Anat Embryol (1990) 182:409-424
Anatomy and Embryology 9 Springer-Verlag 1990
Review article
Hypertrophy of visceral smooth muscle Giorgio Gabeila Department of Anatomy,UniversityCollegeLondon, Gower Street, London WCI E 6BT, United Kingdom Accepted July 7, 1990
Summary. Smooth muscles of viscera undergo a large increase in volume when there is a chronic, partial obstruction impairing the flow of lumenal contents. Hypertrophy of smooth muscle occurs in various medical conditions and several methods are available for inducing it experimentally in laboratory animals, especially in urinary bladder, small intestine and ureter. The hypertrophic response differs somewhat with the type of organ, the animal species, the age of the subject, and the experimental procedure. Ten- to fifteen-fold increases in muscle volume develop within a few weeks in the urinary bladder or the ileum of adult animals, a growth that would not have occurred in the lifespan of the animal without the experimental intervention. The general architecture of the muscle and the boundaries with adjacent tissues are well preserved. In intestinal hypertrophy, muscle cells increase in number: mitoses are found in mature, fully differentiated muscle cells. Cell division by full longitudinal splitting of muscle cells may also occur. Enlargement of muscle cells accounts for most of the muscle hypertrophy. The hypertrophic muscle cell has an irregular profile with deep indentations of the cell membrane, bearing caveolae and dense bands; however, the cell surface grows less than the cell volume (reduction of surface-to-volume ratio). The nucleus is crenated and is much less enlarged than the cell (reduction of the nucleo-plasmatic ratio). Mitochondria grow in number but in some muscles their spatial density decreases; intermediate filaments increase more than myofilaments. The spatial density of sarcoplasmic reticulum is generally increased. In the hypertrophic intestine, gap junctions increase in number and size; in the bladder, gap junctions are absent both in control and in hypertrophy. Thus the hypertrophic muscle cell is not only larger than a control cell, but has a different pattern of its structural components. Extensive neo-angiogenesis maintains a good blood supply to the hypertrophic muscle. The density of innervation is much decreased in the hypertrophic intestine, whereas it appears well maintained in the bladder. Neu-
ronal enlargement is found in the intramural ganglia of the intestine and in the pelvic ganglion. The mechanisms involved in hypertrophic growth are unknown. Three possible factors, mechanical factors, especially stretch, altered nerve discharge, and trophic factors are discussed. Key words: Smooth muscle - Hypertrophy - Intestine - Urinary bladder - Cell division - Cell size - Intermediate filaments - Gap junctions
Organs, tissues and cells grow during development until they reach a standard size, at a stage in life defined as maturity. In individuals of the same species and strain, the standard size varies within a limited range. Little is known of the mechanisms by which growth is slowed down and eventually stopped, though this process is regarded, perhaps wrongly, as a failure of growth rather than an inhibition of growth. Development unfolds according to a plan that regulates and harmonises the relative size of each organ at each stage up to maturity. Growth in excess of this plan is hypertrophic growth, and the process whereby an organ grows above the size it has at maturity, or beyond the relative size it averages at any stage of development, is hypertrophy. A classical introduction to the question of growth, regeneration and hypertrophy in many organs of the body and a general discussion is found in a book by Richard Goss (1964). In the musculature of viscera (including blood vessels, which, however, are not examined in this article) the occurrence of hypertrophy (and atrophy) is widespread, on account of the great variability in the volume and density of the contents of hollow organs, and in the resistance of their outflow. The content can be reduced to nil, or, at the opposite extreme, it can increase many times, while the outflow resistance can become progressively harder to overcome. Smooth muscle is a tissue exquisitely sensitive to the mechanical conditions imposed upon it. It responds to
410 nerve transmitters, to other substances released locally, and to hormones; it is often spontaneously active, and it responds directly to stretch, to changes in temperature, and to other physical factors. Many of these factors also affect the development of the muscle, a development normally occurring when the muscle is already performing mechanical tasks. While there is a limit to the extent of muscle growth, and while there is no complete atrophy even in conditions of total disuse 1, the volume of a smooth muscle is promptly affected by changes in the mechanical conditions imposed upon it. The response to these changes is the basis of hypertrophy; in this sense, hyperplasic or malignant growth is not hypertrophy. Hypertrophy should also be distinguished from growth and development. Although the distinction is not always simple, hypertrophy can occur during development and can be congenital or derive from a congenital condition (e.g. megacolon in Hirschprung's disease). Finally, enlargement of hollow organs, such as stomach and caecum, can occur without hypertrophy. The vas deferens on the epididymal side of a complete obstruction or transection becomes greatly distended and increases in length; its muscle coat, however, is thinned and does not hypertrophy (Gillenwater et al. 1978). Smooth muscle hypertrophy arises when there is a local obstruction or constriction of the lumen in a tubular organ, or when an obstruction impairs the outlet of a hollow organ.
Occurrence in humans Numerous examples of visceral smooth muscle hypertrophy are encountered in human medicine: in segments of the gut orad to a partial obstruction (aganglionosis, either congenital, as in Hirschprung's disease, or acquired, as in Chagas disease, and ensuing spasm and obstruction; congenital strictures; malformation of sphincters; lumenal obstructions by tumoural growths), in the urinary bladder after outlet obstruction (benign prostatism; strictures, tumours and congenital abnormalities of the urethra; autonomic neuropathies), in the bile duct and ureter when partially obstructed; in trachea and bronchi with asthma. The myometrium undergoes extensive hypertrophy during pregnancy. However, this latter special case of hypertrophy (which is surprisingly little studied) will not be considered in this review.
Experimental models in laboratory animals Urinary bladder hypertrophy can be obtained by placing a ligature around the urethra to reduce the lumen (Brent and Stephens 1975; Mattiasson and Uvelius 1982; Levin et al. 1984; Steers and De Groat 1988; Elbadawi et al. 1989). These experiments are unlike human prostatism in that the partial obstruction is established abruptly 1 There is little or no change in intestinal muscle cell sizes in parenteral nutrition (Nygaard 1967; Nemeth et al. 1983). A special case is that of the musculature of maIe genital organs, where an atrophy of up to 90% occurs following castration (Wakade et al. 1975)
at the time of operation. However, the cystometric characteristics are very similar to those of the human hypertrophic bladder, and in both cases there is the important dysfunction known as detrusor instability (Malmgren et al. 1987). By placing a silver ring around the urethra of immature animals (usually pigs), the urethra slowly becomes obstructed following the growth of the animal, and the detrusor muscle hypertrophies (physiological growth and hypertrophy) (Brent and Stephens 1975; Sibley 1985; Mockless et al. 1988). Hypertrophy ensues also after denervation of the bladder (denervation and hypertrophy) (Eltiott 1970; Carpenter and Root 1951; Goss etal. 1973; Sharkey etal. 1983). Other investigators have induced hypertrophy by placing a bolus of paraffin chronically into the bladder (Peterson et al. 1974), or simply by repeatedly injecting fluid into the organ (Carey 1921). Compression or ligature is used to induce hypertrophy of the ureter (Lange 1940). An experimental stenosis (narrowing of the lumen, for example by applying externally a ring of cellophane) of a short segment of small intestine, slows down the transit of ingesta, causes their accumulation on the oral side, followed by gradual distention of the lumen and hypertrophy of all the coats (intestinal hypertrophy), including the muscularis externa (smooth muscle hypertrophy) (Herczel 1886; Gabella 1979a) (Figs. 1, 2, 4-7). In contrast, complete occlusion leads to rapid distension followed by necrosis and perforation before there is a growth response in the wall. Experimental hypertrophy of the ileum in the rat can be obtained by creating a continent self-emptying ileostomy reservoir (Philipson et al. 1985). The colon of the rabbit hypertrophies with anal stenosis (Brent 1973). Resection of a long segment of small intestine is followed by hypertrophy of the remaining segment. Dogs
Figs. 1-7. Guinea-pig hypertrophic intestine (except Fig. 3). Phasecontrast micrographs Fig. 1. Transverse section of the wall, about 15 cm oral to the obstruction, showing the regular architecture of the muscle and the increase in thickness. Both muscle layers are hypertrophic. The a r r o w s point to some of the blood vessels (full of red blood cells) in the circular layer, u, muscutaris mucosae, x 170 Fig. 2. Transverse section of the circular muscle layer. The muscle cells are enlarged, have an irregular and heterogeneous shape, and their tapering ends are often split into two or three processes. The a r r o w points to a layer of small and dark muscle cells lining the submucosal surface of the circular muscle and penetrating for some distance along an intramuscular septum. • 610 Fig. 3. Hypertrophic musculature of rat urinary bladder. Muscle bundle in transverse section. Enlargement of the cell profiles and extensive crenation of the cell surface are apparent, x 560 Fig. 4. Muscle cell in late anaphase in the circular layer, x 720 Fig. 5. Musclecell undergoing mitosis in the circular layer, m, longitudinal muscle layer, x 720 Fig. 6. A bundle of musculature (arrow) oriented circumferentially is present in the serosa, m, longitudinal muscle layer; v, vessel, probably lymphatic, x 450 Fig. 7. A ganglion of the myenteric plexus surrounded by hypertrophic musculature of the circular (top) and logitudinal (bottom) layer. The small, black dots around the ganglion are elastic fibres. x 640
411
412 and rats may show full recovery in weight and metabolism after a resection of 50% of the small intestine (Flint 1912; Touloukian and Spencer 1972). The hypertrophy 2 involves :mainly the mucosa, but in certain cases there is some growth of the muscle too (Booth et al. 1959; Touloukian et at. 1972). The intestinal hypertrophy of lactating animals is very marked, but it is probably due entirely to changes in the mucosa (Fell et al. 1963). A high-fibre-containing diet produces an increase in length and weight of every portion of the gut, especially the large intestine (Addis 1932), where muscle hypertrophy has been documented (Jacobs 1985). Hypertrophy of the small intestine musculature occurs in rats fed a diet rich in pectin (Brown et al. 1979). Enlargement of the caecum is also observed in animals reared in germfree conditions (Philipps et al. 1959; FaIk 1960), in relation to the absence of intestinal flora (Gordon 1959), and in diabetic animals (Pillion et al. 1988). The histology of the caecal musculature has not yet been studied, but we have observed that in rats fed with a high-fibre diet, or born and bred in germ-free conditions, or affected by streptozotocin diabetes, the enlargement of the caecum is accompanied by a moderate thickening of its wall and thus involves a process of hypertrophy. Lack of muscle tone in germ-free rats also contributes to the enlargement (Gordon 1959). Mice with congenital absence of ganglion neurons in the distal colon (lethal spotted mice, Ls/Ls) (Lane 1966; Webster 1973) develop a local obstruction, followed by an enlargement of the colon and a thickening of the walt on the oral side.
Conditions creating hypertrophy In the examples mentioned, hypertrophy is not an acute process: the stimulus that produces it must be intense and persistent, but not such as to cause acute organ failure. Stow-setting but progressive stimuli are the most effective. There is always distension of the lumen and an increase not only in thickness but also in extent of the wall. Hypertrophy is a response of the muscle or an adaptation to an increase in the functional load on the muscle. It is useful for the maintenance of the organ function and the survival of the organism. It may thus be regarded as an acquired mechanism that in evolutionary terms represents an anticipation of the risk of obstruction of a viscus. It remains doubtful whether a smooth muscle can hypertrophy without overload and distension (this is sometimes defined as' work hypertrophy' [see, for example, Bortoff and Sillin 1986]). The hypothesis of a work hypertrophy is used by Nemeth et al. (1983) to explain the intestinal hypertrophy that follows an increase in the bulk of the diet. Goss et al. (1973), in complex experiments involving pairs of rats in parabiosis and bilateral nephrectomy of one of them, concluded that the hypertrophy of the bladder of the non-nephrectomized rat was partly due to the increased amount of urine, and 2 The term 'compensatory hypertrophy' is correctly used in this case
partly to the increased frequency of voiding. The latter component would be an example of work hypertrophy. A hypertrophy of this type might occur in rats with diabetes insipidus, in which the frequency of micturition is much increased; however, the vastly increased flow of urine in this condition also enhances distension of the organ, and it remains unclear to what extent the hypertrophy is due to excessive distension and to what extent to increased work. The occurrence of work hypertrophy, therefore, remains to be proven, and new experimental conditions should be devised to investigate the problem. One wonders, for example, whether an exposure to low temperature, repeated at frequent intervals and over a long period of time, would produce hypertrophy of the arrectores pilorum. Hypertrophy was not found in the tracheal muscle of hyperventilating rats at the end of a three-week sojourn in a hypoxic chamber; in these conditions the increased muscle work was not accompanied by stretch and the muscle did not hypertrophy.
Hypertrophic growth The increase in size (weight) of smooth muscles during hypertrophy in adult animals is very large. The musculature of bladder or small intestine in laboratory rodents increases up to 15 times (Gabella 1979a; Gabella and Uvelius 1990) (Figs. 1, 3, 8-9). It is not clear whether this amount is related to an absolute limit in the capacity of growth, or to a limit in the size a working organ can reach in a given organism. An eightfold increase in muscle volume was found in the rat ureter on the cranial side of an obstruction and in the small intestine orad to a resection (Lange 1940). In the obstructed bladder, muscle hypertrophy appears to occur uniformly over the entire organ, whereas in the intestine the hypertrophy is maximal a few centimetres orad to the stenosis and is progressively less intense more orally (Gabella 1987). At every level, the muscle coat has uniform thickness around the circumference of the intestine (Fig. I). The diameter of the maximally hypertrophic intestine is two to three times that of controls, and the wall is three to four times thicker. The length of the gut is virtually unchanged, probably because the mesentery is inextensible. Therefore, the growth in muscle volume is due to an increase not only in the thickness of the muscle but also in the circumference of the intestine; the latter corresponds to an increase in length (but not in width) of the circular muscle, and to an increase in width (but not in length) of the longitudinal muscle, both changes posing complex structural problems. The longitudinal muscle layer usually hypertrophies to the same extent as the circular layer 3 (Fig. l). In the hypertrophic ureter 3A possible explanation of the fact that both muscle layers of the intestine hypertrophy,is that the increased intraluminal pressure, caused by accumuIationof ingesta orad to the obstruction, produces transmural stresses that are distributed circumferentially and longitudinallyvia the cross-ply arrangement of the collagen bundles in the submucosa, which in mammals may be regarded as the "skeletal component" of the intestinal wall
Fig. 8. General appearance of the bladder musculature of a control rat. Most cell profiles are muscle cells, some of which display the nucleus. The dark structures within the muscle cells are mitochondria. F i s a fibroblast. The arrows point to nerves, x 7800
Fig. 9. Hypertrophic bladder of rat at the same magnification as the control preparation above. The muscle cells are enormously enlarged and display deep invaginations (arrows) of the cell membrane. N, nucleus; n, nerve fibres, x 7800
414 of the rabbit there is a doubling of the external diameter and only a marginal increase in length (Haussman et al. 1979).
and these are large enough to account for the muscle hypertrophy (Gabella and Uvelius 1990) (Figs. 8, 9).
Hyperplasia Time course
Depending on the degree of obstruction, intestinal hypertrophy can develop in a rat in less than two weeks, but it takes twice as long in a guinea-pig or a rabbit. In the rat bladder, the hypertrophy can develop in a matter of days, although morphological and physiological studies are usually carried out after 6 10 weeks. The effect of a high-fibre diet takes several weeks to become manifest (for example, 10 weeks in Jacobs' [1985] experiments). In Lange's (1940) experiments on the ureter, structural changes were observed already three days after the obstruction.
Architecture of the musculature
The hypertrophic growth is harmonious and the general architecture of the muscle well maintained (Figs. 1 3). In the obstructed bladder the basic arrangement of the musculature is the same as in controls, although larger and longer muscle bundles are present. Some muscle bundles lift the mucosa as folds into the lumen and give rise to the so-called trabeculation of the bladder inner surface (Gosling and Dixon 1980). While the bladder is much enlarged, its shape, and the characteristic ratios of its axes, are not altered (Gabella and Uvelius 1990). In the hypertrophic intestine, outer and inner surfaces of the muscle coat remain sharp and planar (Figs. 1, 2), and the arrangement of the muscle cells is almost as orderly as in controls. Generally speaking, the hypertrophic response seems uniformly to affect all the cells of the muscle. Sometimes small bundles of muscle running circumferentially appear in the serosa (Gabella 1979 a) (Fig. 7). Muscle cell hypertrophy
The most obvious aspect of muscle hypertrophy is the increase in size of its muscle cells (cell hypertrophy) (Figs. 2, 3, 8, 9). The detection of this process in histological preparations requires a good control of the conditions of distension or contraction of the muscle at the time of fixation, to distinguish true changes in size from apparent ones. Many authors have assessed muscle cell size by measuring nuclear size (diameter), an approach clearly not accurate. Lange (1940) reported that the nucleo-plasmatic ratio in hypertrophic muscle cells did not change, but this has been denied in more recent studies (Gabella 1979 a; Gabella and Uvelius 1990). In the intestine, muscle cell volume increases from 3500 gm 3 in controls to 13 500 gin 3, a three- to fourfold increase that is quite substantial, but insufficient fully to account for the increase in muscle volume (Gabella 1979a). Even larger increases in muscle cell size occur in the bladder,
The question of muscle cell hyperplasia has been considered by many authors, but there is little firm evidence on this important issue and no quantitative studies. In the obstructed ureter, Gee and Kiviat (1975) estimate that in rabbits there is a doubling of the number of muscle cells, whereas, in the rat, Lange (1940) found no increase in cell number, and thought that the mitoses observed only provided new muscle cells to replace those which had died; Cussen and Tymms (1972) observed in pups a fourfold increase in muscle cells size and a fourfold increase in cell number due to division of preexisting muscle cells. Changes in the structure of the nucleus and the actual detection of dividing muscle cells (see below) indicate that mitoses do occur in hypertrophic muscles.
Nucleus
In the hypertrophic bladder and intestine, nucleated muscle cell profiles are less frequent than in controls (4-5% of all muscle cell profiles, as opposed to 7-8%), and nuclei occupy a lesser proportion of the profile area (about 10%, as opposed to over 30% in controls) and are often eccentrically placed (Gabella and Uvelius 1990). Nuclear transverse sectional area is about twice that in controls, but there is no significant change in nucleus length. Therefore, in hypertrophy the increase in nuclear volume is less than that of the cell, and the nucleus-cytoplasm ratio is much lower than in control muscle cells. Nuclei display a crenate or lobed profile, and many have prominent nucleoli. Binucleate muscle cells (with the two nuclei aligned along the cell length) are not uncommon, whereas they are very rare in normal visceral muscles. Some nuclei become tetraploid, but quantitative studies of this change are not available; tetraploidy is well documented in the pregnant myometrium (Heijden and James 1975) and in hypertensive blood vessels (Owens and Schwartz 1983). Mitosis
In the hypertrophic intestine of the guinea-pig mitoses are observed in both the circular and longitudinal layers (Gabella 1979a) (Figs. 4, 5). The cells undergoing division are differentiated muscle cells, with a full complement of myofilaments and specialized structures such as dense bodies, dense bands, caveolae and cell junctions. The muscle contains no undifferentiated cells, such as myoblasts. Also, there is no evidence that in hypertrophy, muscle cells de-differentiate and then divide, as is usually observed in tissue culture (Chamley-Campbell et al. 1979). In the hypertrophic intestine, dividing muscle cells in the longitudinal muscle are found only in
415 the inner half of the layer; by contrast, in the circular muscle they are predominantly located in the middle portion of the layer, approximately half way between longitudinal muscle and submucosa (Gabella 1979a). The dividing muscle cells display a central portion, ovoid, occupied by chromosomes and at a later stage by the mitotic spindle. On either side of this portion, the cell appears as a fully differentiated muscle cell, indistinguishable from its neighbours. The mitotic equatorial plate lies invariably transverse to the length of the muscle. Brent (1973) observed dividing muscle cells in the hypertrophic large intestine of rabbits; the increase in cell number was much more substantial in immature animals than in adult animals. The occurrence of mitoses has not been recorded
in the hypertrophic bladder, where the increase in cell size may fully account for the muscle enlargement (Figs. 8, 9). However, the total D N A content of the bladder musculature increases nine times after a 6-week obstruction and its concentration decreases only by a factor of 2 (Uvelius et al. 1984). Part of this D N A increase is probably accounted for by enlargement of the nucleus and by polyploidy, as is found in vascular smooth muscle hypertrophy in hypertensive rats (Owens and Schwartz 1983); some of the increase may be the results of mitoses occurring in the early stages of hypertrophy that are not yet documented. Some hypertrophic muscle cells display a nucleus almost completely split into two, the two parts lying side by side and showing no changes in the chromatin pattern
Fig. 10. Photographic montage of the circular muscle layer, in transverse section, and the longitudinal muscle layer, in longitudinal section, from the hypertrophic intestine of a rat. The cell indicated by a r r o w s runs circumferentially for part of its length and longitudinally for another part, and is located in both muscle layers. The insets show details of the same cell at higher magnification. A possible interpretation of this appearance is that the cell
was originally located at the border between longitudinal and circular layers, running in either of them: growth of the tissue and mechanical distortion, such as compression and stretch, pulled a part of the cell with the longitudinal muscle and another part with the circular muscle,j, gap junction; n, nerve bundle. (From Gabella 1987) x 3800
416
(Figs. 13, 14). These appearances suggest the occurrence of cell divisions by a mechanism different from mitosis, that is, a kind of amitotic division (Gabella 1979 a).
along the length of the circular muscle, and reflex contractures arising from the abnormal stimulation of the wall. The tapering ends of the muscle cells are often split into two or more profiles (Fig. 2).
Cell shape Cell surface Many of the hypertrophic muscle cell profiles in bladder and gut are polygonal, crescent-like, very flattened, indented by other cells, or wedge-shaped (Figs. 2, 3), in contrast to the ovoid or simple polygonal shape of control muscle cell profiles. The shape of the profiles suggests that the muscle cells are exposed to an exceptional level of stress (Fig. 10), including compression from the increased content of the lumen, circumferential stretch
Fig. 11. Hypertrophic bladder of rat. Muscle cells, in transverse section, show an irregular surface with numerous, long invaginations, f, collagen fibrils, x 25000
The hypertrophic muscle cells of the intestine and bladder have prominent invaginations of the cell membrane, often radially arranged as wedge-shaped (Fig. 11) or finger-shaped (Figs. 12, 16) inward folds. These invaginations are found at any point along the cell, and seem to be present in some ceils to a greater extent than in others. They reach as far as halfway into the deepest
Fig. 12. Hypertrophic ileum of guinea-pig. The nucleated cell profile (centre) shows five tubular invaginations of the cell membrane; they are lined by a basal lamina, and contain fibrillar material. N, nucleus; n, nerve fibre, x 14500
4t7 point of the cell profile, and their membrane bears dense bands and caveolae and is coated by a basal lamina (Figs. 12, 16). While in some cells these invaginations are evenly distributed around the circumference, in others they occur only over the side facing the edge of a muscle bundle (Fig. 9). The infolding of the cell membrane produces an increase of the cell surface. Geometrically, the surface grows less than the volume, and, in spite of the membrane infoldings, the surface-to-volume ratio falls from 1.4 in control to 0.8 in hypertrophy, i.e. there is 0.8 gm 2 of cell surface per cubic micron of cell volume. The cell membrane is studded with caveolae that are distributed in rows (Fig. 21) between the areas occupied by dense bands (Fig. 20) as in controls, and are similar in shape and size to those of control cells. In intestinal hypertrophic muscle cells, the spatial density of caveolae is marginally smaller than in controls (16.5/gm 2 vs 19.2/ ~m 2) (Gabella 1979a). However, on account of the increased cell size, the number of caveolae per cell must be substantially higher than the control values of about 170 000.
Filaments In the hypertrophic muscle cells of the intestine of rat and guinea-pig, all three types of filament (thin, thick and intermediate) increase in number, on account of the large increase in cell size. Counts of filaments are not available, but thin filaments seem to increase more than thick filaments. Intermediate filaments increase substantially more than either type of myofilament and the cells display large bundles of them (Fig. 15) (Gabella 1979 b). In the hypertrophic bladder, in contrast, changes in the relative proportion of filaments are not readily seen (Gabella and Uvelius 1990) (Fig. 16). Biochemically, however, there is no change in the concentration of actin and a modest fall in the concentration of myosin; the concentration of intermediate filament proteins (90% desmin and 10% vimentin) increases (Uvelius etal. 1989). Active stress of the muscle (force generated/unit transverse area) is unchanged in the hypertrophic ureter of the rabbit (Haussman et al. 1979), whereas in the guinea-pig intestine it is reduced (Gabella 1979 b). While the force generated by the hypertrophic intestinal muscle is much increased, on account of its increased mass, the force generated per unit area is less than in controls. A similar fall in the 'efficiency' in hypertrophy is found also in other muscles (e.g. the portal vein [Johansson 1976]), but the reasons for this change are not yet clear.
Cell organelles A morphometric study of hypertrophic muscle cells of the guinea-pig intestine revealed a halving of the percentage volume of mitochondria from about 7% in the control muscle. Even if the percentage volume decreases, the total mitochondrial volume per cell is about twice
as large as in controls (Gabella 1979a). In the rat intestine, the relative decrease of mitochondria is less marked, and in the rat bladder it is barely noticeable. Decrease in spatial density of mitochondria is a characteristic of most forms of cardiac hypertrophy (Anversa et al. 1971 ; Page et al. 1972) and of the so-called "overload" hypertrophy of skeletal muscles (as opposed to "endurance" hypertrophy, which has the opposite effect on mitochondria) (Goldspink 1971 ; Hoppeter et al. t973). Various forms of sarcoplasmic reticulum are found in smooth muscle cells, probably representing different functional specializations. In hypertrophic muscle cells, sacs and tubules of smooth sarcoptasmic reticulum are more abundant than in controls (Gabella 1979b) (Figs. 15, 20). Fenestrated cisternae of smooth reticulum of great extension and complex shape are common near the surface of intestinal hypertrophic muscle cells (Fig. 21). In control muscle, sarcoplasmic reticulum has been implicated in calcium storage and release (Johansson and Somlyo 1980). It is possible that the expanded sarcoplasmic reticulum of hypertrophic cells, which have a reduced surface-to-volume ratio, helps to maintain adequate calcium supply to the contractile apparatus in the presence of a relative reduction of calcium entry through the cell membrane. In many hypertrophic cells, but not in all of them, at least judging from single transverse sections, there is substantial increase in rough sarcoplasmic reticulum, often in the form of radially arranged cisternae that approach with one edge the cell membrane. Some cisternae are expanded by an amorphous content of medium electron density. The evidence suggests an activation of the synthetic and secretory activity of the muscle cell.
Gap junctions The intestinal musculature (circular layer) is well provided with gap junctions (Fig. 20). With hypertrophy the number of gap junctions per cell unit surface does not change (about 46 junctions per 1000 gm 2, in the guinea-pig); there is, therefore, an increase in the number of junctions per cell, proportional to the cell surface increase. Moreover, the average size of junctions increases (Fig. 19) and the percentage of cell membrane surface occupied by gap junctions doubles (from 0.22% to 0.49%) (GabeUa 1979b). Measurements of tissue impedance in the hypertrophic ileum of the cat reveal an increase in intercellular electrical coupling (Bortoff and Sillin 1986). By contrast, gap junctions are very rare in the longitudinal musculature of the gut, and they remain so in hypertrophy. Gap junctions were not found in the urinary bladder, either in control or in hypertrophy (Gabella and Uvelius 1990).
Nerves In the hypertrophic intestine, intramuscular nerves grow less than the muscle, and nerve bundles are sparser than in controls. There is no loss of nerve fibres; the pre-
418
419
;-.~
o~
\ ~OL)_
Anatomy and Embryology 9 Springer-Verlag 1990
Review article
Hypertrophy of visceral smooth muscle Giorgio Gabeila Department of Anatomy,UniversityCollegeLondon, Gower Street, London WCI E 6BT, United Kingdom Accepted July 7, 1990
Summary. Smooth muscles of viscera undergo a large increase in volume when there is a chronic, partial obstruction impairing the flow of lumenal contents. Hypertrophy of smooth muscle occurs in various medical conditions and several methods are available for inducing it experimentally in laboratory animals, especially in urinary bladder, small intestine and ureter. The hypertrophic response differs somewhat with the type of organ, the animal species, the age of the subject, and the experimental procedure. Ten- to fifteen-fold increases in muscle volume develop within a few weeks in the urinary bladder or the ileum of adult animals, a growth that would not have occurred in the lifespan of the animal without the experimental intervention. The general architecture of the muscle and the boundaries with adjacent tissues are well preserved. In intestinal hypertrophy, muscle cells increase in number: mitoses are found in mature, fully differentiated muscle cells. Cell division by full longitudinal splitting of muscle cells may also occur. Enlargement of muscle cells accounts for most of the muscle hypertrophy. The hypertrophic muscle cell has an irregular profile with deep indentations of the cell membrane, bearing caveolae and dense bands; however, the cell surface grows less than the cell volume (reduction of surface-to-volume ratio). The nucleus is crenated and is much less enlarged than the cell (reduction of the nucleo-plasmatic ratio). Mitochondria grow in number but in some muscles their spatial density decreases; intermediate filaments increase more than myofilaments. The spatial density of sarcoplasmic reticulum is generally increased. In the hypertrophic intestine, gap junctions increase in number and size; in the bladder, gap junctions are absent both in control and in hypertrophy. Thus the hypertrophic muscle cell is not only larger than a control cell, but has a different pattern of its structural components. Extensive neo-angiogenesis maintains a good blood supply to the hypertrophic muscle. The density of innervation is much decreased in the hypertrophic intestine, whereas it appears well maintained in the bladder. Neu-
ronal enlargement is found in the intramural ganglia of the intestine and in the pelvic ganglion. The mechanisms involved in hypertrophic growth are unknown. Three possible factors, mechanical factors, especially stretch, altered nerve discharge, and trophic factors are discussed. Key words: Smooth muscle - Hypertrophy - Intestine - Urinary bladder - Cell division - Cell size - Intermediate filaments - Gap junctions
Organs, tissues and cells grow during development until they reach a standard size, at a stage in life defined as maturity. In individuals of the same species and strain, the standard size varies within a limited range. Little is known of the mechanisms by which growth is slowed down and eventually stopped, though this process is regarded, perhaps wrongly, as a failure of growth rather than an inhibition of growth. Development unfolds according to a plan that regulates and harmonises the relative size of each organ at each stage up to maturity. Growth in excess of this plan is hypertrophic growth, and the process whereby an organ grows above the size it has at maturity, or beyond the relative size it averages at any stage of development, is hypertrophy. A classical introduction to the question of growth, regeneration and hypertrophy in many organs of the body and a general discussion is found in a book by Richard Goss (1964). In the musculature of viscera (including blood vessels, which, however, are not examined in this article) the occurrence of hypertrophy (and atrophy) is widespread, on account of the great variability in the volume and density of the contents of hollow organs, and in the resistance of their outflow. The content can be reduced to nil, or, at the opposite extreme, it can increase many times, while the outflow resistance can become progressively harder to overcome. Smooth muscle is a tissue exquisitely sensitive to the mechanical conditions imposed upon it. It responds to
410 nerve transmitters, to other substances released locally, and to hormones; it is often spontaneously active, and it responds directly to stretch, to changes in temperature, and to other physical factors. Many of these factors also affect the development of the muscle, a development normally occurring when the muscle is already performing mechanical tasks. While there is a limit to the extent of muscle growth, and while there is no complete atrophy even in conditions of total disuse 1, the volume of a smooth muscle is promptly affected by changes in the mechanical conditions imposed upon it. The response to these changes is the basis of hypertrophy; in this sense, hyperplasic or malignant growth is not hypertrophy. Hypertrophy should also be distinguished from growth and development. Although the distinction is not always simple, hypertrophy can occur during development and can be congenital or derive from a congenital condition (e.g. megacolon in Hirschprung's disease). Finally, enlargement of hollow organs, such as stomach and caecum, can occur without hypertrophy. The vas deferens on the epididymal side of a complete obstruction or transection becomes greatly distended and increases in length; its muscle coat, however, is thinned and does not hypertrophy (Gillenwater et al. 1978). Smooth muscle hypertrophy arises when there is a local obstruction or constriction of the lumen in a tubular organ, or when an obstruction impairs the outlet of a hollow organ.
Occurrence in humans Numerous examples of visceral smooth muscle hypertrophy are encountered in human medicine: in segments of the gut orad to a partial obstruction (aganglionosis, either congenital, as in Hirschprung's disease, or acquired, as in Chagas disease, and ensuing spasm and obstruction; congenital strictures; malformation of sphincters; lumenal obstructions by tumoural growths), in the urinary bladder after outlet obstruction (benign prostatism; strictures, tumours and congenital abnormalities of the urethra; autonomic neuropathies), in the bile duct and ureter when partially obstructed; in trachea and bronchi with asthma. The myometrium undergoes extensive hypertrophy during pregnancy. However, this latter special case of hypertrophy (which is surprisingly little studied) will not be considered in this review.
Experimental models in laboratory animals Urinary bladder hypertrophy can be obtained by placing a ligature around the urethra to reduce the lumen (Brent and Stephens 1975; Mattiasson and Uvelius 1982; Levin et al. 1984; Steers and De Groat 1988; Elbadawi et al. 1989). These experiments are unlike human prostatism in that the partial obstruction is established abruptly 1 There is little or no change in intestinal muscle cell sizes in parenteral nutrition (Nygaard 1967; Nemeth et al. 1983). A special case is that of the musculature of maIe genital organs, where an atrophy of up to 90% occurs following castration (Wakade et al. 1975)
at the time of operation. However, the cystometric characteristics are very similar to those of the human hypertrophic bladder, and in both cases there is the important dysfunction known as detrusor instability (Malmgren et al. 1987). By placing a silver ring around the urethra of immature animals (usually pigs), the urethra slowly becomes obstructed following the growth of the animal, and the detrusor muscle hypertrophies (physiological growth and hypertrophy) (Brent and Stephens 1975; Sibley 1985; Mockless et al. 1988). Hypertrophy ensues also after denervation of the bladder (denervation and hypertrophy) (Eltiott 1970; Carpenter and Root 1951; Goss etal. 1973; Sharkey etal. 1983). Other investigators have induced hypertrophy by placing a bolus of paraffin chronically into the bladder (Peterson et al. 1974), or simply by repeatedly injecting fluid into the organ (Carey 1921). Compression or ligature is used to induce hypertrophy of the ureter (Lange 1940). An experimental stenosis (narrowing of the lumen, for example by applying externally a ring of cellophane) of a short segment of small intestine, slows down the transit of ingesta, causes their accumulation on the oral side, followed by gradual distention of the lumen and hypertrophy of all the coats (intestinal hypertrophy), including the muscularis externa (smooth muscle hypertrophy) (Herczel 1886; Gabella 1979a) (Figs. 1, 2, 4-7). In contrast, complete occlusion leads to rapid distension followed by necrosis and perforation before there is a growth response in the wall. Experimental hypertrophy of the ileum in the rat can be obtained by creating a continent self-emptying ileostomy reservoir (Philipson et al. 1985). The colon of the rabbit hypertrophies with anal stenosis (Brent 1973). Resection of a long segment of small intestine is followed by hypertrophy of the remaining segment. Dogs
Figs. 1-7. Guinea-pig hypertrophic intestine (except Fig. 3). Phasecontrast micrographs Fig. 1. Transverse section of the wall, about 15 cm oral to the obstruction, showing the regular architecture of the muscle and the increase in thickness. Both muscle layers are hypertrophic. The a r r o w s point to some of the blood vessels (full of red blood cells) in the circular layer, u, muscutaris mucosae, x 170 Fig. 2. Transverse section of the circular muscle layer. The muscle cells are enlarged, have an irregular and heterogeneous shape, and their tapering ends are often split into two or three processes. The a r r o w points to a layer of small and dark muscle cells lining the submucosal surface of the circular muscle and penetrating for some distance along an intramuscular septum. • 610 Fig. 3. Hypertrophic musculature of rat urinary bladder. Muscle bundle in transverse section. Enlargement of the cell profiles and extensive crenation of the cell surface are apparent, x 560 Fig. 4. Muscle cell in late anaphase in the circular layer, x 720 Fig. 5. Musclecell undergoing mitosis in the circular layer, m, longitudinal muscle layer, x 720 Fig. 6. A bundle of musculature (arrow) oriented circumferentially is present in the serosa, m, longitudinal muscle layer; v, vessel, probably lymphatic, x 450 Fig. 7. A ganglion of the myenteric plexus surrounded by hypertrophic musculature of the circular (top) and logitudinal (bottom) layer. The small, black dots around the ganglion are elastic fibres. x 640
411
412 and rats may show full recovery in weight and metabolism after a resection of 50% of the small intestine (Flint 1912; Touloukian and Spencer 1972). The hypertrophy 2 involves :mainly the mucosa, but in certain cases there is some growth of the muscle too (Booth et al. 1959; Touloukian et at. 1972). The intestinal hypertrophy of lactating animals is very marked, but it is probably due entirely to changes in the mucosa (Fell et al. 1963). A high-fibre-containing diet produces an increase in length and weight of every portion of the gut, especially the large intestine (Addis 1932), where muscle hypertrophy has been documented (Jacobs 1985). Hypertrophy of the small intestine musculature occurs in rats fed a diet rich in pectin (Brown et al. 1979). Enlargement of the caecum is also observed in animals reared in germfree conditions (Philipps et al. 1959; FaIk 1960), in relation to the absence of intestinal flora (Gordon 1959), and in diabetic animals (Pillion et al. 1988). The histology of the caecal musculature has not yet been studied, but we have observed that in rats fed with a high-fibre diet, or born and bred in germ-free conditions, or affected by streptozotocin diabetes, the enlargement of the caecum is accompanied by a moderate thickening of its wall and thus involves a process of hypertrophy. Lack of muscle tone in germ-free rats also contributes to the enlargement (Gordon 1959). Mice with congenital absence of ganglion neurons in the distal colon (lethal spotted mice, Ls/Ls) (Lane 1966; Webster 1973) develop a local obstruction, followed by an enlargement of the colon and a thickening of the walt on the oral side.
Conditions creating hypertrophy In the examples mentioned, hypertrophy is not an acute process: the stimulus that produces it must be intense and persistent, but not such as to cause acute organ failure. Stow-setting but progressive stimuli are the most effective. There is always distension of the lumen and an increase not only in thickness but also in extent of the wall. Hypertrophy is a response of the muscle or an adaptation to an increase in the functional load on the muscle. It is useful for the maintenance of the organ function and the survival of the organism. It may thus be regarded as an acquired mechanism that in evolutionary terms represents an anticipation of the risk of obstruction of a viscus. It remains doubtful whether a smooth muscle can hypertrophy without overload and distension (this is sometimes defined as' work hypertrophy' [see, for example, Bortoff and Sillin 1986]). The hypothesis of a work hypertrophy is used by Nemeth et al. (1983) to explain the intestinal hypertrophy that follows an increase in the bulk of the diet. Goss et al. (1973), in complex experiments involving pairs of rats in parabiosis and bilateral nephrectomy of one of them, concluded that the hypertrophy of the bladder of the non-nephrectomized rat was partly due to the increased amount of urine, and 2 The term 'compensatory hypertrophy' is correctly used in this case
partly to the increased frequency of voiding. The latter component would be an example of work hypertrophy. A hypertrophy of this type might occur in rats with diabetes insipidus, in which the frequency of micturition is much increased; however, the vastly increased flow of urine in this condition also enhances distension of the organ, and it remains unclear to what extent the hypertrophy is due to excessive distension and to what extent to increased work. The occurrence of work hypertrophy, therefore, remains to be proven, and new experimental conditions should be devised to investigate the problem. One wonders, for example, whether an exposure to low temperature, repeated at frequent intervals and over a long period of time, would produce hypertrophy of the arrectores pilorum. Hypertrophy was not found in the tracheal muscle of hyperventilating rats at the end of a three-week sojourn in a hypoxic chamber; in these conditions the increased muscle work was not accompanied by stretch and the muscle did not hypertrophy.
Hypertrophic growth The increase in size (weight) of smooth muscles during hypertrophy in adult animals is very large. The musculature of bladder or small intestine in laboratory rodents increases up to 15 times (Gabella 1979a; Gabella and Uvelius 1990) (Figs. 1, 3, 8-9). It is not clear whether this amount is related to an absolute limit in the capacity of growth, or to a limit in the size a working organ can reach in a given organism. An eightfold increase in muscle volume was found in the rat ureter on the cranial side of an obstruction and in the small intestine orad to a resection (Lange 1940). In the obstructed bladder, muscle hypertrophy appears to occur uniformly over the entire organ, whereas in the intestine the hypertrophy is maximal a few centimetres orad to the stenosis and is progressively less intense more orally (Gabella 1987). At every level, the muscle coat has uniform thickness around the circumference of the intestine (Fig. I). The diameter of the maximally hypertrophic intestine is two to three times that of controls, and the wall is three to four times thicker. The length of the gut is virtually unchanged, probably because the mesentery is inextensible. Therefore, the growth in muscle volume is due to an increase not only in the thickness of the muscle but also in the circumference of the intestine; the latter corresponds to an increase in length (but not in width) of the circular muscle, and to an increase in width (but not in length) of the longitudinal muscle, both changes posing complex structural problems. The longitudinal muscle layer usually hypertrophies to the same extent as the circular layer 3 (Fig. l). In the hypertrophic ureter 3A possible explanation of the fact that both muscle layers of the intestine hypertrophy,is that the increased intraluminal pressure, caused by accumuIationof ingesta orad to the obstruction, produces transmural stresses that are distributed circumferentially and longitudinallyvia the cross-ply arrangement of the collagen bundles in the submucosa, which in mammals may be regarded as the "skeletal component" of the intestinal wall
Fig. 8. General appearance of the bladder musculature of a control rat. Most cell profiles are muscle cells, some of which display the nucleus. The dark structures within the muscle cells are mitochondria. F i s a fibroblast. The arrows point to nerves, x 7800
Fig. 9. Hypertrophic bladder of rat at the same magnification as the control preparation above. The muscle cells are enormously enlarged and display deep invaginations (arrows) of the cell membrane. N, nucleus; n, nerve fibres, x 7800
414 of the rabbit there is a doubling of the external diameter and only a marginal increase in length (Haussman et al. 1979).
and these are large enough to account for the muscle hypertrophy (Gabella and Uvelius 1990) (Figs. 8, 9).
Hyperplasia Time course
Depending on the degree of obstruction, intestinal hypertrophy can develop in a rat in less than two weeks, but it takes twice as long in a guinea-pig or a rabbit. In the rat bladder, the hypertrophy can develop in a matter of days, although morphological and physiological studies are usually carried out after 6 10 weeks. The effect of a high-fibre diet takes several weeks to become manifest (for example, 10 weeks in Jacobs' [1985] experiments). In Lange's (1940) experiments on the ureter, structural changes were observed already three days after the obstruction.
Architecture of the musculature
The hypertrophic growth is harmonious and the general architecture of the muscle well maintained (Figs. 1 3). In the obstructed bladder the basic arrangement of the musculature is the same as in controls, although larger and longer muscle bundles are present. Some muscle bundles lift the mucosa as folds into the lumen and give rise to the so-called trabeculation of the bladder inner surface (Gosling and Dixon 1980). While the bladder is much enlarged, its shape, and the characteristic ratios of its axes, are not altered (Gabella and Uvelius 1990). In the hypertrophic intestine, outer and inner surfaces of the muscle coat remain sharp and planar (Figs. 1, 2), and the arrangement of the muscle cells is almost as orderly as in controls. Generally speaking, the hypertrophic response seems uniformly to affect all the cells of the muscle. Sometimes small bundles of muscle running circumferentially appear in the serosa (Gabella 1979 a) (Fig. 7). Muscle cell hypertrophy
The most obvious aspect of muscle hypertrophy is the increase in size of its muscle cells (cell hypertrophy) (Figs. 2, 3, 8, 9). The detection of this process in histological preparations requires a good control of the conditions of distension or contraction of the muscle at the time of fixation, to distinguish true changes in size from apparent ones. Many authors have assessed muscle cell size by measuring nuclear size (diameter), an approach clearly not accurate. Lange (1940) reported that the nucleo-plasmatic ratio in hypertrophic muscle cells did not change, but this has been denied in more recent studies (Gabella 1979 a; Gabella and Uvelius 1990). In the intestine, muscle cell volume increases from 3500 gm 3 in controls to 13 500 gin 3, a three- to fourfold increase that is quite substantial, but insufficient fully to account for the increase in muscle volume (Gabella 1979a). Even larger increases in muscle cell size occur in the bladder,
The question of muscle cell hyperplasia has been considered by many authors, but there is little firm evidence on this important issue and no quantitative studies. In the obstructed ureter, Gee and Kiviat (1975) estimate that in rabbits there is a doubling of the number of muscle cells, whereas, in the rat, Lange (1940) found no increase in cell number, and thought that the mitoses observed only provided new muscle cells to replace those which had died; Cussen and Tymms (1972) observed in pups a fourfold increase in muscle cells size and a fourfold increase in cell number due to division of preexisting muscle cells. Changes in the structure of the nucleus and the actual detection of dividing muscle cells (see below) indicate that mitoses do occur in hypertrophic muscles.
Nucleus
In the hypertrophic bladder and intestine, nucleated muscle cell profiles are less frequent than in controls (4-5% of all muscle cell profiles, as opposed to 7-8%), and nuclei occupy a lesser proportion of the profile area (about 10%, as opposed to over 30% in controls) and are often eccentrically placed (Gabella and Uvelius 1990). Nuclear transverse sectional area is about twice that in controls, but there is no significant change in nucleus length. Therefore, in hypertrophy the increase in nuclear volume is less than that of the cell, and the nucleus-cytoplasm ratio is much lower than in control muscle cells. Nuclei display a crenate or lobed profile, and many have prominent nucleoli. Binucleate muscle cells (with the two nuclei aligned along the cell length) are not uncommon, whereas they are very rare in normal visceral muscles. Some nuclei become tetraploid, but quantitative studies of this change are not available; tetraploidy is well documented in the pregnant myometrium (Heijden and James 1975) and in hypertensive blood vessels (Owens and Schwartz 1983). Mitosis
In the hypertrophic intestine of the guinea-pig mitoses are observed in both the circular and longitudinal layers (Gabella 1979a) (Figs. 4, 5). The cells undergoing division are differentiated muscle cells, with a full complement of myofilaments and specialized structures such as dense bodies, dense bands, caveolae and cell junctions. The muscle contains no undifferentiated cells, such as myoblasts. Also, there is no evidence that in hypertrophy, muscle cells de-differentiate and then divide, as is usually observed in tissue culture (Chamley-Campbell et al. 1979). In the hypertrophic intestine, dividing muscle cells in the longitudinal muscle are found only in
415 the inner half of the layer; by contrast, in the circular muscle they are predominantly located in the middle portion of the layer, approximately half way between longitudinal muscle and submucosa (Gabella 1979a). The dividing muscle cells display a central portion, ovoid, occupied by chromosomes and at a later stage by the mitotic spindle. On either side of this portion, the cell appears as a fully differentiated muscle cell, indistinguishable from its neighbours. The mitotic equatorial plate lies invariably transverse to the length of the muscle. Brent (1973) observed dividing muscle cells in the hypertrophic large intestine of rabbits; the increase in cell number was much more substantial in immature animals than in adult animals. The occurrence of mitoses has not been recorded
in the hypertrophic bladder, where the increase in cell size may fully account for the muscle enlargement (Figs. 8, 9). However, the total D N A content of the bladder musculature increases nine times after a 6-week obstruction and its concentration decreases only by a factor of 2 (Uvelius et al. 1984). Part of this D N A increase is probably accounted for by enlargement of the nucleus and by polyploidy, as is found in vascular smooth muscle hypertrophy in hypertensive rats (Owens and Schwartz 1983); some of the increase may be the results of mitoses occurring in the early stages of hypertrophy that are not yet documented. Some hypertrophic muscle cells display a nucleus almost completely split into two, the two parts lying side by side and showing no changes in the chromatin pattern
Fig. 10. Photographic montage of the circular muscle layer, in transverse section, and the longitudinal muscle layer, in longitudinal section, from the hypertrophic intestine of a rat. The cell indicated by a r r o w s runs circumferentially for part of its length and longitudinally for another part, and is located in both muscle layers. The insets show details of the same cell at higher magnification. A possible interpretation of this appearance is that the cell
was originally located at the border between longitudinal and circular layers, running in either of them: growth of the tissue and mechanical distortion, such as compression and stretch, pulled a part of the cell with the longitudinal muscle and another part with the circular muscle,j, gap junction; n, nerve bundle. (From Gabella 1987) x 3800
416
(Figs. 13, 14). These appearances suggest the occurrence of cell divisions by a mechanism different from mitosis, that is, a kind of amitotic division (Gabella 1979 a).
along the length of the circular muscle, and reflex contractures arising from the abnormal stimulation of the wall. The tapering ends of the muscle cells are often split into two or more profiles (Fig. 2).
Cell shape Cell surface Many of the hypertrophic muscle cell profiles in bladder and gut are polygonal, crescent-like, very flattened, indented by other cells, or wedge-shaped (Figs. 2, 3), in contrast to the ovoid or simple polygonal shape of control muscle cell profiles. The shape of the profiles suggests that the muscle cells are exposed to an exceptional level of stress (Fig. 10), including compression from the increased content of the lumen, circumferential stretch
Fig. 11. Hypertrophic bladder of rat. Muscle cells, in transverse section, show an irregular surface with numerous, long invaginations, f, collagen fibrils, x 25000
The hypertrophic muscle cells of the intestine and bladder have prominent invaginations of the cell membrane, often radially arranged as wedge-shaped (Fig. 11) or finger-shaped (Figs. 12, 16) inward folds. These invaginations are found at any point along the cell, and seem to be present in some ceils to a greater extent than in others. They reach as far as halfway into the deepest
Fig. 12. Hypertrophic ileum of guinea-pig. The nucleated cell profile (centre) shows five tubular invaginations of the cell membrane; they are lined by a basal lamina, and contain fibrillar material. N, nucleus; n, nerve fibre, x 14500
4t7 point of the cell profile, and their membrane bears dense bands and caveolae and is coated by a basal lamina (Figs. 12, 16). While in some cells these invaginations are evenly distributed around the circumference, in others they occur only over the side facing the edge of a muscle bundle (Fig. 9). The infolding of the cell membrane produces an increase of the cell surface. Geometrically, the surface grows less than the volume, and, in spite of the membrane infoldings, the surface-to-volume ratio falls from 1.4 in control to 0.8 in hypertrophy, i.e. there is 0.8 gm 2 of cell surface per cubic micron of cell volume. The cell membrane is studded with caveolae that are distributed in rows (Fig. 21) between the areas occupied by dense bands (Fig. 20) as in controls, and are similar in shape and size to those of control cells. In intestinal hypertrophic muscle cells, the spatial density of caveolae is marginally smaller than in controls (16.5/gm 2 vs 19.2/ ~m 2) (Gabella 1979a). However, on account of the increased cell size, the number of caveolae per cell must be substantially higher than the control values of about 170 000.
Filaments In the hypertrophic muscle cells of the intestine of rat and guinea-pig, all three types of filament (thin, thick and intermediate) increase in number, on account of the large increase in cell size. Counts of filaments are not available, but thin filaments seem to increase more than thick filaments. Intermediate filaments increase substantially more than either type of myofilament and the cells display large bundles of them (Fig. 15) (Gabella 1979 b). In the hypertrophic bladder, in contrast, changes in the relative proportion of filaments are not readily seen (Gabella and Uvelius 1990) (Fig. 16). Biochemically, however, there is no change in the concentration of actin and a modest fall in the concentration of myosin; the concentration of intermediate filament proteins (90% desmin and 10% vimentin) increases (Uvelius etal. 1989). Active stress of the muscle (force generated/unit transverse area) is unchanged in the hypertrophic ureter of the rabbit (Haussman et al. 1979), whereas in the guinea-pig intestine it is reduced (Gabella 1979 b). While the force generated by the hypertrophic intestinal muscle is much increased, on account of its increased mass, the force generated per unit area is less than in controls. A similar fall in the 'efficiency' in hypertrophy is found also in other muscles (e.g. the portal vein [Johansson 1976]), but the reasons for this change are not yet clear.
Cell organelles A morphometric study of hypertrophic muscle cells of the guinea-pig intestine revealed a halving of the percentage volume of mitochondria from about 7% in the control muscle. Even if the percentage volume decreases, the total mitochondrial volume per cell is about twice
as large as in controls (Gabella 1979a). In the rat intestine, the relative decrease of mitochondria is less marked, and in the rat bladder it is barely noticeable. Decrease in spatial density of mitochondria is a characteristic of most forms of cardiac hypertrophy (Anversa et al. 1971 ; Page et al. 1972) and of the so-called "overload" hypertrophy of skeletal muscles (as opposed to "endurance" hypertrophy, which has the opposite effect on mitochondria) (Goldspink 1971 ; Hoppeter et al. t973). Various forms of sarcoplasmic reticulum are found in smooth muscle cells, probably representing different functional specializations. In hypertrophic muscle cells, sacs and tubules of smooth sarcoptasmic reticulum are more abundant than in controls (Gabella 1979b) (Figs. 15, 20). Fenestrated cisternae of smooth reticulum of great extension and complex shape are common near the surface of intestinal hypertrophic muscle cells (Fig. 21). In control muscle, sarcoplasmic reticulum has been implicated in calcium storage and release (Johansson and Somlyo 1980). It is possible that the expanded sarcoplasmic reticulum of hypertrophic cells, which have a reduced surface-to-volume ratio, helps to maintain adequate calcium supply to the contractile apparatus in the presence of a relative reduction of calcium entry through the cell membrane. In many hypertrophic cells, but not in all of them, at least judging from single transverse sections, there is substantial increase in rough sarcoplasmic reticulum, often in the form of radially arranged cisternae that approach with one edge the cell membrane. Some cisternae are expanded by an amorphous content of medium electron density. The evidence suggests an activation of the synthetic and secretory activity of the muscle cell.
Gap junctions The intestinal musculature (circular layer) is well provided with gap junctions (Fig. 20). With hypertrophy the number of gap junctions per cell unit surface does not change (about 46 junctions per 1000 gm 2, in the guinea-pig); there is, therefore, an increase in the number of junctions per cell, proportional to the cell surface increase. Moreover, the average size of junctions increases (Fig. 19) and the percentage of cell membrane surface occupied by gap junctions doubles (from 0.22% to 0.49%) (GabeUa 1979b). Measurements of tissue impedance in the hypertrophic ileum of the cat reveal an increase in intercellular electrical coupling (Bortoff and Sillin 1986). By contrast, gap junctions are very rare in the longitudinal musculature of the gut, and they remain so in hypertrophy. Gap junctions were not found in the urinary bladder, either in control or in hypertrophy (Gabella and Uvelius 1990).
Nerves In the hypertrophic intestine, intramuscular nerves grow less than the muscle, and nerve bundles are sparser than in controls. There is no loss of nerve fibres; the pre-
418
419
;-.~
o~
\ ~OL)_
