ACTIVE SPECIES AND QUATERNARY STRUCTURE OF BIODEGRADATIVE L - THREONINE DEHYDRASE FROM ESCHERICHIA COLI III.- Dissertation for the Degree of Ph. D. MICHIGAN STATE UNIVERSITY ROBERT CARROLL MENSON 1976 This is to certify that the thesis entitled ACTIVE SPECIES AND QUATERNARY STRUCTURE OF BIODEGRADATIVE L-THREONINE DEHYDRASE FROM ESCHERICHIA COLI presented by Robert Carroll Menson has been accepted towards fulfillment of the requirements for Ph . D . degree in Biochemistry // gem—e776 Major professor Daté‘T/f 2 5 /‘/7% 0-7639 ABSTRACT ACTIVE SPECIES AND QUATERNARY STRUCTURE OF BIODEGRADATIVE L-THREONINE DEHYDRASE FROM ESCHERICHIA COLI BY Robert Carroll Menson Biodegradative L-threonine dehydrase catalyzes the irreversible deamination of L-threonine to form a-keto- butyrate. AMP is an allosteric effector which activates the enzyme through a 16-fold decrease in the Km for L- threonine from 50 mM to 3.1 i 1.0 mM (Dunne, C. P., Gerlt, J. A., Rabinowitz, K. W., and Wood, W. A. (1973) g; Biol. Chem. 248, 8189-8199). This activation is de- pendent on protein concentration to the second order and therefore requires a two-fold change in the state of association (Gerlt, J. A., Rabinowitz, K. W., Dunne, C. P., and Wood, W. A. (1973) g; Biol. Chem. 248, 8200-8206). The active species in the absence of AMP has a limiting sedimentation coefficient of 3.6 S as determined by active enzyme centrifugation using two coupled assay systems. A similar value was obtained by active enzyme centrifugation using an assay system which followed the product of the reaction directly. The active enzyme in Robert Carroll Menson the absence of AMP undergoes a protein dependent associ- ation which is affected by threonine concentration. Dimer to monomer dissociation constants of 4.50 x 10-6 M and 2.34 X 10-6 M were calculated for equilibrium in the presence of 400 mM and 100 mM threonine, respectively. The active enzyme in the presence of its activator, AMP, has a limiting sedimentation coefficient of 5.8 S with higher values observed which are the product of protein concentration or other experimental conditions. The enzyme associates during active enzyme centri- fugation as indicated by a change in the 520 w value from I 3.6 S to 6.86 S when enzyme diluted in a non-AMP containing buffer is layered on an AMP containing assay mix. Upon removal of AMP from the dehydrase during active enzyme centrifugation, the enzyme dissociates as indicated by a decrease in the sedimentation coefficient from about 6.1 S to 3.7 S. A concomitant decrease in activity is observed as would be expected from the 16-fold higher Km observed in the absence of AMP than in the presence of AMP. The native dehydrase has a molecular weight of 70,500 as determined by polyacrylamide gel electrOphoresis, 84,000 1 5,000 as determined by equilibrium ultracentri- fugation and a 3 value of 5.7 S. It is made up of —20 ,W identical monomers consisting of a single polypeptide chain. An average molecular weight of 35,900 was obtained Robert Carroll Menson on sodium dodecyl sulfate polyacrylamide gels under a variety of reducing conditions for both the native dehy- drase and the carboxymethyl derivative. A value of 34,100 was observed with equilibrium ultracentrifugation in 6 M guanidine hydrochloride plus 0.2 M mercaptoethanol. Thus the allosteric control is exhibited through an AMP-induced monomer to dimer inter-conversion. Further asSociation to higher oligomers is observed but does not play a significant role in the allosteric control. Another threonine dehydrase, designated threonine dehydrase II, isolatable from E; 9911 is reported. In contrast to previously reported threonine dehydrases found in §;_ggli it is neither isoleucine inhibited nor AMP activated. This enzyme has a Km for L-threonine of 4.6 mM in the presence or absence of AMP. Threonine dehydrase II copurifies with the biodegradative threonine dehydrase in a nearly constant ratio from the crude ex- tract to the highly purified dehydrase. Ferguson plots indicate that it is a size isomer with electrophoretic properties identical to those of the biodegradative dehydrase. A native molecular weight of 136,000 and a molecular weight of 69,000 for the denatured enzyme was established using polyacrylamide gel electrophoresis. The native molecular weight does not appear to be af- fected by AMP. Robert Carroll Menson As evidenced by the integrity of a peak for thre- onine dehydrase II under a variety of reducing conditions with both native and carboxymethylated enzyme it is not a readily convertible disulfide oligomer of biodegradative threonine dehydrase. Despite the similarity of charge as assessed by electrophoretic mobility and copurification through several absorption chromatography columns initial amino acid analysis indicates that there are significant differ- ences between dehydrase II and the biodegrative dehydrase. ACTIVE SPECIES AND QUATERNARY STRUCTURE OF BIODEGRADATIVE L-THREONINE DEHYDRASE FROM ESCHERICHIA COLI BY Robert Carroll Menson A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Biochemistery 1976 DEDICATION To my wife, Susan, and daughter, Jennifer for their time, faith and support. To my son, Jeremy, whose imminent arrival set the completion date. And to my parents for their many years of giving, I hope this makes up in some small way. ii ACKNOWLEDGMENT I would like to thank Dr. W. A. Wood for his ‘guidance and leadership both as a mentor and as a friend and in whose laboratory all this was possible. I acknow— ledge the early guidance of Drs. Hammerstedt, and Robertson who made a beginning graduate students indoc- trination to the world of research an interesting experi- ence. The many discussions and friendship of Dr. C. P. Dunne during the formative stage and actual thesis work are also recognized. A special thanks to Drs. Kemper and Everse at the University of California, San Diego for tutorship in Active Enzyme Centrifugation and to Dr. N. 0. Kaplan, of the above institution and Dr. A. T. Phillips, Pennsylvania State University, State College for the use of their analytical ultracentrifuge. Thanks go to my committee for discussions, input and time spent developing this thesis research. The financial support of USPHS predoctoral traineeship is gratefully recognized. iii TABLE OF CONTENTS DEDICATION . . . . . . . . . . . . . ACKNOWLEDGMENTS . . . . . . . . . . . LIST OF TABLES . . . . . . . . . . . LIST OF FIGURES . . . . . . . . . . . INTRODUCTION . . . . . . . . . . . . LITERATURE REVIEW . . . . . . . . . . Active Enzyme Structure . . . . . . . . Active Enzyme Centrifugation . . . . . . Calculations . . . . . . . . . . Theoretical Considerations . . . Active Species of Enzymes which Exhibit Association-Dissociation Equilibria . . Active Species of Enzymes which Exhibit Allosteric Control . . . . . Enzymes which Undergo a Substrate Induced Change in the State of Oligomerization . Quaternary Structure of Threonine Dehydrase . Molecular Species in the Absence of AMP . . Molecular Species in the Presence of AMP . Requirements for Association . . . . . Subunit Composition . . . . . . . . Stoichiometry . . . . . Allosteric Regulation of Enzyme Activity . . Concerted Transition Model . . . . Induced Fit Sequential Isomerization Model . Ligand Induced Oligomerization Model . . . Threonine Dehydrases . . . . . . . . Biodegradative Dehydrase of Clostridium Tetanomorphum . . . . . . Biodegradative Dehydrase from Escherchia coli Other Oligomeric Systems . . . . . . . METHODS AND MATERIALS . . . . . . . . . Bacteriological . . . . . . . . . . iv Page ii iii vii ix I7 19 21 22 26 27 28 29 3O 30 33 35 35 39 46 47 47 Page Chemicals . . . . . . . . . . . . 48 Purification of Threonine Dehydrase . . . . . 49 Crude Extract . . . . . . . . . . 49 Protamine Sulfate Treatment . . . . . . . 50 Ammonium Sulfate Precipitation . . . . . . 50 DEAE Sephadex Chromatography--First Column . . 51 Hydroxylapatite Chromatography . . . . . . 52 DEAE Sephadex Chromatography--Second Column . 53 Determinations and Procedures . . . . . . . 53 Protein . . . . . . . . . . . . . 53 Photometric Determinations . . . . . . . 57 Removal of AMP . . . . . . . . . . . 59 Other . . . . . . . . . . . . . . 59 Data Analysis . . . . . . . . . 60 Assays for Threonine Dehydrase . . . . . . 60 Coupled Spectrophotometric Assay . . . . . 60 Direct Spectrophotometric Assay . . . . . 61 Polyacrylamide Gel Electrophoresis . . . . . 61 Equipment . . . . . . . . . . . 61 Analytical Gel Electrophoresis . . . . . . 61 Native Molecular Weights . . . . . . . 62 Polyacrylamide Gel Electrophoresis in Sodium Dodecyl Sulfate . . . . . . . . . . 64 Polyacrylamide Gel Staining Procedures . . . 65 Activity Stain . . . . . . . . . . . 67 Urea Gels . . . . . . . . . . 68 ElectrOphoretic Purification of Threonine Dehydrase . . . . . . . . . . 68 Analytical Ultracentrifugation . . . . . . 69 Active Enzyme Centrifugation . . . . . . . 70 Equipment . . . . . . . . . . . . . 72 Cell Loading . . . . . . . . . . . . 72 Density Gradient . . . . . . . . . . 75 Calculations . . . . . . . . . . . . 78 Assay . . . . . . . . . . . . . . 80 RESULTS . . . . . . . . . . . . . . . 82 Characterization of Highly Purified Threonine Dehydrase . . . 85 AMP Activation of Electrophoretically Purified Threonine Dehydrase . . . . . . . . . 92 Amino Acid Analysis . . . . . . 94 Polyacrylamide Gel Electrophoresis . . . . 94 Active Enzyme Centrifugation . . . . . 99 Active Enzyme Centrifugation in the Presence of AMP . . . . . . . 102 Active Enzyme Centrifugation in the Absence of AMP . . . . . . . . . . . . . 118 Association of the Active Form of Threonine Dehydrase Upon the Addition of AMP . . . . Dissociation of the Active Dehydrase Oligomer Upon Removal of AMP . . . . . . Determination of the Molecular Weight of Native Threonine Dehydrase . . . . Polyacrylamide Gel ElectrOphoresis Molecular Weights of Native Threonine Dehydrase . . . High Speed Equilibrium Ultracentrifugation . . Determinatidn of Native Molecular Weight by Conventional Sedimentation Equilibrium Analysis . . . . . . . . . . . . Sedimentation Velocity . . Determination of the Protomer Molecular Weight and Composition . . . . . . . Investigation of the Protomer Molecular Weight and Structure Using Polyacrylamide Gel Electrophoresis . . . Analysis of the Carboxymethyl Cysteine Derivative of Threonine Dehydrase by Conventional Sedimentation Equilibrium . . . . . . . DISCUSSION . . . . . . . . . . . . . . Protomer Molecular Weight and Composition . . . Molecular Weight of Native Threonine Dehydrase . Molecular Structure as Determined by Active Enzyme Centrifugation . . . . . . . . . SUMMARY . . . . . . . . . . . . . . . BIBLIOGRAPHY . . . . . . . . . . . . . APPENDIX . . . . . . . . . . . . . . vi Page 138 143 148 149 155 170 176 176 179 205 212 212 219 228 238 241 248 Table 10. 11. 12. LIST OF TABLES Comparison of Allosteric Models . . . . . Kinetic Parameters for the Biodegradative L-threonine Dehydrase of E; coli. . . . . Purification of Threonine Dehydrase . . . . A Comparison of Lowry and Fluorescamine Methods for Determination of Threonine Dehydrase Concentration . . . A. . . . Comparison of Protein.Concentrations as Determined by Dry Weight, Lowry and Fringe Number in the Ultracentrifuge . . . . . Relative Amounts of the Two Forms of Threonine Dehydrase as Determined by Polyacrylamide Gel ElectrOphoresis . . . . . . . . Activation of Electrophoretically Purified Threonine Dehydrase by AMP . . . . . . Amino Acid Analysis of Threonine Dehydrase . Sedimentation Coefficients Obtained in the Absence of AMP Using a Crab Lactic Dehydrogenase Coupled Assay . . . . . . Change in Sedimentation Velocity and Dehydrase Activity Resulting from the Removal of AMP from the Enzyme O O O O O O O O O 0 Native Molecular Weights as Determined by High Speed Sedimentation Equilibrium Ultracentrifugation . . . . . . . . High Speed Equilibrium Analysis of Threonine Dehydrase in 20 mM L-homoserine and 10% (w/v) Sucrose . . . . . . . . . . vii Page 31 40 S4 56 58 91 93 95 130 147 161 169 Table 13. 14. 15. 16. 17. 18. 19. 20. Page Low Speed Equilibrium Analysis of Threonine Dehydrase . . . . . . . . . . . 173 Subunit Molecular Weights of Native Threonine Dehydrase from SDS Gel Electrophoresis . . . . . . . . . 189 Determination of the Subunit Molecular Weights of S-carboxymethylated Threonine Dehydrase by SDS Gel Electrophoresis . . 193 Subunit Molecular Weights and the Relative Amounts of the Two Forms of Threonine Dehydrase as Determined by SDS Gel Electrophoresis in the Presence of Mono- and Dithiols . . . . . . . . 194 SDS Gel Electrophoresis of Electrophoreti- cally Purified Threonine Dehydrase . . . 198 Summary of the Molecular Weight Data for the Protomer of Threonine Dehydrase . . 218 Summary of the Molecular Weight Data for Native Threonine Dehydrase I . . . . . 227 Amino Acid Analysis of Threonine Dehydrase II 0 O O O O O I O O O O O O 254 viii 10. 11. 12. LIST OF FIGURES Concentration Dependence of the £20 w Values for Highly Purified Threonine De y rase . A Model for Threonine Dehydrase Involving all Random Ligand Interactions . . . . Vinograd-type Double Sector Centerpiece . . Superposition of Scanner Tracings Using the Outer Reference Hole as a Point of Reference . . . . . . . . . . . Analytical Polyacrylamide Gel Electrophoresis of Highly Purified Threonine Dehydrase . Analytical Polyacrylamide Gel Electrophoresis Analysis of Crude Preparations of Threonine Dehydrase . . . . . . . . The Effect of AMP on the Gel Pattern in Analytical Polyacrylamide Gel Electrophoresis . . . . . . . . . Analytical Polyacrylamide Gel Analysis of Electrophoretically Purified Threonine Dehydrase . . . . . . . . . . . Centrifugation Analysis Using the Coupled Assay at Two Concentrations of Lactic Dehydrogenase . . . . . . . . . . Active Enzyme Centrifugation Analysis of Threonine Dehydrase Using an Assay for a-KetObutyrate O O O O O O O O 0 Concentration Dependence of the Sedimentation Coefficient in the Presence of AMP at Two Concentrations of L-threonine . . . . Centrifugation in the Absence of AMP and Analysis Using Heart Lactic Dehydro- genase . . . . . . . . . . . . ix Page 24 43 74 76 86 89 96 100 106 110 115 121 Figure 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. Sedimentation Coefficient of Horseshoe Crab Lactate Dehydrogenase with a—Ketobutyrate as the Substate . . . . . . . . . Superpositions of Scanner Tracings Using Horseshoe Crab Lactic Dehydrogenase as the Coupling Enzyme . . . . . . . . Centrifugation in the Absence of AMP and Analysis Using Horseshoe Crab Lactic Dehydrogenase . . . . . . . . . . The Effect of L-Threonine Concentration on the State of Association in the Absence Of AMP O O 0 >0 0 O O O O O O 0 Association of the Active Form of Threonine Dehydrase Upon Addition to AMP . . . . Dissociation of the Active Threonine Dehydrase Oligomer Upon Removal of AMP . Determination of the Native Molecular Weight of Threonine Dehydrase by Polyacrylamide Gel Electrophoresis . . . . . . . . Slope Characteristics of Threonine Dehydrase and Bovine Serum Albumin in Polyacrylamide Gels of Varying Percentages . . . . . High Speed Sedimentation Equilibrium Analysis of Threonine Dehydrase . '. . . . . . High Speed Sedimentation Equilibrium Analysis of Native Threonine Dehydrase . . . . High Speed Equilibrium Analysis of Threonine Dehydrase in the Presence of Increased Concentrations of AMP . . . . . . . High Speed Sedimentation Equilibrium Analysis of Threonine Dehydrase in Sucrose . . . Conventional Sedimentation Equilibrium Analysis of Threonine Dehydrase . . . . Molecular Weight of Threonine Dehydrase as a Function of Concentration Determined by Conventional Sedimentation Equilibrium Analysis . . . . . . . . . . . X Page 123 126 128 134 140 144 150 153 156 158 163 166 171 174 Figure Page 27. Sedimentation Coefficients of Threonine Dehydrase as a Function of Dehydrase Concentration . . . . . . . . . . 177 28. Scans of Polyacrylamide Gels after Electro- phoresis of Standard Enzyme Mixtures in Sodium Dodecyl Sulfate . . . . . . . 180 29. Sodium Dodecyl Sulfate Gel Electrophoresis of Threonine Dehydrase . . . . . . . 182 30. Polyacrylamide Gel Electrophoresis of Native Threonine Dehydrase in Sodium Dodecyl Sulfate . . . . . . . . . 185 31. Polyacrylamide Gel Electrophoresis of Native Threonine Dehydrase in Sodium Dodecyl Sulfate . . . . . . . . . . . . 187 32. Sodium Dodecyl Sulfate Gel Electrophoresis of Threonine Dehydrase Incubated Under Various Reducing Conditions . . . . . 195 33. Polyacrylamide Gel Electrophoresis of Electrophoretically Purified Threonine Dehydrase I in Sodium Dodecyl Sulfate . . 199 34. Analysis of Electrophoretically Purified Threonine Dehydrase I by Urea-Gel Electrophoresis . . . . . ,. . . . 203 35. Determination of the Subunit Molecular Weights of Threonine Dehydrase by Conventional Sedimentation Equilibrium in Guanidine Hydrochloride . . . . . 206 36. Low Speed Sedimentation Equilibrium Analysis of Alkylated Threonine Dehydrase in Guanidine Hydrochloride . . . . . . 208 37. Analysis of Electrophoretically Purified Threonine Dehydrase II under Denaturing Conditions . . . . . . . . . . . 250 xi INTRODUCTION Previous studies of the biodegradative L-threonine dehydrase from E; 9911 have shown that L-threonine is dehydrated through a classical pyridoxal phosphate- catalyzed a,B-elimination of water, with a subsequent spontaneous hydrolytic deamination of a-aminocrotonate to form a-ketobutyrate and ammonia. Enzyme catalyzed reaction: L-threonine > a-aminocrotonate + H20 Spontaneous nonenzymatic reaction: a-aminocrotonate + H20 ——————> a-ketobutyrate + NH Threonine dehydrase exhibits allosteric behavior 3 in that AMP acts as a positive effector or activator. AMP exhibits a profound effect on the Km and physical struc- ture of the enzyme. The activation and accompanying structural changes upon AMP binding can be described as follows: 1. It increases the non-covalent binding of threonine, that is it decreases the Km. 2. It alters the state of Oligomerization of the enzyme. 3. It does not participate in the reaction 4. It has no effect on V . max Analysis of presteady state kinetics at low protein concentrations showed that the activation by AMP is second order in protein concentration indicating that the activation step (or decrease in Km) requires a two- fold change in molecular weight. Threonine dehydrase also exhibits a protein-dependent association in the absence of AMP, but this Oligomerization alone, as indicated by Km determinations at varying protein con- centrations is not sufficient to cause the decrease in Km' in fact the Km increases. The kinetic parameters, Km for threonine and Ka for AMP, are dependent on protein concentration. Because of the oligomeric equilibrium involved these parameters are therefore dependent on the state of Oligomerization. Sucrose gradients and column chromatography experiments in the presence and absence of AMP indicate that AMP causes a change in the E20,w value from 3.2 S to 7.6 S at protein concentrations compa- rable to those in the kinetic assay. Based on the as- sumptions of Martin and Ames (1) and assuming an average partial specific volume for a sperical protein, these £20,w Value correspond to molecular weights of about 40,000 and 155,000 respectively (2). This is equivalent to a quadrupling of the molecular weight. Hence AMP causes an apparent tetramerization of threonine dehydrase. Previously reported kinetic and physical data indicate that threonine dehydrase is a complex oligomerizing system involving the effect of at least three ligands on its state of association. They are AMP, threonine and the protein itself. In this thesis the molecularity of the activation process has been examined in order to solve the apparent dilemma between the previously reported four-fold increase in molecular weight observed using physical data and the previously observed two-fold change in molecular weight indicated by kinetic analysis. There are many problems involved in determining the structure of an enzyme during catalysis. For threonine dehydrase these are further complicated by the existence of a protein-dependent Oligomerization process. However, through the use of the relatively new technique of active enzyme centrifugation many of these problems may be resolved. Since the product of the enzyme reaction is measured, it is possible to determine the physical state of the active species under actual kinetic conditions and at low enzyme concentrations. This eliminates the inherent dangers associated with drawing conclusions from experiments done at higher protein concentrations in the presence of a substrate analog. The studies presented herein show that active threonine dehydrase in the presence of AMP has £20,w values of 5.8 - 7.2 S. This suggests that the e;;;;é is in a dimer to tetramer equilibrium. This is substantiated by analysis of equilibrium centrifugation data at higher protein concentrations in the presence of AMP. In the absence of AMP the E20,w 3.5 - 3.6 S indicating that threonine dehydrase is an value of the active enzyme is active protomer under these conditions. This and other data indicate that the protomer to dimer conversion is the required Oligomerization step for activation. A rapidly associating-dissociating system may exhibit a sedimentation coefficient in sedimentation velocity experiments, that does not correspond to an integral multiple of the protomer molecular weight. Therefore, an accurate and independent determination of the protomer molecular weight is needed. The denatured carboxymethylated enzyme was investigated by SDS gel electrophoresis and conventional equilibrium centri- fugation in guanidine-hydrochloride to determine the subunit molecular weight. The data obtained revealed that the protomer consists of a single polypeptide chain with a molecular weight of 35,900. Threonine dehydrase exhibits the classical acti- vation via a decrease in Km ("K" type) as described by Monod, Wyman and Changeux (3). However its other charac- teristics, i.e. the required change of oligomeric state and the dependence of its kinetic parameters on protein concentration, are not consistent with the classical concerted model of Monod 32 a1 (3), or the induced fit model of Koshland £3 31 (4). The molecularity of the process describes a new model of allosterism which has a required state of change of Oligomerization as its control step. A second protein which contains L-threonine dehy- drase activity has been found in §;.Egll° This enzyme co-purifies with biodegradative threonine dehydrase and exhibits a molecular weight of 70,000 under denaturing conditions and a molecular weight of l35-140,000 in its native state. It is not a readily convertible disulfide oligomer of biodegradative threonine dehydrase. It exhibits a low Km, 4.6 mM, in the presence or absence of AMP, and is neither activated by AMP nor inhibited by isoleucine. The amino acid composition of this second threonine dehydrase appears to exhibit significant differ- ences from the amino acid composition of biodegradative threonine dehydrase. Whether this is actually a new enzyme, a proenzyme of biodegradative threonine dehydrase, an extremely hardy disulfide linked form of biodegradative threonine dehydrase, or two polypeptide chains with ester- like links (5) between them needs further study. LI TERATURE REVIEW Active Enzyme Structure Until recently the determination of the active form of an enzyme required the correlation of kinetic data obtained under one set of conditions, and structural data obtained under a different set of conditions. The assumptions necessary can lead to wrong conclusions (6). Kinetic parameters of an enzyme are often determined at low concentrations, about 1.0 ug/ml protein or less, and in a relatively short time period measured in minutes. Structural data are obtained by several "classical" methods involving ultracentrifugation or column chroma- tography. Equilibrium ultracentrifugation (7) experiments require concentrations at least two-fold higher than kinetic experiments, and they require relatively long time periods,generally 16-30 hours. Sedimentation velocity (8) experiments while having the advantage of requiring a shorter time, 60 minutes or less, are also conducted at concentrations in the mg/ml range. The molecular weight of enzymes can be determined at con- centrations cdmparable to those used in kinetic assays by use of sucrose gradients (1) or column chromatography (9). However, these take a relatively long time when compared to kinetic assays. They also rely on the iso- lation of the different forms of an enzyme by a physical separation with a subsequent separate test for activity. One cannot rule out a change in the state of association in the assay (10). In order to more closely approximate kinetic conditions assay components substituting a non- reacting analog for a substrate have been incorporated in all the above methods of structure determination. These experiments still suffer from inherent disadvantages in data correlation. Through use of the "classical" methods of determining protein structure, the molecular weight and quaternary structure can be obtained for the free enzyme or, in some cases, a non-catalytic enzyme- substrate complex. Active Enzyme Centrifugation Of more fundamental significance is a comparison of the structure and activity under closely correlated conditions, i.e. a direct determination of the molecular weight during catalysis. Cohen (11,12) describes a technique where this is possible. He demonstrated the feasibility of determining the sedimentation and diffusion coefficient during catalysis. A thin lamella of enzyme is layered on a column of liquid containing the components normally used in a kinetic assay. As the zone of enzyme migrates down the cell under centrifugal force it may be obserVed spectrophotometrically due to the change in absorbance produced during formation of a product. A more detailed description is provided in the Methods Section and in a methods review by Kemper and Everse (13). Cohen (14) termed this method "active enzyme centri- fugation" because only the active form of the enzyme is observed. The advantages are that active enzyme centri- fugation determines the physical properties of the enzyme substrate complex during catalysis. This nullifies any differences in hydration, polymerization or other para- meters which might exist between either the free enzyme or a non-reacting enzyme complex and the active enzyme. Thus the direct attainment of the physical structure of the active unit eliminates the disadvantages in making the above assumptions. It also allows one to compare the active enzyme structure with the free enzyme struc— ture determined by one of the more "classical" methods. Dilute solutions of enzyme are required which are equiva- lent to the concentration range used in kinetic assays (0.1-10 Hg/ml). Only a small amount of enzyme is required since pl amounts of a dilute enzyme solution are used; and, if a specific assay is available, a purified prepa- ration is not required. A corollary to this is that inactive forms of the enzyme being investigated do not contribute to the molecular weight determination since it is the product of a reaction which is being followed. Three assay methods for use in active enzyme centrifugation have been described in the literature (6,14,15). They are described in detail by Kemper and Everse (13) and will only be briefly listed here: (1) direct measurement of a product in the reaction which undergoes a change in absorption; (2) measurement of the change in pH produced by a reaction using a dye indicator in a weakly buffered solution; and (3) coupling of a reaction product which does not undergo a change in absorbance with one that does, such as enzymatic coupling to NADH, or chemical coupling of a keto compound with phenylhydrazine. According to Le Chateliers Principle multiple sub- units must dissociate at some dilution. For enzymes which are in an association-dissociation equilibrium the equilibrium constant is often greater than 106 M.1 (16). These enzymes would dissociate in the ug range. There— fore determination of the active structure at low con- centrations would eliminate difficulties in correlation of kinetics with the physical structure especially for those enzymes known to undergo association-dissociation. At the same time, the method can be used to study the influence of effectors or substrates on the degree of polymerization without the presence of sucrose. 10 Calculations Since the publication of his original article Cohen gt 31 (17) have described a rigorous mathematical treatment for determination of the sedimentation and diffusion coefficients from active enzyme centrifugation data. Concurrently, Schumaker and Rosenbloom (18) de- scribed a theoretical treatment establishing the validity of determining g values from observation of migrating 20,w zones. Prior to the publication of more detailed papers by Cohen and Mire (6,14) describing the technique and practical application of active enzyme centrifugation several authors had used this technique (19,20,21). Hoagland and Teller (19) were the first to use a photo- electric scanner system for active enzyme centrifugation as suggested by Cohen and Mire (14). Cohen and Mire (14) describe an approximation technique for calculation of the sedimentation coefficient which they have correlated with their more rigorous mathematical approach. This involves determination of a differential curve by sub- traction of the absorbance values for two successive scans. The movement of this curve is assumed to be due only to new product formed and neglects the small con- tribution of diffusion. This assumption imparts a 3-5% uncertainty in the g value. The logarithm of the peak value is then plotted versus time as in the standard method of calculating sedimentation coefficients from 11 sedimentation velocity experiments (7). Schumaker and Rosenbloom (18) point out that for determination of the g value by zone migration the center of gravity of a peak instead of the peak center should be the value plotted versus time. However for a symmetrical peak these posi- tions were nearly identical. When the peak is composed of several species the weight average g value is obtained (18). Although Cohen and Mire (14) state that calcu- lations based on using only the substrate product boundary can lead to erroneous g values, more recently the midpoint of the sedimenting boundary (Figure 4) has been used to determine the weight average sedimentation coefficient (10,13,15,20,22). This is partially due to the laborious and time consuming manual manipulations required in computing the difference curve. However, other more valid reasons are also presented. Often the boundary heights are too low and the signal to noise ratio is too low to ulitize the difference curve but the midpoint is still attainable (15). Because of this poor signal to noise ratio, the expected smooth decrease in the height of a differential curve is not always observed (10). It appears that the shortcomings of the scanner system do not warrant a more rigorous approach. Wampler (22) compared the g values determined by four different methods of data reduction and concluded that essentially identical results 12 are obtained by three of them including the use of the midpoint of the sedimenting boundary. Taylor g£_al (10) note that if asymmetry is imparted in the difference curve due to use of too much enzyme, an unstable product, or product inhibition, the difference plot will give a falsely elevated s value. The use of the boundary mid- point under these conditions would produce a better esti- mate of the g value. A detailed description for deter- mining the boundary midpoint and curve analysis is provided in a review by Kemper and Everse (13). Theoretical Considerations For valid results from active enzyme centrifugation certain theoretical considerations must be accounted for in practical application. Since the publications of Cohen and Mire (6,14) and the review of Kemper and Everse (13) discussing these considerations, several papers have been published concerning these subjects. It seems desirable to review these here. Boundary formation and density gradient.--The enzyme layered on the assay solution must form a thin lamella evenly distributed across the surface of the column (13,14). This requires that a density differential be present between the lamella of enzyme and the assay column. During centrifugation a positive stabilizing gradient is required to counteract a negative density 13 gradient produced by the sedimenting zone (14). If this counteraction is not present a falsely elevated §_value is produced due to "sinking" of the enzyme. Cohen and Mire (l4) accomplish this by layering the enzyme in low salt (0.01 M) on an assay column of a higher salt con- centration. They found the use of 0.1 M buffer sufficient to accomplish this. Recently several authors have sug- gested that D 0 (15,23) or sucrose (16,21,24) be incorpo- 2 rated into the assay to increase its density thereby providing the positive density gradient and also pre- venting forward boundary spreading due to diffusion. The use of either method to increase the density incorporates an unknown factor in determining the active species unless the kinetics are measured under the same condition. Sucrose, although of low molecular weight and sedimentation coefficient, will form a slight density gradient during ultracentrifugation and for this reason should be avoided (13). In addition polyols, such as sucrose and glycerol, have a demonstrated effect on the association-dissociation equilibrium of some enzymes (16,25,26). Experiments with D20 are subject to unknown ef- fects on the kinetics or active structure. It has been demonstrated that D 0 alters the monomer-dimer equilibrium 2 of cytoplasmic malic dehydrogenase (27). One paper (15) reports the use of D20 to increase the density gradient 14 and also attributes a 20% inhibition of enzyme activity to the D20. Therefore its use should also be carefully considered. Substrate cofactor limitations.--Cohen 2E.El (l7) and Cohen and Mire (14) have published detailed analyses of considerations which must be applied to the concentrations of substrates and cofactors in active enzyme centrifu- gation. These are based on the criteria as stated earlier that for valid determination of the g value every molecule must react on the substrate with the same velocity. They routinely used 20 times Km substrate concentrations to insure these criteria were met. A decreased activity by the trailing edge due to substrate depletion, product inhibition, or a reversible reaction would produce falsely elevated g values. They suggest a routine check for erroneous g values can be accomplished by a plot of enzyme concentration versus sedimentation coefficient. However, this conflicts with enzymes which undergo a protein dependent association. As pointed out earlier, Taylor EE.El.(1O) demonstrated that by using the midpoint method for determining the enzyme zone position an increased sedimentation was not observed for their reversible reaction whereas it was observed when the difference curve was used. Kemper (13), taking careful consideration of the kinetics, obtained valid results at low substrate con- centrations equal to Km or less. Shill g£_al (15) have investigated and quantitated the effect of substrate 15 depletion on the s value both empirically and by computer simulation studies. Experimental overloading produced an almost linear increase in §_value with the log of enzyme concentration above a certain critical enzyme concentra- tion. They mimicked this with computer simulation studies. The percent increase in 5 value with substrate depletion was dependent on the magnitude of the original sedimentation coefficient. At 100% substrate depletion a 3.0 S species exhibited a 42% increase to 4.2 S, while a 5.0 S species exhibited only a 15% increase. For a 3.0 S species a 5% error was determined at 5% substrate depletion whereas a 15% increase in the sedimentation coefficient was not observed until 50% substrate depletion. A 15% increase would not obscure a dimerization. Although they routinely used the midpoint calculation method a comparison with the difference curve calculation method produced negligible differences in the g value during both normal and overloaded centrifugation. Active Species of Enzymes which Exhibit Association-Dissociation Equilibria Equilibrium ultracentrifugation studies of bovine heart lactic dehydrogenase (28) and rabbit muscle glycer- aldehyde-B-phosphate dehydrogenase (19) show that these enzymes undergo a concentration dependent association. A plot of molecular weight versus concentration indicates that both enzymes exhibit a dimer-tetramer equilibrium at 16 concentrations below 0.5 mg/ml. Since this concentration is two to three orders of magnitude higher than the enzyme concentrations used in a kinetic assay, Reithel (29) had proposed that the dimer of lactic dehydrogenase was the active Species; the same reasoning held for glyceraldehyde- 3-phosphate dehydrogenase (19). For bovine heart lactic dehydrogenase the sedimen- tation coefficient of the active species, as described by Cohen (11), was determined as a function of pyruvate in the assay (30). Comparable data were obtained for the activity of lactic dehydrogenase in a spectrOphotometer assay at similar protein concentrations. Plots of the sedimentation coefficient versus pyruvate concentration and activity versus pyruvate concentration show that a maximum value of 7.7 S is obtained at the pyruvate con- centration which produces maximal activity. Thus pyruvate shifts the dimer-tetramer equilibrium toward the tetramer and the dimer is at least less active than the tetramer if not inactive. Glyceraldehyde-3-phosphate dehydrogenase is a tetramer with a molecular weight of about 140,000. It exhibits a dimer-tetramer equilibrium at pH 7.0 and 5° at low concentrations as evidenced by (l) the molecular weight distribution across the cell being a function of concentration only; (2) the highest specific activity peak eluted from a column exhibits association-dissociation 17 in equilibrium centrifugation; and (3) the ability to perturb the equilibrium by chemical effectors. This data suggested that the dimer was the active species. Sedimentation velocity studies indicated that the holoenzyme was a tetramer with a sedimentation coeffi- cient of 7.7910.03 S. Sedimentation coefficients of 8.0710.18 S and 7.9310.25 S were obtained for the active species at 20° and 4°, respectively. The upper enzyme concentration limit for valid data was 80 ug/ml at 20° and 140 ug/ml at 4°. Thus a tetramer was the active species and the tetramer was observed at enzyme con- centrations much lower than those which produced a dimer for the free enzyme. Active Species of Enzymes which Exhibit Allosteric Control There are three enzymes in §;_ggli which catalyze the phosphorylation of aspartic acid as the first step in the synthesis of threonine, methionine and lysine, respectively (31). Each is subject to feedback inhibition by one of the products. Two of these enzymes are also homoserine dehydrogenases. Aspartokinase-homoserine dehydrogenase I is subject to inhibition by threonine (32) and is a single protein, molecular weight 360,000 in phosphate buffer plus threonine (33) composed of four subunits with a molecular weight of 90,000 (22). The enzyme rapidly dissociates to dimers in other buffers 18 such as N-tris (hydroxymethyl)-2-aminoethanesulfonate (TES). The enzyme is stabilized against subunit for- mation by threonine (21). Although both the dehydrogenase activity and the kinase activity are sensitive to threonine control it is possible to desensitize the former activity to allosteric control with a concomitant loss of the latter activity (21). This resulted in a change in oligomeric structure of the free enzyme from a tetramer of 11 S to a dimer of 7 S. Thus it seems possible that the allosteric control might be mediated through a change in structure and that one molecular species would exhibit a preferential activity. Active enzyme centrifugation (14) was conducted to determine the active species for both reactions in the presence and absence of the inhibitor, threonine (22). With the forward reaction, the kinase exhibits a sedimen- tation coefficient of 9.2-10.3 8 in the absence of threonine and 9.9-10.4 S in the presence of threonine. The active species for the dehydrogenase reaction was 7.2—7.6 S in the absence and 9.2-10.4 S in the presence of threonine (22). Based on these data, Wampler (22) concluded that the predominant form in the biosynthetic direction was the fully associated tetramer and that it was unlikely that metabolic control was mediated through the associ- ation-dissociation of the enzyme. 19 Bovine liver glutamate dehydrogenase exhibits both glutamate and alanine dehydrogenase activities and undergoes an association-dissociation reaction which is mediated by its effectors. Dissociation is promoted by inhibitors of the glutamate dehydrogenase activity and activators of the alanine dehydrogenase activity. At the same time activators of the former activity and inhibitors of the latter activity promote association (34,35). Thus it was thought that the polymer carried the glutamate dehydrogenase activity and the dissociated form the alanine dehydrogenase activity. Cohen and Mire (6) showed that the dissociated form of 12-13 S exhibited both activities and that the state of Oligomerization was not effected by the effectors ADP and GTP which will effect the state of polymerization of the free enzyme. Enzymes which Undergo a Substrate Induced Change in the State of Oligomerization Phosphoenolpyruvate carboxytransphosphorylase from propionicbacteria catalyzes the formation of oxa- lacetate and pyrophosphate from phosphoenolpyruvate and phosphate in the presence of C02. In the absence of CO2 pyruvate is formed from the above substrates (36,37). Two molecular forms of the enzyme have been isolated. The crystalline enzyme has a sedimentation coefficient of 15.2 S and the enzyme remaining in the mother liquor has an s value of 7.4 S (38). As evidenced by sucrose 20 gradient experiments the enzyme in the presence of sub- strates necessary to form oxalacetate is a dimer of 10-11 S whether one starts with the 15 S or the 7 S species. In the absence of C02 this change of state of Oligomerization does not occur, or the enzyme catalyzing the formation of pyruvate does not change state. On the basis of these data and data obtained from sucrose gradi- ents in the presence of phosphate and C02 Haberland 25.21 (38) concluded that CO2 was essential but not sufficient for optimal formation of the dimer of 10 S. Sedimentation velocity experiments with the two forms demonstrated that this association was not con- centration dependent for the 7 S and 15 S species over a protein range of 0.24-0.06 mg/ml and 5.2-0.04 mg/ml, respectively. A linear 1n y versus £2 plot representing a molecular weight of 408,000 was obtained for the 15 S species and a biphasic plot representing molecular weights of 94,500 and 227,500 was obtained for the 7 S species in equilibrium sedimentation. Investigation of the active species for the oxalacetate reaction by active enzyme centrifugation indicated that the dissociation of the 15 S form to the dimer was concentration dependent. At 70 ug/ml the difference curve indicated a species of 10.0 S with a rather broad faster sedimenting peak appearing at 19 minutes into the run. The nature of the peak was such that it could not be quantitated but the 21 authors attributed it to an active 15 8 species. This was substantiated by finding only a 16 S active species at 140 ug/ml. At the lower concentration of 35 ug/ml enzyme only the 10 S species was observed. No change in the state of Oligomerization was observed for the 15 S species while catalyzing the pyruvate reaction. At an initial protein concentration of 52 ug/ml the main peak of activity for the 7 8 species exhibited a value of 7.4 S with a broad, ill-defined preceding peak attributed to the 10 S species. The initial con- centration in this experiment was almost three times greater than the sucrose gradient experiment in which 78% of the activity was found in the 10 S peak in the presence of the same substrates. This suggests that there is a substantial sucrose effect on the state of oligomeri- zation. Another enzyme which undergoes a substrate induced change in the state of Oligomerization is yeast hexo- kinase (15). The free enzyme exists predominately in the monomeric state at low concentrations while the active species is a dimer. Sodium chloride which inhibits the reaction favors stabilization of the monomer. Quaternary Structure of Threonine Dehydrase There are two threonine dehydrases found in E; coli (39). One, shown to be inhibited by isoleucine, functions 22 in the synthetic pathway for isoleucine and has been termed the biosynthetic threonine dehydrase. The other appears to function in energy metabolism and has long been known to be activated by AMP and requires a sulfhydryl reagent for activity (40). It is not inhibited by isoleu- cine (41). It is termed biodegradative threonine dehydrase. Phillips and Wood (42) were the first to show that AMP did not participate in the reaction mechanism of threonine dehydrase. They also observed that AMP had a marked effect on the oligomeric structure of the enzyme. In the presence of 3 mM AMP the enzyme had a sedimen- tation coefficient of 7.6 8 while in the presence of 3 mM IMP, which does not activate the enzyme, or in the absence of a nucleotide a value of 4.8 S was observed. They con- cluded that the magnitude of the difference between the two sedimentation coefficients was sufficient to rule out a conformational change and implied a link between acti- vation and Oligomerization. Molecular Species in the Absence of AMP More recent evidence showed that threonine dehy- drase undergoes even further dissociation in the absence of AMP. At assay levels of enzyme of the order of 1.0 ug/ml a £20,w value of 3.2 S was obtained by sucrose gradient EeHErifugation (2,43). The inclusion of all components necessary for a coupled assay (see Methods) 23 with the substitution of DL-allothreonine for threonine did not significantly affect the sedimentation coefficient with an observed value of 2.9 S (2). A molecular weight of 40,000 was observed for the enzyme in the absence of AMP and DTT (2) using calibrated Sephadex G200 columns as outlined by Andrews (9). Similar molecular weights were obtained by column chromatography in the presence of DTT. The difference in the limiting §_value obtained by Whanger 2E.El (2) and Phillips and Wood (42) was explained by the observation that threonine dehydrase in the absence of AMP undergoes a protein dependent association (Figure l). A smooth increase in the sedi- mentation coefficient from 3.2 to 7.0 S was observed with increasing protein concentration by Whanger gt_§1 (2) with impure enzyme and substantiated by Gerlt 25 21 (43) with a homogeneous preparation. The latter authors found that a ten-fold higher level of enzyme was required for the protein dependent association. Their data (pre- sented in Figure 1) shows that a value of 6.4 S was reached with 366 ug dehydrase on the gradient while Whanger 2; 31 (2) observed a value of 7.0 S with 22 ug dehydrase on the gradient. This difference is attributed [by Gerlt 23 31 (43)] to the more rigorous removal of AMP. The protein dependent association was also observed on columns with molecular weights of 83,000 at 1 mg/ml 24 Figure l.--Concentration Dependence of the 5 Values —20,w for Highly Purified Threonine Dehydrase. 25 33 Ema/Eu zo mmxmo 000.9 000. 00_ 0.0_ 0.. ..O 0.0 ON QM 153 oz \ \NA 0 \O‘ O \ 03V 0 / EEON.\ 0.0 \ / .2 E E w \ IIEEmNm \ ”122055375 . \ ON . \oi . \ INIIHbld\ a . \ ll. 1.11.141 G” O m b I I SEQ use. Om .LNBIOIJJBOO NOIIVINBINIOBS 26 (2) and 75,000 (44). At 1 mg/ml a single symmetrical peak with a sedimentation coefficient of 8.0 S was observed in sedimentation velocity experiments (45). This association is specific in its dependence on dehydrase concentration alone. Bovine serum albumin had no effect on the state of association in the absence of AMP (2,44). Molecular Species in the Presence of AMP In the presence of AMP biodegradative threonine dehydrase exists in an associated form. At dehydrase concentrations of 1.0 ug/ml or less as used in the coupled assay, a sedimentation coefficient of 7.2-7.4 S was observed (2,43). As was seen in the absence of AMP addition of assay components, substituting DL-allo- threonine for threonine, to the gradient had no effect on the observed g value. Again a dependence of the sedimentation coefficient on dehydrase concentration was observed, although of much less magnitude (Figure 1) than in the absence of AMP. Molecular weights of 123,000-158,000 were obtained on Sephadex G-200 columns and 122,000-162,000 on Bio-Gel P-60 columns with the higher values observed at higher dehydrase concentrations and greater enzyme purity. A £20,w value of 8.16 S was determined by sedimentation velggizy experiments over a range of l.l-12.3 mg/ml protein (4). This correlates well with the upper limit of 8.0 S observed using sucrose gradients (2,43). 27 A smooth increase in the molecular weight was observed using sucrose gradients with increasing AMP 6 3 M to a maximum 5 M to 10'2 M concentration over a range of 10- to 10- value of 7.6 S (42) and over a range of 10- with no further increase in g value observed above 5 X 10.-3 M (44). Whanger gt gt (2) observed two peaks on sucrose gradients in 0.1 mM AMP, corresponding to sedi- mentation coefficients of 3.2 S and 3.9 S. They observed the dual peaks at three dehydrase concentrations, 6.6, 11.3, and 22.6 ug/ml. At the same AMP concentration but a dehydrase concentration of 0.5 mg/ml, Hirata gt gt (46) observed a value of 6.4 S. Requirements for Association As indicated above threonine dehydrase exhibits two types of association, one protein dependent and one AMP dependent, with a large difference in their respective association constants. However, both types of association depend on the integrity of the enzyme. Biodegradative threonine dehydrase is a pyridoxal phosphate enzyme (42). The presence of the cofactor is necessary for association. The apoenzyme exhibits a value of 2.5-2.6 S both in the presence or absence of AMP with a molecular weight of 40,000 as determined on Sephadex columns (2). The restored holoenzyme is active and has a sedimentation coefficient in the absence of 28 AMP of 3.9 S versus a value of 4.4 S for the control and 7.8 S in the presence of AMP. The integrity of specific sulfhydryl groups is also necessary for either protein dependent or AMP de- pendent association. Enzyme oxidized in the absence of AMP does not undergo a protein dependent association (2) and exhibits no difference in sedimentation coefficient in the presence or absence of AMP (44,47). In contrast enzyme treated with p-hydroxymercurobenzoate in the presence of AMP is not active but does associate (44,47). This suggests the presence of at least two distinct types of sulfhydryl groups, one group involved in the catalysis and one group involved in the association mechanism. Air oxidation can be readily reversed by DTT with the restoration of activity and both protein dependent and AMP dependent association (2,47). Subunit Composition Sedimentation velocity analysis of threonine dehydrase at pH 12.0 produced a single peak with a molecu- lar weight of 40,000 and a sedimentation coefficient of 2.6 S (44). This latter value correlates well with a value of 2.6 S for the apoenzyme as reported above (2). Analysis of oxidized and rereduced enzyme (dis- cussed in the previous section) by sucrose gradients and column chromatography produced multiple peaks of activity. 29 Using columns of two different dimensions molecular weight values of 38,000, 49,500, 68,000, and 80,000 or 45,000, 53,500, 67,000, and 85,000 were observed (3). On the basis of these data Whanger gt gt (2) suggested that the subunit might consist of more than one polypeptide chain. This is not consistent with the single peak found by Tokushige (44). Further discussion of this discrepancy can be found in the Results and Discussion sections., Stoichiometry The limiting molecular weight of the protomer is around 40,000 as determined by column chromatography (2) and sedimentation velocity analysis (44). Assuming a globular protein and an average partial specific volume, this is in good agreement with the lower limit value of 3.2 S obtained on sucrose gradients in the absence of AMP. This is consistent with a minimum protomeric molecular weight of 38,00012,000 calculated from the pyridoxal phosphate content and 3.7 AMP molecules per 147,000 or approximately one AMP per 37,000 (45). Threonine dehydrase has 24 sulfhydryl groups per 147,000 and sixteen of them are titratable with sulfhydryl reagents in the presence of AMP without loss of its association characteristics (44,47). This indicates that there are four pairs of intermolecular disulfide bonds involved in the association-dissociation equilibrium. 30 Thus threonine dehydrase exhibits a protein depend- ent monomer-tetramer equilibrium in the absence of AMP and is a tetramer with a molecular weight of 147,000- 160,000 in the presence of AMP. Allosteric Regulation of Enzyme Activity There have been numerous papers and review articles on allosteric control of enzymes including the more recent reviews of Atkinson (48) and Koshland (49). Frieden authored a comprehensive listing of associating enzymes which may be involved in cellular regulation (50) and presented an extension of present models to include a slow change in the conformation state of an enzyme termed "hysteresis" (51). Recently Dunne and Wood (25) have published a model for allosteric control which involves a ligand-induced Oligomerization and included a review of enzymes which undergo this state of change. Only the salient features of the two classical models of allosteric control (3,4) plus the new model as proposed by Dunne and Wood (25) will be presented here. Table I presents an outline of the salient features of each model. Concerted Transition Model The first model of allosteric control to gain wide recognition was the concerted transition model of Monod, wyman and Changeux (3). This model required that the enzyme be an oligomeric species with identical subunits. 31 coaumNHHmEomHHo wooschlvcmmHA unmoamncmflm oaaonuommm no m>wpmuomooo mmflum> AIII HGEO OH lily n m HmEomHHo coaumNHumEOmH HMflucmsvmm TGOZ oaaonuwmmm no m>wumuomoou unnumcoo HmEomHHo COADMNflHofiomH counmocoo @GOZ m>wumnomoou mam>fluamom ucmumcou Hoeomflao coaum>fiuo¢ mo moo: coaumuucmoaou aflmuoum mo uommmm moauocwm can mcflocflm ccmmflq usmwoz umanomaoz ocmmflq mo unsound :H wumum mmmucmnoc unaccouna mm mm eamagmox flm MM coco: .Amwv mambo: UHHmumoaad mo somfiummEou H mflmfifi 32 It assumes a preexisting conformational equilibrium of at least two states with different catalytic efficiencies. Further discussion of this model will assume only two states. All the subunits and therefore, all the ligand binding sites of an enzyme molecule in one state must have the same conformation, i.e. the symmetry of the molecule must be maintained. The ligand binding sites, Neither effector or substrate, have a different affinity between the two states. Binding of a ligand to one site shifts the preexisting equilibrium and enhances the binding of subsequent molecules compared to the binding of the first one. In the case of positive cooperativity, either heterotropic by an effector molecule or homotropic by a substrate molecule, this results in activation of the enzyme. The concerted isomerization model cannot account for negative cooperativity and therefore, the demonstra— tion that some enzymes exhibit negative cooperativity indicates that it could not be applied as a universal model. Induced Fit Sequential Isomerization Model The other widely accepted model of allosteric control is the induced fit model of Koshland, Nemethy and Filmer (4). The fundamental difference between the induced fit model and the concerted isomerization model is the lack of a requirement for a preexisting equilibrium 33 in the former model, i.e. different conformational forms with different catalytic and binding properties are not required. Ligand binding itself causes the conformational change to a state with a different activity. Also, in contrast to the Monod model (3), the binding of the first ligand places no restriction on the next event, i.e. no effect on other binding sites will be produced and co- operativity will be absent; or, at the other extreme, all the other binding sites will change concertedly. However, in this case the concerted isomerization would be caused by ligand binding and would not be preexistent (3). Any variation between the two stated extremes is possible with allowance for the initial ligand bound subunit to affect some or all of the other subunits and not necessarily to the same extreme. A ready extension of this is the allowance of non-identical subunits since all subunits do not have to affect all others to the same extent, if at all. Negative cooperativity would be allowed, with the first ligand bound subunit inducing the other subunits to forms which possess ligand binding sites with decreased affinity. Ligand Induced Oligomerization Model Neither of the above models require a change in the state of Oligomerization nor do they account for a 34 protein dependence of the kinetic constants. The recog- nition of associating systems which are involved in cellular regulation (50) and of enzymes whose kinetic constants depend on protein concentration (52) led Dunne and Wood (25) to prOpose a ligand induced oligomeri- zation model of allosteric control. The essential feature of this model of allosterism is the requirement for a ligand induced Oligomerization as the control step. Although many enzymes involved in regulation have been shown to undergo ligand dependent association (50) simple correlation of physical changes with changes in activity is not sufficient proof for these changes to be acausitive regulatory mechanism (53) . Changes in the oligomeric state are essential but alone are not sufficient criteria. Analysis of the thermo- dynamic model for oligomerizing control systems by linked function analysis (54) and the application of this model to threonine dehydrase have led Dunne and Wood (25) to establish certain criteria for identifying ligand depend- ent Oligomerization as a mechanism of control. These criteria are as follows: (a) dependence of the associ- ation constants for Oligomerization on ligand con- centration; (b) the dependence of steady state kinetic parameters of Km’ Ka’ or Ki on protein concentration with a Hill g_value greater than one at an intermediate protein concentration where the protein is in a finite 35 association-dissociation equilibrium. Criteria which require pre-steady state analysis of the control mecha- nism are: (c) a protein dependent order for activation of greater than 1.0 (i.e. 2.0 for a dimerization); and (d) dependence of the rate of modification on the concen- tration of the modifier. The above criteria have been written for an Oligomerization activation process; how- ever, they also apply to an inhibition or/and deoligomeri- zation process. Threonine Dehydrases There are many known threonine dehydrases and excellent reviews are provided by Wood (39) and Umbarger (55). Bacterial dehydrases are divided into two basic categories, biosynthetic dehydrases and biodegradative dehydrases. The former dehydrases function in amino acid synthesis and are subject to feedback control by a product of the pathway. In contrast the latter dehydrases function in an energy cycle from threonine and are regu- lated by a nucleotide. Data from two well studied biodegradative threonine dehydrases are presented below illustrating their similarities and differences. Biodegradative Dehydrase of Clostridium tetanomorphum Biodegradative threonine dehydrase of C1. tetanomorphum is similar in many respects to the E; coli 36 dehydrase although there are some important differences. For comprehensive reviews see Wood (39) and Umbarger (55). ADP is the nucleotide activator which functions as a regulator of energy from threonine (56). The native homogeneous enzyme has a molecular weight of 184,000 (57) with a sedimentation coefficient of 7.8 S (58) in the presence of ADP. It consists of four identical protomers with a molecular weight of 45,000 and the protomers may contain two polypeptide chains with molecular weights of approximately 28,000 and 17,000 (57). Vanquickenborne gt gt (59) have obtained evidence for separate and distinct substrate and activator binding sites. At a pH of greater than 8.6 in TRIS-HCl or carbon- ate buffers the native enzyme of 7.8 S dissociates to a species with a sedimentation coefficient of 4.5 S (58). This dissociation is reversed by ADP and threonine. In contrast to the §;.22l£ dehydrase theCflu.tetanomorphum dehydrase does not exhibit a rapid association-dissociation equilibrium. Depending on the age of the preparation, Whitely (60) observed multiple peaks of activity in sucrose gradients roughly corresponding to molecular weights of 160,000, 120,000, 80,000, and 40,000. Employing assay conditions in which the enzyme would not associate, an inverse relationship between the Km for threonine and molecular weight was observed (60). Approximately a 37 three-fold increase in the Km was shown between the species isolated as the fully associated enzyme and the species isolated as the fully dissociated enzyme. Thus all species were active pgt gg with the lower molecular weight forms being less active. The inherent disadvan- tages of this type of correlation were discussed earlier; i.e. isolation of a species by physical methods followed by separate activity assays is not proof that the isolated species is active or remains in its isolated form. It would be interesting to repeat the experiment using active enzyme centrifugation analysis of the isolated forms. As with the E; ggtt enzyme the integrity of specific sulfhydryl groups is necessary for association (60). Conflicting data exists in the literature con- cerning the kinetics of the Q; tetanomorphum dehydrase. Vanquickenborne and Phillips (58) reported that the dehydrase exhibits hyperbolic kinetics with and without ADP in the assay mix. However, Whitely and Tahara (61) and Nakazawa and Hayaishi (62) observed sigmoidal plots for threonine in the absence of ADP and hyperbolic plots in the presence of ADP. Obviously determination of the kinetic parameters will depend on the shape of the kinetic plots. Despite this, similar Km values were observed for threonine by Vanquickenborne and Phillips (58) and Nakazawa and Hayaishi (62). Values of 3.5 mM and 54 mM were obtained by the former authors and values of 3.6 mM and 38 37 mM by the latter authors in the presence and absence of ADP, respectively. Whitely and Tahara (61) found much different values which probably reflect the pH difference in the assays of 9.6 versus 8.4 (62) and 8.0 (58). A more detailed investigation of the kinetics was performed by Vanquickenborne and Phillips (58) which provides an explanation for widely different results. They observed that above pH 8.6 biomodal Lineweaver- Burke plots were produced in the absence of ADP. Incu— bation of the mixture for 30 minutes prior to the rate determination provided a linear reciprocal plot indicating that the enzyme exhibits a slow conformational change or hysteresis (51). In phosphate buffer above pH 8.6 or at any pH in TRIS or carbonate buffer, the g_value was 4.5 S and this dissociation was reversible by ADP or threonine (58). Nakazawa and Hayaishi (62) used TRIS-HCl, pH 8.6 and Whitely and Tahara (61) used TRIS-HCl, pH 9.6, while Vanquickenborne and Phillips (58) used phosphate, pH 8.0 similar to conditions used for the g; ggtt enzyme (43). Thus the sigmoidal plots seen in TRIS buffer may be due to a partial dissociation followed by a subsequent threonine induced association. The role of the association- dissociation in control of theCflu tetanomorphum dehydrase is not as clear as it is for the Et coli dehydrase. 39 As would be predicted from linked function analysis (25,54), threonine has an effect on the Km for ADP as well as ADP effecting the Km for threonine (58). A fifteen- fold decrease in Km is observed in the presence of ADP over the absence of ADP. The Ka for ADP is 160 uM in 0.2 mM threonine and 3.3 uM in 10 mM threonine. The conflicting data of various investigators is probably due to the variations in pH, buffer, threonine concentration and ADP concentration. As with the §;.Egli dehydrase it will be necessary to more closely correlate conditions in order to solve the present conflicts and ambiguities. Biodegradative Dehydrase from Escherichia coli The molecular structure of biodegradative threonine dehydrase from E. ggtt and the effect of AMP on the state of association was presented earlier. This section will deal with the kinetics of dehydrase and the relationship of AMP to regulatory control of the enzyme. Table II from Dunne and Wood (25) summarizes the kinetic parameters for threonine dehydrase in the presence and absence of AMP, and the correlation of the kinetic constants with a change in the state of association. A 16-fold decrease in Km for threonine from 56 mM to 3.5 mM was observed upon addition of AMP. In contrast to earlier results (2,63) no significant change in Vmax was observed 40 ommwnoop II II owmmuoop ommonocw mv UHOMIGH omcmzu m ~.R e.Huo.H ~.o o.H owe m.m mza made m N.m In In m.alm.a New mm mad mange SE m mE\:wE\moHOE: SE 4 I I 3 omm c Hafiz mm a Haas xms> am mcoauaeoe mcflsomuaalq Ammv flaoo_dm mo ommutwsoa ocwcoousulq o>HumpmnmmU0Hm map How muopoemumm owuocflm HH mflmda 41 after correction of the velocity to remove artifacts arising from irreversible inactivation in the absence of AMP as described by Dunne gt gt (52). Concomitant with an increased activity resulting from the large decrease in Km is a change in the state of association from a monomer of 3.2 S to tetramer of 7.2-8.0 S. As indicated earlier mere correlation of a change in the state of association with a change in activity is not sufficient evidence to state that the oligomeri- zation causes the activation. In actuality, the protein dependent oligomer exhibits higher Km's for threonine than does the monomer (Figure 1). Therefore, although Oligomerization appears essential for activation, it is not sufficient in and of itself to cause activation. Gerlt gt gt (43) using presteady state analysis upon addition of AMP to a non-AMP-containing assay demonstrated that the activation process was dependent on protein concentration to the second order. Therefore a two-fold change in the state of association was required for activation; or doubling of the molecular weight in the presence of AMP was necessary for activation. Nichol gt gt (64) developed the theoretical con- siderations to show that sigmoidicity, a characteristic of allosteric enzymes, could actually arise from a change in the state of Oligomerization. This explained the apparent COOperativity exhibited by monomeric lamprey 42 hemoglobin (65) which actually existed as an unoxygenated dimer and an oxygenated monomer (66). Wyman (67), using linked functions, developed the theory for the effect of association-dissociation on oxygen binding, which was portrayed in a more pictorial model by Noble (S4). Twelve binding cycles involving the three ligands, composed of threonine, AMP, and dehydrase itself, were analyzed by linked function analysis (25,68). The re- lationship of these cycles are demonstrated pictorially by the cube of Figure 2. Using the thermodynamic principle that the energy required to go from one state to another is independent of the route taken, one can develop equa- tions predicting the interrelationship of dissociation constants. Using the above principles predictions of the interactions of kinetic constants and the state of associ— ation could be made. For example, given that AMP in- creases the state of association, it can be predicted that the Ka for AMP will decrease with increasing protein concentration. This is substantiated by experimental data which show a decrease in the Ka for AMP from 0.35 mM to 0.10 mM as the dehydrase concentration is increased from 10 to 32 mg/ml. Another relationship which was discussed earlier is that the Km for threonine should increase with increasing protein concentration in the absence of AMP (Figure 1). 43 Figure 2.--A Model for Threonine Dehydrase Involving All Random Ligand Interactions (25). Symbols used are: M = monomer, D = dimer, A = AMP, T = threonine, and K = the association constant. 44 G? I 'li'66“ 5 6" . l ‘3‘“ K ’5’ "All" 1”" ‘1'!!!“"”" I "fl'i‘Hl’lr‘.’ IIIIII ""“"y” 45 The interdependence of the kinetic parameters of Ka and Km and the state of association on the ligand concentration and protein concentration probably results in the various values for these constants reported in the literature. Thus as in the case of the El; tetanomorphum enzyme it is important to determine the various values under closely correlated conditions. The use of linked function analysis aids both in new experi- mental design and interpretation of previous data. Biodegradative threonine dehydrase is an inducible enzyme that is completely repressed by oxygen and/or glucose (40). Under anerobic energy poor conditions enzyme synthesis is derepressed and can consist of up to 1% of the total cellular protein (40). The El; tetanomorphum enzyme has been shown to participate in the generation of energy through threonine utilization to proprionate and ATP and its control is probably linked to the energy level through the nucleotide, ADP (56). A similar function has been suggested for the E; ggtt dehydrase (40). In the fully derepressed state enough dehydrase is present so that the AMP regulatory role is not signifi- cant. However, the activation of the initial low levels of enzyme produced in response to anaerobic low energy level growth probably plays a significant role in the energy cycle (25). 46 Other Oligomeric Systems Dunne and Wood (25) have published a comprehensive list of enzymes which undergo a ligand induced change of state. They have applied their criteria for ligand induced Oligomerization as the mechanism of regulation, and many enzymes appear to fit the model. However, no enzyme meets all their criteria. The rate of product formation by deoxycytidylate deaminase is dependent on the square of protein concen- tration (69). This enzyme thus meets one of the vital criteria for allosteric control through Oligomerization (25). Heart lactic dehydrogenase which generally functions as a lactate oxidase exists as a tetramer whereas pyruvate reduction is carried out by the monomer which is most easily formed by the muscle lactic dehydrogenase (70). Several other enzymes which change state of oligomeri- zation as a function of ligands are discussed under Active Enzyme Centrifugation. METHODS AND MATERIALS Bacteriological L-threonine dehydrase was prepared from an isoleucine-requiring mutant of Escherichia coli (ATCC 9739) designated LA9. The mutant was generously donated by Dr. A. T. Phillips, Pennsylvania State University, University Park, Pennsylvania, who showed that the biosynthetic dehydrase was missing. The organism was maintained at 4° in a stab culture of 2% N-Z amine NAK, 1% yeast extract, 0.05% dibasic potassium phosphate and 2% agar with periodic transfer to insure viability. Less frequently the organism was reisolated by taking indi- vidual colonies from Levine EMB agar plates and checking for the isoleucine requirement by growth in Davis and Mingioli medium (71) with and without isoleucine supple- ment. To reduce the risk of selecting an abnormal mutant in this manner the cultures of several individual colonies were combined. Threonine dehydrase was isolated from £1.2911 grown at 37° for 18-24 hours in a 120 liter fermentor (New Brunswick Scientific Co.) without aeration and with just enough agitation to keep the cells suspended. The medium was the same as used for stab cultures except for 47 48 omission of the agar and is the medium of Wood and Gunsalus (40) as modified by Neiderman gt gt (72). A transfer was made from the stab culture into 15 ml of medium supple- mented with 1.0% glucose and grown at 37° with shaking for 8-12 hours. Successive transfers were made into one liter and ten liter amounts of unsupplemented medium. The culture was incubated at 37° for 8-12 hours with sufficient stirring via a magnetic stirring bar to prevent settling. A ten liter culture was used as the final inoculum. This is a modification of the method for growing large cultures outlined by Rabinowitz (63). . Chemicals Catalase, lysozyme, carbonic anhydrase and yeast alcohol dehydrogenase were purchased from Worthington Biochemical Corporation, Freehold, N.J.; bovine hemoglobin from Nutritional Biochemical Company, Cleveland, Ohio; bovine crystalline albumin from Pentex Biochemicals, Kankakee, Ill.; myoglobin, bovine pancreas chymotrypsinogen A, ovalbumin (Grade V), Bacillus subtilis a-amylase, bovine heart and rabbit muscle lactic dehydrogenase, porcine heart fumarase, Escherichia coli alkaline phosphatase, nitrobluetetrazolium, phenazene methosulfate and Coomassie blue R250 from Sigma Chemical Company, St. Louis, Mo.; Coomassie blue G-250 from K & K Laboratories, Plainview, N.Y.; ultrapure guanidine hydrochloride from 49 Schwarz/Mann, Orangeburg, N.Y.; and Fluorescamine from Roche Diagnostics, Nutley, N.J. Lipoyl dehydrogenase and 2-keto-3-deoxy-6-phosphogluconate (KDPG) aldolase were kindly provided by Dr. J. Wilson and Ms. D. Ersfeld respectively of this department. Hydroxyapatite was purchased as Bio-Gel HT from Biorad Laboratories, Richmond, Ca., and DEAE Sephadex and Sephadex G-25 from PharmaCia Fine Chemicals. Piscataway, N.J. Reagents to prepare polyacrylamide gels were obtained from Canal Industrial Company (Canalco), Rockville, Md. N-Z amine NAK was obtained from Sheffield Chemical, Norwich, N.Y.; yeast extract from General Biochemicals, Chagrin Falls, Ohio; and Levine EMB agar from Difco Laboratories, Detroit, Mi. All other chemicals were purchased from commercial suppliers unless otherwise noted. Reagent or analytical grades were routinely used. Purification of Threonine Dehydrase Crude Extract The purification of threonine dehydrase was a modification of that outlined by Rabinowitz (63). In routine purifications approximately 500 g of §t_ggtt_was suspended in an equal volume on a weight basis (500 ml) of 0.1 M potassium phosphate buffer, pH 8.0, 1 mM AMP and 1 mM DTT. The cells were ruptured by three passes in the Manton-Gaulin Laboratory Homogenizer at 7000 psi. 50 A few mg of DNase were added prior to the first pass to prevent the formation of a viscous solution due to release of high molecular weight nucleic acids. Care was taken to keep the suspension cold during this process and all subsequent steps were carried out in an ice bath or in a cold room (4°). All solutions contained 1 mM AMP and 1 mM DTT unless noted otherwise. Protamine Sulfate Treatment After the cell debris had been removed by centri- fugation, the supernatant was diluted to approximately 12 mg/ml of protein with 0.1 M potassium phosphate, pH 8.0. Powdered ammonium sulfate was added to a final concentration of 0.1 M and one fifth volume of 2% protamine sulfate, previously adjusted to a pH of 5.0 with KOH, was added slowly with stirring. The precipitate was removed by centrifugation at 12,000 x g for 30 minutes. Ammonium Sulfate Precipitation The supernatant from the above was made 2.0 M-in ammonium sulfate by the addition of finely powdered ammonium sulfate. The solution was equilibrated with stirring for 30 minutes prior to centrifuging at 12,000 x g for 30 minutes. The precipitate was taken up in 0.1 M potassium phosphate, pH 8.0. 51 DEAE Sephadex Chromatography--First Column Prior to chromatography the enzyme solution was dialyzed overnight versus three changes of 0.5 M potassium phosphate, pH 7.3, plus 0.2 M KCl. The dialysate was checked for conductivity and if necessary, diluted with distilled water containing 1 mM AMP and 1 mM DTT to match the ionic strength of the column effluent. The DEAE- Sephadex column (5 x 25 cm) was equilibrated with the same buffer as used in the dialysis. The enzyme solution was applied to the column at a flow rate regulated at 90-100 ml/hour. The column was washed with 1000 ml of 0.05 M potassium phosphate, pH 7.3, plus 0.2 M potassium chloride. The enzyme was eluted with a linear gradient of potassium chloride (0.2 M to 0.8 M) in 0.05 M potassium phosphate, pH 7.3 employing 1000 m1 of each. The flow rate was regulated at 150 ml per hour and the column effluent was collected in 20 ml fractions. The fractions with higher specific activity were pooled for further purification while those of lower activity were saved to add to subsequent purifications prior to DEAE chromatography. The enzyme was precipitated by the addition of 3.75 M ammonium sulfate to a final concentration of 2.6 M and centrifuged at 20,000 x g for 30 minutes. The precipitate was taken up in a minimal volume of 0.1 M potassium phosphate, pH 6.8. 52 Hydroxylapatite Chromatography The commercial preparation of hydroxylapatite (Bio-Gel HT) was substituted for the laboratory prepared hydroxylapatite (73) used by Rabinowitz (63). The water used in preparation of buffers for hydroxylapatite chromatography was freshly distilled to remove CO2 and reabsorbtion by use of a CO protected from C0 trap. 2 2 This was necessary to prevent the formation of a carbonate crust on top of the hydroxylapatite column. The dialysis buffer and initial column equilibration buffer was 0.01 M potassium phosphate, pH 6.8, plus 1 mM DTT and 1 mM AMP. The use of this lower concentration and pH of the buffer insured consistent final results with different lots of hydroxylapatite. After overnight dialysis against three changes of the dialysis buffer and after the conductivity was matched to that of the column effluent as described earlier, the enzyme was placed on the hydroxylapatite column (4 x 25 cm). The column was washed with a stepwise gradient of 200 m1 of 0.03 M potassium phosphate, pH 6.8, 100 ml each of potassium phosphate (0.05 M, 0.07 M, and 0.09 M, pH 6.8) and finally eluted with 400 ml of 0.11 M potassium phosphate, pH 6.8. All the buffers contained 1 mM DTT and 1 mM AMP. The enzyme was easily located by its yellow color and all fractions visibly yellow were pooled. The pooled enzyme was diluted with an equal volume of 0.4 M KCl plus 1 mM DTT and 1 mM AMP. 53 DEAE Sephadex Chromatography--Second Column This column was adapted from the purification of threonine dehydrase by Shizuta 2E.§£ (45). A (2.5 x 20 cm) column of DEAE Sephadex equilibrated in 0.05 M potassium phSOphate, pH 6.8, 1 mM DTT, 1 mM AMP, and 0.2 M KCl was used. The pooled and diluted fractions from the hydroxyl- apatite column were applied to the DEAE Sephadex column and eluted with a 1000 ml linear gradient of KCl (0.2 M to 0.8 M) in 0.05 M potassium phosphate, pH 6.8, plus 1 mM DTT and 1 mM AMP. The fractions containing the dehydrase were again easily located by their yellow color and pooled on that basis. The pooled fractions were made 2.6 M in ammonium sulfate by the addition of 3.75 M am- monium sulfate containing 1 mM DTT and 1 mM AMP. The solution was centrifuged at 12,500 g and the precipitate was taken up in a minimum volume of 0.1 M potassium phosphate, pH 8.0, plus 1 mM DTT and 1 mM AMP, and stored at -20°. It has recently come to our attention that 0.1 M potassium phosphate, pH 6.8, plus 1 mM DTT and 1 mM AMP may be a better buffer for enzyme storage. Table EEIshows the results of a typical purification. Determinations and Procedures Protein Protein was routinely determined by a semimicro modification of the method of Lowry gt gt (74) using 54 Honuo co oommm .hHo>Huommmou omm one mma hamumeflxonmmm mum mmmum 03D mmonu um mowufi>wuow oawwoomm may maowumoflmansm .m:0fluomum 03¢ mmmnu co omCHEHmuoo uoc mos cflmuoum .% we owe ms.a 0 HH xmsmsdmm mama am « oa.~ he muaumdmasxouemm we . om.~ ems H xmemsdmm mama mm em.s em.m can mummasm ssacoaad Hm H.~ mm.m oemm mammasm mesaauOMM ooa m.m on.m com pomuuxm dunno mE\cHE\HoE: HE w >uw>wuo¢ OH x mEdHo> cowuomnm camas oamaoodm e mums: Hence .mmmupmnoo ocwcomnna mo GOHDMOHmesm HHH mqm<8 55 crystalline bovine albumin as the standard. More recently the Fluorescamine assay of Bohlen gt gt (75) has been used with the above standard. The latter reagent has the advantage of greater sensitivity and non-interference from DTT in the threonine dehydrase solutions. A com- parison of the data for the two methods for both native and aklylated threonine dehydrase is presented in Table IV. No bias was observed in measuring native versus alkylatedsamples and the average protein concentration as determined by the Fluorescamine assay was 91% of the Lowry protein value. The dry weight of the enzyme was determined after exhaustive dialysis against 0.1 M potassium phosphate buffer, pH 8.0, 1 mM DTT and 1 mM AMP. It was necessary to use buffer to prevent protein loss through precipi- tation during dialysis and transfer. Threonine dehydrase (0.3 ml) or buffer (0.3 ml) in tared metal planchets was placed in a 110° oven for 48 hours. The planchets were cooled to room temperature in a dessicator and weighed. The samples were then returned to the 110° oven for another 24 hours and the cooling and weighing was repeated. The values were averaged and the buffer dry weight was subtracted from that of the threonine dehydrase solution. Protein was measured by the Lowry method on a sample of the dialyzed threonine dehydrase solution. 56 .o.s mm .muwsmmonm Snatch z mo.o tmcflmucoo mmHmEmM ommupmnop ocwcoounp toumamxam one .wmmmw mcflamommuosam map CH ommupmsmp mcwcomunu o>Humc onu mo own on Hoaum touflsvon mm3 o.m ma .mpmsdmosd seamed z mo.o as degassed onus a .BBQ as H tam .mzd as H .o.m mm .opmsmmonm Edwmmmuom z H.o Uwcwmucoo mdadfihm mmmntmnmt ocwcoounu m>Humc one an ma.o Amuseme.~ Awusvme.m emummeHm am.e Amusvmo.e Amusvee.e edemasxaa em.o Amncvmm.m Amucvmn.m tmummeag om.o Amucvmm.m Amncvem.m toumamxad om.o Aencvm.aa Amncvm.ma w>Humz HE\mE HE\mE ofiumm ocfleeomouosHm muzoq ommupmnmo ocwcoonna «.GOADMHucmocoo ommuomsmo unaccouna mo coaumcHEHoumo How moonumz ocflEmommuosHm can huzoq mo conflummEou >H magma. 57 The concentrations of serial dilutions of dialyzed threonine dehydrase solution were determined from the number of fringes calculated for go, the concentration at zero time, in an ultracentrifug; equipped with Rayleigh optics. The average value of 4.1 fringes per mg/ml of protein as determined by Babul and Stellwagon (76) was used. Also, the protein concentration was determined on the stock solution by the method of Lowry and by dry weight. The data from the three methods of determining protein concentration are presented in Table V. Based on these data the Lowry and Fluorescamine procedures overestimated the protein concentration by about 25% and 10% respectively. No attempt was made in this thesis tonormalize the protein concentration to a dry weight basis. The method used to determine each protein concentration will be detailed to provide a basis for comparison. Photometric Determinations A Gilford Multiple Sample Absorbance Recorder attached to a Beckman DU monochromator was used for spectrophotometric work. Silica microcuvettes with a 1.0 cm light path were used for the assays. Fluorescence was measured in a filter fluorometer (MicrofluorOphotometer, American Instrument Co.) equipped with a Corning No. CS- 7-51 primary filter having a peak wavelength of 405 nm 58 .Houoom COHDSHHU ouofiumonmmo onu mo own an pouoadoaoo ouoB maaoo Honuo one How mosao> onu can m HonEsn Haoo ca nOHDDHOM on» now oonflauouoo ouoB mosao> euzoq Ugo unmfloz ennLone .npxou oomv Haoo nooo How Achy H.v en .oEHu ouou no coauouucoonoo onu .om mnflofi>flt an tonHEuouop mos AHE\mEV coaponunoocoo neouonm one &. we.o omouo>o me.o Aennvom.~ Amunvme.a em.a m ne.o vm.H av.H av.a v me.o mm.H mo.a vo.a m me.o mm.o oe.o no.0 N we.o mv.o mm.o vm.o H HE\mE HE\mE HE\mE I HonEdz ofluom euzon unmfloz mun Om HHoU e.omDMHuucoooHuHD on» :w Honasz omnwum can muzoq .unmfloz man an tonflfiuouoo mo oGOfluoHpqoonoo omouomnoo mo nomHHomeoo > mflmflB 59 and a Wratten No. 4 secondary filter passing all wave- lengths above 465 nm. Amino acid analysis was performed on a non- commercial ultrasensitive amino acid analyzer. The sample preparation procedure was as outlined by Robertson gt gt (77). Removal of AMP The concentrated threonine dehydrase (10 to 20 mg/ml) was stored in 1.0 mM AMP. Two methods were used to remove AMP from the enzyme in order to investigate the effects of AMP on the structure of the enzyme. To provide for rigorous removal of AMP or for work at high concentrations of threonine dehydrase a Sephadex G-25 column (0.6 x 4.6 cm) was equilibrated in 0.1 M potassium phosphate buffer, pH 8.0, containing 1 mM DTT. The enzyme was displaced with the same buffer. The second method, for use at low concentrations of threonine dehydrase, was based on dilution of the concentrated enzyme solution in nonAMP-containing buffers. The residual AMP concentrations of 1.0 uM or less were at least 50 fold below the lowest reported Ka for AMP of 0.05 mM.(44). Other Lactic dehydrogenase was desalted using a Sephadex G-25 column (1.0 x 5 cm) which was equilibrated in 0.1 M 60 potassium phosphate buffer, pH 6.8. The concentration of the lactic dehydrogenase was determined at 280 nm 1% 1c The iodoacetate used in the carboxymethylation using an E m of 15 (78). experiments was recrystallized twice from a CC14: benzene::l:2 mixture, dryed and stored dessicated at -20° in a vial protected from light. Data Analysis The data were analyzed by an adaptation of the POLFIT computer program supplied by Applied Computer Time Share Inc. (79). It is a least squares program using the method of orthagonal polynomials. Assays for Threonine Dehydrase Coupled Spectrophotometric Assay For routine determination of threonine dehydrase activity the coupled enzyme assay as described by Dunne gt gt (52) was used. The coupled system is described as follows: a ketobutyrate a-hydroxybutyrate + Lactate + NADH Dehydrogenase NAD+ The reaction volume of 0.2 ml contained 1 umole AMP, pH 8.0; 1 umole DTT; 0.08 umole NADH; 10 ug rabbit muscle 61 lactate dehydrogenase; 15 umoles of potassium phosphate buffer, pH 8.0; and 4 umoles of L-threonine, pH 8.0. The reaction was initiated by the addition of an appro- priate dilution of threonine dehydrase. Direct Spectrophotometric Assay For determination of threonine dehydrase activity at high concentrations of enzyme, the production of a- ketobutyrate was followed directly at 320 nm (52). The reaction contents were as outlined above except for the omission of lactic dehydrogenase and NADH. An extinction coefficient of 20 M—1 was used for a-ketobutyrate. A unit of threonine dehydrase is defined as that amount of enzyme which will dehydrate 1.0 umole of L- threonine per minute at 28° in the assay described above. Polyacrylamide Gel Electrophoresis Equipment A Polyacrylamide gel electrOphoresis (PAGE) was performed using an ISCO Model 490 power supply and a non-commercial electrophoresis chamber capable of running from one to ten gels (10 x 0.5 cm) at a time. The con- stant current mode was routinely used. Analytical Gel Electrophoresis Analytical polyacrylamide gel electrOphoresis was conducted according to a modification of the method 62 of Ornstein (80) and Davis (81). The running gel system consisted of a 1:37 ratio of acrylamide to bisacrylamide supplemented to 1.0 mM in AMP (final concentration) unless otherwise indicated. Running gels of various acrylamide percentages were produced by using the appropriate amount of stock solution and making to volume with H20. No spacer or sample gel was used. Instead the sample was applied in 10% (v/v) glycerol containing bromophenol blue as a tracking dye. For Optimal results the "10X" running buffer of Ornstein (80) and Davis (81) was diluted 1:10 and made 1.0 mM in AMP and 1.0 mM in DTT prior to use unless otherwise indicated. The AMP concentration thus produced was sufficient to change the pH of the running buffer if the AMP was not neutralized prior to use. Therefore the pH was routinely checked and adjusted to 8.3 with KOH if needed. The inclusion of DTT eliminated the possibility of artifacts created by residual persulfate in the gels (82). An initial current of 2.0 mA/tube was applied until the sample had entered the gel. The current was then increased to 5.0 mA/tube for the remainder of the run. This resulted in a final voltage of 400-600 volts. Native Molecular Weights The method of Hedrick and Smith (83) was used to determine the native molecular weight of threonine dehydrase. 63 PAGE was performed in the system described by Hedrick and Smith (83) and in the system of Ornstein and Davis (80, 81) in 6, 7, 8, 10, and 12% acrylamide gels. The electro- phoresis pH was 7.9 and 9.5 with running buffers of TRIS- asparagine (pH 7.3) and TRIS-glycine (pH 8.3), respectively. The ratio of acrylamide to bisacrylamide in all gels was kept constant at 30:1 by using an appropriate amount of stock solution and making up to volume with H20. The electrophoresis was at room temperature at 2 mA/tube until the tracking dye had entered the gel, whereupon the current was increased to 4 mA/tube. The electrophore- sis was stopped when the tracking dye had migrated to within 1.0 cm of the gel end. The gels were removed from the column, the leading edge of the dye front was marked mechanically and the gels stained as indicated in results. Plots of log relative mobility (Rm) versus per cent acrylamide concentration were constructed. The slope characteristic of each standard was determined and replotted versus their known molecular weights generating a standard curve for determining the molecular weight of the threonine dehydrase. All slopes were determined by the least squares computer program mentioned earlier. The protein markers for calibration and their assigned molecular weights were Bacillus subtilis a-amylase (48,500); rabbit muscle lactate dehydrogenase (142,000); fumarase (194,000); Escherichia coli alkaline phosphatase (80,000); 64 bovine serum albumin monomer (68,000); and dimer (136,000) 2-keto-3-deoxy-6-phosphogluconate (KDPG) aldolase (72,000) and lipoyl dehydrogenase (100,000) kindly provided by Ms. D. Ersfeld and Dr. J. Wilson respectively of this department. Polyacrylamide Gel Electrophoresis in Sodium Dodecyl Sulfate Electrophoresis in the presence of sodium dodecyl sulfate (SDS) and molecular weight determinations on the denatured threonine dehydrase were carried out according to a modification of the method of Weber and Osborn (84). The standard buffer used for sample incubation consisted of 0.1 M sodium phosphate, pH 7.0, 1% sodium dodecyl sulfate, and 1% mercaptoethanol unless noted otherwise. The gels and running buffer were as detailed by Weber and Osborn with a constant acrylamide to bisacrylamide ratio of 37:1. The various percent acrylamide gels were made by using an appropriate amount of the stock monomer solution and making up to volume with water. The running or separation gels (10 x 0.5 cm) were either 8% or 10% as indicated. Polymerization was routinely allowed to proceed for 24 hours prior to use as outlined in the Canalco Handbook (85). A small amount of Pyronin B was used as the tracking dye and the samples were made 10% in glycerol (v/v) just prior to electrophoresis. Electrophoresis was allowed to continue at 2 mA/tube 65 until the dye front had entered the gel. The current was then increased to 8 mA/tube until the dye was approxi- mately 1 cm from the gel end. The dye front was mechani- cally marked prior to staining. The Rm for each protein was calculated either versus the dye front or in ratio to an internal standard. When an internal standard was used instead of a tracking dye, the gels were allowed to electrophorese for a set time or a parallel gel with a tracking dye was run. The sample preparation was as detailed under results. The protein markers for gel calibration were myoglobin (17,200); chymotrypsinogen A (25,100); oval- bumin (43,000); fumarase (48,500); catalase (57,000); bovine hemoglobin (15,000); lysozyme (14,300); carbonic anhydrase (29,000); yeast alcohol dehydrogenase (37,000); bovine serum albumin (68,000); and 2-keto-3-deoxy-6- phosphogluconate (KDPG) aldolase (24,000). Polyacrylamide Gel Staining Procedures The protein bands were located in the gels with Coomassie blue or threonine dehydrase was located with an activity stain (see below). Three protein staining procedures were used. Method A.--The gels after removal from the electrophoresis tubes were fixed for 30-60 minutes in a 10% trichloroacetic acid solution. They were then 66 stained overnight in a 0.4% solution of Coomassie blue R250, 10% trichloroacetic acid and 33% methanol. Exhaus- tive destaining in several changes of a solution of 10% trichloroacetic acid and 33% methanol took place over a period of several days. This method of staining is very sensitive, approximately 0.1 pg per band could be detected. However, it was often difficult to remove the background stain completely and minor bands were not visible for several days. Method B.--One hundred m1 of a solution of 0.2% Coomassie blue G250 was thoroughly mixed with 100 ml of 2 N H SO 2 4 solution was filtered and 1/10 volume of 10 N KOH was and the precipitate allowed to settle. The added to the filtrate. The solution was made 12% in trichloroacetic acid by the addition of a 100% solution of trichloroacetic acid. The stain was useable several times before it became too weak to use. The gels were routinely stained for 5 to 8 hours although some bands were visible within 30 minutes. The gels were then rinsed with water and destained in 0.2% H2S04. The sensitivity of this method is about 1.0 ug/band. Method B had a disadvantage in staining SDS gels due to the precipitation 0f the potassium salt of SDS within the gel. This did not irrterfere with the protein stain and was used for rapid detection. For a more aesthetic appearing gel, the SDS COLI—Id be removed by soaking the gels in a 20% sulfosalycylic 67 acid at 37° for 24 hours; alternatively Method C could be used. Method B was a modification of the procedure of Malik and Berrie (86). Method C.--In order to eliminate the precipita- tion of the potassium salt of SDS, 1/20 volume of 10 N NaOH was used in place of 1/10 volume of KOH. This staining solution had to be made up immediately prior to use. The gels were placed in the stain for 5 to 8 hours, but the bands were not visible until after 5 to 6 washings with H20 over a several hour period. Activity Stain A threonine specific activity stain was developed in this laboratory and concurrently published by Feldberg and Datta (87). Twenty mg nitroblue tetrazolium and 20 ug phenazine methosulfate were dissolved in 20 ml of 0.08 M potassium phosphate, pH 8.0, 0.1 M L-threonine, and 0.01 M AMP. The gels were placed in the activity stain for 30 to 60 minutes in the dark. This method is extremely sensitive. If DTT was present in the gel buffer, it was necessary to presoak the gels if buffer to remove some DTT'prior to staining because the DTT reacted with the Stain to produce a dark background. The enzyme bands ‘wereestill visible in the presence of DTT if the reaction was carefully controlled. 68 Urea Gels A modification of the method of Burgess (88) was used for gel electrophoresis in urea. The gels (10 x 0.5 cm) were 7.5% in acrylamide and 8.0 M in urea with a acrylamide to bisacrylamide ratio of 37.5:1. The urea was three times recrystallized from ethanol and treated with Biorad AG 501-X8 (20-50) prior to use. No stacking gel was used and the sample was applied directly atop the running gel in 4 M urea. The electrophoresis was run at room temperature at an initial current of l mA/tube until the marker dye had fully entered the gel. The current was then increased to 3 mA/tube until the run was stopped. Electrophoretic Purification of Threonine Dehydrase Small amounts of electrophoretically purified dehydrase were produced by electrophoresis in 7% acrylamide gels. Approximately 100 ug of protein was layered on top of a running gel (10 x 0.5 cm) in a 10% glycerol, 0.005% bromophenol blue solution. The gels were run as described previously. The major band of threonine dehydrase was visually located by its yellow color and was cut from the gel. The gel was macerated in 0.1 M potassium phosphate pH 8.0, 1 mM AMP and 1 mM DTT and allowed to stand overnight at 4°. The location of the second band of threonine dehydrase activity was accomplished either 69 by the activity stain on a parallel gel or by serial slices (0.25 cm) toward the anode from the main band. These slices were treated as above and the dehydrase activity was located by assay of the eluate for activity. Analytical Ultracentrifugation Analytical ultracentrifugation was performed in a Spinco Model E analytical ultracentrifuge equipped with phase plate schlieren optics and Rayleigh inter- ference Optics or with a photometric scanner system. The temperature was controlled with the regulated temper- ature control system. A stock solution of the enzyme was dialyzed versus the described buffer at 4° for at least 24 hours. Dilu- tions were carried out in a solution of the appropriate dialysate. Viscosities were measured in a semimicro Cannon Ubbelohde viscosimeter and densities were measured in a pycnometer. A partial specific volume of 0.738 cc/g was used as reported by Shizuta gt gt (45). Sedimentation velocity measurements were carried out at 60,000 rpm and at 20° using double sectored cells with sapphire windows in the four place An F-Ti rotor. Enzyme sedimentation was followed in an ultracentrifuge equipped with a multiplexed scanner system which allows one to follow each cell individually in a multicelled rotor. The monochromator was set at 413 nm. 70 Low speed equilibrium experiments were conducted according to the method outlined by Chervenka (7). The experiments were allowed to run for 24 to 36 hours at 4° for native molecular weights and 20° for runs in the presence of guanidine hydrochloride. Rayleigh Optics were used and the interference pattern was recorded on Spectrographic LL-G plates (Eastman Kodak Co., Rochester, N.Y.). The values for go were determined using the interference Optics. The data were reduced and molecular weights were calculated as outlined by Chervenka (7). The meniscus depletion technique of thantis (89) as described by Chervenka (7) was utilized for high speed equilibrium runs. The runs were conducted at 4° in a six place An—G rotor using 12 mm double sectored cells equipped with quartz windows and wedge centerpieces. The data were analyzed with a computer program (90) that had been modified for use on a Control Data Corporation 3600 by Mr. John Gerlt of this laboratory. Active Enzyme Centrifugation The active enzyme centrifugation method Of Cohen and Mire (14) as modified by Kemper and Everse (13) was used to determine the sedimentation coefficients of the active species of threonine dehydrase under various conditions. Several recent articles (10,15) including those cited above (13,14) describe the method, theory 71 and practical considerations in detail. Therefore, only a brief synopsis of the method will be given here. The basic technique involves the layering of a thin lamella of enzyme solution on a liquid column in the ultracentrifuge containing the ingredients of the catalytic assay. Only a small amount of enzyme is required to produce concentrations, usually in the microgram range, as used in the spectrOphotometer assay. The migration of the enzyme is followed via the pro- gressive change in absorbance which reflects formation of the reaction product as the enzyme zone sediments down the cell. Thus, the centrifuge acts as a spectro- photometer. In the case of threonine dehydrase, the product formed can be followed directly by an absorbance change at 310 nm which is produced by a-ketobutyrate. Alternatively absorbance changes at 340 nm produced by the utilization of NADH in a coupled assay system can be used. Since the displacement of active enzyme is followed, any inactive forms will not be observed and therefore will not contribute to the average 320 w value. Pure — I enzyme is not required if a specific assay is available. The theoretical bas1s for determ1n1ng g20,w by band sedimentation is detailed by Schumacher gt gt (18) and by Cohen gt gt (1?). 72 Equipment A Beckman model B analytical ultracentrifuge equipped with a monochromotor, multiplexer and scanner output was used. Vinograd Type I band-forming double sector 12 mm cells Of charcoal filled Epon (Figure 3) equipped with plane sapphire upper and lower windows were used. Either an An-D or an An-F Ti rotor was used. Temperature was controlled with the regulated temper- ature control unit set to 20°. Cell Loading The cell, not including the upper windows, gasket, and screw ring, was partially assembled. With the housing plug holes facing away (top of page Figure 3) ten ul of enzyme solution was carefully placed in the upper right hand well using a Hamilton syringe. Care was taken not to produce a bubble in the bottom of the well nor to wet the capillary groove with the enzyme solution. The rest of the cell was then carefully assembled in order not to disturb the enzyme in the well. A one ml tuberculin syringe equipped with a blunt needle was used to transfer 0.30 ml of enzyme assay solution into the right hand sector of the cell and 0.33 ml of reference solution in the left hand sector. When the product of the reaction caused an increasing absorbance, the reference solution and the assay solution were of 73 Figure 3.--Vinograd-type Double Sector Centerpiece (Beckman-Spinco, part-number 331359). Center of Rotation nzyme dm l/x Lw m/l//////// A Reference 75 the same composition. Due to the circuitry of the scanner when the reaction produced a decreasing absorbance it was necessary to omit the absorbing substance from the refer— ence solution so that the absorbance at all locations in the cell in the reference side was less than that in the sample side. The enzyme is quantitatively transferred by centrifugal force at about 10,000 rpm through the capillary groove to the meniscus of the liquid column containing the assay components. Changes in the absorb- ance as the reaction proceeds while the enzyme traverses the liquid column are then measured by the scanner. Figure 4 shows a typical set of traces obtained at increasing time intervals. Density Gradient Several authors (13,14,15) point out the need for a positive self-generating density gradient in band sedimentation. This is to prevent a forward spreading of the boundary and resultant false £20,w value due to limited conviction. Cohen and Mire (I4I_layer the enzyme solution in low salt (0.01 M) on an assay solution of higher salt concentration (0.10 M). Others suggest the use of sucrose (21,24) or D20 (15,23) to increase the viscosity and produce a sharper more stable boundary by limiting forward spreading. In this investigation of an equilibrium among molecular forms, factors known to 76 Figure 4.-—Superposition of scanner tracings using the outer reference hole as a point of reference. Traces representing an assay with decreasing absorbance. Active enzyme centrifugation was conducted using an AnD rotor at 60,000 rpm and at 20° with the monochromator set at 340 nm. The assay mix contained 0.075 M potassium phosphate, pH 8.0, 5 mM AMP, 5 mM DTT, 0.135 mg/ml lactic dehydrogenase and 0.13 mM NADH. The enzyme was diluted in 0.01 M potassium phosphate, pH 8.0, 5 mM AMP, and 5 mM DTT to a concentration of 0.04 ug/ml. The arrow indicates the direction of sedimentation. 77 its seem mm No mum mm _.m _ _ _ n + II 8:029:63 Co cozooto .. I r :4 _ j n x 8533mm .mm mm .vm .8 m. .m .w .0 _ oocototo .. .330 u Coca. _ I EEVEux L I III 53.84 I0. 0. O O wuotg‘eouoqiosqv 9 78 influence this equilibrium such as sucrose or factors whose effect on the equilibrium were unknown, such as D20, were avoided. Thus the technique of Mire and Cohen (14) was used where dehydrase in low salt (0.01 M) was layered on an assay solution of higher salt concentration (0.075 M). There was also a significant contribution to the density gradient by the substrate. Calculations The most critical point in the data analysis is the determination of the enzyme distribution from the scanner tracings. The technique is essentially identical to the technique of data reduction from scanner traces of a conventional sedimentation experiment. To facili- tate reduction of raw data and aid in interpretation, the scanner traces were superimposed upon a common reference point (the outer reference hole) as more recently described by Shill gt gt (15) (Figure 4). The curve was divided into segments as described by Kemper and Everse (13) and the distance, X, of the midpoint of the boundary (measured to the nearest 0.5 mm) from the outer reference hole was determined. The magnification factor (M.F.) was calculated by dividing the distance between reference holes, T (mm), as measured on the scanner output by 16.1, the actual distance in mm between reference holes in the cell (7). The value of t, the 79 distance from the center of rotation, was calculated from the equation: £ = 73.0 - X/M.F. where 73.0 equals the distance from the center of rotation to the outer reference hole in mm (7). A plot of log E versus time was constructed and the slope was calculated from a least squares fit to the data. A computer program using the method of orthogonal polynomials adapted from the POLFIT program of the Applied Computer Time Share Inc. (79) carried out this operation. In the cases where curved plots were Observed, successive slopes through several points were determined. The following equations were then used to calculate g ' 20,w' 3.50 S = —————— x slope obs (rpm)2 nt n l"""2o,w s = s ——— —— -——————— 20,w obs _— . n20 no 1 th in which n/nO corresponds to the relative viscosity of the substrate to that of water, nt/n20 is the relative viscosity of water at the operating temperature to that at 20° (in this case it is one), 020,w and ot are the densities of water and of the substrate solution at t° respectively, and V is the partial specific volume of 80 the protein, in this case 0.738 (45). The Viscosities and densities were determined as described elsewhere in the Methods Section. The justification for using the midpoint of the curve to determine average gobs has been well established. Wampler (22) compared several methods including the mid- point position and the movement of the peak value of a difference plot, which isequivalent to the first deri- vative of the curve, calculated by subtracting the absorb- ance at identical radial positions on successive curves. He concluded that essentially the same results were obtained. Others (10,13) have used only the midpoint. Shill gt gt (15) pointed out that at low enzyme con- centrations the signal to noise ratio was too low to use the difference curve analysis but valid data were still obtainable using the curve midpoint. Taylor gt gt (10) states that use Of the midpoint is a better estimate of the gobs than the difference plot, although he criti- cizes previOus authors that have used the midpoint method without discussing the possible problems inherent in such use. However, a differential plot can be useful because it provides a concentration distribution profile of the sedimenting band. Assay Two basic types of assay were used in the active enzyme centrifugation method; i.e., a coupled assay for 81 a-ketobutyrate using excess lactic dehydrogenase, and a 310 nm assay where the product of the reaction, a- ketobutyrate, is followed directly. The essentials of these methods as used in the spectrophotometer are detailed elsewhere in the Methods Section. The major difference between the two assays is the level of threonine dehydrase concentration required. The coupled assay uses very low protein concentrations, below one ug/ml, and the 310 nm assay is used above one ug/ml although with proper parameter adjustmentl some enzyme concentration overlap can be achieved. Due to absorbance restrictions in the scanner some of the coupled assays were conducted at 366 nm to enable the use of higher NADH concentrations. Lactic dehydrogenase was desalted prior to use with a Sephadex G-25 column (1 x 15 cm) equilibrated in 0.1 M potassium phosphate, pH 8.0. For assays in the absence of AMP, the AMP was removed either by dilution or by passage through a Sephadex G-25 column (Cf. pp. 60, 61). lParameters which can be adjusted include centri- fugal force, concentration of various components, changing of the wave length to allow for a greater change in NADH without exceeding the absorbance limits of the photometric scanner system, and changing the sensitivity of the recorder output. RESULTS Whanger gt gt (2) repOrted that threonine dehydrase in the presence of its activator, AMP, exhibits a value of 7.2 S at very low protein concentrations (approximately 1 ug/ml) as determined by sucrose gradients. There is a slight protein dependent association to about 8.0 S at higher levels of protein (approximately 1 mg/ml). This latter value is in good agreement with the £20 w of 8.16 I S determined by Shizuta gt gt (45) using the technique of sedimentation velocity in the ultracentrifuge at higher protein (2-20 mg/ml) concentrations. The molecular weights calculated by both investigators, 160,000 (2) and 147,000 (45), are in fairly close agreement. In the absence Of AMP the value of 3.2 S, as determined by sucrose gradi- ents, and the molecular weight of 40,000 as determined by calibrated Sephadex columns (2) are in good agreement with the minimum molecular weight of 38,00012,000 calcu- lated on the basis of pyridoxal phosphate content (45). Thus Whanger gt gt (2) have proposed that the activation Of threonine dehydrase occurs through an AMP-caused Oligomerization from a monomer to a tetramer. By analysis of presteady state kinetics Gerlt gt gt (43) have demon- strated that the activation of threonine dehydrase by 82 83 AMP is second order in protein concentration and there- fore, the rate limiting activation process requires a two-fold increase in the molecular weight. Tokushige gt gt (44) indicate that the activation process involves a shift from dimer, 4.4 S, to a tetramer, 8.0 S. This would be consistent with a second order process except that he has neglected the contribution of the protein dependent association which occurs in the absence of AMP (2,43). Gerlt gt gt (43) worked at lower protein concentrations which in earlier work (2) gave a value of 3.2 S in the absence of AMP and 7.2 S in the presence of AMP as determined by sucrose gradients. Two primary objectives were formulated at the onset Of this investigation: (1) to thoroughly investi- gate the quaternary structure of threonine dehydrase and (2) to solve the apparent dilemma between the four- fold change in molecular weight observed by physical means and the two-fold change in molecular weight indicated through kinetic analysis. During the course of this investigation the new technique of active enzyme centrifugation was published by Cohen and Mire (14). This allowed us to investigate the molecular structure of threonine dehydrase at protein concentrations similar to those used in kinetic assays. Since this method of determining sedimentation coefficients depends on the measurement of the reaction product 84 direct correlation between kinetic data and the molecu- larity Of the active forms could be Obtained. The above studies will not dissect the rapid equilibria which threonine dehydrase has been shown to undergo since the basis of the active enzyme centrifu- gation method, sedimentation velocity, will produce a single peak in most instances for a rapid association- dissociating equilibrium (91,92). Thus the more conven- tional techniques of equilibrium centrifugation, sedi- mentation velocity, and polyacrylamide gel electrophoresis were used to determine the molecular weight of the native enzyme. It is Often a difficult task using analytical ultracentrifugation alone to provide direct evidence for a reversibly association system that the macromolecular component consists of a single component in various degrees of association depending on protein concentration. The use of average molecular weights which represent a heterogeneous population of molecular species can compli- cate the interpretation of the quaternary structure. For these systems it is necessary to determine the molecular weight of the monomer under completely dissociative conditions and to use this value as the basis for deter— mining the quaternary structure. 85 Characterization of Highly Purified Threonine Dehydrase The characterization of a protein requires a highly purified enzyme of known purity. Using the criterion of specific activity, the purity of the biodegradative threonine dehydrase used in these studies correlated well with that of Rabinowitz gt gt (93) and Shizuta gt gt (45).2 The enzyme preparations routinely had an initial specific activity of 416-480 umole L— threonine dehydrated per minute as measured by the coupled assay (see Methods). A typical preparation exhibited two bands on polyacrylamide gels (Figure 5) using several different gel concentrations and two buffer systems. The major band was of lower molecular weight, had a characteristic pyridoxal phosphate spectrum and consisted of an average of 78.5% (n=25) of the total protein with values ranging from 70% to 84.2% as deter- mined by the peak area of stained gels. Where it is necessary to differentiate the two bands, threonine 2The Specific activity reported by Shizuta gt gt (4) for crystalline enzyme was obtained at a temperature Of 37° and was approximately twice that which Rabinowitz EE.El (93) obtained at 28°. Using the appropriate correction, assuming that the activity dependence on temperature has a 010 of 2.0, the previously published values are in relatively good agreement with each other and with that of dehydrase used in this thesis. The validity of this assumption is supported by a value for Q3 of 1.51 determined for threonine dehydrase between the activity at 20° and the activity at 28°. This value for Q3 extrapolates to a value for Q10 of 1.89. 86 Figure 5.--Analytical polyacrylamide gel electrophoresis of highly purified threonine dehydrase. The electrode chamber buffer was TRIS-glycine, pH 8.3, plus 1 mM AMP and lerDTT. The gel concentration was 7%. The gel was scanned at 550 nm after staining with Coomassie blue, Method A. 87 2.3 8:230 8:235. v H $0638 oscootf. e 41 - _ a $2233 32885. —I 0.0 l0. 0 wuoog ‘eouoqlosqv ,0. _ 0N 88 dehydrase I will be used to refer to the major band and threonine dehydrase II will refer to the minor band. Both bands exhibited threonine dehydrase activity using the activity stain. These bands coincided with the only activity found when a parallel gel was sliced into 0.25 mM sections, eluted with 0.1 M potassium phosphate buffer, pH 8.0 containing 1 mM DTT and 1 mM AMP and assayed using the coupled assay. The two bands are not an artifact of the enzyme purification procedure or storage. Freshly harvested E; ggtt and E; ggtt which had been stored frozen for one year were subjected to sonication in the presence of glass beads. The crude extract was centrifuged to remove cell debris and a sample of the supernatant was subjected to gel electrophoresis. Parallel gels were run on each sample. One set of gels was subjected to the activity stain containing threonine and the other set of gels to the activity stain without threonine. Two major peaks labeled I and II appeared in the samples containing threonine which were not observed in the control samples (Figure 6). This indicated that these two bands were threonine specific dehydrase activity. The approximate ratio of the two peaks is indicated in Table VI as a percentage of the total threonine specific dehydrase activity. Also listed in the table are per- centages for the two peaks observed with the purified enzyme shown in Figure 5. These data indicate that the 89 Figure 6.--Analytica1 polyacrylamide gel electrophoresis analysis of crude preparations of threonine dehydrase. Electrophoretic conditions were as in Figure 5 except for the exclusion of DTT from the chamber buffer. The left panel was a crude extract from freshly harvested E; coli and the right panel was a crude extract from cells frozen at -20° for one year. After staining with the activity stain with threonine (---) and without threonine(———) the gels were scanned at 600 nm. 90 EB 8:065 e N o n _ _ sttofi. " Ca” \ A. .. an“ v. 10.0 .- _. _. ._ .. .. ... ,2 :1 . 8285-.. 1:1 9.58.274 LI #8 oEcomEeu— LIV 40.. 8285.“ l ode 595 23:8 :5 E: . -m a .< 2.8 emeritus .m 2.8 St“. a o Luuoog ‘eounqmsqv 91 TABLE VI Relative Amounts of the Two Forms of Threonine Dehydrase as Determined by Polyacrylamide Gel Electrophoresis. Threonine Dehydrase Detection Sample Method Band I Band II % % l Coomassie blue 78.4 21.6 2 Activity 77.0 23.0 3 Activity 84.2 15.8 *Sample 1 was highly purified threonine dehydrase. Sample 2 was crude threonine dehydrase Obtained from freshly harvested gt coli and Sample 3 was crude threonine dehydrase obtained from gt coli stored at -20° for one year. The percentages are based on the areas under the two peaks. The area was calculated from the height times the width at half peak height. 92 two bands of activity were present from the beginning of an enzyme purification, that they copurified and were not formed during purification. This is consistent with the data Of Shizuta gt gt (45) who reported a major band with a trailing minor band on polyacrylamide gels using a crystalline enzyme preparation. AMP Activation of Electrophoretically Purified Threonine Dehydrase To establish the integrity of the electrophoreti- cally purified threonine dehydrase, or threonine dehydrase I, kinetic studies were conducted with the eluate Of the major band prepared as described in the Methods Section. The activation by AMP was established by assay in 50 mM threonine with and without AMP (Table VII). Since the Km for threonine in the absence of AMP is 50-70 mM (52) a large decrease in Km would be expected to approximately double the activity. The activation Observed averaged 2.5 fold. To further characterize the threonine dehydrase activity found in the eluted major band, the Km for threonine and the Hill n value were determined on four individual eluates. The average Km of 2.11 mM and a Hill value of 1.2 agreed favorably with a Km of 2.35 mM and a Hill n of 1.1 determined by a control experiment using the same reagents and substituting highly purified threonine dehydrase for the gel eluate. These values are in good agreement with those previously published (52). 93 TABLE VI I Activation of Electrophoretically Purified Threonine Dehydrase by AMP. A Milliabsorbance/min Amount of Fold Eluate -AMP +AMP Activation U1 1 0.8 2.0 2.5 2 2.0 5.6 2 8 3 4.0 8.8 2 2 4 5.2 12.4 2 4 Average 2.5 *The eluate of threonine dehydrase purified by disc gel electrophoresis was sampled into a coupled assay system containing 50 mM threonine. The assay was allowed to run for five minutes and then 2.5 ul of 0.4 M AMP, pH 8.0 was added to a final concentration of 5 mM. 94 Thus the electrophoretically purified enzyme possessed catalytic and regulatory characteristics similar to those of highly purified dehydrase; hence, the integrity of the enzyme was established for use in further struc- tural studies. Amino Acid Analysis The partial amino acid analysis of threonine dehydrase was conducted at a standard hydrolysis time of 24 hours on native threonine dehydrase from two differ- ent preparations and four different alkylated samples of threonine dehydrase. The average values from the six determinations along with the standard deviations are listed in Table VIII. The number of residues was deter- mined on the basis of 72 arginine residues (45). These values agreed quite closely with the previously published amino acid content of threonine dehydrase by Shizuta gt gt (45) except for the isoleucine value which is 7% lower here. This could be due to the use of a single hydrolysis time and incomplete hydrolysis. Polyacrylamide Gel Electrophoresis AMP Requirement.--Using the standard analytical polyacrylamide gel electrophoresis the gel pattern in Figure 5 was observed with a highly purified prep and the pattern in Figure 7A was Observed with less pure threonine dehydrase. The two unlabeled peaks are 95 TABLE VI I I Amino Acid Analysis of Threonine Dehydrase.* Residues per 72 Shizuta Amino Acid Arginine Residues gt gt (45) Cys 28:1.9 24 Asp 15517.9 156 Thr 74:2.8 72 Ser 9113.6 95 Glu 13113.2 127 Pro 43i3.l 45 Gly 14317.7 140 Ala 12417.3 124 Val 11215.8 113 Met 3711.6 37 Ile 11113.5 135 Leu 10012.9 97 Tyr 3412.5 33 Phe 4013.0 38 Lys 7018.9 79 His 23:3.7 27 Arg 72 72 * The am1no ac1d analysis was carried out as indicated in the Methods Section. All samples were hydrolyzed in 6 N HCl for 24 hours at 110°. The raw data were corrected for recovery based on the inclusion of 10 nmoles of norleucine as an internal standard. The number of residues was normalized to 72 residues for arginine as described in the text. The averages and standard deviations are based on six sets of data except for cyéteine which is the average of only two data sets. 96 Figure 7.--The effect of AMP on the gel pattern in analytical polyacryamide gel electrophoresis. ElectrOphoretic conditions were as in Figure 5 except for the exclusion of AMP from the chamber buffer in panel B. The gels were stained by the Coomassie blue Method A and scanned at 550 nm. Absorbance , 600 nm '0! I ES : :=:r Q 01 .O O ‘O .0 U'l ,O O -8. -AMP '\\ V\ v O I L C 2 4 6 Migration Distance (cm) T 8 98 contaminating proteins which contain no threonine dehydrase activity. The broad band under peak II was due to aggre- gation of native threonine dehydrase. A freshly thawed highly purified preparation exhibited a gel pattern similar to that observed in Figure 5. Upon freeze thawing several times or storage at 4° for several days its specific activity decreased and the broad band appeared. Electrophoretically purified peak I rerun immediately showed only a single peak. However if the electrOphoreti- cally purified band was kept at 4° in 0.1 M potassium phosphate buffer, pH 8.0, containing 1 mM DTT and 1 mM AMP for several days or more, the smeared band as evident here was observed along with a peak corresponding to the initial single peak. The peaks labeled I and II exhibited threonine dehydrase activity as outlined earlier. If the running buffer is not supplemented with AMP, Peak I com- pletely disappeared and a shift was seen in the smeared band to higher aggregation resulting in a more defined peak (Figure 7B). This indicated that in the absence of AMP under electrophoretic conditions aggregation occurred, probably through sulfhydryl linkages. This is consistent with the observed protective effect Of AMP against inactivation by pHMB presumably by reaction with sulfhydryl groups (63). There was no change in the relative size or relative mobility of peak II. 99 Sulfhydryl Requirement.--Electrophoretically purified peak I when rerun on polyacrylamide gels in the presence of AMP but in the absence of DTT produced the pattern shown in Figure 8. The main peak is still evident; however, small amounts of several higher molecular weight aggregates were observed. No attempt was made to quantitate the peaks, but they decreased in size with the increase in molecular weight. These minor peaks were probably artifacts caused by excess persulfate (82). Active Enzyme Centrifugation The present methods of determining the relation- ship of quaternary structure to kinetic data suffer from inherent disadvantages. The time period required for a kinetic assay in a spectrophotometer is measured in minutes and uses very low protein concentrations, Often much less than ug amounts. In contrast equilibrium centrifugation requires 16 to 30 hours and much higher protein concentrations (0.25 to 2.5 mg/ml). Sucrose gradients while offering the advantage of working at low protein concentrations require 16 hours in the centrifuge as well as adding the unknown effect of the sucrose, especially when analyzing an associating-dissociating system. Sedimentation velocity while coming closer to being correlated in time uses much higher protein 100 Figure 8.--Analytica1 polyacrylamide gel analysis of electrophoretically purified threonine dehy- drase. Electrophoretic conditions were as in Figure 5 except for the exclusion of DTT from the chamber buffer. The gel was stained by the Coomassie blue Method B and scanned at 600 nm. For enzyme purification by the electrophoretic procedure see Methods. 101 E13 8:055 8:895. 0 V N _ fl 4 _ fl H omEPEoo oEcooEe noESn. 3.8:.ootocaonoom 0.0 In 0 uluoog ‘ eouoqiosqv 0. in. ON 102 concentrations than kinetic assays by at least 103 fold. The uncertainty of whether the reaction itself effects the molecular structure cannot be totally eliminated by the use of a simulated assay system which incorporates a substrate analog. The advantages of determining the quaternary structure Of the active form under actual reacting conditions in time, protein concentration and in the presence of the assay components is then quite Obvious. The technique of active enzyme centrifugation (14) where one follows the product of the enzyme reaction rather than the enzyme itself allows one to determine the molecular form; (1) of the active enzyme under actual assay conditions; (2) at low concentrations of enzyme equivalent to those used in spectrophotometer assays; and (3) in a similar time frame to spectrophometer assays. Thus one can directly correlate the physical structure and the kinetic data. Active Enzyme Centrifugation in the Presence of AMP Coupled Assay.--The E20 w values of threonine __1_ dehydrase in the presence of AMP were determined using a coupled enzyme assay at 340 nm. The dehydrase con- centrations of the layered solutions were about 0.04 ug/ml and 0.025 ug/ml. These give velocities equivalent to the dehydrase concentration range useable in the 103 coupled rate assay in the spectrOphotometer. The approxi- mate dehydrase concentrations were calculated by applying the appropriate dilution factor and making an adjustment for specific activity. This involved multiplying the protein concentration arrived at by dilution by a fraction representing the ratio of the specific activity of the enzyme used to the specific activity of pure threonine dehydrase. A value of 480 umole of threonine dehydrated per minute was assumed for pure enzyme (93). The average value for the protein concentration in the sedimenting band during active enzyme centrifugation is expected to be approximately three--to five-fold lower than the value for the loading concentration (14) due to diffusion. Dunne gt gt (52)have shown that the critical factor in interpreting the kinetic data for threonine dehydrase is to base the calculations on the amount of "active" enzyme. Although the amount of active enzyme was not determined on the enzyme dilution used in these assays, independent tests performed later indicate that in the presence of AMP the protein concentration calculated by dilution, as done here, is a good approximation of the amount of active enzyme. A major concern in using a coupled assay in active enzyme centrifugation is that sufficient coupling enzyme is present to remove any limitation on the rate of NADH oxidation as determined by the rate of threonine 104 dehydration. Cohen and Mire (14) established two criteria in this respect. (1) If lactic dehydrogenase, used as the coupling enzyme, is non-limiting there will be no increase in value for dehydrase in a series of E20,w tests in which there is an increase in the amount of coupling enzyme at a constant level of threonine dehy- drase. Conversely there will be no decrease in the £20,w value if the coupling enzyme remains constant and the amount of threonine dehydrase is decreased. Shill gt gt (15) point out that the latter statement is appli- cable only for an enzyme whose state of aggregation is not protein dependent. (2) The £20 w value of the coupling enzyme must be less than the §20 w value of the enzyme in I question. By using an excess amount of coupling enzyme the application of the coupled enzyme assay has been successfully extended for use in a situation where the apparent g value of the coupling enzyme was approxi- 20,w mately 1 S higher than that of measured enzyme, in this case, threonine dehydrase. Under this circumstance the coupling enzyme can be expected to sediment away from the dehydrase. The following criteria were used to judge the validity of the £20,w value obtained under these conditions: (1) the log t versus t plot is linear indicating that the same point on the sedimenting band is being followed with time assuming no change in the state of association equilibrium occurs during the run; 105 (2) no large decrease in peak height of a differential curve is observed indicating that the sedimenting band catalyzes dehydration at a constant rate; and (3) a flat baseline is Obtained behind the sedimenting band demon- strating the presence of sufficient coupling enzyme in this region. The 8 value for threonine dehydrase was —20,W determined both at a constant level of dehydrase with two levels of coupling enzyme (tOp three lines in Figure 9) and at a constant level of coupling enzyme with two levels of dehydrase (top and bottom lines Figure 9). Figure 9 shows that the first criterion established above was met and an analysis of the scanner tracings indicate that the second and third criteria were also met. The data show that no decrease in s value —20 ,W was Observed for threonine dehydrase when the amount of lactic dehydrogenase was more than doubled. As a matter Of fact a slight decrease in the 3 value is observed, _20 ,W 7.40 S versus 7.26 S. The values of 7.40 S and 7.39 S Obtained for threonine dehydrase in two experiments using the same assay mix but different dilutions of threonine dehydrase to the same final concentration indicate the repeatibility of the data. The average 520 w value of 7.34 S (range 7.26 S to 7.40 S) obtained for the three experiments at a constant level of dehydrase with various levels of coupling enzyme is in good agreement with the 106 Figure 9.--Centrifugation analysis using the coupled assay at two concentrations of lactic dehydrogenase. Active enzyme centrifugation was conducted in an AnD rotor at 60,000 rpm and 20° with the monochromator set at 340 nm. All assays contained 0.075 M potassium phosphate, pH 8.0, 5 mM DTT, 5 mM AMP, 0.13 mM NADH, 50 mM L-threonine, plus lactic dehydrogenase, 0.135 mg/ml (0,3) and 0.054 mg/ml (0,.A). The enzyme was diluted in 0.01 M potassium phosphate, pH 8.0, 5 mM AMP, 5 mM DTT to the following concentrations of threonine dehydrase, 0.04 ug/ml (0,0,A) and 0.025 ug/ml (ET). All sedimentation coefficients were corrected for density and viscosity. 107 Time (min) 108 £20 w value of 7.4 S determined by sucrose gradient I centrifugation using a protein concentration of 0.02 ug/ml, which is a valid level in spectrOphotometric rate assays (52). The decrease in the sedimentation coeffi- cient from 7.26 S to 6.74 S Observed when the amount of threonine dehydrase in the solution layered on the liquid column was decreased from 0.04 ug/ml to 0.025 ug/ml could ggt be attributed to substrate depletion at the higher dehydrase concentrations. NADH is the limiting substrate and an analysis of the total absorbance change after the sedimenting band of threonine dehydrase indicates at the most two thirds of the NADH was consumed and hence was not limiting the threonine dehydrase rate. This would indicate that the decrease observed in the £20 w value I at the lower protein concentration was probably due to further dissociation as dictated by the protein dependent association-dissociation equilibrium. The pitfalls of using a coupled assay system in active enzyme centrifugation have been treated in some detail here. In later active enzyme centrifugation experiments these factors have been taken into account and, where necessary, will be mentioned. The problem of substrate depletion also applies to active dehydrase centrifugation assays where the product is measured directly. This is due to the fact that a greater amount of dehydration is needed because of the low extinction 109 coefficient of a-ketobutyrate. The proper control experi- ments have been conducted and in subsequent data presen- tation neither substrate depletion nor inadequate coupling enzyme is the cause of differences observed in £20 w values. I Sedimentation Analysis Based on Measurement of a- Ketobutyrate.--Since the determination of the kinetic parameters for threonine dehydrase has involved both a coupled assay for a-ketobutyrate as well as direct measure- ment of this end product, it was of interest to investi— gate sedimentation characteristics of the active species by the same two methods. Because of the lower extinction of a-ketobutyrate direct measurement required higher dehydrase concentrations than used in the coupled assay. Threonine dehydrase is known to undergo a concentration dependent association-dissociation equilibrium (2,43) and therefore, the expected £20 w values obtained in I this assay should be greater than the values obtained with the coupled assay. The sedimentation coefficient of threonine dehydrase was determined at two protein concentrations, 3.2 and 1.6 ug/ml, in duplicate experi— ments. The average values of 6.87 S and 7.07 S obtained at the two levels Of threonine dehydrase, respectively (Figure 10), are lower than the £20 w value Of 7.34 S obtained in the coupled assay (Figure 9) at significantly lower dehydrase concentrations. Again the repeatability Of the method is demonstrated by the nearly identical 110 Figure lO.--Active enzyme centrifugation analysis of threonine dehydrase using an assay for a- ketobutyrate. Centrifugation conditions were as in Figure 9 except that the monochromator was set at 310 nm. All assays contained 0.075 M potassium phosphate, pH 8.0, 5 mM AMP, 5 mM DTT and 50 mM L-threonine. The enzyme was diluted in 0.01 M potassium phos- phate, pH 8.0, 5 mM DTT and 5 mM AMP just prior to use to the following concentrations: 3.2 pg/ml (0,0) and 1.6 ug/ml (11,0). All sedimentation coefficients were corrected for density and viscosity. 111 65:: oEF mm H o. n _ _ _ n 1 m OmmN; . 69 $93on . $9589 0 o . a .J . _ . a 68.30am mom . 4 . . . moweam . Ito. J 50'] 112 values obtained from individual experiments at the same levels of threonine dehydrase. Although the differences in the value between the two dehydrase levels are '520,w repeatable, they are not considered significantly different. A lower E20 w value was observed when a- I ketobutyrate was determined directly than when a coupled assay was used. At the higher protein concentration used in an a-ketobutyrate assay a 5 value greater than —20,W 7.4 S was expected due to the protein dependent association established by earlier investigators and substantiated later in this thesis by active enzyme centrifugation. In this connection Shizuta gt gt (45) have reported that a- ketobutyrate inhibits threonine dehydrase and causes a dissociation of the enzyme in the absence of AMP. This effect is partially counteracted by NADH. However, he did not find any effect of a-ketobutyrate on the state of association in the presence of AMP. Dunne gt gt (52) have further investigated the "activation" of AMP-free threonine dehydrase by NADH and have concluded that NADH provided protection against inactivation but did not activate the enzyme. The contradiction Observed here between the higher gzo'w values observed at lower dehydrase concentrations in the coupled assay and the lower values Observed at higher dehydrase concentrations in the a- ketobutyrate assay suggest that the protective effect 113 Observed with NADH may be caused by a positive effect on the association equilibrium. The value obtained by active enzyme centrifugation in the a-ketobutyrate assay is 0.8-1.0 S lower than that seen on sucrose gradients at a comparable protein concentration. This suggests that sucrose has a positive effect on the association constant as is consistent with data to be presented later. The Effect of Increasing Threonine Dehydrase Concentration on the s Value.--In sucrose gradients 20,w it has been observed that dehydrase undergoes a small concentration dependent association over a 103 fold change in protein from 1 pg/ml to 1 mg/ml as seen by the change in g value from 7.2 S to 8.2 S in the presence Of AMP (2,43). It was of interest to investigate whether this behavior holds true for the active form of the enzyme as determined by active enzyme as determined by active enzyme centrifugation. The preparation used in these experiments was different than that used in the preceding active enzyme centrifugation section. There have been some indications that the tendency toward higher oligomeric states is a variable Of each enzyme preparation, possibly related to the integrity of specific sulfhydryl groups. Therefore a control experiment was run in 50 mM L-threonine at 3.2 ug/ml of dehydrase with condi- tions identical to those in Figure 10 in order to establish 114 a baseline for further experiments. The value of 6.37 S observed in this experiment was significantly different from that (6.87 S, n=2, Figure 10) observed in earlier experiments with a different preparation but at identical protein concentrations. Nevertheless, the upper line in Figure 11 shows the expected increase in sedimentation coefficient, i.e. from 6.37 S to 6.87 S, as the protein concentration was increased from 3.2 ug/ml to 11.1 ug/ml. The amount of active dehydrase was back calculated from activity assays. The maximum concentration decrease resulting from the deamination Of threonine in the assay with the highest protein concentration (11.1 ug/ml) was 10 mM. This left 40 mM L-threonine or about 10 fold the Km value behind the sedimenting band. In order to maintain the L-threonine concentration at least ten times Km at the higher protein concentrations to be used in further investigation of the protein depend- ent association, it was necessary to increase the amount of L-threonine in the assay to 100 mM. Dehydrase con- centrations similar to those used in previous experiments with 50 mM L-threonine were repeated at 100 mM L-threonine along with higher dehydrase concentrations. The E20 w I value increased smoothly from 6.08 S to 7.2 S with over an 18 fold increase in protein concentration. A differential threonine effect is observed here (Figure 11) between the sedimentation coefficient observed 115 Figure ll.--Concentration dependence Of the sedimentation coefficient in the presence of AMP at two concentrations of L-threonine. The protein concentration was calculated from separate activity measurements on the dilution used in the ultracentrifugation run (see Methods). Centrifugation conditions were as in Figure 10 except for the use of the An F-Ti rotor. All assays contained 0.075 M potassium phos- phate, pH 8.0, 5 mM DTT, 5 mM AMP, and L- threonine as indicated. The enzyme was diluted in 0.01 M potassium phosphate, pH 8.0, 5 mM DTT, 5 mM AMP just prior to use. All sedimentation coefficients were corrected for density and viscosity. 116 2533 8222880 $06.38 no Ow ov . ON Km 0 Q LO M 4‘ — — — 0. L0 X mecooEeH. 2 600. Mtx\ I\\I IxI oecoonej .2800 [C3 M033 ‘ IUGIOIIIBOO UOIIDIU9UJIP98 117 in 50 mM and 100 mM L-threonine3 with lower values EM observed in the latter assay at comparable dehydrase levels. Others have pointed out that extrapolation to zero concentration to eliminate the effect of the second virial coefficient is unnecessary for sedimentation coefficients determined by active enzyme centrifugation due to the use of low protein concentrations. This is true for systems which do not undergo a protein-dependent association-dissociation; for these systems extrapolation is necessary to determine the limiting g value. Both 20,w plots of 520,w versus dehydrase concentrat1on and 1/320,w versus concentration were extrapolated to infinite dilution. The two methods of extrapolation produced nearly identical limiting values of 5.83 S and 5.90 S respectively for 50 mM L-Threonine and 5.83 S and 5.88 S respectively for 100 mM L-threonine and indicate that the 5 value for the _20 ,W active species in the presence of AMP is 5.8 S to 5.9 S in either 50 mM or 100 mM L-threonine. Threonine dehydrase in the presence of AMP does undergo a further protein 3This dependence of the sedimentation coefficient on threonine concentration in the presence Of AMP was not always observed. In an earlier experiment using a different enzyme preparation no difference in the E20,w was Obtained in active enzyme centrifugation assays containing 50 mM and 100 mM L-threonine at two dehydrase levels. A lower s20 w was Obtained in the 25 mM L-threonine assays at identIEaI- dehydrase levels. Thus the tendency toward higher oligomeric‘ states appears to vary from enzyme preparation to enzyme preparation. 118 dependent association as was observed in sucrose gradients by earlier workers (2,43). It is of interest to note that a reciprocal plot of E20,w versus concentration extrapolated to essentially identical upper limiting values of 7.2 S and 7.3 S for 50 mM and 100 mM threonine respectively. Active Enzyme Centrifugation in the Absence of AMP Coupled AsSay Using Lactic Dehydrogenase.--Threonine dehydrase in the absence of AMP can be expected to exhibit a lower sedimentation coefficient than the value Of 5.8 S or higher obtained in the presence of AMP. Thus there is a large difference between the g value for the coupling enzyme, heart lactic dehydrogenase, and the g value for dehydrase in the absence of AMP, and the coupling enzyme would be expected to sediment away from the dehydrase. Criteria were established to judge the validity of the £20,w value determined under these circumstances (cf. pp. 104). Since all of these criteria were met it was concluded that valid results were obtained using an assay with heart lactic dehydrogenase as the coupling enzyme in the absence of AMP where there is a decrease in the g value from 5.8 S or higher to about 3.6 S (2, 43). 20,w The protein concentrations given are the amounts in the layering solution based on dilution of a stock solution of known protein concentration in a non AMP-containing 119 buffer. Control assays were done at a different time to measure the amount of activity and from these the con- centration of active enzyme was calculated. The results suggested that the values based on dilution were from two- to four—fold too high based on active enzyme levels. To achieve valid results an excess of lactic dehydrogenase was used. Throughout the time period of the assay there appeared to be sufficient overlap of the trailing edge of the coupling enzyme band with the dehydrase band to maintain an excess of lactic dehydrogenase based on a linear log £ versus t curve and flat baseline. A possible explanation of this tailing of heart lactic dehydrogenase is the finding of Kemper (70) that the heart and muscle lactic dehydrogenase isozymes from chicken appear to exist in the monomeric form when re- > lactate + NAD+ acting in the pyruvate + NADH direction. A subOptimal concentration of the coupling enzyme would produce an abnormally low E20,w value (14). The upper line (Figure 12) value of 5.33—S_;epresents a value between the dimer of 5.8 S and the protomer of 3.6 S (as will be shown later). Since the first data point is obtained 15 to 20 minutes after the ultracentrifuge is started, heart lactic dehydrogenase was layered on the liquid column with the lower threonine dehydrase con- centration to insure that excess coupling enzyme would be present at the meniscus. In this case a £20 w value __1_. 120 of 3.52 S was obtained which is in good agreement with the limiting value of 3.2 S to 3.6 S as determined by prior sucrose gradient experiments (2,43). The protein concentration for the lower line in Figure 12 was five- fold lower than that Of the upper line. Other results at intermediate protein concentrations, not presented here, produced decreasing §20 w values with decreasing I protein concentrations between the two E20 w limits as I illustrated in Figure 12. Additional data to be pre- sented below show that threonine dehydrase undergoes a rapid decrease in 5 value with decreasing protein —20,w concentration as is indicated by this experiment. Coupled Assay_Using Horseshoe Crab Lactic Dehydrogenase.--Although some data were obtained using heart lactic dehydrogenase which has a 520 w value of I 7.2 S, another coupling enzyme was sought with a sedi- mentation coefficient closer to that of threonine dehydrase in the absence of AMP. Horseshoe crab (Limulus) lactic dehydrogenase, which has a published molecular weight of around 70,000 (94), was generously supplied by Dr. N. Kaplan and his associates, Department of Chemistry, University of California at San Diego. Deter- mination of the sedimentation coefficient for crab lactic dehydrogenase by active enzyme centrifugation using a- ketobutyrate as substrate gave a 3 value of 5.87 S -20,w (Figure 13). This is in relatively good agreement with 121 Figure 12.--Centrifugation in the absence of AMP and analysis using heart lactic dehydrogenase. Centrifugation conditions were as in Figure 9. All assays contained 0.075 M potassium phosphate, pH 8.0, 5 mM DTT, and 50 mM L- threonine, plus heart lactic dehydrogenase, 0.135 mg/ml (O) and 0.108 mg/ml (O). The enzyme was diluted in 0.01 M potassium phos- phate, pH 8.0, 5 mM DTT just prior to use to the following concentrations: (C)) 2.0 ug/ml and (O) 0.4 rig/ml. The diluent used for the lower threonine dehydrase concentration also contained heart lactic dehydrogenase, 0.108 mg/ml. All sedimentation coefficients were corrected for density and viscosity. 122 NM um 13-: _s\eamo_ $2250 _E\ 034.0 13-: .53: me agate Sodom Es 8.: 0. m _ I m0®.. I __m._ J DO‘| 123 Figure l3.--Sedimentation coefficient of horseshoe crab lactate dehydrogenase with a-ketobutyrate as the substrate. Centrifugation conditions were as in Figure 9 except that the rotor speed was 59,780 rpm. The assay solution consisted of 0.10 M potassium phosphate, pH 8.0, 3.3 mM a-ketobutyrate, and 0.14 mM NADH. The horseshoe crab lactate dehydrogenase was diluted in 0.01 M potassium phosphate, pH 8.0 to a concentration of about 0.1 mg/ml. 124 3.5 .5: 0m gm 0. N. 0 _ . CoEE Im:.:E..: 365083on 2.63 me. _ Now. 000.. 1 00" 9m. .10.. 125 the molecular weight indicated above. Thus crab lactic dehydrogenase was demonstrated to have a molecular weight about one half that of heart lactic dehydrogenase and it reacted with a-ketobutyrate in active enzyme centrifugation. Figure 14 shows a representative set of tracings and Figure 15 shows the linearity of the log t versus time plot indicating that the criteria outlined earlier (cf. pp. 104) for a valid coupled assay have been met. Crab lactic dehydrogenase was desalted with a Sephadex G-25 column prior to its use as a coupling enzyme. The AMP was removed from the dehydrase in the same manner (63). Further dilutions were made in 0.01 M potassium phosphate, pH 8.0, plus 5 mM DTT. The sedimentation coefficient of the active enzyme in the absence of AMP was determined at three concen- trations of dehydrase using varying amounts Of crab lactic dehydrogenase as the coupling enzyme. Data from three loading concentrations of threonine dehydrase are shown in Figure 15 indicating the protein concentration depend- ent change in £20 w value from 4.24 S to 3.64 S with at I least a two-fold decrease in dehydrase concentration. Table IX shows the correlation of the £20 w value for I each dehydrase concentration with the amount of coupling enzyme. For the lowest dehydrase concentration of 1.02 ug/ml no change in the £20 w was observed over a five-fold I change in the concentration of the coupling enzyme in the 126 Figure 14.--Superposition of scanner tracings using horse- shoe crab lactic dehydrogenase as the coupling enzyme. Centrifugation conditions were as in Figure 9 except for the use of an An F-Ti rotor, and the monochromator was set at 366 nm. The reduced data are represented by the middle line in Figure 15. The direction of travel of the sedimenting band is indicated by the arrow. 127 .5. SEE (monotone. toSO mN F .N m0 No no no 1s _ q — _ Jx]. cozoEoEEom Co 8.88.0 K mm .VN m. m o . oocototom .mcc.\\ Q wu 999 ‘ eounqiosqv 128 Figure 15.--Centrifugation in the absence of AMP and analysis using horseshoe crab lactic dehydro— genase. Centrifugation conditions were as in Figure 9. All assays contained 0.075 M potassium phosphate, pH 8.0, 5 mM DTT, 0.3 mM NADH, 100 mM L-threonine, plus horseshoe crab lactic dehydrogenase, 92 ug/ml (O) and 123 ug/ml (C),Z§). AMP was removed from the enzyme by a Sephadex G-25 column and dilu- tions were made to the following concentra- tions: 2.04 ug/ml (O); 1.36 ug/ml (O); and 1.02 ug/ml ([3). The diluent was 0.01 M potassium phosphate, pH 8.0, 5 mM DTT (0) plus horseshoe crab lactic dehydrogenase, 61 ug/ml (C),Z§). The protein was deter- mined on the column eluent by the method Of Lowry (74). The sedimentation coefficients were corrected for density and viscosity. 129 .55. 2.: 1.... .._ . m _ we. Bantam . H . . -mom. . . (Inseam 4 .l 8..."...on #8. .m I _.0._ 10.0.. 130 TABLE IX Sedimentation Coefficients Obtained in the Absence of AMP Using a Crab Lactic Dehydrogenase Coupled Assay. Threonine C-LDH C-LDH Dehydrase* Assay Layering Solution- §ggtg 119/m1 ug/ml ug/ml 2.04 92 -- 4.02 2.04 92 -- 4.24 1.36 205 61 3.93 1.36 123 61 3.82 1.36 41 61 3.75 1.02 92 -- 3.48 1.02 205 61 3.66 1.02 123 61 3.64 1.02 41 61 3.63 * The threonine dehydrase concentration is based on dilution of a more concentrated eluate from Sephadex G-25 using 0.01 M potassium phosphate, pH 8.0, 5 mM DTT as the diluent. Horseshoe crab lactic dehydrogenase was incorporated into the assays and the layering solutions as indicated. The concentration of a stock solution of crab lactic dehydrogenase was determined as indicated in Methods. The values in the table were determined by applying the appropriate dilution factor. 131 assay mix. The lack of dependence of £20,w values on dehydrogenase concentration holds for all data presented for active enzyme centrifugation experiments. However, the comprehensive data are presented here for the value of 3.5 S to 3.6 S as the lower limiting value for the active protomer. Also, a value of 3.48 S was Obtained in an active enzyme centrifugation assay in which the layering solution was not supplemented with crab lactic dehydrogenase indicating the presence of sufficient coupling enzyme at the meniscus with or without crab dehydrogenase in the layering solution. These §20,w values were in good agreement with the value of 3:52 S Obtained at a lower loading concentration with the coupled assay using heart lactic dehydrogenase (Figure 12). It also agrees fairly well with an interpolated value of 3.6 S derived from sucrose density data of Whanger gt gt (2). The AMP-less threonine dehydrase used in this series of experiments with crab lactic dehydrogenase as the coupling enzyme was the same Sephadex G-25 eluent. Since the protein concentrations were not corrected to Obtain the amount of active enzyme and since threonine dehydrase when diluted in the absence of AMP undergoes a rapid and irreversible 30-50% loss in activity followed by a slow decline in activity (52) only a relative correlation between the amounts of active enzyme in each run could be obtained. However, the use 132 of the four cell An F-Ti rotor allowed three active enzyme centrifugation assays to be conducted at the same time using the same dilution of the Sephadex G-25 eluent. Thus for each loading concentration using different amounts of coupling enzyme in the assay, it was assumed that the amount of active dehydrase in each cell within a run was identical. As detailed earlier, control experiments run at a later time indicated that the use of the loading protein concentration in the absence of AMP over estimates the amount of active enzyme by two- to four-fold. The Effect of L-threonine concentration on the State of Association.--Dunne _e_t gt (52) have postulated, on the basis of kinetic experiments and changes in the Hill n value, that L-threonine shifts the state of association of threonine dehydrase in the absence of AMP toward the monomer. In other words the monomer has a lower Km for L-threonine than does the oligomer present at increased protein concentrations. This is consistent with the increase in the Km for L-threonine observed with increasing protein concentrations in the absence of AMP (25). £20,w values for threonine dehydrase were determined in 100 mM and 400 mM L-threonine at several protein concentrations. The AMP was removed from the enzyme using a Sephadex G—25 column and the appropriate 133 dilutions were made in 0.01 M potassium phosphate buffer, pH 8.0 plus 5 mM DTT. Proper correlation of these data with the kinetic data of Dunne gt gt (52) depends upon determining the amount of active threonine dehydrase present in each centrifuge cell. Since, as pointed out earlier, the enzyme undergoes a rapid irreversible loss of 30-50% immediately upon removal of the AMP and then subsequently a slower loss of activity, false results would be obtained if the protein concentration based on dilution was used for results obtained in the absence of AMP. Therefore, the activity Of each dilution was deter- mined in the spectrophotometer simultaneous with the active enzyme centrifugation runs. The standard coupled assay using lactic dehydrogenase in the presence of AMP was used in the spectrophotometer. Since AMP is known to preserve activity even at high dilutions, gg further tggg of activity would occur in the assays. The amount of active enzyme present in the layering solution was then determined by back calculation using the value of 480 umole/min/mg (as outlined earlier) for the specific activity of pure threonine dehydrase. Figure 16 demonstrates the change in the E20,w value as a function of threonine dehydrase concentration at the two levels of L-threonine. At similar protein concentrations a lower 320 w value is obtained in 400 mM I L-threonine than in 100 mM L-threonine. The physical data tend to support the earlier postulation that 134 Figure 16.--The effect of L-threonine concentration on the state of association in the absence of AMP. Centrifugation conditions were as in Figure 9 except for the use of the An-F Ti rotor and for the monochromator setting of 310 nm. All assays contained 0.075 M potassium phosphate, pH 8.0, 5 mM DTT, and L-threonine, 100 mM (0,0) or 400 mM (0). The protein concentrations are as shown and were deter- mined by activity assays as indicated in the text. The lines represent the theoretical curves calculated from the dehydrase oligomer dissociation constants of 4.50 X 10"6 M and 2.34 x 10-6 M for 400 mM and 100 mM L- threonine respectively. See text for calcu- 1ations. All sedimentation coefficients were corrected for density and viscosity. 135 2,503 omotgcoo 00¢ 00m _ 8.88.... OON oo. o _ j 1 8.825.-.. .2E oou/ 0 ‘II Q q- I 0 IO I/oc.8ot........ .2E 00. M‘OZS ‘ iuagogyaog UOIIOIUGWIPGS 136 L-threonine has an effect on the state of association and that higher threonine concentrations favor the lower molecular weight form. The data were extrapolated to infinite dilution by the computer program, POLFIT (79), as described under Methods. A value of 3.98 S O i 20,w in 100 mM L-threonine and 3.6 S in 400 mM threonine was determined. These values are in reasonably good agree- ment and correlate well with the limiting value of 3.6 S observed earlier using the crab lactic dehydrogenase coupled assay system at lower threonine dehydrase con- centrations. The observed sedimentation coefficients fell between the values determined for the dimer4 in the presence of AMP of 5.8 to 5.9 S and the protomer of 4A limiting value of 5.8 S to 5.9 S was deter- mined earlier for the active form of threonine dehydrase in the presence of AMP. As determined here and in the previous experiment a value of about 3.6 S can be assigned as the limiting sedimentation coefficient in the absence of AMP. This s value for dehydrase in the absence of AMP is in relatively good agreement with that of 3.2 S as determined by Whanger 3; El (2) and assuming a globular protein and average partial specific volume the molecular weight of 38,000i2,000 assigned to the monomer by Shizuta gt a; (45).~ As will be developed later this protomer is made up of a single polypeptide chain with a molecular weight of about 35,900. Thus the value of 3.6 S is assigned to the active monomeric species. According to Martin and Ames (25) the relationship between the molecular weight and the sedimentation coefficient of two species is _ 3/2 MWl/MW2 — (SI/82) Thus for a two-fold change in molecular weight 82:1.59 81, giving a calculated value of 5.72 S for the dimeric species of a 3.6 S monomer. This value is in good agree- ment with the observed average value of 5.85 S which has been assigned to the dimer. 137 3.5 to 3.6 S in the absence of AMP as determined earlier. Using a limiting value of 5.85 S for the dimer and 3.6 S for the protomer the proportion of each species can be determined by the method of Kirschner and Tanford (95). The degree of dissociation is determined by where SD refers to the value for the dimer, 5.85 S, Sp the value for the protomer, 3.6 S, and So the experi- bs mentally determined value. The degree of dissociation was calculated for each experimental value, averaged, and the dissociation constant, was calculated K2,1 according to Anderson (70) using the following formula: 2a 2C K =—————— 2'1 (l-a) MW where K is the dissociation constant for the dimer to 2,1 the protomer, as defined earlier, C the concentration in mg/ml and MW the molecular weight of the monomer taken 6M and 2.34 x 10'5M were as 36,000. Values of 4.50 x 10- calculated for dehydrase in 400 mM and 100 mM L-threonine respectively. That these values are in reasonably good agreement with the experimentally determined data can be seen in Figure 16 where the lines represent the theoretical curves as calculated from the dissociation constants. 138 Association of the Active Form of Threonine Dehydrase Upon the Addition of AMP One of the requirements for active enzyme centri- fugation is that there is no significant sedimentation of the assay components. Mire and Cohen (14) have calculated the g value of NADH (MW=663) to be 0.2 S and surmised that this is sufficiently low to be neglected. AMP with molecular weight of 365 would be expected to have an even lower i value and for all practical purposes would then be uniformally distributed throughout the assay solution column. This fact provided a unique opportunity to observe directly the physical effect of AMP on threonine dehydrase. For this experiment we needed the threonine dehy- drase in its monomeric form. Dunne §E_al (52) indicated that dilution in the absence of AMP did not always insure that the dehydrase was a monomer. Earlier active centri- fugation analysis in the absence of AMP tended to sub- stantiate this. By analysis of the kinetic data Dunne gt a; (52) postulated that in the absence of AMP the substrate L-threonine bound preferentially to the monomer, i.e. threonine will promote dissociation to the monomer. The effect of different L-threonine concentrations on the state of aggregation in the absence of AMP was documented earlier and showed that lower g values were obtained in the presence of 400 mM L-threonine than in 100 mM L- threonine at comparable active dehydrase concentrations. 139 Thus 100 mM D-threonine5 was included in the diluent to insure that the dehydrase would be in its monomeric form. Threonine dehydrase was diluted 1:2500 in the absence of AMP. This is sufficient to lower the AMP concentration to 0.4 uM or over 100 fold lower than the lowest reported Ka for AMP (44). The enzyme was layered onto an assay solution containing AMP and the sedimen- tation of the band was observed. By following the for- mation of d-ketobutyrate directly any ambiguities which might arise in the interpretation of the data from a coupled assay were avoided. Care was taken to bring the rotor up to 60,000 rpm as fast as possible and the time between layering of the enzyme, i.e. as it first en- countered AMP and the time the rotor reached operating speed, was noted. The first scan was taken at eight minutes after layering. As can be seen in Figure 17 the enzyme initially exhibited an s20 w value of approxi- I mately 3.6 S which is that of the monomeric form. Twelve minutes after the first scan or twenty minutes after the dehydrase first contacted AMP the log r versus 5 curve was linear and it exhibited an £20,w value of 6.86 S corresponding to a molecular weight value between dimer and tetramer. The enzyme will not be totally saturated with AMP immediately due to some dilution of AMP at the 5Rabinowitz EE,§l (93) have shown that D-threonine binds to the active site with a Ki of 10 mM. 140 Figure l7.--Association of the active form of threonine dehydrase upon addition to AMP. AMP-free dehydrase was layered on an assay column containing AMP. Centrifugation conditions were as in Figure 16. The assay solution consisted of 0.075 M potassium phosphate, pH 8.0, 5 mM DTT, 5 mM AMP, and 100 mM L- threonine. The enzyme was diluted to a con- centration of 8 ug/ml in 0.01 M potassium phosphate, pH 8.0, 5 mM DTT and 100 mM D- threonine. The sedimentation coefficient was corrected for viscosity and density. 141 ES 2.: ¢N m. _ 142 meniscus by the layering solution. A control experiment using identical dilution conditions but with the enzyme activity measured in the standard assay indicated that a maximum velocity and a linear rate is not reached for about twenty minutes. Thus, the time period for change in structure and in activity are directly correlated. Prior investigators combining kinetic data with sucrose gradient data in the presence and absence of AMP have suggested that AMP causes an Oligomerization of threonine dehydrase (2,42) from a monomer of 3.2 S to a tetramer of 8.0 S. As pointed out before, Gerlt gt a; (43) have demonstrated that the activation is second order in protein concentration or that a two-fold in— crease in molecular weight is required for activation. By active enzyme centrifugation we have been able to follow the physical change in the structure of threonine dehydrase from active monomer to an active oligomer upon the addition of AMP. Since we have demonstrated that the monomer of 3.6 S is active and that upon addition of AMP the monomer species undergoes at least a dimeri- zation, the required activation step is the conversion of monomer to dimer. A higher order Oligomerization as the activation step would result in a greater than second order dependence on the protein concentration. 143 Dissociation of the Active Dehydrase Oligomer Upon Removal of AMP The requirement of non-sedimenting assay components pointed out in the previous section provided another unique opportunity to investigate the effect of AMP on threonine dehydrase. The components of the solution in which the enzyme was layered will sediment insignificantly. Therefore, the enzyme may be layered in a solution of one composition and then the composition of the environment around the enzyme can be changed by sedimenting the enzyme out from the original solution to another. Some diffusion takes place and therefore, a sharp transition is not observed. However, during the course of the centrifugation the environment to which the enzyme is exposed may be totally changed. Using this fact we were able to follow the change in physical structure upon removal of AMP from the enzyme. The appropriate dilutions of threonine dehydrase were made in 0.01 M potassium phosphate, pH 8.0, 5 mM DTT and 5 mM AMP. The enzyme was layered onto an assay solution containing 0.075 M potassium phosphate, pH 8.0, 5 mM DTT and 100 mM L-threonine but without AMP. A representative result is plotted in Figure 18. Along the left hand axis is plotted the values for log 5 for the curve represented by the open circles. The initial s20 w value is 6.24 S; this decreased to a lower slope I 144 Figure l8.--Dissociation of the active threonine dehy- drase oligomer upon removal of AMP. The dehydrase in an AMP containing solution was layered on an AMP-free assay column. Certi- fugation conditions were as in Figure 16. The assay solution consisted of 0.075 M potassium phosphate, pH 8.0, 5 mM DTT and 100 mM L-threonine. The enzyme was diluted to 12.9 ug/ml in 0.01 M potassium phosphate, pH 8.0, 5 mM DTT, and 5 mM AMP. The open circles (CJ) represent the sedimentation data and refer to the left hand axis. The relative activity is represented by the closed circles (O) and refers to the right hand axis. The sedimentation coefficients were corrected for density and viscosity. 145 3,5 2:: . mm VN Q m 0 NOT 0. _ _ . _ _ _ _ vow; - vmo "3.on . _V v.01 . o . N5. w o W I o \ w. . U o C . moi. Ennaomm \\ \ 0mm. . \\ 26 23 a QC: \ m . _ .60 - wNm._ J 601 146 representing a final s20 w value of 3.87 S. The "active enzyme" concentration was back calculated from control activity experiments as outlined earlier. The relative peak enzyme activity between any two scans may be deter- mined by measuring the vertical distance between any two scans at the midpoint of the curve (13). This is plotted in arbitrary activity units along the right hand axis with the data represented by the closed circles. There is a good correlation between the observed decrease in value and the decrease in activity. Thus, production §-20,w l of lower molecular weight species as expected in transition into an environment devoid of AMP is accompanied by greatly decreased activity as expected from a greatly increased Km for L-threonine. Similar data were obtained at other loading concentrations. The data is summarized in Table X. The s values obtained both initially in —20,W the presence of AMP and at the end of the absence of AMP are in reasonably good agreement with earlier data at similar protein concentrations in 100 mM L-threonine. Upon removal of AMP from the enzyme, not only was a decrease in value observed, but also a decrease §2o,w in activity. Assuming an irreversible 30-50% loss in activity (52) and a Km in the absence of AMP of about 100 mM, the activity ratio of about 4:1 for the initial activity to the final activity is reasonable. 147 TABLE X Change in Sedimentation Velocity and Dehydrase Activity Resulting from the Removal of AMP from the Enzyme. Initial Final Initial Activity Threonine . . . Ezperiment Dehydrase E20,w £20,w Final ACthltY ug/ml 1 12.9 6.24 3.87 4.5 2 8.2 6.13 3.61 4.0 3 6.6 6.07 3.46 3.3 *The centrifugation conditions were as in Figure 18 except for the dehydrase concentrations as shown in the table. The activity ratio was calculated by dividing the initial Acm/4min at 4 to 8 minutes by the final Acm/ 4min at 32 to 36 minutes. 148 Determination of the Molecular Weight of Native Threonine Dehydrase The sedimentation velocity of the active forms of threonine dehydrase in the presence and absence of AMP has been determined. The interrelationships of the three ligand system involving AMP, threonine, and the dehydrase itself have been investigated as they pertain to the sedimentation behavior of the active enzyme. The data presented indicate that threonine dehydrase is in a rapid association-dissociation equilibrium and with such a system, average values are obtained for the sedi- mentation coefficients which are not representative of a single species (91,92). In order to interpret the active enzyme centrifugation data it is therefore necessary to determine the subunit composition and to determine the molecular weight of the native threonine dehydrase by means independent of sedimentation velocity. The prior knowledge, both from earlier evidence presented in this thesis and from previous publications (2,43), that the dehydrase is in an association-dissociation equilibrium suggested that this should be done by two independent methods. Thus the native molecular weight was investi- gated using polyacrylamide gel electrophoresis and ultracentrifugation. 149 Polyacrylamide Gel Electrophoresis Molecular Weights of Native Threonine Dehydrase It has been shown (Figure 5) that the highly purified threonine dehydrase exists as two distinct bands on polyacrylamide gels. To determine the rela- tionship of the two bands of threonine dehydrase activity, the relative mobilities, Rm, of the enzyme under standard conditions as outlined by Hedrick and Smith (83) were determined. This method can distinguish between change isomers and molecular weight isomers. In the TRIS-asparagine electrode chamber buffer system (pH 7.3), supplemented with 1 mM AMP and 1 mM DTT, two protein bands were observed using Coomassie blue, Method A. The major band consisting of about 80% of the total protein exhibited a molecular weight of 71,000 and the minor band exhibited a molecular weight of 143,000 (Figure 19A). A good correlation to a straight line (r>0.98) was observed for the reference proteins. A second buffer system TRIS-glycine (pH 8.3) was used to eliminate the possibility that the observance of two bands was an artifact of the above gel system. To establish that both bands contained threonine dehydrase activity, the enzyme was located with the activity stain. For this reason AMP but not DTT was included in the running buffer (Figure 19B). Again two bands were observed with the major band exhibiting a molecular weight of 70,000 and the minor band a molecular weight of 132,000. The 150 Figure l9.--Determination of the native molecular weight of threonine dehydrase by polyacrylamide gel electrophoresis. The slope characteristic of each protein was determined in 6, 7, 8, 10, and 12% polyacrylamide gels as in Figure 20. The electrode chamber buffers were: (A) TRIS-asparagine, pH 7.3, 1 mM AMP, and 1 mM DTT; and (B) TRIS-glycine, pH 8.3, plus 1 mM AMP. For the TRIS-asparagine system the gels were stained with Coomassie blue, Method A. For the TRIS—glycine system the reference proteins were located with Coomassie blue, Method B, and the threonine dehydrase was located by the activity stain. The slopes were computer calculated. Slope 0.30 0.20 0.|0 030 0.20 0.l0 151 A. Tris-Asparagine,pH 7.3 J Lactic Dehydrogenase Bovine Serum Albumin Dimer Lipoyl Dehydrogenase KDPG Aldolase T Bovine Serum Albumin Monomer l l l 1 l 1 l L B. Tr is-Glycine . PH 8-3 Fumorose Dehydrogenase ‘Bovine Serum Albumin Dimer -KDPG Aldolase 0 Alkaline Phosphatase \Bovine Serum Albumin Monomer “-Amylase l 1 1 l l 1 IOO I40 Molecular Weight 60 152 data for the TRIS-glycine system is the average of two complete set of gels with 6, 7, 8, 10, and 12% of acrylamide. Hence the minor component is not an artifact of a particular gel system and its molecular weight is consistent with it being a higher molecular weight oligomer of the threonine dehydrase in the major peak. Hedrick and Smith observed that change isomers (such as the lactic dehydrogenase isoenzymes) but not size isomers gave parallel lines in a plot of Rm versus % acrylamide concentration and that molecules which were both size and charge isomers produced lines which were nonparallel and which intersected at some non zero % acrylamide concentration. However, size isomers with different molecular weights (such as bovine serum albumin polymers) produced lines which tended to intersect in the region of zero % acrylamide. They noted empirically, however, that the plots for bovine serum albumin and its dimer intersected between the acrylamide concentrations of 2% and 3%. In both systems used here, the line for the bovine serum albumin monomer intersects with the line for its dimer at a limiting acrylamide concentration .of about 4%; the data for the TRIS—glycine buffer system is presented in Figure 20. The two threonine dehydrase bands which have molecular weights similar to the monomer and dimer of bovine serum albumin also intersect at a limiting gel concentration of about 4% (Figure 20). This 153 Figure 20.--Slope characteristics of threonine dehydrase and bovine serum albumin in polyacrylamide gels of varying percentages. Electrophoretic conditions and staining were as in Figure 19 for the TRIS-glycine buffer system. The data for threonine dehydrase are derived from duplicate determinations except for that from 7% polyacrylamide gel. Threonine dehydrase I refers to the major band, and II refers to the minor band. The slopes were computer calculated. 154 0.8 A. Threonine Dehydrase 0.6 :' g 04— ° 2 1 0.2- 11 E a: O, 1 l n - 1 3 08*- B.Bovine Serum Albumin 0.6- ' . Monomer 0.4t . 0.2- Dimer I l l 6 8 l0 l2 % Polyacrylamide 155 is consistent with a similar, if not identical, charge for the two observed bands and suggests that they are size isomers with one being a dimer of the other. High Speed Equilibrium Ultracentrifugation Molecular Weight Analysis of Native Threonine Dehydrase.--The molecular weight of native threonine dehydrase in 0.1 M potassium phosphate, pH 8.0, 1 mM DTT and 1 mM AMP was determined by the procedure of thantis (89). Non-linear 1J1 E versus £2 plots were observed (Figure 21). This non-linearity is indicative of either multiple forms of threonine dehydrase in an association- dissociation equilibrium or of heterogeneity. If it reflects an association-dissociation, the resulting molecu- lar weights versus concentration curves for different loading concentrations should superimpose (97,98). That this condition is not met can be seen in Figure 22. The decrease in the molecular weight curves versus con- centration observed with increasing loading concentrations is consistent with heterogeneity (97,98). This agrees with data presented earlier using polyacrylamide gel electrOphoresis which showed two distinct molecular weight forms of threonine dehydrase. The molecular weight of the smallest species can be determined graphi- cally by extrapolating the data at the meniscus for different protein concentrations to infinite dilution 156 Figure 21.--High speed sedimentation equilibrium analysis of threonine dehydrase. Plot of 1n 3 versus £2 for 1.0 mg/ml of threonine dehydrase. Conditions were as in Figure 22. 157 3.0 P 49.00 49.50 50.00 48.50 158 Figure 22.--High speed sedimentation equilibrium analysis of native threonine dehydrase. The threonine dehydrase, 2.5 mg/ml, was dialyzed for 30 hours versus 0.1 M potassium phosphate, pH 8.0, 1 mM AMP, and 1 mM DTT. Dilutions were made in the dialysis buffer to the following concentrations: (A), 1.0 mg/ml; (O), 1.5 mg/ml; (I), 2.0 mg/ml; and ( + ), 2.5 mg/ml. The upper graph portrays the weight average molecular weight (MW) versus 9 (mg/ml) and the lower graph shows the number average molecular weight (Mn) versus 9 (mg/ml). The temperature was 7.6° and the speed was 16,205 rpm. 159 I40- 6‘ 0A 0A 0 8 9 A08 9. Q 9 '20” ADAO u + 9 (£2: +0+n+ o l00-A +0“: '90 [3 MW — ._ Al.0 / l : 80 8'2ng m .c o E J '1 +2'5 I l 2 I40— I"; '3'; IZO— Aone ’-‘° 0° 2 owed” 6 5 c1 "' IOO- 6863... + 5 5 a a 8?- + S 0 MN + n D A |.0mg/ml 80~a 0|5 02.0 + 2.5 60 I I l 1 0 048 0.96 L44 L92 Concentration (mg/ml) 160 (97). Table XI summarizes the molecular weights obtained by extrapolation of the curves of Figure 22 to infinite dilution. The average values of Mn=78,90014,100 and Mw=84,80015,200 are in fairly good agreement and are consistent with the presence of a single species at infinite dilution. The data were analyzed by plotting Mw versus 1/Mn according to Teller §t_al (99) (plots not shown). The following equation was then used to solve for the lowest molecular weight species: M1 = 1/2(a-/a2+4b) where Ml equals the molecular weight of the smallest species present, a is the intercept and b is the slope. Although the theory for this analysis was derived for a two species monomer-oligomer of an unknown state of association, the values obtained for threonine dehydrase by this method (see column 3 of Table XI) are in reason- ably good agreement with the extrapolated values obtained above. Effect of Increased AMP.--Reported values for the Ka of AMP range from 0.26 mM at dehydrase concentration <3f 0.021 ug/ml (52) determined kinetically to 0.05 mM at lligher dehydrase concentrations in the mg/ml range deter- Inined by equilibrium dialysis. Dunne gt 31 (52) show 161 TABLE XI Native Molecular Weights as Determined by High Speed Sedimentation Equilibrium Ultracentrifugation. Loading. M * M * Calculated Concentration n w Molecular Weight mg/ml 1.0 82,500 83,500 89,800 1.0 73,500 80,500 68,300 1.5 81,000 93,500 87,100 2.0 75,500 81,500 80,600 2.5 82,000 85,000 90,800 under Methods. The values for the molecular * . The number average (Mn) and weight average (MW) weights were determined by extrapolation of the initial portion of the plots of molecular weights versus concen- tration to infinite dilution. weights at each concentration were determined as outlined The calculated value for the lowest molecular species was determined according to Teller 3; a1 (99) as outlined in the text. 162 that the Ka for AMP decreases as the protein concentration increases and that this may explain the large difference in the reported value for the Ka for AMP. Assuming the higher value for the Ka the level of 1 mM AMP used in the previous equilibrium centrifugation experiment was only four times the Ka and therefore the allosteric site would not be totally saturated. Therefore high speed equilibrium ultracentrifugation was performed in the presence of higher concentrations of AMP, 5 mM and 10 mM, to determine the effect of AMP concentration on molecular weight (Figure 23). Although some difference was observed in the region of higher protein concen- tration the extrapolated values for the two levels of AMP, average 80,800 for Mn and 79,100 for Mw respectively, are in good agreement with those obtained at an AMP concentration of 1 mM in the previous experiment. The higher values for Mw encountered at higher protein concentrations in the cell may be a reflection of a small increase in the amount of hetereogeneity in this enzyme preparation. Effect of Sucrose on the State of Association.-- The values obtained for the molecular weights by equi- librium ultracentrifugation were not in agreement with those reported by previous workers who used sucrose gradient centrifugation (2,43) or sedimentation velocity analysis (45). Kinetic data suggested that sucrose 163 Figure 23.--High speed equilibrium analysis of threonine dehydrase in the presence of increased con- centrations of AMP. Molecular weight moments of threonine dehydrase (2 mg/ml) were plotted versus concentration in the cell. The buffer was 0.1 M potassium phosphate, H 8.0, 1 mM DTT, 5 mMAMP [Mn (0), MW (0) and 10 mM AMP [Mn (I), MW (0)]. The temperature was 4.3° and the speed was 15,214 rpm. ~—_. . 2.595 cozotEoocoo 164 3». mod m¢d O _ _ _ now I n. o w I I no 0 .nl I o n! o o m I I I o :00. c I o n o W 1.24. EEOT _2 I o o n. o .m. u.- l . 0 If I C C 00 x o 124. 2:5. _2 no 10m. NM NV C. o o n. as? 229;: o 1 a 3 03 o an:2< EEO... _2 165 favors higher molecular weight species increasing the state of association6 of the enzyme (52) and that this may have resulted in the higher molecular weights observed in sucrose gradients. Therefore high Speed equilibrium ultracentrifugation was performed in the presence of 10% sucrose at two enzyme concentrations. Again non- linear ln 9 versus £2 plots were obtained. The plots of molecular weight moments versus concentration within a cell, shown in Figure 24, indicate that more than a single species is present. The values for Mn and Mw of the lowest species obtained by extrapolation to infinite dilution are nearly identical for each loading concentration Mn=101,500 and Mw=102,000 at 1.0 mg/ml and Mn=86,000 and Mw=90,000 at 0.5 mg/ml. The average values obtained, Mn=93,750 and Mw=96'000' are higher than values without sucrose in the buffer (Table XI) and are consistent with the hypothesis that the presence of sucrose favors higher states of association. The Effect of L—homoserine and Sucrose on the Molecular Weight.--Since sucrose appeared to have a 6The stabilizing effect of glycerol and other polyalcohols has long been known especially in protection against cold lability. It has been postulated that this is due to the effect of the polyol on the state of Oligomerization. Indeed it has been demonstrated that in the case of glutamate dehydrogenase (84) sucrose has a tendency to stabilize the oligomer and prevent dis- sociation. For threonine dehydrase the rate of inacti- vation by dilution is decreased in the presence of glycerol suggesting that glycerol has an effect on the dissociation process (52). 166 Figure 24.--High speed sedimentation equilibrium analysis of threonine dehydrase in sucrose. Prior to centrifugation the threonine dehydrase was dialyzed for 24 hours versus 0.1 M potassium phosphate, pH 8.0, 1 mM AMP, 1 mM DTT, and sucrose, 10% (w/v). The loading concentration was 1.0 mg/ml [Mn (I), Mw (C1)]. The temper- ature was 5.3° and the speed was 16,198 rpm. -A _- -——~. ._4-.' .— *4‘-— _— 167 2.595 8:828:00 3w. mad mVO O _ _ _ 109 I W C I III a m. I I 0 tom. 1 I o M C . 0 mm. 5— O I 00 W. I o 1. o x o o n.lu.. 0 [OS 2 3—2 . O o o $985+ 168 positive effect on the state of association and L- homoserine appeared to favor a lower state of association at infinite dilution,7 it was of interest to investigate the combined effect of these substances. Sucrose gradient data (2,63) indicated that the presence of assay com- ponents using L-allothreonine as the threonine analog had no effect on the observed sedimentation coefficient at a dehydrase concentration of approximately 0.3 pg/ml. A value of 7.2 S was obtained both in the presence and absence of the analog. The data presented here were not in agreement with the sucrose gradient data, generally, and therefore, high speed equilibrium ultracentrifugation was performed in the presence of both sucrose and homoserine for comparison. The extrapolated values of Mn and Mw at three loading concentrations are nearly identical (Table XII). The average values, Mn=83,700i3,400 and Mw=84,0001 1,000, are similar to those obtained in the standard potassium phosphate buffer without sucrose and homoserine (Table XI) indicating that the presence of an analog did reduce the apparent molecular weight under the conditions of high speed equilibrium ultracentrifugation. Limiting molecular weights obtained using a higher concentration of DL-homoserine, 40 mM, with both forms of homoserine being good competitive inhibitors 7Unpublished observation. R. C. Menson and W. A. Wood, 1970. 169 TABLE XI I High Speed Equilibrium Analysis of Threonine Dehydrase in 20 mM L-Homoserine and 10% (w/v) Sucrose. Loading Concentration Mn Mw mg/ml 0.5 81,000 83,000 1.0 82,500 84,000 1.5 87,500 85,000 * The values for Mn and Mw were obtained as described in Table XI. The dehydrase was dialyzed for 24 hours versus 0.1 M potassium phosphate, pH 8.0, 1 mM AMP, 1 mM DTT, 20 mM L—homoserine, and 10% (w/v) sucrose, prior to centrifugation. The temperature was 4.3° and the speed was 16,026 rpm. 170 of the reaction, were consistent with the results obtained for 20 mM L-homoserine. Determination of Native Molecular Weight by Conventional Sedimentation Equilibrium Analysis Low speed sedimentation equilibrium analysis of threonine dehydrase in 0.1 M potassium phosphate, pH 8.0, 1 mM AMP and 1 mM DTT suggested polydispersity based on the plots of log of the fringe number (log E) versus radius squared (£2) at five different loading concen- trations. Illustrative data are presented in Figure 25. This is consistent with the behavior of threonine dehy- drase at a higher centrifugal force as presented earlier. The limiting values were calculated at the meniscus and at the cell bottom (Table XIII). A plot of the molecular weights obtained at the meniscus versus go, the initial loading concentration, extrapolated to a molecular weight of 97,500 (Figure 26). This value is higher than that obtained by high speed equilibrium studies and may reflect the greater contribution of a higher molecular weight species to the apparent molecular weight at the meniscus; one possibility is the contribution of the second dehydrase observed on polyacrylamide gels. The molecular weights of 325,000 and up observed at the bottom of the cell suggest that considerable amounts of dehydrase exist at higher states of aggregation. 171 Figure 25.--Conventional sedimentation equilibrium analy- sis of threonine dehydrase. Plot of log 3 (fringes) versus £2 for threonine dehydrase. Threonine dehydrase, 2.3 mg/ml, was dialyzed versus 0.1 M potassium phosphate, pH 8.0, 1 mM AMP, and 1 mM DTT for 24 hours and then diluted to 1.38 mg/ml with the dialysis buffer prior to centrifugation. The temperature was 7.6° and the speed was 6,975 rpm. The protein concentration was determined by dividing the fringe number by 4.1 fringes/m1 (76). 172 3.0 _N.0 _ 0.0 an O .03 O :5 0 Log c (fringes) N0 L kfiNOO _ _ _ _ #000 A500 @000 9.00 N w 0.0 m0 mg /ml 173 TABLE XI I I Low Speed Equilibrium Analysis of Threonine Dehydrase. Mapp* at Mapp* at Concentration Meniscus Cell Bottom mg/ml 0.34 100,500 357,000 0.67 86,000 399,900 1.04 89,800 393,200 1.41 85,900 338,100 1.67 89,500 335,400 * Ma p was calculated from the limiting slopes of a plot of tge log of the fringe number versus radius squared (r2). Centrifugation conditions and sample prepa- ration were as in Figure 25. 174 Figure 26.--Molecu1ar weight of threonine dehydrase as a function of concentration determined by conventional sedimentation equilibrium analy- sis. The molecular weights are the limiting values obtained at the meniscus as shown in Table XIII. 175 I l l |00 - 8 e e 29' x lubleM minoalow L50 LOO 050 Co (mg/ml) 176 Sedimentation Velocity The s of threonine dehydrase at three differ- 20,w ent concentrations, 0.5, 1.0 and 2.0 mg/ml, was deter- mined in 0.1 M potassium phosphate, pH 8.0, 1 mM AMP and 1 mM DTT using a photometric scanner system. The sedimentation was followed at 413 nm, the peak absorbance of the pyridoxal phosphate cofactor, to eliminate the interference caused by AMP and DTT at lower wave lengths. The data were corrected for density and viscosity prior to extrapolation to infinite dilution (Figure 27). The s value of 5.75 S suggests that the dimeric form of _20 ,W threonine dehydrase is present and is consistent with the data presented earlier obtained from active enzyme centrifugation. It is not in agreement with the published value of 8.16 S as determined by Shizuta 35 31 (45). Determination of the Protomer Molecular Weight and Composition The previously reported molecular weight of 40,000 for the protomer of threonine dehydrase was based on (a) the limiting value of 3.2 S as determined by sucrose density gradient centrifugation (2,43); and (b) a minimum molecular weight of 38,000i2,000 per mole of pyridoxal phosphate (45). As stated earlier, when analyzing an association- dissociation equilibrium, it is necessary to establish the molecular weight and composition of the protomer 177 Figure 27.--Sedimentation coefficients of threonine dehydrase as a function of dehydrase con- centration. The dehydrase, 2.0 mg/ml, was dialyzed for 24 hours versus 0.1 M potassium phosphate, pH 8.0, 1 mM AMP, and 1 mM DTT. Appropriate dilutions were made in the dialysis buffer prior to centrifugation. Sedimentation of the dehydrase was monitored at 413 nm with a photometric scanner. Centri- fugation was at 59,780 rpm and the temperature was 20°. 178 2.0 Dehydrase (mg /ml) l I J l o. O. o, O. Q [- 0 ID <1- '9 M‘Ozs‘ Iuelouleoo uouowawlpas 179 under completely dissociative conditions in order to interpret the oligomeric structure of the enzyme. This section investigates the molecular weight of threonine dehydrase under denaturing conditions. Investigation of the Protomer Molecular Weight and Structure Using Polyacrylamide Gel ElectrOphoresis SDS Gel Electrophoresis of Standard Proteins.-- SDS gel electrOphoresis was carried out in both 8 and 10% gels at a constant ratio of acrylamide to bisacrylamide. Standard curves were prepared by electrophoresis of mixtures of two to five proteins of known molecular weights (Figure 28). The relative mobility, Rm, of each standard protein was determined individually in a control run prior to mixing. Upon storage in the SDS gel buffer some of the standards were subject to break- down producing small molecular weight fragments. There- fore new mixtures were routinely made up from stock solutions (10 mg/ml) of the individual proteins and the stock solutions were periodically checked for protein integrity. Illustrative results for the 8% gels are shown in Figure 29; all slopes were computer calculated as outlined in Methods. SDS Gel Electrophoresis of Native Threonine Dehydrase.--Threonine dehydrase was incubated in the SDS sample buffer (see Methods) for one minute in a boiling 180 Figure 28.--Scans of polyacrylamide gels after electro- phoresis of standard enzyme mixtures in sodium dodecyl sulfate. The mixtures in SDS sample buffer were incubated for one minute in a boiling water bath, brought to 10% (v/v) in glycerol and electrophoresed in an 8% gel at 8 mA/tube for a set time (see Methods). The gels were stained with Coomassie blue, Method C, and scanned at 600 nm. U1 O 01 ,O O Absorbance , 600nm 181 Myoglobin % KPDG Aldolase Carbonic / ._ Anhydrase\ Bovine fl Serum Ovalbumi . _Albumin “ KDPG Aldolase Fumarase W Hemoglobin .. \ Catalase % Cyto- chrome 1C W U L/J U K I l J l 2 4 6 8 Migration Distance (cm) 182 Figure 29.--Sodium dodecyl sulfate gel electrophoresis of threonine dehydrase. The conditions were identical to those in Figure 28. The rela- tive migration of each protein was related to the migration of the internal standard, KDPG aldolase. 183 3:32). 2:63 S N._ 9 m0 0.0 to 0 _ _ _ a a _ £86066: 5203022 ION $2024. oaox I $225.3 9:09 I too D #00 5252.4 Eamm 9.38. low gal x lubleM JDInoanw 501 184 water bath. KDPG aldolase was included in the sample as an internal standard. As shown by the arrows in Figure 29, two bands were observed on 8% gels for the dehydrase with average molecular weights of 36,000 and 75,500. No low molecular weight components were evident. Although the above conditions should be sufficient to reduce all the disulfide bonds (100), the effect of longer incu- bation times of 1, 5, and 10 minutes was investigated in the above sample buffer without KDPG aldolase. Figure 31 shows that no intermediate peaks representing reduction of additional disulfide bonds were observed, and no indication of a small polypeptide chain was found. The molecular weights for the three incubation times were nearly identical as determined in 10% acrylamide gels (Table XIV). No relative change in the size of the two peaks was observed indicating that no significant con— version of the higher molecular weight band to the lower molecular weight form occurred through further reduction of disulfide bonds. The averaged value obtained for the major band of 34,700 was in good agreement with the value obtained above on the 8% gels. Since the second band was shown to have a molecular weight greater than 60,000 which is the upper limit of linearity for a 10% gel system (101) quantitation of its molecular weight was not carried out. The ten minute incubation time was 185 Figure 30.--Polyacrylamide gel electrophoresis of native threonine dehydrase in sodium dodecyl sulfate. The conditions were the same as in Figure 28 except that 10 ug of KDPG aldolase were included in the sample as an internal standard. Ten pg of threonine dehydrase were layered on the gel. 186 E3 8:220 5:825. v $2026. odox :2 m _ _ _ x233 :: a 686me 65:02,: H 886me 65:08:... l0. 0 wu 009 ‘ eouoqmsqv 0. 187 Figure 31.--Effect of incubation time under denaturing conditions on electrophoretic pattern of threonine dehydrase. The conditions were the same as in Figure 30 except that KDPG aldolase was not present in the incubation mixture, and the incubation times were (A) ten minutes, (B) five minutes, and (C) one minute. 188 0.5 - _ .0 01 Absorbance , 600 nm \ An. I KI: |.0— lminute at l00° C. 05] 11 L V 7 IO minutes at IOO° C. \A l.0- 5minutes at IOO° C. I ll I ll l l 0 2 4 6 Migration Distance 189 III" I..." .HH ammuphswp maficomnnu .Ucmn Homes on» How uanmB uaHsomHOE pmumHsono on» on whammy HH uanmB HaHSUmHOE 6cm .H ammuomnmp mCHcomunu mo pawn Momma may now omumasoaao unmfimz Haasomaoa any on mummmn H unmflm3 Haasomaoz .manme on» CH Umumoflpcfl mm mmmaucmoumm Ham cam mafia» coflumnsocfl wnu How umwoxm om musmflm CH mm mHmB mCOHuHUcoo use: oom.mm oom.mm OH ca Hummus ooo.mo ooo.mm OH OH Himwlz ooo.am ooo.vm OH OH Himhtz ooo.mm ooo.mm m oa Himhtz ooo.ao ooo.mm H OH Himhlz ooo.mn ooo.mm a m Hummus ooo.mn ooo.mm a m Hummus mmuzcflz. w HH H mEHB mofleaaauoaaaom cofluauammum coflumnsocH memucm munmfimS Haasomaoz .mflmmnonmouuomam H00 mam Eoum mmmnomsma unaccouna m>wuaz mo musmflmz Haasomaoz uflcznsm >HX mnmfla 190 repeated using a different enzyme preparation and the data are incorporated in Table XIV. SDS Gel ElectrOphoresis Analysis of Threonine Dehydrase Containing Blocked Sulfydryl Groups.--Dunker and Rueckert (101) point out that the formation of mixed disulfides during electrophoresis can be prevented by incubation of the protein samples with.iodoacetate under non-reducing conditions. For further SDS gel analysis, threonine dehydrase was reacted with iodoacetate in both guanidine hydrochloride and urea using reducing and non- reducing conditions (101). Alkylated threonine dehy- drase, 1.5 mg, was prepared in the following mixtures: (1) 8 M urea, 0.2 M TRIS-HCl, pH 8.2, 1% SDS and 0.002 M iodoacetate; (2) the above solution except for the addi- tion of 0.1 M mercaptoethanol and the use of 0.2 M iodoacetate; (3) and (4) solutions 1 and 2 respectively except for the substitution of 6 M guanidine hydrochloride for 8 M urea. The pH was adjusted to 8.2 with KOH after the addition of 6 M guanidine hydrochloride. The mixtures without the iodoacetate were incubated at 40° for one hour with nitrogen passing through the solution. Threonine dehydrase (1.5 mg) was added to each mixture and allowed to incubate for an additional hour with nitrogen forming the gas-meniscus interface. The apprOpriate amount of iodoacetate was then added to each incubation tube and the reaction was allowed to proceed for one hour. The 8 M urea was then dialyzed versus 0.01 M sodium phosphate, pH 7.1, plus 0.1% SDS. Dehydrase alkylated in the solu- tions containing 6 M guanidine hydrochloride was first 191 threonine dehydrase alkylated in the solutions containing dialyzed versus distilled water to prevent formation of ”w a guanidine hydrochloride precipitate and was then dialyzed versus distilled water to prevent formation of a guanidine hydrochloride precipitate and was then dialyzed against the above buffer. The samples were 1yophilized and taken up in distilled water to one-tenth the original volume. This provided the samples in a buffer solution approxi- mating that of the SDS sample buffer (see Methods). The total number of S-carboxymethyl groups was determined by amino acid analysis. Values of 20.2 and 20.7 for the urea solutions and 16.1 and 15.5 for the guanidine hydrochloride solutions were obtained under non-reducing and reducing conditions respectively. The expected values of 24 (45) to 29 carboxymethyl cysteine residues was presented earlier in Results. Although these results indicate that all of the sulfhydryl groups were not blocked (45) (Table VIII), the near identity of the number of carboxymethyl cysteine residues obtained for each denaturant under both reducing and non-reducing conditions suggest that no intermolecular disulfide bonds remained. Aliquots of each alkylated sample were made 4 M in urea, split in two and mercaptoethanol to a 192 final concentration of 0.31 M was added to one of each pair prior to incubation for five minutes in a boiling water bath. SDS gel electrophoresis was performed using 8% gels. The results from each set, summarized in Table XV, are nearly identical and the averaged values of 35,400il,200 and 67,600i2,900 are consistent with the values obtained for the native threonine dehydrase as determined by SDS gel electrophoresis in the previous section. No evidence of smaller molecular weight com- ponents was observed. Smithies (102) used a monothiol-catalyzed disfulide interchange reaction between the protein being investi- gated and a disulfide reagent such as dithiothrieitol (DTT) to block sulfhydryl groups. Using this method he was able to prevent formation of aggregates of haptoglobin by disulfide interchange during electrophoresis. Alkylated threonine dehydrase samples (20.2 SCM groups) were pre- pared in the mixtures outlined in Table XVI and analyzed in 8% gels by SDS gel electrophoresis. KDPG aldolase was included in each sample as an internal standard. The values obtained for the molecular weights in each case are essentially identical for the two threonine dehydrase species, 38,100i400 and 69,300:l,700 for threonine dehydrase I and threonine dehydrase II respectively. No change in relative amount of each species was observed. As can be seen in Figure 32, no evidence of smaller Determination of the Subunit Molecular Weights S-Carboxymethylated Threonine Dehydrase 193 TABLE XV by SDS Gel Electrophoresis. of Molecular Weight * Mercapto- Sample ethanol I II Urea nonreducing — 35,200 68,500 Urea nonreducing + 34,500 65,000 Urea reducing - 34,500 65,000 Urea reducing + 35,200 68,500 GuHCl nonreducing - 35,200 66,800 GuHCl nonreducing + 33,000 73,500 GuHCl reducing - 34,500 65,000 GuHCl reducing + 36,200 68,500 * Each alkylated sample prepared as described in the text was subjected to SDS gel electrophoresis with and without added mercaptoethanol, 0.31 M to the SDS sample buffer, as indicated. "Urea, nonreducing" and "urea, reducing" refer to the samples alkylated in urea under nonreducing and reducing conditions, respectively. "GuHCl, nonreducing" and "GuHCl, reducing" refer to the samples alkylated in guanidine hydrochloride under non- reducing and reducing conditions, respectively. weight I and II are as in Table XIV. Molecular 194 .>x manna GH mm mum HH cam H unmflw3 Haasomaoz .omumoflocfl mcowuauu Icmocoo any um maoflsu mo coeuappa may owUSHUCH can Houwumam How mom: cwusufiumnzm m nmzounp N mmameam .mposumz CH confluommp ma Hmmwsn coflumnsoafl mom wumpcmum on» mpcmmmummu a wamfiam .mm musmwm ca mm mum3 mcofluflvcoo mammuosmonuomam one fl o.m~ ooo.ee o.ee ooo.mm H.e e mea.o e m.ma ooe.oe e.om eeo.mm H.o e eHo.o e m.m~ eeo.ee p.65 ooo.mm H.o e eaee.o m m.- ooe.oe e.ee ooo.mm I- e em.o e e.e~ ooo.me 0.0m ooe.mm I- e me.o m e.me ooe.ae 0.3m ooo.wm I- e mea.o N e.ea oee.ee e.~m ooo.mm I- I- meo.e H z z z w HH w H 99o may: Hosanna mamemm mmauomnma mmanpmnao loumaoumz « wcflcomnns mcflcomune maceueeem .mHoHQMHQ psalocoz mo mocmmmum may :fl mfimmuonmouuomam How mom an pmcfleumuwo mm mmauoanwn mafiaomuce mo mauom 039 on“ mo mDCSOE< m>flumamm any can mpnmflmz Haasomaoz uflchsm H>X mnmiwfi 195 Figure 32.-~Sodium dodecyl sulfate gel electrOphoresis of threonine dehydrase incubated under various reducing conditions. All samples were in 0.01 M sodium phosphate, pH 7.0, 4 M urea plus: (A) 0.143 M B-mercaptoethanol; (B) 0.86 M B-mercaptoethanol; and (C) 0.143 M B-mercaptoethanOl plus 0.1 M dithiothreitol. The scans are of (A) sample 2, (B) sample 4, and (C) sample 7 of Table XVI. 196 LO” 0.72M fiMe ,O (11 l ,— .0 O O l l Absorbance , 600 nm O 01 ,O O .WUM Threonine Dehydrase I "KDPG Aldolase II L: __ 1.74 v — 0.86M flMe, 4M Urea Threonine Dehydrase I "KDPG Aldolase H 0.l43 M BMe, 4M Urea , DJ M DTT "07 Threonine I ”KDPG Aldolase Dehydrase I] 0.5- U L; I 2 l l J I l . O 2 4 6 8 IO Migration Distance (cm) was I x _ _ ' ‘. -- .. ‘ . 197 molecular weight components was observed despite the stringent reduction conditions. Thus, apparently no smaller polypeptide chains of threonine dehydrase were present and threonine dehydrase I and II are not related by the presence of intermolecular disulfide bonds. SDS Gel Electrophoresis Analysis of Electrophoreti- callerurified Threonine Dehydrase.--The major band of electrophoretically purified threonine dehydrase prepared as outlined in Methods was analyzed by SDS gel electro- phoresis on both 8% and 10% gels. Four different fractions from at least two threonine dehydrase preparations were analyzed. One, sample A, was also used for kinetic analysis (see earlier Results) and for urea gel electro- phoresis (see below). All samples were incubated in the SDS sample buffer (see Methods) in a boiling water bath for one minute. Previous evidence (Table XIV) showed that this condition was sufficient to reduce all inter- molecular disulfide bonds. The data are summarized in Table XVII and illustrative results are presented in Figure 33. In contrast to earlier data, a small differ- ence was observed between the average value obtained for 8% gels of 36,500 and that obtained for 10% gels of 38,800. Since the determination of subunit molecular weight by SDS gel electrophoresis has an accuracy of i 10%, the data are considered to be in substantial agreement, and the overall average of 37,300 is in good 198 TABLE XVII SDS Gel Electrophoresis of Electrophoretically Purified Threonine Dehydrase. * Polyacrylamide Molecular Sample % Weight A 8 36,500 A 8 36,500 A 8 34,500 A 10 39,500 A 10 38,500 A 10 38,500 B 8 38,500 C 8 36,000 C 8 37,000 D 10 38,500 *Conditions as in Figure 28. Samples A, B, C, and D refer to different fractions of electrophoretically purified threonine dehydrase. Sample A was the same eluate used for kinetic analysis as outlined earlier in Results and summarized in Table VII. 199 Figure 33.-~Polyacrylamide Gel ElectrOphoresis of Electro- phoretically Purified Threonine Dehydrase I in Sodium Dodecyl Sulfate. (The conditions were as in Figure 28.) 200 AEov 8:965 5:292 V N 1 _ fl 1 $2024 000v. q _ _ _ l" I n 323:8 65:08:... I Emeonaotomfi .60 m o w L0. 0 wuoog ‘ eouoqiosqv g 201 agreement with other SDS gel electrophoresis data for the molecular weight of the major form of threonine dehydrase (threonine dehydrase I). Analysis of Electrophoretically Purified Threonine Dehydrase I byiUrea-Gel Electrophoresis.--Ana1ysis of the major band of threonine dehydrase by SDS gel electro- phoresis indicated that the protomer is made up of a single polypeptide chain with a molecular weight between 35,000 and 36,000 as evidenced by a single peak. The possibility existed that the oligomer is made up of nonidentical protomers with similar if not identical molecular weights. In this connection SDS gel electro- phoresis will only distinguish molecules on the basis of size and not charge (103). It is also possible that a low molecular weight polypeptide chain, if present, would not have been detectable. Urea gels will separate molecules on the basis of both size and charge (103). They also allow one to identify low molecular weight components which are not readily identifiable on SDS gels. Burgess (88) unequivocally identified the low molecular weight component of 9,000 of ribonucleic acid polymerase with urea gels. This band represented approximately 2% of the protein molecule and was observed as a diffuse weakly stained region on SDS gels. With urea gels Marciani and Kuff (104) resolved into several distinct bands the diffuse band of low molecular weight ”THREE, { 202 components of membranes. In the case of threonine dehy- drase, a small polypeptide chain of 11,000 to 13,000 (2) would represent at least 30% of the dehydrase molecule and should be observed with SDS gel electro- phoresis since the reference proteins with molecular weights in this range were readily observable (Figure 28). However to be sure that a small molecular weight component had not gone undetected and to eliminate the possibility that threonine dehydrase was composed of nonidentical monomers of similar molecular weights, the electrophoretically purified threonine dehydrase I, i.e. the major band, was analyzed on urea gels. The integrity of this purified dehydrase was demonstrated earlier by its typical kinetic properties and the possession of both a catalytic and allosteric site. The same fraction used to obtain the kinetic data and analyzed by SDS gel electrophoresis was used in this experiment. A single peak was observed with no evidence of smaller molecular weight components (Figure 34). Similar results were obtained with other samples of electrophoretically puri— fied dehydrase. This indicated that the monomer of threonine dehydrase I is made up of a single polypeptide chain and that there is a high probability that the oligomer consists of identical monomers. If). {Elm mew H 203 Figure 34.--Analysis of electrophoretically purified threonine dehydrase I by urea-gel electro- phoresis. ElectrOphoretically purified threonine dehydrase prepared as outlined in Methods was electrophoresed in a 7.5% poly- acrylamide gel containing 8 M urea. The electrode chamber buffer consisted of TRIS- glycine, pH 8.7, and the sample was electro- phoresed at 3.5 milliamps per tube for a set time. The gel was stained using Coomassie blue, Method C, and scanned at 600 nm. 204 l... , 4.ch A58 8:955 8:895. N fl m 0 v \f— _Il~ _ H 82:»ch 63:85...- m_8..o::o:8_m _mo 85 LQ 0 am 009 ‘ eouoqmsqv 9 205 Analysis of the Carboxymethyl Cysteine Derivative of Threonine Dehydrase by Conventional Sedimentation Equilibrium The carboxymethyl derivative of threonine dehydrase was investigated by low speed equilibrium analysis in guanidine hydrochloride. A stock solution of the de~ rivatized dehydrase was dialyzed overnight at 4° against two changes of 6 M guanidine hydrochloride, pH 6.0 plus 1% (v/v) freshly distilled B-mercaptoethanol. Appropriate dilutions were made using the dialysis buffer. Non-linear curves were observed for plots of the log of the fringe number (log 9) versus £2 indicating the presence of more than one species (Figure 35). The plot of molecular weight versus initial loading concentration, 90' for the limiting molecular weights of the species at the cell bottom produced a value of 34,500 at infinite dilution (Figure 36). Goldberg and Edelstein (105) point out the difficulty in determining an accurate molecular weight at the bottom of the cell for paucidisperse system. Due to the contribution of the lower molecular weight species to the plot of log 9 versus £2 at the bottom of the cell, the value obtained may be an underestimate of the actual molecular weight, although 34,500 is in reasonable agree- ment with the values obtained for threonine dehydrase I by SDS polyacrylamide gel electrophoresis. Because the amount of low molecular weight material was small the molecular weight values from the limiting 51 f.“ I a." TL; mum.“ i 206 Figure 35.--Determination of the subunit molecular weight of threonine dehydrase by conventional sedi- mentation equilibrium in guanidine hydrochloride. The alkylated threonine dehydrase concentration was 1.62 mg/ml in 6.0 M guanidine hydrochloride and freshly distilled 0.2 M mercaptoethanol, pH 6.0. Centrifugation was performed at 23,131 rpm and the temperature was 23.2°. Log 0 (fringes) 207 N .e .m .00 5 .13 O O O O O O l i l T T £3_ 0 O .b O .00.. 8 I b O O 6* -. .0 .0 01 I .O.. .0 O 8 ., O O .0 91.. ' A) cit 4L '0 '0 '0 '0 mg/ml Wit—w . 208 Figure 36.--Molecular weight of alkylated threonine dehy- drase as a function of concentration. The graph shows molecular weight as a function of concentration at: (A) the cell bottom; and (B) the meniscus. The concentrations were determined by dividing the c value in figures by 4.1 fringes/mg proteifigas outlined in Methods. Conditions are as in Figure 35. 209 SE 95 6o om o._ _ 0 Li _ O 88222 2962, 838.22 626m :00 292s 838.22 10. L9 [1 ll Imm 10m Imm 0| x NfilaM JDanSIOW 2 210 slopes at the meniscus were difficult to determine. A limiting value of about 11,000 to 13,000 was calculated for the smaller species at infinite dilution. This value for a small polypeptide chain is consistent with the data of Whanger gt a1 (3) that the threonine dehydrase protomer consisted of smaller polypeptide subunits. Threonine dehydrase which had been artificially oxidized and then reduced with dithiothreitol gave multiple peaks when analyzed by both sucrose gradients and Sephadex column chromatography (3). The exact values for these peaks was not repetitive between experiments. However the Spacing of the molecular weight values calculated for the multiple peaks suggested that the threonine dehydrase protomer of 40,000 consisted of at least two polypeptide chains with at least one of the chains having a molecular weight of 10,000 to 13,000. Fosmire and Timasheff (106) investigated the sub- unit molecular weights of lactic dehydrogenase from beef heart and dogfish muscle using equilibrium centrifugation in guanidine hydrochloride. Occasionally they observed smaller molecular weights than 35,000 for both lactic dehydrogenase species, and they concluded that this was due to a contaminant in one of the lots of guanidine hydrochloride. Deal (107) has reported similar problems with equilibrium centrifugation in guanidine hydrochloride. Since no evidence of this small polypeptide chain was 211 observed in any of the experiments analyzing the subunit composition of threonine dehydrase using polyacrylamide gel electrophoresis under strongly reducing and denaturing conditions, it must be concluded that the low molecular weight species is an artifact. ‘. . 5.1 emu—nun. DISCUSSION The primary objectives of this thesis were (1) to thoroughly investigate the quaternary structure of threonine dehydrase and (2) to solve the apparent dilemma between the four-fold change in molecular weight indicated by physical means (2,43) and the two-fold change in molecu- lar weight indicated by kinetic analysis (43) upon acti- vation of dehydrase by AMP. In accomplishing this we were also able to investigate the interactions of the three ligand system composed of threonine, AMP, and dehy- drase itself and the effects of ligand binding on the physical structure of the active enzyme under actual catalytic conditions. To understand the physical changes in the active enzyme structure, it was first necessary to establish molecular weight and composition of the subunit and also of the native enzyme. Protomer Molecular Weight and Composition Previous investigators had established a sedi- mentation coefficient of 3.2-3.6 S for threonine dehydrase in the absence of AMP at protein concentrations in the submicrogram range by using sucrose gradients (2). A molecular weight was assigned to the protomer based on 212 213 (a) gel filtration in calibrated columns, (b) the g value as determined on sucrose gradients (2) and assuming an average specific volume for a globular protein, and (c) using the minimum value per molecule of the pyridoxal phosphate cofactor (45). All of these methods agreed quite closely in assigning a molecular weight of 38,000: 2,000 to the protomer of threonine dehydrase as the rmmmmww smallest active species in the absence of AMP. There was some evidence, however, that the protomer was made up of smaller subunits with a limiting molecular weight for the smallest subunit of around 13,000 (2), Initial analysis of the protomer molecular weights of threonine dehydrase by SDS gel electrophoresis using standard published conditions (10,84) produced two distinct bands (Figure 29). The major band consisting of about 80% of the total protein exhibited a molecular weight of 36,000 in 8% acrylamide gels and the minor band a molecular weight of 75,500. Incubation of the dehy- drase in 1% SDS and 1% B-mercaptoethanol for one minute in a boiling water bath should have been sufficient to reduce all the intermolecular disulfide bonds (100). Under these conditions no evidence of a subunit smaller than 34,000-36,000 was observed and the minor band of a higher molecular weight was still prominent. More stringent reduction conditions were used by increasing the reduction times and the mixtures were analyzed on 214 a 10% acrylamide gel which gives more linear results and sharper bands in the lower molecular weight range. This would eliminate the possibility that a smaller polypeptide chain had gone undetected in the 8% gels. Again the minor band was prominent with no detectable change in its relative magnitude and no band smaller than 34,000 to 36,000 was observed. An average molecular weight for the major band of 35,300i700 (Table XIV) was obtained for seven experiments at two acrylamide concentrations, 8% and 10%, and three different incubation times, one, five, and ten minutes. Threonine dehydrase was alkylated under both reducing and nonreducing conditions in urea and guanidine hydrochloride. Values for the number of S-carboxymethyl cysteine groups as determined by amino acid analysis were 20.7 and 20.2 for urea and 15.1 and 15.5 for guanidine hydrochloride for the reducing and nonreducing conditions, respectively. Comparison of these values to the expected value of 24 half cystines per tetramer (45) indicates that complete reduction and alkylation had not occurred, although the near identical agreement of the values under reducing and nonreducing conditions suggest that all intermolecular disulfide bonds had been alkylated. The molecular weight for each alkylated sample was determined by SDS gel electrophoresis. Since the data suggested that complete alkylation had not 215 occurred, split samples were analyzed in the presence and absence of 0.31 M B-mercaptoethanol in the incubation buffer. No difference in either gel pattern or molecular weight was observed for the split samples (Table XV). A major band with a molecular weight of 34,800i900 (n=8) and a minor band with a molecular weight of 67,500f2,900 (n=8) was observed. Again no bands with a molecular weight below that of the protomer were observed. Smithies (102) and others (108) have reported the presence of aggregate bands formed during electrophoresis due to disulfide interchange. One author (10) prevented aggregation by including iodoacetate in the SDS incubation mixture. As demonstrated earlier a higher molecular weight band was observed for threonine dehydrase even with partially alkylated dehydrase in the presence of a large excess of B-mercaptoethanol. Therefore, to eliminate the possibility that the higher molecular weight minor band was an artifact of disulfide inter- change, alkylated dehydrase containing 20.2 S-carboxymethyl cysteine groups was incubated in the presence of various amounts of B-mercaptoethanol with and without dithiothreitol. When free sulfhydryl groups are formed on the enzyme by a catalytic amount of a monothiol in the presence of an excess amount of dithiol, a mixed disulfide will result between the enzyme sulfhydryl and the dithiol (102), thus preventing any artifactual aggregation. The results of 216 this experiment (Table XVI and Figure 32) showed no indication of (1) a decrease in the ratio of threonine dehydrase I, molecular weight 38,100f400, to threonine dehydrase II, molecular weight 69,400il,700; or (2) a smaller subunit than the protomer. Electrophoretically purified threonine dehydrase I had a molecular weight of 37,400:l,500 (n=10) when analyzed by SDS gel electrophoresis. Dehydrase prepared in this manner retained both its catalytic and regulatory properties. The kinetic constants of Km and Hill 2 were essentially identical with those of the control experiment using highly purified threonine dehydrase. As can be seen in Figure 33 no evidence for either the minor band, molecular weight 70,800, or of smaller sub- units was found. The appearance of a single band for electro- phoretically purified threonine dehydrase I on urea gels again indicated that not only were there no smaller subunits than the protomer, but that the protomers were probably identical in charge as well as in molecular weight (107). It also showed that band II was not easily produced from band I. Analysis of alkylated threonine dehydrase by sedimentation equilibrium in 6 M guanidine hydrochloride and 1% B-mercaptoethanol seemed to substantiate the previous investigator's data that the protomer consisted 217 of more than one polypeptide chain. Non-linear plots of 1n 9 versus £2 were obtained with limiting molecular weights of 13,100 and 34,100. The value of 34,100 was in reasonable agreement with the monomer of 35,900 as established by SDS polyacrylamide gel electrophoresis in this thesis and of 38,00012,000 as published by previous investigators (45). The value of 13,100 agreed I. ‘1; )u I a l a well with the molecular weight of the polypeptide sug- gested by Whanger et 21 (2) from analysis of artificially oxidized and rereduced threonine dehydrase on sucrose gradients and Sephadex columns. However, the data pre- sented here indicates that either (1) there was incomplete separation of all dehydrase molecules into smaller sub- units because this would require a limiting molecular weight for the larger fragment to be around 25,000 assuming a protomer molecular weight of 38,000i2,000 as previously published; or (2) that the protomer molecular weight was nearer to 45,000. Thus exhaustive analysis of highly purified threonine dehydrase using polyacrylamide gel electro- phoresis indicated that (l) the two forms of native dehydrase are not related by a disulfide bond; (2) that the protomeric structure of threonine dehydrase I con- sists of a single polypeptide chain with a molecular weight of 35,900 (Table XVIII); and (3) that the oligomeric threonine dehydrase I is made up of identical protomers. 218 .o.m mm .Hocmsumoummonmfi z N.o .Hom mcflvwcasm 2 mm .HH>x magma mom mcofiuflccoo Home .muamasm Hmomuoc Edwcom wa LVommz z Ho.o a .o.e ma .Hoamcumouamoums 2 HM.: .mummasm Hmomcoc eseoom ma .eommz s Ho.oo .H ammucmnmc acficomncu Umflmwusm >HHMOfluwHonmouuowamn .o.e :6 .Hocweumoemmoume z meo.o .mummasm Hmomeoe aseeom .eoemz z He.om ooa.em u .32 aseuaeaesam camam 30a mmmz.aom meeeecmso assumemxonumo oov Hooa.mm mama mom a Hmnumamxonumu oov woom.vm mama mam cmam Hmnumswxonuau eem.aeooe.em memo mom ommz .mem Assamesxoeemo oom.aeooe.em mamo mom mmmz .mom am>epmz eon noom.mm memo mom mmmz .mam m>eumz unmfima “wasomaoz wosumz uqm>aom m>wum>fluma .H mmauvmsmo mcflcomuna mo HmEououm may now anon unmwmz Haasomaoz may no mHmEEDm HHH>x MAMHuod m h.m huHoon> coHumuamEvam oom.on mHmmuoamonuome mUHEmHmHomeom uanwz HMHsomHoz ucm>Hom @ocumz .H mmauwmnmo mchomHSB 0>Humz HOm mama ucmHmz uaHdowHoz may no mHmEEsm XH>x mHm