Eur. J. Immunol. 1990. 20: 437-440
Autoantibodies against snRNP proteins E , F and G
437
Short paper Rolf ReUteroo, Susanne Rothe., Winand HabetsA, Walter J.Van VenrooijA and Reinhard LuhrmannO
Autoantibody production against the U small nuclear ribonucleoprotein particle proteins E, F and G in patients with connective tissue diseases*
Max-Planck-Institut fur molekulare GenetikO, Otto-Warburg-Laboratorium, Berlin, Department of Biochemistry, University of NijmegenA, Geert Grooteplein
The nucleoplasmic U small nuclear ribonucleoprotein particles (snRNP) have a set of seven proteins in common which are designated B', B, D, D ' , E, F and G. Patients suffering from rheumatoid autoimmune diseases such as systemic lupus erythematosus often develop autoantibodies against the proteins B', B, and D. Here we describe a sensitive immunoassay which allows the specific detection of autoantibodies reacting with the E , F or G snRNP proteins. We were able to identify several patient sera containing autoantibodies against one or more of these proteins. This demonstrates that all snRNP proteins described so far are potentially antigenic in systemic rheumatoid diseases.The characterization of the antibodies showed an immunological cross-reactivity between the snRNP protein G and the 70-kDa protein of U1 snRW. Several sera contained autoantibodies which were specific for the F snRNP protein.
1 Introduction All eukaryotic cells contain a class of small nuclear ribonucleoproteins, termed U snRNP,which are involved in various steps of RNA processing (for recent reviews see [1-31). Among the best characterized U snRNP are the four nucleoplasmic RNP U1, U2, U5 and U4/6 from HeLa cells which are essential for the splicing of pre-mRNA.They are composed of one [4], or in the case of U4/U6 of two snRNA [5-71 and a set of proteins with an apparent molecular mass between 200 kDa and 11kDa (for a review see [S]). Some proteins are specific for an snRNP species: the 70-kDa, the A (34-kDa) and the C (22-kDa) protein for U1 RNP, the A' (33 kDa) and the B"(28.5 kDa) for the U2 RNP and a set of at least six proteins of between 200 kDa and 42 kDa for U5 RNp[9-11]. Sevenotherproteins,B', B,D,D', E , F a n d G (29, 28, 16, 15.5, 12, 11 and 9 kDa, respectively), are common to all nucleoplasmic snRNP so far known. Autoantibodies were found in sera of patients suffering from rheumatoid autoimmune diseases against many of the snRNP proteins. For mixed connective tissue disease (MCTD) for example, the so-called anti-RNP autoantibodies, which react with the 70-kDa protein [12], are typical whereas for SLE anti-Sm antibodies, which are directed
[I 78561
* This work was supported by a postdoctoral fellowship from the DECHEMA (R. R.) and by grants from the Deutsche Forschungsgemeinschaft (DFG) to R. L. (Lu 294-2). Present address: Max-Planck-Institut fur Entwicklungsbiologie, Abt. 111, Spemannstr., D-7400 Tiibingen, FRG Present address: Freie Universitat Berlin, Institut fur Humangenetik, Heubnerweg 6, 1 Berlin, FRG Correspondence: Reinhard Liihrmann, Institut fur Molekularbiologie und Tumorforschung der Philipps-Universitat Marburg, Emil-Mannkopff-StraBe 2, D-3550 Marburg, FRG
Abbreviations: MCTD: Mixed connective tissue disease mJG:N2,@, N7-Trimethylguanosine NCS:Newborn calf serum NHS: Normal human serum snRNP: Small nuclear ribonucleoprotein particle(s) 0 VCH Verlagsgesellschaft mbH, D-6940 Weinheim, 1990
against an epitope common to the proteins B', B, and D [13-151, appear to be of diagnostic significance [16]. Beyond that, autoantibodies have been found in sera from patients with SLE, MCTD and less frequently with polymyositis and scleroderma which react with the snRNP proteins A or C and, far less common, with the proteins A', B"or E [ 15,17-22].The origin of the autoantibodies against snRNP proteins and their significance for the pathogenesis of the diseases is not yet known. However, evidence is accumulating that the formation of the autoantibodies against snRNP proteins in rheumatoid diseases is an antigen-driven process and that the U snRNP are recognized as a whole by the immune system ([23-271 see [16] for a review). Here we describe an immunoassay which allows the specific detection of autoantibodies against the three low-molecular mass snRNP proteins, E, Fand G. With this assay we were able to demonstrate for the first time the presence of autoantibodies against the proteins F and G in sera of patients suffering from rheumatoid diseases. The autoantibodies were characterized immunochemically and showed interesting specificities.
2 Materials and methods 2.1 Sera and antibodies Sera were from patients with SLE, diagnosed according to the preliminary ARA criteria.They were obtained from the Nijmegen stock of sera from connective tissue disease patients or kindly provided by Prof. Dr. H. Peter, University of Freiburg. Anti-N2, N2, N7-trimethylguanosine (m3G) antibody used for the purification of U s n R " was purified as described by Bochning et al. [28]. 2.2 Isolation of snRNP proteins U snRNP were purified from HeLa cells by anti-m3G immune affinity chromatography essentially as described by Bringmann et al. [29]. The snRNP proteins were separated by SDS-PAGE and isolated by electroelution from the gel [25].
438
Eur. J. Immunol. 1990.20: 437-440
R. Reuter, S. Rothe,W. Habets et al.
2.3 ELISA (Cl
The ELISA using single gel purified snRNP proteins E, F and G as antigens was essentially performed as described previously for other snRNP proteins [25]. However, polystyrene instead of polyvinylchloride microtiter plates were used. After coating with 10 ng of antigerdwell and saturation with BSA the plates were incubated with patient sera diluted in 0.15 M NaCI, 0.02 M potassium phosphate, pH 7.4,0.1% Tween 20 (PBST) with 5% newborn calf serum (NCS) for 16 h at 4 "C. The plates were then washed four times with PBSTand incubated with a goat anti-human IgG alkaline phosphatase conjugate for 2 h at 22 "C. Bound antibody was detected after a second washing cycle by the color reaction of the phosphatase substrate, p-nitrophenyl phosphate at 22 "C. 2.4 Immunoblot
Proteins of affinity-purified U snRNP (see Sect. 2.2) were separated by electrophoresis in a 15% SDS-polyacrylamide gel and transferred electrophoretically to nitrocellulose [30]. Strips of the nitrocellulose were saturated with 5% NCS in PBS and were incubated with a dilution of the respective patient serum in buffer A (10 mM Tris-HCI, pH 7.4, 150 mM NaCl, 100 mM MgCl2, 0.5% Tween 20, 0.1% Triton X-100, 5% NCS).Then they were washed five times in buffer B (buffer A without Triton X-100 and NCS) and subsequently incubated with an appropriate dilution of an alkaline phosphatase-conjugated goat anti-human IgG antibody in buffer B. After a second washing cycle the strips were stained with a solution of 0.12 mM 5-bromo-4chloro-indolylphosphate, 0.12 mM nitrotetrazolium blue, 0.5 mM MgC12 and 0.1 M diethanolamine, pH 9.8 [31]. 2.5 Affinity purification of antibodies
U snRNP proteins were transferred to a nitrocellulose sheet as described above. The horizontal, 10-cm-wide nitrocellulose strip containing the protein from which antibodies were to be eluted was located relative to two adjacent vertical strips from the same blot that had been immunostained as described above. Then the horizontal strip was saturated, incubated with the serum dilution and washed (see above). Subsequently, the antibodies bound to the strip were eluted with 5 mM glycine, pH 2.5,150 mM NaCl, 0.1% BSA, 0.1% Tween 20, essentially as described by Smith and Fisher [32], and the eluates were immediately neutralized with 1M Tris-HCI, pH 8.
3 Results and discussion 3.1 A sensitive ELISA for the detection of specific autoantibodies against U snRNP proteins We tested 64 sera of patients suffering from systemic rheumatoid diseases for the presence of antibodies against E, F o r G with the ELISA described in Sect. 2.3.Those sera had been pre-selected by their ability to precipitate snRNP U1 to U6 from nuclear extracts. In Fig. 1 examples of titrations of two patient sera are displayed in comparison to a normal human serum (NHS). The graphs show that the
NHS 0.5
0'4 0.3
I\
I\ 6
7
8
9
10
6
7
8
9
10
6
7
8
9
10
-log3 of serum dilution
Figure I. Specificity and sensitivity of the ELISA for the detection of autoantibodies against the snRNP proteins E, F and G. Patient serum M.M. (A), P.B. (B) or NHS (C) were serially diluted (1 : 600, 1 : 1800,etc.) and reacted with snRNP proteins E, For G coated on wells of polystyrene microtiter plates as described in Sect. 2.3. Bound antibody was detected by ELISA techniques and is presented here by AM5nm after a color development for 3 h following the reaction with the snRNP protein E (closed circles), F (open circles), G (closed square) or no snRNP protein (open square).
ELISA is highly specific and that the autoantibodies discriminate well between the three s n R " proteins. The patient serum M.M. for example only gave a signal against the F snRNP protein (Fig. 1A) and the serum P B. exclusively with the protein G (Fig. 1B) when compared to the low background of the NHS (Fig. 1C). The ELISA is also very sensitive; autoantibodies were detectable against the snRNP proteins at serum dilutions of 1:5000 or higher (Fig. 1A and B). Under the conditions chosen the signal of the ELISA depended strongly on the concentration of the autoantibody in the sera, and the titration curves of all patient sera reactive against one respective protein were approximately of the same shape (data not shown). Thus, the signal at a given dilution gave an estimate of the concentration of the autoantibody in the serum and allowed a semiquantitative comparison of its concentration in different sera.
3.2 Differential immune response against the three small U snRNP proteins E, F and G in sera of patients with rheumatic disorders Eight out of the 64 tested patient sera significantly reacted with one or two of the three low-molecular mass snRNP polypeptides E, F, and G (Fig. 2), i.e. the signal was more than three times higher than the control signal obtained with the given serum in microtiter wells without an snRNP protein coat. Three sera (A.D., A.H., R.H.) contained autoantibodies against the E protein, four sera (A.D., S.R., H.T., M.M.) against the F protein and two sera (PB., J.R.) against the G protein. Thus, autoantibodies exist in sera of patients suffering from systemic rheumatoid diseases which react with the low-molecular mass snRNP proteins E, Fand G. Apparently, the immune system of the patients is able to form autoantibodies against any of the known polypeptide components contained in the U snRNF! The autoimmune response of the patients against the three low-molecular mass snRNP polypeptides E , F, and G varies among
Eur. J. Immunol. 1990. 20: 437-440
Autoantibodies against snRNP proteins E, F and G
Y L
1
A 0.
S.R. H.T.
PB.
MM
AH.
R.H.
J.R.
NHS
Figure 2. Differential autoimmune response of SLE patients against snRNP proteins E , Fand G as detected by an ELISA. Serial dilutions of patient sera were reacted with snRNP proteins E , For G in microtiter plates as described in Sect. 2.3. Displayed is the quotient of Am5 nm found for the 1 :600 serum dilution after reaction with the snRNP proteins E , F or G (stippled, open or cross-hatched columns) and of AM^ nm found for the same dilution after reaction in a control microtiter well coated with buffer and BSA only.
individuals (Fig. 2). We do not know whether this differential autoimmune response has a diagnostic relevance. However, we assume that it reflects the individual immunological responsiveness of the patients towards U snRNP. To compare the sensitivity and specificity of ELISA and immunoblotting, we also tested the reactivity of the eight patient sera containing autoantibodies against the s n R ” proteins, E , F or G by immunoblotting (Fig. 3). When snRNP polypeptides which had been immobilized on nitrocellulose after separation in an SDS-polyacrylamide gel were probed with low dilutions (1:25) of the eight sera only five of them showed a detectable reaction with the low-molecular mass snRNP proteins. In good agreement with the ELISA data, patient serum A.D. bound to proteins E and F (Fig. 3, lane l),patient sera S.R., H.T. and M.M. to protein F (lanes 2 , 3 and 5 ) and patient serum P.B.
Figure 3. The differential autoimmune response of SLE patients against snRNP proteins, E, F and G is partially detectable by an immunoblot. U s n W were purified from HeLa cells and their proteins were separated by SDS-PAGE and transferred to nitrocellulose. Subsequently, strips of nitrocellulose were incubated with a 1 :25 dilution of the indicated patient sera, and bound antibodies were detected by immunostaining (see Sect. 2.4).
439
to protein G (lane 4). However, no binding to E , For G was detectable for the patient sera A.H., R.H. and J.R. The lack of binding may indicate the lower sensitivity of the immunoblot compared to the ELISA. The sera from these three patients gave the weakest signal in the ELISA reflecting low concentrations of autoantibodies against E, F or G snRNP proteins which may be not sufficient to be detected by the immunoblot. However, qualitatively the immunoblot gave the same results with respect to the specificity of the autoantibodies contained in the patient sera as the ELISA inspite of its lower sensitivity.There was no indication for systematic differences of the two assays which would lead to the failure to detect autoantibodies against E , F o r G protein in the patient sera A.H., R.H. or J.R. 3.3 Immunochemical properties of the autoantibodies against snRNP protein E, F and G
To determine the specificity of the autoantibodies against the snRNP proteins E, F or G and to detect possible cross-reactivity among the snRNP proteins we affinity purified the respective antibodies. When we eluted autoantibodies of the patient sera M.M. (Fig. 4, lane 3) or H.T. from the F protein immobilized on nitrocellulose and subsequently reacted the antibodies with snRNP proteins on a nitrocellulose strip, they reacted specifically with the F protein only (Fig. 4, lane 4). There was no detectable cross-reactivity with other snRNP proteins.When autoantibodies of the patient sera A.D. (Fig. 4,lane 5), A.H. or R.H. were eluted from the E protein, they did not react detectably with the E protein or any other snRNP protein
Figure 4. The specificity of the autoantibodies against the snRNP proteins Fand G. Autoantibodies reacting with the snRNP protein G were affinity purified from patient serum P.B. (lane l), then reacted with U snRNP proteins transferred to a nitrocellulose strip and detected by immunostaining (lane 2), as described in Sect. 2.4 and 2.5. Correspondingly, autoantibodies against the snRNP protein F were purified from patient serum M.M. (lane 3) and against the snRNP proteins E from patient serum. A.D. (lanes 5) and tested for their specificity (lanes 4 and 6, respectively).
440
R. Reuter, S. Rothe, W. Habets et al.
(Fig. 4, lange 6).Therefore, we could not draw a conclusion as to the specificity of the autoantibodies reactive to the E protein. The autoantibodies against the protein G, however, showed a strong cross-reactivity with the 70-kDa protein (Fig. 4, lanes 1, 2). We are convinced that this property is specific to the antibody against G and reveals an epitope common to the 70-kDa and the G protein since none of the other autoantibodies in patient serum P.B., e.g. against the D protein, showed a comparable cross-reactivity when eluted from the respective protein bands (data not shown). The cross-reactivity of the antibody against the G protein was somewhat surprising since a polyclonal antiserum raised against the 70-kDa protein in rabbits was highly specific for that protein [33]. Presumably, the respective cross-reactive epitope is not particularly immunogenic in the context of the 70-kDa protein. When in the reciprocal experiment autoantibodies of the patient serum P.B. were eluted from the 70-kDa protein they bound only this. We assume that only a minor fraction of the 70-kDa reactive autoantibodies in the serum cross-reacts with the G protein and may, therefore, be under-represented in the antibodies eluted from the 70-kDa protein.
4 Concluding remarks We have introduced an ELISA which allows the specific and sensitive detection of autoantibodies against the three low-molecular mass U snRNP proteins, E, F and G. By these means, we were able to demonstrate for the first time the presence of autoantibodies against these proteins in sera of patients suffering from systemic rheumatoid disease. Thus, autoantibodies can be formed against every individual snRNP polypeptide known so far which supports the notion that the U snRNP are recognized as a whole by the patient's immune system. The antibody against the F protein appears to be highly specific for the polypeptide. In contrast, the autoantibody binding to the G protein showed a strong cross-reactivitywith the 70-kDa protein revealing a hitherto unnoticed immunological relationship between these proteins. Immunological cross-reactivity seems to be a more general phenomenon among snRNP proteins. The U 1 s n R W proteins A and C share at least one epitope with each other and with the human proteins B'/B [27]. The latter two proteins share a different epitope with protein D [13, 151. Furthermore, antibodies cross-reacting with proteins A and B" have been described [34]. As has been shown for other anti-snRNP antibodies, anti-F and anti-G autoantibodies will be helpful tools for structural studies on the U snRNP and for the identification of cDNA clones coding for the snRNP proteins F or G. We are indebted to Dr. H . H. Peter (Institut fur Klinische Immunologie, Freiburg) for providing patients' sera. We thank Dr. D. Andrew for critical reading of the manuscript. R. R. is most grateful to Dr. M. Scott for the opportunity of finishing the manuscript in his laboratory at the University of Colorado.
Received July 31, 1989; in revised form October 30, 1989.
Eur. J. Immunol. 1990.20: 437-440
5 References 1 Sharp, F! A., Science 1987. 235: 766. 2 Maniatis, T. and Reed, R., Nature 1987. 325: 673. 3 Steitz, J. A., Black, D. L., Gerke,V., Parker, K. A., Kramer, A., Frendewey, D. and Keller,W., in Birnstiel, M. L. (Ed.), Structure and Function of Major and Minor Small Nuclear Ribonucleoprotein Particles. Springer-Verlag, Berlin, New York 1988, p. 115. 4 Lerner, M. R. and Steitz, J. A., Proc. Natl. Acad. Sci. USA 1979. 76: 5495. 5 Hashimoto, C. and Steitz, J. A., Nucleic Acids Res. 1984. 12: 3283. 6 Bringmann, P., Appel, B., Rinke, J., Reuter, R.,Theissen, H. and Luhrmann, R., EMBO J. 1984. 3: 1357. 7 Rinke, J., Appel, B., Digweed, M. andLuhrmann, R., J. Mol. Biol. 1985. 185: 721. 8 Liihrmann, R., in Birnstiel, M. L. (Ed.), Structure and Function of Major and Minor Small Nuclear Ribonucleoprotein Particles. Springer-Verlag, Berlin, New York 1988, p. 71. 9 Tazi, J., Alibert, C.,Temsamani, J., Reveillaud, I., Cathala, G., Brunel, C. and Jeanteur, P., Cell 1986. 47: 755. 10 Gerke.V. and Steitz. J. A.. Cell 1986. 47: 973. 11 Bach, M., Winkelmann, G. and Luhrmann, R., Proc. Natl. Acad. Sci. USA 1989. 86: 6038. 12 Pettersson, I. ,Wang, G., Smith, E. I.,Wigzell, H., Hedfors, E., Horn, J. and Sharp, G. C., Arthritis Rheum. 1986.29: 986. 13 Lerner, E. A., Lerner, M. R., Janeway, J. A., Jr. and Steitz, J. A., Proc. Natl. Acad. Sci. USA 1981. 78: 2737. 14 Guldner, H. H., Lakomek, H.-J. and Bautz, F. A., J. Immunol. Methods 1983. 64: 45. 15 Habets,W., Berden, J. H. M., Hoch, S. 0.andVanVenrooij,W., Eur. J. Immunol. 1985. 15: 992. 16 Tan, E. M., Adv. Immunol. 1989. 44: 93. 17 Conner, G. E., Nelson, D., Wisniewolski, R., Lahita, R. G., Blobel, G. and Kunkel, H. G., J. Exp. Med. 1982. 156: 1475. 18 Billings, P. B. and Hoch, S. O., J. Immunol. 1983. 131: 347. 19 Kinlaw, C. S., Robbersen, B. L. and Berget, S. M., J. Biol. Chem. 1983. 258: 7181. 20 Pettersson, I . , Hinterberger, M., Mimori, T., Gottlieb, E. and Steitz, J. A., J. Biol. Chem. 1984. 259: 5907. 21 Mimori, T., Hinterberger, M., Pettersson, I. and Steitz, J. A., J. Biol. Chem. 1984. 259: 560. 22 Wieben, E. D., Rohleder, A. M., Nenninger, J. M. and Pederson, T., Proc. Natl. Acad. Sci. USA 1985. 85: 7914. 23 Cohen, I? L. and Eisenberg, R. A., J. Immunol. 1982. 129: 2142. 24 Bernard, N. F., Eisenberg, R. A. and Cohen, P. L., J. Immunol. 1985. 134: 3812. 25 Reuter, R. and Luhrmann, R., Proc. Natl. Acad. Sci. USA 1986. 83: 8689. 26 Habets,W. J., Sillekens, P.T. G., Hoet, M. H., McAUister, G., Lerner, M. R. and VanVenrooij,W., Proc. Natl. Acad. Sci. USA 1989. 86: 4674. 21 Habets,W. J., Hoet, M. H., De Jong, B. A. W. ,Van der Kemp, A. and Van Venrooii,W. J., J. Immunol. 1989, in press. 28 Bochnig, P., Reuteri R., Bringmann, P. and Liihkann, R., Eur. J. Biochem. 1987. 168: 461. 29 Bringmann, P., Rinke, J., Appel, B., Reuter, R. and Liihrmann, R., EMBO J. 1983. 2: 1129. 30 Towbin, H., Staehelin,T. and Gordon, J., Proc. Natl. Acad. Sci. USA 1979. 76: 4350. 31 McGadey, J., Histochemie 1984. 23: 180. 32 Smith, D. E. and Fisher, P. A., J. Cell Biol. 1984. 99: 20. 33 Reuter, R., Rothe, S. and Luhrmann, R., Nucleic Acids Res. 1987. 15: 4021. 34 Habets,W. J., Hoet, M., Bringmann, P.. Luhrmann. R. and Van Venrooij, W., EMBO J. 1985. 4: 1545.
Autoantibodies against snRNP proteins E , F and G
437
Short paper Rolf ReUteroo, Susanne Rothe., Winand HabetsA, Walter J.Van VenrooijA and Reinhard LuhrmannO
Autoantibody production against the U small nuclear ribonucleoprotein particle proteins E, F and G in patients with connective tissue diseases*
Max-Planck-Institut fur molekulare GenetikO, Otto-Warburg-Laboratorium, Berlin, Department of Biochemistry, University of NijmegenA, Geert Grooteplein
The nucleoplasmic U small nuclear ribonucleoprotein particles (snRNP) have a set of seven proteins in common which are designated B', B, D, D ' , E, F and G. Patients suffering from rheumatoid autoimmune diseases such as systemic lupus erythematosus often develop autoantibodies against the proteins B', B, and D. Here we describe a sensitive immunoassay which allows the specific detection of autoantibodies reacting with the E , F or G snRNP proteins. We were able to identify several patient sera containing autoantibodies against one or more of these proteins. This demonstrates that all snRNP proteins described so far are potentially antigenic in systemic rheumatoid diseases.The characterization of the antibodies showed an immunological cross-reactivity between the snRNP protein G and the 70-kDa protein of U1 snRW. Several sera contained autoantibodies which were specific for the F snRNP protein.
1 Introduction All eukaryotic cells contain a class of small nuclear ribonucleoproteins, termed U snRNP,which are involved in various steps of RNA processing (for recent reviews see [1-31). Among the best characterized U snRNP are the four nucleoplasmic RNP U1, U2, U5 and U4/6 from HeLa cells which are essential for the splicing of pre-mRNA.They are composed of one [4], or in the case of U4/U6 of two snRNA [5-71 and a set of proteins with an apparent molecular mass between 200 kDa and 11kDa (for a review see [S]). Some proteins are specific for an snRNP species: the 70-kDa, the A (34-kDa) and the C (22-kDa) protein for U1 RNP, the A' (33 kDa) and the B"(28.5 kDa) for the U2 RNP and a set of at least six proteins of between 200 kDa and 42 kDa for U5 RNp[9-11]. Sevenotherproteins,B', B,D,D', E , F a n d G (29, 28, 16, 15.5, 12, 11 and 9 kDa, respectively), are common to all nucleoplasmic snRNP so far known. Autoantibodies were found in sera of patients suffering from rheumatoid autoimmune diseases against many of the snRNP proteins. For mixed connective tissue disease (MCTD) for example, the so-called anti-RNP autoantibodies, which react with the 70-kDa protein [12], are typical whereas for SLE anti-Sm antibodies, which are directed
[I 78561
* This work was supported by a postdoctoral fellowship from the DECHEMA (R. R.) and by grants from the Deutsche Forschungsgemeinschaft (DFG) to R. L. (Lu 294-2). Present address: Max-Planck-Institut fur Entwicklungsbiologie, Abt. 111, Spemannstr., D-7400 Tiibingen, FRG Present address: Freie Universitat Berlin, Institut fur Humangenetik, Heubnerweg 6, 1 Berlin, FRG Correspondence: Reinhard Liihrmann, Institut fur Molekularbiologie und Tumorforschung der Philipps-Universitat Marburg, Emil-Mannkopff-StraBe 2, D-3550 Marburg, FRG
Abbreviations: MCTD: Mixed connective tissue disease mJG:N2,@, N7-Trimethylguanosine NCS:Newborn calf serum NHS: Normal human serum snRNP: Small nuclear ribonucleoprotein particle(s) 0 VCH Verlagsgesellschaft mbH, D-6940 Weinheim, 1990
against an epitope common to the proteins B', B, and D [13-151, appear to be of diagnostic significance [16]. Beyond that, autoantibodies have been found in sera from patients with SLE, MCTD and less frequently with polymyositis and scleroderma which react with the snRNP proteins A or C and, far less common, with the proteins A', B"or E [ 15,17-22].The origin of the autoantibodies against snRNP proteins and their significance for the pathogenesis of the diseases is not yet known. However, evidence is accumulating that the formation of the autoantibodies against snRNP proteins in rheumatoid diseases is an antigen-driven process and that the U snRNP are recognized as a whole by the immune system ([23-271 see [16] for a review). Here we describe an immunoassay which allows the specific detection of autoantibodies against the three low-molecular mass snRNP proteins, E, Fand G. With this assay we were able to demonstrate for the first time the presence of autoantibodies against the proteins F and G in sera of patients suffering from rheumatoid diseases. The autoantibodies were characterized immunochemically and showed interesting specificities.
2 Materials and methods 2.1 Sera and antibodies Sera were from patients with SLE, diagnosed according to the preliminary ARA criteria.They were obtained from the Nijmegen stock of sera from connective tissue disease patients or kindly provided by Prof. Dr. H. Peter, University of Freiburg. Anti-N2, N2, N7-trimethylguanosine (m3G) antibody used for the purification of U s n R " was purified as described by Bochning et al. [28]. 2.2 Isolation of snRNP proteins U snRNP were purified from HeLa cells by anti-m3G immune affinity chromatography essentially as described by Bringmann et al. [29]. The snRNP proteins were separated by SDS-PAGE and isolated by electroelution from the gel [25].
438
Eur. J. Immunol. 1990.20: 437-440
R. Reuter, S. Rothe,W. Habets et al.
2.3 ELISA (Cl
The ELISA using single gel purified snRNP proteins E, F and G as antigens was essentially performed as described previously for other snRNP proteins [25]. However, polystyrene instead of polyvinylchloride microtiter plates were used. After coating with 10 ng of antigerdwell and saturation with BSA the plates were incubated with patient sera diluted in 0.15 M NaCI, 0.02 M potassium phosphate, pH 7.4,0.1% Tween 20 (PBST) with 5% newborn calf serum (NCS) for 16 h at 4 "C. The plates were then washed four times with PBSTand incubated with a goat anti-human IgG alkaline phosphatase conjugate for 2 h at 22 "C. Bound antibody was detected after a second washing cycle by the color reaction of the phosphatase substrate, p-nitrophenyl phosphate at 22 "C. 2.4 Immunoblot
Proteins of affinity-purified U snRNP (see Sect. 2.2) were separated by electrophoresis in a 15% SDS-polyacrylamide gel and transferred electrophoretically to nitrocellulose [30]. Strips of the nitrocellulose were saturated with 5% NCS in PBS and were incubated with a dilution of the respective patient serum in buffer A (10 mM Tris-HCI, pH 7.4, 150 mM NaCl, 100 mM MgCl2, 0.5% Tween 20, 0.1% Triton X-100, 5% NCS).Then they were washed five times in buffer B (buffer A without Triton X-100 and NCS) and subsequently incubated with an appropriate dilution of an alkaline phosphatase-conjugated goat anti-human IgG antibody in buffer B. After a second washing cycle the strips were stained with a solution of 0.12 mM 5-bromo-4chloro-indolylphosphate, 0.12 mM nitrotetrazolium blue, 0.5 mM MgC12 and 0.1 M diethanolamine, pH 9.8 [31]. 2.5 Affinity purification of antibodies
U snRNP proteins were transferred to a nitrocellulose sheet as described above. The horizontal, 10-cm-wide nitrocellulose strip containing the protein from which antibodies were to be eluted was located relative to two adjacent vertical strips from the same blot that had been immunostained as described above. Then the horizontal strip was saturated, incubated with the serum dilution and washed (see above). Subsequently, the antibodies bound to the strip were eluted with 5 mM glycine, pH 2.5,150 mM NaCl, 0.1% BSA, 0.1% Tween 20, essentially as described by Smith and Fisher [32], and the eluates were immediately neutralized with 1M Tris-HCI, pH 8.
3 Results and discussion 3.1 A sensitive ELISA for the detection of specific autoantibodies against U snRNP proteins We tested 64 sera of patients suffering from systemic rheumatoid diseases for the presence of antibodies against E, F o r G with the ELISA described in Sect. 2.3.Those sera had been pre-selected by their ability to precipitate snRNP U1 to U6 from nuclear extracts. In Fig. 1 examples of titrations of two patient sera are displayed in comparison to a normal human serum (NHS). The graphs show that the
NHS 0.5
0'4 0.3
I\
I\ 6
7
8
9
10
6
7
8
9
10
6
7
8
9
10
-log3 of serum dilution
Figure I. Specificity and sensitivity of the ELISA for the detection of autoantibodies against the snRNP proteins E, F and G. Patient serum M.M. (A), P.B. (B) or NHS (C) were serially diluted (1 : 600, 1 : 1800,etc.) and reacted with snRNP proteins E, For G coated on wells of polystyrene microtiter plates as described in Sect. 2.3. Bound antibody was detected by ELISA techniques and is presented here by AM5nm after a color development for 3 h following the reaction with the snRNP protein E (closed circles), F (open circles), G (closed square) or no snRNP protein (open square).
ELISA is highly specific and that the autoantibodies discriminate well between the three s n R " proteins. The patient serum M.M. for example only gave a signal against the F snRNP protein (Fig. 1A) and the serum P B. exclusively with the protein G (Fig. 1B) when compared to the low background of the NHS (Fig. 1C). The ELISA is also very sensitive; autoantibodies were detectable against the snRNP proteins at serum dilutions of 1:5000 or higher (Fig. 1A and B). Under the conditions chosen the signal of the ELISA depended strongly on the concentration of the autoantibody in the sera, and the titration curves of all patient sera reactive against one respective protein were approximately of the same shape (data not shown). Thus, the signal at a given dilution gave an estimate of the concentration of the autoantibody in the serum and allowed a semiquantitative comparison of its concentration in different sera.
3.2 Differential immune response against the three small U snRNP proteins E, F and G in sera of patients with rheumatic disorders Eight out of the 64 tested patient sera significantly reacted with one or two of the three low-molecular mass snRNP polypeptides E, F, and G (Fig. 2), i.e. the signal was more than three times higher than the control signal obtained with the given serum in microtiter wells without an snRNP protein coat. Three sera (A.D., A.H., R.H.) contained autoantibodies against the E protein, four sera (A.D., S.R., H.T., M.M.) against the F protein and two sera (PB., J.R.) against the G protein. Thus, autoantibodies exist in sera of patients suffering from systemic rheumatoid diseases which react with the low-molecular mass snRNP proteins E, Fand G. Apparently, the immune system of the patients is able to form autoantibodies against any of the known polypeptide components contained in the U snRNF! The autoimmune response of the patients against the three low-molecular mass snRNP polypeptides E , F, and G varies among
Eur. J. Immunol. 1990. 20: 437-440
Autoantibodies against snRNP proteins E, F and G
Y L
1
A 0.
S.R. H.T.
PB.
MM
AH.
R.H.
J.R.
NHS
Figure 2. Differential autoimmune response of SLE patients against snRNP proteins E , Fand G as detected by an ELISA. Serial dilutions of patient sera were reacted with snRNP proteins E , For G in microtiter plates as described in Sect. 2.3. Displayed is the quotient of Am5 nm found for the 1 :600 serum dilution after reaction with the snRNP proteins E , F or G (stippled, open or cross-hatched columns) and of AM^ nm found for the same dilution after reaction in a control microtiter well coated with buffer and BSA only.
individuals (Fig. 2). We do not know whether this differential autoimmune response has a diagnostic relevance. However, we assume that it reflects the individual immunological responsiveness of the patients towards U snRNP. To compare the sensitivity and specificity of ELISA and immunoblotting, we also tested the reactivity of the eight patient sera containing autoantibodies against the s n R ” proteins, E , F or G by immunoblotting (Fig. 3). When snRNP polypeptides which had been immobilized on nitrocellulose after separation in an SDS-polyacrylamide gel were probed with low dilutions (1:25) of the eight sera only five of them showed a detectable reaction with the low-molecular mass snRNP proteins. In good agreement with the ELISA data, patient serum A.D. bound to proteins E and F (Fig. 3, lane l),patient sera S.R., H.T. and M.M. to protein F (lanes 2 , 3 and 5 ) and patient serum P.B.
Figure 3. The differential autoimmune response of SLE patients against snRNP proteins, E, F and G is partially detectable by an immunoblot. U s n W were purified from HeLa cells and their proteins were separated by SDS-PAGE and transferred to nitrocellulose. Subsequently, strips of nitrocellulose were incubated with a 1 :25 dilution of the indicated patient sera, and bound antibodies were detected by immunostaining (see Sect. 2.4).
439
to protein G (lane 4). However, no binding to E , For G was detectable for the patient sera A.H., R.H. and J.R. The lack of binding may indicate the lower sensitivity of the immunoblot compared to the ELISA. The sera from these three patients gave the weakest signal in the ELISA reflecting low concentrations of autoantibodies against E, F or G snRNP proteins which may be not sufficient to be detected by the immunoblot. However, qualitatively the immunoblot gave the same results with respect to the specificity of the autoantibodies contained in the patient sera as the ELISA inspite of its lower sensitivity.There was no indication for systematic differences of the two assays which would lead to the failure to detect autoantibodies against E , F o r G protein in the patient sera A.H., R.H. or J.R. 3.3 Immunochemical properties of the autoantibodies against snRNP protein E, F and G
To determine the specificity of the autoantibodies against the snRNP proteins E, F or G and to detect possible cross-reactivity among the snRNP proteins we affinity purified the respective antibodies. When we eluted autoantibodies of the patient sera M.M. (Fig. 4, lane 3) or H.T. from the F protein immobilized on nitrocellulose and subsequently reacted the antibodies with snRNP proteins on a nitrocellulose strip, they reacted specifically with the F protein only (Fig. 4, lane 4). There was no detectable cross-reactivity with other snRNP proteins.When autoantibodies of the patient sera A.D. (Fig. 4,lane 5), A.H. or R.H. were eluted from the E protein, they did not react detectably with the E protein or any other snRNP protein
Figure 4. The specificity of the autoantibodies against the snRNP proteins Fand G. Autoantibodies reacting with the snRNP protein G were affinity purified from patient serum P.B. (lane l), then reacted with U snRNP proteins transferred to a nitrocellulose strip and detected by immunostaining (lane 2), as described in Sect. 2.4 and 2.5. Correspondingly, autoantibodies against the snRNP protein F were purified from patient serum M.M. (lane 3) and against the snRNP proteins E from patient serum. A.D. (lanes 5) and tested for their specificity (lanes 4 and 6, respectively).
440
R. Reuter, S. Rothe, W. Habets et al.
(Fig. 4, lange 6).Therefore, we could not draw a conclusion as to the specificity of the autoantibodies reactive to the E protein. The autoantibodies against the protein G, however, showed a strong cross-reactivity with the 70-kDa protein (Fig. 4, lanes 1, 2). We are convinced that this property is specific to the antibody against G and reveals an epitope common to the 70-kDa and the G protein since none of the other autoantibodies in patient serum P.B., e.g. against the D protein, showed a comparable cross-reactivity when eluted from the respective protein bands (data not shown). The cross-reactivity of the antibody against the G protein was somewhat surprising since a polyclonal antiserum raised against the 70-kDa protein in rabbits was highly specific for that protein [33]. Presumably, the respective cross-reactive epitope is not particularly immunogenic in the context of the 70-kDa protein. When in the reciprocal experiment autoantibodies of the patient serum P.B. were eluted from the 70-kDa protein they bound only this. We assume that only a minor fraction of the 70-kDa reactive autoantibodies in the serum cross-reacts with the G protein and may, therefore, be under-represented in the antibodies eluted from the 70-kDa protein.
4 Concluding remarks We have introduced an ELISA which allows the specific and sensitive detection of autoantibodies against the three low-molecular mass U snRNP proteins, E, F and G. By these means, we were able to demonstrate for the first time the presence of autoantibodies against these proteins in sera of patients suffering from systemic rheumatoid disease. Thus, autoantibodies can be formed against every individual snRNP polypeptide known so far which supports the notion that the U snRNP are recognized as a whole by the patient's immune system. The antibody against the F protein appears to be highly specific for the polypeptide. In contrast, the autoantibody binding to the G protein showed a strong cross-reactivitywith the 70-kDa protein revealing a hitherto unnoticed immunological relationship between these proteins. Immunological cross-reactivity seems to be a more general phenomenon among snRNP proteins. The U 1 s n R W proteins A and C share at least one epitope with each other and with the human proteins B'/B [27]. The latter two proteins share a different epitope with protein D [13, 151. Furthermore, antibodies cross-reacting with proteins A and B" have been described [34]. As has been shown for other anti-snRNP antibodies, anti-F and anti-G autoantibodies will be helpful tools for structural studies on the U snRNP and for the identification of cDNA clones coding for the snRNP proteins F or G. We are indebted to Dr. H . H. Peter (Institut fur Klinische Immunologie, Freiburg) for providing patients' sera. We thank Dr. D. Andrew for critical reading of the manuscript. R. R. is most grateful to Dr. M. Scott for the opportunity of finishing the manuscript in his laboratory at the University of Colorado.
Received July 31, 1989; in revised form October 30, 1989.
Eur. J. Immunol. 1990.20: 437-440
5 References 1 Sharp, F! A., Science 1987. 235: 766. 2 Maniatis, T. and Reed, R., Nature 1987. 325: 673. 3 Steitz, J. A., Black, D. L., Gerke,V., Parker, K. A., Kramer, A., Frendewey, D. and Keller,W., in Birnstiel, M. L. (Ed.), Structure and Function of Major and Minor Small Nuclear Ribonucleoprotein Particles. Springer-Verlag, Berlin, New York 1988, p. 115. 4 Lerner, M. R. and Steitz, J. A., Proc. Natl. Acad. Sci. USA 1979. 76: 5495. 5 Hashimoto, C. and Steitz, J. A., Nucleic Acids Res. 1984. 12: 3283. 6 Bringmann, P., Appel, B., Rinke, J., Reuter, R.,Theissen, H. and Luhrmann, R., EMBO J. 1984. 3: 1357. 7 Rinke, J., Appel, B., Digweed, M. andLuhrmann, R., J. Mol. Biol. 1985. 185: 721. 8 Liihrmann, R., in Birnstiel, M. L. (Ed.), Structure and Function of Major and Minor Small Nuclear Ribonucleoprotein Particles. Springer-Verlag, Berlin, New York 1988, p. 71. 9 Tazi, J., Alibert, C.,Temsamani, J., Reveillaud, I., Cathala, G., Brunel, C. and Jeanteur, P., Cell 1986. 47: 755. 10 Gerke.V. and Steitz. J. A.. Cell 1986. 47: 973. 11 Bach, M., Winkelmann, G. and Luhrmann, R., Proc. Natl. Acad. Sci. USA 1989. 86: 6038. 12 Pettersson, I. ,Wang, G., Smith, E. I.,Wigzell, H., Hedfors, E., Horn, J. and Sharp, G. C., Arthritis Rheum. 1986.29: 986. 13 Lerner, E. A., Lerner, M. R., Janeway, J. A., Jr. and Steitz, J. A., Proc. Natl. Acad. Sci. USA 1981. 78: 2737. 14 Guldner, H. H., Lakomek, H.-J. and Bautz, F. A., J. Immunol. Methods 1983. 64: 45. 15 Habets,W., Berden, J. H. M., Hoch, S. 0.andVanVenrooij,W., Eur. J. Immunol. 1985. 15: 992. 16 Tan, E. M., Adv. Immunol. 1989. 44: 93. 17 Conner, G. E., Nelson, D., Wisniewolski, R., Lahita, R. G., Blobel, G. and Kunkel, H. G., J. Exp. Med. 1982. 156: 1475. 18 Billings, P. B. and Hoch, S. O., J. Immunol. 1983. 131: 347. 19 Kinlaw, C. S., Robbersen, B. L. and Berget, S. M., J. Biol. Chem. 1983. 258: 7181. 20 Pettersson, I . , Hinterberger, M., Mimori, T., Gottlieb, E. and Steitz, J. A., J. Biol. Chem. 1984. 259: 5907. 21 Mimori, T., Hinterberger, M., Pettersson, I. and Steitz, J. A., J. Biol. Chem. 1984. 259: 560. 22 Wieben, E. D., Rohleder, A. M., Nenninger, J. M. and Pederson, T., Proc. Natl. Acad. Sci. USA 1985. 85: 7914. 23 Cohen, I? L. and Eisenberg, R. A., J. Immunol. 1982. 129: 2142. 24 Bernard, N. F., Eisenberg, R. A. and Cohen, P. L., J. Immunol. 1985. 134: 3812. 25 Reuter, R. and Luhrmann, R., Proc. Natl. Acad. Sci. USA 1986. 83: 8689. 26 Habets,W. J., Sillekens, P.T. G., Hoet, M. H., McAUister, G., Lerner, M. R. and VanVenrooij,W., Proc. Natl. Acad. Sci. USA 1989. 86: 4674. 21 Habets,W. J., Hoet, M. H., De Jong, B. A. W. ,Van der Kemp, A. and Van Venrooii,W. J., J. Immunol. 1989, in press. 28 Bochnig, P., Reuteri R., Bringmann, P. and Liihkann, R., Eur. J. Biochem. 1987. 168: 461. 29 Bringmann, P., Rinke, J., Appel, B., Reuter, R. and Liihrmann, R., EMBO J. 1983. 2: 1129. 30 Towbin, H., Staehelin,T. and Gordon, J., Proc. Natl. Acad. Sci. USA 1979. 76: 4350. 31 McGadey, J., Histochemie 1984. 23: 180. 32 Smith, D. E. and Fisher, P. A., J. Cell Biol. 1984. 99: 20. 33 Reuter, R., Rothe, S. and Luhrmann, R., Nucleic Acids Res. 1987. 15: 4021. 34 Habets,W. J., Hoet, M., Bringmann, P.. Luhrmann. R. and Van Venrooij, W., EMBO J. 1985. 4: 1545.
