III. Three-dimensional electron microscopy of mammalian cardiac sarcoplasmic reticulum at 80 kV

Evolution of the cardiac dyad

Cardiac dyads are the site of communication between the sarcoplasmic reticulum (SR) and infoldings of the sarcolemma called transverse-tubules (TT). During heart excitation–contraction coupling, Ca²⁺-influx through L-type Ca²⁺ channels in the TT is amplified by release of Ca²⁺-from the SR via type 2 ryanodine receptors, activating the contractile apparatus. Key proteins involved in cardiac dyad function are bridging integrator 1 (BIN1), junctophilin 2 and caveolin 3. The work presented here aims to reconstruct the evolutionary history of the cardiac dyad, by surveying the scientific literature for ultrastructural evidence of these junctions across all animal taxa; phylogenetically reconstructing the evolutionary history of BIN1; and by comparing peptide motifs involved in TT formation by this protein across metazoans. Key findings are that cardiac dyads have been identified in mammals, arthropods and molluscs, but not in other animals. Vertebrate BIN1 does not group with members of this protein family from other taxa, suggesting that invertebrate BINs are paralogues rather orthologues of this gene. Comparisons of BIN1 peptide sequences of mammals with those of other vertebrates reveals novel features that might contribute to TT and dyad formation. The analyses presented here suggest that the cardiac dyad evolved independently several times during metazoan evolution: an unexpected observation given the diversity of heart structure and function between different animal taxa.

This article is part of the theme issue ‘The cardiomyocyte: new revelations on the interplay between architecture and function in growth, health, and disease’.

Direct imaging of dehydrogenase activity within living cells using enzyme-dependent fluorescence recovery after photobleaching (ED-FRAP)

Reduced nicotine adenine dinucleotide (NADH) is a key metabolite involved in cellular energy conversion and many redox reactions. We describe the use of confocal microscopy in conjunction with enzyme-dependent fluorescence recovery after photobleaching (ED-FRAP) of NADH as a topological assay of NADH generation capacity within living cardiac myocytes. Quantitative validation of this approach was performed using a dehydrogenase system, in vitro. In intact cells the NADH ED-FRAP was sensitive to temperature (Q(10) of 2.5) and to dehydrogenase activation by dichloroacetate or cAMP (twofold increase for each). In addition, NADH ED-FRAP was correlated with flavin adenine dinucleotide (FAD(+)) fluorescence. These data, coupled with the cellular patterns of NADH ED-FRAP changes with dehydrogenase stimulation, suggest that NADH ED-FRAP is localized to the mitochondria. These results suggest that ED-FRAP enables measurement of regional dynamics of mitochondrial NADH production in intact cells, thus providing information regarding region-specific intracellular redox reactions and energy metabolism.

Quick-freezing of cultured cardiac cells in situ with special attention to the mitochondrial ultrastructure

A new method has been developed which allows quick-freezing in situ of primary, cardiac cell cultures grown to confluence on gas-permeable membranes (Petriperm dishes). Small pieces of the growth substratum, with rhythmically beating myocardial cells, were slam-frozen, without cryoprotectants, against the surface of a helium-cooled copper block at approximately 16 K. The quality of the cellular cryopreservation, as judged by ultrastructural criteria, was studied in freeze-substituted specimens processed for transmission electron microscopy. The ultrastructure of cryofixed cardiac cells was compared with that of unfrozen, chemically fixed samples. The severity of cryodistortions increased progressively with increasing distance from the point of first impact. Of particular interest were the dramatic alterations of the mitochondrial ultrastructure. The concept that the reticular and the outer mitochondrial membranes are intimately and strongly associated was clearly demonstrated. Optimally frozen material revealed cryopreserved ultrastructure of high quality. The method described appears to offer an ideal model system for correlating the information gained by phase-contrast microscopy of living cell cultures with the ultrastructure of the same samples fixed in situ by chemical or physical techniques. Cryofixation would be particularly useful for studying dynamic cellular processes associated with physiological and pathophysiological conditions, e.g. metabolic inhibition, anoxia and substrate deprivation.

High-resolution scanning electron microscopic studies on the three-dimensional structure of the transverse-axial tubular system, sarcoplasmic reticulum and intercalated disc of the rat myocardium

The three-dimensional structure of the transverse-axial tubular system, sarcoplasmic reticulum (SR), and intercalated disc of the rat left ventricle was examined by high-resolution scanning electron microscopy after removal of the cytoplasmic matrices by the osmium-DMSO-osmium procedure. In the intermyofibrillar space, the transverse tubules (T-tubules) are accompanied by longitudinally oriented axial tubules and together form a transverse-axial system. The junctional SR is usually small but occasionally medium or large in size and couples with the T- or with the axial tubules. On the surface of the junctional SR facing the T- or the axial tubule, tiny junctional processes are seen. One or two sarcotubules, the so-called Z-tubules, frequently run parallel to the T-tubule. The sarcotubules derived from the junctional SR or from the Z-tubule run longitudinally or obliquely and form polygonal meshes around the myofibrils. On the surface of the SR at the H-band level, small fenestrations of 12-40 nm in diameter, and tiny hollows 8-20 nm in diameter are seen. Bulbous swellings of the SR, the corbular SR, are preferentially seen near the Z-band. The large and flat SR, known as the cisternal SR, intercalates among the SR meshes. In the subsarcolemmal space, the sarcotubules form a multilayered network (peripheral SR). The cisternal SR is frequently intercalated in these meshes and closely associated with the inner surface of the sarcolemma. The intercalated disc appears as a prominently undulated membrane demarcating the border between two adjacent heart muscle cells, and occasionally small projections 60-90 nm in diameter and 200-600 nm in length display on its surface.

The atrial myocardial cells of mouse heart: a structural and stereological study

Structural and stereological studies of mouse atrial myocardial cells, carried out in the same fashion as our previous investigations on mouse ventricle, demonstrate an extremely well-developed sarcoplasmic reticulum (SR) in atrial cells. The volume fraction (Vv) of the SR exceeds 12% in mouse atrial cells; perimyofibrillar network SR constitutes the major portion. We have confirmed the findings of Bossen et al. (1981, Tissue Cell 13, 71-77) of a difference between atria in terms of coupling density, the right atrium having a significantly lower incidence of interior junctional SR than the left. The SR of mouse atrium comprises a rich variety of specialized segments, including the IJSR, peripheral junctional SR, corbular SR, cisternal SR (including regions similar to fenestrated collars of striated skeletal muscle SR), as well as a peculiar form of extended junctional SR (EJSR). Although less frequent in occurrence than corbular SR, the EJSR seems closely related, since it occurs in multiple clusters at or near the Z-line regions, contains internal granular densities, and bears surface-connected structures resembling junctional processes. Seen in thin sections, mouse atrial EJSR elements are more complex than corbular SR, being larger in diameter and frequently circular in profile. Thick-section and serial-section analyses reveal that bodies of EJSR are in fact hollow spheroids. The transverse-axial tubular system of mouse atrium is rather poorly developed in comparison to its ventricular counterpart. The Golgi apparatus and associated specific atrial granules are prominent cell components. "Focal ellipsoidal deposits" (FEDs) previously described by Page and co-workers (1986, Amer. J. Physiol.) are consistently located adjacent to the Golgi region, but immunocytochemical staining for two different segments of atrial natriuretic peptide reveals no specific reaction in FEDs, whereas the SAGs are densely labeled for both antibodies.

Morphological study of sarcoplasmic reticulum in the atrioventricular node and bundle cells in guinea pig hearts

The osmium-ferrocyanide method for staining of the sarcoplasmic reticulum (SR) was used for a morphological investigation of the various components of the SR in the atrioventricular node and bundle (AVNB) cells of guinea pig hearts. On the basis of light microscopic observations, the AVNB tissue in guinea pig hearts can be divided into five regions: atrionodal junction, midnode, proximal bundle, distal bundle, and bundle branches. Electron microscopic observations revealed two types of junctional SR (j-SR) saccules in the cells from all the regions of AVNB tissue. One is similar to that seen in the working cardiac cells, i.e., flattened saccules with junctional granules. The second type is dilated and contains electron-dense granular material throughout its lumen. The flattened type is seen more often than the dilated type in atrionodal junctional cells and midnode cells, whereas the dilated type occurs more often in distal bundle cells and bundle branch cells. In most cells from the atrionodal junction and midnode regions, the j-SR saccules are apposed more often to sarcolemmal areas associated with nonspecialized regions of intercellular junctions than to other sarcolemmal areas. This distribution was not found in the distal bundle and bundle branch cells. Free SR tubules around the myofilament bundles are poorly developed in the midnode cells, generally in accord with the extent of development of myofibrils. Z-tubules are found in cells from all regions but are poorly developed in midnode cells. Corbular SR vesicles are found in cells from all the regions of AVNB tissues but are rare in midnode cells. Thus, each of the regions in the AVNB tissue has a different, characteristic distribution of SR components. Because of their possible relationship to the regulation of the intracellular concentrations of calcium, these differences in SR morphology may contribute to the diverse physiological properties of the different regions of the AV node and bundle.

The application of various electron microscopic techniques for ultrastructural characterization of the human papillary heart muscle cell in biopsy material

Various electron microscopical techniques have been applied to biopsy material obtained from patients suffering from mitral stenosis in order to characterize the subcellular organization of the hypertrophied papillary muscle. Small pieces of the same sample were processed for correlative transmission - (TEM) and scanning - (SEM) electron microscopical studies. TEM was carried out on conventionally fixed tissue with or without en bloc staining with a Cu-Pb citrate solution, and on freeze fracture replicas, while cryofractured material was studied by SEM. Stereo electron micrographs of the Cu-Pb impregnated tissue and of the cryofractured material were especially useful for studying the spatial distribution and relationship between various cell organelles. The myofilaments of the hypertrophied cells were arranged in a normal hexagonal pattern. Regions with irregular orientation of the myofibrils were occasionally seen. Accumulations of interfilamentous glycogen particles adjacent to the Z-bands were characteristic patterns of the contracted muscle cells. The extensive nexuses frequently observed in the subsarcolemmal regions may reflect functional alterations of the intercommunication between hypertrophied cells. The T-tubules were relatively few and irregularly distributed, and the complexity of the sarcotubular system (SR) revealed regional variations. Excellent visualization of the interior couplings between the SR and the T-tubules was achieved by studying thick sections of Cu-Pb impregnated tissue in the TEM. The dense staining of the various intracellular membranes when compared with the almost unstained external membranes including the free cell surface, intercalated disc and T-system, strongly indicates differences in chemical and functional properties of the two membrane systems. En bloc staining resulted also in contrasted glycogen as well as components of the nucleolus and the heterochromatin. The biochemical basis for the selective staining remains obscure; it may be a result of binding of heavy metal ions to carboxyl groups of specific proteins, and/or it may represent deposits of lead phosphate.

The sarcoplasmic reticulum of mouse heart: its divisions, configurations, and distribution

The sarcoplasmic reticulum (SR) is a prominent, highly ramified component of mouse myocardial cells. The use of ferrocyanide-reduced osmium tetroxide (OsFeCN) as a postfixative solution facilitates appreciation of both its extent and three-dimensional architecture. We have found that the individual volume fractions (Vv) of myofibrils, mitochondria, and SR are similar in cells of the right and left ventricular walls. Vv(total SR) is approximately 7%, a value considerably larger than previously reported. We attribute this disparity in large part to the recognition factor which comes into play with OsFeCN-treated tissue. Previous observations pertaining to the stereology of myocardial SR have likely substantially underestimated both volume fraction and surface density of this membrane system, since none to this point has utilized specific staining such as that conferred by the OsFeCN regimen. Our stereological measurements of different depths of the ventricular cell indicate that although considerable differences are found between SR configuration at peripheral and deep cell levels, no significant difference exists between the volume fractions of either the total SR or its individual constituents. Two different stereologic regimens gave close agreement on volume fractions of the various SR segments; the majority (approximately 92%) of the total SR is network SR, whereas the remainder is composed of the various categories of junctional SR (peripheral, apposed to the surface sarcolemma; interior, complexed with the transverse-axial tubular system; corbular, existing free of sarcolemmal contact). In the adult mouse, interior junctional SR greatly preponderates the other types of junctional SR; corbular SR is qualitively assessed to be a far more common component of atrial cells than of ventricular cardiomyocytes.

Ultrastructure and architecture of the sarcoplasmic reticulum in frog sino-atrial fibres: a comparative study with various preparatory procedures

Various preparatory procedures were tested to preserve the ultrastructure of the sarcoplasmic reticulum (SR) by the best possible method within frog sino-atrial muscle fibres. These procedures were: conventional aldehyde fixation with or without tannic acid, cryofracture, metallic impregnation and quick-freezing followed by freeze-substitution. Our results illustrated that, when optimally preserved, the SR architecture and ultrastructure of frog sino-atrial fibres were not fundamentally different from those described in many other vertebrate muscle fibres, particularly cardiac fibres. The three-dimensional arrangement of the SR and the structure of its main compartments were situated in a precise fashion: the peripheral SR, located close to the plasma membrane, was made of a tight network of tubules and showed typical couplings; the juxtafibrillar SR was made of a loose network of tubules, small cisternae and some tubules near Z-lines; the intermediary SR, associated with the mitochondria, was made of tubules and fenestrated cisternae. Contacts between SR and mitochondrial membranes were also studied; cryofractures revealed no special intramembrane particles at this level. Collapsed portions of the SR were found after quick-freezing. Because of its relative importance and its three-dimensional arrangement, the SR of frog sino-atrial fibres may have comparable functional significance to the SR of other cardiac muscle fibres.

Excitation-contraction coupling in rested-state contractions of guinea-pig ventricular myocardium

Different types of rested-state contractions were examined under the influence of various inotropic agents. In magnesium-free solution, in low sodium (40 mmol/l) solution or in the presence of dihydroouabain, an "early" rested-state contraction developed without delay after stimulation. A distinctive "late" rested-state contraction was observed under the influence of noradrenaline. It is characterized by a latent period of about 100 ms between stimulation and onset of contraction. This latency was not reduced by increasing the catecholamine concentration, despite a concentration-dependent increase in the height of the "late" rested-state contraction. The late rested-state contraction under the influence of noradrenaline was suppressed by the slow inward current inhibitor nifedipine whether or not the nifedipine-dependent shortening of the action potential duration was prevented by caesium. When the slow inward current was not inhibited, the prolongation of the action potential duration by caesium resulted in an increase of the late rested-state contraction because of a prolongation of the time to peak force. High concentrations of dihydroouabain led to the appearance of an early contraction component without appreciably influencing the noradrenaline-dependent late component. From this it was deduced that the activator calcium for the late rested-state contraction was not stored intracellularly during rest prior to stimulation and, consequently, could not have been released by inflowing calcium. Instead, it is proposed that the activator calcium for the late rested-state contraction entered the sites of the sarcoplasmic reticulum and subsequently released from its release sites as long as the cell was depolarized. The "early" rested-state contractions in Mg2+-free solution, in low sodium solution or in the presence of dihydroouabain were not influenced in their contraction velocity by high concentrations of nifedipine which fully inhibited the late rested-state contractions. Nifedipine caused only a slight reduction in peak force due to a shortening of the time to peak force as a result of a shortening in action potential duration. This indicates that the activator calcium for the "early" rested-state contractions had accumulated in the sarcoplasmic reticulum during rest prior to stimulation and that it was released immediately by depolarization without a participation of the slow inward current.