Invest 50:350-359, 1971 3. Notman DD, Kurata N, Tan EM: Profiles of antinuclear antibodies in systemic rheumatic diseases. Ann Intern Med 83:464469, 1975 4. Farber SJ, Bole GG: Antibodies to components of ex- tractable nuclear antigen. Arch Intern Med 136:42543 1, 1976 5. Sharp GC, Irvin WS, Tan EM: Mixed connective tissue disease-an apparently distinct rheumatic disease syndrome associated with a specific antibody to an extractable nuclear antigen (ENA). Am J Med 52:148-158, 1972 6. Reichlin M, Mattioli M: Correlation of a precipitin reaction to an RNA protein antigen and a low prevalence of nephritis in patients with systemic lupus erythematosus. N Engl J Med 286:908-911, 1972 7. Hamburger M, Hodes S, Barland P: The incidence and clinical significance of antibodies to extractable nuclear antigens. Am J Med Sci 273:21-28, 1977 8. Alspaugh M, Maddison P: Resolution of the identity of certain antigen-antibody systems in systemic lupus erythematosus and Sjogren’s syndrome: an interlaboratory collaboration. Arthritis Rheum 22:796-798, 1979 9. Kassan SS, Akizuki M, Steinberg AD, Reddick RL, Chused TM: Antibody to a soluble acidic nuclear antigen in Sjogren’s syndrome. Am J Med 63:328-335, 1977 10. Akizuki M, Powers R, Holman HR: A soluble acidic protein of the cell nucleus which reacts with serum from patients with systemic lupus erythematosus and Sjogren’s syndrome. J Clin Invest 59:264-272, 1977 1 1. Kurata N, Tan EM: Identification of antibodies to nuclear acidic antigens by counterimmunoelectrophoresis. Arthritis Rheum 19:574-580, 1976 12. Preliminary criteria for the classification of SLE. Bull Rheum Dis 21:643-648, 1971 13. Ginsberg B, Keiser H: A Millipore filter assay for antibodies to native DNA in sera of patients with systemic lupus erythematosus. Arthritis Rheum 16:199-207, 1973 14. Keiser H: Preparation of ‘251-labelednative DNA for use in radioimmunoassays for anti-native DNA antibodies. Arthritis Rheum 16:468470, 1973 15. McCarty DJ, Editor: Arthritis and Allied Conditions. Ninth edition. Philadelphia, Lea & Febiger, 1979, 16. Goodwin GH, Sanders C, Johns EW: A new group of chromatin-associated proteins with a high content of acidic and basic amino acids. Eur J Biochem 38:14-19, 1973 17. Dorsch CA, White GM, Berzofsky RN: The measurement of antibodies to extractable nuclear antigen. Am J Clin Pathol71:333-337, 1979 18. Mattioli M, Reichlin M: Physical association of two nuclear antigens and mutual occurrence of their antibodies: the relationship of the Sm and RNA protein (Mo) systems in SLE sera. J Immunol 110:1318-1324, 1973 19. Lerner MR, Steitz JA: Antibodies to small nuclear RNAs complexed with proteins are produced by patients with systemic lupus erythematosus. Proc Natl Acad Sci USA 7615495-5499, 1979 20. Miyawaki S, Kohmoto K, Kurata N, Ofuji T: Identification and characterization of two new soluble nuclear antigens reactive with sera of patients with connective tissue diseases. Arthritis Rheum 212303-810, 1978 ANTIBODIES TO SALINE ENA 1035 21. Winn DM, Wolf JF, Harmon D, Sharp GC: Characterization of a distinct acidic protein antigen (MA) and clinical findings in systemic lupus erythematosus patients with MA antibodies. J Clin Invest 64:820-823, 1979 22. Miyachi K, Tan EM: Antibodies reacting with ribosomal ribonucleoprotein in connective tissue diseases. Arthritis Rheum 22237-93, 1979 23. Koffler D, Miller TE, Lahita RG: Studies on the specificity and clinical correlation of antiribosomal antibodies in systemic lupus erythematosus sera. Arthritis Rheum 22:463470, 1979 24. Alspaugh MA, Tan EM: Antibodies to cellular antigens in Sjogren’s syndrome. J Clin Invest 55:1067-1073, 1975 25. Parker MD: Ribonucleoprotein antibodies: frequency and clinical significance in systemic lupus erythematosus, scleroderma and mixed connective tissue disease. J Lab Clin Med 82:769-775, 1973 26. McCain GA, Bell DA, Chodirker WB, Komar RR: Antibody to extractable nuclear antigen in the rheumatic diseases. J Rheumatol 5:399406, 1978 27. Dorsch CA, Feinglass EJ, Stevens MB: Clinical significance of antibodies to extractable nuclear antigen (ENA) in systemic lupus erythematosus (SLE). Arthritis Rheum 20:114, 1977 28. Winn DM, Wolfe JF, Lindberg DA, Fristoe FH, Kingsland L, Sharp GC: Identification of a clinical subset of systemic lupus erythematosus by antibodies to the Sm antigen. Arthritis Rheum 22: 1334-1337, 1979 29. Fishbein E, Ramos-Niembro F, Alarcon-Segovia D: Free serum ribonucleoprotein in mixed connective tissue disease and other connective tissue diseases. J Rheumatol 5:384-390, 1978 Home Study Course A home study course in Immunodeficiency Diseases will be presented by the University of Wisconsin School of Medicine, to be completed between September 1 and December 31, 1980. The course is developed and presented by Sheldon Horowitz, MD, Associate Professor of Pediatrics, Division of Immunology, and Member, lmmunobiology Research Center. This program meets the criteria for 36 hours AMA credit. For further information write to Home Study-CME, 481 WARF Building, 610 Walnut Street, Madison, WI 53706.