Desensitization and Sequestration of Human m2 Muscarinic Acetylcholine Receptors by Autoantibodies from Patients with Chagas’ Disease*

Chronic Chagas’ disease is associated with pathologic changes of the cardiovascular, digestive, and autonomic nervous system, culminating in autonomic denervation and congestive heart failure. Previously, circulating autoantibodies that activate signaling by cardiac muscarinic acetylcholine receptors (mAChRs) have been described. However, it remains unclear whether the chagasic IgGs directly interact with the m2 mAChRs (predominant cardiac subtype), and, if so, whether chronic exposure of the mAChRs to such activating IgGs would result in receptor desensitization. Here we performed studies with purified and reconstituted hm2 mAChRs and demonstrate that IgGs from chagasic serum immunoprecipitated the mAChRs in a manner similar to an anti-m2 mAChR monoclonal antibody tested in parallel. The chagasic antibodies did not directly interact with the ligand binding site, because the binding of radiolabeled antagonist was unchanged by the addition of the chagasic IgG. In intact cells stably expressing the hm2 mAChR, the chagasic IgGs, but not normal IgGs, mimicked the ability of the agonist acetylcholine to induce two effects associated with agonist-induced receptor desensitization: a decrease in affinity for agonist binding to m2 mAChR and sequestration of the hm2 mAChRs from the cell surface. The results demonstrate that the chagasic IgGs can directly interact with and desensitize m2 mAChRs and provide support for the hypothesis of autoimmune mechanisms having a role in the pathogenesis of Chagas’ cardioneuromyopathy.

Chagas' disease, one of the most common determinants of congestive heart failure and sudden death in the world, is transmitted by the parasite Trypanosoma cruzi (1,2). Most patients survive the acute phase of the disease asymptomatically; the chronic stage of the disease develops over 20 -30 years and is manifested by cardiovascular, digestive, and autonomic nervous system disorders (3)(4)(5)(6). Several studies have focused on the disease-related loss of function of the autonomic receptors, the ␤-adrenergic and muscarinic cholinergic recep-tors (mAChR) 1 (5)(6)(7)(8)(9)(10)(11). The paradoxical severe involvement of the heart in the absence of any form of the parasite there has prompted proposals that autoimmune mechanisms participate in the pathogenesis of this cardiac neuromyopathy (12)(13)(14). Chagasic patients have been found to possess circulating antibodies that are able to interact with and modulate the activity of cardiac ␤-adrenergic and mAChRs (15)(16)(17)(18)(19). This report focuses on the molecular events initiated by the interaction of the chagasic antibodies with the mAChRs to begin to understand how the antibodies could ultimately account for the abnormalities of cardiac autonomic function characteristic of the late stages of Chagas' disease. Previously, chagasic antibodies were shown to mimic agonists of mAChR to cause inhibition of cardiac contractility, stimulation of cGMP levels, and attenuation of cAMP synthesis (18,19). Thus, the in vivo situation during the course of Chagas' disease could be similar to that of a persistent stimulus, i.e. circulating antibodies that activate cardiac mAChRs could result in receptor desensitization. Such a scenario could cause receptor/G-protein uncoupling and loss of receptors from the cell surface and explain the progressive blockade of receptors that is observed in the disease (20,21). To test this hypothesis, we used both purified human m2 mAChRs and mammalian CHO cells transfected with human m2 mAChRs to assess the ability of chagasic IgG to directly interact with and desensitize the mAChR. The results support a role for the chagasic antibodies in causing loss of mAChR function.

MATERIALS AND METHODS
Sera and IgG Purification-Trypanosoma cruzi-infected patients from metropolitan Buenos Aires were studied. These patients, classified as Group I based on WHO Expert Committee on Chagas' disease criteria (1), had positive T. cruzi serology and the absence of clinical symptoms with normal electrocardiogram and chest x-ray. Among this group, sera were chosen for their previously demonstrated ability to activate rat atrial mAChR in an atropine-sensitive manner (18,19). Normal sera taken as controls were from volunteers who had negative serology for Chagas' disease, presented no evidence of cardiovascular or chronic systemic disease or acute viral or febrile disease, and had normal electrocardiogram and chest x-ray. The age of patients and controls ranged between 25 and 50 years. IgG was isolated from chagasic or normal sera by means of DEAE-cellulose chromatography (Bio-Rad) as shown (18) and tested for purity by immunoelectrophoresis with goat anti-whole serum or anti-IgG antiserum (Sigma). The IgG concentration was determined by radioimmunodifussion assay. F(abЈ) 2 fragments were obtained from normal or chagasic patients by pepsin digestion of IgG fractions in 0.1 M sodium acetate buffer (pH 4.5) for 22 h at 37°C, dialysis against phosphate-buffered saline (PBS), and chromatography on Sephadex G-200. Purification of IgG was assessed by SDS-polyacrylamide gel electrophoresis and immunoelectrophoresis.
Purification of m2 Muscarinic Acetylcholine Receptors-Human mAChRs (hm2 mAChR) from Spodoptera frugiperda (Sf9) insect cells infected with recombinant baculovirus encoding the human m2 mAChR were purified and reconstituted as described (22)(23)(24). The hm2 mAChR purified from Sf9 cells is expressed as a 55-kDa protein, which is smaller than that observed in mammalian cells because it lacks the extensive glycosylation associated with the mammalian expressed receptor (24 (26).
Immunoprecipitation Assays-Purified reconstituted hm2 mAChRs were immunoprecipitated with chagasic antibodies or a monoclonal anti-m2 antibody (27) generously provided by Dr. M. Schimerlik from Oregon State University (Corvallis, OR). Chagasic or normal sera or anti-m2 hybridoma supernatant were precoupled to protein A-agarose beads (50 l, Pierce) overnight at 4°C. The beads were then incubated with ϳ0.7 pmol of reconstituted mAChRs at 4°C for 5 h and washed thoroughly. The immunoprecipitates were subjected to SDS-polyacrylamide gel electrophoresis using 8.5% gels and then transferred to nitrocellulose membranes. Proteins were visualized by means of the anti-m2 monoclonal antibody and a goat anti-mouse IgG conjugated to peroxidase followed by the ECL reaction (Amersham Corp.) as described (26).
Desensitization and Sequestration of the mAChR-The m2 CHO cells were grown in 100-mm dishes, washed with PBS, and then allowed to stabilize for 10 min at 37°C with serum-free fresh DMEM:F-12 containing 10 M eserine (a cholinesterase inhibitor) before the addition of treatments to induce desensitization or sequestration. To assess the effects of chagasic sera to cause decrease in high affinity agonist binding, cells were treated with serum-free medium containing either PBS (as a no drug/no IgG control), the agonist acetylcholine, or chagasic or normal IgGs. Cells were incubated for 1 h at 37°C in a 5% CO 2 atmosphere with one of the above reagents, and after extensive washing, they were harvested in 10 mM phosphate buffer (pH 7.4), 1 mM EDTA, 0.25 M sucrose. All steps were performed at 4°C. Cells were centrifuged at 400 ϫ g for 15 min, and crude membranes were prepared as described previously (27). To test for effects of pretreatments with agonist or IgGs on the affinity of the receptors for agonist, the agonist carbachol was used to compete for [ 3 H]QNB binding to mAChR in the membrane preparations. Membranes were incubated at 37°C for 75 min in 20 mM phosphate buffer (pH 7.4), 1 mM EDTA, 2 mM MgCl 2 , [ 3 H]QNB (0.5-0.6 nM), and varying concentrations of carbachol in the presence or the absence of the guanine nucleotide analogue 5Ј-guanylylimidodiphosphate (Gpp(NH)p, at 100 M). The data were analyzed with the curve-fitting program LIGAND (28).
To assess for sequestration of hm2 mAChRs, experiments were performed on intact m2 CHO cells using the hydrophilic ligand [ 3 H]Nmethylscopolamine (NMS) to determine changes in cell surface m2 mAChR number, whereas the hydrophobic ligand [ 3 H]QNB was used to assess total receptors. After pretreatment with drugs or IgGs as described above, the culture dishes were placed on ice, and the cells were thoroughly washed with cold PBS. Then they were spun down at 400 ϫ g, resuspended in ice-cold DMEM:F-12 and assessed for viability by trypan blue exclusion, and suspensions with more than 95% viable cells were used for binding assays. Saturating concentrations of [ 3 H]NMS or [ 3 H]QNB were incubated with cells in Hepes-buffered DMEM:F-12 in the presence or the absence of 1 M atropine to calculate specific binding for 2 h at 4°C for [ 3 H]NMS and for 75 min at 37°C for [ 3 H]QNB. Cells were filtered onto GF/C glass fiber filters, and radioactivity was counted in a Beckman spectrometer. Control levels of surface and total receptors were defined as the levels in cells treated the same as above but with the corresponding volume of PBS instead of drug or IgG. These values were considered as 0% internalization of each assay. The percentage of internalization due to drug or IgG treatments was calculated as the percentage of loss of surface receptors measured with [ 3 H]NMS in treated samples compared with 0% internalization con-trols. Down-regulation was defined as a net decrease in the total receptor pool as assessed by [ 3 H]QNB binding assays. Variations in cell density per plate were controlled by normalizing all data on the basis of mg protein/plate.

Molecular Interaction of Chagasic Antibodies with Purified
Human m2 mAChRs-To test the ability of chagasic antibodies to directly interact with purified hm2 mAChRs, immunoprecipitation experiments were performed with several chagasic or normal sera and purified hm2 mAChRs reconstituted in vesicles. As a control, an equivalent amount of the hm2 mAChR were also immunoprecipitated with a known monoclonal anti-m2 mAChR antibody (26). The immunoprecipitates were fractionated by SDS-polyacrylamide gel electrophoresis, and immunoprecipitated mAChR was visualized by immunoblotting with the anti-m2 mAChR monoclonal antibody (Fig. 1). The results demonstrated that the chagasic sera immunoprecipitated the hm2 mAChR (Fig. 1, lanes 1 and 2). Two immunoreactive bands were detected by the anti-m2 mAChR mAb, one corresponding to a protein of 55 kDa, the expected size for the hm2 mAChR purified from Sf9 cells, and a second species at ϳ116 kDa that likely represented the receptor dimer (24). The immunoreactive bands that were present in the immunoprecipitates from the chagasic sera migrated identically to those immunoprecipitated by the m2 mAb (Fig. 1, lane 5). The amount of mAChR immunoprecipitated by the chagasic antisera was lower than the fraction of receptors immunoprecipi-  3 and 4), or the m2 monoclonal antibody (lane 5) were precoupled to protein A-agarose beads and then incubated with equivalent amounts of m2 mAChRs as described under "Materials and Methods." Immunoprecipitated proteins were fractionated by SDS-polyacrylamide gel electrophoresis and transferred to nitrocellulose, and the immunoprecipitated receptors were identified by immunoblotting with the anti-m2 monoclonal antibody. The results are representative of six other experiments with different chagasic and normal sera. M r indicates the position of the molecular weight standards. tated by the m2 mAb run in parallel; however, the amount of mAChR immunoprecipitated by the chagasic sera was significant and far greater than the trace amounts (background levels) immunoprecipitated by the normal sera used as control (Fig. 1, lanes 3 and 4). These results provide a direct demonstration of the recognition of the hm2 mAChR by antibodies present in chagasic sera.
To test whether the chagasic sera interacted directly with the ligand binding site of the m2 mAChR or with another domain, we tested whether the chagasic sera could compete with the antagonist [ 3 H]QNB in ligand binding assays. For these studies we prepared membranes from CHO cells expressing the hm2 mAChR and performed saturation binding isotherms in untreated membranes and membranes treated with either normal or chagasic IgG. No differences in the K d or B max values were observed among the three groups (Table I).
Decreased Affinity of Agonist Binding to m2 mAChRs after Interaction with Chagasic IgG-The possibility that the persistent activation of m2 mAChRs by chagasic antibodies could lead to receptor desensitization was evaluated. Desensitization of G-protein-coupled receptors is associated with several events, in particular, receptor/G-protein uncoupling and sequestration of the receptors away from their normal membrane environment. A hallmark of receptor uncoupling from G-proteins is the conversion of receptors from a high to low affinity state. The high affinity state for agonists is considered to be a receptor/G-protein complex, whereas the lower affinity state is thought to represent the free receptor. In agonist binding studies in vitro a loss of high affinity agonist binding can be induced upon the addition of GTP or its analogs, presumably due to the activation of the G-protein and its subsequent dissociation of the G-protein from the receptor. Agonist-induced desensitization of the ␤-adrenergic receptor has been associated with the loss of high affinity agonist binding and conversion of receptors to a low affinity state (29).
To test for desensitizing effects of chagasic antibodies on agonist binding properties, we compared the effects of pretreatment of intact cells with either the agonist acetylcholine or normal or chagasic IgG on agonist affinity. After each pretreatment, membranes were prepared and used in carbachol/ [ 3 H]QNB competition binding experiments in the presence or the absence of a guanine nucleotide analogue. In assays of membranes from control cells (treated with PBS), shallow, biphasic agonist competition curves were obtained in the absence of Gpp(NH)p, reflecting both high and low affinity agonist binding (Fig. 2a). The addition of Gpp(NH)p right shifted and steepened the curves, as expected for the guanine nucleotidedependent conversion of high affinity sites into low affinity sites (Fig. 2a, untreated). The corresponding IC 50 values and Hill coefficients reflected these changes (Table II). In addition, analysis of the data with the curve fitting program LIGAND (28) demonstrated the expected results, in that a high and low affinity state were observed and Gpp(NH)p caused a conversion of receptors from the high to low affinity state (Table III). Pretreatment of cells either with the agonist acetylcholine under conditions that lead to desensitization (27) or with chagasic IgG produced rightward shifts of the concentration-displacement curves of carbachol in the absence of Gpp(NH)p compared with untreated cells (compare Fig. 2, b and d, to Fig. 2a; Table  II). The effects caused by either pretreatment were not as extensive as those induced by the addition of Gpp(NH)p to the in vitro assays. The protease digestion fragment of chagasic IgG, F(abЈ) 2 , had similar effects to the nondigested chagasic IgG (data not shown). The shift induced by chagasic IgG was not observed when incubations were performed with normal IgG (Fig. 2c, Table II). Analysis of the data with the LIGAND program indicated that the decrease in overall affinity caused by the chagasic IgG was due to an increase in the value of K H , whereas there was no detectable change in the percentage of receptors in the higher affinity state (Table III). Treatment of the cells with agonist produced a similar trend (Table III). The LIGAND analysis also suggested that the pretreatment of the   (Table III). The data obtained with the normal IgG were similar to those from the untreated cells (Table III). It is not entirely clear why the effects of the pretreatments with agonist or chagasic IgG produced increases in K H without the expected decrease in R H . One explanation is that the chagasic IgG or agonist only partially desensitized the receptor. Indeed, in studies of the desensitization of the ␤-adrenergic receptor, similar results were obtained (29). With partial desensitization, a decrease in numbers of high affinity ␤-adrenergic receptors was less apparent, and the main effect seemed to be a small increase in K H (29) as was observed here. The present results also were similar to those we previously obtained in chick heart tissue treated with desensitizing doses of acetylcholine, where K H for agonist binding to the mAChR was increased, whereas the percentage of high affinity receptors was less affected (30). In addition, we observed very similar trends in studies with heterologously expressed hm2 mAChR in Sf9 cells (31). It appears that the effects of desensitization on the decreased affinity of agonist binding to the m2 mAChR are less extensive than can be induced in vitro upon the addition of guanine nucleotides to ligand binding assays. Further studies will be necessary to dissect these complex patterns of responses that are elicited by desensitization. Nevertheless, the present results are consistent with the conclusion that acute exposure of the m2 CHO cells to the chagasic IgG resulted in a decreased affinity of the receptor for agonist. These desensitizing effects of the chagasic IgG were similar to those induced by exposure to agonist. The results are consistent with the suggestion that the agonist-like activity of the chagasic sera can induce desensitization of the m2 mAChR in intact cells.
Sequestration of m2 mAChRs by Chagasic IgG-To test if chagasic IgG induced sequestration of receptors, cells expressing m2 mAChRs were incubated for different times with increasing concentrations of chagasic or normal IgGs or F(abЈ) 2 before binding assays were performed. Comparable incubations with acetylcholine were run in parallel. The hydrophilic ligand [ 3 H]NMS and the lipophilic ligand [ 3 H]QNB were used in whole cell binding assays to assess cell surface and total receptors, respectively. Both the chagasic IgG and acetylcholine induced a time-dependent sequestration of the mAChRs to a similar extent, whereas normal IgG had no significant effect (Fig. 3a). The effects of both acetylcholine and the chagasic IgG were concentration-dependent; the maximum effect observed with the chagasic IgG occurred at 0.2 mg/ml and was similar to that caused by 100 nM acetylcholine (Fig. 3b). The chagasic F(abЈ) 2 also induced internalization but with a lower efficacy (Fig. 3b). Normal IgG was ineffective at any concentration (Fig.  3b). In the time periods tested, no down-regulation of receptors by chagasic IgGs, normal IgGs, or acetylcholine was observed based on measurements of total mAChR with [ 3 H]QNB (data not shown). The present report provides evidence of a specific interaction between antibodies from the serum of chagasic patients and hm2 mAChRs and describes some early consequences of the persistent activation of the receptors by the antibodies. The results obtained support the hypothesis that antibodies having an "agonist-like" activity can participate in the regulation of the receptors they bind to, thus influencing their normal function in vivo. The antibody-induced decrease in agonist affinity occurred after 1 h of incubation of m2 CHO cells with chagasic IgGs or F(abЈ) 2 , whereas internalization was extensive by 90 -120 min, similar to the effect of acetylcholine. The data demonstrate that the agonist-like behavior of the antibodies is sufficient to induce the series of reactions associated with desensitization of the receptors. The results are consistent with our previous observations on the ability of chagasic IgG to impair agonist-induced inhibition of contractility of rat atria after prolonged incubation with IgGs (18). It is unlikely that the present results could have arisen from nonspecific interactions, because the data in Fig. 1 demonstrated an ability of the chagasic antibodies to directly interact with the m2 mAChR. In addition, whereas previous studies have suggested effects of chagasic serum on adrenergic signaling (15)(16)(17)(18)(19), the CHO cells used in this study lack adrenergic receptors, making it unlikely that the observed desensitizing effects of the chagasic antibodies on the mAChR were due to cross-talk arising from effects on signaling pathways downstream of the receptors.
Chagasic patients exhibit several signs of impaired parasympathetic function (14); however, the molecular mechanisms  Fig. 2 were analyzed with the curve-fitting program LIGAND (28). K H and K L refer to K i values for carbachol binding to high and low affinity states of the receptor, respectively, whereas R H and R L refer to the percentage of receptors in each affinity state. Shown are the means Ϯ S.E. from 6 -8 experiments similar to the one shown in Fig. 2 3. Internalization of m2 mAChRs by chagasic IgG. a, CHO cells expressing m2 muscarinic receptors were incubated with serumfree medium containing eserine and 1 mM acetylcholine (filled triangles) or 0.2 mg/ml chagasic (filled circles) or normal (empty circles) IgGs. Cells were incubated for the indicated times and then washed and harvested, and radioligand binding assays were performed as described under "Materials and Methods." The data are the means Ϯ S.E. of five experiments performed in triplicate with five different IgGs. b, cells were preincubated with varying concentrations of acetylcholine (ACh), chagasic IgG (ch IgG), chagasic F(abЈ) 2 (ch F(abЈ)2), or normal IgG (n IgG) as indicated. After 2 h of incubation cells were washed and harvested, and radioligand binding was performed. The data are the means Ϯ S.E. of five experiments with different IgGs. involved remain unexplained. Thus, the major point of our report is that early events of agonist-promoted desensitization of m2 mAChR can be initiated by autoantibodies that bind to and persistently activate the receptors. Further studies will be required to address the issue of whether early molecular alterations in the function of the cardiac autonomic nervous system during Chagas' disease are related to the presence of autoantibodies promoting receptor desensitization. Interestingly, the existence of autoantibodies against autonomic receptors has been reported in other cardiomyopathies such as human dilated cardiomyopathy (32)(33)(34) and congenital heart block (35), but evidence is still lacking about the mechanisms underlying their potential pathological role. Whether or not autoantibodies play a general role in other cardiomyopathies remains to be determined.