Antibodies to cardiac Purkinje cells: Further characterization in autoimmune diseases and atrioventricular heart block

CLINICAL

IMMUNOLOGY

AND

IMMUNOPATHOLOGY

42, 14 I- 150 ( 1987)

Antibodies to Cardiac Purkinje Cells: Further Characterization in Autoimmune Diseases and Atrioventricular Heart Block M. OBBIASSI, A. BRUCATO, P. L. MERONI, A. VISMARA, M. LETTINO, F. POLONI, A. FINZI,* M. G. FENINI,? AND L. ROSSIS Istituto

di Clinica Medica II and jlstituto di Anatomia Patologica, Cardiologia, Ospedale Policlinico, Milan, and fServizio Spedali Civili, Brescia, Italy

Universitd di Milan, *SenGzio di Immunologia Clinica.

di

We confirmed the occurrence of IgG antibodies reacting with ox cardiac conducting tissue in the serum of some human subjects. These antibodies failed to react with all ox cardiac conducting tissue cells; they reacted only with the cells defined as Purkinje cells. Having checked 352 sera, we found that the prevalence of antibodies to Purkinje cells was 11% in normal subjects (no correlation with sex and age), 14% in systemic lupus erythematosus, 21% in rheumatoid arthritis, 18% in progressive systemic sclerosis, and 23% in Sjogren syndrome. In 50 patients with permanent pacemakers for chronic nonpostinfarction atrioventricular (AV) block the prevalence was 30% (P = 0.008). In a selected set of 29 patients with clinically idiopathic AV block located at or below the level of the His bundle the prevalence was 34.5% (P = 0.006). The possible role of anti-Purkinje cell antibodies in autoimmune damage of cardiac conduction tissue is discussed.

0 1987 Academic

Press, Inc.

INTRODUCTION

In 1975 Fairfax and Leatham remarked that some autoimmune disorders (vitiligo, pernicious anemia, thyroid diseases, and diabetes mellitus) were more prevalent in a set of 100 patients with chronic heart block than in the general population (1). These results suggested that heart block could be produced by some autoimmune process. To consider this possibility, Fairfax and Doniach found an assay to show possible antibodies specifically reacting with cardiac conducting tissue (2). They tested human sera by the indirect immunofluorescence technique, using as substrate the “false tendon” of ox heart that is rich with typical Purkinje cells. Actually some sera contained antibodies specifically reacting with the substrate; these were called cardiac conducting tissue antibodies (CCTAs). CCTAs were IgG-fixing complement. They were found in the serum of 8 (8.6%) of 93 patients with chronic atrioventricular (AV) blocks and in I (4.5%) of 22 normal control sera; the difference between the two groups was not statistically significant. Using the same assay, in 1983 Villecco et al. (3) remarked that CCTAs were correlated with the right bundle branch block in several patients with rheumatoid arthritis. Recently Lotze et al. (4) found CCTAs in 5.6% of their patients with sick sinus syndrome and in 12.7% of the healthy controls. This study was undertaken to further characterize the CCTAs and their possible clinical significance. We study various sets of patients, including subsets 141 0090-1229187

$1.50

Copyright 0 1987 by Academic Press, Inc. All rights of reproduction in any form reserved

142

OBBIASSI

ET AL.

with autoimmune diseases and patients with AV block, who have undergone electrophysiological studies of the cardiac conduction system. MATERIALS

AND METHODS

Methods CCTA detection. Sera from patients and controls were tested by the standard immunofluorescence technique (IFL) (5) on 5-pm-thick unfixed frozen sections of ox heart false tendon (three animals). The ox heart false tendon was obtained fresh from the slaughterhouse and snap-frozen in isopentan precooled with acetone-solid CO2 mixture within 2 hr as described by Fairfax and Doniach (2). The tissue was kept at - 70°C until cryostat sections were cut. Some cryostat sections from every block of frozen tissue were stained with hematoxylin and eosin, or with toluidine blue, to be certain of the presence of cardiac conducting tissue. It is well known that in the ox heart the false tendon contains the right bundle branch and that in the ungulates Purkinje cells are easily distinguishable from ordinary working myocardial cells on light microscopy. Sera were diluted l/10 with phosphate-buffered saline (PBS), pH 7.2, and FITC-labeled goat anti-human Ig (Cappel, Cochranville, Pa.) was used to assess immunoglobulin binding, as previously described (6, 7). Antibody titers were assessed by doubling dilutions of sera in PBS. Antibody churacterization. The immunoglobulin class was determined by using FITC-labeled goat anti-human IgG, IgA, and IgM antisera (Cappel) (6, 7). To distinguish CCTAs from other circulating antibodies, CCTA-positive sera were also tested on cryostat sections of rat tissues (liver, kidney, heart, and diaphragm) and of human tissues (stomach, thyroid, and pancreas) (6, 7). Sera from patients with immunologic disorders known for containing antinuclear antibodies (ANAs). anti-smooth muscle antibodies (SMAs) type G according to the classification of Bottazzo et (11. (8), anti-mitochondrial antibodies (AMAs) type M2 (9), anti-striated muscle antibodies from polymiositis patients, or anti-reticulin antibodies type Rs, according to Rizzetto and Doniach (lo), all were tested on ox false tendon for comparison of staining pattern. Antibodies to extractable nuclear antigens (ENAs) were investigated by counterimmunoelectrophoresis, according to Bernstein et al. (11); rabbit thymus extract (Pel Freez Biological, Inc., Rogers, Ark.) and human spleen extract, prepared as described by Venables et al. (12), were used as the source of antigens. Identification of the conducting tissue cells reacting with CCTAs. CCTA-positive sera also were tested by IFL on cryostat sections of ox heart including the conducting tissue at the level of the sinoatrial node, atrioventricular node, His bundle, and proximal and distal portions of the right and left bundle branches. One of us (L.R.) controlled adjacent sections (hematoxylinieosin or toluidine blue stained) on light microscopy every time. Subjects Normal controls. One-hundred healthy people were examined as a representative sample for sex and age of Milan citizens according to the general census of

ANTIBODIES

TO CARDIAC

PURKINJE

143

CELLS

1982. At the moment of our study mean age was 38.8 years (range l-84); 47 subjects were males (mean age 36.2 years) and 53 females (mean age 41.7 years). All the subjects recently had undergone physical examinations and standard laboratory tests (including ESR, blood cell count, blood glucose, blood urea nitrogen, urinary sediment). Patients with immunologic disorders. Sera from 193 patients with systemic immunologic disorders were tested. Many sera had been sent from external laboratories. Therefore we could detain a complete clinical documentation (ECG included) for only a portion of them. Table 1 shows the diseases considered and the number of patients for each one. Patients treated by permanent pacemaker implantation. We examined 59 patients with permanent artificial pacemakers followed by the Milan-Policlinico Cardiology Unit staff. Each patient had been submitted to invasive electrophysiologic testing to determine the block site in the cardiac conducting system. Patients could be divided into two groups. The first one included the patients with a permanent AV block arising immediately after an acute myocardial infarction. They were classified as patients with an “ischemic heart block.” The duration of pacemaker implantation was 4.5 ? 3.8 years (mean + SE). The second group included the remaining 50 patients, 31 males and 19 females, mean age 66.6 (range 2 l-84), which had permanent pacemakers for a mean period of 5.1 years. Invasive electrophysiologic testing had shown a conduction block at the sinoatrial or AV node level in 11 subjects, at the His bundle or a lower level in 32, and at both levels above and below the His bundle in 2. The block site could not be determined in 5 subjects. An accurate analysis of the clinical data ruled out any reference to a valvular, TABLE 1 RESULTSFOREACHGROUPOFSUBJECTSAFTER 352 SERAWERETESTEDFORCCTA Group Normal controls SLEb RA’ PSSd Sjogren S .e (with mixed connective tissue disease associated) Subjects with pacemakers (not postinfarction block) (Patients with idiopathic block at or below His bundle level) Patients with pacemakers for AV postinfarction block

No. of subjects

No. of positivities

Percentage

P”

0.79 0.12 0.30 0.30

100 42 67 67 17

II 6 14 12 4

11.0 14.3 20.9 17.9 23.5

(3)

(1)

(33.3)

50

15

30.0

0.008

(IO)

(34.5)

0.006

0

0.64

9

0

LIP vs normal controls using chi-square test with Yates’ correction. b SLE: systemic lupus erythematosus. r RA: rheumatoid arthritis. d PSS: progressive systemic sclerosis. * Sjogren S.: Sjogren syndrome.

144

OBBIASSI

ET AL.

hypertensive, or ischemic heart disease, as well as to a peripheral vasculopathy, in 29 of the 32 patients with heart conduction block located at or below the level of the His bundle. These patients were classified as affected by a “clinically idiopathic heart block”: they had a duration of pacemaker implantation of 5.5 t 1.8 years (mean 2 SE). Statistical

Analysis

Chi-square test with Yates’ correction was used to compare proportions. for unpaired sample was used to compare means.

t test

RESULTS

CCTA Characterization Pattern. Positive sera (62 of 352) reacted with ox cardiac conducting cells showing a bright staining, while the connective tissue and the ordinary working myocardial cells did not stain (Fig. 1). Cellular specificity of CCTAs. Sera from 30 subjects (taken from all the groups) were tested on cryostat sections cut at different levels of the ox conducting tissue. The 17 sera which showed to be CCTA positive reacted only with a certain kind of cell. A thorough histological control (performed by L.R.) of the adjacent sections demonstrated that CCTAs had reacted only with Purkinje-like cells (Figs. 2A, B), which were numerous in the His bundle and its branches whereas they were rare and scattered at the level of sinoatrial and AV node. Moreover no

FIG. 1. IFL whereas

on ox cardiac conduction the myocardial cells do not stain.

tissue. The x 160.

Purkinje

cells

show

a definite

fluorescence,

ANTIBODIES

TO CARDIAC PLJRKINJE CELLS

145

FIG. 2A. A discrete fascicle of clear voluminous Purkinje cells in the His bundle x 130: cryosl ta section; hematoxylin/eosin stain. FIG. 2B. IFL on an adjacent section of the Purkinje cells fascicle in Fig. 2A. X 260.

I46

OBBIASSI

ET AL

serum stained any other kind of the cells which are common in cardiac conducting tissue: P cells, transitional cells, and atria1 or ventricular working cells (13). Antibody titer. The 62 CCTA-positive sera showed a titer from 1:lO to I:80 inclusive. The titer was 1: 10 in 25 cases, 1:20 in 19 cases, 1:40 in 13 cases, and 1:80 in 5 cases. The average titer of CCTA-positive cases was similar in all the groups (about 1:24). Antibody class. The CCTAs showed to be IgG in every case. Testing other substrates. A panel of rat kidney, liver, heart, and diaphragm, and of human thyroid, stomach, and pancreas, was used to test by IFL every group of patients for the presence of organ- and non-organ-specific autoantibodies. The positivity for the different autoantibodies was comparable in CCTApositive and CCTA-negative subjects (see Table 2). This implies that the CCTAs were not associated with organ and with non-organ-specific autoantibodies. In addition, sera positive for CCTAs from patients with AV block were also investigated for the presence of antibodies to extractable nuclear antigens, using rabbit thymus extract and human spleen extract; all the sera were negative. IFL pattern on Purkinje cells of some non-organ-specific autoantibodies. The ANAs reacted with respective antigens both in Purkinje cells and in working myocardial cells. The SMAs and anti-reticulin antibodies reacted neither with Purkinje nor with working myocardial cells. On the contrary, the anti-striated muscle antibodies from subjects with polymiositis, as well as the anti-mitochondrial antibodies from patients with primary biliary cirrhosis, reacted with both kinds of cellular substrates. The anti-striated muscle antibodies gave the characteristic striational immunofluorescence in the Purkinje and myocardial cells (Fig. 3A). The AMAs type M2, moreover, imparted to Purkinje cells a more granular appearance than did CCTAs, with especially bright areas at the periphery of the cytoplasm, and also stained intensely the cytoplasm of working myocardial cells (Fig. 3B). TABLE 2 PREVALENCEOFCOMMONAUTOANTIBODIESINCCTA-POSITIVEANDCCTA-NEGATIVE EVERYGROUPOFPATIENTS SLE

ANAs" AMAb SMA’ str. muscled PCAs' Anti-thyroidf

RA

Sjiigren S.

PSS

CCTA+ (%)

CCTA!%)

CCTA+ (%)

CCTAP (%)

CCTA+ (%)

CCTA(%)

83 17 0 0 0 0

86 II 0 0 3 0

89 0 0 0 0 0

71 0 0 0 0 0

75 0 25 0 0 0

93 0 7 2 0 0

a ANAs: anti-nuclear antibodies. b AMAs: anti-mitochondrial antibodies. c SMAs: smooth muscle antibodies. d Str. muscle: anti-striated muscle antibodies. c PCAs: parietal cell antibodies. f Anti-thyroid: anti-thyroid antibodies.

SERA FROM

CCTA+ (o/o) 7s 0 25 0 0 0

CCTA(%I 46 0 II 0 8 8

Subjects with pacemakers CCTA+ (%) I 0 I 0 0 0

CCTAfro) II 0 6 0 0 0

ANTIBODIES

TO CARDIAC

PURKINJE

CELLS

147

FIG. 3A. Ox conducting tissue stained with a serum containing anti-striated muscle antibodies. The myofibrils of the Purkinje cells show the charactertistic striational immunofluorescence. x 260. FIG. 3B. Section containing working myocardial cells (right) and Purkinje cells (left). Anti-mitochondrial antibodies stain both types of cells. In the Purkinje cells a brighter fluorescence is present at the periphery, corresponding to the location of mitochondria. Anti-nuclear fluorescence is also present. X 260.

148

OBBIASSI

Prevalence

of Anti-Purkinje

ET AL.

Cell Antibodies

in Our Series of Patients

The prevalence in normal controls was 11% (see Table 1). There was no correlation with either sex or age. The prevalence in the patients with various immunologic disorders is shown in Table 2. It was slightly higher than prevalence in the normal controls, but any difference was not statistically significant. Some of the patients with immunologic diseases, whose ECG records were known, had some kind of cardiac AV disorder(s). No association between conduction disorders and presence of CCTAs was found (Table 3). Of the 50 patients with permanent pacemakers for sinoatrial or AV chronic block, 15 (30%) were CCTA positive (P = 0.008 vs normal controls; see Table 1). The statistical significance still remained if only control subjects older than 50 years were considered. In this way 34 normal controls. with a mean age of 63.2 years, were compared to 50 patients with a mean age of 66.6 years. The prevalences of CCTA positivity were 8.8 and 30%, respectively (P = 0.04). To focus attention on the possible pathogenetic role of CCTAs, the patients with pacemakers were selected in accordance with clinical data and with the anatomical localization of cardiac conduction block. Of the 50 patients with pacemakers, 10 suffered from a definitely diagnosed heart disease whereas 40 exhibited a clinically “idiopathic” cardiac conduction block. In 7 of these, electrophysiologic studies demonstrated that the block was located at the sinoatrial or AV node level, and in 4 cases at an indeterminate level, whereas only 29 patients had a cardiac conduction block which could be defined as clinically idiopathic and located at or below the His bundle level, i.e., at a level where the conduction tissue is made mostly by Purkinje cells. Of these 29 patients, 10 (34.5%) had CCTAs in their sera (P = 0.006 vs control group). Finally the sera of 9 patients with AV block from acute myocardial infarction all were CCTA negative. Moreover the relative time that the pacemakers were in place was comparable in post-myocardial infarction patients and in idiopathic heart block patients (4.5 -+ 3.8 years vs 5.5 -+ 1.8 years, mean ? SE: P = 0.41). DISCUSSION

Our data, in agreement with those of Fairfax and Doniach (2) and Villecco et al. (3), confirm the peculiarity of these antibodies, in spite of the fact that it still is CCTA

TABLE 3 POSITIV~IES~NPATIENTSWITHIMMUNOLOGICDISORDERSANDKNOWN Number of patients with known ECG records

Group SLE RA PSS SjGgren Note.

For

Number of patients with some AV conduction disorders

28 31 25 5

S. meanings

of abbreviations

ECGRECORDS

I 3 3 2 see Table

1

Number of CCTA-positive patients 0 I 0 I

ANTIBODIES

TO CARDIAC

PURKINJE

CELLS

149

unknown with which antigens they react. We could, however, observe that they fail to react with all ox cardiac conducting tissue cells and react only with morphologically typical Purkinje cells. Therefore we think it more correct to refer to these antibodies as cardiac Purkinje cell antibodies (CPCAs), and not generalizing the term to a broad and cytologically varied set of specialized myocardial cells (13). In our normal group a prevalence of 11% of CPCAs was observed, whereas the prevalence in Fairfax and Doniach’s controls was 4.5% (2), in Villecco et a/.‘~ 3.3% (3), and in Lotze et al.‘s 12.7% (4). So, in our experience, CPCAs are relatively common, and they are associated neither with the age nor with an overt disease. Having checked the sera of 193 patients with various immunologic disorders (SLE, PSS, RA, Sjogren S.), we found that in every disease there was a prevalence of CPCA-positive subjects higher than that found in normal controls, but in no case was the difference statistically significant. We can assume that the higher prevalence of CPCAs in systemic autoimmune diseases can be related to the polyclonal B-cell activation, which is present in these diseases. Our results in patients with rheumatoid arthritis should be pointed out. Only 3 of the 3 1 cases with electrocardiographic documentation showed cardiac conduction disorders. This percentage, like that of CPCA positivity observed by us in 14 (20.9%) of the 69 patients, is sharply different from that of Villecco et al. (3). Such a discrepancy could be due to a different case selection. Fairfax and Doniach (2, 14) found a prevalence of 8.6% of CPCA positivity in the sera of 93 unselected patients with chronic heart block. This prevalence was higher than the prevalence found in healthy controls, but the difference was not statistically significant. In our study the patients with chronic heart block were selected on the basis of two criteria. The first criterion is that Purkinje cells are relatively rare and dispersed in the upper part of cardiac conducting system (sinoatria1 and AV node), while they are a constant cellular component of the His bundle and its branches. The second consideration is that a remarkable percentage of chronic cardiac blocks escapes a definite etiologic assessment, not only clinically but also pathologically (13, 15, 16). It is also well known that the majority of these heart blocks are intra- or infrahissian (13). Of 59 of our patients with heart block, we selected 29 whose block could be defined as clinically idiopathic and located at or below the level of the His bundle. Ten of them (34.5%) presented CPCAs in their serum. In turn, CPCAs were not found in 3 groups of patients: (I) patients with intra- or infrahissian AV block associated with an overt cardiac disease (valvular, hypertensive, and/or ischemit), (2) patients with clinically idiopathic suprahissian block and (3) patients with AV block from myocardial infarction. These data altogether suggest that CPCAs do not seem to be a mere epiphenomenon of Purkinje cell damage. but that CPCAs can be correlated with AV blocks idiopathic in nature. In addition it should be noted that CPCAs were not influenced by the duration of the pacemaker implantation. Recently. anti-Ro(SS-A) antibodies were held responsible for congenital AV blocks present in malformation free hearts in children born from mothers positive

150

OBBIASSI

ET AL.

for these seric autoantibodies (17-22). However, we found that these antibodies are not related to the CPCAs. Finally, it is very difficult to ascertain whether and to what extent the CPCAs could play a real pathogenetic role in ~ivo. It should be underlined that the low titers of these antibodies raise doubts about their critical pathogenetic role. In any case, further studies are now in progress in our laboratory to investigate the reactivity of the CPCAs with human heart substrates. Indeed, we must stress that we used as substrate classic Purkinje cardiac cells of the ox heart, whereas the specialized elements commonly called Purkinje cells in the human heart are far from typical in most cases, as the Purkinje eponym has been extended to them more for functional than for morphological reasons ( 13, 23, 24). ACKNOWLEDGMENTS We thank Miss Adriana Cercone for her competent technical assistance. This work was supported in part by grants from Minister0 della Pubblica Istruzione.

REFERENCES 1. 2. 3. 4.

Fairfax, A. J., and Leatham A., Brit. Med. J. 4, 322. Fairfax, A. J., and Doniach D.. Clin. Ex~, Imm~lnol. Villecco, A. S., De Liberali, E.. Bianchi. E B.. and Lotze, U., Maisch. B.. Schneider, J., and Kochsiek,

1975.

23, 1, 1976. Pisi. E.. Clin. Exp. Immunol. 53, 536, 1983. K.. Elr~. Heart J. 5 (Abstr. Suppl. I), 255,

1984.

5. Roitt, I. M., and Doniach. D., “WHO Manual of Autoimmune Serology.” WHO, Geneva, 1969. 6. Meroni, P. L.. De Bartolo, G., Barcellini, W., Riboldi. P. S.. Basile, R., Betterle. C., and Zanussi. C., J. Clin.

Immunol.

4, 45. 1984.

7. Caredda, F., De Bartolo. G., Zampini, L., Barcellini. W.. Zavaglia, C., Carella. G., and Meroni, P. L., Digestion 33, 181, 1986. 8. Bottazzo, G., Florin-Christensen, A., Fairfax. A., Swana. G.. Doniach, D., and GroeschelStewart, V., J. Clin. Pat/to/. 29, 403, 1976. 9. Meek, F., Khoury, E. L.. Doniach, D.. and Baum. H.. C/in. Exp. Immunol. 41, 43, 1980. IO. Rizzetto, M., and Doniach, D., J. C/in. P&ho/. 26, 841, 1973. 11. Bernstein, R. M., Bunn. C. C.. and Hughes, G. R. V., Ann. Rheum. Dis. 41, 554, 1982. 12. Venables, P. G. W., Smith, P. R., and Maini, R. N., C/in. Exp. Zmmunol. 54, 731, 1983. 13. Rossi, L.. “Histopathology of Cardiac Arrhythmias,” Lea & Febiger, Philadelphia, 1979. 14. Fairfax, A. J., Proc. R. Sot. Med. 70, 327. 1977. 15. Lenegre, J., Ma/. Cnrdiolwsc. 3, 311, 1962. 16. Rossi. L., Cardiol. C/in. 1, 13, 1983. 17. Carter, J. B., Bhden, L. C., and Edwards, J. E., Arch. Pathol. 97, 51. 1974. 18. Esscher. E., and Scott, J. S., Brit. Med. J. 1, 1235. 1979. 19. Litsey, S. E.. Noonan, J. A., O’Connor, W. N.. Cottril, C. M., and Mitchell, B., Nena Engl. J. Med. 312, 98. 1985. 20. McCue, C. M., Mantakas, M. E.. Tingelstad. J. B., and Ruddy, S., Circulation 56, 82, 1977. 21. Scott, J. S., Maddison, P. J., Taylor, P. V., Esscher, E.. Scott. 0.. and Skinner, R. P., New Eng/. J. Med. 309, 209, 1983. 22. Vetter, V. L.. and Rashkind, W. L., Neul Engl. J. Med. 309, 236, 1983. 23. Chiodi, V., and Bortolami R., “The Conducting System of the Vertebrate Heart,” p. 328. Ed. Calderini, Bologna, 1967. 24. Rossi, L.. PACE 7, 1101, 1984. Received January 23, 1986; accepted with revision August II, 1986