Human monoclonal IgM directed against lamins
Eur. J. Immunol. 1992. 22: 1547-1551
Kaiss LassouedeA, Chantal AndrCO, Franqoise DanonA, Robert Modiglianiv, Daniel DhumeauxO, Jean-Pierre ClauvelA, Jean-Claude BrouetA and Jean-Claude Courvalina Laboratoire d’Immunochimie et ImmunopathologieA, U108 INSERM and DCpartement de Gastro-Ent6rologiev, HGpital Saint-Louis, Paris, Laboratoire d’HCmatologie et ImmunologieO and DCpartement d’HCpatologieO, HGpital Henri Mondor, CrCteil and Centre de GCnCtique MolCculairea, CNRS, Gif sur Yvette
1547
Characterization of two human monoclonal IgM antibodies that recognize nuclear lamins Using immunofluorescence and immunoblotting techniques, we have identified monoclonal IgM h from two patients that are specific for lamins A and C and lamin B, respectively. Lamins A , B, and C are peripheral membrane proteins of the nuclear envelope with structural similarities to cytoplasmic intermediate filament proteins. When studied by indirect immunofluorescence on rat tissues, the serum containing anti-lamin B IgM stained smooth and striated muscles in addition to nuclear envelopes. Lamin B antibodies affinity purified from this serum were able to label muscle cells, suggesting that lamin B shares an epitope(s) with an unidentified muscular component(s). Since in an enzyme-linked immunosorbent assay there was no reactivity with a panel of proteins which are frequent targets of “natural” antibodies, these monoclonal IgM appear to belong to the rare category of IgM that possess a restricted specificity.
1 Introduction
2 Materials and methods
Human monoclonal IgM possess a defined antibody activity in nearly 25% of the patients in which they are found [l]. Identification of the autoantigens, as well as of the IgM repertoire, is of interest since it may help to elucidate the relationship between certain antibody activities and clinical symptoms and the pathogenesis of associated lymphoproliferative disorders. Self antigens reactive with monoclonal IgM include the Fc fragment of IgG [2], the red cell li antigen [3], nuclear antigens such as double- or singlestranded DNA [4, 51, myelin-associated glycoprotein and nerve glycolipids [6, 71 and various polypeptides of the cytoskeleton [8]. Monoclonal IgM specific for actin or tubulin frequently react with other auto- or heteroantigens, as do “natural” or “polyspecific” antibodies in normal sera [9]. In contrast, a monoclonal IgM directed against vimentin, a cytoplasmic intermediate filament protein, has been identified and shown to be of a restricted specificity [S].
2.1 Patients and sera
Lamins A, B and C are the major structural proteins of the nuclear envelope. From their primary and secondary structure and their self-associative properties, the lamins have been shown to belong to a class of intermediate filament polypeptides [lo-131. We report here the characterization of two IgM directed against lamins A and C and lamin B, respectively, which exhibit a restricted specificity.
[I 101831 Recipient of a grant from Le Fonds d’Etudes et de Recherches du Corps MCdical des HBpitaux de Pans.
Correspondence: Jean-Claude Brouet, U. 108 INSERM, HBpital Saint-Louis, 1, avenue Claude Vellefaux, F-75475 Paris Cedex 10, France
0 VCH Verlagsgesellschaft mbH, D-6940 Weinheim, 3992
Patient G suffered from a cryptogenic liver cirrhosis and patient H from a primary exsudative enteropathy caused by intestinal lymphangiectasies. In both sera, immunoelectrophoresis detected a monoclonal IgM. The serum concentration of the IgM was 35 mg/ml and 1.5 mg/ml, respectively. No lymphoplasmocytic proliferation was found on bone marrow biopsies. Ring-like antinuclear antibodies (ANA) were detected using dilutions up to 1: 40000 and 1: 10000, respectively. Anti-double-stranded DNA (dsDNA), antiextractable nuclear antigen (ENA) and anti-centromere antibodies were absent from both sera. Control human sera included normal sera from a pool of 25 donors, purified IgG and IgM molecules from normal serum, a serum containing a monoclonal IgM without known antibody activity, a serum containing anti-smooth muscle antibodies (ASMA) and a serum containing anti-M2 mitochondria1 antibodies. Polyclonal anti-lamins A and C or anti-lamin B sera have been previously characterized [14, 151. A murine monoclonal antibody (CC212) reactive with myosin heavy chain was also used [16].
2.2 Detection of autoantibody specificities ANA were detected by indirect immunofluorescence (IF) on air-dried rat liver tissue sections, Hep2 cells, and HeLa cells using rabbit IgG specific for human Ig heavy and light chains (Dako, Copenhagen, Denmark). Anti-smooth, antistriated muscles and anti-mitochondria1 antibodies were detected by IF on unfixed composite tissue blocks of snap-frozen rat kidney, stomach, diaphragm, heart, sciatic nerve, testis and human salivary glands. Anti-ds-DNA antibodies were detected by IF on the kinetoplast of Crithidia luciliae [17] and by the Farr assay [18]. Antibodies to extractable nuclear antigens were evaluated by immunoblotting on HeLa cell nuclear fractions [15] and by double immunodiffusion [19, 201 using calf thymus (Northeast, Uxbridge, GB) and human spleen saline-soluble extracts as 00 14-2980/92/0606-1547$3.SO + .25/0
1548
K . Lassoued, C. Andr6, F. Danon et al.
a source of antigens and reference antisera to Sm, U1sn-RNP, La/SSB, Ro/SSA. So-called “natural” antibodies directed against actin, tubulin and myosin were evaluated by ELISA as previously described [9]. 2.3 Nuclear envelope fractionation Rat liver nuclei were prepared according to Blobel and Potter [21]. Nuclear envelopes and pore complex-lamina fractions were obtained according to Dwyer and Blobel [22]. Further fractionation of nuclear envelopes was performed as previously described [23]. Lamin B was solubilized in 6 M urea and purified on DEAE-cellulose according to Reeves et al. [24].
2.4 Preparation of muscle homogenates Mouse esophagus, bowel and diaphragm were pulverized with a mortar and pestle in the presence of liquid nitrogen and then homogenized in SDS electrophoresis sample buffer in a glass-glass homogenizer. The homogenate was then sonicated.
2.5 Immunoblotting Protein fractions were subjected to electrophoresis on 8% polyacrylamide gels and proteins were transferred to nitrocellulose sheets by electrophoresis using a semi-dry method [25]. All the following steps were carried out in 10 mM Tris-HC1 (pH 7.4), 0.15 M NaCl, 0.1% Tween-20 and 50 g/l nonfat milk. Nitrocellulose strips were blocked for 1 h, incubated for 2 h with the indicated dilution of serum, washed four times and finally revealed with peroxidase-labeled sheep IgG specific for human Ig or p chains (Institut Pasteur, Paris) or by 1251-proteinA (NEN, Wilmington, DE) followed by autoradiography. 2.6 Affinity purification of IgM and absorption experiments Immunoblots were performed on proteins of rat liver nuclear envelopes using serum of patient G at a 1: 100 dilution. The band corresponding to the 68-kDa lamin B was excised and the antibodies were eluted according to Smith and Fisher [26]. A strip of the same blot in the 150-kDa region was eluted under the same conditions and the eluate was used as a control. Serum of patient G was absorbed with rat liver nuclei as previously described [14]. A serum with a high titer (1 : 20000) in ASMA and devoid of ANA was absorbed under the same conditions and used as a control.
3 Results 3.1 IgM in both patients are directed against lamins Using IFon rat liver tissue sections, sera of patients G and H displayed a ring-like staining pattern down to an IgM concentration of 1 pg/ml, i.e. up to a dilution of 1: 40000
Eur. J. Immunol. 1992. 22: 1547-1551
and 1: 10000, respectively. This labeling pattern is characteristic for components of the nuclear envelope. When studied with monospecific sera to Ig chains, the perinuclear
Mr k D a 1
2
3
5
4
20011692-
A -B
J
66-
-C
45-
31 Figure 1. Immunoblot of rat liver nuclear pore complex-lamina fraction with sera of patients H and G . A pore complex-lamina fraction was prepared as indicated in Sect. 2.5. Polypeptides in these fractions were separated by SDS-PAGE and visualized by Coomassie blue staining (lane 2) or were transferred to nitrocellulose (lanes 3-5) and probed with a 1: 100 dilution of the serum of patient H (lane 3), patient G (lane 4) or a pool of normal human sera (lane 5). A , B, and C refer t o lamins A , B, and C, respectively. Lane 1 contains molecular mass markers (Bio-Rad).
Coomassie Blue 0.1N NaOH MW NE
S
P
LB
3
4
5
lmmunoblot 0.1 N NaOH NE S
P
LB
6
8
9
200 116 -
9’ 66 43
1
2
7
Figure 2. Immunoblotting analysis of purified lamin fractions with serum of patient G. Rat liver nuclear envelopcs were extracted with 0.1 N NaOH (as indicated in the figure) and centrifuged to yield supernatant (S) and pellet (P) fractions. Polypeptides in these fractions or of unfractionated nuclear envelopes (NE), and purified lamin B (LB), were run on SDS-PAGE and visualized by Coomassie blue staining (lanes 2-5) or were transferred to nitrocellulose (lanes 6-9) and probed with serum of patient G. Lane 1 shows migration of molecular mass markcrs (Bio-Rad). In lane 3 and 7, the shift in mobility of lamin B is due to chemical modification of this polypeptide upon NaOH extraction. Lane 1 contains molecular mass markers (Bio-Rad).
Eur. J. Immunol. 1992. 22: 1547-1551
Human monoclonal IgM directed against lamins
1549
labeling was observed with anti-p and anti-h antisera only; such staining was not observed with normal sera or purified IgG and IgM fractions. To identify the autoantigens, the proteins of a rat liver pore-complex nuclear fraction were isolated and separated by SDS-PAGE. The separated proteins were stained with Coomassie blue (Fig. 1, lane 2), or transferred to nitrocellulose sheets and probed with the serum of both patients. Two polypeptides with apparent molecular masses of 74 kDa and 60 kDa were found to react with the serum of patient H (Fig. 1, lane 3). This reaction pattern is characteristic for lamins A and C. A polypeptide with an apparent molecular mass of 68 kDa, similar to that of lamin B, was revealed by the serum of patient G (Fig. 1, lane 4). Since other nuclear envelope proteins migrate in the 68-kDa range, rat liver nuclear envelopes were extracted with alkali to separate peripheral proteins (soluble) (Fig. 2, lanes 3 and 7) from integral membrane proteins (insoluble) (Fig. 2 , lanes 4 and 8). The 68-kDa antigen was solubilized by alkaline extraction (Fig. 2, lane 7) and, therefore, like lamins, fractionates as a peripheral membrane protein. To complete the identification, lamin B was purified from rat liver nuclear envelope to near homogeneity as judged by SDS-PAGE (Fig. 2 , lane 5). Immunoblot analysis (Fig. 2, lane 9) showed that serum from patient G reacted with purified lamin B, further demonstrating that it is the target of this particular autoimmune serum. 3.2 IgM specific for lamin B also reacts with a cytoplasmic muscular component IF performed on rat tissue sections with the serum of patient G showed a staining of smooth-muscle fibers (muscularis mucosae) and muscle fibers of rat stomach and blood vessel walls of different tissues (Fig. 3 a-c) as well as myoepithelial cells of the human salivary glands (Fig. 4 a ) . The titer of these antibodies was high (1 : 20000). Striated muscle (diaphragm and heart) was also labeled using a dilution of 1 : 1000 (Fig. 3 d). Glomeruli, basal membrane of the tubular epithelial cells of kidney, neural tissue and testis were not stained (data not shown). The staining was
Figure 3. IFof rat tissue sections using a 1 :SO dilution of serum of patient G. IF wab performed o n tissue sections of stomach (a), esophagus (b), kidney (c) and diaphragm (d). Magnification: 400 x .
Figure 4. IF using affinity-purified IgM of patient G. IgM was affinity purified using immobilized rat lamin B as indicated in Sect. 2.2. Immunofluoresccnce was performed on human salivary gland scctions (a, c, d) and rat stomach section (b).The antibodies used were either total serum G (a), or affinity-purified IgM (b and c) and control affinity-purified Ig (d). Note that both nuclear envelopes (b) and smooth muscle (c) arc labeled by the affinitypurified IgM. Magnification: a, c, and d X 560; b, X 400.
only obtained with anti-p and anti-h second antibodies, suggesting that both the nuclear and muscular specificities were carried by the same monoclonal IgM. To check this possibility, serum from patient G was absorbed with rat liver nuclei and then tested by I F on rat tissue. Both the nuclear and muscular reactivities disappeared after serum absorption (data not shown). This absorption was specific since anti-smooth muscle antibodies of a control serum were not absorbed under the same experimental conditions. To substantiate further the fact that the same IgM accounted for both nuclear and muscular specificities the patient’s serum was affinity purified using rat lamin B as a ligand and the affinity-purified IgM were tested by IFon rat liver and stomach sections and on human salivary glands. The affinity-purified antibodies were able to stain the nuclear envelope (Fig. 4 b) as well as myoepithelial cells (Fig. 4c). When matcrial eluted from proteins other than lamin B was analyzed by IF, no cellular labeling was observed (Fig. 4d).
To identify the muscular antigen(s), whole proteins from mouse esophagus, bowel and diaphragm were solubilized and separated by SDS-PAGE and transferred to nitrocellulose sheets and probed with serum of patient G. No polypeptide was found to react with the autoantibodies, whereas, the same preparaiion was found to react with control anti-myosin antibodies (data not shown). Several attempts to identify the muscular antigen were unsuccessful.
1550
K. Lassoued, C. AndrC, F. Danon et al.
4 Discussion The nuclear lamina is a protein meshwork situated between the inner nuclear membrane and heterochromatin. In higher eukaryotic cells, the lamina consists primarily of a polymer of proteins termed lamins, which are biochemically and structurally similar to the cytoplasmic intermediate filaments [lo-131. The lamins are divided into two major classes: the acidic B-type lamins and the neutral A-type lamins. In mammalian somatic cells, lamin B consists of a single polypeptide with a M, of 67 kDa, whereas, type A lamins are composed of two closely related polypeptides, lamin A (74 kDa) and lamin C (65 kDa), which are coded by a single gene transcribed into two mRNA by differential splicing [lo, 111. Human autoantibodies to lamins are of restricted specificity, being mainly directed against either lamin B, lamins A and C, or lamin A [14,15,24,27-29].This suggests a production of antibodies restricted to one or a few epitopes. A predominance of either 3c and h chains isotypes has been noted for lamins polyclonal antibodies [29, 301. Polyclonal autoantibodies to lamins are mainly found in certain autoimmune disorders [15,28] featured by the presence of vasculitis, peripheral cytopenia and/or liver diseases. We report here the cytological and biochemical characterization of two human monoclonal IgM h specific for lamins B and lamins A and C, respectively. The IgM specific for lamin B also stained smooth or striated muscle fibers and myoepithelial cells. In view of the structural homologies between the lamins and the intermediate filaments, an appealing hypothesis would be that these antibodies react with either muscle-specific intermediate filaments or associated proteins. However, attempts to identify the muscular antigen were unsuccessful. The peculiar specificity of an IgM for both lamin B and smooth muscle indicates that this IgM is directed to a unique epitope of lamin B since polyclonal antibodies to lamin B do not share this dual reactivity. Interestingly the anti-lamin B monoclonal IgM features h light chains as do predominantly polyclonal IgG autoantibodies directed against lamin B [30]. Polyclonal autoantibodies to lamin B from a patient have been shown previously to be directed against epitopes common to a few peripheral membrane proteins that share with lamin B the characteristic of being intermediate filaments-binding proteins [31, 321. The monoclonal IgM of this study provides a new example of cross-reactivity between lamin B and another cellular component. We are currently trying to identify this crossreactive component with the expectation that the common epitope will also be of functional value. Some insights into the origin of monoclonal antibodies had been gained from the elucidation of their antibody activity. The most frequent is the so-called natural or polyreactive activity [9] also found in the repertoire of fetal and adult B cells. Monoclonal rheumatoid factors may, although to an unknown extent, belong to this category of antibodies. Structural features indicate that the latter antibodies are encoded by a small number of conserved, usually unmutated V genes [33]. On the other hand, some monoclonal IgM, such as those directed to myelin-associated glycoprotein have a unique specificity and use a rather large repertoire of V genes which exhibit an unusual degree of
Eur. J. Immunol. 1992. 22: 1547-1551
mutations [34], possibly indicating a distinct pathogenesis for the underlying lymphoproliferative process. The two IgM specific for lamins reported in this study probably belong to the latter group of monoclonal Ig since they possess a highly restricted specificity. The emergence of anti-lamin antibody may be related either to a preexisting cellular damage or secondary to an abnormal immune response to a processed internal antigen. We would like to thank Dr. N. Chaudhary and for the gift of purified lamin B. We thank Mrs. Yvette Signoret, FranGoise Sanch and Sylvie Bourahla for technical assistance and Muriel Bargis-Touchard for typing the manuscript. We thank Dr. H. Worman for reviewing our manuscript. Received December, 5 , 1991; in final revised form February 26, 1992.
5 References 1 Seligmann, M. and Brouet, J. C., Semin. Hematol. 1973. 10: 163. 2 Kunkel, H. G ., Agnello,V., J o s h , F. G.,Winchester, R. J. and Capra, J. D., J. Exp. Med. 1973. 137: 331. 3 Williams, R. C., Kunkel, H. G. and Capra, J. D., Science 1968. 161: 379. 4 Fermand, J. I?, Danon, F. and Brouet, J. C., Clin. Exp. Immunol. 1985. 59: 467. 5 Zouali, M., Fine, J. M. and Eyquem, A., Eur. J. Irnmunol. 1984. 14: 1085. 6 Braun, I? E . , Frail, D. E. and Latov, N., J. Neurochem. 1982. 39: 1261. 7 Ilyas, A. A., Quarles, R. H., MacIntosh,T. D., Dobersen, M. J. ,Trapp, B. D., Dalakas, M. C. and Brady, R. O., Proc. Natl. Acad. Sci. USA 1984. 81: 1225. 8 Dellagi, K., Brouet, J. C., Perreau, J. and Paulin, D., Proc. Natl. Acad. Sci. USA 1982. 79: 446. 9 Dighiero, G., Guilbert, B., Fermand, J. I?, Lymberi, I?, Danon, F. and Avrameas, S., Blood 1983. 62: 264. 10 McKeon, F. D., Kirschner, M. W. and Caput, D., Nature 1986. 31 9: 463. 11 Fisher, D. Z., Chaudhary, N. and Blobel, G., Proc. Natl. Acad. Sci. USA 1986. 83: 6450. 12 Krohne, G ., Wolin, S. L., McKeon, F. D., Franke, W. W. and Kirschner, M. W., EMBO J. 1987. 6: 3801. 13 Pollard, K. M., Chan, E. K. L., Grant, B. J., Sullivan, K. F., Tan, E. M. and Glass, C. A., Mol. Cell. Biol. 1990. 10: 2164. 14 Guilly, M.-N., Danon, F., Brouet, J. C., Bornens, M. and Courvalin, J. C., Eur. J. Cell Biol. 1987. 43: 266. 15 Lassoued, K., Guilly, M.-N., Danon, F., Andr6, C., Dhumeaux, D., Clauvel, J. I?, Brouet, J. C., Seligmann, M. and Courvalin, J. C., Ann. Intern. Med. 1988. 108: 829. 16 Klotz, C., Bordes, N., LainC, M. C., Sandoz, D. and Bornens, M., J. Cell. Biol. 1986. 103: 613. 17 Aarden, L. A ., De Groot, E. R. and Feltkamp,T. E.W., Ann. N. Y Acad. Sci. 1975. 254: 505. 18 Peltier, A. P., Hughes, G . R. V., Kahn, M. E and Haim, T., Nouv. Presse, Mkd. 1974. 3: 649. 19 Clark, G., Reichlin, M. and Tomasi, Jr., T. B., J. Immunol. 1969. 102: 117. 20 Moore,T. L.,Weiss,T. D., Neucks, S. H., Baldossare, A. R. and Zuckner, J., Semin. Arthritis Rheum. 1981. 10: 309. 21 Blobel, G. and Potter,V. R., Ann. Intern. Med. 1966.108: 829. 22 Dwyer, N. and Blobel, G., J. Cell Biol. 1976. 70: 581. 23 Courvalin, J. C., Lassoued, K., Bartnik, E., Blobel, G. and Wozniak, R . , J. Clin. Invest. 1990. 86: 279. 24 Reeves,W. H., Chaudhary, N. L., Salerno, A. and Blobel, G., J. Exp. Med. 1987. 165: 750.
Eur. J. Immunol. 1992. 22; 1547-1551 25 Kyhse-Andersen, J., J. Biochem. Biophys. Methods 1984. 10: 203. 26 Smith, D. E . and Fisher, P. A., J. Cell Biol. 1984. 99: 20. 27 McKeon, F. D.,Tuffanelli, D. L., Fukuyama, K. and Kirshner, M. W., Proc. Natl. Acad. Sci. USA 1983. 80: 4374. 28 Wesierska-Gadek, J., Penner, E., Hitchman, E. and Sauermann, G., Clin. Immunol. Immunopathol. 1988. 49: 107. 29 Courvalin, J. C., Chaudhary,N. L., Danon,F.,Brouet, J. C. and Lassoued, K . , Biol. Cell. 1990. 69: 93. 30 Reeves, W. H. and Syed Asad, A. P., J. Immunol. 1989. 143: 3614.
Human monoclonal IgM directed against lamins
1551
31 Cartaud, A., Courvalin, J. C.., Ludosky, M. A. and Cartaud, J., J. Cell Riol. 1989. 109: 1745. 32 Cartaud, A ., Ludosky, M. A., Courvalin, J. C. and Cartaud, J., J. Cell Biol. 1990. 111: 581. 33 Chen, P. Pl., Albrandt, K., Orida, N. K., Radoux,V, Chen, E. Y., Schrantz,V., Liu, F. T. and Carson, D. A., Proc. Natl. Acud. Sci. USA 1986. 83: 8318. 34 Mihaesco, E., Ayadi, H ., Congy, N., Gendron, M. C., Roy, J. P., Heyermann, H., Frangione, B. and Brouet, J. C., J. Biol. Chem. 1989. 264: 21481.
Eur. J. Immunol. 1992. 22: 1547-1551
Kaiss LassouedeA, Chantal AndrCO, Franqoise DanonA, Robert Modiglianiv, Daniel DhumeauxO, Jean-Pierre ClauvelA, Jean-Claude BrouetA and Jean-Claude Courvalina Laboratoire d’Immunochimie et ImmunopathologieA, U108 INSERM and DCpartement de Gastro-Ent6rologiev, HGpital Saint-Louis, Paris, Laboratoire d’HCmatologie et ImmunologieO and DCpartement d’HCpatologieO, HGpital Henri Mondor, CrCteil and Centre de GCnCtique MolCculairea, CNRS, Gif sur Yvette
1547
Characterization of two human monoclonal IgM antibodies that recognize nuclear lamins Using immunofluorescence and immunoblotting techniques, we have identified monoclonal IgM h from two patients that are specific for lamins A and C and lamin B, respectively. Lamins A , B, and C are peripheral membrane proteins of the nuclear envelope with structural similarities to cytoplasmic intermediate filament proteins. When studied by indirect immunofluorescence on rat tissues, the serum containing anti-lamin B IgM stained smooth and striated muscles in addition to nuclear envelopes. Lamin B antibodies affinity purified from this serum were able to label muscle cells, suggesting that lamin B shares an epitope(s) with an unidentified muscular component(s). Since in an enzyme-linked immunosorbent assay there was no reactivity with a panel of proteins which are frequent targets of “natural” antibodies, these monoclonal IgM appear to belong to the rare category of IgM that possess a restricted specificity.
1 Introduction
2 Materials and methods
Human monoclonal IgM possess a defined antibody activity in nearly 25% of the patients in which they are found [l]. Identification of the autoantigens, as well as of the IgM repertoire, is of interest since it may help to elucidate the relationship between certain antibody activities and clinical symptoms and the pathogenesis of associated lymphoproliferative disorders. Self antigens reactive with monoclonal IgM include the Fc fragment of IgG [2], the red cell li antigen [3], nuclear antigens such as double- or singlestranded DNA [4, 51, myelin-associated glycoprotein and nerve glycolipids [6, 71 and various polypeptides of the cytoskeleton [8]. Monoclonal IgM specific for actin or tubulin frequently react with other auto- or heteroantigens, as do “natural” or “polyspecific” antibodies in normal sera [9]. In contrast, a monoclonal IgM directed against vimentin, a cytoplasmic intermediate filament protein, has been identified and shown to be of a restricted specificity [S].
2.1 Patients and sera
Lamins A, B and C are the major structural proteins of the nuclear envelope. From their primary and secondary structure and their self-associative properties, the lamins have been shown to belong to a class of intermediate filament polypeptides [lo-131. We report here the characterization of two IgM directed against lamins A and C and lamin B, respectively, which exhibit a restricted specificity.
[I 101831 Recipient of a grant from Le Fonds d’Etudes et de Recherches du Corps MCdical des HBpitaux de Pans.
Correspondence: Jean-Claude Brouet, U. 108 INSERM, HBpital Saint-Louis, 1, avenue Claude Vellefaux, F-75475 Paris Cedex 10, France
0 VCH Verlagsgesellschaft mbH, D-6940 Weinheim, 3992
Patient G suffered from a cryptogenic liver cirrhosis and patient H from a primary exsudative enteropathy caused by intestinal lymphangiectasies. In both sera, immunoelectrophoresis detected a monoclonal IgM. The serum concentration of the IgM was 35 mg/ml and 1.5 mg/ml, respectively. No lymphoplasmocytic proliferation was found on bone marrow biopsies. Ring-like antinuclear antibodies (ANA) were detected using dilutions up to 1: 40000 and 1: 10000, respectively. Anti-double-stranded DNA (dsDNA), antiextractable nuclear antigen (ENA) and anti-centromere antibodies were absent from both sera. Control human sera included normal sera from a pool of 25 donors, purified IgG and IgM molecules from normal serum, a serum containing a monoclonal IgM without known antibody activity, a serum containing anti-smooth muscle antibodies (ASMA) and a serum containing anti-M2 mitochondria1 antibodies. Polyclonal anti-lamins A and C or anti-lamin B sera have been previously characterized [14, 151. A murine monoclonal antibody (CC212) reactive with myosin heavy chain was also used [16].
2.2 Detection of autoantibody specificities ANA were detected by indirect immunofluorescence (IF) on air-dried rat liver tissue sections, Hep2 cells, and HeLa cells using rabbit IgG specific for human Ig heavy and light chains (Dako, Copenhagen, Denmark). Anti-smooth, antistriated muscles and anti-mitochondria1 antibodies were detected by IF on unfixed composite tissue blocks of snap-frozen rat kidney, stomach, diaphragm, heart, sciatic nerve, testis and human salivary glands. Anti-ds-DNA antibodies were detected by IF on the kinetoplast of Crithidia luciliae [17] and by the Farr assay [18]. Antibodies to extractable nuclear antigens were evaluated by immunoblotting on HeLa cell nuclear fractions [15] and by double immunodiffusion [19, 201 using calf thymus (Northeast, Uxbridge, GB) and human spleen saline-soluble extracts as 00 14-2980/92/0606-1547$3.SO + .25/0
1548
K . Lassoued, C. Andr6, F. Danon et al.
a source of antigens and reference antisera to Sm, U1sn-RNP, La/SSB, Ro/SSA. So-called “natural” antibodies directed against actin, tubulin and myosin were evaluated by ELISA as previously described [9]. 2.3 Nuclear envelope fractionation Rat liver nuclei were prepared according to Blobel and Potter [21]. Nuclear envelopes and pore complex-lamina fractions were obtained according to Dwyer and Blobel [22]. Further fractionation of nuclear envelopes was performed as previously described [23]. Lamin B was solubilized in 6 M urea and purified on DEAE-cellulose according to Reeves et al. [24].
2.4 Preparation of muscle homogenates Mouse esophagus, bowel and diaphragm were pulverized with a mortar and pestle in the presence of liquid nitrogen and then homogenized in SDS electrophoresis sample buffer in a glass-glass homogenizer. The homogenate was then sonicated.
2.5 Immunoblotting Protein fractions were subjected to electrophoresis on 8% polyacrylamide gels and proteins were transferred to nitrocellulose sheets by electrophoresis using a semi-dry method [25]. All the following steps were carried out in 10 mM Tris-HC1 (pH 7.4), 0.15 M NaCl, 0.1% Tween-20 and 50 g/l nonfat milk. Nitrocellulose strips were blocked for 1 h, incubated for 2 h with the indicated dilution of serum, washed four times and finally revealed with peroxidase-labeled sheep IgG specific for human Ig or p chains (Institut Pasteur, Paris) or by 1251-proteinA (NEN, Wilmington, DE) followed by autoradiography. 2.6 Affinity purification of IgM and absorption experiments Immunoblots were performed on proteins of rat liver nuclear envelopes using serum of patient G at a 1: 100 dilution. The band corresponding to the 68-kDa lamin B was excised and the antibodies were eluted according to Smith and Fisher [26]. A strip of the same blot in the 150-kDa region was eluted under the same conditions and the eluate was used as a control. Serum of patient G was absorbed with rat liver nuclei as previously described [14]. A serum with a high titer (1 : 20000) in ASMA and devoid of ANA was absorbed under the same conditions and used as a control.
3 Results 3.1 IgM in both patients are directed against lamins Using IFon rat liver tissue sections, sera of patients G and H displayed a ring-like staining pattern down to an IgM concentration of 1 pg/ml, i.e. up to a dilution of 1: 40000
Eur. J. Immunol. 1992. 22: 1547-1551
and 1: 10000, respectively. This labeling pattern is characteristic for components of the nuclear envelope. When studied with monospecific sera to Ig chains, the perinuclear
Mr k D a 1
2
3
5
4
20011692-
A -B
J
66-
-C
45-
31 Figure 1. Immunoblot of rat liver nuclear pore complex-lamina fraction with sera of patients H and G . A pore complex-lamina fraction was prepared as indicated in Sect. 2.5. Polypeptides in these fractions were separated by SDS-PAGE and visualized by Coomassie blue staining (lane 2) or were transferred to nitrocellulose (lanes 3-5) and probed with a 1: 100 dilution of the serum of patient H (lane 3), patient G (lane 4) or a pool of normal human sera (lane 5). A , B, and C refer t o lamins A , B, and C, respectively. Lane 1 contains molecular mass markers (Bio-Rad).
Coomassie Blue 0.1N NaOH MW NE
S
P
LB
3
4
5
lmmunoblot 0.1 N NaOH NE S
P
LB
6
8
9
200 116 -
9’ 66 43
1
2
7
Figure 2. Immunoblotting analysis of purified lamin fractions with serum of patient G. Rat liver nuclear envelopcs were extracted with 0.1 N NaOH (as indicated in the figure) and centrifuged to yield supernatant (S) and pellet (P) fractions. Polypeptides in these fractions or of unfractionated nuclear envelopes (NE), and purified lamin B (LB), were run on SDS-PAGE and visualized by Coomassie blue staining (lanes 2-5) or were transferred to nitrocellulose (lanes 6-9) and probed with serum of patient G. Lane 1 shows migration of molecular mass markcrs (Bio-Rad). In lane 3 and 7, the shift in mobility of lamin B is due to chemical modification of this polypeptide upon NaOH extraction. Lane 1 contains molecular mass markers (Bio-Rad).
Eur. J. Immunol. 1992. 22: 1547-1551
Human monoclonal IgM directed against lamins
1549
labeling was observed with anti-p and anti-h antisera only; such staining was not observed with normal sera or purified IgG and IgM fractions. To identify the autoantigens, the proteins of a rat liver pore-complex nuclear fraction were isolated and separated by SDS-PAGE. The separated proteins were stained with Coomassie blue (Fig. 1, lane 2), or transferred to nitrocellulose sheets and probed with the serum of both patients. Two polypeptides with apparent molecular masses of 74 kDa and 60 kDa were found to react with the serum of patient H (Fig. 1, lane 3). This reaction pattern is characteristic for lamins A and C. A polypeptide with an apparent molecular mass of 68 kDa, similar to that of lamin B, was revealed by the serum of patient G (Fig. 1, lane 4). Since other nuclear envelope proteins migrate in the 68-kDa range, rat liver nuclear envelopes were extracted with alkali to separate peripheral proteins (soluble) (Fig. 2, lanes 3 and 7) from integral membrane proteins (insoluble) (Fig. 2 , lanes 4 and 8). The 68-kDa antigen was solubilized by alkaline extraction (Fig. 2, lane 7) and, therefore, like lamins, fractionates as a peripheral membrane protein. To complete the identification, lamin B was purified from rat liver nuclear envelope to near homogeneity as judged by SDS-PAGE (Fig. 2 , lane 5). Immunoblot analysis (Fig. 2, lane 9) showed that serum from patient G reacted with purified lamin B, further demonstrating that it is the target of this particular autoimmune serum. 3.2 IgM specific for lamin B also reacts with a cytoplasmic muscular component IF performed on rat tissue sections with the serum of patient G showed a staining of smooth-muscle fibers (muscularis mucosae) and muscle fibers of rat stomach and blood vessel walls of different tissues (Fig. 3 a-c) as well as myoepithelial cells of the human salivary glands (Fig. 4 a ) . The titer of these antibodies was high (1 : 20000). Striated muscle (diaphragm and heart) was also labeled using a dilution of 1 : 1000 (Fig. 3 d). Glomeruli, basal membrane of the tubular epithelial cells of kidney, neural tissue and testis were not stained (data not shown). The staining was
Figure 3. IFof rat tissue sections using a 1 :SO dilution of serum of patient G. IF wab performed o n tissue sections of stomach (a), esophagus (b), kidney (c) and diaphragm (d). Magnification: 400 x .
Figure 4. IF using affinity-purified IgM of patient G. IgM was affinity purified using immobilized rat lamin B as indicated in Sect. 2.2. Immunofluoresccnce was performed on human salivary gland scctions (a, c, d) and rat stomach section (b).The antibodies used were either total serum G (a), or affinity-purified IgM (b and c) and control affinity-purified Ig (d). Note that both nuclear envelopes (b) and smooth muscle (c) arc labeled by the affinitypurified IgM. Magnification: a, c, and d X 560; b, X 400.
only obtained with anti-p and anti-h second antibodies, suggesting that both the nuclear and muscular specificities were carried by the same monoclonal IgM. To check this possibility, serum from patient G was absorbed with rat liver nuclei and then tested by I F on rat tissue. Both the nuclear and muscular reactivities disappeared after serum absorption (data not shown). This absorption was specific since anti-smooth muscle antibodies of a control serum were not absorbed under the same experimental conditions. To substantiate further the fact that the same IgM accounted for both nuclear and muscular specificities the patient’s serum was affinity purified using rat lamin B as a ligand and the affinity-purified IgM were tested by IFon rat liver and stomach sections and on human salivary glands. The affinity-purified antibodies were able to stain the nuclear envelope (Fig. 4 b) as well as myoepithelial cells (Fig. 4c). When matcrial eluted from proteins other than lamin B was analyzed by IF, no cellular labeling was observed (Fig. 4d).
To identify the muscular antigen(s), whole proteins from mouse esophagus, bowel and diaphragm were solubilized and separated by SDS-PAGE and transferred to nitrocellulose sheets and probed with serum of patient G. No polypeptide was found to react with the autoantibodies, whereas, the same preparaiion was found to react with control anti-myosin antibodies (data not shown). Several attempts to identify the muscular antigen were unsuccessful.
1550
K. Lassoued, C. AndrC, F. Danon et al.
4 Discussion The nuclear lamina is a protein meshwork situated between the inner nuclear membrane and heterochromatin. In higher eukaryotic cells, the lamina consists primarily of a polymer of proteins termed lamins, which are biochemically and structurally similar to the cytoplasmic intermediate filaments [lo-131. The lamins are divided into two major classes: the acidic B-type lamins and the neutral A-type lamins. In mammalian somatic cells, lamin B consists of a single polypeptide with a M, of 67 kDa, whereas, type A lamins are composed of two closely related polypeptides, lamin A (74 kDa) and lamin C (65 kDa), which are coded by a single gene transcribed into two mRNA by differential splicing [lo, 111. Human autoantibodies to lamins are of restricted specificity, being mainly directed against either lamin B, lamins A and C, or lamin A [14,15,24,27-29].This suggests a production of antibodies restricted to one or a few epitopes. A predominance of either 3c and h chains isotypes has been noted for lamins polyclonal antibodies [29, 301. Polyclonal autoantibodies to lamins are mainly found in certain autoimmune disorders [15,28] featured by the presence of vasculitis, peripheral cytopenia and/or liver diseases. We report here the cytological and biochemical characterization of two human monoclonal IgM h specific for lamins B and lamins A and C, respectively. The IgM specific for lamin B also stained smooth or striated muscle fibers and myoepithelial cells. In view of the structural homologies between the lamins and the intermediate filaments, an appealing hypothesis would be that these antibodies react with either muscle-specific intermediate filaments or associated proteins. However, attempts to identify the muscular antigen were unsuccessful. The peculiar specificity of an IgM for both lamin B and smooth muscle indicates that this IgM is directed to a unique epitope of lamin B since polyclonal antibodies to lamin B do not share this dual reactivity. Interestingly the anti-lamin B monoclonal IgM features h light chains as do predominantly polyclonal IgG autoantibodies directed against lamin B [30]. Polyclonal autoantibodies to lamin B from a patient have been shown previously to be directed against epitopes common to a few peripheral membrane proteins that share with lamin B the characteristic of being intermediate filaments-binding proteins [31, 321. The monoclonal IgM of this study provides a new example of cross-reactivity between lamin B and another cellular component. We are currently trying to identify this crossreactive component with the expectation that the common epitope will also be of functional value. Some insights into the origin of monoclonal antibodies had been gained from the elucidation of their antibody activity. The most frequent is the so-called natural or polyreactive activity [9] also found in the repertoire of fetal and adult B cells. Monoclonal rheumatoid factors may, although to an unknown extent, belong to this category of antibodies. Structural features indicate that the latter antibodies are encoded by a small number of conserved, usually unmutated V genes [33]. On the other hand, some monoclonal IgM, such as those directed to myelin-associated glycoprotein have a unique specificity and use a rather large repertoire of V genes which exhibit an unusual degree of
Eur. J. Immunol. 1992. 22: 1547-1551
mutations [34], possibly indicating a distinct pathogenesis for the underlying lymphoproliferative process. The two IgM specific for lamins reported in this study probably belong to the latter group of monoclonal Ig since they possess a highly restricted specificity. The emergence of anti-lamin antibody may be related either to a preexisting cellular damage or secondary to an abnormal immune response to a processed internal antigen. We would like to thank Dr. N. Chaudhary and for the gift of purified lamin B. We thank Mrs. Yvette Signoret, FranGoise Sanch and Sylvie Bourahla for technical assistance and Muriel Bargis-Touchard for typing the manuscript. We thank Dr. H. Worman for reviewing our manuscript. Received December, 5 , 1991; in final revised form February 26, 1992.
5 References 1 Seligmann, M. and Brouet, J. C., Semin. Hematol. 1973. 10: 163. 2 Kunkel, H. G ., Agnello,V., J o s h , F. G.,Winchester, R. J. and Capra, J. D., J. Exp. Med. 1973. 137: 331. 3 Williams, R. C., Kunkel, H. G. and Capra, J. D., Science 1968. 161: 379. 4 Fermand, J. I?, Danon, F. and Brouet, J. C., Clin. Exp. Immunol. 1985. 59: 467. 5 Zouali, M., Fine, J. M. and Eyquem, A., Eur. J. Irnmunol. 1984. 14: 1085. 6 Braun, I? E . , Frail, D. E. and Latov, N., J. Neurochem. 1982. 39: 1261. 7 Ilyas, A. A., Quarles, R. H., MacIntosh,T. D., Dobersen, M. J. ,Trapp, B. D., Dalakas, M. C. and Brady, R. O., Proc. Natl. Acad. Sci. USA 1984. 81: 1225. 8 Dellagi, K., Brouet, J. C., Perreau, J. and Paulin, D., Proc. Natl. Acad. Sci. USA 1982. 79: 446. 9 Dighiero, G., Guilbert, B., Fermand, J. I?, Lymberi, I?, Danon, F. and Avrameas, S., Blood 1983. 62: 264. 10 McKeon, F. D., Kirschner, M. W. and Caput, D., Nature 1986. 31 9: 463. 11 Fisher, D. Z., Chaudhary, N. and Blobel, G., Proc. Natl. Acad. Sci. USA 1986. 83: 6450. 12 Krohne, G ., Wolin, S. L., McKeon, F. D., Franke, W. W. and Kirschner, M. W., EMBO J. 1987. 6: 3801. 13 Pollard, K. M., Chan, E. K. L., Grant, B. J., Sullivan, K. F., Tan, E. M. and Glass, C. A., Mol. Cell. Biol. 1990. 10: 2164. 14 Guilly, M.-N., Danon, F., Brouet, J. C., Bornens, M. and Courvalin, J. C., Eur. J. Cell Biol. 1987. 43: 266. 15 Lassoued, K., Guilly, M.-N., Danon, F., Andr6, C., Dhumeaux, D., Clauvel, J. I?, Brouet, J. C., Seligmann, M. and Courvalin, J. C., Ann. Intern. Med. 1988. 108: 829. 16 Klotz, C., Bordes, N., LainC, M. C., Sandoz, D. and Bornens, M., J. Cell. Biol. 1986. 103: 613. 17 Aarden, L. A ., De Groot, E. R. and Feltkamp,T. E.W., Ann. N. Y Acad. Sci. 1975. 254: 505. 18 Peltier, A. P., Hughes, G . R. V., Kahn, M. E and Haim, T., Nouv. Presse, Mkd. 1974. 3: 649. 19 Clark, G., Reichlin, M. and Tomasi, Jr., T. B., J. Immunol. 1969. 102: 117. 20 Moore,T. L.,Weiss,T. D., Neucks, S. H., Baldossare, A. R. and Zuckner, J., Semin. Arthritis Rheum. 1981. 10: 309. 21 Blobel, G. and Potter,V. R., Ann. Intern. Med. 1966.108: 829. 22 Dwyer, N. and Blobel, G., J. Cell Biol. 1976. 70: 581. 23 Courvalin, J. C., Lassoued, K., Bartnik, E., Blobel, G. and Wozniak, R . , J. Clin. Invest. 1990. 86: 279. 24 Reeves,W. H., Chaudhary, N. L., Salerno, A. and Blobel, G., J. Exp. Med. 1987. 165: 750.
Eur. J. Immunol. 1992. 22; 1547-1551 25 Kyhse-Andersen, J., J. Biochem. Biophys. Methods 1984. 10: 203. 26 Smith, D. E . and Fisher, P. A., J. Cell Biol. 1984. 99: 20. 27 McKeon, F. D.,Tuffanelli, D. L., Fukuyama, K. and Kirshner, M. W., Proc. Natl. Acad. Sci. USA 1983. 80: 4374. 28 Wesierska-Gadek, J., Penner, E., Hitchman, E. and Sauermann, G., Clin. Immunol. Immunopathol. 1988. 49: 107. 29 Courvalin, J. C., Chaudhary,N. L., Danon,F.,Brouet, J. C. and Lassoued, K . , Biol. Cell. 1990. 69: 93. 30 Reeves, W. H. and Syed Asad, A. P., J. Immunol. 1989. 143: 3614.
Human monoclonal IgM directed against lamins
1551
31 Cartaud, A., Courvalin, J. C.., Ludosky, M. A. and Cartaud, J., J. Cell Riol. 1989. 109: 1745. 32 Cartaud, A ., Ludosky, M. A., Courvalin, J. C. and Cartaud, J., J. Cell Biol. 1990. 111: 581. 33 Chen, P. Pl., Albrandt, K., Orida, N. K., Radoux,V, Chen, E. Y., Schrantz,V., Liu, F. T. and Carson, D. A., Proc. Natl. Acud. Sci. USA 1986. 83: 8318. 34 Mihaesco, E., Ayadi, H ., Congy, N., Gendron, M. C., Roy, J. P., Heyermann, H., Frangione, B. and Brouet, J. C., J. Biol. Chem. 1989. 264: 21481.
