wvflfifi fie.I .II I' {WU .IWIIIIIIIIIIIIIUWII IIIIIIIIII III INIIIIIIIIIIIIII. I, I 1%. II : .I.III'I.II.II ”I” . IIIIIIIIIIIIII.III..' III" III II‘ III; III} II IIIII,IIII.IIIIIIIII‘IIIIIIIIIIIIIIIIIIIIIIIIII’IIIIIIIIII III'II IIIIIIIII III.I:IIII‘I‘I§ I I. IbEIJII’Q‘IILIII IIWWIIMMIIW MummIWII '“M'II,‘1 IIII I I IIIIIIIIIIIIIIIII..I’ IIIIIIII IIII III" 'II‘IIIIIIIIII‘I IIIIIIIIIIIIIIII WWWUWINWI‘. W IIIIIIIM IIII'IIIII IIIIIIIIII IIIIIIIIIl 'I II” IIIIIIIII I.III IIIIIII III. .I’ I I , III 'II II III III IIIII I.,'.' IIIII IIIIIII wII II I" -II III III)II I. I II IIIIIIIIIIII'I II III .. I .I IIII I II I I' I I'I II II I'II‘III. . ' . . .3 II II”II.I'I' 'IIIII'II'II" .II.I ‘I .I .' I.I‘II’II'I'I."I'IIII'III'IIIII'II'IIIII. II.IIII.III I ,"-I II II' IIII'II III I‘IIIIIIIIIIIIIIIII I ' I.II. .II... IIIIII'I IIII II,III IIII .II-I‘I 'I.. II II III IIII IIIJIIII/ WW ;II II..I.;.I'I-I IIIII I)“ I.IIIII‘IIIII II'I IIII‘II.II‘.II I‘IIIIIIIII‘ IIIIIIIIIIII II.II' I'IIIIIIII IIIII IIIIII IIIIIII IIIIIIIIl I II IIII"I.' II) III IIIIII,.I.~I' .. .IIII.‘ I .II.--’I.‘I:II:IIIII 'II'II'IIIIII III II' 'II III' 'I' ...I,IIIIIII IIIIIIIII I I I'IIIII.I.IIIIIII.II‘.IIII.IIII.IIIII IIIIIIIMII.)IIIIIII’II.'.'IIIII.I‘I.IIIII'I.'I I'II.IIII I II II I II'I. l'l I I I I‘ II ‘I II‘ III I I III I 'III‘IIII‘III’IMIIIIIIII-II .' 'IIILIII‘IIII'Iw IIIIII III I . I .. I... IIII I II. I I . . . . . ' . . . II"I I' II. .". ... ' . . I , . I.II I 'I ." .III I I. In. I. I" II' ‘III)I_ I 0’ II J. I ' I ”." ‘ I II III I II 'I'I. .II, ’ILI II "I I 'I . I' III I I I. . I I . . I . l I I . II. I . I. .. II III'III’ II III I III.I‘II.I' . I I I II.I .‘I II 'I' IIIIIII’IIII .I'III‘IIIIIIIIII I III IIHIII’II III IIIIIIIIIII'I . I'I'III 'I, .III “II' I I II I I'IIIII IIIIIIIIII II III'II II I“I III’fIvIIIIII .I III I II; I.II II I III III 'IIIIIIIIIIIII III I I II IIIIIIIII III III. IIIJI‘III‘II II III II .IIIII Iw W... II"'I'I“II'“IIII'II'IIIIII'I' ~ ‘IIHIIIIII IIII‘ IIIIII III’; I IXI I II I‘IIIII III‘I'III I‘IIIIII'IIIIIIII'IIIIIIIIII III_ II .III' II III I ".I W . . IIII.’.’I IIIIIIIIII'IIIIIIIII,3I‘- .III I.II..'.II‘ . '.I'. .- II IIII;I',‘II'II I 'II- ,I I I I. . . II I .. I .. I I ‘II. I .‘ LII r I 'II‘III’IIIII IIIIIII ‘ "I.II I.I’IIIII I m-- a”: I’ In: . I-IIIII‘III II " II I I I I III. II III IIII'III I.I III‘ III I. I.‘ ‘II . I'IIII' III“ IIIIII(.I I IIII I II I IIIIIIIIIIII, “‘II” {IIII \II‘IIIIIIIII II‘III I, .I‘I II I. II II I.I-IIIIIIIIIIIIIII IIIIII III III III.II'IIIIIII.I.,IIIIIIIIII I .l [III II .'. I I:. I I 'IIII'I ’III‘I'I‘ ‘I'II‘II‘IIII'I III".IIII I ‘IIII‘II I II‘ I I ‘., I. II.II. II.“ . VIII!" III IU'III‘II 'IIIII‘IIIIII. ',III'IIIII'; III‘I| IIIII'II III’ II II I I. I. . ‘ IIII II ‘ . ' . .I I I II'I'IIIII‘IIIIIH II III 'IIIIIIHIII III I I? IIII [I I.II {II‘-I.II'IIII ’I II/ II" I. I IIII‘. II If I prn‘n. vaI"'”“MI HIM a." ..w1 -4 I ‘ I; ' ,I. .3 II. III .I -. i‘ . ,I'II" I"IIII "II'I‘I' " ""I I .‘I I II I I I .I‘IIF'.I".::.II'i um.III‘:I..‘III_..IMI. ".I..." .II'III‘.I‘~.I“'?,I’, .I . .. "I In . . . .' I .. ‘ II‘I' "' II‘III III! I I‘III IIIIIII III" "WIII I . “, uwthIV; I. IVW‘.wII I I II " ' III I‘IIII:III.I IIII “I I I‘ll I I I I, IIIIIIIIIII IIIIIIIII‘III IIII‘ III. II‘IIII-I IIIIII\(I F’NI’H‘I/IIH 'I'I‘II' I I! (RI!!!) Ila] .II I. ‘ M I ,;.-; ' .' III' .I. I .I 'I I.. IIIIIIII.II III. . ,' I II'IIIII .’I I III' ., .-I I. " .... .; ...,. ~IIIII'IIII. . II...“"" I I .| I I I I II I II)‘ IIIII IIII II'IIIIII'IIIIII. IA'I INIpII II." I ., I I II‘ IIIIII. .4 l I IIII, I-I I III “II .. . I I I ‘ II 'III' "MI III'II' I“ I I.IIW'IwHII IIIMI #VIIWI IIIIIMIIIZ . ’ 'LIIIIII .I,4‘ .' II II I II I.I‘ “III 'I I I‘ | ILI I I I II I'IIIIIIrIIéI I‘I’IIJI I‘hI'II ""1 3) VIII“ ...h;I III I I.IIIIIIIN. 'IIJL‘I‘ I HIIIHIIII I IIIIIIJI‘ "I',III' IIIIIIIIIIIIIIIIIIIIIIIII IIIIIIIIIIIIIII III'II II.II I.” [II III ‘4'! III'.‘ I I.I'II IIHI *- :z _‘A:_’;:— This is to certify that the thesis entitled Isolation and Partial Characterization of Tubulin from a Higher Plant presented by Narendra Singh Yadav has been accepted towards fulfillment of the requirements for Ph. D. degree inW P1 ant Pathology I n/w 7kgW1 fiajox professor Date_£ebLua_n¥_18,_L9.80 0-7639 OVERDUE FINES; 25¢ per day per item RETUMIING LIBRARY MATERIALS: M- Place in book return to remove charge from circulation records ISOLATION AND PARTIAL CHARACTERIZATION OF TUBULIN FROM A HIGHER PLANT BY Narendra Singh Yadav A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Botany and Plant Pathology 1979 ABSTRACT ISOLATION AND PARTIAL CHARACTERIZATION OF TUBULIN FROM A HIGHER PLANT BY Narendra Singh Yadav The objective of this study was to isolate and characterize plant tubulin. Section I deals with attempts to isolate and study tubulin of Cchhicum autumnale, a plant which accumulates the microtubule poison, colchicine. It was not possible to measure tubulin in extracts of Cchhicum by the conventional colchicine binding assay because the extracts strongly inhibited binding of 3H-colchicine to brain tubulin. Therefore, mobility of the doublet of tubulin polypeptides in sodium dodecyl sulfate (SDS) polyacrylamide gel electrophoresis, using brain tubulin as a marker was adopted as the tubulin assay. Two major polypeptides~in Colchicum leaf extracts behaved as a doublet that migrated close to the doublet of brain tubulin. The abundance of the plant doublet raised the possibility that it was the large subunit of Fraction-I- protein. Therefore, brain tubulin labeled by complexing with 3H—colchicine and spinach Fraction-I-protein labeled by complexing with 2(14C-carboxy)-ribitol-l,5-bisphosphate were used to develop a method for their separation. An apparently associative interaction Narenda Singh Yadav between these two proteins was encountered in experiments involving gel filtration on Sepharose 68. However, ion-exchange chromatography on BEBE-cellulose was effective in their resolution. When the Cblchicum protein was analyzed by this method the prominent protein proved to be ‘Eraction-I-protein. In the absence of a good tubulin assay attempts to isolate tubulin from Cchhicum were abandoned in favor of the more general problem of isolating plant tubulin. Section II deals with studies on the colchicine binding activity in extracts of cultured tobacco XD cells. Extracts of tobacco cells were shown to inhibit partially the colchicine binding activity of brain tubulin. This inhibition was removed most effectively by dithioerythritol. However, tobacco extracts continued to show poor colchicine binding activity under conditions optimal for binding of colchicine to brain tubulin. Section III deals with the isolation of tubulin from cultured tobacco cells by sedimentation of polymerized tubulin. Cow brain tubulin was added to the soluble protein fraction of extract from 35S-labeled cells and subjected to cycles of temperature-dependent assembly-disassembly. When analyzed by SDS polyacrylamide gel electrOphoresis about 70% of the radioactivity in the twice copolymerized protein.was found in a prominent doublet migrating close to the doublet of brain tubulin, detected by Coomassie blue staining. While the leading radioactive band comigrated with the B—subunit of brain tubulin, the trailing radioactive band migrated between the a- and B-subunits of brain tubulin. When analyzed by two dimensional isoelectric-focusing/SDS polyacrylamide gel Narendra Singh Yadav electroPhoresis the radioactive doublet behaved like the doublet of brain tubulin but not like the large subunit of Fraction-I- protein. Limited proteolysis of the individual bands observed by SDS electrophoresis with Staphylococcus aureus protease or a- chymotrypsin showed that, while the peptide maps of the leading radioactive band and of the B-subunit of brain tubulin were virtually indistinguishable, the maps of the trailing radioactive band and of the a-subunit of brain tubulin, though similar, were not identical. Most of the capolymerized 35S-labeled protein also behaved like brain tubulin during gel filtration on Sephadex G-150 and during ion- exchange chromatography on columns of DEAE cellulose or phospho- cellulose. It is concluded that the doublet of radioactive polypeptides isolated by cepolymerization with brain tubulin are tobacco tubulin polypeptides and that these polypeptides, in their native as well as denatured forms, have properties very similar to, but not identical with, cow brain tubulin. Apparently, tubulin has been highly conserved during evolution. Tobacco tubulin was also partially purified from cell extracts by self-assembly. Unlike cow brain tubulin, which has approximately equal amounts of a- and B-subunits, tobacco tubulin, isolated either by copolymerization with brain tubulin or by self-assembly, has a much higher amount of a-subunit relative to the B~subunit, as . . . 35 . . . detected both by Cooma851e blue staining and S—radioact1v1ty. To my parents, for letting me reach the stars. on 9, ACKNOWLEDGMENTS I am grateful to the American taxpayer for financing my graduate study and to the many friends and colleagues who made my stay at the Plant Research Laboratory a very pleasant experience. I wish to express deep gratitude to my advisor and guru, Dr. Philip Filner for his compassion and encouragement throughout my work, for enlightening me in the obvious and the not-so-obvious aspects of scientific thinking and for making science, and this thesis, an adventure for me. I am indebted to the other members of my guidance committee for their support, especially Dr. Norman Good for his criticism of my thesis, and Drs. Peter Carlson and Hans Kende for their wise counsel. I wish to thank John Pierce for providing me with 2(14C- carboxy)ribitol-l,5-bisphosphate, Christian Paech for purified Fraction-I-protein, Dean Gabriel for his help in two-dimensional electrophoresis and last, but not least, Philip Trinity for his invaluable assistance in getting this thesis printed into reality. TABLE OF CONTENTS LIST OF FIGURES. . . . . . . . . . . . . . . . . . . . . LIST OF TABLES . . . . . . . . . . . . . . . . . . . . . LIST OF ABBREVIATIONS. . . . . . . . . . . . . . . . . . GENERAL INTRODUCTION . . . . . . . . . . . . . . . . . . Biochemistry of Tubulin . . . . . . . . . . . . . . 3 1h Vitro Assembly of Microtubules . . . . . . . . . . SECTION I ISOLATION OF TUBULIN FROM COLCHICUM.AUTUMNALE IntrOduction O O O O O I O O O O O O O O O O O O O O O 0 Is Colchicine Binding Activity a Universal Property of T‘lbul in O O I O O O O O O O I O O O O O O O O O O O 0 Resistance of Cells to Colchicine . . . . . . . . . . Materials and Methods. . . . . . . . . . . . . . . . . . Preparation of Cow Brain Tubulin. . . . . . . . . . . Colchicine Binding Assays . . . . . . . . . . . . . . Measurement of Radioactivity. . . . . . . . . . . . . Spectroscopy. . . . . . . . . . . . . . . . . . Preparation of Fraction-l-Protein Labeled with 2(14C-Carboxy)- ribitol-bisphosphate (CRBP) . . . . . . . . . . . . . Preparation of 3H-CLC-Tubulin Complex . . . . . . . . Sucrose Density Gradient Centrifugation . . . . . . . Gel Filtration. . . . . . . . . . . . . . . . . . . . DEAR-Cellulose Ion Exchange Chromatography. . . . . . SDS-Polyacrylamide Gel Electrophoresis. . . . . . . . Protein Determination . . . . . . . . . . . . . . . . Thin Layer Chromatography . . . . . . . . . . . . . . Chemical. . . . . . . . . . . . . . . . . . . . . . . Results. . O O O O O O O O I O O I O O O O I O O O O O 0 Analysis of the (NH ) SO Fractions of Cchhicum Extract on SDS- GEL ElectrophoreSIS. . . . . . . . . . . Absorption Spectra of the (NH4 ) 2SO Fractions of Cchhicum Extracts O O O O O C O I O O O O O 2? O O O O C I O O O Colchicine Binding Activity in the (NH4 ) SO Fractions of Colchicum Leaf Extracts . . . . . . 4.2.. 4 . . . Inhibition of Colchicine Binding Activity of Brain Tubulin by Cchhicum Leaf Extracts . . . . . . . . . . . . iv Page - - vii . . xi 0 O xii . . 4 . . 6 . . 8 . . 8 . . 9 . . 11 . . l3 . . 14 . . 15 . . 15 . . l6 . . 16 . . 17 . . 17 . . 18 . . 18 . . 18 . . 19 . . 19 . . l9 . . 19 . . 23 . . 28 . 28 Identification of the Protein Doublet in SDS-Gel after Electrophoresis of Cblchicum Extract. . . . . . . . . . Sucrose Density Gradient Centrifugation. . . . . . . Gel Filtration . . . . . . . . . . . . . . . . . . . DE-32 Ion Exchange Chromatography . . . . . . . . . . . SECTION II ISOLATION OF COLCHICINE BINDING PROTEIN FROM CULTURED TOBACCO CELLS IntrOduction O O O O O O O O I O O O O O I O O 0 O O O O O O I mterial and MethOdS O O O O O O O I O O O O O O O O O O O O 0 Preparation of Colchicine Binding Protein from Tobacco Cells Results 0 O O O O O O O O O O O O O O O O O C O O O O O O O 0 Preparation of Colchicine Binding Fraction from Tobacco Cells. Colchicine Binding Activity of Brain Tubulin in the Presence of Tobacco Extract and Different Buffers . . . . . . . . . Colchicine Binding Activity in Other Plants. . . . . . . . Binding of 3',5'-Cyclic Adenosine Monophosphate to Brain Tubulin. . . . . . . . . . . . . . . . . . . . . . . . . . Discussion. 0 O O O O O O O O I O O O O O O O O O O O O O 0 SECTION III ISOLATION OF POLYMERIZED TOBACCO TUBULIN BY DIFFERENTIAL CENTRIFUGATION Introduction 0 O O O O C O I O O O O O O O I O O O O O O O O O In Vitro Self-Assembly of Microtubules . . . . . . . . . . Neuronal Tissues. . . . . . . . . . . . . . . . . . . . Non-Neuronal Cells and Tissues. . . . . . . . . . . . . Flagellar and Ciliary Axonemes . . . . . . . . . . . . . . Heterologously Primed Assembly . . . . . . . . . . . . . . Copolymerization . . . . . . . . . . . . . . . . . . . . . Isolation of Intact Microtubules . . . . . . . . . . . . . Materials and Methods . . . . . . . . . . . . . . . . . . . . Cell Culture and Radiolabeling . . . . . . . . . . . . . . Preparation of Cow Brain Tubulin . . . . . . . . . . . . Copolymerization of Labeled Tobacco Cell Extract with Cow Brain Tubulin. . . . . . . . . . . . . . . . . . . . . . . SOS-Polyacrylamide Gel Electrophoresis and Autoradiography . Peptide Mapping by Limited Proteolysis . . . . . . . . . . Quantitation of Radioactivity in Polyacrylamide Gels . . . Two Dimensional Electrophoresis. . . . . . . . . . . . . . Gel Filtration. . . . . . . . . . . . . . . . . . . . . Ion-Exchange Chromatography . . . . . . . . . . . . . . Electron Microscopy. . . . . . . . . . . . . . . . . . . . Determination of Protein Trichloroacetic Acid Stable Radioactivity. . . . . . . . . . . . . . . . . . . . . . . v Page 42 42 55 66 80 84 84 85 85 86 91 96 96 104 104 104 106 108 109 110 110 111 111 112 113 114 115 115 116 116 116 117 117 Determination of Radioactivity. Chemicals and Radiochemicals. Results. . . . . Copolymerization of 35[SJ—Labeled-Proteins of Tobacco Cells with Cow Brain Tubulin. Effect of Cow Brain Tubulin on Sedimentation of735[S]-Counts from.Extracts of 35[s]-Labe1ed Tobacco Cells. . . . . . . . . Effect of Brain Tubulin Concentration on the Sedimentability Of 35 [SJ-counts o o o o o o o o o o o o o o o o o o o o o o 0 Specific Radioactivity of Copolymerized Proteins. . SDS-Polyacrylamide Gel Electrophoresis of 35Isl-Labeled Copolymerized Proteins. Peptide Maps of the Prominent Protein from Tobacco Cells that Copolymerize with Brain Tubulin. Two Dimensional Isoelectric-Focusing/SDS-Electrophoresis of the Proteins . Comparison of the Properties of Native Twice Copolymerized 35[S]-Labeled Protein with 3H-Colchicine-Brain-Tubulin Complex . . . Isolation of Particulate Tubulin from Tobacco Cells . Isolation of Tubulin by Self Assembly . Discussion . . . Identity of the Tobacco Doublet . Significance of the Difference between Tobacco and Brain Tubulin LITERATURE CITED vi Page 117 118 118 118 118 119 121 127 137 142 153 156 175 179 179 184 194 LIST OF FIGURES Figure 1 SA SB 6A 63 7A 78 9A 9B 10 SDS polyacrylamide gel electrophoresis of proteins in Cblchicum leaf extracts. . . . . . . . . . . . . . . . . Absorption spectra of colchicine (CLC). . . . . . . . . . . Absorption spectra of the 350 nm absorbing compounds in Cchhicum extracts. . . . . . . . . . . . . . . . . . . . . Decay of the 350 nmrabsorption of colchicine with time (in min) on irradiation with long ultraviolet irradiation . Inhibition of the 3H-colchicine binding activity of brain tubulin by unlabeled colchicine. . . . . . . . . . . . . . Thin layer chromatography of Cblchicum extract and colchicine. . O C C O I O C O O O . O C O I O . O O I O O O Sucrose density gradient centrifugation of spinach Fraction- I-protein (F-I-P) and colchicine (CLC). . . . . . . . . . . Sucrose density gradient centrifugation of spinach Fraction- I-protein (F—I-P) and dialyzed 0-30% and 30-50% saturated (NH4)ZSO4 fractions of COZchicum leaf extract. . . . . . . Sucrose density gradient centrifugation of spinach Fraction- I-protein (F-I-P) and dialyzed 0-30% and 30—50% saturated (NH4)ZSO4 fractions of Cchhicum leaf extract. . . . . . . Sucrose density gradient centrifugation of spinach Fraction- I-protein (F-I—P) and dialyzed 0-30% and 30-50% saturated (NH4)2504 fractions of Cchhicum leaf extract. . . . . . Sucrose density gradient centrifugation of spinach Fraction- I-protein (F-I-P) and the dialyzed O-30% saturated (NH4)ZSO4 fractions of Cchhicum leaf extract. . . . . . . Sucrose density gradient centrifugation of spinach Fraction- I-protein (F-I-P) and the dialyzed 0-30% saturated (NH4) 80 fractions of Cchhicum leaf extract. . . . . . 2 4 Molecular sieve filtration of the Cchhicum leaf extract on Sepharose 4B. . . . . . . . . . . . . . . . . . . . . . Page 22 25 27 30 38 41 44 46 48 SO 52 54 57 Figure 11 12 13 14 15 16 17 18 19 20 21 22 23 24 Molecular sieve filtration of Fraction-I—protein and 3H-colchicine in Sepharose 68. . . . . . . . . . . . . . . Molecular sieve filtration of 3H-colchicine-brain-tubulin complex in Sepharose 68. . . . . . . . . . . . . . . . . . Molecular sieve filtration of 3H—colchicine-‘brain—tubulin complex and Fraction-I-protein in Sepharose 6B. . . . . . . Molecular sieve filtration of 3H-colchicine-brain-tubulin complex and 2(14C-carboxy)-ribitol-l,5-bisphosphate- Fraction-I-protein complex in Sepharose 68. . . . . . . . . DEAE-cellulose ion-exchange chromatography of 2(14C- carboxy)-ribitol-l,5-bisphosphate-Fraction-I-protein complex and -colchicine-brain-tubulin complex. . . . . . DEAE-cellulose ion-exchange chromatography of 3H-colchicine- brain-tubulin complex and dialyzed 30-50% saturated (NH ) SO fraction of Cchhicum leaf extract. . . . . . . . . . . . DEAE-cellulose ion-exchange of 3H-colchicine-brain-tubulin complex and dialyzed 0-30% saturated (NH ) SO fraction of Colchicum leaf extract. . . . . . . 5 ¥ .4. . . . . . . DEAE-cellulose ion-exchange chromatography of 3H-colchicine- brain-tubulin complex and spinach Fraction-I-protein. . . . Effect of brain tubulin concentrations on the sedimentation of 35S-c.p.m. from 353-labe1ed tobacco XD cells. . . . . . Specific radioactivity (trichloroacetic acid stable radioactivity, c.p.m., per mg protein) in once, twice and thrice copolymerized microtubule pellets. . . . . . . . . . SDS-PAGE of brain tubulin, spinach Fraction-I-Protein (F-I-P) and twice copolymerized microtubule pellet in 15% separating gel. . . . . . . . . . . . . . . . . . . . . . . SDS-PAGE of cow brain tubulin and twice copolymerized microtubule pellet in 8% separating gel. . . . . . . . . . Profile of 35S-radioactivity in strips cut from SOS-PAGE of twice copolymerized microtubule pellet (of the same gel shown in Tract C3 of Figure 22). . . . . . . . . . . . . . SDS-PAGE of cow brain tubulin and 35S-labeled proteins of tobacco cells after self-assembly and copolymerization. . . viii 4 2 4 Page 0 68 . 70 . 74 . 76 . 123 . 126 . 129 . 132 . 134 . 136 Figure Page 25 SDS-PAGE of peptide fragments generated by Staphylococcus aurens protease from cow brain tubulin and the putative tobacco tubulin subunits present in twice copolymerized microtubule pellet. . . . . . . . . . . . . . . . . . . . . 139 26 SDS-PAGE of peptide fragments generated by Staphylococcus aurens protease from cow brain tubulin and the putative tobacco tubulin subunits present in twic copolymerized microtubule pellet. . . . . . . . . . . . . . . . . . . . . 141 27 SDS-PAGE of peptide fragments generated by a-chymotrypsin from cow brain tubulin and the putative tobacco tubulin in the twice copolymerized protein. . . . . . . . . . . . . 144 28 SDS-PAGE of peptide fragments generated by a-chymotrypsin from cow brain tubulin and the putative tobacco tubulin in the twice copolymerized protein. . . . . . . . . . . . . 146 29 SDS-PAGE of peptide fragments generated by Staphylococcus aurens protease from the slower-moving large subunit of Fraction-I-protein (F-I-P) and cow brain tubulin. . . . . . 148 30 Two-dimensional isoelectric focusing/SDS-PAGE of twice copolymerized protein. . . . . . . . . . . . . . . . . . . . 150 31 Two-dimensional isoelectric focusing/SDS-PAGE of brain t‘JbUIino O O O O O O O O O O O O O O O O O O O O O O I O O O 152 32 Two-dimensional isoelectric focusing/SDS-PAGE of spinach FraCtion-I-pIOtein. o o o o o o o o o o o o o o o o o o o o 152 33 Molecular sieve filtration of 35S-tobacco proteins in twice copolymerized microtubule fraction, 3H-CLC-brain tubulin complex and free 3H-CLc on Sephadex 9-150. . . . . . 155 34 DEAR-cellulose ion exchange chromatography of 3SS-tobacco protein in twice copolymerized fraction. . . . . . . . . . . 158 35 DEAE ion exchange chromatography of 3H—colchicine tubulin complex. . . . . . . . . . . . . . . . . . . . . . . . . . . 160 36 Phosphocellulose ion exchange chromatography of twice copolymerized 35S-tobacco protein. . . . . . . . . . . . . . 162 37 Phosphocellulose ion exchange chromatography of cow blood platelet tubulin. . . . . . . . . . . . . . . . . . . . . . 164 38 Electron micrograph of the pellet of high speed sedimentation of extracts of tobacco cells made in a stabilizing medium, described in text. . . . . . . . . . . . . . . . . . . . . . 168 ix Figure Page 39 SDS-PAGE of self-assembled and copolymerized 35s-proteins in high speed pellet of tobacco cells prepared under stabilizing conditions. . . . . . . . . . . . . . . . . . . . 171 40 Electron micrograph of self-assembled proteins in the high speed supernatant of tobacco extracts in the presence of 15% dextran T10 in a buffer containing 100 mM MES, pH 6.5, 1 mM EGTA, 1 mM GTP and SIM DTE. . . . . 174 41 SDS-PAGE of 35S—proteins in high speed supernatant of tobacco extracts, after self-assembly. . . . . . . . . . . . 177 Table 10 11 12 13 14 LIST OF TABLES 3 . . . . . . . . H-Colch1c1ne Binding Act1v1ty 1n (.NH412804 Fractions of Cblchicum Leaf Extracts. . . . . . . . . . . . . . . . Inhibition of the CLO-Binding Activity of Brain Tubulin by Boiled Extract of Cblchicum Leaf. . . . . . . . . . . . Inhibition of the 3H—CLC Binding Activity of Brain Tubulin by the (N84)2804 Fraction and Leaf Extract of Golchicum. . UV-Induced Loss of Absorbance at 350 nm in Extracts of CO ZChicm Leaves 0 O O O O O O O O O O O O O O O O O O O O 0 Binding of 14C-CRBP and 3H-CLC Binding of Brain Tubulin. . 3H-CLC Binding Activity in Different Fractions of Tobacco Extract 0 O O O C 0 O O O O O O I O O O O O C O C O O O O I Effect of (N84)250 on 3H-CLC Binding to Brain Tubulin. . . 4 Effect of Tobacco Extracts in Different Buffers on the 3H-CLC Binding Activity of Brain Tubulin. . . . . . . . . . 3H-CLC Binding Activity in Cauliflower Extracts in Sucrose. 3 . . . . . . H-CLC B1nd1ng Act1v1ty 1n Caul1flower Extracts. . . . . Cyclic-Adenosinemonophosphate and Colchicine Binding Activities of Brain Tubulin. . .1. . . . . . . . . . . . . Effect of Brain Tubulin on Sedimentation of 35[S]-Counts from 3S[SJ-Tobacco Extracts. . . . . . . . . . . . . . . . Specific Radioactivity of Copolymerized Proteins. . . . . . Depolymerization of the High-Speed Pellet and Its Poly- merization in the Absence and Presence of Brain Tubulin. Page .120 .124 .169 230: 350‘ CAMP: DMSO: EDTA: EGTA: F-I-P: GTP: OD: PM: PPO: RnBP: SDS: LIST OF ABBREVIATIONS absorbance at 280 nm absorbance at 350 nm 3',5'-cyclic adenosine monOphosphate column buffer colchicine counts per minute 2(carboxy)-ribitol bisphosphate deuterium oxide diethylaminoethyl dimethylsulfoxide dithioerythritol ethylenediamine tetraacetic acid ethylene glycol-bis(B-aminoethy1 ether),N'-tetraacetic acid Fraction-I-protein guanosine triphosphate dissociation constant microtubule associated protein methyl benzimidazol—Z-yl carbamate 2(N-Morpholino)ethane sulfonic acid microtubule optical density phosphate buffer diphenol oxazol ribulose-l,5-bisphosphate sodium dodecyl sulphate ultraviolet GENERAL INTRODUCTION Eukaryotic cells are not bags of enzymes. On the contrary, they have the capacity to define an intracellular spatial distribution of molecules and organelles and to change that distribution. It is becoming increasingly clear that eukaryotes use at least two types of fibrous structures, microtubules and microfilaments, to perform and regulate movements of intracellular material, of the cell surface and of the whole cell. Knowledge concerning microtubules has been accumulating rapidly and has been the subject of a recent book (Dustin, 1978), of several excellent symposia (Borgers and De Brabander, 1975; Soifer, 1975; Goldman et al, 1976), general reviews (Mohri, 1976; Snyder and McIntosh, 1976; Stephens and Edds, 1976), and reviews concerned with microtubules in the plant kingdom (Pickett-Heaps, 1974; Hepler, 1976; Hepler and Palevitz, 1976; Filner and Yadav, 1979). Microtubules are a class of subcellular structures of eukaryotes which have the appearance of rigid cylinders about 250 A in diameter, with a central lumen about 150 A in diameter. The wall of the cylinder is usually composed of 13 protofilaments which are parallel or nearly parallel to the cylinder axis, each a chain of globular subunits. Center-to-center spacing between subunits is about 40 A within a protofilament, and about 50 A between adjacent protofilaments. Optical diffraction patterns obtained from electron micrographs of microtubules have provided evidence of helical periodicities on the microtubule surface (Amos et al, 1976). When microtubules are negatively stained, the lumen fills with stain, indicating that the microtubule is probably a hollow cylinder (Bryan, 1974). A clear annular zone or halo about 100 A wide is sometimes observed around microtubules in cross section CLedbetter and Porter, 1963). Microtubules are involved in a variety of cellular processes concerned with the form and motility of cells (see Stephens and Edds, 1976). The best known function of microtubules is their involvement in chromosome movement. In addition, plant micro- tubules are associated with movement of vesicules during cell wall deposition and cell plate formation and control of cell shape, directly, in the case of wall-less cells and, indirectly, in the case of walled cells, where they appear to control the orientation of cellulosic microfibrils in the wall (see Filner and Yadav, 1979). A striking feature of the distribution of the of microtubules in vivo is their ephemeral nature. For example, the mitotic spindle appears just before, and disappears just after chromosome movement. Micro~ tubules are believed to be in a state of dynamic turnover and several antimicrotubule agents can shift the "equilibrium" towards depolymerization or polymerization. The fragility of in situ microtubules is evident by the fact that they can be readily dis- rupted in vivo by exposure to several antimicrotubule agents, such as low temperature, high pressure, and several drugs (see Stephens and Edds, 1976) or by merely disrupting the cell in which.they occur. On the other hand, D20 enhances the level of microtubules in viva (Gross and Spindel, 1960). How the cell regulates the time, place and the orientation of microtubule assembly is a major unsolved problem. Although microtubules have a strikingly similar morphology regardless of the organism, microtubules with slightly different geometries do occur. The microtubules in flagella, cilia and sperm tails are found in a cylindrical arrangement, called the axoneme, which consists of nine outer doublet microtubules encircling a central pair of single microtubules. Each outer doublet fiber contains a complete microtubule called the A subfiber, and an incomplete microtubule, the B subfiber, which shares three protofilaments with the A-subfiber. The axonemal microtubules differ from the other cellular microtubules, the so-called cytoplasmic microtubules mentioned above, not only in their morphology but also with respect to their stability to the antimicrotubule agents. Axonemal microtubules are resistant to these treatments. On the basis of their relative stabilities microtubules have been classified into two broad types: labile microtubules (including most of the cytoplasmic microtubules) and stable microtubules (including the axonemal microtubules). However, microtubules of each type also have differing degrees of stability to antimicrotubule agents. The multitude of cellular processes involving microtubules and the diversity in the morphology and stability of.microtubules raise the question of the biochemical basis of this diversity -— another major unresolved question. Biochemistry of Tubulin The major component of microtubules is a 63, 110,000 dalton protein called tubulin. Tubulin exists as a dimer of two poly- peptides of about 55,000 daltons each. The dimension of the globular subunits of microtubules make it likely that they represent the 55,000 dalton polypeptides, i.e. tubulin is thought to be a dimer of globular subunits. Two moles of guanine nucleotide are non-covalently bound to both flagellar and cytoplasmic tubulin dimers. While one mole of the bound nucleotide is readily exchangeable with.free GTP the other is apparently not (Weisenberg et al, 1968). There are several drugs that bind to tubulin and interfere in its function. For example, each tubulin dimer binds approximately one mole of colchicine or podophyllotoxin (Cortese et al, 1977) and two moles of the vinca alkaloid, vin- blastine (Bhattacharya and Wolff, 1976; Wilson et al, 1978). While colchicine and podophyllotoxin have a common binding site (Cortese et al, 1977), the binding sites for guanine nucleotides and vinblastine are distinct from that of the other ligands mentioned above. Both guanine nucleotide and vinblastine stabilze the colchicine binding site. When tubulin is denatured it dissociates into two distinct polypeptides of similar molecular weights. The two polypeptides can be resolved on the basis of their mobility on polyacrylamide gel electrophoresis in the presence of either 8 M urea or 8 M urea and 1% sodium dodecyl sulfate (SDS) (Bryan and Wilson, 1971; Lee et al, 1973). The faster moving polypeptide is called the B-subunit and the slower moving the a-subunit. The resolution of these two subunits has been shown to be based on their charge, and not their size difference (Bryan, 1974). The separation of the a- B-subunits under certain conditions on SDS-electrophoresis has been attributed to the anomalous behavior of the d-subunit on such gel systems (Bryan, 1974). Although the a- and B-subunits have similar amino-acid composition, their peptide-maps are distinct (see Snyder and McIntosh, 1976). Two other definite biochemical differences between these subunits are known: B-subunit is specifically phosphorylated (Eipper, 1972) and the o—subunit can be specifically tyroinsylated at its carboxyl terminus by a tubulin- tyrosine ligase. It is not known if all tubulin dimers are o8 heterodimers. Since the d- and B-subunits occur in an almost equal amount, it is believed that tubulin is an a8 dimer and not do or 88 homodimers. Direct evidence on this question has been obtained by treating tubulin with bifunctional reagents and analyzing the products in a gel system capable of resolving ad, a8, and BB. Sixty to ninety percent of the cross-linked dimer was a8. However, it is not known how much of the cross-linked homodimers do, 88 obtained was due to non-specific aggregation. Axonemal tubulin has properties that are strikingly similar to those of neunonal tubulin, including amino-acid composition, sequence near the amino terminus, peptide fingerprints, interaction with drugs and binding sites for guanine nucleotides (see Stephens and Edds, 1976). Luduena and Woodward (1973) isolated the o— and 8— subunits from the chicken and from the sea urchin. Amino acid sequences were determined for the first twenty-five amino acid residues. While there was no difference in these sequences of a-subunit from the two sources, there was only a single difference in the B-subunits from these sources. In Vitro Assembly of Microtubules Following the discovery of the conditions for in vitro assembly of tubulin into microtubule in brain extracts (Weisenberg, 1972), there has been a major research effort directed towards elucidation of the pathway of microtubule assembly and the nature of its regulation both by physiological effectors and pharmacological agents. 0n the basis of the observation that a critical concentration is required for assembly (Gaskin et al, 1974), it appears that the in vitro assembly of microtubules is a nucleated polymerization with distinct nucleation and elongation steps (Timasheff, 1979; Johnson and Borisy, 1977). In vitro studies on microtubule assembly have also shown that a dynamic turnover of microtubules occurs (Johnson and Borisy, 1977; Margolis and Wilson, 1978) and that the antimicrotubule drugs, colchicine and podophyllotoxin, block the assembly reaction by binding to a small, substoichmetric amount of free tubulin (Margolis and Wilson, 1977). Substoichiometric poisoning of in vitro microtubule assembly is consistent with the estimate that only three to five percent of free tubulin need bind colchicine to block mitosis in K8 cells (Taylor, 1965). Margolis and.Wilson (1977) have proposed that colchicine blocks microtubule assembly by binding to the ends of microtubules as a colchicine- tubulin complex. Although.the minimal requirements for in vitro assembly of microtubule appear to be pure tubulin, GTP, possibly Mg2+ and absence of Ca2+, several other agents can facilitate microtubule assembly. The in vivo significance of these requirements, if any, remains to be determined. Study of tubulin has largely been limited heretofore to the microtubule-rich.sources. More data on the structure of tubulin from a variety of cell types are necessary before the molecular basis of microtubule function can be understood. The present study is an effort to extend the work on tubulin to higher plants. WOrkcnisuch microtubule-poor sources has been hindered by the lack of a good method to assay tubulin, since it has no easily quantified biological properties. Therefore, the present study is largely directed towards development of a method to isolate tubulin from plant cells. SECTION I ISOLATION OF TUBULIN FROM COLCHICUM AUTUMVALE' SECTION I ISOLATION OF TUBULIN FROM COLCHICUM AUTWNALE Introduction Is Colchicine Binding Activity a Universal Property of Tubulin? Colchicine (CLC) binds to soluble tubulin isolated from a variety of animal cells and from stable microtubules (MTs) and prevents their in vitro assembly into MTs. The exact mechanism by which CLC prevents the assembly of MTs is not known (see General Introduction). The available evidence, from studies of brain tubulin in vitro, indicates that MTs depolymerize and polymerize at the ends, with depolymerization occurring perferentially at one end and polymerization at the other (Dentler et al, 1974; Margolis and Wilson, 1978). It has been proposed that the CLC-tubulin complex binds to MT ends and prevents further addition of tubulin subunits (Margolis and Wilson, 1977). Because of the dynamic turnover of labile MTs, inhibition of assembly results in MT disruption. According to the prevailing view, the apparent resistance of MTs (for example, that of the axonemal MTs) to CLC may be attributed not to CLC-insensitive tubulin but to their lack of turnover, either due to their intrinsic stability or to the presence of 8 accessory proteins that stabilize the MTs. Evidence in support of this idea comes from the observations that (a) CLC prevents regeneration of flagella (Rosenbaum et al, 1969): (b) CLC binds to tubulin solubilized from axonemes (Wilson and Meza, 1973; Maekawa and Sakai, 1978); and (c) CLC inhibits in vitro assembly of tubulin solubilized from axonemes (Maekawa and Sakai, 1978). CLC binding and normal MT assembly are, therefore, mutually exclusive properties of the tubulins studied thus far. The question is "how universal is this incompatibility between CLC binding and MT assembly?". The fact that CLC disrupts the cytoplasmic MTs of both plant and animal cells has led to the hypothesis that the CLC binding site on tubulin overlaps with one or more sites of tubulin- tubulin interaction required for precise MT assembly, and, therefore, the CLC-binding site on tubulin is as evolutionarily conserved as are the sites for assembly. The hypothesis that tubulin has a common site for both CLC binding and MT assembly can be tested by looking for mutants for natural genetic variants of tubulin which can assemble into a MT even in the presence of CLC, with or without binding CLC. Furthermore, variants of tublin showing differential sensitivity to CLC could be useful in studying the structure-function relationships of tubulin and also throw light on the mechanism of MT assembly. Resistance of Cells to Colchicine A wide range of concentrations of CLC is required to block . . . . . -6 mitos1s or disrupt MTs 1n d1fferent spec1es, for example, 10 M for most animal cells (see Dustin, 1978) and 10"2 M for Nicotiana 10 callus (WOoding, 1969). It remains to be determined if this range of tolerance is related to differential sensitivity of tubulin to CLC. In fact, in most cases studied so far, CLC resistance of cells appears to be related to other factors, such as: a) permeability of cells to CLC, e.g. in.Amoeba (Camandon and Defonbrune, 1942) and in Chinese hamster cells (Ling and Thompson, 1974); b) metabolic destruction of colchicine, e.g. in golden hamster (Sch3nharting et al, 1974); c) presence of an extra pool of tubulin or other CLC binding sites, e.g. in Syrian hamster cells (Minor and Roscoe, 1975). Historically, MTs have had a special relationship with several drugs of plant origin, known as anti-mitotic agents. The three most common drugs are CLC from Colchicum autumnale L. (Liliaceae). podophyllotoxin from Podophyllum‘peltatum L. (Berberidaceae) and vinblastine from Vinca rosea L. (Apocynaceae). The function of these anti-mitotic drugs like that of other secondary plant products is unknown. However, the obvious question of the mechanism of resistance of these plants to the endogenous drugs is raised. Cblchicum autumnale is a good choice of material in which to look for CLC-resistant MTs. Roots of Cblchicum can grow in 500 mM CLC (Levan, 1940). Cornman (1941, 1942) reported a complete spindle destruction in C. byzantinum with four hours of 125 mM CLC and in C. autumnale with six hours of 250 mM CLC. They concluded that the immunity of Colchicum resides not in a difference in the mitotic process but in some extra-mitotic protective agency. Levan and Steineggar (1947) suggested that the above reaction in Colchicum may 11 be due to traces of chloroform left over during the crystallization of CLC. Mehra et al (1948) reported normal mitosis in roots of C. luteum placed in 50 mM CLC for over two months. However, Takenaka (1950) investigating the sterility of C. autumnale reportedza c-mdtosis effect on the pollen mother cells and attributed this to endogenous colchicine. The CLC content of Cblchicum plant varies greatly with the tissue, stage of growth and environment. It is reported to be around 0.6% by dry weight in the corm and the seed, 0.8% in the mature flower and slightly less in the leaves. The CLC content of the corm.increases with its age and that of the seed almost doubles with fruit ripening. In seeds the alkaloid is concentrated chiefly in the endosperm and the third layer of the seed coat. When intact Allium roots were exposed to the crushed corm of Cblchicum, the root cells showed the CLC effect within 30 min of the exposure (Cornman, 1942). With the hope of finding an interesting genetic variant of tubulin, such as one insensitive to CLC, the author attempted to isolate tubulin from Colchicum autumnale. Materials and Methods Leaves, corm and flowers of Colchicum autumale (Liliaceae) were obtained from plants grown in one of the four places on campus: Horticulture Garden, Plant Research Laboratory field and greenhouse, and Beal Botanical Gardens. Plants grown in the greenhouse were 12 obtained as bulbs from a distributor in Babylon, New York. The tissues were washed in cold running water to remove dirt, surface dried by a paper towel and unless mentioned otherwise homogenized in the homogenizing buffer, H8, 20 mM sodium phosphate, pH 6.8, 5 mM MgCl2 and 5 mM potassium metabisulfite, at three volumes of buffer per 9 fr. wt. with a pestle and mortar in the presence of a little acid-washed sand, at 4°C. Addition of 5 mM potassium metabisulfite was necessary to prevent browning of the extract due to oxidation of phenolic compounds, which was not prevented by up to 5 mM cysteine-HCl. In later experiments, Colchicum leaves were frozen in liquid nitrogen and made into a powder before being homogenized in the homogenizing medium by twenty passes of a motor-driven glass-teflon homogenizer. The total homogenate was sedimented at 40,000 x gmax for 60 min in a Sorvall SS-34 rotor at 4°C. The resultant supernatant fraction, 40KS60, was made to 30% saturated (N84)280 by the slow addition of 4 solid (N84)ZSO4 with constant mixing at 4°C. The pH of the solution was maintained at 6.8 and the solution allowed to stand in cold for one hour. The precipitate was then collected by sedimentation at 12,000 x gmax for 20 min, and resuspended in the least possible volume of H8. The resultant supernatant was made 50% saturated in (N84)2804, as described above. The 30-50% fractions precipitate was collected by sedimentation at 12,000 x gmax for 20 min and resuspend- ed in least possible volume of H8. The resuspended 0-30% and 30-50% fractions were dialyzed overnight against ca.200-fold excess of H8. If 13 necessary, the (NH.4)ZSO4 fractions were lyophilized and stored frozen. Preparation of Cow Brain Tubulin Cow brain was isolated from fresh cow skulls, usually obtained from a local slaughterhouse. In the laboratory, within an hour of slaughter, the brain was cleaned of blood clots and 50 9 portions were minced with scissors and suspended in 150 ml of grinding medium,50 mM MES, pH 6.5, 2 mM MgCLZ, 50% glycerol and 10% DMSO at room temperature. This medium is-a modification of the stabilizing medium devised by Filner and Behnke (1973) for the isolation of pre- existing microtubules from brain. The suspension of minced brain was blended in a Waring blender at room temperature, for 15 sec at low speed. The homogenate was rapidly chilled on ice to prevent soluble tubulin from polymerizing spontaneously. The chilled homogenate was sedimented at 12,000 x gmax for 15 min and the resultant supernatant fraction, beneath the lipid overlay, was carefully collected and then sedimented at 40,000 x gmax for 60 min. This second supernatant fraction was further sedimented at 142,000 x gmax for three hours. This high speed sedimentation ensures removal of all pre-existing microtubules. The soluble tubulin in the high speed supernatant was polymerized by incubation with 0.1 mM GTP, and 1 mM EGTA at 37°C for one hour. The in vitro polymerized microtubules were collected as described above, but at 25°C. The resultant pellet was resuspended in storage buffer,50‘mM MES, pH 6.5, 2 mM MgCl and 50% (v/v) glycerol and stored at 4°C. When required 2 14 this microtubule suspension was diluted four to five times in 100 mM MES, pH 6.5, 4 mM MgC12, 2 mM EGTA and 1 mM GTP and depolymerized at 4°C for about 30 min before use. If necessary the microtubule preparation was carried through one more cycle of polymerization by first sedimenting it at 142,000 x gmax for 60 min at 4°C and incubating the resulting supernatant with 50% (v/v) glycerol at 37°C for one hour. The twice polymerized microtubule pellet was collected and stored as described for the once polymerized pellet. Colchicine Binding Assays Unless mentioned otherwise 3H-CLC was incubated with protein samples (at final concentrations of 2.2 uCi/ml and 0.4 uM in total volume of 450 pl) at 37°C, away from direct light (to prevent the photoconversion of CLC to the inactive lumi-CLC). After varying periods of time the reaction was st0pped by chilling the incubation mixture on ice. Bound colchicine was assayed by a filter assay using either DEAE cellulose (Whatman DE—32) resin, following the method of Nicholson and Veldstra's (1972) modification of that of Weisenberg at al (1968) or DEAE cellulose impregnated filter discs (Whatman DE~ 81 paper). These methods are based on the binding of tubulin, an acidic protein,with cationic DEAR-cellulose resin. For the filter assay using DEAE cellulose resin, Whatman DE-32 powder was equilibrated in phosphate buffer (5 mM sodium phosphate, pH 6.8 and 5 mM MgC12) and suspended at 100 mg powder/ml. Aliquots of the reaction mixture were added to 0.2 ml of the cold DE-32 15 slurry and kept ice-cold for at least 30 min. The assay was diluted with 5 ml of cold phosphate buffer and then filtered through a 19 mm Millipore fibre glass pro-filter, followed by five washes of 5 m1 cold phosphate buffer under vacuum. For the filter assay using the filter discs, aliquots of the reaction mixture were added directly onto two layers of Whatman DE-8l filter discs, equilibrated in cold phosphate buffer and then washed by 30 ml of the same buffer. Measurement of Radioactivity Radioactivity in aqueous samples or on filter-discs was measured in 5 m1 of scintillation fluid using a Beckman LS-133 scintillation counter. In early experiments, the scintillation fluid was composed of 5 g PPO, 0.5 g POPOP, 570 ml Toluene and 330 m1 Triton x-1oo per liter. However, the use of this mixture was later discontinued because of excessive chemiluminescence. Instead, a dioxane based scintillation fluid was used which comprised of 60q;naphthalene, 4.0 g PPO and 0.2 g POPOP, 100 m1 methanol and dioxane to make the final volume of one liter. Spectroscopy The absorption Spectra of different fractions were obtained between 270 and 400 nm, with a Cary model 14 recording spectro- photometer, using a cell with a 1 cm light path. Long wavelength ultraviolet radiation was obtained with Blak-Ray UVL lamp (Ultra- Violet Products, Inc.). UV irradiation was, when necessary, done on ice in the cold room in front of a fan to prevent heat denaturation of sample proteins. 16 Preparation of Fraction-l-Protein Labeled with 2(14C-Carboxyl- ribitol-bisphosphate (CRBP) Purified spinach fraction-l-protein (FlP) was dialyzed over- night against 10 mM Tris-Cl, pH 7.8, 5 mM MgC12, 0.1 mM EDTA and 10 mM 2-mercaptoethanol. The dialysate was sedimented at 10,000 rpm for 20 min in a Sorval SS-34 rotor at 4°C to remove aggregated protein. The resultant supernatant fraction was incubated at room temperature with about 22 moles of 14C-CRBP (ca. 1.5 x 105 cpm/umole), per mole of FlP. Since eight moles of CRBP are expected to bind one mole of FlP, this represented about two-fold excess of 14C-CRBP over FlP. Unbound 14C-CRBP was removed by passing the reaction mixture through a G-25 Sephadex column (2.2 x 36 cm) at 4°C, equilibrated with 10 mM Tris-Cl, pH 8.0, 5 mM MgCl 0.1 mM EDTA and 5 mM 2! B-mercaptoethanol. The bound radioactivity represented the 14C- CRBP-FlP complex. Preparation of 3H-CLC-Tubulin Complex Once polymerized microtubules in 50 mM M88, 1 mM GTP, 1 mM EGTA, 2 mM MgCl 50% (v/v) glycerol, pH 6.5 was diluted five-fold to 2' W11 uM tubuline in 10% (v/v) glycerol, 100 mM MES, 1 mM MgCl 1 mM 2! GTP, pH 6.5. This was incubated with 3H-CLC (sp. act.: 5.5 x 10 cpm/umole) at 37°C for 75' and then sedimented at 18,000 rpm for 60 min in Sorvall 58-34 rotor. The supernatant fraction was then assayed by the DE-32 filter assay. Thirty-five percent of the label- ed CLC was found bound to tubulin at 1.9 x 105 cpm/ml. 17 Sucrose Density Gradient Centrifugation Five percent and 20% sucrose solutions were made in 20 mM sodium phosphate, pH 6.8 and 5 mM MgC12. Aliquots of 17.5 ml of the 5% and of the 20% sucrose solutions were used to prepare a linear gradient of 5-20% sucrose in polyallomer tubes using a gradient maker. Protein samples were sedimented at 12,000 x g for 20 min to remove any denatured aggregates. Sample volumes of 0.5-1.0 ml were carefully overlaid on top of the gradient. The samples were in turn overlaid by a layer of mineral oil. The gradients were centrifuged at 27,000 rpm in a Beckman SW 27 rotor at 4°C for the desired time. Fractions were collected by puncturing a hole at the bottom of the tube and collecting fractions driven out by air pressure from an infusion pump connected to the top of the tube and absorbance measured at 280 nm or 350 nm in a Gilford spectrophotometer. Gel Filtration Different molecular sieve gels (Sepharose 4B and Sepharose 68) were equilibrated in 25 mM MES, pH 7.0, 1 mM EGTA and 1 mM MgC12, pH 6.8 for the required period of time and poured into columns (1.5 x 35 cm for Sepharose 4B and 0.9 x 31 cm for Sepharose 68). After equilibration with the buffer to be used in the run, the column was loaded with 1.0 mlof sample and eluted with the buffer at a flow rate of 5—10 ml per hr generated by a peristaltic pump (LKB Instruments). Fractions of about 2.1 ml were collected and 18 analyzed either for absorption in a Gilford spectrophotometer or for radioactivity. DEAE-Cellulose Ion Exchange Chromatography_ Whatman DE-32 powder was equilibrated in the column buffer C8 (25 mM MES, pH 7.0, 1:mM.MgC12, 0.1 m.GTP and 25% glycerol) at 4°C. Protein samples were applied onto the column and eluted first with C8 and then with a linear salt gradient, as described in the text. SDS—Polyacrylamide Gel Electrophoresis The method of Laemmli (1970) was followed in preparing the disc gels; they were 0.8 cm in diameter and had a separating gel containing 10% acrylamide. The polymerizing solution was overlaid with n—butanol until the gel had formed. SDS was omitted from the lower electrode buffer. Protein samples were also prepared according to Laemmli (1970). Between 10-200 ug protein in 100 pl was applied, then subjected to electrophoresis at 4 mA/gel. The procedures of Weber and Osborn (1967) were followed for staining and destaining with Coomassie blue, except that the gels were diffusion destained at 50°C. Protein Determination The method of Lowry et al (1951) was used after precipitation of protein by 10% trichoroacetic acid and washing with ethanol. Bovine serum albumin was used as the standard. 19 Thin Layer Chromatography . Thin layer chromatography was performed on silica gel plates with chloroform:acetone:diethylamdne (70:20:10) as the solvent system. The UV absorbing spots were detected under long UV light. Chemicals All chemicals were analytical grade. (BH-methoxy) colchicine was purchased from New England Nuclear. l4C-CRBP was the kind gift of John Pierce, Department of Biochemistry, MSU. Purified spinach Fraction-l-protein was obtained from Dr. Tolbert's Laboratory at MSU. Results Analysis of the (NHA)QSO, Fractions of Cblchicum Extract on SDS-Gel Electrophoresis The presence of tubulin in different fractions of Cblchicum extracts was looked for by analyzing them on SDS-polyacrylamide gel electrophoresis (EDS-PAGE) and comparing the Coomassie blue stained bands with those obtained from electrophoresis of brain tubulin. Cblchicum polypeptides that comigrated.with the brain tubulin marker were tentatively identified as Colchicum tubulin. Fresh Cblchicum leaves were harvested and homogenized in ice- cold homogenizing buffer, H8 (20 mM sodium phosphate, pH 6.8, 5 mM MgCl , 0.1 mM GTP, 5 mM potassium meta bisulfite at three volumes 2 per 9 fr. wt. by a pestle and mortar. The homogenate was sedimented 20 at 40,000 x g for 60 min. The resultant supernatant fraction was used to prepare 0—30% and 30-60% saturated (NH4)ZSO4 fractions. Aliquots of the total homogenate and the dialyzed 0-30% and 30-50% (NH ) SO 4 2 4 fractions were analyzed on SDS-PAGE (Fig. 1). A prominent doublet of polypeptides was observed in the 0-30% (NH4)ZSO4 fraction of colchicum extract that migrated similarly to the doublet characteristic of brain tubulin. In different extractions, between 35% and 45% of total soluble protein from colchicum leaves fraction- ated in the 0-30% (NH4)ZSO percipitate. Since the doublet accounted 4 for most of the protein in the 0-30% (NH fraction, it con- 4,2504 stituted roughly 30% of the total soluble leaf protein. If the doublet was in fact tubulin polypeptides then it was possible that Cblchicum was producing a vast excess of tubulin as a means of detoxifying some of the endogenous CLC. There is a precedent for such a mechanism of resistance in certain mammalian cells. For example, levels of dihydrofolate reductase increase about ZOO-fold in the presence of substrate analogs (Alt et al, 1976). HOwever, the prominent doublet was found only in the extracts of leaves and not in extracts of flowers and corm (storage tissue) of Colchicum. This suggested that the doublet in leaf extracts might be derived from the major protein of photosynthetic tissues, Fraction-l-protein. Fraction-l-protein (FlP) or ribulose-l,S-bisphosphate carboxylase often comprises more than 50% of the soluble leaf protein (Jensen and Bahr, 1978). The molecular weight of the native FlP is around 550,000 daltons. It is made up of two kinds of subunits - the N55,000 dalton large subunit and the W16,000 Figure l. 21 SDS polyacrylamide gel electrophoresis of proteins in Colchicum leaf extracts. Total leaf homogenate ("T.H."L dialyzed 0-30% ("30%”) and 30-50% (“50%") saturated (NH4)2SO4 precipitates and cow brain tubulin (B) were electrophoresed in 10% separating gel and stained with Coomassie Blue. 22 at U i‘.‘ 'I a“ f" “'17”ng 50/ B D Figure l. 23 dalton small subunit. Thus it was possible that the major doublet observed in the 0- 30% (NH4)2SO4 fraction was not tubulin polypeptides but those of FlP. However, in no published report had it been shown that the large subunit of FlP migrated as a doublet under denaturing conditions. Furthermore, FlP usually precipitates between 30 and 50% fraction of leaf extracts (Wishnick and Lane, 1971). (NH4)ZSO4 Absorption Spectra of the (NH,)qSO‘ Fractions of Cblchicum Extracts CLC absorbs in the long-wavelength ultra-violet (UV) region, with absorbance peak at about 350 nm (molar extinction coefficinet at 350 nm = 1.585.x 104). When CLC is irradiated with long wave— length UV it is photoconverted into its derivatives, lumnicolchicine, with concomitant loss of absorbance at 350 nm (Fig. 2). To determine if endogenous CLC is bound to the major leaf protein the 0-30% (NH4)2804 fraction was dialyzed overnight, against 200 volumes of HB, and its absorption spectrum obtained before and after two hours of long—wavelength UV irradiation at 4°C. The dialyzed 0-30% (NH4)ZSO4 fraction absorbed in the long UV region, with a peak at about 350 nm, which was partly lost (ZS-50% depending on the preparation) on UV irradiation (similar to the absorption spectrum shown in Fig. 3). When the (NH4)ZSO4 fraction was boiled for 5 min and sedimented at 12,000.x gmax for 15 min to remove the precipitate, practically all A was recovered in the supernatant. This A 350 350 was also partially lost with long UV irradiation for two hours (Fig. 3) . When a CLC solution is irradiatedw'it’h W, 50% of 24 .AmCOHOAE may numcmao>m3 msmnm> pwuuon ma unwed uma0w>amuuas mcoH sufi3 :oflumavmuufl H50: 03» Houmm use mnemmn Uqu 2: Va mo A.Q.og >uwmcm© HMOfiumo .AUQUV msfloanoaoo mo muuowmm GoduQHOmnd .N musmwm 25 zkmzu4w><3 .m musmfib V «I c. c. c. 7.. AV :5 n8 7n Au 15 O 9 O 9 0 9 .x q q — uni: d (u — _,4— q _ — a - a a, _ — .>.D c... N35 0 1| [Nd 0.8 a. .0 I 1'00 O I. 19.0 .b..—....—r...—...-—p..._0.0 26 .Amsouowe say sumcmaw>m3 msmum> cwuuon ma unmfia umaofl> nmupan mcoH :uw3 sofiumwpmunw Hoop osu m Hmumm cam muomwn sofiuomum ucmumcuomsm on» NO A.o.ov >uwmcm© Hmoaumo .coHummzmwuusmo ha cmflmemHo cam moussws m>fim How mmafion was uomuuxm ESUNNUNQQ mo mumuwmflomum vOmNavmzv vmumusumw womno pmN>HMfiQ .muomuuxm Exow£0NOQ ca mocsomsoo msHQHOmnm E: own on» no muuommm cowumuomnd .m onsmflm 27 Ikozu .— w><3 C. C- 9 z 0 9 009 9 1.9 002 .m musmflm 0L2 qudJ—uddq‘uuqu—uu-u—dudfid I .>.:+ - p p p, — - p p p —, p p U- n — p, p p p — p - p .n . 1N6 Il.'nnv mu 0.0 0.0 28 absorbance at 350 nm (A350) is lost in about 22 min (Fig. 4). Therefore two hours of UV irradiation should be sufficient to cause loss of essentially all the A of the dialyzed (NH4)ZSO 350 4 fraction. In fact, a control solution of CLC with the initial A350 comparable to that of the (NH ) SO fraction, lost all its A 4 2 4 350 under the same conditions of irradiation (Fig. 2). The A350 was present in varying amounts in the crude extract and the dialyzed 0-30% and 30-50% (NI-I4)ZSO4 fractions of both leaf and flowers. This indicated that the A350 species was not stoichiometrically related to the major doublet, which occurred only in the leaves. Colchicine Binding Activity in the (NH‘)q§9‘ Fractions of Cchhicum Leaf Extracts CLC can be released from the tubulin-CLC complex by its photo- conversion to lumicolchicine with UV irradiation, at the cost of denaturing two-thirds of the CLC binding sites (Amrhein and Filner, 4)2504 fraction 1973). Therefore, tubulin in the dialyzed 0-30% (NH of Cchhicum leaf extract was assayed by 3H-CLC binding before and after UV irradiation (Table l). The poor CLC binding activity was not enhanced by prior UV irradiation. Inhibition of Colchicine Binding Activity of Brain Tubulin by Cblchicum Leaf Extracts Since extracts of Cblchicum leaves were expected to contain CLC, . . . . . 3 they could be expected to inhibit the binding of exogenous H-CLC. Therefore, experiments were performed to measure the degree of in- hibition of 3H-CLC binding activity of brain tubulin and the nature 29 .cowumwpmuufl uoaofi>muuaa mcoa npfiz cofiuwfipmuuw so ACME adv mEHu :ufl3 mcwoflcoaoo mo coflumuomnmusc 0mm on» NO xmomo .v musmwm 30 Amwpaziv 92:. ow on 0 .v muswflm _.O 31 TABLE 1. 3H-Colchicine Binding Activity in (NH4)ZSO4 Fractions of Colchicum Leaf Extracts. Protein Fraction CPM Bound/mg Protein/Hr 0-30% (N84)2804 Fraction 3611 " + 2 hrs of UV irradiation 1434 " + 0.1 mM CLC 0 30-50% (NH4)ZSO4 Fraction 1885 50 ul of the 0-30% and 30-50% (NH4)ZSO4 fractions were incubated with 100 111 of 3I-I-CLC (20 uCi/ml; 16.03 Ci/mmole) in a total volume of 500 pl of HB.. The 0-30% fraction was also assayed after two hours of UV irradiation in cold and in the presence of 0.1 mM CLC. The final concentration of CLC in the incubation mixture was 0.25 UM. 32 of the inhibitory factor. In the first experiment, Cblchicum leaves were homogenized in HE with 5 mM DTE and the homogenate sedimented at 40,000 x gmax for 60 min. The resultant supernatant fraction, 4OKSGO was divided into three portions, two of which were boiled for 5 min. One of the boiled 40KS60 aliquots was then sedimented at low speed to remove the precipitate and the supernatant collected. The unboiled 40K560, boiled 40K860, and the supernatant of boiled 40K860 were assayed for their inhibition of 3H-CLC binding activity of brain tubulin (Table 2). The binding of 3H-CLC to brain tubulin was completely inhibited by the Cchhicum extract, even after it was boiled. This heat-stable inhibitory factor could not be sedimented after boiling. To determine how much of the inhibitory factor was dialyzable and how much of it was complexed to a non-dialyzable material in the 0—60% (NH4)ZSO precipitate, Cchhicum leaves were homogenized in BB 4 containing 5 mM DTE. The homogenate was sedimented at 40,000 x g for 60 min. Aliquots of the resultant supernatant fraction, 40KS60, were either saved on ice or treated by one of two of the following ways: (i) boiled for 5 min and sedimented at low speed to remove the precipitate; the supernatant fraction was saved, and (ii) dialyzed against ZOO-fold excess of HB with 5 mM DTE for two hours. The remaining 40KS60 was made 60% saturated in (NH ) SO at 4°C. The 4 2 4 0-60% (NH4)ZSO precipitate was collected by sedimentation and 4 dialyzed against HB for four hours. An aliquot of the dialyzed 0—60% fraction was saved on ice and another boiled for 5 min and 33 TABLE 2. Inhibition of the CLC—Binding Activity of Brain Tubulin by Boiled Extract of Colchicum Leaf. Protein Sample CPM Bound/Hr % Inhibition per 100 pl of Incubation Mixture 1. Brain tubulin control 16,400 2. " " + unboiled 40KS60 126 99.99 3. " " + boiled 4OKS60 220 99.99 4. " " + supernatnat of boiled 40KS60 28° 99°98 25 ul of brain tubulin was incubated with 25 ul of 3H-CLC (20 UCi/ ml; 16.03 Ci/m mole) in the presence of 175 pl of HB or one of the differently treated portions of 40KS60. 3H-CLC binding was assayed by the filter assay using 03-32 resin, as described under Materials and Methods. 34 sedimented at low speed to remove the precipitate. The resultant supernatant was saved. Different amounts of the various fractions prepared above were assayed for their inhibition of the 3H-CLC binding activity of brain tubulin (Table 3). It is evident that the binding of 3PIE-CLC to brain tubulin is almost completely inhibited by both the crude extract (the 40,000 x gmax supernatant) and the supernatant of the boiled crude extract of Colchicum leaves. However, the inhibition is reduced to about 50% after dialysis of the crude extract for two hours. In other words, all inhibition was heat stable and at least 50% of the inhibition was due to a low molecular weight dialyzable factor. Similarly, the dialyzed 0-60% (NH precipitate and the supernatant of the 4’2504 boiled precipitate both.strongly inhibited the binding of the 3H-CLC to brain tubulin. The fact that the inhibitor precipitates with (NH4)ZSO4, survives dialysis for two hours and is released from the precipitate on boiling the fraction, suggests that it is tightly but not covalently bound to a high-molecular weight component that precipitates on boiling. It is likely that the inhibitor is CLC or a closely related compound. Estimate of Endogenous Colcicine in Cblchicum Extracts If we assume that the inhibition of the 3H-CLC binding activity of brain tubulin is entirely due to competition with endogenous CLC, then a rough estimate of CLC concentration in Colchicum leaf extract can be obtained by using a standard graph of percent inhibition of the binding of 3H-CLC to brain tubulin versus 3S TABLE 3. Inhibition of the 3H-CLC Binding Activity of Brain Tubulin by the (NH4)ZSO4 Fraction and Leaf Extract of Cblchicum Sample Added Sample Buffer CPM.Bound/Hr A % Volume Volume per 100 pl of Inhibition Incubation ( ul) ( ul) Mixture 1. Control 350 11,500 2. 40KS60 240 150 200 98 3. Dialyzed 40KS60 240 150 6,040 47 4. Supernatant of boiled 4oxseo 24° 15° 19° 98 5. Dialyzed 0-60% (NH4)ZSO4 Fraction 50 300 1,460 87 6. Supernatant of boiled dialyzed 0-60% (NH ) S0 Fraction 50 300 3'200 72 4 2 4 7. Same as #6 100 250 2,200 81 Different volumes of various fractions from Cchhicum extracts were incubated at 37°C with 50 pl of brain tubulin plus 50 ul of 3H-CLC (20 uCi/ml; 16.05 Cifinmole)in.a total volume of 450 pl in HE with 5 mM DTE. The 3H-CLC binding activity was assayed by the filter method using DE-32 resin. Final concentration of CLC in the incubation mixture was 0.14 uM. 36 concentration of unlabeled CLC, prepared under identical incubation conditions. Thus, the 98% inhibition of the CLC binding activity of brain tubulin by 240 pl of Cblchicum extract 40K860 (Table 3) is equivalent to the inhibition caused by about 36 nanomoles un- labeled CLC under identical incubation conditions (Fig. 5). This gives an estimate for the concentration of endogenous CLC in Cchhicum extracts of 0.15 mM. An alternate estimate of the concentration of the endogenous CLC in Cblchicum extract is based on absorbance at 350 nm (Table 4). If we assume all A350 to be due to CLC then, knowing the molar extinction coefficient of CLC at 350 nm (E = 1.585 x 104), the CLC concentration in the plant extract was estimated to be about 2 mM. However, the TLC analysis indicates that when the chloroform extract of the aqueous Cchhicum extract is run on TLC, four major spots of about the same intensity can be detected by long-wavelength UV, only one of which co—migrates with authentic CLC (Fig. 5B). Also only 25-50% of total A350 is UV-sensitive, whereas all of the A350 of CLC is UV sensitive. Therefore a more accurate estimate of the concentration of the endogenous CLC can be obtained if it is assumed that the absorbance at 350 nm that is lost on UV irradiation is due to CLC and related compounds (Table 4), approximately 50% being due to CLC. The concentration of CLC in Colchicum extract by this method is estimated to be about 0.5 mM. Based on the reported CLC content of leaves [0.1-0.6% of the dry weight and assuming a fresh weightzdry weight ratio of 15, the extracts prepared in the above study (i.e. in Table 4); where 3.0 m1 37 .COwumuucoocoo deflownoaoo uncammm mmuuon ma msflcwmsmu xuw>fiuom mcwmcfln mcflowonOUImm mo mmmusooumm one .m wanna on osmmma may ca Umnwuomwp anon» on Hmoflusmofl mcofl0fi©ooo Homo: mcwofinoaoonmm mo unseen cm>flm m suw3 omumndocw mm3 swasnou gamma 300 .ocfloflnoaoo UmHQOHGo an sflaonsu Gamma mo wufl>wuom mswoofln mowoflnoaoolmm man no soauflnflnsH .«m musmwm 38 l 8 I00 flaw/o wag,I [/1on/0 17 Mme-9 79w; mamas! (nmoles) COLCHICINE Figure 5A- 39 TABLE 1%. UV-Induced Loss of Absorbance at 350 nm in Extracts of Cchhicum Leaves. Fraction Dilution Absorbance at UV—Sensitive 350 nm Absorbance at 350 nm in Undiluted Sample* 1. 40,000 x g 80 0.41 supernatant 2. 40,000 x g supernatant + 80 0.21 16 2 hr UV-irradiation 3. boiled supernatant of #1 113 0.35 4. boiled supernatnat of #1 + 2 hr 134 0.16 18.11 UV-irradiation 5. CLC (0.5 mM in PMg buffer) 16 0.435 8 6° + 2 hr 16 -0.065 UV-irradiation *UV-sensitive absorbance at 350 nm was obtained by subtracting the absorbance at 350 nm after UV irradiation from the absorbance prior to UV irradiation. by the dilution factor, the UV sensitive A was obtained. When this absorbance difference was multiplied in undiluted sample Figure 5B: 40 Thin layer chromatography of Cchhicum extract and colchicine. Samples were chromatographed on a glass plate coated with silica gel. The silica gel plate was photographed under long ultraviolet light. Left track, chloroform extract of an aqueous extract of Colchicum leaves; middle track, colchicine; right track, benzene extract of the aqueous extract of Cchhicum leaves. 41 t 4 ).||.)(l||.|bl\.. Figure 5B. 42 or HB/g fr. wt. resulted in the final extract having 3.73 ml/g fr. wt.] would be expected to contain 0.04—0.25 mM CLC. Identification of the Protein Doublet in SDS-Gel after Electrophoresis of Colchicum Extract In order to determine whether the predominant doublet on SDS-PAGE is due to fraction-l-protein (FlP), tubulin or some other protein, the proteins in the 0-30% and 30-50% (NH4)ZSO4 fractions were analyzed on the basis of their native molecular weight both by sucrose density gradient centrifugation and gel filtration. Sucrose Density Gradient Centrifugation. Dialyzed samples of the 0-30% and the 30-50% (NH4)ZSO fractions of the soluble protein 4 from Colchicum leaves were centrifuged on a 5—20% sucrose density gradient at 27,000 rpm in a Beckman SW27 rotor for 20 hrs (Figs. 6a and 6b), seven hours (Fig. 8), six hours (Figs. 9a and 9b) and 12 hrs (Figs. 7a and 7b). A peak of absorbance at 280 nm (A280) that has the same 8 value as FlP is observed in the extracts of Cblchicum leaf protein (Figs. 7a, 8 and 9a). It was tentatively concluded to be the FlP of Cblchicum. However, in two preparations this peak was present predominantly in the 0—30% (NH4)2804 fraction (Figs. 8 and 9a) and another in the 30-50% (NH4)ZSO fraction (Fig. 7c). Such variation 4 could depend on the source and age of the plant leaves. In any case, most of the absorbance at 350 nm (A350) is not associated with this 188 peak (Fig. 7b and 9b). Instead it is present as a large trailing peak. FlP purified from spinach leaves did not absorb at 43 .mcon Ugo .mmumavm pom “macaw mIHIm .mwaouwo Ummoao meHIm msam Ugo .moaouwo ammo .usmwpmum on» no Eouuon .ommav 2: own am can flaw .mnm on» ma H .0: noduomnm .mmnommmfi Amp .mwm .ommdv a: 0mm um mmocmnuomnm pom omuowaaoo wumz mGOwuoon .muoos om How so» ooo.a~ um cwucmefiomm can can mafia manm no oqo .muHum sues ommmem>o mum3 musmflmmum muflmsoo omouoom womlm .AQQUV mcwowsoaoo psm.amaHlmv samuoumlchowuomum nomcwmm mo sowummoMfiuusmo powwomum hufimcmm omouosm "mm com me mmudmwm 44 ZO_._.U<¢u_ 20. PU<¢u .mo EB <6 833m 1N.O 1“ .o v V 8 Z S 8 0 O .46 $6 0.0 0.0 m < Figure 7A: 45 Sucrose density gradient centrifugation of spinach Fraction-I-protein (F-I-P) and dialyzed 0630% and 30- 50% saturated (NH'4)ZSO4 fractions of Cblchicum leaf extract. 5-20% sucrose density gradients were sedimented at 27,000 rpm for seven hours. The fractions were collected (fraction no. 1 is at the bottom) and their absorbances measured at 280 nm. Squares, the 0-30% fraction; open circles, the 30-50% fraction; and closed circles, F-I-P. 46 0.3‘ Figure 7A. Y r I 10 FRACTION Figure 7B: 47 Sucrose density gradient centrifugation of spinach Fraction-I-protein (F-I-P) and dialyzed 0-30% and 30-50% saturated (NH‘4)2S04 fractions of Cblchicum leaf extract. 5-20% sucrose density gradients were sedimented at 27,000 rpm for seven hours. The fractions were collected (fraction no. 1 is at the bottom) and their absorbances measured at 350 nm. Squares, the 0—30% fraction; open circles, the 30-50% fraction; and closed circles, F-I-P. A350 48 O. 6~ 0.2: Figure 7B. I T FRACTION Figure 8: 49 Sucrose density gradient centrifugation of spinach Fraction-I-protein (F-I-P) and dialyzed 0-30% and 30- 50% saturated (NH4)ZSO4 fractions of Colchicum leaf extract. 5-20% sucrose density gradients were sedimented at 27,000 rpm for seven hours. The fractions were collected (fraction no. 1 is at the bottom) and their absorbance measured at 280 (A2 80). Closed circles, F-I-P; open squares, 0-30% fraction; and open circles, 30-50% fraction. SO 0.1 0.61 0.4* 9A280 :Iz/z v—Fl—‘F 1O FRACTION UITfITI 15 .4 51 Figure 9A: Sucrose density gradient centrifugation of spinach Fraction-I-protein (F-I-P) and the dialyzed 0-30% saturated (NH4)2804 fraction of Cblchicum leaf extract. The 5-20% gradients were sedimented at 27,000 rpm for six hours. The fractions were collected and their absorbances measured at 280 nm. Closed circles, F-I-P and Open circles, 0-30% fraction. The arrow marks the position of F-I-P. 1. A A o 0.84 0.6: 0.44 0 fl . ‘EU 0.2‘ 52 Figure 9A. 10 FRACTION 53 .mIHIm mo :ofluwmom on» mxume souum 0:8 .c0wuomuu womno .moamsmfluu some can mIHIm .mmaouwo ommoau .Ed 0mm um communes moocmnHOmnm news» can mmuomaaoo wnm3 moofiwomuw one .muno: xwm How Emu ooo.b~ um pmusmeflpom onw3 muomflpmum womnm one v .uomuuxm mama EsowmoNos mo :ofluomum Ommawmzv owDMH:DMm womso powwamHo mou new AmIHnmv camuoumIHIGOADUMHm nomcwmm mo cowummsmwuucmo usofiomum zuwmswo mmou05m "mm musmflm 54 mmwmiaz 20Fo mmpsHocH we» mxume :H: .maouum xa pmxumfi who mCOADHmom swoop pom >Hmumummmm can some mum3 A390 cwaonou Gamma 300 com AmIHImV aflououmuHICOHuomum somcflmm .Ammaouwo oomoHov 8: 0mm pom Amoaouwo somov Ed omm um posefiuouoo mm3 cowuomum pmuomaaoo some mo A.Q.ov woomnu0moo one .cEsHoo mv mmoumsmmm m v cmsousu mousam mma uomuuxm mama ExowonQQ mo sawuomum Ommavmzv :oHumuoumm womno 039 .me mmoumnmmm co uomuuxm mama EsowsoNOD mnu mo codumuuaaw o>mwm umaoomaoz «OH munmflm 57 20:05": .3 856E 6.4 . o.» . ow. p 0.. LP 9 cu —> d'l-J —> .m... 'IOA pgon—> 58 as an affinity label for F—I-P, although.it has not been used as such until this report. Although 14C-CRBP is not available commercially, as described by Wishnick et al (1970), it is relatively easy to prepare from ribulose-l,S-bisphosphate and 14 Na CN by the cyanohydrin reaction followed by hydrolysis of the 0 nitrile. The ribitol is precipitated as its barium salt, dissolved in 0.1N HCl and converted to its potassium salt by passage over 4. Dowex 50(K ). . . . 14 Purified spinach FlP or C-CRBP-FlP complex (prepared as described under Materials and Methods) and a mixture of 3H—CLC- tubulin complex and free 3H-CLC were run through a Sepharose 68 column singly or together. As a control, 14C-CRBP was incubated with brain tubulin to show that tubulin does not bind 14C-CRBP (Table 5). FlP and tubulin peaks elute at close but distinct positions with K (=glution volume - void volume av total bed volume-void volume ) values of 0.33 and 0.48, respectively (Figs. 11, 12). Theoretically, 560,000 and 120,000 dalton globular proteins should elute with Kav of 0.3 and 0.5, respectively. However, because of possible variations in the packing of the gel bed, the true resolving power of Sepharose 6B was determined by running standard FlP and tubulin together (Figs. 13, 14). Surprisingly, when FlP and tubulin were run together they showed an apparent associative interaction. In one instance, the Kav of FlP was 0.28 and that for tubulin was 0.46, with the front of the tubulin peak markedly skewed towards the FlP peak (Fig. 13). In another instance, the FIP eluted as a double peak with Kav of 0.276 and 0.36, while most of the tubulin eluted with the leading peak of 59 TABLE 5. Binding of 14C-CRBP and 3H-CLC Binding to Brain Tubulin. Tube # Incubation Mixture Radioactivity Bound/50 ul of Incubation Mixture/75 min 3H-CPM 14OCH! 1 3H-CLC + tubulin 2341 -— l4 . 2 C-CRBP + tubulin - 0 20 ul of 3H-CLC (16.03 Ci/mmole; 6.8 x 107 cpm/ml) or l4C-CRBP (1.55 umole/ml; 2 x 105 cpm/ml) was added to 2.0 ml of l x polymerized MTs, diluted five times (N11 0M) and incubated at 37°C and room tempera- ture, respectively, for 75 min. The incubation mixtures were then sedimented at 40,000 x gmax for 60 min at 4°C to remove denatured protein. 50 pl aliquots were taken at 0 time and after the 40,000 x gmax spin and assayed for bound counts by the filter assay using DE-32 resin. Radioactivity bound was obtained after subtracting the counts bound at the 0 time. 60 .xme coquuOmnm 0:» mo cofiuwmom on» mxume sound odomflfi on» com Amowoanoaoo canons: mov QESHO> mmmnaosfl on» mxums .H. .E.m.u:mm .mmaouao pomoHo .A.Q.ov a: omm um moGMQHOmnm .mwaouflo some .mm mmouosmmm ow mcwowcoaoolmm pom :HmuoumaHloowpomum mo coaumuuawm m>ofim Hmaaomaoz .HH wasmfim 61 mi e.O|-X WdO zo_.—.Uwuomofiomu pom pmuomaaoo muw3 mcofluomum m .mw mmoumsmmm :H xoamsoo owHandulcwmunlmsfloflsoaoolmm mo :OAHMHUHHM w>mam Moasomaoz .mH musmnm 63 .9 -)'|OA plon r—F e '7‘ .L e_0|- X H d O 20 Figure 12. FRACTION 64 .osfiofinoaoonmm ossonca mo cofiuwmom map mxumE .H. .E.m.onmm .moaouflo oomoHo “5.0.00 a: 0mm um mononuomnm .moaouwo ammo .mm omoumnmmm cw :kuoum IHIoofluomum nomcflmm mom memEOo seasnduuswmunlmcflowsuaoolmm mo coflumuuaflm m>oflm Hmasomaoz .ma ousmflm 65 2.01 X Wd 0 .MH musmfim z O-PU<¢& h fin _ ._N p 9.— L09 ()0 .«J 66 F1P at Kav = 0.276 CFig. 14). Such variation in the degree of interaction could depend on the relative concentration of the two proteins. If F1P and tubulin do bind to each other rather tightly, say in a 1:1 stoichiometry, then a simple calculation can be done to predict the mobility of the binary complex. Thus, knowing the molecular weight of F1P and tubulin to be about 550,000 and 120,000 daltons, respectively and assuming that the F1P—tubulin complex behaves as a globular protein of 670,000 daltons, then the expected Kav of the complex can be determined (using a standard plot of Kav versus molecular weight, given in the booklet "Beaded Sepharose 2B-4B-GB.” Pharmacia Fine Chemicals). Thus for a binary complex of tubulin and F1P the theoretical Kav would be 0.27. The shifts in the observed Kav values of tubulin and F1P from 0.46 and 0.33, respectively, when run alone, to 0.276 when run together, are consistent with such an interaction. From the point of view of the author's specific goal, gel filtration was concluded to be unsuitable to resolve F1P and tubulin. Therefore, this otherwise interesting observation was not pursued. DE-32 Ion Exchange Chromatography The resolution of the native F1P and tubulin proteins was then attempted by ion-exchange chromatography, on a DE-32 column which was equilibrated with CB (25 mM MES, pH 7.0, 1 mM MgCl 0.1 mM GTP and 2’ 25% glycerol. l4C-CRBP-FlP and 3H-CLC-tubulin complexes were eluted first by C8 and then by a linear gradient of 0—0.8 M KCl in CB (Fig. 15). 14C-CRBP-FlP eluted at 0.15 M KCl and 3H-CLC-tubulin complex at 0.29 M KCl. This was the first successful resolution of 67 .mxmmm mocmnHOmnm on» xumfi mzouum mappflz .wcflownoaoo masonos mo :0wuflwom mxumfi .H. .Ec own an A.a.ov mocmnuomnm .moumsvm x.E.Q.OIUvH .moaouwo ammo x.E.m.onmm .mmflouwo pmmoHU .mo mmoumsmwm ow xoamsoo cwmuoumuHus0wuomumnwumzmmonmmwnnm.Hnaouflnwu Ifl>xonuooluvavm pom xoamaoo :flaonouucamunumcw0fl£OH00Imm mo sofiumuuaflm m>mfim Hoasomaoz .va madman 68 9.0LX was ZO_h0 Ion mm 2 .wa ousmflm .0... 69 .A80\mosfiou0flfi Gav >DH>Huooosoo camaoomm .moumovm x.E.m.oImm .mmaouwo pomoHo x.E.Q.UIUvH .mmaouwo ammo .A=m=v 0m .0: sowuomum um comma mm3 vowflpmum Homofia HUM z m.o I 0 one .memEoo :Haonsunswmunlmcflo“soaooumm pom wamEoo aflououmnHlaowuomum mumnmmosmmfinlm.Huaouwnflul“axonumono HVN mo xsmmumoumaouno omsmnoxmscofl mmoHDHHmolm¢mo v .mH musmflm 7O - l . -d you puoo s N J n: 20..—.0£mmumoumsouno mmsmnoxmlsofl mmOaflaamonmdmo .mH mucosa 76 20..—.0<¢u ,mx 'pUOO'ds ”T 2 mad 2 «No.0 . 0.0 .ma ousmflm e.OI>X "d0 00 77 F1P and tubulin and made it possible to ask if the predominant protein in the Cchhicum extract was F1P or tubulin. Zero to 30% and 30—50% (NH4)2504 fractions of the leaf soluble protein from Colchicum were passed through the DE-32 column with 3H-CLC-tubulin complex, and eluted first by CB without GTP and then by a linear gradient of 0—0.5 M KCl in the same buffer (Figs. 16, 17). In both 0-30% and 30-50% (N114)2804"2 fractions two major peaks of A280 were eluted. The first major A280 peak in both.cases eluted like free colchicine. However it lacked absorption at 350 nm and therefore was unrelated to the A350 absorbing species- The second major A 0 peak eluted at about 0.07 M KCl, well ahead of the 28 3H-CLC—tubulin complex (eluted at about 0.26 M). In fact, it eluted even before the spinach 14C-CRBP-FlP protein, which is known to elute at 0.15 M. In order to test if CRBP binding might have altered the elution behavior of F1P, purified spinach F1P was chromatographed without 14C-CRBP label but with 3H-CLC-tubulin complex (Fig. 18). In this case F1P eluted with 0.072 M KCl. The higher salt concentration requried to elute 14C-CRBP-FlP complex may be related to the negative charge of CRBP. Since eight moles of CRBP are expected to be bound to one mole of F1P this charge difference may be sufficient to alter the ionic interaction of F1P with the cationic DE-32. In any case, the resolution of F1P and tubulin only improved in the absence of CRBP. It was concluded that the major protein in Colchicum extracts is not tubulin but F1P. No A350 eluted from the DE-32 column though a visible yellow colored material was very strongly bound at the top of the column and may contain the A350 species. 78 In summary, extracts of Cblchicum leaves strongly inhibited the binding of 3H—CLC to brain tubulin. The concentration of the inhibitor, presumably endogenous CLC or a related compound, in the leaf extract was sufficiently high to prevent detection of tubulin in it by the 3H-CLC binding assay. Detection of tubulin in the extract by an alternate method, viz, mobility on SDS-PAGE, was hampered by the presence of a large amount of F-I—P, when a large subunit migrated as a doublet of polypeptides similar to that of tubulin. In the absence of a suitable assay for plant tubulin attempts to isolate tubulin from Cblchicum, a seasonal plant, were abandoned in favor of developing a general method to isolate plant tubulin. Nevertheless, this study led to three, and possibly four, findings of significance: a) development of a method to separate F-I-P and tubulin, which may be useful in future isolation of tubulin from green plant tissue; b) demonstration that the large subunit of F-I—P migrates as a doublet of polypeptides on SDS-PAGE; c) use of 14C-CRBP as an affinity label for F—I-P and d) preliminary observation of an apparent interaction between tubulin and F-I-P during gel filtration, which.needs to be reproduced to establish its significance, if any. SECTION II ISOLATION OF COLCHICINE BINDING PROTEIN FROM CULTURED TOBACCO CELLS SECTION II ISOLATION OF COLCHICINE BINDING PROTEIN FROM CULTURED TOBACCO CELLS Introduction Borisy and Taylor (1967) first used radioactively labeled colchicine to demonstrate that it was taken up by cultured mammalian cells and bound to a 68, 120,000 dalton soluble protein. Similar results were obtained in cultured grasshopper embryos (Wilson and Friedkin, 1967). This led Weisenberg et al (1968) to the purification of the colchicine binding protein from porcine brain extract by (NH4)ZSQ4 fractionation and DEAF-Sephadex chromatography. Since then, this classical fractionation method has been used in conjunction with assays of colchicine binding to purify tubulin from several animal tissues, such as thyroid (Bhattacharya and Wolff, 1974), liver (Patzelt et al, 1975), Islets of Langerhans (Montague et al, 1975), blood platelets (Puszkin et al, 1971), and fibroblast cells (Ostlund and Pastan, 1975). The specific affinity of colchicine for the microtubule subunits has not only allowed the isolation of the protein from cell extracts but also its identification and subsequent characterization, when isolated by other methods, such as by cycles of assembly—disassembly and solubilization of flagellar microtubules. 80 81 About one mole of colchicine binds to one mole of tubulin dimen. but apparently not to the intact microtubule (Wilson and Meza, 1973) or to the separated or and Bpsubunits of tubulin. CLC has a high but not absolute specificity for tubulin (Mizel and Wilson, 1972). The dissociation constant KD for the CLC-tubulin complex is 0.024 x 10-6 M, after correcting for denaturation of tubulin during incubation (Sherline et al, 1975). The high affinity of tubulin for colchicine has made it the most commonly used quantitative assay for tubulin (Weisenberg et al, 1968; Borisy, 1972; Wilson et al, 1974). Colchicine binding to tubulin from such diverse sources as brain and sea urchin sperm flagella was found to be qualitatively similar (Wilson et al, 1974). Rigorous analysis of the CLC-tubulin interaction, however, is complicated by the lability of the CLC- binding site, which decays rapidly in an apparent first-order manner (Weisenberg et al, 1968; Wilson, 1970). The half-life, Tl/Z’ of the decay of CLC binding activity varies with experimental conditions. Under optimal conditions the CLC binding activity of chick brain tubulin decays with Tl/2 of about four hours. The stability of the CLC binding activity is a function of the pH and exposure of tubulin to a pH of 4.5 or 10.5 for less than ten seconds irreversibly destroys all CLC binding activity (Wilson et al, 1974). The rate of decay is also dependent on the concentration of tubulin: Tl/Z for decay at 120.'min ina tubulin concentration of 20 ug/ml and T of 270 min at a tubulin concentration of 240 ug/ml (Bamburg 1/2 82 et al, 1973). The CLC binding activity can be stabilized with CLC or GTP (Weisenberg et al, 1968), vinca alkaloids (Wilson, 1970), sucrose (Cortese et al, 1977) or glycerol (P. Filner, unpublished observations). The decay of CLC binding activity is made even more critical by virtue of the fact that the rate of CLC binding reaction is slow. At low concentrations of colchicine, equilibrium is not reached for six to eight hours (Wilson et al, 1974). In the presence of a vinca alkaloid to abolish the decay, equilibrium was not reached for seven to eight hours (Wilsonet al, 1974). Finally, the temperature optimum for the binding of colchicine is not the same for tubulin isolated from different sources (Bryan, 1972). Because of these complications, quantitative comparison of the different parameters of colchicine binding activity in tubulin from different sources has not been possible. Nevertheless, the CLC binding property of tubulin appears to have been highly conserved in evolution,at least in animals. Although colchicine is the reagent of choice for detecting tubulin in animal systems, it does not work as well in plant systems. Colchicine certainly blocks mitosis and other micro- tubule-associated phenomena in plants, but millimolar concentrations are usually required, compared to the micromolar concentrations effective in animals (see Section I, Introduction). Although this may, in part, reflect uptake differences, there have been repeated observations of poor CLC binding activity (Hart and Sabnis, 1976b) by proteins in extracts of fungi (Haber et al, 1972; Davidse and Flach, 1977; Burns, 1973; Heath, 1975; Olson, 1973), an 83 alga Chlamydbmonaa (Burns, 1973), and higher plants (Hart and Sabnis, 1973, 1976a, b; Hotta and Shepard, 1973; Lescure and Filner, 1975; Rubin and Cousin, 1976). The following two reasons warranted further work in attempting to isolate tubulin from plant tissues with the aid of its CLC-binding activity: 1) Plant tubulin is expected to bind colchicine since CLC binding seems to be a universal property of tubulin; plant mitosis is blocked by colchicine and microtubules are disrupted by colchicine in viva, albeit at millimolar external concentrations (see Eigsti and Dustin, 1955). 2) Binding of radioactively labeled colchicine was the only readily available assay for tubulin. The poor CLC-binding activity in extracts of plant cells might be due to one or several of the following factors: 1) intrinsically poor affinity of plant tubulin for colchicine; 2) denaturation of the CLC-binding site during extraction and purification of tubulin; 3) low levels of tubulin in plant extracts; 4) masking of the specific CLC-binding activity of plant tubulin by a high level of non-specific CLC-binding; 5) presence of inhibitors of CLC-binding activity in plant extracts; 6) sub-optimal buffer conditions. Earlier work in this laboratory had shown that a CLC-binding protein could be obtained from cultured tobacco cells by the conventional biochemical method of Weisenberg et a1 (1968). It was composed of the usual two species of about 55,000 dalton subunits but only a small fraction of this protein had a detectable CLC- binding activity (Lescure and Filner, 1975). 84 The author proposed to investigate some of the above-mentioned factors that might be responsible for the poor CLC-binding activ- ities in tobacco XD suspension cultures in an effort to enhance the binding. Materials and Methods Preparation of Colchicine Binding Protein from Tobacco Cells Stock cultures of tobacco cells, Nicotiana tabacum L. cv. Xanthi, strain XD, were grown on MID medium as suspension culture at 28°C, as described previously (Filner, 1965). Exponentially- dividing cells are normally harvested after four days (ca. 4 g fr. wt./L) through two layers of Miracloth and excess liquid was expressed from the cells by twisting the Miracloth. Preparation of soluble CLC-binding protein was essentially by the method of Weisenberg et al (1968), as modified by Lescure and Filner (unpublished). All operations were carried out in cold. The tobacco cells were resuspended in phosphate buffer, PM,, 20 mM 1304-2, pH 6.8, 5 mM MgC12, 2 mM cysteine-HC1 and 0.1 mM GTP, at two to five volumes per gram fresh weight. The suspension was homogenized in 50 m1 portions by ten up and down strokes of a motor driven teflon-glass homogenizer. The homogenate was sedimented at 400,000 x gmax for 60 min at 4°C. The resultant supernatant fraction was made to 30% saturated (NH4)ZSO4, pH 6.8 at 4°C. After one hour the precipitate was collected by sedimentation at 20,000 x g for 20 min. The resultant pellet, 0-30% 85 precipitate,.was dissolved in a minimal volume of PM, and the super- natant fraction was made to 50% saturated (NH4)2804, pH 6.8. The 30—50% precipitate was collected as above and resuspended in the desired volume of PM. The 30-50% (NH4)ZSO4 fraction, resuspended in 0.2-0.7 ml of PM per initial 9 fr. wt. was used for further fractionation on an anion exchanger. Packed DEAE-Sephadex-A-SO, equilibrated in PM, was added to the 30-50% (NH4)2S0 resuspension at 0.1—0.4 ml per initial 4 9 fr. wt. After intermittent stirring for 30 min, the resin was sedimented in 40 ml glass centrifuge tubes in a swinging bucket rotor at 2500 rpm for 10 min. The supernatant was carefully removed and the resin washed two times each with PM, 0.4 M KCl in PM and 0.7 M KCl in PM. The washes were pooled separately, filtered through Whatman No. 1 filter to remove traces of resin and then made to 50% saturated (NH4)ZSO4,LHI6.8. After one hour, the precipitates were collected as above and dialyzed overnight against PM. Isolation of brain tubulin, colchicine binding assay, protein determination and radioactivity measurements were as described under Materials and Methods of Section I. Results Preparation of Colchicine Binding Fraction from Tobacco Cells Based on the methods of biochemical fractionation of tubulin from animal tissues, a procedure was divised to purify CLC-binding 86 protein from tobacco cells, as described under Materials and Methods. It involved fractionation of tobacco soluble proteins by (NH4)2504 precipitation and on the ion exchange resin, DEAE Sephadex. When the 30-50% saturated (NH4)ZSO4 fraction and the different fraction eluted off the DEAE-Sephadex A-50 column were assayed following dialysis, the CLC-binding activity was poor (Table 6). Colchicine Binding Activity of Brain Tubulin in the Presence of Tobacco Extract and Different Buffers The poor CLC-binding in the 30-50% (NH4)ZSO4 fractions of the extracts of tobacco cells might be due to traces of (NH4)2804 in these fractions resulting from incomplete dialysis. The effect of different amounts of (NH4)ZSO on the CLC-binding activity of brain 4 tubulin was determined (Table 7). CLC-binding activity of brain tubulin was not affected by (NH4)ZSO4 concentrations up to 0.1% and it was in fact enhanced by 1% (NH ) SO 4 2 4- was left over after dialysis, Therefore, if (NH4)2SO4 it was not likely to be responsible for the poor CLC-binding in the (NH4)2SO4 fractions of tobacco cell extracts. To determine if the extracts of tobacco cells inhibited CLC- binding activity of brain tubulin, four day old tobacco XD cells were harvested and divided into four equal portions. Each portion was homogenized in one of the following four buffers at 4°C (at two volumes/g fr. wt.): A PM buffer (20 mM po4‘2, 5 mM MgCl pH 6.8), without cysteine 2 and 0.1 mM GTP, B PM buffer, without cysteine + 5 mM dithioerythreitol 87 3 . . . . . . . TABLE 6L H—CLC Binding Act1v1ty in Different Fractions of Tobacco Extract CPM Bound/mg Protein/Hr 30-50% (NH4)ZSO4 fraction 1250 DEAE-Sephadex, unbound 525 DEAE-Sephadex, buffer wash 0 DEAE-Sephadex, 0.4 M KCl wash 325 DEAE-Sephadex, 0.7 M KCl wash 1700 Brain tubulin control 576630 30-50% (NH4)ZSO4 fraction, prepared as described under Materials and Methods, was fractionated on DEAE-Sephadex A-50 column by sequential 4 , 5 mM MgC12, and 0.1 mM GTP) containing, 0, 0.4, 0.7 M KCl. The protein in washes with PMgCGTP buffer (20 mM P0 2 mM cysteine different wash fractions were precipitated with 50% saturated (NH4)2 SO4 and dialyzed overnight resuspension in PMgCGTP. Incubation mixture consisted of 50 pl of 3H-colchicine (20 pCi/ml, 16.05 Ci/mmole), 50 111 of protein sample plus 400 111 of PMgCGTp (with 1 mM GTP). 100 ul aliquots were assayed at various time periods using DE-32 resin. Brain tubulin was isolated by l x polymerization. CPM bound was calculated after subtracting the 0 time 3H-CLC binding. 88 TABLE 7. Effect of (NH4)280 on 3H-CLC Binding to Brain Tubulin. 4 Incubation Mixture CPM Bound/100 pl Aliquot/Hr % Control Brain tubulin control 34,400 100 in 0.01% (NH4)2504 37,000 108 " in 0.1% " 37,400 109 " in 1.0% " 51,000 148 50 p1 of brain tubulin was incubated with 50 ul of 3H-CLC (16.05 Ci/ Inmole; 20 uCi/ml) and 400 1 of PMgCGTP buffer (20 mM PO4-2, 2 mM cysteine HC1; and 0.1 mM GTP, pH 6.8) containing the 5 mM MgC12; different amounts of (NH4)ZSO4. bound 3H-CLC at 0 and 60 min. 100 pl aliquots were assayed for 89 C PM buffer D PM buffer + 1 mM phenylmethyl sulfonyl fluoride (phenylmethyl sulfonyl fluoride is suggested to be and inhibitor of serine proteases, such as trypsin and chymotrypsin). After sedimenting the homogenate at 40,000 x gmax for 60 min, the resultant supernatant fraction, 40K560 was assayed for its effect on the CLC-binding activity of brain tubulin. The CLC-binding activity of brain tubulin is similar in all the three buffers tested. However, it is inhibited to varying extents by tobacco extracts prepared in the different buffers. If we take the colchicine binding activity of brain tubulin in the presence of tobacco extract that was prepared in PM buffer containing 5 mM dithioerythritol, as 100%, then the inhibition of the CLC-binding activity of brain tubulin by tobacco extracts prepared in other buffers varied from 62% (for the extract prepared in PM, without a reducing agent) to 22% (for the extract prepared in PM buffer containing 2 mM cysteine-HCl). Thus, reducing agents, especially dithioerythritol were effective in overcoming the inhibition of tobacco cell extract (since they themselves have little effect on the CLC-binding activity of brain tubulin). Also phenylmethylsulfonyl fluoride, an inhibitor of serine proteases is without effect on the CLC-binding activity of either the tobacco extract or brain tubulin in the presence of the tobacco extract prepared in PM buffer containing 2 mM cysteine-HCl. The CLC-binding activity in the tobacco extract prepared in PM buffer containing 5 mM dithioerythritol was not enhanced (Table 8). Since the poor 90 TABLE 8. Effect of Tobacco Extracts in Different Buffers on the 3 . . . . . . H-CLC Binding ActiVity of Brain Tubulin. Tube # Protein Sample in Incubation Mixture CPM-Bound/30 Min per 100 pl of Incubation Mixture 1 Brain tubulin + 40KS60 (in buffer A) 144 2 Brain tubulin + 40KS60 (in buffer B) 377 3 Brain tubulin + 40K860 (in buffer C) 295 4 Brain tubulin + 40KS60 (in buffer D) 264 5 40KS60 (in buffer B) 65 6 40KS60 (in buffer D) 48 7 Brain tubulin (in buffer A) 678 8 Brain tubulin (in buffer B) 606 9 Brain tubulin (in buffer C) 597 Each supernatant, 40K860, was made 1 mM in GTP. For tubes #1 to 4, 430 pl of each 18KS60 was incubated with 50 pl of 3H-CLC (16.05 Ci/ Inmolen 20 pCi/ml) and 19.6 p1 of brain tubulin. For tubes #5 and 6, 450 pl of 40KS60 in buffers B and D were incubated alone with 50 m1 3H-CLC. For the control tubes, 50 p1 of brain tubulin was incubated with 50 pl of 3H-CLC and 400 pl of buffers A, B, or C, all made 1 mM in GTP. Incubations were alone at 37°C for 30 min. 100 pl aliquots were assayed for bound 3H-CLC at 0 time and after 30 min using the DE-32 resin filter assay. 91 CLC-binding activity in the crude extracts of tobaCCo cells.may be due to the low concentration of protein, these studies.were repeated on the 0-60% saturated (NH4)ZSO fractions, after dialyzing each 4 (NH4)ZSO4 fraction against the same buffer used for its extraction. However there was no enhanced CLC-binding activity in the presence of any medium. Therefore, there was poor colchicine binding activity in (NH.4)ZSO4 fractions of tobacco cell extract under conditions which are optimal for the colchicine binding activity of brain tubulin. Thus either the tubulin in tobacco extracts has an intrinsically low CLC-binding activity or most of it is so denatured by the time the protein is concentrated that it shows insignificant CLC-binding activity. Sucrose has been shown to stabilize CLC—binding activity of brain tubulin (Cortese et al, 1977). However, the 0-30% and 30-50% (NH4)ZSO4 fractions of tobacco cell extract, prepared in the presence of l M sucrose did not show enhanced CLC-binding activity. Colchicine Binding Activity in Other Plants The search for CLC-binding activity was concentrated primarily on tobacco XD cells because of the convenience of working with cultured cells in terms of available material throughout the year and ease of experimental manipulation. However, certain plant tissues are known from ultrastructure studies to be rich in MTs, for example, radish root hairs (Newcomb and Bonnet, 1965), coleoptiles (Heyn, 1972), and the chrysophycean alga, OchromonaS'maZhamensis (Bouck and Brown, 1973). Attempts were made to isolate CLC—binding 92 activity from all of these sources without any success. Since cauliflower heads at an early stage of growth.are expect- ed to be rich in meristematic tissue, an attempt was made to detect CLC-binding activity in different (NH4)2SO4 fractions of cauliflower extracts. Top 1-2 cm portions of fresh cauliflower Bmssica oZeracea var. botrytis) inflorescence obtained from the Horticulture farm, MSU, were frozen in liquid N2 and then the frozen pieces were resuspended in two volumes of the homogenizing medium: 20 mM sodium phosphate, pH 6.8, 5 mM MgC12, 0.1 mM GTP, 5 mM DTE, 1 mM EDTA and 25 mM sucrose per g fr. wt. The suspension was homogenized in a Waring blender (45 sec at high speed). The homogenate was sedimented at 40,000 x gmax for 30 min, and the resulting supernatant fraction, 18K330, used to prepare 0-30% and 30-50% saturated (NH4)ZSO fractions, as described for tobacco 4 cells under Materials and Methods. CLC-binding activity in the total homogenate and the dialyzed 0-30% and 30-50% fractions was determined (Table 9). Significant binding was observed in both these fractions. The above experiment with cauliflower head was also carried out in a slightly different homogenizing buffer: (20 mM sodium phosphate, pH 6.8, 5 mM MgCl 5 mM cysteine HC1 and 0.1 mM GTP). The 2. homogenate was sedimented at 40,000 x gmax for 60 min and the resultant supernatant used to prepare 0-40% and 40-60% saturated (NH4)ZSO4 fractions, as described above. CLC-binding activity in the dialyzed fractions was determined (Table 10), using the following incubation mixtures: 93 .HE\mE m.w u :ofluomum uHmm «onion 0:0 HB\mB o.H u COAUOMHM uamm womno “wcowunhucmocoo :Hmuoum .040 :5 m.vv HO H5\flon hw.o "muauxfia s0aumnaosw on» as mCOWHoHucooooo awash .momflo uwuafim Hmumo xo mHo>umuofl ode om mod 0 us Ugo mason Mom owmummd mums musuxwa cofluoumumxfiwmo muosvflao H: OOH .ummmsn on» «o an own maoo n0fi3 non m>ond mm pouunsocw mos ouocmmoson Houou mo a: 00H “oofluonum womnom no soHuooum acmno no a: ov moad.¢xdsas\fio mo.mH ads\ao: on. oqonmm mo H: on made Amen as m .meo :5 H .maomz as m .m.m mm .m won as one “woman he an oov ”mouspxas cofluonsocHa . . v N v me 0 0mm m oofluomum 0m A mzv woouom . . v N v on o omm «a casuomum om A mzv womno Hm.o oaa.o mumammoeom amuoe Hm\cwmuoum ma\ossom Ugo mace m «Hm\swmuoum mE\p:som Emu emz .mmouosm ca muomuuxm Hoseawaaomu cw mua>wuo¢ mcflooam qunm .m mamas m 94 .21 NH.o mo musuxfia_cowunnsocH :w 040 no cofluowucoosoo Hocwm .oowuomuw AVOmmAvmzv wooaog e omusfluao ca Haxms H.m can nonhuman omNAemzv movuo omusamao ca Ha\me «v.A "mcoaumuucmoaoo camuoum .momflp Hayden amino moan: acmmo umuHfim on» an spousamsp cw .swfi cm com o no chMmmm mumz muoswwam a: om omm omm.a a m wee 2: mm + cofluunuw Om A $20 womlog .v . . v N v . mm o moo NH sofluooum cm A :20 woouov m omm mam.v v memo 2: mm + COAuonum om A mz. wovuo .m . . v N v . 8H H mmb ma cOAuonHm Om A mzv wovno H :Houowm ma :wwuoum ms “om :Hououm ms Mom \OAU maofi m venom Emu owmaommm Hm\o:som Zoo umz mamadm .muomuuxm H030amaaooo mo xufl>flu04 mcfloswm OAOImm .OH mamme 95 Volume (pl) Added sample 1 2 ' 3 4 ~ 5 6 brain tubulin 25 25 — - - — 0-40% (NI-145504 - _ 100 100 _ _ fraction 40-60% (NH ) SO 9 4 2 fraction H-CLC (20 pCi/ml) 16.03 Ci/m mole 1 mM CLC - - - 20 - 20 homogenizing buffer 180 80 105 85 105 85 Total Volume (pl) 225 225 225 225 225 225 4 - 100 - - 100 100 3 20 20 20 20 20 20 There was significant binding of 3H-colchicine in the (NH4)2SO4 fractions of cauliflower inflorescence extracts, even in the presence of 89 pM unlabeled colchicine. In the presence of unlabeled colchicine about 250 pmoles of colchicine was bound per mg protein. In the absence of excess unlabeled colchicine roughly 1 pmole of colchicine was bound per mg protein. This may be related to the low amounts of colchicine in undiluted stock of radioactive colchicine, such that the binding occurred at colchicine concentrations much below saturation. However, in another experiment, poor colchicine binding was observed in a different preparation of cauliflower extract prepared from cauliflower obtained at a local store. The age and the physiological state of the cauliflower appear to be very important in regard to the CLC-binding. Unfortunately, this work on cauliflower was begun as a last resort too late in the growing season of 1978 to follow up. 96 Binding of 3',5'-Cyclic Adenosine Monophosphate to Brain Tubulin It has been reported that 3',5'-cyc1ic adenosine monophosphate (cAMP) binds to tubulin from blood platelets CSteiner, 1978). If cAMP did in fact bind to tubulin it might be used as an alternative to CLC-binding in assaying for plant tubulin. Cow brain tubulin was purified by two cycles of assembly- disassembly following the method of Filner and Behnke (1973). 3H-cyclic AMP and 3H-colchicine binding activities were determined (Table 11). No binding of cAMP was detected even when the highest available specific radioactivity of cyclic AMP was used. In addition unlabeled cAMP did not compete with CLC-binding activity of the brain tubulin. No binding of cyclic-AMP was detected even after purifying cow brain tubulin following Steiner's procedure for blood platelets. It is concluded that brain tubulin does not bind anycAMP. Discussion As mentioned in the Introduction, plant cells require much higher concentrations of colchicine than do animal cells to inhibit mitosis. For example, 10"4 M colchicine is required to block mitosis in cultured Haemanthus endosperm cells (Hepler and Jackson. 1969), whereas 10-7 M colchicine can block mitosis in HeLa cells (Taylor, 1965). The lowest concentration of colchicine reported to inhibit mitosis in plants if 1.25 x 10“5 M (Jeffs and Northcote. 1967), and even this is an order of magnitude greater than the levels 97 TABLE 111. Cyclic-Adenosinemonophosphate and Colchicine Binding Activities of Brain Tubulin. Incubation Mixture CPM Bound/Hr/SO pl Aliquot Brain tubulin + 3H-cA‘MP 0 Brain tubulin + 3H-cAMP + 1‘mM cAMP 0 Brain tubulin + 3H-cAMP + 0.1 mM CLC 0 Brain tubulin + 9230 Brain tubulin + 1.0 mM cAMP 9300 50 pl of twice polymerized brain tubulin was incubated in a total volume of 450 p1 as follows: a) with 5.56 nM 3H-cyclic-adenosinemonOphosphate (cAMP). b) same as in a) plus 1 mM unlabelled cAMP. c) same as in A0 plus 0.1 mM unlabelled colchicine (CLC). d) with 0.2 pCi of 3H-CLC at 16.05 m Ci/pmole. e) same as in d) plus 1 mM unlabelled cAMP. Incubations were made in 20 mM PO4—2, 5 mM MgC12, pH 6.8 at 37°C for various time intervals. 50 p1 aliquots were assayed for bound radioactivity by the filter assay using DEe-81 filter disks. 98 used to block mitosis in animal cells. In callus cultures of Nficotina 12.5 to 25 mM colchicine is required to disrupt 70-80% microtubules (Wooding, 1969). Hotta and Shepard (1973) demonstrated CLC-binding activity in the nuclear membrane fraction and the cytoplasmic fraction of extracts of Lilium anthers prepared in the presence of 10'.5 M vinblastine and sucrose. While, trypsin abolished all particulate binding, various detergents solubilized about 70-80% of it. The molecular weights of the soluble and particulate binding components was 110,000 and 100,000 daltons, respectively. Although vinblastine stabilized both the soluble and particulate binding it precipitated only the soluble binding. In a series of papers Hart and Sabnis (1973, 1976a, b) have investigated CLC-binding activity in higher plants. In the first report (Hart and Sabnis, 1973) they reported CLC-binding activity in 30-50% (NH4)ZSO4 fraction of Heracleum extract. The binding was very labile and the lability of the binding was not stabilized by vinblastine or GTP. In later studies they showed that the crude extracts of various higher plants (Heracleum, Myrrhis, pea and mustard) lacked CLC-binding activity and inhibited CLC-binding activity of brain tubulin to varying extents (4-5% inhibition). 4)ZSO4 fractions of Heracleum extract and in the DEAE-Sephadex eluates between 0.5 and 0.8 m KCl CLC-binding was detected in 30-50% (NH of this (NH4)ZSO4 fraction. However, colchicine was shown to bind even to denatured proteins and lumi-derivatives of both colchicine and colcemid could bind to the 30-50% (NH4)ZSO precipitate of the 4 99 plant extract. They concluded that although a specific CLC-binding site may be present in plant extract due to plant tubulin, the presence of a variety of other nonspecific binding sites for colchicine makes it difficult to identify tubulin by CLC-binding activity. They concluded that "...previous work on colchicine binding may need reevaluation in terms of the degree of specificity and component involved." On the other hand, Burns (1973) failed to detect any CLC-binding activity in extracts of peas and corn. The apparent tolerance of plant microtubules to relatively high levels of exogenous colchicine appears to be a general plant character. Although this may, in part, reflect uptake differences, the poor CLC-binding activity of plant extracts, reported above and confirmed in this study, suggests that plant tubulin may also have a lower affinity to colchicine. Lower plants also exhibit high tolerance to colchicine: Thus, 12.5 to 25 mM colchicine is required to disrupt the cytoplasmic microtubules of the chrysophycean alga, Ochromonas (Bouck and Brown, 1973). About 3 mM colchicine is required to inhibit flagellar regeneration in Ciamydomonas (Rosenbaum et al, 1969). Burns (1973) found no CLC-binding activity in the extracts of Clamydomonas. Neither colchicine nor colcemid (N-methyl N-deacetyl-colchicine) inhibited growth of several species of yeasts (Lederberg and Stetten, 1970). A protein has been isolated from Saccharomyces cerevisiae which binds colcemid ten times more effectively than it does colchicine; it has been suggested that this is the tubulin of these 100 colchicine-insensitive cells (Haber et al, 1972). However more recently other workers (Baum et al, 1978) have questioned the conclusion. Although they too found a colcemid-binding activity, it was excluded both on Sephadex G-200 and Sepharose CL-6B and was insensitive to pronase. They demonstrated that lumi-colcemid, which does not bind to tubulin but does bind to membranes, also binds to the CLC-binding fraction. It was concluded that the colcemid binding was not to tubulin but to a membrane fraction and that yeast extracts had little or no affinity for either colchicine or colcemid (Baum et al, 1978). The phycomycete AZZomyces is Colchicine-sensitive but its spindle microtubules apparently are not. A trichloroacetic acid stable CLC-binding protein of about 30,000 daltons has been isolated from this fungus (Olson, 1972). The microtubules of the water mold Baprolegnia are resistant to several anti-microtubule agents, including near lethal dose of colchicine. Two components of Shprolegnia bind colchicine and colcemid, one of these components is trichloroacetic acid stable (Heath, 1975). The identity of the radiolabel that was bound to the trichloroacetic acid stable component in the above two studies was not determined. An impurity in 3H-colchicine solution was shown to bind to a perchloric acid stable component isolated from Aspergillus nidfilans (Davidse and Flach, 1977). Although CLC-binding has been reported in extracts of different fungi the binding component is not characterized well enough to be identified. No CLC-binding was found in SChizosaccharomyces pambe (Burns, 1973), nor in Physarum polycephalum (Jockusch, 1973). Davidse and Flach 101 (1977) have carried out a much more thorough characterization of putative tubulin from Aspergillus nidhlans. A benzimidazole derivative, methyl benzimidazol-Z-yl carbamate or MBC, was shown to bind to a protein of 110,000 dalton which possessed some properties characteristic of tubulin. The binding was stabilized with glycerol, sucrose and benomyl. It was competitively inhibited by another benzimidazole derivative, oncodazole, and by colchicine. Colchicine was also shown to bind to the protein, though very poorly. The CLC-binding was unusual in being rapid and temperature independent. MBC did not affect brain tubulin assembly. However, oncodazole, the related derivative and a strong competitive inhibitor of MBC binding to the putative tubulin from yeast, strongly arrested mammalian mitosis, binds to brain tubulin at the CLC-binding site and inhibits its in vitro assembly (Hoebeke et al, 1976). MBC was also shown to selectively disrupt the microtubules of parasitic worms without affecting those of the host (Borgers et al, 1975). There is preliminary evidence that yeast tubulin also binds MBC. If the MBC binding protein is indeed the fungal tubulin it represents a major difference between tubulins from fungi and animal sources and may explain the resistance of fungi to colchicine. Resistance to high levels of colchicine is not unique to plants. Ciliary regeneration and cell division of Tetrahymena is known to be resistant to high concentrations of colchicine (Rosenbaum and Carlson, 1969; Williams and Jackel-Williams, 1976). No CLC-binding was detected in extracts of Tetrahymena (Burns, 1973). More recently, Maekawa (1978) has characterized the binding of 102 colchicine to cytoplasmic tubulin from Tetrahymena. He reported the dissociation constant (determined kinetically) to be 2.7 x 10-3. This is ten thousand times greater than the dissociation constant determined for brain tubulin. Thus, there exists in nature a tubulin with one ten-thousandth the affinity for colchicine that has been measured for brain tubulin. This example tends to refute the hypothesis, made in the Introduction to Section I, that CLC-binding site is as conserved as the sites required for microtubule assembly. If plant tubulin also has such a low affinity for colchicine then this fact would explain the lack of good CLC-binding activity in plant extracts. SECTION III ISOLATION OF POLYMERIZED TOBACCO TUBULIN BY DIFFERENTIAL CENTRIFUGATION SECTION III ISOLATION OF POLYMERIZED TOBACCO TUBULIN BY DIFFERENTIAL CENTRIFUGATION Introduction There are two obvious methods of choice that come to mind for isolation of any structural protein: a) isolation of the intact structure by differential centrifugation after its stabilization in vivo; and b) in vitro assembly of the structure from the soluble subunits in the extract, followed by its collection by differential centrifugation. These methods of isolation are convenient, rapid and, in the case of the labile tubulin protein of animal tissues, result in higher yields of the native protein than by the conventional method of biochemical fractionation (Ikeda and Steiner, 1976). In vitro Self-Assembly of Microtubules Neuronal Tissues. Weisenberg (1972) was first to demonstrate cell-free assembly of microtubules in an extract of rat brain that was incubated at 37°C in the presence of GTP and EGTA, a calcium chelating agent. Within one year of this discovery, Shelanski at al (1973) applied it to the isolation of brain tubulin. Since then, this method, or one of its several variants, based on cycles of assembly-disassembly, has virtually replaced the conventional 104 105 protein fractionation method (which employs (NH4)ZSO4 fractionation and DEAE-Sephadex chromatography) for isolating brain tubulin (Weisenberg et al, 1968). Whereas the tubulin that is isolated by the protein fraction- ation method is 99% pure on SDS gel electrophoresis, that isolated by the cycle procedure is 75-90% pure, the remainder being composed of 15-30 non-tubulin proteins, collectively termed microtubule- associated proteins (or MAPs), that co-purify with tubulin through several cycles of assembly-disassembly (Timasheff, 1979; Sloboda et al, 1976; Cleveland et al, 1977; Borisy et al, 1975). Besides differeing in their purity, tubulin prepared by these two methods also differ in some features of self-assembly. For instance, tubulin purified by the cycle procedure has a critical tubulin concentration (that is, the minimum concentration of tubulin required for self-assembly), C0 of about 0.1-0.2 mg/ml (Olmsted et al, 1974; Gaskin et al, 1974; and Burns and Pollard, 1974). On the other hand, pure tubulin purified by DEAE-Sephadex or phospho- cellulose chromatography has a Co of about 6-8 mg/ml (Himes et al, 1977). However the Co for pure tubulin can be reduced, in the presence of 10 mM Mgz+, to 2.5 mg/ml (Herzog and Weber, 1977), in the presence of 10-16 mM Mg2+ and 31% (v/v) glycerol, to 0.5-1.2 mg/ml (Lee and Timasheff, 1975; Gaskin et al, 1974; Lee and Timasheff, 1977), in the presence of 10% DMSO, to 0.8 mg/ml (Himes et al, 1977) and in the presence of 7.5-10% Dextran T10 or 2-4.5% poly(ethy1ene glycol) type 6000, to as low as 0.25 mg/ml (Herzog and Weber, 1978). 106 The low Co values for tubulin purified by the cycle procedure have been attributed to the presence of MAPS. MAPs, unlike tubulin, are polycationic and are removed from tubulin on ion exchange chromatography. Addition of the MAP fraction to pure tubulin enhances MT assembly (Weingarten et al, 1975; Sloboda et al, 1976; Whitmanet al, 1976). Addition of other basic proteins, such as histone, RNAse A and poly-L-lysine and also non-biological polycations, such as DEAE-Dextran, can substitute for the MAP fraction in facilitating self-assembly of pure tubulin (Erickson and VOter, 1976). There are two important points to bear in mind in extending these findings to the isolation of tubulin from other tissues: first, virtually all this work has been based on tubulin isolated from neuronal tissue, in which tubulin constitutes ca 20% of the total soluble protein, and second, these in vitro studies are based on highly purified tubulin and that too, at relatively high concentrations. Non-Neuronal Cells and Tissues. Reports of successful in vitro polymerization of microtubules from extracts of tissues other than brain have been relatively meager. Spontaneous in vitro assembly of cytoplasmic tubulin has been reported in extracts of blood platelets (Castle and Crawford, 1975, 1977; Ikeda and Steiner, 1976), renal medulla (Barnes et al, 1975), bovine anterior pituitary (Sheterline and Schofield, 1975), animal cell cultures (Wiche and Cole, 1976; Wiche et al, 1977; Nagle et al, 1977 and Fuller et al, 1975). On the other hand spontaneous assembly of tubulin in 107 extracts of other non-neuronal tissues, including Chinese hamster ovary, HeLa, neuroblastoma, mouse ascite cell lines, under conditions that are optimal for brain tubulin assembly, were unsuccessful (Farrell and Burns, 1975; Bryan, 1975; Rebhun et al, 1975; Bryan et al, 1975; Kane, 1975, 1976; Wiche and Cole, 1976). One major reason for limited success in cell-free assembly of MTs in the extracts of most non-neuronal tissues may be their low concentration of tubulin. Thus the in vitro assembly of MTs was possible in extracts of blood platelets, whiCh, next to brain, are probably one of the richest sources of tubulin (Castle and Crawford, 1977), in renal medullary extracts after concentration of the total protein by ultrafiltration (Barnes et al, 1975) and in Ehrlich ascite tumor cells when the protein concentration was higher than 20 mg/ml in the extracts (Doenges et al, 1977). However, it is not sufficient to have the concentration of tubulin higher than the critical concentration required for in vitro assembly. In some cases, the lack of self-assembly of MTs has been attributed to a heat-stable, non-dialyzable inhibitary factor that was sensitive to RNAases (Bryan, 1975; Bryan et al, 1975). It was further shown that RNA, synthetic polynucleotides and other polyanions inhibited the spontaneous assembly of purified brain tubulin, by removing from the tubulin solution a heat-stable basic protein, possibly a MAP fraction (Bryan et al, 1975). The poor spontaneous assembly of MTs in extracts of a variety of cultured cells that contained tubulin at two to five times the critical concentration required for assembly of brain tubulin, was shown to 108 be due to the presence of inhibitors in the extracts. The efficiency of the inhibition was attenuated by the addition of glycerol (Nagle et al, 1977). In at least one instance, the failure to isolate tubulin from an animal cell'culture was shown to be due,run:to the lack of in vitro polymerization, but to resistance of these MTs to subsequent depolymerization (Wiche and Cole, 1976). Chlamydbmonas extracts not only failed to form MTs under a variety of cell-free conditions (even when the concentration of tubulin was around 5 mg/mD but also inhibited the assembly of MTs in gerbil brain extracts (Farrel and Burns, 1975). It was suggested that Chlamydbmonas tubulin itself was the inhibitory factor and that Chlamydbmonas tubulin requires to be modified prior to its assembly into MTs. Flagellar and Ciliary Axonemes Flagella and cilia, like brain, are rich in microtubules and repeated attempts had been made for their in vitro polymerization even before the discovery of the conditions for in vitro assembly of brain tubulin (Stephens, 1968; see Stephens and Edds, 1976). However, until recently, normal MT assembly was not observed with tubulin solubilized from the stable axonemal MTs. This failure has been attributed to the denaturation of the labile tubulin by the harsh treatments (use of detergents, organic mecurials, thermal melting, high pH and acetone extractions) required to stabilize tubulin from these stable MTs. Recently, Kuriyama (1976) used sonication to partially solubilize tubulin from the outer-fiber doublets that was able to self assemble into MTs. The in vitro reassembly of tubulin derived from the stable outer doublet 109 microtubules has been demonstrated in sea urchin sperm flagella (Kuriyama, 1976; Binder and Rosenbaum, 1973; Farrell and.Wilson, 1978, 1979), TBtrahymena cilia (Kuriyama, 1976) and also Chlamydbmonas (Binder and Rosenbaum, 1978). The overall similarities between outer doublet and brain tubulin reassembly are very remarkable: the polymerization of outer doublet tubulin is optimal at 37°C, dependent on GTP, cold- and CLC-sensitive, and the critical tubulin concentration required for assembly is around 0.6 mg/ml which could be decreased in the presence of small amounts of either outer-fiber fragments or MAP fraction obtained from brain MTs. In other words these MTs behave like labile MTs. Heterologously Primed Assembly Polymerization of tubulin from one organism onto pre-existing MTs or a MT-organizing sites (such as basal bodies) from another, i.e. heterologous, primed assembly, has been demonstrated by the assembly of brain tubulin onto flagellar axonemes of Chlamydomonas (Allen and Borisy, 1974; Snell et al, 1974; Binder and Rosenbaum, 1973) and sea urchin sperm (Binder and Rosenbaum, 1973; Binder et al, 1975; Kuriyama, 197 ), and of soluble tubulin from Tetrahymena onto outer-fiber doublets of Tatrahymena cilia (Maekawa and Sakai, 1978). Homogenates of unfertilized eggs of Spisula would not form MTs unless provided with centriole-like material from fertilized eggs (Weisenberg and Rosenfeld, 1975). Basal bodies purified from Chlamydbmonas or Tetrahymena would induce aster formation in Xenopus eggs (Heidemann and Kirschner, 1975). Crude extracts of sea urchin 110 eggs failed to form.MTs unless provided with.brain MTS (Burns and Starling, 1974). Brain tubulin has been used to stabilize isolated mitotic spindles (Rebhun et al, 1974; Inoué et al, 1974; Cande et al, 1974). On the other hand, cytoplasmic tubulin from Chlamydbmonas purified by colchicine-affinity chromatography failed to polymerize onto added brain MTs (Farrell and Burns, 1975). Copolymerization Copolymerization with exogenous brain tubulin has been used to isolate tubulin from radioactively labeled CHO cell line (Spiegelman et al, 1977) and to detect tubulin in labeled products of in vitro translation of mRNA isolated from deciliated sea urchin embryos (Kleinsmith et al, 1978) and deflagellated Chlamydomonas cells (Weeks and Collis, 1976). Copolymerization has also been used to isolate radiolabeled tubulin from yeast cells (Sheir-Neiss et al, Water and Kleinsmith, 1976; Baum et al, 1978). This review of literature on assembly, heterologous, primed assembly and copolymerization of tubulin from very different organisms shows that the intrinsic properties of tubulins which are important for polymerization in vitro are highly conserved. These precedents provided a rationale for the attempts described in this thesis to isolate tubulin from tobacco cells by polymerization and copolymerization. Isolation of Intact Microtubules Very little work has been done on the isolation of intact MTS other than a few early reports on the purification of intact MTs 111 from brain (Kirkpatrick et al, 1970). Filner and Behnke (1973) developed a medium containing 50% glycerol and 10% DMSO to stabilize in viva MTs in brain. This technique has been applied by others to isolate intact MTs (Pipeleers at al, 1977). However, more work has been done on the isolation of mitotic spindles using media containing ethanol-digitonin (Mazia, 1955), hexylene glycol (Kane, 1965) , thiodiglycdl (Mazia at al, 1961) , dithiodipropanol (Sakai, 1966), glycerol (Sakai and Kuriyama, 1974) or glycerol and dimethyl sulfoxide (Forer and Zimmerman, 1974 ). It would be particularly appropriate to purify intact MTs from tissues which have tubulin concentrations lower than that required for in vitra self-assembly. Unfortunately, as mentioned above, few attempts have been made towards developing such procedures. An agent known to stabilize MTs in viva deuterium oxide (D20) (Gross and Spindel, 1960; Marsland and Zimmerman, 1963). Burgess and Northcote (1969) showed that D20 treatment of wheat roots increased the number of MTs in the preprophase band and the mitotic spindle. Such an increase in number of MTs was also reported in Sphagnum (Schnepf and Deichraber, 1976). Materials and Methods Cell Culture and Radiolabeling Tobacco XD cells in suspension culture were grown on MID medium containing 100 pM sulfate as described previously (Filner, 1965). 112 Two days after inoculation, filter-sterilized sodium 3SS-sulfate (715 Ci/mmole) was added to obtain the desired final specific radioactivity. Preparation of Cow Brain Tubulin Cow brain tubulin was prepared by cycles of assembly-disassembly essentially by the method of Shelanski et al (1973). Fresh cow brain obtained from the Michigan State University Meat Laboratory within 30 min of slaughter was cleaned of blood clots and 50 9 portions were minced with scissors and suspended in 100 ml of ice cold homogenizing medium, 100 mM.MES, pH 6.5 at room temperature, 4 mM MgCl 2 mM EGTA and 0.2 mM GTP and homogenized in a Waring 2: blender (15 sec, low speed). The homogenate was sedimented at 16,000 3 gmax for 15 min to remove cell debris. The supernatant fraction was decanted and sedimented at 510,000 x gmax for 60 min at 4°C. The high-speed supernatnat fraction was made 29% (v/v) in glycerol and incubated at 37°C for 30 min. The microtubules formed in vitra were collected by sedimentation at 510,000 x gmax for 60 min at 25°C. The once polymerized MT pellet obtained was either resuspended in storage buffer [50 mM MES, pH 6.5, 2 mM MgC12, and 50% (v/v) glycerol] and stored in a freezer or immediately cold- depolymerized in a medium identical to the homoganizing medium except that it contained 1 mM GTP and sedimented at 510,000 x gmax at 4°C to remove cold-insensitive aggregates. The supernatant was carried through one more cycle of assembly as described above, to yield twice polymerized MT pellet. This was stored in the storage buffer in a freezer. Immediately, before use, the stored once or 113 twice polymerized MTs were depolymerized by dilution (1:5) in the cold homogenizing medium, containing 1 mM.GTP, for 30.min. After sedimentation at 510,000 x gmax for 60 min at 4°C the supernatant was repolymerized as above and the freshly polymerized.MTs were col- lected as described above and used in copolymerization experiments. Copolymerization of Labeled Tobacco Cell Extract with Cow Brain Tubulin 35 [SJ-'SO4 -labeled tobacco cells were harvested after the desired period of time on two layers of Whatman No. 1 filter discs on a buchner funnel under vacuum. After rinsing once in deionized H20, the harvested cells were weighed (ca 7 g fr. wt./500 ml of seven-day-old culture) and homogenized in ice-cold cycle medium (100 mM MES, pH 6.8, at room temperature, 2 mM MgC12, 0.5-1.0 mM GTP, 1 mM EGTA, 5 mM DTE and 0.1 mM EDTA) by 20 passes of a motor- driven teflon-glass homogenizer. After an initial sedimentation at 12,000 x gmax for 15 min to remove cellular debris, the supernatant was sedimented at 106,000 x gmax for 80 min. The high speed supernatant was mixed with the desired amount of brain tubulin, prepared as described above. The mixture was made 1 mM in GTP and incubated with 29% (v/v) or 50% (v/v) glycerol, (ca.3h4and 5.4 M glycerol, respectively) in a water bath maintained at 37°C, for 30 min. The MTs are collected by sedimenting the incubation mixture at 6.3 x 106 x g 'min (usually 1.06 x 105 x g for 60 min in max max Beckman type 30 rotor) at 25°C. The resultant supernatant fraction is decanted and the once copolymerized MT pellet resuspended in ice-cold cycle medium, containing 1 mM GTP, and chilled on ice for 114 30 min to depolymerize the MTs. The cold-insensitive aggregates in this resuspension are sedimented as above, but at 4°C. The pellet is discarded and the supernatant fraction was passed through one or two additional cycles of temperature-dependent assembly- disassembly, if necessary, to yield pellets of twice and thrice copolymerized MTs, respectively. No additional brain tubulin was added for the second and third cycles of assembly. SDS-Polyacrylamide Gel ElectrOphoresis and Autoradiography SDS-polyacrylamide gel electrophoresis was performed on slab gels (14 cm x 12 cm x 0.075 or 0.15 cm) using the discontinuous system described by Laemmli (1970) on 8% separating gel, unless mentioned otherwise. Samples were prepared by resuspension in sample buffer [50 mM Tris-HCl, pH 6.8, 1.05% (w/v) SDS, 5% (v/v) -mercapto-ethanol, 0.0015% bromophenol blue and 10% glycerol]. Immediately before loading on gels, samples were heated for 3 min in a boiling water bath and cooled. Electrophoresis was performed at 10 mA, after which the gels were stained in a solution containing final concentrations of 0.25% (w/v) Coomassie blue, 45% (v/v) methanol and 9% (v/v) acetic acid, and destained in a solution of 5% (v/v) methanol and 7.5% (v/v) acetic acid. Gels were dried on Whatman 3 mm paper under vacuum, using steam heat. Dried gels were, if required, autoradiographed for about a week using Kodak SB-5 x-ray film. 115 Peptide Mapping by Limited Proteolysis Peptide maps were prepared by the method of Cleveland at al (1977). Protein samples were run on SDS-PAGE in 8% separating gel (1.5 mm thick), as described above, and the bands of interest were detected by Coomassie blue staining. To avoid possible hydrolysis, the gels were stained for 30 min and then destained for about an hour. After washing once in cold water, the wet slab gels were placed on a Mylar sheet over a light box. Each individual band of interest was excised with a scalpel, trimmed to about 4 mm width and washed in 10 m1 of wash buffer (125 mM Tris-HCl, pH 6.8, 0.1% SDS and 1 mM EDTA). After about one hour the excised bands were loaded ontoaisecond SDS gel, similar to the first one, except that it was made in 15% acrylamide (1.5 mm) and had about 5 cm long stacking gel. Space around each gel slice was filled with 10 pl of wash buffer containing 25% (v/v) glycerol. Finally 10 pl of wash buffer containing 10% (v/v) glycerol and 50 pg/ml of either Staphylococcus aureus V8 protease or a-—chymotrypsin was added to each gel slice, except in the controls, to which only 10 p1 of wash buffer containing 10% glycerol was added. Electrophoresis was performed at about 10 mA as above. Quantitation of Radioactivity in Polyacrylamide Gels The dried gel was sliced into 36 strips of equal width and each strip was placed in a scintillation vial with 5 ml of dioxane based scintillation fluid. The radioactivity was counted in a Beckman LS-l33 scintillation spectrometer. 116 Two Dimensional Electrophoresis Protein samples were prepared by the method of Ames and Nikaido (1976) and the two dimensional isoelectric focussing/SDS-PAGE was run essentially by the method of O'Farrell (1975) using the discontinuous gel systsn of Laemmli (1970). The final pH.gradient in the first (isoelectric focussing) dimension was from ca 7.6 to ca 3.8. The gels were stained, destained and dried as described above. . 35 Gel Filtration. [SJ-labeled tobacco extract that copolymer- ized with brain tubulin through two cycles of assembly-disassembly and 3H-colchicine-brain tubulin complex were applied to a Sephadex G-150 column (1.6 x 27 cm) that was pre-equilibrated and eluted with 25 mM MES, pH 6.8, 1 mM EGTA, 1 mM M9012 and 0.1 mM GTP at 4°C. The 35[S]- and 13IC]-CPMwasmeasured in 0.5 ml aliquots of each fraction. Ion-Exchange Chromatography. Twice copolymerized mixture of 35Isl-labeled tobacco extract and brain tubulin was applied both to DEAE-cellulose (Whatman DE-32) column (2.5 x 15 cm) and to phosphocellulose (Whatman P11) column (2.5 x 15 cm), each equilibrated with CB [25 mM MES, pH 6.8, 1 mM MgCl 1 mM EGTA, 2. 0.1 mM GTP and 25% (v/v) glycerol]. The 35[Sl-radioactivity applied to the DEAE-cellulose column was eluted firstwith moml of CB and then with.a 500 ml linear gradient of 0-0.5 M KCl in CB. The 35[SJ-radioactivity applied to the phosphocellulose column was eluted first with 200 m1 of CB, than with 200 ml of 0.1 M KCl in CB, and finally with 200 m1 of l M KCl in CB. 117 Electron Microscopy Routine negative staining was performed by placing a drop of the preparation on a horizontal copper grid (200 mesh, Pelco EM Supplies) coated with formvar and carbon. After ca 15 sec the forceps holding the grid.was tilted at 45° angle and four drOps of 1% uranyl acetate solution were dropped on the specimen. After 30-60 sec, excess stain was removed by a filter paper. The grids were viewed under a Siemens Elmiskopllelectron microscope operating at 80 KV and up to 40,000 x magnification. Determination of Protein and Trichloroacetic Acid Stable Radio- activity Protein was precipitated by 10% trichloroacetic acid. The precipitate was washed in ethanol and then resuspended in 1N NaOH. An aliquot of this resuspension was assayed for protein by the method of Lowry et al (1951). To determine trichloroacetic acid stable radioactivity another aliquot of the same resuspension was used to measure the radioactivity. Determination of Radioactivity Radioactivity in samples was determined by adding 10-50 p1 of the sample to a 5 ml of a dioxane based scintillation cocktail [6% (w/v)naphthalene, 0.4% (w/v) PPO, 0.02% (w/v) POPOP, 10% ethanol in dioxane]. 118 Chemicals and Radiochemicals 4 specific radioactivity available) was obtained from New England All chemicals were analytical grade. 35[SJ-SO (highest Nuclear. Staphylococcus aureus V8 protease and a-chymotrypsin were obtained from Miles and Sigma Chemical 00., respectively. Acrylamide (Mallinckrodt) and N,N'-methy1ene bisacrylamide (Eastman) were recrystallized before use. Results Copolymerization of 35[SJ-Labeled-Proteins of Tobacco Cells with Cow Brain Tubulin Earlier attempts to obtain self-assembly of tubulin in extracts of tubulin were unsuccessful. It was possible that these failures were due to a low concentration of tubulin in the cell extract, possibly lower than the critical concentration required for self- assembly of tubulin. Since tubulins from a wide group of organisms could copolymerize (see Introduction) it was assumed that tubulin from higher plants could copolymerize with brain tubulin and the concentration of tubulin in tobacco cell extracts could thus be raised by adding brain tubulin to the extract. Isolation of tubulin from extracts of 3S[S]-1abeled tobacco cells by copolymerization with brain tubulin was therefore attempted. . 35 Effect of Cow Brain Tubulin on Sedimentation of [S]-Counts from Extracts of 3S[SJ-Labeled Tobacco Cells To determine if any protein from tobacco extract will sediment 119 with cow brain tubulin by the temperature dependent cycles of assembly-disassembly, tobacco cells were grown for five days on 35 - 100 pM [Sl-SO 2 (3 x 107 cpm per pmole). The cells were harvested 4 (ca 9 9 fr. wt.) and homogenized in 1.7 volumes of cycle medium (100 mM MES, pH 6.8, 2 mM MgC12, 1 mM GTP, 1.2 mM EGTA, 0.1 mM EDTA). The homogenate was sedimented at 40,000 x gmax for 60 min. Aliquots of the supernatant fraction were incubated with 29% (v/v) glycerol, either in the absence or in the presence of brain tubulin, at two concentrations, at 37°C for 30 min. The preparation of brain tubulin and the conditions for temperature-dependent cycles of assembly-disassembly are described under Materials and Methods. After two cycles of assembly-disassembly the sedimentable counts in the clarified supernatant fraction after depolymerization of once copolymerized MT pellet and in the twice copolymerized pellet were determined (Table 12). Significant 35S-radioactivity was found in the copolymerized fractions. Effect of Brain Tubulin Concentration on the Sedimentability of 35[SI-Counts To quantitate the dependence of copolymerization of 35[8]- counts on brain tubulin concentration, tobacco cells were grown for five days in the presence of 100 pM 35[S]-SO4"2 (8.9 x 106 cpm/pmole) and harvested (ca 13 g fr. wt.) in late exponential phase. The cells were homogenized in 0.5 volume of cycle medium (see page ) 9 fr. wt. After sedimentation at 12,000 x gmax for 15 min, the supernatant fraction was collected and sedimented at 105,700 x gmax for 60 min. A 0.5 m1 aliquot of the high-speed supernatant fraction (0.86 mg 120 H .m m UoNHHoEMHommp mo coHususoEHpmm pHoo nouns coHuomwm assumcuomsm n mHU .uoHHmm a: pmNHHmE>Homoo coco u mHm« AmH.oVom.H Am>.ovHo.m om.HH o.o H o.m m AmH.ova.o Avm.ovmm.m mn.v o.m H o.m N AHo.ovvo.o AHm.ocHe.~ mn.v o.m I o.m H cowuomwm :oHuolo mmaatz sHHsnsB mHm mHU onadeUzH GH HouoUAHu :Hmum usmumsuwmdm .Oz mmDB «AS ommmaoomm mIoH x 2.0 70H x 2.8 egos SE omens moron? .muomuuxm ooomnoeIHmAmm scum mucsOOIHmHmm mo cOHumucmsHpmm :0 cHHcose chwm mo uommmm .mH momma 121 protein; 2.34 x 107 trichloroacetic acid precipitable 3SISJ-cpm/mg protein) was incubated with varying amounts of brain tubulin (0.12 to 7.46 mg) in a total incubation volume of 2.83 ml. After one cycle of polymerization-depolymerization, the fractions obtained were used to determine radioactivity (see Fig. 23). The sedimentability of 35[SJ-labeled proteins into once copolymerized pellet increased with increasing concentration of brain tubulin (Fig. 19). Since microtubules formed from brain tubulin in vitra were cold-labile, the radioactivity not solubilized from the once copolymerized protein by cold represents either non-specific aggregation of plant proteins or denaturation of plant tubulin. The 35[SJ-proteins reversibly associated with brain tubulin (i.e. the cold soluble fraction) also increased linearly with brain tubulin concentration in the incubation mixture, up to the highest concentration of brain tubulin tested, 2.64 mg/ml (Fig. 19). Specific Radioactivity of Copolymerized Proteins The specific radioactivity (trichloroacetic acid-precipitable 35[SJ-CPM per mg protein) of the copolymerized proteins decreased markedly between once and twice copolymerized MT fractions, but changed much less between second and thrice copolymerized fractions (Table 13, Fig. 20). It is also evident that there was great loss of both trichloroacetic acid-precipitable CPM and protein at each step of copolymerization. Since brain tubulin constituted about 74% of the total protein in the incubation mixture, even brain tubulin is lost in large amounts. Figure 19. 122 Effect of brain tubulin concentrations on the sedimentation of 35S-c.p.m. from 35S-labeled tobacco XD cells. A given amount of the high speed supernatant from 35S-labeled tobacco cells was incubated with varying amounts (mls) of brain tubulin (1.24 mg protein per m1) under polymerizing conditions. 35S-c.p.m. that sediments in the once copolymerized pellet, HlP (closed circles) is resuspended under depolymerizing conditions and resedimented in cold to yield the supernatant fraction, C S (open circles). 1 123 H1? 3— 2—: AC‘s o “3 01- - '_ a X Z O. U C I I r I f I 0.5 1.0 1.5 Milliliiers Figure 19. .as\caaaoaa uaafla as ma.o mafia sedans» awmwn as v~.a.. .Aauec neon oauoooowoA:0awe. 1JZ4 uoHHom no.m vm.o w~.o oH.o m.m h>.o m~.o mh.~ powHuwi I>Homoo moHuca uoHHmm w~.m 0.0 com.c ~m.0 m.h no.0 Nv.o Hm.N poNHuoe I>Homou ooHss uoHHom mm.H ~.oH mm.m oo.m mm ov.v oH.H o.v vowHuoe I>Homoo 00:0 I . . . . . . ousuxHx mH o ooH On no em 0H OCH on mH ache H o w coHuenaocm Acimacws oIoH x oIoA x Ava. AHs\osc AHs. as was oIcA x zao zao As was :ao oanuAaHoouQIuA>Huo< onHooam mo >wo>00wm v I499 choe ¢Iuw>ooom a Heuoa :Houoam mesHo> onEum .mcwwuous ponwumE>Homou wo huH>HuomoHomm oHuHomsm .MH manta 125 Figure 20. Specific radioactivity (trichloroacetic acid stable radioactivity, c.p.m., per mg protein) in once, twice and thrice copolymerized microtubule pellets. Specilic AcIivin(CPM x10’6/MG Protein) 126 Figure 20. Number of cycles of capolymerisoiion 127 Since the specific radioactivity of 35[SJ-labeled tobacco proteins is 2.34 x 107 cpm/mg protein, about 12 pg of plant proteins are present in the thrice copolymerized MT fraction. This ' I fraction has a total of 770 pg protein. Therefore, 758 pg is as brain tubulin. This represents a recovery of about 7.6% of brain tubulin. If 70% of the plant proteins in the thrice copolymerized MT fraction is tobacco tubulin (Fig. ) and if tobacco tubulin is recovered with the same yield as brain tubulin, then we can obtain a rough estimate of plant tubulin in tabacco extracts: ca 3.2% of total soluble protein. SDS-Polyacrylamide Gel Electrophoresis of 3S[SJ-Labeled Copolymerized Proteins To determine if there is an enrichment for tobacco protein by sedimentation with brain tubulin, the proteins in the twice copolymerized fraction, purified spinach Fraction-l-protein (F1P), and twice polymerized MTs from brain were analyzed by SDS-polyacryl- amide gel electrophoresis (SDS-PAGE) on 15% separating gels. While cow brain and F1P polypeptides were detected by Coomassie blue staining, the much lower levels of 3s[SJ-labeled tobacco proteins were detected by autoradiography (Fig. 21). Brain tubulin migrated as a doublet of a faster moving B-subunit and a slower moving a-subunit. The large subunit of spinach F1P migrated as a doublet of two very closely migrating polypeptides, at least one of which comigrated with the B-subunit of brain tubulin. There was a single major band of 35[S]-labeled polypeptide that also co-migrated with the B-subunit of brain tubulin. For a better resolution of major Figure 21. 128 SDS-PAGE of brain tubulin, spinach Fraction-I-Protein (F-I-P) and twice copolymerized microtubule pellet in 15% separating gel. A: cow brain tubulin; 8: lower amount of brain tubulin plus F-I-P; C and D: higher and lower amounts of F-I-P, respectively; E and F: higher and lower amounts of twice copolymerized pellet, respectively; e and f: radioautogram of same gel shown in E and F. A to F: Coomassie blue stained gels. 129 Figure 21. 130 35S-polypeptides the twice copolymerized fraction and brain tubulin marker were analyzed by SDS-PAGE on 8% separating gels (Fig. 22). Although several minor 35[SJ-labeled polypeptides can be observed in the autoradiogram of the twice copolymerized protein about 70% of total radioactivity is found in a prominent doublet (Fig. 23). In order to make evident the relationship between the brain tubulin doublet and the radioactive tobacco doublet, the gels stained with Coomassie blue were compared side by side with their corresponding autoradiograms (Figs. 22, 24c). Two major characteristics of the tobacco doublet were evident. First, while the faster moving band in the radioactive doublet co-migrated with the B-subunit of brain tubulin, the slower moving radioactive band had a migration rate during SDS-PAGE intermediate between those of a- and B-subunits of brain tubulin. Although the faster moving band in the tobacco doublet cannot be detected in the gels stained with Coomassie blue, due to the presence of brain tubulin, the slower moving tobacco band was detected, in some gels, as a faintly staining band between the a- and B-subunits of brain tubulin (Figs. 22, 24c). Second, the relative intensities of the images of the slower and faster moving radioactive bands on the autoradiogram were different: the a-band appeared to have more radioactivity (Fig. 24c). The lack of such a difference in some gels may be related to the fact that the intensity of autoradiographic images are not directly proportional to the radioactivity (Laskey and Mills, 1975). Figure 22. 131 SDS-PAGE of cow brain tubulin and twice copolymerized microtubule pellet in 8% separating gel. B: brain tubulin alone, C ; twice copolymerized pellet, C and 1 2 C3; same as C1 but in a higher amount. c1, c2 and c3: 1, C2, and C3 , C2 and C3 were stained by Coomassie Blue. autoradiograms of C , respectively. Tracks B, C1 132 Figure 22. 133 Figure 23. Profile of 35S-radioactivity in strips cut from SDS-PAGE of twice copolymerized microtubule pellet (of the same gel shown in Tract C of Figure 22). 3 134 CPM x10'3 T WA 1O 20 Fraction Number Figure 23. 3'0 135 N .H .osHm onmMEooo cqu poonum mem "U ou m .>H0>Huoommmw .O can m N .H m . 4 m mo manna NH IoHpmwousm no use a .N .H n m m .uoHHom mHsnsuOMOHE pmNHHoE>Hom00 o0H3u mo ucaoEm meow mo mxomuu mourn no .uomuuxm ooomnou mo coHuomum ucmumcuwmsm modem an3 mmHmanImmm mo uoHHmm coHnEmmmmIMHom coco mo >Ho>Huommmou .mu:508m HosoH pom woann "mm pom Hm .mHHao ooomnou paHmanIm mo meannuxc mo coHumucoEHpmm pmmmm 50H: nouns coHuomum mm accumcummsm mo .>H0>Huoommou .mucsoem wmzmHn pom wm3oH "we can He .coHumeumawHomoo pom >HnaommMIMHom nouns mHHmo ooomnou mo mchuoum ponanImmm mam cHHonou :Hmun 300 no mommImom .vm musmHm 136 NAH FD .vm wHDmHh 137 Peptide Maps of the Prominent Protein from Tobacco Cells that Co- polymerizes with Brain Tubulin The identity of the prominent doublet observed on SDS-PAGE of copolymerized tobacco proteins (Fig. 24c) was investigated by subjecting each band of the doublet to limited proteolysis and separating the resulting peptide fragments on SDS-PAGE by the method of Cleveland et al (1977), which is outlined below. The twice copolymerized proteins were first resolved on SDS- PAGE. After a brief period of staining with Coomassie blue and destaining, the positions of the tubulin subunits were detected. The a- and the B-subunits of brain tubulin and the portion of the gel in between that contains the slower moving band of the tobacco doublet, were carefully excised and then reloaded onto a SDS-PAGE with 15% separating gel. Electrophoresis was begun after adding the desired protease onto each excised band. Limited proteolysis of the individual band occurs as the protease and the substrate move through the stacking gel. The resultant peptide fragmentS‘then get separated from each other and the protease protein, according to their molecular weight, in the separating gel. Using Staphylococcus aureus V8 protease, the peptide maps of the a- and B-subunits of brain tubulin, detected by staining with Coomassie blue, were compared with those of the slower- and faster- moving radioactive bands of the tobacco doublet, detected by autoradiography of the same gel (Figs. 25, 26). The peptide map of the B-subunit of brain tubulin and that of the faster moving band of tobacco doublet were virtually indistinguishable (Figs. 25, 26). Figure 25. 138 SDS-PAGE of peptide fragments generated by Staphylococcus aurens protease from cow brain tubulin and the putative tobacco tubulin subunits present in twice copolymerized microtubule pellet. B a-subunit of brain tubulin; 1: t1: a-subunit of tobacco tubulin; 82: B-subunit of brain tubulin and t2: B-subunit of tobacco tubulin; B1 and B2: gels stained with Coomassie Blue; t1 and t2: radioautograms of B1 and 82 gels, respectively. 139 4. . . 4.2.. we I}? .. 1 _ in? ....2 1.]?! 1. i an. .anfikbtf .- _——_-._ ._.... —.—-—— _ _-._--..._. Figure 25 140 .mHmm mo mEmHmoHpmuouam N H u u one u «osHm mHmmmEooo nqu pmcHMUm mHam "N m can .mo .Hm .Ho acHHabau aamwn N m a m acHHsnsu mo uHcsnsmIm u U «coHuomwm poNHHoshHomoo cH sHHsosu chun mo uHcansmIm N H ooomnou mo uHsaoamIm “ u xcHHansu ooomnou mo uHcsnsmIo u u acoHuomnm ooNHumshHomoo cH ceases» camwn mo swaanamua "Hm u3on sadness aamwn wo pagansmua ”Ho .umaama mHsnss IouoHa pwwHuosmHomoo moHau cH usomoum muHcsnom sHHsnsu ooomnou m>Humusm one new sHHsnsu cwmun 300 scum ommmuoum breeze exooooonsdeem an psychoses musasmmum opwummm mo mummImom .om musmHm 141 .oN ousmHm 142 On the other hand, the peptide map of the a-subunit of brain tubulin was distinct from that of the sibwer moving band of tobacco doublet (Figs. 25, 26). Using another protease, a-chymotrypsin, the above observations were confirmed. The peptide maps of the B-subunit of brain tubulin and the faster moving band of tobacco doublet generated by a-chymotrypsin were still indistinguishable,although S. aureus V8 protease and a-chymotrypsin produce different fragment patterns (Fig. 27). However, the peptide maps of the a-subunit of brain tubulin and the slower moving band of tobacco doublet, though still distinguishable show a much more striking resemblance than that evident in the maps generated by S. aureus V8 protease (Fig. 27, 28). As a control, the large subunit of F-l-P was also subjected to limited proteolysis, which resulted in a peptide map distinct from each band in the brain tubulin doublet (Fig. 29). Two Dimensional Isoelectric-Focusing/SDS-Electrophoresis of the Proteins The behavior of the prominent tobacco doublet observed on SDS- PAGE was also determined on two dimensional electrophoresis, with isoelectric focusing of the denatured proteins in one direction and SDS-PAGE in another. When the twice c0polymerized proteins were subjected to the two-dimensional electrOphoresis, the behavior of the brain tubulin doublet, detected by Coomassie blue staining of the gel (not shown), was compared with that of the tobacco doublet detected by autoradiogram of the same gel (Fig. 30). As controls 143 .msHm memmEOOO suHB pochum mHmm one No was mm .H membHuoommou .mm pom Hm mHom mo EmumoHpmuousm was my use He xcHHansu chwn mo uHconsmIm ”No use «coHuomum poNHuosaHomoo :H sHHsnsu chun mo uHcsnomIm "mm xcHHsnsu ooumnou m .Hu .0 mo uHssnamIm “Nu xcHHsnsu Cabana» mo uHcsnsmIo "Hp «coHuomuw omNHumshHomoo cH :HHsnsu swung mo uw::fldmld "Hm «GHHsnflu CHMHQ mo UHCSQSmIo ”H U “Houuooo moummmHocs .cHHsnsu :Hmwn mo uHchSmIo "O .sHmuoum moNHumESHomoo oOHsu asp sH sHHsnsu ooomnou m>Humusm on» new :HHsnsu sHmwQ zoo Baum sHmmwuuoehsoIo >0 oauoumcmm mucosmmnm opHumam mo mummImom .am magmas 144 .hN wusmHm 145 .osHm memmEooo QuHs pochum mHom mum muosuo HHm .sz>Huowmmon .me use He mHmm mo mEmHmoHpmwousm mum Nu pom Hu xcHHsnsu chwn mo uHssnsmIm "mm “coHuomum pmNHwaemHomoo on» :H cHHsnau chnn «0 uHssnamIm .Ns stHsnau ooomnou mo uHssnsmIm "Nu acHHsnau ooomnou mo uHcsndmIo “Hu acHHansu Gaucho» mo UHssnsmIo "Ha AcOHuomum paNHuosaHomoo as» cH sHHansu :Hmun mo uHCSQSmIo "Hm stHonsu sHmun mo uHssnsmIo "U .cHououm poNHume>Homoo onsu onu sH cHHsosu ooomn0u m>Humusm on» use oHHsnsu sHmHn :00 Beam :Hmmhuuoa>sOIo he poumnocmm mucosmmwm opHumom mo mommImom .mm magmas 146 .mm museum Figure 29. 147 SDS-PAGE of peptide fragments generated by Staphylococcus aurens protease from the slower-moving large subunit of Fraction-I-protein (F-I-P) and cow brain tubulin. C1: a-subunit of brain tubulin; Bl: a-subunit of brain tubulin in copolymerized fraction; F: the large subunit of F-I-P; C B-subunit of 2: brain tubulin and 82: B-subunit of brain tubulin in copolymerized fraction. All gels were stained with Coomassie Blue. Figure 29. 149 Figure 30. Two-dimensional isoelectric focusing/SDS-PAGE of twice copolymerized protein. Autoradiogram of gel. Arrow indicates the dimension of isoelectric focusing (acidic end at the right). 150 Figure 30. Figure 31. Figure 32. 151 Two-dimensional isoelectric focusing/SDS-PAGE of brain tubulin. Arrow indicates the dimension of isoelectric focusing (acidic end at the right). Gel stained with Coomassie Blue. Two-dimensional isoelectric focusing/SDS-PAGE of spinach Fraction-I-Protein. Arrow indicates the dimension of isoelectric focusing (acidic end at the right). Gel stained with Coomassie Blue. 152 Figure 31 9 ' Figure 32. 153 brain tubulin by itself (Fig. 31) and purified spinach.FlP (fig. 32) were also run on separate gels under the same conditions. It is evident that the labeled tobacco doublet and the brain tubulin doublet behave similarly in isoelectric focusing and both migrate in the acidic region of the two-dimensional gel system. The lack of resolution of the tobacco doublet and the poor resolution of the brain tubulin doublet in SDS-PAGE dimension can be attributed to the high acrylamide concentration (15%) in this gel. In contrast,FlP subunits migrate in very different regions. Comparison of the Properties of Native Twice Copolymerized 35IS]- Labeled Protein with 3H-Colchicine-Brain-Tubulin Complex 35[Sl-labeled tobacco protein that copolymerized two times with cow brain tubulin and 3H-CLC-tubulin complex (with free 3H-CLC) were applied together to a Sephadex G-150 column (1.6 x 27 cm) that was equilibrated and eluted with 25 mM MES, pH 6.8, 1 mM EGTA, 1 mM MgCl2 and 0.1 mM GTP at 4°C (Fig. 33). Since about 70% of the 35[SJ-labeled protein in the twice c0polymerized fraction is in the doublet of tobacco tubulin polypeptides (Fig. 23), the major peak of 35[SJ-protein is native tobacco tubulin, which comigrates with 3H-CLC-brain tubulin complex through Sephadex G-150 is a further indication of their close similarity. When the twice c0polymerized 3S[SJ-labeled tobacco tubulin was applied to a DE-32 ion-exchange column (2.5 x 15 cm), which was equilibrated with 25 mM MES, pH 6.8, 1 mM MgC12, 1 mM EGTA, 0.1 mM GTP, and 25% (v/v) glycerol, and eluted first with 200 ml of the same buffer and then with a 500 m1 linear gradient of 0-0.5 M KCl in the same buffer 154 . . . . 5 . Figure 33. Molecular Sieve filtration of 3 S-tobacco proteins in twice copolymerized microtubule fraction, . . 3 3H-CLC-brain tubulin complex and free H-CLC on Sephadex G-l 50 . 155 \ v‘lll \:MChMM€ 1 I 7O 50 10 2-4 Number Frocflon Figure 33. 156 (Fig. 34), there were two major peaks and several small ones of 3S[SJ-radioactivity. One major peak eluted with the buffer alone and may have been unbound material. The other major peak eluted around 0.31 M KCl. When 3H-colchicine-tubulin complex was chroma- tographed separately, it eluted with peaks between 0.26 and 0.315 M KCl (Figs. 15, 18 and 35). Thus, the major peak of 35S-radioactivity in tobacco protein that bound to DEAE cellulose eluted in the region where 3H-CLC-tubulin complex eluted. When the twice copolymerized 3S[SJ-labeled tobacco protein was applied to a phosphocellulose ion-exchange column equilibrated with the same buffer used for DEAE cellulose chromatography, a large peak eluted in the void volume with the buffer and a smaller one eluted with 1 M KCl added to the buffer (Fig. 36). Again, since ca 70% of total radioactivity in this protein fraction is found in the doublet of tobacco tubulin polypeptides, one can tentatively conclude that the behavior of 3SIS]-radioactivity when passed through a phosphocellulose column is attributable to native tobacco tubulin. It behaves as does brain tubulin which is also not retained on a phosphocellulose column (Spiegelman et al, 1977; also see Fig. 37). Isolation of Particulate Tubulin from Tobacco Cells Although copolymerization of 35[SJ-labeled tobacco extract with brain tubulin has been valuable in the identification and partial characterization of tobacco tubulin, the constraints of this method for further characterization of tobacco tubulin. such as drug binding or self assembly, are obvious. 157 .coHumuucoocoo HUM msmuo> >uH>Huoapcoo mo smmwm mumpcmum o co momma msoHuomuw xmom cH coHumuucoocoo UHmm .80 Mom mosaouoHS cH mosmuoapcoo OHMHoomm .mmHowHo sumo x.s.m.0Immm .onowHo owmoHU .mcHououm on» ousHm on can: mmz HUM z m.o I 0 mo usoHpmHm HmmcHH s .soHuomum pmeHmE>Homoo ooHBu :H chuoum ooomnouImmm mo wsmmumoumsouso omcmnoxo coH omoHoHHaOImdmo .vm owsmHm 158 ,pI x was 118 N F l 03H! o—> I) «m FRACTION Figure 34. 159 .ssoon mm3 usoHomsm as» can: soHuomwm as» mxHME .0. .A.U.v soHumwusmosoo HUM momwm> wuH>Huospsoo mo smoum pammsmum m AD posflfiwouop mm3 xmom use so soHumuuswosoo uHmm .50 was moseouoHS sH muscuOSpsoo UHMHoomm .moHoHHo sumo a.8.m.0Imm .mmHouHo momoHU HmmsHH s .msHououm as» ousHo ou poms mm3 HUM z m.o I o no usoHpmum .memEoo sHHsnsu sHmun osHOHsoHooImm mo >sfimwmoumsowso omsmsoxo soH memo .mm ousmHm mwmgaz ZOE-04ml:— . mm 9»:on on. 00. On I 160 r- a/ x lamp/7pm.? Q N d u - 3.0/ X “60 1 000. I OOON .2900 000m _1 161 Figure 36. Phosphocellulose ion exchange chromatography of . . 35 . . twice copolymerized S-tobacco protein. The proteins were first eluted with buffer and then with 0.1 M and 1.0 M KCl in buffer at the indicated fractions. 162 CPM x10‘4 Figure 36. Fraction number Figure 37. 163 Phosphocellulose ion exchange chromatography of cow blood platelet tubulin. Tubulin isolated by a cycle of assembly-disassembly, was eluted first with buffer alone and then with 0.1 M and 1.0 M KCl in buffer at the indicated fractions. Aliquots of the eluted fraction were assayed for 3H-colchicine binding activity by Whatman DE-8l filter assay. 164 N l CPM IIIO'3 I I l T l 10 30 50 Fraction Number Figure 37. 165 Attempts were made in the past in this laboratory to obtain intact MTs from tobacco XD cells by homogenizing the cells at room temperature in a medium, containing 50% glycerol and 10% dimethylsulfoxide, DMSO, that had been used to isolate intact MTs from brain (Filner and Behnke, 1973). However, no MT images were seen when pellet after high-speed sedimentation (40,000.x gmax for 60 min) was examined in an electron microscope after negative staining with uranyl acetate (Lescure and Filner, unpublished). A possible reason for the failure to stabilize MTs in situ in these cells may have been that the presence of 50% glycerol in the medium caused severe plasmolysis of the cells, leading to a drastic change in the cellular milieu, which in turn may have disrupted the labile :MTs before they had a chance to be stabilized by the permeating medium. To prevent the possibility of MT disruption by osmotic shock, two major changes were made in the procedure used above to isolate intact MTs. First since heavy water D20 is known to stabilize MTs in situ (see Introduction), the tobacco cells were first suspended in D20 for about 30 min. Second, since DMSO makes the plasma membrane of tobacco XD cells leaky (Delmer, 1979), DMSO was added a few minutes prior to the addition of glycerol, in order to facilitate the permeability of glycerol. When the modified procedure to isolate intact MTs was used to homogenize tobacco cells (as described in detail in legend to Fig. 38), the resultant high speed pellet was resuspended in the stabilizing medium (recipe defined in the procedure in legend 166 to Fig. 38) and examined in the electron microscope after negative staining with uranyl acetate (Fig. 38). A few rod-like structures were observed. The diameter of some of these structures was around 235 nm, based on beads of standard size. However, substructure was not sufficiently resolved to permit their identification as MTS. To see if microtubules were in fact present in the 40,000 x g pellet, the resuspended pellet was dialyzed against cold depoly- merizing buffer (100' mM MES, 98 6.8, 4 m1 MgCl 2 mM EGTA, 0.1 mM 2: GTP, 0.1 mM DTE and 0.1 mM EDTA) to remove traces of glycerol, DMSO and D20 from the resuspended pellet. If MTs were present in the 40,000 x gmax pellet they would be expected to depolymerize by dialysis in cold. The dialyzed pellet was resedimented at 40,000 x gmax for 120 min, and the supernatant fraction was analyzed by SDS-PAGE. A very faint tubulin doublet was observed after staining with Coomassie blue (not shown). In order to increase the sensitivity of the above system, 35[SJ-labeled cells were used and the presence of tubulin in the cooled depolymerized supernatant from the high speed pellet was "assayed" by either self assembly or copolymerization with brain tubulin, followed by SDS-PAGE and autoradiography (Fig. 39) as described in detail in the legend of Table 14. Proteins obtained by cold clarification of the dialyzed high- speed pellet were self assembled or copolymerized with brain tubulin. Comparison of the autoradiograms after SDS-PAGE of the self assembled and copolymerized pellets of once polymerized MTs shows- Figure 38. 167 Electron micrograph of the pellet of high speed sedimentation of extracts of tobacco cells made in a stabilizing medium, described in the text. 40,000 X. Seven day old tobacco XD cells were harvested by vacuum filtration to almost dryness (ca. 10 g) and resuspended in 10 ml of D 0 containing 25 mM MES, 2 “pH" 6.8 (this is the reading obtained by the pH meter; however the use of the term 'pH' is incorrect with D20). After 30 min at room temperature for equilibration, 20 ml of D 0 containing the remaining ingredients of the stabilizing buffer (SM) were added, to give the final concentration of 50 mM MES, 2 mM Mg C12, 1 mM GTP, 1 mM EGTA, 5 mM DTE, 0.1 mM EDTA and 10% DMSO, "p " 6.8. Finally after a few minutes the medium was made 50% in glycerol. After an additional 30 min the homogenate chilled on ice. After an initial sedimentation at 12,000 x gmax for 15 min to remove cellular debris, the supernatant was sedimented at 40,000 x gmax for 120 min at 4°C. 168 Figure 38. 169 TABLE 14. Depolymerization of the High-Speed Pellet and Its Poly- merization in the Absence and Presence of Brain Tubulin. VOL. CPM/ML TOTAL CPM % (m1) Dialyzed high-speed pellet «42.0) 1.24x109 $2.5x109 100 supernatant °f °°ld' hlgh' m(1.5) 2.06x107 m3.lx107 1.2 speed spin of the dialysate Self-Assembly (without added brain tubulin) Incubation mixture (2.0) 1.03x107 2.06x107 100 1 x MT pellet (0.2) 2.15x106 4.30x105 2.1 Copolymerization (with added brain tubulin) Incubation mixture (1.413) 4.37x106 6.18x106 100 1 x °°P°lymerlzed Pellet' (0.75) 1.20x106 9.00x105 14.6 Tobacco gells were grown for six days on 100 pM 3551-504“2 (ca 8.8 x 10 cpm/pmole). The cells were harvested (ca 15 g), as above, and resuspended in 15 ml of 020 containing 25 mM MES, "pH" 6.8, for 30 min. Then 11.2 ml of D20 containing other buffer ingredients, (including 30% DMSO, was added, followed by .28 ml of glycerol and 2.8 ml of DMSO, in that order. This staggered addition of DMSO:may decrease the possibility of bad cellular effects due to sudden exposure to a high concentration (> 10%) of DMSO. Thus, the cells were resuspended in 3.9 volumes of the buffer per 9 fr. wt. The final concentrationsixithe homogenizing medium were: 19 mM MES, 2 mM MgC12, 1 mM GTP, 1 mM EGTA, 5 mM DTE, 0.1 mM EDTA, 9.8% DMSO and 50% glyberol in 100% D20, "pH" 6.8 (the lower concentration of MES used is due to the insolubility of higher concentrations in the medium). After an additional 60 min at room temperature, the cells were homogenized and the homogenate sedimented at 12,000 x gmax for 15 min to remove cellular debris. The resultant supernatant was sedimented at 142,000 x Qmax for 45 min at 4°C. The resulting high-speed pellet was resuspended in the smallest possible volume of cold D.P. and dialyzed against 500 m1 of cold D.P. for 2 hrs. The dialysate was resedimented at 368,000 x gmax for 17 min at 4°C. One ml of the resultant supernatant (2.06 x 107 cpm/ml) was incubated with 1 ml of glycerol and another aliquot of 0.3 m1 of the supernatant was diluted to 0.5 ml with D.P. and then incubated with 0.5 m1 of 3 x polymerized brain tubulin and 0.413 ml glyverol [29%(v/v)]. After 30 min at 37°C the incubation mixtures were sedimented at 368,000 gmax for 17 min at 25°C, to yield pellets of once self-polymerized and once copoly- merized proteins. 170 .xHo>Huoommon .9 use m .m .0 mHoo mo weeanHuenon:e one u use m .n .o “AnHoEommeIMHom noume sownoenm uoHHom use soHuoenm useuesnomsm on» one m use we uoHHom uoomm smHs uowHHHn:H0m uHoo on» no >Ho6ommeIMHom .m “AsoHneanos>Homoo on» nonme .aHo>Huoommon .wsoHuoenm uoHHom on» use usenesnomsm on» one m use my sHH:Q:n sHenn 300 suHs noHHom uoomm smHs uowHHHn:H0quHoo mo sOHneNHnoE>Homoo .O nonesooosos Henou on» no useuesnom:m uoomm 30H: .9 nosoHe sHH:n:n sHenn .m .AmHHenou now .eH oHneB oomv msoHuHusoo msHNHHHnenm nous: uonemonm mHHoo oouenou mo uoHHom uoomm 30H: sH msHouonmImmm uoanosaHomoo use uoHnEommeIMHom mo mosmImom .mm on:mHm 171 I an: Ij '"~ I Figure 39. 172 that the self-assembly of the cold-solubilized high-speed pellet enriches for the tubulin doublet in a manner similar to its copolymerization with brain tubulin. It is also evident, especially in gels stained in Coomassie blue (Fig. 39), that there is more a-subunit than B-subunit in self-assembled MTs, confirming the earlier observation with both copolymerized and self-polymerized tobacco tubulin (Fig. 24). Having established that plant tubulin exists in a sedimentable form when tobacco cell extract is prepared in conditions which stabilize intact MTs in viva, attempts were made to use this method to concentrate tobacco tubulin by resuspending the sedimentable tubulin in the least volume of buffer possible. If the concentration is sufficiently high, then tubulin might be further purified by its self assembly under polymerizing conditions. Therefore, the above experiment using 35(SJ-labeled cells was essentially repeated with unlabeled cells, except that water was used instead of heavy water, which is expensive, and dialysis of the high-speed pellet was replaced by mere resuspension of the pellet in cold depolymerizing buffer for 30 min. After sedimenting the cold-resuspended pellet at the high speed at 4°C, the supernatant was made 50% (v/v) in glycerol and incubated at 37°C for 30 min. The pellet fraction, in which the self-assembled tubulin was expected, was collected by sedimentation at 25°C. The visible pellet was much smaller than that observed using D20. The proteins in the pellet and supernatant were separated by SDS-PAGE and visualized with.Coomassie blue stain (Fig. 40). There was a definite enrichment of the a-subunit of plant tubulin in the Figure 40. 173 SDS-PAGE of 35S-proteins in high speed supernatant of tobacco extracts, after self-assembly. B: cow brain tubulin; S: once self-assembled pellet of high speed extract of tobacco; and, s: autoradiogram of gel 8. B and S are Coomassie Blue stained gels. 174 Figure 40 . 175 supernatant of the cold-solubilizing high-speed pellet, but the enrichment was not noticeably improved in once self assembled MT pellet. Again the B-subunit is markedly lower in concentration than the a-subunit. However the yield of plant tubulin was too low, especially in the absence of D O, to use this method for a large 2 scale isolation of tobacco tubulin. In order to determine if any MT-like structures were formed in the above studies of self assembly, aliquots of the incubation mixtures were observed in the electron microscope after negative staining. Although fibrous structures were observed (Fig. 41) there were no microtubule-like images present, even after varying the concentration of "92+, GTP and EGTA. When 15% dextran T10 was used to facilitate microtubule assembly, striking fibrous materials were observed, which also appeared to have a substructure, but again no microtubule image was observed. Isolation of Tubulin by Self Assembly In one of the copolymerizhugexperiments described earlier, an aliquot of the supernatant resulting from the high speed sedimentation of the 35S-labeled tobacco extract was incubated at 37°C in the presence of 50% (v/v) glycerol without brain tubulin, as a control. The once polymerized MT pellet was collected by high speed sedimentation of this incubation mixture, and analyzed in SDS-electrophoresis, followed by Coomassie blue staining and autoradiography. There was an apparent enrichment for the a-subunit 176 Figure 41. Electron micrograph of self-assembled proteins in the high speed supernatant of tobacco extracts in the presence of 15% dextran T10 in a buffer containing 100 mM MES, pH 6.5, 1 mM EGTA, 1 mM GTP and 5 mM DTE. 20,000 X. 177 Figure 41. 178 of brain tubulin and also for another polypeptide of higher molecular weight (Fig. 24). Hawever, the amount of protein in the pellet was too small to be purified by additional cycles of assembly- disassembly. As already mentioned earlier, tubulin obtained in the particulate fraction from high speed sedimentation of extracts of tobacco cells prepared in 50% (v/v) glycerol and 10% DMSO either in D20 or in H20 could be self assembled after its cold solubilization (Fig. 39). In addition, the supernatants from the same high speed centrifugation of the cell extracts were also carried through two cycles of assembly-disassembly. When the fractions were analyzed on SDS-electrophoresis followed by either Coomassie blue staining or autoradiography, it was evident from the enrichment of the tobacco tubulin subunits, that reversible aggregation and that self assembly had occurred (Fig. 41). These results suggest that a large scale purification of tobacco tubulin is possible by cycles of spontaneous assembly-disassembly without adding carrier tubulin or primer. There are two prominent characteristics that are evident in SDS-electrophoresis of self assembled tobacco tublin: First, the amounts of the B-subunit are very much lower than the a-subunit. Since the differences are observed both in autoradiographs of 3SS- labeled polypeptides and in gels with proteins stained by Coomassie blue it most probably reflects a true difference in the amounts of the two polypeptides at least in the "assembled“ form. Second, the enrichment of high molecular weight polypeptide was also apparent. 179 Both of these characteristics albeit less pronounced, were also evident in 35S-labeled'polypeptides isolated by two cycles of copolymerization with brain tubulin. Discussion Identity of the Tobacco Doublet When purified cow brain tubulin was carried through two cycles of assembly-disassembly, in the presence of 35[SI-labeled tobacco cell extract, less than 0.2% of the total radioactivity and ca 0.6% of the trichloroacetic acid-precipitable radioactivity present in the extract copurified with the brain tubulin (Tables 12, 13). When the twice c0polymerized protein was analyzed by SDS-PAGE on 15% separating gels it was apparent that there was a selective enrichment for radiolabeled polypeptide that comigrated with the faster moving B-subunit of the doublet of brain tubulin polypeptides. However, the slower moving band of the doublet of the large amount of purified spinach F1P migrated to the same position in the gels (Fig. 21). Heterogeneity of the large subunit of F1P from the bacterium, Hydrogenomas eutrapa has been observed on SDS-PAGE, where the large subunit migrated as 56,000- and 52,000-da1ton polypeptides (Purohit and McFadden, 1976). In higher plants, although the large subunit has been resolved into three or four species showing different isoelectric points (Kung at al, 1975; Kung, 1977), the observation made in the present study is apparently the first evidence of hetero- geneity of the large subunit obtained by SDS-PAGE. This 180 heterogeneity seems to be a general plant character since it was also found in the large subunit of F1P from Cblchicum leaf extract. It is obvious, therefore, that identification of tubulin in plant extracts merely by its mobility is SDS-PAGE is inadequate. This evidence must be accompanied by evidence that it is not the large subunit of F1P. This holds true also for non-green cells and tissues, since F1P is also found in some non-photosynthetic tissues, including storage tissues (Dockerty et al, 1977). The following evidence ruled out the possibility that the radioactive doublet in twice polymerized protein was related to the large subunit of F1P: (i) the radioactive doublet from tobacco cells and the large subunit doublet of spinach F1P behaved differently in SDS-PAGE (Fig. 21); (ii) There was no prominent radioactive band in twice copolymerized protein attributable to the small subunit of F1P which migrates close to the bottom.of the gel (compare tracks C and F in Fig. 21); (iii) the peptide fragment generated from the slower moving band of the large subunit doublet was very different from the patterns generated from the a- and B-subunits of brain tubulin (Fig. 29) and also very different from the patterns generated from the slower moving radioactive band in twice copolymerized protein (not shown due to difficulty in photographing the very faint auto- radiographic image); finally and most convincingly, (v) the radioactive doublet in twice copolymerized protein migrated in the acidic end on isoelectric-focusing (Fig. 30). The large subunit, on the other hand, migrated as four or five isoelectric species, in a more basic region on isoelectric focusing (Fig. 32). 181 Resolution of polypeptides on SDS-PAGE is independent of the other polypeptides in the gel. However, the possibility that the prominent radioactive bands arose as a result of their non-specific adherence to the excess of brain tubulin is rendered unlikely because of l) the non-identical behavior of the slower moving of the prominent radioactive polypeptides from tobacco and the B-subunit of brain tubulin on both 15% and 8% SDS gels (Fig. 21, 22); and 2) the different amounts of radioactivity in the two radioactive poly- peptides such that there is more radioactivity in the slower moving band than in the faster moving band which comigrates with the B-subunit of brain tubulin. If non-specific adherence were occurring the opposite relationship would be expected (Fig. 24c). It is therefore unlikely that the radioactive tobacco doublet is from tobacco tubulin which copolymerizes with brain tubulin. The main evidence regarding the identity of the two radioactive poly- peptides of the copolymerized tobacco doublet however comes from a comparison of the sizes of their peptide fragments with those from the a- and B-subunits of brain tubulin. The fragment patterns of the faster moving radioactive band and the B-subunit of brain tubulin generated by two different proteases were virtually indistinguishable when analyzed on SDS-PAGE (Figs. 25, 27). On this basis, the faster moving radioactive band was identified as the B-subunit of tobacco tubulin. On the other hand, the slower moving radioactive band and the a-subunit of brain tubulin yielded different patterns when treated with S. aureus protease (Figs. 25, 26). When the same bands were treated with a-chymotrypsin, although the fragment patterns were 182 not identical, several fragment sizes were common to the brain and tobacco polypeptides (Figs. 27, 28). There appears to be sufficient homology between the slower moving radioactive band from tobacco and the a-subunit of brain tubulin to justify designation of the former as the a-subunit of tobacco tubulin. The fragment pattern generated from limited proteolysis with a particular combination of enzyme and substrate depends on the relative concentrations of the two or length of incubation. With an increasing ratio of the enzyme to the substrate there is a greater degree of proteolysis and consequently a greater abundance of low- molecular weight peptides in the fragment pattern. Therefore, the same amount of polypeptide can result in different fragment patterns depending on the amount of protease added. This is not a problem when comparing the fragment patterns of the faster moving radio- active band and the B-subunit of brain tubulin, since they are digested together. However, since the slower moving radioactive band has much less protein in it then the o-subunit of brain tubulin in the copolymerized protein and since they are digested separately by the same amount of the enzyme, the difference in their fragment patterns could merely reflect differences in the amounts of protein. If this were true than the fragment pattern of the slower moving radioactive band should have more low molecular weight fragments than the fragment pattern from a-subunit of brain tubulin. However, in most cases, especially in Figure 28, it is clear that the fragment pattern of the o-subunit of brain tubulin actually has more low molecular weight fragments than in the fragment pattern of the slower 183 moving radioactive band. Neither can the difference in the fragment patterns be related to the higher sensitivity of autoradiography, since all radioactive fragments are eqully strong. Therefore, the' differences in the peptide maps of the slower moving radioactive band and the a-subunit of brain tubulin result from qualitative and not quantitative differences. Additional evidence for the identification of the radioactive doublet as tobacco tubulin come from (a) two dimensional gel electrophoresis (isoelectric-focusing/SDS-PAGE) where the radioactive polypeptide comigrate with the corresponding subunits of brain tubulin (Figs. 30,31); (b) 3H-colchicine-tubulin complex and 3SS-labeled copolymerized proteins comigrate on gel filtration with Sephadex G-150 (Fig. 33) and (c) ion-exchange chromatography on DEAE-cellulose (Fig. 34), where most of the twice copolymerized 35S-labeled proteins elutes at 0.31 M KCl, whereas 3H-colchicine- tubulin complex usually elutes between 0.26 M and 0.315 M KCl (Figs. 15, 18, 35). Similarly, most of the 3s[S]-1abeled counts in twice copolymerized protein do not bind to the phosphocellulose column (Fig. 36) consistent with the behavior of brain tubulin (Fig. 37). Since about 70% of the total radioactivity in the twice copolymerized protein is in the doublet (Fig. 23), it can be concluded that the native properties of the tobacco doublet and the brain tubulin are similar. Thus, the putative plant tubulin and brain tubulin have similar properties, both in their native as well as SDS-denatured forms. It is concluded that the 35S-protein from tobacco cells 184 isolated by copolymerization with brain tubulin is tobacco tubulin, which is very similar to, but not identical with, cow brain tubulin. Significance of the Difference between Tobacco and Brain Tubulins Although there is a definite difference between tobacco and brain a-subunits, both.in their mobility on SDS-PAGE and in their peptide fragment patterns, these techniques do not allow us to put a quantitative value on the difference. When reduced and carboxymethylated the two subunits of tubulin can be resolved by polyacrylamide gel electrophoresis in the presence of 8 M urea and 8 M urea and 0.1% SDS (Lee et al, 1973; Bryan and Wilson, 1971; Eipper, 1972). By using different gel concentrations and constructing Ferguson plots, it has been shown that the separation of the a-subunit and the B-subunit is on the basis of charge (Bryan, 1974). Not all SDS-PAGE systems resolve the o- and B-subunits of tubulin. When SDS-PAGE is carried out according to the conditions (high ionic strength and low pH) of Weber and Osborn (1969) the subunits behave as a single protein. This confirms the results in the above mentioned studies in 8 M urea that the tubulin subunits are charge isomers and not size isomers. The large resolution of the two subunits under conditions of high pH and low ionic strength, like the ones used in this study following Laemmli (1970), has been attributed to anomalous behavior of the a-subunit of tubulin in these SDS gels. The reasons for the anamolous behavior are not clear (Bryan, 1974). The a- and B-subunits of tubulin from several different species, including chicken, sea urchin and Tetrahymena co- migrate on polyacrylamide gel electrophoresis in the presence of urea; 185 suggesting evolutionary conservation of charge properties. Similarly the tubulin subunits from chicken or sea urchin are resolved on SDS-PAGE; these phylogenetically divergent a- and B-subunits comigrate. This suggests that the SDS anomaly is also a property of tubulins conserved in evolution. Thus, the subunits of tubulin from different animal species appear to be conserved in size, charge and the property responsible for the anomalous behavior on SDS. The question of conservation of the tubulin subunits is, however, more complex, at least in the lower organisms. The d-subunit of Chlamydomonas showed a slightly increased mobility compared to brain tubulin on a SDS-urea-gel system (Olmsted et al, 1971). The a- subunit of tubulin from mitotic apparatus and from the A tubules of Ciliary outer doublet of sea urchin but not from the flagellar outer doublet of the same organism migrate as a doublet in SDS-urea system but migrate as a single polypeptide in urea- SDS-systems (Bibring et al, 1976). When tubulin from radiolabeled Aspergillus nidulans was isolated by copolymerization with brain tubulin and then analyzed on a SDS-urea gel system, it exhibited two polypeptides. In the same gel system brain tubulin migrated as a single protein species between the two fungal polypeptides. However, when analyzed in a SDS gel system both fungal and brain tubulin migrated as doublets one fungal polypeptide comigrating with the a-subunit of brain tubulin and the other fungal polypeptide migrating slightly faster than the B-subunit of brain tubulin (Sheir-Neiss at al, 1976). Both subunits of the putative tubulin doublets isolated from radiolabeled yeast 186 cells, by copolymerization with brain tubulin, migrated faster than the subunits of brain tubulin on a SDS gel system (Baum.et al, 1978; Shrivers and Byers, 1977) but the fungal tubulin doublet comigrated with the brain tubulin doublet in a slightly different SDS-gel system.(Water and Kleinsmith, 1978). When radiolabeled subunits of tubulin were isolated from the in vitro translation-products of polysomes from sea urchin (Merlino at al, 1978) and Chlamydomonas (Weeks and Collis, 1976), by copolymerization with.brain tubulin, the labeled tubulin subunits comigrated with the subunits of brain tubulin. Finally, the comparison of the mobilities of brain tubulin and cytoplasmic tubulin from Tétrahymena, best exemplifies the dependence of the mobilities of tubulin subunits on the gel conditions: when analyzed on a SDS gel following Laemmli (1970), without carboxymethylation, brain tubulin migrated as a single protein species, while the Tetrahymena tubulin migrated as a doublet, with the slower band comigrating with brain tubulin. However, when analyzed on the same SDS gel system after carboxymethylation, the brain tubulin migrated as a doublet but the Tetrahymena tubulin now migrated as a single band which comigrated with the 8- subunit of brain tubulin. When analyzed on a SDS-gel system following Weber and Osborn (1969), both brain and Tetrahymena tubulins migrated as doublets. When carboxymethylated, the 8- subunit of brain tubulin comigrated with the B-subunit of Tetrahymena tubulin, while the a-subunit of Tetrahymena tubulin migrated slightly faster than the o-subunit of brain tubulin (Maekawa and Sakai, 1978). This discussion illustrates the 187 apparent heterogeneity in tubulin subunits isolated from different sources based on their mobility in polyacrylamide gel electro- phoresis. However, it is also clear that the mobility of the tubulin subunits on SDS-gels is not a simple function of their molecular weights. Heterogeneity of tubulin subunits is also present in the same organism. Witman et al (1972) showed five different Chlamydomanas tubulin species in isoelectric focusing. Cytoplasmic rat brain tubulin has also been resolved into seven to nine components by isoelectric focusing (Gozes and Littauer, 1978), while the a- subunit was preferentially associated with presynaptic membrane (Gozes and Littauer, 1978). Unlike the numerous comparative studies on the mobilities of tubulin subunits from different sources on different gel systems discussed above, little comparative work has been done in comparing the amino acid sequences or peptide fragments of subunits of tubulins from different species. Partial sequence data on the o- and B-subunits from sea urchin and chicken brain tubulin (see General Introduction) shows that the subunits are evolutionarily conserved (Luduena and Woodward, 1973). Similarly, the amino acid composition and the fingerprints of 15I-labeled tryptic peptides of platelet and brain tubulins showed considerable similarity (Castle and Crawford, 1977). The peptide fragments of tubulin from brain and Chinese hamster ovary cells were indistinguishable (Spiegelman at al, 1977). When the peptide fragments of the putative tubulin doublet from yeast which, as 188 already mentioned earlier, migrated slightly faster than the brain tubulin doublet on SDS-electrophoresis, was analyzed by two dimensional electrophoresis/chromatography the maps showed little homology with brain tubulin (Shrivers and Byer, 1977). It is possible that fungal tubulin is different from tubulins in other organisms. However, much more comparative biochemistry of the tubulin subunits especially amino acid sequencing, needs to be done before we know how highly conserved tubulins are in evolution. An intriguing difference has been observed between the brain tubulin and tobacco tubulin purified either by copolymerization with brain tubulin and by its self assembly. Whereas equal amounts of the a- and B-subunits appear to be present in brain tubulin, the ratio of a-subunit to B-subunit in tobacco tubulin is much higher. Although the amounts of protein in the two subunits of tobacco tubulin have not been directly measured, the intensity of both Coomassie blue staining and autoradiographic image suggest an excess of a-subunit over B-subunit. This is surprising since the a- and.B-subunits are found in close to equimolar amounts, regardless of the source of tubulin (Bryan and Wilson, 1971; Feit at al, 1971; see Bryan, 1974). Although there are a few reports of unequal amounts of tubulin subunits (Jacobs and McVittae, 1970; Witman et al, 1972) but these results could also be interpreted as artifacts of protein quantitation in polyacrylamide gel electrophoresis as demonstrated by Bibring and Baxandall (1974). Using the same sample of tubulin from outer doublet of sea urchin sperm tail, they demonstrated a range of ratios of the d-subunit to the B-subunit from about equimolar at high loading to complete loss of B-subunit at low‘ 189 loading, could be achieved. Although failure of quantitative staining may explain the contradictory data regarding the relative amounts of the a- and 8- subunits obtained from outer doublet tubulin it is clear from the present study that both Coomassie blue staining as well as autoradiography show the presence of lower levels of B-subunit relative to the a-subunit in tobacco tubulin. It is therefore unlikely that the difference is due to unequal staining. It most probably represents a real difference in the relative amount of the B-subunit at least in polymerized tobacco tubulin. The low level of B-subunit observed in SDS-gels is neither due to the failure of this polypeptide to enter the gel, nor due to its proteolysis, since the B-subunit of brain tubulin, in copolymerized protein, is not preferentially lost. Cleveland et al (1978) have shown that 67% of the a- and 13% of the B-subunits of tubulin produced by in vitro translation of chick brain mRNAs, were competent to copolymerize with carrier tubulin. They also showed that all of the newly synthesized a-subunit of tubulin, recovered after passing over phosphocellulose column, was a competent to polymerize as freshly cycled carrier tubulin. This suggests that a rapid exchange of subunits into dimers occurs. Detrich and Williams (1978) recently demonstrated concentration dependent reversible dissociation of the a8 dimer of tubulin (with an approximate dissociation constant of 8 x 10”7 M). They also showed that reconstitution of the dimer following its dissociation resulted in about 30% of loss of its original activity. If tobacco tubulin is present both as a dimer 190 and as the a- and B-subunits and if the a-subunit c0polymerizes more efficiently than B-subunit then the a-subunit can be preferentially enriched by cycles of copolymerization. The poor efficiency of copolymerization of B-subunit may be related to a requirement for its covalent modification prior to assembly. Phosphorylation has been shown to be specific for the B-subunit of brain tubulin (Eipper, 1972). Besides the prominent doublet several radioactive polypeptides, of both lower and higher molecular weights than the doublet, were detected in autoradiograms of the twice copolymerized protein. There are two major non-tubulin 35S-polypeptides of a higher molecular weight than tubulin that are prominent in most preparations of copolymerized tobacco proteins (Figs. 24). It is not known whether these polypeptides represent non-specific contaminants that are copolymerized inefficiently, relative to tubulin, and would be lost after a few more cycles of assembly- disassembly, or if they represent polypeptides that copurify with tubulin in a constant manner. HOwever, if there are such proteins (microtubule associated proteins) they have molecular weights different from those associated with brain tubulin. There is also enrichment for the high molecular weight proteins during self assembly of tubulin. One of these copurified proteins is present in as high an amount as the a-subunit of tubulin. One wonders if it is in any way related to the lack of a B-subunit of tubulin in such preparations (Fig. 24c). 191 Finally, the structure and the nature of polymers formed by in vitro assembly or coassembly of tobacco tubulin is not known. Although it is assumed that hybrid microtubules made up of tobacco and brain tubulins are formed when brain tubulin was incubated in the presence of 35S-labeled tobacco proteins, there is no direct evidence for this. However, it is likely that copolymers of some sort were formed, even if no hybrid microtubules are formed. This is suggested by the data in Table 13 which shows that a constant specific radioactivity was approached after three cycles of copolymerization. The specific radioactivity in the copolymerized proteins should have been determined for further cycles of assembly-disassembly to ascertain the constancy of the specific radioactivity. However, this was not possible due to large losses of both plant and brain proteins during copolymerization. If the specific radioactivity, in fact, remains constant after a few cycles of copolymerization than it most likely means that at least some plant and brain tubulins assemble into a copolymer with the same efficiency. In summary, the enrichment of tobacco tubulin by copolymerization with brain tubulin has led to the first significant, though limited, characterization of tubulin from a higher plant. Tubulins from tobacco and cow brain behave similarly on gel filtration, ion-exchange chromatography and two-dimensional isoelectric focusing/SDS-PAGE. Whereas the B-subunit of tubulins from tobacco and cow brain were indistinguishable on the basis of both their mobility on SDS-PAGE and their peptide fragments after limited proteolysis, the a-subunits 192 from these sources were distinguishable on the basis of the same two criteria. Although the significance of the observed differences in the a-subunits from tobacco and brain tubulins is not known, one can speculate that the differences may be related to the apparent poor CLC binding activity in plant tubulin, as was apparent in section II of this thesis. In the final analysis, the similarities in the properties of tubulins from brain and tobacco are more striking than the observed differences between them. It appears that tubulin has been highly conserved during evolution. The actual degree of evolutionary conservation of tubulin will be established following a more detailed characterization of tubulin from higher plants. For instance, what is the mechanism of and the conditions for assembly-disassembly of plant tubulin? Is plant tubulin also associated with enzymatic activities, such as kinase, tyrosinase and GTPase? Does plant tubulin have proteins that copurify with it during cycles of assembly-disassembly, and if so, are they different from the so-called MAPS from brain tissue? Such a characterization of plant tubulin will, of course, require its isolation independent of brain tubulin. One approach would be to "force" the self- assembly of tubulin in plant extracts by agents, such as D O, 2 Dextran T-10, polyethylene glycol, MAPS, etc. Some success in this direction has been reported in this section of this thesis. Another approach would be use of antibodies specific for tubulin to pull out tubulin from plant extracts. A comparative study of tubulin from such diverse sources as a higher plant and mammalian brain, may allow us to test the generalization of a given tubulin property. This is presently not possible, since most models of tubulin function are 193 based entirely on the brain tubulin. LITERATURE CITED LITERATURE CITED Allen, C. and Borisy, G. G. (1974) J. Mol. Biol. 89: 381-402. Alt, F. W., Kellems, R. E. and Schimke, R. E. (1976) J. Biol. Chem. 251: 3063-3074. Ames, G. F. and Nikaido, K. (1976) Biochem. 15: 616-623. Amrhein, N. and Filner, P. (1973) FEBS Lett. 33: 139-142. Bamburg, J. R., Shooter, E. M. and Wilson, L. (1973) Biochem. 12: 1476-1482. Barnes, L. D., Engel, A. G. and Dousa, T. P. (1975) Biochim. Biophy. Acta 405: 422-433. Baum, P., Thorner, J. and Honig, L. (1978) Proc. Natl. Acad. Sci. US 75: 4962-4966. Bhattacharya, B. and Wolff, J. (1974) Biochem. 13: 2364-2369. Bhattacharya, B. and WOlff, J. (1976) Proc. Natl. Acad. Sci. US 13: 2375-2378. Bibring, T. and Baxandall, J. (1974) Exptl. Cell Res. 86; 120-126. Bibring, T., Baxandall, J., Denslow, S. and Walker, 8. (1976) J. Cell Biol. 62; 301-312. Binder, L. I., Dentler, W. L. and Rosenbaum, J. L. (1975) Proc. Natl. Acad. Sci. US 22; 1122-1126. Binder, L. I. and Rosenbaum, J. L. (1976) Biol. Bull. (Woods Hole) 145: 424-425. Borgers, M., De Brabander, M. (eds.) (1975) Microtubules and Microtubule Inhibitors. Amsterdam-Oxford: North Holland Publ. Co. 1975, New York: American Elsevier Publ. Co., Inc. 1975. Borgers, M., De Nollin, S., De Brabander, M. and Thienpont, D. (1975) Am. J. Vet. Res. 26: 1153-1166. 194 195 Borisy, G. G. (1972) Anal. Biochem..50: 373-385. Borisy, G. G., Marcum, J. M., Olmsted, J. 8., Murphy, D. B. and Johnson, K. A. (1975) Ann. N.Y. Acad. Sci. 253 : 107-132. Borisy, G. G. and Taylor, E. W. (1967) J. B. C. 34; 525-533. Bouck, G. B. and Brown,D. L. (1973) J. Cell Biol. 56: 340-59. Bryan, J. (1972) Biochem. 11; 2611-2616. Bryan, J. (1974) Fed. Proc 33; 152-157. Bryan, J. (1975) Am. 2001. 15: 649-660. Bryan, J., Nagle, B. W., and Donges, K. H. (1975) Proc. Natl. Acad. Sci. 12: 3570-3574. Bryan, J. and Wilson, L. (1971) Proc. Natl. Acad. Sci. US 8; 1762-1766. Burgess, J. and Northcote, D. H. (1969) J. Cell Sci 5: 433-451. Burns, R. G. (1973) Exptl. Cell Res. 81: 285-292. Burns, R. G. & Pollard, T. D. (1974) FEBS Lett. 49; 274-280. Burns, R. G. and Starling, D. (1974) J. Cell Sci. 14: 411-419. Cande, W. Z., Snyder, J., Smith, D., Summers, K. and McIntosh, J. (1974) Proc. Natl. Acad. Sci. US 21; 1559-1563. Castle, A. G. and Crawford, N. (1975) FEBS Lett. 51; 195-200. Castle, A. G. and Crawford, N. (1977) Biochim. Biophy. Acta 494: 76-91. Cleveland, D. W., Fischer, 8. G., Kirschner, M. W. and Laemmli, U. K. (1977) J. B. C. 252: 1102-1106. Cleveland, D. W., Hwo, S. Y. and Kirschner, M. W. (1977) J. Mol. Biol. 116: 207-225. Cleveland, D. W., Kirschner, M. W. and Cowan, N. J. (1978) Cell _1_§_: 1021-1031 . Comandon, J. and Defonbrune, P. C. (1942) C. R. Soc. Biol. Paris 136: 410-411, 460-461, 746-748. Cornman, I. (1941) Biol. Bull. 81: 297-98. Cornman, I. (1942) Botanical Gazett 104: 50-61. 196 Cortese, F., Bhattacharya, B. and Wolff, S. (1977) J. B. C. 252: 1134-1140. Davidse, L. C. and Flach, W. (1977) J. Cell Biol. 12: 174-193. Delmer, D. P. (1979) Plant Physiol. 63: 144a. Detrich, III, H. W. and Williams, R. C., Jr. (1978) Biochem. 11: 3900-3907. Dustin, P. (1978) Microtubules. Springer-Verlag. Eigsti, O. J. and Dustin, P., Jr. (1955) Colchicine In: Agriculture, Medicine, Biology and Chemistry. Ames, Iowa: Iowa State College Press. Eipper, B. A. (1972) Proc. Natl. Acad. Sci. US 69: 2283-2287. Erickson, H. P. and Voter, W. A. (1976) Proc. Natl. Acad. Sci. US 13: 2813-2817. Farrel, K. W. and Burns, R. G. (1975) J. Cell Sci. 11; 669-681. Farrel, K. W., Morse, A. and Wilson, L. (1979) Biochem. 18: 905-911. Farrel, K. W. and Wilson, L. (1978) J. Mol. Biol. 121: 393-410. Feit, H., Slusarek, L. and Shelanski, M. L. (1971) Proc. Natl. Acad. Sci. 68; 2028-2031. Filner, P. (1965) Exptl. Cell Research.§2; 33-39. Filner, P. and Behnke, O. (1973) J. Cell Biol. 52; 99a. Filner, P. and Yadav, N. S. (1979) In: Physiology of Movements, New Series. Vol. 7 Enclyclopedia of Plant Physiology (eds. Haupt, W. and Feinleib, M. E) Springer-Verlag, Berlin Heidelberg. Forer, A. and Zimmerman, A. M. (1974) J. Cell Sci. 16; 481-497. Fuller, G. M., Brinkley, B. R. and Boughter, J. M. (1975) Science 187: 948-950. Gaskin, F., Cantor, C. R. and Shelanski, M. L. (1974) J. Mol. Biol. 89: 737-758. Goldman, R., Pollard, T. and Rosenbaum, J. L. (eds.) (1975) Cell Motility. Cold Spring Harbor Laboratory. New York. Gozes, I. and Littauer, U. Z. (1978) Nature 276: 411-413. Gozes, I. and Littauer, U. Z. (1979) FEBS Lett. 22; 86-90. 197 Gross, P. R. and Spindel, W. (1960) Ann. N.Y. Acad. Sci. 29; 500- 522. Haber, J. E., Peloquin, J. G., Halvorson, H. O. and Borisy, G. E. (1972) J. Cell Biol. 55; 355-367. Hart, J. W. and Sabnis, D. D. (1973) Planta 109: 147-152. Hart, J. W. and Sabnis, D. D. (1976a) J. Exp. Bot. 21; 1353-1360. Hart, J. W. and Sabnis, D. D. (1976b) Curr. Adv. in Plant Sci. 36; 1095-1104. Heath, 1. B. (1975) Protoplasma 85; 177-192. Heidemann, S. R. and Kirschner, M. W. (1975) J. Cell Biol. 62; 105-117. Hepler, P. K. (1976) Plant Microtubules In: Plant Biochemistry 3rd Ed. (eds. Bonner, J. and Varner, J. E.) Acad. Press, N. Y. pp. 147-187. Hepler, P. K. and Jackson, W. T. (1969) J. Cell Sci. 5; 727-741. Hepler, P. K. and Palevitz, B. A. (1974) Ann. Rev. Plt. Physiol. 25; 309-62. Herzog, W. and Weber, K. (1977) Proc. Natl. Acad. Sci. US 24: 1860-1864. Herzog, W. and Weber, K. (1978) Eur. J. Biochem. 21: 249-254. Heyn, A. N. J. (1972) J. Ultrastruct. Res. 49; 433-57. Himes, R. H., Burton, P. R., and Gaito, J. M. (1977) J. Biol. Chem. 252: 6222-6228. Hoebeke, J.,‘fithijen, G. and De Brabander, M. (1976) Biochem. Biophys. Res. Commun. 92; 319-324. Hotta, Y. and Shepard, J. (1973) Molec. Gen. Genet. 122: 243-260. Ikeda, Y. and Steiner, M. (1976) J. Biol. Chem. 251: 6135-6141. Ihoue, S., Borisy, G. G. and Kiehart, D. P. (1974) J. Cell Biol. 62; 175-184. Jacobs, M. and.McVittie, A. (1970) Exptl. Cell Res. 63; 53- Jeffs, R. A. and Northcote, D. H. (1967) J. Cell Sci. 2; 77-78. Jenson, R. G. and Bahr, J. T. (1977) Ann. Rev. Plant Physiol. 28; 379-400. 198 Jockusch, B. M. (1973) Ber. Dtsch. Bot. Ges. 86; 39. Johnson, K. A. and Borisy, G. G. (1977) J. Mol. Biol. 111: 1-31. Kane, R. E. (1965) J. Cell Biol. 25; 137-144. Kane, R. E. (1975) J. Cell Biol. 66; 305. Kane, R. E. (1976) J. Cell Biol. 11; 704. Kirkpatrick, J. B., Hyams, L., Thomas, V. L. and Howley, P. M. (1970) J. Cell Biol. 41: 384-394. Kung, S. D. (1977) Am. Rev. Plant Physiol. 28: 401-437. Kung, S. D., Gray, J. C., Wildman, S. G. and Carlson, P. S. (1975) Science 187: 353-355. Kuriyama, R. (1976) J. Biochem. 80: 153-165. Laemmli, U. K. (1970) Nature 222; 680-85. Laskey, R. A. and Mills, A. D. (1975) Eur. J. Biochem. 56; 335-341. Ledbetter, M. C. and Porter, K. R. (1963) J. Cell Biol. 12; 239-50. Lederberg, S. and Setten, G. (1970) Science 168; 485-87. Lee, J. C., Frigon, R. P. and Timasheff, S. N. (1973) J. Biol. Chem. 248: 7253-7262. Lee, J. C. and Timasheff, S. N. (1975) Biochem. 14: 5183-5187. Lee, J. C. and Timasheff, S. N. (1977) Biochem. 16; 1754-1764. Lescure, A. M. and Filner, P. (1974) Plant Research, p. 126. Levan, A. (1940) Hereditas 22; 471-486. Levan, A. and Steinegger, E. (1948) Hereditas 33: 552-566. Ling, V. and Thompson, L. H. (1974) J. Cell Physiol. 83: 103-116. Lowry, O. H., Rosebrough, N. J., Farr, A. L. and Randall, R. J. (1951) J. Biol. Chem. 193: 265-275. Luduena, R. F. and Woodward, D. O. (1973) Proc. Natl. Acad. Sci. US 19; 3594-3598. Maekawa, s. (1978) J. Biochem. 34; 641-646. Maekawa, S. and Sakai, H. (1978) J. Biochem. 83: 1065-1075. 199 Margolis, R. L. and Wilson, L. (1977) Proc. Natl. Acad. Sci. US 22; 3466-3470. Margolis, R. L. and Wilson, L. (1978) Cell 33; 1-8. Marsland, D. and Zimmerman, L. (1963) Exptl. Cell Res. 39; 23-25. Mazia, D. (1955) Symp. Soc. Exptl. Biol. 3; 335-357. Mazia, D., Mitchison, J. M., Medina, H. and Harris, P. (1961) J. Cell Biol. 39: 467-474. Mehra, P. N. and Khusoo, T. N. (1948) Curr. Sci. Bangalore 31: 242-243. Merlino, G. T., Chamberlain, J. P. and Kleinsmith, L. J. (1978) J. Biol. Chem. 253: 7078-7085. Minor, P. D. and Roscoe, D. H. (1975) J. Cell Sci. 31; 381-396. Mizel, S. B. and Wilson, L. (1972) Biochem. 33; 2573-2578. Mohri, H. (1976) Biochim. Biophys. Acta 456: 85-127. Montague, W., Howell, S. L. and Green, I. C. (1975) Biochem. J. 148: 237-243. Nagle, B. W., Doenges, K. H. and Bryan, J. (1977) Cell 33; 573-86. Newcomb, E. H. and Bonnett, H. T. (1965) J. Cell Biol. 32; 575-89. Nickolson, V. J. and Veldstra, H. (1972) FEBS Lett. 33; 309-313. O'Farrel, P. H. (1975) J. Biol. Chem 339; 4005-4021. Olmsted, J. 8., Marcum, J. M., Johnson, K. A., Allen, C. and Borisy, G. G. (1974) J. Supraniol. Struct. 3; 429-450. Olson, L. W. (1973) Arch. Mikrobiol. 23; 281-288. Ostlund, R. and Pastan, I. (1975) Biochem. 35; 4064-68. Patzelt, C., Singh, A., Marchand, Y. L., Orci, L. and Jeanvenaud, B. (1975) J. Cell Biol. 33; 609-620. Pickett, Heaps, J. D. (1974) Plant Microtubules In: Dynamic Aspects of Plant Ultrastructure. (ed. A. W. Robards). McGraw-Hill Book Co. (U.K.) pp. 219-255. Pipelleers, D. G., Pipeleers-Marichal, M. A., Sherline, P. and Kipnis, D. M. (1977) J. Cell Biol. 22; 341-350. 200 Purohit, K. and McFadden, B. A. (1976) Biochem. Biophys. Res. Commun. 33; 1220-1227. Puszkin, E., Puszkin, S. and Aledort, L. M. (1971) J. Biol. Chem. 246: 271—276. Rebhun, L. I., Jemiolo, D., Ivy, N., Mellon, M. and Nath, J. (1975) Ann. N.Y. Acad. Sci. 253: 362-377. Rebhun, L. I., Mellon, M., Jemiolo, D., Nath, J. and Ivy, N. (1974) J. Supramol. Struct. 3; 466-485. Rosenbaum, J. L. and Carlson, K. (1969) J. Cell Biol. 39; 61-77. Rosenbaum, J. L., Moulder, J. B. and Ringo, D. L. (1969) J. Cell Biol. 33: 400. Rosenbaum, J. L., Moulder, J. B. and Ringo, D. L. (1969) J. Cell Biol . 33: 600-619. Sakai, H. (1966) Biochim. Biophys. Acta 112: 132-145. Sakai, H. and Kuriyama, R. (1974) Development, Growth and Differentiation 33; 123-134. Schnepf, E. and Deichgraber, A. (1976) Cytobiologie 33; 341-353. Schonharting, M., Mende, G., and Siebert, G. (1974) Hoppe-Seyler's Z. Phys. Chem. 355: 1391-1399. Sheir-Neiss, G., Nardi, R. V., Gealt, M. A. and Morris, N. R. (1976) Biochem. Biophys. Res. Commun. 33; 285. Shelanski, M. L., Gasking, F. and Cantor, C. R. (1973) Proc. Natl. Acad. Sci. US 19; 765-768. Sherline, P., Leung, T. and Kipnis, D. M. (1975) J. Biol. Chem. 250: 5481-5486. Sheterline, P. and Schofield, J. G. (1975) FEBS Lett. 39; 297-302. Shriver, K. and Byers, B. (1977) J. Cell Biol. 33; 297a. Sloboda, R. D., Deutler, W. L. and Rosenbaum, J. L. (1976) Biochem. 33: 4497-4505. Snell, W. J., Deutler, W. L., Haimo, L. T., Binder, L. I. and Rosenbaum, J. L. (1974) Science 185: 357-360. Snyder, J. A. and McIntosh, J. R. (1976) Ann. Rev. Biochem. 33; 699-720. 201 Soifer, D. (ed.) (1975) The Biology of Cytoplasmic Microtubules. New York Academy of Sciences, Annals, Vol. 253. New York, NY. Spiegelman, B. M., Penningworth, S. M. and Kirschner, M. W. (1977) Cell 33: 587-600. Steiner, M. (1978) Nature 333; 834-835. Stephens, R. E. (1968) J. Mol. Biol. 33:: 517-519. Stephens, R. E. and Edds, K. T. (1976) Physiol. Rev. 33; 709-777. Takenaka, Y. (1950) Cytologia 33; 95-99. Taylor, E. W. (1965) J. Cell Biol. 33;, 145-60. Timasheff, S. N. (1979) Trends in Biochemical Sciences;g: 61-65. Water, R. D. and Kleinsmith, L. J. (1976) Biochem. Biophys. Commun. 33; 704-708. Weber, K. and Osborn, M. (1969) J. Biol. Chem. 244: 4406-4412. Weeks, D. P. and Collis, P. S. (1976) Cell 2; 15-27. Weingarten, M., Lockwood, A. H., Hwo, S. Y. and Kirschner, M. W. (1975) Proc. Natl. Acad. Sci. US 33; 1858-1862. Weisenberg, R. C. (1972) Science 127: 1104-1105. Weisenberg, R. C., G. C. Borisy, and Taylor, E. W. (1968) Biochem- istry 2; 4466-4479. Weisenberg, R. C. and Rosenfeld, A. C. (1975) J. Cell Biol. 33; 146-158. Wiche, G. and Cole, R. D. (1976) Proc. Natl. Acad. Sci. US 33; 1227-1231. Wiche, G., Lundblad, V. J. and Cole, R. D. (1977) J. Biol. Chem. 252: 794-796. Williams, N. E. and Jackel-Williams, R. (1976) J. Cell Sci. 20: 61-77. Wilson, L. (1970) Biochemistry 3: 4999-5007. Wilson, L., Bamburg, J. R., Mizel, S. 3., Grisham, L. M. and Creswell, K. M. (1976) Fed. Proced. 33; 158-66. 202 Wilson, L. and U. Friedkim. (1967) Biochem. 3: 3126-3135. Wilson, L. and Meza, I. (1973) J. Cell Biol. 33; 709-719. Wilson, L., Morse, A.N.C. and Bryan, J. (1978) J. Mol. Biol. 121: 255-268. Wishnick, M. and Lane, M.D. (1971) Methods in Enzymology XXIII: 570-577. Wishnick, M., Lane, M. D. and Scrutton, M. C. (1970) J. Biol. Chem. 245: 4939-4947. Witman, G. B., Carlson, K., Berliner, J. and Rosenbaum, J. L. (1972) J. Cell Biol. 33; 507-539. Witman, G. B., Cleveland, D. W., Weingarten, M. D. and Kirschner, M. W. (1976) Proc. Natl. Acad. Sci. US 13; 4070-4074. Wooding, F. B. p. (1969) Planta 35: 284-298.