m5? ’ h'“’fi JEN" ’NHJ CI'TH ":71fi.’11'7""} 3‘ qfilél [Peri .. .‘ ‘1.’ ‘ ,I$-I “I 1‘~.7 'l"\v:n 5' , ’-'~7‘:I.‘{F“"~ ! ,1 \IJI .‘Ili” _ , 1," '1’}! 0‘ ' I 7. "7‘1"" “$79.77;. -._-__._ Hum "‘f 1r. , _‘ . J! : I ' . I: 5" 3 1, . 5511,4311 7 3' 1:133: {‘ I '1, 4 ‘3:1;§3’:;CG:':1.‘ Y 1.2. ' ”1"." -’V" D-z.e-:I7FVIUJ {1, ' p -1“ :‘Ifi‘inll‘ ' {A l . .. 1.17.117 ., 7 a - 1151177“: ', 7 ’ 42312111101 1 1 . ‘ ' “7‘17: ’1 '1‘ ‘ . i I: I. ‘ 71,. .1“? I L1 f ‘1 1 "1" 3| i“ ”1,1131!“ 1}" 11"?!“2‘1'2737” 7- 7 I 1 , 11‘9”; C'A'u 7 ; JV " 1r: 711711 3.1 I 1’3 £51175! . y) -7. ~ 11.. ‘ 1151:1111: f} - . . 7.7. .7 11:11:11 1 - - |-.. :1 3.731,.” b}? . ‘, .. ' - .1 171 ' 11:111.. ,9 . -7- _- _ f. . . - , VH1" “£31 .1 7.1 w , 4“:- _' lrfifg‘bnip 41;; “£..._:.' ':' 1‘17}: “3' 7"; " _ ‘, ,3: ,2; . ;’a ~11?‘.h§‘ 3: 1 1114’ .‘ 35.9“ 3. (5 . 3'! ,. 11,5.37‘2‘75'. I tr." .fi’é‘g ‘ ‘ ‘ C . ’(:f.“ I. I H‘ .I’n If” i}: ' ~43"! [v.5 3&1“. OC" .’-(‘§‘§l~5 " 2:4’“." 5 I_" .1 L -1 " ., " Li" $13, r3113 ' 7~~;‘ = “‘35111-. . . ‘7‘ " ' {.114 '41,; 5 '53]. 1'21 ‘ "745.4. 1:117 .171 7 ' 7 F-z’tgifga‘a ”('1‘ ,3}. (17?" ”.7 Jim . -7 1.7213,.1'771,’ 17171 ~11? £51111 1 £77 77 7; )9. A. ‘ .. §_ 1 1 “g, g7, 9" 111141 ‘ 1439-. vgfgyu fir’i’uififi ‘ 557‘7'777 ”its 0.; . h. ‘ fl - ' .Nfg“: 3:55.. .11731‘14'73 11 ‘ 2‘3""- f‘ c I - 5.” < 7:115; .25., 71:: "7’ 1.15;, N “L ‘ \ figv-Q‘V‘ {3" .. Y a ., ?2,.-'.,. , 3:973:26“ ' NAT: .1": '2‘ daw- . . A Il'u 5 If“ 1. A; A-‘v ,;_ 7, I . . .1. _... 511;: mp” 31533.1 E3 -'_.‘" ‘:. '7' finitil‘? ' 11177;: ' :‘:"F7‘.::f«11?§7".$3’.7§§411*7'1'1' ’17r3’7"r’a1fx‘:£ " \E" I. J . . :7] J ‘- .. }-_.,-¢ . "1 ‘71] " {v . 1; 1:33.. 146' 7 " .1 711731923717 ’“7‘x7 ‘ F‘J‘ 1".“31"? L517 {@3151 1.25 , , ,{L 711;” “£1931: {”7 1:51:21: I’ué*;t‘;lf;: 2:": "uisfi‘: “I: ;.':;?S]: fvifxli’jp’ufiéhlj 31(1)! 31:11,” (fifl ”IT???" {5 3151173131111;1:31:31: :1"; ~Mf’urfi 1:121 7“ 1. c If“? .1 . ,1 f3“. .4“ ‘ .I} _ :I 7711“ fig'!3;fu?:. 3"”. tank“ {VAL} 113: :53 Cd»; ffil‘w 411.134- 11:3'V21‘ff 1:7,: ?:f . Muff}? ((6.4%? Ié'il. 7“;""lit‘f‘-7?J‘ 3:3‘ 11:75“ 171.” 1 If '7” {‘fiid v _-_1 fix: tlflq‘tllzd‘l'”! “C" l €fi:h}:’:‘l.;"v01; ”‘3'?” ("5 C52. ‘1... .f' ‘C‘i.:l :w till”? 3,?“ '[ “th‘ci‘fifib I “(L - . ,‘r1 :- (I: ‘;:'*L'--“ I‘LL’R} ’17! ML") '1: ’7": 9‘ éfihpug €11 , ‘1 '1 III-.1 7"!" 5177-111 :jfh' 17:17PM “11%) 7:397:51? E. {133‘ ICC-'7". 7%;101g15qu7r13117‘i 1.31. “W“ 11131175.“ 1119;171:341 . . 7 {k ' "Y 1 ‘3’. 1:31; C'l‘n 1;}; I‘: CC,"I “3"T 9:111" g , . ~ '1 7771‘ 7'71". "”11 11117177711 1““77121. =. If“ $1.1” C: 1.33:: [2%) 71 1171'”..1 ‘5‘“ :5 ”‘3” ring». ’1 gr! II Nut.“ 14'“ 1’3: _77' 1 ”Cum ‘oh‘m “[1111fo‘Wmgfl'dkvlrfiimfiiflw‘: 3 73-2: ' m: __ 4":‘2’2"? - ..’ ’3’9* (‘5 1/3} -L?‘ '- n“ ‘ -...‘ 1.1: ~.... “y:- ‘FI ‘:!1 L... w‘. :v 77‘ '\ .21. 17% if)? .3 :17." 33%“. :{7'11 63“.. ”3"? :5:_:‘: '.v‘~o¢?_ . L',’ V; W a 13%? :é~ 1‘4", LIeRARY # EAichigen State University This is to certify that the dissertation entitled Studies on the Consequences of Cell Shape For Membrane Dynamics: Relationship to Cytoskeletal Organization presented by Mark Harold Swaisgood has been accepted towards fulfillment of the requirements for Ph.D . degree in _.Bio_ehemi§_c_ry Wm Major professor Date February 9, 1987 MS U is an Affirmative Action/Equal Opportunity limitation 0-12771 Agm‘ .__—4* - MSU LIBRARIES __ % RETURNING MATERIALS: Place in book drop to remove this checkout from your record. FINES will be charged if book is returned after the date stamped below. STUDIES ON THE CONSEQUENCES OF CELL SHAPE FOR MEMBRANE DYNAMICS: RELATIONSHIP TO CYTOSKELETAL ORGANIZATION BY Mark Harold Swaisgood A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Biochemistry 1987 ABSTRACT STUDIES ON THE CONSEQUENCES OF CELL SHAPE FOR MEMBRANE DYNAMICS: RELATIONSHIP TO CYTOSKELETAL ORGANIZATION BY Mark Harold Swaisgood A large body of evidence suggests the alteration of cell shape influences the biological activity of anchorage dependent fibroblasts. The shape transition is also associated with changes in the structural organization of cytoskeletal elements. The importance of these structures to surface receptor organization, biosynthetic processes, and gene expression implies the involvement of cytoskeletal elements in an intracellular communication pathway that can be disrupted by cell shape alteration. On the basis of evidence suggesting the control of protein lateral mobility by cytoskeletal structures, protein lateral mobility was measured as a function of cell shape to determine the effect of shape induced cytoskeletal alteration. The role of protein lateral mobility in growth factor action and the association of surface receptors with the cytoskeleton indicate the significance of using protein lateral mobility to monitor cell shape induced alterations in the association of membrane proteins with the cytoskeleton. Transformation provides a means to perturb the cytoskeleton in addition to cell shape alteration. The major findings of this thesis include a) shape induced alteration of the cell cytoskeleton coincides with a partial release of restriction for lateral diffusion of wheat germ agglutinin (WGA) and succinyl concanavalin A receptors in untransformed BALB/c 3T3 cells; b) the observation of a bimodal population of fast and slow mobility cells after the alteration of cell shape in K-MSV transformed BALE/c 3T3 cells; c) the detection and isolation of a clone from the K-MSV BALE/c 3T3 fibroblasts exhibiting fast mobility; d) the loss of global modulation, a process affecting protein lateral mobility that requires an intact cytoskeleton, in spherical BALE/c 3T3 cells and in the "fast" mobility clone regardless of cell shape. In addition, a study done by others in the laboratory found correlative evidence that cell shape alteration affects nucleo-cytoplasmic transport in BALB/c 3T3 fibroblasts. These observations support the evidence for a role of the cytoskeleton in restriction of protein lateral mobility and support the possibility that the cytoskeleton serves as an integral element in an intracellular communication pathway. This thesis is dedicated to Harold and Janet Swaisgood for instilling in me a love of learning and being a constant source of support. 11 ACKNOWLEDGEMENTS I am grateful to Dr. John Wang and his laboratory who provided critical insight into my research. In his laboratory several friends and colleagues who deserve mention for their insightful comments include Ioannis Moustatsos, Marco Villanuevo, and Jim Brauker. I am also proud to share a special bond of joy and misery with fellow FRAPPERS: Tom Metcalf, Jerome Hochman, and Lian Wei Jiang. I am also grateful to Margaret Wade of Meridian Instruments for showing me the fine art of using the ACAS 470. Margaret Hogan, Doug DeGaetano, and Sally Meiners: friends and rabble-rousers, offered valued advice and aid. I appreciate their patient listening, friendship, and assistance. For her unfaltering faith in me, constant support of my goals, and ready willingness to become my tissue culture technician in times of need I lovingly thank my wife, Aileen. Finally, and most importantly, I gratefully acknowledge the guidance, expertise, support, and good times provided by my mentor, Dr. Melvin Schindler. iii TABLE OF CONTENTS List of Tables List of Figures . . . . . List of Abbreviations . . . . CHAPTER I: OVERVIEW OF THE THESIS CHAPTER II: LITERATURE REVIEW Signal Transmission . . . A Cytoskeletal Communication Pathway Cytoskeletal/Membrane Protein Interaction Cytoskeletal/Nucleus Interaction The Cytoskeleton Cytoskeletal Function Cytoskeletal Structure Microfilaments . Microtubules Intermediate Filaments Spectrin History of Protein Lateral Diffusion Fluid Mosaic Model . . . . . . Lateral Diffusion . . . . . . . The Technique of FRAP . . . . . . Diffusional Control by the Cytoskeleton Diffusional Control: Other Models Protein Density . . Lipid Domains Extracellular Matrix The System: Alteration of Cell Shape The cell line . Suspension of Cells in Methyl Cellulose iv Page viii xii Shape Induced Alteration of Metabolic Activities Shape Induced Disruption of the Cytoskeleton Cytoskeletal Disrupting Drugs Disruption of the Cytoskeleton by Transformation Summmary of Objectives . . . . . . . . . . . . References CHAPTER III PROTEIN LATERAL DIFFUSION IN SPREAD AND SPHERICAL FIBROBLASTS Introduction Materials and Methods Cell Culture FRAP . . . . . . Fluorescence Microscopy Cell Population Histograms WGA Binding . . Comparison of WGA Receptors Characterization of Glycolipid Receptors for WGA Characterization of Cell Growth . . . . . . Thymidine Uptake Glucose Utilization Determination of ATP Levels Scanning Electron Microscopy . Results Consequence of Shape and Transformation on the Number of WGA Receptors in 3T3 Fibroblasts FITC-WGA Binding . . . . . . Radio- iodinated WGA Binding . . Population Distribution of FITC-WGA per Cell Distribution of Fluorescence over the Cell Surface WGA Glycoprotein and Glycolipid Receptors Lateral Mobility in 3T3 Fibroblasts Lipid Mobility . . . . Cell Surface Morphology as a Factor . . . . WGA Receptor Mobility . Consequences of the Removal and Suspension Process Cell Growth as a Function of Shape 31 32 33 34 37 39 46 47 49 49 49 50 51 51 52 52 54 54 54 55 56 58 58 58 6O 6O 6O 64 68 68 68 7O 72 73 Thymidine Uptake Glucose Utilization ATP Levels Actin and Fibronectin Distribution as a Function of Cell Shape and Transformation Discussion References 0 O 0 O O O O I C O O O O O O O O O O 0 CHAPTER IV IDENTIFICATION AND ISOLATION OF A FAST MOBILITY POPULATION FROM TRANSFORMED 3T3 FIBROBLASTS Introduction Materials and Methods Cell Culture Lateral Mobility Measurements Fluorescence Microscopy Results Characterization of the E761 Clone Number and Type of WGA Receptors Lipid Lateral Mobility . . WGA Receptor Lateral Mobility Actin and Fibronectin Patterns Cell Growth . . Discussion References CHAPTER V GLOBAL MODULATION AND CELL SHAPE: EVIDENCE FOR CYTOSKELETAL CONTROL OF PROTEIN LATERAL MOBILITY Introduction Methods and Materials Cell Culture vi 73 76 77 79 82 87 90 91 93 93 93 93 94 94 94 94 100 102 102 106 110 111 112 114 114 Con A Platelets . Measurement of Lateral Mobility Binding of 8- Con A . . . . . . Results . . . . . . . . 8— Con A Receptor Lateral Mobility Consequences of Shape Alteration for Global Modulation . . . . . . Discussion References CHAPTER VI: SUMMARY Appendix A: FRAP THEORY Appendix B: PUBLICATIONS Glossary vii 114 114 115 116 116 118 120 125 127 131 145 146 Table LIST OF TABLES CHAPTER III Phospholipid and Protein Lateral Mobility Effect of Preparing 3T3 Cells for Suspension Glucose Utilization in 3T3 Fibroblasts CHAPTER IV Phospholipid and Protein Lateral Mobility CHAPTER V Protein Lateral Mobility viii Page 71 74 78 101 117 LIST OF FIGURES Figure Page CHAPTER III 1 The binding of FITC-WGA to untransformed 3T3 and K-MSV transformed 3T3 fibroblasts in the spread and spherical state was determined at various concentrations of FITC-WGA. . . . . . . . . . . . 59 2 The binding of radio-iodinated WGA to untrans— formed and K-MSV transformed 3T3 fibroblasts in the spread and spherical stated was determined at various concentrations of WGA. . . . . . . . . 61 3 Population analysis of the amount of fluorescence per cell which is related to the amount of FITC-WGA bound by the cell. . . . . . . . . . . . 62 4 Fluorescence micrographs of spread and spherical 3T3 fibroblasts of spread and spherical 3T3 fibroblasts labeled with FITC-WGA . . . . . . . . 63 5 Western transfer blot of WGA glycoprotein recep- tors from spread and spherical 3T3 fibroblasts. . 65 6 Autoradiogram of blycolipids separated by TLC and probed with radio- iodinated WGA . . . . . . . . 67 7 Scanning electron micrographs of spread and spherical 3T3 fibroblasts . . . . . . . . . . 69 8 Uptake of 3H-thymidine into 3T3 and K-MSV transformed 3T3 fibroblasts . . . . . . . . . . . 75 9 Fluorescence micrographs of spread and spherical 3T3 fibroblasts labeled with rhodamine-phalloidin, an F-actin specific probe, and spread 3T3 fibro- blasts incubated with anti-fibronectin antibodies and then labeled with rhodamine- goat anti-rabbit antibodies. . . . . . . . . . . . . . . . . . . . 80 ix CHAPTER IV The binding of FITC-WGA to spread E761 clone cells and spherical E761 clone cells trypsinized and suspended in methyl cellulose for one hour, one day, and two days. . . . . . . . . . . . . . . . . . . 95 Population analysis of the amount of fluorescence per cell shown in the form of frequency histograms 96 Fluorescence micrographs of spread and spherical E761 clone cells. . . . . . . . . . . . . . . . . 97 Western transfer blot of WGA glycoprotein receptors from spread and spherical E761 clone cells. . . . 98 Autoradiogram of WGA glycolipid receptors obtained from spread and spherical E761 clone cells, separated by TLC, and probed with radio-iodinated WGA . . . 99 Fluorescence micrographs of spread and spherical E761 clone cells labeled with rhodamine- -phalloidin, an F- actin specific probe 103 Uptake of 3H-thymidine into spread and spherical E761 clone cells. . . . 104 CHAPTER VI A model for the interaction of membrane proteins with the cytoskeleton as a function of cell shape and transformation 129 APPENDIX Diagram of the basic components of the FRAP instrument . . . . . . . . . . . . . 133 Representative postbleach scans of spread fibroblasts labeled with rhodamine WGA 134 Representative recovery curve after a photobleach on a spread fibroblast 135 A comparison of the diffusion coefficients of FITC—dextrans measured by the FRAP multipoint analysis and chromatography methodology 138 Representative prebleach and postbleach scans of spherical cells labeled with rhodamine WGA 140 The exponential decay of u(t) as the fluorescence redistributes from the initial azimuthal distribution set up by the photobleach . . . . . . . . . . . . . 142 xi ATP Con A DME DNA EDTA EGF F—actin FITC FRAP GDP GTP HBSS HEPES K-MSV MAPS MC mRNA NBD PBS PMSF PVP LIST OF ABBREVIATIONS adenosine-s'—triphosphate concanavalin A Dulbecco's modified Eagle's media deoxyribonucleic acid ethylenediaminetetraacetate epidermal growth factor filamentous actin fluorescein isothiocyanate fluorescence redistribution after photobleaching guanosine-5'-diphosphate guanosine-5'-triphosphate Hank's balanced salt solution 4-(2—hydroxyethyl)-1-piperazine— ethanesulfonic acid Kirsten murine sarcoma virus transformed microtubule associated proteins methyl cellulose messenger RNA 4-nitrobenzo-2-oxa-1,3-diazole phosphate buffered saline phenylmethylsulfonyl fluoride polyvinylpyrrolidone xii RNA rRNA S-Con A SDS SEM WGA ribonucleic acid ribosomal RNA succinyl concanavalin A sodium dodecyl sulfate scanning electron microscopy wheat germ agglutinin xiii CHAPTER I: OVERVIEW OF THE THESIS The involvement of cell shape as a controlling influence for DNA synthesis and growth in anchorage dependent cells was vividly demonstrated by the seminal work of Folkman and Moscona (1978). The consequences of cell shape alteration were not only limited to effects on DNA synthesis, but also resulted in the inhibition of mRNA and rRNA production (Benecke gt gt., 1978; Ben Ze'ev gt gt., 1980). Translation, on the other hand, merely required cell surface contact but not extensive spreading. The processes of cell detachment, resulting in cell rounding, and reattachment followed by spreading are accompanied by disassembly and reassembly of stress fibers, microfilaments, and microtubules (Badley gt gt., 1980). Considering the importance of these cytoskeletal structures for protein biosynthesis (Penman gt gt., 1981; Cervera gt gt., 1981), transmembrane signal initiation (Geiger, 1983; Landreth gt gt., 1985), and signal transmission (Ingber & Jamieson, 1985), the possibility that the flat to spherical shape transition interrupts a cytoskeletal communication pathway from the cell surface to the nucleus, which is required for nuclear function in untransformed fibroblasts, 1 2 is plausible and has been suggested (Ben Ze'ev, 1985). A signal transduction pathway can be divided into three components: a) signal initiation, triggered by receptor/ligand binding and the resulting receptor aggregation (Schlessinger, 1980) or by alteration of cell shape (Ben Ze'ev, 1985); b) signal transmission, e.g. soluble factors (Sutherland, 1972; Joseph, 1986) or cytoskeletal organization (Ben Ze'ev, 1985; Ingber & Jamieson, 1985); and at the nuclear level, c) signal acquisition or reception as mirrored by changes in the rate of nuclear transport (Schindler & Jiang, 1985) or replication competency (Stiles gt gt., 1979). The role of membrane protein lateral mobility in polypeptide growth factor (P6P) signal initiation (Schlessinger, 1980) and the demonstration of the modulation and restriction of protein lateral mobility by the cytoskeleton (Edelman, 1976; Koppel gt gt., 1981; Tank gt gt., 1982) indicates the significance of the link between the control of protein lateral mobility and the cytoskeleton. This thesis seeks to focus on events occuring at the cell surface and to examine the role of cell shape on the initial event in signal transduction, the membrane lateral diffusion of cell surface receptors. Lateral mobility in spread and spherical fibroblasts was examined for changes in the rate and extent of wheat germ agglutinin receptor lateral mobility. (I will use the term receptor for lectin binding proteins but do not necessarily imply a response 3 from binding as classicly inferred from the term receptor. For example, there is no immediately detectable response for WGA binding in 3T3 fibroblasts. Some lectins do initiate a response after binding and so lectin binding proteins are commonly referred to as receptors.) Other probes for membrane glycoproteins and lipids were also employed to examine the generality of potential dynamic variations in membrane lateral diffusion. A link between membrane lateral diffusion phenomenon and cytoskeletal changes was assessed by examining the potential for concanavalin A induced changes in WGA receptor lateral mobility. Such a phenomenon, termed global modulation, has been ascribed to cytoskeletal linkage of receptors. Kirsten murine sarcoma virus transformed (K-MSV) BALB/c 3T3 fibroblasts were also employed in this study to provide another form of perturbed cytoskeletal structure to examine in regard to plasma membrane lateral mobility. The state of the actin stress fiber system was analyzed and used to monitor the changes in cytoskeletal organization after alteration of cell shape and transformation state. Lateral mobility, then, is used as a dynamic, non-destructive probe to ascertain whether cell shape changes or transformation can play a role in uncoupling the putative initial event in transmembrane signaling. In addition, these experiments provide a new means to examine the suggested role of the cytoskeleton in restriction of protein lateral diffusion. 4 The major findings of this thesis include a) shape induced alteration of the cell cytoskeleton coincides with a partial release of restriction for lateral diffusion of wheat germ agglutinin (WGA) and succinyl concanavalin A receptors in untransformed BALB/c 3T3 cells; b) the observation of a bimodal population of cells with fast and slow protein lateral mobility after the alteration of cell shape in K-MSV BALB/c 3T3 cells; c) the detection and isolation of a clone with fast protein lateral mobility from the K-MSV BALB/c 3T3 fibroblasts; d) the loss of global modulation, a process affecting protein lateral mobility that requires an intact cytoskeleton (Edelman, 1976), in spherical BALB/c 3T3 cells and in the "fast" mobility clone regardless of cell shape. While this project has pursued dynamic events at the plasma membrane that could be involved in transmembrane signal transduction, the work of others in the laboratory on dynamic events at the nuclear membrane provides correlative evidence for a transduction connection between plasma membrane and nuclear activity. The research on nuclear pore mediated transport gives evidence that actin and myosin may play a role in the control of nucleocytoplasmic transport (Schindler & Jiang, 1986). These mechanochemical proteins may serve as potential sites of linkage between the nucleus and the cytoplasmic/submembranous cytoskeleton. Subsequent studies paralleling my investigations of the role of shape induced 5 changes on plasma membrane protein lateral mobility indicate that alteration of cell shape is accompanied by an effect on the rate of nuclear transport in untransformed 3T3 cells but not in the "fast" mobility clone. The relationship of this linkage to hormone action may be inferred from investigations demonstrating that hormone pretreatment with epidermal growth factor increases the rate of nuclear transport in spread cells but not spherical cells as might be predicted by the original observations of Folkman and Moscona (1978). These studies of a dynamic linkage between the plasma membrane receptors and nuclear transport may provide the first experimental evidence for the cell cytoskeleton as the "telegraphic" mediator of plasma membrane originated signals to the nuclear surface and ultimately the chromatin. REFERENCES Badley, R. A., Woods, A., Carruthers, L., & Rees, D. A. (1980) J. Cett:Sci. 3, 370-390. Ben Ze'ev, A. (1985) Cell and Muscle Motility g, 23-53. Ben Ze'ev, A., Farmer, S. R., & Penman, S. (1980) Cell gt, 365-372. Benecke, B.-J., Ben-Ze'ev, A., & Penman, S. (1978) Cell ti, 931-939. Cervera, M., Dreyfuss, 6., & Penman, S. (1981) Cell gt, 113-120. Edelman, G. M. (1976) Science 192, 218-226. Folkman, J. H. & Moscona, A. (1978) Nature 273, 345-349. Geiger, B. (1983) Biochim.. Biophys. Acta 737, 305-341. Ingber, D. E. and Jamieson, J. D. (1985) In Gene Expression During Normal and Malignant Differentiation (Andersson, L. C., Gahmberg, C. 6., and Ekblom, P., eds.), Academic Press, New York. Jiang, L.-W. & Schindler, M. (manuscript in prep) Joseph, S. K. (1985) TIBS t2, 297-298. Koppel, D. E., Sheetz, M. P., & Schindler, M. (1981) Proc. Natl. Acad. Sci. USA 1g, 3576-3580. Landreth, G. E., Williams, L. K., & Rieser, G. D. (1985) J. Cell Biol. 101, 1341-1350. Penman, S., Fulton, A., Capco, D., Ben Ze'ev, A., Wittelsberger, S., and Tse, C. F. (1981) Symp. Quant. Biol. gt, 1013-1028. Schindler, M. & Jiang, L.-W. in Comments on Mgtgcular and Cellular Biophysic (in press) Schlessinger, J. (1980) TIBS g, 210-214. Sutherland, E. W. (1972) Science 177. 401-408. 6 Stiles, C. D., Capone, 6. T., Scher, C. D., Antoniades, H. N., Van Wyck, J. J., & Pledger, W. J. (1979) Proc. Natl. Acad. Sci. USA 1g, 1279-1283. Tank, D. W., Wu, E.-S., & Webb, W. W. (1982) J. Cell Biol. 22. 207-212. CHAPTER II: LITERATURE REVIEW SIGNAL TRANSMISSION Recently, a different pathway for communication between the plasma membrane and the nucleus via the cytoskeleton has garnered interest as indicated by the growing number of reviews on the subject. Ingber and Jamieson (1985) were intrigued by the reported effects of basal membrane substances on gene expression (Bissel gt gt., 1982) correlating with changes in the cytoskeleton (Sugrue and Hay, 1981; Brown gt gt., 1983; Ali gt gt., 1977), and in light of the apparent interaction of the cytoskeleton with the nucleus (Lazarides, 1980), they suggested that the cytoskeleton could act as a transmitter of signals from the outside of the cell to nucleus. Packard (1986) echoed the growing interest in a cytoskeletal pathway between the plasma membrane and the nucleus. Schindler and Jiang (in press) reviewed nuclear-cell communications: the linkages, controllers, and pathways, and discussed the evidence for the role of the cytoskeleton as a controller of nucleo-cytosplasmic transport and as a possible highway for directing a fraction of internalized 8' 9 growth factor-receptor complexes to the nucleus. The model of a signal transmission pathway via the cytoskeleton suggests that there should be three parts: a connection between plasma membrane proteins and the cytoskeleton, a cytoskeletal network spanning the cytoplasm, and an interaction between the cytoskeleton and the nucleus. The connection between plasma membrane proteins and the cytoskeleton has been demonstrated by dynamic evidence from studies on the restriction of protein lateral mobility (Edelman, 1976; Koppel gt gt., 1981; Tank _t _t,, 1982); structural evidence from immunofluorescence microscopic studies (Ash gt gl., 1977; Gabbiani, 1977; Geiger gt gt., 1984; Rogalski & Singer, 1985); and biochemical evidence from studies examining the association of membrane proteins with detergent insoluble cytoskeletal structures (Painter & Ginsberg, 1982; Sheterline & Hopkins, 1981; Koch & Smith, 1978; Streuli gt gt., 1981). Other studies using intact cells have suggested that the binding of crosslinking agents leads to aggregation of the bound membrane receptors (referred to as patching). The coalescence of these patches into a polar cap is known as capping. The binding of Con A and other crosslinking agents such as antibodies enhanced the binding of surface receptors to a detergent insoluble cytoskeleton (Flanagan & Koch, 1978). This occurred only with divalent antibodies and not with antibody fragments that do not crosslink membrane receptors; furthermore, association of membrane 10 receptors with the cytoskeleton still occurred under conditions where crosslinking resulted in patching, but capping and internalization were prevented by cold treatment. Thus, aggregation appears to be a key step in enhancing the association between cell surface receptors and the cytoskeleton as well as being a key step in growth factor action (Schlessinger, 1980). The significance of these interactions between the cytoskeleton and membrane proteins to growth factor action was indicated by the report that epidermal growth factor receptors are associated with the detergent insoluble cytoskeleton (Landreth gt gt., 1985). These lines of evidence will be discussed further in the section on the cytoskeleton. The spanning of the cytoplasm by the cytoskeleton was most readily visualized in electron micrographs from whole mounts of cells (for example see Penman gt gt., 1981). Evidence for an interaction between the cytoskeleton and the nucleus is both structural and dynamic. Electron micrographs suggested that cytoskeletal elements come, at least, in close apposition to the nuclear membrane (Franke, 1971). Even more suggestive was the report that actin and myosin play a role in the control of the rate of nucleo- cytoplasmic transport (Schindler and Jiang, 1986). In this study it was shown that an enhancement of nucleocytoplasmic transport by ATP could be blocked by treatment with microfilament disrupting drugs, anti-actin antibodies, or anti-myosin antibodies. Putting the pieces together 11 indicates the plausibilty of a linkage between plasma membrane proteins and the nuclear structures. In support of this, a correlation between the position of plasma membrane antigens and nuclear antigens has been noted in certain transformed cells treated with cytochalasin B (Berke and Fishelson, 1976; Otteskog gt gt., 1981). Cytochalasin B induced capping of membrane antigens in L cells, SV40-3T3 cells, and NRK La 334 cells. Capping did not occur in enucleated cells suggesting a requirement for the presence of the nucleus. Correlating with the capping of the plasma membrane antigens, a capping of nuclear antigens was detected on the nuclear membrane on the side of the nucleus closest to the plasma membrane cap, even though the nucleus was located on the opposite side of the cell from the cap. In another recent study a cytochalasin sensitive link between acetylcholine receptor clusters and the nucleus was observed in muscle fibers (Englander & Rubin, 1987). Nuclei of muscle fibers having a homogeneous distribution of acetylcholine receptors were observed to move through the myotube. A clustering factor reorganized the acetylcholine receptors into clusters; nuclei near the clusters became stationary. If clusters were dispersed, the nuclei in the proximity became mobile. Addition of cytochalasin with the clustering factor prevented the formation of clusters and nuclei remained mobile. Cytochalasin did not affect preformed clusters and associated nuclei. In a demonstration of the ability of nuclear structures to 12 affect the organization of the cytoskeleton, it has been observed that during polar body formation in mouse oocyte that chromosomal clustering induced a reorganization of subcortical microfilaments resulting in increased accumulation of actin in some areas below the membrane associated with a decrease in microvilli (Maro gt gl., 1986). THE CYTOSKELETON The cytoskeleton consists of several types of tension bearing and tension producing filaments and networks. These networks provide the structural support that gives cells their shape. The cytoskeleton is also involved in a variety of cell motile events, for example, cell movement, membrane ruffling, flagellar movement, endocytosis, exocytosis, adhesion, spreading, chromosomal separation, and cell division (Goldman and Knipe, 1972; Birchmeier, 1984; Tomasek and Hay, 1984). The cytoskeleton has even been implicated in a role of providing an organizational scaffolding to localize and regulate translation as well as many cellular enzymatic activities (Cervera gt gt., 1981; Clegg, 1984). The intricate cytoskeletal networks pervading the cytoplasm provide a means by which cellular proteins and activities can be localized to specific regions of the cell resulting, in effect, in compartmentalization of 13 cellular functions. For example, in one study it was observed that translation of mRNA occured only when the mRNA was attached to the cytoskeleton (Cervera gt gt., 1981). Furthermore, it has been observed that the localization resulted in the insertion of some newly synthesized cytoskeletal proteins into the cytoskeleton at the site of synthesis (Penman gt gt., 1981). One way that association/dissociation of enzymes and substrates to the cytoskeleton or membranes could aid or regulate cellular activities is that during the process of association and dissociation, molecules tend to remain in the vicinity for a longer period of time resulting in a quasi-two dimensional diffusion. Such hopping on and off could limit the distance of diffusion required for collision and, perhaps more importantly, allow more attempts at binding (McCloskey & Poo, 1984). The cytoskeleton, apparently, is a strange mixture of static states and dynamic processes. One group has described it as a system in equilibrium composed of components independently in disequlibria (Ingber & Jamieson, 1985). They represented the cytoskeleton as a system of tension and compression yielding a dynamic, yet stable, structure in a model termed tensegrity. Such a model would allow cells to maintain a form, sense changes in the environment, and quickly react accordingly. In support of this model it was found that in PC12 neurites, depolymerization of microtubules results in retraction of 14 neurites and that retraction of the neurites can be prevented by disruption of microfilaments (Joshi gt gt., 1985). Microfilament disrupting drugs also increased the dose of microtubule disrupting drugs required to cause microtubule depolymerization. These results support their contention that microtubules are under compression arising from tension produced by microfilaments and suggest that variation of compression may serve as a local regulator of microtubule assembly. To better understand how the cytoskeleton works, one needs to examine the components. The cytoskeleton consists of several types of filaments and networks composed of microfilaments, microtubules, intermediate filaments, and spectrin and their associated proteins. Microfilaments consist of actin and actin binding proteins which regulate the form of the microfilaments. Actin monomers have a polarity which can be distinguished by the possession of a barbed end and a pointed end. The monomers polymerize under appropriate conditions to form long filaments in a head to tail configuration. The steps of polymerization are: a) activation of the monomer resulting in a conformational change by cations such as magnesium; b) nucleation to form short polymers (apparently the rate limiting step); c) elongation of the nuclei by a head to tail addition of monomers; and d) longer filaments can be generated by annealing of fragments (Pollard and Craig, 1982). In solutions of purified monomer the extent 15 and rate of polymerization depends on the concentration of monomer present. Such filaments can, apparently, have addition of monomers at both ends simultaneously; however, the association constants appear to differ at the two ends resulting in a difference in the concentrations necessary for polymerization to occur. In vitro polymerization may occur initially at both ends until the concentration of the monomer pool drops below the critical concentration for one of the ends. At this point net depolymerization can occur at one end while net polymerization continues at the other end. Eventually, a steady state is reached where the polymerization at one end is matched by depolymerization at the other end. In such a state a phenomenon, called treadmilling, can occur (Wegner, 1976). In treadmilling a monomer is added to one end, moves through the polymer, and comes off the far end. This is the result of net polymerization at one end and net depolymerization at the far end. In vivo, however, this is believed to be a rare event since actin is usually not in a pure form but bound and regulated by many actin binding proteins (Craig and Pollard, 1982; Birchmeier, 1984). These proteins can block or cap an end of the filament resulting in polymerization or depolymerization depending on the end. Examples of these capping and length regulating actin binding proteins include g-actinin, fragmin, gelsolin, villin, and brevin. Some of these proteins, such as actinin and gelsolin, cap the filaments; others such as fragmin can fragment polymers 16 of actin. Actin binding proteins, profilin for example, can depolymerize polymers of actin instead of merely blocking further addition or fragmenting the longer polymers. Actin polymers can be found in various forms including long single filaments, bundles of filaments, and a meshlike network. To process actin into these various forms there is a host of actin binding proteins that can crosslink actin into bundles or networks. Membrane proteins can also be actin binding proteins, e.g. ankyrin and vinculin, and are believed to anchor actin filaments to the membrane (Birchmeier, 1984). Thus, the microfilament system can take on various forms, be stabilized, and be destabilized rapidly by interaction with a variety of actin binding proteins. Adhesion and interaction with the extracellular matrix resulting in aggregation of membrane actin binding proteins may provide another means of controlling the formation of cytoskeletal assemblies (Bissel gt gt., 1982). The distribution of actin in cells generally is described in two forms, both near the membrane. Studies using electron and fluoresence microscopy suggested that actin is found mainly in stress fiber bundles and membrane ruffles (Glacy, 1983; Goldman gt 31., 1975; Lazarides, 1975). Often the stress fibers were observed to run the length of the cell, under the nucleus and adjacent to the portion of the plasma membrane attached to the growth substrate. Stress fibers were observed only in attached and spread cells. In the membrane ruffles the actin appeared to 17 be in networklike gels that could rapidly form and disappear. Other studies have suggested that actin forms a submembranous sheath entirely enclosing the membrane that may undergo rapid gel-sol transitions during membrane motile events (Davies and Stossel, 1977; Stossel gt gt., 1981). In one study the actin cortical sheath appeared to enclose most of the cytoplasm and the microtubule system as well (Henderson and Weber, 1979). This does not preclude the possibility of actin structures being part of the cytoplasmic cytoskeleton as most of these studies also observed weaker, general staining throughout the cytoplasm. Actin filaments are generally considered tension producing filaments through their interaction with myosin. Although it has been suggested that stress fibers might be tension bearing structures, they are usually considered to be tension producing structures (Burridge, 1981). Microtubules are another polymeric structure of the cytoskeleton and consist of a heterodimer of o and E3 tubulin. GTP bound to the tubulin enhances assembly and the hydrolysis of the GTP to GDP in the microtubule stabilizes the microtubule (Bonne & Pantaloni, 1982). Assembly of microtubules is similar to that of actin in that assembly can occur at both ends with a difference in dissociation constants and critical concentrations for each end. As for microfilaments, this difference in critical concentrations at the ends could result in treadmilling of microtubules (Margolis & Wilson, 1978); however, this does not appear to 18 occur in vivo (Scherson gt gt., 1984). Microtubules also contain various microtubule associated proteins, known as MAPs, which assist and regulate microtubule polymerization. One of these MAPs, Tau, is required for microtubule assembly (Weingarten gt gt., 1975). It appears unlikely that microtubules exist free in the cytoplasm for any significant period of time; instead, microtubules appear to be initiated from microtubule organizing centers and grow toward the plasma membrane (Osborn & Weber, 1976; Solomon, 1980). The microtubules approach the plasma membrane but do not usually reach it; the network of microtubules appears to be encircled by and interact with a submembranous network of actin (Henderson & Weber, 1979). In support of this contention, some MAPs are also actin binding proteins (Craig & Pollard, 1982). Microtubules, however, have been demonstrated to interact with membranes in algae (Murray, 1983), hepatic cells (Reaven & Azhar, 1981), and brain and thyroid tissue (Bhattacharyya & Wolff, 1975); therefore, a role for microtubules in cell surface events can not be eliminated. The microtubules play an important role in the maintenance of cell architecture and cell motility (Goldman, 1971). Microtubules appear to be tension bearing elements in the cell, perhaps acting as the two by fours of the cell (Ingber & Jamieson, 1985; Joshi gt gt., 1985). There is evidence that compression of the microtubules may even play a role in depolymerization and length regulation of microtubules (Joshi gt gt, 1985). Microtubules also 19 appear to be involved in several other cellular activities including protein (Busson-Mabillot gt gt., 1982) and lysosomal (Hoffstein gt gt., 1977) secretion; localization of mitochondria (Bernier-Valentin & Rousset, 1982); organelle movement via kinesin, a microtubule locomotor equivalent of myosin in acto-myosin based filaments (Vale t 1., 1985); and may even interact with DNA (Corces gt gt., 1978). A third type of filamentous structures is the intermediate filament, so called because its thickness is intermediate that of microfilaments and microtubules. Intermediate filaments are composed of polymerized monomers as are microfilaments and microtubules. The monomers however differ among cell types and fall into five general classes: desmin in myogenic cells; neurofilaments in neuronal cells; glial filaments in astroglial cells; vimentin in mesenchymal cells; and a complex class of keratins from epithelial cells (Steinart gt gt., 1984; Lazarides, 1982). Since this project uses cultured fibroblasts, the rest of this discussion will be concerned with vimentin, the primary type of monomer found in these cells. The intermediate filaments form a highly insoluble network extending from the nucleus to out near the periphery of the cell. The intermediate filaments appear to form a meshwork of filaments that have been described as a basket or a cage. The intermediate filament system appears to contain its own arsenal of intermediate filament 20 associated proteins to control assembly and disassembly. This network of intermediate filaments appears to be much more static than the dynamic microfilament and microtubule systems. Even when cells round up, the intermediate filament network fails to depolymerize; instead, it collapses about the nucleus forming a cap or ring. The network appears to also be linked to the microtubule system, since the network will collapse about the nucleus when cells are treated with microtubule disrupting agents (Steinart gt gt., 1984). The intermediate filament network has also been suggested to interact with a cortical actin system on the ventral surface of the cell (Henderson and Weber, 1979). Because of its insolubility and close association with the nucleus, one of its major roles is considered to be centration of the nucleus in addition to other structural and organizational duties. Spectrin is another cytoskeletal element that until recently was thought to be limited to erythrocytes. Spectrin is a large asymmetric molecule composed of a heterodimer (composed of Band 1 and Band 2) that can associate to form a tetramer (for reviews see Steck, 1974; Branton et al., 1981). Spectrin is attached to the membrane via a protein referred to as ankyrin, which attaches it to the membrane by binding to a transmembrane protein known as band 3. Spectrin also interacts with actin oligomers through band 4.1. In models of the erythrocyte membrane, spectrin is portrayed as two dimers linked head on and 21 interacting with actin oligomers at either end. The actin oligomers would thus appear to act as nodes for spectrin tetramers to bind to. Ankryin and band 3 serve to tack the system to the membrane. Recently, the counterparts to spectrin, ankryin, and band 4.1 have been identified in several nonerythroid cell types (Burridge _t gt., 1982; Moon gt gt., 1985; Cohen gt gt., 1982). Whether the structures formed by these components in nonerythroid cells are the same as in the erythroid cells is still not clear. HISTORY OF PROTEIN LATERAL DIFFUSION According to the Singer and Nicolson (1972) model, the plasma membrane consists of a phospholipid bilayer forming a sea where proteins can float embedded in the bilayer or associated with the bilayer. Proteins associated with only the external or cytoplasmic face of the bilayer are called peripheral membrane proteins. Proteins embedded in one half of the bilayer are considered to be integral proteins and proteins traversing the entire membrane are referred to as integral, transmembrane proteins. The intermixing of labeled antigens after cell fusion demonstrated the ability of membrane proteins to laterally diffuse (Frye and Edidin, 1970). Saffman and Delbruck (1975) developed diffusion equations to predict the rate of protein lateral diffusion. They predicted that the rate of protein lateral diffusion would be only weakly affected by the size of the molecule 22 and limited by the viscosity of the lipid bilayer. Subsequent measurement of protein lateral diffusion such as rhodopsin in rod outer segments (Poo and Cone, 1974) or various proteins in artificial vesicles prepared from plasma membranes (reviewed in McCloskey and Poo, 1984) supported the Saffman-Delbruck model. The measurement of protein lateral mobility in natural membranes, however, demonstrated a much slower rate of diffusion than expected by a factor of 10-100 (for reviews see Cherry, 1979; McCloskey and Poo, 1984). The measurement of the rate of protein lateral diffusion was facilitated by the development of fluorescence redistribution of photobleaching (FRAP). In FRAP a sample is homogeneously labeled with a fluorescent probe. Diffusion is detected by photochemically destroying the fluorescent label in a small region of the sample with a focussed laser beam. Monitoring the redistribution offluorescent probe from outside the bleached region into the bleached region over a period of time allows the calculation of the rate of diffusion and extent of recovery (see appendix A). Initially, the use of intense laser light to photobleach the label caused concern over the possibility that crosslinking of membrane components could result in artifactual measurements. Several experiments have been carried out that suggest this is not the case (reviewed in Wolf gt gt., 1980): a) localized heating from exposure to the laser beam is less than 0.03°C; b) the 23 measurement of diffusion by FRAP on model membranes were independently confirmed by measurments using fluoresence correlation spectroscopy; c) FRAP results are generally consistent with estimates obtained from intermixing of surface antigens after heterokaryon fusion; d) FRAP measurements are generally independent of the degree of photobleaching; e) multiple bleaches in the same region of a cell do not affect diffusion rates and result in increases in the extent of recovery (after successive bleaches the immobile components will have been bleached and would no longer be detectable and so only mobile components are seen causing the apparent increase in the extent of recovery), significant crosslinking could be observed as an increase in immobile components replacing those bleached in the previous bleach; f) FRAP does not alter membrane morphology in scanning electron micrographs; 9) free radical quenchers do not alter FRAP results; and h) dual labeling of FC receptors with two labels demonstrated that extensive bleaching of one label does not affect FRAP measurements made with the other label. It has been demonstrated that extremely long bleaching times using low power, in the range of seconds instead of the milliseconds typically used in FRAP, can result in oxidation of lipids (Sheetz & Koppel, 1979) probably through the production of singlet oxygen. Thus caution is called for in the selection of bleaching times. Under the typical conditions of FRAP cellular damage is unlikely to occur since the fluorescent 24 dyes chosen for FRAP are excited at wavelengths above the region where cellular molecules typically absorb. The development of FRAP facilitated the measurement of diffusion and as mentioned above, resulted in the discovery of the restriction of protein lateral diffusion in natural plasma membranes. This sparked a search for the cause of the restriction of protein lateral diffusion. The first clue came from studies of the effect of concanavalin A on patching and capping of membrane antigens in lymphocytes. The observation that concanavalin A restricted cap formation, followed by the finding that the restriction could be partially reversed by cytoskeletal disrupting agents (Edelman gt gt., 1973), led to the proposal that a Con A induced reorganization of the cytoskeleton resulted in an increased interaction of membrane proteins with the cytoskeleton (Edelman, 1976). Studies, in which FRAP was used to measure the rate of protein lateral diffusion, confirmed that concanavalin A treatment resulted in a decrease in protein mobility and found that the combination of a microtubule disrupting drug (colchicine) with a microfilament disrupting drug (cytochalasin B) completely reversed the restriction of protein lateral mobility induced by concanavalin A (Henis and Elson, 1981). On this basis the cytoskeleton became a leading candidate for the restriction of protein lateral mobility. An examination of protein lateral diffusion in spectrin deficient mouse erythrocytes (spherocytes) found that protein lateral 25 diffusion was 50 fold faster in the spherocytes than in the normal erythrocytes (Koppel gt gt., 1981). These results were interpreted in the context of spectrin interacting with membrane proteins resulting in the slow diffusion observed in normal erythrocytes. Further evidence for the role of cytoskeletal interaction in restriction of protein lateral diffusion was provided by the investigation of lateral diffusion in membrane blebs (Tank gt gt., 1982). Membrane blebs are regions of the plasma membrane pulled away from the underlying cytoskeleton. Protein lateral diffusion in these blebs was found to be 50 fold higher than in unaffected regions of the membraneand similar to that of lipid diffusion. In addition to the dynamic evidence for an interaction between membrane proteins and the cytoskeleton there is physical evidence from numerous studies. Some studies using electron microscopy indicated a proximity of cytoskeletal structures to the membrane (Henderson & Weber, 1979; Stossel gt gt., 1981). Other studies have found an association between surface antigens and cytoskeletal elements at the fluorescent or electron microscopic level in chicken fibroblasts (Rogalski and Singer, 1985), rat mammary ascites (Carraway, et a1, 1982), neutrophil leukocytes (Sheterline and Hopkins, 1981), and a murine mastocytoma (Koch and Smith, 1978) to name a few. In light of the use of wheat germ agglutinin as the probe for protein lateral mobility in this thesis, it is significant 26 to note that one ultrastructural study found WGA and Con A receptors localized to regions of the membrane overlying the filaments of the cortical network (Roos gt gt., 1985). In several of these studies it was also determined that detergent extraction of the cells had no effect on the distribution of the antigens or the cytoskeleton. Furthermore, concanavalin A has been demonstrated to increase the interaction of membrane proteins with the cytoskeleton in platelets (Painter and Ginsberg, 1982) and to recruit actin to the membrane in lymphocytes (Michaels _t gt., 1984). Crosslinking of Ig receptors by Ig enhanced the attachment of the Ig receptors to actin structures even in the absence of capping suggesting that aggregation is a critical step in initiating a link between the receptor and the cytoskeleton (Flanagan & Koch, 1978). While numerous studies have suggested an association between membrane proteins and actin in many cell types and between spectrin and ankyrin in erythrocytes, only a few have suggested an association between membrane proteins and other cytoskeletal elements such as microtubules and intermediate filaments. Microtubules have been shown to bind to plasma membranes (Reaven & Azhar, 1981) and a membrane glycoprotein has been found to remain with intermediate filaments upon detergent extraction (Lehto, 1983). However, ultrastructural studies would suggest that microtubules and intermediate filaments are generally excluded from the cortical submembranous region which appears to be occupied 27 mainly by actin networks (Henderson & Weber, 1979). Because of the effect of colchicine on global modulation, microtubules are thought to play a significant role in the interaction of the cytoskeleton with membrane proteins. In light of the evidence for interaction between the two cytoskeletal systems (Craig & Pollard, 1982; Henderson & Weber, 1979), it is possible that microtubules play a role in organizing or supporting a cortical actin system. Interaction of membrane proteins with the cytoskeleton was not the only model proposed to explain the restriction of protein lateral diffusion. Density of protein in the membrane has been shown to affect lateral diffusion probably by increasing the viscosity of the membrane (Golan _t gt., 1984; McCloskey & Poo, 1984). The amount of protein in natural membranes may slightly affect lateral diffusion, but is almost certainly not enough to explain the observed slower rate of protein lateral diffusion. It has been proposed that segregation of lipids into solid and fluid domains, or even two immiscible fluid domains, could slow protein lateral diffusion as proteins diffuse through solid domains or cross boundary regions (Webb gt gt., 1981; Axelrod, 1983; McCloskey and Poo, 1984). While such domains have been observed in plant cells (Metcalf gt _t, 1986), the evidence of their presence in animal cells is still limited. A more likely candidate is the interaction of membrane proteins with the extracellular matrix. The extracellular matrix is a complex system of proteins, 28 proteoglycans, and glycoproteins that form a matrix outside the cell and is believed to be involved in attachment and perhaps even gene expression (Bissel gt _t., 1982) and differentiation (Ingber & Jamieson, 1985). Many of the components, such as fibronectin, are known to interact with membrane proteins as part of their role in attachment (Yamada & Olden, 1978) Fibronectin has even been suggested to be associated with and perhaps organize actin structures (Hynes and Destree, 1978; Burridge & Feramisco, 1980; Virtanen gt gt., 1982). It seems likely, then, that the extracellular matrix could play a role in the restriction of protein lateral diffusion. It is surprising how little experimental evidence exists for such a role. One component of the extracellular matrix, fibronectin, has been demonstrated to be immobile on the time scale of the experiment, but the lateral mobility of cell surface antigens was not affected by the presence or absence of fibronectin (Schlessinger gt gt., 1977). The main evidence for such a model was provided by a study on the effects of cell density on lateral mobility and the effects of plating cells on dishes precoated with extracellular matrix components (Weir and Edidin, 1986). Protein lateral mobility was compared on cells from sparse cultures with cells from dense cultures. No effect on the rate of protein lateral diffusion was observed; however, the extent of recovery decreased in the dense cultures. When cells were sparsely plated on dishes precoated with extracellular 29‘ matrix components and compared to cells plated onto proteolyzed extracellular matrix components, it was observed that the presence of extracellular matrix components decreased the extent of recovery, but not the rate of recovery. Thus, it seems likely that the extracellular matrix associates with its receptors so tightly that those proteins appear immobile on the time scale of the experiment, while the unbound proteins remain unaffected as detected by the lack of effect on the rate of lateral diffusion. This differs, then, from the results of experiments investigating the interaction of the cytoskeleton with membrane proteins noted above where the rate of diffusion is affected. It has been suggested that the interaction with the cytoskeleton may occur with an on/off rate fast enough to allow diffusion to be detected on the time scale measurable by FRAP (Koppel gt gt., 1981; Koppel, 1981; Elson and Reidler, 1979). The results from the extracellular matrix experiment could then be interpreted to mean that membrane proteins associate with the extracellular matrix with too slow an on/off rate to detect an effect on the rate of lateral diffusion, but those already bound would appear to be immobile. 30 THE SYSTEM: ALTERATION OF CELL SHAPE BALB/c 3T3 fibroblasts are an aneuploid culture line derived from mouse embryos. Untransformed BALB/c 3T3 cells demonstrate cell growth regulated by the density of the cell population and a requirement for serum growth factors (Holley, 1975; Holley & Kiernan, 1968). When cells are plated sparsely onto a suitable growth substratum in the presence of serum, the cells will attach and spread into a characteristic elongated shape. As the density of the cells increase and the cells come into contact, the growth rate slows and the cells become slightly rounded until the cells form a monolayer and assume a cobblestone pattern. At confluence the cells become quiescent in a phenomenon called density dependent inhibition of cell growth. Attachment and spreading is also required for cell growth. Cells suspended in agar or methyl cellulose, a common test for transformation, become quiescent but remain viable for a few days (Otsuka & Moskowitz, 1975; MacPherson & Montagnier, 1964). Suspension of cells in methyl cellulose has been a useful means for suspending cells in way that allows recovery of the cells by simple dilution and centrifugation. The methyl cellulose is dissolved into the normal media and serum used for culturing the cells and merely acts as a thickener to suspend the cells and decrease aggregation. The cells will grow in methyl cellulose if they are provided a large enough substratum, 31 such as glass fibrils, indicating that the methyl cellulose is nontoxic to the cells (Stoker gt gt., 1968). Cell shape and adhesion was demonstrated to be vitally important to cell growth in anchorage dependent fibroblasts by Folkman and Moscona (1978) in a landmark study investigating the effect of varying substrate adhesiveness on cell shape and DNA synthesis. Fibroblasts suspended in methyl cellulose demonstrated a rapid decrease in DNA and mRNA synthesis followed by a more gradual decrease in protein synthesis (Benecke gt gt., 1978; Ben Ze'ev gt gt. 1980; Farmer gt gt., 1978). Reattachment of the cells where spreading was prevented, was followed by a rapid recovery of protein synthesis, but not DNA or mRNA synthesis. mRNA was still present in suspended cells but demonstrated aberrant translation properties in in vitro translation systems. Protein translation resumed shortly after attachment before mRNA synthesis reumed and even in the presence of actinomycin (blocks mRNA synthesis). From these results it was concluded that mRNA is stored in an unusable form during suspension. DNA and mRNA synthesis resumed only when extensive cell spreading was allowed to occur. Interestingly, a major protein comigrating with actin in two dimensional gel electrophoresis showed enhanced synthesis during the recovery from suspension. Clearly, cell shape is important to several key cellular metabolic processes in anchorage dependent fibroblasts. It has been shown that doses of growth factors, normally sufficient for 32 stimulation of cell growth in spread cells, had no effect on suspended, spherical cells (Schubert & Lacorbiere, 1976). A relation between cell shape and the mitogenic response has been noted in various systems (Ben-Ze'ev, 1985). Furthermore, a recent study has demonstrated the association of EGF-receptor complexes with the detergent insoluble cytoskeleton was decreased by the alteration of cell shape in an anchorage dependent cell line (Landreth gt gt, 1985). This raises the possibility that the normal signal transduction pathway has been disrupted in suspended cells. Alteration of cell shape also has a dramatic effect upon the organization of the cytoskeleton. The cytoskeleton in attached and spread fibroblasts is a highly organized network of microfilaments encircling the cytoplasm, stress fibers running underneath the nucleus, a few microfilaments running through the cytoplasm, bundles of microtubules radiating throughout the cytoplasm and enclosed by the microfilaments, and an intermediate filament network surrounding the nucleus and extending through the cytoplasm (Henderson & Weber, 1979). When cell shape is altered by detachment of the cells, the stress fibers, microfilaments in the cytoplasm, and microtubules disappear as observed by electron microscopy (Badley gt gt., 1980). It seems probable that the intermediate filament network collapses about the nucleus when the microtubules disappear as happens in colchicine treated cells (Steinart gt gt., 33 1984). The fate of the cortical actin network is unknown. In HeLa cells, alteration of cell shape does not alter the monomer to polymer actin ratio suggesting that cell shape alteration results in a reorganization of and/or possible shortening of microfilaments, but it does not seem to result in a depolymerization to monomers (Blikstad & Carlsson, 1982). The alteration of the cytoskeleton is related to the change in cell morphology and has even been suggested to be linked to the change in gene expression that occurs upon cell shape change (Ben Ze'ev, 1985). The use of anchorage dependent fibroblasts, such as BALB/c 3T3 cells, provides a convenient system in which to study phenomena accompanying cytoskeletal alteration upon changes in cell shape including putative effects on protein lateral mobility, communication pathways, and nuclear activities. True, the cytoskeleton can be disrupted in other ways, such as the use of cytoskeletal disrupting drugs. The cytoskeletal disrupting drugs, colchicine (microtubule disrupting) and cytochalasin B (microfilament disrupting) were in fact used to demonstrate the involvement of the cytoskeleton with the restriction of membrane protein lateral diffusion via global modulation, as discussed above. The effects of the cytoskeletal disrupting drugs, however, produce conflicting results. Attempts to detect an effect of cytoskeletal disrupting drugs on protein lateral mobility without global modulation, i.e. cells not treated with concanavalin A, has 34 produced mixed results. In some instances the drugs appeared to enhance lateral mobility only slightly or not at all; in others it seemed to even decrease mobility to some extent (Elson & Reidler, 1979; Elson & Schlessinger, 1980; Eldridge gt gt., 1980; McCloskey & Poo, 1984; Webb gt gt., 1981). The effects on the cytoskeleton also appear to vary. Usually considered to cause microfilament break down, cytochalasin B has been observed to cause local increases in microfilament concentration underneath the membrane in some cells (Miranda gt gt., 1974). Even the effects of cytoskeletal disruption by drugs on cellular metabolism varies. Microtubule disruption by colchicine initiates DNA synthesis (Crossin & Carney, 1981); whereas, microfilament disruption by cytochalasin blocks initiation of DNA synthesis (Maness & Walsh, 1982) but can also enhance the effect of growth factors in initiation of DNA synthesis (Otto gt gt., 1979). In addition, the cytoskeletal disrupting drugs have some side effects such as the inhibition of glucose uptake by cytochalasin B (Estensen & Plagemann, 1972). Transformation of fibroblasts is another factor that alters the organization of the cytoskeleton. Transformed fibroblasts lack the thick bundles of stress fibers readily apparent in untransformed cells. Instead, the transformed cells contain fewer, thinner stress fibers and in some instances no apparent stress fibers (Ambrose gt gt., 1970; Pollack gt gt., 1975). Filamentous actin aggregated into 35 patches described as rosettes near the ventral surface associated with adhesion plaques has also been observed (David-Pfeuty & Singer, 1980; Carley gt gt. 1981). It is intriguing that these F-actin aggregates are insensitive to actin disrupting drugs (Carley gt gt., 1983). As with the microfilaments, microtubules are scarce in the cytoplasm of transformed cells (Brinkley gt gt., 1975). Despite these differences in the organization of the cytoskeleton, the lack of effect on the lateral mobility of membrane proteins has been noted (Eldridge gt gt., 1980). While this lack of effect is difficult to understand, it must be noted that transformed cells do appear to have some functional form of cytoskeleton to maintain cell shape and mobility. In addition, cytochalasin B in some transformed cells, but not untransformed cells, has been observed to initiate a co—capping of plasma membrane and nuclear membrane antigens and therefore hinting at the possibility of a linkage between the plasma membrane and the nuclear membrane in these treated, transformed cells (Otteskog gt gt., 1981). Another example shows that transformation can affect a dynamic process in which the cytoskeleton plays an active role. An examination of the effect of transformation on global modulation found that transformation inhibited the ability of concanavalin A to restrict protein lateral mobility (Edelman & Yahara, 1976). The story behind the effects of transformation does seem to be complicated, as seen for the cytoskeletal disrupting drugs; however, the 36 ability of the transformed cell to grow in suspension (Freedman & Shin, 1974) suggests that the comparison of the effects of the alteration of cell shape in transformed cells to untransformed cells could prove useful. The experiments discussed in my thesis refer to studies done on fibroblastic cells and thus the conclusions drawn from these studies are limited to fibroblastic cells. Other cell types, such as endothelial cells, have a different shape and cytoskeletal organization and so may have different responses to the conditions of these experiments which has not been tested. In summary, the alteration of cell shape in anchorage dependent fibroblasts provides a system with known effects on gene expression and the cytoskeleton. The development of such a system would be useful in examining the role of the cytoskeleton in restriction of protein lateral mobility and putative communication pathways. The availability of transformed cells from the same cell line provides another means of examining lateral diffusion and putative communication pathways in cells with a perturbed cytoskeleton as well as an opportunity for the comparison of the effects of cell shape alteration in a cell system in which cell growth is unaffected by suspension. 1. 2. 37 Summary o§,Objectives Examine the role of cell shape in the putative initial Test event in transmembrane signaling, the lateral mobility of membrane receptors and its link to the cytoskeleton by: the investigation of WGA receptor lateral mobility in spread and spherical cells monitoring the state of the actin stress fiber system correlating these events with studies by others in this laboratory on nucleocytoplasmic transport for differences in receptors and cell surface morphology characterize lectin receptor binding examine glycoprotein and glycolipid receptors scanning electron microscopy 3. Monitor the consequences of shape change on glucose utilization and DNA Synthesis 4. Determine the extent of the increased lateral mobility in spherical cells by extending the measurements to lipid mobility b. 38 other glycoproteins that bind succinyl concanavalin A receptors. 5. Examine the effect of transformation as another means of 6. Test perturbing the cytoskeleton on . WGA receptor lateral diffusion the actin stress fiber system in conjunction with the alteration of cell shape the relationship between the increased protein lateral diffusion after cell shape alteration and the cytoskeleton by determining the effect of cell shape alteration on global modulation examining the relation between the altered stress fiber patterns and global modulation in the E761 clone. REFERENCES Ali, I. U., Mautner, V., Lanza, R., & Hynes, R. O. (1977) Cell tt, 115-126. Ambrose, E. J., Batzdorf, U., Osborn, J. S., & Stuart, P. R. (1970) Nature 227, 397-398. Ash, J. F., Louvard, D., & Singer, S. J. (1977) Proc. Natl. Acad. Sci. USA 1t, 5584-5588. Axelrod, D. (1983) J. Membrane Biol. 1;, 1-10. Badley, R. A., Woods, A., Carruthers, L., & Rees, D. A. (1980) J. Cell Sci. 43, 370-390. Ben Ze'ev, A. (1985) Cell & Muscle Motility g. 23-53. Ben Ze'ev, A., Farmer, S. R., & Penman, S. (1980) Cell gt, 365-372. Benecke, B.-J., Ben Ze'ev, A., & Penman, S. (1978) Cell ttfl 931-939. Berke, 6. and Fishelson, Z. (1976) Proc. Natl. Acad. Sci. USA 1;, 4580-4583. Bernier-Valentin, F. & Rousset, B. (1982) J. Biol. Chem. 257, 7092-7099. Bhattacharyya, B. & Wolff, J. (1975) J. Biol. Chem. 50, 7639-7646. Birchmeier, W. (1984) TIBS 2, 192-195. Bissel, M. J., Hall, H.6., and Parry, 6. (1982) J. Theor. Biol. 2g, 31-68. Blikstad, I. & Carlsson, L. (1982) J. Cell Biol. gt, 122-128. Bonne, D. & Pantaloni, D. (1982) Biochemistry gt, 1075-1081. Branton, D., Cohen, C. M., & Tyler, J. (1981) Cell gt, 24-32. 39 4O Brinkley, B. R., Fuller, 6. M., & Highfield, D. P. (1975) Proc. Natl. Acad. Sci. USA 1g, 4981-4985. Brown, S. S., Malinoff, H. L., and Wicha, M. S. (1983) Proc. Natl. Acad. Sci. USA g9, 5927-5930. Burridge, K. (1981) Nature 294, 691-692. Burridge, K. & Feramisco, J. R. (1980) Cell tg, 587-595. Burridge, K., Kelly, T., & Mangeat, P. (1982) J. Cell Biol. 2;, 478-486. Busson-Mabillot, S., Chambaut-Guerin, A.-M., Ovtracht, L., Muller, P., & Rossignol, B. (1982) J. Cell Biol. 2;, 105-117. Carley, W. W., Barak, L. S., & Webb, W. W. (1981) J. Cell Biol. 29, 797-802. Carley, W. W Physiol. ., Lipsky, M. 6., & Webb, W. W. (1983) J. Cell 117, 257-265. Carraway, K. L., Cerra, R. F., Jung, 6., & Carraway, C. A. C. (1982) J. Cell Biol. 22. 624-630. Cervera, M., Dreyfuss, 6., & Penman, S. (1981) Cell gg, 113-120. Cherry, R.J. (1979) Biochim. Biophys. Acta 559, 289-327. Clegg, J. S. (1984) Am. J. Physiol. 246 (Regulatory Integrative Comp. Physiol. 15), R133-R151. Cohen, C. M., Foley, S. F., & Korsgren, C. (1982) Nature 299, 648-650. Corces, V. 6., Salas, J., Salas, M. L., Avila, J. (1978) Eur. J. Biochem. gt, 473-479. Craig, S. W. & Pollard, T. D. (1982) TIBS 1, 88-92. Crossin, K. C. & Carney, D. H. (1981) Cell gg, 61-71. David-Pfeuty, T. & Singer, S. J. (1980) Proc. Natl. Sci. Acad. Sci. USA 11, 6687-6691. Davies, W. A. & Stossel, T. P. (1977) J. Cell Biol. 1;, 941-955. Edelman, 6. M. (1976) Science 192, 218-226. Edelman, 6. M., Yahara, I., and Wang, J. L. (1973) Proc. Natl. Acad. Sci. USA 19, 1442-1446. 41 Edelman, G. M. & Yahara, I. (1976) Proc. Natl. Acad. Sci. USA 1g. 2047-2051. Eldridge, C. A., Elson, E. L., & Webb, W. W. (1980) Biochemistry 19, 2075-2079. Elson, E.L. and Reidler, J.A. (1979) J. Supramol. Struct. 12. 481-489. Elson, E. L. & Schlessinger, J. (1980) In The Neurosciences: Fourth Study Program (Schmitt, F. O. & Worden, F. 6., eds.) pp 691-701. MIT Press, Cambridge, Mass. Englander, L. L. & Rubin, L. L. (1987) J. Cell Biol. 1 4, 87-95. Estensen, R. D. & Plagemann, P. G. W. (1972) Proc. Natl. Acad. Sci. USA g2, 1430-1434. Farmer, S., Ben Ze'ev, A., Benecke, B.-J., & Penman, S. (1978) Cell tg. 627-637. Flanagan, J. & Koch, G. L. E. (1978) Nature g_g, 278-281. Folkman, J. H. & Moscona, A. (1978) Nature ggg, 345-349. Franke, W. W. (1971) Protoplasma 1g, 236-292. Freedman, V. H. & Shin, S. (1874) Cell g, 355-359. Frye, L. D. & Edidin, M. (1970) J. Cell Sci. 1, 319-335. Gabbiani, 6., Chaponnier, C., Zumbe, A., & Vassalli, P. (1977) Nature 269, 697-698. Geiger, B., Avnur, 2., Rinnerthaler, 6., Hinssen, H., & Small, V. J. (1984) J. Cell Biol. 9, 83s-91s. Glacy, S. D. (1983) J. Cell Biol. 21. 1207-1213. Golan, D. E., Alecio, M. R., Veatch, W. R., & Rando, R. R. (1984) Biochem. gg. 332-339. Goldman, R. D. (1971) J. Cell Biol. gt, 752-762. Goldman, R. D. & Knipe, D. M. (1972) Cold Spring Harbor Symp. Quant. Biol. g1, 523-534. Goldman, R. D., Lazarides, E., Pollack, R., & Weber, K. (1975) Exp. Cell Res. g9, 333-344. Henderson, D. & Weber, K. (1979) Expt. Cell Res. 124, 301- 42 316. Henis, Y. I. & Elson, E. L. (1981) Proc. Natl. Acad. Sci. USA gt, 1072—1076. Hoffstein, S., Goldstein, I. M., & Weissman, 6. (1977) J. Cell Biol. 73, 242-256. Holley, R. W., (1975) Nature 258, 487—490. Holley, R. W. & Kiernan, J. A. (1968) Proc. Natl. Acad. Sci. USA t9, 300-304. Hynes, R. O. & Destree, A. T. (1978) Cell tt, 875-886. Ingber, D. E. and Jamieson, J. D. (1985) In Gene Expression tggiggiNormal and Mattgnant Differentiation (Andersson, L. C., Gahmberg, C. 6., and Ekblom, P., eds.), Academic Press, New York. Joshi, H. C., Chu, P., Buxbaum, R. E., & Heidemann, S. (1985) J. Cell Biol. 101, 697-705. Koch, G. L. E. & Smith, M. J. (1978) Nature 273, 274-277. Koppel, D. E. (1979) Biophys. J. gt, 281-292. Koppel, D. E. (1981) J. Supramol. Struct. and Cell. Biochem. t1, 61-67. ., Sheetz, M. P., & Schindler, M. (1980) . t9, 187-192. Koppel, D. Biophys. E J Koppel, D. E., Sheetz, M. P., & Schindler, M. (1981) Proc. Natl. Acad Sci. USA gt, 3576-3580. Landreth, 6. E., Williams, L. K., & Rieser, G. D. (1985) J. Cell Biol. 101, 1341-1350. Lazarides, E. (1975) J. Histochem. & Cytochem. gt, 507-528. Lazarides, E. (1980) Nature g_t, 249-256. Lazarides, E. (1982) Ann. Rev. Biochem. tt, 219-250. Lehto, V.—P. (1983) Exp. Cell Res. ttt, 271-286. Maness, P. F. & Walsh, R. C. (1982) Cell tt, 253-262. Margolis, R. L. & Wilson, L. (1978) Cell 13, 1-8. Maro, B., Johnson, M. H., Webb, M., & Flach, 6. (1986) J. Embryol. Exp. Morph. 22. 11-32. 43 MacPherson, I. & Montagnier, L. (1964) Virology gt, 291-294. McCloskey, M. & Poo, M.-M. (1984) Int. Rev. Cytol. £1: 19- 81. McNutt, N. S., Culp, L. A., & Black, P. H. (1973) J. Cell Biol. §§. 412-428. Metcalf, T. N., Wang, J. L., & Schindler, M. (1986) Proc. Natl. Acad. Sci. USA tt, 95-99. Michaels, R. L., Pasternak, C., Young, J.-L., Elson, E. L., & Heuser, J. E. (1984) J. Cell Biol. tt, 32a. Miranda, A. F., Godman, 6. C., & Tanenbaum, S. W. (1974) J. Cell Biol. 62, 406-423. Moon, R. T., Ngai, J., Wold, B. J., & Lazarides, E. (1985) J. Cell Biol. 100, 152-160. Murray, J. M. (1983) J. Cell Biol. 2Q, 283-295. Osborn, M. & Weber, K. (1975) Proc. Natl. Acad. Sci. USA 1;: 867‘871. Otsuka, H. & Moskowitz, M. (1975) J. Cell Physiol. gt, 213-220. Otteskog, P., Ege, T., & Sundqvist, K.-G. (1981) Exp. Cell Res. 136, 203-213. Otto, A. M., Zumbe, A., Gibson, L., Kubler, A.-M., & Jimenez de Asua, L. (1979) Proc. Natl. Acad. Sci. USA 1t, 6435-6438. Packard, B. (1986) TIBS tt, 154. Painter, R. G. & Ginsberg, M. (1982) J. Cell Biol. tg, 565-573. Penman, S., Fulton, A., Capco, D., Ben Ze'ev, A., Wittelsberger, S., and Tse, C. F. (1981) Symp. Quant. Biol. gt, 1013—1028. Pollack, R., Osborn, M., & Weber, K. (1975) Proc. Natl. Acad. Sci. USA 1g, 994-998. Pollard, T. D. & Craig, S. W. (1982) TIBS 1, 55-58. Poo, M. M. & Cone, R. A. (1974) Nature 247, 438-441. Reaven, E. & Azhar, S. (1981) J. Cell Biol. tt, 300-308. 44 Rogalski, A. A. & Singer, S. J. (1985) J. Cell Biol. 1 1, 785-801. Roos, E., Spiele, H., Feltkamp, C. A., Huisman, H., Wiegart, F. A. C., Traas, J., & Mesland, D. A. M. (1985) J. Cell Biol. 101, 1817-1825. Saffman, P. 6. & Delbruck, M. (1975) Proc. Natl. Acad. Sci. USA 1g. 3111-3113. Scherson, T., Kreis, T. E., Schlessinger, J., Littauer, U. 2., Borisey, G. 6., & Geiger, B. (1984) J. Cell Biol. gg. 425-434. Schindler, M. & Jiang, L.-W. (1986) J. Cell Biol. 1 2, 859-862. Schindler, M. & Jiang, L.-W. in Comments on Motecular and Cellular Btophystg (in press) Schlessinger, J. (1980) TIBS t, 210-214. Schlessinger, J., Barak, L. S., Hammes, 6. 6., Yamada, K. M., Pastan, I.. Webb, W. W., & Elson, E. L. (1977) Proc. Natl. Acad. Sci. USA gt, 2909-2913. Schubert, D. & Lacorbiere, M. (1976) Nature 264, 266-267. Sheetz, M. P. & Koppel, D. E. (1979) Proc. Natl. Acad. Sci. USA gt, 3314-3317. Sheterline, P. & Hopkins, C. R. (1981) J. Cell Biol. gt, 743-754. Singer, S. J. & Nicolson, 6. L. (1972) Science 175, 720-731. Solomon, F. (1980) Cell g2. 331-332. Sugrue, S. P. and Hay, E. D. (1981) J. Cell Biol. gt. 45-54. Steck, T. (1974) J. Cell Biol. tg, 1-19. Steinart, P. M., Jones, J. C. R., & Goldman, R. D. (1984) J. Cell Biol. gg, 22s-27s. Stoker, M., O'Neill, C., Berryman, S., & Waxman, V. (1968) Int. J. Cancer t, 683-693. Stossel, T. P., Hartwig, J. H., & Yin, H. L. (1981) in Cytosketgtal Etements and Plasma Membrane Organization (Poste, G. E. & Nicolson, 6. L., Eds.) pp 139-168, Elsevier/North-Holland Biomedical Press, New York. 45 Streuli, C. H., Patel, B., & Critchley, D. R. (1981) J. Exp. Cell Res. 136, 247-254. Tank, D. W., Wu, E.-S., & Webb, W. W. (1982) J. Cell Biol. a: 207-212. Tomasek, J. J. & Hay, a. D. (1984) gg, 536-549. Vale, R. D., Reese, T. S., & Sheetz, M. P. (1985) Cell tg, 39-50. Virtanen, I., Vartio, T., Badley, R. A., & Lehto, V.-P. (1980) Nature 298, 660-663. Webb, W.W., Barak, L.S., Tank, D.W., and Wu, E.-S. (1981) Biochem. Soc. Symp. tt, 191-205. Wegner, A. (1976) J. Mem. Biol. 1 8, 139-150. Weingarten, M. D., Lockwood, A. H., Hwo, S.-Y., & Kirschner, M. W. (1975) Proc. Natl. Acad. Sci. USA 1g, 1858-1862. Wier, M.J. and Edidin, M. (1986) J. Cell Biol. 1 3, 215-222. Wolf, D. E., Edidin, M., & Dragsten, P. R. (1980) Proc. Natl. Acad. Sci. USA 11. 2043-2045. Yamada, K. M. & Olden, K. (1978) Nature 275, 179-184. CHAPTER III PROTEIN LATERAL DIFFUSION IN SPREAD AND SPHERICAL FIBROBLASTS 46 INTRODUCTION The involvement of cell shape as a controlling influence for DNA synthesis and growth in anchorage dependent cells was originally demonstrated by the work of Folkman and Moscona (1978). Further efforts showed that while mRNA production, rRNA and DNA synthesis, and cell growth require an extensively spread cell, protein synthesis occurs upon cell attachment (Benecke gt gt., 1978; Ben Ze'ev gt gt., 1980). These results and other work (Otsuka & Moskowitz, 1975; MacPherson & Montagnier, 1964) suggest that the spherical to spread transition of anchorage dependent cells which follows cell attachment to growth surfaces, influences the subsequent biological activity of these cells. The changes in cell architecture associated with the spherical to flat structural transition have been correlated with the reassembly of stress fibers, microfilaments, and microtubules (Badley _t _t., 1980). In light of the importance of these cytoskeletal structures for protein biosynthesis (Penman gt gt., 1981; Cervera gt gt., 1981), and transmembrane signal aquisition (Geiger, 1983; Landreth gt gt., 1985) and transmission (Ingber & Jamieson, 1985), the possibility that the flat to spherical 47 48 shape transition interrupts a cytoskeletal communication pathway from the cell surface to the nucleus has been suggested (Ben Ze'ev, 1985). Since the lateral mobility and topology of membrane receptors has been demonstrated to be modulated by the cell cytoskeleton with consequences for transmembrane signaling (Edelman, 1976; Koppel t l., 1981; Tank _t _t., 1982), we have attempted to determine the role of shape induced cytoskeletal changes in the lateral mobility of BALB/c 3T3 fibroblast plasma membrane phospholipids and glycoprotein receptors. In an effort to examine whether shape changes can play a role in uncoupling receptor mediated signaling, changes in receptor lateral mobility are monitored to serve event. Transformed cells, then, represent a system in which the normal communication pathway is bypassed and this is reflected in the cells ability to grow in suspension. The experiments to be discussed were performed on spherical and spread untransformed, and Kirsten murine sarcoma virus transformed (K-MSV) BALB/c 3T3 fibroblasts. The major result of this work is that the apparent diffusion coefficient for both spherical untransformed and transformed cells was 12-fold greater than for spread cells. This suggests that in adhering, flat cells the membrane receptors were associated more closely with assembled cytoskeletal elements which may serve to transmit transmembrane signals from the cell membrane to the interior, particularly the nucleus. MATERIALS AND METHODS Celt Culture BALB/c A31 3T3 (3T3) and Kirsten mouse sarcoma virus transformed (K-MSV) BALB/c 3T3 fibroblasts were obtained from American Type Culture Collection and cultured in Dulbecco's modified Eagle's medium (DME) plus 10% calf serum. Spread cells were plated on tissue culture dishes on the day before use. Spherical cells were prepared by trypsinization and suspension in DME/10% calf serum containing methyl cellulose (MC, 1.5% final concentration) for one hour, one day, or two days. DME/MC was prepared in a fashion similar to that described elsewhere (Rheinwald and Green, 1974). Briefly, methyl cellulose was autoclaved and dissolved in sterile water at 60’C to a concentration of 2.3% and stirred at room temperature for 20 minutes. The methyl cellulose was diluted to 1.6% by the addition of 2x concentrated DME and calf serum (10% final concentration). The DME/MC was stirred overnight at 4°C before use. Spherical cells were also prepared for some studies by EDTA removal and suspension in methyl cellulose for one hour. Fluorescence Redistributtgn After Photobtgaching (FRAP) The instrument used was similar to that described by Koppel (1979). Diffusion measurements on spread cells were obtained using a point bleach and multipoint analysis 49 50 described elsewhere (Koppel, 1979) and reviewed in the appendix A. Lateral diffusion was measured on spherical cells using an edge bleach and normal mode analysis as presented previously (Koppel gt gt., 1980). Protein mobility was monitored using rhodamine-labeled wheat germ agglutinin (WGA) (Vector, Burlingame, CA). Lipid mobility was monitored using NBD-phosphatidylcholine. ttuorescence Microscopy Actin filaments in cells were labeled with rhodamine-phalloidin (Molecular Probes, Junction City, OR) and fixed using the one step method described in the package insert. The distribution of WGA receptors was examined on cells labeled with 100 ug/ml rhodamine-WGA and fixed with 3.7% formaldehyde. Control samples contained 0.2 M N-acetylglucosamine in the incubation and wash buffers. Fibronectin patterns were determined by fixing the cells with 3.7% formaldehyde for 30 minutes at room temperature, washing the cells with phosphate buffered saline (PBS) for 20 minutes at room temperature, and labeling the cells with anti-fibronectin antibodies (1:100 dilution). Control samples were incubated with normal rabbit serum. Samples were washed with PBS and incubated with rhodamine-labeled goat anti-rabbit antibodies (1:30 dilution) for 30 minutes at 37’C. Antibodies were obtained from Miles Scientific (Naperville, IL). Samples were washed 3 times with PBS/5% goat serum. All samples were mounted in PBS/90% glycerol 51 (v/v)/5% N-propylgallate (w/v). Cell Population Htstggrams Cells, labeled with fluorescein-WGA and fixed, were scanned with a laser beam in 2 dimensions using the ACAS 470 Fluoresence Workstation (Meridian Instruments, Okemos, MI). The resulting fluorescence was collected, measured by a photomultiplier tube, and digitized to produce 2-dimensional images of the labeled cells. Cell population histograms were obtained through computer analysis of the digital images. WGAgttnding The amount of WGA bound to the cells was estimated using fluorescein-WGA. Various concentrations of FITC-WGA were incubated with the cells for 30 minutes at 4°C. After washing, the cells were lysed in HEPES (12.5 mM) buffered Hank's Balanced Salt Solution (HBSS/HEPES) containing 1% sodium dodecyl sulfate (SDS). The fluorescence was determined in a Perkin-Elmer spectrofluorimeter using an excitation wavelength of 460 nm, an emission wavelength of 520 nm, and 5 nm slit widths. The amount of fluorescein-WGA bound to the cells was estimated by comparison to a standard curve of FITC-WGA in HBSS/HEPES/1% SDS and was normalized to the amount of protein. The amount of protein was determined using a modified Lowry procedure (Markwell t l., 1978). The standard curve could also be prepared by 52 incrementally adding FITC-WGA to previously measured cell samples with no noticeable effect which suggested that the fluorescence was not absorbed by the sample. gggpggtson qttWGA Receptors Equal number of cells were lysed in 1% SDS/1 mM PMSF. Equal amounts were loaded and separated on duplicate 10% polyacrylamide gels. Proteins were transferred onto nitrocellulose sheets by the procedure of Towbin et a1. (1979) for 3-4 hours at 400 mA. The blots were incubated overnight in PBS/2%PVP (phosphate buffered saline, pH 7.4, polyvinylpyrrolidone) as suggested by Bartles and Hubbard (1984) to block excess protein binding sites. The blots were incubated for 3-5 hours at 4°C, shaking with 2 ug/ml WGA-horse radish peroxidase in PBS/PVP. The blots were washed 5 times for 5 minutes with PBS/PVP and rinsed briefly with Tris buffered saline. Buffers used for replicate blots included 0.2 M N-acetylglucosamine plus or minus 0.2 M AMP as controls. Color development was carried out as described in the Bio-Rad procedure using Biorad color development reagent. Charactettzation of Glycolipid Receptors for WGA Spherical cells were prepared by overnight suspension in DME/MC. Immediately before use, spread cells were removed from their substrate and counted. Spherical cells were removed from the methyl cellulose and counted. 5 x 106 53 cells were placed into centrifuge tubes and centrifuged for 5 minutes at low speed to pellet the cells. The cells were extracted as described by Folch gt gt. (1957). Briefly, 3 mls of methanol were added to the cell pellets and mixed extensively. 6 mls of chloroform were added to the samples and again mixed extensively. To the methanol:chloroform mixture was added 0.5 mls of 4 M potassium chloride and 2.5 mls of water followed by mixing. Phases were separated by low speed centrifugation. The chloroform phase was removed to a tared tube. The chloroform extraction was repeated with a second 6 mls and added to the first 6 mls. The organic phases were evaporated under nitrogen, redissolved in 2:1 chloroform:methanol, and stored at -20°C until use in sealed ampules under nitrogen. The aqueous phase was lyophilized in a tared tube and redissolved in 2:1 chloroform:methanol at time of use. Glycolipids were separated by thin-layer chromatography (TLC) using 5:4:1 chloroform:methanol:water. WGA receptors were probed with radioiodinated WGA as described previously (Smith, 1983). WGA binding could be inhibited by addition of 0.2 M N-acetylglucosamine. WGA was labeled using chloramine T as described elsewhere (Greenwood gt gt., 1963). 54 Characterization of Cgtt Growth Thygidine uptake: Spread cells were plated on 12 well plates (4 x 105 cells/well). Spherical cells were grown in 150 cm2 flasks until 0 hour. To synchronize the cells, all cells were blocked twice with 3 mM thymidine as described elsewhere (Adams, 1980). Cells were released by washing and adding fresh media. Cells to be made spherical were removed with trypsin and placed in DME/MC; this was considered 0 hour. At 1, 5, 9, 13, 17, 21, 25, and 33 hours, samples were removed and incubated for one hour with 10 uCi 3H-thymidine/sample (spherical cells were removed from DME/MC before labeling). Cells were washed twice with PBS, lysed in 0.5 ml 0.05 N NaOH/1% SDS, and counted by liquid scintillation methodology. Glucose utilization: The utilization of glucose was determined using a novel technique described by Pollard gt gt. (1981). The method uses the exchange of tritiated hydrogen with water from the 2 position of D-glyceraldehyde-3-phosphate (originally the 5 position of glucose) during its isomerization to dihydroxyacetone, catalyzed by triosephosphateisomerase. In this way, tritiated water is essentially the only radioactive product of the metabolism of D-[5-3Hngucose. The tritium is extracted from the water 55 into an organic fraction containing scintillants. The tritium from the water exchanges with hydrogen on isoamyl alcohol in the scintillation fluid, while the unmetabolized radioactive glucose remains in the aqueous fraction and immiscible with the scintillation fluid. Briefly, the cells were incubated with various concentrations of glucose spiked with 1.5 pCi/ml D-[5-3H1glucose in glucose-free HBSS (1 ml/sample). At 0 hour and at 4 hours triplicate 50 p1 aliquots were removed to scintillation vials containing 200 pl of Benedict's reagent (43.25 g sodium citrate mixed with 25.0 g anhydrous sodium carbonate in 200 ml of water; add 4.325 g cupric sulfate in 25 ml water and adjust volume to 250 ml with water). 80 pl of 1 M carrier glucose was added and samples were boiled for 30 minutes in a water bath. After cooling for 10 minutes at room temperature, 5 ml of scintillation fluid (750 ml toluene, 250 ml isoamyl alcohol, 18.95 g PPO (2,5-diphenyloxazole), and 1.14 g POPOP (p-bis[2-(5-phenyloxazoyl)Jbenzene) was added and mixed. Phases were allowed to separate for at least 5 minutes before scintillation counting. Aliquots were also analyzed for protein concentration by the method of Markwell gt gt. (1978). geterminatton of ATP tevels The amount of ATP in spread untransformed and transformed fibroblasts was determined using an enzyme assay with Luciferin/Luciferase (Sigma, St. Louis) based on 56 that described elsewhere (package insert and Webster gt gt., 1981). Cells were plated on 60 mm dishes at 3 x 105 cells/dish. Cells were removed with EDTA, pelleted, suspended in 300 pl PBS, and injected into 700 pl boiling water for 15 minutes to prevent degradation of ATP and cooled on ice. To 200 pl sample was added 75 pl Luciferin/Luciferase (40 mg/ml) and buffer (0.025 M Tricine, 5 mM MgSO4, 0.5 mM EDTA, 0.5 mM dithiothreitol, pH 7.8) to 1 ml. After rapid mixing, the amount of light emitted over 1 minute was recorded on a chart recorder with a spectrofluorimeter set at emission wavelength 555 nm and 20 nm slit width. Extrapolation to 0 time allowed estimation of the initial emission. Comparison of light emitted from samples to that from known concentrations of ATP allowed estimation of ATP levels. Protein concentration was determined from aliquots using a modified Lowry procedure described by Markwell gt gt. (1978). Scanning_§lectron Microscopy Spread cells were fixed with 4% glutaraldehyde for two hours, post-fixed with 0.5% osmium tetraoxide for 15 minutes, dehydrated through a series of ethanol washes, and critical point dried. Spherical cells were removed from DME/MC, fixed and post-fixed as for spread cells. Fixed spherical cells were attached to SEM grids or disks with poly-L-lysine, dehydrated through a series of ethanol washes, and critical point dried. SEM was performed on a JEOL 350 at 15 kV. 57 RESULTS Consequence of Shape and Transformation on the Number of WGA Recgptors in 3T3 Fibrobtasts Glycoprotein receptors for WGA served as a general probe to monitor lateral mobility of plasma membrane proteins in both untransformed and transformed cells. To assure appropriate comparisons between different cell shapes and transformation states, a series of investigations were directed to characterizing the number and type of WGA receptors under all measuring conditions. These experiments were performed both in bulk and on a single cell basis. It was of particular importance to insure that the techniques required to remove cells for spherical analysis did not greatly alter the number or type of WGA receptors on the cell surface. In this context, we determined the binding of WGA to spread cells, EDTA removed cells, and trypsin removed cells. The binding to trypsin removed cells was determined after one hour, one day, and two days in suspension. The removal of 3T3 cells (untransformed) with EDTA or trypsin had little effect on the number of WGA receptors (see figure 1A, differences were within standard deviations). As seen in figure 1A, cells suspended in DME/MC for one or two days also bound an amount of WGA similar to that of the spread cells. The binding of WGA to transformed K-MSV 3T3 cells appeared to 58 59 .maamo Hmoaumcam mnu ca umHHEam mm: gowns maaou vacuum man you coauma>mc vumccwum on» muocov mums nouns one .vchEuoumv omam mma Alv use: oco no“ popcoamam vcm osou coma can umcu mHHmo Hmofiumzdm ou mcuvcan «o3 .2uv mama oau vcm .AOV amp oco .A4V use: oco LOM mmoasaamo assume cw coamcmamam 6cm coaumuficwmdzuu umuum cocfiEumuoc was maaoo Hmuapondm ou weapons use .AOV >3 vmuocmv ma mHHmu emmudm cu <03 mo wchCAB one . um emcfisuoumv was mumum Hmuauocdm cam vacuum ago as Amv mummanounfiu mam noEHOmemuu >mznx new A um vaca8umumc mma oumum Hmofiumcdm cam vacuum ecu ca mummanounfim mam coauommcmuu >mzux pom cmEuowmcmuuca ou <03 woumcacowuoavmu mo weavcan och .N muswam :E\01V<03 JON— 00¢ com com 00. o u oo.~ 00.? 0.06 (1-0001139 / ONflOG SB'lflOS'lOW 62 .c3ch 1w m_ioc new amanoumcmau >mzax Aav Hmofiumsdm vcm on ammcam mm Haw: mm magma mem Amv Hmuwpozdw can A$OAOposuoE wzawoafl acaflxfiz an aUCHEumuwa mmB pmccmow Have some EoLe oucwumwuosau mo ucaosm ash .Hamo ecu An assoc <53|o+Hm Lo Lzzoem mzu cu cwumfiop ma sown: Haoo uma mucoummLCJAL Lo ucsoEm ecu do mammamcm coHumH2a3; .m unsuL; _.ou coa oucuumoco:_u _oaok __00 can aucoomoco:_u _o.oh o. n o o_ m o o . o m m m w m m... m 0 t 0 m m N 8 w. o 8. 0 3 __mo .oa mocoummco:_u _o.0h ._oo con oucoumoco=_u _o.op o_ n o o. n o o N m a w. 0. w 9 q .. n m... 9 o s. m a n n s - M m ON. < 0n 7 a a .A u ' ~ . .- Figure 4. Fluorescence micrographs of spread and spherical 3T3 fibroblasts labeled with FITC-WGA. (A) spread BALE/c 3T3 cells; (B) spherical BALB/c 3T3 cells; (C) spread K-MSV BALB/c 3T3 cells; (D) spherical K-MSV BALE/c 3T3 cells 64 homogeneously distributed over the cell surface in all cell types. A comparison of the WGA receptors from whole cell lysates from spread and spherical cells shown in Figure 5 examined the effect of the removal and suspension process on the WGA binding proteins. The majority of the specific WGA binding proteins appeared to be in the upper half of the Western transfer blot since many of those in the bottom half were not removed by N-acetylglucosamine (the hapten inhibitor for WGA) as shown in lane I. The nonspecific binding appears to be a charge dependent phenomenon that can be blocked by charged species such as AMP (this lab, unpublished observations) as shown in lane J. In general the major bands on the transfer blot specific for WGA appeared similar regardless of cell shape, length of time in suspension, or transformation. One of the major glycoproteins on murine fibroblasts is an 80,000 dalton WGA binding glycoprotein (Jacobson gt gt., 1984). A band in the 80,000 dalton region in Figure 5 is one of the WGA binding proteins that appeared to be unaffected by suspension or transformation. Since FRAP measures an ensemble diffusion coefficient it would require major differences in the WGA surface receptors to account for the increase in protein lateral mobility and such differences were not apparent in the WGA receptor number or whole cell WGA binding glycoproteins. 65 200K> 116K» 92.5K» ‘. Hill} "I Figure 5. Western transfer blot of WGA glycoprotein receptors from spread and spherical 3T3 fibroblasts. WGA receptors from spread BALB/c and K—MSV BALB/c 3T3 cells are shown in lanes A and B, respectively; from spherical 3T3 cells suspended for 1 hour in methyl cellulose are shown in lanes B (BALB/c) and F (K—MSV BALB/c); 3T3 cells suspended for 1 day are shown in lanes C (BALE/c) and G (K-MSV BALE/c); 3T3 cells suspended for 2 days are shown in lanes D (BALE/c) and H (K—MSV BALB/c); CODCrols for blots incubated with 0.2 M N—acetylglucosamine (I) and 0.2 M N—acetylglucosamine plus 0.2 M AMI‘ (.1). Molecular weight markers are as indicated. 66 The possibility of WGA binding to glycolipids was also considered. The autoradiogram from thin-layer 125I-WGA is presented in Figure chromatographs probed with 6. The most apparent bands on the autoradiogram were found in both the aqueous and organic fractions and are most likely protein contaminants. There were only faint bands observed in the organic fraction which had a relative mobility similar to that of the ganglioside standards (which bound WGA and acted as a positive control) despite the fact that five times as much starting material was used for these TLC plates as for the western transfer blots shown in Figure 5. These bands appeared in approximately equal amounts in 3T3 and K-MSV 3T3 cells, regardless of cell shape. Of particular importance for the interpretation of our results was that the amount of glycolipid was considerably less than that of the protein contaminants seen in both the organic and aqueous fractions. Furthermore, in the aqueous fractions no bands were detected in the region of the gangliosides standards indicating all the detectable glycolipid receptors had been extracted with the organic phase. Thus, it seems unlikely that the binding of WGA to glycolipids is significant in our diffusion measurements. These results suggest that WGA binding glycoproteins are sufficiently similar in amount and type under the differing shape and transformation conditions so that they may be used as comparative probes of lateral mobility. 67 Figure 6. Autoradiogram of glycolipids separated by TLC and probed with radio-iodinated WGA. WGA glycolipid receptors from spread BALB/c and IC—MSV BALB/c cells are shown in lanes A and C, respectively; receptors fronlspherical BALB/c and K-MSV BALE/c cells are shown in lanes B and D. respectively. 68 ggtgrgl Mobittty tn 3T3 Fibroblgsts Lateral mobility of lipids and WGA receptors was measured in spread and spherical untransformed and K-MSV transformed 3T3 fibroblasts. Table I is a compilation of those results. Using NBD-phosphatidylcholine as a probe for phospholipid mobility, no major differences could be observed as a function of cell shape or transformation. The values observed for NBD-phosphatidylcholine in the 9 untransformed (about 2 x 10- cm2/s)and transformed (about 4 x 10.9 cmz/s) 3T3 cells were similar to those reported 9 cmz/s by Tank et 1. (1982) and 9 for FDB cells (1 x 10- various transformed cells (about 4 x 10- cmz/s) reported by Aroeti & Henis (1986). The factor of two observed between spread cells and spherical cells for untransformed and transformed cells may reflect the kind of cell surface irregularities such as small microvillar structures that were suggested by Aizenbud and Gershon (1982) to be able to affect diffusion measurements by no more than a factor of two. Scanning electron micrographs of the fibroblasts demonstrated the differences in the amount of microvilli (Figure 7). In contrast, WGA receptor mobility appeared to be markedly affected by the alteration of cell shape. WGA binds to glycoproteins containing N-acetylglucosamine and sialic acid residues and so binds to an ensemble of proteins in the membrane. The diffusion coefficient measured by FRAP will therefore be an average 69 .aam>Auuoamou .a cam m as :Bocm one mummapounaw mam vuauoumcmuu >mz|x can muuwanoppaN mam vaHOmemuucz Hmuauwcam .xao>auuuammu .0 use < ca caonm one mummanounau men umEuOumcauu vac quuowmcmuucs vmmuam .mumMHBOHBAM mum HmUprcam can ammuam mo mcamumouuae couuumao wcficcmum .5 muswfim meow: :imfi of us di ti. se me di he CC Cll N 1'} I“: 3! re 70 of the diffusion coefficients of all the WGA receptors. The use of the term diffusion coefficient refers to an apparent diffusion coefficient measured in a biological system and therefore is not a constant in the classical physical sense. It is, however, a useful measure of the mobility of membrane components and I will continue to use the term diffusion coefficient for the measured lateral diffusion as has been commonly used in the literature for FRAP studies. The measured WGA receptor lateral diffusion coefficient for spread 3T3 cells was slow at 0.33 x 10—10 cmz/s (Table I). The restriction of WGA receptor lateral diffusion is obvious when the diffusion coefficient for WGA receptors in spread 3T3 cells is compared to the diffusion coefficient for lipids in the same cells. WGA receptor mobility was 50- to 100-fold slower than lipid mobility in spread 3T3 cells. Alteration of cell shape by suspension in DME/MC for one or two days was accompanied by a dramatic increase in WGA receptor lateral mobility. The lateral diffusion coefficient for WGA receptors was about 12-fold faster in spherical 3T3 cells than in spread 3T3 cells. 10 cmz/s) for Although the diffusion coefficient (3.9 x 10- spherical 3T3 cells was an order of magnitude faster than in spread cells, it was still about 10-fold slower than lipid diffusion coefficients. The restriction of WGA receptor diffusion appeared to have been partially eased in spherical cells, yet some shape-independent means of restricting protein lateral mobility remained. mucmsfinmmxm no nmnEDZO u>nm>oomn scooped may me man um\wao aw Aofisonxv unmfiowwmmoo cofimSMMAU may ma mm 71 mHva m.~Mm.m Away sawmm o~.oMm~.o <63 e-mem auHms a.~+m.m Emu. cmwvm mm.o+mm.o <02 mamsm sawmm nswoo “was mwos mawnm omcmz elmam m~+ov m~+mm Essa m+sm Ha+ma omomz mumsm ma 0 n as m o mmomm mass game maamo aaonmmam maamo cammmm VHfiQHmOZ Anm>oomn unmonmm may ma mxo um\m50 as Aofi-ofixi bcmaosmceoo cofimsccwe may ma am Ame Amwas mfi.oflmm.o maamo nammmm zo oz\chmz isms smwns H.Nwm.m oz\ozmm szawms class s~+ms m.m+m.m om>ozmm aaom n ma m o szmzsamms ZOHmzmmmDm mom muqmo mhm Oszdmmmm m0 BOthm .N mqmdfi 75 200. I50. CPM (MO‘H IOO _ HOURS AF TER RELEASE 200; ISO, CPM Duo”) 5 o 50. 0 I 5 9 l3 l7 2| 25 29 33 HOURS AFTER RELEASE Figure 8. Uptake of 3H-thymidine into 3T3 (A) and K-MSV transformed 3T3 (B) fibroblasts. Uptake was measured in spread (solid line) and spherical (dashed line) cells. Horizontal bars mark the uptake during the one hour periods in which cells were pulsed with labeled thymidine and are the average of duplicate samples. 76 suspension, whereas untransformed fibroblasts do not (MacPherson & Montagnier, 1964; Stoker gt gt., 1968; Freedman & Shin, 1974; Otsuka & Moskowitz, 1975). As expected, spherical K-MSV 3T3 cells continued through the cell cycle and showed a second peak of thymidine uptake at 17 hours, similar to that of the spread cells. The untransformed 3T3 cells, blocked at the 61/5 interphase and suspended in DME/MC, took up thymidine at the first time point 1 hour later but did not take up thymidine at 20 hours, apparently remaining in 6 phase (Otsuka & l Moskowitz, 1975). These results demonstrated that the transformed cells were proliferating in suspension at the time FRAP experiments were carried out; whereas the untransformed 3T3 cells in suspension were not. The consequences of cell shape alteration on metabolic processes, such as RNA, DNA, and protein synthesis have been investigated as detailed above; however, little work has been done on the role of cell shape in carbohydrate metabolism. One study reported that suspension of chick embryo fibroblasts resulted in a decrease in hexose transport (Bissell gt gt., 1977). Hexose transport has also been reported to be enhanced by transformation (Eliam & Vinkler, 1976; Bissell gt gt., 1977); however, other studies suggested that the apparent enhancement in glucose transport by transformation is a reflection of the increased phosphorylation of glucose and that it is not a specific function of the viral genome, rather it reflects 77 the overall growth rate of the transformed cells which are not affected by cell density or suspension (Romano & Colby, 1973; Romano, 1976). Furthermore, the rate of glucose uptake is not a primary determinant of cell growth rate under the usual conditions of cell culture (Romano & Connell, 1982). Instead of monitoring the uptake of glucose, the utilization of glucose was monitored as a function of cell shape. The method used provides an estimate of the apparent K and Vm for the overall M ax process of glucose metabolism from uptake to the isomerization of D-glyceraldehyde-S-phosphate to dihydroxyacetone phosphate in the glycolysis pathway. As shown in Table III, neither alteration of cell shape nor transformation had much effect on the metabolism of glucose in 3T3 fibroblasts under conditions where neither cell density nor serum would limit cell growth in the untransformed cells. Since glucose is not the only metabolic means of producing energy for cellular use, the level of ATP in untransformed and transformed 3T3 fibroblasts was determined. Using a Luciferin/Luciferase assay the amount of ATP was found to be similar in EDTA removed untransformed 3T3 (9.77:2.90 pg ATP/mg protein) and K-MSV transformed 3T3 (7.02:1.30 pg ATP/mg protein) cells. According to these assays, then, transformation does not affect glucose metabolism or the levels of ATP significantly in these cells. 78 mucmfiwnwmxm Mo amnesm 0 cae\msamooofl\floee .5 saw “as 6:0a x o.oMm a s.fim~.u men >m23e aaonmmmm Ame ones x o.on.fi p.0Hm.fi mew >m21e cammmm Ame muofi x ~.oHfl.H m.omm.H mam qaonmmmm via. one” x m.o+n.o m.o+H.H men oammmm XME E n > w e moms Hamo mam¢amommHh mam 2H ZOthNHQHBD mmoosnw .0 wqm¢9 79 Actin and Fibronectin Distribution as a Function of Cell tggpe and Transformation The state of the actin stress fiber system was monitored using rhodamine-labeled phalloidin, an actin specific probe. Alteration of cell shape in 3T3 cells was accompanied by a disappearance of stress fibers as reported previously (Badley gt gt., 1980) (see Figures 9A and QB). Transformation was also associated with microfilament reorganization (see Figures 9D and 9E) as reported previously (McNutt gt gt., 1973; Pollack gt gt., 1975; Carley gt gt., 1981). Stress fibers were thick and numerous in the spread untransformed 3T3 cells. The spread transformed K-MSV 3T3 cells were of two types. About 30% had a few wispy stress fibers and rosette structures resembling those described by David-Pfeuty and Singer (1980). These rosette structures appear to contain F-actin aggregates near the ventral surface (Carley _t _t., 1981); whereas the remainder had no apparent stress fibers only the rosettes. Fibronectin patterns on 3T3 fibroblasts were examined by immunofluorescence microscopy. Fibronectin is a major attachment protein in the extracellular matrix, important for cell adhesion and spreading. Transformation of fibroblasts has been demonstrated to affect fibronectin associated with the cell surface (Yamada gt gt., 1976; Ali t l., 1977; Hynes gt gt., 1978). As shown in Figure 9, there was less fibronectin associated with spread K-MSV 3T3 80 .Aao>fiuuoamou .a ace 0 as csosm cum cauumcounfiu new vwaanma mHHou 2m u\mqmzux can “33.7.5 vmmpam “33/362393 .m can a as gozm mum cfiuumlm new nmamnma mflnmu men 6:35“ >mz|x vcm 023/5 Hmofipmcam “333333 .a cam < a.“ c396 mum cauumnm now 630an mHHoo nan “(mg/E >mZIx can gag/E ammuam .moavonfiucm ufianmulflcm umom mcHEmponu :33 swamps.“ :93 can mwfipoflnucm chocouszlaucn :33 532365 mummasoz: 2m vacuum was .393 3:0me 53.7.; cm .cwgoaamxa 19:15.55.“ :33 632:3 3330.5: 2% Haste—Em _Em vmupam mo wsemumouuae mucmuwwucaam .m gnaw: 81 cells (9F) than untransformed 3T3 cells (90). These results indicate that the transformed cells differ from the untransformed cells in at least one extracellular matrix component. Despite this difference, WGA receptor lateral diffusion was similar in spread 3T3 and K-MSV 3T3 cells, suggesting that at least fibronectin of the extracellular matrix does not control WGA receptor lateral mobility under the conditions of our experiment in agreement with the results of Schlessinger gt gt. (1977a). DISCUSSION Cell shape in anchorage dependent cells has been demonstrated to have a dominant influence on nuclear events such as DNA, rRNA, and mRNA production, while protein synthesis requires cell-surface contact. In both cellular activities, nucleoplasmic and cytoplasmic, the organization of the "cytoskeletal framework" has been implicated; an extended spread morphology is necessary for nuclear activity, while a point attachment appears to be sufficient to trigger translation of mRNA attached to detergent insoluble structures (Farmer gt gt., 1978; Cervera gt gt., 1981). Considering the recent attempts to examine the organization of the cytoskeleton and its influence on the lateral mobility of plasma membrane proteins, and the rheology of the cell, we have pursued a series of investigations to examine potential links between the metabolic consequences of cell shape change and the concomitant changes in the mobility of plasma membrane glycoprotein receptors. The cytoskeleton, in this regard, may serve to integrate and transmit signals acquired at the cell surface to intracellular sites where the initiation of biosynthetic processes may require specific types of cytoskeletal interactions or organization. In this context, the major observation to be reported is that the dynamic properties of a class of plasma membrane glycoproteins is 82 83 dramatically altered in a shape transition from flat and adherent to spherical and detached (a 12-fold increase in the ensemble diffusion coefficient). Under these conditions, phospholipid mobility in 3T3 fibroblasts was not significantly affected. That these changes were a result of shape change and not significant differences in attachment to extracellular matrix may be inferred from diffusion results in adhering transformed fibroblasts which secrete considerably less fibronectin. This is noticeable in the decreased degree of spreading and in the ease of detaching the K-MSV 3T3 fibroblasts. A comparison of the lateral diffusion data for the WGA receptors in untransformed and transformed adherent, spread cells demonstrated no significant differences. This has also been observed by Eldridge gt gt. (1980) in a comparison of SV40 transformed and untransformed cells. The observed shape dependent changes in lateral mobility are reminiscent of the phenomenon of global modulation. As presented by Edelman (1976) and characterized by Schlessinger gt gt. (1977b), Henis & Elson (1981), Edelman gt gt. (1976), Michaels gt gt. (1984), and Pasternak & Elson, (1983); cells exposed to tetravalent Con A, even when only a fraction of the cell surface was bound by Con A-platelets, have a 6- to 8—fold decrease in the diffusion coefficient for a number of classes of cell surface receptors, do not form a cap, demonstrate recruitment of actin to the plasma membrane, and are 84 significantly more resistant to deformation. That these properties are related to the cytoskeletal state or reorganization is suggested by the reversal of these Con A induced changes by cytochalasins or colchicine. We believe our results provide evidence that the fibroblasts in the adherent, spread state are analogous to the Con A attached lymphocytes in that the aggregation of cell surface proteins by Con A or surface attachment initiates changes in the cytoskeleton that increase cortical cytoskeleton/membrane interactions detected by Pasternak & Elson (1983) as enhanced resistance to deformation. A cortical cytoskeleton, by adherence or Con A binding, would be under greater tension provided through recruitment and contraction of cytoskeletal elements at points of attachment. The increased interaction with the membrane and increased tension on the membrane could affect the diffusion of membrane proteins by either passively slowing down diffusion through a more resistant polymeric network or varying the on/off rates of proteins capable of associating with the cortical cytoskeleton as has been suggested previously (Koppel gt gt., 1981). Upon release from spreading and significant attachment, the microfilament, microtubule, and intermediate filament networks are affected and the diffusion of membrane receptors is increased as observed for cytochalasin/colchicine reversal of global modulation in lymphocytes. This is concomitant with a 85 cytochalasin/colchicine mediated loss of resistance to deformation reported to occur in lymphocytes (Pasternak & Elson, 1983). More support for these contentions is provided in Chapter V which examines global modulation in spread and spherical cells. Medullary gpd Corttcat Cytosketeton Although a major enhancement was observed in the diffusion coefficient of spherical cells, the rates observed were still approximately 10-fold less than reported for membrane proteins in which the membrane associated cytoskeletal network has been either removed or disorganized (Koppel gt gt., 1981; Tank gt gt., 1982). This suggests that in the spherical cells, a constraint to unhindered lateral diffusion still exists. In consideration of the global modulation results (Henis & Elson, 1981), the deformation studies which suggest biphasic characteristics of the force-compression curves for lymphocytes, and cytochalasin B sensitive resistance to compression (Petersen gt gt., 1982), it may be appropiate to suggest at least two controlling actin based networks for membrane surface receptor movement. A cytoskeleton anchored to the membrane in the manner found in mammalian erythrocytes (cortical), and a cytoskeleton that exists in the cytoplasm and may, in fact, be anchored or be organized from the nucleus (medullary). In this linked system, tension may be altered in the cortical cytoskeleton by changes in the 86 organization or structure of medullary cytoskeletal elements and vice versa. Tension itself, on the other hand, may be a local mediator or regulator of cytoskeletal reorganization as has been suggested for P012 neurites (Joshi, _t _t., 1985). These changes are induced by contact mediated recruitment of actin, myosin, and other cytoskeletal molecules to membrane/surface contact sites or by reorganization of these structural elements at the plasma membrane initiated by surface receptor aggregation as reported to occur in lymphocytes after Con A treatment (Michaels gt gt., 1984). In this manner, a bi-directional communication system may be modulated by changes at its surface which are transmitted to the cytoplasm where medullary cytoskeletal reorganizations then cause rearrangements at the cell surface, modifying the physical properties of the membrane and perhaps relaying a signal to the nucleus. REFERENCES Adams, R. L. P. (1960) In tgtl Cplture for Biochemtsts (Work, J. S. & Burdon, R. H., ed.) pp. 136-203, Elsevier/North-Holland and Biomedical Press, New York. Aizenbud, B. M. and Gershon, N. D. (1982) Biophys. J. tg, 287—293. Ali, I. U., Mautner, V., Lanza, R., & Hynes, R. 0. (1977) Cell tt. 115-126. Aroeti, B. & Henis, Y. I. (1986) Biochemistry gt, 4588-4596. Badley, R. A., Woods, A. (1980) J. Cell Sci. , Carruthers, L., & Rees, D. A. 3, 370-390. Bartles, J. R. & Hubbard, A. L. (1984) Anal. Biochem. 140, 284-292. Ben Ze'ev, A. (1985) Cell & Muscle Motility t, 23-53. Benecke, B.-J., Ben Ze'ev, A., & Penman, S. (1978) Cell tt, 931-939. Bissell, M. J., Farson, D., & Tung, A. S. C. (1977) J. Supramol. Struct. t, 1-12. David-Pfeuty, T. & Singer, S. J. (1980) Proc. Natl. Sci. Acad. Sci. USA 11, 6687-6691. Carley, W. W., Barak, L. S., & Webb, W. W. (1981) J. Cell Biol. gg, 797-802. Cervera, M., Dreyfuss, 6., & Penman, S. (1981) Cell gt, 113-120. Edelman, 6. M. (1976) Science 192, 218-226. Eldridge, C. A., Elson, E. L., & Webb, W. W. (1980) Biochemistry tg, 2075-2079. Eliam, Y. & Vinkler, C. (1976) Biochim. Biophys. Acta 4 3, 393-403. 87 88 Farmer, S., Ben Ze'ev, A., Benecke, B.-J., & Penman, S. (1978) Cell tt, 627-637. Folkman, J. H. & Moscona, A. (1978) Nature 273, 345-349. Freedman, V. H. & Shin, S. (1874) Cell t, 355-359. Geiger, B. (1983) Biochim. Biophys. Acta 737, 305-341. Greenwood, F. 0., Hunter, W. M., & Glover, J. S. (1963) Biochem. J. gg, 114-123. Henis, Y. I. & Elson, E. L. (1981) Proc. Natl. Acad. Sci. USA 1t, 1072-1076. Hynes, R. 0. & Destree, A. T. (1978) Cell tt, 875-886. Ingber, D. E. and Jamieson, J. D. (1985) In Gene Expresstgp During Normal and Malignant Differentiation (Andersson, L. C., Gahmberg, C. 6., and Ekblom, P., eds.), Academic Press, New York. Jacobson, K., O'Dell, D., & August, J. T. (1984) J. Cell Biol. gg, 1624-1633. Joshi, H. C., Chu, P., Buxbaum, R. E., & Heidemann, S. (1985) J. Cell Biol. 101, 697-705. Koppel, D. E. (1979) Biophys. J. gt, 281-292. Koppel, D. E., Sheetz, M. P., & Schindler, M. (1980) Biophys. J. tg, 187-192. Koppel, D. E., Sheetz, M. P., & Schindler, M. (1981) Proc. Natl. Acad. Sci. USA 1t, 3576-3580. Landreth, G. E., Williams, L. K., & Rieser, G. D. (1985) J. Cell Biol. 10 , 1341-1350. MacPherson, I. Montagnier, L. (1964) Virology gt, 291-294. Markwell, M. A. K., Haas, S. M., Bieber, L. L., & Tolbert, N. E. (1978) Anal. Biochem. t1, 206-210. McNutt, N. s., Culp, L. A., & Black, P. H. (1973) J. Cell Biol. gt, 412-428. Michaels, R. L., Pasternak, C., Young, J.-L., Elson, E. L., & Heuser, J. E. (1984) J. Cell Biol. gg, 32a. Otsuka, H. & Moskowitz, M. (1975) J. Cell Physiol. t1, 213-220. 89 Pasternak, C., & Elson, E. L. (1983) J. Cell Biol. gg, 270a. Penman, S., Fulton, A., Capco, D., Ben Ze'ev, A., Wittelsberger, S., and Tse, C. F. (1981) Symp. Quant. Biol. tt. 1013-1028. Petersen, N. 0., McConnaughey, W. B., & Elson, E. L. (1982) Proc. Natl. Acad. Sci. USA gg, 5327-5331. Pollack, R., Osborn, M., & Weber, K. (1975) Proc. Natl. Acad. Sci. USA gg, 994-998. Pollard, H. B., Stopack, S. S., Pazoles, C. J., & Creutz, C. E. (1981) Anal. Biochem. 110, 424-430. Rheinwald, J. G. & Green, H. (1974) Cell g, 287-293. Romano, A. H. (1976) J. Cell. Physiol. gg, 737-744. Romano, A. H. & Colby, C. (1973) Science 179, 1238-1240. Romano, A. H. & Connell, N. D. (1982) J. Cell. Physiol. 111, 195-200. Schlessinger, J., Barak, L. S., Hammes, 6. 6., Yamada, K. M., Pastan, I., Webb, W. W., & Elson, E. L. (1977a) Proc. Natl. Acad. Sci. USA gt, 2909-2913. Schlessinger, J., Elson, E. L., Webb, W. W., Yahara, I., Rutihauser, U., & Edelman, 6. M. (1977b) Proc. Natl. Acad. Sci. USA gt, 1110—1114. Stoker, M., O'Neill, C., Berryman, S., & Waxman, V. (1968) Int. J. Cancer t, 683-693. Tank, D. W., Wu, E.-S., & Webb, W. W. (1982) J. Cell Biol. gg, 207-212. Towbin, H., Staehelin, T., & Gordon, J. (1979) Proc. Natl. Acad. Sci. USA gt, 4350-4354. Webster, J. J., Taylor, P. B., & Leach, F. R. (1981) In Bioluminescence and Chemiluminescence (DeLuca, M. & McElroy, W. D., ed.) pp. 497-504, Academic Press, New York. Yamada, K. M., Yamada, S. S., & Pastan, I. (1976) Proc. Nat. Acad. Sci. USA gt. 1217-1221. CHAPTER IV IDENTIFICATION AND ISOLATION OF A FAST MOBILITY POPULATION FROM TRANSFORMED 3T3 FIBROBLASTS 9O INTRODUCTION In an examination of the effect of cell shape on the lateral mobility of membrane proteins, it was observed that a bimodal distribution existed with respect to measured diffusion coefficients for Kirsten murine sarcoma virus transformed BALB/c 3T3 fibroblasts (K-MSV 3T3) maintained in the spherical state. An analysis of the data suggested the existence of a subpopulation of cells having an abnormally high rate of diffusion for wheat germ agglutinin (WGA) glycoprotein receptors. Using the technique of fluorescence redistribution after photobleaching (FRAP), an effort was initiated to identify and then isolate a clonal population of "fast" mobility cells. These cells would be employed to examine a number of cellular properties and ascertain the cortical and medullary cytoskeletal elements that are potentially involved in the control of protein lateral mobility. In addition, this mutant could be used to examine a number of mechanisms requiring cytoskeletal links 'between cell surface receptors and nuclear structures. This communcation discusses the isolation of a clonal POpulation of transformed 3T3 fibroblasts that demonstrate IAnhindered lateral mobility for membrane proteins ((2.1 i 91 92 9 1.6) x 10- cm2/s) when cell shape is altered from flat and adherent to spherical and detached. The observed "fast" mobility is consistent with the fastest rates reported for proteins in cytoskeletal perturbed blebs from muscle cells 9 ((4.0 i 1.0) x 10- cm2/s) (Tank gt al., 1982), spectrin deficient spherocytes ((2.5 1 0.5) x 10-9 cm2/s) (Koppel gt al., 1981), and rod outer segment localized rhodopsin 9 ((2-4) x 19- cm2/s) (Poo & Cone, 1974). The actin stress fiber patterns in the "fast" mobility population were similar to the parent K-MSV 3T3 cells in that the majority had no visible stress fibers and the rest had fewer and thinner stress fiber cables than the untransformed BALB/c 3T3 cells. About half of the K-MSV 3T3 cells contained rosette structures as described in the preceding paper; however, only a few percent of the clonal cells contained the rosette structures. The increased rate of protein lateral diffusion in spherical cells of the "fast" mobility population and the subtle differences in the actin stress fiber structures correlated with a difference in the control of the rate of nucleo-cytoplasmic transport in these cells (Jiang and Schindler, unpublished observations). Significantly, the "fast" mobility trait appeared to be stably inherited, at least for 23 passages. MATERIALS AND METHODS Cells: BALB/c 3T3 and Kirsten murine sarcoma virus transformed BALB/c 3T3 cells were obtained and cared for as noted in the preceding chapter. A "fast" mobility population was derived through a series of subclonings on 96 well tissue culture plates and from colonies grown in soft agar from single cells. Clones having fast protein lateral diffusion were further subcloned. After a total of 3 subclonings on 96 well plates and 2 selections from soft agar, the "fast" mobility population (designated E761) was obtained and maintained through common tissue culture methods. Lateral Mobility Measurements: Lateral mobility of lipids and glycoproteins was determined in the plasma membrane using the technique of Fluorescence Redistribution After Photobleaching (FRAP) as described in the preceding chapter. NBD-phosphatidylcholine was used to monitor lipid mobility. Rhodamine-labeled wheat germ agglutinin (WGA) was used to monitor protein lateral mobility. Cells were labeled for 30 minutes at 4°C with 100 UQ/ml rhodamine-WGA in HBSS/HEPES. Fluorescence Microscopy: Stress fibers in cells were labeled with rhodamine-phalloidin and photographed as described in the preceding chapter. 93 RESULTS In the preceding chapter, we characterized the number and type of WGA receptors in BALB/c 3T3 and K—MSV BALB/c 3T3 cells (parent population for the E761 clone cells) since rhodamine-WGA was used to monitor protein lateral diffusion. Similar experiments, done with E761 clone cells, found no significant differences between the various cell populations. A comparison of WGA binding by bulk and single cell methods of analysis found no significant differences between E761 clone cells (Figures 1 & 2) and BALB/c 3T3 and parent K-MSV BALB/c 3T3 cells (Chapter III, Figures 1-3) in the spread and spherical state. As found for the parent population (Chapter III, Figure 4), E761 clone cells (Figure 3) had WGA homogeneously bound to the cell surface. Glycoprotein and glycolipid receptors from E761 clone cells (Figures 4 & 5) also were similar to those in the parent population (Chapter III, Figures 5 & 6) regardless of cell shape. It was concluded that the E761 clone cells do not have an anomalous WGA receptor population which account for the differences in protein lateral diffusion. The lateral mobility of phospholipids was monitored with NBD-phosphatidylcholine. Again, little difference in lipid mobility was found when cell shape was altered from spread and adherent to spherical and detached in E761 clone cells despite the large increase in protein lateral 94 WGA BOUND 95 2500 l— g 2000 — DJ p. 8 O. 1500 - ‘E” O \ 44 2' - ’0” \ l \ . fi - 500 -/ 0 I J l J l 25 50 IOO 200 300 [we A] ( pg/ ml 1 Figure 1. The binding of FITC-WGA to spread E761 clone cells 0.) and spherical E761 clone cells trypsinized and suspended in methyl cellulose for one hour (A), one day (0), and two days (0). WGA binding to spherical cells that had been removed with EDTA and suspended in methyl cellulose for one hour (I) was also determined. The error bars denote the standard deviation for the spread cells which was similar in the spherical cells. 96 .c3ocm ma wHHmo mcoau Honm Amy Hmouumnam pom Anm>oown unmonom may as fine uEwan as on-onv ucmaofiammoo coamsmoflo may ma om less oAMom oAMmo less mMoo oAMmu omamz Hons Amos mmwno oflwau kale oawss ma.oHoH.o <63 Hops loos oHHne m.~Hm.m floss oawmo o~.on~.o so: ezmsm clone o~+oo H.~+m.m lows o~+om mm.o+oo.o <63 muoso as a n ma m o mmomm mm>a a mo maamo aaonmmmm magma nammmm whHAHmOZ Aflfiflhdq 2Hmeomm Qz< QHquommwomm .H mqmdh 102 mobility phenotype. The status of the cytoskeletal element, actin, which has been demonstrated to be linked to membrane proteins (for a review see Geiger, 1983) was also investigated. The stress fiber actin (stained with the F actin specific probe, rhodamine phalloidin) of the clonal population in the spread state resembled that of the parent transformed BALB/c 3T3 cells in some ways (see Figure 6). As in the parent population there were two types observed. One type (about 17%) had some visible stress fibers, but again fewer and thinner than in the untransformed cells; the other type (about 83%) had no visible stress fibers. As noted in the preceding chapter, the parent K-MSV 3T3 cells displayed rosette structures, especially in cells lacking visible stress fibers. In contrast, only 5% of the E761 clone cells contained visible rosette structures. E761 clone cells in the spherical state had no visible stress fibers as seen for the untransformed and parent populations. Fibronectin in the E761 clone cells (Figure 6C) resembled that of the parent population (Chapter III, Figure 9F) i.e. has little fibronectin on the cell surface. The consequence of such dramatic changes in cells' dynamic and structural properties on cell growth was also examined. To determine cell growth rates, the synthesis of DNA was monitored for a period of 36 hours. The pattern of thymidine uptake in spread and spherical E761 clone cells (Figure 7) was similar to that observed in the parent 103 .Auv cfi czocm mum cfluomcounaw new vmaonma mHHmu mcoao acmm vacuum .onoum ofiwfiumam cHuumnm cm .cflvfioaamnalocfiemconu cuss pmamnma mHHmo wcoau Hunm Amy Hegemonmm cam Am who mum use ocavashcu vmaopma ecu Sofia comasm mums maamo scans ca mpoauma moon mco ecu wcwusc oxmuda ecu xums mums Hmucoufiuo: .maamo ocoau Acum Amcwa cocmmcv Hmoaumcdm new AmcHH caaomv pmmuam oucH ocwpasxnulmm mo oxmud: .m shaman mmdNDwm mMPud mmDOI mm mm mm _N N_ m_ m m _ O 1 q I ‘ 10000000. I .000600100'0000 II. :71... C) C) (3 IO N ‘— (POIX) was 1 C) Q’ 105 population (Chapter III, Figure 8). DNA synthesis occured about every 16 hours. While the E761 population was not synchronized as well as the parent population, DNA synthesis continued in the E761 clone cells after cell shape alteration. The "fast" mobility population, therefore, appears to be transformed as was the parent population. DISCUSSION In the preceding chapter a model was proposed for a communication pathway between the plasma membrane and the nucleus via the cytoskeleton. Evidence was presented to suggest that this pathway could be cell shape sensitive and its status could be monitored through measurement of protein lateral diffusion in the plasma membrane. Based on the observation that release of the restriction on protein lateral diffusion in untransformed BALB/c 3T3 cells was only partial and on Pasternak and Elson's (1983) finding of a biphasic compression curve for lymphocytes, it was suggested that the pathway consisted of two interconnected cytoskeletal systems: a cortical and a medullary cytoskeleton. In this chapter we report the isolation of a subpopulation of cells from Kirsten murine sarcoma virus transformed BALB/c 3T3 fibroblasts in which WGA receptors are completely released from lateral diffusion restrictions when cell shape is altered from flat and adherent to spherical and detached. A simple way to integrate these results would be to suggest that in untransformed BALB/c 3T3 cells the cortical cytoskeleton is reorganized during attachment and spreading as a result of recruitment of cytoskeletal elements to the membrane via cytoskeletal organizing proteins activated by attachment and spreading. This process may be analogous to 106 107 the binding of concanavalin A to lymphocytes and 3T3 fibroblasts which has been demonstrated to restrict protein lateral diffusion (Schlessinger gt gtt, 1977; Henis & Elson, 1981), recruit actin to the plasma membrane (Michaels gt gt., 1984), and increase the resistance of cells to deformation (Pasternak & Elson, 1983). Several membrane proteins have been shown to bind actin and might serve to seed filament formation. Other membrane proteins or, perhaps, the reorganization of the cortical cytoskeleton might initiate the formation of a medullary cytoskeleton completing the link between the plasma membrane and the nucleus. During detachment and the process of rounding, a reorganization or partial dissolution of the cortical cytoskeleton could account for the partial easing of the restriction of protein lateral diffusion. In the "fast" mobility population from the transformed cells, membrane proteins in spherical cells appear to be uncoupled from cytoskeletal interactions allowing unrestricted lateral diffusion to occur. Attachment and spreading seems to be accompanied by reassociation of membrane proteins and cytoskeletal elements as detected by the observation of restricted protein lateral diffusion in spread E761 clone cells. This suggests that the cortical cytoskeleton is functional in spread E761 clone cells if not in the spherical cells. We would suggest, however, that the medullary cytoskeleton is not functional in E761 clone cells, or at least not connected with the cortical 108 cytoskeleton, based on evidence presented in Chapter V which examines global modulation in spread and spherical cells. A key question yet to be addressed concerns the finding of a change in nuclear function or structure correlating with the change in protein lateral diffusion and cytoskeletal organization that occurs concomitant with cell shape alteration. The model proposes a communication pathway from the plasma membrane to the nucleus via the cytoskeleton. Schindler and Jiang (1986) reported that the cytoskeletal elments, actin and myosin, play a role controlling the rate of nucleo-cytoplasmic transport. The rate of nucleo-cytoplasmic transport decreased in 3T3 and K-MSV 3T3 cells after alteration of cell shape (Jiang and Schindler, manuscript in preparation), but not in the clonal population demonstrating the "fast" protein lateral diffusion. These results suggest a loss of control of a nuclear activity correlating with a loss of control of protein lateral diffusion in the plasma membrane. The biological significance of these studies becomes further apparent in consideration of the finding that while insulin or EGF can enhance nucleo-cytoplasmic transport in untransformed spread, adherent BALB/c 3T3 cells, there is no hormone effect upon transport in spherical BALB/c 3T3 cells (Jiang and Schindler, unpublished observations). 109 In summary, we have proposed a model for communication between the cell surface and the nucleus via the cytoskeleton similar to those suggested by Ingber and Jamieson (1985), Ben Ze'ev (1985), and Packard (1986). In this model it is suggested that adhesion or receptor mediated stimulation could affect the interaction of membrane proteins with cortical cytoskeletal elements and relay positional or stimulatory information to the nucleus via a medullary cytoskeletal system. Consistent with a functionally divided cytoskeleton, we observed a partial increase in protein lateral mobility after perturbing the cytoskeleton by alteration of cell shape in BALB/c 3T3 cells; furthermore, we found a complete release of diffusional constraints on membrane proteins in the "fast" mobility clone. The coincidence of the changes in protein lateral mobility, in cytoskeletal organization, in the control of nucleo-cytoplasmic transport, and the relation to cell proliferation suggests that altering cell shape in anchorage dependent fibroblasts provides a model system in which to examine the proposed communication pathway, especially when used in conjunction with the "fast" mobility population. REFERENCES Ben Ze'ev, A. (1985) Cell & Muscle Motility t, 23-53. Geiger, B. (1983) Biochim. Biophys. Acta 737, 305-341. Henis, Y. I. & Elson, E. L. (1981) Proc. Natl. Acad. Sci. USA gt, 1072-1076. Ingber, D. E. and Jamieson, J. D. (1985) In Gene Expression DuringtNormal and Maltgnant Differentiation (Andersson, L. C., Gahmberg, C. 6., and Ekblom, P., eds.), Academic Press, New York. Koppel, D. E., Sheetz, M. P., & Schindler, M. (1981) Proc. Natl. Acad. Sci. USA gt, 3576-3580. Michaels, R. L., Pasternak, C., Young, J.-L., Elson, E. L., & Heuser, J. E. (1984) J. Cell Biol. 99, 32a. Packard, s. (1986) TIBS 11. 154. Pasternak, C., & Elson, E. L. (1983) J. Cell Biol. gg, 270a. Poo, M. M. & Cone, R. A. (1974) Nature 247, 438-441. Schindler, M. & Jiang, L.-W. (1986) J. Cell Biol. 1 2, 859-862. Schlessinger, J., Elson, E. L., Webb, W. W., Yahara, I., Rutihauser, U., & Edelman, G. M. (1977b) Proc. Natl. Acad. Sci. USA gt, 1110-1114. Tank, D. W., Wu, E.-S., & Webb, W. W. (1982) J. Cell Biol. 22: 207-212. 110 CHAPTER V GLOBAL MODULATION AND CELL SHAPE: EVIDENCE FOR CYTOSKELETAL INVOLVEMENT IN THE CONTROL OF PROTEIN LATERAL MOBILITY 111 INTRODUCTION The model for a signal transmission pathway discussed in the previous chapters suggests that the cytoskeleton transmits messages between membrane receptors and the nucleus. The cytoskeleton is a likely candidate for such a pathway, since it spans the region between the plasma and nuclear membranes and is capable of conformational changes as well as translocation of vesicles and organelles along its length (Spudich gt gt., 1985; Vale gt gt., 1985). In the Chapters III and IV, it was suggested that lateral mobility could be used as an indicator for alterations in the cytoskeleton when cell shape was altered in anchorage-dependent fibroblasts. In this chapter further evidence is presented for a link between the change in protein lateral mobility that occurs concomitant with a change in cell shape and a change in the association of Inembrane proteins and the cytoskeleton. Initial evidence suggesting a link between the restriction of protein lateral mobility and the cytoskeleton first came from studies on the effect of concanavalin A (Con A) on lymphocytes (Edelman gt gt., 1973). Binding of Con A restricted protein lateral mobility 112 113 in an effect called global modulation (Edelman, 1976), because binding in even a small localized region of the membrane resulted in propagation of the effect over the entire surface. Both soluble factors and cytoskeletal elements were proposed initally to be the propagators of the effect. It was found that colchicine or cytochalasin B could partially reverse the effect. When used in combination, cytochalasin B and colchicine completely. reversed the effect (Henis & Elson,1981) suggesting that the cytoskeleton played a major role in regulation of protein lateral diffusion. It is clear that cytoskeletal integrity is necessary for global modulation. The consequence of alteration of cell shape, known to alter cytoskeletal organization, for Con A induced global modulation was examined. Global modulation was found to occur in flat, adherent BALB/c 3T3 fibroblasts, but not in spherical, detached fibroblasts. Furthermore, global modulation did not occur in flat, adherent or spherical, detached cells from the "fast" clone (E761) discussed earlier. The preceding chapters examined the changes in ‘wheat germ agglutinin receptor lateral diffusion upon cell shape alteration. In this chapter, the phenomenon is shown to also occur with succinyl Con A receptors indicating the general nature of the phenomenon. MATERIALS AND METHODS Cells: BALB/c 3T3 fibroblasts were obtained and maintained as stated previously. E761 clone cells are a "fast" mobility population of fibroblasts derived from Kirsten murine sarcoma virus transformed BALB/c 3T3 cells as detailed in Chapter IV. Spherical cells were suspended in media containing methyl cellulose as discussed in Chapter III. Con A-platelets: Platelets were obtained from the American Red Cross, washed, fixed, and bound with Con A as described elsewhere (Henis and Elson, 1981). Measurement of lateral mobility: The technique of fluorescence redistribution after photobleaching was used to monitor the lateral mobility of wheat germ agglutinin (WGA) receptors on the plasma membrane as described in Chapter III. To assess the effect of Con A on lateral mobility, cells were pretreated with Con A (100 pg/ml) or Con A—platelets for 15 minutes at 37°C and then incubated with rhodamine WGA (100 pg/ml) for 15 minutes at 4°C. Cells luere washed three times before use. Control samples were incubated with succinyl concanavalin A (S-Con A) instead of Con A. In addition, the measurement of protein lateral mobility was extended to include S-Con A receptors. Cells were incubated with 100 pg rhodamine labeled S-Con A/ml HBSS-HEPES for 15 minutes at 37°C, washed, and measured as 114 115 for cells labeled with WGA. Binding of S-Con A: The amount of S-Con A was measured on cells incubated for 15 minutes at 37°C with 100 UQ/ml FITC labeled S-Con A and washed. The amount of fluorescence was determined using an excitation wavelength of 460 nm and an emission wavelength of 520 nm (5 nm slit widths) and compared to a standard curve. RESULTS The monitoring of protein lateral mobility with a lectin, such as wheat germ agglutinin, provides an average diffusion coefficient for the variety of membrane proteins it binds. We extended the measurements of protein lateral mobility to glycoproteins that bind succinyl Con A. S-Con A has a mannose specificity as does Con A but does not induce global modulation. As seen in Table I, the results for the lateral diffusion of S-Con A receptors paralleled the results for WGA receptors. A slower, restricted S-Con A receptor mobility was seen in flat, adherent BALB/c 3T3 and 3761 ("fast" clone) cells. When the cell shape was altered from flat and adherent to spherical and detached; the S-Con A receptor mobility increased an order of magnitude for the BALE/c 3T3 and two orders of magnitude for the E761 clone Cells. As for the WGA receptors, there was a similar amount of S-Con A bound to flat, adherent cells and spherical, detached cells suspended in methyl cellulose for one day. After a 15 minute incubation, BALB/c 3T3 cells bound about 6 cells and E761 clone cells bound about 3 4 ug S-Con A/lO 119/105 cells, regardless of cell shape. These results suQgest that the increase in WGA and S-Con A receptor 1ateral mobility after cell shape alteration is a general suI'face phenomenon involving a broad range of membrane proteins . 110 mumfimumamus coo u amnmzoo “mucmefinmmxm mo amnesz Amhm>oomn ucmonmm any ma mxo 117 “3&6 so no-0Hxs ocmauwaumoo coamscoao may ms om mumps o~.amO~.~ Away ouMoo omo.mno~o.o azoo 483 “ohm «swam Hu.oHHm.o nos «swam mmoo.oHNooo.o qmaazoo <03 mamas «swam oo.owos.o Away mHHmo mooo.o+Mooo.o azoo cos mumem fimwmo om.HHo~.~ less mHHNo smo.owooo.o azOOIm see seam mfi+oo wfi.o+om.o Acme mfi+oo mfio.o+0mo.o «zoozm <63 mamem oumoe om.~mOs.m ANHV maMms mmo.om~uo.o mzoz «zoonm «ohm mmwoe oo.fiHoH.~ lass ofiwss oao.owofio.o mzoz <03 ”one mmwoo om.oHHo.o Ame mmwmo mmo.owmmo.o mzoz «zoosm mnmam e~+oo Hm.o+om.o lows o~+vo nmo.o+mmo.o mzoz <03 mlmam as a n as m a mmomm mass ammo ezmz maamo aaonmmmm magma newsmm -eammsmmm VHHQHmOZ Q¢KMH¢Q ZHMHomm .fi mamdfi 118 The requirement of an intact cytoskeleton for global modulation suggests that global modulation may be used to demonstrate the role of the cytoskeleton in restriction of protein lateral mobility. Global modulation occured in Spread BALB/c 3T3 fibroblasts where pretreatment with Con A resulted in a 5-fold decrease in the rate of WGA receptor lateral diffusion (see Table I) similar to that reported elsewhere (Schlessinger, gt gt., 1977; Henis & Elson, 1981) . No significant effect was observed when cells were pretreated with S-Con A as a control. Con A bound to platelets has been used previously to demonstrate that local binding of Con A is sufficient to propagate the modulatory effect over the entire membrane. Con A platelets bound to a localized region of the cell membrane was Sufficient to induce global modulation in spread BALB/c 3T3 68118 as has been reported by others (Schlessinger _e__t_ _a_1_=_. 1977; Henis & Elson, 1981). This demonstrated the effect can be propagated and was not simply the result of Crosslinking the entire surface of the membrane. In cOntrast, Con A or Con A-platelets did not induce global modulation in spherical BALB/c 3T3 cells. These results Paralleled the reported effects of cytochalasin B plus colchicine on global modulation (Henis & Elson, 1981) and °°incided with the known change in cytoskeletal °rganization after cell shape alteration (Badley gt g_lt, 1980). Especially intriguing was the finding that no Con A induced global modulation occurred in the "fast" mobility 119 clone, E761, regardless of cell shape. The E761 clone cells, derived from a transformed cell line, have a pattern of stress fibers differing from that of the untransformed BALB/c 3T3 cells. The E761 clone cells had few, if any, visible stress fibers as noted in the preceding chapter characterizing the clone. While no correlation could be detected between stress fiber organization and restriction of protein lateral mobility in spread cells, a coincidence of stress fiber disruption and absence of global modulation was noted . DISCUSSION A signal transmission pathway from the plasma membrane to the nucleus via the cytoskeleton requires demonstration of the importance of cytoskeletal integrity to the sending of the message. The studies of Folkman and Moscona (1978), .Benecke gt gtt (1978), and Cervera gt gt., (1981) suggest the importance of cytoskeletal integrity to cell growth in 'untransformed cells. These studies demonstrated the effect :of cell shape alteration on DNA and RNA synthesis and :showed in HeLa cells that mRNA being actively translated Ivere attached to the cytoskeleton. Schlessinger (1980) suggested the importance of lateral mobility to epidermal szrowth factor and insulin stimulation. Furthermore, Ilandreth gt gtt (1985) reported a shape sensitive aassociation of epidermal growth factor receptor with the <2ytoskeleton. These observations indicate the importance of <2ytoskeletal associations with membrane receptors in <:ellular function and suggest the significance of mnonitoring changes in lateral mobility as a means of (ietecting changes in the interaction of the receptors with ‘the cytoskeleton. The preceding chapters contained evidence for an increase in protein lateral diffusion coinciding ‘with a shape induced alteration of the cytoskeleton. The results presented in this chapter provide evidence that the increase in lateral mobility after cell shape alteration is 120 121 a general phenomenon and that alteration of cell shape coincides with the inhibition of a cytoskeletal dependent process (global modulation) that controls protein lateral mobility. The theory of global modulation is based on evidence that Con A reorganizes cytoskeletal elements thereby further restricting plasma membrane protein lateral diffusion. The disruption of actin and microtubules jprevents the occurence of global modulation. Alteration of cell shape is another means of disrupting the cytoskeleton; therefore, inhibition of global modulation by alteration of cell shape would provide evidence that the increase in protein lateral mobility after cell shape alteration (reported in the preceding chapters) is related to the (iisruption of the cytoskeleton. The lateral diffusion of PVGA receptors on flat, adherent BALB/c 3T3 fibroblasts was Irestricted by pretreatment with Con A as has been reported :Eor other proteins on lymphocytes (Henis & Elson, 1981) and :fibroblasts (Schlessinger §£.§l;: 1977). WGA receptor ilateral diffusion was not affected by pretreatment with asuccinyl Con A, which has the same binding specificity as (Ion A, in agreement with the literature (Schlessinger gt Elli: 1977). This suggests that the effect on protein lateral diffusion is specific to Con A and is not the result of an effect of Con A binding on WGA binding. The Con A induced global modulation observed in lymphocytes and fibroblasts may, in fact, use part of the 122 signal transmission pathway. There are a few intriguing similarities between Con A induced global modulation and hormone action of EGF which suggests a relavence for the use of global modulation in the study of the putative cytoskeletal mediated signal transduction pathway. Con A shows mitogenic activity (Edelman gt gt., 1973), caused aggregation of surface receptors, reorganizes the cortical cytoskeleton (Edelman, 1976), increases the association between the cytoskeleton and surface receptors (Painter & Ginsberg, 1982), and is inhibited by alteration of cell shape in anchorage dependent fibroblasts. Similarly, EGF action requires microaggregation of EGF receptors (Schlessinger, 1980), affects cytoskeletal organization (Schlessinger & Geiger, 1981), and EGF receptors are associated with the cytoskeleton in a shape sensitive fashion (Landreth gt gt., 1985). This is plausible in consideration of the following observations. One, global modulation requires a functional cytoskeleton (Edelman, 1976); two, Con A binding is followed by recruitment of actin to the membrane (Michaels gt gt,1984); and three, low doses of Con A, which do not restrict lateral mobility, are mitogenic (Edelman gt gt., 1973). The monitoring, then, of the state of the proposed signal transmission pathway by testing for global modulation may be especially appropriate. In untransformed flat adherent BALB/c 3T3 cells the restricted protein lateral diffusion, highly structured cytoskeleton, global modulation, and cell growth 123 all coincided. In contrast, alteration of cell shape was accompanied by an increase in protein lateral diffusion by a factor of 12, disruption of cytoskeletal elements, inhibition of global modulation, and inhibition of cell growth. Transformation, however, results in different reponses to cell shape alteration. The transformed, flat, adherent E761 clone, a "fast" mobility population, also demonstrated restricted protein lateral mobility but had a disorganized pattern of stress fibers indicating that stress fibers do not play a significant role in the restriction of protein lateral mobility in spread cells. The E761 clone cells, however, demonstrated a loss of global modulation correlating with the lack of stress fibers. In conjunction with the observations of Pasternak & Elson (1983) demonstrating a biphasic curve for the resistance of cell to deformation, these observations suggest that the cytoskeleton may be divided into at least two functional domains. One domain would be under the plasma membrane in association with membrane proteins and would function more or less independently of a medullary cytoskeleton. Another domain would be in the medullary region of the cell would be necessary for global modulation to occur. The alteration of cell shape in the E761 clone cells also coincided with an increase in protein lateral diffusion; however, the increase was 5-10 fold greater than in the untransformed cells and similar to that of lipid lateral diffusion. This 124 suggests that the E761 clone cells have a functional cortical cytoskeleton when attached and spread, but which becomes nonfunctional in regard to restriction of protein lateral diffusion when cell shape is altered. Finally, the alteration of cell shape was not accompanied by an inhibition of cell growth in spite of the increased protein lateral mobility and the disruption of the cytoskeleton. REFERENCES Badley, R. A., Woods, A., Carruthers, L., & Rees, D. A. (1980) J. Cell Sci. 3, 370-390. Benecke, B.-J., Ben Ze'ev, A., & Penman, S. (1978) Cell tt, 931-939. Cervera, M., Dreyfuss, G., & Penman, S. (1981) Cell gt, 113-120. Edelman, G. M. (1976) Science 192, 218-226. Edelman, G. M., Yahara, I., and Wang, J. L. (1973) Proc. Natl. Acad. Sci. USA gt, 1442-1446. Folkman, J. H. & Moscona, A. (1978) Nature 273, 345-349. Henis, Y. I. & Elson, E. L. (1981) Proc. Natl. Acad. Sci. USA gg, 1072-1076. Landreth, G. E., Williams, L. K., & Rieser, G. D. (1985) J. Cell Biol. 10 , 1341-1350. Michaels, R. l., Pasternak, C., Young, J.-L., Elson, E. L., & Heuser, J. E. (1984) J. Cell Biol. gg, 32a. Painter, R. G. & Ginsberg, M. (1982) J. Cell Biol. gg, 565-573. Pasternak, C., & Elson, E. L. (1983) J. Cell Biol. gg, 270a. Schlessinger, (1980) TIBS t, 210-214. J. Schlessinger, J., Elson, E. L., Webb, W. W., Yahara, I., Rutihauser, U., & J. & Geiger, B. (1981) Exp. Cell Res. 1 4, Schlessinger, 273-279. Schlessinger, J., Elson, E. L., Webb, W. W., Yahara, I., Rutihauser, U., & Edelman, G. M. (1977) Proc. Natl. Acad. Sci. USA gt, 1110-1114. Spudich, J. A., Kron, S. J. & Sheetz, M. P. (1985) Nature 315, 584-586. 125 126 Vale, R. D., Reese, T. S., & Sheetz, M. P. (1985) Cell tg, 39-50. CHAPTER VI: SUMMARY Protein lateral mobility decreased in the plasma membrane of BALB/c 3T3 fibroblasts following cell attachment and spreading. In contrast to the spread cells which demonstrated a further decrease in protein lateral mobility after Con A treatment, no global modulation was observed in spherical cells. The apparent shape dependency of plasma membrane protein lateral mobility was also observed in Kirsten murine sarcoma virus transformed BALB/c 3T3 fibroblasts which demonstrated two subpopulations of protein lateral mobility in the spherical state. In one subpopulation lateral mobility was approximately equal to that of the untransformed cells, while the other had a fast, apparently unhindered protein lateral mobility (approximately equal to lipid mobility; this is termed the "fast" mobility clone). In contrast to the untransformed cells, Con A mediated global modulation was absent in E761 spread cells as well as spherical cells. Other studies have indeed suggested that the absence of global modulation is characteristic of transformed cells. Investigation of the different cell states (shape and transformation), using an F-actin specific probe to detect actin based cytoskeletal 127 128 rearrangements, indicated a strong correlation between loss of global modulation and disruption of stress fibers. An interpretation of these results was presented that views plasma membrane protein receptor lateral diffusion under the control of at least two interacting cytoskeletal organizations, a cortical (submembranous) and a medullary (cytoplasmic) cytoskeleton (see Figure 1). In this model spherical cells have a disassembled medullary cytoskeleton and a loose cortical network which interacts with membrane receptors partially restricting their diffusion. Diffusionial control is mediated by either direct or indirect interactions with the cortical cytoskeleton. During cell attachment membrane proteins involved in adhesion are crosslinked, tension is generated, and the cortical cytoskeleton is reorganized into a tighter network. In the process, the rate of protein lateral diffusion decreases. Global modulation results as Con A crosslinks its receptors and triggers increased medullary/cortical interaction. The enhanced cortical density and tension is detected as a further decrease in protein lateral mobility and an increase in membrane stiffness. We believe the crosslinking of membrane receptors and increase in tension are the driving forces in the assembly of actin based cortical structures that result in enhanced interactions with membrane receptors. The "fast" mobility clone would be an example of cells with a loose cortical network that has an insignificant 1'1 129 m on ecumaoxmoumo osu co.ooaaool nones- as oooau. ooo..o on ooolooooo oooooaooooo oooo cooogoaoooso nooaoooo .ooooo oo.o¢o. oooooooo. oooagooooo coo..- .co93o3o 33...; 0.3;... :3) :OZo—oCooo econ—.23....“ co.o.oaooouo ~ou...oo oooo— co.ooo. o: .eotzoo..o o: oooa.~ oooooo oo --VW¢NNH%:- Jihad. . .cowumEuommcmuu one woman Haoo mo coauocau cuwa mcaououd accesses mo coauumumuca on» now ampoe < .s ouawam coaoo.ooooo oooo oooso ooaoooo .ooooo coco—on oofloeou sooooooc. oeoaguoooo sooaao 3.6.0; 3:17.. :3) cosozuoooo ooo“ . co.o_.4.o.su _oo_..oo oooo. so.o:o. o: ”cos—oouuo oe ouopu_ coo... cc eaoo.o‘oo..o s.o~aasoo coconogoooxo MK“ .oo.o.ou eouoo—aooanao ” H w » ounoo suhu oaaoo ool.a.oco.. "on..ozoo oouoouaooc noaoa. ooooau oooooo o.ooo _o ocoooaasooo ooaoooo oooooooo. .ooazooooo .. ...\ o . KI .Ioo osdmo n ooaoouooooo ouoo eooouonooouo ~oo«.~oo oooooo ooaoooo voooooooq oooagooooo \W vooaoo oe.ooo.o ocooaloc so.) so..o.o0ooo ooou cooouoxoooao “oosoooo oooo_ coqoco. o: 238.93: 2. oooaqc oooooo on coco-oaoooou soosaaooe I: ..u....u oqaoo ooooocoeocuo: .oosoozoo 130 interaction with membrane receptors. In this fashion the "fast" clone is reminiscent of spectrin deficient erythrocytes. The changes in protein lateral diffusion concomitant with the changes in the cytoskeleton and cellular metabolic activities take on new significance in light of the studies by others in the laboratory demonstrating shape induced 1 alterations in nucleocytoplasmic transport. The role of actin and myosin in the control of nucleocytoplasmic a transport provides suggestive evidence for a relationship I between cell shape alteration and the decreased rate of nucleocytoplasmic transport in spherical cells. In conclusion, this thesis presents a first step in connecting plasma membrane activation to nuclear response through the cytoskeleton. APPENDIX A The difference in the geometry of spread and spherical cells resulted in the development of diffusion equations and methods to detect diffusion in cells of planar or spherical geometries. For the measurement of lateral diffusion on planar cells such as spread fibroblasts, a solution of the diffusion equation for an infinite plane was derived. Such equations can be used for a point bleach or a multipoint method as described previously (Axelrod, gt gt., 1976; Koppel, 1979). In the multipoint bleach method the laser beam is scanned by an optical mirror in one direction (x) over the cell. The advantages of the multipoint method is that flow versus diffusion can be more easily detected and measurement of the bleach width necessary for the calculation of diffusion can be determined for each experiment instead of assuming it remains constant. The assumptions of the diffusion equations are that diffusion is occuring in an infinite plane and the labeled molecules have a uniform distribution, i.e. the prebleach fluorescence intensity, F(-), is independent of position. c(p,t) denotes the concentration of fluorescence molecules at position 2 at 131 132 time t. It is assumed that the bleach is centered at £=0 and that redistribution, not chemical recovery, accounts for the changes in c(t,t). The fluorescence intensity at time t and at position tr relative to the bleaching point is described by the equation: F(lr.t) = QII(£-Ar)6(£.t)dzg (1) where q is the product of all the quantum efficiencies of light absorption, emission, and detection, and I(g-tr) describes the monitoring beam profile which is assumed to be Gaussian. The solution of this equation for a circularly symmetric Gaussian beam results in the equation: F(Ar.t)/F(-) = 1-a(t)eXPI-IAr-2tl2/w2(t)1 (2) where v is flow velocity w, is the effective laser beam size, the 1/e radius 2 2 w (t) = w. (1+t/TD) (3) a one .<03 ocfiempocu cuss poaonma mummHQOEBHm emoudw mo mcmom commacumOQ m>wumucmmoudcm .N ouswfim Amuse: sumpuqcumv mflxm coon wcoam coHuHmom no ow mm on um om B on o o —op»-—->op—pppo—ooo._~> h—Pp-p—pbpp—pp-o—b-po O .182 Iooom m _- O Iooon a r S Ioooo m . w [coon a 188 m, 1 q Tooon fl. .- J Ioooo m _- rA jun-0000, m :88. m I Wooo: f Foooma .Ho>oa summacoua ecu mmumUAocfi ooom uaonm um wumcficuo can so won one .ummflnounfim cmmuam m co cumoac0uocd m nouum o>uso muo>oomu m>Humucomouaom .m ouswfim Acofima>ao\ooac mass on“ on“ on“ on“ 03 o: on« cue o: co“ oo on on on on ov on cm 3 o m. o I 82 . .... .38 l coon 1.. . w I 89. .x r 3 I oooo ( m r > I 88 m t q .... 82 m. I oooo u Il._r. 88 m I I. I 88. m .. ooo: ... oooma 136 fluorescent microscope by a series of mirrors (M). The last mirror before entering the microscope is mounted on a motor permitting scanning of the laser beam (SM). The beam is focused on the samples by two lenses (L). Fluorescence from the sample passes through the dichroic mirror (DM), scattered light is filtered out with a barrier filter (BF), and the fluorescence is measured by a photomultiplier tube (PMT). The fluorescence from several points along the scan of the sample is quantitated and a representative plot of the fluorescence versus position, x, is shown in Figure 2. The two scans shown superimposed in Figure 2 are postbleach scans normalized to the prebleach scan and demonstrate the recovery process. The vertical line represents the x=0 position of an arbitrary scale centered on the bleach. Fax x,t) plotted versus time is shown in Figure 3 for 180 postbleach scans and demonstrates the recovery of fluorescence due to redistribution of labeled molecules. A curve fitting algorithm is used to fit the data on the curve to a form of equation 2: F(Ax.t) = F(-){1-Ba(t) epr-(AX)2/w2(t)]-(1-B) w.exp[-(Ax>2/w.21) (6) where 8 gives the mobile fraction or extent of recovery. Using the curve fitting routine for equation (6) and and D can be computed. equations (3), (4), and (5) TD 137 To test the validity of the equations and the capability of the instrument, the diffusion of fluorescent dextrans in a thin layer of aqueous solution was measured. The diffusion coefficients obtained by the FRAP system for various molecular weight dextrans are shown in Figure 4 compared to diffusion coefficients obtained for various molecular weight dextrans obtained by chromatographic methods (Granath and Kvist, 1967). The values obtained by FRAP closely follow those from chromatographic methodology suggesting that the instrument and method of analysis are capable of producing reasonable results. Furthermore, the values obtained for lipid lateral diffusion and protein lateral diffusion fall within the values reported in the literature (Tank gt gt., 1982; Aroeti & Henis, 1986; Cherry, 1979). The equations shown above are appropiate for adherent, spread cells where the membrane approximates a plane. Koppel gt gt., (1980) presented equations for analysis of diffusion on spheres such as lymphocytes, detached fibroblasts, and vesicles. The normal mode of analysis has (two main advantages. It works for a general class of concentration distributions, with its only requirement being an initial azimuthal symmetry, and the relative ease of analyzing the exponential decays occuring from these initial distributions. The general solution for diffusion on a sphere with radius r, is: 138 .mump mop pumccmum ecu oumuaccfi mums uouum .mumc accommoumEOLLU och cwaounu cameo ma mafia one .Anu zcdmquUmEOHco can A09 mammamcm bosonsuaae moo\mocc 9549 as «V on «n on «m 0“ «a o « bbPP—bPbbLDPPP—PbbP_bbhh—|.PPPleI-Lbb_bbbbt—bLbkhDubb— a mo swamp HmHucocOme one .o musmfim ooo.vu oom.VI oov.VI oom.wl ooo.VI oom.mn oom.mu oov.nl oom.mn ooo.mu oom.~t oom.ml oov.ml (MIMI 143 D = (rzr)/2 (11). While there has been some doubt of the validity of the equations used and some problems in implementing them, several things indicate that systematic deviations in the calculations are not significant in the type of experiment performed. Analysis of the extent of recovery and diffusion coefficients found no evidence for a correlation between the extent of recovery and the calculated diffusion coefficient which would indicate a systematic error. Finally, the measured diffusion coefficients for lipids in the spherical fibroblast cells were close to those measured in spread fibroblasts using the multipoint analysis and were within the range for lipids reported in the literature (Tank, gt gt, 1982; Aroeti & Henis, 1986; Cherry, 1979). REFERENCES Aroeti, B., & Henis, Y. I., (1986) Biochemistry gt, 4588-4596. Axelrod, D., Koppel, D. E., Schlessinger, J., Elson, Webb, W. W. (1976) Biophys. J. tt, 1055-1069. Cherry, R. J. (1979) Biochim. Biophys. Acta 559, 289-327. Koppel. D. E. (1979) Biophys. J. gt, 281-292. Koppel, D. E., Sheetz, M. P., & Schindler, M. (1980) Biophys. J. tg, 187-192. E. I & Tank, D. W., Wu, E.-S., & Webb, W. W. (1982) J. Cell Biol. 22. 207-212. 144 APPENDIX B Publications Swaisgood, M. & Schindler, M. "Lateral Diffusion of Proteins in Fibroblast Membranes as a Function of Cell Shape" (to be submitted to Biochemistry) Swaisgood, M. & Schindler, M. "A Fast Mobility Mutant in Transformed 3T3 Fibroblast - Clonal Selection by Fluorescence Redistribution After Photobleaching (FRAP)" (to be submitted to Biochemistry) Swaisgood, M. & Schindler, M. "Global Modulation of Proteins in Fibroblast Membranes as a Function of Cell Shape" (to be submitted to Biochemistry) 145 [r 1‘72". 5'"! GLOSSARY Aggregation- crosslinking of surface receptors to form regions of closely associated receptors (variable in size). Anchorage dependent- refers to the cell growth requirement for the attachment and spreading (anchorage) of untransformed fibroblasts. Capping- the coalescence of patches into one large aggregate called a cap. Contact mediated recruitment- adhesion or contact with the substratum resulting in increased accumulation of actin under the membrane. Cortical cytoskeleton- a submembranous cytoskeletal assembly (e.g. mammalian erythrocytes) encircling the cytoplasm which is probably actin based and interacts with the medullary cytoskeleton. E761 clone, "fast" mobility population- is a population of cells isolated from K-MSV transformed 3T3 fibroblasts that demonstrate unhindered protein lateral mobility when cell shape is spherical. F-actin- actin in filamentous form. Global modulation- a cytoskeleton dependent process reported to restrict protein lateral mobility after crosslinking of surface receptors. The restriction is believed to occur through reorganization of the cytoskeletal systems. K-MSV 3T3- Kirsten murine sarcoma virus transformed BALB/c 3T3 cell line. Lateral diffusion, lateral mobility- refers to the random movement of membrane components in the plane of the plasma membrane and the rate is represented by the diffusion coefficient Medullary cytoskeleton- a cytoskeletal system pervading the cytoplasm composed of microfilaments, stress fibers, microtubules, and intermediate filaments and interacts withlgge cortical cytoskeleton. 147 Nucleocytoplasmic transport- the pore mediated movement of particles (dextrans in this instance) between the cytoplasm and the nucleoplasm. Patching- the aggregation of labeled receptors on the plasma membrane in response to crosslinking agents. Signal tansduction- the initiation, transmission, and nuclear acquisition of environmental signals or messages. Tensegrity- a structural support system composed of tension-producing elements and non-interacting compression-resistant elements and are independent of gravity. The model was derived from the highly efficient architectural design for the geodesic dome. 3T3 used alone refers to the untransformed cell line of BALB/c 3T3 fibroblasts. AN “711111.11 312 R 11111111111111 RIES in