Effect of exogenous extracellular matrices on proteoglycan synthesis by cultured rabbit costal chondrocytes.

We examined the effect of an extracellular matrix (ECM), produced by either bovine corneal endothelial (BCE) cells or mouse PF HR-9 teratocarcinoma cells, on the ability of rabbit costal chondrocytes to re-express their phenotype once confluent. Rabbit chondrocytes seeded at low densities and grown on plastic tissue culture dishes produced a heterogeneous cell population composed of both overtly differentiated and poorly differentiated chondrocytes, as well as fibroblastic cells. On the other hand, cultures grown on BCE-ECM- or HR-9-ECM-coated dishes reorganized into a homogeneous cartilage-like tissue composed of round cells surrounded by a refractile matrix that stained intensely with alcian green. The cell ultrastructure and that of their pericellular matrix were similar to those seen in vivo. The differentiation of chondrocyte cultures grown on the ECMs vs. plastic was reflected by a two- to three-fold increase in the maximal rate of incorporation of [35S]sulfate and [3H]glucosamine into proteoglycans. Furthermore, the ratio of 35S-labeled proteoglycans incorporated in the cell layer vs. those released into the medium was 1.5-2.5-fold higher when cultures were grown on the ECMs than on plastic. This suggests that the ECMs stimulate the incorporation of newly synthesized proteoglycans into a cartilaginous matrix. Since chondrocyte cultures grown on BCE-ECM or HR-9-ECM give rise to a homogeneous cartilage-like tissue even when seeded at low cell densities, they provide a model for the study of cell-substrate interactions that are responsible for the maintenance of the differentiated phenotype of chondrocytes.

Sulfated proteoglycan synthesis by confluent cultures of rabbit costal chondrocytes grown in the presence of fibroblast growth factor.

We examined the effect of fibroblast growth factor (FGF) on proteoglycan synthesis by rabbit costal chondrocyte cultures maintained on plastic tissue culture dishes. Low density rabbit costal chondrocyte cultures grown in the absence of FGF gave rise at confluency to a heterogeneous cell population composed of fibroblastic cells and poorly differentiated chondrocytes. When similar cultures were grown in the presence of FGF, the confluent cultures organized into a homogenous cartilage-like tissue composed of rounded cells surrounded by a refractile matrix. The cell ultrastructure and that of the pericellular matrix were similar to those seen in vivo. The expression of the cartilage phenotype in confluent chondrocyte cultures grown from the sparse stage in the presence vs. absence of FGF was reflected by a fivefold increase in the rate of incorporation of [35S]sulfate into proteoglycans. These FGF effects were only observed when FGF was present during the cell logarithmic growth phase, but not when it was added after chondrocyte cultures became confluent. High molecular weight, chondroitin sulfate proteoglycans synthesized by confluent chondrocyte cultures grown in the presence of FGF were slightly larger in size than that produced by confluent cultures grown in the absence of FGF. The major sulfated glycosaminoglycans associated with low molecular weight proteoglycan in FGF-exposed cultures were chondroitin sulfate, while in cultures not exposed to FGF they were chondroitin sulfate and dermatan sulfate. Regardless of whether or not cells were grown in the presence or absence of FGF, the 6S/4S disaccharide ratio of chondroitin sulfate chains associated with high and low molecular weight proteoglycans synthesized by confluent cultures was the same. These results provide evidence that when low density chondrocyte cultures maintained on plastic tissue culture dishes are grown in the presence of FGF, it results in a stimulation of the expression and stabilization of the chondrocyte phenotype once cultures become confluent.

Mitogenic effect of fibroblast growth factor on early passage cultures of human and murine fibroblasts.

Fibroblast growth factor (FGF), a polypeptide that has been shown to stimulate division in 3T3 cells, was tested for mitogenic effects on diploid, early-passage cells from human and murine sources. The quantitative assay of [3H]thymidine incorporation into acid-insoluble material showed that FGF at low concentrations (10 minus 9 M) was more effective than additional serum for provoking the initiation of DNA synthesis in human foreskin fibroblasts or mouse fibroblasts maintained in 5 or 10% serum, respectively. The growth of the human fibroblasts was twice as fast in the presence of FGF plus 10% calf serum as it was in the presence of 10% calf serum or 20% fetal calf serum alone. The addition of FGF to primary cultures of mouse fibroblasts in 0.4% serum resulted in a twofold increase in cell number compared to controls. In contrast to results obtained with 3T3 cells, neither insulin nor a glucocorticoid potentiated the effects of FGF on either human or mouse cells.

Bovine brain and pituitary fibroblast growth factors: comparison of their abilities to support the proliferation of human and bovine vascular endothelial cells.

The mitogenic effects of brain and pituitary fibroblast growth factors (FGF) on vascular endothelial cells derived from either human umbilical vein or bovine aortic arch have been compared. Both brain and pituitary FGF are mitogenic for low density human umbilical endothelial (HUE) cell cultures maintained on either fibronectin- or laminin-coated dishes or on biomatrices produced by cultured cells such as bovine corneal endothelial cells or the teratocarcinoma cell line PF-HR-9. Pituitary FGF triggered the proliferation of HUE cells at concentrations as low as 0.25 ng/ml, with a half-maximal response at 0.55 ng/ml and optimal effect at 2.5 to 5 ng/ml. It was 50,000-fold more potent than commercial preparations of endothelial cell growth factor and 40 times more potent than commercial preparations of pituitary FGF. Similar results were observed when the effect of pituitary FGF was tested on low density cultures of adult bovine aortic endothelial cells. When the activity of brain and pituitary FGF on low density HUE cell cultures was compared, both mitogens were active. To confirm the presence in brain extract of both acidic and neutral, as well as of basic mitogen, for HUE cells, brain tissues were extracted at acidic (4.5), neutral (7.2), and basic (8.5) pH. The three types of extracts were equally potent in supporting the proliferation of either HUE or adult bovine aortic endothelial cells. When the various extracts were absorbed at pH 6.0 on a carboxymethyl Sephadex C-50 column, the neutral and basic extracts had an activity after adsorption similar to that of unadsorbed extracts. In contrast, extracts prepared at pH 4.5 lost 90-95% of their activity which was recovered in the adsorbed fraction containing FGF.

Expression of human basic fibroblast growth factor cDNA in baby hamster kidney-derived cells results in autonomous cell growth.

Growth factor over-production by responsive cells might contribute to their autonomous proliferation as well as their acquisition of a transformed phenotype in culture. Basic fibroblast growth factor (bFGF) has been shown to induce transient changes in cell behavior that resemble those encountered in transformed cells. In addition, several types of human tumor cells have been shown to produce bFGF. To determine directly the role that bFGF might play in the induction of the transformed phenotype, we have introduced a human bFGF cDNA expression vector into baby hamster kidney-derived (BHK-21) fibroblasts. One of the BHK transfectants, termed clone 19, expresses the bFGF mRNA and produces biologically active bFGF that accumulates to a high concentration inside the cells. These properties correlate with the ability of the cells to grow in serum-free medium without the addition of exogenous bFGF. Clone 19 cells also proliferated in soft agar, indicating that constitutive expression of the bFGF gene results in a loss of anchorage-dependent growth.

Comparison of the ability of basement membranes produced by corneal endothelial and mouse-derived Endodermal PF-HR-9 cells to support the proliferation and differentiation of bovine kidney tubule epithelial cells in vitro.

The proliferation and morphological differentiation of bovine kidney collecting-tubule epithelial cells has been examined as a function of substrata and plasma factors. Collecting kidney tubule explant maintained in vitro gave rise to two distinct cell populations; one was composed mostly of fibroblastic cells whereas the other was epithelioid (EP cells). The proliferation of fibroblastic cells when exposed to serum-supplemented medium was best expressed when cells were maintained on a basement membrane produced by bovine corneal endothelial cells. This basement membrane has a composition, which in previous studies has been shown to favor the proliferation of mesenchymal cells. In contrast, the proliferation of EP cells was best expressed when cells were maintained on a basement membrane produced by the mouse-derived endodermal cell line PF-HR-9 (HR-9-BM). This basement membrane has a biochemical composition very similar to the basement membrane underlying the kidney tubules. Although the fibroblast confluent monolayer maintained on bovine corneal endothelial cell extracellular matrix did not undergo morphogenesis, the confluent monolayer of EP cells maintained on HR-9-BM shows hemicyst formation, suggesting that they were capable of vectorial fluid transport. They also built a complex three-dimensional kidney tubulelike network. Some tubules became grossly visible and floated into the tissue culture medium, remaining tethered to the cell monolayer at either end of the tubule. On an ultrastructural level, the tubules consisted of cells held together with junctional complexes arranged so as to form a lumen. The smallest lumen were bordered by 2-3 cells, and the largest ones by 8-15 cells. The lumens of the larger tubules did contain granular fibrillar and amorphous debris. Low-density EP cell cultures maintained on HR-9-BM could be induced to proliferate at a rate approaching that of cultures exposed to serum when they were exposed to medium supplemented with high-density lipoprotein (HDL, 750 micrograms protein/ml) and transferrin (50 micrograms/ml). When exposed to HDL concentrations equal or lower than 250 micrograms protein/ml, low-density cultures proliferated at a slow rate and readily formed tubulelike structures. This observation indicates that EP cells do not need to reach confluence to undergo morphogenesis, and that HDL, which in the presence of transferrin supports the cell proliferation, can favor their differentiation into tubulelike structures once its concentration becomes limiting for mitogenesis.

Vascular endothelial cells maintained in the absence of fibroblast growth factor undergo structural and functional alterations that are incompatible with their in vivo differentiated properties.

Vascular endothelial cells cultured in the presence of fibroblast growth factor (FGF) adopt at confluence a morphological appearance similar to that of the vascular endothelium in vivo. Similarly, their apical cell surface is, as in vivo, nonthrombogenic. In contrast, when the cultures are maintained in the absence of FGF, the cells undergo within two to three passages structural and functional alterations that are incompatible with their in vivo morphological appearance and physiological function. Cultures maintained in the absence of FGF no longer adopt, upon reaching confluence, the configuration of a monolayer composed of small closely apposed and nonoverlapping, cuboidal cells. Instead, confluent cultures deprived of FGF consist of large, overlapping cells which have lost the polarity of cell surface characteristic of the vascular endothelium. The apical cell surface becomes thrombogenic, as reflected by its ability to bind platelets, whereas fibronectin, which at confluence is normally associated only with the basal cell surface, can be found both on top of and underneath the cell layer. Among other changes, both sparse and confluent cultures maintained in the absence of FGF showed a greatly increased production of fibronectin. CSP-60, a cell surface protein whose appearance is correlative with the adoption of a cell monolayer configuration, can no longer be detected in cultures maintained in the absence of FGF. Overlapping endothelial cells maintained in the absence of FGF can also no longer function as a protective barrier against the uptake of ligands such as low density lipoprotein. Exposure of the culture to FGF induces a restoration of the normal endothelial characteristics concomitant with the adoption of a flattened cell monolayer morphology. These results demonstrate that, in addition to being a mitogen. FGF is involved in controlling the differentiation and phenotypic expression of the vascular endothelium. This is reflected by its effect on the morphological appearance, polarity of cell surfaces, platelet binding capacity, and barrier function of the vascular endothelium.

Factors controlling the proliferative rate, final cell density, and life span of bovine vascular smooth muscle cells in culture.

Low density vascular smooth muscle (VSM) cell cultures maintained on extracellular-matrix(ECM)-coated dishes and plated in the presence of either plasma or serum will proliferate actively when serum-containing medium is replaced by a synthetic medium supplemented with three factors: high density lipoprotein (HDL, 250 micrograms protein/ml); insulin (2.5 micrograms/ml) or somatomedin C (10 ng/ml); and fibroblast growth factor (FGF, 100 ng/ml) or epidermal growth factor (EGF, 50 ng/ml). The omission of any of these three factors from the synthetic medium results in a lower growth rate of the cultures, as well as in a lower final cell density once cultures reach confluence. When cells are plated in the total absence of serum, transferrin (10 micrograms/ml) is also required to induce optimal cell growth. The effects of the substrate and medium supplements on the life span of VSM cultures have also been analyzed. Cultures maintained on plastic and exposed to medium supplemented with 5% bovine serum underwent 15 generations. However, when maintained on ECM-coated dishes the serum-fed cultures had a life span of at least 88 generations. Likewise, when cultures were maintained in a synthetic medium supplemented with HDL and either FGF or EGF, an effect on the tissue culture life span by the substrate was observed. Cultures maintained on plastic underwent 24 generations, whereas those maintained on ECM-coated dishes could be passaged repeatedly for 58 generations. These experiments demonstrate the influence of the ECM-substrate only in promoting cell growth but also in increasing the longevity of the cultures.

Effect of cell density on thrombin binding to a specific site on bovine vascular endothelial cells.

We studied thrombin binding to proliferating and confluent endothelial cells derived from bovine vascular endothelium. [125]thrombin was incubated with nonconfluent or confluent endothelial cells and both the total amount bound and the amount linked in a 77,000-dalton thrombin-cell complex were determined. Approximately 230,000 molecules of thrombin bound per cell in nonconfluent cultures compared to 12,800 molecules per cell in confluent cultures. Approximately 67,7000 thrombin molecules were bound in an apparently covalent complex, Mr = 77,000, with each cell in sparse cultures, whereas only 4,600 thrombin molecules per cell were bound in this complex with confluent cultures. Similar studies with [125I]thrombin and endothelial cells derived from bovine cornea revealed no difference either in the total amount of thrombin bound or in the amount bound in the 77,000-dalton complex using sparse or confluent cultures. When confluent vascular endothelial cultures were wounded, additional cellular binding sites for the 77,000-dalton complex with thrombin appeared within 24 h. A 237% increase in the amount of thrombin bound to these sites was induced by a wound which resulted in a 20% decrease in cell number in the monolayer. There was no significant increase in thrombin binding to other cellular sites at 24 h. These experiments provide evidence that the first change in thrombin binding after injury is an increase in the cellular sites involved in the 77,000-dalton complex, and suggest that thrombin binding to endothelial cells may be important in the vascular response to injury.

Modulation of collagen synthesis by a growth factor and by the extracellular matrix: comparison of cellular response to two different stimuli.

Cultured bovine corneal endothelial cells can be grown in three ways: on plastic, on plastic with fibroblast growth factor present in the media, and on their own preformed extracellular matrix. On plastic alone, cells grow in a disorderly fashion and secrete matrix on all cell surfaces. Cells grown on plastic with growth factor or on a matrix, at confluence, have matrix deposition only on the basal surface of the cells and an orderly contact-inhibited pattern of growth. This correlates with the polarity they demonstrate histologically. This cell-matrix pattern resembles the pattern observed in vivo. Both the soluble growth factor and the extracellular matrix are able to modulate the pattern of collagen synthesis and deposition by cells, but they do so in two entirely different ways. In cells grown on the extracellular matrix, total collagen synthesis is lower but more efficient. Collagen is deposited primarily into the cell layer even at the early sparse stage of culture. In cells grown on plastic with growth factor in the media, collagen is initially secreted into the media and does not become incorporated into the matrix. The deposition of collagen on the basal surface of cell occurs only late in the culture, and is achieved by increments in a stepwise manner. The in vivo-like pattern is not manifest until confluence has been reached. Thus, the extracellular matrix functions not only as a structural support, but is also instructional to the cells plated on it. In this case, the matrix regulates the level of collagen synthesis in the cells and modulates the pattern of collagen deposition. Soluble growth factors may act in part by enhancing a cell's ability to elaborate an appropriate matrix pattern necessary for the cell's own growth and accurate function.

Effect of fibroblast growth factor on the division and fusion of bovine myoblasts.

The effect of fibroblast growth factor (FGF) on the rate of proliferation and fusion of bovine myoblast has been examined. Addition to the cultures of 0.1 mug-1 mug/ml of FGF stimulates the rate of proliferation and delays the fusion of primary cultures of bovine myoblasts cultured in 10% serum. Final cell densities reached in the presence of 0.1 mug/ml of FGF were fivefold higher than in controls; with 1 mug/ml, they were 10-fold higher. Increases in cell density were paralleled by increases in acetylcholine receptor sites as measured by the binding of 125I-alpha-bungarotoxin. Both fusion and the appearance of acetylcholine receptor sites were delayed in the presence of FGF. Growth hormone, insulin and testosterone, which have been reported to be mitogenic for rat and chick embryo myoblasts, did not have significant effects on DNA synthesis in bovine myoblasts when compared to the FGF. Conversely, FGF did not stimulate the proliferation of chick embryo myoblasts, indicating that it is not active in all vertebrate species.

Control of proliferation of human vascular endothelial cells. Characterization of the response of human umbilical vein endothelial cells to fibroblast growth factor, epidermal growth factor, and thrombin.

Because the response of human endothelial cells to growth factors and conditioning agents has broad implications for our understanding of wound healing angiogenesis, and human atherogenesis, we have investigated the responses of these cells to the fibroblast (FGF) and epidermal growth factors (EGF), as well as to the protease thrombin, which has been previously shown to potentiate the growth response of other cell types of FGF and EGF. Because the vascular endothelial cells that form the inner lining of blood vessels may be expected to be exposed to high thrombin concentrations after trauma or in pathological states associated with thrombosis, they are of particular interest with respect to the physiological role of this protease in potentiating cell proliferation. Our results indicate that human vascular endothelial cells respond poorly to either FGF or thrombin alone. In contrast, when cells are maintained in the presence of thrombin, their proliferative response to FGF is greatly increased even in cultures seeded at a density as low as 3 cells/mm2. Human vascular endothelial cells also respond to EGF and thrombin, although their rate of proliferation is much slower than when maintained with FGF and thrombin. In contrast, bovine vascular endothelial cells derived from vascular territories as diverse as the bovine heart, aortic arch, and umbilical vein respond maximally to FGF alone and neither respond to nor bind EGF. Furthermore, the response of bovine vascular endothelial cells to FGF was not potentiated by thrombin, indicating that the set of factors controlling the proliferation of vascular endothelial cells could be species-dependent. The requirement of cultured human vascular endothelial cells for thrombin could explain why the human cells, in contrast to bovine endothelial cells, are so difficult to maintain in tissue culture. Our results demonstrate that by using FGF and thrombin one can develop cultures of human vascular endothelial cells capable of being passage repeatedly while maintaining a high mitotic index. The stock cultures used for these studies have been passed weekly with a split ratio of 1 to 10 and are currently in their 30th passage. These cultures are indistinguishable from earlier passages when examined for the presence of Weibel-Palade bodies or Factor VIII antigen. We conclude that the use of FGF and thrombin can prevent the precocious senescence observed in most human endothelial cells cultures previously described.

Characterization of a factor that promotes neurite outgrowth: evidence linking activity to a heparan sulfate proteoglycan.

Rat sympathetic neurons, plated onto extracellular matrix produced by cultured bovine corneal endothelial cells, rapidly extended neurites in the absence of nerve growth factor (NGF). The response was unaffected by antiserum to NGF. Rapid outgrowth also occurred when sympathetic neurons were plated onto polylysine-coated surfaces that had been exposed to serum-free medium conditioned by corneal endothelial cells (CMSF). A response was seen even when the neurons were cultured without serum. When plated onto a polylysine-coated dish treated with CMSF over half its surface, only the neurons on the treated half extended neurites. The active factor in CMSF was destroyed by trypsin, acid (pH 1.6), base (pH 12.7), or heating to 80 degrees C; it was stable to heating to 60 degrees C, collagenase, deoxyribonuclease, and neuraminidase. The factor elutes just after the void volume of a Sepharose 6B column. In associative cesium chloride gradients, it sediments as a peak centered at a density of 1.36-1.37, corresponding to a peak of material that can be biosynthetically labeled with [35S]sulfate or [3H]leucine. Material from this fraction was inactivated by heparinase, but not chondroitinase ABC, implying that a heparin sulfate proteoglycan is essential for the factor's activity. Inactivation by contaminants in the heparinase preparation was ruled out. Further purification indicated that the active factor may exist as an aggregate containing a heparin sulfate proteoglycan and other molecules. CMSF also promoted neurite outgrowth by other types of neurons. Furthermore, a variety of cell types were shown to produce factors similar to that in CMSF.

Nucleotide sequence of a bovine clone encoding the angiogenic protein, basic fibroblast growth factor.

Basic and acidic fibroblast growth factors (FGF's) are potent mitogens for capillary endothelial cells in vitro, stimulate angiogenesis in vivo, and may participate in tissue repair. An oligonucleotide probe for bovine basic FGF was designed from the nucleotide sequence of the amino-terminal exon of bovine acidic FGF, taking into account the 55 percent amino acid sequence homology between the two factors. With this oligonucleotide probe, a full length complementary DNA for basic FGF was isolated from bovine pituitary. Basic FGF in bovine hypothalamus was shown to be encoded by a single 5.0-kilobase messenger RNA; in a human hepatoma cell line, both 4.6- and 2.2-kilobase basic FGF messenger RNA's were present. Both growth factors seem to be synthesized with short amino-terminal extensions that are not found on the isolated forms for which the amino acid sequences have been determined. Neither basic nor acidic FGF has a classic signal peptide.

Extracellular matrix and control of proliferation of vascular endothelial cells.

Bovine vascular endothelial cells plated at low cell density in the presence of high (10%) concentrations of serum and maintained on plastic tissue culture dishes proliferate slowly. If the cultures were exposed to fibroblast growth factors (FGF), the cells proliferated actively and, after a week, a monolayer composed of closely apposed and highly contact-inhibited mononucleated cells formed. In contrast to cultures maintained on plastic, cultures maintained on dishes coated with an extracellular matrix produced by corneal endothelial cells proliferated rapidly and no longer required FGF to reach confluence. Addition of FGF to such cultures did not decrease the mean doubling time, which was already at a minimum (18 h), nor did it result in a higher final cell density, which was already at a maximum (700-1,000 cells/mm(2)). Likewise, although human umbilical vein endothelial cells plated at low density on plastic did not proliferate, they proliferated rapidly when plated on dishes coated with an extracellular matrix. However, unlike bovine vascular endothelial cells, they still required FGF if the cultures were to become confluent. The ability of plasma vs. serum to sustain cell proliferation was analyzed using low density bovine-vascular endothelial cell cultures maintained either on plastic or on dishes coated with an extracellular matrix. Cells plated on plastic had a lower growth rate when exposed to plasma than to serum. In both cases, FGF was required for the cultures to become confluent. In contrast, when cells were plated on an extracellular matrix, they proliferated equally well, regardless of whether they were exposed to plasma or serum, and no longer required FGF to become confluent. Because the growth rate of the cultures maintained on an extracellular matrix was a direct function of the serum or plasma concentrations to which they were exposed, it is likely that the extracellular matrix had a permissive rather than a direct mitogenic effect on the cells. Therefore, one can conclude that the simple change of substrate from plastic to extracellular matrix will restore the sensitivity of vascular endothelial cells to physiological agents present in plasma or serum.

Effect of high and low density lipoproteins on proliferation of cultured bovine vascular endothelial cells.

Bovine vascular endothelial cells maintained on dishes coated with an extracellular matrix and exposedto medium supplemented with lipoprotein-deficient serum (LPDS) require the presence of lipoprotein to proliferate optimally. High density lipoprotein (HDL) seems to be the major factor involved in the proliferation of vascular endothelial cells. This is mostly due to its lack of toxicity when added at high concentration, as well as to its nondependence on LPDS to exhibit its mitogenic properties. Therefore, HDL at physiological concentrations (1,000--1,500 microgram protein/ml) can fully replace serum. Low density lipoprotein, unlike HDL, has a biphasic effect. Although mitogenic for vascular endothelial cells when added at low concentration, once physiological concentrations are reached it becomes toxic for the cells. Moreover, and in contrast with HDL, the mitogenic effect of low density lipoprotein was found to be a function of the LPDS concentration to which cultures were exposed. The substrate upon which cultures are maintained has been found to be an important factor if a mitogenic effect of HDL is to be observed. When maintained on plastic, cells proliferate poorly in response to HDL unless fibroblast growth factor is added to the medium. In contrast, when maintained on extracellular matrix, an optimal growth rate is induced by HDL, even in the absence of fibroblast growth factor. This suggests that, in vivo, the integrity of the basement membrane upon which endothelial cells rest and migrate is an important factor in determining the cells response to lipoproteins present in plasma.

Basic fibroblast growth factor as a growth inhibitor for cultured human tumor cells.

Basic fibroblast growth factor (bFGF) stimulates the proliferation of many cells and it is found in a wide variety of normal or transformed tissues. As demonstrated here, bFGF is also present in cultured human Ewing's sarcoma cells. Unexpectedly, however, bFGF isolated from these cells inhibits their own proliferation, indicating that bFGF can act as an endogenous (autocrine) growth inhibitor for cultured Ewing's sarcoma cells. Since bFGF also inhibits the proliferation of some further tumor cells, but stimulates that of others, it can be considered a bifunctional regulator of tumor cell proliferation. The autocrine growth-inhibitory effect of bFGF in Ewing's sarcoma cells may explain the low mitotic activity of Ewing's sarcomas.

Effect of rous sarcoma virus transformation of rat-1 fibroblasts upon their growth factor and anchorage requirements in serum-free medium.

The proliferative response of nontransformed rat embryo (Rat-1) cells and avian sarcoma virus-transformed B31 cells to high-density lipoprotein (HDL), transferrin, insulin, epidermal growth factor (EGF), and fibroblast growth factor has been compared. HDL, added in combination with transferrin, supported the active proliferation of low-density cultures of both Rat-1 and B31 cells. No major difference in the sensitivity of Rat-1 or B31 cells to HDL and transferrin was observed when cells were maintained on dishes coated with an extracellular matrix (ECM) obtained from bovine corneal endothelial cells. The two cell types differed in their response to the other known growth-promoting agents, however, in contrast to Rat-1 cells, transformed B31 cells no longer respond to EGF and fibroblast growth factor and respond only inconsistently to the mitogenic stimulus of insulin. Nontransformed Rat-1 cells and transformed B31 cells grown in the presence of medium containing, respectively, HDL, transferrin, insulin, EGF, and dexamethasone or HDL, transferrin, and insulin could be subcultured for more than 50 generations in the complete absence of serum without significant alteration in morphology, growth rate, or tumorigenicity (B31 cells). When plastic or collagen-coated dishes were used as the substrate instead of ECM-coated dishes, nontransformed Rat-1 cells grew very slowly in the serum-free medium described above. Dishes coated with collagen were not more efficient than was plastic in supporting growth of Rat-1 cells under these conditions. Coating dishes with fibronectin, however, clearly improved their growth, bringing the final cell density of the cultures up to 50% of that obtained on ECM-coated dishes. In contrast, transformed B31 cells grew significantly in serum-free medium when seeded on plastic or collagen-coated dishes, and the final cell density reached by cells on these substrates was 50% of that of cells maintained on ECM-coated dishes. In addition, B31 cells grew equally well when seeded on fibronectin- or ECM-coated dishes. The transformed cells thus showed less stringent substrate requirements when grown under serum-free conditions than did nontransformed Rat-1 cells. Our data also indicate that HDL, in combination with transferrin, supported efficient anchorage-independent growth of B31 cells. Fibroblast growth factor, but not insulin or EGF, further improved anchorage-independent growth of these cells. The capacity of cells to form colonies in semisolid medium when exposed to HDL and transferrin seems to correlate with high tumorigenic potential.

Determination of cellular shape by the extracellular matrix and its correlation with the control of cellular growth.

Although the problem of cellular proliferation may seem at first glance to be tremendously complex, the mechanisms which control it may be extremely simple. One of the primary factors which regulates the mitogenic response of a given cell type to a given class of mitogenic agents seems to be the cellular shape. We have found that corneal epithelial cells, for example, adopt a flattened configuration when maintained in vitro on plastic and are very sensitive to fibroblast growth factor, but not to epidermal growth factor. When maintained on collagen, on the other hand, they become tall and columnar and respond primarily to epidermal growth factor. The cellular shape is dictated in vivo by the extracellular material upon which the cells rest and in vitro by the substrate upon which the cells are maintained. The substrate itself may, in turn, induce the cells to manufacture their extracellular material and specific cell surface proteins which control the cellular shape.

High-density lipoproteins and the proliferation of human tumor cells maintained on extracellular matrix-coated dishes and exposed to defined medium.

The ability of high-density lipoprotein (HDL) to support the growth of an established tumor cell line exposed to defined medium supplemented with transferrin has been examined. Low-density A-431 carcinoma cells maintained on extracellular matrix- or fibronectin-coated dishes proliferated actively when exposed to a synthetic medium supplemented with HDL, 500 micrograms protein per ml. Epidermal growth factor added at concentrations above 0.5 ng/ml inhibited cell growth, while at concentrations above 5 ng/ml it was cytotoxic. Among the various substrata tested for their ability to support the active proliferation of low-density A-431 cells when exposed to transferrin and HDL, plastic was the least efficient. On fibronectin-coated dishes, cells ceased to proliferate after 8 population doublings, while on extracellular matrix-coated dishes cells could be passaged for 50 population doublings. In the case of colon carcinoma, rhabdomyosarcoma, and Ewing's sarcoma cells exposed to medium supplemented with transferrin, the addition to the cultures of HDL alone resulted in a growth rate and final cell density which were similar to those observed when cells were exposed to serum-supplemented medium. In the case of the mammary carcinoma cell lines MCF-7 and ZR-75-1, HDL also supported cell growth, although to a lesser extent than did serum. The present study therefore indicates that HDL is capable of supporting, either totally or partially, the in vitro proliferation of tumor cells.