CELL SURFACE PROPERTIES IN RELATION TO GROWTH AND FORM IN CHINESE HAMSTER OVARY CELLS IN CULTURE By Jacques van Veen A DISSERTATION PRESENTED TO THE GR^vDUATE COIRICIL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLIIENT OF THE PvEOUIIEKENTS FOR THE DEGREE OF DOCICR OF PHILOSOPHY LTJIVEP.SITY OF FLORIDA 19 75 ACKNOWLEDGEIIENTS I am grateful to Dr. Robert Michael Roberts for his encouragement and efforts on my behalf during my studies as a graduate student, his guidance and support in the conduct of this research and his help in the preparation of this dissertation. « I would like to thank Dr. Kenneth D. Noonan for his continued help and suggestions during the course of this research. I also thank Drs. Eugene Sander and Bryan Gebhardt for their critical evaluation of this dissertation. 1 thank the government of the Netherlands Antilles, the Department of Biochemistry and my parents for their financial support. I would also like to express gratitude for the continuing encouragement and support of my wife, Brenda. TABLE OF CONTENTS Acknowledgements il List of Tables iv List of Figures vl Key to Abbreviations lx Abstract . ..... x Introduction 1 Materials and Methods 16 Results 52 Discussion ^ . . 126 Bibliography 133 Appendix 145 Biographical Sketch 149 LIST OF TABLES 1. Sources of chemicals 17 2. Sources and specific activities of radiochemicals 18 3. Buffers used in polyacrylamide gel electrophoresis 29 4. Preparation and composition of the two phases used in the Brunette and Till method of membrane preparation . . , . 44 5. Effect of 3':5'-cyclic AMP, dB-cAMP and Sq 20009 on the morphology of the CHO cell lines 53 6. The intracellular 3':5'-cyclic AMP levels in the CHO cell lines 7. Specific radioactivity of plasma membranes labeled with [^H]-sodium borohydrlde 93 8. The effect of dB-cAMP, 3' :5'— cyclic AMP, butyric acid and Sq 20009 on the agglutination of CHO cell lines by concanavalin A and WGA 95 9. The effect of dB-cAMP and Sq 20009 on the binding of [%] -concanavalin A by CHO cell lines 97 10. The effect of temperature, enzyme treatment and colchicine on lectin induced agglutination of CHO- H-7 by concanavalin A and WGA 98 11. The temporal effects of addition and removal of dB-cAMP on the agglutination of CHO-H-7 by concanavalin A and WGA 100 12. The effect of inhibitors of protein and RNA synthesis on dB-cAMP induced loss of agglutination of CHO-H-7 by concanavalin A and WGA 102 13. The effect of cyclohexamlde and actinoraycin D on the Sq 20009 induced loss of agglutination of CHO-K-1-24-2 by concanavalin A and WGA 104 iv 105 14. The effects of cyclohexamide and actinomycin D on the dB-cAMP induced loss of agglutination of CHO-K-1-24-2 by concanavalin A and WGA 15. A summary of the effects of 3':5'-cyclic AMP, dB-cAMP and Sq 20009 on the surface glycopeptides , agglutination and morphology of the' CHO cell lines studied V LIST OF FIGURES 1. A schematic representation of the "Fluid Mosaic Model" of plasma membranes 7 2. Flow chart demonstrating the procedure followed in measuring thymidine uptake and incorporation in CHO cells . . 22 3. A diagram of the Buchler analytical temperature regu- lated polyacrylamide gel electrophoresis apparatus 28 4. Flow chart demonstrating the procedure for lacto- peroxidase mediated iodination of cell surfaces 32 5. Flow chart demonstrating the procedure for galactose oxidase mediated labeling of cell surfaces 34 6. a) Reactions involved in the pyridoxal phosphate mediated labeling of cell surfaces 36 b) Flow chart demonstrating the pyridoxal phosphate mediated labeling of cell surfaces 37 7. Flow chart demonstrating the procedure followed in the isolation of plasma membranes, by the method of Barland and Schroeder 41 8. Flow chart demonstrating the procedure followed in the isolation of plasma membranes, by the method of Brunette and Till 42 9. Standard curves for the determination of protein concentration 46 10. Experimental procedure used in the determination of 3':5'-cyclic AMP concentrations 50 11. Standard curve for the determination of 3’:5'-cyclic AMP concentrations ■ . . . . 51 12. The effect of dB-cAMP or Sq 20009 on the morphology of CHO cells 54 13. a) The effect of dB-cA>fP on the synchronization of CHO-H-7 . . 58 b) The effect of dB-cAMP or Sq 20009 on the growth rate of ' he CHO cell lines 59 vi 14. The effect of dB-cAMP on the thymidine uptake in CHO cells 62 15. Reproducibility of membrane preparations . . .' 64 16. Coomassie blue staining pattern of isolated plasma membranes of CHO cells analyzed on 5% (w/v) poly- acrylamide gels 65 17. a) Coomassie blue staining pattern of isolated plasma membranes of CHO cells analyzed on 7.5% (w/v) polyacrylamide gels 66 b) Standard curve for the estimation of molecular weights of proteins on 7.5% (w/v) polyacrylamide gels . . 67 18. Coomassie blue staining pattern of isolated plasma membranes of CHO cells analyzed on 10% (w/v) poly- acrylamide gels 68 19. Profile of incorporation of radiolabeled leucine in the plasma membranes 70 20. Comparison of incorporation of radiolabeled leucine in the plasma membranes 71 21. The effect of dB-cAMP on the incorporation of leucine in the plasma membranes analyzed on 10% (w/v) poly- acrylamide gels 73 22. The effect of dB-cAMP on the incorporation of leucine in the plasma membranes analyzed on 7.5% (w/v) poly- acrylamide gels 74 23. lodinatable surface polypeptides of different CHO cell lines grown in the presence or absence of 1 mM dB-cAMP ... 77 24. lodinatable surface polypeptides of CHO-H-7 grown in the presence or absence of 1 mM dB-cAMP 79 25. lodinatable surface polypeptides of CHO-K-1-24-2 grown in the presence or absence of 1 mM dB-cAMP 81 26. lodinatable surface polypeptides of CHO-K-l-M-7 grown in the presence or absence of 1 mM dB-cAMP 83 27. lodinatable surface polypeptides of CHO-K-1 grown, in the presence or absence of 1 mM dB-cAMP 85 28. Effect of trypsin treatment on the iodinatable surface polypeptides of CHO-M-7 87 vii 29. The non-specific labeling of CHO-H-7 surface poly- peptides by [^Hj-sodium-borohydride 90 30. Galactose oxidase mediated labeling of CHO-H-7 surface polypeptides 92 31. Paper chromatographic separation of hydrochloric acid hydrolysate of fucose labeled glycopeptides 107 32. G-50 Sephadex elution profile of pronase digested mixed trypsinated of CHO-H-7 108 33. The effect of hydrochloric acid hydrolysis on fucose containing glycopeptides HO 34. The effect of neuraminidase on fucose containing surface glycopeptides • H2 35. Fucose containing glycopeptides released from the surface of CHO cells by trypsinization after different conditions of growth 114 36. Fucose containing glycopeptides released from the surface of CHO cells by trypsinization after different labeling conditions 12Q 37. Fucose containing glycopeptides released from the surface of CHO-H-7 after removal of dB-cAMP from the growth medium 2.22 38. The effect of colchicine on the fucose containing surface glycopeptides of CHO-H-7 123 39. The effect of cyclohexamide on the fucose containing surface glycopeptides of CHO-H-7 125 viii KEY TO ABBREVIATIONS CHO cells Chinese Hamster Ovary cells dB-cAMP dibutyryl 3':5'-cyclic adenosine monophosphate Sq 20009 l-Ethyl-4- (isopropylidenehydrazino)-l H-pyrazolo- [3,4-b] pyridine-5-carboxylic acid, ethyl ester HCl PB .01 M sodium phosphate buffer, pH = 7.4 PBS Phosphate buffered saline HBSS Hanks balanced salt solution SDS sodium dodecyl sulfate WGA Wheat germ agglutinin ix Abstract of Dissertation Presented to the Graduate Council of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy CELL SURFACE PROPERTIES IN RELATION TO GROWTH AND FORM IN CHINESE HAMSTER OVARY CELLS IN CULTURE By Jacques van Veen ' August, 1975 Chairman: R. M. Roberts Major Department: Biochemistry The addition of dibutyryl 3':5'-cyclic AMP (dB-cAMP) to cultures of Chinese Hamster Ovary (CHO) cells induces changes in the appearance and properties of these cells which seem to be the reverse of those that normally occur following transformation of fibroblasts by onco- genic virus. In this study this phenomenon of "reverse transf oimation" has been examined in detail in four lines cloned from an L-proline- requiring CHO line. The four lines K-1, M-7, 24-2 and H-7 , show contrasting, rapid morphological response both to dB-cAMP and to the phosphodiesterase inhibitor Sq 20009. Whereas K-1 and H-7 elongate in presence of dB-cAMP, M-7 does not change in morphology, although it does respond to Sq 20009. 24-2 is normally fibroblast-like in * appearance but shows some slight response to both drugs. This system X is therefore an excellent one for comparing cells which show contrasting morphologies and social behavior, yet which are genotypically very similar. An added advantage was that the growth rate and final cell densities of each of the lines were not influenced by the drugs. The susceptibility of certain of the lines to agglutination by the plant lectins wheat germ agglutinin and concanavalin A was also affected by growth in the presence of Sq 20009 and dB-cAMP although the cells bound similar amounts of lectin under both growth conditions. Each of • the clones was agglutinated by both lectins in their normal state. However, whereas K-1, which is morphologically responsive, remained fully agglutinable after treatment with dB-cAMP, H-7 and 24-2 quickly lost their agglutinability . Line M-7 was unaffected by dBcAMP but lost its susceptibility to the lectins following treatment with Sq 20009. In addition, H-7 showed a marked drop in agglutinability in presence of the natural nucleotide 3' :5' -cyclic AMP to which it does not res- pond morphologically. There was not, therefore, an absolute correla- tion between the morphological responsiveness of the cells and lectin- induced agglutinability. Therefore, these features of the transformed phenotype, morphology and agglutinability, are not necessarily related phenomena . The rapid losses of agglutinability are prevented by actinomycin D, cordycepin and cyclohexamide, even if dB-cAMP is present. The return to the agglutinable state, however, which also occurs very rapidly is not prevented by these inhibitors. It is concluded that the components responsible for maintaining the non-agglutinable state turnover rapidly. xi Neither the plasma membrane protein composition, as determined by electrophoretic analysis on polyacrylamide gels, nor the surface pro- teins available for lactoperoxidase mediated iodination showed a gross modification of the composition or orientation of surface polypeptides of the CHO cells in the response to dB-cAMP, which could be correlated with any of the- other phenomena studied. In presence of dB-cAMP and Sq 20009, certain of the lines demon- strate an overall reduction in the size of a class of fucose-containing glycopeptides that can be released from their surface with trypsin. r This decrease in the size of the glycopeptide is due to a lower content of sialic acid. The change correlated with the losses in agglutinability but not with the morphological responsiveness of the cells. Double- labeling experiments with L-fucose have shown that upon addition of dB-cAMP existing glycopeptides are not modified, but new material » reaching the surface is incompletely glycosylated. After removal of the drug, sialic acid can be added to these incomplete glycopeptides. It is proposed that the loss in protein mobility induced by dB-cAMP results in a lowered efficiency in glycosylatlon at the surface since substrate and enzyme will have less opportunity for contact, and a loss in agglutinability by plant lectins. Chairman XXI INTRODUCTION Chinese hamster ovary cells Most CHO cells in culture originate from Dr. T. T. Puck's laboratory. They are derived from an original explant and are pre- sumably fibroblast in origin. However, the parental- line, CHO-K-1 which was isolated by Kao and Puck (1) does not show a strict fibro— blast-like morphology. CHO cells are not believed to be tumor igenic, but they show many of the characteristics of transformed fibroblasts. For example, they have an apparent capacity for unlimited growth, and Stow in soft agar. Furthermore, they show little tendency to orientate in parallel arrays and they do not show contact inhibition of growth (2,3). The effect of dB-cAllP and Sq 20009 on the intracellular 3 ' : 5 ' -cyclic AMP level dB-cAMP, an analog of 3':5'-cyclic AMP, recognized as the "second messenger in the activity of many hormones (4) , has been used exten- sively to induce changes in growth and metabolism of numerous cell lines, including the CHO cells. The latter cells undergo morphological and physiological changes which resemble the reverse of those which occur when cells are transformed by chemical carcinogens or viruses. For example, upon viral transformation, fibroblasts typically lose their elongated shape and assume a more compact morphology (5). They also demonstrate reduced contact inhibition of grow’th (6,7), become 2 more agglutinable by plant lectins (8) and show a change in fucose- containing surface glycopeptides (9-12). It is believed that dB-cAMP acts by increasing the effective intracellular concentrations of 3’;5'- cyclic AMP . dB— cAMP is reported to be relatively less sensitive to degradation by serum enzymes as compared to 3': 5' -cyclic AMP (13-18). It is also an inhibitor of 3':5'-cyclic nucleotide phosphodiesterase (1A,19). Further, the less charged dB— analog may be more easily trans- ported across the membrane although this has been disputed (20) . It is therefore not surprising that the analog is generally but not universally (21) more effective in inducing "reverse transformation" than the natural nucleotide itself. That uptake is an important factor in determining the effectiveness of 3':5'-cyclic AMP was demonstrated by fusion of lipid-vesicles containing 3':5'-cyclic AMP with cell- membranes (22). The cyclic nucleotide concentration effective in causing a reduction in growth rate was three orders of magnitude smaller when the cyclic nucleotide was introduced into the cell by fusion of lipid- vesicles with the cell-membranes than when it was added to the growth medium. The differences in effectiveness in different cell lines to increase the intracellular 3' :5’— cyclic AMP concentrations observed between the cyclic nucleotide and its dibutyryl analog, may well reflect differences in uptake and phosphodiesterase activities between the cell lines . Some workers have suggested that the effect of dB-cAMP is not due to its acting as an analog of 3' :5' -cyclic AMP (21). However, a number of diesterase inhibitors such as thebphyline and papaverine and hormones such as prostaglandins and testosterone, which due to their stimulating effect on adenyl cyclase might be expected to increase '.ntraceliular 3 3':5’-cyclic AMP levels, have an effect similar to dB-cAMP (3,23,24). Furthermore, growth stimulation of untransformed cells by protease treatment (25) or by insulin (26,27) decreases intracellular 3':5'- cyclic AMP levels, and these effects can be prevented by dB-cAMP. Hsie and coworkers have recently published an extensive study on f the mode of action of dB-cAMP and its degradation products by CHO-K-1 (20,28). It was shown that dB-cAMP was largely metabolized to two 2 major products: 1) 0 -monobutyryl-cyclic AMP, which in turn was rapidly hydrolyzed, and 2) N^-monobutyryl-cyclic AMP, which was rela- tively stable and acted as a competitive inhibitor of one of the 3':5'- cyclic AMP-phosphodiesterases causing intracellular 3':5'-cyclic AMP concentrations to rise. However, even after 4 hours' incubation signi- ficant amounts of dB-cAMP were still present. Their work did not rule out the possibility therefore that dB-cAMP was not acting directly as ah analog of 3' :5'-cyclic AMP. Although its effectiveness may vary from cell line to cell line, it seems unequivocal that dB-cAMP acts by increasing effective intracellular 3':5'-cyclic AMP concentrations. The effect of Intracellular 3':5'-cycllc AMP concentrations on growth rate and pattern The relationship between 3':5'-cyclic AMP concentrations and cell growth which was alluded to earlier has been investigated in many cell lines. The available evidence strongly suggests that there is a direct relationship between intracellular 3':5'-cyclic AMP levels and cell growth. 3' :5'-cyclic AMP levels seem to vary in relation to degree of confluency, growth rate and the physiological state of the cultures. For example, cyclic nucleotide levels rise when the cel] reaches con- fluency, in cells that exhibit contact inhibition of growth, but not 4 in cells that do not show this phenomenon (29-32). By increasing the intracellular 3': 5' -cyclic AMP concentrations through addition of dB-cAMP (33,34) an apparent contact inhibition of growth could be induced in transformed fibroblasts. Some investigators suggested that the rise in 3':5'-cyclic AMP concentrations noted at confluency only reflected a depletion of serum factors and is not correlated with contact inhibition of growth (35). However, upon addition of trypsin (25,31), insulin (25) or serum stimulation (35), 3':5'-cyclic AMP levels first fall before cell division occurs. When 3';5'-cyclic AMP levels were measured in « cells at different stages of the cell cycle it was found that the cAMP levels were lowest during mitosis and rose again during the G-1 phase (25,36). 3':5'-cyclic AMP has also been implicated in the regulation of mobility of fibroblasts. \*7hen these cells were grown in the presence of dB-cAMP or prostaglandin E, which also increases intracellular 3':5'- cyclic AMP (37,38) levels the cells became less mobile (39). Further- more, the adhesion of these cells to the substratum was increased (40) when grown in the presence of dB-cAMP or theophyline. The ordered growth patterns observed in CHO cells upon addition of dB-cAMP to the growth medium (2) may also be due to increased cell-cell recognition and .to reduced mobility and stronger adhesion to the substratum of dB-cAMP treated cells. The opposite effects were seen in fibroblasts grown in the presence of insulin. These lost their ordered growth pattern and were freed from contact inhibition of growth while their 3':5'-cyclic AMP levels fell (30) . Another aspect of cell growth in the presence or absence of drugs increasing 3’:5'-cyclic AMP concentrations, is the change in transport observed for many nutrients (41-44). Especially important in relation 5 to this study is the fact that different cell lines show a contrasting • effect of these drugs on thymidine uptake. The uptake of thymidine was reported to increase in SV40-CV-1 (43), a monkey kidney cell transformed by SV40 while it decreased in CHO cells (44) upon addition of cyclic AMP to the medium. From this observation it is clear that incorporation of 3 . . . . H-thymidine in DNA is not a good measure of the growth rate of cells treated with different drugs. When CHO cells are grown in the presence of dB-cAMP or other drugs or hormones which increase intracellular 3' :5' -cyclic AMP levels, they « undergo a process which was termed "reverse transformation" by Hsie and Puck (2) . They typically revert back from a compact morphology to a more elongated form which resembles "normal" fibroblasts. They align in parallel arrays, they show a decreased agglutinability by plant lectins and an increase in synthesis of collagen (3) and of acid muco- polysaccharides (45). There is also a shift in fucose-containing glyco- peptides at the cell surface towards components of lower molecular weight as estimated by chromatography on Sephadex G-50 (46,47). CHO cells can be enucleated by treating them with cytochalasin B, a drug which dis- rupts microfilaments in the cells (48-50). These enucleated CHO cells contain all cell-organelles, regain their normal shape and display some motility after enucleation. They also elongate upon treatment with dB-cAMP (50). These observations suggest that cellular elongation in response to dB-cAMP is independent of nuclear events (i.e. transcription of RNA) . This is also consistent with experiments indicating that elonga- tion induced by dB-cAMP is independent of protein or RNA synthesis (51) . The role of 3':5'-cyclic AMP in cellular growth and morphology prompted investigators to study the possible role of an altered 3':5'- 6 cyclic AMP metabolism in tumor cells which have abnormal growth patterns. Lower levels of cyclic AMP were indeed found in several transformed lines as compared to their normal counterparts (52,53) and it was suggested that this might be the basis of their altered growth properties. The most convincing results were obtained with several cell lines transformed by a temperature-sensitive mutant of the Rous Sarcoma Virus. When these cells are grown at the permissive temperature, i.e. « when the viral function is expressed, but not at the nonpermissive temperature, there is a decrease in 3':5'-cyclic AMP content because of a decrease in adenyl cyclase activity (54,55). In another study using a temperature sensitive mutant of SV40, the growth of the transformed cell was temperature sensitive, however, although the 3':5'-cyclic AMP levels were lower in these cells than in the parental line, they were not temperature sensitive (56). dB-cAMP also prevents the induction of DNA synthesis induced by adenovirus type 12 in BHK cells arrested in G-1 (57) and reduces the tumorigenicity of Celovirus transformed CHO cells (24). Adenyl cyclase, the enzyme which catalyzes the synthesis of 3':5'-cyclic AMP from ATP, is part of the plasma membrane (58). Its activity can be modified by extracellular agents which bind to plasma membrane receptors (58-60). Insulin, which affects adenyl cyclase, can t be immobilized on Sepharose beads and still retain its effectiveness even though it cannot enter the cell. This indicates that the inter- action of insulin with superficial membrane structures alone may suffice to affect adenyl cyclase activity (59). The fluid mosaic model" of the plasma membrane A number of models have been suggested to explain the structures and properties of the plasma membrane (see reviews 61-67). The one most 7 f accepted has been proposed by Singer and Nicolson (68). They recognized that the membrane was not a rigid structure, but a dynamic, fluid organelle. Figure 1 shows the model they proposed with modifications and additions accepted since its publication. Basically the membrane is believed to consist of a lipid bilayer in which the lipid molecules are arranged with their hydrophobic aliphatic chains pointing inwards towards each other and the polar heads forming the outer edges. This structure resembles a flattened micelle. Globular proteins are distributed along and within this bilayer. Some only penetrate partially into the bilayer, others span the entire membrane (69,70). Although it was elegantly shown that membranes are fluid (71), other experiments provide evidence that membrane proteins are not com- pletely free to move and can also assume a non-random distribution that is characteristic for a particular cell type or physiological state and which can be modified (72-75). For example, different topographic changes can be induced in transformed cells as compared to the normal parental cell lines using plant lectins such as concanavalin A (76-79). For while the Con A binding sites on transformed cells formed clusters, there was a more random distribution on the normal cell types. Where cells were in contact (i.e. agglutination sites) clustering was particu- larly evident. It was proposed that the lectin molecules were involved in forming multiple cross-links in these regions. It seems, therefore, that the lectin receptors on normal cells migrate less readily in the plane of the membrane than in their transformed counterparts. This implies the existence of forces that limit free diffusion of membrane proteins and suggests that there may be differences in the nature or effectiveness of restraining or limiting forces on membrane proteins. 8 * D E Figure 1. A schematic representation of the Fluid Mosaic Model of plasma membranes. t * A - Proteins carrying carbohydrate moieties. B “ Trypsin labile protein, may be connected in some way to A. C - Membrane particles. D - Lipid bilayer. E - Microtubular framework. F - Carbohydrate moieties. G - Microtubules or micro filaments connecting A with E. 9 There were not sufficient differences in lipid composition to produce overall differences in membrane fluidity. A molecule with a polar region probably cannot move from the inside to the outside of the bilayer or vice versa (80). This would be thermo- dynamically unfavorable since the polar region would have to pass through the non-polar bilayer. However, as seen before, the lipid bilayer allows 4 the globular proteins to diffuse laterally in the plane of membrane, subject to the restraints mentioned above. Some membrane proteins extend outward from the plane of the membrane These proteins often carry carbohydrate branches on a protein backbone, while lipids are completely embedded within the bilayer, with only the polar head, which carries the carbohydrate moieties extending outward The different saccharide residues are localized mainly on the outer face of the membrane (81,82). These cell surface molecules are believed to carry some of the antigenic determinants and many hormone receptors of the cell (64) . There is some evidence that these glycoproteins are linked on the inner face of the plasma membrane to parts of an internal framework of microtubules and microfilaments, as indicated in Figure 1. Microtubules and microfilaments are structural components visible only by electron microscopy. Both are present in cultured fibroblasts (83—87) . Microtubules are composed of polypeptide subunits called tubulin. Colchicine and vinblastin disrupt these microtubules. dB-cAMP on the other hand appears to induce the polymerization of tubulin into microtubules without increasing the total amount of tubulin present in the cell (52) . The exact mechanism by which dB— cAMP induces polymeriza- tion is not known, although it may involve phosphorylation of the protein 10 . dimer (88) . Both actin and myosin were found in the microfilament bundles, and it is believed that they can provide the motility and con- tractility of the cell (87). In fibroblasts microtubules are mainly ^li§riGd in parallel to the long axis of the cell. When cells are treated with compounds which disrupt microtubular structure, the cells lose their elongated shape, suggesting that microtubules are involved in the maintenance of cell morphology. binlcages between microtubules and membrane— bounded vesicles have been detected by electron microscopy (89). This may explain the decrease in collagen secretion and a reported increased number of deleted Golgi vesicles, agents providing the directive channels for movement of secretory vesicles to the cell surface upon disruption of the micro- tubules (90) . Microtubules have been implicated as the restraining forces involved in agglutinin receptor-protein mobility (91-97), and in phago- cytosis (97). Upon transformation a decreased amount of membrane associated actin in fibroblasts have been reported, although the total amount of actin in the cell remained constant (98). The myosin content of transformed rat kidney cells was decreased to half that of normal cells. However, dB-cAMP did not affect the myosin content, although it did affect the morphology of the cell (85) . It seems that the micro- tubules and microfilaments form a filamentous subcellular micro- architecture which may be very important in several cell functions. However, the exact mode of involvement is not yet knmm. Complex saccharides at the cell surface Under electron microscope many cells have a fuzzy region around their peiiphary. Histochemical observations suggested that was largely carbohydrate in nature (99,100,70,71). Evidence suggests that much of 11 the carbohydrate is located at the cell surface and protrudes out from the membrane (71,101). However, in the literature there is no adequate distinction between the true cell membrane proteins discussed above and the extracellular coat. The latter is probably loosely attached, secreted material rich in mucopolysaccharides such as hyaluronic acid, chondroitin sulfate, and heparan sulfates (100). A number of changes in the cell surface carbohydrates have been reported to accompany trans- formation, the most well defined are discussed below. a) Mucopolysaccharides. Using staining techniques it was observed that the surface coat of transformed cells was thicker than that of normal cells (102-104). On the other hand, a decreased rate of synthesis of acid mucopolysaccharides was observed, which could be counteracted I by addition of dB-cAMP and theophylline to the growth medium. Others have reported a tenfold increase in [H] -glucosamine incorporation into the hyaluronic acid fraction following viral trans- formation but no increased incorporation in the sulfated fraction of the cell coat (105). The same authors reported an increase in the average molecular size of cell coat molecules. By contrast, other authors using different cell lines have reported contradictory results (106). It seems that there is no consistent pattern of altered mucopolysaccharides when cells become malignant. b) Glycolipids. Changes observed in this class of molecules ' include a general trend toward simplification, e.g. size reduction of carbohydrate moieties upon transformation (107-110) and quantitative changes in normal cells but not transformed cells during the cell cycle (111). However, some investigators report no. changes upon transformation (112-114). 12 Recently it was shown that upon labeling the cells by cell sur- face specific methods, transformed cells lacked label in "galactoprotein" present in normal cells but did have label in a glycolipid which was not labeled in normal cells (115-116) . It was also shown that when cells were grown in the presence of dB-cAMP or dextran sulfate, the glycolipid labeling pattern of a transformed ceil became indistinguishable from that of a normal cell (117). Results of comparative experiments on glycolipid content of normal and transformed cells are difficult to compare because of the effect of growth condition and cell density on glycolipid content (118). Furthermore, an analysis of glycolipid content in a number of contact inhibited clones showed that there were no consistent changes in lipid patterns related to degree of malignancy (119) . Again results have been confusing and it is not justifiable at the present time to make any absolute correlations between surface glycolipid composition and the transformed state. c) Glycoproteins. Several differences were observed in membrane teins after transformation. These include an increase in apparent molecular weight of the carbohydrate moieties of certain sur- face glycoproteins, those that contained L-fucose (9-12), the disap- pearance of a high molecular weight glycoprotein (120,121), and altered glycosyl transferase activities at the cell surface (122-125) . Transformed cells have certain glycoproteins of average higher molecular weight at their cell surface than do normal cells. It was shown that this difference is due to an increased amount of sialic acid on the f ucose-containing glycoproteins (10) . Warren and his associates also reported that transformed cells have a fivefold higher sialic acid 13 transferase activity than do normal cells. This enzyme was specific for particular desialated glycopeptides and was not active towards desialyzed fetuin, the normal substrate for such assays. Other workers, in contrast, have shown that the overall glycosyl transferase activities of normal cells were increased as compared to transformed cells (122,123). Furthermore, normal cells could only transfer galactose to adjacent cells, whereas transformed cells (126) and mitotic normal cells (127) can transfer galactose to receptors on the same cell, i.e. endogenous receptors. It has been suggested that these surface enzymes play a role in contact inhibition of growth. Complexes between a glycosyl transferase on one cell and an incompleted carbohydrate chain, the substrate, on adjacent cells are visualized as forming the corsS“links . This has been disputed by other investigators (128) . Finally it has been reported that transformed cells lacked a high molecular weight glycoprotein when plasma membrane polypeptide compositions of these and normal cells were compared by polyacrylamide gel electrophoresis (120,121). This protein had a molecular weight Sweater than 200,000 and was removed from normal cells by brief treat- ment with trypsin. In cells transformed by temperature sensitive virus it appeared only at the permissive temperature. A similar protein could be Induced in neuroblastoma cells grown in the presence of dB-cAMP (129) . Under these conditions the neuroblastoma cells underwent morphological differentiation, i.e. growth of long processes resembling neurites was induced. At present the functional significance of such a protein is unclear , d) .Lhe architecture of the cell surface. The orientation and dxstribution of the cell surface glycoproteins in the plasma membrane 14 may be very important in determining cell surface characteristics. Useful methods of studying these parameters are: 1) the agglutination of cells by plant lectins. Glycoproteins characteristically bind to plant lectins such as concanavalin A (130,131). Although it was deter- mined that transformed cells do not bind significantly more lectins (132,133), transformed cells are more agglutinable than are their normal counterparts. As was pointed out earlier, investigators believe that this is due to a higher mobility of the binding sites, and hence glyco- proteins in the membranes of transformed cells; 2) chemical labeling techniques. In these methods the radiolabel was supplied either in an appropriate precursor, through membrane impermeable reagents which reacted with available groups on surface molecules (134,135) or by using enzymes which presumably do not enter the cell to oxidize surface molecules, and subsequently reducing these same molecules with [ H]- sodium borohydride (136,137). An advantage of supplying the label through precursors is that one can use contrastingly labeled compounds. The advantages, disadvantages and molecular bases for the methods employed in this study are detailed in the relevant section of the Materials and Methods chapter. Concluding statement It is clear from the preceding discussion that although differences have been obtained between nomial cells and their transformed counter- parts, results have often been contradictory. In part this may be due to "genetic drift," since transformed cells in culture are genetically unstable. Therefore, many invalid comparisons have probably been made between cells which are genetically quite dissimilar. The best systems devised up to now for studying changes accompanying transformation have 15 been cells transformed by temperature sensitive oncogenic viruses. In this case differences in cell surface composition, agglutinability and morphology have consistently been observed between cells grown at the permissive and non-permissive temperatures. However, even in these experiments it requires at least a full cell cycle, frequently 24-30 h, before the surface becomes "modulated" from the transformed to the normal phenotype. In the experiments described in this dissertation I have chosen to work with a series of Chinese Hamster Ovary lines, some of which respond morphologically to the dlbutyryl derivative of 3':5'-cyclic AMP. The main aim of these experiments was to study the changes in cell surface which accompany these morphological changes. In particular, these lines seemed to be an excellent model for comparing groups of cells which were genetically similar, but whose growth and pattern of growth had been dramatically modified. In addition it seemed that they might provide a system for studying the surface events that normally accompany transfor- mation, since the treated cells have been reported to: a) lose their agglu- tinability by plant lectins, b) show changes in cell surface glycopeptide composition reminiscent of the reverse of those that normally occur fol- lowing transformation. In addition, there were the following advantages to this system: a) some of the lines used do not respond morphologically to dB-cAMP , so that the biochemical events responsible for the change in morphology might be distinguished from the cell surface events using the different cell variants, b) as we shall show later, some of the changes occur both rapidly and synchronously, c) the growth rate and final cell densities of the CHO cells are unaffected by the presence of dB-cAMP, so that comparisons can be made between cells in a similar state of growth. MATERIALS AND METHODS Materials Chemicals and radiochemicals The chemicals and radiochemicals used in this study are listed with their source in Table 1 and Table 2 respectively. Table 2 also lists the specific radioactivity and labeled position of the radio- chemicals. Purification of plant lectins Wheat germ agglutinin was prepared by a modification of the method of Nagata and Burger (138) from commercially available wheat germ. Unprocessed whole wheat germ was ground fine in a mill. The rest of this procedure assumed a starting quantity of about 100 grams of wheat germ. The powder was extracted with a liter of 0.05N hydrochloric acid for 1 hour at room temperature. The. residue was centrifuged at 15,000 x G for 20 minutes. The extract, brought to 35% ammonium sulfate saturation by slow addition of the solid salt, was kept stirring for 1 hour in an ice slurry (4°C) and centrifuged for 20 min at 15,000 x G. The precipitate was resuspended in 200 ml of 0.05N hydrochloric acid. Insoluble material was removed by centrifugation. n-Butanol was added dropwise to give a final concentration of 20% (v/v) with constant stirring at room temperature. After one hour of stirring, the mixture 17 TABLE I. Sources of chemicals Compound Manufacturer N,N,N' ,N' ,-tetramethyl6thylene diamine Ammonium persulfate Cyanogum 41 Eastman Kodak Company Rochester, New York ibid ibid l-ethyl-4- (isopropylidene- hydrazino) -IH-pyrazola- [ 3 , 4-b ] pyridine-5-carboxylic acid, ethyl ester Squib Institute for Medical Research Princeton, New Jersey Cordycepin Cyclohexamide Sigma Chemical Co. St. Louis, Missouri Actinomycin D N^,o2-dibutyryl adenosine 3':5'-cyclic monophosphoric acid Galactose oxidase Fluoresceine mercuric acetate Streptomyces protease Trypsin Chymotrypsin Ovalbumin Bovine serum albumin Neuraminidase ^ • Lac toper oxidase Calbiochem International La Jolla, California Sephadex (G-50) Diethyl amino ethyl cellulose Pharmacia Fine Chemicals Piscataway, New Jersey Concanavalin-A-sepharose Biorad P-2 Bio-Rad Laboratories Richmond , California Glutamine Me Coy's 5 A medium Fetal calf serum Antibiotic-antimycotic Trypsin Grand Island Biochemical Co. New York 18 TABLE 2. Sources and specific activities of radiochemicals Compound Label Specific Activity Manufacturer L-fucose [1-^^C] 57 Ci/mol Amersham-Searle [1-^H] 1 Ci/mmol Arlington Heights, 111, L-leucine [4,5-\] 38 Ci/mmol 348 Ci/mol Sodiumboro- hydride [^H] 570 Ci/mol Acetic anhydride [1-^^C] 24 Ci/mol Acetic anhydride [^H] 51 Ci/mol Thymidine Methyl [^H] 48.5 Ci/mmol New England Nuclear Boston, Massachusetts Sodium iodide 17 Ci/mg International Chemical and Nuclear Waltown, Massachusetts 19 was centrifuged at 5,000 x G for 30 min. The uppermost n-butanol phase and fluffy interphase were removed by suction followed by filtration. Lower clear water-phase was dialyzed against 2 liters of 0.05N hydro- chloric acid overnight. Ammonium sulfate was added to 35% saturation as in step 3, and centrifuged as above. The precipitate was resuspended in 20 ml of 0.05N hydrochloric acid and dialyzed extensively against O.OIM Tris-HCl buffer, pH 8.5 overnight. The dialysate was centrifuged to remove insoluble material and applied to a DEAE-cellulose column (2.5 x 20 cm) which was equilibrated with O.OIM Trls-HCl buffer, pH 8.5. Elution was performed with approxi- mately 300 ml of the buffer, with a flow rate of 0.5 ml/min. Wheat germ ^8§lubinin was not adsorbed on the cellulose. Active fractions were pooled, dialyzed against distilled water and lyophilized. Concanavalin A was prepared from jack bean meal (Sigma Chemical Company, St, Louis, Mo.) by a modified procedure of Agrawal and Goldstein (139). In this procedure 500 g of jack bean were stored overnight at -90°C in order to make the seeds brittle. 200 g aliquots of the frozen bean were then ground (at 4°C) to a fine white powder in a mechanical tissue grinder. The powder was passed through a series of fine mesh screens to remove any large pieces of the seed coat. The powder was extracted for 12 hours in 0.15M NaCl (4:1 (w/w) , NaCl:powder) at 4 C. After 12 hours the residue was spun down at 3400G and the supernatant collected. The remaining residue was re— extracted with 0.15M NaCl at 4°C for 12 hours and the supernatant decanted after centrifugation. 20 The extract was then brought to 40% saturation with solid ammonium sulfate. The precipitate was spun down at 10,000G and discarded. The supernatant was then brought to 60% saturated ammonium sulfate, the precipitate collected and resuspended in l.OM NaCl. The precipitate was then dialyzed against distilled water for 12 hours at 4°C. The dialysate was added to 1 liter of Sephadex G-75 in PBS, and stirred for 1 hour at 4°C. This solution was poured into a column, which had a void volumn of 120 ml. The column was eluted with O.IM phosphate buffered saline (pH 7.2). 2.0 ml fractions were collected with an Isco fraction collector and monitored for protein at 280 my in a Gilford Spectrophotometer. Most of the protein was removed at the void volume. After the column had been washed free of extraneous protein, the elutant was changed to O.IM glucose in O.IM phosphate buffered saline (pH 7.2). Con A, which had been bound to the dextran, came off as a single, symmetrical peak. The fractions containing the concanavalin A were collected, dialyzed overnight against O.IM phosphate buffered saline (pH 7.2) and stored desicated below 0°C. 500 g of jack bean gave a final yield of approximately 3.0 g of concanavalin A. Cells Chinese Hamster Ovary (CHO) cell lines K-1, K-l-M-7 and K-1-24-2 were obtained from Dr. Abraham Hsie, Oak Ridge National Laboratory. The other line (clone H-7) originated in the laboratory of Dr. R. M. Humphrey, Texas Medical Center, Houston. All these cells were derived from the original L-proline requiring CHO clone K-1 of Kao and Puck (1). 21 Maintenance of Cell Lines The cells were grown in 75 cm^ Falcon flasks, Bioquest, Cockeysville MD, in 20 ml McCoys 5A medium, supplemented by 10% v/v fetal calf serum, 2 mM Glutamine and an antibiotic antibicotic mixture with a final con- centration of 10^/liter penicillin, 83.3 meg/1 Fungionone and 10^ meg/1 streptomycin. The cultures were routinely passaged 1:10 upon reaching approximately 90% confluency. Trypsin solution, (0.25% w/v Hank's balanced salt solution) was used to free the cells from the plastic. For growth, the temperature was maintained at 37°C in a moist incubator in which the CO2 tension was 5%. Histological and Microscopic Methods Cells were grown in Falcon dishes. The drugs were added at the specified concentrations, under sterile conditions. When the cells reached 60-80% confluency , the medium was discarded and the plates -were washed free of medium with PB. The cells were fixed for 15 min in 10% (v/v) f ormaldehyde-PB, and were dehydrated by sequentially incubating them for 5 min in 20, 40 and 60% ethanol— water . The cells were then stained for 15 min with 0.1% (w/v) Toluidine Blue in 70% ethanol-water. The excess stain was washed off with PB and the monolayers were allowed to air dry Overnight. A drop of immersion oil was placed on the cells and covered with a coverslip. The cells were photographed under a Wild Heerburg microscope, with camera attachment. A Mamayia/Sekoz TL-500 camera and Kodak Tri-X 135 Pan film were used. Measurement of the Incorporation and Uptake of [ ^H] -Thymidine Cells were gro’.m in the presence or absence of 1 uJH dB-cAMP in T-20 Falcon flasks. The procedure used to determine incorporation of 22 Fig. 2. Flow chart demonstrating the procedure followed in measuring thymidine uptake and incorporation in CHO-cells.' 3 0.5 ml PB containing 12.5 pCi [ H]-thymidine were added to 2 ml of medium on which the cells were growing. after 1 h V The monolayer was washed 3 times with 3 ml PBS. 1 The cells were incubated for 1 min in 1.0 ml of 0.25% Trypsin , 1 Trypsin solution was poured off and cells were allowed to stand for 1 min. - 1 ml of cell solution was used for determining number of cells, the remaining 4 ml was centrifuged for 5 min at 1200 rpm. I Pellet was resuspended in 5 ml cold TCA (5%) and allowed to stand at 4°C overnight. (The pellet was broken-up with a syringe with a fine needle.) I The solution was centrifuged for 5 min at 1200 rpm. I 1 ml of supernatant was counted for radioactivity,^ pellet was washed with 5 ml cold TCA (5%) . I • The pellet was resuspended in 1 ml water, and the radioactivity was determined. 23 3 [ H] -thymidine is detailed in Figure 2 . When the culture reached the desired density, a certain time after plating, [^H] -thymidine was added for 1 hour. The radioactivity was washed off and the cells were removed from the flask. An aliquot was used for counting cell number, the remainder of the cells were broken and the DNA precipitated with cold trichloroacetic acid, and radioactivity in DNA was determined. An aliquot of the supernatant was counted as a measure of soluble thymidine 3 and uptake of [ H] -thymidine . Radiochemical Labeling Technique with Precursors In order to compare the proteins or glycoproteins of the plasma membrane of the cell lines, one line was grown on an appropriate [^H]- labeled precursor, while the other line was grown on the contrasting [ C] -precursor . For studying protein composition radiolabeled L-leucine was employed. For glycoproteins either L— fucose or D-glucosamine were the chosen precursors. When the cultures reached approximately 80% confluency, either plasma membranes were Isolated, or the cells were trypsinized for the study of the surface components released by the enzyme. Contrastingly labeled samples from the different lines, or from the same cell line grown under different experimental conditions (e.g. - dB-c-AMP) were then mixed and processed together. Trypsinization of cells Cells were grown in roller bottles containing 50 ml McCoy's 5A medium, and on medium containing either dB-c-AMP, Sq 20009 or other drugs at the concentrations indicated, in the presence of either 3 1 A [ H]- or [ C] -fucose. These bottles were rotated at 2 rpm and main- tained at 37°C. Upon reaching 80% confluency, the. cells were rinsed 24 with 20 ml of PBS, and rotated at 2 rpm for 8 min at 37 °C in 10 ml of 0.25% (w/v) Trypsin-PBS. The cells were centrifuged for 3 min at 800 rpm and the supernatants of the two samples with contrasting labels were mixed. The mixed solution was heated at 100°C for 2 min to inactivate the trypsin and digested with pronase to hydrolyze the remaining peptide material. The digest was dried, redissolved in 4 ml water and filtered through glass wool to remove large particles of insoluble precipitate. .Column-Chromatography of Trypsinates • Molecular sieve chromatography ' The filtered trypsinate was desalted on a column of Bio-gel P-2. The eluant was water, and the elution rate 50 ml h~^. In each case 55-60% of the total radioactivity appeared as a peak coinciding with the void volume of the column. The remaining radioactivity was largely free in the form of L-fucose and eluted in the included volume of the column. The material collected at the void volume was concentrated to 2 ml and chromatographed on a column of Sephadex G-50 (90 x 1.5 cm; void volume 55 ml). The elution rate was 0.52 ml min ^ and each fraction had a volume of 2.6 ml. The column was equilibrated with PB and this buffer also served as eluant. Ion-exchange chromatography The fractions 20 to 50 obtained through G-50 molecular sieve chromatography were concentrated and desalted on the Bio-gel P-2 column described in the previous section. The fraction eluting in the excluded volume was collected and concentrated to 4 ml. This sample was applied to a diethylamino-ethyl-sephadex column (7 x 0.9 cm) in water. The column was washed with 50 ml of water after which a salt 25 gradient was applied. This gradient ranged from 0 to 0.5M, sodium chloride in O.OIM Tris-HCl at pH 8.2 and was non-linear. The gradient was made by connecting 3 chambers in series. The first two chambers contained only 100 ml of buffer , while the last chamber contained 100 ml 0.5M sodium chloride in the buffer. Five ml fractions were collected and 2 ml of each fraction were used to determine radioactivity. Hydrolytic Procedures Used in Analysis of Carbohydrates Removal of N-acetyl neuraminic acid using neuraminidase Hydrolysis of whole cells. A confluent roller bottle containing 3 cells labeled with [ H]-fucose was washed with PB and incubated for 45 min at 37°C with 20 ml PB containing 2 units of neuraminidase. The cells were then trypsinized. This trypsinate was mixed with the tryp- sinate of a control culture labeled with [ C]-fucose and co-chromato— graphed as described. Hydrolysis of carbohydrate moieties. The fractions 20 to 50 obtained from G-50 molecular sieve chromatography were dried under vacuum, desalted on a Bio-gel P-2 column as before, and evaporated to dryness again. The sample was then dissolved in 3 ml. of O.OIM, sodium acetate buffer, pH 5.4. 0.2 units of neuraminidase were added and the mixture was incubated for 20 hours at 37°C. This sample was re- chromatographed on the G-50 column, or fractionated by ion-exchange chromatography. Removal of N-acetylneuraminic acid using hydrochloric acid • ’ The fractions 20 to 50 obtained from G-50 molecular sieve chroma- tography were dried under vacuum, desalted on a Bio-gel P-2 column as before, and dried again. The sample was then incubated in 1 ml O.IN hydrochloric acid at SO^’C for 1 hour. 3 ml of PB was added at this 26 « time. This final mixture was re-chromatographed by gel filtration on the G-50 column, or by ion-exchange. Analysis of label The fractions 20 to 50 obtained from G— 50 molecular sieve chroma- tography were dried under vacuum, desalted on a Bio-gel P-2 column and dried as before. The sample was then incubated in 5 ml O.IN or 0.5N hydrochloric acid at 100°C for 24 or 48 h. The hydrolyzate was evaporated to dryness and dissolved in 200 pi water. This sample and several reference standards were spotted 1 inch apart on a sheet of Whatman number 1 chromatography paper. Descending chromatography was performed using a solvent system of butanol, acetic acid and water in a ratio of * 37:25:9, by volume. A strip containing the sample was cut from the chromatogram. The rest was stained to determine the relative mobility of the standards. The strip containing the sample was scanned using a Packard Model 7201 Radiochromatogram Scanner, then cut in 1 cm strips perpendicular to the direction of separation, and the radioactivity in each was determined by liquid scintillation counting. Gel Electrophoretic Techniques Electrophoresis Polyacrylamide gels were prepared according to the method of Laemmli (140), using the buffer systems of Davis (141). The polyacryla- mide gel consisted of two sections: (1) a large-pore gel, 5% (w/v) polyacrylamide with 5% (w/w) bis-acrylamide-polyacrylamide in which electrophoretic concentration takes place (stacking-gel) , and (2) a small- pore gel in which electrophoretic separation takes place (running gel). The gels varied in polyacrylamide concentration but always contained 5% (w/w) bis-acrylamide-polyacrylamide. A pxiritied tiiixtsire of polyacrylamide 27 and bis-acrylamide at this ratio (Cyanagura) was purchased from Eastman Kodak, Inc. The diameter of the gels was always 5 mm i.d. Therefore, the length of the gel varied with the volume of gel solution in the glass tubes. For the large-pore gel, 0.3 ml gel solution was used. The gels were polymerized chemically using tetramethylene-diamine and ammonium persulfate. The buffers used are given in Table 3. The glass tubes were tightly closed at one end with flat topped rubber stoppers. These tubes were arranged vertically and filled with the appropriate solution volume of running gel. A layer of water was carefully layered on top of this solution and the gel was allowed to polymerize. The water was then removed, and 0.3 ml of stacking— gel (large pore gel) solution was added, carefully overlaid with water, and allowed to polymerize. A Buchler Analytical Temperature Regulation Polyacrylamide Gel Electrophoresis Apparatus was used for the electrophoretic* separation (see Figure 3). The same buffer, compartment buffer in Table 3, was used in both upper and lower chambers. Solid crystals of sucrose were added to the sample solution in order to increase its density and 100 yl to 200 yl of the sample, containing approximately 100 yg of protein was layered on the upper surface of the stacking gel. Electrophoresis was carried out by applying a current of 3 mA per tube, until the sample had entered the stacking gel. The current was then increased to 6 mA per tube. One gel, a control, was loaded only with the dye marker bromphenol blue. When the dye reached a point approximately 0.5 cm from the bottom of tube, the power v/as switched off. This took approximately 60 min for a 6.5 cm gel and 90 min for a 10.0 cm gel. 28 D Figure 3. A diagram of the Buchler Analytical Temperature * Regulated Polyacrylamide Gel Electrophoresis Apparatus. A - Stacking gel, B - Running gel C - Inlet for temperature regulating water D - Electrical connections 29 TABLE 3. Buffers used in polyacrylamide gel electrophoresis Compartment Buffer pH 8.3 .38M Glycine .1% (w/v) SDS .049M Tris .1% (v/v) 6~mercap- toethanol Stacking Gel Buffer pH 7.6 .00612M Tris .1% (w/v) SDS .0032M Phosphate Running Gel Buffer pH 8.9 .375M Tris .1% (w/v) SDS .006M Chloride I 30 The gels were immediately removed from the glass tube by gently rimming them with a fine needle through which a thin stream of water was passed. Staining and destaining procedures Prior to staining the gel, the SDS wah removed by shaking the gel in ethanol-acetic acid-water (40:10:50 v/v) overnight. This procedure also pre-"fixed" the protein, thus eliminating diffusion. The gels were stained for protein using 0.125% (w/v) Coomassie blue in the same ethanol-acetic acid— water mixture for 1 hour and diffusion destained in ethanol-acetic acid-water (10:7:83 v/v) until all excess stain was removed . Gel slicing Two methods were used for fractionation of the gels. Each gave comparable results. (1) The gels were fractionated using an "Autogeldivider " (Savant Instruments, Hicksville, NY), into approximately 60-80 fractions per gel. The crushed gel fractions were left overnight in 0.5 ml of water and their radioactivity content determined after addition of 5 ml of toluene-Triton X-100 scintillant (142) . (2) The gels were stained with Coomassie blue. They could then be photographed if desired. They were then frozen in a Revco ultra low temperature freezer at —70 C, before being sliced into 1 mm pieces, using a guillotine of razor blades. Each slice was incubated overnight in 0.5 ml of 30% hydrogen peroxide-water at 60°C in glass scintillation vials in order to solubilize the gel. Their radioactive content was determined after addition of 5 ml of 'toluene Triton X-100 scintillant. 31 External Labeling of Cells These methods are based on a reaction of non-penetrating reagents with reactive groups exposed at the outer surface of the cell. In all of the experiments the cells were grown until approximately 90% con- fluent, i.e. they were still in log phase growth, before treatment. lodination of the lactoperoxidase technique This is based on a modification of the method first described by Phillips and Morrison (137), and is relatively specific for tyrosine residues. The detailed mechanism of the reaction is still not under- stood, but leads mainly to the formation of monoiodo— tyrosine . Since lactoperoxidase is a relatively large molecule, it does not 125 easily penetrate the membrane and I is believed to be incorporated into proteins projecting out of the lipid bilayer. The procedure is outlir^sd in the flow chart shown in Figure 4. After thoroughly washing the monolayer, the cells were covered with buffered saline containing the enzyme and 50 pCi Na I. Hydrogen peroxide was then added to initiate the reaction. After incubation, the cells were washed and plasma membranes isolated by the method of Barland and Schroeder (143). Iti later experiments I attempted to reduce the amount of non-specific 125^ . _ j . ■ . , . . I introduction into lipid by first treating the cells with sodium borohydride. and Nal can react with the double bonds of unsaturated fatty acids (144) . I reasoned that reduction of these bonds might render the labeling procedure more specific for proteins. It should be emphasized that the above reactions seem sufficiently mild that labeled cells continue to grow with no measurable reduction in generation time after treatment. 32 Fig. 4: Flow chart demonstrating the procedure for lactoperoxidase mediated iodination of cell surfaces. Cell Monolayer (in Blake bottles, 400 cm^) wash 3 X with PBS Add 20 ml PBS containing 50 mg lactoperoxidase; 50 yCi Na i 20 pi aliquots of .06 (v/v) hydrogen peroxide added every min for 10 min. I Incubate at 37°C for 10 min with occasional agitation. wash 3 X with PBS Isolate plasma membranes. X X Labeling by the galactose oxidase technique This method is based on that originally described by Gahmberg and « Hakoraori (136) and is outlined in the flow chart shown in Figure 5. The cells were grown on Blake bottles until almost confluent, washed thoroughly and then treated with 2 x 10 sodium borohydride in order to reduce the amount of non-specific radioactivity incorporated by the latter treatment with [ H]— sodium borohydride. The monolayer of cells was then incubated with galactose oxidase, a treatment which leads to the oxidation of the primary hydroxyl group on exposed galactosyl groups of glycoproteins and glycolipids. 34 Figure 5: Flow chart demonstrating the procedure for galactose oxidase mediated labeling of cell surfaces. O Cell monolayers (in Blake bottles, 400 cm ) wash 3 X with PBS (20 ml) ▼ _5 Add 20 ml PBS containing 2 x 10 M sodium borohydride 1 Incubate for 15 min wash 3 X with PBS (20 ml) Add 10 ml PBS containing 10 units galactose oxidase I Incubate for 90 min 1 Collect suspension; remove remaining cells on bottle by • incubating with 15 ml of .02 w/v disodium ethylenediamine tetraacetate. Mix supernatant and cells. wash 3 X with PBS (20 ml) O Add 10 ml PBS containing 1 mCi [ H]-sodium borohydride 1 Incubate for 15 min wash 3 X with PBS (20 ml) V Isolate plasma membranes • 35 I found that under these conditions many of the cells detach from the substratum. Therefore, I chose to remove all of the cells using a solution of disodium ethylenediamino tetra acetate (0.02% in PBS) . The suspension was then washed with PBS and treated with 1 mCi of 3 [ H] -sodium borohydride in PBS in order to reduce the aldehyde groups on the C— 6 position of the galactosialdose. Finally plasma membranes were prepared by the method of Brunette and Till (145). Reduction of Schiff's bases formed between surface amino groups and pyridoxal phosphate This reaction is based on the reaction between surface amino groups and pyridoxal phosphate, and the subsequent reduction of the Schiff 3 bases with [ H]-sodium borohydride (135). This reaction is shown in Figure 6a. The procedure is outlined in the flow chart sho^m in Figure 6b. After thoroughly washing the monolayer, the cells were covered with PBS containing 0.002 mM pyridoxal phosphate, and incubated for 15 min. The monolayer was then washed with PBS and finally covered with PBS 3 containing 1 mCi [ H]-sodium borohydride. After 15 min the cells were • washed with PBS and plasma membranes isolated by the procedure of Barland and Schroeder (143). Radioactivity Counting Procedures Counting method of g-emitters Samples were counted in the soluene-Triton X-100 3:1 by volume) scintillation fluid described by Turner (143), using a Beckman LS-300 or Nuclear Chicago Isocap scintillation spectrometer. In single label ■ 3 4 'i experiments ' C was counted with about 95% efficiency, and H with 40% 36 'W H-C— fl jV^.3H, 3\ V H,i1 Remix supernatant in its entirety and recentrifuge at 3000 rpm for 10 min at 4°C i 43 Figure Continued Collect interphase using a syringe with a bent needle and dilute with water to 10 ml 1 Centrifuge for 15 min at 8500 rpm at 4°C i Boil pellet for 3 min in 100 pi of 0.04M Tris-HCl buffer, pH = 8.3, containing 2% (w/v) SDS and .1% (v/v) 3-mercap- toethanol Dialyze overnight in 3, 11 changes of 0.04M Tris-HCl buffer, pH = 8.3, containing 0.1% (w/v) SDS and 0.1% (v/v) 3-mercaptoethanol 44 TABLE 4. Preparation and composition of the two phases used in the Brunette and Till method of membrane preparation Mix : i - 200 g of 20% (w/w) Dextran 500 in water - 103 g of 30% (w/w) polyethylene glycol in water - 99 ml of double distilled water - 333 ml of . 22M phosphate buffer, pH = 6.5 _2 - 80 ml of 10 M zinc chloride This is mixed thoroughly and allowed to settle overnight in a separa- tory funnel. The two phases can be collected and are stored at 4°C until used. 45 . diaminotetraacetate. They were then washed with isotonic sodium chloride in water and the pellet of cells resuspended in hypotonic zinc chloride for 15 min at room temperature and subsequently for 5 min in an ice-bath. At this stage the cells were^ swollen to several times their normal size. They were ruptured in a Bounce homogenizer with a tight fitting pestle. The extent of cell breakage could be followed by phase contrast micros- copy. The plasma membranes were purified by centrifugation in a two phase polymer system. The preparation and composition of these phases are given in Table 4. The material which collected at the interphase between the two phases was collected, resuspended in water and centrifuged. The pellet was dissolved in SDS solution, and analyzed on SDS-polyacrylamide gels. Determination of Protein Concentrations determinations, except for those samples containing 3-mercaptoethanol, were performed by the procedure of Lowry et al. (146). There was no difference in standard curve for samples without SDS or those with 0.1% (w/v) SDS, see Figure 9. When the SDS concentration was 3% (w/v) there was a slight, but definite, difference in the standard curve. When 3-mercaptoethanol was present, the modified pro- cedure of Ross and Schatz (147) was used, since thiol groups interfere strongly with the standard Lowy assay. In the procedure of Ross and , Schatz the interfering thiol groups were precipitated prior to color development using iodoacetic acid. 46 ugr B5.A. Figure 9. Standard curves for the determination of protein concentration. 0 Determined by the method of Lowry et al. • “ethod of Lowry et al. , in the presence of 3A (w/v) SDS. X Determined by the modified procedure of Ross and Schatz. Bovine serum albumin was used as standard. 47 Determination of Agglutination by Wheat Germ Agglutinin and Concanavalin A In all the experiments in which lectin mediated cell-agglutination was measured, the cells were grown in 100 cm Falcon dishes, Bioquest, Cockeysville, MD, on medium defined in the section on cell maintenance. The cells were plated at a density of approximately 8 x 10^ cells per plate, and grown for at least 48 h to between 60 and 80% confluency. The cells were washed 3 times with 5 ml calcium magnesium free phosphate balanced saline, CMF-PBS, and subsequently 3 times with CMF-PBS con- taining 0.02% EDTA, and incubated for 15 min at 37°C in 5 ml of CMF-PBS containing 0.02% EDTA. The cells were washed off the plate, centri- fuged at 700 rpm for 3 min, washed twice with 5 ml PBS, and resuspended in 1.0 ml PBS. Cells from 5 dishes were pooled for each measurement. 0.2 ml of this cell suspension containing about 1.0 x 10^ cells/cm^ was placed in a spot plate maintained at 22°C. The desired concentra- tion of lectin was then added and the cells were incubated for 15 min. The total number of single cells, and of cells in aggregates of 3 or more were counted for various regions of the microscope slide. From this the average percent agglutination was determined, as the total num- ber of cells in aggregates, divided by the total number of cells, both free and in aggregates. Drugs were added, or removed from the medium at intervals and concentrations indicated for each experiment. ’ Determination of Intracellular Concentrations of 3':5'-Cyclic Adenosine Monophosphate The 3' : 5 '-cyclic AMP concentration in the cells was estimated using a modification of the protein binding assay of Gilman (148). A special kit containing the rucnropriate reagents was purchased from 48 Amersham Searle, Arlington Heights, 111, The assay is based on the competition for a binding site on a 3' : 5 ' -cyclic-AMP binding protein from bovine muscle between a known amount of [^H] -3 5 ' -cyclic-AMP added during the assay, and the 3 5 ' -cyclic-AMP in the sample. The 3 unbound [ H]— 3 :5 -cyclic-AMP is removed with charcoal and the amount 3 of [ H]-3' :5'-cyclic-AMP bound to the protein is determined by liquid scintillation counting. This is, of course, inversly proportional to the amount of 3 :5' -cyclic-AMP in the cell. The assay is most sensitive between .5 and 8 pmoles 3 5 ' -cyclic-AMP in the sample. 8 x 10^ cells 2 were xnoculated into 100 cm Falcon dishes containing 15 ml medium, as defined in the section on cell maintenance. By allowing the cells to grow for different lengths of time, the 3 5' -cyclic-AMP concentra- tions could be determined for cell cultures at different cell densities. After the cells had reached the desired density, the cells were washed thoroughly with PB and solubilized by incubation with 5 ml .IN sodium hydroxide for 5 min. This solution was transferred to a test tube, and 5 ml 5% trichloroacetic acid was added. The mixture was allowed to stand for 30 min at 4°C. This precipitated material was centrifuged, using the maximum setting of a clinical table-top centrifuge. The amount of protein in the precipitate was determined and used as an indirect measure of cell density. The supernatant was mixed 5 x with 10 ml water saturated ether and allowed to reseparate. The residual ether was removed by passing a nitrogen stream through the solution. The solution was Ij'opholyzed and the sample resuspended in 0,15 ml .05m Tris— EDTA buffer p}> = 7.4. The determination of the 3 .5 -cyclic-AMP concentration in the samples prepared as described above was performed at 0°C. 49 The procedure used is outlined in Figure 10. A known amount of 3 [ H]-3' :5'-cyclic AMP and binding protein was added to 50 pi of the samples. This mixture was allowed to equilibrate and the excess labeled and unlabeled 3':5'-cyclic AMP was removed with charcoal. The amount of bound [ H]-3' :5'-cyclic AMP was determined by liquid scintillation counting. Each determination was performed in duplicate and the average used in the calculations. To determine the blank value, tubes 1 and 2, which did not contain binding protein, were averaged. The radioactivity bound to binding protein in the absence of unlabeled 3' :5'-cyclic AMP (C^) was determined. Figure 11 shows a typical standard curve obtained by this method. 50 Figure 10: Experimental procedure used in the determination of 3 ' : 5 ' -cyclic-AMP concentrations. Pipette 150 pi .05M Tris-EDTA pH = 7.5 into assay tube 1 and 2. These tubes are for the determination of the blank cpm for the assay i Pipette 50 pi of the buffer into assay tubes 3 and 4 for the determination of the binding in the absence of unlabeled 3 ' : 5 ' -cyclic-AMP . * i Add 1, 2, 4, 8 and 16 pm 3 5 ' -cyclic 'AMP in duplicate to the next ten test tubes to obtain a standard curve i 50 pi of the unknown samples are added to the appropriate test tubes Add 50 pi of [^H]-cyclic AMP (25 pCi) to each test tube Add 100 pi of the binding protein to all test tubes except tube 1 and 2 • i Vortex each tube for about 5 seconds Allow the mixtures to equilibrate for 2 hrs at 2-4°C Add 100 pi of charcoal suspension to all tubes, and vortex for 10-12 seconds. Do not add charcoal to more tubes than can be centrifuged in one batch Centrifuge all tubes to sediment the charcoal i Remove 200 pi of samples from each tube and determine radioactivity 51 pM 3;5-cydic A^MP Figure 11. Standard curve for the determination of 3' :5' -cyclic AMP concentration. Co = The radioactivity bound to the binding protein in the absence ■ of unlabeled 3' :5'-cyclic-AMP. - . Cx = The radioactivity bound to binding proteins in sample number x. RESULTS Morphology of the Cell Lines When normal fibroblasts are transformed either by oncogenic viruses or by carcinogenic agents, they generally lose their elongated shape and assume a more compact morphology, similar to that of epithe- lial cells. Moreover, they no longer tend to align in parallel arrays, but show a more random distribution on the growth plate. The reverse of these changes were observed when cell lines K-1 and H-7 were grown on medium containing ImM dB-cAMP. The fibroblast-like line 24-2 was much less affected by the drug, but nevertheless, the cells did appear to be somewhat more elongated . M-7 which closely resembles K-1 and H-7 under normal growth conditions, showed no morphological or visible growth responses to dB-cAMP. Butyric acid, which would arise from any decomposition of dB-cAMP in the growth medium did not have any influence on the morphology of any of the cells when present at a concentration of ImM. The effects of 3' :5'-cyclic AMP on the morphology of the cells were much less pronounced than dB-cAMP. The 3':5'-cyclic AMP dies- terase inhibitor Sq 20009 had a similar, though less pronounced effect on line M-7, a line which was morphologically unaffected by dB-cAMP. .These results are summai'ized in Table 5 and Figure 12. Although the untreated cells vary in morphology but do not have significantly different 3':5'-cyclic AMP levels (Table 6), these results suggest that the morpho.logical responsiveness of the CHO cells does relate to a 53 TABLE 5. Effect of 3':5'-cyclic AMP, dB-c-AMP and Sq 20009 on the morphology of the CHO cell lines Cell Line No Addition 10 % c-AMP 10 % dB-c-AMP 10~'^M Sq 20009 K-1 Irregular + 4-f+ ± M-7 Irregular 0 0 + 24-2 Elongated + + + H-7 Irregular + ++ ± 0 no elongation; + - little elongation; I I I = pronounced elongation - ~ response could not be established unequivocally 54 Figure 12. The effect of dB-c-AMP or SQ 20009 on the morphology of CHO cells. The CHO lines K-1-24-2 (1) , K-1 (2) , H-7 (3) and K-l-M-7 (4) were grown on 10% fetal calf serum containing McCoy's 5A medium (A) in the presence of 10"\ dB-c-AMP (B) or 10“^M SQ 20009 (C). After staining with Toluidine Blue they were photographed on the culture dish. 56 TABLE 6. The intracellular 3':5'-cyclic AMP levels in the CHO cell lines Cell Line ■Number of Cells -(x 10^) yg Protein/Dish pM cAMP/mg Protein H-7 12.4 2688 • 1.79 15.4 3360 9.0 K-1 12.1 2640 2.4 18.2 3960 11.4 K-l-M-7 12.7 2760 3.1 16.0 3480 12.4 K-1-24-2 12.8 2784 4.8 18.7 4080 6.4 57 change in internal 3 ' : 5 ' -cyclic-AMP concentration within the cell. As discussed in the introduction, the lack of potency of 3 5 ' -cyclic- AMP itself possibly relates to either its failure to penetrate the cells (15-17) or to its greater instability (16,17) as compared to the dibutyryl analogue. The above results are in general agreement with those of others using a wide number of transformed fibroblast lines, as discussed in the introduction. The Effect of dB-cAMP on the Growth Rate of the Cells and on the Uptake and IncorporaTion of [^H] -Thymidine A number of other studies have revealed that dB-cAMP reduces the ' growth rate of transformed fibroblasts grown in culture (32-34). In the cells used in this study this was not the case. That all cells do continue through the cell cycle in the presence of dB-cAMP was shown by low serum synchronization of the cells. Cell cultures at approximately 30% confluency were transferred to 0.5% (v/v) fetal calf serum containing up to 36 h. Serum was then added to a final concentration of 10% (v/v) . Figure 13a shows that all cells do indeed divide syn- chronously after approximately 11 h. In the experiment shown in Figure 13b, I examined the growth rate of the four CHO cell lines in the presence and absence of ImM dB-cAMP. Cells (2 x 10^) were transferred to 25 cm Falcon dishes containing 5 ml of medium. After allowing the cells 12 h to attach to the substratum, cells from duplicate plates were removed by trypsinization at 12 h intervals, and counted using a Coulter counter. It can be seen clearly from Figure 13b that the growth rates of the ugII lines were not affected by dB— cAMP in the medium. In each case the generation tim.e was around ]6 n wnether or not the drug was present. 2 O Ul _J IjJ O (C V— o Or: ijJ m 3 2: 'O X 1 0 4 8 12 time (hJ o 16 Figure 13a. The effect of dB-cAMP on the synchronization of CHO-H-7 Highly confluent cells were subcultured and grown for 22 h in the presence or absence of 1 mM dB-cAMP. They were then transferred to medium containing 0.5% (v/v) fetal calf serum for 16 h. dB-cAMP was present in the same cultures as in the previous incubation. At 2 h intervals duplicate cultures were used for both sets of cells to determine cell number. The cells were removed from the plate by a 1/2 min incubation with 0.25% (v/v) trypsin-PB and counted with a Coulter counter. No dB-cAMP (0), 1 mM dB-cAMP (0). 59 Figure 13b. The effect of dB-cAMP on the growth rate of the CHO cells. The Cells were grown in the presence or absence of ImM dB-cAMP. At 12 h intervals cells were removed from a flask by trypsin and counted using a Coulter counter. ;M-7; control (fi) ; + ImM dB-cAMP (0) and + O.lmM Sq 20009 (x) K-1; control (0); + ImM dB-cAMP (0) ■ 24-2; control (6); + ImM dB-cAM? (0) H-7; control («) ; + ImM dB-cAMP (0) 60 The growth rate of the treated cells did appear to fall away in the last 12 h of the experiment, so that the final cell density was reduced by about 30% compared with the controls. However, it should be emphasized that the medium was not changed in these experiments, and might be more "depleted" in one set of cells than in the other. The final cell density achieved by the untreated CHO cells in an experiment of this kind was 6 x 10^ cells/plate. Addition of new medium does not increase this value significantly since mitotic cells are shed from the monolayer and remain in suspension. However, it seems clear from these experiments that dB— cAMP does not reduce growth rate significantly nor promote a condition that might be regarded as "contact inhibition of growth." In the same experiment in which I measured cell number, I also measured the ability of the cells to incorporate [^H] -thymidine into trichloracetic acid-soluble and -insoluble fractions. Every 10-12 h a pulse of radioactivity (2.5 pCi/ml [ H] -thjTnidine) was administered to duplicate flasks of cells. The cells were then trypsini'zed, washed with PBS and resuspended in 5 ml PBS. One ml of the cell solution was used for the counting of cells in a Coulter counter while the remainder was centrifuged. The pellet was resuspended in 5 ml cold 10% (w/v) tri- chloracetic acid-water and allowed to stand at 4°C overnight. The sus- pension was centrifuged and the radioactive content of the supernatant fraction and pellet was measured. Figure 14 shows the amount of H recovered from the cells. It is clear that the presence of dB-cAMP significantly increases the total uptake of thymidine by the cells. Therefore, a comparison of the amount of radioactivity incorporated into the cells As not a useful measure of Figure 14. The effect of dB-cAMP on the thymidine uptake in CHO cells. At 12 h intervals cells in log phase of growth were pulsed with 3 12.5 pCi [ H]-thymidine for 2 h.. The cells were extensively washed and then disrupted in 5% trichloroacetic acid. Incorporation of into trichloroacetic acid insoluble material was used as a measure of DNA synthesis, while the radioactivity in the supernatant was used as a measure of the uptake of thymidine. A H 7 Incorporation into DNA in the presence (2) or absence (1) of 1 mM dB-cAMP. Uptake in the presence (3) or absence (4) of 1 mM dB-cAMP. B K-1 Incorporation into DNA in the presence (1) or absence (2) of 1 mM dB-cAMP. Uptake in the presence (3) or absence (4) of 1 mM dB-cAMP. C 24-2 Incorporation into DNA in the presence (1) or absence (2) of 1 mM dB-cAMP. Uptake in the presence (3) or absence (4) of 1 mM dB-cAMP. D M-7 Incorporation into DNA in the presence (3) or absence (1) of 1 mM dB-cAMP or in the presence of 0.1 mM Sq 20009 (2). Uptake into DNA in the presence (6) or absence (4) of 1 mM dB-cAMP or in the presence of 0.1 mM Sq 20009 (5). TiME (h) 62 Cpm/CELL (xlO^) ln_.ro ui_.o o 63 the relative amounts of DNA synthesized. Secondly, Figure 14 indicates that as cell density increases the amount of thymidine incorporated declines even though the rate of cell division had not fallen (Figure 13b). Clearly, great care has to be taken in interpreting experiments m which rates of DNA synthesis are measured by determining radioactive thymidine incorporation by whole cells. Comparison of Plasma Membrane Polypeptides of the Different Cell Lines A comparative study of the plasma membrane of the different cell lines revealed no major differences in the polypeptides which are present as analyzed by gel electrophoresis in SDS-containing polyacryla- mide gels. In these experiments, the plasma membrane polypeptides were isolated according to the procedure of Barland and Schroeder (143). The method is highly reproducible, and the isolated plasma membrane fraction can be stored at -4°C for extended periods of time. From Figure 15 it can be seen that no marked differences were observable between two preparations which had been isolated as much as three months apart, and subjected to electrophoresis on the same polyacrylamide gel. Figures 16, 17a and 18 show gels of plasma membrane preparations stained with Coomassie Blue. Three concentrations of polyacrylamide were employed in order to allow a more complete resolution of the component polypeptides. In Figure 17a (7.5% (w/v) polyacrylamide) the migration of the polypeptides has been compared with four standards of known molecular weight. Figure 17b also shows a typical calibration cui've obtained for this gel concentration. Polypeptides ranged in molecular weight from around 120,000 to less than 15,000. Very little material was located at the very top of the ge] suggesting that the ®^^^6ilization in detergent, \v.au .■ lu:. '.lUd that.', i.ev* ptoteins of 64 Figure 15. Reproducibility of membrane preparations. Two different membrane preparations of CHO-K-1 labeled with 3 14 either H or C, and prepared approximately 3 months apart, were co-electrophoresed on 7,5% (w/v) polyacrylamide gels. The gels were sliced and radioactivity in each slice was determined. 3 X A ( L-[ H]-leucine newer preparation; L- [ ^C]-leucine older preparation) . Figure 16. Coomassle blue staining pattern after polyacrylamide gel electrophoresis of the plasma membrane polypeptides from CHO lines K-1 (a) , 24-2 (b) , H-7 (c) and M-7 (d) . Membranes were prepared by the method of Barland and Schroeder. Electrophoresis was carried out in 5% (w/v) polyacrylamide gels. 66 Figure 17a. Coomassie blue staining pattern after polyacrylamide gel electrophoresis of the plasma membrane polypeptides from CHO lines K-1 (a) , 24-2 (b) , H-7 (c) and M-7 (d) . Membranes were prepared by the method of Barland and Schroeder. Electrophoresis was carried out in 7.5% (vt/v) polyacry].amide gels. mcLwt. xIO 1 J ^ 0 .2 .4 .6 Mobility Figure 17b. Standard curve for molecular weight determinations using the 7.5% (w/v) polyacrylamide gel -system. Standards are BSA (Ml«/ 68,000); ovalbumin (tM 44,000); trypsin (tW 23,800) and chymotrypsin- ogen (tW 25,000). Figure 18. Coomassie blue staining pattern after polyacrylamide gel electrophoresis of the plasma membrane polypeptides from CHO lines K-1 (a), 24-2 (b) , H-7 (c) and M-7 (d) . Membranes were prepared by the method of Barland and Schroeder. Electrophoresis was carried out in 10% (w/v) polyacrylamide gels. 69 high molecular weight analogous to spectrin of erythrocytes (149) were present. The sharp band at the bottom of the gel corresponded with the dye front in the 7.5%. (w/v) and 5% (w/v) gels. It probably represents polypeptides of relatively low molecular weight which can be resolved only in the 10% (w/v) gel system (Figure, 18) . Note that in Figure 18 when 10% (w/v) polyacrylamide concentrations were used, this band is not present. Five major polypeptide species (molecular weights 38,000, 47,500, 54,000, 58,000, 68,000) can be identified in each CHO cell line and are indicated by arrows on Figure 17a. No qualitative differences appear to distinguish the preparations from each other at any of the gel concentrations employed. Similar numbers of bands with approximately similar staining intensities can be observed. . In order to make quantitative comparisons in the polypeptide compositions of the plasma membranes derived from the different cell lines I grew the cells to 90% confluency on radioactive L-leucine as described in the Materials and Methods section. In this experiment, the line CHO-K-1 was provided with L- [ ]-leucine (0.16 pCi/ml medium) for 48 h, and the other lines with L-[\]-leucine (0.9 yCi/ml medium). Portions of labeled membrane material were elec trophoresed in 7 . 5% (w/v) gels (Figure 19) . Clearly when the samples were run separately it was impossible to make accurate quantitative comparisons due to slight variations in the gel-slicing procedure. In order to overcome this, contrastingly labeled samples were mixed and co-electrophoresed . Figure 20 shows the result of one such experiment using 10% (w/v) polyacrylamide gels. In this experiment, the line K-1 labeled with 70 .5% (w/w) polyacrylamide gels. TOTAL D.RM. 71 Figure 20. Profile of radioactivity after polyacrylamide gel electro- phoresis of the plasma membrane polypeptides from CHO line K-1 labeled with 5 yCi [^^C]-leucine ( ) mixed with membrane polypeptides from « H-7 (a), M-7 (b) and 24-2 (c) labeled with 10 pCi [^H]-leucine ( ) Membranes were prepared by the method of Barland and Schroeder. Electrophoresis was carried out in 10% (w/v) polyacrylamide gels. 72 C provided the basis for comparisons. From this experiment, it appears that while M-7 and 24-2 show great similarity to the K-1 strain •the H-7 line shows some quantitative differences from the others. These results are not entirely unexpected, since M-7 and 24-2 were immediate derivatives of. K-1 while H-7 originated from a different laboratory. Clearly although the same polypeptide species were present in H-7 as the other lines, these were not present in exactly similar proportions. of dB-cAMP on the Plasma Membrane Polypeptide Composition In these experiments comparisons were made between the treated cells and their untreated counterparts by growing them on contrastingly labeled L-leucine, isolating plasma membranes from each, mixing the ' preparations and subjecting the mixture to polyacrylamide gel electro- phoresis. The results of a typical series of experiments are shown in Figure 21. In these, the treated cells had been grown on ImM dB-cAMP in the presence of 10 yCi of [^H]-L-leucine for 2 days, while the normal cells had been provided with 5 yCi of [^^C]-L-leucine. This growth period constituted approximately 3 to 4 cell genera- tions and was probably sufficient to allow all of the membrane proteins to reach similar specific radioactivities. Two different polyacrylamide preparations (7.5/ and 10%) were employed in order to resolve more completely, the different size classes of polypeptide. It is clear from Figure 21 and Figure 22 that no major qualitative or quantitative differences could be observed between the controls and the treated cells. Similar components were present in almost identical proportions. Nevertheless, SDS-electrophoresis separates proteins b'y size and we 73 O FRACTION N0M3ER Figure 21. Profile of radioactivity after polyacrylamide gel e ectrophoresis of the plasma membrane polypeptides of the CHO lines grown in the presence ( ) or absence ( ) of 1 mM dB-cA^. Lines H-7 (a) and 24-2 (c) were labeled with 10 yCi I HJ leucine in the absence and with 5 yCi [I'+Cj-leucine in the presence of dB-cAMP. Lines M-7 (b) and K-1 (d) were labeled with b yCi [ C] leucine in the absence and with 10 yCi [%]- leicine in the presence of dB-cAMP. Membranes were prepared by the method of Barland and Schroeder Electrophoresis was carried out- in 10% (w/v) polyacrylamide TOTAL D.P.M. 74 Figure 22. Profile of radioactivity after polyacrylamide gel electro- _ phoresis of the plasma membrane polypeptides of the CHO lines grown in the presence ( ) or absence ( ) of ImM dB-cAMP. Lines H— 7 (a) and 24-2 (c) were labeled with 10 pCi [^H]-leucine in the absence and with 5 pCi [ ]-leucine in the presence of dB-cAI-IP. Lines M-7 (b) and K-1 (d) were labeled with 5 pCi [14c]_peucine in the absence and with 10 pCi [^H]-leucine in the presence of dB-cAMP. Membranes were prepared by the method of Bar land and Schroeder. Electrophoresis was carried out in 7.5% (w/v) polyacrylamide gals. 75 cannot rule out the possibility that certain proteins have not become modified by phosphorylation or other types of substitution reactions during the course of treatment with dB-cAMP. Lactoperoxidase mediated iodination of the cell surface of cells grown in the presence or absence of dB-cAMP Because the loss in agglutinability of the CHO cells induced by dB- cAMP, which will be discussed later, might be due to a conformational rearrangement of the plasma membrane which restricts the movement of recep 125 tor molecules, I used the lactoperoxidase 1-labeling method to tag proteins exposed at the cell surface after growth in presence or absence of the nucleotide. Plasma membranes w’ere prepared by the method of Bar land and Schroeder. The radioactive polypeptides of the plasma mem- brane were then separated by electrophoresis in SDS-polyacrylamide gels. Figure 23 shows the results achieved using a 7.5% (w/v) polyacrylamide stacking gel and a 10% (w/v) polyacrylamide running gel (7.5/10% w/v gel)’. In neither of the cell lines which respond to dB-cAMP, i.e. H-7 (Figure 24) or 24-2 (Figure 25) was there any detectable change in the pattern of radioactivity along the gels. Moreover the profiles of incorporated radioactivity of the two were indistinguishable. However, since much of the radioactivity was concentrated at the top of the stacking gel, at the interphase between the stacking and running gels and at the dye front, we used a number of other electro- phoretic separation conditions. Figures 24-27 show the results of these studies for lines H-7, 24-2, M_-7 and K-1 respectively. Probably the most complete separation was achieved using a 7.5% (w/v) running gel in conjunction with a 5% (w/v) stacking gel. For each line studied a very high molecular weight component was again located at the very top of Figure 23. lodinatable surface polypeptides of different CHO cell lines grown in the presence or absence of ImM dB-cAMP. Cells were labeled using the lactoperoxidase technique, membranes prepared by the method of Barland and Schroeder, and polypeptides analyzed on SDS-polyacrylamide gels consisting of a 7.5% (w/v) stacking gels and a 10% (w/v) running gel. The gels wore sliced and the radioactive content of each slice was determined. 77 Figure 24. lodinatable surface polypeptides of CHO-H-7 grow in the presence or absence of dB-cAMP. Cells were labeled using the lacto- 125 peroxidase I technique. Membranes were prepared by the method of Barland and Schroeder and analyzed on three gel systems: Top - a gel consisting of a 7.5% (w/v) stacking gel and a 10% (w/v) running gel. Middle - a gel consisting of a 5% (w/v) stacking gel and a 7.5% (w/v) running gel. In these graphs molecular weight markers are indicated by arrows above the graph. Bottom - a gel consisting of a 3.75% (w/v) stacking gel and a 7.5% running gel. OTAf. 79 2 a: u I i- * M. o FRACTION NUMBER Figure 25. lodinatable surface polypeptides of CHO-K-1-24-2 grown in the presence or absence of ImM dB-cAMP, Cells were labeled using the 125 lactoperoxidase 1 technique. Membranes were prepared by the method of Barland and Schroeder and analyzed on three gel systems: Top - a gel consisting of a 7.5% (w/v) stacking gel and a 10% (w/v) running gel. Middle - a gel consisting of a 5% (w/v) stacking gel and a 7.5% (w/v) running gel. In these graphs molecular weight markers are indicated by arrows above the graph. Bottom - a gel consisting of a 3.75% (w/v) stacking gel and a 7.5% running gel. •»/o of TOTAL CP.M. 81 Figure 26. lodinatable surface polypeptides of CllO-K-l-M-7 grown in the presence or absence of dB-cAllP. Cells were labeled using the lactoperoxi- 125 dase I technique. Membranes were prepared by the method of Barland and Schroeder. and analyzed on three gel systems: lop - a gel consisting of a 7.5% (w/v) stacking gel and a 10% (w/v) running gel. f yiiddle - a gel consisting of a 5% (w/v) stacking gel and a 7.5% (w/v) running gel. ' In these graphs molecular weight markers are indi- cated by arrows above the graph. Bottom - a gel consisting of a 3.75% (w/v) stacking gel and a 7.5% running gel. d^e'A’HN NOIlDVijd 09 Ot^ 03 I 09 Ob' 03 L es of TOTAL C.RM. Figure 27. lodinatable surface polypeptides of CHO-K-1 grown in the presence or absence of dB-cAMP. Cells were labeled using the lacto- 125 peroxidase I technique. Membranes were prepared by the method of Barland and Schroeder and analyzed on three gel systems: Top - a gel consisting of a 7.5% (w/v) stacking gel and a 10% (w/v) running gel. Middle - a gel consisting of a 5% (w/v) stacking gel and a 7.5% (w/v) minning gel. In these graphs molecular weight markers are indicated by arrows above the graph. Bottom - a gel consisting of a 3.75% (w/v) stacking gel and a 7.5% running gel. dBSlAinN NOilDVyd. V 8 SL O 8 2v- O P 9 51 58 "/o ct TOTAL C.P.M. 86 the stacking gel. A protein of molecular weight greater than 130,000 was found close to the interphase region. A third band (molecular « weight approximately 110,000) was located one-third the way down the gel. Most of the remaining radioactivity collected at the dye front and was in proteins of molecular weight less than 25,000. By using lower concentrations of polyacrylamide in the stacking gel it was possible for the very high molecular weight material to pass through the stacking gel (Figures 24-27) and to band at the top of the running gel. Examination of these results indicate a) that all of the CHO lines have similar size classes of iodinatable proteins at their surface, b) that as in other cell species (120,129) only a limited number of proteins in the plasma membrane are accessible to the labeling agent, and are sensitive to trypsin (Figure 28), c) that the lines H-7, M-7 and 24-2 show no changes in lodination pattern following growth on dB-cAMP even though H-7 and 24-2 show complete loss of agglutinability under these conditions. We conclude therefore, that there has not been a major rearrangement or masking of surface proteins in response to the dB- cAMP-induced loss of agglutination, d) only line K-1 shows any change in lodination pattern (Figure 26) . Here additional bands of molecular weight greater than 130,000 and 72,000 are detectable after growth on dB-cAMP . These changes have noti been investigated further since this line does not lose its agglutinability in response to the nucleotide. The significance of the observation is not understood, e) there was no evidence in the treated H-7 line, which both loses its agglutinability and which assumes a fibroblast-like morphology, for the appearance of a new, iodinatable protein of high molecujar weight greater than 200, COO. 87 FRACTION NUMBER Figure 28. The effect of trypsin treatment on the iodinatable surface polypeptides of CHO-K-l-M-7 . The cells were incubated for 1 min with 0.25% trypsin. Subsequently the surface polypeptides of the cells were labeled using the lactoperoxidase technique. Membranes were prepared by the method of Barland and Schroeder, and analyzed on SDS-polyacrylamide gels consisting of a 5% (w/v) stacking gel and a 7.5% (w/v) running gel. 88 Such a protein has been reported as a persistent component of the plasma membranes of fibroblast-like cells, but is apparently absent in their transformed counterparts (120). Other methods studied as membrane selective probes » 1 studied the galactose oxidase and pyridoxal phosphate mediated labeling of the surface components of the plasma membrane. Both these methods finally depend on a reduction by [ H]-sodium borohydride, for the actual introduction of label into the plasma membrane components. We shall show that these methods did not have the specificity required for this study, presumably due to non-specific labeling of protein and 3 other compounds by [ H]-sodium borohydride. 3 For example, if cells are treated with [ H]-sodium borohydride alone (ImCi, 15 min), their plasma membranes isolated, and polypeptides ana- lyzed by electrophoresis on SDS polyacrylamide gels, a complex pattern of radioactivity was found along the gels (Figure 29) indicating that com- ponents of both high and low molecular weight became labeled. If the cells were first treated with unlabeled borohydride in order to react with readily reducible compounds, the pattern of radioactivity was some- what changed. Most of the label in the higher molecular weight compo- nents was lost. A very similar pattern was obtained after CHO cells were incubated with unlabeled sodium borohydride for 15 min' and then treated with 10 units of galactose oxidase for 90 min before reduction 3 with ImCi [ H]-sodium borohydride (5 min). In addition, it can be seen from Table 7 that specific radioactivity of the membranes only Increased by a factor of about 2 after the cells had been pretreated with galactose ... 3 oxxdase prior to the reduction with [ H]-sodium borohydride. Figure 29 and 30 show that the pattern of labeling on SDS-polyacrylamide gels is Figure 29. The non-specific labeling of CHO-H-7 surface polypeptides 3 by [ H]-sodium borohydride. The cells were disrupted in 10% (w/v) trichloroacetic acid, and label incorporated in pellet was analyzed on SDS-polyacrylamide gels consisting of a 5% (w/v) stacking gel and a 7.5% (w/v) running gel. The gels were sliced and the radioactive content of each slice was determined. 3 Top - cells were incubated for 15 min with 1 mCi [ H] -sodium boro- hydride, and disrupted in 10% trichloroacetic acid. Middle - cells were incubated for 15 min with 2 x 10 sodium boro- hydride, after a 90 min incubation with PBS to mimick incubation with galactose oxidase, the cells were incubated for 15 min 3 with ImCi [ H] -sodium borohydride and membrane proteins analyzed Bottom - cells were incubated for 15 min with 2 x 10~^M sodium boro- hydride, 90 min with 10 units galactose oxidase and finally 3 with ImCi [ H] -sodium borohydride and membrane proteins analyzed rOTAL CP.M. 90 4^. Figure 30. Galactose oxidase mediated labeling of CHO-H-7 surface polypeptides. After specified pretreatments the surface polypeptides were labeled. Membranes were prepared by the method of Brunette and Till, and analyzed on SDS-polyacrylamide gels consisting of a 7.5% (w/v) stacking gel and 10% (w/v) running gel. The gels were sliced and the radioactive content of each slice was determined. Top - cells were incubated with 2 x 10 sodium borohydride for 15 min, but not incubated with galactose oxidase. Middle - incubation with sodium borohydride was followed by a 90 min incubation with galactose oxidase. Bottom - incubation with sodium borohydride was followed by a 2 min incubation with 0.25% trypsin and a subsequent 90 min incubation with galactose oxidase. 8 6 4 2 8 6 4 O 8 6 4 2 O 92 93 TABLE 7 . Specific activity of H-7 membranes prepared by the Brunette 3nd Till method after labeling with the galactose— oxidase— sodium-borotitride technique Treatment Before Incubation with ImCi [^H]-sodium borohydride cpm/yg Protein Cells were incubated with 2 x 10~^ M sodiumboro- hydride for 15 min Cells were incubated with 2 x 10~^ M sodiumboro- hydride for 15 min and subsequently for 90 min with 10 units of galactose oxidase, after they were cultured with: - no addition to the growth medium - ImM dB-cAMP added to the growth medium for 48 h - O.lmM Sq 20009 added to the growth medium for 48 h Cells were incubated with 2 x 10“'^ M sodium boro- hydride for 15 min, then for 2 min with 0.25% (w/v) trypsin, and a subsequent incubation with 10 units of galactose-oxidase for 90 min 116 229 283 289 136 94 not very different either after incubation with galactose oxidase and analysis on two different polyacrylamide gel concentrations. Figure 30, furthermore, shows that only a fraction of labeled molecules can be removed from the cell surface by incubation with trypsin. Therefore, we felt that the galactose oxidase technique in which 3 H is introduced into dialdose groups of glycoproteins following oxida- tion with galactose oxidase (p. 33 of Materials and Methods), or the technique involving reduction of Schiff's bases to pyridoxal phosphate 3 (Figure 6a), both of which require the use of [ H]-sodium borohydride, 3 cannot be successful because other groups, readily reducible by [ H]- sodium borohydride are already present in the membrane. Neither of the techniques was used in the subsequent study. The Effects of dB-cAMP and Sq 20009 on the Agglutinability of the CHO Lines The ability of concanavalin A and WGA to agglutinate the different CHO cell lines was tested on cells that had been grown in the presence _3 and absence of ImM 3':5'-cyclic AMP, ImM dB-cAMP, 1 x 10 butyric acid -4 and 10 M Sq 20009 (Table 8). In absence of any additives, all of the lines, including the elongated 24-2 were completely agglutinable by both lectins tested at concentrations of 25 and 125 yg/ml. The presence of dB-cAMP, however caused a marked loss in agglutinability of the H-7 and 24-2 lines to Con A, but had no measurable effect on K-1 and M-7. This was surprising since K-1 is the line most responsive to dB-cAMP in terms of morphological change; M-7 of course does not elongate under these conditions. Only H-7 showed a .marked decreased agglutinability towards VJGA and then only at the lowest concentration of lectin. 3':5'- cyclic A?'5P also caused a los.s in concanavalin A-induced agglutinability cf !i.-7 , but liaci lit lie influence, on the other lines. Even in relation 95 TABLE 8. The effect of dB-cAMP, 3':5'-cyclic AMP butyric acid and Sq 20009 on the , agglutination of the CHO cell lines by concanavalin A and WGA TREATMENT* % AGGLUTINATION BY WGA CON A (yg/ml) (yg/ml) 25 125 25 125 H-7 No additions 100 100 100 100 1 mM dB-cAMP 0 90 0 0 1 mM c-AMP — — 25 65 1 mM Butyric acid 85 100 90 100 0.1 mM Sq 20009 M-7 No additions 100 100 100 100 1 mM dB-cAMP 90 100 100 100 1 mM cAMP — — 100 100 0.1 mM Sq 20009 10 100 10 100 24-2 No additions 100 100 90 100 1 mM dB-cAMP 100 100 30 60 1 mM cAMP — — 75 100 0.1 mM Sq 20009 15 100 20 20 K-1 No additions 90 100 100 100 1 mM dB-cAMP 90 100 100. 100 1 mM cAMP — — 100 100 0.1 mM Sq 20009 60 — 80 — * Description of treatment of cells after they reached the desired density. Unless otherwise indicated the cells were grown without additions to the growth medium except those indicated in the section on cell maintenance. 96 to H-7,’ it was less potent than dB-cAMP. This may be due to the greater stability of the butyryl analog in the cell's growth medium. Sq 20009 induced some loss in both WGA and concanavalin A agglutinability of the 24-2 and M-7 lines but had no measurable influence on K-1 or H-7. Butyric acid, which was included as a control to test the effect of possible decomposition products of dB-cAMP, did not alter the agglutina- bilit}' of CHO-H-7 and can be eliminated as a non-specific cause of changes observed in the presence of dB-cAMP. The changes in agglutinability observed are not due to an increased number *of binding sites upon treatment with the drugs, as is shown in Table 9. In fact, the 24-2 line has twice as many binding sites upon treatment with dB-cAMP, whereas it is less agglutinable. It is clear from the results that there is no satisfactory correlation between the morphological responsiveness of the different cells to various drugs and their ability to undergo drug-induced losses in agglutinability. We conclude, therefore, that the two phenomena are separable. Effect of colchicine, low temperature and enzyme treatment on agglutinability of CHO cells Colchicine is known to disrupt intracellular microtubules and has a direct effect on cellular processes in which microtubules are involved. CHO cells (H-7) were therefore gro^ra in presence of 5 x 10 colchicine for 5 h before being tested for agglutinability by low con- centrations of concanavalin A and WCA. This treatment caused a marked loss in agglutinability when compared with controls (Table 10). The change was of about the same magnitude as that induced by dB-cAMP alone. If colchicine is added to cells growing in presence of dB-cAMP, the cells lose their elongated shape, suggesting that microtubules are 97 Table 9. The effect of dB— cAMP and Sq 20009 on the binding of [^H]-concanavalin A by the CHO cell lines. Cell Line Drug Added to Medium cpm/10^ Cells H-7 No 67227 10~^ M dB-cAMP 51915 K-1 No 61228 10"^ M dB-cAMP 51513 K-l-M-7 No 79875 • 10“'^ M Sq 20009 53460 K-1-24-2 No 37443 10"^ M dB-cAMP 81897 lO”'^ M Sq 20009 46055 * Average of 5 determinations. 98 TABLE 10. The effect of temperature, enzyme treatment and colchicine on lec tin— induced agglutination of CHO H— 7 by concanavalin A and WGA. % Agglutination by 25 yg/ml JL Treatment" Con A WGA Untreated cells 90 90 Cells incubated at 0° for 15 min before assay , 10 10 Cells incubated at 0° for 15 min and then returned to 37° for 15 min before assay 90 85 Cells grown on 10 M dB-cAMP 20 10 _3 Cells grown on 10 M dB-cAMP and treated with 10 yg/ml chemotrypsin for 5 min at 22°C before assay 100 100 _3 Cells grown on 10 M dB-cAMP and treated with 5 x 10~^ M colchicine for 4 h before assay 20 0 Cells grown on normal medium and treated with 5 x 10~^ M colchicine 4 h before assay 30 20 Cells grown on normal medium and treated with 10. yg/ml neuraminidase for 15 min before assay 90 90 99 involved in the morphological response. The agglutinability of these cells is very low. The influence of low temperature on the agglutinability of CHO cells was similar to that observed on other transformed mammalian cell lines (132). A reversible decrease in both concanavalin A and WGA- induced agglutinability was observed if the cells were cooled to 0°C for 15 min before testing (Table 10). Brief exposure of the cells to low concentrations of chymotrypsin reversed the loss in agglutinability induced by dB-cAMP. Similar results have been observed on untransformed fibroblasts which are normally not agglutinable but which will respond to the lectins after brief trypsinization (132). The presence of sialic acid on certain glycopeptides is known to enhance the binding of the natural ligand N-acetylgalactosamine to WGA (149). In addition, a decrease in sialic acid on certain glycopeptides has been observed after the H-7 line were exposed to 3': 5' -cyclic AMP (46,47). However, removal of sialic acid groups from the surface of CHO-cells by means of neuraminidase did not mimic the effects of dB-cAMP and the cells remained completely agglutinable. Time course for changes in agglutination and effects of inhibitors of protein and RNA synthesis The agglutinability of cells was determined at fixed time intervals after dB-cAMP was added to or removed from the growth medium. From the con- trol values in Table 11 it is clear that the response of the cells to the drug was extremely rapid. Within 4 h.of adding dB-cAMP, line H-7 com- pletely lost its agglutinability towards both concanavalin A and WGA when ti'.e lectins were tested at concentrations of 50 and 125 yg/ml. In the 100 TABLE 11. The temporal effects of addition and removal of dB-cAMP on the agglutination of CHO-H-7 by concanavalin A and WGA % AGGLUTINATION BY WGA CON A TREATMENT (yg/^i) 30 125 50 125 No additions 95 100. 100 100 1 mM dB-cAMP was added: 4 hrs before assay 10 10 10 10 2 hrs before assay 20 80 70 80 1 hr before assay 25 90 70 90 Cells were grown on 1 mM dB-cAMP for 48 hrs. The medium was changed to normal medium: 4 hrs before assay 90 95 95 100 2 hrs before assay 80 95 100 100 1 hr before assay 50 90 75 100 101 reverse-type of experiment, when the cells were transferred to normal medium after being maintained on dB-cAMP, complete agglutinability was restored within 1 to 2 h. To see if protein or RNA synthesis were required for the changes in agglutinability to occur we added a number of metabolic inhibitors to the cells at intervals before assaying for agglutination (Table 12) . Cyclohexamide , actinomycin D and cordycepin all inhibited the loss in agglutinability normally caused by dB-cAMP if they were introduced to the cells simultaneously with the nucleotide. It seems likely that both RNA and protein synthesis are required in order for dB-cAMP to induce the non-agglutinable state. Cells were also observed to become fully agglutiriable after they had been grown continuously on dB-cAMP but cyclohexamide, actinomycin D or cordycepin added 4 h before assay. It would appear that even after the non-agglutinable state had been achieved, protein synthesis had to continue for the condition to be maintained. We conclude that trans- • • criptive events are required not only to induce the non-agglutinable state but probably also to maintain it. In the final set of experiments reported in Table 12 we tested the ability of inhibitors to counteract the return of the cells to the fully agglutinable state after they had been removed from medium con- taining dB-cAMP. This reversion is normally very rapid (see control values in Table 11) and it was not prevented by cyclohexamide, actino- mycin D or cordycepin. Similar experiments to the ones described above have been carried out to determine the effects of cycloheximide and actinomycin D on the loss oi agglutinability induced by dB-cAMP or SQ 20009 on line 24-2. TABLE 12. The effects of inhibitors of protein and RNA synthesis on dB-cAMP induced loss of agglutination of CHO-H-7 by concanavalin A and WGA 102 c c ^ o •H r— ! c- E > 00 C B P o H c m U •H • cd c •H <3 (1) 4-» o P :s •H rH fH B B OO Cu 00 X 00 < 0) p o CM ' — ✓ < r-H ' — ^ u c o CJ u o < 4-» o iH •H s o U H p: iH o o o o o Os o CTi m o m o o^ o^ o m m o^ o o o iH 0> o o o m ctn o> o o o m 00 ctn mom r'- o c^^ I P >> P rH no iH d 4-) CO p CO P fH • • In »H P p P P P u cd P • • P p P o P o E o p P p CO E CO CO CO p JO CO p jp u u CO p CO CO •H CO CO CO p CO p o •H CO CO CO •U cd CO p Cd p •H CO p •H 00 p H p P P c • • a AJ 4 4 4-» •H QJ no 0) 1 0) O 0) P •d p P o JO p P P c O’ P DO H u M o u 0) H o u P u u u H no o TO nd o c o a o d O o o •H o o o cd y-i no U-i 14-i d iH tw d 4-1 4h 4-1 0) Cd 0) . ^ JO 0) 0) 0) P p P p p CO p p P Hi H H Ph JO P. 00 P JO JO J2 H •H o o p rH o rP O :s H _r; H u < o »P u <3 p D. JO JO JO iJ 00 u OO a JO •rH •H >'» iH •H mT u E fH CM mT 4’> fH to CP JO CO CP •H •H CC/ .“•■N .<■ — V y-N, «H '•d *H /•~N rH d p •H y~S y— S y— S .J-. oi JO P H o fH P rH p d P JO u c o 200,000), iodinatable, trypsin-sensitive protein which has been con- sidered to be almost invariable characteristic of non-transformed fibro- blasts, but which is absent in transformed ceils (120). Again tills 128 suggests that dB-cAMP is not eliminating all features of the transformed phenotype in these CHO cells. Pastan and his collaborators have recently suggested that high levels of intracellular 3' :5 '-cyclic AMP will render a cell minimally agglutinable with Con A (150). However, it will be noted that high levels of dB-cAMP may be added to two of the lines investigated in this work, K-1 and M-7 , without modifying the agglutinability of the cell line with concanavalin A. Of special Interest in this case is the line K-1 which elongates dramatically but does not lose its agglutinability upon addition of dB-cAMP to its growth medium. Thus these lines also demon- strate that there is no necessary correlation between the morphology of a cell line and its relative agglutinability with concanavalin A or WGA. This conclusion is also reinforced by line 24-2 which is permanently 4 elongated yet fully agglutinable. Similarly H-7 which does not respond • morphologically to 3':5'-cyclic AMP does lose its agglutinability in its presence. Further, others have shown that the morphological changes induced in CHO cells (48,50,51) do not require nuclear events, whereas we have demonstrated that the loss of agglutinability requires both pro- tein and RNA synthesis. Clearly these two features normally characteristic of the transformed phenotype, a compact epithelial-like shape and sensi- tivity to agglutination by plant lectins, are separable phenomena. Modification in agglutinability of these CHO cell lines by concana- valin A following the addition or removal of dB-cAMP. to the growth medium was: (1) not dependent on a change in the number of lectin receptors, (2) not dependent on a modification of the fatty acid composition of the plasma membrane, and (3) not dependent on a change in the total amount « of surface sialic acid, since normal cells remained fui ly agglutinable 129 following treatment with neuraminidase. However, a correlation does exist between loss in agglutinability and loss of sialic acid from certain cell surface glycoproteins, as will be discussed later. Maintenance of the non-agglutinable state by cells grown in the presence of dB-cAMP, or the ability of cells to attain this state following addition of dB-cAMP was: (1) dependent on protein and RNA synthesis and (2) dependent on a trypsin-labile and presumably surface membrane protein. On the other hand, the change from the non-agglutinable state to the agglutinable state following removal of dB-cAMP or Sq 20009 was not dependent on protein or RNA synthesis. Since the change to the non-agglutinable state and its reversal happen quickly, it must be assumed that very rapid changes, presumably in the mobility of the lectin receptors occur when cells are confronted with altered levels of 3':5'-cyclic AMP. Moreover, it has to be assumed that the components that are responsible for maintaining the non- agglutinable state turn over very rapidly. As stated earlier, however, we have not been able to demonstrate major differences in the pattern of plasma membrane proteins between the agglutinable and non-agglutinable states or between the lines which dif- fer in their responsiveness to dB-cAMP. Neither have we been able to detect a protein which turns over more rapidly than other components^ in the membrane of treated cells (J. van Veen, R. M. Roberts and K. D. Noonan, unpublished results). Therefore, it has not been possible to decide what precise components are involved in maintaining the non- agglutinable state in these treated CHO cells. It must be assumed that a new structural protein, possibly, a very minor one, or an induced enzyme is involved. There is some evidence that a cellular micro- 130 architecture comprised of microtubules or microfilaments (or possibly both) has a role in restricting the lateral movement of proteins in the plasma membrane since drugs which disrupt these structures also affect such processes as agglutinability of transformed cells and cap- ping in lymphocytes (92-98). It is possible that the new protein is a "linkage" protein between the membrane polypeptides and such restraining structures in the cytoplasm. Alternatively, the induction of the non- agglutinable state may require such events as phosphorylation or other modifications of existing proteins, processes which in turn may require, the induction of specific enzymes with short half-lives. One surface modification which seems to occur concomitantly with the change in agglutinability is the decrease in sialic acid associated with the f ucose-containing glycopeptides that can be released from the cell surface by incubation with trypsin. It seems likely that this associa- tion of two modifications in surface characteristics is not a chance event. On the other hand, it is most unlikely that these sialic acid- containing glycopeptides are the lectin receptors for two reasons: (1) % cells treated with neuraminidase and which have lost their sialic acid from these glycopeptides remain fully agglutinable by concanavalin A, (2) the sialic acid glycopeptides themselves do not bind efficiently to concanavalin A immobilized on Sepharose (R. M. Roberts, J. Van Veen, unpublished results). Similar differences in f ucose-containing glyco- peptides have been shown to distinguish transformed fibroblasts from their normal counterparts (9-12). I believe that these results are also consistent with the hypothesis that the protein components of the dB-Ci\MP treated cells are less mobile, i.e. less able to migrate laterally than those of controls. I propose that in the untreated cell, incomplete 131 glycopeptides and appropriate glycosyl transferases are able to make ready contact so that terminal sialic acid groups are added to complete the carbohydrate chain. If protein movement is restricted, however, this event may occur less readily so that a population of incomplete macromolecules results. Glycosyl transferases capable of utilizing exogenous sugar nucleo- tides including CMP-N-acetyl neuraminic acid are present at the surface of cultured mammalian cells. Their presence there may be the result of movement of intracellular membrane to the surface via an assembly line process in which case their location may have no functional significance. On the other hand, others have proposed that they continue to function ■ (122-124). However, it is generally assumed that highly charged, labile molecules such as sugar nucleotides are unlikely to persist outside the environment of the cell. Nevertheless, it cannot be ruled out that the glycosyl donor in these reactions is not a glycolipid or that reaction I does not involve transglycosylation from other glycoproteins. Our results certainly indicate that sialic acid can be ‘added to existing surface glycopeptides in the absence of dB-cAMP but not in its presence. The results with colchicine are also consistent with the idea that the degree of mobility of membrane proteins affects glycosylation reactions. This drug disrupts microtubules and causes a single cap-like patch of lectin, and presumably other proteins, receptors to form in treated cells (96). Presumably all of the restraints on mobility are relieved. As expected completion of the glycopeptides occurs under these circumstances even if dB-cAilP is present. On the other hand, agglutination is inhibited presumably because the opportunity for multiple cross-links to form between cells is reduced. 132 In conclusion, therefore, we believe that in these experiments dB-cAMP induces the formation of some unstable component (s) which has a role in controlling the lateral mobility of lectin receptors and probably other proteins in the plasma membrane of susceptible cells. We believe that the altered composition of fucose-containing glycopeptides and changed susceptibility of the cells to lectin-induced agglutination are related phenomena and a direct result of these 3':5'-cyclic AMP mediated events. Moreover, since these changes also occur following cell transformation, identical processes may be involved. , BIBLIOGRAPHY 1. Kao, F. T. and Puck, T. T. (1968) Genetics of somatic mammalian cells, VII. Induction and isolation of nutritional mutants in Chinese hamster cells. Proc. Natl. Acad. Sci. 60:1275-1281. 2. Hsie, A. W. and Puck, T. T. (1971) Morphological transformation of Chinese hamster cells by dibutyryl adenosine cyclic 3':5'-mono- phosphate and testosterone. Proc. Natl. Acad. Sci. 68:358-361. 3. Hsie, A. W. , Jones, C. and Puck, T. T. (1971) Further changes in differentiation state accompanying the conversion of Chinese hamster cells to fibroblastic form by dibutyryl adenosine cyclic 3' :5'- monophosphate and hormones. Proc. Natl. Acad. Sci. 68:1648-1652. 4. Hershko, A., Mamont, P. , Shield, R. and Tomkins, G. M. (1971) Pleiotypic response. Nature New Biol. 232:206-211. 5. Bader, J. P. and Brown, N. R. (1971) Induction of mutations in an RNA tumor virus by an analog of a DN.A precursor. Nature New Biol. 234:11-12. 6. Aaron, S. A. and Todaro, G. J. (1968) Basis for the acquisition of malignant potential by mouse cells cultivated in .vitro. Science 162:1024-1026. 7. Black, P. H. (1968) The oncogenic DNA viruses: a review of in vitro transformation studies. Ann. Rev. Microbiol. 22:391-426. 8. Zarling, J. M. and Tevethia, S. S. (1971) Expression of concanavalin A binding sites in rabbit kidney cells infected with vaccinia virus. Virology 45:313-316. 9. Buck, C. A., Click, M. C. and Warren, L. (1970) Comparative study of glycoproteins from the surface of control and R S virus trans- formed hamster cells. Biochem. 9:4567-4576. 10. W’arren, L. , Fuhrer, J. P. and Buck, C. A. (1972) Surface glycopro- teins of normal and transformed cells: a difference determined by sialic acid and a growth-dependent sialyl transferase. Proc. Natl. Acad. Sci. 69:1838-1842. 11. Warren, L. , Fuhrer, J. P. and Buck, C. A. (1973) Surface glyco- proteins of cells before and after transformation by oncogenic viruses. Fed. Proc. 32:80-85. 133 134 12. Buck, C. A., Fuhrer, J. P. , Soslau, G. and Warren, L. *(1974) Membrane glycopeptides from subcellular fractions of control and virus-transformed cells. J. Biol. Chem. 249:1541-1550. 13. Pasternak, T. , Sutherland, E, W. and Henion, W. F. (1962) Deriva- tives of cyclic 3' : 5 ' -adenosine monophosphate. Biochim. Biophys. Acta 65:558-560. 14. Robinson, G. A., Butcher, R. W. and Sutherland, E. W. (1968) Cyclic AMP. Ann. Rev. Biochem. 37:149-174. 15. Menaham, L. A., Hepp, K. D. and Wiesland, 0. (1969) Liver 3':5'- nucleotide phosphodiesterase and its activity in rat liver per- fused with insulin. Eur. J. Biochem. 8:435-443. 16. Heersche, J. N. M. , Fedak, S. A. and Aurback, G. D. (1971) The mode of action of dibutyryl adenosine 3 ': 5 ' -monophosphate on bone tissue in vitro. J. Biol. Chem. 246:6770-6775. 17. Kaukel, E. and Hilz, H. (1972) Permeation of dibutyryl cAMP into HeLa cells and its conversion to monobutyryl cAMP. Biochem. Biophys. Res. Commun. 46:1011-1018. 18. Kaukel, E. , Fuhrmann, U. and Hilz, H. (1972) Divergent action of cAMP and dibutyryl cAMP on macromolecular synthesis in HeLa S3 cultures. Biochem. Biophys. Res. Commun. 48:1516-1524. 19. d’Armiento, M. , Johnson, G. S. and Pastan, I. (1972) Regulation of adenosine 3':5'-cyclic monophosphate phosphodiesterase activity in fibroblasts by intracellular concentrations of cyclic adenosine monophosphate. Proc. Natl. Acad. Sci. 69:459-462. 20. Hsie, A. W. , Kawashima, K. , O'Neill, P. J. and Schroder, C. H. (1975) Possible role of adenosine cyclic 3' :5' monophosphate phos- phodiesterase in the morphological transformation of CHO cells mediated by N^,o2-dibutyryl adenosine cyclic 3 ': 5 ' -monophosphate. J. Biol. Chem. 250:984-989. 21. Ryan, W. L. and Kurick, M. A. (1972) Adenosine 3 ': 5 ' -monophosphate and N^-2 ' -0-dibutyryl adenosine 3 ': 5 ' -monophosphate transport in cells. Science 177:1002-1003. 22. Papahadj opoulos , D. , Poste, G. and Mayhew, E. (1974) Cellular uptake of cyclic AMP captured within phospholipid vesicles and effect on cell growth behavior. Biochim. Biophys. Acta 363:404-418. V 23. De Asua, L. , Surian, S. , Flawia, M. and Torres, N. (1973) Effect of insulin on the growth pattern and adenyl cyclase activity of BHK fibroblasts. Proc. Natl. Acad. Sci. 70:1388-1392. 24. Reddi, P. K. and Constantinides , S. M. (1972) Partial suppression of tumor production by dibutyryl cyclic AMP and theophylline. Nature 238:286-287, 135 25. Burger, M. M. , Bombik, B. M. , Breckenridge , B. M. and Sheppard, J. R. (1972) Growth control and cyclic alterations of cAMP in the cell cycle. Nature New Biol. 239:161-165. 26. Lieberman, I. (1959) Growth factors for mammalian cells in culture. J. Biol. Chem. 234:2754-2758. 27. Illiano, G. and Guatrecasas, P. (1972) Modulation of adenylate cyclase activities in liver and fat cell membranes by insulin. Science 175:906-908. 28. O'Neill, P. J. , Shchroder, C. H. and Hsie, A. W. (1975) Hydrolysis of butyryl derivatives of adenosine cyclic 3':5' monophosphate by CHO cell extract and characterization of the products. J. Biol. Chem. 250:990-995. 29. Seifert, W. and Paul, D. (1972) Levels of cyclic AMP in sparse and dense cultures of growing and quiescent 3T3 cells. Nature New Biol. 240:281-283. 30. Otten, J. , Johnson, G. S. and Pastan, I. (1971) Cyclic AMP levels in fibroblasts: relationship to growth rate and contact inhibition of growth. Biochem. Biophys. Res. Commun. 44:1192-1198. 31. Otten, J. , Johnson, G. S. and Pastan, I. (1972) Regulation of cell growth by cyclic adenosine 3' :5'-monophosphate. J. Biol. Chem. 247:7082-7087. 32. Johnson, G. S. and Pastan, I. (1972) Role of 3 ': 5 ' -adenosine mono- phosphate in regulation of morphology and growth of transformed and normal fibroblasts. J. Natl. Cane. Inst. 48:1377-1387. 33. Teel, R. W. and Hall, R. G. (1973) Effect of dibutyryl cyclic AMP on the restoration of contact inhibition in tumor cells and its relationship to cell density and the cell cycle. Expt. Cell Res. 76:390-394. 34. Sheppard, J. R. (1971) Restoration of contact-inhibited growth to transformed cells by dibutyryl adenosine 3':5'-cyclic monophosphate. Proc. Natl. Acad. Sci. 68:1316-1320. 35. Dey, J. , Vogel, A. and Pollack, R. (1974) Intracellular cyclic AMP concentration responds specifically to growth regulation by serum. Proc. Natl. Acad. Sci. 71:694-698. 36. Sheppard, J. R. and Prescott, D. M. (1972) Cyclic AMP levels in synchronized mammalian cells. Expt. Cell Res. 75:293-296. . 37. Maganiello, V. and Vaughan, M. (1972) Prostaglandin E effects on adenosine 3':5'-cyclic monophosphate concentration and phosphodies- terase activities in fibroblasts. Proc. Natl. Acad. Sci. 69:269-273. 38. Strohl, W. A. (1969) The response of BHK21 cells to infection with type 12 adenovirus. II. Relationship of virus-stimulated DNA synthe- sis to other viral functions. Virology 39:653-665. 136 39. Johnson, G. S. , Morgan, W. D. and Pastan, I. (1972) Regulation of cell motility by cyclic AMP. Nature 235:54-56. 40. Johnson, G. S. and Pastan, I. (1972) Cyclic AMP increases the adhe- sion of fibroblasts to substratum. Nature New Biol. 236:247-249. 41. Weber, M. J. (1973) Hexose transport in normal and in Rous Sarcoma virus-transformed cells. J. Biol. Chem. 248:2978-2983. 42. Venlita, S. and Rubin, H. (1973) Sugar transport in normal and Rous Sarcoma virus-transformed chick-embryo fibroblasts. Proc. Natl. Acad. Sci. 70:653-657. 43. Roller, B. , Hirai, K. and Defend!, V. (1974) Effect of cAMP on nucleoside metabolism. I. Effect of thymidine transport and incor- poration in monkey cells (CV-1). J. Cell PHys. 83:163-176. 44. Hauschka, P. V. , Everhart, L. P. and Rubin, R. W. (1972) Alteration of nucleoside transport of Chinese hamster cells by dibutyryl cyclic monophosphate. Proc. Natl. Acad. Sci. 69:3542-3546. 45. Coggins, J. F. , Johnson, G. S. and Pastan, I. (1972) The effect of dibutyryl cyclic adenosine monophosphate on synthesis of sulfated acid mucopolysaccharides by transformed fibroblasts. J. Biol. Chem. 247:5759-5764. 46. Roberts, R. M. , Citorelli, J. J. and Walker, A. (1973) Fucose con- taining glycopeptides from the cell surface of Chinese hamster ovary cells grown in the presence or absence of cyclic AMP. Nat. New Biol. 244:86-89. 47. Roberts, R. M. and Baig, M. M. (1973) Comparative studies on the carbohydrate containing membrane components of normal and cAMP treated CHO-cells. Biochemical J. 134:329-339. 48. Schroder, C. H. and Hsie, A. W. (1973) Morphological transforma- tion of enucleated Chinese hamster ovary cells by dibutyryl adenosine 3' :5 '-monophosphate. Nature New Biol. 246:58-60. 49. Carter, S. B. (1967) Effects of cytochalasins on mammalian cells. Nature 213:261-264. 50. Shay, J. W. , Porter, K. R. and Prescott, D. M. (1974) The surface morphology and five structure of CHO-cells following enucleation. Proc. Natl. Acad. Sci. 71:3059-3063. 51. Patterson, D. and Waldren, C. A. (1973) The effect of inhibitors of RNA and protein synthesis on dibutyryl cyclic AMP mediated morpliolo- gical transformations of CHO cells in order. Biochem. Biophys. Res. Commun. 50:566-573. 52. Rein, A. , Carchman, R. A. , Johnson, G. S. and Pastan, I. (1973) Simian virus 40 rapidly lowers cAMP levels in mouse cells. Biochem. Biopliys. Res. Commun. 52:899-904. 137 53. Sheppard, G. R. (1973) Difference in cAMP levels in normal and transformed cells. Nature New Biol. 236:14-16. 54. Anderson, W. B. , Lovelace, E. and Pastan, I, (1973) Adenylate cyclase activity is decreased in chick embryo fibroblasts trans- formed by wild type and temperature sensitive Schmidt-Ruppin Rous Sarcoma virus. Biochem. Biophys. Res. Commun. -52:1293-1299. 55. Anderson, W. B. , Johnson, G. S. and Pastan, I. (1973) Transformation of chick-embryo fibroblasts by wild-type and temperature sensitive Rous Sarcoma virus alters adenylate cyclase activity. Proc. Natl. Acad. Sci. 70:1055-1059. 56. Burstin, S. J. , Renger, H. C. and Basilico, C. (1974) Cyclic AMP levels in temperature sensitive SV40 transformed cell lines. J. Cell Phys. 84:69-74. 57. Zimmerman, J. E. and Raska, K, (1972) Inhibition of adenovirus type 12 induced DNA synthesis in G1 arrested BHK 21 cells by dibutyryl adenosine cyclic 3 ': 5 ' -monophosphate. Nature New Biol. 239:145-147. 58. Criss, W. E. (1974) Second messenger system in neoplasia. Oncology 30:43-80, 59. Cuatrecasas, P. (1969) Interaction of insulin with the cell membrane the primary action of insulin. Proc. Natl. Acad. Sci. 63:450-457. 60. Cuatrecasas, P. (1973) Insulin receptors of liver and fat cell membranes. Fed. Proc. 32:1838-1846. 61. Stoeckenius, W. and Engelman, D. M. (1969) Current models for the structure of biol membranes. J. Cell Biol. 42:613-646; 62. Winzler, R. J. (1969) Carbohydrates in cell surfaces. Int. Rev. of Cytol, 29:77-125. 63. Hendler, R. W. (1971) Biological membrane ultra structure. Physiological Rev. 51:66-96. 64. Wallach, D. F. H. (1972) The plasma membrane: dynamic perspectives genetics and pathology. The English Universities Press Ltd., London. 65. Steck, T. L. (1974) The organization of proteins in the human red blood cell membrane. J. Cell Biol. 62:1-19. 66. Singer, S. T. (1974) The molecular organization of membranes. Ann. Rev. of Biochem. 43:805-833. 67. Berlin, R. D. , Oliver, J. N. , Ukena, T. E. and Yin, H. il. (1975) The cell surface; New England J. of Med. 292:515-520. 138 68. Singer, S. J. and Nicolson, G. L, (1972) The fluid mosaic model of the structure of cell membranes. Science 175:720-730. 69. Steck, T. L. , Fairbanks, G. and Wallach, D. F. H. (1971) Disposi- tion of major proteins in the isolated erythrocyte membrane: proteolytic dissection. Biochem. 10:2617-2624. 70. Wallach, D. F. H. (1972) The dispositions of proteins in the plasma membranes of animal cells: analytical approaches using controlled peptidolysis and protein labels. Biochim. Biophys. Acta 265:61-83. 71. Frye, L. D. and Edidin, M. (1970) The rapid" intermixing of cell surface antigens after formation of mouse-human heterokaryons . J. Cell Science 7:319-335. 72. Edidin, M. and Weiss, A. (1972) Antigen can fornmtion in cultured fibroblasts: a reflection of membrane fluidity and cell motility. Proc. Natl. Acad. Sci. 69:2456-2459. ' 73. Edelman, G. M. (1973) Receptor mobility and receptor-cytoplasmic interactions in lymphocytes. Proc. Natl. Acad. Sci. 70:1442-1446. 74. Albrecht-Buhler-, G. and Solomon, F. (1974) Properties of particle movement in the plasma membrane of 3T3 mouse fibroblasts. Exp. Cell Res. 85:225-233. 75. Nicolson, G. L. and Yanadimachi, P. (1974) Mobility and restriction of mobility of plasma membrane lectin-binding components. Science 184:1294-1296. 76. Nicolson, G. L. (1971) Difference in topology of normal and tumor cell membranes shown by different surface distributions of ferretin conjugated Con A. Nature New Biol. 233:244-246. 77. de Petris, S. and Raff, M. C. (1973) Ligand-induced redistribution of concanavalin A receptors on normal, trypsinized and transformed fibroblasts. Nature New Biol. 244:275-278. 78. Rosenblith, J. L. , Ukena, T. E. , Yin, H. H. , Berlin, R. D. and Kamovsky, M. J. (1973) A comparative evaluation of the distribution of concanavalin A-binding sites on the surfaces of normal, virally-transformed , and protease- treated fibroblasts. Proc. Natl. Acad. Sci. 70:1625-1629. 79. Inbar, M. , Ben-Bassat, H. , Fibach, E. and Sachs, L. (1973) Mobility of carbohydrate containing structures on the surface membrane and the normal differentiation of myeloid leukemic cells to macrophages and granulocytes. Proc. Natl. Acad. Sci. 70:2577-2581. 80. Kornberg, R. D. and McConnell, H. M. (1971) Inside-outside transi- tions of phospholipids in vesicle membranes. Biochem. 10:1111-1120. 139 81. Hiramo, H. , Parkhouse, B. , Nicolson, G. L. , Lennox, E. S. and Singer, S. J. (1972) Distribution of saccharides residues on mem- brane fragments from myeloma cell homogenates. Proc. Natl. Acad. Sci. 69:2945-2949. 82. Steck, T. L. and Dawson, G. (1974) Topographical distribution of complex carbohydrates in the erythrocyte membrane. J. Biol. Ghem. 249:2135-2142. 83. Byers, B. and Porter, K. R. (1964) Oriented microtubules in the elongation of cells of the developing lens rudiment after induction Proc. Natl. Acad. Sci. 52:1091-1099. 84. Weber, K. , Pollack, R. and Bibring, T. (1975) Antibody against tubulin: the specific visualization of cytoplasmic microtubules in tissue culture cells. Proc. Natl. Acad. Sci. 72:459-463. 85. Asthund, R. E. and Pastan, J. (1974) Myosin in cultured fibroblasts J. Biol. Ghem. 249:3903-3907. 86. Lazarides, E. and Weber,. K. (1974) Actin Antibody: the specific visualization of actin filaments in non-muscle cells. Proc. Natl. Acad. Sci. 71:2268-2272. 87. Weber, K. and Groeschel-Stewart , U. (1974) Antibody to myosin: the specific visualization of myosin-containing filaments in nonmuscle cells. Proc. Natl. Acad. Sci. 71:4561-4564. 88. Piras, M. M. and Piras, R. (1974) Phosphorylation of vinblastin isolated microtubules from chick-embryonic muscles. Eur. J. Bioch. 47:443-452. 89. Allen, R. D, (1975) Evidence for linkages between microtubules and membrane bound vesicles. J. Gell. Biol. 64:497-503. 90. Ehrlich, H. P. , Ross, R. and Bornstein, P. (1974) Effects of anti- microtubular agents on the secretion of collagen. A biochemical and morphological study. J. Cell. Biol. 62:390-405. 91. Berlin, R. D. and Uhena, T. E. (1972) Effect of colchicine and vin- blastine on the agglutination of polymorphonuclear leucocytes by concanavalin A. Nature New Biol. 238:120-122. 92. Yin, H. H. , Ukena, T. E. and Berlin, R. D. (1972) Effect of col- chicine, colcemid and vinblastine on the agglutination by concana- valin A of transformed cells. Science 178:867-868. 93. de Petris, S. (1974) Inhibition and reversal of capina by cytoch. B. and vinblastine and colchiciiie . Nature 250:54-55. 94. Ryan, G. B. , Borysenko, J. Z. and Karnovsky, M. J. (1974) Factors affecting the redistribution of surface-bound concanavalin A on human pol>Tnorphonuc lear leukocytes. J. Cel].. Bioi. 62:251-365. lAO 95. Nakamura, J. and Terayama, H. (1975) Colchicine affects kinetics of concanavalin A-mediated agglutination of hepatoma cells and plasma membranes from liver and hepatoma cells. Proc. Natl. Acad. Sci. 72:498-502. 96. Ukena, T. E. , Borysenko, J. Z. , Karnovsky, M. J. and Berlin, R. D. (1974) Effects of colchicine, cytochalasin B, and 2-deoxyglucose on the topographical organization of surface-bound concanavalin A in normal and transformed fibroblasts. J. Cell Biol. 61:70-82. 97. Oliver, J. M. , Ukena, T. E. and Berlin, R. D. (1974) Effects of phagocytosis and colchicine on the distribution of lectin-binding sites on cell surfaces. Proc. Natl. Acad. Sci. 71:394-398. 98. Wickus , G. , Gruenstein, E. , Robbins, P. W. and Rich, A. (1975) Decrease in membrane-associated actin of fibroblasts after trans- formation by Rous Sarcoma Virus. Proc. Natl. Acad. Sci. 72:746- 749. 99. Rambourg, A. (1971) Morphological and histochemlcal aspects of glycoproteins at the surface of animal cells. Int. Rev. of Cytol. 31:57. 100. Kraemer, P. M. (1971) Heparan sulfates of cultured cells. 1. Membrane-associated and cell-sap species in Chinese hamster cells. Biochem. 10:1437-1445. 101. Nicolson, G. L. and Singer, S. J. (1974) The distribution and asymmetry of mammalian cell surface saccharides utilizing ferritin conjugated plant agglutinins as specific saccharide stains. J. Cell. Biol. 60:236-248. 102. Martinez-Palomo , A., Braislovsky, C. and Bernhard, W. (1969) Ultra- structural modifications at the cell surface and intracellular contacts of some transformed cell strains. Cancer'Res. 29:925-937. 103. Torpier, G. and Montaguier, L. (1970) Ultrastructural changes of the surface of BHK 21/13 transformed cells, depending on adenine nucleotides. Int. J. Cancer 6:529-535. 104. Malucci, L. , Poste, G. H. and Wells, V. (1972) Synthesis of cell coat in normal and transformed cells. Nature New Biol. 235:222- 223. 105. Satoh, C. , Duff, R. , Rapp, F. and Davidson, E. A. (1973) Production of mucopolysaccharides by normal and transformed cells. Proc. Natl. Acad. Sci. 70:54-56. 106. Chiarugi, V. P, , Vannucchi, S.-and Urbano, P. (1974) Exposure of trypsin-removable sulphated polyanions on the surface of normal and cirally transformed BHK 21/C13 cells. Biochim. Biophys. Acta 345:283-293. 141 107. Hakomori, S. and Murakami, W. T. (1968) Glycolipids of hamster fibroblasts and derived malignant-transf ormed cell lines. Proc. Natl. Acad. Sci. 59:254-261. 108. Brady, R. 0. , Borek, C. and Bradley, R. M. (1969) Composition and synthesis of gangliosides in rat hepatocyte and hepatoma cell lines. J. Biol. Chem. 244:6552-6554. 109. Brady, R. 0., Fishman, P. H. and Mora, P. T. (1973) Membrane components and enzymes in virally transformed cells. Fed. Proc. 32:102. 110. Brady, R. 0. and Fishman, P. H. (1974) Biosynthesis of glucolipids in virus-transformed cells. Biochim. Biophys. Acta 355:121-148. 111. Hakomori, S. (1970) Cell density-dependent changes of glycolipid concentrations in fibroblasts, and loss of this response in virus transformed cells. Proc. Natl. Acad. Sci. 67:1741-1747. 112. Mora, P. T. , Brady, R. 0., Bradley, R. M. and McFarland, V. W. (1969) Cangliosides in DNA virus-transformed and spontaneously transformed tumorigenic mouse cell lines. Proc. Natl. Acad. Sci. 63:1290-1296. 113. Cumar , F. A. (1970) Enzymatic block in the synthesis of ganglio- sides in DNA virus-transformed tumorigenic mouse cell lines. Proc. Natl. Acad. Sci. 67:757-764. 114. Mora, P. T. , Cumar, F. A. and Brady, R. 0. (1971) A common biochemi cal change in SV40 and polyoma virus transformed mouse cells coupled to control of cell growth in culture. Virology 46:60-72. 115. Gahmberg, C. C. and Hakomori, S. (1974) Organization of glycolipids and glycoproteins in surface membranes: dependency on cell cycle and on transformation. Biochem. Biophys. Res. Commun. 59:283-291. 116. Cahmberg, C. G. and Hakomori, S. (1974) Surface carbohydrates of hamster fibroblasts. J. Biol. Chem. 250:2438-2446. 117. Gahmberg, C. G. and Hakomori, S. (1973) Altered growth behavior • of malignant cells associated with changes in externally labeled glycoproteins and glycolipids. Proc. Natl. Acad. Sci. 70:3329-3333. 118. Wisnieksi, B. J. , Williams,. R. E. and Fox, C. F. (1973) Manipula- tion of F. A. composition in animal cells grown in culture. Proc. Natl. Acad. Sci. 70:3669-3673. 119. Sakiyama, H. , Gross, S. K. and Robbins, P. W. (1972) Glycolipid synthesis in normal and virus-transformed hamster cell lines. Proc. Natl. Acad. Sci. 69:872-876. 120. Hynes, R. 0. and Humphrys , K. C. (1974) Characterization of the external proteins of hamster fibroblasts. J. Cell Biol. 62:438- 448. 142 121. Hynes, R. 0. and Maepherson, J. (1974) ^n "Membrane Transformation in Neoplasia" eds. J. Schultz and R. E. Block, Miami White Symposium, Academic Press. 122. Grimes, W. J. (1970) Sialic acid transferases and sialic acid Ivels in normal and transformed cells. Biochem. 9:5083-5092. 123. Den, H. , Schultz, A. M. , Basu, M. and Roseman, S. ’(1971) Glyco- syltransferase activities in normal .and polyoma-transformed BHK cells. J. Biol. Chem. 246:2721-2723. 124. Patt, L. M. and Grimes, W. J. (1974) Gell Surface glycolipid and glycoprotein glycosyltransf erases of normal and transformed cells. J. Biol. Chem. 249:4159-4165. 125. Saito, M. , Satoh, H. and Ukita, T. (1974) Sialyl transferase activities of rat ascites hepatoma cells and rat liver. Biochim. Biophys. Acta 362:549-557. 126. Roth, S. and \7hite, D, (1972) Intercellular contact and cell- surface galactosyl transferase activity. Proc. Natl. Acad. Sci. 69:485-489. 127. W^bb, G. C. and Roth, S. (1974) Cell contact dependence of surface galactosyltransf erase activity as a function of the cell cycle. J. Cell Biol. 63:796-805. 128. Deppert, W. , Werchau, H. and Walter, G. (1974) Differentiation between intracellular and cell surface glycosyl transferases: Gal transferase activity in intact cells and in cell homogenate. Proc. Natl. Acad. Sci. 71:3068-3072. 129. Truding, R. , Shelanski, M. L. , Daniels, M. P. and Morell, P. (1974) Comparison of surface membranes isolated from cultured murine neuroblastoma cells in the differentiated or undifferentiated state. J. Biol. Chem. 249:3973-3982. 130. Sharon, N. and Lis, H. (1972) Lectins: cell-agglutinating and sugar-specific proteins. Science 177:949-959. 131. Inbar, M. and Sachs, L. (1969) Interaction of the carbohydrate- binding protein concanavalin A with normal and transformed cells. Proc. Natl. Acad. Sci. 63:1418-1425. 132. Noonan, K. D. and Burger, M. M. (1973) The relationship of concana- , ' valin A binding to lectin-initiated cell agglutination. J. Cell Biol. 59:134-142. 133. Noonan, K, D. and Burger, M. Ml (1973) Binding of con A to normal and transformed cells. J. Biol. Chem. 248:4286-4292. 134. Bretscher, M. S. (1973) On labeling membranes. Nature New Biol. 245:114. 143 135. Rifkin, D. B. , Compans, R. W. and Reich, E. (1972) A specific labeling procedure for proteins on the outer surface of membranes. J. Biol. Chem. 247:6432-6437. 136. Gahmberg, C. G. and Hakomori, S. (1973) External labeling of cell surface galactose and galactosamine in glycolipid and glycoprotein of human erythrocytes. J. Biol. Chem. 248:4311-4317. 137. Phillips, D. and Morrison, M. (1970) The arrangement of proteins in the human erythrocyte membrane. Biochem. Biophys. Res. Commun. 40:284-289. 138. Nagata, Y. and Burger, M. M. (1970) \7heat germ agglutinin isolation and chrystalization. J. Biol. Chem. 247:2248-2250. 139. Agrawal, B. I. and Goldstein, I. J. (1965) Specific binding of concanavalin A to cross linked dextran gels. Biochem. J. 96:23c- 25c. 140. Leakly, U. K. (1970) Cleavage of structural proteins during the assembly of the head of the bacteriophage T4. Nature 227:680-685. 141. Davis, B. J. (1964) Disc electrophoresis II method and application to human serum proteins. Ann. N. Y. Acad. Sci. 121:404-427. 142. Turner, J. C. (1968) Triton X-100 scintillant for carbon-14 labeled materials. Int. J. of applied Rad. and Isot. 19:557-563. 143. Barland, P. and Schroeder, E. A. (1970) A new rapid method for the isolation of surface membranes from tissue culture cells. J. Cell Biol. 45:662-668. 144. Morrison, R. T. and Boyd, R. N. , ^n "Organic Chemistry", 2nd edition. Section 6:6 and 28:18. Allyn and Bacon, Inc., Boston. 145. Brunette, D. M. and Till, J. E. (1971) A rapid method for the isolation of L-cell surface membranes using an aqueous two-phase polymer system. J. Membr. Biol. 5:215-224. 146. Lowry, 0. H. , Rosebrough, N. J. , Farr, A. L. and Randall, R. J. (1951) Protein measurement with Folin phenol reagent. J. Biol. Chem. 193:265-275. 147. Ross, E. and Schatz, G. (1973) Assay of proteins in the presence of high concentrations of sulfhydryl compounds. Anal. Biochem. 54:304-306. 148. Gilman, A. G. (1970) A protein binding assay for adenosine 3':5'- cyclic monophosphate. Proc. Natl. Acad. Sci. 67:305-312. 149. Burger, M. M. and Goldberg, A. R. (1967) Identification of a tumor- specific determinant on neoplastic cell surfaces. Proc. Natl. Acad. Sci. 57:359-366. 144 150. Willingham', M. C. and Pastan, I. (1974) C-AIIP mediates the Con A agglutinability of mouse fibroblasts. J. Cell Biol. 63:288-294. APPENDIX Program used for the calculation of percent of total DPM in 3 14 samples labeled with H and C. This program was written for a Wang 600 series calculator with a 3 model 602 plotting output writer. This program will correct the H 14 3 counts for the overlap of C counts into the H channel, calculate and print DPM in each sample and will print and plot percent of total DPM for each sample. Program: STEP CODE KEY(S) STEP CODE KEY(S) 0000 09 00 MARK 0026 03 06 -6 0001 00 00 EO 0027 00 15 CLEAR DISP 0002 00 14 CLEAR ALL 0028 09 03 STOP 0003 09 03 STOP 0029 06 06 ST 6 0004 06 00 ST 0 0030 07 00 RE 0 0005 09 03 STOP 0031 03 06 -6 0006 06 01 ST 1 0032 06 06 ST 6 0007 09 00 MARK 0033 05 07 t7 0008 00 01 El 0034 06 06 ST 6 0009 00 15 CLEAR DISP 0035 07 07 RE 7 0010 09 03 STOP 0036 03 07 -7 0011 06 05 ST 5 0037 00 15 CLEAR DISP 0012 07 00 RE 0 0038 09 03 STOP 0013 03 05 -5 0039 06 07 ST 7 0014 06 05 ST 5 0040 07 01 RE 1 0015 00 15 CLEAR DISP 0041 03 07 -7 0016 09 03 STOP 0042 06 07 ST 7 0017 06 06 ST 6 0043 00 15 CLEAR DISP 0018 07 00 RE 0 0044 09 03 STOP 0019 03 06 -6 0045 06 08 ST 8 0020 06 06 ST 6 ' 0046 07 01 RE 1 0021 06 07 ST 7 0047 03 08 -8 0022 07 05 RE 5 0048 06 08 ST 8 0023 05 06 ib 0049 07 07 RE 7 0024 06 05 ST 5 0050 05 08 v8 0025 07 06 RE 6 14 > STEP 0051 0052 0053 0054 0055 0056 0057 0058 0059 0060 0061 0062 0063 0064 0065 0066 0067 0068 0069 0070 0071 0072 0073 0074 0075 0076 0077 0078 0079 0080 0081 0082 0083 0084 0085 0086 0087 0088 0089 0090 0091 0092 0093 0094 0095 0096 0097 0098 0099 0100 146 ■ CODE KEY(S) STEP CODE KEY(S) 06 07 ST 7 0101 07 09 RE 9 00 15 CLEAR DISP 0102 03 08 -8 09 03 STOP 0103 06 08 ST 8 06 02 ST 2 0104 . 07 07 RE 7 06 03 ST 3 0105 05 08 ^8 00 01 E 1 0106 08 05 Jif+ 00 06 E 6 0107 00 15 CLEAR DISP 06 04 . ST 4 0108 00 00 EO 09 00 ■ MARK 0109 08 02 PRINT 00 03 E 3 0110 10 02 F 2 00 15 CLEAR DISP 0111 15 11 Dll 09 03 STOP 0112 06 04 ST 4 06 08 ST 8 0113 06 15 ST 15 07 00 RE 0 0114 06 14 ST 14 03 08 08 0115 07 12 RE 12 06 08 ST 8 0116 06 13 ST 13 06 09 ST 9 0117 07 14 RE 14 07 05 RE 5 0118 03 . 12 -12 05 08 t8 0119 08 05 Jif+ 08 05 Jif+ 0120 07 14 RE 14 00 15 CLEAR DISP 0121 06 13 ST 13 00 00 EO 0122 07 13 RE 13 08 02 PRINT 0123 06 12 ST 12 05 02 v2 0124 07 15 RE 15 15 11 INDIR 0125 02 11 +11 06 04 ST 4 0126 00 01 El 06 15 ST 15 0127 02 04 +4 06 14 ST 14 0128 00 01 El 07 12 RE 12. 0129 03 02 -2 06 13 ST 13 0130 08 04 JifO 07 14 RE 14 0131 08 00 SEARCH 03 12 -12 0132 00 03 E3 08 05 J lf+ 0133 09 00 MARK 07 14 RE 14 0134 00 04 E4 06 13 ST 13 0135 07 10 RE 10 07 13 RE 13 0136 08 02 PRINT 06 12 ST 12 0137 06 02 ST 2 07 15 RE 15 0138 06 10 ST 10 02 10 +10 0139 00 01 El 00 01 E 1 0140 00 00 EO 02 04 + 4 0141 00 00 EO 07 06 RE 6 0142 05 10 tIO 05 09 v9 0143 06 10 ST 10 06 09 ST 9 0144 07 11 RE 11 00 15 E 15 0145 08 02 PRINT 09 03 STOP 0146 07 02 RE 2 06 08 ST 8 0147 06 11 ST 11 07 01 RE 1 0148 00 01 E 1 03 08 -8 0149 00 00 E 0 06 08 ST 8 0150 00 00 E 0 147 STEP CODE KEY(S) STEP CODE KEY(S) 0151 05 11 t11 0201 04 00 xO 0152 06 11 ST 11 0202 00 01 El 0153 00 01 El 0203 00 12 El 2 0154 00 05 E6 0204 06 01 STl 0155 06 04 St 4 0205 09 02 a 0156 09 00 MARK 0206 07 08 RE8 0157 00 05 E5 0207 02 02 +2 0158 07 11 . RE 11 0208 07 00 REO 0159 06 00 . ST 0 0209 03 00 -0 0160 06 11 ST 11 0210 07 01 . REl 0161 07 00 RE 0 0211 03 01 ‘ -1 0162 05 12 -12 0212 00 01 El 0163 06 12 ST 12 0213 02 04 +4 0164 06 13 ST 13 0214 07 11 REll 0165 00 04 E 4 0215 15 11 Dll 0166 00 00 EO 0216 05 04 -4 0167 00 00 EO 0217 06 06 ST6 0168 06 00 ST 0 0218 08 02 PRINT 0169 07 12 RE12 0219 10 02 +2 0170 05 .. 00 -0 0220 07 12 RE12 0171 06 12 ST12 0221 04 06 x6 0172 09 00 MARK 0222 07 06 RE6 0173 00 06 E6 0223 06 00 STO 0174 00 00 EO 0224 00 00 EO 0175 06 00 STO 0225 06 01 STl 0176 00 00 EO 0226 09 02 a 0177 06 01 STl 0227 05 > 09 -9 0178 09 02 a 0228 02 02 +2 0179 05 06 -0 0229 00 01 El 0180 02 02 +2 0230 00 12 E12 0181 09 00 MARK 0231 04 00 xO 0182 00 07 E7 0232 00 01 El 0183 07 10 RE 10 0233 00 12 E12 0184 15 11 INDIR 0234 06 01 STl 0185 05 04 -4 0235 09 02 a 0186 06 05 ST5 0236 07 03 RE8 0187 08 02 PRINT 0237 02 02 +2 0188 05 02 -2 0238 07 00 REO 0189 07 12 RE12 0239 03 00 -0 0190 04 15 x5 0240 07 01 REl 0191 07 15 RE5 0241 03 01 -1 0192 06 00 STO 0242 07 05 RE5 0193 00 01 El 0243 03 05 “5 0194 00 00 EO 0244 07 06 PvE6 0195 06 01 STl 0245 03 06 -6 0196 09 02 a ■ 0246 00 01 El 0197 05 06 v6 0247 02 04 +4 0198 02 02 +2 0248 00 01 El 0199 00 01 El 0249 03 03 -3 0200 00 12 E12 0250 08 04 JifO 148 STEP , CODE KEY(S) 0251 08 00 SEARCH 0252 00 07 E7 0253 09 00 MARK 0254 00 08 E8 0255 00 00 EO 0256 06 00 STO 0257 00 09 E9 0258 06 01 STl 0259 09 02 a 0260 05 06 t6 0261 02 02 +2 0262 00 01 El 0263 08 02 PRINT 0264 05 00 -0 0265 00 00 EO 0266 08 02 PRINT 0267 10 00 FO 0268 00 14 CLEAR ALL 0269 09 14 END PROG Once the program is in the memory of the calculator the following steps are required for the calculator to perform the calculations properly; PRIME SEARCH 0 14 C blank . go blank go DPM IN STAND. 14 3 C overlap into H channel go DPM IN STAND, go CPM go total number of samples go go CPM for STAND, go ENTER ^'^C CPM for sample go ENTER CPM for same- sample as ^'^C entered above. BIOGRAPHICAL SKETCH Jacques van Veen was born on December 19, 1949, on Aruba, in the Netherlands Antilles, v;here he attended elementary and secondary schools. He was awarded a scholarship by the government of the Netherlands Antilles to study biochemistry and medicine in 1969, and enrolled at the University of South Carolina. , In 1970 he transferred to the University of Florida, where in June, 1972, he received a Bachelor of Science degree in Chemistry. He received an award from the American Chemical Society's division of analytical chemistry. He entered graduate school in the Department of Biochemistry at the University of Florida in September, 1972. He will begin his studies in medicine at the Erasmus Universiteit in Rotterdam, the Netherlands, in September, 1975. The author is married to the former Brenda M. Every, who will receive the degree of Master of Arts in Economics from the University of Florida in August, 1975. j49 I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Robert M. Roberts, Chairman Associate Professor of Biochemistry I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Cenneth D . Noonan 'Assistant Profe/ssor Biochemistry I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Eug^e G. Sander As3p4:iate Professor of Biochemistry I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Br}ian Gebhardt Associate Professor of Pathology This dissertation was submitted to the Graduate Faculty of the Department of Biochemistry in the College of Arts and Sciences and to the Graduate Council, and was accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy. August, 1975 Dean, Graduate School