THE M-ECHANESM OF ACHON OF 2-KE'I'O-3- DEOXY~6—PHOSPHOGLUCONATE ALDOLASE OF PSEUDQMONAS FLUORESCENS Thesis for £419 Degree e§ Ph. D. MICHIGAN STA'E'E ‘JNW‘ERSEW Jordan M. Engram 1965 IHEQS LIBRARY Michigan Sta. University This is to certify that the thesis entitled The Mechanism of Action of 2-Keto-fi-Deoxy-6-Phosphogluconate Aldolase of Pseudomonas fluorescens presented by Jordan M. Ingram has been accepted towards fulfillment of the requirements for Ph-D. degree in Biochemistry Major professor Date October 6, 1965 0-169 till" “ ABSTRACT THE MECHANISM OF ACTION OF 2— KETO—B-DEOXY- 6-PHOSPHOGLUCONATE ALDOLASE OF PSEUDOMONAS FLUORESCENS by Jordan M. Ingram 6-Phospho-2—keto—3—deoxy-D-gluconate—D—glyceraldehyde- 3-phosphate lyase (KDPG-aldolase) was purified from glucose— grown Pseudomonas fluorescens. In the presence of sodium borohydride and pyruvate—luC an inactive, labeled enzyme derivative was obtained. Upon complete acid hydrolysis of the LiAlHu—reduced derivative, a new component identical to N6-a— (l—hydroxypropyl) lysine was isolated. This consti- tutes firm evidence that the aldolase forms an azomethine between the carbonyl carbon of the substrate and an enzyme bound lysine e—amino group. Upon treatment with 2, A-dinitrofluorobenzene, (FDNB), KDPG-aldolase is inactivated by reaction with four moles of FDNB per mole of enzyme. The reaction is prevented by phos- phate ion, KDPG, and partially by D, L-glyceraldehyde —3— phosphate. After complete acid hydrolysis of the treated enzyme, only e—dinitrOphenyl lysine was found. These results imply that four lysine Eramino groups function as a phos- phate binding sites lhe specificity for aZuYerhlne formation was studied by incubating various a—carbonyl compounds with the enzyme in the presence of sodium borohydride. Only pyruvate and a—ketobutyrate gave complete inactivation. Mo no. ydroxyac e— tone, a—ketoisovalerate, a-ketoglutarate, 5-Keto-A— A deexY€1Ut arat 3 2-ketO-u—hydroxyg1utarate, and 2-keto-j- (1‘ \II deoxygluconate inactivated between 2 and 75 per cent. Hydroxypyruvate, 2-ketog1uconate, and dihydroxyacetone gs ve ivaticn. These results suggested that the only 5 \) l'h I3 0.) ) C? P limitation for azomethine format on is the absence of a pclar func ion, 1.6. a hydroxyl grou T) , at the C3—position. a—Keto— butyrate did not bind to the aldolase previously rendered inactive with pyruvate and borohydride, and it compete y .. - . . 4s :avorably with pyruvate-l t for inactivation of the aide lase suggesting that both compounds bind at the same lysiin ne an presence 0: borcrydride and a- refirrityrate is the aldolase- alyzed tritium ex;.hange into a—ketobutyrate. The aldolase (T) 'atalvzed the exchange of one proton from tritiated water, presumably in a (I) (‘1 ereospecific manner, into a-ketobutyrate. CT The rate of exch nge wit a-Ketobutyrate was at least 37 En tim:s slower than the rate obtained with pyruvate. Neither hydroxypyruvate, nor a—ketoglutarate exchanged to this degree even after prolonged incubation. The results of these Jordan M. Ingram experiments imply that azomethine formation is a reaction distinct from exchange. The enzyme catalyzed the cleavage of 2-keto-A-hydroxy- glutarate and the B-decarboxylation of oxalacetate, but did not cleave 2-keto-3-deoxyg1uconate to any extent. These results together with the borohydride inactivation studies and the observation that FDNB-treated enzyme binds pyruvate suggests that the phosphate group of KDPG not only increases the binding efficiency of the substrate, but also aids catalysis by providing stress upon the electronically weak- ened C3 -04 bond. These results suggest that cleavage is not the direct result of azomethine formation. Hence, the com— plete aldolase reaction may be visualized as proceeding from azomethine formation to cleavage and exchange. A radioactive peptide containing the lysine-azomethine forming site was isolated and the amino acid composition of the peptide and the native protein was determined. THE MECHANISM OF ACTION OF 2—KETO-3-DEOXY— 6—PHCSPHOGLUflONATE ALDOLASE or PSEUDQMONAS FLUDRESCENS By L Ar Jordan M, Ingram A THES H U) Eibmitted t3 flichigan State UniverSity ‘r ;;rfi1al fuif:-lmert of the r quirements for the degree of Department of Biochemistry i965 k \ support ‘fl10‘” ACKNOWLEDGMENTS , 7 ~ ,.4 \ ,lA ; ,_ ,C ., i .A it i s a}, i is i A. f , ~ V_v‘ :rr ’ 7 A: ’, ' , :: ' 1L, Ami , Jordan M. Ingram was Ontario, Canada. He He v A in May 1956. in Agriculture from McGill Univers:ty, Montreal, He received the Master tution in May, 1961. antship in the Department University from September, Mr. Ingram is the American Asszciaticn for and the Canadian 5:c_e1" for and has one Child. i1 eceived the degree 19 aduate m w the American the Advancement ,; v 1_.C- obiologists. [,k degree from the same on August 2%, L9j6 in Ottawa, fvzm Glebe Collegiate, Ot.awa, of Ba:hel:r of Scien2e in May 1959, 1.13t1" research assist— 1 until the present time. Chemical Society, of Science, He is married TABLE OF CONTENTS ACKNOWLEDGMENTS VITA. LIST OF LIST OF Chapter I. II. III. TABLES FIGURES INTRODUCTION LITERATURE REVIEW. Occurrence of KDPG—Aldolase in Metabolism. Properties of the Purified Aldolase. Mechanism of the Aldol Reaction Biological Aldol Reactions. Biochemical Mechanism of the Aldol Reaction Nature of the Combining Site of Aldolases. The Role 0: Active Centers in Aldolase Bull/”[95 Physical Properties of Aldolase EnZymes METHODS AND MATERIALS Det fliinations and Procedures. Limited Digest of KDPG— Aldolase . Amino Ac;d Composition of KDPG— Aldolase Isoropic. Enzymatic Assay and Preparation of KDPG-Aldolase. iv {\, K) R) (A) LA) [v (2 f1 R.) 00 D-4 0 (VA) (EA) LI.) [\A) LA} I LI UL) ‘J Chapter IV. REFEFE RESULTS . . . . . . . . . . . . . Purification of KDPG— Aldolase. . . . . . Ac tive Sit es of KDPG— Aldolase. . . . . . Amino Acid Involved in Azomethine r,rmati3n. . . . . . . . . .C;,Vi.jlldfi oi KDPG—Aldolase by 2,4— Dinitroiiuorobenzene . . 2 . . . IidthLiILZQCIOfl of Amino Acid Involved in Dini-ropneryii-ion. . ina wLivation o1 KDPG— Aldolase by I, 5— Diiiuorodinitrobenzene . . . “:,nanisti; Studies of the KDPG—Aidoiase ogtaiyzed Reaction. . . . . . . 5.1u ti.e Bi hdi rg of o- -Keto Compounds in the Pres enoe of Sodium Borohydride P;-tcn Ex hange Reaction of a— Ketobutyrate :—De.arboxyiation of Oxalacetic Acid ~tics oi the Peptide Containing the idue Res po nsible for Azomethine "‘131’1. . . . . . a ,t e Sites of KDPG—Aidolase. . . . a: iati: SLudIES oi the KDPG—Aldolase ‘4 :ZEQ Rea;tit:n. . , -id Jlfpl;;Cl: n of Native KDPG— 1-1'Iuee and NO—a—Propionyi lysine Peptide. Page 36 q < 37 \j‘ C17 1‘ (I\ If] x \ Ii 71 Table 10. ll. l2. l3. LIST OF TABLES Purification of KDPG—aldolase from P. fluorescens . . . . . . . . . . . Characteristics of radioactive derivatives obtained by hydrolysis of KDPG-aldolase. Stoichiometry of fluorodinitrobenzene inactiva— tion of KDPG—aldolase. . . Stoichiometry of pyruvic-l—luC and fluorodini— tropenzene treatments to yield doubly labeled aldolase . . . . . . . . . . . Identification of dinitrophenyl amino acids obtained from KDPG-aldolase. . . . The reaction of l, 5— difluorodinitrobenzene with KDPG— aldolase. - . . . . Chromatographic characteristics of the product obtained after treatment of KDPG—aldolase with l,5—difluorodinitrobenzene . . Inactivation of KDPG—aldolase in the presence of sodium borohydride and a—carbonyl compounds The competition of a—ketobutyrate and pyruvate in azomethine formation . . . . . . Competition of a—ketobutyrate and pyruvate—l-luC for azomethine formation. Chromatographic characteristics of the IAC_ labeled peptide obtained from KDPG-aldolase Amino acid composition of luC—labeled peptide obtained by pepsin hydrolysis of KDPG- aldolase . . . . . . . . Amino acid composition of KDPG-aldolase vi Page 37 39 143 “9 51 53 EU 55 57 59 72 73 Figure LIST OF FIGURES Reaction of KDPG-aldolase With fluorodini— trobenzene and chlorodinitrobenzene Protection of KDPC—aldolase against dinitro— phenylation by potassium phOSphate and KDPG Sodium borohydride inactivation of KDPG—aldolase in the presence of pyruvate and a—ketobutyrate Exchange reactions catalyzed by KDPG—aldolase KDPG—aldolase catalyzed B—decarboxylation of oxalacetate . . . . Sephadex G—25 chromatography of pepsin digested KDPG—aldolase . . . . . . . . . Chemical sequence for the identification of the amino aCid involved in azomethine formation Reaction sequence catalyzed by KDPC—aldclase leading to exchange or cleavage . vii Page CHAPTER I INTRODUCTION Enzymes which catalyze aldol-type reactions are found in mammalian, microbial, and plant systems. The enzyme under present investigation, 2-keto-3-deoxy—é-phosphogluconate (KDPG-aldolase) 6—phospho-2-keto-3—deoxy—D-gluconate—D‘ glyceraldehyde-3-phosphate lyase (E. C. A.l.2.lu) is charac— teristic of microorganisms which metabolize substrates, such as glucose, via the Entner-Doudoroff pathway. With the crystallization of KDPG-aldolase from Pseudomonas fluorescens by Meloche and Wood (1), an aldolase with unique properties became available for study. Although the mechanism for aldol cleavage by this enzyme is similar to that reported for muscle fructose diphosphate (FDP) aldolase (2), the activity of the KDPG—aldolase is completely pre— served under acid conditions which inactivated FDP—aldolase by dissociation into subunits (3). Further, lack of agree- ment between molecular weights derived from sedimentation constants (assuming an average partial specific volume and spherical shape) and sedimentation equilibrium data indi— rated a rather large axial dissymmetry. From equilibrium sedimentation data the molecular weight was calculated to be approximately 90,000 (1). As with other aldolases, l incubation with sodium borohydride and the carbonyl substrate, pyruvate, caused inactivation. Based upon the amount of stably bound pyruvate-l-luc, one mole of pyruvate was bound per “3,500 g of aldolase (I). These data reveal the exist- ence of two active sites per aldolase molecule. Although the detailed reaction mechanism of KDPG— cleavage is known (I), there is little evidence available con— cerning the residue involved in the reductive binding of pyruvate-l“ C with borohydride. Also, information regarding other residues which might be important in other phases of the reaction is lacking. The investigations to be discussed were begun with the original intent of elucidating the residue involved in the binding of the carbonyl portion of pyruvate of KDPG. Methods were later developed to obtain a peptide fragment containing the residue responsible for carbonyl binding. In order to pursue the latter study, a modified procedure was deveIOped for the preparation, in good yield, of highly purified KDPG-aldolase. The amino acid com— position of the aldolase was determined to facilitate a com— parison of the structure of the peptide with that of the native protein. The reaction of a non-specific reagent under specified conditions was used to ascertain the existence and role of additional groups concerned with catalysis. The ability to inhibit enzyme action in the presence of such a reagent, 2, A-dinitrofluorobenzene, suggested the presence of an anionic binding site for the substrate phosphate group. Methods were developed to doubly label the enzyme with pyruvate-IMO and FDNB to eventually study the spatial relationship of these two important sitesc ' The reaction mechanism was clarified further by the development of a technique involving the borohydride reduc- tion reaction: The reaction of substrate analogs of pyruvate and KDPG with the aldolase in the presence of borohydride indicated the importance of the presence or absence of certain substrate group substituents in forming the azomethine with the enzyme. During the course of this investigation a novel B-decarboxylase activity of KDPG—aldolase was discoveredc The results of the forementioned experiments have been interpreted with a View towards the importance of the catalytically active centers of the enzyme and the requirements of substrate structure as each relate to the detailed reaction mechanism. CHAPTER II LITERATURE REVIEW Occurrence of KDPG-Aldolase in Metabolism Early studies on the growth of Pseudomonas saccharophila in the presence of glucose—l— l“C showed that two moles of pyruvic acid were obtained and that the pyruvate carboxyl was labeled to the extent of 50 per cent of that of the glucose— 1- 1“ca Thus, Entner and Doudoroff (N) postulated a new route for the metabolic utilization of glucose by this organ- ism. The apparent sequence of reactions was believed to be: D—glucose + glucose-6-PO,4 + 6-P gluconate + glyceraldehyde—3-P0u + <- [X] pyruvic acid In later reports, which appeared almost simultaneously, Macgee and Doudoroff (5) and Kovachevich and Wood (6) described the occurrence of a new phosphorylated derivative, 2-keto-3—deoxy-6-phosphogluconate (KDPG), in the reaction sequences of §.saccharophila and §.f1uorescens, respectively. It was definitively shown by the latter investigators that 6- P gluconate was converted to KDPG by a dehydrase enzyme and subsequently the KDPG was cleaved to D-G-3-P and pyruvate by the newly discovered KDPG-aldolase. The C to 1 C3 portion of KDPG thus became pyruvic acid: COOH COOH H-C=O | I l C=O C=O H-C-OH I I I KDPG + 3H2 ¢ i CH3 H2-C-OP H-C-OH ALDOLASE l H‘fi‘OH pyruvate D—G—3—P H2-C-OP KDPG In addition to the two organisms just described the aldolase has been postulated, or shown, to occur in other microorganisms as well. Gibbs and DeMoss (7) studying the anaerobic dissimilation of glucose by 3. lindnerii concluded that the formation of ethanol and carbon dioxide occurred by a mechanism analogous to the aerobic pathway of §.saccharophila. Later, Ashwell and co—workers (8) studying the metabolism of uronic acids by Escherichia coli showed that both glucuronic and galacturonic acids were dehydrated in the presence of specific dehydrases to yield a common intermediate, 2-keto-3- deoxygluconic acid. This latter intermediate was subse- quently phosphorylated by a specific kinase to yield KDPG. The KDPG was then cleaved by an aldolase presumably similar to that already reported to be present in Pseudomonas species. Prior to these studies, DeLey and Doudoroff (9) demonstrated that galactose was metabolized by §.saccharophila through a route similar to that for glucose. Galactose yielded KD-galactonate which was phosphonylated by a specific kinase and subsequently cleaved by an aldolase to yield D-G-3-P and pyruvic acid. The general pattern appears to be, therefore, that KDPG-aldolase is found in organisms which metabolize hexose sugars either by an initial phosphorylation, such as the case of glucose, to ultimately yield KDPG; or the enzyme also occurs in organisms which convert the hexose to an "onic" acid which is dehydrated to a 2—keto-3-deoxy-deriva- tive. This compound is phosphorylated and then cleaved by a KDPG—aldolase type enzyme. PrOperties of the Purified KDPG-Aldolase The first attempts to purify KDPG-aldolase were those of Kovachevich and Wood (6) with E. flugrescens as the source of the enzyme. The aldolase was purified about 20-fold by treating a crude extract with protamine sulfate, ammonium sulfate, and calcium phosphate gel absorption and elution. In this manner the aldolase was purified free of the 6-P gluconate dehydrase. In the presence of the aldolase, KDPG, excess NADH, and crystalline lactic acid dehydrogenase a decrease in the absorbency at 340 mu was recorded. This assay system was utilized as a measure of KDPG—aldolase activity. In the presence of the appropriate coupling enzyme systems and the aldolase, KDPG was converted quantitatively to pyruvate and D-G-3-P. No cofactor or metal requirements could be demonstrated for maximal activity. Later studies by Meloche and Wood (1) produced crystalline KDPG-aldolase. With these preparations the reaction was shown to be freely reversible in the presence of imidazole buffer. Mechanism of the Aldol Reaction Biological Aldol Reactions The aldol condensation classically involves the forma- tion of a carbon—carbon bond between a hydroxymethyl or methyl carbon and an aldehyde. Biologically this reaction has been found to occur in a wide variety of microbial and mammalian systems. The types of reactions which occur biologically may be clarified to some extent by summarizing the characteristics of some known aldol enzymes. The characteristics of KDPG—aldolase were described in the previous section. 1. Muscle FDP-Aldolase Fructose-l, 6—diPOu I D-glyceraldehyde-3-P0u + dihydroxyacetone—3-P0u This reaction was demonstrated in yeast and muscle by Warburg and co-workers (10, II). The yeast enzyme has since been shown to differ with respect to that of muscle in that it contains one mole of zinc per 60,000 gm of protein and also that the reaction is considerably stimu- lated by potassium ions. No cofactor requirements have been found for the muscle enzyme. 2. Liver FDP-Aldolase Fructose-l-POM 3 D-Glyceraldehyde + dihydroxyacetone -3-PO J: This enzyme was first implicated to play an important role in fructose metabolism by the liver (12). Peanasky and Lardy (13) crystallized the enzymefimmibovine liver and showed that the rate of cleavage of F—l-P was 32 per cent as the rate 01 FDP cleavage. The liver enzyme was later prepared by Butter and co-workers (IN) and shown to be unaffected by treatment with carboxypeptidase. On the other hand, muscle aldolase treated with carboxypeptidase yielded a preparation which cleaved F-l—P as well as the liver enzyme, but which cleaved FDP at a much slower rate. Based upon the above observations, these investigators concluded that muscle aldolase was an enzyme distinct from liver aldolase. 3- Transaldolase. Fructose-6-P + Erythrose-U-P : Sedeoheptulose-7-P + glyceraldehyde-3-P The enzyme transaldolase, which catalyzes the above reaction was discovered by Horecker and co-workers (15) and shown to use F-6—P or S—7-P as carbonyl donors. The enzyme differs from those mentioned previously because detectable cleavage does not occur in the absence of an aldehyde acceptor. In addition a net synthesis of F—6—P from dihydroxyacetone and glyceraldehyde-B-P occurs with difficulty indicating the inability of this enzyme to bind free dihydroxyacetone (16). A. 2-Deoxyribose—S-Phosphate aldolase 2-Deoxyribose-5—P I acetaldehyde + glyceraldehyde —3—P Deoxyribose-S-P aldolase was purified from crude ex- tracts of Lactobacillus plantarum by Pricer and Horecker (l7)° An interesting point of this reaction is the strict requirement for a dicarboxylic acid. A recent report by Hoffee, Rosen, and Horecker (18) describes the preparation of a crystalline zinc salt of DR-aldolase. 5. Mono- and Dicarboxylic Acid Aldolases Blumenthal and Fish (19) described the occurrence of an aldolase, present in §.cgli, which cleaves both 2-keto- 3-deoxyg1ucarate and 5—keto-A—deoxy-g1ucarate to yield pyruvate and tartronic semialdehyde. The enzyme is present in cells grown on D-glucarate or galactarate. A similar enzyme, present in D—arabinose-grown §.sacchar9phila, was described by Weimberg and Doudoroff (20). The enzyme cleaves D-arabonate to yield pyruvate and glycolaldehyde as Y 10 products. Maitra and Dekker (21) isolated a partially purified enzyme from rat liver which catalyzes the cleavage of 2-keto-N-hydroxyglutarate. The enzyme is apparently in- volved in the metabolic reactions of G—hydroxyglutamic acid. On the basis of amination of KH-glutarate, both erythro and threo G-OH glutamic acids resulted suggesting the ability of the aldolase to utilize each of the two possible stereoisomers of KH-glutarate. 6. Aldol Condensation Involving Activated Substrates Aldolase enzymes which require activated substrates have also been found in biological systems. It is suggested that these enzymes are involved in biosynthetic rather than degradative reactions. An enzyme involved in the synthesis of 2-keto-3-deoxy-D-araboheptonic-7-POu, an intermediate in the shikimic acid pathway, was purified and characterized from E.ggli by Srinivasan and Sprinsin (22). The enzyme utilized phosphoenol-pyruvic acid and erythrose -u-POu as substrates. The loss of the a-phospho group of PEP renders the reaction irreversible. A similar type enzyme was also described by Levin and Racker (23) which catalyzed an apparent aldol condensation of PEP and D-ribose-S-POu to yield KD-octanoic-B-POQ. However, further study indicated that D-R-S—POu was first converted to D-arabinose-S-POM which then served as the aldehyde acceptor. ll 7. Amino Acid Aldolases The isolation of a threonine and allothreonine aldolase from sheep liver has been described by Karasek and Greenberg (24). With threonine as substrate, the reaction rate is some twenty times slower than with allothreonine. The enzyme has a strict dependence upon the presence of pyridoxal phosphate, presumably to facilitate the formation of a Schiff base intermediate between the enzyme and the a-amino group of the substrate. The products of the cleavage reaction were identified as glycine and acetaldehyde. Biochemical Mechanism of the Aldol Reaction Mechanistic studies regarding dealdolization in model systems was supplied by Westheimer and Cohen (25). Using diacetone alcohol, these workers showed that catalysis by primary and secondary amines i.e. methylamine and dimethyla— mine yielded rate equations which differed from those obtained for classical base catalyzed reactions. In addition, it was pointed out that tertiary amines did not catalyze the reaction and the requirement for at least one nitrogen bound proton was implicated. These workers con— cluded that ketimine formation was most probably involved as an integral part of catalysis. Later studies by Spock and Forist (26) demonstrated dealdolization catalysis of diacetone alcohol by amino acids. Although glycine was not as effective as methylamine 12 an appreciable rate increase was observed. Glycine was superior to the other two amino acids, a and B—alanine, which were also tried. This observation Lead to a pro— posal that an amino group present in the aldolase could conceivably assist the dealdolization reaction. Early work by Rose and Rieder (27) with mastic aldolase showed that the enzyme catalyzed the exchange of tritium from tritiated water into dihydroxyacetone phos- phate. The rate of this exchange was decreased in the presence of the substate, FDP. It was postulated, therefore, that the enzyme bound the substrate in some fashior as to cause the displacement of a proton from DHAP. The enzyme- bound DHAP was displaced by neutralization with arother proton from the medium or by the acceptor aldehyde, D—G—3-P. Later studies by Topper, Mehler and Bloom (28) showed that in the presence of DHAP, muscle aldolase exhibited a new absorption peak at 235 + 250 mu. The peak did not appear on the addition of D—G-3-P to the aldolase preparation. These experiments suggested the formation of an enol—like bond between the substrate and the enzyme. Rose and Rieder (2) in a further study on the mechanism of muscle and yeast aldolases demonstrated that in tritiated water the FDP synthesized from DHAP and D-G-3—P was not labeled. This observation suggested that the proton incorporated, or lost, by DHAP was always in the same location and hence the reaction proceeded in a stereo- specific manner. A mono-tritiated species of DHAP could also be synthe- sized via the triosephosphate isomerase reaction. The product of this incorporation reaction did not lose radio- activity when incubated with muscle or yeast aldolase. Degradation of the FDP-aldolase labeled DHAP-H3 and the use of enzymes of known stereospecificity for proton exchange indicated that the triton gained in the aldolase reaction differed from that incorporated in the triosephos- phate isomerase reaction. Therefore the absolute stereospecificity for proton incorporation in the aldolase roaction proceeded as shown in the following sequence. T O a 2 a H - C - OH +——— T H - C - OH I I H H2O T 1“C into FDP was shown The rate of exchange of DHAP— to be more rapid than the rate of proton exchange. How- ever, the intimate association of the exchange reaction with the aldol reaction was established when it was demonstrated that as the levels of DHAP were increased the rate of tritium exchange approximated the rate of cleavage. There- fore, as the concentration of the enzyme-DHAP complex increases the concentration of the free enzyme becomes limiting in the reaction. The results indicated that aldolase forms an enzyme-DHAP complex in the presence of either DHAP or FDP and that the complex must later be dissociated to release free enzyme. 1H The concept of product dissociation was extended further by the studies of Venkataraman and Racker (16) with transaldolase. In the presence of enzyme and F—6-P a new component, later identified as the tansaldolase—DHA complex, was separated from the native enzyme by carboxymethyl cellulose column chromatography. Upon enzyme denaturation the bound radioactive DHA was released. The net synthesis of F-6—P from DHA and D—G—3—P was also observed indicating that the enzyme-DHA complex dissociated to some extent. Quoting unpublished experiments these investigators state that transaldolase catalyzes a stereospecific exchange of solvent protons into DHA in a manner analogous to the muscle aldolase reaction. The rate of exchange for transaldolase was not given and therefore a comparison with the FDP-aldolase reaction is not possible. In view of later experiments with transaldolase, this reviewer questions whether this exchange is significant. Pontremoli and associates (29) later described the preparation of crystalline transaldolase from Candida utilis. In the presence of F—6—PluC aldolase-DHA-luc complex was isolated by careful ammonium sulfate precipitation. The stabilized enzyme—bound-luC , a stable trans- 15 was transferred to erythrose -4-P or was dissociated from the enzyme by warming to 80°C. A further observation was the fact that in the presence of F-6-P, sodium borohydride and transaldolase, an inactive enzyme preparation was obtained. The mechanism of KDPG-aldolase was shown by Meloche and Wood (1) to be similar to muscle aldolase. However, KDPG-aldolase catalyzed a non-stereospecific exchange of solvent protons into all three methyl hydrogens of pyruvate. In addition, chemically synthesized B-ditritio KDPG, in the presence of the aldolase, did not decrease in specific activity indicating that cleavage was more rapid than the exchange reaction. The proposed reaction mechanism envisioned the formation of an enzyme bound pyruvyl anion which was neutralized by a proton from water, to release pyruvate, or neutralized by D-G-3-P to release KDPG. Nature of the Combining Site of Aldolases A major clarification concerning the combining site of aldolase enzymes was made by Pontremoli and associates (29), that in the presence of F-6-P and NaBHu, transaldolase was completely inactivated. In a similar study by Grazi, Cheng and Horecker (30) it was demonstrated that in the presence of DHAP-32P and NaBHu radioactivity was stably bound to muscle aldolase with an accompanied complete loss of enzyme activity. A later communication demonstrated 16 that NaBHu probably reduced an azomethine formed between the carbonyl carbon of the substrate and an amino group of the enzyme (31). The product obtained by complete acid hydrolysis of transaldolase or muscle aldolase following substrate inactivation was indistinguishable. When the radioactive derivative was subjected to periodate oxidation at alkaline pH, there was a release of radioactivity indicting the presence of vicinal hydroxyl or hydroxyl and amino groups. Following this treatment, a new compound, identified as lysine, was found. The chemical synthesis of authentic N6—8—glyoeryl lysine by Speck and co—workers (32) and the comparison of this compound with that obtained following acid hydrolysis of NaBHu —inactivated transaldolase or muscle aldolase positively identified the enzyme product as the stable secondary amine. The amine could arise only by reduc— tion of an azomethine formed between DHAP or DHA and muscle aldolase or transaldolase respectively: H H I I H—C—OH (P) H-C-OH (P) I I C=N—R NaBHu H—C—N—R l ' —.+ l H—C-OH H a H—C—OH H H These experiments indicated that an e-amino group of an enzyme lysine residue was responsible for azomethine forma— tion with the substrate. 17 Following the above experiments with muscle aldolase and transaldolase; Grazi, Meloche, Martinez, Wood, and Horecker (33) described the inactivation by NaBHu of both DR-aldolase and KDPG-aldolase in the presence of the respective substrates, acetaldehyde and pyruvate. These experiments constituted presumptive evidence for azomethine formation between a lysine c-amino group of the respective enzymes and the carbonyl carbon of the substrate. For KDPG aldolase it was calculated that one mole of l” C- pyruvate was bound per 50,000 gm of protein. The initial report by Grazi §£_al. (31) indicated that one mole of DHAP was stably bound per mole of FDP-aldolase. An independent study by Westhead, Butler, and Boyer (3A) showed by ultracentrifugation and equilibrium dialysis experiments that one mole of DHAP was bound per mole of aldolase. A more recent communication by Lai and collaborators (35), however, indicated that two moles of DHAP per mole of aldolase were stably bound in the presence of NaBHu. Similar studies to those reported above have recently been undertaken with DR-aldolase. N6—ethyl lysine was synthesized by Rosen, Hoffee, Horecker and Speck (36) and shown to be identical to the product obtained from a complete acid hydrolysis of DR-aldolase inactivated in the presence of 1A0 and NaBHu. Results with l“Cwacetaldehyde acetaldehyde— indicate that one mole of substrate is bound per mole of the aldolase (18). It appears that yeast aldolase is not 18 inactivated by NaBHu in either the presence or absence of substrate (37). On the basis of these data it was concluded that the mechanism of yeast aldolase differed in a signifi- cant manner from that of the other aldolases discussed thus far. Butter and co-workers (37) speculated that the divalent metal required for activity by yeast aldolase most probably facilitates a true enolization of the carbonyl portion of the substrate, The enolization would, therefore, facilitate the loss of a proton from the B-carbon of the substrate° Studies available on KDPG-aldolase to the present time are restricted to those of Meloche and Wood (1) concerning the tritium exchange reactions into pyruvate and the pre— liminary report by Grazi et_al. (33) that KDPG—aldolase most probably catalyzes azomethine formation between an enzyme bound amino group and the carbonyl carbon of the substrate, Data concerning the positive identification of this group are still lacking. The Role of Active Centers in Aldolase Enzymes Evidence concerning the participation of other active centers in the aldol reaction is much more limited than that available for the azomethine site, thus making generali- zations for aldolase enzymes more difficult. One study performed by Swenson and Boyer (38) showed that muscle aldolase was partially inhibited in the presence of l9 p—mercuribenzoate° However, the rates of reaction of various thiol groups varied depending upon the conditions employed. For example, up to ten thiol groups reacted without an appreciable loss of activity. However, in U.8 M urea, 28 thiol groups reacted resulting in an irreversible loss of activity. Reaction of the thiol groups was independent of the presence of substrates. A later report by Westhead, Butler and Boyer (3H) showed that reduction of muscle aldolase liberated only one thiol group rather than two as expected if a disulfide bond was cleaved. No experimentally substantiated conclusion was offered for this anomaly. In a further study Rowley, Tchola and Horecker (39) demonstrated that FDP-aldolase from muscle was inactivated in the presence of 2,4—dinitrofluoro— or 2,U—dinitrochloro- benzene. It appeared 5 to 6 residues were dinitrophenylated before inactivation of the enzyme. In a later report, Cremona, Kowal and Horecker (“0) showed that the consumption of one mole of ClDNB resulted in a three—fold stimulation of enzyme activity. The further uptake of two moles of ClDNB diminished the elevated activity to the original level. In all cases thiol groups of cysteine residues were involved in the reaction. Upon single reaction with ClDNB and treatment with mild alkali, a dehydroaldolase derivative was obtained. These workers suggested that the cysteine residue participating in the latter reaction may constitute an allosteric site for FDP—aldolase. Similar studies were 20 also reported by Rowley §t_al. for transaldolase. The enzyme was inactivated by the uptake of two moles of ClDNB per mole of enzyme. It was suggested that the reaction occurred at the e-amino groups of two lysine residues and that these groups possibly functioned as an anionic binding site. The latter conclusion was further substantiated when it was observed that phosphate ion prevented dinitrophenylation. The effect of other inactivating compounds on aldolase enzymes is limited to one report by Rosen, Hoffee, and Horecker (#1) with crystalline DR-aldolase. The absence of cysteine residues in this enzyme probably accounts for its extreme stability in the presence of thiol inactivating agents. The enzyme is susceptible, however, to inactivation by methylene blue and light, and diazo-p-nitroaniline suggesting the possible importance of tyrosine and histidine residues for catalysis. Similar studies to those discussed above are not avail- able for KDPG-aldolase. Therefore generalizations of reaction mechanism and enzyme site structure is difficult because of the lack of data. ghysical Properties of Aldolase Enzymes The elucidation of physical parameters of aldolase enzymes is limited to a few isolated studies. Reports by Deal, Rutter and van Holde (3) and Stellwagen and Schachman (“2) concerning the gross structure of muscle FDP-aldolase appeared simultaneously. Under mildly acidic conditions, dl muscle aldolase dissociates into three subunits of approximate equal size. The subunits are reassociated, in vitro, when neutralized to pH 5.5. A complete recovery of activity was not realized; however, physical properties such as S20,w, catalytic activity, and immunological response of the recovered enzyme were identical to the native protein. The studies also indicated the existence of an intermediate structure believed to be a trimer of the unfolded units. Dreschler, Boyer, and Kowalsky (M3) found that muscle FDP-aldolase was severely inhibited by treatment with carboxypeptidase. The loss of activity was associated with the loss of three carboxyl terminal tyrosine residues. The degraded aldolase was crystallized and shown to have physical properties identical with the native, unaltered enzyme. In this connection it was noted by Rutter and co—workers (14) that as the aldolase underwent carboxypeptidase degradation the rate of proton exchange into DHAP decreased to a greater extent than the rate of dealdolization, while with the native enzyme the rate of cleavage is more rapid than tritium exchange. Rose, O'Connell, and Mehler (MA) demonstrated later that under these conditions of degradation there was an alteration of the rate determining step. The rate determining step now became the proton neutralization of an enzyme-bound DHAP carbanion and the net result was reflected in a decreased rate of proton exchange. 22 A new concept concerning the physical structure of muscle FDP aldolase was advanced recently by Lai and co- workers'(u5). These investigators described the isolation of a 28 amino acid peptide containing the "active center" lysine residue which forms the azomethine linkage with the carbonyl carbon of DHAP. After treatment with DHAP and NaBHu, the derivatized protein was treated with trypsin to yield a number of peptide fragments, one of which was labeled. Through a series of purification procedures the peptide was isolated and its primary sequence deduced. The "active center" lysine residue is placed almost equidistant from the N and C-terminal ends of the peptide. To date the study of the physical characteristics of KDPG—aldolase have been minimal and limited to a few pre— liminary observations made by Meloche and Wood (1). The molecular weight of the crystalline enzyme is 90,000 as determined by equilibrium ultracentrifugation analysis and the stable reduction of pyruvate-lac. The enzyme displayed a very unique property of extreme stability towards 0.1 N HCl for at least four hours at room temperature. No evidence was found for molecular unfolding under these conditions and refolding upon neutralization. Preliminary data also indicated that upon treatment in guanidine— HCl, the enzyme was split into two subunits. No attempt has yet been made towards the isolation and characterization of the azomethine lysine peptide from KDPG-aldolase. CHAPTER III METHODS AND MATERIALS Bacteriological Pseudgmpnas fluorescens strain A 3. l2 Stanier was maintained on agar slants composed of 1.2% glucose, 0.6% (NHu)2HP0u, 0.3% KH2P0u, 0.05% MgSOu.7H20, 0.0005% FeCl3.6H20, and 0.5% yeast extract. The cultures were incubated at 30°C for growth. Large batches of cells were grown in a 10 gallon fermenter (New Brunswick Scientific) or a similar 30 gallon fermenter, on a medium as described above without the added agar. The glucose and magnesium salts were sterilized separately. Growth was from a 10% inoculum at 30°C until the cells Just reached the stationary phase. Chemical Materials Amino acid preparations, D, L-glyceraldehyde-3—P0u, protamine sulfate, a—ketoisovalerate, a-ketoglutarate, and dihydroxyacetone were obtained from the California Corp. for Biochemical Research. Sodium borohydride, and a-ketobutyrate were preparations of the Sigma Chemical Co. Hydroxypyruvate was obtained from Nutritional Biochemicals 23 "' 1"}W .. __. 2A and monohydroxyacetone from K and K Laboratories. 2,4- Dinitrofluorobenzene and 2,4-dinitrochlorobenzene were obtained from Distillation Products and lithium aluminum hydride from Metal Hydrides Incorporated. NE, NE, dinitrophenylene-bis-L-lysine was a gift from Dr. Peter Marfey Walker Biochemical Lab. Boston, Mass. 5-Keto-A- deoxygluconate was a gift from Dr. H. Blumenthal, Loyola University (Chicago) and 2-keto-N-hydroxyglutarate was a gift from Dr. E. Dekker, the University of Michigan. All ion exchange resins were obtained from The California Corporation for Biochemical Research. Sephadex was purchased from Pharmacia, Limited. KDP-gluconate was prepared by incubating a four-fold excess of pyruvic acid with D,L-g1ycera1dehyde—3-P04 and KDPG-aldolase in 0.1 M imidazole buffer, pH 8.0. (Specif- ically, the quantities used were: 11.2 mmoles of D—G-3—P0u, 22.A mmoles of Na pyruvate, and 5,800 units of the aldolase). The reaction was followed to completion by determining the disappearance of pyruvic acid on aliquots with a modified lactic dehydrogenase assay. Eight equivalents of barium acetate were added and the pH of the mixture adjusted to 3.5 with HCl. After the addition of four volumes of ethanol, the mixture was allowed to stand overnight in the cold and then was centrifuged. The precipitated barium salts were redissolved at pH3.5, reprecipitated with four volumes of ethanol, and centrifuged as before. The precipitate was 25 washed well with absolute ethanol, ether, and dried over P205 in vacuo. After two precipitations, KDP—gluconate was obtained in 90 per cent yield, with 4.5 per cent pyruvic acid content. The purity, without correction for moisture, was 67 per cent as Judged by an assay with KDPG aldolase (l) .KD—Gluconate was prepared by incubating KDP- gluconate with alkaline phosphatase (5). The reaction was followed to completion by measuring the disappearance of KDP-gluconate with KDPG-aldolase. The reaction mixture was then chromatographed on Dowex-l—Cl with a 0 to 0.05M HCl linear gradient. The semicarbazide positive (1) fractions were pooled and concentrated on a rotary evaporator in yagug. DL—Alaninol was synthesized by reducing DLealanine with lithium aluminum hydride as described by Vogl and P6hm (46). The alaninol was recovered by distillation at 68 to 70°C and 7 to 8 mm pressure. From 7.0 gm of DL-alanine, 3.03 gm of DL-alaninol was obtained as a colorless oil. The oxalate salt of DL-alaninol had a metling point of 138 to 140°C, uncorrected; the reported value is 141 to 142°C (46). 5-6-Bromobuty1hydantoin was prepared by the method of Gaudry (47). The yields for two separate preparations were 55 per cent and 60 per cent of theory rather than 80 per cent as reported. Authentic N6 -a-(1-hydroxypropy1) lysine was synthesized by reacting a ten-fold excess of DL-alaninol with 5-6—bromobuty1hydantoin under reflux in absolute ethanol for 40 hours as described by Speck et a1. 26 (32). The compound was purified in a similar manner except that the reaction mixture was deionzed by passage through Bio-Rad AGwll A 8 ion retardant resin. After adjusting the pH to 5.8 with HCl, the material crystallized from an ethanol—acetone-water mixture in the cold. The majority of the material was precipitited from solution by the addition of an excess of acetone. Both the crystals and the amorphous material were washed immediately with water-free ether and dried in vacuo over P205. The compound.was very hygroscopic and was stored over P205. The material was essentially homogeneous on paper chromatography in butano1:acetic acid: water (60:15:25 v/v), but some preparations revealed a second minor component which moved Just off the origin. In all likelihood this compound is the disubstituted N6—1ysine derivative. For the pure N6—a-(l-hydroxypropy1) lysine Calculated C, 41.87%; H, 8.90%; N, 10.84% Foundl c, 41.08%; H, 7.44%; N, 10.94% The compound had an indistinct melting point range of 185 to 188°C (uncorrected). lThe elemental analysis was performed by Spang Microanalytical Laboratory, Ann Arbor, Michigan. 27 All attempts to synthesize N6-a-propionyl lysine, the suspected product of borohydride reduction of KDPG-aldolase in the presence of pyruvate, by procedures similar to that described above for N6-a—(l—hydroxypropyl) lysine have failed. The reaction of alanine with 5—6-bromobutylhydantoin in ethanol probably failed because of the insolubility of the amino acid. However, no detectable reaction occurred when methyl alaninate was used instead of alanine. Another procedure involving the reaction of a-bromopropionic acid with N—a—carbobenzoxylysine in water yielded only lysine after the removal of the protective carbobenzoxy group. Authentic e-DNP lysine was prepared from the copper complex as described by Leggett-Bailey (48). All other DNP-amino acids used as standards were obtained from Mann Research Laboratories. Enzymes used as reagent chemicals were obtained as follows: crystalline lactic acid dehydrogenase, pepsin, and trypsin from Worthington BiOChemical Corporation, and alkaline phosphatase of calf mucosa from the Sigma Chemical Company. The c-glycerophosphate dehydrogenase-triose phosphate isomerase was a product of C. F. Boerhinger and Son. All radioactive materials were purchased from the Volk Radiochemical Company. 28 Determinations and Procedures Alpha—keto acid determinations were performed by the semicarbazide method of Macgee and Doudoroff (5). KDPG was also determined spectrophotometrically by a procedure similar to that used to assay KDPG-aldolase. The separation of a— keto acids from reaction mixtures was achieved by column chromatography on analytical Dowex-l—Cl columns with a linear 0 to 0.05 M HCl gradient as described by Meloche and Wood (1). Sodium borohydride reductions were performed by the pro— cedure of Grazi et_al. (33). The reaction mixture was made to a final volume of 1.0 m1 and sodium borohydride (0.01 ml of a l M solution) and acetic acid (0.005 ml of a 2 M solution) were added alternately every three minute interval for four additions each. The protein was precipitated by the addition of 600 mg of ammonium sulfate per m1 of reaction volume, centrifuged, redissolved and precipitated in a similar manner, this procedure was repeated twice. Lithium aluminum hydride reduction of the derivatized protein was performed as described by Leggett-Bailey (48). Following reaction of pyruvate-lac and sodium borohydride with the aldolase and the isolation of the inactive protein all of the available carboxyl groups of the enzyme were esterified and reduced with LiAlHu as follows. The deriva~ tized enzyme, 5.0 mg, was thoroughly dried over P205 and methanol and methanolic HCl were added to the solid residue 29 to give 3 ml of 0.1N HCl. The suspension was then shaken gently for 24 hours at 25°C. Eight ml of ether were added and the mixture allowed to stand in the cold for several hours. The precipitate was suspended in 5 ml of tetrahydro— furan (redistilled), 20 mg of LiAlHM were added and the mixture refluxed gently for 6 hr. Upon cooling, excess methanolic HCl was added to destroy the excess LiAlHu. After one hr, 20 ml of an acetone—ether mixture (1:1) were added and the precipitate which formed was removed by centrifugation. After drying over P the residue 205’ was suspended in constant boiling HCl and hydrolyzed at 110°C. The mixture, after hydrolysis, was deionized by passage through a column of Bio—Rad AG—ll A 8 ion retardant resin before further use. The radioactive peak was concen- trated prior to chromatography. Reactions with 2,4-dinitrof1uorobenzene were followed spectrophotometrically at 360 mu. The reactions, usually in 1.0 ml total volume, were performed with the aid of a Gilford modified Beckman DU spectrophotometer fitted with an auxillary dwell timer. After each time interval the optical density was recorded for a 30 sec period. A control was run in parallel to determine the optical density change in the absence of the enzyme. To determine the degree of dintro— phenylation the optical density of the control cuvette was subtracted from the value obtained in the presence of enzyme. The uptake of FDNB was calculated using an extinction coeffecient of 17,700 for e-DNP lysine (39). 30 Limited Digest of KDP-gluconate Aldolase KDP-gluconate aldolase, 40.4 mg of a 62% pure prepara- tion, was treated with pyruvate—l-luC and sodium borohydride as previously described. The protein was precipitated from solution by the addition of one—fifth volume of 50 per cent trichloroacetic acid. The precipitate was removed by cen- trifugation and washed with 5 per cent trichloroacetic acid. This washing was repeated twice. The mixture was resuSpehded in 0 02N HCl and 2 mg of pepsin dissolved in 0.02N HCl were added. The mixture was inzubated for four hours at 378C. After 35 minutes the suspension was solubilized. After completion of the digestion, the pepsin was removed by boiling and centrifugation. The hydrolyzate was adjusted to pH 6flland plazed on a Sephadex C—25 column (3 x 110 cm) which had previously been equilibrated with 0.05M NHQOH in deionized water. The hydrolyzate was eluted with 0.05M NH:CH at the rate of 40 ml per hour. Ten ml fractions were collected and tested for radioactivity and ninhydrin reactive material (48). The tubes containing radioactixity were combined and concentrated for further study. Complete protein hydrolysis was performed in evacuated, sealed Pyrex tubes in the presence of constant boiling HCl for 20 hours at ilC°C in an autoclave. Paper chromatography of all amino acids or peptides was carried out in glass tanks with various solvent systems; butanol: acetic acid: 31 water (60:15:25 v/v), butanol: pyridine: water (1:1:1v/v) or phenol:water:ammonia (100:20:l: v/v). Analytical paper chromatography was performed with Whatman 1 mm filter paper and preparative chromatography of peptide digests was done with Whatman 3 mm filter paper. Amino Acid Composition of KDPG-Aldolase Crystalline KDPG—aldolase used for amino acid compo— sition determinations was crystallized by the conventional procedure (1). Samples were hydrolysed in sealed, evacuated pyrex tubes in 6N HCl at 110°C for 24, 48, and 108 hours. Upon completion of hydrolysis, the HCl was removed on a rotary evaporator. Water was added and the mixture evapor- ated again. This procedure was repeated three times. The final material was dissolved in 0.1 N HCl and made to 1.0 ml of 12.5 per cent sucrose. Amino acid analyses were performed with a modified AutoTechnicon Amino Acid Analyzer. The conventional 1.5 cm flow cell was fitted into a Beckman model DU spectrophoto— meter fitted with a Gilford log converter. The optical density changes were recorded on a Sargent model MRA recorder. Readings were taken at 570 mu except in the region where proline eluted in which case the wavelength was changed manually to 440 mu. For most chromatograms 50 ug of hydrolysed protein sufficed with a recorder setting of 0.4 0.D. full scale. Norleucine was used as an internal standard to correct 32 for pump tube variations. The citrate buffers were identical to those described except that they were passed through a column of Dowex-50-Na+ after final adjustment of the pH to remove small traces of ammonia. All other aSpects of the equipment including the gradient were as described in the AutoTechnicon Manual. The area of each peak was quantitated by determining the half-height in terms of optical density units and calcu- lating the width at the half-height with the aid of micro- calipers. Conversion factors for each amino acid were obtained from chromatograms of standard mixtures. Cysteic acid determinations were performed as described by Moore (49) after treatment of the enzyme with performic acid. Tryptophan analyses were performed by the spectro- photometric method of Bencze and Schmid (50). Isotopic All quantitative radioisotope determinations, either tritium or carbon-14, were performed with the aid of a Packard Tri-Carb liquid scintillometer. The scintillation system of Kinard (51) was used for 1.0 ml of aqueous sample. The efficiency of counting and the degree of quenching was determined by adding standard radioactivity to a previously counted sample. Radioactivity on paper chromatograms was located with a Nuclear-Chicago strip counter generously supplied by Dr. R. L. Anderson. 33 Enzymatic Assay and Preparation of KDP— gluconate aldolase A coupled assay based on the oxidation of NADH as developed by Kovachevich and Wood (6) and modified by Meloche and Wood (1) was used for the estimation of KDP- gluconate aldolase activity. The assay is based on the following reaction sequence: KDPG D—glyceraldehyde _3'PO4 + pyruvic acid 2-keto-3—deoxy—6—phosphog1uconate aldolase Pyruvic acid + NADH.H+ —————+excess lactate + NAD lactic dehydrogenase The reaction was followed spectrophotometrically in silica glass cuvettes (b = 1cm) in a Gilford Modified Beckman DU spectrophotometer (52). One unit of activity is described as an absorbance change of 1.0 per minute in a total reaction volume of 0.15 ml. Specific activity is the number of units per mg of protein. Protein was determined by the 2802260 ratio method of Warburg and Christian (53). Crystalline enzyme was prepared by the method of Meloche and Wood (1). Usually, 200 gm of fermenter—grown cells were suspended in 400 m1 of pH 6.0, 0.1M phosphate buffer. The cells were subjected to sonic disruption in.alO-KC oscillator for 15 minutes. All debris was removed by centrifugation at 30,000 x g. The remainder of the purification procedure was similar to that described (1). 34 A more rapid procedure for enzyme purification was also developed to acquire large amounts of aldolase for exploratory experiments. The efficiency of this procedure is based upon the fact that the enzyme is stable in 0.l M HCl for at least four hr (1). Fifty ml portions of a cell suspension prepared as above were disrupted in a sonic oscillator as previously described. The cellular debris was removed by centrifugation at 30,000 x g. The supernatant was adJusted to 0.2 N HCl by the slow addition of 4 N HCl, with stirring, at 4°C and the precipitate removed by two successive centrifugations. The subsequent steps were carried out without neutralization of the acidic solution. Solid ammonium sulfate was added successively to 1.0M, 1 5M, 2.0M, and 3.0M concentration and the precipitate which formed at each stage was removed by centrifugation. The precipitate collected between 2.0M and 3.0M ammonium sulfate was dissolved in 20 ml of distilled water. Calcium phosphate gel (8 mg of gel per mg of protein) was added with stirring and collected by centrifugation. The activity was eluted from the gel with an equal volume of 0.2M phosphate buffer, pH 6.0. This rapid and efficient pro— cedure gave 75 to 95 per cent of the specific activity of crystals (1) with a yield of approximately 30 per cent. Recent preparations have not fractionated with ammonium sulfate as described above. Rather, the enzyme precipitated between 1.3 and 2.0M ammonium sulfate. By the careful addition of ammonium sulfate in a step—wise manner, the SpeQifiC activity of the In addition, crystalline aldolase crystallized by (l) was obtained through 35 enzyme after this step is maintained. enzyme characteristic of KDPG- the procedure of Meloche and Wood this rapid procedure. CHAPTER IV RESULTS Purification of KDPG—Aldolase Crystalline KDPG-aldolase used in critical experiments was prepared by the method of Meloche and Wood (1). The total recovery by this procedure was usually of the order of two or three per cent of the total units present initially. The procedure was slightly modified by concen- trating the enzyme solution, after dialysis, on a Swissco evaporator under reduced pressure rather than lyophilizing the dialyzate as described in the procedure of Meloche and Wood° A small amount of Dow Corning antifoam agent was used to prevent excessive foaming while concentrating. KDPG-aldolase used in exploratory experiments or used for the large scale synthesis of N6—propionyl—luC peptide was prepared by a modification of the procedure used to obtain crystalline material. The detailed procedure is described under Materials and Methods. It must be pointed out that the steps described must be worked through in a stepewhfiefashion to achieve the maximum yield and fold purification at each stage. A typical purification is shown in Table 1. This rapid and efficient procedure gave 75 to 95 per cent of the specific activity of crystals with a 30 per cent yield. 36 37 TABLE l.-—Purification of KDPG—aldolase from £.fluorescens. Total Unit§ Specific Recovery Step (X10 ) Activity (Per Cent) Fold Crude extract 433 23.7 100.0 1 Ammonium sulfate (2.0M to 3.0M) 303.6 4,060 70.0 172 Calcium phosphate gel (elution) 200.0 13,000 46.0 548 Active Sites of KDPG-Aldolase Amino Acid Involved in Azomethine Formation Radioactive KDPG—aldolase was prepared by reduction of the pyruvyl-enzyme complex with sodium borohydride in the following manner: KDPG—aldolase, 2.25 mg (specific activity 12,500) was dissolved in 0.05 M phosphate buffer, pH 6.0. Four umoles of pyruvate—l-luC (specific activity 2.44 ucuries per pmole) were added and the mixture allowed to incubate for 5 minutes at 4°C. Sodium borohydride (0.01 ml of a 1.0M solution) and 0.005 ml of 2 M acetic acid were added alter— nately at 3 minute intervals. After four additions of each, 99 per cent of the activity was lost. The radioactive derivative of the enzyme was precipitated by the addition of 600 mg of ammonium sulfate per ml of reaction volume. The precipitate was washed by resuspension in three successive portions of 5 per cent tricholoroacetic acid. An aliquot 38 of the final precipitate was dissolved in 1.0 ml of 0.75 N NaOH and an aliquot placed into scintillation fluid to determine the extent of binding of pyruvate-l-luC. The amount of stably bound 1“C was found to be 1.25 x 105 dpm per mg of protein. This corresponded to 1.96 moles of pyruvate-14C bound per mole of aldolase and is in accord with the value obtained by Grazi _e_t_:__a_1_.l (33). The remainder of the radioactive protein was suspended in constant boiling HCl and hydrolyzed in a sealed tube for 22 hours at llO°C.V After hydrolysis the excess HCl was removed by repeated lyophilization and the residue dissolved in a minimal amount of water. Samples of the hydrolyzate were chromatographed on paper in various solvent systems and chromatographed on Dowex-SO-Na+ at pH 5.28 (48) (Table 2). As shown in column three of Table 2, the radioactive product behaved differently than lysine in all tests. By analogy with the studies of Horecker et_al. (54) the sodium borohydride—reduced adduct should be N6 — a - propionyl lysine. However, because of the inability to snythesize this derivative of lysine, a direct comparison of the radioactive compound with authentic N6 - a — propionyl lysine was not possible. Accordingly, steps were taken to lspegific activity of pyruvate = 2.44 ucuries/pmole = 5.5 x 10 dpm/umole. Amount of aldolase treated = 2.25 mg = 0.026 umoles (based upon a m.w. of 87,000) Total dpm bound to aldolase = 2.82 x %@5 me per umole of aldolase = 10.8 x'lO 10 8 x lo6 uMoles of pyruvate bound per pmole of aldolase = "545‘§_l06 = 1.96 39 .mufi>flpomowpoh an Uouoooqn .zmhom cflpozzcflc mo Uopmooqw 0.00 o.ma 0.0m o.m© Atomsao mo fiev mm.m me .+az-om-xezoo mm.o ms.o mm.o mw.o Aauomnomv choEEmnaopmznaocwzm mfi.o swo.o ma.o :a.o Ammumauomv scams "oflom capoomuaocmpsm Acme teams ama<flq mopm< ama¢fiq chomom oCfimzfl Aahdood mocfimhqla anammemeOLzo Q Q ImeLUmmIHvlol z m ommfiooa< mo opmmmaoppzm mopmwcwom .omoaocflmuwmmm do mflmhaopohn mo UoCHMpoo mo>wpm>whoo o>Hpom0HomL mo moapmflsopomem:011.m mqmds 4O prepare the primary alcoholic derivative of the enzyme- pyruvate adduct (48) and to compare the fragment from hydrolysis with authentic N6 - a - (1—hydroxypropyl) lysine. Following reaction of pyruvate-l—luc and borohydride with 5.0 mg of aldolase and isolation of the inactive protein as described above, all of the available carboxyl groups of the enzyme were esterified and reduced with LiAlHl4 as described under Methods. The residue was hydrolyzed in constant boiling HCl and concentrated before further use. As shown in columns three and four of Table 2, the LiAlHl4 - reduced derivative migrated more rapidly on paper and more slowly on Dowex-SO—Na+ than did the radioactive derivative (presumably N6—a-propionyl lysine) isolated before treatment with LiAlHu. Further, the mobility did not correspond to that of authentic lysine. By comparison with column two of Table 2, it can be seen that the new derivative behaved identically to authentic N6 - a n (l-hydroxyprOpyl) lysine. The chemically synthesized lysine derivative, when treated with periodic acid at pH 8.6, as described by Horecker gt_al. (54), gave a positive test for formaldehyde in the chromotropic acid procedure (55). The lysine deriva— tive obtained by hydrolysis of the LiAlHu—reduced aldolase was treated with periodic acid and then with dimedon, as described by Horecker gtgal. (54).. The dimedon derivative was isolated by filtration and counted. Fifty per cent of the 4l radioactivity present in the lysine derivative after acid hydrolysis was recovered in the dimedon derivative. The radioactive derivative isolated after treating the aldolase with NaBHr (before LiAlH 4 4 to Dowex—50—Na+ at pH 5.28 as did the derivative obtained reduction) did not adhere after treatment with LiAlHu. This indicated an acidic or neutral character and a liklihood that the acidity of the carboxyl group of Néwrpropionyl lysine counteracted the basicity of the secondary c-amino group of this lysine derivative. However, after reduction with LiAlH,, the primary alcoholic function did not offset the basicity of the substituted c—amino group of lysine with the result that this more reduced and more basic derivative adhered to the resin. It continued to elute ahead of lysine, however. The identification of this derivative is considered defiri— tive proof that KDPG-aldolase forms an azomethine between an e—amino group of an enzyme-lysine residue and the carbonyl carbon of the substrates pyruvate, or KDPG. Inactivation of KDPG-Aldolase by 2,4- Dinitrofluorobenzene KDPG-aldolase was inactivated by FDNB or ClDNB when the reaction was carried out under strictly specified conditions. The enzyme preparation studied had been stored in 0.2 M phosphate buffer, pH 6.0, and was subsequently dialyzed against 0.025 M bicarbonate buffer, pH 9.2, prior to treatment with FDNB or ClDNB. Figure 1 shows the course FIGURE l.—-Reaction of KDPG—aldolase with fluorodinitro- benzene and chlorodinitrcbenzene. The time scale is expanded two—fold for ClDNB at pH 8.0 and the optical density scale was compressed ten-fold for FDNB at pH 9.2. The open circles relate to loss of enzyme activity and the closed circles to the change in Optical density. ach cuvette contained in 1.0 ml: 200 umoles of bicarbonate buffer épH 9.2) or imidazole buffer (pH 8.0), about 2.5x10‘ umoles of dialyzed enzyme for FDNB at pH 9.2 or 2.5x10‘3 umoles for the other reactions, and 0.5 Umoles of FDNB or ClDNB as indicated; Samples were withdrawn prior to each reading at 360 mu for the determination of KDPG—aldolase activity. O) O .b O 20 PERCENT OF ORIGINAL ACTIVITY DINITROPHENYLATION KDP-GLUCONATE ALDOLASE __1 T/~+ 0.l2 0F 40.10 0 d I PH-B-CL 0.08 EH 9.2 .BH 8.0 - T 0.06 pH 9.2 'pH 8.0 EH 8.0 0.04 l \ 0.02 TIME IN MINUTES OPTICAL DENSITY CHANGE - 360 mp 44 of dinitrophenylation and the parallel loss of enzyme activity. At pH 8.0, and in the presence of a lO-fold molar excess of FDNB, dinitrophenylation, as measured as increased absorbancy at 360 mu, proceeded until the time of complete inactivation and then ceased. These results suggested that only a few groups of the enzyme were dini- trophenylated with a corresponding inactivation of the enzyme. The stoichiometry of dinitrophenylation at pH 9.2 was investigated using the conditions described in Table 3. The reaction was allowed to proceed until 86.7% of the enzyme activity was lost and the absorbancy change at 360 mu was recorded. The absorbancy change corresponded to 3.50 moles of FDNB added per mole of aldolase. From this, it was calculated that for 100% inactivation, 4.08 moles of FDNB were required per mole of aldolase. This corres- f‘ ponds to 2.04 moles of FDNB consumed per mole of pyruvate-140 stably bound to the enzyme. At pH 8.0, the stoichiometry of dinitrophenylation was similar to that observed at pH 9.2. Preliminary experiments indicated that the dini- trophenylation reaction did not occur if the aldolase was dissolved in phosphate buffer. With a dialyzed (phos- phate-free) preparation, which was sensitive to FDNB as above, the addition of potassium phosphate to 0 01M concen— tration preserved enzyme activity (Figure 2); arylation was also inhibited to almost the same degree. Although in this 45 TABLE 3.--Stoichiometry of fluoriodinitrobenzene inactiva- tion of KDPG-aldolase. The reaction mixture contained: 4m umoles of crystalline aldolase, prepared by Dr. H. P. Meloche (l); 200 umoles of bicarbonate buffer, pH 9.2; 0.5 mmoles of FDNB and water to 1.0 ml. The uptake of FDNB was followed at 360 mu and the enzyme activity was assayed prior to each reading. All optical density changes were corrected for changes in a control without enzyme. At 86.7% inactivation, the optical density was determined and the optical density corresponding to 100% inactivation was calculated. Change in Optical Moles of DNP Bound a Inactivation Density at 360 mu per Mole of Aldolase 86.7% 0.250 3.50 100. % 0.288 4.08 __.;_ aBased upon a molecular weight of KDPG—aldolase of 87,000 (1). particular experiment with a low concentration of aldolase the consumption of FDNB appears not to have been completely inhibited by phosphate, subsequent experiments with higher phosphate concentrations have shown complete inhibition of arylation. As shown in the right-hand portion of Figure 2, at pH 8.0 (the optimum for activity), 0.02M KDPG preserved activity and inhibited arylation completely. 0.02M DL- Glyceraldehyde-B-Pou also afforded about 75% protection. Under similar conditions at pH 8.0, pyruvate did not prevent dinitrophenylation or protect against loss in activity. In a control at pH 8.0, the dinitrophenylation of lysine was not inhibited by the same concentration of phosphate ion or of 46 FIGURE 2.—-Protection of KDPG-aldolase against dinitro- phenylation by potassium phosphate and KDPG. The open circles relate to the loss of enzyme activity and the closed circles to the change in optical density at 360 mu. Each reaction cuvette contained in 1.0 ml: 200 umoles of imidazole buffer (pH 8.0), about 2.5x10“3 umoles of dialyzed enzyme, potassium phosphate, 0.01M, or KDPG, 0.02M as labeled, and 0.5 umoles of FDNB. Samples were withdrawn prior to each reading at 360 mu for the determination of KDPG—aldolase activity. 47 M 092 - 39NVH0 AiISNBO wouao A~00 00.0I 00.0 N_.0 - 00.52.: 2. 05:... 00 0¢ ON 0 00 0¢ 0N IIWIIJ! IIAV 1|...qu I i _ \Cl‘lfi ._ \ m OIIOIIIIOIIIOIIIIO 0.5200340 00x .2 .N0.0 wkzmzn_0mtz_o 0.);sz “.0 20_._._m_Iz_ 0 d 3 H m mm N I. m Om S S O f. a V O H. 09m l A 48 KDPG. At pH 9.2, under the same reaction conditions, KDPG did not prevent the arylation of KDPG-aldolase. The preceding experiments suggested strongly that FDNB reacted with groups specifically required for binding of the phosphate moiety of KDPG and that it did not react with the lysine residue involved in azomethine formation. The validity of this assumption was tested by following the dinitrophenylation of an aldolase preparation which had been treated with pyruvate-lac and NaBHl4 to render unavailable the lysine residue involved in azomethine forma- tion. In the first experiment, aldolase was fully inacti- vated by reaction with pyruvate-14C and borohydride as described above. After precipitation and dialysis, the aldolase was incubated at pH 9.2 with FDNB, as described. The results in Table 4 show that N6-a-pr0pionyl aldolase was capable of reacting with FDNB in the same manner and the same extent (4.30 moles of FDNB added per mole of enzyme) as the normal enzyme. Based upon the virtually complete dinitrophenylation of N6—o—propionyl aldolase, it is apparent that the amino acid residues involved in dinitrophenylation are intact after reduction of the azomethine, and hence, distinct from those involved in azomethine formation. In the second experiment, native aldolase was fully dinitrophenylated at pH 9.2 (4.08 moles of DNP per mole of aldolase) and then incubated with pyruvate-l-luc and .opCCHE Coo mCOHpmwouCHmHo mm commoooxo zpfl>fiuom0Hommm om.: oom.o mo.m o.mma Coapoamcooooauacflo mo oozofiaom oofinomzopoo moan o Ianopm>ohmm 3H oofimomsopon moan mo.: mmm.o no.0 eH.OH 02HIHutem>sese so emzoaaoe CoapmHmCoCQOCpfiCHQ oemucm 00 as com maanm mo moa x mpH>Hpoo pCoEuooCB oaoz poo oCsom oowno zpfimCoQ 0H0: Coo oCoom Iofiomm Hmooe 00 Coooo mzm mo moaoz Hoofipoo Hopoe opm>spmm mo wofioz ID .o>onm mm mzom Cows oopmosoCfl Con» oCm «m.m mo .Commso oumCooCmoHo z mmo.o meHmwm pcwflCCo>o commamfio mm: oEzNCo HzCOHQ0polmz one .opfinomnopoo Esfioom oCm AoHoE: Coo mofinoon m amufl>fiuom ofimfiooomv oxalalouo>ooho Cpfiz oopooap who; osmnCo mo onoE:E mm.n poop poooxo CoHHEHm 0C0; mpCoE . tumoop on» apCoEHCooxo oCooom on CH .oooCooo oCo UHCHM CofipmHHHpCHom CH oo>H0mmHo Coappoo m oCm mowz 2m~.o CH oo>aommfio mm: uoaaoo one .oHow owpoomozoanowhp mm Sufi: moeflp oonCp Cosmos oCm mammasm ECHCoEEm Cows oopoufiowoooo mos Campooo one .ooHCo lacooon Esfioom oCm AoHoEn moo woflmso: ::.m mow>wuom ownaooomv ovqatatopm>oomo mo moHoE: m Cows Uopmohp UCo «0.0 mg «Common opmnomono mo moHoE: OCH CH Uo>aowmfi© mos Campoaoumza one .oEHp oCooom m oopmpfiofioono oCm “00>Hommfiooo nopmmaom SCHCosew Cpfiz oopmpfiofloopo mm: Campooo one .mopCCHE om pom Doom um oopmosoCH who: HE o.H mo oECHo> m CH mzam mo moHoE: m.o oCm .m.m mo .Commoo omeonCooHo moHoE: com .oEmNCo go moaoene : .pCoEHCooxo pmofim 0C» CH .ommaooam ooaooma zaosoo came» on moCoEpmopo oConCooopuHCHUOCOCHm oCm oqasalofiom oH>3Czo mo mouoEOHCOHOmeI.3H mqmmson ommHoonlomom Soon uCoCOQEoo one .efie om emcee eoapesapoaea Rm .CHE om Copmm COHuo>HpomCH mo .eae om Cocoa Coeeasaeomea “mm .CHE om Locum oopm>fluooCH haopoHoEoo Cope; uofiom ofiuoom "HOCmpCC CH opommaoaohn ofiom Cm mo zCowaoumEomCo :om moHoE; om mo ooCowoCQ 0C0 CH o.m mo pm uCoEpooCB coax moaoe: om mo ooComoCo oCo CH o.m mo no oCoEpwoCe o.w mo um uCoEuwohB m.m mo pm pCoEpmoCB seaflooaaIemoe one: Coaseemm soaeommm pCoEumoCB .aCoEHCooxo Como CH com: one: ommaooam mo mpHC: ooo.m haoposfixoooom non» poooxo m manta CH Confinomoo omonp op HooHoCoofl who: mCOHpHUCoo Cofipooon 0C9 .ommaooHMIomox Cpfis oCoNCoooopHCHUOCosamflovm“H no CoauoooC oCBII.m mquB 54 TABLE 7.——Chromatographic characteristics of product obtained after treatment of KDPG-aldolase with 1,5- difluorodinitrobenzene. Compound Chromatography in Butanol:Acetic Acid:Water e—DNP lysine NE, N€'-dinitrophenylene— bis-lysine Product from aldolase hydrolysis Rf 0.71 0.33 0-35 Mechanistic Studies of the KDPG-Aldolase Catalyzed Reaction Reductive Binding of a-Keto Compounds in the Presence of Sodium Borohydride Since KDPG-aldolase is completely inactivated by incubation With pyruvate in the presence of borohydride (33), this test could be used to measure azomethine formation by other compounds. at 3 x 10'2 aldolase and NaBHu, specificity of this phase of the aldolase reaction. Accordingly, a—carbonyl compounds, usually M concentrations, were incubated with the as previously described, to measure the As shown in Table 8, a number of o-keto acids and monohydroxy- acetone were able to inactivate in the presence of NaBHu. Of the compounds tested, only pyruvate and a—keto-butyrate caused complete inactivation of aldolase, whereas a-keto- isovalerate, monohydroxyacetone, a-ketoglutarate, 55 TABLE 8.-—Inactivation of KDPG—aldolase in the presence of sodium borohydride and a—carbonyl compounds. Each reaction mixture contained substrate in the concentration indicated and 3,000 to 4,000 units of aldolase in a total volume of 1.0 ml. NaBHu (0.01 ml of a 1.0 M solution) was added alter— nately with 0.005 ml of 2 M acetic acid at three min inter— vals. Enzyme activity was determined before and after sodium borohydride reduction as previously described (33). Addition Concentration Cleavagea Inactivation Mx10_2 Per Cent Per Cent None 0 Pyruvic acid 0 4 --— 100 Hydroxpyruvic acid 3.0 —-— 0 Acetone 3.0 ___ 0 Dihydroxyacetone 3.0 ——_ 0 Monohydroxyacetone 3.0 -—- 25.7 aertobutyrate 1.0 --, 100 3.0 0 100 atKetoisovalerate 3.0 0 43.4 a—Ketoglutarate 3.0 0 62.0 2-Ketog1utonate 3.0 0 0 5—Keto—4- deoxyglucarate 3.0 0 66.3 2—Keto—4— hydroxyglutarate 3.0 0.1 68.0 2—Keto—3- deoxygluconate 2.0 <0.1 69.0 aExpressed as dash indicates that % of the rate for KDP-gluconate. A the test was not run. 56 5—keto-4-deoxyg1ucarate, 2—keto-4—hydroxyg1utarate, and 2—keto-3—deoxyg1uconate inactivated to the extent of 25 to 70 per cent. Hydroxypyruvate, acetone, dihydroxyacetone, and 2—ketog1uconate were without effect. In all cases, there was no loss of activity with NaBHu or with the carbonyl compound alone. Of the compounds tested, only 2-keto—4- hydroxyglutarate and KD—gluconate were cleaved (Table 8), as determined by the production of pyruvate. Activity with 2-keto—4-hydroxyg1utarate was approximately 0.1 per cent compared to KDP-gluconate. Activity with KD-gluconate was somewhat lower than 0.1 per cent, but upon exceedingly long incubation the formation of pyruvate was measurable. The inactivation of KDPG—aldolase in the presence of NaBHu and a-ketobutyrate suggested that this compound was stably bound to the active site lysine e—amino group normally occupied by pyruvate or KDP—gluconate. If this were true, aldolase inactivated with a—ketobutyrate and NaBHu should be incapable of binding pyruvate—l—luc. The results of Table 9 show that after treatment with a-ketobutyrate and NaBHu, only 2 per cent of the usual amount of pyruvate was bound. Thus, it would follow that if a—ketobutyrate and pyruvate were competing for the same site, the presence of o-ketobutyrate should depress the amount of lac-pyruvate which would be reductively bound by NaBH,4 when both were present. In a preliminary experiment no such depression was noted when a ten-fold excess of a—ketobutyrate was used. However, as 00.0 moH x m0.0 0.00 0mH 000.0 0 HIHIepmsstse 0 sh oezOHHoo oudpzponoquIu 00.H mOH x mm.m m.wm 00H 000.m 00HIHIopesstsm ommHooHo m“ oHoE moo moHoz we Coo Eon R Copm< oCOHom COHuMLoohooCH oanHIopm>dhhm. COHum>HpooCH mpHCb oEzNCm . COHpHvo< .oopCooo posoHHo Co oCm Cope: mo HE o.H CH ©o>HommHo ImHo mo: oompHoHooho HMCHH oCB .hopmz CH oo>HomchoC oCm opmHHCm ECHCoEEm Ho HE Coo me 000 Csz moEHu oops» popmpHoHooho mos CHopOCQ on mono Como CH .HHm50H>oCo oooHComoo mo coshomuoo one: mCOHposooC oCH .oopoOHUCH mm .AoHoE: Coo moHCSo: m .HpH>Hpom OHHHooomv ozHHIopm>CCHQ Ho oHoE: H Co ouoCHpCCOooxIe 00 moHoE: om 00.0mo «Consn opMCowOCQ Ho moHoE: oom mommHooHoIomQM ”ooCHmpCoo COHpomoL HEo.H Comm .COHmeCoH oCHCuoEonm CH oum>ChHQ oCm oCMCHpCQOCoxis HOCOHpHpoQEoooCHII.m mqmshzm 0H m0.H 00H x 0.m 00H 0 000.0 00HVoem>0esm ommHoon Ho oHoE we ComIzmQ N nopm< oCOHom mod moHoz COHpmhoohooCH o :Hlopm>oozm COpr>HpooCH mpHCD oEHNCm COHuHU©¢ .ConHmomoo HHm50H>oCQ mm onoz,mCOHumpHQHooCQ CHopopd CCm mCOHuosooh 0C9 .AoHoE: Com moHCSoa m nHui/Hoot oHMHooomv o IHIopo>CCHQ mo onoE: m.o abouMOHUCH ohonz omeHCCCOCoxle mo moHoEa om mmHo mm ”Comwsn opMCQmOCQ mo moHoE: oom mommHoonIGmQM "ooCHopCoo oCCprE COHpomoC HE o.H Comm .COHmeCOH oCHCpoEoum Com ozHIHIopo>CCHQ UCm opmnzpsooooxIe mo COHoHpooEooII.OH mqm._. 300.55. I .0 Iv / sz I wz_m»4I IT _ Im I00_0 _ _ _ exmoz _ _ m I/I/ onxIIIII. o- o 1. N12. 0- 02600- INzL///// G zooo _ _ _ _ m0<4004< m._.<2003I_0I max “.0 20_.r<>_._.0FDmOwa I 0f . _ O u 0 _ ; LIIIIIIIIIIII . O \ wk<>3m>a _O 0 v / \ 0041.001? 0._.4200340Idox >0 owN>I_<._I<0 m0zNz0 III I 002000-202 01. 002000-202 010 0 .10 290.004 00 290.090 .3 I/ I 0 II \I /0 / I /.. < I/._ _ /. IIIII I. WV/“r fiV/ I 02020 . .. 002000 IV1002 it? 05520 00.80 0108/. 02>Nzw 032:2 mFmIfiEOo ./ ms;sz 032:2 ,m._.mI_ _n=200 I I _ I I I . I 01002 030 0250 I I I >a I I 0 22>sz 00.350 I 0:203 00358003000 20 003000,... 022830-03 002000 .00 000.100 v.0 AJJSNEICI "IVOlidO N._ 0._ M 0172 - 0.N 68 peptide mixture was placed on a Sephadex G-25 column previously equilabrated with 0.05M NHMOH. The column was eluted with the same concentration of NHMOH at a flow rate of 40 ml per hr. Seven m1 fractions were collected and assayed for lu—C content and ninhydrin reactivity (48). The elution pattern of a typical chromatogram is shown in Fig. 6. In this particular chromatogram, 2.2 x 105 cpm were placed on the column and 2.4 x 105 cpm were recovered. 0f the two peaks containing radioactivity there was approximately ten times more lu-C in peak one than peak two. The fractions of peak one were pooled and concentrated in vaggg and then chromatographed in the solvent systems described in Table 11. The migratory characteristics of the labeled peptide in both solvent systems indicate that it moved more rapidly than the majority of the ninhydrin reactive TABLE ll.——Chromatographic characteristics of the lLIC-labeled peptide obtained from KDPG—aldolase. The aldolase, 2.5 mg (specific activity, 10,000) was reduced with NaBH in the presence of pyruvate—l—l C The protein was precipitated and washed three times with 5 per cent trichloroacetic acid, re- suspended in 0.01 N HCl and hydrolyzed with 0.1 mg of pepsin for one hour at 37°C. The digest was eluted from Sephadex G-25 as described above. Solvent System Rf Value Butanolzacetic acid:water (60:15:25) 0'81 P : henol water 0.86 (100:20) FIGURE 6.--Sephadex G—25 chromatography of pepsin digested I KDPG—aldolase. V mum—232 20_I_I0II u m 00000 00:00: 5:00.00 02.000300000006000:109-02-... 0-:z-o-: oIEyIv 0-:z-o-: % :ooo-:oIm0~:00IN:0-I~iz-o-: m . . N . 4 . m :oooso :o :30 :o :30 .: :Om:o 02:20 00002800 02000202 0030004 2. 02.05 053000 0:0 0:0 :z m:o 0-:z-oI: I 0-2.0 TI I:z-wI:o-mI~:oIIN:oIm:z + one . o . :oozo :002 5900 o :ooso 77 native aldolase covalently binds pyruvate or KDPG. These results are in accord with those obtained with muscle FDP- aldolase, transaldolase, and DR-aldolase (5H, 36). Treatment of KDPG-aldolase with FDNB, as described for ribonuclease by Hirs (58) and for transaldolase by Rowley et_al. (39) caused complete inactivation and yielded exclusively e-DNP lysine upon hydrolysis. No stimulation of activity was observed as for FDP-aldolase (40) and no evidence for S—DNP cysteine was obtained. The arylation reaction was prevented by inorganic phosphate, KDPG, and DL-glyceraldehyde-3—P0u. Since no protection was afforded by pyruvate, another substrate for the aldolase reaction, it is concluded that these lysine residues are concerned with binding the phosphate group of KDPG. The same amount of dinitrophenylation occurred with aldolase previously rendered completely inactive by reductive binding with borohydride of two moles of pyruvate per mole of aldolase. That the lysine residues concerned with the dinitrophenyla— tion are located relatively near the lysine involved in azomethine formation is indicated by the fact that after dinitrophenylation of four lysine residues per mole of aldolase, reductive binding of pyruvate-lac with borohydride never exceeded 50 per cent of the value obtained with the native enzyme, presumably due to steric effects of the dinitrophenyl groups. It is considered unlikely that some arylation of the lysine residue involved in azomethine 78 formation had occurred because the stoichiometry and kinetics of arylation are the same with both the native and N6-a— propionyl enzyme. The fact that dinitrophenylated enzyme is capable of binding pyruvate—luC, although having lost the ability to catalyze the cleavage reaction indicates that, by this means, cleavage and binding of pyruvate may be artificially separated into distinct reactions. A similar separation was noted by Rutter, Richards, and Woodfin (1“) with muscle aldolase. Upon treatment with carboxypeptidase, the cleavage reaction decreased 20-fold, whereas the exchange reaction between dihydroxyacetone phosphate and water pro— tons decreased SOO—fold. On this basis cleavage must be distinct from exchange. The formation of NE, Ne'-dinitrophenylene-bis-lysine upon treatment of the aldolase with 1,5—difluorodinitroben— zene shows that reactive lysine groups are sufficiently close together so that two arylation reactions are possible from a single reagent as follows. First, the diFDNB may react with one lysine residue involved in the phosphate binding site and thereafter with a nearby lysine residue not specifically involved in catalysis. The reverse seems unlikely because mono- FDNB reacts only with substrate protectable lysine residues. Second, the reagent may react with both lysine residues producing the anionic binding site by either of two mechanisms. For either of the two latter possiblities, it must be assumed that the lysine 79 residues are sufficiently together so as to successfully bind the relatively small phosphate moiety. A bridged NE, NE'—dinitrophenylene-bis—lysine would result if the lysine residues were in close proximity by virtue of their situa- tion in the primary sequence, that is, separated by only a few amino acids within a linear portion of a peptide chain. Also, the bridged derivative would be formed if the two e-amino groups of the lysine residues were ideally situated by virtue of their spatial orientation while actually residing on two chains or on unrelated portions of the same chain. If the latter were the case an unlimited number of amino acid residues might separate the important groups. The reaction postulated to occur is illustrated below with- out regard for the spatial relationships of the groups involved. h F \ N02 S’€,w%2 XES’EJS , E<1§J + F - NO E/ + 2H}? ‘ZJ’S~ 2 \‘ZJ’S / E‘NH 02 “W 8 1 NO ,9 2 An interesting speculation arising from the latter two possibilities discussed above concerns the distance which actually separates the residues. The isolation of a pure peptide from KDPG-aldolase containing the bridged double derivative might answer whether the groups are closely related in primary sequence, or closely related due to enzyme conformation 80 To function as a phosphate binding site, the lysine residue involved must be protonated because at the optimum pH for the reaction the phosphate group of KDPG exists as a dianion. However, at the same pH, the e-amino group of the lysine residue involved in azomethine formation must be unprotonated because it is believed that a non— protonated amino group is required for azomethine forma- tion (59). It is possible, due to local inductive effects, that at the same pH, an e—amino group involved in azomethine formation can exist in the non—protonated form, whereas two others concerned with binding can exist in protonated forms. Further, in a pH region near the pK of the lysine e-amino group, interconversion of the protonated and non-protonated forms by mass action would be predicted when one of these is removed by reaction or binding. Mechanistic Studies of the KDPG—Aldolase Catalyzed Reaction Previous studies concerning the reaction mechanisms of FDP—aldolase, DR—aldolase, and KDPG-aldolase have demon- strated the importance of Schiff base formation. The reaction occurs between the carbonyl carbon of the substrate and an e-amino group of an enzyme lysine residue. Whether Schiff base formation completely accounts for the detailed reaction mechanism is a problem of current research. Pre— sumably as a result of Schiff base formation FDP—aldolase catalyzes a stereospecific proton exchange between dihydroxy— acetone phosphate and the medium (2) while DR-aldolase and 8l KDPG-aldolase (Al, 1) catalyze non-specific exchanges of solvent protons with all three methyl hydrogens of acetalde- hyde and pyruvate, respectively. The present investigation demonstrates that KDPG—aldolase also exchanges one hydrogen, presumably the methylene hydrogen, of a-ketobutyrate with the protons of water. Thus, a degree of assymetry, as com- pared to pyruvate, is introduced by the additional methyl group of a-ketobutyrate. During the present investigation, ‘ Rosen and co-workers (Ml) reported that DR-aldolase also E exchanges, stereospecifically, a single methylene hydrogen of propionaldehyde with water. An important observation in the present investigation is the fact that a-ketobutyrate exchanges solvent protons at a rate at least 37-times slower than pyruvate (or 5,000—fold slower based upon the total exchange capacity of the aldolase). The point may be argued that perhaps this is a Km effect because the Km of pyruvate and a—ketobutyrate with KDPG— aldolase is not known. The results of the borohydride inactivation experiment (Fig. 3) in the presence of various concentrations of pyruvate or a—ketobutyrate suggest that the Km for a-ketobutyrate is not significantly different than that of the pyruvate since an essentially similar inactiva- tion occurred with the same levels of each substrate. Complete inactivation with pyruvate was achieved at a concentration of 3 pmoles per ml and at 10 pmoles per ml with a-ketobutyrate. The manner in which the reduction was 82 performed is based upon the assumption that the enzyme must be fully charged with substrate prior to the addition of NaBH“. As a result of these considerations it is assumed that the concentration of a-ketobutyrate used in the exchange reactions (20 umoles/ml) was saturating and there— fore its binding (reaction K Fig. 8) was not a limiting 1’ factor. Therefore, the decreased rate of proton exchange into a-ketobutyrate must be due to a slower release of the methylene hydrogen, or a slower incorporation of a triton of Fig. 8. On a statisti- as shown in reactions K and K_ 2 2 cal basis one would expect that the rate of exchange into a-ketobutyrate would be three times slower than the rate with pyruvate since the three methyl hydrogens of pyruvate are essentially equal. It is obvious that this value does not account for the discrepancy found for the exchange rates of a—ketobutyrate and pyruvate. The above experiments suggest that the degree and rate of azomethine formation for pyruvate and a-ketobutyrate is similar but the rate of proton exchange is slower with a—ketobutyrate. Since the proton exchange reaction is slower than azomethine formation then it must be a process distinct from it. This situation is analogous to that observed for carboxypeptidase treated FDP-aldolase (14) wherein the rate of dealdolization is decreased but the rate of proton exchange suffers a considerably greater attentuation. The explanation offered is that following carboxypeptidase 83 FIGURE 8.——Reaction sequence catalyzed by KDPGvaldolase leading to exchange or cleavage. :00-m-0-0 co 0:0 ;lu : m0z<:0xm 00000-: 00<0000<-000x :0 054 uz=2_ON< msgmmm H.025»:— >xomouo._.mx m m m m 0:: 1.0.: 1.0.: J- : mw _ TX N _ sz-2no TII NZmTZuo Tnllllv sz-z I + Ono _ N: L _ _x _ 1000 1000 1000 09340.6 0-x: 0: : :-0-: N:2.sz + 0.0 _ Iooo 85 treatment the rate of proton neutralization becomes the rate determining step in the cleavage of FDP. Since the making, or breaking, of the C—H bond is rate determining then a pronounced isotope effect is revealed by the drastic rate decrease of the tritium exchange reaction. Therefore, the rate of proton neutralization of the enzyme-bound DHAP carbanion is slower than azomethine formation, as compared to the native enzyme, and hence exchange must be a reaction distinct from azomethine formation. From the reported experiments with KDPG-aldolase, it has now been demonstrated that the rate of proton exchange may be decreased relative to the rate of azomethine formation by substituting a—ketobutyrate for pyruvate. Thus, azomethine formation can be visualized as a process distinct from the exchange reaction (Fig. 8); that is azomethine formation is independent of exchange. The fact that KD—gluconate binds to the enzyme, as demonstrated by borohydride inactivation (Table 8), and that the cleavage of this compound is greatly decreased relative to KDPG, indicates the important role of the substrate phosphate group in cleavage as well as in binding (60). The importance of the substrate phosphate group has also been discussed in the previous section concerning the inactivation of the enzyme in the presence of FDNB. Although a direct determination of the Km of KDG is not possible an 86 approximate value may be arrived at. It has been shown by Hartman and Barker (60) that the binding due to the phosphate groups on the substrate, FDP, for muscle FDP- aldolase could account for the Km value of FDP. The Km values for FDP—aldolase and KDPG-aldolase for FDP and KDPG are essentially equal at 6 x 10.5 M and 10 x 10-5 M respectively (37,6). The Km value for dihydroxyacetone phosphate is 2 x 10‘3 M and, while not determined rigor- ously, the value for pyruvate deduced from the effect of pyruvate concentration on borohydride inactivation is approximately 1 x lO_3 M. On this basis it is assumed that the Km for KDG would approximate the Km value of F-l-POu for FDP—aldolase which is l x lO_2 M (37). While the production of pyruvate was measurable after prolonged incu- bation of KDG with the aldolase, there was no appreciable rate as determined spectrophotometrically, although the level of substrate was at least lO—fold greater than the suspected Km value. Since KDG is bound to the aldolase in the presence of borohydride, but exhibits no appreciable cleavage, it is concluded that the phosphate group of KDPG contributes more than increased binding capacity and that cleavage is not a direct result of azomethine formation. Further support for this latter view is the fact that FDNB reacts with the e—amino groups of four lysine residues which function as a phosphate binding site. With arylation of these residues the enzyme lost the ability to cleave d- aim F... I. 87 KDPG while the ability to form an azomethine with pyruvate was readily apparent (Table 5). In this connection the B-decarboxylation of oxalace— tate is believed to be the direct result of Schiff base formation between the substrate and the enzyme with the subsequent formation of the eneamine species (Fig. 8). This would result in a decreased electron density about the C3 - Cu bond and result in B-decarboxylation presum- ably in the absence of additional forces supplied by the enzyme. These data also suggest that carbon dioxide is an excellent leaving group subsequent to azomethine formation. As demonstrated with KDG, cleavage is not the direct result of azomethine formation. It is proposed that the phosphate group of KDPG, in addition to promoting binding, may aid catalysis by imposing stress upon the 03 - C)4 bond of KDPG by inducing an enzyme conformation change or by substrate stretching. The possibility that the A—carboxyl group of 2-keto-M-hydroxyglutarate binds in the same fashion as the phosphate group of KDPG may explain the reasonable rate of cleavage of this compound. A separation of the various phases of the KDPG- aldolase catalyzed reaction is possible through the use of substrate analogs. The results obtained with a-ketobutyrate suggest that binding and azomethine forma- tion are distinct from proton removal or neutralization. 88 The results with KDG suggest that cleavage is not a direct result of azomethine formation and that the C6—phosphate group of KDPG contributes more to catalysis than increased binding efficiency. A reaction mechanism based upon these results is shown in Fig. 8. Whether any or all of these steps are enzyme catalyzed has not been determined rigor- ously. A recent report by Meloche (6i) suggests that bromopyruvate reacts with a catalytically active base within the active site of the aldolase. The evidence suggests, therefore, that the proton exchange step may be enzyme catalyzed. The evidence that KDPG—aldolase pro- ceeds via Schiff base formation is further clarified by the discovery of oxalacetate decarboxylase activity. Rutter and co—workers (37) predicted that Class I aldolases may structurally resemble pre-existing enzyme forms, such as a B—decarboxylase. This is the first demonstration of the fa:t that an aldolase does have B—decarboxylase activity. It is evident from the variety of substrates which inactivate KDPG-aldolase in the presence of NaBH“, that the specificity for azomethine formation is not high. How— ever, the substitution of a polar function, such as a hydroxyl group on the C3 position, prevents azomethine formation. Steric hindrance on the C3 position of pyruvate must be ruled out as an inhibitor of azomethine formation since the addition of a methyl group rather than a hydroxyl results in a pronounced inactivation in the presence of borohydride. 89 Substrates which maintain non-polarity at the C of pyruvate, 3 such as a-ketobutyrate, a—ketoisovalerate, a-ketoglutarate, 2—keto-H—hydroxyglutarate, and 2—keto—3—deoxygluconate inacti- vated in the presence of NaBH“. The above observations clearly indicate that a non-polar C position is an essen- 3 tial requirement for the binding of a substrate. Perhaps of relevance in this connection is the isolation of the 30 amino acid peptide from the active site region of KDPG- aldolase which was shown to contain approximately 65 per cent non-polar amino acids. If the non-polar amino acids are concentrated in the azomethine region a pronounced exclusion of protons necessary to neutralize an enzyme-bound carbanion may well explain the decreased rate of proton exchange into a-ketobutyrate. For example, the addition of a nonnpolar methyl group could conceivably prohibit the necessary unmasking of the site, thereby preventing the rapid entrance of solvent protons needed for the neutralization of the carbanion. The fact that other compounds, such as a—ketoglu— tarate, and 2-keto-u-hydroxyglutarate also inhibit the aldolase in the presence of borohydride indicates that hydroxyl groups and a ring configuration are probably not required for azomethine formation. In the case of FDP- aldolase, however, hydroxyl groups at carbons 3 and A may play an important role (60). In contrast, for KDPG—aldolase, the presence of a hydroxyl group at the C position of the 3 90 substrate or substrate analog renders the compound inactive for azomethine formation. This is seen from the complete inability of 2-ketogluccnate, hydroxypyruvate and dihydroxy- acetone to form azomethines as measured by borohydride inactivation. Amino Acid Compogition of Native KDPG- Aldolase and Nb—a—Propion llysine F I | | Peptide The amino acid analysis of native KDPG—aldolase indicates a molecular weight of 88,348 as compared to 87,000 determined by pyruvate—lac binding and 89,500 determined by equilibrium ultracentrifugation (l). The data also show low values for tyrosine, histidine, and methionine residues. The absence of free thiol groups, viz cysteine, may explain a preliminary observation that KDPG—aldolase is extremely resistant to thiol inactivating agents such as p-mercuribenzoate. The amino acid composition offers no obvious explanation for the extreme acid stability of KDPG— aldolase. The amino acid composition of the N6-a—propionyl lysine peptide reflects to some degree the overall amino acid composition of the native protein. The marked absence of tyrosine, methionine, and cystine residues is indicative of the low values for these residues found in the native molecule. ""' 2‘5“ CHAPTER VI SUMMARY The formation of an azomethine invoIVing an e—amino group of lysine and pyruvate—l-luC was established by chemi- cal treatment of the aldolase and by hydrolysis to yield a radioactive component whose properties were identical to authentic N6—a—(l-hydroxypropyl) lysine. Four moles of fluorodinitrobenzene reacted with the aldolase with accom- panying inactivation. The reaction was inhibited by phos- phate, 2—keto-3—deoxy—6-phosphogluconate, and DL—glyceralde- hyde-3-phosphate, but not by pyruvate. Only e-dinitrophenyl lysine was found upon complete acid hydrolysis. The same amount of dinitrophenylation occurred with aldolase already inactivated by reductive binding with borohydride of two moles of pyruvate per mole of enzyme. These experiments constitute evidence for a role of three lysine residues per active site in the action of KDPG—aldolase The Specificity for azomethine formation was studied by incubating the aldolase in the presence of a—carbonyl compounds and sodium borohydride. Both pyruvate and d-ketobutyrate pro— duced complete inactivation whereas monohydroxyacetone, d-keto— isovalerate, a-ketoglutarate, B—keto-A—deoxyglucarate, 2—keto- u—hydroxyglutarate, and 2-keto-3-deoxygluconate were partially 9l 92 inhibitory. Hydroxypyruvate, dihydroxyacetone, and 2- ketogluconate were not inhibitory. It was concluded from these data that the only restriction against azomethine formation was the presence of a hydroxyl group at carbon three. The formation of a stabilized azomethine with d— ketobutyrate prevented subsequent azomethine formation with pyruvate and similarly c—ketobutyrate competed favorably with pyruvate in azomethine formation indicating that each substrate reacted with the same lysine e—amino group. The aldolase catalyzed the exchange of one proton from water into o—ketobutyrate. The initial rate of exchange was at least 37-times slower than with pyruvate. The aldolase also catalyzed the cleavage of 2—keto—4—hydroxy- glutarate at 0.1% the rate of KDPG cleavage whereas KD— .gluconate was cleaved at a definite, but much slower rate. Oxalacetate was decarboxylated at a rate 0.5% of the rate of KDPG cleavage. The data obtained with the exchange and cleavage reactions were interpreted as demonstrating a dis— sociation of azomethine formation, exchange, and cleavage phases of the KDPG—aldolase catalyzed reaction. Methods are described for the isolation of a radio— active peptide containing the lysine-azomethine forming active site. The amino acid composition of this peptide together with the compos1tion of the native protein is described. 10. 11. l2. 13. 14. 15. REFERENCES Meloche, H. P., and Wood, W. A. J. Biol. Chem., 239, 3515 (1964). ' Rose, I. A., and Rieder, S. V. J. Biol. Chem., 231, 315 (1958). Deal, W. C., Rutter, W. J., and van Holde, K. E. Biochemistry, 2, 2H6 (1963). Entner, N., and Doudoroff, M. J. Biol. Chem., 196, 853 (1952). '““ Macgee, J., and Doudoroff, M.' J. Biol. Chem., 210, 617 (1954). Kovachevich, R., and Wood, W. A. J. Biol. Chem., 213, 4757 (1955) Gibbs, M., and DeMoss, R. J. Biol. Chem., 207, 689 (1954). Ashwell, G., Wahba, A. J., and Hickman, J. J. Biol. Chem., 235, 1559 (1960). ‘DeLey, J., and Doudoroff, M. J. Biol. Chem., 227, 745 (1957)- Warburg, 0., and Gewehn, K. Z. Naturforsh. 96, 206 (1954). Warburg, 0., and Christian, W. Biochem. Z., 319, 1&9 (1943). Hers, H. G., and Kusaka, T. Biochem. Biophys. Acta, 1;, 427 (1953). Peanasky, R. J., and Lardy, H. A. J. Biol. Chem., 233, 365 (1958). Rutter, W. J., Richards, 0. C., and Woodfin, B. M., J. Biol. Chem., 236, 3193 (1961). Horecker, B. L., and Smyrniotis, P. 2., J. Amer. Chem. Soc., ;5, 2021 (1953). 93 8" i \C) 20. 9A Venkataraman, R., and Racker, E. J. Biol. Chem , 236, 1883 (1961). Pricer, W. E., and Horecker, B. L. J. Biol. Chem., 235, 1292 (1960). Hoffee, P., Rosen, O. M., and Horecker, B. L. J. Biol. Chem., 240, 1512 (1965). Blumenthal, H. A., and Fish, D. Biochem. Biophys. Research Commun., 11, 239 (1963). Weimberg, R. and Doudoroff, M. J. Biol. Chem., 217, 607 (19553. Maitra, U., and Dekker, E. E. J. Biol. Chem., 239, 1485 (1964). Srinivasan, P. R., and Sprinson, D. B., J. Biol. Chem., 234. 716 (1959). Levin, D. H., and Racker, E., J. Biol. Chem., 234, 2532 (1959). Karasek, M. A., and Greenberg, D. M. J. Biol. Chem., 227, 191 (1957). Westheime r, F. H., and Cohen, H. J. Amer. Chem. Soc., 60 90 (1938). __3 Speck, J. C., Jr., and Forist, A. A. J. Amer. Chem. Soc , 19, 4659 (1957). e, I. A., and Rieder, S. V., J. Amer. Chem. Soc., 77, 576A (1955). Ros Y. J., Mehler, A. H., and Bloom, B. Science, 1287 (1957). Pontromoli, S., Prandini, B. D., Bonsignore, A., and Horecker, B. L. Proc. Natl. Acad. Sci., 51, 1942 (1961). Grazi, E., Cheng, T., and Horecker, B. L. Biochem. Biophys. Research Commun., 1, 250 (1962). Grazi, E., Rowley, P. T., Cheng, T., Tchola, 0., and Horecker, B. L. Biochem. Biophys. ResearCh Commun., 2, 38 (1962). (A) L0 34. 35. 95 ) (D y, P. T., and Horecker, B. L., . 1012 (1963). (1.40 a 'peck, J. C., Jr., Rowl Amer. Chem. Soc., 85 Grazi, E., Meloche, H., Martinez, G., Wood, W. A., and Horecker,_B. L. Biochem. Biophys. Research Commun., 10, A (1963). Westhead, E. W., Butler, L., and Boyer, P. D. Biochemistry, 2, 927 (1963) Lai, C. Y., Tchola, 0., Cheng, T., and Horecker, B. L. J. Biol. Chem., 290, 1347 (1965). Rosen, 0. M., Hoffee, P., Horecker, B. L., and Speck, J. C., Jr. J. Amer. Chem. Soc., 86, 2092 (1965). —— Rutter, W. J. Federation Proc., 23, 1248 (1969). Swenscn, A. D., and Boyer, P. D. J. Amer. Chem. Soc., 12. 217a (1957). Rowley, P. T., Tchola, 0., and Horecker, B. L., Arch. Biochem. Bigphys., 107, 305 (196A). Cremona, T., Kowal, J., and Horecker, B. L. Federation Proc., 24, 2968 (1965). Rosen, O. M., Hoffee, P., anquorecker, B. L. J; Biol. Chem., 240, 1517 (1965). Stellwagen, E., and Schachman, H. K. Biochemistry, 1, 1056 (1962). Drechsler, E. R., Boyer, P. D., and Kowalsky, A. G. J. Biol. Chem., 234, 2627 (1959). Ros , I. A., O'Connell, E. L., and Mehler, A. H., ‘3 J. Biol. Chem., 2uo, 1758 (1965). Lai, C. Y., Hoffee, P., and Morse, D. Federation Proc., 24, 2967 (1965). ———— Vogl, 0., and Pohm, M., Monatsh, 83, 591 (1952). Gaudry, R., Can. J. Research, 268, 387 (1948). Leggett—Bailey, J., in Techniques in Protein Chemistry, New York: Elsevier Publishing Co., 1962, pp. (3, 216. Jts‘ \O U: C) FL U7 U: [U 53. 55. 56. 57. \TI \j] \0 (I) 60. 96 (I) Moor , S. J. Biol. Chem , 238, 2‘5 (1963). Bencze, W. L., and Schmid, K., Anal. Chem., 29, 1193 (1957). LU ,.. FA \0 U: \‘l V Kinard, F. E. Rev. Sci. Instruments, 28, 29 .—-——u Wood, W. A., and Gilford, S. R. Anal. Chem , 599 (1961). Warburg, 0., and Christian, W. Biochem. Zn, 310, .8 . ,- ' Horecker, B. L., Rowley, P. T., Grazi, E., and Tchola, 0. Biochem. 2., 338, 36 (1963). West, C. D., and Rapport, S. Proc. Soc. Exptl. Biol. Med., 10, 141 (1959). ’ Hamilton, 0. A., and Westheimer, F. H. J. Amer. Chem. Soc., 81, 6332 (1959). Fischer, E. H., in T. W. Goodwin, J. 1. Harris, and B. S. Hartley (eds.), Structure and Activity of Enzymes, New York: Academic Press, Inc., 1964, p. 119. Hirs, C. H. W. Brookhaven Symposium, 15, 154 (1962). Jencks, W. P. in S. 0. Cohen, A. Streitweish, Jr., and R. W. Taft (eds.) Progress in Physical Organic Chemistry, Vol. 2, New York: lnterscience Publishers, Inc., 1964, pp. 63. Hartman, F. C-, and Barker, R. Biochemistry, 4, 1068 \1965). Meloche, H. P. BioChem. Biophys. Research Commun., ié. 277 (1965). 1 UllillllIMHHHNIHIII 61 940 ”II’I‘I'fwnnu(mum