Sera From Chronic Chagasic Patients With Complex Cardiac Arrhythmias Depress Electrogenesis and Conduction in Isolated Rabbit Hearts
Background Immune dysfunction has long been proposed as a mechanism for the etiopathogenesis of the chronic phase of Chagas’ disease. Antibodies of chagasic patients have been shown to interfere with electric and mechanical activity of embryonic myocardial cells in culture. Here, we demonstrate that antibodies derived from a group of chronic chagasic patients are able to induce disturbances in the electrogenesis and conduction in isolated adult rabbit hearts.
Methods and Results Sera from chronic chagasic patients with complex cardiac arrhythmias (ChA+) decreased heart rate (from 131±26 to 98±37 bpm [mean±SD]; n=6; P<.05) in isolated rabbit hearts when perfused at a dilution of 1:100 (vol:vol) by the Langendorff method. Sera from another experimental group of four chronic chagasic patients without complex arrhythmias (ChA−) and two control groups composed of five Wolff-Parkinson-White (WPW) syndrome patients and five orthopedic surgery patients did not affect heart rate when tested under similar conditions. In addition, sera from five of six ChA+ patients and from one WPW patient induced AV conduction blockade. Effects of the sera from ChA+ patients are due to their IgG fractions. Both serum and IgG effects are blocked by atropine (10 μmol/L).
Conclusions Antibodies of ChA+ patients decrease heart rate and induce AV conduction block in isolated adult rabbit hearts through activation of muscarinic receptors.
Chagas’ disease is caused by the protozoan parasite Trypanosoma cruzi and presents both acute and chronic phases. These phases are separated by an indeterminate period that can last from months to decades, during which patients are relatively asymptomatic. The disease affects primarily the heart and gastrointestinal tract. The acute phase is characterized by active infection and is accompanied by inflammation and myocardial damage.1 2 In the cardiac form of the disease, myocarditis occurs and may lead to heart failure.3 4 The presence of parasites in the circulating blood and within several tissues, including myocardium, is characteristic of the acute phase and usually results in the sequential activation of T and B cells, which produce antibodies to the circulating forms of the parasite.5 6 7 Myocardial dysfunctions resulting from chronic Chagas’ disease include arrhythmias, ventricular ectopy, conduction defects, cardiomegaly, congestive heart failure, and sudden death.8 9
Although subject to extensive investigation, the cause of cardiac disturbances present in the chronic phase has not yet been clearly established. However, the contribution of autoimmune processes in chronic chagasic cardiomyopathy was hypothesized even in early studies of this disease because of the virtual absence of parasites either in myocardium or in circulating blood. In support of this hypothesis, Sterin-Borda and collaborators10 11 12 13 have demonstrated the presence of IgG components in the sera from some chagasic patients that interact with cardiac β-adrenergic and muscarinic membrane receptors.
Thus, we decided to investigate the involvement of antibodies from chronic chagasic patients with defined cardiac dysfunctions14 15 on the electrogenesis and conduction in the isolated rabbit hearts. For these studies, we compared the effects of sera from two groups of chronic chagasic patients with cardiomyopathy, ChA+ and ChA−, and two control groups, WPW patients and patients undergoing orthopedic surgery (NBD). Consistent with the autoimmune hypothesis, both sera and IgG from ChA+ patients slowed spontaneous firing rate and disturbed AV conduction in isolated rabbit hearts, whereas sera from other patients showed no such effects.
ECG Recordings in Isolated Rabbit Hearts
The method for ECG recording from isolated hearts has been previously described in detail.16 In brief, young rabbits weighing 1.5 to 2.0 kg were killed by cervical dislocation, and their hearts were rapidly removed and cannulated through the aorta for continuous perfusion of the coronary circulation with Tyrode’s solution (in mmol/L: NaCl 127, KCl 2.7, NaHCO3 12, MgCl2 0.5, glucose 10, and CaCl2 2.7, pH 7.2) at 36±0.2°C. The hearts were oriented vertically, with the atria up, and immersed in warmed Tyrode’s solution contained in a water-jacketed glass flask. Three glass electrodes filled with 1 mol/L NaCl were positioned inside the flask to obtain optimal ECG recordings. Two of the electrodes were connected to the differential input of a high-gain amplifier (3A9, TEKTRONIX Inc), and the third was connected to ground. In each isolated heart, serum, IgG+, or IgG− fractions from individual patients were tested separately. The experimental protocol consisted of control recordings for 30 minutes in Tyrode’s solution, at least a 30-minute perfusion with Tyrode’s solution containing total human serum (diluted 1:100) or IgG+ (0.02 to 0.06 mg protein/mL) or IgG− (0.2 to 0.6 mg protein/mL) fractions, and then a 30-minute washout with Tyrode’s perfusion. Effects of IgG− and IgG+ fractions obtained from the same patient were tested in the same heart preparation, with a 30-minute washout between application of the two fractions. The ECG was continuously monitored on an oscilloscope (561 TEKTRONIX, Inc) and recorded on a chart recorder (2200, Gould Inc); expanded records were taken every 5 minutes. Experiments were carried out only if no significant changes in the ECG parameters were observed for the 30-minute duration of the control recordings. The ECG analysis included P-wave frequency and the presence of AV conduction block and other arrhythmias. P-wave frequency and AV conduction disturbance under different experimental maneuvers were evaluated at the moment of maximal alteration, which ranged from 5 to 30 minutes during IgG+/serum perfusion.
Patient Follow-up and Selection Procedures
Sera were obtained from 10 cardiopathic patients, all of whom were serum-positive for Chagas’ disease, as tested by hemagglutination, cruzipain-ELISA, and indirect immunofluorescence. Patients were followed in the Cardiomyopathy Research Clinic every 2 months. At each visit, they underwent clinical and laboratory evaluations, including physical examinations, 12-lead ECGs, M-mode and two-dimensional echocardiographies, 24-hour dynamic ECGs, and when necessary, exercise stress testing, besides a complete biochemical evaluation, including thyroid function tests. Patients with concomitant arterial hypertension, chronic obstructive pulmonary disease, cardiomyopathy of any other origin than Chagas’ disease, valvular heart disease, congenital cardiomyopathy, obstructive coronary disease, thyroid dysfunction, excessive alcohol consumption, known immunologic dysfunctions, or systemic disease were excluded from the study.
All patients were in stable clinical condition for at least 3 months. Some were being treated with drugs (eg, digitalis, diuretic, and/or vasodilator agents) to prevent peripheral edema. During the follow-up period, body weight was constant (within ±1 kg), urinary flow was >1.5 L/d, and plasma electrolyte composition was normal. Drug administration was interrupted 48 hours before laboratory and clinical measurements. No clinical problems related to the above procedures were observed throughout this study. World Health Organization and Helsinki Treaty regulations (1963), reviewed in Venice (1983), were followed, and the study was approved by the institutional Committee on Research Using Human Subjects.
Chagasic patients were divided into two experimental groups, depending on whether they displayed complex cardiac arrhythmias.17 18 Table 1⇓ summarizes the ECG and echocardiographic findings of the patients. Under echocardiographic examination, left ventricular dilation was considered present when left ventricular end-diastolic diameter index was ≥15% above the upper normal range (32 mm/m2). The ChA+ group included patients 1 through 6. Clinically, three patients complained of palpitation; the other three had syncopal attacks. One patient had a pacemaker implanted and another died during the follow-up period (the diagnosed terminal event was systemic thromboembolism). The ChA− group included patients 7 through 10. Two patients complained of atypical chest pain, but none had syncopal attacks. No patients had left ventricular dilation, but two demonstrated hypokinesis in the apical and inferomedial segments. The heart rate was 62±13 bpm (mean±SD) in the ChA+ group and 68±11 bpm in the ChA− group as evaluated by a 24-hour dynamic ECG. Both groups displayed sinus rhythm, and heart rates in both chagasic groups were significantly lower than in control groups.
The control groups were constituted by sex- and age-matched patients, all negative for Chagas serology, from two groups; five were WPW syndrome patients, diagnosed by electrophysiological techniques after complaints of palpitation. Their rhythm was sinusal with intermittent periods of preexcitation and supraventricular tachyarrhythmias. The other five, the NBD group, were orthopedic surgery patients with normal cardiovascular function (evaluated by physical examination, ECG, and chest x-ray). The mean heart rate was 102±11 bpm in the WPW group and 92±13 bpm in the NBD group as calculated from data obtained from medical records. Each patient’s heart rate represents a mean value obtained from six measurements taken every 4 hours during 24 hours. All conventional blood biochemical parameters (urea, glucose, creatinine, and electrolytes) in these patients were normal, and they were not subjected to echocardiography.
Table 2⇓ gives information about the clinical conditions of all experimental groups.
Serum fractionation was performed in two steps. The first step consisted of antibody precipitation with ammonium sulfate solution, followed by overnight dialysis against phosphate buffer (pH 8.0). In the second step, the dialyzed solution was fed into a DEAE-matrix column, and 1.5-mL fractions were collected during continuous addition of phosphate buffer, pH 8.0, to the column. IgG was usually present from fractions 4 through 15 as detected by spectrophotometry at 280 nm. After all IgG had been collected, the remaining proteins were eluted from the column by 500 mmol/L NaCl solution, thus giving the IgG depleted (IgG−) fraction. Both IgG+ and IgG− fractions were dialyzed overnight against phosphate buffer. After dialysis, protein concentration in the fractions was determined by the Bradford19 method.
The adequacy of the method used to fractionate the serum was evaluated by immunodiffusion of IgG+ and IgG− fractions against anti-total human immunoglobin (Fig 1A⇓) and anti-human IgG (Fig 1B⇓) antibodies (Cappel). The IgG+ fraction produced one clear reaction line both with anti-human Ig antibody (Fig 1A⇓, three wells at right) and with anti-human IgG antibody (Fig 1B⇓, three wells at right). The multiple reaction lines on the left side of Fig 1A⇓ indicate the presence of several immunoglobulins in the IgG− fractions of the same patients, whereas the undetectable or faint reaction lines on the left side of Fig 1B⇓ show the virtual absence of IgG in the IgG− fractions of the three patients tested.
Statistical comparison of the effects of sera from different experimental groups on sinus rate and of heart rate of these four groups of patients was by use of ANOVA complemented by the Student-Newman-Keuls test. ANOVA for repeated measures was used to compare the effects of IgG+ and IgG− fractions on sinus rate. The effectiveness of sera (or IgG+ and IgG− fractions) from different experimental groups in inducing AV conduction block was compared by use of Fisher’s exact test. In all cases, differences were considered significant at a value of P<.05.
Sera From ChA+ Patients Reduced Sinus Firing Rate and Blocked AV Conduction
Sera from five of six ChA+ patients induced sinus bradycardia in normal isolated rabbit hearts. Fig 2A⇓ shows the results of one such experiment: sinus rate decreased from 113 bpm (control, record 1) to 80 bpm after 23 minutes of serum perfusion (record 2) and recovered to 107 bpm after 10 minutes of washout with Tyrode’s solution (record 3). Pooled results of experiments performed in different rabbit hearts with sera from the six patients of this same group (Table 3⇓) indicate a significant decrease in sinus rate during serum perfusion. Both the absolute decrease of the sinus rate and the percentage change in rate were significantly larger in rabbit hearts treated with sera from the ChA+ patients than with sera from the other three groups (ChA−, WPW, and NBD). Sinus firing rate was not significantly affected by sera from patients in these three groups.
Sera from each of the five ChA+ patients that induced bradycardia in perfused rabbit hearts were also found to block AV conduction. Serum from the ChA+ patient that did not cause bradycardia also did not lead to AV conduction block at the dilution used. The intensity of the AV conduction block varied for sera of different patients. Two patients’ sera caused first-degree block, two other sera caused second-degree block, and one serum induced complete AV block. A second-degree AV conduction block can be seen in Fig 2A⇑ (the second tracing) after 23 minutes of perfusion with serum from patient 3 (Table 1⇑) diluted 1:100 in Tyrode’s solution. As shown in this figure and distinct from the depressing effect on heart rate in which recovery was usually only partial, AV conduction block promptly reverted upon washout with Tyrode’s solution in all experiments.
Sera From ChA− Patients Did Not Block AV Conduction
Among the sera from ChA− patients, three of four decreased heart rate in isolated perfused rabbit hearts. In the example shown in Fig 2B⇑, sinus rate decreased from 105 bpm under control conditions (record 1) to 95 bpm after 30 minutes of serum perfusion (record 2), returning to 105 bpm after 25 minutes of washout (record 3). When pooled data from all ChA− patients’ sera were analyzed, however, changes in heart rate were found not to be significant (Table 3⇑). AV conduction block was never observed in hearts treated with ChA− sera, and the appearance of the ECG was unaltered except for serum from one patient that caused ventricular premature beats and few episodes of ventricular tachycardia.
Sera from the WPW and NBD groups had no significant effect either on the sinus rate or on AV conduction, except for one serum sample obtained from a WPW patient that induced AV conduction block (Table 3⇑).
IgG Fraction Is Responsible for the Observed Effects of Sera From ChA+ Patients on Cardiac Electrogenesis
IgG+ and IgG− fractions were obtained from sera of only five of the six original ChA+ patients because one of the patients whose serum depressed both heart rate and AV conduction did not return to the clinic. The fractions were assayed for bradycardia and conduction defects in isolated rabbit hearts. Fig 3⇓ illustrates results of one such experiment in which perfusion with IgG+ fraction of one effective serum induced second-degree AV block within 5 minutes. Importantly, the IgG− fraction obtained from the same serum did not induce conduction block or bradycardia even after 30 minutes of perfusion with protein concentrations 10-fold higher than that used in the IgG+ fraction. In fact, one of the IgG− fractions tested increased heart rate by 27%, whereas two others increased it by 9%. Table 4⇓ shows that pooled data obtained from all five IgG+ fractions decreased heart rate. It should be noted that the IgG+ fraction isolated from the serum of the ChA+ patient that did not induce AV block or bradycardia also failed to produce an effect, as did IgG− fractions from all five ChA+ patients.
Muscarinic Mediation of Effects of the ChA+ Sera
Because of previous evidence that IgG of chronic chagasic patients binds to membrane receptors,10 11 13 we tested the effect of atropine, a muscarinic antagonist, on the IgG+ response. Addition of atropine (10 μmol/L) alone to the perfusion solution did not significantly change heart rate (152±19 versus 154±8 bpm under control condition, n=4). However, atropine completely antagonized the bradycardia produced by IgG+ fractions of serum of ChA+ patients. As indicated earlier, perfusion of hearts with ChA+ IgG (0.02 to 0.06 mg/mL) decreased heart rate and led to conduction block. By contrast, addition of 10 μmol/L atropine with the ChA+ IgG did not change the heart rate (158±9 bpm, n=4). Additionally, when IgG+ fractions of four ChA+ patients previously shown to induce AV block in isolated hearts were perfused in the presence of atropine, IgG+ fractions from three patients no longer induced the AV block (not shown). The IgG+ fraction from one patient retained its ability to induce AV block in the presence of atropine, even though the bradycardial effect of this same fraction was blocked by the muscarinic antagonist.
We report in this article that sera obtained from a specific group of chronic chagasic cardiopathic patients, ChA+, contain antibodies that reduce cardiac rate and induce AV conduction block in Langendorff-perfused rabbit hearts. Both effects appear to be mediated through interactions of antibodies with muscarinic receptors because they are blocked by atropine. By contrast, sera from ChA− patients did not exert a significant effect on the electrogenesis of isolated rabbit hearts. Similar lack of effect was obtained with sera of two nonchagasic control groups: WPW patients and NBD subjects.
That antibodies of chagasic patients interact with membrane receptors in mammalian cardiomyocytes has been shown in studies on beat rate and/or force of contraction in isolated rat atria by Sterin-Borda and collaborators, first for β-adrenergic10 12 20 and more recently for muscarinic11 13 receptors. In preliminary experiments, we found that sera (diluted 1:100 in Tyrode’s solution) and IgG from rabbits chronically infected with T. cruzi induced sinus bradycardia in isolated rabbit hearts from noninfected animals.21 Serum from one of these rabbits was also found to induce third-degree AV block. The present study provides the functional demonstration that antibodies from human chagasic patients with cardiac symptoms can affect cardiac electrogenesis and conduction when perfused in the rabbit heart.
Although ChA+ and ChA− groups both include chagasic patients in the chronic phase of the disease, only sera and IgG+ fractions from the ChA+ group were effective in altering electrogenesis and conduction in our experimental model. In that regard, the differences and similarities among the patients in these two groups, given in the “Methods” section, may be extremely informative. The ChA− samples include patients of Los Andes groups Ib and II; ie, some had normal ECGs and little myocardial damage (group Ib), and others had altered ECGs and more pronounced myocardial damage (group II). However, none of the patients included in this group had symptoms of heart failure, and all of them were below class IV in the modified Lown and Wolf22 scale; ie, none exhibited complex cardiac arrhythmias. ChA+ samples, in contrast, included two patients from Los Andes class III with congestive heart failure, whereas the other four were class II patients, as in the ChA− group, without congestive heart failure. Thus, although some ChA+ patients had end-stage myocardial damage, others had much less advanced myocardial alterations. Nevertheless, all six patients were class IV or above in the Lown and Wolf22 classification; ie, all patients in the ChA+ group presented complex cardiac arrhythmias in conventional ECG, during 24-hour dynamic ECG monitoring, or during exercise stress testing. Therefore, the main difference between the two groups of chagasic patients appears to be the presence of complex arrhythmias in the ChA+ group, although more pronounced myocardial damage was also generally present in this group.
The muscarinic-like effect of IgG from ChA+ patients may result from molecular mimicry between T. cruzi antigens and functional epitopes in the mammalian M2 receptor. Thus, ChA+ patients may contain an IgG directed at a “T. cruzi-antigen” that activates M2 receptors in the rabbit heart, triggering a cascade of reactions with corresponding functional consequences. One clear difference between the activation of muscarinic receptors by its natural ligand, acetylcholine, and by the antibodies present in these patients’ sera is the absence of desensitization of the receptor by sera. In patients with idiopathic DCM, a disease in which autoimmunity has been recognized as one of the pathogenic mechanisms,23 Magnusson et al24 reported the absence of desensitization in the interaction of serum antibodies with β1-adrenergic receptors. In that case, adrenergic agonists and antibodies were shown to bind to distinct sites in the receptor; such an interaction between the ChA+ antibodies and the muscarinic receptor would also explain our findings. Antibodies in sera of a significant proportion of DCM patients recognize epitopes present in the extracellular loops of both the human β1-adrenergic receptor25 and the human muscarinic acetylcholine receptor M2.26 Thus, in a growing list of pathogenic disease states, autoantibodies that functionally mimic the action of neurotransmitters on their receptors appear to be generated.
We did not check any of the sera used in the present study for the presence of β-adrenergic effect. For the ChA+ group, however, we have indirect evidence that no such effect is present in the IgG+ fraction because this fraction did not increase heart rate when perfused in atropinized hearts.
Recently, molecular mimicry between a specific T. cruzi antigen and a membrane protein in humans was reported by Ferrari et al.27 In that study, IgG from chronic chagasic patients were shown to recognize both the C-terminal part of the ribosomal P0 protein of T. cruzi and the second extracellular loop of the human β1-adrenergic receptor. This latter interaction was shown to be functional in that chagasic serum increased the beat rate of cultured cardiomyocytes from neonatal rats.
It would be interesting to investigate to what extent IgG of ChA+ and DCM patients are related. As mentioned above, myocardial damage in most ChA+ patients was not much more extensive than in most ChA− patients, whose serum did not have clear muscarinic-like action. Thus, we cannot directly correlate myocardial damage with the appearance of antibodies with muscarinic-like effects. Nevertheless, antibodies of ChA+ (this article) and DCM patients26 might interact with different epitopes in the muscarinic receptor. Such a situation was described for β1 receptors by Ferrari et al27 ; the positive chronotropic effect produced by IgG from DCM was not blocked by a peptide derived from the P0 ribosomal protein of T. cruzi that efficiently blocks the same effect produced by sera from some chronic chagasic patients. Because the chronotropic actions of both antibodies were blocked by conventional β1-blockers, the authors concluded that the antibodies acted on β-adrenergic receptors, although at distinct sites.
Selected Abbreviations and Acronyms
|ChA−||=||chronic chagasic patients without complex cardiac arrhythmias|
|ChA+||=||chronic chagasic patients with complex cardiac arrhythmias|
|NBD||=||normal blood donor|
Financial support was provided by CEPG-UFRJ, CNPq, FINEP, FAPERJ, and FUJB. Drs Campos de Carvalho and Masuda are career investigators for CNPq, and author Farias de Oliveira is a recipient of a fellowship for technical assistance from FAPERJ. We wish to acknowledge the valuable contributions of Cristina Borges e Sá and Paulo R. da Costa, from Hospital Universitário Clementino Fraga Filho, for collecting blood from the patients and of Dr Doris Rosenthal for help with the statistical analysis.
- Received September 30, 1996.
- Revision received November 22, 1996.
- Accepted December 16, 1996.
- Copyright © 1997 by American Heart Association
Andrade ZA. Mechanisms of myocardial damage on T. cruzi infection. In: Ciba Foundation Symposium 99: Cytopathology of Parasitic Diseases. London, UK: Pitmann Books; 1983:214-233.
Dias JCP. Acute Chagas disease. Mem Inst Oswaldo Cruz. 1984;79(suppl):85-92.
Krettli AU, Brener Z. Protective effects of specific antibodies in Trypanosoma cruzi infections. J Immunol. 1976;116:755-760.
Laranja FS, Dias E, Nóbrega G, Miranda A. Chagas disease: a clinical, epidemiologic, and pathologic study. Circulation. 1965;14:1035-1060.
Lopes ER, Chapadeiro E, Almeida HO, Rocha A. Contribuição ao estudo da anatomia patológica dos corações de chagásicos falecidos súbitamente. Rev Soc Bras Med Trop. 1975;6:269-282.
Sterin-Borda L, Borda ES. Interaction of the autonomic nervous system with lymphocytes and antibodies in the pathogenesis of Chagas’ disease. In: Chagas’ Disease and the Nervous System. Washington, DC: Pan American Health Organization; 1994:295-231.
Carrasco HA. Diagnostico de dano miocardico en la enfermidad de Chagas. In: Textos de la Universidad de Los Andes. Merida, Venezuela: Consejo de Publicaciones de la Universidad de Los Andes; 1983.
Lown B, Wolf M. Approaches to sudden death from coronary heart disease. Circulation. 1971;44:130-142.
Goodwin JF. Mechanisms in cardiomyopathies. J Mol Cell Cardiol. 1985;17:5-7.
Magnusson Y, Wallukat G, Waagstein F, Hjalmarson A, Hoebeke J. Autoimmunity in idiopathic dilated cardiomyopathy: characterization of antibodies against the β1-adrenoceptor with positive chronotropic effect. Circulation. 1994;89:2760-2767.
Magnusson Y, Marullo S, Hoyer S, Waagstein F, Andersson B, Vahlne A, Guillet JG, Strosber AD, Hjalmarson A, Hoebeke J. Mapping of a functional autoimmune epitope on the β1 adrenergic receptor in patients with idiopathic dilated cardiomyopathy. J Clin Invest. 1990;86:1658-1663.
Fu LX, Magnusson Y, Bergh C-H, Liljeqvist JA, Waagstein F, Hjalmarson A, Hoebeke J. Localization of a functional autoimmune epitope on the muscarinic acetylcholine receptor-2 in patients with idiopathic dilated cardiomyopathy. J Clin Invest. 1993;91:1964-1968.
Ferrari I, Levin MJ, Wallukat G, Elies R, Lebesgue D, Chiale P, Elizari M, Rosenbaum M, Hoebeke J. Molecular mimicry between the immunodominant ribosomal protein PO of Trypanosoma cruzi and a functional epitope on the human β1-adrenergic receptor. J Exp Med. 1995;182:59-65.