The transverse-axial tubular system (tats) of mouse myocardium Its morphology in the developing and adult animal.код для вставкиСкачать
THE AMERICAN JOURNAL OF ANATOMY 170:143-162 (1984) The Transverse-Axial Tubular System (TATS) of Mouse Myocardium: Its Morphology in the Developing and Adult Animal M.S. FORBES, L.A. HAWKEY, AND N. SPERELAKIS Department of Physiology, university of Virginia School of Medicine, Charlottesville, Virginia 22908 ABSTRACT Invaginations of the sarcolemma that generate the transverseaxial tubular system (TATS) of the ventricular myocardial cells have begun to develop in the mouse by the time of birth. The formation of the TATS appears to be derived from the repetitive generation of caveolae, which forms “beaded tubules”. Beaded tubules are retained in the adult, in which they frequently present a spiraled topography. Development of the TATS progresses so rapidly that complex systems are already present in the cardiac muscle cells of young mice; by 10-14 days of age, the ultrastructure is essentially identical to that of the adult. The mouse myocardial TATS is composed of anastomosed elements that are directed transversely and axially (longitudinally).Many tubules have an oblique orientation, however, and most elements of the TATS are highly pleiomorphic. In this respect the TATS of the mouse heart is relatively primitive in appearance in comparison with the more ordered TATS latticeworks typical of the ventricular cells of other mammals. Stereological analysis of the mouse TATS indicates that the volume fraction (VV)and surface density (SV) are considerably greater than previously reported (3.24%and 0.5028 pm-’, respectively). The most complex ramifications of the TATS are embodied in the subsarcolemmal caveolar system and the deeper tubulovesicular “labyrinths”, both of which can be found in early postnatal and adult ventricular cells. In atrial cells, TATS development is initiated several days later than in the ventricular cells. The TATS of adult atrial myocardial cells is less prominent than the ventricular TATS and consists largely of axial elements; the incidence of the TATS, furthermore, is more pronounced in the left than in the right atrium. The mouse heart is often used to present typical examples of myocardial cell ultrastructure. Nevertheless, pronounced differences exist among the cardiac muscle cells of different species. In particular, the transverse-axial tubular system (TATS), a collection of membrane-bound cavities that pervade all levels of mammalian myocardial cells, is particularly elaborate in mouse heart. In addition, the mouse TATS presents a phylogenetically primitive appearance in terms of its components, compared t o the more uniform, “finished” morphology displayed by myocardial TATSs of species such as guinea pig, dog, and monkey (Sperelakis and Rubio, 1971; Forbes and Sperelakis, 1983, 1984). 0 1984 ALAN R. LISS, INC. Several published micrographs of mouse ventricle, prepared by high-voltage electron microscopy (Sommer and Waugh, 1976; Sommer and Johnson, 1979; Yamada and Ishikawa, 1981), indicate a predominantly transverse orientation for the elements of the TATS, thus underscoring the concept of the “T” (transverse) tubular system long ascribed to cardiac and skeletal muscle cells alike. A significant contribution of longituN. Sperelakis is now at the Department of Physiology, College of Medicine, University of Cincinnati, Cincinnati, OH 45267. Address reprint requests to Michael S. Forbes, Ph.D., Department of Physiology, Electron Microscopy Laboratory, University of Virginia School of Medicine,Jordan Medical Education Building, Charlottesville, VA 22908. Received November 28, 1983. Accepted February 16,1984. 144 M.S. FORBES, L.A. HAWKEY, AND N. SPERELAKIS dinally aligned (axial) tubules to the TATS is found in guinea pig ventricular myocytes (Rubio and Sperelakis, 1971; Sperelakis and Rubio, 1971). In addition, a substantial axial component of the TATS has been reported for mouse ventricular cells (Forbes and Sperelakis, 1977, 1980a, 1983). The present communication is addressed specifically to the description of TATS anatomy in the various regions of normal mouse heart, including the relationship of the components of the TATS to the surface sarcolemma and to the surface-connected caveolae. In addition, the development of the TATS has been examined. We have found that the TATS is considerably more complex, its volume and surface area contributions far greater, and its clevelopment far earlier than heretofore thought. MATERIALS AND METHODS Electron microscopy Young (newborn to 17-day-old) and adult mice of the C57BL and ICR strains were anesthetized with intraperitoneal injections of pentobaxbital, thoracotomized, and wholebody perfused with glutaraldehyde fixative (3%glutaraldehyde made up in a solution of 3% dextran [79,000 M.W.] and 3% dextrose, with or without 50 mM CaC12 added, adjusted to pH 7.6). To demonstrate the extracellular spaces of the heart, most tissues were postfixed for 2 hr in a solution consisting of 2% osmium tetroxide and 0.8% potassium ferrocyanide in 0.1 M sodium cacodylate, pH 7.6 (Forbes et al., 1977). For conventional preparation of tissue for thin sectioning, osmication was carried out for 2 hr with 2% osmium tetroxide in 0.1 M sodium cacodylate (pH 7.2). Samples from adult animals were stained en bloc for 30 min in saturated aqueous uranyl acetate solution; most tissues taken from neonatal animals were prepared by a regimen which omitted this step, in view of its documented removal of pooled glycogen Wye and Fischman, 1970.). All tissues intended for thin and thick sectioning were dehydrated in ethanol solutions, passed through propylene oxide, and embedded in Epon 812 or PolyiBed 812 epoxy resin (Polysciences, Inc., Warrington, PA). Thin (50-80 nm) and thick (250 nm-2 pm) sections were cut with diamond knives on Sorvall MT-2 ultramicrotomes, and mounted on bare or Formvar-coated 150-mesh copper grids. The thin sections were stained sequentially with saturated uranyl acetate in 50% acetone (2 min) and 0.4% lead citrate (1min), whereas thick sections were left unstained (see Forbes and Sperelakis, 1980b, 1983, for details of the preparation of thick sections for “pseudo high-voltage electron microscopy”). For freeze-fracture replication, glutaraldehyde-fixed tissues were washed in dextrosedextran vehicle and gradually infiltrated with glycerol to a final 30% glycerol concentration. The glycerinated tissues were frozen in liquid Freon 22, fractured at - 100°C in a Balzers BAF 300 freeze-fracture apparatus, and allowed to etch for 2 min before coating with platinum (to a depth of 2.0-2.5 nm, a t a shadowing angle of 45”)and carbon. The replicas were freed from the underlying tissues by digestion in bleach solution, and then collected on bare copper grids. Specimens were examined and photographed in a transmission electron microscope (either Zeiss EM-9A, Philips EM 200, or JEOL 100s) operated a t a n accelerating voltage of 60 keV. Stereology For estimation of the contribution of the TATS to the ventricular myocardium of adult mouse heart, hearts from ten male ICR mice, 49 days of age, were prepared so that the system of extracellular spaces was selectively opacified, according to the procedure described above. A single thin section was examined from each of four randomly oriented tissue blocks from the right ventricular wall of each mouse heart; from each section ten micrographs were prepared a t a final magnification of x 14,500. On each section, a n area in the upper left-hand corner of each grid square that contained suitably traced myocardial cells was photographed. If the particular section did not provide a s a i cient number of micrographs, we collected data next from the upper right corners, then the lower right corners, and then the centers of the grid squares. The micrographs were printed together with a grid lattice overlay having a scale spacing of 0.65 pm. Pointcounting was used to establish the VV (volume fraction) of the TATS relative to the volume of the myocardial cells (including the nuclei). Another important stereological parameter measured was the quantity of surface area contributed by the TATS. The SV~ATS) (surface density) was determined by use of a semicircular test grid overlay (this, in addition to the use of randomly oriented sections, TATS OF MOUSE MYOCARDIUM eliminates the effects of anisotropy [Weibel, 1969; Bossen et al., 19781)printed onto micrographs having a final magnification of x 25,000 (6.Bossen et al., 1978). In this instance, photographically enlarged copies of many of the same micrographs utilized for VV determinations were prepared and analyzed. The SVwas determined from 200 micrographs, representing the collection of ten negatives from each of two blocks of the ventricular tissues of ten mice. The number of line intersections with TATS membrane profiles constituted the value Ii. The total test length (LT)that fell inside myocardial cells was measured with a Microplan I1 digital planimeter (Laboratory Computer Systems, Inc., Cambridge, MA) and converted into units of micrometers. The Sv value was therefore com uted in units of pm2 TATS membranelpmf: myocardial cell volume by the following formula (Weibel, 1969, 1973): sv= 2 . riLT. OBSERVATIONS Terminology The transverse-axial tubular system (TATS)consists of the interconnected threedimensional latticework of transverse (T) tubules and axial tubules, the latter directed more or less longitudinally (parallel to the long axis of the cell). The lumina of all components of the TATS are open to the extracellular fluid space, and so can be marked by their filling with opaque substances, such as the elemental osmium precipitate employed in this study. The outer cell membrane of myocardial cells-that portion of the “skin” of the cells that does not make deep divergence into the myoplasm-is termed the surface sarcolemma. The flask- or alveolarshaped vesicular invaginations that emanate from the surface sarcolemma, and whose lumina also admit extracellular fluid, are referred to as caueolae (“little caves”). The TATS in adult myocardium Ventricular cells Thin sections of mouse ventricular myocardium usually allow the visualization only of isolated profiles of the TATS. In processed tissue fragments properly exposed to a n extracellular tracer material, such profiles frequently appear in survey micrographs as opaque dots located at or near the Z lines of the nearest myofibrils; less often they assume the form of longitudinally or obliquely 145 directed tubular bbdies (Fig. 1). Substantially thicker slices of the same specimens confirm the existence of a transverse component of the TATS but also make evident numerous axial tubules that interconnect adjacent transverse tubules (Figs. 2, 3, 7). In longitudinal view, thick sections of the opacified TATS demonstrate that it forms a framework whose elements occupy the myoplasmic spaces among the rods or pleiomorphic columns of contractile filaments (the myofibrils or myofibrillar masses) (Figs. 3,7). The numerous “axial” tubules of the mouse TATS may run longitudinally or obliquely with respect to the cell axis (Figs. 2 , 3 , 7 ) and are frequently but not always thinner than their transversely oriented counterparts (Figs. 2-4, 7). The diameters of both axial and transverse tubules vary along their respective lengths. The TATS is characterized by distended pockets a t the points of T-axial intersection, which usually occur at the Z lines (Figs. 2-4, 7); similar dilatations also form adjacent to the A- and I-bands of the sarcomeres (Figs. 3,5). The pervasive nature of the TATS is obvious in transverse sections of the myocardial cell, and its pleiomorphism in this plane is evident as well (Figs. 5, 6). Values for the volume fraction and surface density of the TATS in cells of the right ventricular wall are given in Table 1. A common pattern of the TATS is one in which its most peripheral components are composed of transverse tubules, the system thence ramifying toward the center of the cell, a t various points and in different directions, to form a profuse anastomotic tangle (Figs. 2, 3, 7). Many of the transverse elements originate at surface sarcolemmal levels that lie nearly opposite the Z lines of the outermost myofibrils (Figs. 7-9). This is not a specific site of emanation, however, since T tubules may form from sarcolemmal regions that appose other segments of the sarcomere and then veer, along their approximately transverse course, into alignment with the nearest Z lines. The sarcoplasmic reticulum (SR) enwraps the myofibrils in a collection of tubules, saccules, bulbules, and cisternae. This SR complex is continuous across and around the elements of the TATS, and the junctional saccules form couplings of variable configuration with the TATS (Figs. 8-10). The TATS component of couplings may be severely flattened (Fig. 8), but in some instances remains relatively unchanged in diameter where it 146 M.S. FORBES, L.A. HAWKEY, AND N. SPERELAKIS TATS OF MOUSE MYOCARDIUM 147 Fig. 3. Stereoscopic pair of electron micrographs of a thick (ca. 2 pm) section made longitudinally through OsFeCN-traced TATS in myocardial cell of mouse right ventricular wall. The banding pattern of the myofibriis can be made out in these micrographs; although trans- verse TATS elements are aligned with the Z lines of the sarcomeres, a considerable portion of the tubules in this field are longitudinally oriented. Large dilatations are common along both the transverse and axial tubules. Stereo angle = 12". ~6,000. Fig. 1. Thin (70-nm) longitudinal section of mouse right ventricular myocardium treated with osmium-ferrocyanide (OsFeCN) postfixation to fill the system of extracellular spaces, which includes the transverse-axial tubular system (TATS). TATS elements appear in thin section for the most part as punctate profiles located near the Z-line levels of myofibrils, thus apparently representing transversely oriented tubules ("). The thick- ness of this section captures only a few longitudinally directed elements of the TATS (axial tubules: AxT). ~6,500. Fig. 2. Section, ca. 0.5 pm thick, cut successive to the thin section shown in Figure 1. The TATS is revealed as interconnected, pleiomorphic tubules, many of which run axially or obliquely with respect to the long axis of the myocardial cell. ~6,500. Fig. 4. Freeze-fracture replica. Plane of fracture passes longitudinally through this myocardial cell (left ventricular wall). Both transverse ('IT) and axial tubules (AxT: both oblique and truly longitudinal) appear in relief (only their E faces are represented in this field), along with their interconnections at several points. The irregular contours of the TATS are evident in the beaded segments of the axial tubules and in distensions in regions of T-axial anastomosis (arrows). Two caveolar evaginations (C) from one T tubule have been broken off at their necks. Other recognizable myocardialcell components include network SR (N-SR), junctional SR (J-SR), and mitochondria N).x68,OOO. Fig. 5. Thick (0.5 pm) section of transversely cut, TATS-traced myocardial cell in right ventricular wall. An extensive interconnection between transverse elements of the TATS can be traced for great distances. One connection between the TATS and the surface sat-colemma is seen (arrow); subsarcolemmal “beaded tubules” (BT) are also part of the TATS. The numerous dilatations (*) of the TATS also display corrugated surfaces. X25,500. Fig. 6. Stereo micrograph pair of 1-pm-thick transverse section of right ventricular wall. The profile of a nucleus RJ) is seen below some of the TATS elements, the rest of which are rather evenly distributed within the myoplasm. Stereo angle = 12”. ~ 5 , 0 0 0 . 150 M.S. FORBES, L.A. HAWKEY, AND N. SPERELAKIS TABLE 1. Stereological measurements ofthe transuerse-axial tubular system in the myocardial cells of right ventricular wall of4Bday-old male ICR mice Parameter Volume fractiona = Vv Surface densityb = Sv Units (%) G) Mean + S.E. 3.24 + 0.10 0.5028 + 0.0126 "Volumeof TATS relative to total myocardial cell volume. bSurfacearea of TATS membrane per unit volume of myocardial cell cytoplasm. comes into contact with the junctional SR (Figs. 9, 10). The openings of T tubules a t the level of the surface sarcolemma in many instances are only 49-55 nm in diameter (Figs. 8, 9). As a result, a regular array of recognizable T-tubule apertures is seldom present in replicas of the surface sarcolemma (Fig. 12). Large-bore transverse tubules (up to 300 nm diameter) occur in some cells, but their apparently correspondent ostia-although clearly distinguishable (Figs. 11, 13)-are of widely differing size and shape and do not always overlie the Z lines of the subsarcolemma1 myofibrils (Fig. 11). In the majority of mouse ventricular myocytes, the surface openings of T tubules are not distinguishable from caveolar apertures (Fig. 12). Such surface invaginations appear on the sarcolemma a t any point over the underlying contractile material (Fig. 12). Adjacent myocytes may display a striking disparity in the relative degree of caveolar population, as well as in number of definitive T-tubule apertures. caveolar openings range from 46-62 nm in diameter in freeze-fracture replicas, whereas the caveolar bodies themselves open up to 70-90 nm (measured in thin sections). Caveolae are found both singly and in alveolar or racemose clusters composed of three, four, or more individual caveolae that fuse a t a point below the cell surface and open to the extracellular space through a single neck (Fig. 14). The clusters of caveolae also lead into reticular formations whose elements may display either beaded profiles or evencontoured tubules (Fig. 14). Junctional SR saccules may form couplings with the subsarcolemmal caveolar complex (Fig. 15). When replicated, the caveolae display small numbers of intramembranous particles (IMPS)on their P faces as compared to the P face of the surface sarcolemma proper (Fig. 16). A distinct resemblance exists between the subsarcolemmal ramifications and the Zuby rinths, proliferated portions of the TATS that may incorporate vesicular, tubular, and lamellar elements. Labyrinths may be found at any depth within the ventricular myocardial cell (Forbes and Sperelakis, 1973, 1977, 1983). Thin and thick sections exemplify the sinuosity and corrugation of extensive segments of the TATS (Figs. 2, 3, 5-9). Stretches of small-diameter tubules having a beaded or spiral appearance are characteristic of both transverse and axial constituents (Fig. 5). The topography of these segments is best appreciated in freeze-fracture replicas (Figs. Fig. 7. Stereo pair of longitudinal section (ca. 2 pm) through right ventricular myocardial wall. At the lefthand side, a regular array of T tubules (TT)is seen to project into the cell from the surface sarcolemma. The system ramifies profusely deeper within the cell to form a latticework of interconnected tubules, many of which display beaded or saccular profiles. Stereo angle = 12". ~6,500. Fig. 8. Right ventricular wall cell. Thin section of OsFeCN-traced transverse tubule, whose course can be followed from the origin (arrow) at the surface sarcolemma (this ostium is approximately 49 nm in diameter), through a convoluted "beaded" segment, and into a flattened portion, only 30 nm in thickness, associated with junctional SR saccules (J-SR) to form an interior coupling which is juxtaposed to the Z line (Z) of the nearest myofibril. ~81,500. Fig. 9. Thin section of untraced myocardial cell in right ventricular myocardium, in which the transverse tubule opens to the extracellular fluid space via a 50-nrn ostium (arrow) and follows a course toward the 2-level of the most peripheral myofibril. The T tubule is 106 nm thick at its widest point, and it participates in a multipartite coupling that consists of three junctional SR (JSR) profiles and an additional T-tubule profile. X 107,000. Fig. 10. Coupling in right ventricular cell, composed of a n unflattened T tubule completely encircled by junctional SR. ~77,500. TATS OF MOUSE MYOCARDIUM 151 152 M.S. FORBES, L.A. HAWKEY, AND N. SPERELAKIS TATS OF MOUSE MYOCARDIUM 4,17, 18).In these “spiral tubules”, the contours of the E faces (that class of unit membrane leaflets closest to the lumen and extracellular fluid) appear as twisted rods having varying degrees of symmetry along their lengths (Figs. 4, 17). These membrane faces do not exhibit a noticeably different distribution of IMPS than is present on the E faces of tubules having greater diameters and more even contours (Fig. 17). However, the P faces of the spiral tubule segments, which resemble a n interconnected series of slanted caveolar profiles, bear considerably fewer IMPS than do the P faces of larger TATS elements (Fig. 18). Atrial cells It is established that mammalian atrial myocardial cells are thinner and shorterand hence considerably less voluminousthan the ventricular muscle cells (Fig. 19). In addition, the TATS content of atrial cells is distinctly lower than that observed in ventricular myocytes. Muscle cells of the right atrium are typified by considerable proliferation of caveolae in the subsarcolemmal cytoplasm, but few tubules bear extracellular fluid to levels beyond 1-2 pm of the surface sarcolemma (Fig. 19). In contrast, the TATS of the left atrium, although sparse in comparison with the ventricular myocardial cells, is in the adult animal substantially further developed than that of the right atrium and Fig. 11. Freeze-fracture replica of the P face of the surface sarcolemma of a myocardial cell in the left ventricular wall. The long axis of the cell is parallel with the horizontal axis of the micrograph. A number of large depressions, which represent the openings of transverse tubules W), are evident; not all of these are located directly over the Z-line levels of the underlying myofibrils (Z). The T-tubule ostia vary considerably in their shape. Smaller openings, most of which probably represent caveolae (C), are scattered over the sarcolemmal surface. ~ 2 5 , 5 0 0 . Fig. 12. E face of sarcolemma of another left ventricular myocardial cell. Numerous caveolae (C) appear in this field but are not limited in occurrence to any particular sarcolemmal regions (Z-line levels are indicated). The fracturing process has broken off many of the surface sarcolemmal invaginations P),and it is not clear which if any of these are the openings of T tubules. ~25,500. Fig. 13. T-tubule opening on the sarcolemmal E face of a cell adjacent to the one shown in Figure 12. Here the T tubule can be clearly distinguished from caveolae (C). ~ 6 5 , 0 0 0 . 153 consists largely of axial components (Fig. 20). In atrial cells, the beaded and often distended silhouettes of axial and transverse tubules display their evolution from caveolae (Fig. 20). Developing myocardial cells During the first several days of postpartum life of the mouse, the individual ventricular myocardial cells exhibit a highly variable degree of caveolar presence immediately beneath their surface sarcolemmata (Fig. 21). Through the ages of 0-3 days, numerous ventricular cells possess smooth sarcolemmal profiles; scattered among these cells, however, are cells whose borders are distinguished by pockets of caveolation (Figs. 21, 22). These caveolae may exist as individual profiles just beneath the sarcolemma (Fig. 21). Beaded tubules (Figs. 21-24) appear as well, both as individual simple tubules and as complex racemose arrays from which several caveolar projections radiate (Fig. 22). The specific levels of caveolation may face the Z bands of the outermost myofibrils but, like T tubules in the adult cell, do not observe a strict sarcomere-related origin. The connections between caveolar systems and the surface sarcolemma may be thin corridors (Fig. 23); but more often they are openings which have diameters similar to those of caveolae (Fig. 221, like the ostia of many T tubules of adult myocardium (Figs. €49). The interaction of the forming TATS and the elements of the SR is evident in the neonatal animal. Perimyofibrillar networks of SR are well developed in the newborn animal (Forbes and Sperelakis, 19831, and certain differentiated portions of this system-the junctional SR saccules-appear in close contact with the pleiomorphic profiles of groups of fused caveolae (Fig. 25). These caveolar clusters are destined to constitute the definitive TATS, including labyrinths (Figs. 25,26). Atrial cells in newborn mice are thin, but already contain specific atrial granules. They exhibit numerous caveolae a t their sarcolemma1 borders, but few deep incursions are present in muscle cells of either atrium at this stage. By 5 to 7 days of age, a few TATS elements are beginning to appear in cells of both atria. After 10-14 days of postnatal development, the muscle cells of the various regions of mouse heart display a fully formed sarcomeric banding pattern, associated in the ventricular regions with arrays of transverse 154 M.S. FORBES, L.A. HAWKEY, AND N. SPERELAKIS TATS OF MOUSE MYOCARDIUM tubules that appear at the Z-line levels. Axial tubules and labyrinths are encountered as well in ventricular cells, and the atrial cells exhibit small numbers of TATS profiles and couplings. Hence, the ultrastructure of the hearts at this age is qualitatively indistinguishable from that of the fully adult animal. DISCUSSION Formation of the TATS A considerable amount of evidence has implicated surface-connected caveolae as cell components which are vital t o the format‘ion of transverse and axial tubules, both in skeletal muscle (Ezerman and Ishikawa, 1967; Ishikawa, 1968; Kelly, 1971) and myocardial cells (Forbes and Sperelakis, 1973; Ishikawa and Yamada, 1975; Forbes and Sperelakis, 1976).The TATS of adult mouse myocardium is a particularly convincing example of this relationship, since its typical components consist primarily of beaded and spiraled seg- Fig. 14. Grazing thin section that passes along the sarcolemma of a myocardial cell in right ventricular wall. The subsarcolemmal myoplasm contains numerous racemose and tubular bodies, many of which are obviously derived from caveolae. Lucent, 50-nm-diameter circular profiles in two of the bodies (3indicate the caveolar neck, which opens to the extracellular fluid space at the level of the cell surface (cf. Figures 11-13). ~77,500. Fig. 15. OsFeCN-traced left ventricular wall. A saccule of junctional SR (J-SR) forms a coupling with a subsarcolemmal collection of fused caveolae. ~89,000. Fig. 16. Freeze-fracture replica of the left ventricular wall in which the plane of fracture has revealed the P face of the surface sarcolemma (SL) and then diverted downward into the myoplasm. A single caveolar opening is seen at the cell surface (arrow), and six caveolar profiles are evident just beneath the sarcolemma. The E face of one caveola (CE) displays a smooth surface, and only a few intramembranous particles appear on the caveolar P faces (Cp). ~94,000. Fig. 17. Freeze-fracture replica of axial tubule in left ventricular wall cell. The topography of this tubule includes a segment of smooth contour (at the left) and a narrow, spiralled length (at right). The E face of this tubule contains few intramembranous particles. ~82,500. Fig. 18. P face of an axial tubule in left ventricular wall. The larger, distended luminal portion bears numerous intramembranous particles (IMP) in a distribution similar to that seen in the sarcolemma (Fig. 16);but the spiraled segment, which resembles a series of fused caveolae, displays a considerably lower density of such particles. x 102,500. 155 ments. The low incidence of intramembranous particles within these segments is also typical of the structure of subsarcolemmal caveolae. The considerable ramification of the subsurface caveolae to form complex racemose structures and beaded tubules, which lie just beneath the sarcolemma, indicate that all caveola-based bodies might be considered as part of the TATS. The extensive labyrinthine proliferations of the TATS, which are continuous with transverse and axial tubules (Forbes and Sperelakis, 1973), occur at various depths within mouse ventricular myocardial cells and closely resemble, on a grand scale, the subsarcolemmal caveolar complexes. A further parallel is evident among all the various subcategories of the TATS; namely, the saccules of junctional SR form definitive couplings with transverse and axial tubules (Forbes and Sperelakis, 19771, labyrinths (Forbes and Sperelakis, 19731, and subsarcolemmal caveolar systems alike (Forbes and Sperelakis, 1982). It is possible that many of the caveolar complexes are not blind-ended structures, but may form a continuum with the deeper transverse tubules, axial tubules, and labyrinths. In addition to being the means of generation of T tubules in developing skeletal muscle fibers, caveolae are retained in the adult fibers in the form of short beaded segments, interposed between the surface sarcolemma and the smooth-contouredtransverse tubules (Zampighi et al., 1975). This might impart some flexibility to the T-tubule segments nearest the fiber surface, thus protecting their integrity during contraction (Forbes and Sperelakis, 198Ob). Similarly, the coiled TATS segments of mouse heart could act as reservoirs of membrane which expand and flex during the cycles of cell shortening and lengthening. The TATS of mouse heart exhibits a plasticity not commonly encountered in other mammalian myocardial cells. The retention of the archetypal form may allow caveolation to proceed along many vectors, thus creating exotic configurations such as those seen in labyrinths (Forbes and Sperelakis, 1973, 1977, 1983).The TATS participates as well in such unusual structures as “T-tubule desmosomes”, in which desmosomal plaques face one another across the lumen of a TATS element, thus forming internal, reflexive junctions (Myklebust and Jensen, 1978; Forbes and Sperelakis, 1980a). The SR is particu- 156 M.S. FORBES, L.A. HAWKEY, AND N. SPERELAKIS TATS OF MOUSE MYOCARDIUM larly plastic as well. Multipartite couplings are prominent constituents of mouse heart, usually manifested in thin sections as multiple alternations of J-SR saccules and TATS profiles (Forbes and Sperelakis, 19771, but more likely being the product of interdigitated S-and C-shaped membrane system elements (Forbes and Sperelakis, 1980a, 1983). The incidence of longitudinal and oblique (axial) tubules, and their contribution to the TATS, is high in the mouse heart. This is not in agreement with reports by Sommer and Waugh (1976) and Sommer and Johnson (1979). These authors suggested that‘ many profiles of presumed axial tubules result from the plane of section passing through folds of the surface sarcolemma or through limited longitudinal excursions of transverse tubules (Waugh and Sommer, 1975). This latter caveat has been eliminated from consideration in our study of mouse heart, because of the use of thick TATS-traced sections. Axial tubules are likely derived from caveolation which occurs a t right angles to previously formed transverse tubules. Much of the sarcolemma a t the cell ends is occupied by the insertions of myofibrils and by the intermembranous junctions (e.g., Forbes and Sperelakis, 1980a, 1984). Although some axial elements can be seen to emanate from the cell ends in some developing and adult myocardial cells, for the most part only restricted spaces are available, in the regions of cell-tocell apposition, to accommodate the openings of axial tubules. 157 atrium possess T tubules, whereas some of the right atrial myocytes lack a T system. This opinion is based on the finding that the VV of the TATS is similar in the two atria, but that values for the VV and SVof interior junctional SR-which forms couplings with the TATS-are twice a s great in the left atrium as in the right atrium. Our own examinations of regions of mouse atria whose extracellular spaces had been traced out by OsFeCN postfixation have found a definite disparity in the distribution of the TATS between the right and left atria, thus supporting the interpretation of Bossen and colleagues. Bossen et al. (1981)and Sommer (1982) consider the two muscle-cell populations of mouse right atrium to represent: 1) working myocytes (those possessing the TATS), and 2) vestigial or transitional cells that lack the TATS and are related more closely to the atrioventricular conducting system (AVCS) (concentrated in the right atrium). Many cardiac muscle cells of murine right atrium appear to be devoid of transverse and axial tubules; the absence of the TATS, however, is a doubtful criterion for the identification of AVCS cells (Osculati and Garibaldi, 1974; Osculati et al., 1978; Rybicka, 1977; Forbes and Sperelakis, 1984). Furthermore, the TATS is absent from muscle cells of both right and left atria of guinea pig heart (Sperelakis and Rubio, 1971). Without doubt, then, the TATS of mouse atrial myocytes is far less notable in incidence and complexity than its ventricular counterpart. Furthermore, the concept of a The TATS in atrial cells “T system”, composed of membranous incurBossen et al. (1981)have drawn the conclu- sions which have a predominantly transsion, based on morphometric determinations, verse orientation, is vitiated for atrial tissue, that most muscle cells of the mouse left both by the present investigation and by Forssmann and Girardier’s (1970) study of the rat. When present, the mouse atrial TATS nevertheless is composed of intricately anastomosed bodies composed of caveola-derived Fig. 19. Right atrial cells in oblique longitudinal sec- units, as is clearly the case for the ventricution; extracellular spaces have been traced out with OsFeCN postfixation. Only a few TATS elements (TT) lar TATS (Forbes and Sperelakis, 1973, 1976, 1977, 1982, 1983; Forbes et al., 1977; Ishiappear deep within the cells, although there is profuse caveolation (C) at the sarcolemmal border. x 10,500. kawa and Yamada, 1975; Yamada and Ishikawa, 1981). Pleiomorphism of the TATS Fig. 20. Thick longitudinal section (ca. 0.3 pm) similar t o that of the ventricular myocytes is through myocardial cells of mouse left atrium. Though also typical in atrial muscle cells. It appears, the TATS complement is less extensive than that of the ventricular cells (cf. Figures 2, 3, 7), it is better devel- therefore, that the same generative mechaoped than that of the right atrium (Fig. 19). It consists of nism for the TATS is in force in all regions of prominent axial (AxT) elements and some transverse the mouse heart, although the onset of TATS components (’IT),all of which display corrugated and development appears to occur later in the distended segments similar to those of the ventricular TATS. X23,500. atrial than in the ventricular cells. 158 M.S. FORBES, L.A. HAWKEY, AND N. SPERELAKIS Fig. 21. Survey view of a transverse section through OsFeCN-traced right ventricular wall from a 3-day-old mouse. Long stretches of surface sarcolemma are devoid of caveolae, but at some points caveolar chains (arrows) and racemose vesicular complexes (*) project from the sarcolemma into the myoplasm. x 16,500. TATS OF MOUSE MYOCARDIUM Volume and surface area contribution of the TATS There is a fourfold difference between the VV values derived for the TATS of mouse ventricular cells by Bossen et al. (1978) and by us (respective values of 0.81% and 3.24%). It seems likely that this disparity has resulted from the preparation of tissue samples in the two studies. The clear identification of TATS elements is of paramount importance; the properties of OsFeCN postfixation (Forbes et al., 1977)have made our regimen ideal for this purpose. The procedure utilized by Bossen and colleagues may be less effective for two reasons. First, the lanthanum hydroxide colloid used to trace the extracellular fluid space can be washed out during the postfixation and dehydration steps (Revel and Karnovsky, 1967). Compounding the difficulty of accurate identification of the finer ramifications of the TATS is the fact that Bossen et al. (1978) examined sections from both lanthanum-impregnated and conventionally prepared (lanthanum-free) blocks of tissue and used this admixture for the generation of stereological data, Given our VV determinations, it is not surprising that our value for the surface density (Sv)of mouse ventricular TATS is also considerably greater (0.5028 pm-’, as opposed to 0.2234 pm-l found by Bossen and colleagues), indicating a 2.25-fold larger surface area for the mouse TATS than previously reported. The large Vv and SVvalues of mouse heart TATS that we have reported are not particularly surprising in consideration of the qualitative view of this system which is readily obtained through the use of such techniques as selective staining and thick-section electron microscopy. The VVnATS)determined for certain other mammals (rat,cat, pig) achieves values of approximately 1-2% (Denoit and Coraboeuf, 1965; Pager, 1971; Page and McCallister, 1973; Mall et al., 1978; Hirakow et al., 1980; Singh et al., 1982, Gotoh, 1983). The guinea pig ventricular myocardial cell is notable in that its TATS is far more regular in distribution, contour, and average diameter of its elements than that of the mouse (Rubio and Sperelakis, 1971; Sperelakis and Rubio, 1971; Forbes and Sperelakis, 1983).In this way, the guinea pig TATS may represent one of the most advanced stages in development of the TATS, whereas the mouse TATS is among the most primitive of such systems. Values given to date for the VV(TATS)of guinea pig ventricular cells vary widely. Hir- 159 akow and Gotoh (1980) have obtained a value of 1.7-2.1% in the 2-week-old animal in guinea pig (the TATS begins its development during the embryonic period Forbes and Sperelakis, 1976). A study of the fully adult animal has determined the volume contribution to be 3.34% (Denoit and Coraboeuf, 1965), and other estimates range from 7.7% to 15% (Rubio and Sperelakis, 1971; Sperelakis and Rubio, 1971). The presence of the TATS, a n “excitatory network” (Hoyle, 1965) having substantial volume and surface area, is compatible with the properties of the mouse heart, most notably its high rate of contraction (up to 600 beatdmin; Geddes, 1970).In consideration, however, of the lower heart rate of the guinea pig (ca. 300 beats/ min), the comparative morphometric values thus far reported for mouse and guinea pig TATSs do not support a n absolute correspondence between relative TATS population and physiological properties (e.g., speed of shortening of myocardial cells). Other parameters, such as average diameter of TATS elements, volume and surface contributions of couplings, and over-all geometry of muscle cells and cell bundles, are likely to play a part in the physiological performance of different hearts. Function of the TATS We have demonstrated that the transverseaxial tubular system of mouse ventricle is a n intricate and pervasive entity, although it lacks the geometric precision and large luminal diameters of the latticeworks which predominate in myocardial cells of many other mammals. It has recently been argued that the role of all such membrane-limited extensions of the extracellular fluid space in heart is not obvious (Sommer and Waugh, 1976; Sommer and Johnson, 1979). Although the majority of mammalian hearts possess the TATS, the thinner, less voluminous cardiomyocytes of lower vertebrates generally lack it. The greater volume and diameter of mammalian heart cells may account for the need for the TATS to provide the optimum diffusion distance requisite to the even distribution of excitation and the simultaneous initiation of contraction at all depths of the myocardial cell. The voltage- and time-dependent Ca-Na slow channels of the myocardial cells are probably located in the TATS membrane as well as in the surface sarcolemma. Activation of these cation channels during the cardiac action potential would allow a n influx of 160 M.S. FORBES, L.A. HAWKEY, AND N. SPERELAKIS 161 TATS OF MOUSE MYOCARDIUM Ca+ ions down their electrochemical gradient into the myoplasm and would serve to couple contraction with excitation. Thus, the TATS would promote the influx of C a f + a t multiple levels within each myocardial cell, thereby shortening the diffusion distance and time required to activate the contractile proteins. The great ramification of the mammalian myocardial TATS may be necessary to provide ready access by all regions of each individual “working” cardiac muscle cell to electrically excitable membrane and extracellular fluid (e.g., Sperelakis et al., 1974).In this regard, a notable role of the TATS might be the maintenance of a constant surface:volume ratio, especially under conditions of induced hypertrophy (Page and McCallister, 1974; Anversa et al., 1979; Tomanek, 1979). This capability of the TATS to provide membrane augmentation in response to cellular enlargement may also be of use in the natural volume increase that accompanies maturation and aging of myocardial cells. + ACKNOWLEDGMENTS The work reported here was supported by grants awarded to Dr. Forbes by the American Heart Association (78-753) and the Na- Figs. 22-26. Thin sections of newborn mouse right ventricular wall, treated with OsFeCN postfixation to opacify the system of extracellular spaces and all channels open to it within the myocardial cells. Fig. 22. Several caveolar complexes, presumably forming transverse tubules, are filled with osmium precipitate. The connection of the rightmost complex to the surface sarcolemma can be discerned (arrow). X 35,000. Fig. 23. A collection of approximately six fused caveolar elements is open to the extracellular space through a narrow channel (arrow). ~ 8 1 , 5 0 0 . Fig. 24. A transverse tubule including beaded and spiraled profiles extends deep into the myocardial cell. ~81,500. Fig. 25. A complex caveolar body that partially surrounds a forming saccule of junctional SR (J-SR) whose lumen contains wisps of opaque material (junctional granules). In the gap between the J-SR and the forming T tubule, opaque projections (arrows)are visible; these appear to be junctional processes. x 104,500. Fig. 26. Complicated caveolar assemblage (*) that resembles a “labyrinth” (tubulovesicular proliferation of TATS elements) such as that seen in adult heart. Deeper within the myocardial cell, a smooth-surfaced T tubule (TI?) forms a coupling with a saccule of junctional SR (JSR). ~ 8 5 , 0 0 0 . tional Institutes of Health (HL-28329 and Research Career Development Award 5 KO4 HL-00550)and by NIH grant HL-18711to Dr. Sperelakis. Freeze-fracture replicas were prepared by Ms. Margaretta Allietta (Department of Pathology) and Ms. Bonnie Sheppard (Central Electron Microscope Facility of the University of Virginia School of Medicine). LITERATURE CITED Anversa, P., G. Olivetti, M. Melissari, and A.V. Loud 1979 Morphometric study of myocardial hypertrophy induced by abdominal aortic stenosis. Lab. Invest., 40r341-349. Bossen, E.H., J.R. Sommer, and R.A. Waugh 1978 Comparative stereology of the mouse and finch left ventricle. Tissue Cell, 1Or773-784. Bossen, E.H., J.R. Sommer, and R.A. Waugh 1981 Comparative stereology of mouse atria. Tissue Cell, 13t7177. Denoit, F., and E. Coraboeuf 1965 Etude comparative de I’ultrastructure du myocarde chez le rat et le cobaye. C. R. Soc. Biol., 159r2118-2121. Ezerman, E.G., and H. Ishikawa 1967 Differentiation of the sarcoplasmic reticulum and T system in developing chick skeletal muscle in uitro. J. Cell Biol., 35405420. Forbes, M.S., and N. Sperelakis 1973 A labyrinthine structure formed from a transverse tubule of mouse ventricular myocardium. J. Cell Biol., 562365-869. Forbes, M.S., and N. Sperelakis 1976 The presence of transverse and axial tubules in the ventricular myocardium of embryonic and neonatal guinea pigs. Cell Tissue Res., t66r83-90. Forbes, M.S., and N. Sperelakis 1977 Myocardial couplings: their structural variations in the mouse. J. Ultrastruct. Res., 58t50-65. Forbes, M.S., and N. Sperelakis 1980a Structures located a t the level of the Z bands in mouse ventricular myocardial cells. Tissue Cell, 12r467-489. Forbes, M.S., and N. Sperelakis 1980b Membrane systems in skeletal muscle of the lizard Anolis carolinensis.J. Ultrastruct. Res., 73:245-261. Forbes, M.S., and N. Sperelakis 1982 Bridging junctional processes in couplings of skeletal, cardiac, and smooth muscle. Muscle Nerve, 5;674-681. Forbes, M.S., and N. Sperelakis 1983 The membrane systems and cytoskeletal elements of mammalian myocardial cells. In: Cell and Muscle Motility, Vol. 3. R.M. Dowben and J.W. Shay, eds. Plenum, New York, pp. 89-155. Forbes, M.S., and N. Sperelakis 1984 Ultrastructure of mammalian cardiac muscle. In: Function of the Heart in Normal and Pathological States. N. Sperelakis, ed. Martinus Nijhoff, The Hague, pp. 3-42. Forbes, M.S., B.A. Plantholt, and N. Sperelakis 1977 ’ Cytochemical staining procedures selective for sarcotubular systems of muscle: Applications and modifications. J. Ultrastruct. Res., 60:306-327. Forssmann, W.G., and L. Girardier 1970 A study of the T system in rat heart. J. Cell Biol., 44t1-19. Geddes, L.A. 1970 The Direct and Indirect Measurement of Blood Pressure. Year Book Medical Publishers, Chicago. Gotoh, T. 1983 Quantitative studies on the ultrastructural differentiation and growth of mammalian cardiac muscle cells. The atria and ventricles of the cat. Acta Anat., 115:168-177. 162 M.S. FORBES, L.A. HAWKEY, AND N. SPERELAKIS Hirakow, R., and T. Gotoh 1980 Quantitative studies on the ultrastructural differentiation and growth of mammalian cardiac muscle cells. 11. The atria and ventricles of the guinea pig. Acta Anat., 108~230-237. Hirakow, R., T. Gotoh, and T. Watanabe 1980 Quantitative studies on the ultrastructural differentiation and growth of mammalian cardiac muscle cells. I. The atria and ventricles of the rat. Acta Anat., 108:144-152. Hoyle, G. 1965 Nature of the excitatory sarcoplasmic reticular junction. Science, 149~70-72. Ishikawa, H. 1968 Formation of elaborate networks of Tsystem tubules in cultured skeletal muscle with special reference to the T-system formation. J. Cell Biol., 3851-66. Ishikawa, H., and E. Yamada 1975 Differentiation of the sarcoplasmic reticulum and T-system in developing mouse cardiac muscle. In: Developmental and Physiological Correlates of Cardiac Muscle. M. Lieberman and T. Sano, eds. Raven Press, New York, pp. 21-35. Kelly, A.M. 1971 Sarcoplasmic reticulum and T tubules in differentiating rat skeletal muscle. J. Cell Biol., 49:335-344. Mall, G., H. b i n h a r d , K. Kayser, and J.A. Rossner 1978 An effective morphometric method for electron microscopic studies on papillary muscles. Virchows Arch. [Pathol. Anat.], 379~219-228. Myklebust, R., and H. Jensen 1978 Leptomeric fibrils and T-tubule desmosomes in the 2-band region of the mouse heart papillary muscle. Cell Tissue Res., 188.205-215. Osculati, F., and E. Garibaldi 1974 Fine structural aspects of the Purkinje fibres of the dog’s heart. J. Submicrosc. Cytol., 6~39-53. Osculati, F., S. Amati, E. Petrini, M. Marelli, and G. Gazzanelli 1978 A study on the organization of the tubular endoplasmic system in the rat heart conduction fibres. J. Submicrosc. Cytol., 10:371-380. Page, E., and L.P. McCalIister 1973 Quantitative electron microscopic description of heart muscle cells. Application to normal, hypertrophied and thyroxinstimulated hearts. Am. J. Cardiol., 31~172-181. Pager, J. 1971 Etude morphometrique du systeme tubulaire transverse du myocarde ventriculaire de rat. J. Cell Biol., 50:233-237. ReveI, J.-P., and M.J. Karnovsky 1967 Hexagonal array of subunits in intercellular junctions of the mouse heart and liver. J. Cell Biol., 33tC7-Cl2. Rubio, R., and N. Sperelakis 1971 Entrance of colloidal Thoz tracer into the T tubules and longitudinal tubules of the guinea pig heart. 2. Zellforsch. Mikrosk. Anat., 116:20-36. Rybicka, K. 1977 Sarcoplasmic reticulum in the conduct- ing fibers of the dog heart. Anat. Rec., 189:237-262. Singh, S., F.C. White, and C.M. Bloor 1982 Effect of acute exercise stress in cardiac hypertrophy. 11. Quantitative ultrastructural changes in the myocardial cell. Virchows Arch. [Cell Pathol.], 39t293-303. Sommer, J.R. 1982 Ultrastructural considerations concerning cardiac muscle. J. Mol. Cell. Cardiol. 14,(Suppl. 3):77-83. Sommer, J.R., and E.A. Johnson 1979 Ultrastructure of cardiac muscle. In: Handbook of Physiology, Section 2: The Cardiovascular System, Vol. I The Heart. R.M. Berne, N. Sperelakis, and S.R. Geiger, eds. Am. Physiol. SOC., Bethesda, pp. 113-186. Sommer, J.R., and R.A. Waugh 1976 The ultrastructure of the mammalian cardiac muscle cell-with special emphasis on the tubular membrane systems. Am. J. Pathol., 82:191-232. Sperelakis, N., and R. Rubio 1971 An orderly lattice of axial tubules which interconnect adjacent transverse tubules in guinea-pig ventricular myocardium. J. Mol. Cell. Cardiol., 2.211-220. Sperelakis, N., M.S. Forbes, and R. Rubio 1974 The tubular systems of myocardial cells: ultrastructure and possible function. In: Recent Advances in Studies on Cardiac Structure and Metabolism, Vol. IV,Myocardial Biology. N.S. Dhalla, ed. University Park Press, Baltimore, pp. 163-194. Tomanek, R.J. 1979 Quantitative ultrastructural aspects of cardiac hypertrophy. Texas Rep. Biol. Med., 39:111122. Vye, M.V., and D.A. Fischman 1970 The morphological alteration of particulate glycogen by en bloc staining with uranyl acetate. J. Ultrastruct. Res., 33:278-291. Waugh, R.A., and J.R. Sommer 1975 Longitudinal transverse tubules in cardiac muscle. J. Cell Biol., 67~449a. Weibel, E.R. 1969 Stereological principles for morphometry in electron microscopic cytology. Int. Rev. Cytol., 26~235-302. Weibel, E.R. 1973 Stereological techniques for electron microscopic morphometry. In: Principles and Techniques of Electron Microscopy, Vol. 3. M.A. Hayat, ed. Van Nostrand Reinhold Co., New York, pp. 237-296. Yamada, E., and H. Ishikawa 1981 Dense tissue and special stains. In: Methods in Cell Biology, Vol. 22: Three-dimensional Ultrastructure in Biology. J.N. Turner, ed. Academic Press, New York, pp. 123-145. Zampighi, G., J. Vergara, and F. Ramon 1975 On the connection between the T-tubules and the plasma membrane in frog semitendinosus skeletal muscle. Are caveolae the mouths of the transverse tubule system? J. Cell Biol., 64~734-740.