Author + information
- Received November 5, 2001
- Revision received April 25, 2002
- Accepted June 19, 2002
- Published online September 18, 2002.
- ↵*Reprint requests and correspondence:
Akiyasu Baba, MD, Department of Medicine, Kitasato Institute Hospital, 5-9-1 Shirokane Minato-ku, Tokyo 108-8642, Japan.
Objectives We sought to test the hypothesis that autoantibodies (Abs) produced against sarcolemmal Na-K-ATPase play a role in the development of ventricular tachycardia (VT) and cardiac sudden death in patients with dilated cardiomyopathy (DCM).
Background Autoimmunity is one of the mechanisms of pathogenesis of DCM as well as virus infection and genetic predisposition.
Methods One hundred patients with DCM and age-matched control subjects (CTL) were screened for Abs produced against Na-K-ATPase by using enzyme-linked immunosorbent assay.
Results Abs were detected in 26 DCM and 2 CTL patients. Na-K-ATPase activity in the presence of patient IgG was lower in DCM with Abs than without Abs, but there was no difference between two groups in CTL. Western blots showed that Abs recognized the alpha subunit of Na-K-ATPase, and 3H-ouabain bindings in the presence of patient IgG showed that dissociation constant was higher in DCM with Abs than without Abs. No difference existed between subjects with regard to age, gender, New York Heart Association functional class, cardiac function, or neurohormone levels, except for plasma norepinephrine, which was higher in patients with Abs than without Abs. VTs were more common in patients with Abs than without Abs, and multiple logistic regression analysis demonstrated that the presence of Abs, but not plasma norepinephrine, was an independent predictor for the occurrence of VT. Cardiac sudden death was independently predicted by the presence of Abs, as well as poor systolic function.
Conclusions We conclude that there are Abs produced against sarcolemmal Na-K-ATPase in patients with DCM and that Abs could be responsible for the electrical instability in some cases.
Numerous prognostic indicators, including cardiac function and neurohormonal activation, are reported to be useful in predicting the long-term prognosis of patients with congestive heart failure (CHF) (1–3). Although these parameters are helpful in predicting heart failure death, there are few parameters that predict sudden cardiac death caused by ventricular arrhythmias. Several predisposing factors for the development of ventricular arrhythmias in patients with CHF have been proposed, including autonomic nervous system imbalance, myocardial ischemia, focal fibrosis, stretch-activation of ion channels and electrolytes disturbance (4). Sarcolemmal Na-K-ATPase activity has been reported to be decreased in patients with idiopathic dilated cardiomyopathy (DCM) (5). Preliminary data from our laboratory have shown that erythrocyte Na-K-ATPase activity was decreased and inversely correlated with plasma norepinephrine in CHF patients with coronary artery diseases. However, in patients with DCM, other circulating substances, except for plasma norepinephrine or digitalis, also were suggested as decreasing Na-K-ATPase activity, which was related to the occurrence of ventricular tachycardias (6).
Autoimmunity, viral infection and genetic predisposition have been proposed as causes of DCM (7). The hypothesis that autoimmunity may cause cardiomyopathy is supported by the improvement in subjective symptoms and hemodynamics in patients after the removal of autoantibodies (Abs) by immunoadsorption (8). The use of high-dose immunoglobulin therapy in patients with myocarditis and acute cardiomyopathy also supports the role of autoimmunity in the pathogenesis of cardiomyopathy (9).
In the present study, we tested the hypothesis that Abs produced against Na-K-ATPase cause a decrease in enzymatic activity, increasing the risk of ventricular arrhythmias and sudden death in patients with DCM who have Abs produced against Na-K-ATPase.
We studied 100 patients diagnosed with DCM based on the World Health Organization/International Society and Federation of Cardiology Task Force definition and classification of cardiomyopathies report of 1995 (10). At the time of sample acquisition, all patients were stable under conventional therapy, including digitalis, diuretics, angiotensin-converting enzyme inhibitors, or beta-blockers. Eighty-seven patients underwent cardiac catheterization, which revealed normal coronary arteries in all of the patients. Endomyocardial biopsy was performed in 78 patients and revealed no inflammatory cell infiltrates or granulomas. Healthy control subjects (n = 100) were matched for gender and age. The study protocol was approved by an institutional review committee, and all patients gave written, informed consent to participate in the present study.
Antigen and substrate
As an antigen used in enzyme-linked immunosorbent assay (ELISA) and as a substrate used in enzyme activity measurement, porcine cerebral cortex Na-K-ATPase was obtained from Sigma Chemical Co (St. Louis, Missouri). It was dissolved in phosphate-buffered saline (PBS: 10 mM phosphate, 140 mM NaCl, pH 7.4) to 2 mg/ml and stocked in −80°C until the ELISA was performed. For the enzyme activity measurement, porcine cerebral cortex Na-K-ATPase was dissolved in 50 mM histidine buffer (pH 7.2) containing 100 mM NaCl, 10 mM KCl, 3 mM MgCl2 and 1 mM EGTA in the same concentration. As an antigen used in Western blots and as a substrate used in radioligand binding assay, human myocardium was obtained at the time of autopsy, all of which exhibited underlying heart diseases. Cardiac membrane fractions were prepared as 50,000 g pellet by the previous methods (11). The final membrane pellet was resuspended in 50 mM Tris-HCl buffer (pH 7.5) containing 250 mM sucrose and 1 mM EGTA and stored at −80°C.
Enzyme-linked immunosorbent assay
Ninety-six-well ELISA plates (Iwaki Glass, Japan) was coated with 50 μl of stocked Na-K-ATPase solution plus 0.5% gelatin. Blocking was performed by PBS supplemented with 3% bovine serum albumin and 0.1% Tween-20. Fifty microliters of each patient’s serum, diluted from 1:20 to 1:160, were added to the microtiter plates and incubated overnight at 4°C. After washing, a peroxidase-conjugated goat anti-human IgG (H + L) antibody (1:1000 dilution, Biosource, Camarillo, California) was added and allowed to react for 1 h. After washing, the bound peroxidase-conjugated antibody was detected by incubating the plates for 1 h with orthophenylenediamine dihydrochloride (Organon Teknika, West Chester, Pennsylvania). Sulfuric acid (2.5 M) was added to stop the reaction, and the optical density was read at 492 nm in a microtiter reader with a positive finding defined as an optical density that was 2.5 times the background density (12).
Functional enzyme activity measurement with purified IgG
The IgG purification from patients’ serum was performed by protein A-agarose methods using the Affi-Gel Protein-A MAPS II Kit (Bio-Rad Laboratories, Hercules, California). One microgram of purified IgG was added to 50 μl of the stocked Na-K-ATPase solution with 3 mM ATP. For each sample, triplicate reaction tubes with or without 2 mM ouabain was processed. The reactions were conducted for 3 h at 37°C. Inorganic phosphate was determined by the colorimetric method (6). Na-K-ATPase activities in the presence of patient IgG were compared between patients with positive and negative antibodies for both DCM and control subjects.
Fifty micrograms of human cardiac membrane was subjected to electrophoresis on a 10% polyacrylamide gel in SDS and transferred to nitrocellulose membrane. Immunopurified IgG from each positive and negative patients with DCM was used as primary antibodies and were incubated overnight at 4°C. After the strips were washed by TBST (10 mM Tris, 150 mM NaCl, 0.1% Tween 20, pH 8), they were incubated with 1:1000 dilution of rabbit anti-human IgG for 1 h at room temperature. The specificity of the bands was checked by blocking effect of porcine cerebral Na-K-ATPase after preincubation with antibody overnight at 4°C. To further clarify the specific bands, polyclonal antibody produced against Na-K-ATPase catalytic subunit (LSL) were used in another experiment.
Radioligand binding assay
The ouabain-binding assay was performed by the modified method described by Kim et al. (13). Eighty micrograms of human cardiac membrane was incubated in duplicate with six concentration (5 to 100 nmol/l) of [3H]ouabain (15 Ci/mmol; New England Nuclear) in the presence of either unlabeled ouabain (final concentration 1 mmol/l) or vehicle at 37°C for 1 h. Each reaction tube was added to 1 μg of patient IgG. To separate bound and unbound [3H]ouabain, the sample was filtered through Whatman GF/C filters of 96-well UniFilter plates by MicroMate cell harvester (Packard Japan Co.) followed by washing with the distilled water. The filters were measured for 3H radioactivity by Microplate Scintillation Counter (TopCount, Packard Japan Co.). Specific binding was defined as the difference between binding of [3H]ouabain in the absence and presence of 1 mmol/l unlabeled ouabain. 3H-ouabain binding sites were compared between positive and negative IgG from patients with DCM.
Clinical features of autoantibody (AB)-positive patients
We compared age, gender, New York Heart Association functional class (NYHA), left ventricular ejection fraction (LVEF) and other clinical parameters between patients with or without the Na-K-ATPase Abs. Clinical variables, including findings from ultrasonic echocardiography and Holter monitoring, were recorded for all patients within one month of serum sampling for measurement of Abs. Echocardiography was performed using a commercially available probe and system to assess left ventricular end-diastolic dimensions, end-systolic dimensions and fractional shortening. The Holter electrocardiogram (ECG) was analyzed on a dual-channel by using the Marquette system (SXP 8000 or MARS 8000) to assess the total number of premature ventricular contractions (PVCs) and occurrence of ventricular tachycardia (VT). Premature ventricular contractions were defined as any beat of ventricular origin faster than the sinus rate and VT as three or more consecutive PVCs with a heart rate > 100 beats/min, which was noted in at least one of the tests. Data were averaged from two or more sets of recordings. Blood samples for measurement of serum potassium and digitalis concentration, plasma norepinephrine, renin and atrial natriuretic peptide concentrations were collected from a peripheral vein after the patient had rested in the supine position for at least 30 min. Blood samples were placed in ice and spun in a refrigerated centrifuge to separate the plasma then stored at −80°C until analysis by adequate techniques.
The patients studied were followed for a mean of 31 months (range 0.2 to 52.5 months). Deaths were classified as follows: sudden death was defined as death within 1 h without previous worsening of symptoms of heart failure. Unexpected death occurring during sleep was considered to be sudden when patients were found dead in the morning by family members sharing the same room. When death occurred as a consequence of progressive deterioration of heart failure, acute pulmonary edema or cardiogenic shock, we classified it as a result of pump failure.
Data are expressed as the mean ± SD. Optical densities of ELISA and enzyme activities and radioligand binding assay of Na-K-ATPase were evaluated using two-way repeated-measures analysis of variance (ANOVA). For variables with p < 0.05 regarding the effect of interaction, we have analyzed the data among groups by one-way ANOVA with Bonferroni post hoc test for multiple comparisons. In clinical features of patients, comparison between two groups was performed by using the nonpaired two-tailed t test or the nonparametric means test for continuous variables and by the chi-square test for discrete variables. Logistic regression analysis was used to assess predictors of VT. Ten variables, including those with p values <0.10 by univariate analyses, were further assessed by multivariate analysis. Kaplan-Meier survival curve for cardiac sudden death was calculated according to presence or absence of Abs, and the differences were analyzed by the log-rank test. Long-term prognosis was assessed using the Cox proportional hazards model. Statistical significance was defined as p < 0.05. StatView 5.0 software (SAS Institute Inc.) was used for statistical analyses.
Enzyme-linked immunosorbent assay detected Abs produced against the Na-K-ATPase in 26 patients with DCM and 2 control subjects in our study group. A substantial reduction in optical densities was produced by dilution in control subjects who were judged positive (data not shown) but not in patients with DCM who were similarly positive for Ab even in 1:160 dilution (Fig. 1).
Figure 2 shows activity of Na-K-ATPase from porcine cerebral cortex in the presence of patient and control IgG fractions. Na-K-ATPase activity was lower in the presence of positive sera determined by ELISA than the presence of negative sera in patients with DCM (48 ± 16 vs. 79 ± 19 μM Pi/h/mg protein). There was no difference in the activity between the two groups for control subjects (102 ± 22 vs. 99 ± 21 μM Pi/h/mg protein).
In Western blots analyses (Fig. 3), no band was detected in the presence of negative IgG determined by ELISA from patients with DCM. A single distinct band of approximately 110 kd was evident in the presence of positive IgG determined by ELISA from patients with DCM. This band was absorbed by preincubation with porcine Na-K-ATPase. The molecular size was identical with that of the catalytic alpha subunit of the enzyme.
In ouabain binding experiments, there was no difference in the maximal binding sites between the presence of positive and negative sera determined by ELISA (12.7 ± 1.8 vs. 15.3 ± 1.8 pmol/mg protein). However, the dissociation constant was higher in the presence of positive than negative sera determined by ELISA (21.5 ± 0.8 vs. 10.7 ± 0.7 nmol/l, p < 0.01).
No differences existed in age, gender, NYHA functional class, plasma potassium concentration, concomitant medications, including digitalis, and cardiac function between patients with or without Abs. Furthermore, digitalis concentration, plasma renin activity and atrial natriuretic peptide concentration were similar between the two groups (Table 1). However, the PNE concentration was significantly higher in Ab-positive than in Ab-negative patients (579 ± 254 vs. 370 ± 285 pg/ml, p < 0.001). The frequency of PVC on Holter monitoring was greater in Ab-positive than in Ab-negative patients (3,505 ± 4,482/day vs. 1,669 ± 4,601/day, p < 0.001; Fig. 4). In addition, multifocal PVC (100 vs. 68%, p < 0.01) and nonsustained VT (81% vs. 32%, p < 0.001) were more common in Ab-positive than in Ab-negative patients. Both the frequency of VT (52 ± 202/day vs. 45 ± 62/day, p < 0.05) and the maximum length of runs of VT (11 ± 8 vs. 7 ± 5, p < 0.03) were greater in Ab-positive than in Ab-negative patients. Univariate analyses were performed for 22 variables, including the PNE concentration, as well as for the variables listed in Table 1, with the result that the presence of Abs and PNE concentration was related to the presence of VT. Multivariate logistic regression analyses were performed for the 10 variables listed in Table 2, which showed the presence of Abs, but not PNE concentration, to be the only independent predictor of VT.
After a mean follow-up period of 31 ± 15 months, 80 patients were alive. The 20 deaths included 11 sudden deaths, 9 deaths resulting from pump failure, with no deaths from unknown causes. Ten cardiac deaths occurred in Ab-negative patients (four sudden deaths and six pump failure deaths), and 10 cardiac deaths occurred in Ab-positive patients (seven sudden deaths and three pump failure deaths). Figure 5 compares the Kaplan-Meier survival curves for cardiac sudden death according to the presence or absence of Abs. Sudden deaths were more common among autoantibody-positive as compared with Ab-negative patients. A Cox proportional hazards model for cardiac sudden death was used to assess the significance of Abs and other clinical variables in patients with DCM. First, univariate analyses were performed for 20 variables, including the presence of VT and each concomitant medication, as well as for the variables listed in Table 2. Eight variables, including those with p < 0.10 by univariate, were further assessed by multivariate analysis (Table 3). The presence of Abs and LVEF <21% were independent predictors for cardiac sudden death.
Abs in idiopathic DCM
Several functional Abs have been described in patients with DCM, including antibodies against myosin (14), ADP/ATP transporter (15) and G-protein–coupled receptors (12,16–18). We reported previously the frequency of antimyocardial Abs in patients with DCM by either the method of standard indirect immunofluorescence and immunoblotting (19). There were some patients who had unknown Abs that were related to the occurrence of arrhythmias. Particularly, the presence of a 110-kd autoantibody was significantly associated with the risk of VT by multiple logistic regression analysis. Our preliminary data by immunoscreening of human heart cDNA library indicated that the Abs directed against 110-kd antigens reacted with sarcolemmal Na-K-ATPase pump.
Na-K-ATPase in CHF
A number of studies have shown that ouabain binding sites were decreased in the myocardium of patients with CHF (5,20). The decrease in the alpha-3 isoform expression of Na-K-ATPase was reported in the setting of pacing-induced heart failure, with similar alterations induced by the chronic administration of norepinephrine (13). We also reported that Na-K-ATPase activity was decreased in the erythrocyte membranes of patients with CHF (6). A significant inverse correlation existed between PNE and erythrocyte Na-K-ATPase activity, even in the absence of digitalis glycosides. However, in some patients with DCM, the Na-K-ATPase activity also was decreased without an increase of PNE. These findings suggested that other circulating substances might influence the Na-K-ATPase activity in the pathophysiologic states of DCM.
In the present study, 26% of patients with DCM and 2% of control subjects had Abs produced against Na-K-ATPase by the ELISA method, which have not been described previously. External constant Na-K-ATPase activities in vitro were lower in the presence of purified IgG from DCM patients with Abs than without Abs. Therefore, this Ab (IgG) was biologically active. In fact, Western blots using human myocardium demonstrated that the Abs specifically target at the alpha subunit of the enzyme. Furthermore, radioligand-binding assay indicated surmountable antagonistic effect of Abs. These results suggested that the autoantibodies bound the catalytic subunit of Na-K-ATPase, resulting in a conformational change from a high-affinity to low-affinity state. Preliminary data from our laboratory show that human cardiac membranes immunoprecipitated by the Abs were directed against the alpha3-, but not alpha1-, isoform of the Na-K-ATPase.
Clinical features of patients with Abs
In the present study, the presence of Abs produced against Na-K-ATPase was associated with more frequent PVCs and nonsustained VT, which appear to be mediated by non–re-entrant mechanism in patients with nonischemic heart failure (21). Although the electrophysiologic function of the Abs is not known, it is speculated that abnormal intracellular Ca2+ handling by reduced Na-K-ATPase activity may cause delayed afterdepolarizations via reverse-mode operation of the Na+/Ca2+ exchanger resulting from increased intracellular Na+ concentrations (22). Preliminary data using chick embryonic ECGs (23) indicated that various ventricular arrhythmias ranging from PVCs to VT, which were terminated by verapamil, could be induced by affinity-purified IgG from patients with the autoantibodies.
Functional Abs produced against Na-K-ATPase may exert an arrhythmogenic effect via an autoimmune response in certain patients with DCM. Characterization of this Ab could be helpful in identifying patients with electrical instability whom potentially susceptible to sudden cardiac death.
☆ Supported by Grant-in-Aid for Encouragement of Young Scientists (No.12770376, 2000) from Japan Society for the Promotion of Science and by funds from the Idiopathic Cardiomyopathy Research Group of the Ministry of Health and Welfare of Japan.
- congestive heart failure
- dilated cardiomyopathy
- enzyme-linked immunosorbent assay
- left ventricular ejection fraction
- non-sustained ventricular tachycardia
- New York Heart Association
- plasma norepinephrine
- premature ventricular complex
- ventricular tachycardia
- Received November 5, 2001.
- Revision received April 25, 2002.
- Accepted June 19, 2002.
- American College of Cardiology Foundation
- Swedberg K.,
- Eneroth P.,
- Kjekshus J.,
- et al.
- Wei C.M.,
- Lerman A.,
- Rodeheffer R.J.,
- et al.
- Muller J.,
- Wallukat G.,
- Dandel M.,
- et al.
- McNamara D.M.,
- Holubkov R.,
- Starling R.C.,
- et al.
- World Health Organization/International Society and Federation of Cardiology
- Anzai T.,
- Yoshikawa T.,
- Baba A.,
- et al.
- Kim C.H.,
- Fan T.H.,
- Kelly P.F.,
- et al.
- Caforio A.L.,
- Grazzini M.,
- Mann J.M.,
- et al.
- Schulze K.,
- Becker B.F.,
- Schauer R.,
- et al.
- Limas C.J.,
- Limas C.
- Chiale P.A.,
- Ferrari I.,
- Mahler E.,
- et al.
- Schwinger R.H.G.,
- Wang J.,
- Frank K.,
- et al.
- Hariman R.J.,
- Zeiler R.H.,
- Gough W.B.,
- et al.