CELL PROLIFERATION AND COLLAGEN SYNTHESIS IN EXPLANT CULTURES OF ARTERIOSCLEROTIC RABBIT AORTA

Fisher, Douglas Otto. Ph.D., University of Rhode Island. 1979. Cell Proliferation and Collagen Synthesis in Explant Cultures of Arteriosclerotic Rabbit Aorta. Major Professor: Dr. George c. Fuller. The rate of cell proliferation and the ratio of type I/ III collagen s ynthesized by aorta organ culture and cultured smooth muscle cel l s (SMC) was determined in cells derived from rabbits with injury induced (daily epinephrine and thyroxine administration) arteriosclerosis and increased va.scular collagen s ynthesis. Tissue was taken from medial smooth muscle of normal and diseased rabbits and incubated in organ or cell culture systems with 2,33H-proline, ascorbate and· beta-aminopropionitrile. Collagen types were separated by SDS-polyacrylamide electrophoresis and CMC-chromatography and quantitated radiometrically. 1icroscopic observation at 12 days in culture and 3H-thyrnidine in~orporation at 18 days in culture indicated a greater rate of proliferation of SMC from explants of arterioschlerotic tissue compared to control. In organ culture the ratio of collagen type III : type I was 1 : 1 in the control group and 1 : 1.7 for the arteriosclero.tic group . Collagen type III : type I for daughter cells at the end of the 4th passage was 1 : 1.4 and 1 : 2.0 for control and ar~eriosclerotic, respectively . By the 10th passage in culture SMC in both groups were synthesizing almost exclusively type I collagen. This study indicates that cells from arteriosclerotic smooth muscle ere fundamentally unique since the ratio of collagen t ype is controlled by gene expression. This property affecting collagen deposition may play a role in t he pathogenesis of vascular disease.

Arteriosclerosis is a complex disease involving many intrinsic and extrinsic factors which influence the susceptibility of the vasculature to insult and the subsequent clinical sequelae.
In arteriosclerosis a variable combination of metabolic changes seen in the intima and media of arteries is characterized by" .... the focal accumulation of modified smooth muscle cells, lipids, complex carbohydrates, blood and blood products, fibrous tissue and calcium deposits" (Wo rld Health Organization, 1958). These changes predominate in the major arteries and resistance vessels and by compromising blood flow, account for the clinical appearance of heart attack s, strokes, senility and memory loss, angina and impaired peripheral circulation.
The major current theories on the genesis of arterio-scl~rosis share the belief that the lesions begin as localized foci of hyperactive smooth muscle cells in the intima and medial regions of the blood vessels. Authors from this laboratory have reported evidence, using various techniques, which establish the existence of increased collagen synthesis in experimentally induced plaques in animals. The deposition of collagen in the arterial intima is largely responsible for the occlusive and irreversible nature of arteriosclerosis (Benditt, 1977). It is believed that the defect in collagen metabolism which precedes the formation of fibrous vascular plaques is due to a change in the growth characteristic and function of smooth muscle cells, the predominant cell type in the arteriosclerotic intima.
Four genetically distinct types of collagen are currently recognized and, although their distribution is tissue specific, various disease processes and aging can be correlated to altered distribution of collagen type. For example, type I collagen is the predominant extractable collagen species in human fibrous atheromatous plaques whereas type III appears to be the major aortic medial collagen (Mccullagh and Balian, 1975). Little is known about the control of collagen heterogeneity but it could be proposed that a change in collagen gene expression, as a result of smooth muscle cell transformation, may be associated with the development of arteriosclerosis.
This investigation was conducted to examine the hypothesis that the arteriosclerotic lesion in rabbits is associated with a distinct population of smooth muscle cells with altered growth characteristics and function, suggestive of cellular transformation. To examine this hypothesis the growth characteristic of cells from primary cultures of control and arteriosclerotic rabbit aorta were examined. Also, studies were conducted to determine if daughter cells derived from medial tissue explants of arteriosclerotic plaques continue to synthesize a similarly abnormal ratio of type I 3 to type III collagen compared to smooth muscle cells derived from explants of normal aorta. It is believed that these studies are valuable to gaining an understanding of the normal regulation of collagen synthesis and perhaps in screening for useful drugs in the treatment of abnormalities in collagen metabolism which occur in various disease states.

LITERATURE SURVEY
The Vascular Connective Tissue Matrix The intercellular matrix of the arterial wall contains four major types of macromolecules (collagen , glycosaminoglycans, elastin and glycoproteins) . Most chemical studies have been carried out using the aortas of different animal species; only limited information is available on other arteries or veins. Several reviews discuss the chemical and biological characteristics of these macromolecules (Balazs, 1970;Robert, 1970;Slavkin, 1972). Much of the early knowledge on the chemistry and biosynthesis of elastin was gained from experimental work by Partridge (1970Partridge ( , 1972. Several laboratories have studied the composition and biosynthesis of glycosaminoglycans and glycoproteins in normal and pathological arterial walls (Engel, 1971;Kumar et al., 1967;Srinivasan et al., 1971). Considerable interest in arterial collagen has been generated by research which demonstrated its dynamic role in experimental atherosclerosis (Levene, 1962;Smith, 1965;Fuller et al., 1970Fuller et al., , i972, 1973Fuller et al., , 1976Crossley et al., 1972;Ooshima et al., 1974 ) . As a result various laboratories began i so l ating and characterizing aortic collagen genetic types (Chung et al., 1974;Epstein et al., 1975;Trelstad, 1974;Mccullagh and Balian, 1975).
Collagen synthesis occurs in a series of sequential steps consisting of assembly of a praline-rich and lysinerich polypeptide precursor of collagen (procollagen alpha chains} enzymatic hydroxylation of some of the prolyl and lysyl residues and glycosylation of some of the hydroxylysyl residues. The cellular processing of the procollagen alpha chains also includes formation of the triple helix with stability provided by disulfide bonds in the nonhelical regions of the molecule (Bornstein and Ehrlich, 1973).
Following secretion, these nonhelical regions are removed by the extracellular enzyme procollagen peptidase (Lapiere et al., 1971;Goldberg , et al., 1975). The crosslinking of collagen, which imparts stability to helical and .fibrillar collagen, occurs through further extracellular processing 5 by the enzyme lysyl oxidase which oxidatively deaminates specific lysine residues leaving aldehyde moieties which form cross links through condensation reactions (Siegel et al., 1970 The experimental studies also indicated that a determination of total content as well as concentrations is necessary for a correct evaluation of the metabolic alterations of the macromolecules in studies on vascular injury (Helin et al., 1971). In a recent study of canine atherosclerosis the percentage of protein synthesis represented by collagen rose from a mean of less than 5 % in normal aortic branch arteries to 14 % in severely atherosclerotic branch arteries, showing that the increases in collagen synthesis were not associated simply with an overall increase in protein synthesis (McCullagh and Ehrhart 1974). Langner and Fuller (1973), reported increased total collagen and 0.45M NaCl soluble collagen in thoracic aorta of rabbits with epinephrine-thyroxine induced arteriosclerosis. The collagen content of an apparently normal human coronary artery approximates 25 mg/100 mg of dry, defatted tissue (Tamai et al., 1978). In arteriosclerosis the wet weight of a vessel increases and therefore content of vascular components can be misleading when expressed on the basis of organ wet weight (Tammi et al., 1978) . In another study collagen was reported to comprise 30 % of the dry weight of human fibrous atherosclerotic plaques (Levene and Poole, 1962).
Aortic prolyl hydroxylase has been used as a marker of collagen synthesis (Fuller and Langner , 1970;Fuller et al., 1972) and has been shown to increase up to six-fold in an experimental model of arteriosclerosis in rabbits and in aorta of miniature pigs fed a lipid-rich diet. Kinetic properties of this enzyme and the existence of immunologically cross-reacting forms of the prolyl hydroxylase have been investigated for their potential role in regulating the rate of collagen synthesis (Fuller et al., 1976;Ooshima et al., 1974). The Vmax of prolyl hydroxylase in an organ affected by arteriosclerosis increased four-fold, compared to controls, while the Km remained unchanged (Fuller et al., 1976). Various parameters of collagen synthesis including inununologically related protein are elevated in rats with hypertension induced by deoxycorticosterone acetate-salt (Ooshima et al ., 1974).
During the mid-seventies advancements in connective tissue methodology led to the separation and ident ification of three distinct species of collagen (gene products) from human aorta (Trelstad , 1974 (Miller et al.,1 971 ) . There are at least five structural genes involved in the synt hesis of the different types and the transcription of these various genes seems to be tissue specific. Mccullagh and Belian (1975 )  ing the structure and biosynthesis of these substances in the aorta of several species. The methods used in current investigations i n vol v e separation and fractionation by digestion with hyaluronidase and CFC-cellulose microchromatography l Thunnel, 1967 ) . The GAG fractions can be quantitated by measuring hexosamine content (Boas,19 3 3 ) .
Great variations between species hav e been reported for both total GAGs per mg. dry aorta or mg . DNA and in the relative distribution of different GAGs l Engel, 1971 ) , Reports on the changes of these substances in human arteriosclerosis do not completely agree (Helin et al., 1970} (Tammi et al., 1978, However , most agree that chondroitin 4-6 sulfate is the principle GAG in the large .human arteries . The interactions of collagen with glycosaminoglycans has been a subject of investigation for a number of years (Mathews, 19651. By electron microscopy a regularly distri- in rat aorta (Kajikawa , 1970 Glycoproteins have been extracted by several authors from aortic tissue and some of the glycoproteins were purified and characterized ). An elevated level of glycoproteins was reported in arteriosclerotic intima by Hullinger and Manley (1969) , confirming preceding studies by Nakamura (1968) and Schoenbeck et al. (1962) . Srinivasan (1971) (Hoffman, 1971) . It is not within the scope of this literature survey to thoroughly address the most current investigations on the physiochemical properties and biosynthesis of elastin ( see Ross and Bornstein, 1969;Abraham et al., 1975). However, it would be appropriate to stress that the fragmentation-degradation of elastic fibers and laminae is one of the most conspicuous findings of the atherosclerotic changes in aortas. This is observed in the human atherosclerotic plaque and it seems to be independent of the method used to induce experimental arteriosclerosis. Neutrophil granulocytes (Janoff , 1970) and blood platelets were shown to contain proteases with elastolytic activity (Legrand et al., 1973) and it is quite probable that similar enzymes may be isolated from aorta, where the degradation of elastin during atherosclerosis has been demonstrated.

Response of the Vasculature to In~
The repair processes have been proposed to be characteristic features in various vascular diseases, among these arteriosclerosis and vasculitis (Helin et al., 1972). _1echanical strain and the subsequent inflammation has been shmm to be an important initiating factor in the development of aortic arteriosclerosis in rabbits subjected to systemic hypoxia and intravenous injections of catecholamine (Helin and Lorenzen, 1969;Crossley et al., 1972). The investigations by these and other authors have established significant relationships between vascular repair following injury and atherogenesis (Helin et al . , 1971;Hartman, 1977;Kobaysi, 1969). The nature of the repair and the ultimate outcome of the entire process depends, among other factors, upon the characteristics of the injurious agent, duration of exposure to it, the type of initial local manifestations and mural reactions and the status of the host.
Owing to the peculiarities of structure and function, the arterial wall has limited versatility in defense mechanisms and the defense forces are hampered not only by the lack of mural capillaries but also by the fact that the artery is never at rest and thus an important healing-promoting factor is absent.
A number of the established processes in inflammation and repair are operative following vascular injury and have been reviewed in detail (Robbins, 1974). Two tissue reactions are characteristically obse~ved in atherosc~erosis and appear to be unique for this disease process . One is the Evidence that intimal smooth muscle, or a closely related cellular derivative, is involved in the formation of connective tissue during atherosclerotic intimal thickening of elastic and muscular artiers has been reported in both human and experimentally induced atherosclerosis (Haust et al., 1960;Geer and Haust, 1972;Ross and Glomset, 1973 ) It is not clear whether the modified smooth muscle found in the intima orginates in the intima or migrates from the \ media.
At birth most of ~he smooth muscle cells are located in the medial layer of arteries and the i ntirna is comparatively thinner and contains only a few smooth muscle cells .
The internal elastic lamina represents a morphological border between the intima and the media. Fragmentation of the elastic lamina has been shown to be associated with va~ious forms of mechanical injury to the vasculature and would thus permit the migration of medial smooth muscle into the intima (Ross and Glomset, 1973). It has also been suggested that the intact arterial endothelium normally acts as a barrier to some substances present in the plasma which upon exposure to vascular smooth muscle promote cell proliferation (Ross and Glomset, 1973) .

Smooth muscle cells have been reported
to accumulate in the intima at those sites in the arterial vascular bed where endothelial permeability appears to be increased (Helin et al., 1972) . Benditt (1977) proposed that cells comprising atherosclerotic plaques have undergone mutational changes which are analogous to the transformation process occurring in benign tumor cells. This premise is based on the observation that cells within the plaques are monoclonal in origin with respect to glucose-6-phosphate dehydrogenase i soenz yme patterns (Benditt and Benditt, 1973 ) . These cells also have morphological and functional features which are different from normal smooth. mu scle cells in the arterial wall (Somlyo and Somlyo, 1968).

The Smooth Muscle Cell in Culture
Tissue culture, since it was introduced at the turn of this century, has undergone several stages in its evolution.
In its present phase emphasis is placed on the analysis of cellular interaction, cell differentiation, and cell function. Initial experiments have been conducted, establishing the growth characteristics of normal rabbit smooth muscle cells in tissue culture (Ro ss, 1973;Doaud et al., 1964).
A pure population of smooth muscle cells can be isolated from the tunica media of the aorta or large blood vessels of most mammals. Therefore, if the adventitia and intima are stripped off such ves sels, the remaining tunica media is a source of pure smooth muscle cells for culture.
In culture, smooth muscle cells first acquire a fibroblastlike appearance. During this stage many cells degenerate but others proliferate. After a few weeks in culture the daughter cells acquire the appearance of smooth muscle cells (Ross, 1971). Their identity can be easily established by electron microscopy because by then, they again contain considerable amounts of myofilarnents in their cyt~ plasm. The actornyosin in these cells can be demonstrated by immunohistochemical methods (Knieriem et al., 1968 ) .
The similarity of these cells to cells in atherosclerotic le sions make them a potentially important tool for atherogenesis research.
tissue from cholesterol fed rabbits showed better cellular outgrowth, after being placed in culture, than normal tissue.
In contra s t to this Wexler and Thomas (1967) reported that explants from arteriosclerotic breeder rats had a lower growth curve than those from virgin rats . Fritz and co-workers Another study reported that in suitable medium, chick chondroc y tes retain the morphology characteristic of cartilage tissue and synthesize a matrix whose main constituent is chondroitin sulfate . ~hen chick embryo serum was added to the culture medium the cells attained a fibroblastlike morphology and stopped synthesizing chondroitin sulfate (Marzullo and Lash, 1970). These authors proposed that under suboptimal conditions, competition between two kinds of synthesis for limiting levels of energy metabolites may favor cell division over the production of tissue specific mole-c~les.
These tissue specific molecules could be considered luxury molecules which are expendable and non-essential for survival of a cell (Holtzer and Abbot, 1968) . This is con - It is therefore not surprising that somewhat similar variability in the synthesis of collagen types has been repo rted for medial smooth muscle cells in culture (Layman and Titus, 1975;Barnes et al., 1976;Scott et al., 1977).
The difference s in the proportion of type I and III collagens by these cells in culture may be a reflection of cell line variability or differences in growth conditions employed.
Although it may be argued that such in vitro condition s hardly resemble the in vivo situation (Fowler et al., 1977) jt seems apparent that further studies if properly designed, would be helpful in learning more about the regulation of collagen heterogeneity and influences by the extracellular matrix on gene expression. An understanding of those factors controlling collagen synthesis is basic to ou~ understanding of several disease processes including atherosclerosis.
Collagen Heterogeneity: Pathological Implications Changes in the relative proportions of collagen types within a tissue may occur in various disease states (Pope et al . , 1975;Mc c ullagh and Belian, 1975;Seyer et al . , 1976;Pent tinen et al . , 1975) .
In one of these studies the importance of type III collagen was indicqted by the susceptibility of large arteries to rupture in patients with Ehler's Danlos Syndrome (Type IV) who exhibited reduced levels of type III collagen in these tissues (Pope et al., 1975). It has been suggested that the arteriosclerotic plaques represent scar tissue, because a mixture of type I and II collagens with a predominance of type I appears to be typical of wounds .
During fetal development, the proportion of type I and type III collagen molecules forming the dermis shifts from a large proportion of type III to increasing amounts of type I (Epstein, 1974 ) . Concomitant with this are changes in mechanical properties and coarsening of the bundle fibers (Pierard et al., 1976 ) . Such variations in bundle organization observed in vivo occur in the same range of con-  (Lapiere and Pierard, 1977) .
Whatever the mechanism of the interaction of the collagen types, it appears important during development and in pathological conditions. It would explain the resistance to · stretching observed for thick bundles of type I collagen in tendon in contrast to the laxity and the distensibility of the blood vessel wall, which contains a mixture of type I and type III collagens (Lapiere and Pierard, 1977) . Also, Hughes et al. (1976)   Labelled collagen was extracted by first homogenizing the smooth muscle tissue in its 2 ml incubation media using a conical ground-glass homogenizer. A 400 ul portion of this homogenate was taken for DNA determination according to the method of Burton (1956).
The remainder of the sample was dialyzed against 0.5M acetic acid and then treated with pepsin (100 ug/ml ) at 4°C for 12 hours. After limited proteolysis of noncollagenous protein, tissue debris was removed by centrifugation at 30,000 rpm for 20 minutes.
The supernate was then dialyzed for 12 hours, against 0.05M phosphate buffer (pH 7.2) and the v olume adjusted to 1 . 6 ml.
Aliquots of these sam?les could be combined with electrophoresis buffer, heat denatured (56°C x 20 minutes) and the labelled alpha chains separated on 5% polyacrylamide gels according to the method of Neville (1971).

Culturing of Smooth Muscle Cells
A portion of the medial tissue was also used as explant culture to derive colonies of smooth muscle cells for each of the experimental animals. The 60mm Falcon petri di shes, each containing approximatel~ 5 to 10 self-adhering explants from a specific animal, were flooded with 10 ml of DMEM.

Polyacrylamide Gel Electrophoresis
The 5 % polyacrylamide gels and electrophoresis buffers used in this part of the study were prepared by the methods described by Neville (1971) . To a 50 ul. aliquot of each The radioactivity in ~2 was not affected by reduction.

RESULTS
This investigation used an established method of epinehrine infusion with thyroxine administration to produce ~acroscopic fibrous lesions in rabbit aorta (Fuller and La ngner, 1970 ;Mickulicich and Oester, 1970;Lorenzen, 1962) Histopathological changes in the aorta included among other ?rominent features (Fuller et al., 1976): pronounced intimal thickening and the proliferation of a specialized cell type in the media. Vascular tissue for this investigation was obtasined from the medial region of the aorta which is composed predominantly of smooth muscle cells (Ross and Glomsett, 1973;Bierring and Kobayasi, 1963) . There are no fibroblasts present in the media of mammalian arteries in contrast to the arteries of other species such as birds (French, 1966).  ing the absolute number of explants assigned to their appropriate growth categories (e.g . , GOOD, SOME, and NO growth) .
This analysis also indicated significant differences between control and arteriosclerotic in the GOOD and NO growth cate-      ""' I\.) o(-region due to heat denaturation in the presence of 2% 2-mercaptoethanol . Panel (C) in Figure 3 shows loss in radioactivity due to prior incub~tion of the collagen preparation with purified clostridial collagena s e (Peterkofsky & Diegelnann, 1971) .
The increased radioactivity in the Q(.-lregion after reduction with rnercaptoethanol indicates that control rabbit vascular smooth muscle synthesized 50 % type I collagen and 50% type III collagen ( Thoracic medial smoot h muscle explants from normal and arteriosclerotic rabbits were placed in culture as described Aorta medial smooth muscle was incubated for 72 hours at 37°c in 2 ml of DMEM containing 2, 3-3 H-proline, BA.PN, and ascorbate. Collagen types separated on polyacrylamide ge ls and quantitated radiometrically . Values are the mean + S.E. *Significantly different from control (p <. .05) Student "t" test. and ascorbate (400 ug/ml) . Labelled collagen was extracted from the culture medium and DNA determined in the cell layer. Collagen type s I and III were quantitated on gels according to the method previously described . Table 4 shows the relative and total amounts of type I and type III cultures ~5 X 10 cells/75 cm ) at the fourth passage. Incubated with 2,3-II-praline in the presence of BAPN ascorbate and 10% fetal Collagen types separated on polyacrylamide gels and quantitated radio-Values are the mean + S.E. *Significantly different from control (p < .05) Stude nt " t" test .

PASSAGE NUMBER
between age in culture and the relative synthesis of type I and type III collagen. The cells from arteriosclerotic tissue make a larger proportion of type I (vs type III) compared to control at all passage levels. By the tenth passage type III synthesis represents no more than 10% of the collagen synthesized by cells grown from normal or arteriosclerotic tissue.
It was also apparent that by the tenth passage the cells in both groups were layi~g down less matrix compared to earlier passages which allows these cells to be more easily dissociated (enzymatic) upon subculturing, and to grow in confluent monolayers. In contrast, cells at early passages grow on top of one another, resembling hills and valleys, before reaching confluency.

Collagen Heterogeneity by CMC-Chromatography
The electrophroeticseparation of labelled collagen types was developed for this study in order to permit anal y sis of individual cultures in sufficient numbers for statistical comparison . In order to validate and confirm the data from this gel-electrophoresis system, pools w~re prepared from the primary organ culture system and also from the cell culture system to obtain sufficient radioactive material to anal y ze collagen t ypes on carboxymeth y lcellulose as described in the ~ethods.
After loading the reduced col_agen extracts, the collagen was eluted from the CMC with 400 ml of sodium acetate buffer (0.06M, pH 4.8) over a continuous NaCl gradient (0.0 to O.lM) and 10 ml. fractions were collected.
The elution pattern is reflected by the radioactivity in aliquots from each fraction ( Figure 5). Moving from left to right on the chromatogram in Figure 5 the three peaks after the gradient correspond to ~l(I); ~l (III) ; and ~2.
In Figure 5  The ratio of type III: type I collagen was calculated from the total radioactivity found in each peak and is compared with the values obtained by gel electrophoresis (Table 5) . For the organ culture system the ratios of collagen type III:I determined by either method were nearly identical. For the cell culture system both methods indicated an increase in type I collagen, relative to type III, for the arteriosclerotic group; however, the ratios of III:I calculated by CMC-chrornatography were generally higher than those obtained by the electrophoretic method (Table 5) . A significant difference in the results by the two analytical methods appeared only in the control cell culture group . This difference can be accounted for by the function of the medial smooth muscle (Ross and Glomset, 1973).
It has been proposed that medial smooth cells, responding to intimal injury, migrate into the intirna and begin to proliferate establi s hing a colony rese~bling a benign neoplasm which synthesize a generous amol1.nt of collagen (Benditt, 1977). McCullagh and Belian (1975)  a marked increase in the ratio of type I to type III collagen, compared to that in the nonarteriosclerotic aortic media .
In their procedure, less than 50% of hlli~an aorta collagen was recovered in the extracts and subsequently characterized.
No studies have reported the relative rate of synthesis of type I and III collagen in diseased vascular tissue.
The data described in this investigation establishes that a significantly greater population of smooth muscle cells migrate out of the medial explants dissected from tissue subjected to an atherogenic condition compared to medial explants from the nonarteriosclerotic group. This enhanced outgrowth from tissue explants supports the current hypothesis that in atherogenesis modified smooth muscle cells in the medial layer migrate from the media, through the internal elastic lamella and into the intimal region.
Conditioned media from rapidly growing medial explants when transferred to slower growing explants did not stimulate migration of cells from slower growing explants. This observation would rule out the possibility that the rapidly proliferating and migrating modified smooth muscle cells liberate soluble growth promoting factors. However, there may be a higher concentration of platelet factors, insulin or lipoproteins, due to increased vascular permeability, in the explants of arteriosclerotic tissue which could enhance SMC outgrowth (Ross and Glomset, 1973;Friedman et al., 1976) Also the accumulation of collagen, elastin and mucopol ysaccharides in arterioscler0tic tissue may present an extracellular matrix more supportive of cellular outgrowth and migration.
Another explanation for the more rapid outgrowth could be based on decreased protein degradation and higher metabolic activity in the tissue explant. The first evidence that decreased protein degradation may contribute to rapid growth of mammalian tissues came from studies of work induced hypertrophy of skeletal muscles (Goldberg et al., 1975). Similar reports exist in the literature correlating necreased proteolysis with rapid compensatory growth in other tissues (Scornik, 1972 adapt to and proliferate more rapidly in vitro is suggestive that they have undergone fundamental changes and could be considered analogous to those cells comprising the benign smooth muscle tu~or described by Moss and Benditt (1975) and . Rounds et al. (1976).
Morphologically , S~C grown from normal tissue and arteriosclerotic tissue were indistinguishable and they both exhibited t ypical features such as abundant myofilaments and surface vesicles. These observations agree with those of Ross (1971) who has used identical culture methods.
Further characterization of these cells was not within the scope of this study , but has been explored by other investigators (Mauger et al., 1975;Ki~es and Brandt, 1976) who obtained SMC by the same technique . However, it was observed in this study that SMC f rom normal tissue enter senescence sooner than SMC from a rterioscler otic tissue.  (Fuller et al., 1976) .
The daughter cells growing in the tissue culture system proved to be more efficient in producing labelled collagen.
In these cultures as much as 20% of the radioactive protein released into the medium was collagenase digestible; however, an average of about 10% was collagenase digestible. It was still necessary to add carrier collagen to the extracted 3H-proline labelled collagen in order to visualize the characteristic collagen bands after separat ion on po l y acrylamid e gel s stained with Coomassie blue. The total disappearance of radioacti vity from t~e collagen protein bands by prior treatment of the sample with purified clostridial collagenase (protease-free) consistentl y confirmed the identity of these characteristic bands.
Daughter cells derived from control aorta were s ynthesizing 60% type I and 40% t ype III collagen at the end of their 4th passage (Table 4). Cells derived from arteriosclerotic aorta at the same passage and density (5 X 10 6 / 75 cm 2 ) synthesized 80% type I and 20% type III collagen The synthesis of collagen t ype I / ug. DNA was the same in each g roup. The synthesis of type III was significantly (p <0.05) decreased in the arteriosclerotic groupl however, this decrease was not enough to produce a significant (p <0.05) difference in total collagen s y nthesized b y the experimental group of smooth muscle cells. The significant decrease in type III s y nthesis was confirmed by CMCchromatography ( Figure 5). As mentioned in the Results section there was variation in the ratio of collagen types determined by the different analytical methods for the cell culture system. However, there was no observed disagreewent in the ratio of these types s ynth esized by the organ culture system. This ma y be due to the differential influ-  (Brown et al., 1978). Thus, these chains are not synthesized in large enough quantity to interfere in the detection of a s ignifi cant shift in the ratio of type I and type III collagen. However, they may account for some of the variabilit y between the electrophoretic and chromatographic results, since the A chain migrates in the same position as the ~1-c hain on sodium dodecyl-sulfate polyacrylamide gels. It has been suggested that the formation of type I trimer is associated with the shift in gene expression responsible for a change in the type of collagen produced (Mayne et al., 1976). It is evident from this study that atherogenic factors that modify the growth characteristics of vascular smooth muscle cells also modify t::.e gene expression for different collagen types.
By the tenth passage in culture SMC derived from both control and arteriosclerotic groups were synthesizing exelusively type I collagen. Considerable difficulty was experienced in carrying SMC in the control group to the 10th passage, due to an apparent decrease in plating efficiency and growth rate, suggestive of senescence. Coliagen heterogeneity was measured in only two flasks of control cells at the 10th passage and found to be about 90% t y pe I for each.
The effect of time in culture on the s y nthesis of type I collagen can be explained b y what has been reported regarding "collagen switching." Cellular senescence appears to be delayed in the arteriosclerotic cell group, which is consistent with the other indication of dedifferentiation mentioned above . The SMC in the control group proliferate at a slower rate which suggests that their growth req~ire rnents are more stringent than those for the cells in the arteriosclerotic group . The SMC in the control group may be more dependent than SMC in the arteriosclerotic group, 63 on species-specific growth factors not present in the heterologous , ~eat inactivated, fetal calf serum . A decrease in the requirement for multiplication stimulating factors in serum has been observed in viral transformed chick embryo fibroblasts (Smith et al., 1971) . These types of factors could modulate the onset of senescence.
CONCLUSI0NS 1) . The migration of smooth muscle cells from aorta extracts is more rapid in tissue derived from arteriosclerotic lesions than from control aorta. This supports the hypothesis that there is a greater population of these cells with enhanced migratory potential in the arteriosclerotic lesions. It is suggested that these modified smooth muscle cells would similarly be more attracted to the injured intima in the in vivo situation.
2). Daughter cells from arteriosclerotic smooth muscle proliferate in culture at a higher rate than their counterparts from normal tissue. These modified smooth muscle cells resemble transformed cells which are generally considered to have exaggerated or higher growth kinetics than untransformed cells. 3) The population of smooth muscle cells derived from arteriosclerotic aortas contain cells which are phenotypically different fro!"\ smooth muscle cells from control aorta with respect to collagen heterogeneity. This disease related change in collagen heterogeneity has been reported in vivo and now has been shown to be characteristic in the genome of daughter cells maintained in vitro.

4).
Derepression of collagen type I synthesis has been linked with cellular processes associated with aging, scars and transformation as observed for chondroc y tes in culture. It appears that arteriosclerosis enhances this derepression process for vascular smooth muscle cells permitting the accumulation of fibrous collagen components resembling scar tissues which become fibrous vascular plaques. 5). These data also support a clonal selection hypothesis.
If a cell only synthesizes one type of collagen at a time, tissues synthesizing more than one type of collagen have more than one, type-specific 909ulation, of coll~gen producing cells. The numerous population doublings and subsequent subculturing procedures may favor for one subpo pulation of SMC (e . g . type I collagen producing) and not the other subpopulation (e . g . type III collagen producing) .