Sarcolemmal fragility secondary to the degradation of dystrophin in dilated cardiomyopathy, as estimated by electron microscopy

Abstract

A common gene deletion or mutation of delta-sarcoglycan (δ-SG) in dystrophin-related proteins (DRPs) is identified in both TO-2 strain hamsters and human families with dilated cardiomyopathy. We have succeeded in the long-lasting in vivo supplementation of a normal δ-SG gene by recombinant adeno-associated virus vector, restoration of the morphological and functional degeneration, and improvement in the prognosis of the TO-2 hamster. To evaluate the integrity of the sarcolemma (SL) and the subsequent change of organelles in cardiomyocytes of the TO-2 strain hamster, we examined electron microscopy (EM) images focusing on the sarcolemmal stability at the end stage of heart failure. Two types of sarcolemmal degradation were detected: the widened and locally thickened SL, and blurred and discontinuous SL. Bizarrely formed mitochondria of varying sizes were also observed. Immuno-EM revealed clear expression of dystrophin in the SL and intense expression at the costamere as well as at the T-tubules in the control F1B strain hearts, but a patchy deposition of dystrophin was observed along the SL without the transgene of δ-SG. In contrast to the previous reports that dystrophin’s integrity was intact, the present results suggest that the gene deletion of δ-SG and the loss of δ-SG protein in the SL cardioselectively cause the morphological and functional deterioration of dystrophin and the resultant instability of the SL. The sarcolemmal fragility may be similar to Duchenne-type progressive muscular dystrophy in skeletal muscle. In addition to the mechanical role, another aspect of DRPs for the intracellular signal transmission is also discussed.

The pathological process of dilated cardiomyopathy (DCM) to the advanced stage is of great importance for developing both the prevention and the treatment of heart failure. A cardiomyopathic hamster is a representative model of human hereditary cardiomyopathy (CM) (1). We have identified a gene defect of delta-sarcoglycan (δ-SG), a component of dystrophin-related proteins (DRPs) (Figure 1), in both the BIO-14.6 strain, which shows hypertrophic CM followed by DCM, and the TO-2 strain, which reveals DCM from the onset (1–3). This δ-SG gene defect has also been reported in four human families with DCM and sudden death (4). The gene abnormality and the corresponding transgene disruption in DRPs may commonly induce muscle degeneration, especially in the sarcolemma (SL), because DRPs link intracellular contractile machinery, F-actin, with the extracellular matrix, laminin alpha-2, via dystrophin. In addition to the major role of stabilizing the SL during the repeated heart beating (Figure 1) (5), DRPs are now presumed to have another function of transmitting a mechanical stimulus to an intracellular signal.

Figure 1).

A scheme of dystrophin-related proteins (DRPs). DRP is made of a protein complex consisting of alpha- and beta-dystroglycans (DG); alpha-, beta-, gamma- and delta-sarcoglycans (SG); sarcospan (SPN); and syntrophin (Syn). Note that the complex connects with the intracellular contractile machinery actin via dystrophin (Dys) and with the basement membrane at the extracellular matrix via laminin alpha-2 in addition to neuronal nitric oxide synthase (nNOS) and caveolin (Cav). This illustration was kindly presented by Dr M Yoshida, Department of Cell Biology, National Institute of Neuroscience, National Center of Neurology and Psychiatry, Japan. DB-1 Dystrobrevin-1

Although many pathohistological studies have been performed in cardiomyopathic hamster hearts (1,6–8), no report has addressed the SL except our results using immunohistological analyses (3,9,10). In the present study, to evaluate the integrity of the SL and the morphological change of organelles in cardiomyocytes of TO-2 strain hamsters, we examined the fine architecture at the end stage of heart failure with conventional electron microscopy (EM). Furthermore, we focused on the exact location of δ-SG and dystrophin using an immuno-EM.

MATERIALS AND METHODS

Hereditary cardiomyopathic hamsters

TO-2 strain hamsters with hereditary DCM were used as the DCM model at age 25 weeks, when clinical symptoms of advanced heart failure became evident (3,9,10). The age-matched F1B strain hamsters were used as controls. Both male strains were purchased from Bio Breeder (Fitchburg, USA) and kept in the Infection Research Laboratory under conditions following the Japanese Pharmacological Society’s Guidance for Animal Facility.

EM

The hamsters were deeply anesthetized with intraperitoneal injections of sodium pentobarbital (Dainippon Pharmaceutical Co, Osaka, Japan). The hearts were carefully removed and the apex was excised for EM, as described previously (11). Myocardial cubes of 1 mm were prefixed with 2.5% of glutalaldehyde in 0.1 M sodium cacodylate buffer (pH 7.4), postfixed with 1% osmium tetroxide and dehydrated in graded alcohol. The tissue was embedded in Epon mixture (Resolution Performance Products, USA), and ultra-thin sections (60 nm thick) were made with an ultramicrotome (Porter-Blum MT-II, Sorvall, USA). The sections were double-stained on copper grids with uranyl acetate for 10 min to 15 min and lead nitrate for 5 min at room temperature. Sections were photographed on an electron microscope (Hitachi H-7000, Tokyo, Japan) in a magnification range from ×5000 to ×25,000.

Immuno-EM

The hearts were excised and fixed with 2% paraformaldehyde in saline including 10 mM ethylenediamine tetraacetic acid disodium salt (pH 7.4) for 2 h at 4°C. They were then sliced (2 mm thickness) and fixed for an additional 4 h. The frozen specimens in OCT compound were prepared for the immunohistological assessment, as described previously (9,10). The samples were incubated overnight with primary monoclonal antibodies to dystrophin or with polyclonal antibodies to δ-SG at 4°C. After washing with Tris-buffered saline and incubating for 20 min at 37°C with the second antibody, the section was incubated with diaminobenzidine for 10 min at room temperature. Ultra-thin sections (60 nm thick) mounted on a copper grid and counterstained with uranyl acetate were examined with EM, as described earlier.

Antibodies

Site-directed polyclonal antibodies to δ-SG, of which an epitope was selected from a unique amino acid sequence deduced from the cloned cDNA, were raised in rabbits and purified by affinity chromatography, as described previously (2). Monoclonal antibodies to the rod domain of dystrophin were from Novocastra (Newcastle, United Kingdom). The second antibodies to the monoclonal and polyclonal primary antibodies from the mouse and rabbit, respectively, were labelled with horseradish peroxidase.

RESULTS AND DISCUSSION

EM

The ultrastructure of normal F1B hamster hearts at low-power magnification demonstrated tightly arranged cardiac muscle without noticeable fibrosis (Figure 2A). Mitochondria were homogeneous in size and neatly located between myofibrils or under the SL. In contrast, the TO-2 heart at the same age (25 weeks) was markedly dilated, and the myocardium showed interstitial fibrosis or calcification (12). Disarray and thinning of myofibrils were also noted (data not shown). The principle outcome of the EM studies was a multi-elemental pathology at subcellular levels. TO-2 samples showed that the myofibrils were unarranged, disrupted and degenerated. As described by Schaper et al (13) in human cases, the reduction of myofilaments was obvious. The interstitial space was widened and contained macrophages, fibroblasts, collagen fibres and cell debris in TO-2 hamsters. The I- and Z-bands showed disintegration. Bizarrely formed mitochondria were also observed and the size varied with the vacuolar degeneration and the proliferation. The cristae were loosely arranged or disrupted (Figure 2B). A wide variety of intra- and intercellular findings among myocardial cells were apparent (data not shown), similar to the results of previous reports (7,8).

Figure 2).

Electron microscopy of a hamster heart of normal F1B (A,C) or hereditary DCM TO-2 strain (B,D) at the low-power magnification (original magnification ×5000; A,B) and the high-power magnification (original magnification ×25,000; C,D). A In the F1B heart, myofibrils were tightly arranged in parallel, mitochondria were homogeneous in size and located between myofibrils, and fibrotic change was minimal. B In a TO-2 strain sample, the myofibrils were disarranged, numerous bizarrely formed mitochondria were observed and the mitochondrial size was variable (closed arrows). C In the F1B heart, the sarcolemma was sharply demarcated (closed arrowheads) and the electron density of costameres was increased (closed arrows). The structure of myofilaments and cristae was well preserved. D In a TO-2 strain sample, the sarcolemma was widened and locally thickened (closed arrowheads) or disrupted without preserving the structure of the lipid bilayer (open arrowheads). Note that the increment of electron density was lost at costameres. Bar length indicates 0.2 μm (A,B) or 1 μm (C,D)

We noticed the alteration of the SL at high magnification. EM of the control heart demonstrated sharply demarcated SL and tightly arranged myofibrils in parallel (Figure 2C). We identified two types of sarcolemmal degradation in TO-2 hearts (Figure 2D): widened and locally thickened SL, and a blurred and discontinuous lipid bilayer. The SL was poorly demarcated in some portions and discontinuous in other parts, where the basal lamina was not clearly observed. Myofibrils demonstrated frying of myofilaments with blurring of the I- and Z-band regions. As described by Jasmin and Eu (8), contraction bands adjacent to an intercalated disk were often encountered. Such severe involvement of peripheral fibrils is characterized by deep infoldings of the SL that enclose aberrant mitochondria to be eventually phagocytosed.

Immuno-EM

Immunostaining of δ-SG revealed homogenous deposits along the SL in F1B hearts (Figure 3A), whereas it was not detected in TO-2 hearts at all (Figure 3B). These results were in agreement with our report that the δ-SG gene showed a large deletion including the promoter region, exon 1 and intron 1 (2), and caused both genetic and morphological findings, as published previously (2,3,9,10). Figure 4A shows that immunostaining of dystrophin revealed deposits along the SL in the normal hamster, but the deposits were more densely detected just above Z-discs, at the costamere. We could not detect immunoproduct on Z-discs, as was reported by Kaprielian et al (14).

Figure 3).

Immuno-electron microscopy of delta-sarcoglycan (δ-SG) in F1B (A) and TO-2 (B) strain hearts at high-power magnification (original magnification ×25,000). Immunostaining of δ-SG revealed homogenous deposits along the sarcolemma in the normal heart (A, arrowheads), while it was not detected in the TO-2 heart (B, arrowheads). Bar length indicates 1 μm

Figure 4).

Immuno-electron microscopy of dystrophin in F1B (A) and TO-2 (B) hamster hearts at high-power magnification (original magnification ×25,000). A In the normal heart, dystrophin revealed homogenous deposits along the sarcolemma (closed arrowheads) and the T-tubular system (open arrowheads). B In the TO-2 heart, immunode-posits of dystrophin were heterogeneously expressed (closed arrowheads). Bar length indicates 1 μm

There has been both agreement and controversy over the precise location of dystrophin at the subcellular level. Many studies using immuno-EM or confocal microscopy confirmed that cardiac muscle shows dystrophin expression not only at the SL but also at the T-tubules (14–16); these findings were not observed in skeletal muscle, which demonstrates selective expression on the SL alone (17). However, the expression on Z-discs did not reach an agreement on the precise location of dystrophin. Present results in hamsters (Figure 4A) and recent reports in animals and humans reveal no expression on Z-discs (14,16). Meng et al (18) revealed the selective labelling of Z-discs with dystrophin antibodies. Because T-tubules are located close to Z-discs, more exact and careful evaluation is required.

As was distinctly demonstrated in Figure 4B, dystrophin appeared to be heterogeneously stained along the SL in TO-2 hearts. The immunoproduct showed patchy deposition along the fragile SL. The specific location of dystrophin to T-tubules and costameres was not evident in the present study. The amount of dystrophin appeared to be preserved at 15 weeks of age (9) but seemed to be decreased in the failing stage at 25 weeks of age. Accordingly, the age-dependent reduction rate of dystrophin was more gradual than SG complex (9).

In addition to dystrophin, other sarcolemmal proteins including L-type calcium channels (19,20), sodium-calcium exchangers (21) and sodium-potassium ATPases (22) have been reported to be reduced in failing hearts. These findings suggest that biosynthesis declined and/or the degradation rate of these proteins was accelerated under the pathological conditions. We need to be cautious of whether the decreases of these sarcolemmal proteins are directly related to the cause of the myocardial dysfunction or result from it.

The function of dystrophin may not be restricted to mechanical support for preventing overexpansion of the SL in repetitive systole to stabilize the SL (5). Considering the connection between dystrophin and neuronal nitric oxide synthase or caveolin (Figure 1), it is conceivable that dystrophin together with DRPs may serve signal transmission, such as myocardial hypertrophy secondary to cell stretching, volume overload or pressure loading. The latest report that the gene mutation of intranuclear protein, lamin A/C, also accompanies DCM, severe arrhythmias as well as mental retardation (23) would strongly suggest another aspect of DRP, because DRP is indirectly linked to lamin A/C via dystrophin and actin (24). The leaky SL has already been reported in skeletal muscle in human patients with Duchenne-type muscular dystrophy and in mdx mice (25). Evans blue uptake, the marker of a leaky SL, was detected in cardiomyocytes in the cardiomyopathic hamster heart (10,26). We have already succeeded in the supplementation of a normal sequence δ-SG gene using long-lasting and harmless recombinant adeno-associated virus vector to the TO-2 strain hamster, and verified the restoration of both functional and morphological degeneration (9,10,12), amelioration of the permeability of the SL (10) and improvement in the animal’s prognosis (10). These data provide evidence that the gene therapy is actually beneficial for fundamental treatment of advanced heart failure, although more detailed analysis is required to establish our scheme.