Autoimmunity against M2 muscarinic acetylcholine receptor induces myocarditis and leads to a dilated cardiomyopathy-like phenotype

Authors


Abstract

Patients with dilated cardiomyopathy (DCM) often have autoantibodies against cardiac antigens including the M2 muscarinic acetylcholine receptor (M2R). To elucidate the role of autoimmunity against M2R in disease development, we induced an immune response against M2R by adoptive transfer into Rag2–/– mice of splenocytes from M2R–/– mice immunized with a recombinant M2R protein. T lymphocytes transiently infiltrated the heart in recipient mice followed by morphological changes in cardiomyocytes. These mice produced IgG antibodies against M2R, which bound to cardiomyocytes in vivo and decreased the amplitude of calcium signals in isolated rat cardiomyocytes in vitro. Recipient mice showed increased heart weights associated with increased intraventricular diameter, decreased systolic function, and increased action potential duration, which are characteristics of DCM. Our results suggest that myocarditis and DCM associated with the presence of anti-M2R antibodies are autoimmune diseases with a risk of progressing to the terminal stage. Our mouse model will be useful in the analysis of the molecular mechanisms of disease progression and the development of new therapies for DCM.

Introduction

Dilated cardiomyopathy (DCM) causes severe heart failure in young adults including occasional sudden cardiac death. Mutations in genes encoding myocyte structural proteins are found in about 30–40% of DCM, but the etiology of the remaining 60–70% of cases is poorly understood [[1-3]]. Advances in the field of immunology have shed light on the close relationship between immunological disturbances and the progression of heart failure [[4-7]]. Adenovirus or enterovirus-specific RNA sequences are often detected in the myocardium of patients with DCM and myocarditis [[7-9]]. Myocardial-reactive antibodies such as those against β1 adrenergic receptor (β1AR), M2 muscarinic acetylcholine receptor (M2R), and cardiac myosin have also been observed in the sera of these patients [[10-14]]. As a result, it has been hypothesized that prolonged immunological abnormalities after myocarditis, presumably induced by a subclinical viral infection or autoimmunity, could lead to DCM [[5, 6]]. Evidence supporting this hypothesis has accumulated through clinical and experimental studies [[15-17]]. For example, immunization of animals with peptides derived from proteins expressed in cardiomyocytes such as β1AR [[16, 17]] or cardiac myosin [[15]] or the transfer of antibodies against cardiac antigens mimics the prolonged immune abnormalities observed after acute myocarditis [[18, 19]]. However, the progression from myocarditis to DCM has not been investigated in detail. The causes of the initial myocardial damage and subsequent cardiac dysfunction along with the subsequent role of autoimmunity in DCM are still unclear.

The pathophysiological role of autoantibodies in disease progression has been established in only some autoimmune diseases such as pemphigus vulgaris (PV), involving anti-desmoglein 3 (Dsg3) antibodies [[20]], and myasthenia gravis with anti-nicotinic acetylcholine receptor antibodies [[21]]. Certain types of antibodies may affect cardiomyocyte functions by binding membrane receptors on heart cells in DCM patients [[10-14]] and in animal models [[15-17]]. For example, agonistic anti-β1AR autoantibodies induce cardiomyocyte apoptosis in vitro [[22]] and the passive transfer of β1AR-specific antisera induced a DCM-like phenotype in recipient rodents in vivo [[17]]. Functional studies on heart failure in humans or experimental animals have suggested the involvement of altered intracellular calcium levels in disease onset and progression [[23]]. Since signals through β1AR increase the intracellular Ca2+ concentration ((Ca2+)i), thereby prolonging the action potential duration (APD), it has been suggested that anti-β1AR antibodies are involved in disease progression.

On the other hand, other studies have suggested the involvement of autoantibodies distinct from anti-β1AR antibodies. Immunoadsorption (IA) to remove IgG class anti-cardiac antibodies using anti-IgG columns has been used to treat DCM patients [[24, 25]]. Acute hemodynamic improvement including increased cardiac index and decreased systemic vascular resistance was observed following repeated IAs in DCM patients [[25]]. Interestingly, however, the eluant from IA columns, which contained the adsorbed IgG antibodies, caused an immediate decrease rather than increase of (Ca2+)i as well as cell shortening when added to cultured cardiomyocytes [[26]]. Consistent with these results, a study using a patch-clump technique showed that antibodies from DCM patients negatively modulated the L-type calcium current in isolated cardiomyocytes [[27]]. Since β1AR-specific autoantibodies usually increase (Ca2+)i, the cardiodepressant IgGs removed by IA were likely of a different type. These studies suggest that antibodies other than anti-β1AR antibodies play a significant role in disease progression; for example, anti-M2R IgGs are also observed in DCM patients [[27]]. Accumulating evidence suggests that IA can a potential therapeutic approach for DCM patients [[28]]. In a small group of severe heart failure patients, significant correlation between IgG3 anti-M2R antibodies titer and disease severity was observed and IA significantly improved cardiac functions [[29]]. We thus focus in this study on the role of autoimmunity against M2R in DCM disease progression. Our results suggest that M2R is a target of chronic autoimmunity leading to DCM and that the presence of anti-M2R antibodies observed in myocarditis patients [[11, 14]] is a sign of disease progression to DCM.

Results

Lymphocyte infiltration and cardiomyocyte destruction upon induction of an immune response against M2R

More than 90% of the muscarinic acetylcholine receptors expressed by cardiomyocytes are encoded by the M2R gene, one of five distinct genes for the muscarinic acetylcholine receptor family [[30]]. To investigate the potential role of autoimmunity against M2R in the chronic progression from myocarditis to DCM, we attempted to induce an autoimmune reaction against M2R in wild-type mice. However, little sign of autoimmunity was observed several weeks after immunization (data not shown). Previous studies have shown that it takes several months or more to detect signs of autoimmunity in the heart after immunizing rats with cardiac components [[17]]. To overcome these difficulties, we employed a method previously established to induce an autoimmune skin disease, PV. In this model, splenocytes of Dsg3–/– mice immunized with a recombinant Dsg3 protein were adoptively transferred into Rag2–/– recipient mice [[20]]. We thus immunized M2R–/– mice with a recombinant mouse M2R protein consisting of the 2nd- extracellular loop (ECL) fused with glutathione-S-transferase (GST). M2R-immunized M2R–/– mice produced antibodies in the sera reactive with recombinant M2R proteins (A450 = 1.00 ± 0.44 at 1:10,000 dilution). Immunized splenocytes were then transferred into Rag2–/– mice in which donor lymphocytes migrate rapidly to secondary lymphoid organs due to the lack of endogenous T and B cells [[20]].

As early as 2 weeks after splenocyte transfer, CD4+ and CD8+ T cells, but not B cells or neutrophils, infiltrated the myocardium of the recipient Rag2–/– mice (Fig. 1, A–E). By 3 weeks after transfer, lymphocyte infiltration had progressed, becoming more diffuse, and cardiomyocyte necrosis became apparent (Fig. 1, F–I). Immunohistochemical analyses showed that CD4+ and CD8+ T cells but not B cells were observed around necrotized cardiomyocytes (Fig. 1, H and I, and data not shown). Nuclear swelling was observed in some necrotized cardiomyocytes (Fig. 1, F–I, arrows and pink circles), which was different from that observed in physiologically hypertrophied cardiomyocytes (data not shown).

Figure 1.

Lymphocyte infiltration and cardiomyocyte destruction upon induction of an immune response against M2R. (A–E) Hearts of Rag2–/– recipient mice were harvested 2 weeks after splenocyte transfer. Sections of the (A) right ventricular wall, (B) right atrium wall, and (C–E) infiltrated lymphocytes were stained with (A, B) H&E and (C) anti-CD4 (GK 1.5), (D) anti-CD8 (53–6.7), or (E) anti-B220 antibodies. Data shown are representative of n = 4 mice. (A, B) Scale bars = 200 μm, (C–E) scale bars = 50 μm. (F–I) Hearts of Rag2–/– recipient mice were harvested 3 weeks after splenocyte transfer and sections were stained with (F, G) H&E and (H) anti-CD4 or (I) anti-CD8 antibodies. Data shown are representative of n = 5 mice. Scale bars = 100 μm. (F) The arrow indicates necrotic cardiomyocytes surrounded by lymphocytes. (G) The arrow indicates a degenerative necrotic cell with nuclear swelling. (H and I) Immunohistochemical staining of infiltrated lymphocytes (CD4+ and CD8+ cells shown by black circles) around necrotized cardiomyocytes (pink circles). (J, K) Abnormal morphological changes of nuclei at 6 weeks after transfer were evaluated by histological staining with H&E. Data shown are representative of n = 5 mice. Scale bars = 50 μm. (J) Arrows indicate multinuclear cardiomyocytes. (K) The arrow indicates the abnormal swelling of a nucleus. (L, M) Morphological changes in hearts at 10 weeks post transfer were evaluated after H&E staining. Data shown are representative of n = 5 mice. Scale bars = 200 μm. (M) Red lines indicate the disordered alignment of cardiomyocytes such as abnormal branching and disruption. (N) Heart sections at 13 weeks after splenocyte transfer into Rag2–/– mice were examined and instances of abnormal swelling of nuclei (over twofold larger in size than healthy nuclei) and multinuclear cardiomyocytes were counted to quantify the morphological changes. These parameters were counted at 40× magnification in five different fields. Rag2–/– recipient mice transferred with M2R-immunized M2R–/– splenocytes (experimental group, n = 8) and Rag2–/– recipient mice transferred with M2R-immunized M2R+/+ splenocytes, GST-immunized M2R–/– splenocytes, GST-immunized M2R+/+ splenocytes (control groups n = 6, 6, and 7, respectively) were examined. Data are shown as mean + SEM and are pooled from n = 3 experiments. *p < 0.05, one-way ANOVA followed by Tukey–Kramer multiple comparison test.

Six weeks after transfer, inflammatory cell infiltration into heart tissues declined and only small numbers of infiltrating inflammatory cells were observed. As shown in Fig. 1 (J and K), necrosis, nuclear enlargement, and disordered alignment were widely observed in the cardiomyocytes of recipient mice. Ten weeks after transfer, the disordered alignment of cardiomyocytes, as exemplified by abnormal branching and disruption, was readily observed but few infiltrating lymphocytes were found (Fig. 1, L and M). Such cardiomyocyte morphological changes are also characteristic of DCM. The number of cardiomyocytes with abnormal morphology was significantly higher in Rag2–/– mice transferred with M2R-immunized M2R–/– splenocytes than control recipient mice (Fig. 1 N). We noted little fibrosis in the heart tissues as examined by Masson's trichrome staining (data not shown). Contrary to the heart tissue, we observed little evidence of T-cell infiltration into the liver, lungs, and salivary glands of Rag2–/– mice transferred with splenocytes of M2R–/– mice immunized with M2R 6 and 8 weeks after splenocyte transfer (data not shown).

Anti-M2R antibody production and IgG deposition on the myocardium

Using direct immuofluorescence analysis, we noted IgG deposition on cardiomyocytes in recipient mice 15 weeks after the transfer of M2R-immunized M2R–/– splenocytes but not in mice receiving GST-immunized M2R–/– splenocytes (Fig. 2, A and B). Diffuse IgG deposition was observed throughout the entire cardiac wall regardless of the presence of tissue damage. Complement deposition on the myocardium was not found in these mice (data not shown). To examine whether the antibody was reactive with M2R, we stained a frozen section of heart tissue with IgG purified from the sera of recipient mice. Indirect immunofluorescence analysis showed that the extracted IgG bound to cardiomyocytes of wild type but not M2R–/– mice (Fig. 2, C and D), indicating that most of the IgG antibodies reactive with cardiomyocytes are specific for M2R.

Figure 2.

Anti-M2R antibody production and IgG deposition on the myocardium. (A, B) Hearts were obtained from (A) Rag2–/– mice transferred with splenocytes from M2R-immunized M2R–/– mice and (B) Rag2–/– mice transferred with splenocytes from GST-immunized M2R–/– mice 15 weeks after transfer. Cryosections were stained with FITC-conjugated antimouse IgG to determine IgG deposition on cardiomyocytes. (C, D) The specificity of IgG isolated from the sera of Rag2–/– mice transferred with splenocytes from M2R immunized M2R–/– mice was tested on frozen sections of hearts from untreated (C) M2R–/– or (D) wild-type mice. Sections were stained with indirect immunofluorescence using the IgG fraction as the primary antibody and Alexa Fluor 488-conjugated antimouse IgG as the secondary antibody. Results shown are representative of five sections in each staining and are representative of n = 4 mice/experiments.

We next examined the time course of antibody production using ELISA analyses of serum samples. Antibodies reactive with the 2nd-ECL of M2R were readily detected in the sera of recipient mice as early as 1 week after transfer (data not shown) and the increase in antibody titers was maintained afterwards (Fig. 3 A: closed symbols). M2R-specific IgGs contained IgG1 and IgG2b but not IgG3 subclasses (data not shown). Such antibodies were not observed in mice transferred with splenocytes of M2R–/– mice immunized with GST (Fig. 3 A: open symbols and data not shown). We further examined the presence of antibodies against the cytoplasmic portion of M2R using a fusion protein consisting of the 3rd-intracellular loop (ICL) of the M2R and GST. Antibodies against the 3rd-ICL of M2R were not observed in the sera of recipient mice during early time points but became apparent 10 weeks after splenocyte transfer and the titers increased thereafter (Fig. 3 B: closed symbols and data not shown). Such epitope spreading is likely due to the destruction of cardiomyocytes as observed in Fig. 1 and the resulting exposure of cytoplasmic components to the immune system. No antibody reactive with β1AR or cardiac myosin was detected in the sera of these mice even after tissue damage (data not shown). These results collectively suggest that myocarditis in this model is induced by a specific immune response against M2R.

Figure 3.

IgG production from recipient Rag2–/– mice transferred with splenocytes from M2R-immunized M2R–/– mice. (A) Sera collected at the indicated time points from Rag2–/– mice transferred with splenocytes from M2R-immunized M2R–/– mice (closed diamonds, n = 9) or Rag2–/– mice transferred with splenocytes from GST-immunized M2R–/– mice (open squares, n = 11) were examined for anti-M2R-2nd-ECL IgG by ELISA. (B) Sera collected on the indicated time points from Rag2–/– mice transferred with splenocytes from M2R-immunized M2R–/– mice (closed diamonds, n = 9) or Rag2–/– mice transferred with splenocytes from GST-immunized M2R–/– mice (open squares, n = 11) were examined for anti-M2R-3rd ICL IgG by ELISA. Data are shown as mean ± SEM of the indicated values and are pooled from n = 2 experiments. *p < 0.05, **p < 0.01, ***p < 0.001, unpaired Student's t-test.

Increased heart weight with decreased cardiac function

Heart weight normalized by body weight (mg/g) increased significantly (p < 0.05) in Rag2–/– mice transferred with splenocytes of M2R-immunized M2R–/– mice (M2R-immunized M2R–/– group) compared with those transferred with splenocytes of GST-immunized M2R–/– mice (GST-immunized M2R–/– group) 15 weeks after splenocyte transfer (Table 1). Normalization by tibial length (mg/mm) showed the same results (Table 1, p < 0.01). No increase in heart weight was observed in Rag2–/– mice transferred with splenocytes of M2R-immunized M2R+/− mice (M2R-immunized M2R+/− group) when compared with those transferred with splenocytes of GST-immunized M2R+/− mice (GST-immunized M2R+/− group) (Table 1). These results indicate that myo-carditis induced by autoimmunity against M2R leads to an increase in heart weight.

Table 1. Changes in heart weight 15 weeks after splenocyte transfera
Splenocyte transferHW/BWHW/TL
from(mg/g)(mg/mm)
  1. a

    HW: heart weight; BW: body weight; TL: tibial length.

  2. *p < 0.05 versus GST-immunized M2R–/–, unpaired Student's t-test.

  3. **p < 0.01 versus GST-immunized M2R–/–, unpaired Student's t-test.

M2R-immunized M2R–/– (n = 9)4.8 ± 0.6*7.7 ± 0.7**
GST-immunized M2R–/– (n = 11)4.4 ± 0.36.9 ± 0.6
M2R-immunized M2R+/− (n = 10)5.0 ± 0.38.5 ± 1.8
GST-immunized M2R+/− (n = 10)4.8 ± 0.38.1 ± 1.0

We then examined cardiac function in mice with myocarditis by echocardiographic analysis 15 weeks after splenocyte transfer. As shown in Fig. 4 and Table 2, we observed dilatation of the left ventricular end-diastolic dimension (LVEDD) (3.5 ± 0.3 mm in control GST-immunized M2R–/– group (n = 11) versus 3.9 ± 0.5 mm in M2R-immunized M2R–/– group (n = 9); p < 0.05) and decreased % fractional shortening (%FS) (41 ± 4% in GST-immunized M2R–/– group versus 33 ± 5% in M2R-immunized M2R–/– group; p < 0.005) in recipient mice transferred with M2R-immunized M2R–/– splenocytes. These results indicate that myocarditis induced by autoimmunity against M2R resulted in an increase in heart weight accompanied by ventricular dysfunction and dilation.

Table 2. Changes in left ventricular diameter and systolic function 15 weeks after splenocyte transfer
Splenocyte transfer fromLVEDD (mm)%FS
  1. *p < 0.05 versus GST-immunized M2R–/–, unpaired Student's t-test.

  2. **p < 0.005 versus GST-immunized M2R–/–, unpaired Student's t-test.

M2R-immunized M2R–/– (n = 9)3.9 ± 0.5*33 ± 5**
GST-immunized M2R–/– (n = 11)3.5 ± 0.341 ± 4
Figure 4.

Echocardiograms of hearts of recipient mice. Representative echocardiograms of (A) recipient Rag2–/– mice transferred with splenocytes from M2R-immunized M2R–/– mice (n = 9) or (B) those of Rag2–/– mice transferred with splenocytes from GST-immunized M2R–/– mice (n = 11) are shown 15 weeks after transfer.

Effect of anti-M2R autoantibody on cardiomyocytes and cardiac tissues

We next examined the pathophysiological role of anti-M2R antibodies and IgG deposition in disease progression. We collected sera from recipient Rag2–/– mice and prepared IgG fractions from the sera. The effect of IgG antibodies on calcium homeostasis was measured in cultured neonatal rat ventricular cardiomyocytes by monitoring the transient changes of (Ca2+)i using a fluorescent calcium indicator, Fluo-4, under field stimulation. As shown in Fig. 5 A, IgG from the M2R-immunized M2R–/– group significantly (p < 0.05) decreased (-39 ± 8%, n = 6) the amplitude of the transient change in (Ca2+)i in isolated cardiomyocytes, whereas control IgG fractions from the M2R-immunized M2R+/− group showed little affect (Fig. 5 B; -15 ± 12%, n = 6). The effect of IgG from the M2R-immunized M2R–/– group was strongly blocked by AF-DX 116, an M2R selective antagonist. Figure 5 C shows that AF-DX 116-pretreated cardiomyocytes did not show a significant decrease in the amplitude of (Ca2+)i (-21 ± 9%, n = 4) upon treatment with IgG from the M2R-immunized M2R–/– group. Similarly, a synthetic peptide derived from the 2nd-ECL of M2R significantly blocked the effect of anti-M2R IgG (Fig. 5 D; -11 ± 12%, n = 6). As expected, the synthetic peptide itself did not significantly affect the amplitude of (Ca2+)i (Fig. 5 E; -11 ± 11%, n = 6). These results suggest that the anti-M2R antibodies do react with M2R and actually have an electrophysiological role in cardiomyocyte contraction.

Figure 5.

Anti-M2R IgG causes a decrease in the amplitude of (Ca2+)i in isolated rat ventricular cardiomyocytes. The change in (Ca2+)i of cardiomyocytes cultured in 1 mL media was monitored using a fluorescent calcium indicator Fluo-4 AM (scale bars = 700 ms). (A) A representative (Ca2+)i tracing at baseline (open circle) and 30 min (filled circle) after IgG administration. After establishing the baseline, 2 μg of IgG from Rag2–/– recipient mice transferred with splenocytes from M2R-immunized M2R–/– mice (anti-M2R IgG) were added to the culture and calcium signals examined. The average of the amplitude of (Ca2+)i was −39 ± 8% (n = 6). (B) A representative (Ca2+)i tracing at baseline (open circle) and 30 min after (filled circle) IgG administration. Similarly, after establishing the baseline, 2 μg IgG from Rag2–/– mice transferred with splenocytes from M2R-immunized M2R+/− mice (control IgG) were added to the culture and calcium signals examined. The average of the amplitude of (Ca2+)i was −15 ± 12% (n = 6). (C) Similarly, after pretreatment with AF-DX 116 (2 μM) (open circle), 2 μg of anti-M2R IgG were added (filled circle) (−21 ± 9%, n = 4). (D) Activity of anti-M2R IgG was blocked by administration of 500 μg of synthetic peptide derived from 2nd-ECL of M2R was added (−11 ± 12%, n = 6). (E) The synthetic peptide (500 μg) only was added (−11 ± 11%, n = 6). All data shown are representative of the indicated number of samples.

In order to observe the chronic effect of the IgG on cardiac tissues, action potentials were recorded using a standard microelectrode technique. A prolongation of APD was observed in all six Rag2–/– mice receiving splenocytes from the M2R-immunized M2R–/– group. As shown in Fig. 6 and Table 3, APDs at 20, 50, and 90% repolarization (APD20, APD50, and APD90, respectively) in the left ventricular papillary muscle were significantly (p < 0.005) prolonged in samples isolated from the M2R-immunized M2R–/– group compared with values observed in samples from mice receiving splenocytes from the M2R-immunized M2R+/− group. These results collectively suggest that antibodies against M2R lead to cardiac electrical remodeling.

Table 3. Action potential parameters of the left ventricular papillary muscle isolated from Rag2–/– recipient mice 15 weeks after splenocyte transfera
Splenocyte transfer fromRP (mV)AMP (mV)APD90 (ms)APD50 (ms)APD20 (ms)
  1. a

    RP: resting membrane potential; AMP: action potential amplitude; APD90,50,20: ADP at 90, 50, and 20% reporlarization.

  2. *p < 0.005 versus M2R-immunized M2R+/−, unpaired Student's t-test.

M2R-immunized M2R–/– (n = 6)−63.8 ± 8.580.4 ± 9.458.5 ± 10.3*14.0 ± 2.8*7.0 ± 1.4*
M2R-immunized M2R+/− (n = 6)−65.3 ± 6.072.1 ± 8.932.8 ± 8.99.0 ± 2.34.7 ± 0.8
Figure 6.

Effect of antibodies on action potential. The action potential from the endomyocardium of the left ventricle prepared 15 weeks after splenocyte transfer was recorded by standard microelectrodes. Representative action potential tracings obtained with Rag2–/– recipient mice transferred with M2R-immunized M2R–/– splenocytes (black line, n = 6) and Rag2–/– recipient mice transferred with M2R-immunized M2R+/− splenocytes (gray line, n = 6) are shown. All data shown are representative of the indicated number of samples.

Discussion

DCM is a myocardial disease characterized by ventricular dilatation and the progressive decline of myocardial contractile function [[31]]. It is assumed that progressive inflammation leads to DCM beginning from symptomatic or asymptomatic myocarditis. Although the presence of cellular and humoral immunity against cardiac self-antigens, including α- and β-myosin heavy chains, heart mitochondrial proteins, β1AR and M2R, have been described in patients with DCM [[27, 32]], it has been difficult to demonstrate a pathological role for these autoantibodies in the process of DCM development. Although animal models have contributed to the elucidation of disease mechanisms, previously reported DCM models required an extensive time period in order to induce the phenotype of cardiomyopathy using repeated immunizations with cardiac components [[15, 17]], presumably due to self-tolerance mechanisms. To overcome the barrier of self-tolerance, in this study we utilized a previously established method, which was successfully applied to develop a PV mouse model with an autoimmune reaction against Dsg3, the target antigen of PV [[20]]. Our present results prove that the method used to establish the PV model in mice can be applied to other autoimmune disease models. Using M2R–/– mice immunized with the 2nd-ECL of M2R, we have herein established a mouse myocarditis model producing IgG antibodies reactive with M2R. The advantage of our method is the development of acute myocarditis in a short time period induced by a single adoptive transfer of splenocytes from M2R–/– mice immunized with M2R. Recipient mice go on to exhibit a DCM-like phenotype in the later phase of the disease.

After acute myocarditis, the lymphocyte infiltration gradually decreased and the morphological changes in cardiomyocytes such as nuclear enlargement and disordered alignment developed. In addition, the recipient mice display an increase in heart weight and marked prolongation of the APD in the left ventricular myocardium, which is associated with repolarization abnormalities. Such morphological and functional changes are consistent with the characteristics of DCM. Whereas the thinning of the left ventricle wall, a common finding in DCM patients, was not observed in this model, echocardiograms showed contractile dysfunction along with chamber dilatation, which may represent an early stage of DCM. It has been unclear whether the production of autoantibodies against M2R in DCM patients is a result of inflammatory responses or whether such antibodies are pathogenic and actively participate in the disease progression [[26-29]]. Our present findings strongly suggest that the anti-M2R antibodies participate in the electrical remodeling of left ventricular muscles and subsequently lead to DCM-like cardiac dysfunction.

The fact that the adoptive transfer of M2R-immunized M2R–/– splenocytes, but not M2R-immunized M2R+/− splenocytes, induced myocarditis, a subsequent increase in heart weight and electrical remodeling of the left ventricular papillary muscles suggests that the myocarditis is induced in an autoantigen- (in this case, M2R) specific manner as has been shown in PV model mice [[20]]. However, unlike the PV model in which the transfer of naïve Dsg3–/– splenocytes was sufficient to induce the PV phenotypes in recipient Rag2–/– mice, the transfer of naïve M2R–/– splenocytes or GST immunized M2R–/– splenocytes did not induce myocarditis in recipient Rag2–/– mice. The onset of an autoimmune reaction against M2R may require events in addition to a break of tolerance, such as a strong priming effect of adjuvants, which might be provided by microbial infection. Another difference between the PV and myocarditis models is that while autoantibody is the only cause of disease development in both PV patients and PV model mice and no lymphocyte infiltration is observed in the affected skin lesions [[20]]. In contrast, transient but significant T lymphocyte infiltration was observed in our myocarditis model. Although homeostatic proliferation in lymphopenic mice sometimes induces autoimmunity [[33, 34]], we observed little evidence of T-cell infiltration into the liver, lungs, and salivary glands of recipient Rag2–/– mice in both PV model previously reported [[20]] and myocarditis model reported here. Adoptive transfer of relatively large number of splenocytes (50 million/mouse) seems sufficient to block homeostatic proliferation of transferred lymphocytes in recipient Rag2–/– mice.

In our myocarditis model, cardiomyocytes undergoing destruction were surrounded by CD4+ and CD8+ lymphocytes. IgG antibodies against the 3rd-ICL of M2R, an antigen to which the immune system had never been exposed prior to the destruction of cardiomyocytes, were produced at later times. Intense cell damage induced by M2R-reactive T-cell infiltration likely led to the exposure of the cytoplasmic portion of M2R and triggered further autoimmune responses and epitope spreading. It is possible that the T-cell-mediated immunity initiates myocarditis and the B-cell immunity with the production of anti-M2R IgG antibodies induces the cardiomyopathic changes in DCM disease in humans as well.

Active immunization of animals with synthetic peptides derived from the 2nd-ECL of β1AR resulted in a significant increase in heart weight and the phenotype of cardiomyopathy 9 months after the immunization [[17]]. However, anti-β1AR autoantibodies were not detected in the sera of Rag2–/– mice transferred with splenocytes from M2R-immunized M2R–/– mice, nor were anti-cardiac myosin autoantibodies identified in the same sera. These facts exclude the involvement of anti-β1AR and anti-cardiac myosin antibodies in our model. Rather than these antibodies, anti-M2R antibodies are likely involved in the disease process. Anti-M2R IgG decreased the amplitude of transient (Ca2+)i in neonatal rat ventricular cardiomyocytes, which is consistent with the results obtained with the IgGs purified by IA columns from DCM patients’ sera [[26]] and the effects of anti-M2R antibodies on guinea pig cardiomyocytes [[35]]. The antibodies had specificity for M2R because a selective M2R antagonist AF-DX 116 as well as a synthetic M2R peptide blocked the activity. Because a decrease in (Ca2+)i is expected to induce APD shortening [[36]], the APD prolongation observed in our model was unexpected. M2R is coupled with an inhibitory G protein, Gi. In contrast, β1AR is coupled with a stimulatory G protein, Gs. It is intriguing that autoimmunity against these functionally opposite autonomic membrane receptors resulted in a similar phenotype. It will be important to elucidate in future studies the precise molecular mechanism how antibodies against these receptors induced a similar phenotypes. Nevertheless, anti-M2R antibodies have also been implicated in two additional human diseases, sinus node dysfunction [[37]] and atrial fibrillation [[38, 39]]. It is thus possible that mechanisms similar to what we observe in our mouse model are operative in humans.

The appearance of antibodies against M2R in patients with myocarditis or with an early stage of DCM could be a warning sign of a risk for disease progression. In this context, myocarditis and DCM with anti-M2R antibodies can be considered to be an autoimmune disease and treatment to modulate such antibodies would be beneficial for those patients in order to prevent the progression of the disease to a fatal stage. Furthermore, since it is nearly impossible to monitor the presymptomatic phase of myocarditis at molecular and cellular levels or disease progression even after the onset of DCM in humans, our model mice provide novel experimental tools with which to study the pathophysiology and molecular mechanisms of DCM.

Materials and Methods

Mice

M2R–/– mice on a C57BL/6 background were described previously [[40]] and were maintained by mating female M2R+/− and male M2R–/– mice. C57BL/6-Rag2–/– mice were obtained from Taconic. All animals were maintained at Taconic or in our animal facility under specific pathogen free conditions and experiments were performed in accordance with our institutional guidelines. Tail DNA was extracted by phenol and chloroform from proteinase K-digested samples. PCR amplifications for genotyping were performed in a single reaction mixture using the following primers: MF26 (5′-GGTTGGGTGCATTGGTTAGT-3′), PR1 (5′-CAGACTGCCTTGGGAAAAGC-3′), and MR22 (5′-GTGTTCAGTAGTCAAGTGGC-3′).

Construction of recombinant mouse M2R and β1AR proteins

cDNAs encoding the 2nd-ECL of mouse M2R (GenBank AF264049; corresponding amino acid sequence: VRTVEDGECYIQFFSNAAVT FGTAI) and the 3rd-ICL of mouse M2R (GenBank AF264049; corresponding amino acid sequence: RASKSRIKKEKKEPVANQDPVSPSLVQGRIVKPNNNNMPGGDGGLEHNKIQNGKALRDGGTENCVQGEEKESSNDSTSVSAVASNMRDDEITQDENTVSTSLGHSKDDNSRQTCIKIVTKTQKGDACTPTSTTVELVGSSGQNGDEKQNIVARKIVKMTKQPAKKKPPPS) were inserted into the pGEX-4T-3 vector (GE Healthcare) to prepare GST fusion proteins. The recombinant fusion proteins were expressed in Escherichia coli JM109 and purified on a Glutathione-Sepharose column. A cDNA encoding the 2nd-ECL of mouse β1AR (GenBank L10084; corresponding amino acid sequence: HWWRAESDEARRCYNDPKCCDFVTNR) was inserted into the pEGFP-C1 vector (Clontech Laboratories). The cDNA encoding the enhanced green fluorescence protein (EGFP)-fused 2nd-ECL of the β1AR gene was subsequently inserted into the pQE-30 vector (QIAGEN) to add a His-tag to the EGFP-β1AR fusion protein. The fusion protein was expressed in E. coli JM109 and purified on a Ni2+ affinity column.

ELISA

Circulating anti-M2R-2nd-ECL IgG and anti-M2R-3rd-ICL IgG antibodies were measured by ELISA using mouse GST-fused M2R-2nd-ECL and GST-fused M2R-3rd-ICL proteins as coating antigens, respectively. Each serum sample was diluted 100- to 10,000-fold with PBS and examined in duplicate. To exclude anti-GST IgG antibodies from serum samples, sera were preincubated with 1 mg/mL recombinant GST protein for 30 min. Circulating anti-β1AR IgG and anti-cardiac myosin antibodies were also measured by ELISA using a mouse His-tagged EGFP-β1AR protein and purified porcine cardiac myosin (Sigma) as coating antigens. Each serum sample was diluted and examined as described above after a 1,000-fold dilution with PBS. Class-specific Ig concentrations were measured by Clonotyping System HRP (Southern Bioscience).

Adoptive transfer of splenocytes from immunized mice into Rag2–/– mice

M2R–/– and M2R+/− mice were immunized and boosted twice with a mixture of 10 μg of the recombinant M2-2nd-ECL protein fused with GST and ImmuneEasy (QIAGEN). After 8 week, the splenocytes of immunized mice were dispersed through a 100 μm nylon mesh to prepare single-cell suspensions and erythrocytes were lysed with an Ack lysis buffer solution (150 mM NH4Cl, 1 mM KHCO3, 0.1 mM Na-EDTA). Fifty million total splenocytes in 250 μL PBS were transferred to each Rag2–/– mouse by intravenous injection. Cellular compositions of donor splenocytes were similar with each other.

Histological analysis and immunohistochemistry

Mice were sacrificed and their hearts dissected out and immersed in PBS on ice. After depletion of blood and thrombi in the ventricles, hearts were weighed and fixed in 10% formalin in PBS for sectioning. Paraffin-embedded tissue sections were stained with H&E for examination by light microscopy. Tissues for immunohistochemistry were frozen immediately in OCT compound (Sakura Finetek). The sections (5 μm thick) were fixed by acetone and then treated with 0.6% hydrogen peroxide in PBS for 10 min to quench endogenous peroxidase activities. The sections were then washed with PBS and immersed in a dish containing a blocking buffer solution for 1 h. Sections were then incubated with primary antibodies diluted in the blocking buffer solution at room temperature for 1 h. The primary antibodies used here were monoclonal antibodies specific for CD4 (GK 1.5), CD8 (53-6.7), and B220 (RA3-6B2). Sections were rinsed three times in PBS for 2 min per wash. Sections were then covered with a secondary antibody specific for rat IgG diluted in the blocking buffer solution. After incubation for 1 h, the slides were washed three times with PBS and then developed with the VECTASTAIN Elite ABC kit and Vector DAB substrate (Vector Laboratories). After substrate development, the sections were washed in water and counterstained with hematoxylin. Slides were rinsed in water, and cover slips were mounted using Entellan New (Merck KGaA) [[41]].

Purification of polyclonal IgG

We transferred M2R-immunized M2R–/– splenocytes or M2R-immunized M2R+/– splenocytes into five Rag2–/– recipient mice each. Sera were then collected twice a week from 6 to 10 weeks after splenocyte transfer. IgG fractions were purified from the pooled sera using HiTrap Protein G HP (GE Healthcare Bio-Sciences) and concentrated using Amicon Ultra-15 100K filters (Millipore). Concentrations of IgG were 46.1 mg/mL for the M2R-immunized M2R–/– group and 25.3 mg/mL for the M2R-immunized M2R+/– group, respectively.

Immunoflourescence staining

IgG deposition was detected by direct immunofluorescence performed on frozen sections of heart from recipient mice, as previously described [[42]]. Briefly, 6 μm cryostat sections were incubated in cold acetone, then incubated with FITC-conjugated goat antimouse IgG (Abcam). Specimens were analyzed under a Zeiss fluorescence microscope. Frozen heart sections from untreated M2R–/– or wild-type mice were stained for indirect immunofluorescence as previously described [[43]]. The sections were fixed in acetone and blocked in 4% Block Ace (Yukijirushi) for 60 min. They were incubated with purified IgG as the primary antibody for 90 min at 37°C followed by Alexa Fluor 488-conjugated goat antimouse IgG (Molecular Probes) as the secondary antibody for 60 min. Specimens were observed under a Zeiss LSM 510 META confocal laser microscope.

Echocardiography

Transthoracic echocardiography examinations were performed as follows. The mice were lightly anesthetized with ketamine (100 mg/kg) and xylazine-HCl (4 mg/kg), shaved (chest only), and placed prone on a specially designed apparatus. Echocardiograms were obtained by an EnVisor C echocardiographic system (Royal Philips Electronics) equipped with a 12-MHz phased array transducer. M-mode tracings were recorded in the short-axis views through the anterior and posterior left ventricular wall at the level of the papillary muscles. The wall thickness and left ventricular internal dimensions, namely LVEDD and left ventricular end-systolic dimension (LVESD) were measured directly on the screen and %FS values were calculated according to the following formula: (%FS = 100 × (LVEDD - LVESD) / LVEDD).

Cardiomyocyte culture

Primary cultures of cardiomyocytes were prepared from the ventricles of 1-day-old Wistar rats (Japan CLEA) as described previously [[44]]. Cardiomyocytes were separately prepared by differential adhesiveness. Nonattached cells were collected and seeded at a density of 5 × 105 cells/cm2 as a cardiomyocyte-rich culture on gelatin-coated dishes.

Measurement of (Ca2+)i changes

The changes in (Ca2+)i were monitored using a fluorescent calcium indicator Fluo-4 AM (Molecular Probes, Eugene, USA), as described previously [[45]]. The (Ca2+)i changes in cultured cardiomyocytes on a gelatin-coated Cover Glass M-22 (Matsunami Glass) were recorded for 30 min after the addition of IgG to the medium. To ascertain whether anti-M2R IgG have a functional role in cardiomyocyte contraction through M2R, we added synthetic peptide derived from the 2nd-ECL of M2R synthesized by Operon Biotechnologies to anti-M2R IgG 30 min before measurement. Alternatively, we added the selective M2R antagonist, AF-DX 116 (Tocris Bioscience) [[46]] to the medium 10 min before IgG administration.

Action potential recording

The action potential from the endomyocardium of the left ventricle was recorded by standard microelectrodes as described previously [[47, 48]]. The resting membrane potential, the action potential overshoot, the action potential amplitude, and the APD at 20, 50, and 90% repolarization (APD20, APD50, and APD90) were measured at 1,000 ms pacing cycle length.

Statistical analysis

All data were expressed as means ± SEM. Differences between experimental groups were evaluated for statistical significance using the Student's t-test for unpaired data or for multiple groups using both the two-tailed Student's t-test and the one-way or two-way ANOVA followed by Tukey–Kramer multiple comparison test. The data were distributed normally in all cases. p values of less than 0.05 were considered to be statistically significant.

Acknowledgments

We thank Dr. A. Anzai for help in some experiments and M. Motouchi, N. Yumoto, and K. Takei for animal care. This work was in part supported by a Health Sciences Research Grant for Research on Specific Diseases from the Ministry of Health, Labour and Welfare, Japan, a National Grant-in-Aid for the Establishment of a High-Tech Research Center in a private University, a grant for the Promotion of the Advancement of Education and Research in Graduate Schools, and a Scientific Frontier Research Grant from the Ministry of Education, Culture, Sports, Science and Technology, Japan.

Conflict of interest

S.K. is a consultant for Medical and Biological Laboratories, Co. Ltd. The authors otherwise declare no financial conflict of interest.

Abbreviations
APD

action potential duration

(Ca2+)i

intracellular Ca2+ concentration

DCM

dilated cardiomyopathy

Dsg3

desmoglein 3

ECL

extracellular loop

EGFP

enhanced green fluorescence protein

FS

fractional shortening

GST

glutathione S-transferase

IA

immunoadsorption

ICL

intracellular loop

LVEDD

left ventricular end-diastolic diameter

LVESD

left ventricular end-systolic diameter

M2R

M2 muscarinic acetylcholine receptor

PV

pemphigus vulgaris ·β1AR: β1 adrenergic receptor

Ancillary