E V A G I N A T I O N O F VASCULAR SMOOTH MUSCLE CELLS D U R I N G THE EARLY STAGES O F C R O T A L A R I A PULM O N A R Y HYPERTENSION PAULSMITHAND DONALDHEATH The Department of Pathology, University of Liverpool
PLATESCVII-CXII THEadministration of the seeds of CrotaIaria spectabilis is a reliable and rapid method of inducing pulmonary hypertension and hypertensive pulmonary vascular disease in rats. We have used this method for short periods of time so that we might determine by electron microscopy which part of the pulmonary vasculature reacts initially to poisoning by the pyrrolizidine alkaloids contained within the seeds. This proved to be the smooth muscle cells of the pulmonary veins which showed striking ultrastructural changes. We describe our findings in this paper. MATEFUALS AND METHODS Eighteen adult female Wistar albino rats weighing between 215-275 g were divided into a group of 15 test animals and a group of 3 controls. The animals were housed singly and given free access to food and water. The animals were fed a powdered diet which, in the case of the test rats, was thoroughly mixed with powdered Crotalaria spectabilis seeds at a concentration of 0.1 per cent. by weight. Three test rats were killed with ether at weekly intervals giving a maximum period on the diet of 5 wk. A control rat was killed on the second, fourth and fifth wk. The survival of the various animals is shown in the table. Immediately after death thoracotomy was performed and the thoracic organs removed intact. Thin slices of lung were removed from each lobe of the right lung, 5 mm from the hilum in each instance. The slices were laid on a tile and covered with a pool of chilled buffered 6 per cent. glutaraldehyde. Using a razor blade the relatively avascular areas of the lung periphery were trimmed off. The remainder of the slice was cut into 1 mm cubes and stored in buffered glutaraldehyde at 4°C overnight. The right main bronchus was ligated and the left lung distended via the trachea with 10 per cent. buffered formalin until the pleural surface was smooth. After fixation blocks of lung were taken for histology and sections 5 pm in thickness were stained with Miller’s Elastic/Van Gieson method. The heart was also dissected free from its attached vessels and atria and the free wall of the right ventricle carefully removed. The weights of the two cardiac ventricles were determined t o confirm whether or not hypertrophy of the right ventricle had occurred. The weight of the right ventricle (RV) was expressed as an inverse ratio of the weight of the left ventricle plus interventricular septum (LV+ s). Unequivocal right ventricular hypertrophy was considered to have occurred if this ratio was less than 2.0. For electron microscopy, numerous blocks of the glutaraldehyde-fixed tissue were embedded in Araldite to ensure that a representative number of pulmonary blood vessels was available for study. Suitable vessels were selected from semi-thin sections stained with toluidine blue. Approximately 5 fine sections from each rat were examined with an AEI EM 6 B electron microscope. The fine sections were stained with lead citrate. Received 20th Aug. 1977; accepted 6th September 1977. I. PATH.-VOL.
124 (1978)
177
N
PAUL SMITH AND DONALD HEATH
178
RESULTS Cardiac weight The weights of the right and left cardiac ventricles and the ratios of LV+S/RV are shown in the table. In no case was the ratio of LV+S/RV less than 2.0 indicating that right ventricular hypertrophy had not occurred. The heart ratios of rats T l l , T14 and T15 were slightly lower than those of the other test rats and this was associated with dilatation of their right ventricle. TABLE Survival, body weights and weights of cardiac ventricles in test and control rats Rat No. T1 T2 T3 T4 T5 T6 T7 T8 T9 T 10 T 11 T 12 T 13 T 14 T 15
c1 c 2 c 3
Survival Initial Final in days body wt. (8) body wt. (g) 7 225 240 7 225 240 7 245 250 14 230 235 240 14 .. 260 14 220 215 21 230 230 21 230 230 21 230 230 28 250 250 28 275 275 28 220 225 35 245 245 35 245 225 225 35 215 ~
14 28 35 LV+S RV
250 260 255 = Weight = Weight
260 295 285
LV+S (9)
0513 0494 0.524 0.505 0.570 0417 0.502 0.434 0494 0520 0.575 0.507 0.487 0.556
RV (8)
0.448
0.140 0150 0154 0140 0.195 0.123 0.158 0.138 0 144 0.157 0.210 0.136 0.154 0.243 0.161
0.565 0.576 0.640
0.168 0.156 0190
LV+S __ RV 3.66 3.29 3.40 3.61
2.92 3.39 3.18 3.14 3.43 3.31 2.74 3.73 3.16 2.29 2.78
3.36 3.69 3.37
of left ventricle and interventricular septum in g. of free wall of right ventricle in g.
Light microscopy The lungs of rats killed between the first and fourth week showed no abnormality which could be detected by the light microscope. In animals killed after 5 wk some of the pulmonary arterioles showed early changes of muscularization in which there was a thin discontinuous layer of smooth muscle internal to the original elastic lamina, and internal to this a very thin newly-formed internal elastic lamina. The lung parenchyma contained foci of oedema, fibrin and macrophages within the alveolar spaces. Ultrastructure of the pulmonary vasculature In control rats. Elsewhere we have described the ultrastructure of the various classes of pulmonary blood vessel in the rat (Smith and Heath, 1977). A few features need emphasising here. The small pulmonary veins have an ill-defined, discontinuous internal elastic lamina. In the region of the fibromuscular pads the muscle and endothelium are separated merely by a thin basement membrane (fig. 1).
CROTALARIA PULMONARY HYPERTENSION
179
In veins which had contracted during fixation the elastic laminae assumed an undulated configuration. Small protuberances of the contiguous muscle cells in the media filled out the folds and crevices of the undulations (fig. 1). The cytoplasm of these protuberances had the same electron density as that of the parent muscle cell (fig. 1). The protuberances occurred between attachment points in the sarcolemma to which the actin filaments were attached (fig. 1). In test ruts. A striking change occurred in the pulmonary veins of most of the rats from the first week onwards. At first sight the endothelial cells appeared to be distended by cysts (fig. 2). On closer examination, however, these vesicles proved to be bounded by two membranes (figs 3 and 4). Some had been fortuitously sectioned to reveal that they were in fact evaginations of muscle cells and the connections with the body of the parent cell were unequivocally demonstrated (figs 3 and 4). The evaginations pressed on the undersurface of the endothelial cells distorting them so that they fitted like a cap over the protruding muscular vesicle. The double membrane referred to above consisted of the limiting membrane of muscle on the inner surface, limiting plasma membranes of endothelial cell on the outside, and interstitial space between the two. Vesicles which appeared to be free within the endothelial cells were without doubt similar evaginations of muscle which had been sectioned distal to their point of attachment to their parent cell. The evaginations of smooth muscle arose between attachment points on the sarcolemmal membrane of the parent muscle cell to which the actin filaments are attached (figs 3 and 4). The contents of the evaginations were electron-lucent and appeared to contain pale, flocculent material. No myofilaments or organelles were seen within them and in many instances the evaginations appeared almost empty (fig. 4). The line of demarcation between the dense cytoplasm of the parent muscle cell and the clear contents of the evagination was both sharp and striking (figs 3 and 4). In some instances the contents of the evaginations had the appearance of degenerate cytoplasm. Similar appearances were found in the muscle cells of fibromuscular pads (fig. 5). Occasionally, vesicles were found in endothelial cells which were bounded only by a single membrane (fig. 3). This membrane was continuous with that of the endothelial cell beneath which the vesicle was situated. The cavity of these vesicles was continuous with the basement membrane and they appeared to have been formed by invagination of the base of the cell to produce a deep fold containing the substance of the basement membrane (fig. 3). Pulmonary veins which contained evaginations of smooth muscle were extremely constricted as indicated by the sharply undulating borders of the muscle cells within the media (fig. 2). Nevertheless evaginations of muscle were restricted to the intimal surface. In contrast the muscular pulmonary arteries showed no muscular evaginations. After 4 wk the extensions of muscle into folds in the internal elastic lamina were abnormally pale and devoid of organelles. In some of the arteries of rats killed after 4 wk, the endothelial cells were enlarged and caused considerable occlusion of the arterial lumen. From the fourth week onwards the pulmonary arterioles showed early changes of muscularization such as we have described previously (Smith and
180
PAUL SMITH AND DONALD HEATH
Heath, 1977). These comprised hypertrophy of existing muscle, the presence of immature muscle cells internal to the elastic lamina and the deposition of elastin in the endothelial basement membrane to create a new, thin, internal elastic lamina. In all rats killed after the third week the alveolar capillaries contained numerous, prominent endothelial vesicles (fig. 6). These were large and often completely filled the affected capillary. In two of the rats killed after 5 wk there were foci in which the alveolar capillaries, arterioles and small muscular pulmonary arteries contained platelet thrombi (fig. 7). DISCUSSION The most striking ultrastructural feature of the pulmonary vasculature of rats fed on Crotaluriu spectubilis seeds was the development of evaginations of smooth muscle into the intima of pulmonary veins. Such lesions have been reported before by Dingemans and Wagenvoort (1976) in the pulmonary veins of rats given fulvine, a closely related pyrrolizidine alkaloid found in the plant Crotalariafulva. They are not specific for intoxication by alkaloids, however, since we have seen the same lesion in the intima of pulmonary veins and the pulmonary trunk in hypoxic rats (Smith, Heath and Padula, 1978). Since both hypoxia and Crotuluriu alkaloids affect the pulmonary vasculature by inducing constriction of pulmonary blood vessels and pulmonary hypertension, the likeliest explanation is that the evaginations of muscle are the result of pulmonary vasoconstriction. This conclusion was also reached by Dingemans and Wagenvoort (1976) following their experimental work with fulvine. This view is moreover, suggested by earlier studies in which evaginations of the sarcolemma were described in contracted smooth muscle cells from the stomach of Bufo murinus (Fay and Delise, 1973), and from the small intestine (Lane, 1965). When living, isolated smooth muscle cells contract, numerous protuberances appear on their surfaces (Fay and Delise, 1973). Under the electron microscope these evaginations are seen to be paler than the rest of the cytoplasm and to lack myofilaments. They always develop from the regions of sarcolemma between dense attachment points and are believed to develop by an inwardly directed pull at these sites by the myofilaments, forcing the intervening areas outwards to form evaginations (Kelly and Rice, 1969; Fay and Delise, 1973). Evagination of smooth muscle thus appears to be a normal response to contraction and in blood vessels it occurs most commonly in veins because they have an interrupted internal elastic lamina so that little resistance is offered to the formation of the protuberances. Their formation may be aided by the infolding of the basement membrane into the endothelium which provides a natural pocket into which the evaginations can expand. In the pulmonary trunk the muscular evaginations occur through the numerous small gaps which exist in the internal elastic lamina (Smith, Heath and Padula, 1978). In muscular pulmonary arteries evagination is a rarer phenomenon since the internal elastic lamina is continuous. Dingemans and Wagenvoort (1976)
PLATECVII
PAULSMITH AND DONALD HEATH
crotalaria PULMONARY
HYPERTENSION
FIG.1 .-Pulmonary vein from a control rat. The lumen of the vessel (L) is lined by the thin cytoplasmic extensions of endothelial cells (e) and is separated from the underlying smooth muscle (m) by only a thin basement membrane. Protuberances of the smooth muscle (p) occupy folds and crevices in the undulated basement membrane which has resulted from fixation of the vein in contraction. The protuberances occur between attachment points in the sarcolemma to which the actin filaments are attached (arrows). They have the same electron density as the cytoplasm of the remainder of the muscle cells. Electron micrograph (EM). x 25,000.
PAULSMITH AND DONALD HEATH
PLATECVIII
crotalaria PULMONARY HYPERTENSION
FIG.2.-Small pulmonary vein from a test rat fed on Crotalaria spectabilis seeds for 1 wk. The endothelial cells appear at first sight to be distended by cysts (arrows) but closer examinationreveals that these clear cystic structures are in fact evaginations of muscle cells of the underlying media pressing onto the underface of the endothelial cells. Note the highly folded profile of the smooth muscle cells and the internal elastic lamina indicating that the vessel is constricted. EM. x 7,000.
PAULSMITHAND DONALB HEATH crotalaria PULMONARY
FIG.3.-Detail
PLATECIX HYPERTENSION
of part of the previous figure showing a large evagination of smooth muscle (v) pressing on the undersurface of an endothelial cell (e). Note that the inner membrane of the evagination is continuous with the sarcolemmal membrane of the underlying smooth muscle cell (m). Its outer membrane is continuous with the plasma membrane of the endothelial cell (e). The gap between the membranes is interstitial space. The evagination has occurred between two attachment points on the sarcolemmal membrane (arrows). Pressing onto the undersurface of an adjacent endothelial cell is a large vesicular infolding of the basement membrane (b). EM. x50,OOO.
PAUL SMITH AND DONALD HEATH
PLATECX
crotalaria PULMONARY m E R T e N s I o N
FIG.4.-Pulmonary vein from a test rat after 1 wk. An evagination (v) arises from a muscle cell (m) in the underlying media between two attachment points (arrows). It presses on the underlying endothelial cell (e) which is deformed to fit like a cap over it. The evagination shows an almost total loss of cytoplasm. EM. x 37,500.
FIG.5.-Test rat after 3 wk. Part of a fibro-muscular pad from a pulmonary vein showing bulbous
evagination (v) of smooth muscle cells (m) adjacent to the endothelium. The evaginations show degeneration of their cytoplasm, EM. x 12,500.
PAULSMITHAND DONALD HEATH
PLATECXI
crotalaria PULMONARY HYPERTENSION
FIG.6.-Test rat after 5 wk. An alveolar capillary which is filled by two large endothelial vesicles. Such vesicles are produced by invagination of the tightly-stretched capillary endothelium (arrow) with subsequent distension with oedema fluid. They are attached to the fused basement membrane of the alveolar wall hut this has not been included in the section. EM. x 25,000.
P A U L SMITH AND
DONALD HEATH
PLATECXII
crotalaria PULMONARY HYPERTENSION
FIG.7.-Test
rat after 5 wk. The endothelium (e) and media (m) of a small muscular pulmonary artery is shown to the left of the figure. To the right, the lumen is filled with blood platelets. EM. ~ 7 , 5 0 0 .
FIG.8.-Test
rat after 3 wk. An endothelial cell of a muscular pulmonary artery contains a large discoid body (d). This has been bent into the shape of a horseshoe by constriction of the vessel. The limiting membrane of the discoid body is lined by ribosomes. These cyst-like dilatations of the rough endoplasmic reticulum are a normal feature in the rat. EM. x 25,000.
CROTALARIA PULMONAR Y HYPERTENSION
181
also described outward evaginations of muscle into the adventitia adjacent to the external elastic lamina but we saw this in only one pulmonary arteriole from a test rat. In the present study we saw two forms of muscular evagination. In the control rats these consisted merely of small protuberances which extended into folds in the endothelium of pulmonary veins. Such protuberances are a common feature of normal pulmonary blood vessels, and occur in both the intima of veins and adventitia of arteries (Smith, Heath and Mooi, 1978). They have a cytoplasmic density similar to or paler than the cytoplasm of the parent cell. However, only in rare cases are they large and cystic (Smith, Heath and Mooi, 1978). These protuberances are produced by evagination of the smooth muscle cells as they shorten when the lung collapses at thoracotomy. They may be aided by a mild degree of muscular contraction caused by charged particles in the fixative solution When lungs are fixed in distension, evaginations are absent (Dingemans and Wagenvoort, 1976; Smith, Heath and Mooi, 1978). By contrast, the evaginations in test rats consisted of numerous, large, bulbous prominences with a strikingly electron-lucent cytoplasm. Such features are not produced merely by collapse of vessels but by active constriction of vascular smooth muscle. They are not abolished by distending the lung with fixative (Dingemans and Wagenvoort, 1976). Furthermore, acute pulmonary vasoconstriction in guinea pigs induced by histamine produces small protuberances similar to those in control rats except that they occur in distended lungs (Dingemans and Wagenvoort, 1976). It seems likely that the pale evaginations which we found in test rats are brought about by a sustained and severe vasoconstriction. In fact, so pronounced are the changes that the evaginated part of the muscle appears to undergo degeneration and comes to resemble a cyst (fig. 2). Care should, therefore, be exercised before identifying cystic spaces in endothelium as vacuoles. The cystic evagination of smooth muscle should not be confused with the ‘‘ discoid body ” (Weibel and Palade, 1964; Smith and Heath, 1977) which consists of dilated rough endoplasmic reticulum (fig. 8) and is a normal feature of pulmonary vascular endothelium in the rat. Although the phenomenon of evagination of constricting smooth muscle has been known to physiologists for some years, it has not been so widely recognised by pathologists. So far as we are aware the effects and fate of these cysts of degenerate muscle cytoplasm in the intima are unknown. It is conceivable that this process of muscular evagination may be of some significance in the pathogenesis of hypertensive pulmonary vascular disease. So far as the induction of pulmonary arterial hypertension by pyrrolizidine alkaloids is concerned, the muscular evaginations in the present study occurred in the absence of right ventricular hypertrophy or any significant muscularization of pulmonary arterioles. This supports the contention of Wagenvoort, Wagenvoort and Dijk (1974) and Dingemans and Wagenvoort (1976) that one of the earliest effects of these alkaloids is to induce constriction of the pulmonary veins. We observed numerous endothelial vesicles in alveolar capillaries after only
182
PAUL SMITH AND DONALD HEATH
3 wk. There has been considerable argument as to whether these lesions cause pulmonary hypertension by obstructing the capillary bed (Valdivia, Lalich, Hayashi and Sonnad, 1967) or whether they are part of the exudative changes secondary to pulmonary hypertension (Kay, Smith and Heath, 1969). Their presence in capillaries before the development of right ventricular hypertrophy, medial hypertrophy of pulmonary arteries or muscularization of arterioles is consistent with their involvement in the pathogenesis of pulmonary arterial hypertension. Equally well, however, their development in pulmonary capillaries could be the result of increased resistance to flow in the pulmonary veins due to the constriction for which the present study provided ultrastructural evidence. SUMMARY
Fifteen adult female Wistar albino rats were fed on a diet containing powdered Crotalaria spectabilis seeds for periods of up to 5 wk. Electron microscopic studies were carried out on the lungs of these animals and also of three control rats. Both groups of animals showed protuberances of smooth muscle cells. In the control rats such protuberances were small and filled out spaces created by undulation of the internal elastic lamina produced by collapse of the vessel. These protuberances could be prevented by fixing the lung in distension. Evaginations of smooth muscle cells in the test rats were larger, devoid of myofilaments and organelles and arose from the parent cell between dense attachment points on the sarcolemma. Frequently they arose through a narrow cytoplasmic isthmus and had such electron-lucent contents as to resemble a cyst within the endothelium. In fact they pressed onto the undersurface of endothelial cells which fitted over them like a cap. Such evaginations are thought to arise as a result of sustained vasoconstriction. This work was carried out with the aid of a grant from the British Heart Foundation.
REFERENCES K. P., AND WAGENVOORT, C. A. 1976. Ultrastructural study of contraction of DINGEMANS, pulmonary vascular smooth muscle cells. Lab. Invest.,35, 205. FAY, F. S., AND DELISE, C. M. 1973. Contraction of isolated smooth-musclecells-Structural changes. Proc. Nat. Acad. Sci. U.S.A., 70, 641. Kay, J. M., SMITH,P., AND HEATH, D. 1969. Electron microscopy of Crotalaria pulmonary hypertension. Thorax, 24, 51 1, KELLY, R. E., AND RICE,R. V. 1969. Ultrastructural studies on the contractile mechanism of smooth muscle. J. Cell Biol., 42, 683. LANE,B. P. 1965. Alterations in the cytologic detail of intestinal smooth muscle cells in various stages of contraction. J. Cell Biol., 27, 199. SMITH,P., AND HEATH, D. 1977. Ultrastructure of hypoxic hypertensive pulmonary vascular disease. J. Path., 121, 93. SMITH, P., HEATH,D., AND MOOI,W. 1978. Observations on some ultrastructural features of normal pulmonary blood vessels in collapsed and distended lungs. J. Anat. (in press). S m , P., HEATH,D., AND PADULA, F. 1977. Evagination of smooth muscle cells in the hypoxic pulmonary trunk. Thorax (in press).
CROTALARIA PULMONAR Y HYPERTENSION
183
VALDIVIA, E., LALICH, J. J., HAYASHI, Y . , AND SONNAD, Y . 1967. Alterations in pulmonary alveoli after a single injection of monocrotaline. Arch. Path., 84, 64. WAGENVOORT, C. A., WAGENVOORT, N., AND DIJK, H. J. 1974. Effect of fulvine on pulmonary arteries and veins of the rat. Thorax, 29, 522. WEIBEL, E. R., AND PALADE, G. E. 1964. New cytoplasmic components in arterial endothelia. J. Cell Biol., 23, 101.
PLATESCVII-CXII THEadministration of the seeds of CrotaIaria spectabilis is a reliable and rapid method of inducing pulmonary hypertension and hypertensive pulmonary vascular disease in rats. We have used this method for short periods of time so that we might determine by electron microscopy which part of the pulmonary vasculature reacts initially to poisoning by the pyrrolizidine alkaloids contained within the seeds. This proved to be the smooth muscle cells of the pulmonary veins which showed striking ultrastructural changes. We describe our findings in this paper. MATEFUALS AND METHODS Eighteen adult female Wistar albino rats weighing between 215-275 g were divided into a group of 15 test animals and a group of 3 controls. The animals were housed singly and given free access to food and water. The animals were fed a powdered diet which, in the case of the test rats, was thoroughly mixed with powdered Crotalaria spectabilis seeds at a concentration of 0.1 per cent. by weight. Three test rats were killed with ether at weekly intervals giving a maximum period on the diet of 5 wk. A control rat was killed on the second, fourth and fifth wk. The survival of the various animals is shown in the table. Immediately after death thoracotomy was performed and the thoracic organs removed intact. Thin slices of lung were removed from each lobe of the right lung, 5 mm from the hilum in each instance. The slices were laid on a tile and covered with a pool of chilled buffered 6 per cent. glutaraldehyde. Using a razor blade the relatively avascular areas of the lung periphery were trimmed off. The remainder of the slice was cut into 1 mm cubes and stored in buffered glutaraldehyde at 4°C overnight. The right main bronchus was ligated and the left lung distended via the trachea with 10 per cent. buffered formalin until the pleural surface was smooth. After fixation blocks of lung were taken for histology and sections 5 pm in thickness were stained with Miller’s Elastic/Van Gieson method. The heart was also dissected free from its attached vessels and atria and the free wall of the right ventricle carefully removed. The weights of the two cardiac ventricles were determined t o confirm whether or not hypertrophy of the right ventricle had occurred. The weight of the right ventricle (RV) was expressed as an inverse ratio of the weight of the left ventricle plus interventricular septum (LV+ s). Unequivocal right ventricular hypertrophy was considered to have occurred if this ratio was less than 2.0. For electron microscopy, numerous blocks of the glutaraldehyde-fixed tissue were embedded in Araldite to ensure that a representative number of pulmonary blood vessels was available for study. Suitable vessels were selected from semi-thin sections stained with toluidine blue. Approximately 5 fine sections from each rat were examined with an AEI EM 6 B electron microscope. The fine sections were stained with lead citrate. Received 20th Aug. 1977; accepted 6th September 1977. I. PATH.-VOL.
124 (1978)
177
N
PAUL SMITH AND DONALD HEATH
178
RESULTS Cardiac weight The weights of the right and left cardiac ventricles and the ratios of LV+S/RV are shown in the table. In no case was the ratio of LV+S/RV less than 2.0 indicating that right ventricular hypertrophy had not occurred. The heart ratios of rats T l l , T14 and T15 were slightly lower than those of the other test rats and this was associated with dilatation of their right ventricle. TABLE Survival, body weights and weights of cardiac ventricles in test and control rats Rat No. T1 T2 T3 T4 T5 T6 T7 T8 T9 T 10 T 11 T 12 T 13 T 14 T 15
c1 c 2 c 3
Survival Initial Final in days body wt. (8) body wt. (g) 7 225 240 7 225 240 7 245 250 14 230 235 240 14 .. 260 14 220 215 21 230 230 21 230 230 21 230 230 28 250 250 28 275 275 28 220 225 35 245 245 35 245 225 225 35 215 ~
14 28 35 LV+S RV
250 260 255 = Weight = Weight
260 295 285
LV+S (9)
0513 0494 0.524 0.505 0.570 0417 0.502 0.434 0494 0520 0.575 0.507 0.487 0.556
RV (8)
0.448
0.140 0150 0154 0140 0.195 0.123 0.158 0.138 0 144 0.157 0.210 0.136 0.154 0.243 0.161
0.565 0.576 0.640
0.168 0.156 0190
LV+S __ RV 3.66 3.29 3.40 3.61
2.92 3.39 3.18 3.14 3.43 3.31 2.74 3.73 3.16 2.29 2.78
3.36 3.69 3.37
of left ventricle and interventricular septum in g. of free wall of right ventricle in g.
Light microscopy The lungs of rats killed between the first and fourth week showed no abnormality which could be detected by the light microscope. In animals killed after 5 wk some of the pulmonary arterioles showed early changes of muscularization in which there was a thin discontinuous layer of smooth muscle internal to the original elastic lamina, and internal to this a very thin newly-formed internal elastic lamina. The lung parenchyma contained foci of oedema, fibrin and macrophages within the alveolar spaces. Ultrastructure of the pulmonary vasculature In control rats. Elsewhere we have described the ultrastructure of the various classes of pulmonary blood vessel in the rat (Smith and Heath, 1977). A few features need emphasising here. The small pulmonary veins have an ill-defined, discontinuous internal elastic lamina. In the region of the fibromuscular pads the muscle and endothelium are separated merely by a thin basement membrane (fig. 1).
CROTALARIA PULMONARY HYPERTENSION
179
In veins which had contracted during fixation the elastic laminae assumed an undulated configuration. Small protuberances of the contiguous muscle cells in the media filled out the folds and crevices of the undulations (fig. 1). The cytoplasm of these protuberances had the same electron density as that of the parent muscle cell (fig. 1). The protuberances occurred between attachment points in the sarcolemma to which the actin filaments were attached (fig. 1). In test ruts. A striking change occurred in the pulmonary veins of most of the rats from the first week onwards. At first sight the endothelial cells appeared to be distended by cysts (fig. 2). On closer examination, however, these vesicles proved to be bounded by two membranes (figs 3 and 4). Some had been fortuitously sectioned to reveal that they were in fact evaginations of muscle cells and the connections with the body of the parent cell were unequivocally demonstrated (figs 3 and 4). The evaginations pressed on the undersurface of the endothelial cells distorting them so that they fitted like a cap over the protruding muscular vesicle. The double membrane referred to above consisted of the limiting membrane of muscle on the inner surface, limiting plasma membranes of endothelial cell on the outside, and interstitial space between the two. Vesicles which appeared to be free within the endothelial cells were without doubt similar evaginations of muscle which had been sectioned distal to their point of attachment to their parent cell. The evaginations of smooth muscle arose between attachment points on the sarcolemmal membrane of the parent muscle cell to which the actin filaments are attached (figs 3 and 4). The contents of the evaginations were electron-lucent and appeared to contain pale, flocculent material. No myofilaments or organelles were seen within them and in many instances the evaginations appeared almost empty (fig. 4). The line of demarcation between the dense cytoplasm of the parent muscle cell and the clear contents of the evagination was both sharp and striking (figs 3 and 4). In some instances the contents of the evaginations had the appearance of degenerate cytoplasm. Similar appearances were found in the muscle cells of fibromuscular pads (fig. 5). Occasionally, vesicles were found in endothelial cells which were bounded only by a single membrane (fig. 3). This membrane was continuous with that of the endothelial cell beneath which the vesicle was situated. The cavity of these vesicles was continuous with the basement membrane and they appeared to have been formed by invagination of the base of the cell to produce a deep fold containing the substance of the basement membrane (fig. 3). Pulmonary veins which contained evaginations of smooth muscle were extremely constricted as indicated by the sharply undulating borders of the muscle cells within the media (fig. 2). Nevertheless evaginations of muscle were restricted to the intimal surface. In contrast the muscular pulmonary arteries showed no muscular evaginations. After 4 wk the extensions of muscle into folds in the internal elastic lamina were abnormally pale and devoid of organelles. In some of the arteries of rats killed after 4 wk, the endothelial cells were enlarged and caused considerable occlusion of the arterial lumen. From the fourth week onwards the pulmonary arterioles showed early changes of muscularization such as we have described previously (Smith and
180
PAUL SMITH AND DONALD HEATH
Heath, 1977). These comprised hypertrophy of existing muscle, the presence of immature muscle cells internal to the elastic lamina and the deposition of elastin in the endothelial basement membrane to create a new, thin, internal elastic lamina. In all rats killed after the third week the alveolar capillaries contained numerous, prominent endothelial vesicles (fig. 6). These were large and often completely filled the affected capillary. In two of the rats killed after 5 wk there were foci in which the alveolar capillaries, arterioles and small muscular pulmonary arteries contained platelet thrombi (fig. 7). DISCUSSION The most striking ultrastructural feature of the pulmonary vasculature of rats fed on Crotaluriu spectubilis seeds was the development of evaginations of smooth muscle into the intima of pulmonary veins. Such lesions have been reported before by Dingemans and Wagenvoort (1976) in the pulmonary veins of rats given fulvine, a closely related pyrrolizidine alkaloid found in the plant Crotalariafulva. They are not specific for intoxication by alkaloids, however, since we have seen the same lesion in the intima of pulmonary veins and the pulmonary trunk in hypoxic rats (Smith, Heath and Padula, 1978). Since both hypoxia and Crotuluriu alkaloids affect the pulmonary vasculature by inducing constriction of pulmonary blood vessels and pulmonary hypertension, the likeliest explanation is that the evaginations of muscle are the result of pulmonary vasoconstriction. This conclusion was also reached by Dingemans and Wagenvoort (1976) following their experimental work with fulvine. This view is moreover, suggested by earlier studies in which evaginations of the sarcolemma were described in contracted smooth muscle cells from the stomach of Bufo murinus (Fay and Delise, 1973), and from the small intestine (Lane, 1965). When living, isolated smooth muscle cells contract, numerous protuberances appear on their surfaces (Fay and Delise, 1973). Under the electron microscope these evaginations are seen to be paler than the rest of the cytoplasm and to lack myofilaments. They always develop from the regions of sarcolemma between dense attachment points and are believed to develop by an inwardly directed pull at these sites by the myofilaments, forcing the intervening areas outwards to form evaginations (Kelly and Rice, 1969; Fay and Delise, 1973). Evagination of smooth muscle thus appears to be a normal response to contraction and in blood vessels it occurs most commonly in veins because they have an interrupted internal elastic lamina so that little resistance is offered to the formation of the protuberances. Their formation may be aided by the infolding of the basement membrane into the endothelium which provides a natural pocket into which the evaginations can expand. In the pulmonary trunk the muscular evaginations occur through the numerous small gaps which exist in the internal elastic lamina (Smith, Heath and Padula, 1978). In muscular pulmonary arteries evagination is a rarer phenomenon since the internal elastic lamina is continuous. Dingemans and Wagenvoort (1976)
PLATECVII
PAULSMITH AND DONALD HEATH
crotalaria PULMONARY
HYPERTENSION
FIG.1 .-Pulmonary vein from a control rat. The lumen of the vessel (L) is lined by the thin cytoplasmic extensions of endothelial cells (e) and is separated from the underlying smooth muscle (m) by only a thin basement membrane. Protuberances of the smooth muscle (p) occupy folds and crevices in the undulated basement membrane which has resulted from fixation of the vein in contraction. The protuberances occur between attachment points in the sarcolemma to which the actin filaments are attached (arrows). They have the same electron density as the cytoplasm of the remainder of the muscle cells. Electron micrograph (EM). x 25,000.
PAULSMITH AND DONALD HEATH
PLATECVIII
crotalaria PULMONARY HYPERTENSION
FIG.2.-Small pulmonary vein from a test rat fed on Crotalaria spectabilis seeds for 1 wk. The endothelial cells appear at first sight to be distended by cysts (arrows) but closer examinationreveals that these clear cystic structures are in fact evaginations of muscle cells of the underlying media pressing onto the underface of the endothelial cells. Note the highly folded profile of the smooth muscle cells and the internal elastic lamina indicating that the vessel is constricted. EM. x 7,000.
PAULSMITHAND DONALB HEATH crotalaria PULMONARY
FIG.3.-Detail
PLATECIX HYPERTENSION
of part of the previous figure showing a large evagination of smooth muscle (v) pressing on the undersurface of an endothelial cell (e). Note that the inner membrane of the evagination is continuous with the sarcolemmal membrane of the underlying smooth muscle cell (m). Its outer membrane is continuous with the plasma membrane of the endothelial cell (e). The gap between the membranes is interstitial space. The evagination has occurred between two attachment points on the sarcolemmal membrane (arrows). Pressing onto the undersurface of an adjacent endothelial cell is a large vesicular infolding of the basement membrane (b). EM. x50,OOO.
PAUL SMITH AND DONALD HEATH
PLATECX
crotalaria PULMONARY m E R T e N s I o N
FIG.4.-Pulmonary vein from a test rat after 1 wk. An evagination (v) arises from a muscle cell (m) in the underlying media between two attachment points (arrows). It presses on the underlying endothelial cell (e) which is deformed to fit like a cap over it. The evagination shows an almost total loss of cytoplasm. EM. x 37,500.
FIG.5.-Test rat after 3 wk. Part of a fibro-muscular pad from a pulmonary vein showing bulbous
evagination (v) of smooth muscle cells (m) adjacent to the endothelium. The evaginations show degeneration of their cytoplasm, EM. x 12,500.
PAULSMITHAND DONALD HEATH
PLATECXI
crotalaria PULMONARY HYPERTENSION
FIG.6.-Test rat after 5 wk. An alveolar capillary which is filled by two large endothelial vesicles. Such vesicles are produced by invagination of the tightly-stretched capillary endothelium (arrow) with subsequent distension with oedema fluid. They are attached to the fused basement membrane of the alveolar wall hut this has not been included in the section. EM. x 25,000.
P A U L SMITH AND
DONALD HEATH
PLATECXII
crotalaria PULMONARY HYPERTENSION
FIG.7.-Test
rat after 5 wk. The endothelium (e) and media (m) of a small muscular pulmonary artery is shown to the left of the figure. To the right, the lumen is filled with blood platelets. EM. ~ 7 , 5 0 0 .
FIG.8.-Test
rat after 3 wk. An endothelial cell of a muscular pulmonary artery contains a large discoid body (d). This has been bent into the shape of a horseshoe by constriction of the vessel. The limiting membrane of the discoid body is lined by ribosomes. These cyst-like dilatations of the rough endoplasmic reticulum are a normal feature in the rat. EM. x 25,000.
CROTALARIA PULMONAR Y HYPERTENSION
181
also described outward evaginations of muscle into the adventitia adjacent to the external elastic lamina but we saw this in only one pulmonary arteriole from a test rat. In the present study we saw two forms of muscular evagination. In the control rats these consisted merely of small protuberances which extended into folds in the endothelium of pulmonary veins. Such protuberances are a common feature of normal pulmonary blood vessels, and occur in both the intima of veins and adventitia of arteries (Smith, Heath and Mooi, 1978). They have a cytoplasmic density similar to or paler than the cytoplasm of the parent cell. However, only in rare cases are they large and cystic (Smith, Heath and Mooi, 1978). These protuberances are produced by evagination of the smooth muscle cells as they shorten when the lung collapses at thoracotomy. They may be aided by a mild degree of muscular contraction caused by charged particles in the fixative solution When lungs are fixed in distension, evaginations are absent (Dingemans and Wagenvoort, 1976; Smith, Heath and Mooi, 1978). By contrast, the evaginations in test rats consisted of numerous, large, bulbous prominences with a strikingly electron-lucent cytoplasm. Such features are not produced merely by collapse of vessels but by active constriction of vascular smooth muscle. They are not abolished by distending the lung with fixative (Dingemans and Wagenvoort, 1976). Furthermore, acute pulmonary vasoconstriction in guinea pigs induced by histamine produces small protuberances similar to those in control rats except that they occur in distended lungs (Dingemans and Wagenvoort, 1976). It seems likely that the pale evaginations which we found in test rats are brought about by a sustained and severe vasoconstriction. In fact, so pronounced are the changes that the evaginated part of the muscle appears to undergo degeneration and comes to resemble a cyst (fig. 2). Care should, therefore, be exercised before identifying cystic spaces in endothelium as vacuoles. The cystic evagination of smooth muscle should not be confused with the ‘‘ discoid body ” (Weibel and Palade, 1964; Smith and Heath, 1977) which consists of dilated rough endoplasmic reticulum (fig. 8) and is a normal feature of pulmonary vascular endothelium in the rat. Although the phenomenon of evagination of constricting smooth muscle has been known to physiologists for some years, it has not been so widely recognised by pathologists. So far as we are aware the effects and fate of these cysts of degenerate muscle cytoplasm in the intima are unknown. It is conceivable that this process of muscular evagination may be of some significance in the pathogenesis of hypertensive pulmonary vascular disease. So far as the induction of pulmonary arterial hypertension by pyrrolizidine alkaloids is concerned, the muscular evaginations in the present study occurred in the absence of right ventricular hypertrophy or any significant muscularization of pulmonary arterioles. This supports the contention of Wagenvoort, Wagenvoort and Dijk (1974) and Dingemans and Wagenvoort (1976) that one of the earliest effects of these alkaloids is to induce constriction of the pulmonary veins. We observed numerous endothelial vesicles in alveolar capillaries after only
182
PAUL SMITH AND DONALD HEATH
3 wk. There has been considerable argument as to whether these lesions cause pulmonary hypertension by obstructing the capillary bed (Valdivia, Lalich, Hayashi and Sonnad, 1967) or whether they are part of the exudative changes secondary to pulmonary hypertension (Kay, Smith and Heath, 1969). Their presence in capillaries before the development of right ventricular hypertrophy, medial hypertrophy of pulmonary arteries or muscularization of arterioles is consistent with their involvement in the pathogenesis of pulmonary arterial hypertension. Equally well, however, their development in pulmonary capillaries could be the result of increased resistance to flow in the pulmonary veins due to the constriction for which the present study provided ultrastructural evidence. SUMMARY
Fifteen adult female Wistar albino rats were fed on a diet containing powdered Crotalaria spectabilis seeds for periods of up to 5 wk. Electron microscopic studies were carried out on the lungs of these animals and also of three control rats. Both groups of animals showed protuberances of smooth muscle cells. In the control rats such protuberances were small and filled out spaces created by undulation of the internal elastic lamina produced by collapse of the vessel. These protuberances could be prevented by fixing the lung in distension. Evaginations of smooth muscle cells in the test rats were larger, devoid of myofilaments and organelles and arose from the parent cell between dense attachment points on the sarcolemma. Frequently they arose through a narrow cytoplasmic isthmus and had such electron-lucent contents as to resemble a cyst within the endothelium. In fact they pressed onto the undersurface of endothelial cells which fitted over them like a cap. Such evaginations are thought to arise as a result of sustained vasoconstriction. This work was carried out with the aid of a grant from the British Heart Foundation.
REFERENCES K. P., AND WAGENVOORT, C. A. 1976. Ultrastructural study of contraction of DINGEMANS, pulmonary vascular smooth muscle cells. Lab. Invest.,35, 205. FAY, F. S., AND DELISE, C. M. 1973. Contraction of isolated smooth-musclecells-Structural changes. Proc. Nat. Acad. Sci. U.S.A., 70, 641. Kay, J. M., SMITH,P., AND HEATH, D. 1969. Electron microscopy of Crotalaria pulmonary hypertension. Thorax, 24, 51 1, KELLY, R. E., AND RICE,R. V. 1969. Ultrastructural studies on the contractile mechanism of smooth muscle. J. Cell Biol., 42, 683. LANE,B. P. 1965. Alterations in the cytologic detail of intestinal smooth muscle cells in various stages of contraction. J. Cell Biol., 27, 199. SMITH,P., AND HEATH, D. 1977. Ultrastructure of hypoxic hypertensive pulmonary vascular disease. J. Path., 121, 93. SMITH, P., HEATH,D., AND MOOI,W. 1978. Observations on some ultrastructural features of normal pulmonary blood vessels in collapsed and distended lungs. J. Anat. (in press). S m , P., HEATH,D., AND PADULA, F. 1977. Evagination of smooth muscle cells in the hypoxic pulmonary trunk. Thorax (in press).
CROTALARIA PULMONAR Y HYPERTENSION
183
VALDIVIA, E., LALICH, J. J., HAYASHI, Y . , AND SONNAD, Y . 1967. Alterations in pulmonary alveoli after a single injection of monocrotaline. Arch. Path., 84, 64. WAGENVOORT, C. A., WAGENVOORT, N., AND DIJK, H. J. 1974. Effect of fulvine on pulmonary arteries and veins of the rat. Thorax, 29, 522. WEIBEL, E. R., AND PALADE, G. E. 1964. New cytoplasmic components in arterial endothelia. J. Cell Biol., 23, 101.
