THE MECHANISM OF THREONINE DEHYDRASE OF ESCHERICHIA COLI The‘sls lor flue Degree J pll. D. MICHIGAN STATE UNIVERSlTY. Allen Thurman Phillips 1964 This is to certify that the thesis entitled The Mechanism of Threonine Dehydrase of Escherichia coli presented by Allen Thurman Phillips has been accepted towards fulfillment of the requirements for Ph . D . degree in Biochemis try // /V///%»zz{/ Major professor Date September 25, 1961+ 0-169 LIBRARY Michigan State University n “his ill! in pro COIT but\ ABSTRACT THE MECHANISM OF THREONINE DEHYDRASE OF ESCHERICHIA COLI by Allen Thurman Phillips L-threonine dehydrase from Escherichia coli grown anaerobically on an amino acid medium is activated by adenosine monophosphate (AMP) and reduced glutathione (GSH). This enzyme is presumed to function in the biodegradation of threonine. The enzyme has been purified 50-fold from crude cell-free extracts and its prOperties investigated. Inhibition studies with KCN A and hydroxylamine suggested that pyridoxal phosphate might also be required for activity. A partial reactivation could be observed upon addition of pyridoxal phosphate to the inhibited enzyme. The purified enzyme exhibited an absorption maximum at 402 mu and this absorption was considerably decreased by borohydride treat- ment. Hydrolysis of enzyme reduced with tritium-labeled borohydride allowed the isolation of Nb-pyridoxyllysine, indicating that pyridoxal phosphate was bound as a Schiff base to a lysyl residue in the native enzyme. In order to elucidate the mechanism of activation by AMP and GSH, a detailed study of the reaction mechanism was undertaken. Experi- ments with D30, H3018 and a-tritiothreonine confirmed that the reaction proceeds first through dehydration to yield a-aminocrotonate, and this compound subsequently tautomerizes and hydrolyzes to yield a-keto- butyrate and ammonia. The observation of deuterium and 018 ”dqfl,ghsl.lflfl Allen Thurman Phillips back- incorporation into threonine during dehydration was taken as evidence that certain of the intermediate steps in the dehydration are reversible. Further evidence for a pathway of elimination involving unsaturated intermediates came from the isolation and identification of c-aminobutyrate from threonine dehydrase reaction mixtures which had been treated with borohydride. This d-aminobutyrate was formed by the reduction of a-iminobutyrate, an intermediate in the spontaneous conversion of a-aminoc rotonate to a-ketobutyrate and ammonia. Threonine dehydrase is also active on serine but this substrate results in a slow inactivation of the enzyme toward both substrates. When dehydrase was inactivated with L-serine-C”, radioactivity was found associated with the protein after removal of the excess radio- activity by gel filtration. Total enzymatic hydrolysis of the labeled protein fraction, followed by paper chromatography, permitted the detection of Chi-lanthionine. This lanthionine would be formed by the addition of a-aminoacrylate, an intermediate in serine dehydration, to the SH group of a cysteinyl residue around the active site of the enzyme. The role of AMP in threonine dehydrase was clarified by the following findings. The Michaelis constant for threonine decreased and the2 maximum velocity increased 6-fold in the presence of AMP at pH 8. Sucrose gradient centrifugations in the presence and absence of AMP revealed threonine dehydrase to have 520 values of 7.6 S and 5. 0 S, respectively. If a spherical protein is assumed, these 520 values correspond to molecular weights of 153,000 and 83, 000 respectively. Intermediate 52.0 values were noted as the AMP concentration was decreased from 3 x 10'3 M to zero. A gradient containing cytidine monophosphate, a less effective activator of this enzyme, showed threonine dehydrase to have an 520 of 6. 9 S. ”it lE'E r Hal”- 7.” «.3 E1344 _ . i. r Allen Thurman Phillips Assays conducted on the 7. 6 S form of threonine dehydrase revealed no requirement for GSH, whereas the 5. 0 S species required both AMP and GSH for activity. These data are consistent with the idea that one or more SH groups on threonine dehydrase are essential for the increases in sedimentation coefficient and activity observed in the presence of AMP. Thus AMP and CSH activate threonine dehydrase by promoting the formation of a highly active species with an 520 of 7. 6 S from a lesser active 5.0 S species. Because no evidence was obtained for AMP participation in the reaction mechanism and because no structural similarity exists between AMP and threonine, this compound must act as an allosteric effector in the regulation of threonine catabolism. THE MECHANISM OF THREONINE DEHYDRASE OF ESCHERICHIA COLI BY Allen Thurman Phillips A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOC TOR OF PHILOSOPHY Depa rtment of Biochemistry 1964 ACKNOWLEDGMENT The author wishes to extend his sincere appreci- ation to Dr. W. A. Wood for his guidance, encouragement and confidence throughout the course of this research. The support of a National Institutes of Health predoctoral fellowship is gratefully acknowledged. ’1‘ z}: >:< >:< >1: 3:: 3:: >3 :3 ::< >§< >:< >:< ii VITA Allen Thurman Phillips was born on October 30, 1938, in Vicksburg, Mississippi. He graduated from Menard Memorial High School in Alexandria, Louisiana, in May, 1956. He received the degree of Bachelor of Science in biochemistry from Louisiana State University in January, 1960, and a Master of Science degree from the same department in August, 1961. He accepted a graduate teaching assistantship in the Department of Biochemistry at Michigan State University, and in September, 1962, received a National Institutes of Health predoctoral fellowship for the remainder of his graduate training. The requirements for the Ph. D. degree will be completed in the fall of 1964. Mr. Phillips is a member of the Division of Biological Chemistry of the American Chemical Society, the American Associ- ation for the Advancement of Science, the American Society of Microbiology, and Sigma Xi. He is married and has one child. iii TO KIKI iv RES TA BLE OF CONTENTS Page INTRODUCTION ............ . . . .......... 1 LITERATURE REVIEW . . . . ................ . 2 Serine and Threonine in Intermediary Metabolism . . . 2 Discovery, Distribution and Characteristics of Serine and Threonine Dehydrases ....... Z Pr0perties of the Purified Enzymes . . ...... 5 Mechanistic Aspects of the Hydroxyamino Acid Dehydrases . ..... . . . . . . . . . . . . 8 Nucleotide Activation in Other Systems and the Allo- steric Effector Concept ............... 12 METHODS AND MATERIALS . ................. 20 Bacteriological . . ..... . . . . .......... 20 Chemical . . . .................. . . . . 20 Materials ...................... 20 Determinations and Procedures ........... Z4 Isotopic .......................... 25 Enzymatic ............. . .......... 27 Assay and Preparation of Threonine Dehydrase . . 27 RESULTS.. ......... .......... 31 Characteristics of Purified Threonine Dehydrase . . . 31 Purification ............ . . . . . . . 31 Effect of pH and Buffers . ............. 32 Stability ....................... 32 Inhibitors . ....... . . . . . . ....... . 32 Activation of Adenosine Monophosphate ...... 33 Substrate Specificity . . . ............. 35 Activation by Pyridoxal Phosphate . . . . . . . . . 36 Mechanistic Studies of Threonine Dehydration ..... 44 Tritium Incorporation from TZO . . . . . . . . . . 44 The Fate of Tritium in the Dehydration of a-tritio threonine . . .................. 47 F). (J) (7 SUM} TABLE OF CONTENTS - Continued Experiments in Deuterium oxide ......... Back-incorporation of Deuterium from D30 and from O18 from HZO18 into Threonine ..... Isolation of an Intermediate in Threonine Dehy- dration . . . . ....... . ........ Identification of a Cysteinyl Residue in Proximity to the Active Site of Threonine Dehydrase ..... The Mechanism of AMP Activation. ......... AMP Activation as a Metabolic Control Mechanism . . DISCUSSION .......................... SUMMARY ........................... REFERENCES . ........................ vi Page 48 50 52. 56 6O 73 77 85 86 TABLE II. III. IV. VI. VII. VIII. IX. XI. XII. XIII. XIV. XV. LIST OF TABLES Purification of threonine dehydrase from E. coli. . . Specificity of activation of threonine dehydrase. . . . Chromatography of commercial AMP and activating ability of the various components ........... Substrate specificity of threonine dehydrase . . Resolution and pyridoxal phosphate reactivation of threonine dehydrase .................. Threonine dehydrase reactions in tritiated water. . . Enzymatic dehydration of a-tritio threonine ..... Threonine dehydrase reactions in deuterium oxide . . Enzymatic back-incorporation into threonine of O18 and deuterium from H2018 and D20 . . ....... . . Oxidation of tritiated a-aminobutyrate ........ Oxidation of H3, CH-a-aminobutyrate ......... Rf values of Cu-labeled products from enzymatic hydrolyzates of Cu-serine-inactivated threonine dehydrase. .. . OOOOOOOOOOOOOOOOOOOO Sedimentation coefficients and molecular weights of threonine dehydrase in the presence and absence of AMP ................. Activities exhibited by the 7. 6 S and 5. 0 S species of threonine dehydrase . . . Adenylate kinase content of extracts of E. coli and g. tetanomorphum ................... vii Page 31 33 34 35 36 47 48 49 51 54 55 58 67 71 76 FIGURE LIST OF FIGURES Spectra of native and reduced threonine dehydrase. Identification of borotritide reduction product from threonine dehydrase. . . . . . . . . . ....... Alternative pathways for threonine dehydration . . . Km and Vmax determinations at pH 7, 8 and 8.6 Sucrose gradient centrifugation of threonine dehydrase Sedimentation rate of threonine dehydrase as a function of AMP concentration ............ Proposed model for activation of threonine dehydrase. The mechanism of enzymatic threonine dehydration. viii Page 38 41 45 62 65 68 74 79 m h ’0 m INTRODUCTION Hydroxyamino acid dehydrases have a widespread occurrence among microbial and mammalian species. Relatively little work has been conducted, however, on the details of catalysis by these enzymes. The reasons for this neglect of such a biologically im- portant set of reactions are threefold: firstly, the mechanistic aspects of the reaction seemed to be straightforward from an organo- chemical point of view; secondly, many investigators chose to study the physiological and genetic factors which influence the formation and activity of these enzymes; and thirdly, the techniques for con- ducting such a detailed study of the reaction mechanism and enzyme chemistry were not available until recent years. The investigations to be discussed were begun with the original intent of elucidating the mechanism of AMP activation of the L-threonine dehydrase from anaerobically grown Escherichia coli. In order to accomplish this, a procedure was developed for the partial purification of this enzyme and studies were initiated on the mechanistic aspects of the reaction; these were aimed at either implicating or eliminating a role for AMP at this level. Later, an investigation of enzyme con- formational changes in the presence of AMP was undertaken. The results obtained from these two experimental approaches have been further utilized in a preliminary way to identify and study the nature of some catalytically important regions of this enzyme. fig“... L- 10 LITERATURE R EVIEW Serine and Threonine in Intermediary Metabolism The synthesis and metabolism of the hydroxyamino acids, L-serine and L-threonine, have been subjects of considerable interest to biochemists over the last twenty years. Because recent and compre- hensive reviews (1-4) on these broad topics are available, this dis- cussion will be mainly concerned with the catabolism of these amino acids, and more specifically, with their catabolism via dehydrative deamination. This particular route of amino acid degradation assumes considerable importance in most biological systems, because serine and threonine are utilized poorly by many amino acid oxidases and amino transferases (transaminases). Catabolism of the hydroxyamino acids is thus dependent largely upon the action of specific dehydrases and aldolases, and to some extent, dehydrogenases. Discovery, Distribution and Characteristics of Serine and Threonine Dehydrases Following Gale and Stephenson's initial discovery of an anaerobic deamination of serine by Escherichia coli (5), Chargaff and Sprinson (6, 7) found that several other species of microorganisms (Proteus OX- 19, Pseudomenas pyocyanae, Clostridium welchii), as well as rat, mouse, and rabbit liver extracts, were capable of producing an anaerobic deamination of serine and/or threonine. They characterized the products as ammonia and pyruvic acid and a-ketobutyric acid, respectively. Substitution of the hydroxylic hydrogen by various alkyl groups (ethyl or methyl) or phosphate or phosphatidic acid prevented .Fu C. .7. :3 n... 1' «\v deamination. On the basis of such evidence, the authors proposed a mechanism for these reactions which involves the loss of water from serine or threonine and results in the formation of an a, B-un- saturated amino acid; this subsequently rearranges to the corresponding imino acid which then hydrolyzes to yield ammonia and an u-keto acid. H H H H l I 9 'H20 9 l 9 R-Cf-CII—COO ———> R-C=C—COO -—>R-C-C|3l—COO Ho NH2 NHz H NH H +HZO ' e NI-l3 + R-C-(l‘T—COO - | H (R = H for serine, and CH3 for threonine) In this way, an enzymatic reaction which is in reality a dehydration results only in a deamination. A more detailed mechanism whereby this dehydration, tautomerism, and hydrolysis could occur was not postulated until sometime later. Extensive study of this reaction was soon undertaken and in the next few years numerous reports appeared of anaerobic deamination of serine and/or threonine in a wide variety of organisms. Although the overall picture was far from clear, it soon became evident that there existed several types of dehydrase which differed markedly in substrate specificity and activator requirements. Gale and Stephenson (5) had noted that the loss of serine dehydrase activity in whole cell preparations of E. co_1i could be prevented by the addition of reduced glutathione (GSH), by 5'-adenylic acid (AMP), or by boiled cell extracts. A brief report by Binkley (8) stated that serine dehydrase in cell-free extracts of E. coli and mouse liver was inactivated by dialysis, but could be restored by certain divalent cations, notably zinc ions. Lichstein and Umbreit (9) reported that the L-serine and the DL-threonine dehydrases of resting suspensions of E. coli and Bacterium cadaveris were stimulated by yeast extract and biotin. Christman and Lichstein (10) found that AMP or biotin could stimulate serine and threonine dehydrases in E. coli or E. cadaveris cells which had been aged in phosphate buffer at pH 4 to inactivate these dehydrases. Certain of these observations were considerably clarified by the work of Wood and Gunsalus (11), who studied a partially purified dehydrase from anaerobically grown E. coli. This investigation indicated a marked stimulation of enzymatic activity by AMP and GSH. Furthermore, these authors showed that a single enzyme was responsible for the dehydration of both serine and threonine, and that serine, although a good substrate, caused a time- dependent total inactivation of the enzyme toward both substrates. Research on the hydroxyamino acid dehydrases was soon ex- tended to studies of the formation of these enzymes, and for E. c_g_l_i_, this eventually resulted in a clarification of the function of such de- hydrases in amino acid metabolism. Pardee and Prestidge (12) found that a specific L-serine dehydrase was induced by serine, glycine, L-leucine and certain related compounds. Furthermore, they reported that L-threonine dehydrase behaved like a constitutive enzyme, but that additional activity could be induced by threonine as well. The inducibility of threonine dehydrase was also reported by Boyd and Lichstein (13). The apparent inconsistency between inducible and constitutive dehydrases was resolved by Umbarger and Brown (14, 15) with the finding of two L-threonine dehydrases in E. coli. One, a constitutive enzyme, was formed only in the absence of isoleucine, was inhibited by isoleucine, and was considered to function in the bio- synthesis of isoleucine (hence its name “biosynthetic” dehydrase). r l b b (n :4. (D e- 54 C) U‘ .4 Pro: The enzyme was absent in isoleucineless mutants which could utilize a—ketobutyrate. They obtained indications of a requirement for pyridoxal phosphate for activity. The second dehydrase was induced only under anaerobic conditions during growth on an amino acid medium. This enzyme, designated the "biodegradative" dehydrase or the Wood and Gunsalus enzyme, required AMP and GSH, and there were indications that pyridoxal phosphate functioned as a coenzyme. It was not inhibited by isoleucine and hence was considered to function in the biodegradation of threonine. Both of these enzymes dehydrated serine and became inactivated during the process. In addition, they obtained sketchy evidence for a separate L-serine dehydrase. Properties of the Purified Enzymes The first attempts to purify threonine dehydrase were those of Wood and Gunsalus (11) using anaerobically grown E. coli as source of enzyme. They were able to show a requirement only for AMP and GSH; biotin, metals and pyridoxal phosphate did not activate. Shortly thereafter, Reissig (16) found that the L-threonine dehydrase from Neurospora required pyridoxal phosphate for activity. Later, Umbarger and Brown (15) utilized hydroxylamine inhibition to resolve their crude preparations of the ”biosynthetic" and "biodegradative" dehydrases and thus were able to show that pyridoxal phosphate would partially reactivate these enzymes. Sayre and Greenberg (17) and then Nishimura and Greenberg (18) partially purified the threonine dehydrase from sheep liver (final specific activity of 13 moles of a-ketobutyric acid formed per mg. of protein per minute). The yield through the seven step procedure was 10%. This enzyme was active on L-threonine, L-ilio threonine and L-serine. Serine inactivated the enzyme, but this inactivation was overcome by the addition of pyridoxal phosphate. The latter authors (18) in“. n) (I) l H Ihe (s dehy Ct. hE‘ e c1L by th speculated that an oxazolidine ring conjugate might be formed by reaction of the aldehydic group of pyridoxal phosphate with the hydroxyl and amino group of serine. No evidence, however, was obtained for this hypothesis. This threonine dehydrase was also stimulated by potassium or ammonium ions but not by any divalent cations. Mercuric ions and p- chloromercuribenzoate (PCMB) were inhibitory. The pH optimum was 8. 2 to 8.6 and the Michaelis constant (Km) for threonine was 8. O x 10‘3 M. Carbonyl reagents such as hydroxylamine and semicarbazide were inhibitory due to their reaction with the aldehydic group of pyridoxal phosphate. These reagents were used in resolving the holoenzyme free of coenzyme and in showing that pyridoxal phos- phate was indeed the coenzyme of threonine dehydrase. Activation of the enzyme by AMP and GSH, first reported by Sayre and Greenberg, was not observed in the further purified preparations of Nishimura and Greenberg. Walker (19) has reported the purification and properties of the L-threonine dehydrase of the rumen microorganism LC, now known as ngtostreptococcus elsdenii. In a lOO-fold purified preparation, the, Km for threonine was 3 x 10'3 M whereas that for serine was 7 x 10"3 M; the optimum pH was 9.5. During purification the ratio of activities on threonine and serine remained constant, suggesting the existence of one enzyme. Freeze-drying in the presence of (NH4)ZSO4 or GSH produced a partial resolution of the enzyme-coenzyme complex. Activity could be restored only by the addition of pyridoxal phosphate. GSH was required for optimal activity. Goldstein e_t a_._1. (20) found that rat liver serine and threonine dehydrase activities were associated with the same protein, as judged by the similar inactivation rates obtained with ultraviolet irradiation, heat denaturation and proteolytic attack. Inducibility of the enzyme by threonine administration in the diet or in isolated perfused livers {—1. (D In (3. could not be shown, but a high protein intake or cortisol administration did result in increased activity. The failure of these workers to show inducibility of threonine dehydrase is in conflict with the findings of Sayre, Jensen and Greenberg (21). Considerable attention has recently been focused on the threonine dehydrase of Clostridium tetanomorphum. Hayaishi e_t all. (22-24) demonstrated a marked dependence on adenosine diphosphate (ADP), especially at low substrate concentrations. Yeast extract was also stimulatory in assays below pH 8. 0. Their preparations were purified some 35-fold, but this specific activity was only equal to 0.1 mole of u-ketobutyric acid produced per minute per mg. protein. Km for threonine was found to be 6 x 10'3 M in the presence of ADP and 3 x 10'2 M in the absence of ADP. Experiments on the role of ADP will be discussed in a later section. Most recently, Holzer and his co-workers (25) have studied a threonine dehydrase from the yeast, Saccharomyces cerevisiae. This enzyme was stimulated 6-fold by NH4+ ions. In addition to its stimulatory effect on activity, ammonia was found to repress the synthe- sis of the enzyme. An earlier publication of Holzer (26) had shown that this dehydrase was concerned with the biosynthesis of isoleucine. The "biosynthetic" threonine dehydrase of E. c__01_i_, first studied by Umbarger (15), has recently been extensively investigated by Changeux (27). Purification of this enzyme from a 10—fold derepressed mutant of E. e_cii K12 has resulted in a ZOO-fold purification over the specific activity of extracts of the wild type. The partially purified enzyme is no longer competitively inhibited by isoleucine as was the crude enzyme. This "desensitization" was brought about by heating in the presence of high concentrations of isoleucine. A change in the kinetic properties was also noted upon purification. Whereas the native dehydrase did not follow simple Michaelis-Menten kinetics, the altered dehydrase had normalized kinetics and a Km of 5 x 10"3 M was calcu- lated for threonine. Desensitization could also be accomplished by PCMB treatment; a protection against PCMB action was noted with L-threonine or L-isoleucine. Further studies by Changeux (28) have indicated that at pH values above 10, the normal enzyme becomes insensitive to isoleucine and exhibits the usual type of kinetics observed for the dependence of velocity upon threonine concentration. In addition, 1"ng threonine, which is not a substrate for the enzyme, inhibits threonine dehydration in a competitive manner even at pH values above 10. The interpre- tations given for the action of isoleucine and _a_l_l_o threonine will be presented in the later section on allosteric effectors. Changeux has also investigated the substrate specificity of the "biosynthetic" threonine dehydrase (29). Besides L-threonine and L—serine, the dehydrase slowly utilizes L-threonine amide and L-threonine methyl ester. An AMP-stimulated threonine dehydrase was reported to be present in Streptomyces rimosus (30). This enzyme, however, was present in aerobic, glucose-grown mycelia and thus may play a dif- ferent role than that assigned the nucleotide-activated enzymes found in anaerobic or facultatively anaerobic bacteria, 1. e. , anaerobic energy generation by substrate phosphorylation. Mechanistic Aspects of the Hydroxyamino Acid Dehydrases Little effort has been made to prove conclusively the scheme for serine and threonine dehydration proposed in 1954 by Metzler, Ikawa and Snell (31). Their original proposal was based on nonenzymatic experiments wherein serine, pyridoxal and certain divalent cations were incubated at elevated temperature. Evidence for Schiff base formation between serine and pyridoxal was found. Therefore, the electron-withdrawing properties of the pyridine ring were invoked to form a long conjugated double bond system which could facilitate the loss of a particular atom or groups of atoms which were in close proximity to the a-carbon atom. The mechanism for serine dehy— dration which would be in accord with this theory is that shown below (4): OH H 0 OH 0 o I I // ‘I // // H-C-CD-C H-C-fi-C H-C=C-C | l \ _ I K \ - I I \ - H ,N o H ,N o H ’,N o /// \\ ’// \ ’l \ H CH + H QCH - I-[I CH l -H \ I -OH 0 ———> o ————-> o | k) CHZOPO3: l K. CHZOPO3: CHZOPo; HBC' H3C' \ H3C —— j N‘) N + + H H +I-IZO H Pyruvic Acid + NH3 + Pyridoxal phosphate Braunstein and his collaborators (32), from consideration of enzymatic reactions, have independently arrived at a scheme which is very similar to that described above. Some related aspects of pyridoxal catalysis should be mentioned at this point. The success of the Schiff base idea in explaining the wide variety of enzymatic pyridoxal-catalyzed reactions and the demonstration that nonenzymatically, pyridoxal does catalyze these reactions via a Schiff base has served to direct attention toward the additional contribution of the enzyme protein to the catalytic action. Metzler (33) has found that Schiff bases between pyridoxal and many amino acids exhibit an absorption peak at 410-415 mp. 10 By analogy, several highly purified pyridoxal enzymes which exhibit absorption peaks in the absence of substrate in the range of 410-430 mp. are considered to contain a Schiff base-like structure formed between pyridoxal and some amino group on the enzyme. Homoserine deaminase (34) and serine transhydroxymethylase (35) exhibit peaks at 427 mu and 415 mu, respectively, and these peaks are unchanged in the pH range of 5 to 10. On the other hand, phosphorylase (36) and glutamate-aspartate aminotransferase (37) exhibit maxima at 415 mu and 430 mu, respectively, but only at pH values below 5. These differences may reflect slight variations in the mode of binding of pyridoxal to the enzyme, but the interpretation still remains that the 415-430 mu absorption corresponds to Schiff base formation. The nature of the protein amino group participating in Schiff base formation has been investigated in the two best characterized pyridoxal enzymes, glycogen phosphorylase (38) and glutamate- aspartate aminotransferase (59). Upon reduction of both enzymes with borohydride, covalently bound pyridoxal phosphate was found associated with the proteins. Hydrolysis of the proteins yielded the fragment, Nb-pyridoxyllysine, indicating that the e-amino group of a lysyl residue was the amino participant. Although two instances do not pro- vide sufficient evidence for a generalization, it is considered likely that many other pyridoxal phosphate-containing enzymes will prove to have pyridoxal phosphate bound to a lysyl residue. The reason for this possibility will now be discussed. It has been generally considered essential that the aldehydic group of pyridoxal phosphate be intact so that Schiff base formation With the amino group of the substrate will occur. This view was favored in spite of the above mentioned instances where the aldehydic group was not free, but rather was bound in an internal Schiff base. It was presumed that either the incoming amino group of the substrate f1— To n1: .: ll displaced the protein amino group or reacted with the small amount of free aldehyde form present, thereby shifting the equilibrium towards the free group and away from the Schiff base. Recent work by Jencks and Cordes (40) has indicated, however, that non- enzymatically, the pyridoxal-valine Schiff base reacts more rapidly with semicarbazide than does free pyridoxal. This and other similar model experiments led them to conclude that the Schiff base is more reactive than free pyridoxal, and thus, in the case of a pyridoxal phosphate enzyme, it would be energetically more favorable for the coenzyme to be bound as a Schiff base prior to reaction. While the requirement for pyridoxal phosphate in hydroxyamino acid dehydration is reasonably well established, no information is available from the literature as to the groups involved in binding of coenzyme to the apo-dehydrase. Only two studies have been conducted and these only indirectly suggest the nature of the groups involved in dehydration. Davis and Metzler (41) have studied the pH dependence of the kinetic parameters of threonine dehydrase from sheep liver. They concluded from the variation in Km with pH that the unprotonated amino group of threonine combines with the enzyme. The lack of any signifi- cant variation in the maximum velocity over the pH range 6 to 10 was interpreted as indicating that no groups with pK values in the range 6 to 10 participate in the catalysis of the rate-limiting step. Another possibility would be that some step subsequent to the dehydration, such as release of bound a-aminocrotonate, is the slowest step in the re- action sequence. Further investigation of the pH dependence of Km revealed that the free enzyme is reversibly converted to an inactive or non-binding form by simultaneous dissociation of three protons at about pH 9. 1 and at 370 C. The authors speculated that this might reflect an alteration in the binding of pyridoxal phosphate. 12 I Russell 2‘. 3:1. (42) have conducted similar studies on the vari- ation of maximum velocity and Michaelis constant with pH for the threonine dehydrase reaction. Their data confirm that the unpro- tonated amino group of threonine is the form bound to the enzyme, but in addition indicate that some basic group on the enzyme ionizes around pH 8. 0. No indication was given as to the nature of this group. After consideration of current knowledge and theories of the mechanism of dehydration, the results of these kinetic experiments remain largely uninterpretable. Nucleotide Activation in Other SJrstems and the Allosteric Effector Concep_t A review by Monod, Changeux and Jacob (43) has thoroughly dis- cussed the subject of allosteric proteins as metabolic regulators. The model proposed by these authors for such proteins features two or more stereospecifically different, non-overlapping receptor sites. The regions of most interest are the "active site, " where substrate is bound, and the "allosteric site, ” where some regulatory metabolite is specifically and reversibly bound. The binding of the "allosteric effector" is thought to produce some sort of reversible change in the protein structure so altering the substrate interactions as to produce either a more or a less favorable situation for catalysis. Furthermore, since the allosteric effector plays no role in the reaction brought about by the protein, this effector need bear no structural relationship to the substrate. This model serves well to explain a variety of activations and inhibitions observed in a number of enzymes where the effects are produced by compounds which bear little if any resemblance to the sub- strate. Included in this category are certain enzymes which have l3 nucleotide activators or inhibitors. Some of the more thoroughly studied examples of such enzymes will now be discussed along with the isoleucine-inhibited threonine dehydrase which has served as a model for allosteric effector studies. Prior to the inception of the present study, three enzymes which exhibited a nucleotide activation had been extensively studied. These enzymes were glycogen phosphorylase from muscle, glutamate dehydrogenase from liver and aspartate transcarbamylase (ATCase) from E. coli. Because it would not be possible to discuss thoroughly the research conducted on the role of nucleotides in the activation or inhibition of these enzymes, only a brief summary of recent findings will be attempted. Two excellent reviews on glycogen phosphorylase have been written by Krebs and Fischer (44,45). In these reviews the phos— phorylase molecule is pictured as a tetramer of molecular weight 500, 000 (referred to as phosphorylase a) held together by at least two disulfide bridges and unknown interactions between two serine phosphate pairs. This form of the enzyme is only slightly activated by AMP. Cleavage of the serine phosphate bonds (achieved by a specific phos- phatase) results in the formation of a dimer of molecular weight 250, 000 (referred to as phosphorylase b). This latter form of phos- phorylase is virtually inactive in the absence of AMP. Work by Kent 9‘. E!" (36) indicated on the basis of sedimentation equilibrium experi- ments that phosphorylase b in the presence of AMP reverted to the tetrameric state with a doubling of the 250, 000 molecular weight. This process was likened to the reformation of active phosphorylase a with the exception that removal of the loosely bound AMP resulted in a dissociation back to the dimeric or phosphorylase b form. In ad- dition to a structural-induced change, AMP decreased the Km for glucose- 1 -phosphate or glycogen. 14 In recent months, however, new evidence concerning the function of AMP has been presented. Monod gt a__1. (43) cite unpublished sucrose gradient experiments by Vagelos in their laboratory which showed that formation of the tetramer in the presence of AMP did not occur. This writer has also learned from Dr. Carl Frieden (personal communication) that the original sedimentation experiments of Kent e_t a}. (36) on the aggregation of phosphorylase b in the presence of AMP were in error and that no such phenomenon is observed in light-scattering experiments. Thus no specific structural change can yet be attributed to AMP for phosphorylase b. Further kinetic effects produced by AMP have been reported by Helmreich and Cori (46) and also by Madsen (47). Both groups have observed that AMP and glucose—l-phosphate have reciprocal abilities to decrease each other's Km values, as do AMP and the other two sub- strates, glycogen and inorganic phosphate. In addition, it has been found (46) that the degree of stimulation of phosphorylase a by AMP is dependent upon substrate concentration; thus it, too, is dependent upon AMP but only at low substrate concentrations. Glutamate dehydrogenase has been extensively studied, expecially with regard to the relationship between enzymatic activity and molecular structure. A recent paper by Frieden (48) summarizes much of what is known in this area. Glutamate dehydrogenase undergoes a reversible, concentration-dependent association to form an enzymatically active tetramer of molecular weight one million from 250, 000 molecular weight sub-units which are also active. A number of nucleotides includ- ing guanosine triphosphate (GTP), adenosine triphosphate (ATP) and adenosine diphosphate (ADP) are activators or inhibitors. Generally GTP and ATP are inhibitors while ADP is an activator. The mechanism whereby the inhibition or activation is accomplished has not been elucidated, but it is clear that the nucleotides have a specific binding 15 site on the enzyme separate from the substrate binding site and that they can influence enzyme activity as well as enzyme structure. Inhibition of the enzyme leads to a configurationally-altered subunit which cannot undergo the concentration-dependent association reaction. Tomkins e_t a_1_1. (49) have contended that the state of aggregation of glutamate dehydrogenase does play an important role in metabolic regulation. They cite work which indicates that the 250, 000 molecular weight species is active in catalyzing an alanine dehydrogenation reaction, whereas the one million weight species is inactive towards alanine. Agents such as GTP, ATP, reduced nicotinamide adenine dinucleotide (NADH) and diethylstilbesterol which favors disaggregation stimulate alanine dehydrogenase activity, and ADP which protects against disaggregation prevents alanine dehydrogenase activation. Even dilution-induced disaggregation favors alanine dehydrogenase activity. Immunological experiments by Talal (49) have detected three antigenic forms of glutamate dehydrogenase, and these forms each have a characteristic substrate specificity; one has glutamate dehydro- genase activity, one has alanine dehydrogenase activity, and a third has both. This and other evidence seem strongly to favor the concept that the type of activity displayed is a direct function of the state of aggre- gation. Aspartate transcarbamylase from E. c_o_li is inhibited by cytidine triphosphate (CTP) in a manner analogous to the isoleucine inhibition exhibited by the ”biosynthetic” threonine dehydrase of E. 30$. CTP is bound to ATCase at a site which is kinetically distinguishable from the binding site for aspartate and carbamyl phosphate. The most recent investigations of Gerhart and Pardee (50) indicate that the binding of CTP at its "regulatory” or allosteric site strengthens subunit inter- actions in the enzyme and distorts the active site, resulting in a decreased substrate affinity. ATP is capable of antagonizing the CTP 16 inhibition (51); this effect is presumed to be achieved by ATP binding at the regulatory site without a concomitant increase in subunit inter- action. Only a brief investigation of induced conformation changes has been made, but simultaneous sedimentation of two enzyme solutions, one containing carbamyl phosphate plus an aspartate analogue, and the second containing only enzyme, indicated that a 5% decrease in sedimentation occurred in the presence of substrates. This finding was considered indicative of a weakening of subunit interactions in the presence of substrate. Although these examples are the most thoroughly studied for such nucleotide allosteric effectors, several other enzymes have recently been investigated for their nucleotide effectors. NAD-linked isocitrate dehydrogenase from yeast has been studied by Atkinson and his co-workers (52). This enzyme is stimulated by AMP, but only at low isocitrate concentrations. The Km for isocitrate was markedly affected by the presence of AMP. A very similar situ- ation exists in the case of liver NAD-isocitrate dehydrogenase studied by Chen e_t a_1. (53). This enzyme was activated by ADP rather than AMP, but the activation was qualitatively similar in other respects. A third NAD-requiring isocitrate dehydrogenase has been investi- gated by Sanwal gt a_l. (54). This enzyme was partially purified from Neurospora crassa and was found to be markedly activated by AMP. At pH 6. 5, however, the AMP stimulation decreased as the substrate concentration increased. Furthermore, at this pH, the enzyme ex- hibited normal kinetic behavior; whereas at higher pH values, abnormal kinetics were observed. These findings were considered indicative that two sites, a substrate site and an activator site, were present, but that the activator site was either modified or destroyed at pH 6. 5. 17 A study most pertinent to the present work is that reported by Hayaishi e_t a;l. (22) on an ADP-activated threonine dehydrase from Clostridium tetanomorphum. The 35-fold purified enzyme showed a marked dependence on ADP at low substrate levels, but the effect decreased as substrate level increased. Only guanosine diphosphate and inosine diphosphate were able to substitute for ADP; AMP was only 5% as active as ADP. Experiments with CH-labeled ADP failed to show any utilization or breakdown of ADP during activation. Whitely and Hayaishi (24) also found that ADP protected threonine dehydrase against inactivation by heat and dilution. In addition, they found that ADP was loosely bound to the enzyme, but was associated with the enzyme fraction after passage through a Sephadex G-50 column. A control mechanism has been proposed by Tokushige e_t a_._l. (23) for energy production in E. tetanomorphum in which threonine metabol- ism is regulated by the ADP concentration. This postulate is based on experiments which indicate that CH-threonine is converted to CH-propionate via a-ketobutyrate and propionyl phosphate. Their scheme is as follows: NH, C0; + H2 L-threonine L d-ketobutyrate —4 Propionyl phosphate Threonine dehydrase€§ § P It. Stimulatory P ADP ' l s1gna Acyl kinase Energy ATP utilizing reactions PrOpionate 18 Thus they consider that the degradation of threonine in C. tetanomorphum is inversely linked to ATP synthesis; that is, that as ATP is utilized, the ADP level increases to activate threonine dehydrase so as to provide the means for more ATP synthesis. No discussion of allosteric effectors and their protein receptors would be complete without mention of the work of Changeux (27-29) on the allosteric interactions in the biosynthetic threonine dehydrase. As discussed earlier, considerable evidence has been accumulated to show that isoleucine can bind to at least one site on this enzyme which is kinetically and physically distinct from the substrate binding site. The findings that isoleucine inhibition can be destroyed by various treatments without corresponding destruction of enzymic activity, and that al_I9_ threonine, a substrate analogue, can inhibit enzyme activity under conditions where isoleucine sensitivity is present or absent are considered to support strongly the hypothesis of separate regulator and substrate sites (29). Still further evidence for the presence of a ”regulatory” site in addition to the active site has been described by Monod e_t a_._l. (43) and Freundlich and Umbarger (55). Valine, presumably an isoleucine analog, activates rather than inhibits the reaction by increasing the affinity of the enzyme for threonine. When valine is incorporated into dehydrase reaction mixtures containing isoleucine, the inhibitory effect of isoleucine is prevented. A similar behavior is observed with nor- leucine or methionine in place of valine. Because this enzyme has resisted purification, the more common techniques for observing a conformational alteration have been unsuit- able. Sucrose gradient centrifugation experiments (29) have indicated that no detectable change in sedimentation velocity occurs in the presence of isoleucine or a_llc_1 threonine. Urea inactivation, however, has produced a kinetically detectable disaggregation of the native enzyme and this . ‘9 8-; m In 19 effect is reversed by isoleucine and stimulated by allo threonine-- a situation somewhat similar to that found in ATCase, where subunit interactions are weakened in the presence of substrate. METHODS AND MATERIALS Bacteriological Escherichia coli, Crookes strain, ATCC 8739, was maintained in stationary liquid culture at 370C. and was transferred biweekly. The culture medium consisted of 1% tryptone, 0. 5% yeast extract and 0.5% KZHPO4. For the isolation of threonine dehydrase, E. (£1 was grown in 18 liters of the above medium dispensed in 20 liter glass carboys, or in the maximum volume of a 10 gallon stainless steel fermenter (New Brunswick Scientific Company). Growth from a 10% inoculurn proceeded for 20-24 hours at 370C. without agitation or aeration. Clostridium tetanomorphum, ATCC 3606, was maintained anaerobically at 30°C. on a medium containing 500 g. of beef liver, 10 g. of peptone, l g. of KZHPO4, 5 g. of yeast extract and 1 1. of distilled water; the final pH of the medium was adjusted to 7. 0. Larger quantities of cells were obtained from a glutamate medium as described by Tokushige e_t a_l. (23). This procedure also includes the method for preparing cell-free extracts. Chemical Materials CH-L-threonine, sodium glutathione, 5'-AMP, 2'-AMP and 3'-AMP were purchased from Schwarz Bioresearch, Incorporated. Polyadenylic acid was obtained from the Miles Chemical Company. ADP, ATP, cytidine monophosphate (CMP), inosine mon0phosphate 20 1'9 (“(1 21 (IMP), NAD, NADH, and NADPH were products of Pabst Laboratories. All amino acids, protamine sulfate, vitamin B6 derivatives, and other nucleotides or nucleotide analogs were from the California Corporation for Biochemical Research. Bovine serunn albumin and u-ketobutyric acid were products of the Sigma Chemical Company. Enzymes used as excess reagents were obtained as follows: crystalline lactic dehydrogenase, D-amino acid oxidase and catalase were from the Worthington Biochemical Corporation; fungal protease Type VI (Pronase), L-amino acid oxidase, Type III, alkaline phospha- tase, Type I, of calf mucosa, and yeast hexokinase were purchased from the Sigma Chemical Company; triose phosphate isomerase-c- glycerophosphate dehydrogenase (mixture of crystals) was from C. F. Boehringer and Sons; L. Light and Company supplied the glucose-6- phosphate dehydrogenase. CM-L-serine, NaBH4-H3 and a Cu-benzoic acid standard were purchased from New England Nuclear Corporation. Tritiated water, specific activity = 1 curie/g. , was purchased from Volk Radiochemical Company; a tritiated water standard was obtained from the Packard Instrument Company. Deuterium oxide, 99. 7%, was purchased from the Liquid Carbonic Division of the General Dynamics Corporation. Oxygen-18 enriched water (5. 58 atom % 018) was purchased from the Yeda Research and Development Company, Weizmann Institute, Rehovoth, Israel. All ion-exchange resins were obtained from the California Corporation for Biochemical Research. Sephadex was purchased from Pharmacia, Limited. Nb-pyridoxyllysine was first prepared by adapting the procedure for the synthesis of Né-B-glyceryllysine (56):25 mg. of pyridoxamine dihydrochloride and 165 mg. of 5-6 -bromobutylhydantoin (57) were dissolved in 10 ml. of 70% ethanol, adjusted to pH 9. 5 with saturated 22 NaOH, and refluxed for 40 hours. The pH was checked at 12-hour intervals and solid NazCO3 added to restore the pH to approximately 9 when necessary. After 40 hours, the solvent was removed in vacuo, and the residue resuspended in 1 ml. of cold water. The precipitate which did not go into solution (unreacted bromobutylhydantoin) was removed by centrifugation and discarded. The supernatant was then passed through a Dowex-50 (H+) column, and the column was washed with water until no material absorbing at 290 mg was eluted. This eluate was concentrated i_n vacuo, made 2 M in NaOH and hydrolyzed at 800 C. for 8 hours under nitrogen in a Thunberg tube. The solution was then neutralized with 6 M HCl and passed through a column of AG 11A8 ion-retardant resin to remove NaCl. The ninhydrin-positive fraction from this column was concentrated in vacuo and characterized. This material was homogeneous on cellulose acetate electro- phoresis at two pH values, but paper chromatography of larger amounts (n-butanol-acetic acid-water, 60:15:25) revealed two minor components in addition to the major spot. This preparation of Nb-pyridoxyllysine was strongly fluorescent under ultraviolet light and had maximum absorption at 325 mu at pH 7. 0 (45). An alternative and more convenient procedure for the prepara- tion of Nb-pyridoxyl lysine was that recently described by Krebs and Fischer (45). This procedure involved the NaBH4 reduction of a Schiff base formed between pyridoxal and N-u-carbobenzoxy-L-lysine (58). Removal of the carbobenzoxy group by treatment with 2 N HBr in glacial acetic acid at room temperature for 2 hours and subsequent evaporation i3 w yielded Nb-pyridoxyllysine. This material had the same electrophoretic and spectral characteristics as that described above, but paper chromatography revealed a small lysine contami- nation. 23 Lanthionine sulfoxide and sulfone were prepared as described by Zahn and Osterloh (59). Over-oxidation to unknown products seemed to cause some difficulty, but this was largely overcome by careful temperature control. For the preparation of N6-(DL-2-amino-2-carboxyethyl)-L-lysine, commonly referred to as lysalanine, N-acetyldehydroalanine was prepared (60, 61) and 60 mg. were reacted with excess cupric complex of L-lysine at pH 10, 500 C. , and under flowing N2 gas. After 12 hours, the solution was acidified with concentrated HC1 and H25 bubbled through to remove Cu++. After removal of CuS by centrifugation, the solution was evaporated to 0. 1 volume and the precipitates of any unreacted N-acetyldehydroalanine and L-lysine were centrifuged off and discarded. The final supernatant contained N-acetyl lysalanine and some lysine. Concentrated HCl was added to a final concentration of 2 N_ and the solu- tion was refluxed for 2 hours to remove the N-acetyl group. This solu- tion was evaporated to dryness i_n w and redissolved in water. Lysine-copper complex was replaced in later syntheses by N-a-carbobenzoxy-L-lysine (58). Removal of the carbobenzoxy group was achieved by treating with 2 N HBr in glacial acetic acid at room temperature for 2 hours. Alpha-tritio threonine was prepared by performing the racemi- zation of DL'EEE threonine as described by Akabori e_t a_._l. (62), in the presence of TZO. The preparation was carried out on a 2 mmole scale in the presence of tritiated water of specific activity of 0. 6 mcurie per mmole of H20. The thrice-recrystallized mixture of threonine isomers was obtained in 23% yield and had a specific activity of approximately 0. 06 mcurie per mmole. DL-threonine was separated from DL-fl threonine by preparative paper chromatography on Whatman 3 MM paper in the solvent system 2-butanone, n-butanol, water and cyclohexylamine (5:522. 5:1), as recommended by Flavin and Slaughter (63). Strip Chan: Cir15in 24 Determinations and Procedures Threonine in cultureimedia was determined by the colorimetric procedure of Desnuelle and Naudet (64). For a more sensitive threonine determination, the procedure of Flavin and Slaughter (65) was used. This involved oxidation of threonine to acetaldehyde with periodic ac id, followed by Spectrophotometric determination of acetaldehyde as NADH oxidation in the presence of excess yeast alcohol dehydrogenase. All threonine isomers are measured equally. A specific method for L-threonine was developed utilizing purified threonine dehydrase which was shown to have high specificity (Table IV). This assay was similar to that used for determination of threonine dehydrase activity, but employed excess dehydrase. Thus the amount of NADH oxidized was a measure of the amount of L-threonine present (see below). Serine was estimated colorimetrically by the method of Lambert and Neish (66). Alpha-keto acid determinations were performed either by the semicarbazide method of Macgee and Doudoroff (67) or by a spectro- photometric-enzymatic method similar to that used for a-ketobutyrate determination in threonine dehydrase assays. When desired, the separation of a-ketobutyrate from reaction mixtures was achieved by column chromatography on Dowex-l-chloride. Following a brief wash- ing with water, the d-ketobutyrate was eluted with a linear gradient of 0 to 0.15 M HCl. Threonine was isolated from reaction mixtures, purified and crystallized as described by Flavin and Kono (68). Electr0phoresis experiments were performed on cellulose acetate strips (2. 2 x 12 cm. ). The equipment included an electrophoresis chamber (Colab) and a power supply (Heath Co. ). Electrophoresis, drying and staining were carried out as described by Scherr (69). .__a a 25 Additional experiments were conducted with a Veronal buffer, pH 7. 5, in place of the acetic acid-formic acid buffer, pH 2. 0, of Scherr. Spectrophotometric assays were conducted in silica cuvettes of 1. 0 cm. light path. Absorbance changes were measured in a Gilford Multiple Sample Absorbance Recorder attached to a Beckrnan DU monochromator. Linear sucrose gradients (5 - 20% w/w) were prepared with a gradient maker constructed from Lucite as described by Martin and Ames (70). The total volume of each gradient was 4. 3 m1. , and 0.10 ml. of enzyme sample was usually layered carefully on each gradient contained in 0. 5 x 2 inch lusteroid centrifuge tubes. Centrifugation was carried out in a pre-cooled Spinco SW-39 rotor in the Spinco Model L ultracentrifuge. After a slow manual acceleration, the speed was set at 35, 000 rpm and centrifugation continued for 15-17 hours. The sample temperature was maintained between 10 and 150 C. by adjust- ing the coolant temperature to 130 F. The rotor was allowed to coast to a stop and the gradients removed; 30-35 fractions of 10 drops each were collected by puncturing the centrifuge tubes with a 22-gauge hypodermic needle without Luer fitting mounted on a ring stand sup- port (70). The fractions were stored in ice until all the assays were completed. Isotopic All quantitative radioisotope measurements, either tritium or carbon-l4, were performed in a Packard Tri-Carb liquid scintillometer. Three scintillation systems were used, depending largely on the isotope to be measured and the volume to be sampled. The system of Kinard (71) was used primarily for counting C14 in aqueous samples of up to one ml.; the system of Werbin (72) was used for counting tritium and C” when the volume of aqueous samples was 1 to 3 ml. The latter amOL DOIIII desir Radic stand by Rit on the in the nation itin a emp103 then a enrichE q ds “yexw ment. 26 system was also employed when radioactivity on cellulose acetate strips was to be determined, because the strips were soluble in this solvent system and geometric considerations were eliminated; the same system was used for counting tritium on paper strips because of its high efficiency. However, reproducible positioning of the paper was essential in this case. For routine counting of C14 on paper strips, the system and procedure of Wang and Jones (73) was used because of its simplicity and low cost. Efficiencies for C14 and H3 were determined by adding known amounts of radioactive standards to already counted samples in bottles and then recounting. When a preliminary location of C14 on paper chromatograms was desirable, a Nuclear Chicago thin-window gas flow counter1 was used. Radioautography was performed on Kodak No-Screen X-ray film in standard film holders. Combustion of samples and analyses for deuterium content by the falling drop method were performed by Mr. Josef Nemeth, 303 W. Washington Street, Urbana, Illinois. Samples for 018 analysis were combusted with HgClz as described by Rittenberg and Ponticorvo (74). The CO; obtained was then analyzed on the Michigan State University mass spectrometer. The 018 content in the H3018 used in these experiments was determined at the termi- nation of each incubation by removing the water i_n w and trapping it in a dry ice-acetone bath. The method of Boyer e_t all. (75) was employed for the conversion of 018 in H2018 to C0218, and the CO; was then analyzed as described above. Several standard samples of 018- enriched KHZPO4 were generously supplied by Dr. Clarence Suelter, as were instructions and assistance in performing these combustions. Om 1Dr. N. E. Tolbert graciously permitted the use of this instru- ment. 27 Enzymatic A number of commercially available enzymes were employed as standards in this study. Yeast and horse liver alcohol dehydrogenases were obtained as ly0philized powders from Worthington Biochemical Corporation and were assayed as described by Vallee and Hoch (76), but with a 15-fold reduction in volume. Malate dehydrogenase and fructose diphosphate aldolase from muscle were obtained from Worth- ington Biochemical Corporation as suspensions in (NH4)ZSO4. Malate dehydrogenase was assayed by the method of Mehler e_t a_._l. (77); aldolase was assayed by the triose phosphate isomerase-a-glycerophosphate dehydrogenase method of Baranowski and Neiderland (78). A 15-fold reduction in volume was used in both malate dehydrogenase and aldolase assays. 2-Keto-3-deoxy-6-phosphogluconic (KDPG) aldolase was supplied as a crystalline suspension in (NHAZSO4 by Dr. H. P. Meloche and Mr. J. M. Ingram of this laboratory (79). Assays were conducted as described by Kovachevich and Wood (80). Adenylate kinase was assayed by the rate of formation of reduced nicotinamide adenine dinucleotide phosphate (NADPH) in a coupled assay consisting of ADP, glucose, excess hexokinase from yeast and glucose- 6-phosphate dehydrogenase (81). This assay was also scaled down 15-fold from the published procedure. Assay and Preparation of Threonine Dehydrase A coupled assay based on the oxidation of NADH was used for estimation of threonine dehydrase activity: Threonine Dehyd ra s e Threonine > a-ketobutyrate + NH3 excess lactic -k t b t t + NAD fx— - o. e o u yra e H dehydrogenase o. hydroxybutyrate + NAD 28 The assay mixture for threonine dehydrase which was used for the major part of this work contained the following: 1. 0 prnole of AMP, l. 0 mole of GSH, 0. 08 mole of NADH, 0.025 mg. of lactic dehydro- genase (rabbit muscle, Worthington Biochemical Corporation), 15 moles of potassium phosphate buffer, pH 8. 0, and a suitable dilution of the dehydrase. After a 5 minute pre-incubation of all components, 4. 0 moles of L-threonine, pH 8. 0, were added to initiate the re- action. All assays were run in a total reaction volume of 0. 20 ml. Absorbance changes were measured at 340 mp; the reaction tempera- ture was usually 250 C. A unit of threonine dehydrase is defined as that amount of enzyme which will produce an absorbance change of l. 0 per minute under the prescribed conditions. Based on a molar extinction coefficient of 6. 22 x 103 for NADH, this unit equals 0. 032 moles of threonine dehydrated per minute. Specific activity is defined as units per mg. of protein (protein determined by the method of Lowry e_t a}. (82) with crystalline bovine serum albumin as standard). The assay was shown to be highly reliable and linear with enzyme content for rates up to 0. 2 OD change per minute; dilutions for assay were usually made in 0. l Mphosphate buffer, pH 8. 0, containing bovine serum albumin, AMP, and GSH at 1 mg. per ml. concentration each. For purification of threonine dehydrase, anaerobically grown E. 99}; were collected by continuous centrifugation, washed once with 0. 1 M KCl, and resuspended in an equal weight of 0. 1 Mphosphate buffer, pH 8.0, containing 3 x 10-3 1:4 each of AMP and GSH. The sus- pension was then subjected to sonic oscillation for 10 minutes at 10 Kc and 200 watts. The crude extract thus obtained was centrifuged at 30, 000 x g and 30 C. for 30 minutes. Purification of the enzyme was conducted in the following manner. All steps were carried out at 0 to 30 in the presence of 3 x 10"3 M 2-mercaptoethanol. 29 Step 1. The protein content of the crude extract was diluted to 12 mg. per ml. with distilled water and (NH4)ZSO4 was added to 0.1 M concentration. One-fifth volume of 2% protamine sulfate was added with stirring, the precipitate collected by centrifugation and discarded. M. To the supernatent from Step 1, (NH4)ZSO4 was added to 1. 2 M, and any precipitate which formed was collected and discarded. A further addition of (NH4)ZSO4 to l. 8 M precipitated the enzyme. This precipitate was redissolved in distilled water and the salt content was adjusted to 0. 04 M (as determined from conductivity measurements) by further dilution with distilled water. M. Ca3(PO4)Z gel was added to the diluted solution from Step 2 until nearly all of the enzyme was adsorbed by the gel; this usually required 1 g. Ca3(PO4)7_ per 100, 000 units of enzyme. The gel containing the enzyme was washed once with 0. 04 M (NH4)ZSO4 and then the enzyme was eluted from the gel by 2 washings with 0. 10 M P04, pH 8. 0. Approximately 50 ml. of eluant were required to elute 100, 000 units . Step 4. The eluate from Ca3(PO4)2 was again treated with (NH4)ZSO4 and the fraction precipitating between 1. 2 and 1.8 M was collected, dissolved in water, and the solution diluted with HZO to a salt content of 0. 02 M. _S_t_ep£. Alumina C7 gel was added (350 mg. per 100, 000 units). When the major portion of the enzyme was absorbed, the gel was collected by centrifugation and the enzyme immediately eluted with 3 portions of 0.05 M phosphate, pH 7. 0 (120 ml. per 100, 000 units). When the majority of the enzyme was not adsorbed, the gel was col- lected and discarded. Only one batch of alumina C7 gel tested failed to adsorb the enzyme, but in that instance, negative adsorption produced a purification comparable to that obtained with the positive adsorption. 30 55316. Protamine sulfate (2%) was added dropwise with stirring to the enzyme solution from Step 5 (1.25 ml. per 10 ml. of the enzyme solution). The sticky yellow precipitate which formed within 30 min. was collected by centrifugation and transferred to a beaker containing 25 ml. of 0. 25 M phosphate buffer, pH 7. 0. This suspension was slowly stirred at 00 C. for 12 hours. Step 7. The slightly turbid solution from Step 6 was centrifuged and the clear supernatent was treated with (NH4)ZSO4. The fraction precipitating between 1. 2 and 1. 8 M was dissolved in 0. 05 M phosphate buffer, pH 7.0, containing 3 x 10‘3 M each of AMP and GSH. This solution was either used in the experiments to be described, or fractionated fu rthe r . Step 8. (Optional). The enzyme solution from the preceding step was diluted with water to a salt concentration of 0. 05 M. This solution was transferred to the t0p of an ECTEOLA-cellulose column which had been previously equilibrated with 0. 05 M phosphate buffer, pH 7. 0. The column was washed with 20 ml. of the equilibration buffer, and then gradient elution was begun. All buffers contained 1 x 10-3 M AMP and 2-mercaptoethanol. A linear gradient (0 to 0. 5 M phosphate, pH 7. 0) was used as eluant. Threonine dehydrase eluted at a molarity of approximately 0. 25 M phosphate. Threonine dehydrase purified through Step 7 was quite stable when rapidly frozen in a dry ice-acetone bath and stored at -150 C. Activity remained essentially unchanged for at least 3 months, and over 50% of the original activity was present after 1 year. However, freez- ing of the enzyme in the presence of mercaptoethanol consistently destroyed the activity. Thus GSH was used in any step just prior to freezing. RESULTS Characteristics of Purified Threonine Dehydrase Purification A summary of the purification of threonine dehydrase is pre- sented in Table I. A 54-fold purification was obtained with a 51% recovery of activity. No reliable values for the purification achieved by ECTEOLA-cellulose chromatography were obtained because the Lowry method for protein determination was unsuitable in the presence of high concentrations of AMP and 2-mercaptoethanol; chromatography in their absence resulted in considerable loss of activity. The data in this table were obtained from the fractionation of a crude extract of cells grown in 40 liters of medium, but essentially similar results were obtained on 2-fold larger or smaller scales. Table I. Purification of threonine dehydrase from E. coli. ‘— _ Total Specific Recovery Step Units Activity (Percent) Fold A Crude extract 67, 000 35 100 -- Protamine treatment 85, 300 -- 127 -- (NH4)ZSO4 ppt. 1. 2-1. 8 M 60, 000 62 90 1. 8 Ca3(PO4)7_ gel eluate 45, 000 188 68 5. 5 (NH4)ZSO4 ppt. 1.2-1.8 M 48, 500 346 72 10 Alumina C7 gel eluate 39, 600 1400 57 41 Protamine precipitate 40, 000 -- 60 -- (NH4)ZSO4 ppt. 1. 2-1. 8 M 34, 000 1900 51 54 ECTEOLA chromatography 20, 000 -- 33 -- 31 32 Effect of pH and Buffers Threonine dehydrase exhibits a fairly sharp pH optimum near pH 8. 0 in 0. 06 Mphosphate buffer. Of a variety of buffers tested, only cacodylate afforded more activity than phosphate. However, since cacodylate has a low buffering capacity at pH 8. 0, phosphate was used for almost all of the tests. The enzyme stability optimum in phosphate buffer is near pH 7. 0. Stability Stability of the enzyme in dilute solutions was markedly enhanced by AMP, GSH or 2-mercaptoethanol, and bovine serum albumin. Heat treatment, charcoal, organic solvents and pH values below 5 all had adverse effects on activity. The insoluble addition product of protamine and threonine dehydrase was very stable at pH 7. 0. Dehydrase activity was reasonably stable upon dialysis when 2-mercaptoethanol was present in all solutions. Inhibitors KCN at 5 x 10‘3 M concentration inhibited threonine dehydrase activity 90%. Hydroxylamine, pH 8, 5 x 10" Mconcentration, destroyed 95% of the dehydrase activity as measured in an enzymatic assay. These compounds were inhibitory only toward the dehydrase and, at the concentrations employed, did not interfere with the measure- ment of very small amounts of a-ketobutyrate by the lactate dehydro- genase assay system. This conclusion has been verified by comparison of the rates of ammonia release with and without inhibitor. Parachloro- mercuribenzoate (PCMB) inhibited threonine dehydrase 90% at 2. 5 x 10"5 M concentration when tested in the absence of added GSH and compared to a similar assay containing no PCMB. 33 Activity was neither inhibited nor stimulated by the following: + ++ ++ , Mg , orMn , 5x10"M; _ - + Fe++, 1x103M;F , 1x10‘zM;Ca arsenite, 1 x 10"4 M; isoleucine, 1 x 10'2 M; and propionate, 1 x 10'2 M. Activation by Adenosine monophosphate A comparison of the ability of various nucleotides and nucleotide analogs to activate threonine dehydrase is shown in Table II. Table II. Specificity of activation of threonine dehydrase. (Assays were performed as described in the text. The final concen- tration of activator was 5 x 10'3 M except where noted.) #— * Compound Tested Percent of Activity with AMP 5'-AMP 100 CMP 48 Deoxy AMP 38 2'-AMP or 3'-AMP 27 ADP or ATP 20 GMP l7 UMP or IMP 4 3', 5'-cyclic AMP 0 Adenosine monosulfate 0 Adenosine monoac etate 0 Polyadenylic acid (2 mg/ml) 0 The values for ADP and ATP have been roughly corrected for the amount of AMP contamination as determined by chromatography (83). Of the nucleotides tested, only CMP and deoxy AMP exhibited appreciable activating ability. Lesser activation was displayed by 2'-AMP, 3'-AMP, ADP, ATP and GMP. It is worthy of mention that, unlike the situation with phosphorylase b (44), IMP has virtually no activating ability. It is evident from the table that certain groups are required for activity, 34 notably an amino group on a purine or pyrimidine ring and a phosphoric acid residue esterified at position 5' on the nucleoside moiety. Although a number of nucleotides could activate threonine dehydrase to varying extents, it was nevertheless considered desirable to examine commercial AMP for possible contamination by a highly active trace component. For this, 50 mg. of AMP were neutralized with KOH and streaked on Whatman 3 MM paper. Development of the chromatogram as described by Cormier (83) permitted the detection of two trace components in addition to AMP. Each of the three compounds was eluted and concentrated i_n v_a_c_1_1_c_> prior to tests for activating ability. The re- sults, presented in Table III, indicate that the AMP fraction was effective in activating threonine dehydrase and that the trace components were inactive at approximately 1/20 the AMP concentration (a concentration many fold greater than that present in AMP preparations). The exact identity of the trace components was not determined, but from other work (83), it is likely that one of the unknowns was 3', 5'-diphospho- adenosine. Table III. Chromatography of commercial AMP and activating ability of the various components. (The AMP examined was Schwarz Bioresearch Lot No. 6025. Chromatography was carried out as described in the text; concentrations are based on a molar extinction coefficient of 15.4 x 103 at 259 mu. Each assay contained the same amount of dehydrase and rates were measured after 5 minutes preincubation.) Component Concentration Activity (A A/min.) .-. -5 - RAMP 0. 30 5 x 10 M 0.002 - -5 RAMP— 0.41 6x10 M 0.002 = -3 RAMP 1.0 1x10 M 0.015 AMP 1 x 10-3 M 0.016 (commercial) None --- 0.002 35 Substrate Specificity The ability of threonine dehydrase to dehydrate other hydroxy- amino acids and analogs is shown in Table IV. In agreement with the specificity of the biosynthetic threonine dehydrase of E. 99E (29), the biodegradative dehydrase was found to be active only on L-serine and L-threonine. A trace of activity was observed on DL-_a_lio threonine, but this could result from the presence of traces of L-threonine. In contrast, however, the threonine dehydrase from sheep liver is appreciably active on L-illp threonine (18). The Km for threonine and serine were identical, 5 x 10"3 M, and very nearly identical maximum velocities were obtained. Since serine inactivates the dehydrase, it is important to consider only initial rates; otherwise it would be concluded (as has been done in the past) that serine is a poorer substrate (11,15,18). Table IV. Substrate specificity of threonine dehydrase. (Excess crystalline malate dehydrogenase replaced lactate dehydro- genase in the usual assay when the (S-hydroxyaspartate isomers were tested. All other compounds were evaluated in the normal assay system.) Percent of Compound Concentration Threonine Activity L-threonine 1 x 10-?- M 100 D-threonine 1 x 10-2 M 0 DL-3._11_o threonine 2 x 10-3 171 3 L-serine 1 x 10-2 171 100 DL-homo serinea 2 x 10"2 M 0 DL-erythro-fi-hydroxy- — aspartate 2 x 10'2 M 0 DL-threo-[S-hydroxy- aspartate 2 x 10-2 M 0 (DL + MM) - cystathionine 2 x 10"2 171 0 L-penicillamine 1 x 10-2 M 0 DL-homocysteine 2 x 10'2 M 0 alnhibitory to normal threonine dehydration. 36 Activation by Pyridoxal Phosphate A requirement for pyridoxal phosphate in threonine dehydration has been previously indicated (see Literature Review). The ability of pyridoxal phosphate to reactivate hydroxylamine-inactivated threonine dehydrase but not serine-inactivated enzyme is seen in Table V. Table V. Resolution and pyridoxal phosphate reactivation of threonine dehydrase. Activity Activity after Incubation Stimulation Inhibitor, after Without With by . . . . . 3- . Concentration d1aly31s pyr1doxal p pyr1doxal p Hmdoxal p NHZOH b: Percent Perc ent Pe rcent Fold 1 x 10-4 M 77 81 72 < 1 1x10-3 M 35 15 52. 3.5 5 x 10'3 M 5 4 27 6. 7 1x10'ZM 0.1 0.1 6.3 63 None 100d 100d 95 < 1 Serine c 5 x 10'2 M 53 35 31 < 1 1 x 10'1 M 37 25 25 1 None 100(1 100d 100 1 :The pyridoxal phosphate concentration was equal to 2 x 10" M. The incubation time was 2 hours. The incubation time was 4 hours. Arbitrarily set at 100%. an These experiments were performed in individual tubes, each containing 1 mole of AMP, 1 mole of NaGSH, 10 moles of potassium phosphate buffer, pH 8. 0, approximately 50 units of dehydrase, and neutralized NHZOH or L-serine at varying concentrations. The final volume was 0. 2 ml. After 2 or 4 hours of incubation at 150 C. , the dehydrase 37 activity was determined on aliquots, and then the contents of each tube were dialyzed at O0 C. for 2 hours against 2 liters of 0. 05 M phosphate buffer, pH 7. 0, containing AMP and 2-mercaptoethanol, each 1 x 10’3 M. Aliquots of the dialyzed solutions were assayed, and then incubated with 2 x 10‘4 Mpyridoxal phosphate, and reassayed. The extent of reactivation was arbitrarily measured after 1 hour incubation at 150 C. The results indicate that reactivation with pyridoxal phosphate was appreciable but by no means complete. Similar partial reactivation was observed by Umbarger (15). Attempts at complete resolution and reactivation at 370 C. as described by Nishimura and Greenberg (18) for the sheep liver threonine dehydrase were consistently unsuccessful, presumably because of enzyme instability at 370 C. The difference between the inactivating ability of NHZOH in an enzymatic assay (95% inactivation at 5 x 10'4 M reported above) and in these mixtures (95% inactivation at 5 x 10-3 MNHZOH) is ascribed to the differences in enzyme concentration under these two conditions. Because the foregoing results were considered inconclusive proof of pyridoxal phosphate participation, additional experiments were per- formed. Fifty-fold purified threonine dehydrase was observed to have a pronounced yellow color at neutral pH. A limited spectrum analysis revealed an absorption maximum at 402 mu (Figure 1, upper curve). This absorption maximum falls between that normally seen for free pyridoxal phosphate (388 mg) and pyridoxal phosphate bound to protein as a hydrogen-bonded Schiff base (415-430 ml.;). The free aldehyde form of pyridoxal phosphate bound to a protein has a maximum around 360 mp. (39). Threonine dehydrase, (4.1 mg. of protein, specific activity of l, 900 units per mg. ), was treated at pH 7 and 30 C. with a total of 10 pxnoles of NaBHyH3 (150 uc per mole), followed by 8 moles of acetic acid. These were added in 4 equal portions at 3 minute intervals, 38 .HE mm .0 Mo @8305. m 5 Uocwmucoo 530nm mo .wE N :v do 06mg one? 930on ”Geououm mo .wE you miss oomV .H mo eff/Sum 330on .m was poms winds 23. .acoguummEoo #533 of. a: flood was? og>uoo «mooxo mucocomgoo 2w 9:53:00 COESHOm < .HE mm .0 mo v§Ho> .m 4:3 mvfio>soouofi5 .58 OH x m .H mo omd o5. poufiauom £033 839$ wcwxmmg Ewen .m gore, pommflswo Hopoaouonmonooomm wcfipuooon ma Hopoz >HmU m :0 mode who? mucogoHSmmoE Hahn—ovum .ommup>£op 05:00.2: poodpou «muopfinvsfkouoo. pom. Swims mo ouuoomm .H “:5th 39 is. 0.3V ONV 00¢ 0mm 090 0.3.. d - A q q q omoaomm LsIm m2>sz\! m>_.—.1mo mZ_ZOmmI.r omoaomm oz< m>_._. com 5%? m .N. In Hm Ava: USOm >335 omdub>£op poodpouuogfihouon signing: rooms .393 mo muouflou 38 092:. pognofiom one? 3:05:0de omosh mo :oSmowfiU ofisfiooaofim 50.3 oocflouno podoonm vigomofipmu pom $55.93 oopmnmv ocwmefifnxopflucwmuoz venomouan>admoflofic>m mo mfimononmonuoofio oumuoom omoHSZoD .omme>£op ocficoonzo 80.3 “360.2“ coflodpou ooflfloouon mo nofimoflflaog .N ondmflh Fzmthmmh mmdbqrdmozd awkud mNIo TV L00m SEQ 100w 42 mNIQ TV 00m sic R§ mmImo mz_zomm_.:. 20mm Hoaoomd ZO_FUDowm mo_._._m._.0mom do ZO_._.£ud on» ”093 U33 Ga :39? who ucv>H0m 80.3 mcwmflum mnemoupewm £3839»on ocflcoonfiu HOW m>m3£umm o>5mcnofi< .m oudmwh 46 ZO .Pdmaquwo m a Q n :2 .08 n o 1w- :o I ON: Q .08 nwawunzo :2 I mmumzI 47 Table VI. Threonine dehydrase reactions in tritiated water. (Each incubation mixture contained 40 moles of L-threonine, 5 prnoles of AMP, 5 prnoles of NaGSH, 40 moles of phosphate buffer, pH 8. 0, and 40 units of dehydrase in a final volume of 0.40 :mlj.j) _— Specific Radioactivity pgatoms T Incubation T20 ' Keto Acid Keto Acid Incorporated Time (Hrs) dpm/“atom dpm/mole Recovery per mole 0.25 1.25x106 1.55x104 90%21 0.013 20.0 1.25 x 106 8.75 x 104 80 0.070 3'Based on 20 moles of a-ketobutyrate present at 50% reaction. of a-ketobutyrate. As judged by this lack of incorporation, the a-hydro- gen of threonine should be transferred by the hydride transfer mechanism (Figure 3, left). The Fate of Tritium in the Dehydration of a-Tritio threonine DL-a-tri’tio threonine, synthesized as described in ”Methods and Materials, “ was incubated with threonine dehydrase until the L-isomer was completely dehydrated as determined by a-ketobutyrate assay (semicarbazide method). When dehydration was complete, the incubation mixture was poured onto a Dowex-l-chloride column (1. 1 x 5 cm.) and washed with 50 ml. of water. The wash was then concentrated to dry- ness using an acetone-dry ice trap to yield a volatile fraction (water) and a residue (D-threonine). The a-ketobutyrate was eluted from the column in the usual fashion. Radioactivity was determined in the water, threonine and the a-ketobutyrate (Table VII). These results clearly show that the (1.-hydrogen of threonine was lost to the water, and virtually no transfer of this hydrogen to the 48 Table VII. Enzymatic dehydration of a-tritio threonine. (Each reaction contained 5 moles of AMP, 5 prnoles of NaGSH, 100 moles of phosphate buffer, pH 8. 0, 250 units of dehydrase (specific activity = 1,300), 2. 5 moles of L-threonine and 5 urnoles of DL-a-tritiothreonine in a final volume of l. 0 ml.) Disintegrations per Minute Isotope Enzyme Substrate Keto Acid HTO Residue Recovery (Percent) Native 362,000 650 172, 000 168,000 96 Boiled 362, 000 0 l, 950 358, 000 99 8-position of a-ketobutyrate occurred. A control with boiled dehydrase handled in the same manner showed that no spontaneous exchange took place between a—tritio threonine and water. The conflict between the lack of incorporation of tritium from T20 and the release of tritium from a-tritio threonine appeared to result from an isotope discrimination in which hydrogen incorporation was considerably favored over that of tritium because of the extremely low molar quantity of tritium present. This disc rirnination could be eliminat- ed by running the dehydration in 99. 7% D20. In this case, if sources of exchangeable H were minimized, the rate of dehydration could be lower but discrimination in favor of H would be impossible or greatly reduced. Experiments with Deuterium oxide After enzymatic dehydration of 1. 5 mmoles of threonine in 99. 7% D20 for 20 hours, approximately 1 mmole of a-ketobutyrate had been formed. Alpha-ketobutyrate was isolated by chromatography on a 1.4 x 14 cm. column of Dowex-l-chloride. The pooled fractions con- taining a-ketobutyrate were treated with 0. 33 volume of 1% 2, 4-dinitro- phenylhydrazine in 2 I! HCl and the a-ketobutyric acid dinitrophenylhydrazone 49 recovered. One hundred thirty mg. of compound were obtained after four recrystallizations from a 3:1 methanol-water mixture. The melting point was 200.5-2010 C. (uncorrected). A control reaction containing 1. 0 mmole of potassium a-keto- butyrate (86) in place of L-threonine was carried through the same procedure in order to partially correct for enzymatic and nonenzymatic enolization of the keto acid. After four rec rystallizations, 99 mg. of the authentic a-ketobutyric acid dinitrophenylhydrazone were obtained. The melting point was 199. 51-2000 C. (uncorrected). Table VIII tabulates the results of the deuterium analyses and indicates that after subtraction of a rather substantial control value, approximately 0. 79 gram atom of deuterium was incorporated per mole of a-ketobutyrate. This incorporation is in basic agreement with a route involving desaturation, and tautomeric rearrangement of a-amino- crotonate as was also the release of tritium from o-tritio threonine shown in Table VII. Table VIII. Threonine dehydrase reactions in deuterium oxide. (Aqueous solutions containing L-threonine, 1. 5 mmoles, 40 moles of AMP, 60 prnoles of NaGSH and 750 moles of phosphate buffer were ly0philized together and redis- solved in 7. 0 ml of 99. 7% D20. Approximately 1, 000 units of dehydrase (specific activity =1, 300) were precipitated with (NH4,)?,SO4r and the well-drained precipitate dissolved in 0. 2 m . of 99. 7% D20. Alpha-ketobutyrate, l. 0 mmole of the K salt, replaced L—threonine in the control reaction.) Compound Atom Percent Hydrogen Atoms Atoms D Analyzed Excess D Per Mole Per Mole A a-Ketobutyrate DNP-hydrazone 14.12 10 1. 41 enz ' ( lec) 0. 79 a-Ketobutyrate DNP-hydrazone 6.18 10 0. 62 (control) 50 The value of 0. 79 gram atom per mole of a-ketobutyrate is considered to be a minimum figure with the true value approaching one gram atom of deuterium incorporated per mole. This conclusion is based on the assumption that the amount of incorporation in this type of control would be higher than the corresponding spontaneous incorpor- ation in the enzymatically produced a-ketobutyrate because: (a) the concentration of a-ketobutyrate available for enolization was initially higher in the control than in the enzymatic reaction which started at zero a-ketobutyrate, and (b) the rate of deuterium incorporation by enolization into the added B-diprotio-a—ketobutyrate of the control would be faster than deuterium incorporation by enolization with the enzymatically produced B-protio-8-deuterio-a-ketobutyrate, which already contained 50% of the maximum deuterium content. In the latter case, a substantial proportion of the exchanges would only involve an exchange of D for D. Back- incorporation of Deuterium from D30 and and 013 from H3018 into Threonine The data presented in Table IX for deuterium and 018 back- incorporation were obtained in separate experiments. Each experiment was interrupted when exactly 50% of the 2 mmoles of threonine had been utilized as measured by threonine disappearance by the alcohol dehydro- genase assay. Reisolation of the remaining threonine and crystallization (68) yielded from 40 to 90 mg. of the nearly pure dipolar ion of threonine. For the experiments with deuterium, 99. 7% D30 was used; for experi- ments with H3018 the 018 final concentrations were 4. 77 atom percent excess for the native enzyme incubation and 4. 53 atom percent excess for the boiled control. The results in Table IX indicate that about 25% of the threonine molecules had incorporated deuterium and an equal amount had incorporated 018. These data indicate that some dehydrated species such as a-aminocrotonate was rehydrated to form threonine containing deuterium or O”. 51 3.3.80 03 20 o o m .26 A 2 .o I- .1 8o: 05:8an OQN .o . . . u: .. ONO + 05>Ncm~ NS o o m no A 8 m 88: oaaoooace . .. u .. u i u . . 00cmpadn< Sm o 308 o Honaooz zeoo . . . . . . Hotcoo 20am 050 o m mm w so 0 0mm o o>¢oo o 80: 05:00.33. wvw .o O~E + magnum o o o o o ”A QQN O m hm. an NN¢ O ONQ O OONHO O EOHH vaGOmHamH. 4 032 non.“ 302 H0nm H0003 g mm0oxm 00.00qu av 0003.93. vasomcuou p0umnomuoonH I .Ho 0 mm0oxm— 9:00 acoouonm 834 “W 083.4 05004 Lona 800.4 83.4. 030m 0002 Tommupcwfiop 00303 00:30:00 0003000.” #03000 0:23-20 00900qu 0003 m0830> pom mucmuomou :0 “@0008 .HHH> 0308 GM 0005. 0» $03003 0H0? 00.35035 mcofiodomv .ONQ 0:0 SONS 50.5 gwuousop pad 20 m0 0:?0005 005 GOSMHomnoochomn oflmgucm .XH 030A. 52 Non-enzymatic rehydration of a-aminoc rotonate was ruled out in the case of O18 incorporation by the observation that the isolated Ola-threonine was pure L-threonine as measured by the excess threonine dehydrase assay, and contained no D-threonine or DL-allo threonine which would be expected if non-enzymatic rehydration were occurring. Isolation of an Intermediate in Threonine Dehydration Further evidence that the dehydration proceeds via an unsaturated intermediate was obtained by treating dehydrase reaction mixtures con- taining Cu-L-threonine with borohydride. The reaction mixtures contained 1 tunole of AMP, 1 mole of NaGSH, 30 moles of potassium phosphate buffer, pH 8. 0, 10 units of dehydrase and 4 moles of C“-L-threonine (specific activity = 0.1 ucurie per mole) in a final volume of 0. 25 m1. During the reaction at 3° C., 10 moles of KBH4, followed by 8 moles of acetic acid, were added as described previously for borotritide reduction. The reaction mixture was then chroma- tographed on Whatman No. 1 paper in 77% ethanol - water or t-butanol - 2-butanone - water - cyclohexylamine (40:20:40:4) and amino acids located by ninhydrin spray or radioautographs developed after 72 hours exposure. A major spot corresponding to threonine was observed, but there was in addition a new ninhydrin-positive and radioactive compound which subsequently co-chromatographed exactly with authentic DL-a- aminobutyrate in the above two solvent systems. The only other labeled compound was identified as CH-a-ketobutyrate or Cu-a-hydroxybutyrate. No (1.-aminobutyrate could be detected after incubation of mixtures of a-ketobutyrate, ammonia, enzyme and KBH4 and no dilution of radio- activity inCH-a-aminobutyrate was observed when CM-L-threonine was dehydrated enzymatically in the presence of a large pool of unlabeled 53 o-ketobutyrate. Both of these results are consistent with the postulate that a-aminobutyrate was not formed from a-ketobutyrate, but rather was formed from an intermediate produced enzymatically. Since the moles of (1.-aminobutyrate formed far exceeded the moles of enzyme present, the reduction of a free intermediate was indicated. The intermediates postulated to exist between threonine and a-ketobutyrate are a-aminocrotonate and a-iminobutyrate. Although either of these compounds could be reduced by suitable agents to a-aminobutyrate, the nature of borohydride as a reducing agent favored a-iminobutyrate as the immediate precursor of (1.-aminobutyrate. To establish the nature of the precursor, reduction was carried out with H3-labe1ed NaBH4. The tritiated aminobutyrate was isolated and purified by chromatography as above. This material was then oxidized to a-ketobutyrate with a mixture of D- and L-amino acid oxidases. When the oxidation was com- plete, the mixture was poured on a column of Dowex-l-chloride and washed with 50 ml. of water. The wash fraction was lyophilized and the radioactivity estimated in both the volatile fraction (water) and the non-volatile residue (unreacted substrate). Keto acid was recovered by the usual gradient technique and analyzed for radioactivity. A control with boiled oxidases was treated in the same manner. I If a-aminoc rotonate were reduced by borotritide, the tritium would be present in the a and [3 carbons of a-aminobutyrate and would yield B-tritio-o-ketobutyrate upon oxidation. If (1.—iminobutyrate were reduced to a-tritio-a-aminobutyrate, oxidation would yield unlabeled a-ketobutyrate. Table X shows that after oxidation by D- and L-amino acid oxidases, tritium was found only in water, indicating that a-tritio- a-aminobutyrate was oxidized. Although only a small part of the substrate was oxidized, these results indicate that the precursor of a-aminobutyrate was (1.-iminobutyrate rather than a-aminoc rotonate. The control reaction showed that no tritium exchange with water had occurred in the absence of active oxidases. 54 Table X. Oxidation of tritiated a-aminobutyrate. (The tritium-labeled (1.-aminobutyrate, isolated from chromatograms of reaction mixture treated with NaBH4-H3 as described in the text, was incubated with 1. 5 units of L-amino acid oxidase, 0.4 unit of D-amino acid oxidase, 30 units of catalase, 25 moles of tris-chloride buffer, pH 7.8, 25 moles of KCl, arC1)d 2.5 prnoles of unlabeled DL-o-aminobutyrate for 4 hours at 38 C. The above enzyme units from the manufacturer‘s labels are equal to 1 mole of substrate decomposed per minute.) Isotope Disintegrations per Minute Recovery Enzymes Substrate Keto Acid HTO Residue (Percent) Native 254, 000 0 68, 000 148, 000 85 Boiled 254, 000 0 3,100 162, 000 65 In a second experiment of the same type, a dehydrase reaction mixture containing Cu-L-threonine as substrate was treated with NaBH4-H3 as described previously. The H3, Che-aminobutyrate was isolated and purified by paper chromatography in an ethanol - water (77%) solvent system and a n-butanol - acetic acid - water solvent (60:15:25). Oxidase treatment was performed as before but a more active prepara- tion of D-amino acid oxidase was available for this experiment. At the completion of the reaction, the flask contents were ly0philized directly and the reaction solvent (water) recovered. The dry residue was then redissolved in HZO, placed on a Dowex-l-chloride column, and washed with 25 ml. of H30 to yield a fraction containing unreacted materials. The a-ketobutyrate was eluted from the column by the usual gradient technique, and both C“ and H3 radioactivities were estimated in all fractions. The results of this oxidation are shown in Table XI. These results confirmed those obtained from the first experiment, and indicated that almost all of the tritium is released from o-aminobutyrate into water 55 N0 00 80.2 80;: Se oom.o o: o 8mg: 25.03 oozom on 0.2. 80.2 08;: on; coon: coma 80.0 oomoa 25.03 9:32 350000an :0 «3 LFU «3 :0 am SD ME SO «Luz 005>Ncm >00>000m 0:0500m OHS 050< 000M 000.3055 0&0000H 05552 009 05030Hw0p50fifli 00.003.90.005 id 505 0 0 000m 0500.059 70003500 003 50305005 0050 05355 #00500 050 00 00000 003 0505.1 o .3 45 o .H 003 0553c, 05m 059. .0 00m 00 0.3.05 m 500 000n>5£o550tdtdn 0505005 no 00505.1 0 .m 050 .m .N. In £30.8— 0050301053 00 005051 ooH 0050000 m0 00.5: CM 0000503 0500 05501Q mo 0550 N 6000508 0500 0550-4 mo 05:: N 553 0300.305 00B 545 A .8 05305 05H. .580 0a.: 5 0050 0000 00 005355 5030000 503 000505 003 0005353055010 05.5 .0u0u>50.055010130 .nm m0 50300MXO .Un 030B 56 by oxidase treatment, and thus is in agreement with an o-tritio-o-amino- butyrate structure. The use of both H3 and C” labels and the ratio of activities served to indicate the purity of the o-aminobutyrate used in this experiment. In addition, the location of the various compounds during the chromatographic phase of the experiment was considerably simplified by the use of the c” label. Identification of a Cysteinyl Residue in Proximity to the Active Site of Threonine Dehydrase The nature of the inactivation of threonine dehydrase during dehy- dration of serine has been a mystery in view of the fact that dehydration of threonine exhibits no such inactivation. A recent postulate (18) that an oxazolidine ring conjugate was formed between serine and pyridoxal phosphate does not explain the inability of threonine to behave similarly. The only support for such a postulate was the observation that added pyridoxal phosphate overcame the inactivation. This finding, however, could not be confirmed in the present study (see Table V). A more plausible explanation could be predicted from the results of treating dehydrase reaction mixtures with borohydride. The hydrolysis of o-iminobutyrate is sufficiently slow to allow a substantial reduction of the a-imi'no acid to o-aminobutyrate with borohydride as shown above. Since addition of borohydride during the dehydration of serine did not yield detectable alanine, it is apparent that a-iminopropionate is much more rapidly hydrolyzed. Thus it could be predicted also that a-amino- ac rylate would be much more reactive than a-aminocrotonate. The propensity for nucleophilic addition to the double bond of a-aminoacrylate derivatives is well-known (87, 88). Consequently, it was considered likely that serine inactivation could result from the addition of a-aminoacrylate to some nucleophilic group on the enzyme. 57 The most likely candidates as nucle0philes are SH groups of cysteinyl residues and e-NHZ groups of lysyl residues. In the former case, the expected product of aminoacrylate addition would be lanthionine; the latter condensation would yield lysalanine. To investigate this possibility, the dehydrase was inactivated by dehydration of CH-L-serine and after removal of the free C14, hydrolysis and chromatographic identification of any radioactive components was undertaken. For inactivation, the reaction mixture contained 0. 5 ml. of CM-L-serine (50 ucuries, 0.4 mole), 50 moles of Clz-L-serine, 300 moles of phosphate buffer, pH 8.0, 1 mg. of bovine serum albumin and 60 units of AMP-free threonine dehydrase (specific activity = 1, 000 units per mg. ). Further additions were made as follows: after 10 minutes reaction, 1 prnole of AMP was added; at 30 minutes, 60 additional units of threonine dehydrase were added; at 40 minutes, 10 prnoles of Clz-serine were added; at 45 minutes, 1 mole 'of AMP was added; at 50 minutes, 30 units of threonine dehydrase were added. After 1 hour of reaction, the entire mixture (final volume = 1. 0 ml.) was transferred to a 1.1 x 26 cm. Sephadex G-25 column. The column was developed with water and 10-drop fractions were collected. The effluent equal to the void volume from the column was pooled and evaporated to 1 ml. The protein was precipitated with 1 ml. of 10% trichloroacetic acid (TCA) containing 1 mg. per ml. Clz-serine, centri- fuged and the precipitate washed 3 times with 2 ml. of 5% TCA contain- ing 1 mg. per ml. Clz-serine and then washed twice with 2 m1. of ether. The washed and dried pellet was resuspended in water and treated with 5.0 pg, of fungal protease as described previously. After 24 hours, the mixture was evaporated to dryness and taken up in 10% isopropanol. For further characterization beyond chromatography, certain chromat- ograms were eluted and subjected to a 6 hour hydrolysis at 121° C. in 6 g HCl in a sealed, evacuated tube. This material was then evaporated to dryness i_n vacuo and the chromatographic identification repeated. 58 Paper chromatography was conducted in the following solvent systems: n-butanol - acetic acid - water, 60:15:25; n-butanol- pyridine - water, 1:1:1; phenol - water, 100:20, with and without a 0. 3% NH, atmosphere; and 77% ethanol - water. Radioactivity was located by direct counting of cut paper sections (1 x 2 cm. ). Table XII presents the Rf values obtained for unknown radio- active compounds and for authentic standard samples. Table XII. Rf values of Cu-labeled products from enzymatic hydroly- zates of CH-serine-inactivated threonine dehydrase. Rf in Solvent a Compound 1 2 3 4 5 Cysteic acid 0.070 0.24 0.10 0.045 0.070 Cystathionine 0.11 0.15 0.20 0.185 0.12 Lanthionine 0.10 0. 090 0. 060 0. 080 0.18 Serine 0.22 0.30 0.48 0.27 0.32 Lanthionine sulfoxide 0. 075 0. 14 0.14 0. 17 0. 15 Lanthionine sulfone 0. 040 0. 10 0. 12 0. 15 0.40 Lysalanine 0.17 0.19 0.32 0.57 0.33 Cu-enzyme hydrolyzate . 0. 055 0. 090 0.095 0.095 + + -b 0.12 0.17 HCI hydrolyzate of 0. 050 CH-material from + 0. l7 Solvent 3 0.10 + + 0.74 0.80 a'Solvent No. 1 = n-butanol : acetic acid : H20; 60:15:25 No. 2 = n-butanol ; pyridine ; H20; 1:1:1 No. 3 = phenol : H20; 100:20 No. 4 = same as No. 3 but with NH3 atmosphere, 0. 3% No. 5 = Ethanol: H30; 77:23 Not suitable for crude hydrolyzate because of considerable streaking. 59 The Cu-containing material migrated to essentially the same position as lanthionine in the butanol - acetic acid — water and butanol - pyridine - water solvents (systems 1 and 2 of Table XII). No radio- activity was found to correspond with authentic lysalanine or serine, indicating that the labeled material had undergone chemical modifica- tion and that the e-NHZ group of a lysyl residue did not add across the double bond of a-aminoac rylate. Chromatography in phenol - water - ammonia (system 4), however, gave rise to two radioactive components; one could be identified as lanthionine and the other had an Rf value similar to that of lanthionine sulfoxide. When the phenol solvent was used without the NH3 atmosphere (system 3), again two radioactive peaks were present. In this instance, one component was identified as lanthionine while the second component migrated to a location which did not permit a decision among lanthionine sulfoxide, lanthionine sulfone and cysteic acid. Elution of this radioactive material, followed by a brief HCl hydrolysis, yielded components identified as cysteic acid and lanthionine when rechromatographed in phenol — water (system 3) or lanthionine when chromatographed in ethanol - water (system 5). In addition, a third compound was observed on the chromatograms. This compound had a high Rf in phenol - water and ethanol - water sol— vents, as would be predicted for S-ethylcysteine. However, positive identification could not be made due to the lack of a standard. On the basis of the foregoing chromatographic evidence, the radioactive compound was considered to be lanthionine formed by the addition of a-aminoacrylate to the SH group of a cysteinyl residue in peptide linkage. The appearance of lanthionine sulfoxide and cysteic acid on phenol-water chromatograms is not surprising because this solvent often causes oxidation of thiols and thioethers (89). The formation of S-ethylcysteine or similar compound must be due to a decomposition of lanthionine during acid hydrolysis, in spite of the precautions taken to prevent such destruction. 60 The incorporation of radioactivity from Cu-serine into the protein fraction could further be utilized for an estimation of the number of active sites of the enzyme, provided that each site had one cysteinyl SH group in proximity. On the assumption that each site inactivated binds serine, the number of moles of serine-Cl4 bound per mole of dehydrase could be determined from the total counts bound and the specific activity of serine-C14. The number of moles of serine-CM bound would then equal the number of moles of active sites. For this, serine inactivation of threonine dehydrase was carried out in the usual manner with the exception that the loss of dehydrase activity was followed. When 72 units of the 430 units of dehydrase had been lost, the reaction mixture was passed through a Sephadex G-25 column and fractions collected as usual. The protein—containing fractions were pooled, evaporated to 0. 35 ml. and dialyzed against distilled water. The protein-bound radioactivity was then measured and calculated to be 18, 700 dpm total.1 If the assumptions of 100% recovery of radio- active protein from the column, a molecular weight of 80, 000 and a purity of 33% are made, a total of 2. 9 moles of (1.-aminoacrylate (serine-C”) was bound per mole of threonine dehydrase. Because of the large number of assumptions, the value of 3 moles of a-aminoacrylate per mole of dehydrase is probably only a rough estimate for the number of active sites. The Mechanism of AMP Activation The preceding investigations of the mechanism of dehydration, the mode of pyridoxal phosphate binding, and the serine inactivation offer no insight as to the role of AMP activation of threonine dehydrase. 1The radioactivity bound was approximately 0. 0017% of the input radioactivity, 1.1 x 108 dpm. 61 It is quite likely, then, that AMP does not directly participate in the catalytic action in a manner analogous to pyridoxal phosphate or any other normal coenzyme, but rather functions less directly in a con- trolling capacity. A kinetic basis for the activation by AMP is presented in Figure 4. Double reciprocal plots of velocities observed with a range of substrate concentrations at pH 7. 0, 8. 0, and 8.6, and in the presence or absence of AMP, indicate that when AMP is present, there are marked decreases in Km as well as increases in Vmax' At pH 8. 0, Km and Vmax values in the presence of AMP were 5 x 10"3 l_\_/1 and 0. 27 A A per minute, respectively; in the absence of AMP, Km and Vmax values were 3 x 10'7‘ M and 0. 034 AA per minute, respectively. Qualitatively similar Km and vmax changes were noted at other pH values. It is interesting to observe that in the absence of AMP, Km values were identical at pH 7. 0, 8. 0, and 8.6, suggesting that Km is independent of pH over this range. In the presence of AMP, however, the Km values were markedly pH dependent. More conclusive evidence for a role of AMP was found through sucrose gradient centrifugation experiments conducted in the presence and absence of AMP. This technique was ideally suited for the study of fairly drastic conformational changes because only microgram quantities of enzyme were required to locate activity peaks and com- plete purity was not essential. Sedimentation patterns of threonine dehydrasel in sucrose 1Marker enzymes as well as threonine dehydrase were incorpor- ated into the 0. 10 ml. sample that was placed on each gradient. Because of the different specific activities and stabilities of the enzymes, the actual amount of each enzyme placed on the gradient was determined empirically. The following amounts of enzymes were used per gradient: j 2 units of yeast alcohol dehydrogenase, 1 unit of liver alcohol dehydro- genase, 1 unit of KDPG aldolase, 1 unit of muscle aldolase and 2 units of malate dehydrogenase. Generally only 2 or 3 marker enzymes were 62 0020830950 :0 00m 0009 003 00:03 00.00000an 0050.? En 00000300 000 0.0 30030». 5 0000000300 00w 0000000 00 00000.”."00 0000. 000 0>0£ 000D $0000 003300 000G0w000>£00 0300TEQ 0:0 SHE 3v 005wwh 63 lm |m 0 A _ _ 000 00. 00 o 000 om. 00 o 000 om. 00 o q _ - _ _ _ _ _ a _2 0.315.. _ u x . n. 124 ILL; $.90 xm F. QW‘M‘IF—fi 2Nu0_x_nEv_ \.\0\ L111 \\o\p \§ 1:3- 8 110111110 0 o i ow 1 00 20-02015. 20.0.3.5. 20-03815. d 20. .SOIEE n:>_< .SOI._._>> n:2< HDOIEB now 0.0 In 0.0 In 0 ON In 0 m < _|> 64 gradients containing 3 x 10"3 M AMP or 3 x 10'3 M IMP yielded results shown in Figure 5. When no nucleotide was added to the gradient and care taken to remove traces of AMP from the enzyme preparation, the dehydrase migrated as in an IMP gradient. The average sedimentation coefficient for threonine dehydrase in the presence of 3 x10'3 M AMP was 7.6 S; in the presence of 3 x 10’3 l\_/IIMP or no nucleotide, the average 520 was 5. 0 S (Table XIII). It was thus concluded that the presence of AMP induced a conform- ational change of sufficient magnitude to produce a drastic increase in sedimentation velocity. If the assumptions of identical partial specific volumes and spherical shapes are made for marker enzymes and threonine dehydrase, a rough estimate of molecular weight can be obtained from sedimentation coefficientsz alone (70). When this calculation was performed, a molecular weight of 153, 000 was found for the 7.6 S species, and 83, 000 for the 5.0 S species. This cor- responds to an apparent dimerization in the presence of AMP. In a similar experiment in which cytidine mon0phosphate (CMP), a less efficient activator, was substituted for AMP at 3 x 10'3 M concentration, threonine dehydrase had an 520 of 6. 9 S; this is in accord with the idea that ability to activate the enzyme is related to the ability to produce a dimerization. Figure 6 shows the sedimentation velocity (expressed as the peak position after 16 hours of centrifugation) versus AMP concentration used in each gradient. Malate dehydrogenase was discontinued in later experiments because the particular preparation available had a sedi- mentation coefficient (520) of 4.8 S rather than 3. 6 S as reported (90); this increased 520 was confirmed in independent experiments using the Spinco Model E analytical ultracentrifuge. The enzyme units referred to above are expressed in prnoles of substrate decomposed per minute. 2520 of unknown : ( Molecular weight of unknown )% 320 of known Molecular weight of known 65 003035000000 mo 000300000 00% #0000 00m .8 30000 002030 0000 ”ma 0030000 .00000w000>£00 0000000 00>: “Md 0030000 .000c0w000>£00 0030000 0009» "0003 00030000 x000 0 000002 00000500 05000000 00 0030990000000 0000000w 00000.9m .7 .m 000m0h 66 m30m_zw2 \ on em 00 mmmzaz 20.00405 _N 0. m. m. 0.0 0.. m; (‘lW/SllNfl) All/\ILOV 67 .mmoa E :830 .35 .H .< .3 683 98 .m .m .2832“. 5:02:08 93 o» mvaommmhuou mm nvngz cowuumnhu .wmmfiogm UnHQM mo ommgm ado: ogmmunoa 04.. mo mwdduon @3250de «022’ .ucvflumnm use 5 9530.93 mm mo H30» m 93 SHozuuommmn .nm2< mo @25QO find oucvmoum 93 G“ A: wad NH Goflomum mo mcofifimom Mmom no vmmmmm 80.3 o.m coca? 9s mmfieé n--- o .m n--- o .N. v ooo .2. m .w 8. 322:4 omnm ace .3 o .m 25 .me m. .5 No coo .2: N .m S 33%? £832 28 .3. m .w ooo .3; m .N. 3 25 .$ N .m 2 En? $34 ace .3 N .m ooo .1: o 4. 2. cos .2: mg 2 E: 38» £5 £02 2m is .202 8m gum £5 .332 gm .833; Magnum ms? 38:; ms? :33 .m omdnv>£ofl mawcomugh. mogncm Hvxumz Unmvcmum 5324 m0 vocomnm tam mocomvum 2i 5 ommHv>£vU 93:00.33 mo muamwoB Haddoofiog Una mucowoflmooo cofimucoewvmm .3an 3an 68 .533 xdom 05 3.353 mom: 33:0». 2:. 638:2: mm we? m5 mm? cofimufihmucoo n32< map #588 m madman a“ mm 05mm 0:» mum? $533500 .coEmHucvucou @644 we cofiuadw m mm ommuw>£vv 05:00.2: mo 8mg cofimucongom .c ondmflh 69 i i Io-4 Io-5 MOLARITY OF AM P IN GRADIENT 1 IO‘5 1 IO'5 v N l 2 >W3d :IO HBBWHN 20" 1 (1] m NOLLOVHd 70 in the sucrose gradient (enzyme concentration was nearly constant for all experiments). There was a series of intermediate 520 values at intermediate AMP concentrations and a pronounced broadening of the peak (as indicated by the vertical lines on Figure 6) was apparent. The intermediate 520 values and the absence of multiple peaks corres- ponding to varying ratios of 7.6 S and 5. O S species suggest that the two species are rapidly interconverted at limiting concentrations of AMP. Gilbert (93) has discussed the theoretical sedimentation patterns for a monomer-dimer system in rapid equilibrium and concluded that no resolution of the two components is possible and that only a single, asymmetric peak would be obtained when an equal concentration of monomer and dimer are present. On this basis, it is proposed that the following equilibrium exists: 2 [5.0 S "monomer"] + nAMP €=’ [7.6 S "Dimer"— nAMP] With the knowledge that dimerization occurred in the presence of AMP and that reversible dissociation was favored by lowered con- centrations of AMP, experiments were designed which would provide further insight into the mechanism for this structural alteration. The sucrose gradient technique was utilized to obtain small quantities of both the "monomeric" or ”dimeric" species in their pure forms. These fractions were first assayed in four ways: (a) with both AMP and GSH in the cuvette, (b) with AMP only, (c) with GSH only, and (d) with both omitted. When a valid rate had been obtained, any omitted component was then added and the new rate observed. The results of assays conducted with these species are tabulated in the upper part of Table XIV. The more prominent facts gathered from assays of the "dimeric" or 7.6 5 fraction and the "monomeric" or 5. 0 S fraction are as follows: 7l .0GEO0ufi 53> 003.955 om H03»? 60300 0:209:33 0H0m0o. 003235 m p0umndug0unf 2 as .o 300 .02.? m coo .o 0000+ o moo .o @331 m go .o 0000+ om mmod 300+ w Sod 008 024+ . u s 0000 o a. - m z . 0:0 + . . . 9v omo o z 3 omo o mmo m2<+ 2: mm; .o 282+ 2: N2 .o $8 .02.? 2: 93 .o 282+ 2: as .o £00... mm to .o 300+ 2 :55 m2? o.m u Sm m» v8 .o $5: 2 go .o $9 3 N 8 .o £00.. s moo .o 282 H 2: 3.6 282+ 2: 36 £05 2. mm .o mmo+ mm on .o m2<+ 0;. u Sm 3 mm .o .22.? S 35.0 mmo+ «m E .o .381 3 OS .o 282 3:00H0n5 A.GME\<30H0m pnoo0m 033.20% pmnmh H ll {1" .000up>£0p 0CEO0§3 mo m0fiu0mm m o .m paw m “v.5 05. >n U03££X0 0033304 .>HX 030R. 72 l) The 7. 6 S fraction exhibited nearly the same activity with or without GSH, but had a greatly reduced activity in the absence of added AMP, suggesting that a rapid dissociation of the dimer may occur in the enzymatic assay in the absence of AMP. A sizable reactivation occurred when AMP was added to a reaction mixture which had been preincubated with only GSH. Z) The 5. 0 S fraction exhibited dependence on both AMP and GSH. Assay of this fraction in the presence of only GSH yielded a low activity; partial activation was seen upon the addition of AMP. In the reverse incubation, considerably less activation was seen upon GSH addition, suggesting that activation is accomplished best by GSH treatment first, followed by addition of AMP. This fact is considered to indicate that the 5. O S fraction as obtained from the sucrose gradients contains at least one disulfide bond which must be reduced before AMP can induce dimer formation. 3) Assay of the 7. 6 S fraction without both AMP and GSH resulted in very low rates; addition of both of these components failed to activate to the same extent as when only GSH was present, and suggests that the dimer form in the absence of both AMP and GSH is first converted to a monomer and then to an additional inactive form. This latter form is quite likely identical with the 5. 0 5 fraction isolated in sucrose gradients containing no AMP. This form of the enzyme requires both AMP and GSH for activation, and can be activated only some 30% when AMP and GSH are added ific}; threonine. Assay of the 7. 6 S fraction with oxidized glutathione (GSSG) at 3 m1_\_/1 concentration in the absence of AMP and GSH revealed little activity (Table XIV, lower part). The addition of AMP to this mixture did not produce appreciable activation unless GSH was also added. Simultaneous preincubation of AMP and GSSG, followed by GSH addition upon assay, resulted in a slightly greater activation. These findings are considered evidence that the 73 monomer—dimer equilibrium can be upset in favor of the monomeric species by the presence of an oxidizing agent (GSSG), but that AMP can partially antagonize this action, presumably by maintaining the dimeric structure. These facts formed the basis for the tentative model proposed in Figure 7. Although additional proof of the accuracy of this model has not been obtained, the model is open to experimental confirmation. One point yet to be experimentally verified is the prediction that a 5. O S fraction isolated from a gradient containing GSH but no AMP would not require further GSH treatment in order to be activated by AMP. Furthermore, p—chloromercuribenzoate treatment of this fraction in the absence of GSH should prevent AMP activation due to the combination of PCMB with the SH groups which are essential for the formation and maintenance of the dimer structure. A further prediction of this model is a second-order dependence upon protein concentration for dimer formation in the presence of excess AMP. AMP Activation as a Metabolic Control Mechanism Having found a basis for the AMP activation, consideration was given its significance as a possible control mechanism in the anaerobic metabolism of E_I_. fl' Tokushige e_t a_._l. (Z3) concluded from experi- ments with the ADP-activated threonine dehydrase of the anaerobe, E. tetanomorphum, that ATP synthesis depressed threonine dehydrase activity by the removal of ADP and inversely, that ADP build-up activated threonine dehydrase (see a discussion of their proposal in the Literature Review). This mechanism could be feasible for anaerobically-grown E. C__°1_i where ATP synthesis via substrate phos- phorylation rather than oxidative phosphorylation would predominate. 74 .ommnp>£op 05:00.23 mo cofiumzuom no“ Hopoe pomomoum . p 0.19me '75 n X 5:: team 51 ”Em 2550.34 m z _ zommzk + 2,2 mEEmmam mo... - iv ”Em ozaza ImImm 1m rammed * .J . 1' \mJ “:24 4 m an Al mm x Imam $2 + m :52 mm 20.58%. Ammva mmszo AmOhV mmmEOZOE mmImo mz_zowmI._. “.0 20:43:04 mo... Ammo—2 QmmOdomd 76 A major point of variance, however, was that ADP rather than AMP activates the dehydrase of g. tetanomorphum. Thus it was postulated that E. coli possesses an adenylate kinase which links the AMP, ADP and ATP levels through the reaction: Adenylate 2ADP4_ > ATP + AMP Kinas e In this case, AMP rather than ADP could accumulate and thus be available to activate threonine dehydrase. Through the action of the dehydrase, a-ketobutyrate would be formed and this keto acid could produce ATP through a dismutation of a-ketobutyrate to propionyl phosphate and a—hydroxybutyrate. A corollary of this postulate would be the fact that (_3_. tetanomorphum is lacking in adenylate kinase. The results presented in Table XV show that adenylate kinase is found in reasonable amounts in both organisms. Thus this postulate for the difference in nucleotide specificity is inadequate. No further work on the i_n .vivo control ofthreonine dehydration has beenconducted. Table XV. Adenylate kinase content of extracts of 12.3% and E. tetanomorphum. (The preparation of extracts and assays were performed as described in the text. Ten pl. of a 1:10 dilution of the crude extracts was used for the assays.) Extract Unitsa/ml Protein (mg/ml) Specific Activity E. coli 164 14.8 11.1 E. tetanomorphum 188 19.4 9. 7 a . . . . A unit is defined as that amount of enzyme which produces an absorb- ance change of l. 0 O. D. per minute under the described assay condi- tions. DISC USSION The results obtained from isotope experiments on the mechanism of threonine dehydration confirm that an 0., fi-elimination mechanism involving unsaturated intermediates is operative. The data which support this conclusion are as follows: (a) (b) (C) (d) Dehydration in the presence of D20 resulted in the formation of a-ketobutyrate with nearly one atom of D per molecule of ketobuty rate . When a-tritio threonine was used as substrate, the tritium was recovered in water while the u-ketobutyrate was un- labeled. Treatment of dehydrase reaction mixtures with borohydride produced CM-a-aminobutyrate from Cu-threonine. Reduction with H3-labeled borohydride resulted in the formation of a-tritio-a-aminobutyrate, indicating that a-iminobutyrate was the immediate precursor of (1.-aminobutyrate. The results of O18 and D back-incorporation into threonine during dehydration suggest a reversibility of the dehydration, at least prior to the release of u-aminoc rotonate from the enzyme. These data also indicate that hydrolysis of the imino acid might be the only irreversible step in the entire sequence. In addition, the presence of pyridoxal phosphate bound as a Schiff base to the e-amino group of a protein lysyl residue strongly suggests that this compound plays an important role in dehydration, presumably through a transaldimination reaction between the pyridoxal-lysine Schiff base and the amino group of threonine. 77 78 A detailed mechanism incorporating all these facts is presented in Figure 8. The unprotonated amino group of threonine presents the second substrate for a transaldimination reaction from the enzyme- pyridoxal phosphate Schiff base; support for this assumption comes by analogy to threonine dehydration in sheep liver (41, 42), where kinetic studies showed that the reactive species has an uncharged amino group. The existence of the azomethines shown in brackets on Figure 8 is somewhat speculative but such structures are postulated as logical intermediates in the removal of an (1.-hydrogen and fi-hydroxyl by a pyridoxal enzyme. Because all these azomethines should be present during dehydration and all contain bonds reducible by borohydride, it seems reasonable to expect that borohydride treatment of reaction mixtures might allow the isolation of these enzyme-bound intermediates in their stable reduced state. This would provide definitive evidence for the existence of such structures. 1 The quite similar rates of back-incorporation of D and 018 into threonine suggest that a stepwise elimination of proton and hydroxyl group may occur in which the initial slow formation of a carbanion is followed by a rapid conversion to an unsaturated intermediate. Although this possibility is hardly distinguishable from a concerted elimination process and has been of quite limited importance in elimination re- actions (94), it is an attractive mechanism for the pyridoxal phosphate- catalyzed enzymatic elimination. A concerted mechanism has never been pr0posed for the pyridoxal phosphate type of elimination and a step- wise elimination is in basic agreement with the generally accepted mechanism for pyridoxal catalysis (31). The same intermediates as shown in Figure 8 were proposed for the non-enzymatic dehydrations catalyzed by pyridoxal (31), and emphasize the virtual identity of the enzymatic and non-enzymatic pyridoxal mechanisms for dehydration. Because of this similarity of 79 .Coflmnprwnvp vcficoofifi QUEEN“; mo Emflcmfloog 09H. .w ohsmwh 80 = $2 + -ooouo ~10 are i O .. u ..m NI7__ I000 w H 10 r m2>szu@umm .. Tom IIOI IIO+ .. 1w 4 m .. .. I . :0 «14. m 08 m. m I \ r m2>szu@u muro L +I+ 1 +7 45 I I In NI7__ IOOU w ”_v I0 rmz>szu@ummqu\z I 81 mechanisms, the role of the enzyme must be that of conferring specificity toward reactants and of increasing the rate of the pyridoxal- facilitated dehydration. This latter role may be accomplished in at least three ways: (a) by enzyme-promoted specific orientation effects which serve to bring all reactants together in the most favorable positions for reaction; (b) by catalysis of the transaldimination re- action, both in the formation of a pyridoxal-substrate Schiff base and in the cleavage of a pyridoxal-product Schiff base; and (c) by facilitating the loss of H+ and OH- from the substrate through an attraction of these groups by various amino acid residues in the enzyme active site. The effect which AMP exerts on threonine dehydrase catalysis is not indicated in mechanistic studies of the dehydration reaction, because the results presented are in keeping with a normal pyridoxal phosphate-catalyzed dehydration mechanism. But since it is evident from kinetic studies that the binding of threonine is markedly enhanced by AMP, and that Vmax is increased, one must conclude that AMP functions as an allosteric effector in controlling threonine dehydrase action; that is, AMP binds at a site somewhat removed from that for substrate but still exerts an effect on substrate binding and on the velocity of catalysis. The apparent dime rization of the enzyme protein in the presence of AMP, under conditions which simulate those existing in an enzymatic assay, strongly suggests that the increased catalytic effectiveness pro— duced by dimerization in the presence of AMP may be due to an assembling of additional groups which further facilitate substrate bind- ing and catalysis, or in other words furnishes a more effective active site than is present in the monomer dehydrase. The manner in which activation is brought about is uncertain, principally because the nature of the protein-AMP interaction is unclear. 82 It is clear, however, that AMP is not a completely specific agent. X-ray diffraction studies (95) indicate a close spatial relationship between the phosphate and primary amino group of 5'-AMP. Since several related compounds possessing these same groups in roughly the same spatial relationship are also capable of activating threonine dehydrase (CMP, dAMP, 3'-AMP) and since CMP, at least, can also induce a dimerization and activate, these two groups must play an important role in activation as well as in determining the specificity of the protein-nucleotide interaction. The requirement for a reducing agent such as GSH or Z-mer- captoethanol has long been known for threonine dehydrase (5, 11). The present work indicates that one or more SH groups are essential for AMP activation, and therefore that reducing agent functions to maintain the integrity of these groups. A direct test of this hypothesis would be to show that dime rization cannot take place in the absence of these SH groups. The role of SH groups in threonine dehydrase is strengthened by the demonstration that lanthionine is formed by incubation of serine with the dehydrase. This lanthionine is presumably formed by the addition of the SH group of a cysteinyl residue to the double bond of a-aminoacrylate, an intermediate in serine dehydration. Since it is likely that the particular SH groups involved are near the site (or sites) where d-aminoacrylate is released from the protein, it would be interesting to see whether these groups are also essential for AMP activation. That this could be the case comes from observations that serine inactivation is most efficient when the AMP concentration is relatively low (see protocol for serine inactivation experiments), indi- cating that excess AMP may partially protect these groups from (1.-aminoacrylate addition. 83 The model for threonine dehydrase activation and activity, illustrated in Figure 7, is consistent with all the data obtained, but only illustrates one of the active sites. It also features several details which have not been experimentally established. While there is evidence bearing on the number of binding sites for substrate, there is no indi- cation of the number of binding sites for AMP. Also, because the mono- meric form exhibits a small degree of enzymatic activity, a threonine binding site on each monomer is illustrated. Whether this same active site is merely modified in some way in the dimer (as shown in Figure 7) or whether a new and different active site is brought into being in the dimer cannot be decided on the basis of the present data. It must be mentioned that the present data suggest but do not conclusively establish that a monomer-dirner relationship exists in threonine dehydrase. The observed 520 changes could well be a reflection of only extensive intramolecular conformational changes. This possi- bility will remain unchallenged until such time as more direct molecular weight measurements are made. Attempts to induce biodegradative threonine dehydrase in rest- ing cells grown under non— inducing conditions have consistently failed. Despite testing of a variety of carbon and nitrogen sources for growth and several different possible inducers (serine, threonine, casein hydrolyzate, tryptone), no conclusive induction was obtained. Although this failure is not completely understood, the difficulty may lie in the failure to remove the repression of this enzyme which occurs in cultures of E. 9211 grown aerobically on a tryptone-yeast extract medium. An additional point remaining to be clarified is the i_n 1112 signifi- cance of AMP activation of threonine dehydrase. This question will remain unanswered until the exact fate of a-ketobutyrate is known. When the pathway for a-ketobutyrate catabolism is established, it should 84 be possible to explain how the rate of utilization of threonine affects the AMP:ATP ratio. It should be emphasized that the demonstration of adenylate kinase in both E. coli and g. tetanomorphum does not invalidate the control mechanism proposed by Tokushige gt a_.l. (23) but does make it difficult to understand how different nucleotides (i. e. , ADP and AMP) could regulate threonine catabolism in these two organisms by the same mechanism. SUMMARY Fifty-fold purified threonine dehydrase from E. col grown anaerobically on an amino acid medium has been found to require pyridoxal phosphate for activity. In the native enzyme, pyridoxal phosphate is bound as a Schiff base to the e-amino group of a lysyl residue. Studies with deuterium oxide, a-tritiothreonine and H3018 have confirmed that dehydration proceeds through unsaturated intermediates such as a-aminoc rotonate and a-iminobutyrate. Further evidence for these intermediates was the production of (1.-aminobutyrate from (1.-iminobutyrate upon borohydride treatment of dehydrase reaction mixtures. Adenosine monophosphate (AMP) increases the catalytic effective- ness of threonine dehydrase by decreasing the Km for threonine and increasing the maximum velocity. Furthermore, threonine dehydrase undergoes an increase in sedimentation coefficient, corre5ponding roughly to a dimerization, in the presence of AMP. The nature of serine inactivation of this enzyme has been investi- gated and found to result from the addition of a-aminoac rylate, the unstable product of serine dehydration, to an essential SH group near the active site of threonine dehydrase. 85 10.. 11. 12. 13. 14. 15. 16. 17. REFER ENC ES . Greenberg, D. M., in Metabolic pathways, Vol. II, Academic Press, Inc., New York, 1957, p. 88. . Greenberg, D. M., ibid., p. 174. . Cohen, P. R., and Sallach, H. J., ibid., p. 1. . Greenberg, D. M., in P. D. Boyer, H. Lardy and K. Myrback (Editors), The enzymes, Vol. V, Academic Press, Inc., New York, 1961, p. 563. . Gale, E. F., and Stephenson, M., Biochem. J., 3_2, 392 (1938). . Chargaff, E., and Sprinson, D. B., J. Biol. Chem., 148, 249 (1943). . Chargaff, E., and Sprinson, D. B., J. Biol. Chem., 151, 273 (1943). . Binkley, F., J. Biol. Chem., 150, 261 (1943). . Lichstein, H. C., and Umbreit, W. W., J. Biol. Chem., 170, 423 (1947). Lichstein, H. C., and Christman, J. F., J. Biol. Chem., 175, 649 (1948). Wood, W. A., and Gunsalus, I. C., J. Biol. Chem., 181, 171 (1949). Pardee, A. B., and Prestidge, L. S., J. Bacteriol., 7_0, 667 (1955). Boyd, W. L., and Lichstein, H. C., J. Bacteriol., Q, 545 (1955). Umbarger, H. E. and Brown, B., J. Bacteriol., 2, 443 (1956). Umbarger, H. E. and Brown, B., J. Bacteriol., 13, 105 (1957). Reissig, J. L., Arch. Biochem. and Biophys., 26, 234 (1952). Sayre, F. W., and Greenberg, D. M., J. Biol. Chem., 220, 787 86 87 18. Nishimura, J. S., and Greenberg, D. M., J. Biol. Chem., 236, 2684 (1961). 19. Walker, D. J., Biochem. J., 62, 524 (1958). 20. Goldstein, L., Knox, W. E., and Behrman, E. J., J. Biol. Chem., 237, 2855 (1962). 21. Sayre, F. W., Jensen, D., and Greenberg, D. M., J. Biol. Chem., 219, 111(1956). 22. Hayaishi, o., Gefter, M., and Weissbach, H., J. 13101. Chem., 238, 2040 (1963). 23. Tokushige, M., Whiteley, H. R., and Hayaishi, 0., Biochem. Biophys. Res. Commun., _1_3_, 380 (1963). 24. Whiteley, H. R., and Hayaishi, 0., Biochem. Biophys. Res. Commun., E, 143 (1964). 25. Holzer, H., Cennamo, C., and Boll, M., Biochem. Biophys. Res. Commun., 1_4_, 487 (1964). 26. Holzer, H., Boll, M., and Cennamo, C., Angew. Chem., 7_5_, 894 (1963). 27. Changeux, J. -P. , Cold Spring Harbor Symposium on Quant. Biology, 26, 313 (1961). 28. Changeux, J.-P., J. Mol. Biol., 4, 220 (1962). 29. Changeux, J. -P. , Cold Spring Harbor Symposium on Quant. Biology, 28, 497 (1963). 30. Szentirmai, A., and Horvath, 1., Acta. Microbiol. Acad. Sci. Hung., 9, 31 (1962). 31. Metzler, D. E., Ikawa, M., and Snell, E. E., J. Am. Chem. Soc., 7_6_, 648 (1954). 32. Braunstein, A. E., in P. D. Boyer, H. Lardy and K. Myrback (Editors), The enzmes, Vol. 11, Academic Press, Inc., New York, 1960, chapt. 6. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 88 Metzler, D. E., J. Am. Chem. Soc., 19, 485 (1957). Matsuo, Y., and Greenberg, D. M., J. Biol. Chem., 230, 561 (1958). Schirch, L. and Mason, M., J. Biol. Chem., 237, 2578 (1962). Kent, A. B., Krebs, E. C., and Fischer, E. H., J. Biol. Chem., 232, 549 (1958). Jenkins, W. T., thantis, D. A., and Sizer, I. W., J. Biol. Chem., 234, 51 (1959). Fischer, E. H., Kent, A. 13., Snyder, E. R., and Krebs, E. G..- J. Am. Chem. Soc., 82, 2906 (1958). Sizer, I. W., and Jenkins, W. T., in E. E. Snell, P. M. Fasella, A. E. Braunstein and A. R. Fanelli (Editors), Chemical and biological aspects of pyridoxal catalysis, Pergamon Press, New York, 1963, p. 125. Jencks, W. P., and Cordes, E., ibid., p. 57. Davis, L., and Metzler, D. E., J. Biol. Chem., 237, 1883 (1962). Russell, D. H., Thain, E. M., and Vernon, C. A., Proc. of Chem. Soc., 255 (1960). Monod, J., Changeux, J.-P., and Jacob, F., J. Mol. Biol., _6_, 306 (1963). Krebs, E. C., and Fischer, E. H., in F. F. Nord (Editor), Advances in enzymology, Vol. 24, Interscience Publishers, New York, 1962, p. 263. Krebs, E. C., and Fischer, E. H., in E. E. Snell, P. M. Fasella, A. E. Braunstein and A. R. Fanelli (Editors), Chemical and biological aspects of pyridoxal catalysis, Pergamon Press, New York, 1963, p. 543. Helmreich, E., and Cori, C. F., Proc. Nat. Acad. Sci. U. S., 21, 131 (1964). 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 89 Madsen, N. B., Biochem. Biophys. Res. Commun., 1_5_, 390 (1964). Frieden, C., J. Biol. Chem.,238, 3286 (1963). Tomkins, G. M., Yielding, K. L., Talal, N., and Curran, J., Cold Sprig Harbor Symposium on Quant. Biology, _2_8_, 461 (1963). Gerhart, J. C., and Pardee, A. B., Cold Spring Harbor Symposium on Quant. Biology, 28, 491 (1963). Gerhart, J. C., and Pardee, A. B., J. Biol. Chem., 237, 891 (1962). Hathaway, J. A., and Atkinson, D. E., J. Biol. Chem., 238, 2875 (1963). Chen, R. F., Brown, D. M., and Plaut, G. W. E., Biochemistry, 3, 552 (1964). Sanwal, B. D., Zink, M. W., and Stachow, C. S., J. Biol. Chem., 239, 1597 (1964). Freundlich, M., and Umbarger, H. E., Cold Spring Harbor Symposium on Quant. Biology, _28, 505 (1963). Speck, J. C., Jr., Rowley, P. T., and Horecker, B. L., J. Am. Chem. Soc., _8_5_, 1012 (1963). Gaudry, R., Can. J. Research, 26B, 387 (1948). Bezas, B., and Zervas, L., J. Am. Chem. Soc., _8_3, 719 (1961). Zahn, H., and Osterloh, F., Ann., _5_9_5, 237 (1955). Bergmann, M., and Grafe, K., Zeit. fur Physiol. Chem., 187, 187 (1930). Greenstein, J. P., and Carter, C. E., J. Nat. Cancer Institute, 1, 57 (1946). Akabori, S., Otani, T. T., Marshall, M., Winitz, M., and Greenstein, J. P., Arch. Biochem. Biophys., _8_3_, 1 (1959). Flavin, M., and Slaughter, C., J. Biol. Chem., 235, 1113 (1960). 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 90 Desnuelle, P., and Naudet, M., Bull. Soc. Chim. France, _1_2_, 871 (1945). Flavin, M., and Slaughter, C., Anal. Chem., _3_1, 1983 (1959). Lambert, M., and Neish, A. C., Can. J. Research, 28B, 83 (1950). Macgee, J., and Doudoroff, M., J. Biol. Chem., 210, 617 (1954). Flavin, M., and Kono, T., J. Biol. Chem., 235, 1109 (1960). Scherr, G. H., Anal. Chem., _3_4_, 777 (1962). Martin, R. C., and Ames, B. N., J. Biol. Chem., 236, 1372 (1961). Kinard, F. E., Rev. Sci. Instruments, 28, 293 (1957). Werbin, H., Chaikoff, I. L., and Irnada, M. R., Proc. Soc. Exp. Biol. Med., 102, 8 (1959). Wang, C. H., and Jones, D. E., Biochem. Biophys. Res. Commun., _1_, 203 (1959). Rittenberg, D., and Ponticorvo, L., Intern. J. Appl. Rad. and Isotopes, 1, 208 (1956). Boyer, P. D., Graves, D. J., Suelter, C. H., and Dempsey, M. E., Anal. Chem., _3_3_, 1906 (1961). Vallee, B. L., and Hoch, F. L., Proc. Nat. Acad. Sci., (U.S.), 11, 327 (1955). Mehler, A. H., Kornberg, A., Grisolia, S., and Ochoa, S., J. Biol. Chem., 174, 961 (1948). Baranowski, T., and Niederland, T. R., J. Biol. Chem., 180, 543 (1949). Meloche, H. P., and Wood, W. A., Fed. Proc., 22, 411 (1963). Kovachevich, R., and Wood, W. A., J. Biol. Chem., 213, 745 (1955). 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 91 Kornberg, A., in S. P. Colowick and N. 0. Kaplan (Editors), Methods in enzymology, Vol. 11, Academic Press, Inc., New York, 1955, p. 497. Lowry, 0. H., Roseborough, N. J., Farr, A. L., and Randall, R. J., J. Biol. Chem., 193, 265 (1951). Cormier, M. J., J. Biol. Chem., 237, 2032 (1962). Grazi, E., Meloche, H. P., Martinez, (3., Wood, W. A., and Horecker, B. L., Biochem. Biophys. Res. Commun., L93 4 (1963). Abeles, R. H., and Lee, H. A., Brookhaven Symposia in Biology,- Number 15, Brookhaven National Laboratory, 1962, p. 310. Umbreit, W. W., Burris, R. H., and Stauffer, J. F., Manometric techniques, Burgess Publishing Co. , Minneapolis, 1959, p. 295. Strumeyer, D. H., White, W. N., and Koshland, D. E., Jr., Proc. Natl. Acad. Sci. U. S., _5_Q, 931 (1963). Patchornik, A., and Sokolovsky, M., J. Am. Chem. Soc., 86, 1860 (1964). Smith, I. , Chromatographic and electrophoretic techniques, Vol. 1, Interscience Publishers, New York, p. 101. Wolfe, R. B., and Neilands, J. B., J. Biol. Chem., 221, 61 (1956). Sund, H., and Theorell, H., in P. D. Boyer, H. Lardy and K. Myrback (Editors), The engrmes, Vol. VII, Academic Press, Inc., New York, 1961, p. 25. Glinkina, M. W., and Finogenov, P. A., Biokhimya, 1_5_, 457 (1950). Gilbert, G. A., Proc. Royal Soc., A., 250, 377 (1959). Gould, E. S. , Mechanism and structure in organic chemistry, Henry Holt and Company, Inc. , 1959, p. 479. Kraut, J., and Jensen, L. H., Nature, 186, 798 (1960). ’5" ran —-- q if“ f—‘j ‘ h ”-4 r5 (.3 3 2 5 4 1 3 o 3 9 2 1 m .”