The role of histones as nuclear autoantigens in drug-related lupus erythematosus.код для вставкиСкачать
1064 THE ROLE OF HISTONES AS NUCLEAR AUTOANTIGENS IN DRUG-RELATED LUPUS ERYTHEMATOSUS ENG M. TAN and JOSEPH P. PORTANOVA Patients with drug-related lupus erythematosus produce antibodies to nuclear histones which can be detected by a three-step indirect immunofluorescent technique. Procainamide-related antinuclear antibodies were detected by this technique, but hydralazine-related antinuclear antibodies were not. Certain evidence suggests that antibodies induced by the two drugs are reactive with Merent subclasses of histones. Hydralazine was shown to interact with a soluble DNA-histone complex, and the resulting interaction rendered the histone moiety resistant to trypsin digestion. This mechanism may help to maintain DNA-histone complexes in a potentially immunogenic form and result in the production of autoantibodies. ities, including anti-DNA and antibodies to nonhistone antigens, whereas in drug LE antibodies to native DNA and nonhistone nuclear antigens were not detectable (1). These two observations, i.e. presence of antibodies to histones and absence of antibodies to other nuclear antigens, have proved to be useful laboratory parameters in establishing the diagnosis of drug-induced lupus and in distinguishing it from SLE. In this report we will review some of the previous observations and report the new finding of differences in the serologic profiles of lupus induced by procainamide and hydralazine. Antibodies to histones have been found in a number of systemic rheumatic diseases including the idiopathic form of systemic lupus erythematosus (SLE), drug-induced lupus (drug LE), and rheumatoid arthritis (1-4). In a previous study an unusual feature in drug-induced lupus was that antibodies to histones were present in almost all patients with drug LE (1). In contrast, in the naturally occurring idiopathic SLE, antibodies to histones were observed in approximately 25% of such patients. A similar frequency appears to be present in patients with rheumatoid arthritis (3). In addition, another distinguishing feature between the two types of lupus was that SLE was characterized by the presence of antinuclear antibodies (ANA) of multiple specific- The materials and methods used in the studies reported here have been published in detail previously (5). Briefly, a three-step immunofluorescent method was used for the detection of antibodies to histones. Mouse kidney cryostat sections fixed in acetone but otherwise untreated were reacted with serum in the indirect immunofluorescent method. The fluorescein conjugate consisted of antiserum to human IgG labeled with fluorescein isothiocyanate. Serum containing antibodies to histones demonstrate a homogeneous nuclear staining pattern. The second step consisted of a mouse kidney section which was extracted with 0.1N HCl for 30 minutes at room temperature. This treatment resulted in the extraction of histones from the nuclei in the tissue section. The HC1-extracted section was then reacted with an aliquot of the same serum sample. Because histones are absent from this acid-extracted tissue section, the serum now becomes negative for ANA if only antibodies to histones are present. The third step consists of another acid-extracted mouse kidney section to which purified histones have been added. The added histones complex with the DNA present in the nuclei, and this substrate is now used as a histone-reconstituted tissue section for reaction with another aliquot of the same serum. the serum becomes positive again for ANA. Thus, in the three-step method, a serum containing antibodies to histones would be positive for ANA on the first From the Division of Rheumatic Diseases, Department of Medicine, University of Colorado Health Sciences Center, Denver, Colorado 80262. Supported by NIH Grant #AM20705 and a Clinical Research Center Grant from the Arthritis Foundation. Address reprint requests to Eng M. Tan, MD, Division of Rheumatic Diseases, University of Colorado Health Sciences Center, Denver, Colorado 80262. Arthritis and Rheumatism, Vol. 24, No. 8 (August 1981) MATERIALS AND METHODS ROLE OF HISTONES 1065 section, negative for ANA on the acid-extracted section, and become positive again on the histone-reconstituted section. The sera examined in this study were taken from patients who had a lupus-like clinical illness which could be directly related to ingestion of hydralazine or procainamide. RESULTS Figure 1 shows the detection of anti-histone antibodies by the three-step immunofluorescent method. A consistent observation was that with the histone-reconstituted section, the pattern of ANA is frequently nodular and not homogeneous. In many sera, the titer of the ANA on histone-reconstituted sections is equal to the titer on the untreated section. In some sera, the titer on histone-reconstituted sections could be a small amount lower. In a previously reported study, the observations on 23 patients with drug-induced LE were compared with those on 20 patients with SLE. The results are depicted in Figure 2 for the three-step immunofluorescent method. It is seen that all patients with drug LE became negative for ANA on acid-extracted sections and with the exception of one patient, the other 22 patients became positive for ANA again on histone-reconstituted sections. The findings with SLE were strikingly different. Many sera became negative for ANA on acid-extracted sections, but there were several where the ANA remained at the same titer or were only reduced in titer. Of further interest was the finding that although a few of these sera became positive for ANA on the histonereconstituted sections, the majority remained negative. It could be demonstrated that the difference in the findings between drug LE and SLE was related to the fact that the sera of SLE patients contained antibodies of multiple specificities, whereas drug LE sera contained mainly antibodies to histones (Table 1). In addition to antibodies to histones, SLE sera contained antibodies to Sm and nRNP antigens as well as to double-stranded and single-stranded DNA. On the other hand, 3 of 23 patients with drug LE had antibodies to single-stranded DNA, but antibodies for other nuclear antigens except histones were not detected. 2 was exThe diversity ofpatterns Seen in ~i~~~~ plained on the following basis. Extraction of tissue sections with acid resulted in the removal of all or most histones and several nonhistone protein antigens. HOWever, double-stranded DNA remained in the nuclei of antibody to doublethese sections. sera stranded DNA would remain positive on acid-extracted sections, which was observed with several SLE sera. On histone-reconstituted sections, the antigens present in A B C Figure 1. A three-step indirect immunofluorescent test for the detection of antibodies to nuclear histones. The substrate consisted of sections of mouse kidney. In A, sections were fixed in acetone and reacted with the serum from a patient with drug-related LE containing to histones. A homogeneous pattern of nuclear staining was observed. In the second step (B), a companion tissue section was extracted with 0.1 N HCl to remove nuclear histones. The ANA test of the same serum became negative. A third tissue section (C) was extracted with HCI to remove histones, and subsequently purified histones were added to the acid-extracted tissue section. Purified histones complex with the DNA remaining on the acid-extracted tissue settion, and the reconstituted “section” shows positive ANA again with the same serum. TAN AND PORTANOVA 1066 - 1024- 0 I IDIOPATHIC SLE DRUG LE 512256- .-c 128- t U z e 64- \ c 3216- ... &*A AAb Contra' Acid Extracted Histone Reconstitutet Control Acid Extracted I I .A 8 A. A Histone Reconstituted Figure 2. The three-step indirect immunofluorescent technique was used to analyze sera of patients with drug-related LE and idiopathic SLE. Striking differences were observed. In drug LE, all sera became negative for ANA on acid-extracted tissue sections, and except for one serum, they were all positive again for ANA on histone-reconstituted sections. In contrast, some SLE sera remained positive for ANA on acid-extracted sections and the majority did not regain positive ANA on histone-reconstituted sections. The reasons for these differences are explained in the text. nuclei now consisted of DNA and histones but were still devoid of nonhistone antigens such as Sm and nRNP. Therefore, sera containing antibodies to Sm or nRNP would not become positive for ANA on the histone-reconstituted sections. This method could also lend itself to the detection of antibodies to specific subclasses of histones. In the study shown in Table 2, it was observed that the majority of drug LE patients had antibodies directed against the H2A-H2B histone complex. There were 3 of 22 sera with antibodies to H3-H4 complex and 2 of 22 sera with antibodies to H1. Only 5 of the SLE sera could be analyzed for antibodies to histone subclasses in this fashion since the other 2 contained antibodies with titers too low for this analysis. Of the 5 SLE sera, the distribution of antibodies to histone subclasses appeared to be more heterogeneous than that found in the sera of the drug LE patients. These observations were made with sera mainly from patients with lupus syndrome induced by procainamide. In our current studies, we have examined the reaction of hydralazine-induced LE sera. These new findings are shown in Tables 3 and 4. In Table 3, 10 patients with pronestyl LE were studied for anti-histone antibodies. It is seen that all sera became negative at a titer of <4 for ANA on acid-extracted sections. With the Table 1. Antibody profiles of drug-related LE and idiopathic SLE Idiopathic SLE Drug-related LE Antibodies to Total tested Sm nRNP Histones dsDNA ssDNA 20 23 9 0 3 0 7 22 7 5 0 3 ROLE OF HISTONES 1067 Table 4. Hydralazine-induced IgG antinuclear antibodies Table 2. Reactivity with histone subclasses ANA titer on mouse kidney Number of sera reactive with Idiopathic SLE Drug-relatedLE Total histones HI H2A-H2B H3-H4 5 22 2 2 3 21 I 3 exception of two sera, the other eight became ANA positive again on histone-reconstituted sections. In contrast, the reaction of hydralazine-induced ANA was quite different. First, the titers of ANA on acetone-fixed sections were generally lower than the ANA titers of pronestyl-induced ANA. The hydralazine sera all became negative for ANA on acid-extracted sections, but the interesting finding was that they all remained negative on histone-reconstituted sections. On untreated tissue sections (i.e. acetone-fixed only), the ANA patterns of hydralazine sera were all homogeneous and looked identical to that of pronestyl-induced AN A. Likewise, the hydralazine sera were negative for ANA on acid-extracted sections and, in these two characteristics, were similar to the pronestyl-induced ANA and different from SLE sera. The negative test for ANA on histone-reconstituted sections therefore appeared to contrast pronestyl-induced LE from hydralazine-induced LE. In preliminary studies, we have evidence that hydralazine-induced ANA contains antibodies to histones but is reactive with histones of Werent subclasses from that of pronestyl-induced ANA. The three-step immunofluorescent method appears to be detecting histone specificities associated with the H2A-H2B complex and not other histone subclasses. Some previous studies have been directed at elucidating the possible mechanism relating to induction of anti-histone antibodies. It has been difficult to analyze for precipitating antibodies to nuclear histones Table 3. Pronestyl-induced IgG antinuclear antibodies ANA titer on mouse kidney Patient - Ta Wh Er Ga Gu Ho Mo No Pr Om Acetone fixed O.lNHC1 extracted Histone reconstituted 4096 4096 4096 2048 2048 2048 1024 1024 512 32 <4 t 4 <4 t 4 t4 <4 <4 <4 <4 t4 512 2048 512 1024 256 2048 256 <4 64 <4 Patient Ra we Bo Ma Ba Wh Jo Mi LY K1 Acetone fixed 0.1N HC1 extracted Histone reconstituted 2048 512 256 256 I28 128 128 128 I28 4 <4 <4 <4 <4 t4 <4 <4 <4 t4 t4 t4 t4 <4 <4 t4 <4 <4 <4 t4 t4 by immunodiffusion methods because of nonimmunologic reactions between serum proteins and free histones. However, certain forms of histones can be isolated from nuclei in a soluble nucleoprotein (sNP) complex, which has been shown to precipitate with antibodies in the sera of patients with SLE and hydralazineinduced LE (6). It was shown that the sNP antigen consists of DNA-histone complexes. The antigenicity of sNP can be abolished by treatment with either DNAse alone or trypsin alone. In experiments previously reported (7), sNP antigen was isolated from calf thymus nuclei and increasing amounts of hydralazine were added to a standard amount of sNP. The findings are shown in Figure 3. There was a physicuchemical interaction between sNP and hydralazine in that the viscosity of this mixture increased with increasing concentrations of hydralazine. Experiments were also performed with purified DNA and purified histones as illustrated in Figure 3. There was some slight increase in viscosity when hydralazine was added to purified DNA but no increase in viscosity when hydralazine was added to purified (free) histones. It was of interest to study some of the immunochemical characteristics of the sNP-hydralazine complex. As mentioned previously, the sNP antigen by itself was labile to treatment with either DNAse or trypsin. On the other hand, when sNP-hydralazine complex was subjected to the same enzyme treatments, DNAse continued to destroy precipitating activity, but trypsin did not. Therefore, the complexing of hydralazine with sNP in some way protected the histone moiety of the sNP complex from digestion with trypsin (Figure 4). It was postulated from these studies that hydralazine interaction with nucleoprotein complexes in vivo might protect nucleoprotein from breakdown by proteolytic enzymes and thus create or maintain a potentially immunogenic autoantigen. TAN AND PORTANOVA 1068 3.0 2.6 0 Is= .s 2.2 v) 0 u .-w v) aa .P w -maa 1.8 a 1.4 Histones ._... ....... ...... 1.oJ 2 4 6 8 1 0 Final Concentration Hydralazine A This finer differentiation of antibodies to histone subclasses and their corresponding relationships with procainamide or hydralazine bring up intriguing questions concerning the mechanisms responsible for their induction. Many investigators at this conference have discussed and analyzed the chemical interactions between procainamide or hydralazine (and their metabolites) and nuclear macromolecules. If these interactions should occur in vivo in patients ingesting these drugs and this interaction stimulates an immunologic response, our studies indicate that the nuclear macromolecules of interest should be histones. This is based on the finding that most patients with drug-related LE have antibodies to nuclear histones. On the other hand, it is possible that these drugs react in vivo with other nuclear macromolecules such as DNA and nonhistone proteins, and such interactions may promote subsequent interaction with histones. If drug-DNA interaction occurs, it does not appear to be very immunogenic since antibodies to double- or even single-stranded DNA are infrequently detected. Antibodies to certain nonhistone nuclear proteins, however, have been detected in patients treated with pro- [mg/mll Figure 3. Hydralazine was added to soluble nuclear protein (sNP) and to calf thymus DNA and calf thymus histones. The relative viscosity of the mixtures was measured in an Ostwald viscosimeter and plotted relative to viscosity without addition of hydralazine. The viscosity of sNP-hydralazine mixtures increased significantly with increasing concentration of hydralazine. There was a slight increase in the viscosity of DNA-hydralazine mixtures but no increase in viscosity with the histone-hydralazine mixtures. DISCUSSION Recent studies have confirmed the usefulness of the three-step immunofluorescent technique for the detection of antibodies to histones. In almost all patients with procainamide-induced lupus, this antibody is present to the exclusion of antibodies to native:DNA and nonhistone proteins. The recent studies comparing procainamide-induced ANA with hydralazine-induced ANA have uncovered some interesting differences. Hydralazine-induced ANA is not detected with the immunofluorescent technique, but other evidence points to the conclusion that hydrdazine-induced ANA are also antibodies to histones. However, these antibodies are not directed against the H2A-HZB complex, which is the antigenic determinant for the procainamide-induced ANA. Figure 4. The precipitating activity of sNP is destroyed by either DNAse or trypsin treatments as demonstrated in the immunodiffusion study shown in the upper frame. The center well contained seTum from a patient with SLE who had precipitating antibody to sNP. The observations with sNP-hydralazine mixtures are shown in the lower frame. The sNP antigen in the hydralazine mixture continued to be labile to DNAse treatment but was now resistant to trypsin. ROLE OF HISTONES cainamide (8). This has some features of a secondary immune response since antibodies are detected very shortly after the drug is started, but the immune response is short-lived and antibodies are not detectable in patients continued on the medication. 1069 4. 5. REFERENCES Fritzler MJ, Tan EM: Antibodies to histones in drug-induced and idiopathic lupus erythematosus. J Clin Invest 62560-567, 1978 Stollar BD: Reactions of systemic lupus erythematosus sera with histone fractions and histone-DNA complexes. Arthritis Rheum 14:485492, 1971 Aitcheson C, Peebles C, Joslin F, Tan EM: Characteristics 6. 7. 8. of antinuclear antibodies in rheumatoid arthritis. Arthritis Rheum 23528-538, 1980 Agnello V, Arbetter A, Ibanez de Kasep R, Powell R, Tan EM, Joslin F: Evidence for a subset of rheumatoid factors that cross-react with DNA-histone and have a distinct cross-idiotype. J Exp Med 151:1514-1527, 1980 Tan EM, Robinson J, Robitaille P: Studies on antibodies to histones by immunofluorescence. Scand J Immunol 5:811818, 1976 Tan EM: An immunologic precipitin system between soluble nucleoprotein and serum antibody in systemic lupus erythematosus. J Clin Invest 46:735-742, 1967 Tan EM: Drug-induced autoimmune diseases. Fed Proc 33:1894-1897, 1974 Winfield JB, Koffler D, Kunkel HG: Development of antibodies to ribonucleoprotein following short-term therapy with procainamide. Arthritis Rheum 18531-534, 1975 DISCUSSION Dr. Shulman: Of the 20 patients with SLE, 6 had glomerulonephritis and the others did not. Were the 3 in the SLE category in the nonrenal group? Did they have an illness that might resemble drug-related lupus? Dr. Tan: No, they didn’t seem to. The incidence of antiDNA antibodies in that group of 20 patients was not inconsistent with that of a larger group of patients with systemic lupus erythematosus. We had about 8 or 9 patients with anti-Sm antibody. In general, there was nothing in the idiopathic lupus that we could ascribe to any particular antibody specificity. Dr. Hess: How do you explain the two pronestyl patients who reacted similarly to the hydralazine patients? Dr. Tan: Not all procainamides are highly reactive with 2A-2B complex. Some are still reactive with H3-H4 complex, and we have 1 or 2 patients who are reactive with H1 complex. Dr. Parker: Is there any indication that the pattern changes with time during the course of the illness? Dr. Tan: We haven’t had a chance to follow these patients. Many have been referred only for a one-time ANA analysis. Dr. Alarcon-Segovia: It has been said that the LE cells result from interaction of antibodies to nucleoprotein with this material in disrupting cells. Does this antibody form LE cells? Dr. Tan: I think that is true. We performed the indirect LE cell test first described by Lachmann in which only the serum is needed. From these studies we concluded that the LE cell phenomenon is related to antibody to nucleoprotein. If antibodies to nucleoprotein are removed, the LE cell phenomenon is negative. Dr. Stollar: Did either or both of those react with isolated nucleosomes? The hydralazine-induced antibodies reacted with chromatin and with isolated histone. They must be able to react with some form of complexed histone unless there are free histones. Dr. Tan: It is clear that native tissue sections are positive for ANA with hydralazine sera. Whatever 2A determinant they are reacting with is available on the tissue section. In our reconstituted sections, the 2A2B complex reassociates with the DNA in a clumpy fashion, and the 2A determinant may be hidden so that it is not available for reaction with hydralazine sera. Dr. Stollar: Have you tried salt-extracted histones that have never seen acid? Dr. Tan: Yes, but it made no difference.