A novel autoantibody reactive with carbonic anhydrase in sera from patients with systemic lupus erythematosus and Sjögren's syndrome

Jourrrul ofDernratological Science, 2 (199 1) 147-154

c

199 1 Elsevier Science Publishers

147

B.V. 0923-l 8 I l/9 l/$03.50

DESC 00075

A novel autoantibody reactive with carbonic anhydrase in sera from patients with systemic lupus erythematosus and Sjiigren’s syndrome Yasunori

Inagaki, Yoko Jinno-Yoshida,

Youichirou

Hamasaki

and Hiroaki

Ueki

Depurtment of Dermatology. Kawasaki Medical School, Kurashiki. Okayama, Japan

(Received 6 September

Key words: Carbonic anhydrase;

Autoantibody:

1990; accepted 20 December

1990)

Sjogren’s syndrome; Systemic lupus erythematosus; Carbonate dehydratase

Renal tubular acidoses:

Abstract

Carbonic anhydrase (CA) is an extremely basic zinc metalloenzyme with a wide phyletic distribution, and the enzyme 1simportant for the regulation ofacid-base status. A novel autoantibody reactive with carbonic anhydrase was demonstrated. Several different classes of CA are known in mammals. Using the immuno blotting method and immun-dot analysis, we found this autoantibody to be reactive with CA in the sera from patients with Sjiigren’s syndrome (20.80,), including a patient with SjBgren”s syndrome and renal tubular acidosis, and in patients with systemic lupus erythematosus (31.63;). The autoantibody varied in the extent of its cross-reactivity among human CA I (or B), human CA II (or C), bovine CA I, bovine CA II. rabbit CA, and dog CA. The titers continued to float and tended to parallel disease activity. Positive reactivity of autoantibody was observed on eccrine sweat glands and the distal tubules of the kidney by the indirect immunofluorescent method.

Introduction Carbonic anhydrase (CA: EC 4,2,1,1) is an extremely basic zinc metalloenzyme with a wide phyletic distribution [ 1,2 1, and the enzyme is imCorrespondence to: Yasunori Inagaki, Department of Dermatology. Kawasaki Medical School, 577 Matsushima. Kurashiki. Okayama, 701-01, Japan, Abbreviations: CA: Carbonic anhydrase; SLE: Systemic lupus erythematosus; h4,: Relative molecular weight; SDS: Sodium dodecyl sulfate; PAGE: Polyacrylamide gel electrophoresis; RTA: Renal tubular acidosis; FITC: Fluorescein isothiocyanate; kDa: Kilo dalton; RA: Rheumatoid arthritis; PBS: Phosphate buffered saline; BSA: Bovine serum albumin.

portant for the regulation of acid-base status in vivo. Several different classes of this enzyme are known, and at least seven enzymic forms (CA I - CA VII) have been designated in amniotes (mammals, birds, reptiles) [ 1.31. However, some of their amino acid sequences show a great deal of homology [ 1,2,4]. These isozymes have also been evolving in a quite conservative style for over 300 million years [ 11. CAs are widely distributed in mammalian tissues [2,4], such as erythrocytes (CA I, II), gastric mucosa (CA I, II), renal cells (CA II, CA IV, CA V), epithelial cells [5] (CA I, II) an d muscle cells (CA III). Using the immunoblotting method and immun-dot analysis, we found an autoantibody reactive with the CA in

148

sera from patients with several autoimmune eases.

dis-

Materials and Methods Carbonic anhydrases as antigens The human CAs used in this study were pur-

chased from two companies. Human CA (HCA) I, II, bovine CA (BCA) I, II, rabbit CA and dog CA were obtained from the Sigma Chemical Company (St. Louis, MO) and HCA I + II were purchased from the Green Cross Corporation (Osaka, Japan). Sera

The sera used in this study were obtained from 62 patients, 38 with SLE and 24 with Sjbgren’s syndrome. Based on counterimmunoelectrophoresis and the immunoblotting procedure [6], the sera from the Sjogren’s syndrome patients contained at least anti SSA/Ro and anti SSB/La antibodies. The sera were not only antinucleic factor (ANF)-positive in a speckled immunofluorescence pattern on KB cells, but also were positive for cytoplasmic antibody. Sera from 20 normal healthy volunteers were also included as controls. The SLE patients satisfied the American Rheumatism Association (ARA) criteria for SLE established in 1982 [7]. Indirect immunofuorescence

test

Skin biopsies were obtained from five healthy volunteers and chilled at - 70 ‘C in a deep freezer. Then the tissue fragments were embedded in O.C.T. Compound and sectioned with a Cryostat microtome (Bright Co., U.K.) to a thickness of 5 to 10 pm at - 30 oC. The sectioned tissues were left standing at room temperature and dried with clean air for 15 min and then were placed in acetone at - 20 “C for 15 min. Sections of human renal tissue without histological change were also used for fluorescence studies. The tissue substrates were incubated with patient sera found to exhibit anti carbonic anhydrase activity by the immunoblotting method or with normal human volunteers’ sera at 25 “C for one hour in a moist

chamber. The reacted antibody (human IgG) was detected by the reaction of FITC-labelled antihuman IgG antibodies (MBL Japan), which were examined under a Fluophot microscope (NIKON, Japan). Specific polyclonal antibodies (rabbit) for HCA (Boeringer Manheim Biochemicals, Green Cross Corporation) were used for positive controls, and the manner in which they were stained was compared. Immunoblotting method

The antigens (CA) were subjected to electrophoresis on 10% SDS polyacrylamide gel slabs [8] in duplicate. Samples were reduced in SDSsample buffer containing 2% SDS and 5 mM /?-mercaptoethanol. One gel slab was stained with Coomassie brilliant blue R-250 and the other slab was used for electroblotting according to Towbin’s procedure [ 6,9]. Electrotransfer from the gel slab to a nitrocellulose membrane was carried out at a constant voltage of 100 V for one hour using a Bio-Rad mini-protean II cell. The blotted nitrocellulose membrane was incubated in 10 mM PBS with 4% BSA, 1% gelatin and 0.05% Tween 20 (Sigma Chemical, St. Louis, MO) to saturate additional protein binding sites of the membrane completely. Diluted (1 : 100) sera were allowed to react with antigens on the membrane at 25 “C for one hour. After extensive washing with TPBS (in 10 mM PBS with 0.05% Tween 20: TPB S), human IgG fixed on the membrane were detected by the avidin-biotinperoxidase complex (ABC) method (Vector Laboratories, Burlingame, CA). Immun-dot analysis

The antigens in PBS were fixed on nitrocellulose membranes by only gravity flow. They were treated neither by boiling in SDS sample buffer nor by any methods involving electrophoresis. Therefore, they were fixed on the membrane in their native form. Blocking procedures, reaction with antibodies, and detection of fixed antibodies were performed in accordance with the immunoblotting method (see Materials and Methods) in a Bio-Dot apparatus (Bio-Rad Lab. Richmond,

method. ( A) Pos itive Fig. 1. R eat :tivity of anti carbonic anhydrase autoantibody detected by the indirect immunofluorescence staining on epidermis. Perinuclear and cytoplasmic staining can be seen. (B) Positive staining on swf:at glamds. (original magnification, x 200).

150

CA). Step dilution of the sera from 1: 8 to 1 : 1024 was used. Immunoabsorbance test An immunoabsorbance test was performed using affinity column chromatography. The immunoabsorbent column was prepared by coupling HCA I or HCA II to cyanogen bromide (CNBr) activated Sepharose 4B (Pharmacia Fine Chemicals, Discatoway, NJ) in the following manner. CNBr sepharose 4B in coupling buffer (0.5 M NaCl, 0.1 M NaHCO, buffer (pH 8.3)) was used in this experiment. The HCA I or HCA II as ligands were allowed to react with the sepharose 4B for 10 h at 4 “C with gentle agitation. The reacted column chromedia was packed and additional protein binding sites were saturated by reaction with 0.2 M glycine (pH 8.0) for 2 h at 25 “C. The IgG fraction prepared by DEAE-cellulose (DE 52, Whatman Paper, U.K.) column chromatography from positively reacted sera by the immunoblotting procedure and immun-dot analysis were applied to the HCA I and HCA II-bound tinity column. Selected affinity purified isolates and unbound fractions were concentrated to 30 mg/ml and examined for specific anti HCA antibody activity by immuno-dot analysis. Results

Presence and localization of CA on normal human skin and kidney Epidermal cells bound to affinity purified anti CA antibody, but dermal connective tissue did not react (Fig. 1A). The autoantibody also bound to hair follicles and the eccrine sweat glands (Fig. 1B). On non-diseased renal tissue, bound to renal tubular cells (not shown). These staining patterns were the same as those of specific antibodies which were raised against HCA. All these findings were diminished by the absorption of the sera with CA fixed column. Anti human CA antibody immunoblot analysis The immunoblotting patterns of the serum containing IgG reactive with HCA I and II are

-

Fig. 2. Immunoblotting pattern. The blotted antigens were human carbonic anhydrase (HCA) I + II for lane 1, HCA I for lane 2, HCA II for lane 3, and KB cell extract for lane 4. Normal human serum reacted on lane 1 for a negative control. Lanes 2 to 4 reacted with SjBgren’s syndrome patient serum.

shown in Fig. 2. Clear bands were noted on a nitrocellulose membrane at the sites of HCA I (lane 2) and HCA II (lane 3 ; arrow). In addition, the serum from a patient with Sjbgren’s syndrome complicated by RA and renal tubular acidosis also contained both SSA/Ro and SSB/La antibodies (No. 2 in Table I). Therefore, they showed reacted bands against KB cell extract (lane 4). Reactivity of anti human CA antibody Immun-dot analysis using sera from normal volunteers, autoimmune disease patients of antiCA positive serum and an aflinity purified IgG fraction was performed. By using this method, rather than SDS-PAGE followed by the electroblotting method, the antigens were not denatured by heating in the SDS sample mixture. The reactivity of the autoantibody against the antigens was not changed by either denaturing by heating in SDS sample buffer or by any non-dena-

151 TABLE I Patient profiles No.

Age

Sex

Clinical diagnosis

Before TX.

After TX. x x x x

I

OM

15

M

SLE

2 3 4 5

AI

18

F

SS, RTA, RA

MS IY KT

35 21 13

F F M

SS, Hashimoto SLE. LN

x x x x

SLE. NS

x 128

256 256 64 128

RTA: renal tubular acidosis; SS: Sjiigren’s syndrome: LN: Lupus nephritis NS: nephrotic thyroiditis.

turing method. Fig. 3 shows representative results for anti human CA on the immun-dot analysis. A strongly reacted dot, No. 2, was made from the affinity purified IgG fraction shown in Fig. 2 (lane 2-4) from a patient with Sjiigren’s syndrome (No. 2 in Table I). No. 8 showed much clearer reactions with serum from a boy with SLE and a

Fig. 3. Immun-dot analysis. HCA was fixed on a nitrocellulose membrane as antigen. Serum dilution was performed from 1: 8 (top) to 1 : 1028 (bottom). The reacted sera were obtained from a normal volunteer (No. 5), a patient with Sjiigren’s syndrome (No. 6), a patient with Sjogren’s Syndrome + RA + renal tubular acidosis (No. 2), a patient with SLE alone (No. 1). and two with SLE with renal involvement (No. 7, No. 8). No. 3 and No. 4 showed reduced reactivity by immunoabsorption.

No. in Fig. 3

Ab titres

1 2 6 I

64 64 32 64

8

x8

syndrome;

Hashimoto:

Hashimoto’s

related nephrotic syndrome (No. 5 in Table I). In both, the titers of the autoantibody were judged to be 1 : 128 or more. The patient profiles and reactivity of anti CA antibody are summarized in Table I. No. 1 and No. 7 revealed positive reactions up to 1 : 64 dilutions. Both of lhem were from patients with SLE in remission, who were being treated with oral corticosteroids. No. 6 from a patient with Sjbgren’s syndrome showed moderated reactivity (1 : 64). No. 5 in Fig. 3 showed an absence of reacted dots in the sera from normal volunteers. Specificity of the anti CA antibod_i The immunoabsorbent test was used to determine the specificity of the autoantibody. The serum (case No. 2, in Table I) was applied to a column in which CA was fixed. The unbound fraction was used to compare the antibody reactivity with the IgG fraction, which was not absorbed by this column. The result using the absorbed positively reacted serum with HCA I can be seen in No. 3 in Fig. 3 and with HCA II in No. 4. The reactivities were markedly diminished in both of them as compared with No. 2. Species specificity of the anti CA antibo&+ Representative results with the immunoblotting method are shown in Fig. 4. Si.x proteins from a molecular weight marker kit (Pharmacia) were used as antigens for the immunoblotting

152 1

2

3

4

5

6

7

8

30Kd

Fig. 4. Cross-reactivity among different species was recognized by the immunoblotting method. The antigens were as follows: molecular weight markers for No. 1, HCA I for No. 2, HCA II for No. 3, BCA I + II for No. 4, BCA I for No. 5, BCA II for No. 6, Rabbit CA for No. 7, and Dog CA for No. 8. (Molecular weight markers: 94 kDa phosphorylase b; 67 kDa bovine serum albumin; 43 kDa ovalbumin; 30 kDa bovine carbonic anhydrase; 20.1 kDa soybean trypsin inhibitor; 14.4 kDa cc-lactalbumin).

method in lane 1, in which one clear reacted band can be seen at the position of bovine CA (M, 30 kDa). In lanes 2 to 8, HCA I, II, BCA I + II, I, II, rabbit CA and dog CA, respectively, were used as antigens. The reacted serum was identical to that of No. 6 in Fig. 3. Each of them revealed distinct reacted bands at the site of antigens on the blotted membrane. Discussion

Carbonic anhydrase occurs in a wide variety of organisms fundamental and necessary to life. To date, at least seven different types of CA isozymes have been found in the human body [4]. Some of these isozymes show a great deal of structural homology. Furthermore, the fact that antibodies raised against one CA may cross-react with other CAs has been reported [lo]. The cross-reaction should occur among different species [ 10,111. By chance, we have found an autoantibody reactive with CAs. Our novel autoantibody might react on the homologous regions among the isozymes in contrast to the specific sites which were recog-

nized by the established monospecific antibodies. The autoantibody appeared in sera from patients with some autoimmune conditions. At the time, we tried to detect other autoantibodies by the immunoblotting method. A blotted nitrocellulose membrane, for which there were not only cellular antigens but also BCA in a molecular weight calibration kit, was reacted with serum from a patient with Sjbgren’s syndrome. Reacted bands for SSA/Ro and SSB/La ]6] were noted on the membrane, but another reacted band was also identified at the position of BCA. As stated above, the autoantibody revealed a wide crossreactivity across not only different types of isozymes in the same species but also, like antinDNA autoantibody, with CAs in different species. This basic enzyme participates in respiration, calcification and regulation of the pH balance through the transportation of H’ and HCO; [2,4,12], which are directly related to maintaining life. Furthermore, they are also found in plants [ 131 and the multi-locus CA isozyme system has become an attractive model for the estimation of molecular evolution over 300 million years [ 11. Some of these isozymes are known to be present on the cell membrane and in extracellular space [4,14], in which the anti CA antibody is able to react with them. This is very important to an understanding of the pathogenic roles of this autoantibody. CAs have appeared widely in the human body [ 15,16,17], and especially ones with a high reactivity in the exocrine organs [ 2,4]. One of the anti CA antibody positive sera we obtained was from a patient with Sjdgren’s syndrome and related renal tubular acidosis (RTA). Sjogren’s syndrome is known as an autoimmune disease which often disturbs the exocrine system [ 181. In addition, many immunological and serological abnormalities, including hyper y-globulinemia, the presence of multiple autoantibodies in sera, and immunoglobulin deposition on the salivary glands, renal tubuli, renal glomeruli, gastric mucosa, and muscles [ 18,191 have been reported. In our case No. 2 (Table l), immunoglobulin deposition was noted on the renal tubuli, renal glome-

153 TABLE II Incidence

of anti CA antibody

Clinical diagnosis

Number of subjects

Positive number

Positive rate

SLE Sjogren’s Syndrome Normal volunteers

38 24 20

12 5 0

31.69, 20.8”, 0”;

C.4: Carbonic Anhydrase;

SLE: Systemic lupus erythematosus.

ruli and sweat glands. A case of RTA with defective CA activity has also been reported [ 201. Some of these conditions might be caused by an anti HCA antibody. In addition, in a case of atrophic gastritis, which is a frequent complication of Sjbgren’s syndrome, gastric parietal cell antibody (PCA) was frequently noted in the sera [31]. No one, however, has examined the relationship between PCA and anti-CA antibody. Another report suggested this PCA inhibits CA activity and other authors concluded that a part of the antibody reacts with gastrin receptors [ 2 11. PCA, however, is thought to be a heterogenous antibody. Some scientists have speculated on the presence of inhibitors of the factors related to control of the acid-base balance in autoimmune disease sera [ 22,231, but, until now, to the best of our knowledge, no one has clearly proven the presence of autoantibodies reactive with CAs themselves. Therefore, this may be the first such report. We believe that, based upon our development and demonstration of anti CA antibody, many questions regarding acid base imbalance and exocrinopathy in autoimmune diseases may be answered. It is relatively easy to detect autoantibodies and they were sometimes found in sera from patients with autoimmune diseases (Table II). Furthermore, titers continued to float and tended to parallel disease activity (Table I). Our new autoantibody was found, based on its cross-reactivity among different species, by means of the immunoblotting method and immun-dot analysis, which are specific and highly sensitive techniques. Since erythrocytes are rich in CAs, hemolysis should be avoided when taking

blood samples. CAs released from the erythrocytes will react with the autoantibody in the serum and this reaction will reduce the reactivity of the antibody at detection. We speculate that this autoantibody will be found, like anti-nDNA antibody in the field of life sciences, in other autoimmune sera besides those of our reported diseases. Therefore, this is an important and significant autoantibody., and these findings may shed light on the pathogenesis of autoimmune diseases. Acknowledgements The authors wish to express special appreciation to Professor Keiichi Hosokawa of the Department of Biochemistry of our school for many helpful suggestions, Miss Akiko Ono for her kind support in preparing the manuscript, and Miss Junko Kazahaya for her skilled technical assistance. We also thank Professor Gengo Osawa and Dr. Toru Shindo of the Department of Internal Medicine of Kawasaki Medical School for providing us with serum from a patient with renal tubular acidosis and with renal specimens. This work was supported in part by Kawasaki Medical School Grant (2-50.5) for Project Research.

References Tashian RE, Hewett-Emmett D, Goodman M: On the evolution of carbonic anhydrases I, II and III, in Isozymes: current topics in biological and medical research. volume 7: molecular structure and regulation. Liss, Inc., New York, 1983. pp. 79-100.

154 2 Deutsch HF: Carbonic anhydrases. Int J Biochem 19: 101-l 13, 1987. 3 Shapiro LH, Venta PJ, Yu Y-HL, Tashian RE: Carbonic anhydrase II is induced in HL-60 cells by 1,25_dihydroxyvitamin D, : a model for osteoclast gene regulation. FEBS Lett 249: 307-310, 1989. 4 Fernley RT: Non-cytoplasmic carbonic anhydrase. TIBS 13: 356-359. 1988. 5 Braun-Falco 0, Rathjens B: Uber die histochemische Darstellung der Kohlenslureanhydratase in normaler Haut. Arch Klin Exper Dermat 201: 73-82, 1955. 6 Inagaki Y, Jinno Y, Hamasaki Y, Ueki H: Higher incidences ofanti-SS-A/Ro and anti-SS-B/La autoantibodies in Japanese patients with autoimmune disorders-studies of antigens and antibodies using the immunoblotting method. Arch Dermatol Res 281: 89-94, 1989. 7 Tan EM, Cohen AS, Fries JF, Masi AT, McShane DJ, Rothtield NJ, Schller JG, Talal N, Winchester RJ: The 1982 resived criteria of the classification of systemic lupus erythematosus. Arthritis Rheum 25: 1271-1277, 1982. 8 Laemmli UK: Cleavage of structual proteins during the assembly of the head of bacteriophage T4. Nature 227: 680-685, 1970. 9 Towbin H, Staehelin T, Gordon J: Electrophoretic transfer of proteins from polyarcrylamide gels to nitrocellulose sheets: procedure and some applications. Proc Nat1 Acad Sci USA 76: 4350-4354, 1979. 10 Erickson RP, Kay G, Hewett-Emmett D, Tashian RE, Claflin JL: Cross-reactions among carbonic anhydrase I, II, and III studied by binding tests and with monoclonal antibodies. Biochem Genet 20: 809-819, 1982. 11 Linser OJ, Perkins MS, Fitch FW, Moscona AA: Comparative characterization of monoclonal antibodies to carbonic anhydrase. Biochem Biophys Res Commun 125: 690-697, 1984. 12 Wistrand PJ, Knuuttila K-G: Renal membrane-bound carbonic anhydrase. Purification and properties. Kidney Int 35: 851-859, 1989. 13 Graham D, Reed ML, Patterson BD, Hockley DG: Chemical properties, distribution, and physiology of plant and algal carbonic anhydrases. Ann NY Acad Sci 429: 222-237. 1984.

14 Fernley RT, Darling P, Aldred P, Wright RD, Coghlan JP: Tissue and species distribution of the secreted carbonic anhydrase isoenzyme. Biochem J 259: 91-96,1989. 15 Spicer SS, Sens MA, Tashian RE: Immunocytochemical demonstration of carbonic anhydrase in human epithelial cells. J Histochem Cytochem 30: 864-873, 1982. 16 Carter MJ: Carbonic anhydrase: isoenzymes, properties, distribution, and functional significance. Biol Rev 47: 465-513, 1972. 17 Carter MJ, Parsons DS: The isoenzymes of carbonic anhydrase: tissue, subcellular, distribution and functional significance, with particular reference to the intestinal tract. J Physiol 215: 71-94, 1971. 18 Talal N, Zisman E, Schur PH: Renal tubular acidosis, glomerulonephritis and immunologic factors in Sjiigren’s syndrome. Arthritis Rheum 11: 774-786, 1968. 19 Pasternack A, Linder E: Renal tubular acidosis: an immunopathological study on four patients. Clin Exp Immuno1 7: 115-123, 1970. 20 Donckerwolcke RA, Van Stekelenburg GJ, Tiddens HA: A case of bicarbonate-losing renal tubular acidosis with defective carboanhydrase activity. Arch Dis Child 45: 769-773, 1970. 21 Masala C, Smurra G, Prima MAD, Amendolea MA, Celestino D, Salsano F: Gastric parietal cell antibodies: demonstration by immunofluorescence of their reactivity with the surface of the gastric parietal cells. Clin Exp Immunol 41: 271-280, 1980. 22 Loveridge N, Bitensky L, Chayen J, Hausamen TU, Fisher JM, Taylor KB, Gardner JD, Bottazzo GF, Doniach D: Inhibition of parietal cell function by human gammaglobulin containing gastric parietal cell antibodies. Clin Exp Immunol 41: 264-270, 1980. 23 Yagita M, Hayashi I, Hasui N, Kurinaga A, Taira A, Kanazawa K, Nakahira H, Takuma M, Tamura K, Miyoshi K: Inhibition of carbonic anhydrase activity of human kidney by sera of a patient with distal renal tubular acidosis. Shikoku Igakkai Shi (in Japanese) 38: 26-31, 1982.