ABSTRACT THE ISOLATION AND CHARACTERIZATION OF A NEW CLASS OF LACTIC DEHYDBOGENASE INHIBITORS: SUBSTITUTED PHENOLS by Mark Chasin In the course of studies with Krebs cycle metabolites, a potent inhibition of lactic dehydrogenase was observed in commercial malic acid solutions. This inhibitor material was isolated and purified not only to characterize its inhibition of the enzyme but also to test the hypothesis, which stemmed from Warburg's conclusions, that compounds which inhibit lactate production should inhibit cancer. The inhibition of lactic dehydrogenase by a commercial malic acid preparation was traced to the presence of an impur- ity. Maple syrup, the commercial source of malic acid, was extracted with methylene chloride. This extract was dried and further extracted with water. Lyophilization of this water extract produced a Maple Syrup Fraction (MSF), which contained the inhibitory substance(s). Using thin layer and gas chromatography, as well as other chemical and instrumen- tal analytical techniques, the following four active lactic dehydrogenase inhibitors were separated and identified from the MSF: (i) vanillin, (2), syringaldehyde, (3) p-hydroxy- ‘benzaldehyde and (4) cyclotene. These compounds, as well as those in the following section, were shown to be reversible inhibitors of lactic dehydrogenase, and to be competitive Mark Chasin with pyruvate and non-competitive with DPNH by Lineweaver- Burke analysis. Since it has previously been shown that only anionic substances inhibit lactic dehydrogenase, it was of particular interest that the anionic form of these substi- tuted phenols was found to be the active inhibitory Species. In an effort to obtain a greater inhibition than that achieved with the naturally occuring inhibitors, the follow- ing series of seven analogues was tested (the Ki's, in uM, are given in parenthesis): p-hydroxybenzaldehyde (#14), protocatachualdehyde (174), "methyl" vanillin (270), "ethyl" vanillin (2&1), "methoxy" vanillin (= vanillin) (96), ”ethoxy" vanillin (108), and the bisulfite adduct of vanil- lin (14). The potency of the inhibitors was found to be a function of the relative negative charge on the alpha atom on the 3 position of the aldehyde ring, rather than a func- tion of the length of the substituent, as would be pre- dicted from purely steric considerations. The aldehyde group increased the activity of the inhibitors over and above lower- ing their pK, since nitro and nitroso groups cannot fully substitute for the aldehyde moiety. Vanillin, chosen as a representative inhibitor, inhibited the H4 isozyme of lactic dehydrogenase approximately twice as much as the M4 isozyme. In tests of the ability of these lactic dehydrogenase inhibitors to inhibit cancer, none showed any reproducible inhibition of Walker 256 intramuscular, leukemia 1210 or sarcoma 180 carcinomas. A study of the rate and pathway of metabolism of both vanillin and its bisulfite adduct in mice Mark Chasin gave LD '8 of 760 and 1850 mg/kg reSpectively. The meta- bolic rgges were 525 and 310 mg/kg/hour respectively. lApparently the bisulfite adduct induced enzymes which pro- moted its metabolism, since the metabolic rate was more than doubled after a 12 day eXposure (to 660 mg/kg/hour). Both compounds were mainly metabolized to vanillic acid, with minor amounts of several different conjugates also found. The rapid rate of metabolism offers a possible explanation of the lack of effect of vanillin on cancer. Moreover, despite the fact that the bisulfite adduct was both a more potent inhibitor and much less rapidly metabolized, at the doses used for carcinostatic testing, the level of even the bisulfite adduct is reduced to the base line level within 3 hours. Therefore it is not surprising that it too did not Show any significant carcinostatic activity. In conclusion, several potent non-toxic lactic dehydrogenase inhibitors have been isolated, but their rapid rates of metabolism pre- clude any meaningful conclusions on the theory of inhibiting cancer by inhibiting lactic dehydrogenase. During.the course of this investigation, it was found that lactic dehydrogenase shows a marked inactivation when left standing at low concentrations (about 5.6 ug/ml.). but that the activity remained constant at 56 ug/ml. for up to #8 hours. THE ISOLATION AND CHARACTERIZATION OF A NEW CLASS OF LACTIC DEHYDROGENASE INHIBITORS: SUBSTITUTED PHENOLS By Mark Chasin A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Biochemistry 1967 G UAW-‘4 «’3 3-7—QQ ACKNOWLEDGMENTS The author wishes to eXpress his apprecia- tion to Dr. w. C. Deal, Jr. for his guidance and assistance throughout the course of this research. We are very much indebted to Dean T. K. Cowden of the College of Agriculture for substantial support at a crucial stage in this research. The finan- cial support of the Michigan State University Department of Biochemistry and an Institutional Cancer Grant from the National Cancer Institute is also gratefully acknowledged. 11 TO MY FAMILY 111 TABLE OF CONTENTS Literature Review Part. A) B) C) D) E) F) G) Enzyme I) Rabbit Muscle Pyruvate Kinase . . . II) Lactic Dehydrogenase . . . . . III) Various Other Glycolytic Enzymes IV) Polyphenol Oxidase . . . . . . . . I, Properties of Lactic Dehydrogenase Reaction Catalyzed and Physiological Function Preparation Of the Enzyme o o o o o o o o o 0 Physical Characteristics . . .'. . . Dissociation and Subunit Composition of the Native 1401601113 0 o o o o o o o o o o o o o Renaturation Studies 0 o o o o o o o o o o Isozymes o o o o o o o o o o o 0 Physiological Significance of the Isozymes AOtive Site StUdieS o o o o o o o o o o o o Mechanisms for the Reaction . . . . . . . Equilibrium Constants and Enthalpy of Reac- tionoo oooooooooooo SUbStrate SpGCiflCity o o o o o o o o Cofactor Specificity . . . . . . . . Inhibitors of the Reaction 1 Substrate Analogues . . . . . . . . . . . 2 Cofactor Analogues . . . . . 3) Inhibitors Which are not Substrate or Cofactor Analogues . . . A Model for Enzymatic Action and Inhibition . II. Lactic Dehydrogenase and Cancer IntrOdUCtlon o o o o o o o o o o o o o Intracellular Location of Cofactors . . . The a-glycerolphoSphate+ Shunt as a Mechanism for Regeneration Of DPN+ o o o o o o o o o o The B-hydroxybutyrate Shunt . . . . . . . Absence of a-glycerolphoSphate Dehydrogenase mmorsoooooooooooooooooo Results of a Non-Operative Shunt in Tumors . Inhibitor Studies . . . . . . . . . . . . . Preparations Materials0.0000000000000000... iv Page CDCDflUll—‘N NHH HHH H00 HH cue HH \nU‘t 21 23 24 27 27 28 30 TABLE OF CONTENTS - Continued Methods I) Assays A) Pyruvate Kinase - Spectrophotometric and Poten- tiometric . . . . . . . . . . . . . B) 103013.10 Dehyergenase o o o o o o o C) Polyphenol Oxidase . . . . . . . . . II) Spectra - Ultraviolet, Infrared, Nuclear Mag- netic Resonance and Mass Spectrometry Analysis . III) Titrations and pK Determinations . . . . . . . IV) Chromatographic and Electrophoretic Techniques A) Gas Chromatography of Maple Syrup Fraction ( MSF O O O O O O O O O O O C O I O B) Column Chromatography of MSF . . . C) Thin Layer Chromatography of Both MSF and the .Compounds to be Tested for Inhibitory Activity D) Paper Chromatography and Electrophoresis of ' MSF O O O O O C O O O O O O E) Polyacrylamide Gel Electrophoresis . . . . . . Results I) Study-of the Effect of Krebs Cycle Metabolites Upon Pyruvate Kinase and Discovery of a Potent Lactic Dehydrogeanse Inhibitor Present as an Impurity in Commercial Malic Acid A) IntrOductlon o o o o o o o o o o o o o o B) Purity of the Pyruvate Kinase . . . . . . . C) Lack of Inhibition of Pyruvate Kinase by Krebs Cycle M3t3b011tes o o o o o o o o o o o o o o‘ D) Impurity Of the M3110 ACld o o o o o o o o o 0 II) The Isolation, Purification and Identification of-a Class of Potent Lactic Dehydrogenase Inhibitors Including Vanillin, Syringaldehyde, p-hydroxybenzaldehyde and Cyclotene from Maple Syrup, the Commercial Source of Malic Acid A) IntrOdUCtion o o o o o o o o o o o o B) The Maple Syrup Fraction, MSF 1) Preparation of the MSF from Maple Syrup . . 2) Physical Properties of MSF . . . . . . . . . 3) Stability Of the MSF o o o o o o o o o o C) Degree of Inhibition of Lactic Dehydrogenase by MSF and Demonstration of the Reversibility Of the Inhibition o o o o o o o o o o o D) Lack of Effect of MSF on Other Enzymes . . . . 'V Page . 36 37 38 38 39 40 40 41 #1 #2 43 45 46 TABLE OF CONTENTS - Continued E) F) G) H) I) J) F) Initial Chromatography Which Failed to Show the Heterogeneity of the MSF . . . . . . Preliminary Instrumental Analysis of MSF . . Chemical Analysis of MSF Suggesting a Phenol Test of the Phenolic Character of MSF by Its Action as a Polyphenol Oxidase Substrate 1)Intr0du0t10noooo 0000000000. 2) Polyphenol Oxidase Assay . . 3) Relation of Structure of Phenols to Their Activity as Polyphenol Oxidase Substrates . 4) Reaction of MSF with the Multiple Forms of Polyphenol Oxidase . . . . . . . . Separation of MSF into Several Components Based on Its Phenolic Structure 1) Thin Layer Chromatography . . . . . . . . 2) Silicic Acid Column Chromatography . . . . Identification _of the Components of MSF 1) Compounds Likely to be Found in MSF . . . . 2) Identification of Lignin in the MSF . . . . 3) Comparison Thin Layer Chromatography of Standards 0 o o o o ' o 4) Identification of the Active and Inactive Components by Assay . . . . . . . . . . . . 5) Organic Fractionation for Gas Chromatography 6) Gas Chromatography to Establish the Presence of the Active Components in MSF . . . . . . 7 ) summary 0 O O O O O O O 0‘ O O O O O O O O O Elucidation of the Mechanism of Inhibition and Preparation and Testing of Analogues to Syn- thetically DevelOp an Improved Inhibitor Introduction . . . . L . . . . Screening of Compounds Similar in Structure to the Naturally Occuring Inhibitors . . . . Determination of the pK of the Active and Inactive Compounds . . . . Synthesis of "Methyl" Vanillin, "Ethyl" Vanil- lin and the Bisulfite Adduct of Vanillin . . . Synthesis of the p-nitroso and p-nitro Ana— logues Of Syringaldehyde o o o o o o o o o 0 Activity of the Nitrogen Containing Analogues IV) Kinetic Evaluation of the Natural Inhibitors and A) the Synthetic Analogues and Proof for the Struc- tural Basis of Inhibition Requirement for a Modified Lactic Dehydrogenase Assay.. oooooooooo ) Stability Of the Inhibitors o o o o o o o Reversibility of the Observed Inhibition . . . . Lineweaver-Burke Reciprocal Plots for the Inhibitors at Various DPNH Concentrations . . . vi Page 56 57 58 61 62 63 66 67 72 73 75 75 77 79 8o 83 84 85 86 89 91 99 103 105 106 TABLE OF CONTENTS - Continued E) F) G) Lineweaver-Burke Reciprocal Plots for the Inhibitors at Various Pyruvate Concentrations . The Molecular Basis of the Inhibition 1) The K 's of the Inhibitors as a Function of the Relative Negative Charge on the 3 Position of the Aldehyde Ring . . . . . 2) Effect of pH on the Observed Inhibition by Vanillin . . . . . . . . Comparison of Vanillin. Inhibition of the M and H4 Isozymes of Lactic Dehydrogenase « . , 1)Intr0du0t10no o o o o o o o o o 2) Determination of the Purity of the Isozymes. 3) Vanillin Exhibits Greater Inhibition of the HuISOZyme....oo........... V) Direct Testing of the Theory of Inhibition of Cancer by Inhibition of Lactic Dehydrogenase A) Testing Procedures . . . . . B) Early Results Demonstrating MSF Inhibition of C) D) Sarcoma 180 in Mice 1) Requirement for High Levels of Dietary Glucose for Inhibition o o o o o o o o o o o 2) LD for the MSF in Mice . . . . . 3) Prégounced Effect of the MSF on Sarcoma 180. in M103 0 o o o 0 0 Slight Effectiveness of the Inhibitors Against Three Different Tumor Systems as Determined by CCNSC . . . Why is Vanillin Essentially Ineffective Against Cancer? 1) The LD and the Rapid Rate of Metabolism OfVQIléRliHiHMiceoooooooooooo 2) Metabolism of Vanillin in Mice a.) IntrOduCtion o o o o o o o o o o o o b) Spectral Analysis of Urinary Vanillin MetabOliteS o o o o o o o o c) Chromatography of Ether Extracts of Both Hydrolyzed and Nonhydrolyzed Urine from Vanillin Treated. M103 0 o o o o o o E) Decreasing the Rate of Metabolism of Vanillin yForming its Bisulfite Addition Product Introduction . . . . . The LD and the Rate of Metabolism of the Bisulfége Adduct in Mice . . . . . . . . . . Metabolism of the Adduct by Mice . . . . . Induction of the Drug Metabolizing Enzymes . Lack of Effectiveness of the Vanillin Bisulfite Adduct in the Treatment of cancer 0 O O O O O O O O O O O O O O O O O O KIT-pm NHO‘ VVV vv‘d vii Page 123 140 141 147 147 148 152 154 154 157 157 167 167 168 171 172 173 173 177 177 TABLE OF CONTENTS - Continued Page Discussion Testing Inhibition of Cancer by Lactic Dehydrogen- aSe Inhibition o o o o o o o o o o o o o o o o o o o 179 Mechanism of the Inhibition of Lactic Dehydro- genase.......................181 summary 0 O O O O O O O O O O O 0 O O O O O O O O O O 185 Bibliography 0 O O O O O O O O I O O O O O O O O O O O 188 Append1XOoooooooooooooooooooooo199 viii FIGURE 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. LIST OF FIGURES Enzyme-DPNH-Substrate Complex . . . . . . . . . EnzymeéDPNH-Inhibitor Complex . . . . . . . . . Alternate Pathways Of DPNH OX1dat10n o o o o 0 Activity of Pyruvate Kinase as a Function of the Concentration of the Krebs Cycle Metabo- lites D O O O O O O O O O O O O O O O O O O O 0 Effect of MgCl on the Inhibition of Pyruvate Kinase by Citrate and Isocitrate . . . . . . . Inhibition of Lactic Dehydrogenase by Silicic Acid Chromatographic Fractions of MSF . . . . . A Typical Gas Chromatographic Pattern of Ether801ubleMSFoooooooooooocoo Comparison of Vanillin and 2-chloro-4_n1tro- phenol as Lactic Dehydrogenase Inhibitors . . . Inactivation of Lactic Dehydrogenase Upon Standing at Low Concentrations . . . . . . . . Lineweaver-Burke Reciprocal Plot Varying DPNH and Vanillin O O O O O O . O O O O O O O O O O O Lineweaver-Burke Reciprocal Plot varying DPNH and. "Ethoxy" Vanillin o o o o o o o o o o o o o Lineweaver-Burke Reciprocal Plot Varying DPNH and. PrOtOCataChUIC AldehYde o o o o o o o o o o Lineweaver-Burke Reciprocal Plot Varying DPNH and. p-hYdroxybenzaldehyde o o o o o o o o o o o Lineweaver-Burke Reciprocal Plot Varying DPNH and. "Methyl" Vanillin o o o o o o o o o o o o o Lineweaver-Burke Reciprocal Plot Varying DPNH and ”Ethyl"Van1111n ooooooooooooo Lineweaver-Burke Reciprocal Plot varying DPNH and the Bisulfite Addition Product of Vanillin. Lineweaver-Burke Reciprocal Plot Varying Pyruvate and Vanillin . . . . . . . . . . . . . ix Page 18 18 26 48 48 7O 82 98 101 110 112 114 116 118 120 122 125 LIST OF FIGURES - Continued FIGURE 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. Lineweaver-Burke Reciprocal Plot Varying Pyruvate and. "Ethoxy" Vanillin o o o o o o o Lineweaver-Burke Reciprocal Plot Varying Pyruvate and Protocatachuic Aldehyde . . . . Lineweaver-Burke Reciprocal Plot Varying Pyruvate and p-hydroxybenzaldehyde . . . . . Lineweaver-Burke Reciprocal Plot Varying Pyruvate and "Methyl" Vanillin . . . . . . . Lineweaver-Burke Reciprocal Plot Varying Pyruvate and."Ethy1" Vanillin o o o o o o o Lineweaver-Burke Reciprocal Plot Varying Pyruvate and the Bisulfite Addition Product OfVanlllin........'........ Eadie-Hofstee Plot of Lactic Dehydrogenase . The Effect of Ring Substituent on the Inhi- bition of Lactic Dehydrogenase . . . . . . . The Effect of the Assay pH on the Vanillin Inhibition of Lactic Dehydrogenase . . . . . Comparison of the Vanillin Inhibition of the H4 and M4 Lactic Dehydrogenase Isozymes . . Comparison of the Sarcoma 180 Tumors Taken From Mice on High and Low Glucose Diets . . External Appearance of Sarcoma 180 Bearing Miceo’ooooooooooooooooooo Sarcoma 180 Tumors Excised From Mice Treated With Various Doses of MSF (Group I) . . . . Sarcoma 180 Tumors Excised From Mice Treated With various Doses of MSF (Group II) . . . . Dose ReSponse Curve for Sarcoma 180 in Mice Treated- With MSF O O O O O O 0 O O O O O O 0 Pattern of Urinary Excretion of the Metabolic Products of Vanillin in the Mouse . . . . . Page 127 129 131 133 135 137 139 143 146 150 156 159 161 163 165 170 LIST OF FIGURES - Continued FIGURE 34. 35. 36. Page Pattern of Urinary Excretion of the Metabolic Products of the Bisulfite AdditiOn Product of Vanillin in the Mouse . . . . . . . . . . . . . 176 The Computer Program in Fortran 63 . . . . . . 201 The Flow Sheet for the Computer Program . . . . 209 xi TABLE 1. 9. LIST OF TABLES Page Solubility of Crude MSF in Various Organic 30 lvents O O O O O O O O O O O O O O O O O O O 52 Stability of MSF under Various Conditions . . 54 Summary of the Chemical Tests Performed to Elucidate the Structure of the MSF . . . . . . 60 Structure of Some Phenol—like Compounds and Their Relative Rates as Substrates for Poly- thHOlOXidaSeooooooococo.oooo 65 The}! s of Several Migrating Species in the Met :&.Ethyl Ketone Solvent System . . . . . . 77 Inhibitory Activity of Several Compounds FoundlnMSF................. 78 The Structure, Length of Substituent Group, pK and Approximate K1 of the Active Naturally Occuring Inhibitors of Lactic Dehydrogenase . 88 Decrease in Activity of Three Naturally Occur- ing Inhibitors Stored at pK 7.50 at Room Temperature While EXposed to Light . . . . . . 103 The Structure, Length of Substituent Group, pK and K of the Inhibitors Used in the Kinetic Studies . . . . . . . . . . . . . . . 108 xii h -' Literature Review Part I, Properties of Lactic Dehydrogenase A) Reaction Catalyzed and Physiological Function Lactic dehydrogenase occurs ubiquitously in animal cells and catalyzes the final reaction of glycolysis, as shown below: pyruvate + DPNH + H+g==2'lactate + DPN+ Physiologically, its two major functions are: (1) to convert pyruvate to lactate under energy utilizing conditions in tissues such as muscle, brain, eye and heart. (2) to convert lactate to pyruvate in the gluconeo- genic organ, liver. B) Preparation of the Enzyme The enzyme was first found in animal tissue by Tununberg (1920). In the early 1930's Bunga, g§_gl. (1932), ahalyzed cell free extracts for activity. The first crys- talline enzyme was prepared from beef heart by Straub (1940) , a'1‘lizlthis has become the standard procedure for the prepara- tdxbnof this enzyme. The.crystalline enzyme has been pre- Pared from rat skeletal muscle (Kubowitz and Ott, 1943; 1 2 Racker, 1951; and Beisenherz, §t_§l., 1953), rat liver (Gibson, 23 21., 1953), rat heart (Wieland, 22 a_1., 1959), human heart (Nisselbaum and Bodansky, 1961a), pig heart (Meister, 1952) and beef heart (Takenaka and Schwert, i956; Schwert, Miller and Takenaka, 1962). C) Physical Characteristics The sedimentation coefficient of the beef heart enzyme has been reported as s = 6.368 and 6.468 for two different preparations (Meister, 1950), although the conditions of the experiment were not given. However, Neilands (1952) reported a sedimentation coefficient of $20,w = 7.058, a diffusion coefficient of D20.W = 5.26 x 10"7 cmZ/sec and a partial Specific volume of 0.754 cc/g for the beef heart enzyme, but experimental details were not described. He calculated the molecular weight to be 135,000 1 15,000. The enzyme from hog heart has a sedimentation coefficient of 3°20,w = 7.658 (Kegeles and Gutter, 1951). Assuming a molecular weight of 135,000 (Neilands, 1952) fVor the native molecule, Velick (1958) estimated the minimal C‘ocambining weight of the beef heart enzyme with DPNH to be 37,500. I’) Dissociation and Subunit Composition of the Native Molecule Results of Millar (1962) suggest that electrophoreti- 1 r 3 cally homogeneous beef heart enzyme dissociates at concentra- tions below 0.2%. The sedimentation coefficient extrapolated from concentrations above 0.2% was reported as $020,w = 7.718. The Archibald method (1947) yielded a molecular weight of 72,000 at infinite dilution, and although the exact extrapo- lation seems questionable, the data do suggest that a marked decrease in molecular weight occurs. However, Appella (1964) reported that the molecular weight of the beef heart enzyme remained at 140,000 in dilutions to 0.04%, as measured both by light scattering and sedimentation-diffusion. Recently, Hathaway and Criddle (1966) reported the beef heart enzyme in the absence of substrate to have a sedimentation coefficient of $20,w = 5.68 at 3.0 ug/ml. However, they found that 10-3M pyruvate could raise the sedimentation coefficient under otherwise identical conditions to $20,w = 7.78, and suggested that substrate can form active tetramers from inactive dimers. At present, the molecular weight of the native beef heart lactic dehydrogenase is still open to question. Appella and Markert (1961) treated the H4 isozyme (see IDelow) from crystalline beef heart lactic dehydrogenase with 53 guanidine hydrochloride and 0.1M 2-mercaptoethanol. This 1treatment dissociated the enzyme into four inactive subunits vFiltha molecular weight of 34,000, a sedimentation coefficient or s° = 1.753, a diffusion coefficient of 13° 20,w 20,w 1C)‘7 cmz/sec., and a partial specific volume of 0.740 cc/g. = 6.7“ X TWuese results suggesting four subunits per native molecule 1'ltaive recently been questioned. Stegink and Vestling (1966) 1“ Id L). report eight amino terminal acetate residues for rat liver M4 enzyme and 7-8 moles of acetate per mole of beef heart H1+ and MR3 isozymes (see below). Appella (1964) reported eight amino terminal valines per mole of beef heart lactic dehydro- genase, and indicated fingerprint studies also showed eight polypeptide chains per molecule. The observation of fifteen isozyme bands rather than the predicted five (see F below) places further doubt on the four subunit models of lactic dehydrogenase. E) Renaturation Studies Reversible disaggregation of beef heart lactic dehyd- rogenase was demonstrated by Markert (1963) by freezing in 1! NaCl. Later, Epstein gt a_l;. (1964) achieved up to 60% reactivation after subjecting the enzyme to 10.5M urea and 0.12M_2-mercaptoethanol, conditions shown to cause complete dissociation of the protein into unfolded polypeptide chains. Chilson, 22 al. (1965a) used 7.5M_guanidine hydrochloride to 3 dissociate the chicken H enzyme. The presence of 1.3 x 10- M L; DPNH in their renaturation medium yielded 30% recovery of tactivity, rather than the 15% observed when DPNH was not Lincluded in the reversal mixture. It had previously been Eshown that DPNH will protect beef heart lactic dehydrogenase :from.inactivation by heat, urea (Pfleiderer, gt‘§;., 1957) tand sodium dodecyl sulfate (Di Sabato and Kaplan, 1964). F) Isozymes The procedure of Straub (1940) using heart tissue yields crystalline enzyme which exhibits two components upon electrophoresis at pH near neutrality (Meister, 1950; Neilands, 1952a and b; Pfleiderer and Jeckel, 1957). These and other workers (Vesell and Bearn, 1957 and 1958; Hill, 1958; Hess, 1958) found the several electrophoretic forms active. However, Schwert and coworkers (1962) separated the two forms.on hydroxylapatite and suggested that the slower electrophoretic component was inactive, although the inter- pretation of their results is open to question, since their Specific activity curve indicated both forms may be active. Results described in this thesis indicate both forms are active. In 1957, Wieland and Pfleiderer reported that elec- trophoresis of lactic dehydrogenase from various tissues gave one to six components exhibiting enzymatic activity. iMarkert and Moller (1959) extended the analysis and defined tune molecular basis for the observation of the several active f‘orms. They coined the term "isozyme" to describe different IIlolecular forms with the same enzymatic activity. Fritz and Jémcobson (1963a, b, 1965) have shown twelve to fifteen enzy- Emitically active bands by electrophoresis of mouse muscle, Ihouse heart and rat liver lactic dehydrogenase in 2-mercapto- e‘tzhanol. Appella and.Markert (1961) found that the subunits 6 produced by 5M guanidine hydrochloride could be separated on the basis of charge into two different kinds and were desig- nated A and B lactic dehydrogenase. Later, Dawson, 23 El. (1964) named these H and M after the tissue sources in which they predominated, namely, heart and muscle. Theoretically. combining these two types of subunits in all possible permu- tations of four subunits per native tetramer predicts the existence of the following five isozymes: H4, H3M, H2M2' HM3 and M4' and these are observed experimentally upon elec- trophoresis. Salthe, gt El- (1965), Emerson, g§,§l. (1964) and Chilson. 22.31. (1965a) have shown that when frozen together in high salt at neutral pH, a mixture of the pure M4 and H“ lactic dehydrogenases from different Species will produce all five isozymes. The lactic dehydrogenase isozymes have been found to differ widely in properties other than migration in an elec— tric field, including amino acid composition (Wachsmith, gt .2;,, 1964; Fondy and Kaplan, 1965; Kaplan, 1964), immunologi- calreactivity (Nisselbaum.and Bodansky, 1961; Cahn, 22 21., 1962), kinetic behavior (Vesell and Bearn, 1961; Kaplan, 1964), thermal stability (Hill, 1958; Pfleiderer, gt a_1_.. 15957; Wroblewski and Gregory, 1961; Zondag, 1963), subunit c(Imposition (Appella and Markert, 1961), and inhibition by 1£actate (Brody, 1964), pyruvate (Plagemann, 23 31., 1960; Stambaugh and Post, 1966), urea (Brand, 23 11,, 1962), OXalate (Emerson, 31; 51., 1964), sulfite (Wieland and Pfleiderer, 1957) and d-hydroxybutyrate (Elliot, 93; 9.1.” 1962). 7 Assays for directly determining the fraction of each parent type in a sample of lactic dehydrogenase have been developed by Dawson, g£,§l.. (1964) and Stanbaugh and Post (1966), based on differential inhibition by either pyruvate or lactate, and Kaplan and Cahn, (1962), based on differences in their activity toward the hypoxanthine cofactor analogue. It is now well established (Shaw and Barto, 1963; .Markert, 1963; Goodfriend and Kaplan, 1964; Fine, §§_gl.. :1963 and Cahn, Kaplan and Zwilling, 1962) that the synthesis (of the two isozymes of lactic hydrogenase is under control (of two separate genes. A third isozyme distinct from H or Dd has recently been demonstrated in Sperm (Blanco and Zinkham, 1963). G) Physiological Significance of the Isozymes Studies by Dawson, g§_§;. (1964), Cahn, gt_gl. (1962), and Kaplan and Cahn (1962) indicate that the H and M forms ‘3f’ lactic dehydrogenase may have significantly different Physiological roles. The H form is inhibited by excess pyruvate to a greater degree than the M form. In aerobic tissue pyruvate inhibits the H form and forces the cell to oxldize pyruvate by the mitochondria, yielding large amounts (If. energy needed by such aerobic tissue. In tissues such as 8133ll‘iated muscle, where M is found, energy is needed in short buJi‘sts under anaerobic conditions, and can be produced despite temporarily high levels of pyruvate. Dawson, 91; 2.3;. 8 (1964) have correlated the M subunit concentration with the apparent degree of dependance of the tissues on aerobic gly- colysis as an energy source. The results of Goodfriend, g£_g;. (1966) demonstrating the differential repression of synthesis of M subunits under increasing oxygen tension support this concept of a functional role for the 2 differ- ent subunits. H) Active Site Studies Nygaard (1956), measured the binding of pyruvate as a function of pH and concluded that an imidazolium group could be at the binding site for pyruvate. Winer and Schwert (1958) concluded two dissociable groups were involved in binding DPNH - one with a pK of 6.8, possiblyuimidazolium, and the other, with a pK of 9.8, a sulfhydryl. Measurements by fluorescence (Velick, 1958; Winer, 25.51,, 1959), binding of DPNH analogues (van Eys, 22.2l-c 1958) and ultracentrifugation (Takenaka and Schwert, 1956) show each mole of enzyme to bind four moles of coenzyme. I) Mechanisms for the Reaction From the work of Alberty (1953) and their work, Hakala gt 9;. (1956) suggest the following possible mechanisms 1' or the action of lactic dehydrogenase: 1. Either substrate or coenzyme may be bound to the 9 enzyme, but the binding sites are not independent. The binding of the second reactant is influenced by the presence of the reactant which is already bound. 2. There is a compulsory order of binding of reac- tants. Substrate, for example, is not bound by the enzyme but only by the enzyme-coenzyme com- plex. Various ternary complexes may be postu- lated without altering the form of the equations relating initial velocity to initial concentra- tions of reactants, 3. There is a compulsory sequence of interaction of enzyme with reactants but in which ternary com- plexes are so Short-lived as to be without kinetic significance - the so-called "Theorell- Chance" mechanism. (Theorell and Chance, 1951) This third possibility is unlikely since there is a direct stereoSpecific transfer of hydrogen from coenzyme to substrate (Loewus and Stafford, 1960). Work with inhibitors (Anderson. gtflgl.. 1964; Novoa and Schwert, 1961), centri- fugal separation eXperiments (Takenaka and Schwert, 1956), rapid equilibrium analysis (Silverstein and Boyer, 1964), (and Spectroscopic measurements (Chance and Neilands, 1952) afawbr a compulsory order of binding, cofactor before sub- Ertrate, but there is no definite proof of mechanism as yet. 10 J) Equilibrium Constants and Enthalpy of Reaction The equilibrium constant (concentrations in moles/ liter) for the enzymatic reaction catalyzed by lactic dehydro- 12 (Racker, genase has been variously reported as 4.4 x 10- 1950). 3.3 x 10.12 (Neilands, 1952) and 2.76 x 10-12 (Hakala, 22.21,. 1956), all at 25.0°C. Hakala further estimated the 43H for the oxidation of lactate by DPN+ to be 10.3 i 0.3 kcal/mole by direct calorimetry. K) Substrate Specificity Generally, multicellular organism lactic dehydrogenase is Specific forI;(+) lactate, with no detectable activity for D (-) lactate, nor is D (-) lactate an inhibitor. However, Dennis, §t_gl, (1959, 1960, 1965) have reported a lactic dehydrogenase isolated from Lactobacillus plantarum Specific for the D (-) lactate. Haugaard (1959) and Bennett, £3.21. (1966) reported the presence of both D (-) and L (+) lactic dehydrogenase in E. coli. Snoswell (1959) reported the pur- ification of a D (-) lactate Specific lactic dehydrogenase from Lactobacillus arabinosus, and Gleason, gt 2;. (1966) f'ound a D (-) lactic dehydrogenase in two lower fungi, SEipromyces'elongatus and Mindeniella Spinosora. Meister (1950) found that a series of a.y-diketo a<31ds, varying in chain length from 5-11 carbons, can be readueed at about one-tenth the rate of pyruvate. Reduction 11 in this case occurs at the a-keto group only. oxidizing one mole of DPNH per mole substrate. Beef heart lactic dehydro- genase will also reduce a,y-diketo-b-methylcaproic acid and a,Y-diketo-e~methylheptanoic acid. 2-mercaptopyruvate (Kun, 1957) and mesoxalic acid (Winer, gt,gl., 1959) will also act as substrate but at a reduced rate. In contrast to these, the rate of reduction of the d-keto series Cu-C decreases markedly with chain length (Nisselbaum and Boiansky, 1961a). The enzyme is inactive against trichloroacetic acid, 3-phOSphoglyceric acid (Frank and H012, 1959), acetaldehyde, acetone, methyl ethyl ketone, 3,6-dimethyl-2,5-p-dioxane, methyl lactate and glycolic acid (Neilands, 1954). It should be noted that both Stegink and Vestling (1966) and Dabich (1960) have found that although rat liver lactic dehydrogenase is resistant to hydrolysis by leucine aminopeptidase, up to 20% of the molecule can be digested by carboxypeptidase without effecting the catalytic activity of the enzyme. .L) Cofactor Specificity , AS has been found for most dehydrogenases, lactic <1ehydrogenase requires DPNH with a beta nicotinamide ribosidic 3.1nkage and shows no activity toward the alpha isomer (Kaplan, it; a” 1955). The enzyme is also Specific for the alpha 1’lydrogen in the para position of the nicotinamide ring, as 12 Shown by Loewus and his coworkers, (1953, 1960). See Noller (1957) for nomenclature system. Although TPNH will serve as cofactor for the reaction, DPNH is 100-380 times as effective (Meister, 1950). Anderson and Kaplan (1959) determined the ability of a series of DPN+ analogues to function as cofactor for the reaction. Using beef heart enzyme, substitution of the 3-amide group in the pyridine ring with a hydroxamide acid, hydrazide, formaldox- ime, thioamide or isobutyryl gave an active cofactor, but the activity was generally less than that of DPN+. Substitution dfbenzoyl. amino, acetamide or 2-butenylamide groups gave an inactive analogue. It Should be noted that Lactobacillus arabinosus has been Shown to contain a DPNH independent lac- tic dehydrogenase (Snoswell, 1959). M) Inhibitors of the Reaction 1) Substrate analogues For reasons described below, inhibitors of lactic aom soapmoaxo mzmm do mamasP®aopmsaopH¢ .m ossmam 26 m oasmfim ouwsamona mumnamona msouoom macs pace uaoucyMHWHMII! umwoupmnwp . ow>9u>a ow ooH +zmmn sesamwoupmsop omwcmmoupmcop iiwzmn eumnamonafiouoomamio owuosH E zma KIT pace ownoomawonamonawpnnsanw mewsowoupmsop ouusamosaunnopmnotawumomaw Hm + mumnamosduMumpmnmpamuoomfim 2222: 22.3 “8 355:5 “22%;: 27 DPNH + dihydroxyacetone phoSphate-———>d-glycerolphOSphate + DPN+ (3) O + a-glycerolphoSphate-——e>energy via the phOSphorylat- 2 ing electron tranSport chain (4) SUM: 02 + fructose-1,6-diphoSphate-——e’1,3-diphOSphoglyceric acid + energy The only obvious alternative scheme for regeneration of DPN+ from DPNH is by reduction of pyruvate to lactate using lactic dehydrogenase. The lactic acid then produced must then be tranSported to the liver prior to any further metabolism. This scheme leads to the prediction that under anaer- obic conditions, the levels of both lactate and d-glycerol- phOSphate would increase and this has been observed by Ciaccio, §£.gl. (1960) in a number of tissues. That d—glycerolphos— phate dehydrogenase may be indiSpensible to the efficient growth of a cell, namely through maximum oxidative pathways, is thus evident. D) The B-hydroxybutyrate Shunt An alternative mitochondrial shuttle similar to that given above utilizes acetoacetate and B-hydroxybutyrate dehyd- rogenase and has been described by Boxer and Devlin (1961). Present evidence is insufficient to evaluate its existence and operation as a physiologically Significant process. 28 E) Absence of daglycerolphOSphate Dehydrogenase in Tumors In most malignant tissues, d=glycerolphoSphate dehyd— rogenase activity is either missing or very low (Holzer, st, 21.. 1958; Boxer and Shonk, 1960; Delbruck, gt 21., 1959; Sacktor and Dick, 1960; Foster and Taylor, 1965). This is not due to an inhibitor, but rather to a decreased amount of enzyme present (Boxer and Shonk, 1960). This lack of d- glycerolphOSphate dehydrogenase results in an almost complete halt in production of deglycerolphoSphate (Ciaccio, 22 91., 1960). The only exceptions to this rule are the Ehrlich- Lettre tumor of the mouse (Boxer and Devlin, 1961) and the Morris hepatoma 5123 of the rat (Morris, at al., 1960). Even though it has normal levels of d-glycerolphoSphate dehydroe genase activity, the Ehrlich-Lettre tumor does not produce a-glycerolphOSphate during glycolysis, for some as yet unknown reason. Both of these tumors completely lack the acetacetate shunt (Boxer and Devlin, 1961). F) Results of a Non=0perative Shunt in Tumors If shuttles other than the d-glycerolphoSphate dehydro- genase shuttle are known or assumed to be absent, this leads inexorably to the conclusion that a cell lacking daglycerol- phoSphate dehydrogenase will be absolutely dependent upon lactic dehydrogenase to regenerate the DPN+ necessary for 29 14 glycolysis. Results by Busch (1955), showing that 2-C pyruvate injected into tumors was primarily converted to lactic acid, support this conclusion. All energy, there- fore, comes from glycolysis, even under aerobic conditions, and the absence of daglycerolphOSphate dehydrogenase pro- vides a biochemical basis for an eXplanation of Warburg's earlier observations (1924). Recently, a detailed study by Goldman, gt 3;. (1964) of a number of different human neoplasms showed that tumors generally have a higher concentration of lactic dehydrogen~ ase M isozyme than surrounding noninvolved tissue. Since the M isozyme is more suitable for high rates of lactic acid production, this observation fits very well the total picture of neoplastic glycolysis. The results have been confirmed and extended by Peznanska-Linde, §£.§A° (1966). Further evidence for the absolute requirement of tumors for lactic dehydrogenase has been furnished by Gregory, 23,31. (1966) and Ng and Gregory (1966). By prepare ing Specific antilactic dehydrogenase antibodies and testing them both against liver and tumor homogenates and tumor cell culture, they demonstrated complete abolition of aerobic glycolysis in tumors, and almost complete inhibition of tumor cell reproduction, while normal cells were almost unaffected by these antibodies. The above picture of neoplastic glycolysis makes one fact very clear a lactic dehydrogenase is an absolutely indispensible enzyme for tumor survival. Without it, the 30 glycolytic pathway would soon be inoperative through a lack of DPN+, and the cell‘s energy supply would be fatally depleted. Since the normal tissue has daglyoerolphoSphate dehydrogenase and thus has no such dependence, this suggests an approach to the chemotherapy of cancer with a fundamental biochemical basis. In the words of Busch (1961), this con- clusion that "... the carbohydrate metabolism of neoplastic tissues was deranged ... was one of the most important which has been developed in the cane oer field, and has served as the focal point of a very large number of important eXperi= ments which have been relevant both to the field of cancer research and general biochem- istry; it still remains one of the most impor- tant differences between many tumor and non- tumor tissues." Selective inhibition of lactic dehydrogenase may then kill a cancer cell without harming normal cells. G) Inhibitor Studies Based upon the above conclusions, inhibitors of lace tic dehydrogenase have been tested for activity against cane oer. Several inhibitors, including oxamate (Novoa, gg’al., 1959; Colowick and coworkers, 1961a,b, 1965a,b), oxalate (Busch and Nair, 1957; Novoa, g£_g;., 1959), tartronate (Ottolenghi and Denstedt, 1958), and fluorOpyruvate (Busch and Nair, 1957) have been found. Fluoropyruvate, tested on the Walker 256 tumor (Davis and Busch, 1958), was shown to inhibit both the reduction and oxidation of pyruvate and to 31 have central nervous system toxicity. Tartronate, tested on the Yoshida ascites hepatoma (Flume, 1960), had very little effect on tumor growth and was shown not to be Specific for lactic dehydrogenase (Fiume, 1959). Hydroxamic acid, a more potent inhibitor of lactic dehydrogenase, also proved to be ineffective against the Yoshida hepatoma. Colowick and coworkers (1961a,b, 1965a,b) have intensely studied both oxamate and oxalate in both Ehrlich ascites tumor and in HeLa 83 cell culture. They have found that oxamate can inhibit both glycolysis and tumor growth, and have demonstrated the site of action as being lactic dehydrogenase. To get these results they had to use an energy supply high in glucose to force the use of the glym colytic pathway by repressing other pathways. They found that oxamate exhibits nonSpecifio toxicity which eliminated it from further consideration as an antitumor chemothera- putic agent. Many of these inhibitors have been toxic and this has discouraged and Slowed the efforts in this direction almost to a complete halt. As a result, the theory has not yet been tested in an unequivocal manner because of the lack of a good non-toxic lactic dehydrogenase inhibitor system. A major objective of this thesis was to provide such a system and test this hypothesis. Enzyme Preparations I) Rabbit Muscle Pyruvate Kinase Rabbit muscle pyruvate kinase was prepared by a slight modification of the method of Tietz and Ochoa (1958). The enzyme was kept concentrated during the ethanol fraction— ation ,, rather than being diluted to 20 mg/ml. Following the final heat step, an ammonium sulfate fractionation was carried out. The enzyme finally used in these studies was the 40-55% ammonium sulfate fraction. It was stored in the cold at a concentration of 90-100 mg/ml, in 0.02M imidazole buffer, pH 7.00, containing 0.001M_EDTA. II) Lactic Dehydrogenase Beef heart lactic dehydrogenase was purchased from Worthington Biochemicals, lot BHaLDH 6010, as a crystalline ammonium sulfate suSpension. It was found to be completely free of pyruvate kinase activity in the range of the assay concentrations. The rabbit muscle lactic dehydrogenase used in the isozyme studies was a crystalline ammonium sul— fate suSpension from Sigma. III) Various Other Glycolytic Enzymes Rabbit muscle glyceraldehyde~3uphosphate dehydrogenase, 32 33 phOSphoglyceric acid kinase, phOSphoglucomutase and glucose- 6~phOSphate dehydrogenase were purchased from Sigma. AMP deaminase, prepared by the method of Smiley, §£,§l. (1967), and UDPG pyrophOSphorylase, prepared by the method of Albrecht, §£_§l. (1966), were kindly donated by the reSpec- tive authors. IV) Polyphenol Oxidase The mushroom polyphenol oxidase used in these studies was of two kinds - the first, lot TY-602, was purchased from Worthington Biochemicals, and the second, prepared as described below, was kindly donated by Dr. S. Constantinides. In a typical preparation, each gram of Agaricus campestris was gently crushed with 5 ml of 0.05fl.ph08phate buffer, pH 7.20. using a mortar and pestle. The resulting suSpension was centrifuged at 300 x g to remove the cellular debris, then at 10,000 x g for 15 minutes to remove the mitochondrial fraction, and finally at 100,000 x g for one hour to remove the microsomes. The supernatant fraction from this final centrifugation was used for the studies described below. Materials The pyruvic acid used in the assays was twice distilled fPI~cnm Matheson, Coleman and Bell lot PX~2125. The ether, metha ylene chloride, acetone and lactic acid were Baker analyzed reagents. Eastman Organic Chemical products included vanillin, 1>—=luzvdroxybenzaldehyde, protocatachuic acid, pmhydroxybenzoic eac:fi.di, gallic acid, pyrogallic acid, o-, m» and pmmethoxy- jpduesrlol, syringic acid, o~ethylphenol and coumarin. Chemicals Fnllrcxhased from Aldrich Chemicals included o-cresol, 3eethoxy- “r-Ilsndroxybenzaldehyde ("ethoxy" vanillin), 2,6udimethylphenol, 2cés-«limethoxyphenol,_2mmethyl—4enitrosophenol, 2-chloro-4- r11tl“0phenol, guaiacol, 2~methylcyclohexanone, a duplicate :saunlble of vanillin to check the Eastman sample, and 2~hydroxy~ 1“‘Tfl€3thyl-l-cyclopentenem3mone hydrate (cyclotene). A sepa- :rE‘t3EB sample of the latter was kindly provided by Dr. Riley or Dow Chemicals. K and K Labs products included syringaldehyde, vanil- lic acid, ferulic acid, 3,5udihydroxyebenzoic acid, salicyl alcohol, 2-methylcyclopentanone, protocatachualdehyde, .p‘qfiydroxyphenylacetic acid, penitrophenol, pecoumaric achi, QTSIchicine, colchiceine and 3emethyl~1,2ucyclopentanediol. c“linic acid and Lumalic acid were purchased from Nutritional Biochemicals Corporation. Caffeic acid was donated by Dr. H. M. Sell of this department, to whom it had been 34 35 provided by the California Foundation for Biochemical B 8 search . The following chemicals were obtained from Sigma; disodium B-DPNH‘3H20, lot 4‘3Be608, phosphoenolpyruvate, tricyclohexylammonium salt, lot 94Bm5200, dwketoglutario acid, lot K42Be217, DL-isocitric acid, lot I107-87, fumaric ac 16., lot 15B~a0760, and adenosine diphOSphate, lot 45Bm7000. Chemicals obtained from Calbiochem included ciss aconitio acid, lot 30074. Mallinckrodt succinic acid and Fisher citric acid and methylethyl ketone were also used. The maple syrup used for the preparation of the Maple Syrup Fraction was obtained from several areas of the Country, and differed both in appearance and in yield of the NSF produced per liter. Michigan syrup gave the high- est yield and appeared darkest. The Vermont and New York ) Syrups gave the lowest yield, and appeared thin and light. All syrup used was 100% grade A (r grade Fancy maple syrup. The Vermont and New York syrup was supplied through the Se:':“Tices of Sugarbush Industries, Lansing, Michigan in bulk. The Michigan syrup was purchased locally, mostly from Carl Gearhart in Charlotte, Michigan, in five gallon lots. ,.~. v. d a: sra .s .‘v ‘0': -- h :I \L t” -.~ ch- Methods I ) Assays A) Pyruvate Kinase «- Spectrophotometric and Potentiometric The coupled Spectrophotometric assay with lactic dehydrogenase measured the decrease in absorbance at 340 mm as DPNH was converted to DPN+. The assay contained 0.01M MgClz, 0.10M KCl, 0.02M adenosine diphOSphate, 3.4 x 10"“M pho Sphoenolpyruvate, 5.6 ug beef heart lactic dehydrogenase, 0.103 Tris-HCl, pH 7.50, and 1.8 x 10"”;4 DPNH, at 25.000 in a. total volume of 0.40 ml. The conditions were found to give maximum velocities and the velocities were proportional to enzyme concentrations. The enzyme and compound to be tested were each added in 5 (1L volumes. The Specific activ- ity was measured in units of (M of phoSphoenolpyruvate reQCted/min/mg of pyruvate kinase. All assays were performed ona Gilford Spectrophotometer equipped with an automatic cuVette positioner. Protein concentrations were measured by abs()rbance at 280 m on the Gilford Spectrophotometer, using an extinction coefficient for pyruvate kinase of 0.54 OD/mg/ cm after Bucher and Pfleiderer (1955). The potentiometric a"sisay involved measuring the rate of uptake of 1.035 x 10-321 EC31 in a Radiometer automatic pH stat, model TTTl. The assay Vessel contained 0.008M_ MgClZ, 0.00211; adenosine diphOSphate, 36 37 0. 001g; phoSphoenolpyruvate and 0.1013; KCl, pH 7.50 at 25.00C. Pyruvate kinase was added in one 5 uL sample, as was the sample to be tested. B) Lactic Dehydrogenase For the Spectrophotometric assay of lactic dehydro- genase, the decrease in DPNH concentration was followed Spectrophotometrically at 340 mm. The assay contained 3- 3 x 104'}; sodium pyruvate, 0.03M potassium phoSphate buffer, 6.7 x 10.613; DPNH, 3.3 x 10'3M_ NaCl and was at pH 7-1+O. The compound to be tested and the lactic dehydro- genase (0.1 pg) were each added in 5 11L volumes, bringing 1Zhe final assay volume to 0.30 ml. Only initial rates were measured. These conditions were found to give accu- rate and reproducible results, in which the velocities were proportional to enzyme concentration. The Specific activ- ity was measured in units of uM of pyruvate reacted/min/mg or enzyme. For this assay, the change in optical density Per the minute, , when multiplied by the factor 1090.0, yields Specific activity. All assays were performed on a GilI‘ord Spectrophotometer equipped with a constant tempera— ture water bath at 25.0°c and an automatic cuvette posi- tioner. Protein concentrations were measured by absorbance at 280 mu. using an extinction coefficient for lactic dehydrogenase of 0.698 OD/mg/cm (Schwert and Winer, 1963). As described under results, the assay was changed 38 after finding that in this assay the enzymatic activity decreased with time, and the following assay was used for all of the kinetic work described under results. For the new assay, the enzyme was kept at enzyme stock concentra- tion of 56 ug/ml. and the cuvette contained the following: 6. 9 x 10'”); NaCl, 1.4 x 10.412 sodium pyruvate, 0.10M phosphate buffer, pH 7.40 and 2.6 x 10-5M DPNH. All other conditions were as above. C) Polyphenol Oxidase The assay of polyphenol oxidase contained pH 6.50, 0.1673; sodium phosphate buffer, 3.3 x 10" i4. tyrosine and approximately 0.025 mg enzyme in a final volume of 3.0 ml. The enzyme was added by layering it on a close-fitting plurlger with small holes in it. The plunger was rapidly moved up and down in the cuvette, serving both to aerate and mix the contents. Although the rate of this reaction can be measured by the rate of increase in optical density at 280 mp. for the reasons outlined under results, the in<‘->::~easing optical density was more routinely measured at “’60 mu. All assays were run as is described above for pyruvate kinase . 11) Spectra - Ultraviolet, Infrared, Nuclear Magnetic Resonance and Mass Spectrometry Analysis The ultraviolet absorbance Spectra were run on a 39 Beckman DB recording Spectrophotometer. Infrared Spectra were taken in KBr pellets in the Beckman 5 infrared Spec- trophotometer. The mass Spectrum was measured in a Con- solidated Electrodynamics Corporation model 21-1030 mass Spectrometer. NMR Spectra were taken in D20 using a Varian A-60-A nuclear magnetic resonance spectrophotometer. III) Titrations and pK Determinations Titrations were performed automatically using the Radiometer Corporation model TTTlc automatic titrator, with 1 . 0 ml.‘ samples. The inflection points of the pK determin- ations were calculated as follows: 1) The linear portions of the curve on either side of the inflection point were extended. 2) The Slope line for the linear region at the inflec- tion point was drawn to intersect the above two lines. . 3) The inflection point falls exactly halfway between the two intersections plotted above. The hydrochloric acid was standardized against 0.97611}; NaOH, which was standardized against potassium acid phthalate. A 3°1ution was prepared which was exactly 1.0000M HCl. and tale acid for titration was prepared from dilutions of this 8took, so that quantitative titrations could be performed. For the pyruvate kinase assay, 0.00111 HCl was used, and for the pK determinations, 0.10M HCl was most generally employed. 40 IV) Chromatographic and Electrophoretic Techniques A.) Gas Chromatography of Maple Syrup Fraction (MSF) Initially, the gas chromatography was performed on the Aerograph 660, varying the column temperature but keep- ing the injector port at 215°C. An ionizing flame detector was used, and the collector portzdetector port ratio was kept at 9:1 throughout the eXperiments. In later gas chromatography work, a model 400 F and M gas chromatograph with a 5% carbowax on chromosorb W column measuring six feet by one-quarter inch was used. The oven was maintained at 250°C, the flash heater at 300°C and the flame detector at 360°C. 13) Column Chromatography of NSF Sephadex G-10. G—200. LH-20, Dowex 50 and Amberlite CG‘SO were all prepared and used according to the instruc- ti one provided by the supplier. Silicic acid columns were prepared using Mallinckrodt 100 mesh silicic acid well waehed with both distilled water and the elution solvent. with the methylethyl ketone solvent, the silicic acid tn~1‘ned yellow at first, but after many consecutive washings a{ppeared completely white. 41 C) Thin Layer Chromatography of Both MSF and the Compounds to be Tested for Inhibitory Activity Since many solvent systems were used, they will be described under results. The solid phase in all cases was silicic acid, 250 1: thick for analytical work and 1 or 2 mm thi ck for preparative procedures. Two types of plates were Jrc>t1tinely prepared using the DeSaga-Brinkmann apparatus; carles measuring 2 inches by 8 inches, and the second 8 by 8. £33r 'using a fluorescent binder in'the silicic acid, it was possible to visualize all Spots present after chromatography by a masking of the fluorescence of the binder everywhere any material was present. Ascending chromatography was used in all cases, and the chromatograms were run until the Selvent front was one-half inch from the top of the plate. The samples to be chromatographed were Spoted one inch from the bottom of the plate, using either water, ether or ace- tone solutions, depending upon the solubility of the material to be run. D) Paper Chromatography and Electrophoresis of MSF The techniques and equipment used for these studies were reported in detail by Katz and co-workers (1959). “hatman number 1 paper, 181- inches by 22% inches, was used for both chromatography and electrophoresis. Desending chromatography utilized n-butanolzacetic acidzwater, 4:1 :5, 42 while the electrophoresis was conducted in a pyridine:acetic acidzwater buffer, 1:10:289, pH 3.70. E) Polyacrylamide Gel Electrophoresis The polyacrylamide gel electrophoresis was performed according to the method of Ornstein (1964) and Davis (1964), using 7% polyacrylamide gels prepared using chemicals pur- chased from Canalco and a Tris-glycine buffer at pH 8.6. The bands were visualized after electrophoresis in two ways. The first was by staining with Amido-Schwartz black and destaining either electrophoretically or by soak- ing in 7%% acetic acid. The second was by using an activ- ity assay in the gels. For polyphenol oxidase, the gels were soaked in a solution of substrate, and the bands of enzyme converted the colorless substrate into colored p35‘<-><1uct. Although the MSF itself was colored, its enzymatic pr<>€1uct could easily be distinguished from the background °°lor. For lactic dehydrogenase, the gel activity assay of MarIl-etert and UrSprung (1962) was used, omitting the hydrazine trajpping agent to achieve a colorless background (Kaplan and Cahn, 1962 ). Results I. Study of the Effect of Krebs Cycle Metabolites Upon Pyruvate Kinase and Discovery of a Potent Lactic Dehyd- rogenase Inhibitor Present as an Impurity in Commercial Malic Acid A) Introduction "A living cell consists in large part of a concentrated mixture of hundreds of different enzymes, each a highly effective catalyst for one or more chemical reactions involving other components of the cell. The paradox of intense and highly diverse chemical activity on the one hand and strongly poised chemical stability (biological homeostasis) on the other is one of the most challenging problems of biology." (Atkinson, 1965) The original purpose of this research was to gain some insight into the systems of controls in the glycoly- tic pathway utilized by the higher vertebrates to enable them to more efficiently reSpond to the energy require- ments placed upon them in their environments. This research led to the discovery of a natural inhibitor of lactic dehydrogenase from maple syrup. Although the major emphasis in this thesis is on the lactic dehydrogenase inhibitor, the studies on the control of pyruvate kinase serve as a good introduction to the discovery of these inhibitors. 43 44 In controlled pathways, regulation has generally been found to be accomplished by a simple feedback mechanism, in which the end product of the system controls the entire pathway through inhibition of the first enzyme involved. Pathways involving threonine deaminase (Umbarger and Brown, 1957) and aSpartate transcarbamylase (Yates.and Pardee, 1957) are two well studied examples of this type of control. Glycolysis, on the other hand, does not seem to pos- sess such a simple control system. It is clear that a very highly complex system of controls exists to determine the state of the system. In an attempt to simplify the problem of defining the control points, Pye and Eddy (1965) studied the levels of the glycolytic intermediates and suggested four probably controlled steps. These were glucose entry and glycogen degradation, 3-phoSphoglyceric acid kinase, phOSphofructokinase and pyruvate kinase. Of these, only phoSphofructokinase and phoSphorylase have been studied in any detail. PhoSphorylase has been shown to be activated by AMP (Cori, Colowick and Cori, 1938; Brown and Cori, 1961; Krebs and Fischer, 1962), and inhibited by ATP, glucose-6-phos- phate (Morgan and Parmeggiani, 1964) and UDPG (Madsen, 1963). The citrate inhibition of phosphofructokinase has been very well documented (Parmeggiani and Bowman, 1963; Passonneau and Lowry, 1963; Vinuela, gthgl.. 1963; Mansour, 1963; Garland, gthgl., 1963; and Williamson, g§_gl.. 1964) for several systems under a variety of conditions, as has 45 its actiyation by AMP (Mansour, 1963; Passonneau and Lowry, 1962; Ranaiah, Hathaway and Atkinson, 1964). Pyravate kinase is known to be noncompetitively inhibited by diethylstilbestrol (Kimberg and Yielding, 1962) and ATP (Boyer, 1963), but to date there has been no work done on the effect of the tricarboxylic acid cycle inter- mediates upon the enzyme. Results, section I, C describes a systematic study of the control of pyruvate kinase by the Krebs cycle substrates. B) Purity of the Pyruvate Kinase The pyruvate kinase used for the studies described below was first physically characterized to insure that it was essentially pure and homogeneous. It was found to be homogeneous in sedimentation velocity experiments in the Spinco model E analytical ultracentrifuge. Chromatography on Sephadex G-200 yielded activity and protein curves which coincided. Polyacrylamide gel electrophoresis revealed one major band and only trace contamination. The pattern was similar to that which had been reported for pure pyruvate kinase previously (Tietz and Ochoa, 1958; Warner, 1958). 0) Lack of Inhibition of Pyruvate Kinase by Krebs Cycle Metabolites The pyruvate kinase was tested for inhibition by the 46 Krebs cycle intermediates, including citrate, cis-aconitate, isocitrate, a-ketoglutarate, succinate, fumarate and malate. AS Shown in figure 4, malate, citrate and isocitrate gave the only apparent significant inhibitions at low concentra- tions of metabolites; cis-aconitate, a-ketoglutarate, suc- cinate and fumarate gave no Significant inhibition. Oxalo- acetate Spontaneously decarboxylates in aqueous solution to pyruvate, the.substrate-forithe lactic dehydrogenase in the coupled assay employed. This eliminated consideration of this compound. Due to solubility problems in the assay system. fumarate was tested at a maximum concentration of 1.5 x 10'3M. Citrate and isocitrate gave appreciable inhibition at higher concentrations than malate. To test the possibility that this apparent inhibition might be due to chelating effects, eXperiments were run in which excess MgCl2 was added to the assay medium along with citrate or isocitrate. As shown in figure 5, the apparent inhibition caused by the citrate and isocitrate was essentially abolished, indicat- ing that a chelating effect had caused the apparent inhibi- tion. The malate curve was totally unaffected by such addi- tion. D) Impurity of the Malic Acid Since malic acid was the only Krebs cycle substrate which gave significant inhibition not ascribable to a chelating effect, the purity of this sample was checked by 4? .mcoapoz Hoods confluence mm whence oahpoaopoaaohpoodm .eopmop mason essoaaoo one on Heads msodpwHPSoosoo dd codes was Ndowz .opmhpaoomd ens opmhpao an emanax assessed ho SodpanHSSa one so «Hows co possum .m enemas .pwmp one s« encodes encams5m an godsendssH can so once 4 .meoSpoz House debauched we mashed odapoSoposa nonpocdm .mopaaonmpoa cacao mnona one mo SoapmnpSooSoo one no Soap0s5m m we omssdx opdhbhha mo hpabapod. .e enemas 48 m mhsmfim as .1 ES: on em 3 ea 4 a a . 1 1 cu .. .. 3 0/0 mzod d Augean .. mzo: < .. 2:53. _- in 3 «as: n I / I IS ‘ /‘ «3.2. v 35:82 MIMI!" "BO 83:! r. 7 a .. /w. -e s enemas as .1 E53: 2 s s 14H 4 mks—to whack—cow. O/A mh<¢ o-diphenol.+ 0= 2 2————> o-quinone + 2H20 The enzyme then catalyzes the formation of colored 2 o-diphenol + 0 products such as dopachrome and melanin from the o—quinone. 62 Dressler and Dawson (1960) report that the oxidation of monophenols does not proceed through a diphenol intermedi- ate. The mushroom enzyme exists in multiple forms (Constantinides, 1966). These forms, separated by poly- acrylamide gel electrophoresis, have been found to exhibit differences in substrate Specifity. The phenolic nature of the MSF indicated by the analytical data presented above could be confirmed if the MSF could serve as a substrate for polyphenol oxidase. Furthermore, by determining the Specificity of the poly- phenol oxidase, structural features of such phenolic com- pdunds might be revealed. Therefore, studies were under- taken to determine both the Specificity of the mushroom polyphenol oxidase and its ability to utilize MSF as a substrate. 2) Polyphenol Oxidase'Assay The activity of polyphenol oxidase is usually mea- sured by following the increase in optical density at 280 ; mg as the o—quinone is produced. This was not possible with the MSF because it had such a strong ultraviolet absorption that it was impossible to zero the Spectro- photometer at this wavelength. Since polyphenol oxidase catalyzes the further reac- tion of o-quinones to highly colored products, it seemed likely that the assay might be conducted in the visible 63 region of the Spectrum to avoid the previous difficulties. To test this a product was made in the following way. Two samples of MSF were mixed with buffer and incubated at 25°C, one with Worthington polyphenol oxidase and one without. After fifteen minutes, the sample containing the enzyme was considerably darker than the control. A difference Spec- trum of these two samples was run on a Beckman DB Spectro- photometer to determine the exact wavelength at which the product of the reaction absorbed. This difference Spectrum peaked at 460 mu. Therefore, this wavelength was selected for the assays of the polyphenol oxidase. A control assay comparison.of.the tyrosine reaction at 280 mo and 460 mu showed the rates to be identical, within experimental error. The major difference was longer lag in the assay at 460 mu: this is to be expected from the reaction mechanism described in Section II H l. 3) Relation of Structure of Phenols to Their Activity as Polyphenol Oxidase Substrates The relative reaction rate of several phenol-like compounds were measured and compared to tyrosine which was arbitrarily assigned a rate of 100. Two milligrams of each substrate were present in each assay and the Worthington enzyme was used. The results are summarized in table 4 along with the structures of each of the tested substrates. The MSF had a relative rate of 85, indicating a high content 64 Table 4. Structure of some phenol-like compounds and their relative rates as substrates for poly- phenol oxidase. See text for discussion and details. * Tyrosine arbitrarily set to a rate of 100 and the other rates are relative to this as a standard. .65 RELATIVE compouun smucwn: ,m . new, Vanillin one 0" . 0 0H quinic aCid "00:00" 0 no H ' 0H p-hydroxy- benzoic acid HOMO—on 0 ' , om:3 syringic acid HOOCQ-Ofl . 1 ' ocu, p-hydroxy- t benzaidehyde Oflc-O—ou 4 ‘ ”.“2 A tyros ine moc—cu— "In on 100 4. 0H protocatachuic acid H000 OH 189 ON caffeic acid uooc-cu- on 592 N 2 0H dihydroxy- - _ . phenylalanine H000 c" "2° 0" 1733 catechol Wm 7600 Table 4 66 of phenolic material, and confirming the previous analytical results. However, the most active inhibitory fraction (see Figure 6) had a relative rate of only 28. This surprising observation is discussed in the following section. 4) Reaction of MSF with the Multiple Forms of Polyphenol Oxidase A member of this laboratory worked on the separation and Specificities of the multiple forms of polyphenol oxi- dase from mushroom (Constantinides, 1966). He found that although 20-25 protein bands were seen when the gels were stained with amido-Schwartz black using the mushroom enzyme (see Enzyme Preparations IV), only a few of these bands were Shown to be active as polyphenol oxidase. Using MSF as sub- strate for the color reaction (see Methods, IV E) three bands could be seen. DOPA gave 7 bands, tyrosine 2, cate- chol 5, catechine 2, caffeic acid 2 and chlorogenic acid gave only 1. Most Significant was the fact that the three bands seen with MSF were the same bands generally seen with the other substrates. That is, the MSF reacted only with the most nonspecific isozymes of polyphenol oxidase. The data presented above and in table 4 Show that the most rapid rates of oxidation by polyphenol oxidase -were achieved with the o-dihydroxyphenols, although a small, but measurable, rate was seen with several monohydro- xyphenols. The lack of reaction with quinic acid suggested 67 that substrates for polyphenol oxidase must be aromatic. That the crude MSF had a high relative rate of 85 may indi- cate the presence of one or more dihydroxyphenols. However, since the most active inhibitory fraction had a relative rate of only one-quarter that of the crude MSF, it would seemthat the inhibitory component or components are not o-dihydroxyphenols. Indeed, if they are phenols at all, the possibility exists that the position ortho to the hydroxy group may be blocked, as in vanillin or syringic acid, thus accounting for their reduced activity with polyphenol oxidase. I) Separation of MSF into Several Components Based on its Phenolic Structure 1) Thin Layer Chromatography Now thatthe basic chemical nature of the MSF had been determined, it was possible to resolve it into its components. Block (1958) describes two solvent systems designed to separate mixtures of phenols: (1) 20% KCl in water and (2) methyl ethyl ketone:2N ammonium hydroxide, 2:1, pH 11. Thin layer chromatography on silicic acid using the 20% KCl solvent system gave only two Spots. However, the methyl ethyl ketone system resolved the MSF into about twenty Spots, several of which fluoresced when excited by ultraviolet light. The most prominent of these 68 was a Spot which fluoresced a very bright blue. It had an Rf of about 0.3, but this varied greatly depending upon the concentration of material applied to the plate. This Spot served as an excellent marker in later work, since by vir- tue of its intense fluorescence, it was detectable at very low concentrations. The next problem was to determine which of the com- ponents was active. Preparative thin layer chromatography was carried out to obtain enough of each of the components to assay. The fractions were eluted from the silicic acid using acetone. This procedure yielded fractions which were not completely soluble in water, probably due to some resi- due from the acetone or from the chromatography solvent. Therefore, as much as possible of each fraction was first dissolved in a small volume of water. It was then filtered through a millipore filter, lyophilized to dryness, weighed and redissolved in water to a known concentration of 5 mg/ml and assayed at an assay concentration of 80 ug/ml. A plot of per cent inhibition of lactic dehydrogenase versus Rf in the methyl ethyl ketone solvent is shown in figure 6. Con- veniently for later detection, the maximum inhibition cen- tered on the bright blue fluorescent band. Although the spots using this solvent system are rather distinct, the inhibition is still spread out over quite a large area of the chromotograph. This probably indicates the presence of more-than one active Species, although this fact was not appreciated for quite some time: initially, the simplest 69 Figure 6. Inhibition of lactic dehydrogenase by silicic acid chromatographic fractions of MSF. The fractions were eluded from thin layer plates developed in the methyl ethyl ketone solvent system described in the text. Spectrophoto- metric assay for lactic dehydrogenase activity is described under Methods. 'IIIHIBITION PER cmr' ~30. 7O INHIBITION OF lDI-I'BY SlllClc ACID CHROMATOGRAPHIC FRACTIONS ,OF MSF 50 40 20 III 0 0-2 0.4 0.6 0.8 II) a, or Manon 71 SXplanation, suSpected breakdown of a single active Species, was thought to account for the spreading. Since the inhibitory activity peak was centered on the blue fluorescent band, attempts were made to purify this band by repeated chromatography in the methyl ethyl ketone solvent system. After six consecutive thin layer chromato- graphs, a sample was obtained which, upon further chromatog- raphy in the same solvent system, showed only the blue fluorescent band. An assay of this sample proved it to be inactive as a lactic dehydrogenase inhibitor. This suggested that the active species probably chromatographed just ahead or Just behind this blue spot. The fluorescent material probably has a very complex structure, since an ultraviolet Spectrum in ether shows peaks at 338, 296, 284, 258 and 223 mu. An infrared Spectrum indicated the presence of hydroxyl, and carbonyl groups with a prominent CH stretch. No signifi- cant infrared long wavelength absorption was seen. Compari- son of the infrared spectrum using the Sadtler index pro- duced no positive identification of the fluorescent material. .Attempts to purify the spots on either Side_of the fluorescent band by repeated chromatography failed. Upon each rechromatography, a larger and larger fraction of the material eluted from the middle of the chromatograph moved near the solvent front. This result indicated that the material was probably decomposing during the time needed for several consecutive chromatographs. An assay of the material moving near the solvent front showed it to be 72 devoid of inhibitory activity toward lactic dehydrogenase. 2) Silicic Acid Column Chromatography In an attempt to collect a larger amount of the active chromatographic region in a Shorter period of time, a silicic acid column was run using the methyl ethyl ketone solvent to elute the fractions of MSF. By examining one microliter portions of consecutive fractions of the eluant for fluores- cence by Spotting them on filter paper and exposing the Spots to ultraviolet light, it could easily be determined when the fluorescent band was eluted. The lactic dehydrogenase inhibitory activity was centered mainly about this region, as shown above. However, a good separation on a column required a very slow flow rate. This resulted in conversion of the active fraction moving near the center to the inactive fraction moving near the solvent.front. This was shown by immediate rechromatography of the column fractions on thin layer using the usual system. This solvent System was there- fore abandoned. From the previous experiments one positive result was obtained. This was evidence for more than one active inhib- itor. While rechromatographing the column fractions on thin layer, it was Seen that the active fractions had several compounds that the inactive fractions did not. These com- pounds chromatographed directly above and below the fluores- cent one, right where the presence of the inhibitor was 73 strongly suSpected. In addition, a slightly active fraction eluted from the column well ahead of the bright blue fluores- cent material contained a third spot which was not present in any other fraction, having an R of 0.71, within the same f limitations noted for R 's in this thin layer system. The f evidence for there being more than one active inhibitory Species continued to mount. J) Identification of the Components of MSF 1) Compounds Likely to be Found in MSF The problem now was having a large number of com- pounds with the possibility that several might be active. Since the commercial value of maple syrup has stimulated considerable chemical research upon its components, this was turned to as a source of information. Also, since maple syrup, the source of MSF, iS a wood product, a search was made in the literature for compounds known to be present in wood and sap. One Such list which proved to be very helpful was found in Pearl and McCoy (1960). In addition to the list of twenty compounds, their Rf's in two different solvent systems and the color reactions of each compound with ten different stable diazo spray reagents were detailed. From the work of Pearl and Dickey, (1951), five common color reactions for wood derived phenolic compounds were obtained. About twenty other compounds were subsequently added to this 74 list from other sources, resulting in a list too long for it to be feasible to assay all compounds for inhibitory activity against lactic acid dehydrogenase. ‘ A series of papers by Underwood and coworkers (1961, 1963, 1964, 1965) described the isolation and identifica- tion of some of the components of a chloroform extract of maple syrup. This chloroform extract had properties con- Siderably different from the CH2C12 extract, MSF, but it was thought that their results might form a starting point for identification of the components of MSF. After organic fractionation of the chloroform extract, they identified by gasuchrOmatography’the following components of maple syrup (see Underwood and coworkers, 1961, 1963, 1964, 1965): 1) ethyl acetate . 11) syringaldehyde 2) ethyl alcohol 12) dihydroconiferyl alcohol 3) acetoin 13) syringoyl methyl ketone 4) acetol 14) guaiacol 5) cyclotene 15) 2,6-dimethoxyphenol 6) phenol 16) coumarin 7) diacetin 17) cbniferyl aldehyde 8) a long chain fatty ‘ 18) 2,6-dimethoxybenzoquinone acid 19)‘lignin 9) vanillin 10) ethyl vanillate Since lignin is a very common constituent of plant extracts, its presence was sought in MSF. 75 2) Identification of Lignin in the MSF The classic phloroglucinol reaction for lignin, treat- ment of the sample with a saturated solution of phloroglucinol in concentrated HCl, was performed on the MSF, and it produced a strongly positive reaction. Lignin is insoluble in diethyl ether, and when the MSF sample was repeatedly treated with ether until no further precipitation occured, the ether sol- uble fraction gave no phloroglucinol reaction, while the precipitate still gave a strongly positive reaction. This was taken as conclusive proof of the presence of lignin in the MSF, and lignin became the first component to be identi- fied in the MSF. Lignin samples from the MSF were assayed for inhibi- tory activity against lactic dehydrogenase in the usual assay system. .Its activity increased with time after isolation from MSF, although it never approached the activity of the MSF preparations. Lignin is known to oxidize in air and undergo decomposition into many products. From the above data, it seemed possible that the active components in the MSF might be degradation products of the lignin known to be present. Such products are generally aromatic structures ‘with many methoxy substituents, which would be consistent I ‘with previous structural data. 3) Comparison Thin Layer Chromatography of Standards 'Since the R of the active inhibitory region had been f 76 well defined, thin layer Chromatography was performed on a series of compounds related to wood, wood products and 11g- nin degradation products from the above lists in an attempt to reduce the number of compounds to assay. The following compounds had an Rf near zero in the methyl ethyl ketone solvent system, and so were eliminated from consideration: 1) syringic acid ‘ 7) coumaric acid 2) protocatechuic acid 8) 3,5-dihydroxybenzoic acid 3) caffeic acid 9) vanillic acid 4) quinic acid 10) gallic acid 5) p-hydroxybenzoic acid 11) DPN+ 6) p-hydroxyphenylacetic acid 12) DPNH These last two were tested since they are involved in the activity of the enzyme, but their Rf's eliminate them from consideration as causing the observed inhibition. Since lactic acid can inhibit lactic dehydrogenase by product inhibition, a control to test the ability of lactic acid to inhibit this system was run; At a lactic acid con- centration of 1.0 x 10'3M, the enZyme had 100% of the con- trol activity. Therefore, lactic acid cannot be reSponsible for the observed inhibition by MSF. It Should be noted that oxamate, a well known potent inhibitor of lactic dehydrogen- ase, cannot be the active factor either, Since it exhibits a positive ninhydrin reaction, while the MSF reaction is negative (see Table 3). Table 5 lists the compounds which did move in the above solvent system and their R 's, with the bright blue f 77 fluorescent band included as a standard Rf marker. Compound Rf syringaldehyde 0.27 blue band 0.35 vanillin 0.50 cyclotene 0.64 p-hydroxybenzaldehyde 0.67 guaiacol 0.7)+ coumarin 0.83 Table 5. The R 's of several migrating Specigs in the methyl ethyl ketone solvent system described under Methods. Two things Should be noted about this table. First, vanillin (moved just ahead of) and syringaldehyde (moved Just behind) are the two compounds moving closest to the blue fluorescent band which acted as a marker for the inhibitory region. Second, the Rf values for guaiacol, cyclotene and p-hydroxy- benzaldehyde are near 0.7: this is approximately the Rf value of the faster moving inhibitory region seen in the silicic acid column chromatography. 4) Identification of the Active and Inactive Components by Assay Now that the large list of possible compounds had been narrowed down to a more workable size, it was feasible to 78 assay them for inhibitory activity against lactic dehydro- genase. The six compounds in Table 5 were tested first at a concentration of 4.5 x io'fifl in the usual assay system. The results are shown in Table 6. Compound k inhibition coumarin 5 guaiacol 13 cyclotene 39 syringaldehyde 62 p-hydroxybenzaldehyde 71 'vanillin 100 Table 6. Inhibitory activity of several compounds found in MSF. inhibition is expressed as per cent inhibition of the standard assay of lactic dehydrogenase by a 4.5 x 10' M solution of material. The results of 13% inhibition by guaiacol and 5% by coumarin were not considered significant, being almost within eXperi- mental error. The other four gave very significant inhibi- tion at this concentration. No combination of them in pairs gave any indication of synergistic inhibition. Proof of the actual presence of these four good inhibitor compounds in the MSF was next undertaken. The first method of proof consiSted essentially of running thin layer chromatogramscfi‘the MSF and the known compounds for comparison. Thin layer chromatography was carried out with 79 vanillin and syringaldehyde on one plate along with MSF. After development in the usual solvent, the plate was Sprayed with a saturated solution of 2,4-dinitrophenyl- hydrazine, in 2N HCl, after Pearl and Dickey (1951). Vanillin gave a yellow-orange Spot correSponding in color and position to a spot in the MSF sample. The syringalde- hyde gave a red-brown Spot which likewise had a similar spot in the MSF sample. No color reaction could be found for cyclotene: the region near where p-hydroxybenzaldehyde moved gave a very dense color with all of the Sprays tested, obviating confirmation of their presence by this method. 5) Organic Fractionation for Gas Chromatography The second method used for proof of the presence of the active inhibitors in MSF was gas chromatography. In Underwood's (1961) gas chromatography scheme, the crude extract was subjected to a simple organic fractionation before gas chromatography was attempted. A similar but far more detailed organic fractionation was attempted on the MSF. However, when the fractions produced by such fractionation were subjected to thin layer chromatography, it could be seen that the separation achieved left much to be desired. For instance, the fraction supposedly contain- ing only the strong acids present in the MSF proved to con- tain at least 13 compounds. These were shown by thin layer 8O chromatography in the methyl ethyl ketone solvent system not to be acidic at all. The problem with this type of fractionation is that many of the compounds in the MSF seem to have at least moderate solubility in both organic and non-organic solvents, so that very little fractiona- tion is actually achieved by organic extraction of either acidic or basic solutions of MSF. Therefore, the gas chromatography described in the next section was performed on unfractionated ether soluble MSF. 6) Gas Chromatography to Establish the Presence of the Active Components in MSF Using the conditions outlined under Methods for the F & M gas chromatograph, attempts were made to conclusively prove that the active inhibitors were actually present in MSF. In earlier attempts to use gas chromatography with the Aerograph 660, the lignin in the MSF was not removed and only one large peak was observed, which was originally erroneously taken as evidence of the homogeneity of the MSF sample.' This time the lignin was removed as above before gas chromatography was performed to avoid fouling the column with non-volatile material. Figure 7 shows a typi- cal gas chromatogram of ether soluble MSF, using the condi- tions described under Methods. The peaks which have been identified and their corresponding numbers on figure 7 are as follows: 81 Figure 7. A typical gas chromatographic pattern of ether soluble MSF, using an F and M model 400 gas chromatography apparatus as described under Methods. The active species are labelled. Identification was both by retention time and peak coincidence when both MSF and the compound were injected Simultaneously. 82 p—HYDROXY— 7 f Figure 7 83 1) ether, the solvent peak 10) coumarin 2) cyclotene 11) vanillin 3) p-hydroxybenzoic acid 15) p-hydroxybenzaldehyde 6) syringic acid 16) syringaldehyde These compounds were identified in two ways: 1) by reten- tion time: and 2) by peak coincidence after applying both ether soluble MSF and compound to be tested together into the column. The active Species are labelled on the figure. In addition to the compounds above whiCh were proved to be present in the MSF, there were several compounds which were shown not to be present in significant amounts, and these included protocatechuic acid, caffeic acid and quinic acid. 7) Summary All four compounds, vanillin, cyclotene, p-hydroxy- benzaldehyde and syringaldehyde, found to be active as inhibitors, have thus been shown’to be present in the MSF in at least one way. They have also been Shown to inhibit lactic dehydrogenase. It was therefore concluded that the observed inhibition of lactic dehydrogenase by MSF was mainly due to vanillin, syringaldehyde, cyclotene and p-hydroxybenzaldehyde. III. Elucidation of the Mechanism of Inhibition and Prepara- tion and Testing of Analogues to Synthetically Develop an Improved Inhibitor .A) Introduction At this stage of development, the next major questions were: 1) What are the critical structural features reSpon- sible for the inhibition? 2) Can a better inhibitor be prepared? Since an answer to the first Question is a prerequisite to a logical approach to the second, it was considered first. Inspection of the active inhibitors isolated from maple syrup revealed that they possessed the following common Structural features (see Table 7): 1) A hydroxyl group, which could bind to the enzyme either through hydrogen bonding or an ionic bond, provided the hydroxyl is ionized. They are thus phenols and phenol analogs. 2) Generally, a substituent ortho to the phenolic hydroxyl such as methyl, methoxy or hydroxy, which may also bind through hydrogen bonding to the enzyme, or which may indirectly affect the hydroxyl, through electronic effects on the aromatic ring system. 84 85 3) Generally, an aldehyde group para to the hydroxyl, which may either bind directly to the enzyme, such as in a Schiff base, or decrease the pH of the hydroxyl group and aid binding to the enzyme in this manner. These observations provided the starting point to determine the relation between structure and inhibition. By assaying other compounds having various combinations of these common structural features, it was expected that elimination of one or more of the required features could be accomplished. B) Screening of Compounds Similar in Structure to the Naturally Occuring Inhibitors All compounds to be tested were assayed in the orig- inal lactic dehydrogenase assay described under Methods, using a concentration of 4.5 x io-ufi, the same concentration used for the experiments in table 6. Under these conditions pyrogallic acid was found to inhibit the enzyme 41%. All of the following compounds were found to be completely inactive as inhibitors. 1) vanillic acid (7) 6) ferulic acid* 2) syringic acid (4) 7) protocatachuic acid (4) 3) o-methoxyphenol 8) p-hydroxybenzoic acid (4) 4) m-methoxyphenol 9) gallic acid (7) 51 p-methoxyphenol 10) quinic acid (4) L *The structure of ferulic acid is o-methoxy-p-acrylyl phenol. 86 Note: The numbers following these compounds refer to the table in which their structures are given. It can easily be seen from the above list that adja- cent hydroxyl and methyl or methoxy groups are not suffi- cient to create an active inhibitor, since o-methoxyphenol, which contains this arrangement, does not inhibit, while pyrogallic acid and p-hydroxybenzaldehyde, which lack any methyl or methoxy groups, inhibit quite well. Therefore, structural similarity to lactate was not the common denom- inator of the active Species. Since the pK of phenolic hydroxyl groups varies widely depending upon the substituents on the ring (see Noller, 1957, chapter 24), the possibility that the pK might be a critical factor merited consideration. In view of the fact that only anionic substances have previously been found to inhibit lactic dehydrogenase (see literature review), this seemed an especially attractive possibility. C) Determination of the pK of the Active and Inactive Compounds Consistent with the above reasoning, all of the inac- tive compounds had pK's above 9.6, and so would be uncharged in the lactic dehydrogenase assay at pH 7.40. Furthermore, the active Species had pK's of about 7.9 or below as summar- ized in Table 7, so that almost 50% of the molecules were ionized. In two cases, pyrogallic acid and cyclotene, the 87 Table 7. The structure, length of substituent group, pH and approximate K1 of the active naturally occuring inhibitors of lactic dehyd- rogenase. The pH values were determined experimentally as described under Methods. 88 SIDE COMPOUND STRUCTURE 12:81)?" I”) vanillin p-hydroxy- benzaldehyde 1-084 7.90 6.90 cyclotene 2.639 10.00 syringaldehyde 3-959 7.50 6.60 pyrogallic acid 2-392 8 80 Table 7 89 hydroxyl group appears to have two pK's. For pyrogallic acid the values may well represent titration of two dif- ferent hydroxyl groups. The two pK's of cyclotene may well be due to a di-enol form of the molecule whose two hydroxyl groups titrate at different pH'S. Since a major common denominator of the active Species was their low pK, synthesis of analogues having a low pK was undertaken. D) Synthesis of "Methyl" Vanillin, "Ethyl" Vanillin and the Bisulfite Adduct of Vanillin The two most active naturally occuring inhibitors, vanillin and p-hydroxybenzaldehyde, differ in structure tnly in the substituent ortho to the hydroxyl. To evaluate the effect of the substituent, the following series of Six com- pounds was prepared with gradually increasing length of this substituent: Substituent Trival Name' H p-hydroxybenzaldehyde OH protocatachualdehyde CH3 “methyl" vanillin OCH3 "methoxy" vanillin = vanillin C235 "ethyl" vanillin OCZHS ”ethoxy” vanillin "Ethyl” and "methyl" vanillin were the only compounds in this list which could not be purchased commercially in pure form. 90 They were therefore synthesized by a slightly modified ver- sion of Adams revision (1923, 1924) of the Gatterman alde- hyde synthesis. After the steam distillation, it was found necessary to pass the highly colored liquid through several consecutive charcoal adsorptions at 100°C to produce a clear, colorless supernatant from which the final product could be prepared. The "methyl" vanillin formed an amorphous preci- pitate which when left undisturbed for several days crys- tallized into small, flat white plates with a Sharp melting point at 117.5°C, as compared to the value of 118°C of Adams (1924). Quantitative titration, thin layer chromato- graphy in three solvent systems (methyl ethyl ketone:2N NHuOH, 2:1: 20% KCl in H20: n-butanolzpyridine:dioxane: water, 70:20:5x5) designed to separate phenols, ultraviolet and infrared Spectrophotometry all were consistent with a pure compound, and so it was used Without further purifica- tion. The "methyl" vanillin procedure was also used to prepare "ethyl" vanillin. Upon cooling the clear superna- tent a white, amorphous precipitate formed much as above. Since conditions to form crystals from this amorphous pre- cipitate could not easily be found, the suSpension was lyophilized, yielding a fine white powder. Quantitative titration, infrared and ultraviolet Spectrophotometry and thin layer chromatography in the above three solvent sys- tems for phenols all were consistent with a pure prepara- tion of "ethyl" vanillin and although it was not crystal- 91 line, it was used without further purification. For reasons described in Section V E 1, it was desir- .able to prepare the sodium bisulfite addition product of vanillin, which was synthesized as follows. To 10 ml. of saturated aqueous sodium bisulfite was added crystalline 'vanillin with vigorous mixing. This order of addition was found to produce a purer product. A white precipitate was Anoted after some time, but addition of the vanillin was con- tinued until a twofold molar excess of vanillin to bisulfite ihad been added, to insure complete reaction. The precipitate ‘was collected by filtration and exhaustively washed with 100% lethanol, removing all unreacted vanillin. The adduct was “then dried for 48 hours at 55°C, and was used without fur- ‘bher purification. Yield was essentially quantitative. The pK's of the three new inhibitor analogues were as follows: ”methyl" vanillin --------------- 7.90 "ethyl” vanillin - -------------- r-- 7.90 bisulfite adduct of vanillin ----L- 7.65 {Plrey were therefore predicted to be good inhibitors of lac- ti o dehydrogenase . E) Synthesis of the p-nitroso and p-nitro Analogues of Syringaldehyde In addition to the effect of the substituent ortho ‘t‘3 the phenolic hydroxyl, the pH of aromatic phenols can be Jtr11Fluenced by the substituents in the position para to the 92 hydroxyl. The aldehyde para to the hydroxyl groups in the naturally occuring inhibitors functions to lower the pH of the phenol. Therefore a more strongly electron-withdrawing group should theoretically produce a lower pK and as a result, a better inhibitor, unless the how required for activity. It duce a lower pH and provide an tiality of the aldehyde. (The ment for carboxyl). Since the aldehyde group itself is some- would be eXpected both to pro- excellent test for the essen- bisulfite tests the require- nitroso and nitro groups are good electron withdrawing groups, the synthesis of the p-nitroso and p-nitro derivatives of certain of the inhib- itors was begun. The analogue of syringaldehyde was chosen due to the ready availability of the starting material, 2,6-dimethoxyphenol. Also, this analogue could be produced in greater yield and purified far more easily than the ana- logues of the other naturally occuring inhibitors. The procedure of Cervinka and Kavka (1957) was con- siderably revised to synthesize 4-nitroso—2,6-dimethoxy- phenol, since the limited solubility of the 2,6-dimethoxy- phenol caused the reaction mixture to freeze at the low temperatures used by the above authors. The synthesis pro- cedure was as follows. To 25 g. of 2,6-dimethoxyphenol was added a solution containing 48 g. NaOH in 450 ml. of deion- ized, distilled water. The suSpension was mixed at 5°C until dissolved. Then a solution containing 16 g. NaNO in 2 450 ml. deionized, distilled water was added slowly, care being taken to maintain the temperature of 5°C. After 10 93 minutes, the dropwise addition of a solution containing 45 g. of H2804 in 100 ml. of water was begun, while pass- ing a stream of 002 into the reaction mixture. Beginning five hours later 50 ml. of concentrated H280“ were Slowly added over two hours, during which time a heavy precipi- tate formed. This precipitate was dissolved in concen- trated NaOH, giving a very dark purple solution, and fil- tered. Acidification with HZSOH to pH 3.0 was followed by extraction with n-butanol, which was subsequently air dried by placing in the hood overnight. This yielded the crude product, a dark brown in color. This crude product was dissolved in 1M_NaOH and the solution was adjusted to pH 4. The solution was then decolorized with charcoal, which, although drastically decreasing the yield, was the only way found to produce reasonably pure product. The clear filtrate from the char- coal step was further acidified'to pH 2.0 and extracted with ether. Upon drying the ether solution, a small quan- tity of very light tan crystals was produced, which melted at 126-127°C. Quantitative titration of the crystals Showed them to be at least 98% pure. They were used with- out further purification because of the limited quantity of material available. Oxidation of a very small quantity of the 4-nitroso- 2,6-dimethoxyphenol with H 02 by the method of Travagli 2 (1950) gave light tan crystals of 4—nitro-2,6-dimethoxy- phenol. These titrated 99$ pure and melted sharply at 116°C. 94 This compound was also assayed for inhibitor activity with- out further purification as described below. AS was predicted above, the pK's of these two ana- logues was considerably lower than either the naturally occuring inhibitors or the analogues described in section III D. The 4-nitroso-2,6-dimethoxyphenol had a pK of 6.55, while the 4-nitro-2,6-dimethoxyphenol titrated at 6.30. Since the lactic dehydrogenase assay is at pH 7.40, the effective anionic concentration of these analogues at a given molarity should be approximately twice that of a com- pound which titrates at pH 7.6 or above, as do the natural inhibitors. Therefore, all other factors being equal, these two compounds should be almost twice as effective inhibitors as those isolated. The results of the inhibi- tion assays on these and the other inhibitor analogues are discussed below.” F) Activity of the Nitrogen Containing Analogues The p-nitro and p-nitroso analogues of syringalde- -hyde were prepared in an effort to determine the importance and Specificity of the aldehyde group for the observed inhibition. In addition to these, 4-nitroso-2-methy1phenol and p-nitrophenol were commercially available (see Materials). In all cases, for the reasons discussed in Section III E, if the aldehyde group only functions to lower the pH of the phenolic hydroxyl, all of these compounds 95 should inhibit lactic dehydrogenase to a greater degree than their correSponding aldehydes. At a concentration of 4.5 x 10°?M, at which syring- aldehyde inhibits lactic dehydrogenase 62%, the p-nitroso and p-nitro analogues inhibit 27% and 32% respectively. Unfortunately, not enough of these inhibitors could be pre- pared for a.more detailed examination of their inhibition. However, it is apparent that rather than being better inhibi- tors than syringaldehyde, they fall short of being even as good, despite their lower pH. The 4-n1troso-2-methylphenol purchased from.Aldrich was technical grade, and dSSpite attempts to purify it, the preparation remained somewhat impure. It was of interest to note, however, that 2.0 x io'fifl "methyl" vanillin will inhibit lactic dehydrogenase approximately 80%, while this nitroso analogue inhibited only 26%. Although admittedly impure, the results with this compound support those cited above for the syringaldehyde analogues. Since p-nitrophenol and p-hydroxybenzaldehyde contain only the phenolic hydroxyl and the two functional groups being compared, aldehyde and nitro, these two compounds are ideal to test the hypothesis concerning the importance of the aldehyde group in the inhibitors. At twice the concen- tration of p-hydroxybenzaldehyde necessary for 100% inhibi- tion of lactic dehydrogenase, p-nitrophenol inhibited the enzyme only 40%. It was therefore concluded that the alde- hyde portion of the inhibitor molecule is Specifically 96 important structurally for activity, although the exact mechanism for this is an yet unknown. As will be shown in Figure 25, the potency of the inhibitors increases with increasing electronegativity of the substituent in the 3 position of the ring. It was therefore concluded that the ideal compounds to test would be the 3-fluoro or 3-chloro p-hydroxybenzaldehydes. Unfor- tunately, these are very difficult to synthesize and are not available commercially. However, 2-chloro-4-nitro- phenol was commercially available, and its activity is com- pared to that of vanillin in Figure 8. It can be seen that this compound is also considerably less potent than the aldehydehyde containing inhibitor, vanillin. This confirms the conclusions that the aldehyde group is required for maximum inhibitor potency. It is clear, then, that the aldehyde portion of the naturally occurring inhibitors and the non-nitrogen con- taining analogues functions in some as yet unknown way to increase the activity of the inhibitors. This effect is in addition to merely functioning to withdraw electrons from the ring and thereby lower the pK of the phenolic hydroxyl group. However, it is not irreplaceable, since cyclotene has no aldehyde, and the most potent inhibitor found, the bisul- fite addition product of vanillin, has its aldehyde group altered, although it does contain a hydroxyl group in the same position which may serve the same function as the alde- hyde in binding to the enzyme surface. 97 .d >H soapoom ca use recaps: Hoes: confiscate mm cow: was momma omosowoathoo caposa oomabom .maopapa22a enmcewoaoaSSU capoma ms Honona uoap«:n:n0noasolm can caaadsmb ho somdamaaoo .m oaomdm 98 m ossmaa as x >523: : _= a Q ~ 9 .c. «a a dJ-JIJIIIJIJ'J . qN _ 355.2..— :— lelazcdzulw z_._.__z§ 33.2.12. :3 a 3552:: 1.1.2351 e: 23.; 3 2325.28 2 a a e s e 2 HBO Did NOIiIOlHNI IV. Kinetic Evaluation of the Natural Inhibitors and the Synthetic Analogues and Proof for the Structural Basis of Inhibition A) Requirement for a Modified Lactic Dehydrogenase Assay Upon measuring enzyme control rates as a function of time in connection with the testing of inhibition by the MSF, it soon became apparent that the measured specific activity of-the lactic dehydrogenase was decreasing with time. This loss in Specific activity was found to be cor- related with the length of time the enzyme was kept at the low concentration used for the pre-assay stock solution (about 5:6 ug/ml). Thus a new assay had to be developed to insure a stable native enzyme control for the study of the kinetics of the lactic dehydrogenase inhibitors. AS indicated in the literature review, there is evidence that at low concentrations lactic dehydrogenase may be inactivated due to dissociation (Millar, 1962) or due to surface denaturation depending on the type of ves- sel in which it is stored (Epstein, at 21., 1964). A study of this inactivation is described below. The enzyme was brought to the pre-assay concentration of 5.6 ug/ml in either glass or plastic test tubes at 0°C and assayed at various times: As shown in figure 9, it is apparent that although glass may adversely affect dehydrogenase renatura- 99 100 .madspoe new use» com Jaopmmononp mead» msoaamb pm crammed one towns» caveman one madam secs 5 .83 case on Sea $85 :a HS\w1 w.m op pounded mos oShuSC one .msoapmeSoosoo 30H pm wadeampm sons ommSow nonsense canosa ac soapobdpoooH .m oaswdm 101 o oaowam . $52.: .2223:— SE< as: can 2.3— .5 8 8 on ° d d d o/E/o as "EVE /0/0 .1114 3.: mafia . El- mzozszzmuzoo 33 E @2325” :2: =3 .3 222255.. IELII: / 9%., Dr. A F D 2 3 S— illMiOV mo and . 102 tion, for dilute solutions the enzyme is more stable in glass. After a very rapid loss of enzymatic activity over the first twenty minutes, the remaining activity seemed to remain.stable in glass for at least two hours, whereas the enzyme stored in the plastic tubes continued to deteriorate. These results demonstrating a decrease in Specific activity of lactic dehydrogenase at low concentration may support those of Hathaway and Griddle (1966) and Millar (1962) demonstrating a molecular change for the enzyme from tet- ramers to presumably inactive dimers at low concentration. {A preliminary sucrose density gradient eXperiment raises some doubts about this, though, Since no enzyme activity ‘was found at the dimer position. (This assumes the disso- ciation is a "cold requiring process" and is reversible lipon warming, as has been shown for several other enzymes .1n.this laboratory (8. Constantinides and S. Blatti, Irrivate communication). Since the primary objective was to Stablize the enzyme and not to explain its instability this irrteresting aspect of the study was not pursued further. This decrease in Specific activity could be overcome 'bor increasing the concentration of the pre-assay enzyme solu- ‘biuon tenfold, to 56 ug/ml. At this concentration, the Speci- fic activity of lactic dehydrogenase was found to remain ‘3<>Ilstant for at least 48 hours. The increased amount of eFlayme present in the cuvette under these conditions Ire<1uired a new assay protocol. This new assay which was used for all of the kinetic results described below is 103 described under Assays. B) Stability of the Inhibitors Prior to conducting detailed kinetic studies, it seemed advisable to test the inhibitors for stability, as the native enzyme had been tested. Solutions of the natur- ally occurring inhibitors which had been Stored for several weeks were found to decrease in activity, as shown in Table 8. [ % Inhibition Inhibitor Initially After 1 week After 2 weeks vanillin 54 42 0 Syringaldehyde 62 59 52 cYolotene 39 30 21 Table 8. Decrease in activity of three naturally occurring ixrhibitors stored at pH 7.50 at room temperature, while elltjposed to light. V nillin was tested at an assay concen- tration of 14101 x 10" _, while the other two compounds were at 4. 5 x 10" Also, both the naturally occuring and the synthetic inhibitors Were very unstable in basic solution, even at pH's between 7 auuc1 8. This observation explains the earlier loss of activ- 11:3, when the active region of the NSF thin layer chromato- E53l!‘£:.phsswas subjected to rechromatography in the same-methyl e1'31'1yl ketone solvent system, which had a pH of 11.0. Vanillin, 104 eXposed to basic solutions, gave a Spot which rechromato- graphed with an R of almost 1.0 in this solvent system, f which correlated exactly with the R of the decomposition f product of rechromatographed MSF. The mechanism of this base catalyzed decomposition is most likely an attack of the ionized hydroxyl group upon the reactive aldehyde portion of the inhibitors, forming an inactive polymeriza- tion product. In molecules with a sterically hindered hydroxyl or with the aldehyde absent, such as in syring- aldehyde or cyclotene, the observed decomposition is slowed, as seen in Table 8. It was also of intereSt to test the effect of air and light on the decomposition of the inhibitors. Solu- tions of the compounds in Table 8 were prepared at pH 5.5 in distilled water to minimize the base-catalyzed decompo- sition. It was found that over a period of two weeks, the samples kept in air exposed to light lost some activity, but that by storing the samples either under nitrogen or in the dark, full activity could be maintained for months. In addition, it was later found that solutions of the inhibitors kept at a pH between 7.0 and 7.5 were completely stable, provided that oxygen was excluded and the solutions were kept in the dark. These conditions were used to routinely store the stock solutions of the inhibitors in the animal testing described in Section V. V 105 C) Reversibility of the Observed Inhibition Although the previous results suggested that the inhibition by MSF was reversible, more direct proof of this reversibility was sought. Therefore, solutions of lactic dehydrogenase containing sufficient inhibitor to completely abolish all enzymatic activity were prepared. The medium contained 0.01M’NaCl and potassium phoSphate buffer at pH 7.40. Vanillin, p-hydroxybenzaldehyde and the bisulfite addition product of vanillin were chosen as representative inhibitors to be tested. The inhibitors were added to the enzyme solutions to yield a final inhibitor concentration of 1.0 x 10-3M, The resulting solutions were divided into two portions. One was dialyzed against the phoSphate buf- fer medium for 48 hours at 4°C, while the other remained in the test tube for 48 hours at 4°C. Assays showed that the nondialyzed controls were still completely inactive. However, the dialyzed samples had regained full activity. Subsequent addition of inhibitor to the reactivated enzyme once again inhibited its activity, demonstrating that no irreversible changes had occured from the eXposure to high concentrations of the inhibitors. A final check of this conclusion was made. A sample of lactic dehydrogenase was stored for 48 hours at-4°C with sufficient vanillin to inhibit 75$ of its activity at zero time. After 48 hours at 4°C it was inhibited to the same degree, which indicated that there were no further reactionS, such as irreversible 106 inactivation, occuring with time. It was therefore concluded that the action of these inhibitors was reversible, and that they were apparently forming a non-covalent bond at the enzyme surface. D) Lineweaver-Burke Reciprocal Plots for the Inhibitors at Various DPNH Concentrations The next experiments were designed to kinetically characterize the mechanism of the inhibition. Figures 10-16 depict the Lineweaver-Burke reciprocal plots at various DPNH concentrations for the inhibitors shown in Table 9. It should be noted that_the zero concentration inhibitor line on figures 11-16 has been copied from that of figure 10, which was determined very carefully and precisely. There- fore, the actual eXperimental points for this line are shown only on Figure 10. AS can be seen from the shape of these plots, the inhibitors are non-competitive with reSpect to DPNH. It was Shown in the literature review that in the case of substrate, there is a compulsory order of addition to the enzyme, pyruvate following DPNH. If this were the case for the inhibitors, one would eXpect uncompetitive inhibition. The observation of non-competitive inhibition indicates that the inhibitors may bind either to the enzyme- DPNH complex or to the enzyme alone, and may serve to eXplain in part why the inhibitors have such low Ki's. The Km for DPNH for lactic dehydrogenase calculated 107 Table 9. The structure, length of substituent group in ring position 3, pH and K1 of the inhibitors used in the kinetic studies described in detail in the text. The pH values were determined experimentally as described under Methods. COMPOUND p-hydroxy- benzaldehyde protocatachuic aldehyde "methyl" vanillin vanillin "ethyl" vanillin u ethoxy n vanillin bisulfite adduct of vanillin 103 STRUCTURE Table 9 SIDE CHAIN, lENCTN 1.084 2-392 2.639 3-959 4.175 5-495 3-959 7-90 7-75 7-90 7.60 7.90 7-65 7-65 109 .4 .5 aoapoom SH confluence momma eomabom .sdflaacm> an mammowoaohgoe oapodfl mo mzmn op pooamoa spas soapandSSa obapapoaaoosos moapmapmsCaoe poaa Hmooaaaooa omaomuaobmozosdq .oa thMam 110 ad 3 0H oasmfia we 2 case: as: a S 3 . d _, a... 3. =3 .3 22:25: 23:2; .2. o.— 3 3 an A 2°” /l 111 .4 .3 soapoom as confluence momma domabom .sdaaaomb :hHOSpos an ems Inowoaohnoo caused mo mzmn on pooamoh and: sodpana::H obapapooaoosOs moapoapmSCEoe soda Hmooaaaooa oxasmlnobmozosdq .aa oaowam 112 we .1 Ease. =Ee\_ and HH mazmah =.~ . o.— a a... a... o.~ =3 .5 22:22:. z_._.=z<> 33:5. 9.— a... 113 .4 .>H Soapoom ad oopaaomoe humus oomabom .oeamoeas odosompsoopona no one mnemonehsoe Canada we mama op pooamoa spas Soapapassd obapdpoaaoosO: moapmapmsoaoe poaa Hoooaaaooa omnomiaobooSooaA ..NH shaman l 3. NH chem: me 2 Ease: see) 3 . ca . c.— _ . Ill! 9: a. QN. — . .3.— 3 . zeta—=2. 3:392 2:23:88”... o.— a II c°"‘/| 3 115 .4 .>H Sodpoom :« confiscate homes oomabom .oehsoeamuaonhwonohnia an odd Inowohemsoe capoma mo mzmn on economy Spas Soapananaa obdpapoaaooSos moapohpmsano poaa Hmoohadooa oxaomlho>SSSCSaA .ma madman 116 no ensues we a :53: 5115\— 3 3 a S a a... a... . a a . _ _ .\. . o { .. \\\\\\\mxmmn. AV\\2V .2 ea. .axuux no no A“ \\\MV\AW\ AV\\AV a .e a 3 . X AV ve_seu zoo do zap—:==z. 3:353sz . >355...— 117 .4 .>H monsoon as sooawomoo homes oomabom .SaHHacmb cahsnoas an one icowoaoasoo oapoSH mo mzmn op accomoa and: soapdnda:a obapdpoaaoosos moapmapmooaoe poaa Hmooaaaooa oxaomlaobooSoSaA .ea oaswam 118 ad 3. as a :53: .223 3.. ea oaomfim 3 c.— 3. a... 9.“. =3 .3 2222::— z_._.=z<> LEGS: a... a.— 9m 119 .4 .3 zodpomm 5 33983 manna doma>mm .caaad:d> =thpm= an 0mm uzowonuhsmc capoma mo mzmn op pomnmon spas nodpanunca mbupapomaoonos mcaymapmnoamd poam ascengaomn exasmlnobmozoqu .ma mafiwam 120 ad 9+ #1: x m.— Ams g 23:2; EEK— ad QN a; .w: 9;. .wu. ==d me zo::==z_ 23:2, 2.5.. an 121 .¢ .>H nod» noom ca dmnahomwd hamma ummabmm .nafida:m> Mo posuoug noddeda madmasmdn on» an own Icemandhnod capoma mo mzmn on pomnmon suds soapanasnd mbapapogaooco: wcapdnum:oamd poam Haoonmaooh mxnfimlnm>aosonda .ma onswdm 122 9... ‘3. e: x 5.32.. 5:33 a.” 0H mHSme 3 o.— \.. so 44 .2 x 9.2 a... a... .2. a... a.— as nhvnl .9.» X MMW. ad =9.— 3 . 22.22:! 522 :53... -235: S _ 123 from the intercept of the zero concentration inhibitor line is 5.0 x 10'6fl, agreeing well with the value reported by Hakala, g§_g;, (1953) of 2.6 x 10‘6fl, The Ki's for the inhibitors are essentially the same as those calculated from both sets of reciprocal plots for each inhibitor. E) Lineweaver-Burke Reciprocal Plots for the Inhibitors at Various Pyruvate Concentrations Figures 17-23 depict the reciprocal plots at various pyruvate concentrations for the series of natural and syn- thetic inhibitors whose structures are given in Table 9. As in the DPNH plots above, the zero inhibitor concentration line was determined for only the first figure, and was copied for the others. The shape of the plots shows that all of the inhibitors are competitive with reSpect to pyru- vate, which agrees quite well with the known inhibition of lactic dehydrogenase as discussed in the literature review. An Eadie-Hofstee plot of lactic dehydrogenase with varying concentrations of pyruvate as shown in Figure 2# yields a Km of 1.10 x 10'5fl,for pyruvate. The intercept of the zero concentration inhibitor line on Figure 17 yields a Km for pyruvate of 1:05 x iO-fifl. Both of these values are consis- tent with those in thegliterature: 5.2 x 10'5fl,(Meister, 5.121. (Hakala. 3.1:. gr. 1953). 1950) and 1;? x 10- The length of the substituent side chain in position 3 of the aldehyde ring is also given in Table 9. It is 12# .d .>H soapoom ca donahomoo momma oomdbom .sdaaasmb an and Isowohohnoo capoofl mo opm>snha op poommon Spa: soapandssa obapaponaoo wsdpsnpmsoaoo poaa Hooonaaooh ostmnnoboozosaA .BH onswdm 125 m. H mhsmam me x :53: :55: am am a“. 3. 3. \ W 0 oo\m“@\ «a \wum q \.wn_\o K . \m\o\ No —x «.n =3 3 \Fe . 3 22:25: a 5:2: _ _ Ill—l1 _ as u, 0'. 126 .4 .5 838m in 25283 momma oomabom .saaaasob :hxospoz an mow IsowonomSoo oapodd ho opsbshhm op poodwoh Spaz Soapandsnd mbapaponaoo wcapwnpmnoaod poaa Hsoohaaooh mansmlaobooSosaq .ma ohfiMam 127 3 ma oszmam me x :25; “Eat: S S 3 3. 3. . 2. u u d, V\ . \‘ nu n- \ o \o 4 =3 3 . 22:21.2. 2...:— x N." z_._.__z<> =>x==1u= 128 .4 .>H soapoom ca dondnomoo humus oomdbom .oohnoodm oassodpdoopOHm kn omd Isowoaohsoo capooa mo opobsham op pooamon and: sodpapdnsa obapapogaoo wsapdnpmsoaoo poHn Hooonadoon mansmlnoboozoSaq .mH onswdm 12C) ma mnowdm , in... x :53: ”52:: . 3.. 3 am am a“. 2. . a. 2. J - AV n x =3 3 o 27... 3 22:21.2. .. . 3:3: 2 a x 2 4\ 2:55:22... 1H... 130 .d .>H soduoom Ga dondnomoo humus oomdbom .ouhnooamusophxoaohalm an coo Isowonu52oo capooa mo opofishhm op pooamon Spas soapanaSSa obapapoqaoo wsdponposoaoo poaa Hooonaaoon oxnsmlhobmoSosaq .om onswdm 131 .m. 3. om mcsmdm me 1 case. Essa: . 3. on 3 3 q d . =.=‘\\\_ L w :3 .3 22:55: “5535sz |>xoma>=nm 132 .< .sH ooduoom ed confluence hommo oomdbom .saaadssb :HhSpoa: hp and IQQMOHohsoo capoma mo opsbshha op pocamoh Spas coapdpassd chapaponado weaponpmsoaoo poam Hooonndooh oxhsmlnobsozosaa .flm enemas 133 3. Hm masmdm me x :23: 5:5: 3 3 3 3. 3. d a d d V\ :3 3 22:25: 23:2; £5.52: 13h .4 .3 838m 5 .ooflnomoo humus oomabom .zdnadscb :Hhsuoa hp and unowonohsoo oauooa yo opobnhha op poonmon sud: soapapdssa chapduogaoo wsdpwhpmnoaoo moan HdoonnaoOH oxhsmlhobdozosdq .NN oaswam mm madmam ing s :33: SEE: " 135 a S 3 . . 3. . a. _ a. a a u + I s . \oNJo . o \ \ o \D q DOD a a... \ D O \ o\ .. n\u K 2 .2 x 3 . \ zogmifi I a 2.. 3.23.? 7 3 23:5; 4 2.5.. _ 136 .4 .>H nod» noom ad donanomow humus oomabom .saaaasdb eo gaseous godpaoem oudofismdp on» an emu Icowohohsoo capooa no oudbflhhn op poonmoh Spa: soapandssa obdpapoaaoo wadponumsoaod poam HooohnaoOH exasmlhobookosaq .mm onswdm 137 3 3 mm magmas in... x :53: “SEEK a 3 3 3. a _ d =3 3 22:22:. :52: 3.333 Iz_._.=z<> 138 Figure 24. Eadie-Horstee plot of lactic dehydro- genase as an alternate determination of the Km for pyruvate. See text for details. 139 —— ' T I I I I4 12 10- v x 10'2 u l r EADIE—HOFSTEE PLOT OF [DH o 25 5'0 15 (v/PYRUVATE moumm x 165 Figure 24 150 125 1HO readily apparent that the K 's are not a direct function of i the size of the substituent, and it was therefore suspected that the differences in the Ki's must be due to the electron withdrawing or donating effects of the various substituents. F) The Molecular Basis of the Inhibition 1) The Kl's of the Inhibitors as a Function of the Relative Negative Charge on the 3 Position of the Aldehyde Ring Such effects as electron withdrawing or donating are very difficult to quantitate, since aliphatic systems such as these substituents are not generally considered in the lengthy tables of such effects on aromatic systems. An additional problem is posed by the presence of a charged group in the ring, whose effect on the 3 position to which the substituents are attached cannot be accurately calcu- lated without molecular orbital calculations. which have not been sufficiently sephisticated to even approach as complicated a problem as an aromatic ring system. These difficulties not withstanding, an approach to the quantiti- zation of the electronic effects of the substituent groups can be made by the method of Taft (1956), who estimated the electronic effects of such groups in aliphatic systems. If we~first assume that the contribution to the 3 position of the aldehyde ring from the ionized phenol to be some arbit- rary value x. then we can base the effects of the substituent 141 groups on X. Furthermore, as was described in the litera- ture review, it has previously been found that the presence of an electronegative group alpha to the anionic portion of the inhibitor produced a more potent inhibitor. One would therefore eXpect naively that the higher the relative charge on the atom directly attached to the 3 position of the ring, the better would be the inhibition. Calculating such a relative negative charge using the values of Taft (1956), where the greater negative charge is more negative in sign. and plotting them against the observed Kl's for the first six inhibitors in Table 9 yields the curve depicted in Figure 25. As can be seen, the Ki's are a function of the relative negative charge en the 3 position of the ring, although the interpretation of the exact nature of the curve is difficult, if not impossible. 2) Effect of pH on the Observed Inhibition by Vanillin Since these phenolic inhibitors all have pK's between 7.60 and 7.90, they are not completely ionized at the pH routinely employed for the assay of lactic dehydro— genase, pH 7.40. The inhibition observed at a given con- centration of the inhibitors should therefore increase with increasing pH of the assay. That is, at higher pH a larger fraction of the ionized anionic form, which is pre- sumably the active inhibitor form,will be present. Vanil- lin, with a pK of 7.60, was chosen as a representative 1H2 .m cedpamoa mafia no aopm Sade on» so czaob to 2.3 no gapgsoamo on» no coapa«nomoo ovoaaaoo a Mom use» com .0 canoe ca sobdw omonp one mosao> «M.osa ..omo:omohoh£od capooa mo soapandssd on» so mucoSpapmpzm mafia mo poommo 0:9 .mm ohswdm mm mysmfim m 2228.. 2:: .562 5:: Lb maid.— ..i 0+ _.+ x ._r n... 1L8 - u a fi A - u d d o 22:25: :3 2.: SEE 535.33 2:: - — n n — p - p n n ! '(II! )I 1&4 inhibitor, since it was readily available and could be assayed at low concentration, due to its high activity. Assays were performed at pH 7.00, 7.00, and 7.80, which corresponds to 20, #0 and 60% ionized vanillin (Figure 26). As can be seen, the increase from pH 7.00 to 7.h0 resulted in a doubling of the inhibition as would be quantitatively predicted from the ionization of vanillin. In apparent contradiction to this working hypothesis, how- ever, the further increase from pH 7.h0 to 7.80 had no effect on the inhibition. It was therefore concluded that in this pH range, titration of groups on the enzyme surface was occuring. Presumably, these could be those groups responsible for the binding of the inhibitors. As discussed in the literature review, lysine and histidine are the residues presumed to be reSponsible for this bind- ing. 0f the two residues, it seems more likely that his- tidine would.be titrated between pH 7.h0 and 7.80, although the environmental electrostatic interactions could lower the pH of lysine sufficiently for it to titrate in this range. It should be noted that maximum inhibition was achieved at pH 7.00, which is very close to physiological pH and is the pH at which all assays of the inhibitors have been performed, so the inhibitors would be eXpected to be effective:ig.vivo as well as in,vitro. 1&5 Figure 26. Effect of the assay pH on the vanillin inhibition of lactic dehydrogenase. Assays were as described under Methods, with the substitution of 0.10fl,potassium phOSphate buffers at the indi- cated pH's for the standard assay buffer. PER cm mmambn 70 60 ’ '50 '40 20 IO 1&6 mm or ASSAY pH on THE vmum mmamou or Lou ' -/° C) /’ J“ , //8 D/B/ / . C) III-D pH 7.00 0-0 pH 7.40 I {D-4I' p" 180 . 0.5 1.0 - 1.5 2.0 vnmum nounm x 10‘ Figure 26 107 G) Comparison of the Vanillin Inhibition of the M# and H“ Isozymes of Lactic Dehydrogenase 1) Introduction Many of the previously discovered inhibitors of lac- tic dehydrogenase have different effects upon the two dif- ferent lactic dehydrogenase isozymes, Mn and H4 (see litera- ture review). It was therefore of interest to determine if these phenolic inhibitors exhibited such a selective inhibi- tion. Again vanillin was chosen as a typical representative because it was the most active inhibitor. 2) Determination of the Purity of the Isozymes Before such a study could be performed, it was neces- sary to demonstrate the purity of the isozyme samples to be used. Polyacrylamide gel electrophoresis of beef heart lac- tic dehydrogenase revealed the presence of two protein com- ponents. when these gels were stained for activity as described under Methods, both bands were shown to be active. These two components of beef heart lactic dehydrogenase are generally found to be active, (see literature review) and these results support this conclusion. The two components were shown to be the H4 and HBM isozymes by comparing their electrophoretic mobility with hybrids produced by freezing rabbit muscle and beef heart 1h8 lactic dehydrogenase together in 1M.NaCl at pH 7.00, after Salthe, gt,§;, (1965). A densitometer tracing and integra- tion of the beef heart enzyme electrophoresis showed the. an and $3M isozymes to be present in a ratio of approxi- mately 5:1. Previous work (Kaplan and Cahn, 1962) has shown that the individual isozymes in a mixture exhibit properties identical to those found for the pure parent type isozyme, and such activity is directly proportional to the amount of that isozyme present in the mixture. 0n Vthis basis, this sample of beef heart lactib dehydrogenase icontained one M subunit for each 15 subunits of the H type, and therefore was 94% pure heart isozyme. This enzyme was used to determine the effect of vanillin on the heart iso- zyme without further purification. The conditions for polyacrylamide gel electrophor- esis described under Methods could not be used to determine the purity of the rabbit muscle enzyme since at pH 8.60, this isozyme migrates toward the cathode and does not enter the gel. Therefore, polyacetate electrophoresis at this pH was employed. A densitometer tracing of the resulting pat- tern as above showed it to be essentially homogeneous Mu lactic dehydrogenase, so it was used without further purifi- cation. 3) Vanillin Exhibits Greater Inhibition of the H” Isozyme Figure 2? depicts the inhibition of both isozymes of 1M9 Figure 27. Comparison of the vanillin inhibi- tion of the Hu and Mn lactic dehydrogenase isozymes. Assays were described under Methods. PER CENT INHIBITION 70 50 30 20 I0 150 COMPARISON OF VANllllN INHIBITION OF THE 2 [DH ISOZYMES as 1.0 1.5 muum vacuum x 10‘ Figure 27 2'0 151 lactic dehydrogenase by vanillin. As can be seen, the H” isozyme is inhibited to a somewhat greater degree than the Mn isozyme. These results are in good agreement with those discussed in the literature review, since differ- ences in either inhibition or activity of the two isozymes with various compounds are generally found to be small (Fondy and Kaplan, 1965; Kaplan, gtmg;.. 1960). V. Direct Testing of the Theory of Inhibition of Cancer by Inhibition of Lactic Dehydrogenase A) Testing Procedures All testing performed in this laboratory followed the protocols described by the Cancer Chemotherapeutic National Service Center (CCNSC). The sarcoma 180 was from a line of tumor approved by CCNSC for use in a screening program of antitumor agents, and was kindly donated by Dr. Everett 8. Beneke of the Michigan State University botany department. For the eXperiments described below, white, Swiss-Webster mice 15-25 days old and weighing approximately 20 grams were used. They were kept in metabolism cages with free access to common laboratory chow, with tap water for drink- ing, except where 13% glucose was specified. The mice were first accustomed to the metabolism cages for at least two days prior to the beginning of the test. Tumors from carrier mice were excised and immediately placed in a solution of saline containing 1.0 mg/ml chloromycetin to prevent bacterial growth, where they were cut into small pieces approximately 1-2 mm square. These pieces were trans- planted subcutaneously in the axilla region of a pre-weighed mice by trocar. These test animals were given a single intraperitoneal dose of compound to be tested daily on days .1-7. They were weighed again on day 8: weight change was 152 153 taken to be a measure of the toxicity of the compound under test. The mice were then sacrificed by cervical disloca- tion and the tumors excised and weighed on the Mettler balance as soon as possible. The LDSO (the lethal dose for 50% of the animals) was used to provide a measure of toxicity. Determination of this quantity was as follows. The LDSO for each of the compounds described below was determined by using groups of six mice each and plotting the per cent mortality against dose to find the single dose level at which 50% of the ani- mals died. The metabolic rate of the compounds was determined as suggested by Koppanyi and Avery (1966). The procedure is to accurately determine the LD50 for a single dose and then to divide this dose evenly into several doses admin- istered over several hours, and to calculate the extra amount necessary then for 50% mortality. The additional amount divided by the number of hours over which the admin- istration proceded yields the metabolic rate of the com- pound in mg of compound metabolized/kg body weight/hour. It should be noted that this is only an approximate value, Since compartmentalization and other factors can influence this value. However, the determination of a more exact metabolic rate requires an extensive study with radioactive tracers, and the eXpense for labeling the compounds described (below was prohibitive in view of the limited value of the information to be gained. 154 B) Early Results Demonstrating MSF inhibition of Sarcoma 180 in Mice 1) Requirement for High Levels of Dietary Glucose for Inhibition It had been reported that in order to effect a can- cer cell 11:1. mg by inhibition of lactic dehydrogenase, it was necessary for the organism to be maintained on a diet high in glucose, to force the use of glucose as the primary source of energy via glycolysis (Goldberg, Nitowsky and Colowick, 1965). To test this requirement in the present system, two groups of mice were inoculated with sarcoma 180. One group was given 13% glucose and the other tap water, while both were daily dosed with 50 mg/kg of MSF. After excision and weighing of the tumors, it was found that the group drinking plain tap water had tumors four times larger than the group on glucose (see Figure 28), supporting those results cited above. It should be noted that glucose had no apparent effect on a group of mice with no tumors, so the results described below all were obtained from mice maintained on 13% glucose rather than plain tap ”at 91' e 2) LDSO for the MSF in Mice Using MSF doses of 5, 20, 50, 200, 500 and 1000 mg/kg 155 .nopu: an» ahead wcaaeano .m .onoosaw mma weaanano .H . .pswaoz hoop wM\mms we cm and: added deacon» ones mASoaw spom .pwop on» ad confluence mm .mpoao omoosam 30H use swag so coda Scam some» whose» oma maooamm 0:» mo condhmaaov .mm madman 157 body weight, the LD50 was determined to be 750 mg/kg. The mice alive after ten days were sacrificed by cervical dis- location and the viscera grossly examined for evidence of toxicity. No gross effects on the viscera were observed with any of the compounds tested. 3) Pronounced Effect of the MSF on Sarcoma 180 in Mice This section describes the direct testing of the MSF as an inhibitor of the $180 tumor. Seven groups of mice bearing sarcoma 180 were given the doses of MSF: namely 10, 20, 25, 35, so, 50 and 100 mg/kg body weight/day. The external appearance of such mice is depicted in Figure 29. The excised tumors from these dosed animals are shown in Figures 30 and 31. Figure 32 depicts the dose response curve for the MSF. These results were quite encouraging, but as is described below, were not reproducible. C) Slight Effectiveness of the Inhibitors Against Three Different Tumor Systems as Determined by CCNSC Since all proSpective cancer chemotherpeutic agents must be screened through the Cancer Chemotherapeutic National Service Center (CCNSC), the MSF was submitted for testing against three different test systems. In contrast to our observed inhibition of approximately 80%, CCNSC observed an inhibition of sarcoma 180 of only 6-20%. Furthermore, the 158 Figure 29. External appearance of sarcoma 180 bear- ing mice. Upper mouse was treated daily with 150 mg MSF/kg body weight. Alcohol was swabbed on the abdomen.to facilitate visualization of the tumor. The lower mouse is a water treated control. The photograph was taken five days after transplan- tation of the tumors. Figure 29 160 .pwop us» ca confluence one madopoo Hopsosanoawm haso Hopes and: oopoonsa .Honpsoo .: aoo\psmaox moon ma\am= ms mm .m aoo\armaox econ ma\mmz we mm .m aoo\o:mnos aoop wx\am= ms oma .H ”mzoaaow mm .mm: no momoo msodnmb Spa: condemn coda scam condoms ences» oma maooamm .om oaswam 162 .pNov 0:» ad confluence one cadence Hopnoaanoawm haso Hope: and: depooflsd .Honpsoo .m aoe\oamaox hoop mx\cn= we ca .3 aoo\pamdor soon wa\ams ms mm .m aoo\uswaox aoon mA\mm= we mm .m aoo\psmaox anon ma\mm= we on .H .msoaaom mm .mm: mo mowed msoaaob and: smooch» coda scam domaowo muoasp oma maoonmm .am onsmdm Hm minim 164 ..puop one ca confluence he an: and: cocoon» code as oma daooamm you chase omsoamon omoa .Nm onswam 165 mm oasmaa 33:3 58 .8. \mm: .25 m8: 2: 2 2 ES O , “as. E 83 .5 zeta—=2. a m. a a a." a Elam; as. e_\az a: as umummu ‘ was an 166 inhibition was not proportional to dose, and so this was felt not to be significant inhibition. No toxicity was noted. It should be noted that the metabolic studies described in Section V CZ were performed after the cancer inhibition studies. The leukemia 1210 system was inhibited only 3-8%, and once again the inhibition was not proportional to the dose. Therefore, the MSF seemed to be ineffective in this test system as well. Although the Walker 256 intramuscular carcinoma was inhibited 10% at 100 mg/kg, the rats lost weight excessively and so the slight therapeutic value was outweighed by its toxicity. At this point, efforts were intensified to purify the inhibitors and assume their sta- bility in the cancer test section (see Section.IV B). After the active inhibitors of lactic dehydrogenase had been isolated from the MSF, it was felt that they merited a test in the above systems, and they were tested against both the sarcoma 180 and leukemia 1210 systems. Vanillin, the most potent natural inhibitor, and cyclotene, another of the original naturally occurring inhibitors, were tested. Cyclotene was completely without effect against sarcoma 180 and gave between 5-15% inhibition of the leukemia 1210. However, the leukemia bearing mice treated with cyclotene lost considerable weight, indicat- ing toxicity, and so the cyclotene was not tested further. Vanillin inhibited the Sarcoma 180 and leukemia 1210 sys- tems 1-16Z and 10-111 respectively, but neither of these 167 are considered significant inhibition in these test sys- tems e D) Why is Vanillin Essentially Ineffective Against Cancer? 1) The LDSO and the Rapid Rate of Metabolism of Vanillin in Mice In an effort to answer this question, it was felt that a knowledge of the rate of metabolism of vanillin in mice would be helpful. Using the techniques outlined in Section V.A above, the L135o for vanillin was determined to be 760 mg/kg, and the metabolic rate was estimated at 525 mg/kg/hour. It can easily be seen that since the mice can metabolize almost four times the test dosages each hour, and are dosed only once every 24 hours, vanillin could not be expected to inhibit cancer in such a test. Increasing the dose could not be effective either, since the metabolic rate approaches the LDSO' and the toxicity would be expected to outweigh the theraputic value achieved by such a proce- dure. 2) Metabolism of Vanillin in Mice a) Introduction Since it was now known that vanillin is rapidly 168 metabolized by mice, it was of interest to determine its metabolic products. Bray and coworkers (1952) determined the metabolic products of ortho, meta, and para hydroxy- benzaldehyde. For p-hydroxybenzaldehyde, they found 67% excreted as the ether soluble acid, 4% as the ester glu- curonide, 16% as the ether glucuronide and 9% as the ethereal sulfate. Since vanillin is quite similar in structure to p-hydroxybenzaldehyde, it was felt these results would probably be quite similar to those found for vanillin. b) Spectral Analysis of Urinary Vanillin Metabolites Urine was collected from 12 mice, after which they were each given a 600 mg/kg dose of vanillin. Urine samples were then taken every hour until the color of the urine again appeared the same as the control (four hours for vanillin). Figure 33 shows the pattern of excretion of the metabolites. This pattern remained the same if the urine optical density is measured at 260, 290 or 480 mu, although the major metabolic product, vanillic acid (see below), absorbs only at 260 and 290 mu. The metabo- lism of the mice must be affected in such a way as to cause an excretion of some colored product together with the metabolic products of vanillin. It should be noted that at the third hour, the peak of absorption in the urine had shifted from near 270 mm to 250 mu, as if the 169 Figure 33. Pattern of urinary excretion of the metabolic products of vanillin in the mouse. Spectral analysis of the urine is described in the text. 170 F'" I I 1: I EXCRETION PATTERN OF VANILLIN METABINJTES 350*- '* aoo- ,' ' - zsor- .' ‘ '. - g e if, u_ ZOOI- a E g ISO- 5 100 h b 95 . o—o zoom» (1! so) If: D- D «o my (1 IN)~ 50h A-«A zoo mu (1 0.1) “a i i nouns mm nos: Figure 33 171 product had been altered in some way. The largest increase in the amount of conjugates was noted at this time. The curve observed in Figure 33 should decrease logarithmically, but apparently the products are excreted so quickly that this is not seen. By 4 hours, the optical density has returned to the pre-dose level, and excretion of metabolites is apparently complete. c) Chromatography of Ether Extracts of Both Hydrolyzed and Nonhydrolyzed Urine from Vanillin Treated Mice To further characterize the metabolism of vanillin in mice, chromatography was employed. The following scheme was used to determine the products of the metabolism of vanillin in mice. 0.2 m1. of each urine sample were mixed with 0.04 ml. of concentrated HCl and extracted with six consecutive 0.5 ml. portions of ether. This ether extract was concen- trated to 100 uL, and 25 uL portions of each urinary extract were subjected to thin layer chromatography on silicic acid in three different solvent systems: 1) n-butanol:acetic acidxwater, 8:2:2 (v/v/v), pH 2.80 2) n-butanolzpyridine:dioxane:water, 70:20:5:5 (v/v/v/v). pH 7.50 3) methyl ethyl ketone:2N ammonium hydroxide, 2:1 (v/v, organic phase), pH 11.00 The phenolic compounds were revealed by spraying the air 172 dried chromatography plates with 5% FeClB. The extracted, acidified urine was hydrolyzed anaerobically for 12 hours at 110°C, and then re-extracted with another six consecu- tive 0.5 m1. portions of ether, which were then concentrated and chromatographed as above. Standards of vanillin and vanillic acid were run with each chromatogram. By far the major product of vanillin metabolism was vanillic acid. Three spots present in the nonhydrolyzed extracts disappeared after hydrolysis, and it was presumed that these were the conjugates as found by Bray and his coworkers (1952) for p-hydroxybenzaldehyde. These conju- gates greatly increased in amount at the third hour,,which agrees with the shift seen in the spectrum of the third hour urine. It is interesting to note that although no free vanillin was ever excreted, hydrolysis revealed the release of vanillin as well as vanillic acid from conju- gates, indicating that many different conjugates are formed, although only three were detected by chromatography before hydrolySiB. E) Decreasing the Rate of Metabolism of Vanillin by Forming Its Bisulfite Addition Product 1) Introduction Although vanillin is a potent inhibitor of lactic dehydrogenase, it seems unlikely that it can inhibit cancer 173 because it is metabolized so rapidly. Since it is apparent that the major site of the metabolism of vanillin is at the aldehyde group, attempts were made to alter this part of the molecule to decrease the metabolic.rate. The nitro and nitroso analogues were not further tested because of their decreased inhibitory activity for lactic dehydrogenase. 0n the other hand, the bisulfite addition product of vanillin both altered the aldehyde functional group and increased the activity. It was therefore of interest to determine the metabolism of this compound. 2) The LD and Rate of Metabolism of the Bisulfite Adduct 50 in Mice The L'DSo for the bisulfite addition product of vanil- lin was determined to be 1850 mg/kg, and the metabolic rate to be 310 mg/kg/hour. In addition to an increased LD50 as compared to vanillin, it has a considerably slower metabolic rate, as would be predicted from the knowledge that it is at the aldehyde group where metabolism occurs, as shown above. Since the bisulfite adduct has no aldehyde, its metabolism would be necessarily slower. 3) Metabolism of the Adduct by Mice Urinary metabolites of the adduct were analyzed in the same way as described for vanillin. Doses of 400 and 174 1400 mg/kg were used. Figure 34 depicts the pattern of excretion following these doses. The delayed 400 mg/kg dose shows the effect of 400 mg/kg for five days prior to the collection of the urine. A drastic change in the pat- tern of metabolism is seen. Whereas the initial doses both yield the predicted exponential decrease in urinary output of metabolites, the delayed dose shows a slight rise at one hour, followed by a later slow rice (see Section V E 4). It should be noted that the initial 400 mg/kg dose peaks one hour earlier than the initial 1400 mg/kg dose. This can be most easily explained by noting that at 1400 mg/kg, the mice are unconscious for about one hour, and appear to be completely concerned with survival, exhibiting greatly increased respiration and pulse, with complete loss of coordination and control of bodily func- tions. The metabolites are the same as vanillin, even to the shift noted from the acid to the conjugates after three hours, although the shift occured at four hours at the 1400 mg/kg dose, probably because of the reasons outlined above. Since the curves in Figure 34 both return to the base line by 10 hours, it was assumed that the entire dose had been eliminated by then. However, a second 1400 mg/kg dose 24 hours later gave almost 60% mortality. Apparently, some small fraction of the dose is stored, and a second high dose raises the stored amount to a lethal level. How- ever, 400 mg/kg can be given for up to two weeks with no 175 Figure 34. Pattern of urinary excretion of the metabolic products of the bisulfite addition product of vanillin in the mouse. Spectral analysis of the urine is described in the text. The delayed 400 mg/kg dose curve repre- sents the pattern of excretion following the sixth daily dose of 400 mg/kg, demonstrating a drastic change from the first dose of 400 Ins/ks. I I I I I T r I :f—r] o Excnmou PATTERN- uooq- II I vmnum BISULFITE AoouoI .- Minimum 120m- . a £5 ,2 IoooI- . i a / = on’ )5 woo too/no " E 600!- y ,5 - g s a . f? n 2' 400 I' ‘. d 2 ee .— g- a —A s. \ zoo A‘7A.’ c'\n\ .. DELAYED ' °""'° D"""D 4NINNb/KG. I l l I I l l “o I 2 3 4 5 615510 nouns mm nos: Figure 34 17? mortality, so even the small stored amount must be slowly metabolized. 4) Induction of the Drug Metabolizing Enzymes Since the pattern of metabolism seen for an initial dose of 400 mg/kg and that seen after five days of this dose is so different, it seemed likely that the change was due to induction of some enzymes reSponsible for the metabolism of the adduct. To test this conclusion, a group of mice were given 400 mg/kg once a day for 12 days, and then were given no dose for two days to be sure all remaining adduct had been excreted. The metabolic rate of the adduct was then re—determined for this group, and it was found to be 660 mg/kg/hour, more than double the value obtained with mice which have never been exposed to such compounds. It was therefore concluded that these compounds induce enzymes responsible for their own metabolism. This is obviously an undesirable characteristic for a compound which is to be used for long term treatment of cancer. 5) Lack of Effectiveness of the Vanillin Bisulfite Adduct in the Treatment of Cancer Tested in sarcoma 180, leukemia 1210, and Walker 256 intramuscular carcinomas, the vanillin bisulfite adduct exhibited the same disappointing lack of inhibition described 178 above for the other tested compounds. The rate of metabo- lism may still be too high to allow reSponse at the tested dosages (100, 200 and 300 mg/kg/day). Discussion Testing Inhibition of Cancer by Lactic Dehydrogenase Inhibition In the course of studies with the Krebs cycle metabo- lites, a potent inhibitor of lactic dehydrogenase was observed. This inhibitor was isolated and purified not only to characterize its inhibition of the enzyme, but also to test the hypothesis that compounds which inhibit lactate production should inhibit cancer. The three elements of this theory are as follows: 1) Warburg (1924) found that cancer cells have an abnormally high production of lactic acid, even under aerobic conditions. 2) Warburg (1956) showed that of all tissues tested, only cancer cells demonstrated a high production of lactate under aerobic conditions (so-called aerobic glycolysis). 3) Boxer and Devlin (1961) showed that the high rate of aerobic glycolysis in tumor cells is caused by the absence in such cells of d-glycerolphosphate dehydrogenase. If conditions could be found under which the glycolysis of tumors was specifically inhibited, the growth of tumors might be specifically inhibited without interfering with the 179 180 growth of cells which depend solely on oxidative metabolism (Papaconstantinou and Colowick, 1961a). Such specific inhibition of glycolysis can only be accomplished by inhi- bition of lactic dehydrogenase, as discussed by the above authors. Any inhibitors to be tested for carcinostatic activity should have the following characteristics maxi- mized: 1) Potency as a lactic dehydrogenase inhibitor. 2) Lack of toxicity in 33.-.19.- 3) A slow rate of metabolism, to allow sufficient time for the compound to affect the tumor before metabolism removes it. Previous workers have tested many compounds for carcino- static activity (Novoa, gthgl., 1959: Papaconstantinou and Colowick, 1961a; Goldberg andColowick, 1965; Busch and Nair, 1957: Davis and Busch, 1958: Ottolenghi and Denstedt, 1958: Fiume, 1960). These compounds have all failed to exhibit signifioalt carcinostatic activity, since they were generally rather inactive lactic dehydrogenase inhibitors (Busch and Nair, 1957), and exhibited high toxicity in ani- mals. Metabolic rates were, in general, not measured. The substituted phenolic inhibitors of lactic dehydro- genase described in this thesis are quite potent as enzymatic inhibitors, having Kl's as low as 14 uM. Two were also shown to have quite low toxicities (LD 's of 760 and 1850 mg/kg) 50 compared to those previously tested. However, the metabolic rates for these two compounds, vanillin and its bisulfite 181 adduct, were 525 and 310 mg/kg/hour. It should be noted that the adduct has excellent characteristics in two of the three categories: 1) It is the most potent lactic dehydrogenase inhibitor reported to date (except for the peptide inhibitors of Wacker and Schoenberger, 1966a, 1966b). 2) It has the lowest toxicity reported to date of any lactic dehydrogenase inhibitor. It fails to meet the third criterion, however, since its rate of metabolism is so high. As might be expected from this high metabolic rate, the compounds are without repro- ducible effect on cancer, thus making impossible direct confirmation of the above theory using these substituted phenolic inhibitors of lactic dehydrogenase. Mechanism of the Inhibition of Lactic Dehydrogenase During the course of this investigation, four inhi- bitors of lactic dehydrogenase were found to be present in a methylene chloride extract of maple syrup, namely vanil- lin, syringaldehyde, cyclotene and p-hydroxybenzaldehyde. It was shown that structural similarity to either lactate or pyruvate was not the common denominator ofsthese active species, since many other structural analogues did not function as inhibitbrs. . Each of the active inhibitors was found to have a pK of 7.9 or below, while the inactive 182 compounds all titrated above pH 9.6. It was therefore con- cluded that in order to function as lactic dehydrogenase inhibitors, these compounds were required to function as anions. This conclusion agrees well with that of Winer and Schwert (1959), who concluded that "The information which is available on inhibitors for the lactic dehydro- genase system indicates that gnly anionic substances £3227 t_i_o_r_1 22 inhibitors." ‘ I The two most active naturally occuring inhibitors were vanillin and p-hydroxybenzaldehyde, which differ in structure only in the substituent on the 3 position of the aldehyde ring. In an effort to obtain a greater inhibe ition, as well as to define the mode of action of these inhibitors, analogues of these two compounds were tested, differing in structure only in the substituent on position 3, as follows (the Ki's are in uM in parenthesis): p-hy- droxybenzaldehyde (414), protocatachualdehyde (174), ”methyl" vanillin (270), "ethyl" vanillin (241), "methoxy" vanillin (= vanillin) (96), "ethoxyfi vanillin (108), and the bisulfite adduct of vanillin (14). These compounds were found to be completely reversible inhibitors of lac- tic dehydrogenase. The kinetic properties of this series of inhibitors were examined. They were found to be competitive with reSpect to pyruvate and non-competitive with respect to DPNH. From purely steric considerations, the Ki's would be expected to be proportional to the size of the substitu- 183 ent groups, but this was found not to be the case. This suggested that the differences in the Ki's must be due to the electron withdrawing or donating effects of the vari- ous substituents. Ottolenghi and Denstedt (1958) found that most substrate analogue inhibitors of lactic dehydro- genase contain both an anionic moiety and an alpha electro- negative group, which could correSpond to the 3 position of the aldehyde ring in the series of inhibitors under investigation here. It was found that the Ki's of this series was proportional to the relative negative charge on the atom directly attached to the 3 position of the ring, which supports the results of Ottolenghi and Denstedt (1958). From pH 7.00 to pH 7.40, the amount of vanillin in the ionized form increases from 20 to 40%, and the observed inhibition of lactic dehydrogenase similarly increases two-fold, suggesting that the anionic form of the molecule is the active inhibitory species. Since further raising the pH from 7.40 to 7.80 does not increase the inhibition, it was concluded that the functional groups on the enzyme surface reSponsible for the binding of the inhibitors, are titrated in this pH range. Nygaard (1956) and Winer and Schwert (1959) postulated these groups to be imidazole and lysine, of which it is more likely the imida- zole which is titrated between pH 7.40 and 7.80, although local electrostatic interactions could lower the pK of lysine sufficiently for it to titrate in this pH range. Since the aldehyde moiety of the inhibitors may 184 function merely to lower the pK of the phenolic hydroxyl group, the nitro and nitroso derivatives of syringaldehyde, "methyl” vanillin and p-hydroxybenzaldehyde were tested, since they should have a lower pK than their aldehyde-con- taining analogues. Since they were not as active, it was concluded that the aldehyde group functions in some way to increase the potency of the inhibitors, in addition to with- drawing electrons from the ring to lower the pH of the phenolic hydroxyl group. Vanillin, chosen as a representative inhibitor of the series, exhibited greater inhibition of the H4 isozyme, although there was less than a two-fold difference between the inhibition of the two lactic dehydrogenase isozymes. During the course of this investigation, it was found that samples of lactic dehydrogenase stored at low concentrations (about 5.6 ug/ml.), showed a rapid loss of activity to approximately 40% of the activity of a control sample stored at higher concentration. This observation may support those of Hathaway and Criddle (1966) and Millar (1962) demonstrating a molecular change for the enzyme from tetramers to presumably inactive dimers at low concentrations. Summary 1) The bisulfite adduct of vanillin was both the most potent (K1 = 14 uM) low molecular weight lactic dehydrogenase inhibitor yet reported (the specific poly- peptide inhibitors are more potent) and the least toxic (LDSO = 1850 ms/ks). 2) The rate of metabolism in mice of vanillin and of the bisulfite adduct of vanillin was very great (525 and 310 mg/kg/hour). As result, at the 300 mg/kg level used for testing inhibition of cancer, the drug is reduced to base line levels‘about 3 hours after adminis- tration,_based on metabolic products in the urine as determined both spectrophotometrically and using thin layer chromatography. 3) For this reason, they could not be expected to and did not show any significant reproducible carcinosta- tic activity against Walker 256 intramuscular, leukemia 1210 or sarcoma 180 carcinomas 1n tests performed by CCNSC. 4) However, 3 series of tests on the MSF in our laboratory gave 75-80% inhibition of sarcoma 180 growth, although tests on MSF by CCNSC did not show significant inhibition. By any standards of scientific objectivity, the testing performed in this laboratory was reproducible. Previous workers have also had considerable difficulty in 185 186 achieving reproducible results in this field. In this case, the explanation may lie in the use of different sarcoma 180 tumor lines in this laboratory and at CCNSC. This conclu- sion is impossible to test, since the tumor line used for testing in this laboratory is no longer available. 5) The inhibition of lactic dehydrogenase by a com- mercial malic acid preparation was traced to the presence of an impurity. 6) Four compounds, vanillin, p-hydroxybenzaldehyde, cyclotene and syringaldehyde, were isolated from maple syrup, the commercial source of malic acid, and found to be active as lactic dehydrogenase inhibitors. 7) In an effort to obtain a greater inhibition than that achieved with the naturally occuring inhibitors, the following series of seven analogues was tested (the K 's, i in uM, are given in parenthesis): p-hydroxybenzaldehyde (414), protocatachualdehyde (174), "methyl" vanillin (270), ”ethyl” vanillin (241), "methoxy" vanillin (= vanillin) (96), ”ethoxy" vanillin (108) and the bisulfite adduct of vanil- lin (14). 8) The kinetic properties of this series of inhib- itors was examined: a) They were reversible inhibitors of lactic dehydrogenase. b) They were competitive with pyruvate and non- competitive with DPNH. c) The K1 was not proportional to the length of 187 the substituents on the ring, as would be eXpected from purely steric considerations. d) The K1 was found to be a function of the rela- tive negative charge on the 3 position of the aldehyde ring. e) The inhibition increased when the assay pH was raised from 7.00 to 7.40, due to an increased concentration of the anionic form of the inhibitors, presumed to be the active form. f) Vanillin, chosen as a representative inhibi- tor, inhibited the H4 isozyme of lactic dehydrogenase approximately twice as much as the M4 isozyme. 9) Nitro or nitroso groups could not fully substi- tute for the aldehyde moiety of the inhibitors. It was therefore concluded that the aldehyde group, although not indiSpensible (since cyclotene and the most potent inhib- itor, the bisulfite-addition product of vanillin, lack any aldehyde group), apparently functions to increase the activ~ ity of the inhibitors, in addition to withdrawing electrons from the ring to lower the pK of the phenolic hydroxyl group. 10) It was found that lactic dehydrogenase shows a marked loss of activity to about 40% of the control when left standing at low concentrations (about 5.6 ug/ml.). but is quite stable for up to 48 hours at higher concentra- tions (about 56 ug/ml.). Bibliography Adams, R. and Levine, 1.. J. Am. Chem. Soc.. 25. 2373 (1923) Adams, R. a2? Montgomery, E., J. Am. Chem. Soc., 46, 1519 (192 Alberty, R. A.. J. Am. Chem. Soc.. 25, 1928 (1953) Albrecht, G. J., Bass, S. T., Seifert, L. L. and Hansen, R. G., J. Biol. Chem., 241, 2968 (1966) ' Anderson, B. M., and Kaplan, N. 0.. J. Biol. Chem., 234 1226 (1959) Anderson, S. H.. Florini, J. R. and Vestling, C. S., J. Biol. Chem., 229, 2991 (1964) Appella, E. T., Brookhaven Symp. Biol., 11. 151 (1964) Appella, E. T., and Markert, C. L., Biochem. Biophys. Res. Comm., 6, 171 (1961) Archibald, W. J., J. Phys. and Colloid Chem., 51, 1204 (1947) Atkinson, D. E., Science, 1 0, 851 (1965) Banga, 1.. Szent-Gyorgyt A., and Von Vargha, L., z. Physiol. Chem., 210, 228 (1932) Beisenherz, G. von, Boltze, H. J., Bucher, Th.. Czok, H.. Garbade, K. H.. MeyerqArendt, E., and Pfleiderer, G., Z. Naturforsch., 86. 555 (1953) Bennett, R., Taylor, D. a. and Hurst, A., Biochim. Biophys. Acta, 118, 512 (1966) Blanca, A. and Zinkham, W. H.. Science, 132, 601 (1963) Block, R. J., Durrum, E. L., and Zweig, G., A,Manual 22 Pa er Chromato ra h and Pa er Electrophoresis, (Academic ress, N. Y., 19 Bolling, D., Sober, H..A., and Block, R. J., Fed. Proc., 8, 185 (1949) Boxer, G. E. and Devlin, T. M., Science, 134, 1495 (1961) Boxer, G. E. and Shonk, C. E., Cancer Research, 29, 85 (1960) 188 189 Boyer. P. D., Proc. Intern. Congr. Biochem., 5,,4, 18 (1963) Brand, L., Everse, J. and Kaplan, N. 0.. Biochem., 1, 423 (1962) Bray, H. G., Thorpe, W. V. and White, K., Biochem. J., 52, 423 (1952) Brody, I. A., Nature, 201, 685 (1964) Brown, D. H., and Cori, C. F. in The Enzymes, P. D. Boyer, H. Lardy, K. Myrback, Eds. (Academic Press, N. E., 1961) V010 59 pe 207 Bucher, T. and Pfleiderer, G., in Math. in Enz., S. P. Colowick and N. D. Kaplan, Eds. (Academic Press, Ne Ye. 1955) VOIO 1! p' ”35 Hertz” D. et al., as cited in Warburg, 0.. Science, 124, 26971936) ""' Burk, D. and Schade, A. L., Science, 124, 270 (1956) Busch, H.. 59 Introduction to the Biochemistry of the Cancer Cell. (Academic Press, N.'Y., 1962) p. 344 Busch, H., Cancer Research,‘;5, 365 (1955) Busch, H. and Nair, P. V.. J. Biol. Chem., 232, 377 (1957) Cahn, R. D., Kaplan, N. 0., Levine, L. and Zwilling, E., Science, 136, 962 (1962) Cahn, R. D., Kaplan, N. 0., and Zwilling, E., Science, 136, 392 (1952) Cervinka, 0. and Kavka, F., Chem. Listy, 51, 1517 (1957) Chance, B. and Hess, E., Science, 122, 700 (1959a) Chance, B. and Hess, B.. J. Biol. Chem., 2 4, 2421 (1959b) Chance,(B. and Neilands, J. B., J. Biol. Chem., 122, 383 1952 Chilson. 0. P.. Costello, L. A. and Kaplan, N. 0., Biochem., .4. 271 (1965b) Chilson, 0. P.. Kitto, G. B. and Kaplan, N.-0., Proc. Natl. Acade 8010 U0 Se. jlfl 1006 (19653) Ciaccio, E. I.. J. Biol. Chem., 241, 1581 (1966) 190 Ciaccio, E. 1., Keller, D. L. and Boxer, G. E., Biochim. Biophys. Acta, 32, 191 (1960) Constantinides, S.. Ph.D. Thesis, Michigan State Univer- sity, East Lansing, Michigan (1966) Davis, J. B.. Ann. N. I. Acad. Sci., 121, 404 (1964) Davis, J. R. and Busch, H., Cancer Research, ;§, 718 (1958) Dawson, D. M., Goodfriend, T. L. and Kaplan, N. 0., Science, 1325 929 (1964) Delbruck, A., Schimassek, H.. Bartsch, K. and Bucher, T., Biochem. Z.. 321. 297'(1959) DeLey, J. and Schel, J., Biochim. Biophys. Acta, 35, 154 1959 Dennis, D. and Kaplan. N. 0.. Fed; Proc.. lée 213 (1959) Dennis,(D.6and Kaplan, N. 0., J. Biol. Chem., 225. 810 19 0 Dennis, D., Reichlin, M. and Kaplan, N. 0., Ann. N. Y. Acad. Sci., 112, 868 (1965) Di Sabato,6fi3 and Kaplan, N. 0., J. Biol. Chem., 232, 438 19 Dressler, H., and Dawson, C. R., Biochim. Biophys. Acta, 32. 508 (1960) Elliot, B. A..1Jepson, E. M. and Wilkinson, J. H., Clin. Sci., 22, 305 (1962) Emerson, P. M.,.Wilkinson, J. H., and Withycombe, W. A., Nature, 202, 1337 (1964) Epstein, C. J., Carter, M. M., and Goldberger, R. F., Biochim. Biophys. Acta, 2, 391 (19 4) Estabrook, R. w. and Sacktor, B.. J. Biol. Chem., 222, 1014 (1958) Feigl, F., 8 ct Tests ég Organic Anal sis (Elseveir Pub- lish ng Company,-N. E., 1966) seventh edition Fine, I. H., Kaplan, N. 0. and Kuftinec, D., Biochem., 2, ~ ' 116 (1963) Fiume, L.._Boll. Soc. Ital. Biol. Sper.. 4. 437 (1958) 191 Fiume, L., Nature, ;§1, 792 (1960) Fondy, T. P. and Kaplan, N. 0., Ann. N. Y. Acad. Sci., 112, 888 (1965) Foster, D. W. and Taylor, J. B., J. Biol. Chem., 241, 38 (1965) Frank, W. and Holz, E., Z. Physiol. Chem., 3&2, 1 (1959) Fritz, P. J., Fed. Proc., 22, 241 (1963b) Fritz, P. J., and Jacobson, K. B.. Biochem.,‘fi, 282 (1965) Fritz, P. J., and Jacobson, K. E., Science, 329, 64 (1963a) Garland, P. B., Handle, P. J., and Newsholme, E. A., Nature, 200, 169 (1963) Gibson, D., Davisson, E. 0., Bachkawat, B. H., Ray, B. H., and Vestling, C. 8., J. Biol. Chem., 203, 397 (1953) Gleason, F. H., Nolan, R. A,, Wilson, A. C. and Emerson, H., Science, 152, 1272 (1966) Goldberg, E. B., and Colowick, S. P., J. Biol. Chem., 240, 2786 (1965) "“ Goldberg, E. B., Nitowskz, H. M., and Colowick, S. P.. J. Biol. Chem., 2 0, 2791 (1965) Goldman, R. D., Kafilan, N. 0., and Hall, T. C., Cancer 0 Research, 2__ 389 (1964) Goodfriend T. L. and Kaplan N. o. J. Biol. Chem., 229 130.(1964) ' ' ' Goodfriend, T. L., Sokol, D. M. and Kaplan, N. 0., J. Mol. Biol;. 1.5,, 18 (1966) Gregory, K. F., Ng, C. W., and Pantekoek, J. F. C. A., Biochim. Biophys. Acta, 1 0, 469 (1966) Hakala, M. T., Glaid, A. J. and Schwert, G. W., Fed. Proc., ‘ 12. 213 (1963) Hakala, M. T., Glaid, A. J. and Schwert, G. W., J. Biol. Chem., 221, 191 (1956) Hammett, 3., Protoplasma, z, 535 (1929) Hathaway, G. and Griddle, R. 8., Proc. Nat. Acad. Sci. U. SJ, 55, 680 (1966) 192 Haugaard, N.. Biochim. Biophys. Acta, 3;, 66 (1959) Hess, 3.. Ann. N. Y. Acad. Sci., 23, 292 (1958) Hill, B. H., Ann. N. Y. Acad. Sci., 23, 304 (1958) Hodgman, C. D., Weast, R. C. and Selby, S. M., Eds., Hfigiigfiin§£CSNTméizigl§%%,PigsgcS' (Chemical Rubber Holzer, P., Glogner, P. and Sedlmayr, G., Biochem. Z., 330, 59 (1958) Jacobson, K. B. and Kaplan, N. 0., J. Biol. Chem., 226, 603 (1957) Kalz, 8., Biochim. Biophys. Acta, 1 , 226 (1955) Kaplan, N. 0., Brookhaven Symp. Biol., 22, 131 (1964) Kaplan, N. 0., in The Enz es P. D. Boyer H. Lardy, K. Myrback, Eds. (Académic Press, N: Y.. 1961) vol. 3, pt. B, 105 Kaplan, N. 0. and Cahn, R. D., Proc. Natl. Acad. Sci. U. S., 23, 2123 (1962) ‘ Kaplan, N. 0., Ciotti, M. M., Hanolsky, M. and Bieber, R. E., Science, 131, 392 (1960) Kaplan, N. 0., Ciotti, M. M., Stolzenbach, R. E. and Bachur, N. H., J. Am. Chem. Soc.. 22, 815 (1955) Karnes, C. Y., Ihnen, E. D., and Vestling, C. 8., Fed. Proc.. $2. 29 (1960) Katz, A. M., Dre er, w. J. and Anfinsen, c. B., J. Biol. Chem.. 22 . 2897 (1959) Kegeles, G. and Gutter, F. J., J. Am. Chem. 800.. 22. 3771 (1951) Kennedy, E. P. and Lehninger, A. L., J. Biol. Chem.. 122, 957 (1949) Kimberg, 0. V.. and Yielding, K. L., J. Biol. Chem., 232, 3233 (1962) Koppanyi, T. and Avery, M. A., Clin. Pharm. Therap.. 2, 250 (1966) Krebs, E. G., and Fischer, E. H., Advanc. Enzymol.. 22, 263 (1962) 193 Kubowitz, F. and Ott, P.. Biochem. Z., 14, 94 (1943) Lehninger, A. L., J. Biol. Chem., 222, 345 (1951) LePage, G. A., and Schneider, W. C., J. Biol. Chem., 126, 1021 (1948) Loewus, F. A., Ofner, P., Fisher, H. F., Westheimer, F. H. and Vennesland, B.. J. Biol. Chem., 202, 699 (1953) Loewus, F. A. and Stafford, H. A., J. Biol. Chem., 235, 3317 (1960) Lynen, F., Proceedings International Symp. on Enz. Chem. (Maruzen, Tokyo, 1958) p. 25 Madsen, N. B., Can. J. Biochem. Physiol., 2;, 561 (1963) Mansour, T. E., J. Biol. Chem., 238, 2285 (1963) Markert, C. L., Science, 140, 1329 (1963) Markert, C. L. and Moller, F., Proc. Natl. Acad. Sci. U. 8., 25, 753 (1959) Markert(,_Cé E. and trsprung, H., Develop. Biol., 5, 363 19 2 Martin, W. H., Abdulian, D. H.. Unna, K. R. and Busch, H., J. Pharm. Exptl. Therap., 124, 64 (1958) Mason, H. 8., Adv. In Enzymol.. £2, 79 (1957) Meister, A., J. Biol. Chem., g§fl, 117 (1950) Meister, A., Biochem. Preps., 2, 18 (1952) Millar, D. B. 3.. J. Biol. Chem., 231, 2135 (1962) Morgan H. E. and Parmeggiani A. J. Biol. Chem. 232 '2u35 (i964) ' ' ' ' Morris, H. P.. Sidransky, H., Wagner, B. P. and Dyer, H. M., » Cancer Research, 22, 1252 (1960) Neifakh, E. A., Dokl. Akad. Nauk USSR, 24%, 1405 (1962) Neilands, J. B.. J. Biol. Chem., _22, 373 (1952a) Neilands, J. B., Science, 225, 143 (1952b) Neilands, J. B.. J. Biol. Chem., ggg, 225 (1954) 194 Nisselbaum, J. S. and Bodansky, 0., J. Biol. Chem.. 236, 323 (1961a) Nisselbaum, J. S. and Bodansky, 0., J. Biol. Chem., 236, 401 (1961b) ‘ Ng, C. W. and Gregory, K. F., Biochim. Biophys. Acta, 130, 477 (1966) Noller, C. H.. Chemistr of 0r anic Compounds, (W. B. Company, Saunders P—iladelphia, Pennsylvania, 1957) second edition Novoa, W. B.)and Schwert, G. W., J. Biol. Chem., 236, 2150 (1961 Novoa, W. B., Winer, A. D., Glaid, A. J. and Schwert, G. W., J. Biol. Chem.,2341143 (1959) Nygaard, A. P., Acta Chem. Scand., 22, 408 (1956) Ornstein, L., Ann. N. Y. Acad. Sci., 121, 321 (1964) Ottolenghi, g. and Denstedt, 0.. Can. J. Biochem., 3_, 1075 (1958 Papaconstantinou, J., and Colowick, S. P., J. Biol. Chem., 26, 278 (1961a) Papaconstantinou, J., and Colowick, S. P., J. Biol. Chem., 2 6, 285 (1961b) Parmeggiani, A., and Bowman, R. H3. Biochem, Biophys. Res. Comm., 22, 268 (1963) Passonneau, J. V., and Lowry, 0. H., Biochem. Biophys. Res. Comm., 2, 10 (1962) Passonneau, J.-V., and Lowr , 0. H., Biochem. Biophys. Res. COMe. $2. 372 (1963¥ Pearl, {. A.$ and Dickey, E. E., J. Am. Chem. Soc., 23, 893 1951 Pearl, I. A., and McCoy. P. F., Anal. Chem., 32, 1407 (1960) Pfleiderer, G. and Jeckel, D., Biochem. Z.. 322, 370 (1957) Pfleiderer, G., Jeckel, D. and Wieland, T., Biochem. Z.. 222. 104 (1957) Plagemann, P. G. W., Gregory, K. F. and Wroblewski, F., J. Biol. Chem., 235, 2288 (1960) 195 Peznanska-Linde, H., Wilkinson, J. H. and Withycombe, W. A., Nature, 2 2, 727 (1966) Pye, E. K. and Eddy, A. A., Fed. Proc., 22, 537 (1965) Quastel, J. H. and Wooldridge, W. R., Biochem. J., 22, 689 (1928 Hacker, E., J. Biol. Chem., $23, 313 (1950) Hacker, E., J. Biol. Chem., $95. 347 (1951) Ramaiah, A., Hathaway, J. A., and Atkinson, D. E., J. Biol. Chem., 232, 3619 (1964) Rapkine, L., Anal. Physiol. Physichim. Biol., 2, 382 (1931) Ringler, R. L. and Singer, T. P., Biochim. Biophys. Acta, 2.29 661 (1958) Sacktor, B. and Cochran, D. G., Arch. Biochem. Biophys.. 223 266 (1958) Sacktor, B. and Dick, A., Cancer Research, 22, 1408 (1960) Salthe, S. N., Chilson, 0. P., and Kaplan, N. 0.. Nature. 221. 723 (1965) Schwert, G. W., Millar, D. B. S., and Takenaka, Y., J. Biol. Chem., 232, 2131 (1962) Schwert, G. W. and Winer, A. D., in The Enzymes, P. D. Boyer, H. Lardy, K. Myrback, Eds. (Academic Press, N. Y., 1963). V01. 79 132 Shaw, C. R. and Barto, E., Proc. Nat. Acad. Sci. U. S., 52, 211 (1963) Shriner, R. L., Fuson, R. C., and Curtin, D. Y., T e Syste- matic Identiiication 22 Or anic Compounds, (John ey and sens, N. Y., 19 fifth edition Silverste12£)E. and Boyer, P. D., J. Biol. Chem., 232, 3901 19 Smiley, K. L., Berry, A. J. and Suelter, C. H., J. Biol. Chem., 242, 2502 (1967) Snoswell, A. M., Biochim. Biophys. Acta, 35, 574 (1959) 'Stambaugh. R. and Post, D., Anal. Biochem.. 25, 470 (1966) Stambaugh, R. and Post, D., J. Biol. Chem., 241, 1462 (i966) 196 Stegink, L. D. and Vestling, C. S., J. Biol. Chem., 241, 4923 (1966) Straub, F. B., Biochem. J., 4, 483 (1940) Szent-Gy3rgyi, A., Science, 222, 34 (1965) Szent-Gy3rgyi, A., Biochem. J., 8, 641 (1966) Szent-Gyorgyi, A., Egm "d, L. G., and McLaughlin, J. A., Science, 155 539 (1967) Taft, R. W. Jr., in Steric Effects in Organic Chemistry, M. S. Newman, Ed., (John Wiley and Sons, Inc.. N. Y., 1956) Chapter 13 Takenaka, Yé)and Schwert, c. w.. J. Biol. Chem., 223, 157 (195 ' Theorel%, H.)and Chance, B., Acta Chem. Scand.. 5, 1127 . 1951 Thunberg, T., Scand. Arch. Physiol., 22, 1 (1920) Tietz, A. and Ochoa, 8.. Arch. Biochem. Biophys., 2_, 477 1958 Tomita. K. and. Helling, Se. Fed. PTOCe. all. 351 (1965) Travagli, G., Atti. Accad. Sci. Ferrara, 22, 3 (1950) Umbarger, H.)E., and Brown, R., J. Biol. Chem., 233, 415 1957 Underwood, J. C., and Filipic, V. J., J. Associa. Off. Agric. Chem., 46, 334 (1963) Underwood J. C., and Filipic, V. J., J. Food Sci., 22, %£ (196u> Underwood, J. c., Filipic, v. J., and Willits, c. 0., J. Food Sci., 3_, 1008 (1965) Underwood, J. C., Willits, C. 0., and Lento, H. G., J. Food Sci., 26, 288 (1961) van Eys, J., Stolzenbach, F. E., Sherwood, L. and Kaplan, N. 0., Biochim. Biophys. Acta, 2 , 63 (1958) Velick, S., J. Biol. Chem., 233, 1455 (1958) Vesell, E. S. and Bearn, A. G., Ann. N. Y. Acad. Sci., 25, 286 (1958) 197 Vesell,eE. S. and Bearn, A. G., Proc. Soc. Exptl. Biol. Med., Si. 96 (1957) Vesell, E.6S. and Bearn, A. G., J. Clin. Invest.,‘22, 586 (19 1) Vinuela, E., Salas, M. L., and $018, A., Biochem. Biophys. Res. Comm., 22. 371 (1963) Wachsmuth, E. D., Pfleiderer, G. and Wieland, T., Biochem. z., 340, 80 (1964) Wacker, W. E. C. and Schoenberger, G. A.,Biochem. Biophys. Res. Comm., 22, 291 (1966a) Wacker, W. E. C. and Schoenberger, G. A., Biochem., 5, 1375 (1966b) Warburg, 0., Uber den Stoffwechse; der Tumoren (Springer, Berlin, 1924) Warburg, 0., Science, 223, 309 (1956) Warburg, 0., Science, 222, 270 (1956) Warner, R. 0.. Arch. Biochem. Biophys., 22, 494 (1958) Webb, J. L., Enz e and Metabolic Inhibitors (Academic Press, Ne Ye. 9 878 I Wenner,-C. E., Spirtes, M. A., Weinhouse, 3., Cancer Research, 12. 44 (1952) Wieland, T. and Pfleiderer, G., Biochem. Z., 322, 112 (1957) Wieland, T., Pfleiderer, G., and 0rtanderl, F., Biochem. Z., 22. 103 (1959) ~ Williamson, J. R., Jones, E. A., and Azzone, G. F., Biochem. Biophys. Res. Comm., 22. 371 (1964) Winer, A. D. and Schwert, G. W., J. Biol. Chem., 231, 1065 (1958) Winer, A. D. and Schwert, G. W., J. Biol. Chem., 232, 1155 ' (1959) ' Winer, A. D., Schwert, G. W. and Millar, D. B. 5.. J. B101. Chem.. 224. 1149 (1959) Wroblewski, F. and Gregory, K. F., Ann. N. Y. Acad. Sci., 4, 912 (1961) 198 Wu, R. and Racker, E., J. Biol. Chem.. .224. 1029 (1959) Yates, R. A6) and Pardee, A. B., J. Biol. Chem., 221. 757 195 Zelitch, 1., J. Biol. Chem. .2_2_5+_. 251 (1957) Zondag, H. A., Science, L42, 965 (1963) APPENDIX The research described in the body of this thesis was performed in the laboratory of Dr. W. C. Deal, Whose imajor interests lie in the field of physical biochemistry, specifically the structure of the glycolytic enzymes. 'Therefore, much work in this laboratory utilizes the Spinco analytical ultracentrifuge. both for determining Wm)! HP” 1!" ‘ molecular weights by equilibrium centrifugation techniques and for determining sedimentation coefficients, both for ‘ native molecules and for dissociated subunits. Both molec- ular weight and sedimentation coefficient are generally found to be dependent upon concentration, so that to achieve accuracy, the values must be determined at several protein concentrations and then extrapolated to zero concen- tration. It is immediately apparent that the accuracy of the extrapolation is of vital importance in determining the correct value at zero concentration. It was for this rea- son that the following computer program was developed. The program is written in Fortran 63, an advanced form of-the basic Fortran, and is presented as Figure 35. The first six comment cards describe the required form of the input data. The intercept must have the highest Y value and the Nth, or last, point must have the greatest X and least I value. In other words, the line must have a nega- tive slope. The program was designed this way to handle 199 200 Figure 35. The computer program as it is described in the body of the appendix, written in Fortran 63. The following pages are a continuation of the program and therefore have no legends. 000000000 101 1 201 PROGRAM LE‘S‘I‘CHI TERCEPT MUST HAVE HIGHEST Y + NTR POINT mST HAVE GREATET X. LEAST Y THE RUN NUMBER mST ONLY CONTAIN THE DIGITS 0.9, NO LETTERS AT ALL isT DATA CARD HAS THE TOTAL N IN COLUMNS 1+2. RUN NUMBER IN COLUMNS 3-8 ALSO 9-15 RAS CENTRI. RUN NO.. 1623 DATE OF RUN. 24-31 DATE OF 3600 RUN, 32 HAS 1 OR 2. 1 FOR SED. COEFF., 2 FOR raw. THE OTHER DATA CARDS RAVE X IN 1.4. Y IN 5-10(N0 DECINAL POINTS) Y AXIS BEGINS AT 7000 8mm Y OF FINAL POINT. ENDS AT 12000 ABOVE 1(1) X AXIS BEGINS AT ZERO, ENDS AT X OF FINAL POINT EXCLUSION LIMITS 3.7.11.15 ARE LINES AT 12.3.6.9 O CLOCK REPECTIVFLY DIMENSION X00).Y(30).NUM(10).YC(30).XC(30).ZUH(10).V(30).NNUH(5). 1 YD(30), CRITVAL(3O).IINCRS(30).SUBSCRPT(3O) COMMON X,Y TYPE REAL MOVE TYPE INTEGER SUBSCRPT CALL PAULT(1) CALL PLOT(200.0,3) READ 1.N.(NUN(xx).xx=1.6).NULTRA.NDU.NDC.IC.(X(J).I(J).J=1.N) IF(EOF.6,0) 999.998 FORMAT(IZ,6(I1.A7,A8,A8,Ii/(F4.2,F6.0)) 998 BLIP 3 -005 370 371 IN = O NNN I: X(N) DIV = 9.0/x(N) MITION I: 1.0 CALL PLOT(0,0,0,100..100.) CALL PLOT(0.-11.0.2) CALL PLOT(0.0.0) CALL PLOT(O.POSITION.2) MOVE e POSITION D0 370 J1 = 1.NNN NOVE e MOVE + DIV CALL PLOT(0.NOVE.1) CALL PLOT(BLIP) CALL PLOT(0) CALL PLOT(0.POSITION.2) NNN = (19000.0 + Y(1) - I(N))/1000.0 DIV = 10.0/FLOW(NNN) MOVE = 0.0 BLIP e POSITION - 0.05 no 371 J2 s: 1.NNN NOVE a MOVE + DIV CALL PLOT(mVE.POSITION.1) CALI. PLOT(IDVE.BLIP) CALL PLOT(NOVE.1.0) SLY e (1000.)/((Y(1)) - (Y(N)) + 19000.) SLX -= (900.)/(X(H)) P 3 [Te/SLY QQ = 4./SLX CALL PLOT(0.POSITION.2) CALL PLOT(0,0,0,SLY,SLX) D0 500 J =1.N v(J) a Y(J) .. I(N) + 7000. J IINCEU) e (V(J) - P)/(100./SLY) 202 CALL PLOT((V(J) 4 P).X(J).2.SLY.SLX) CALL PLOT((V(J) 4 P).(X(J) 4 OO).1.SLY.SLx) CALL PLOT<(V(J) - P).(X(J) 4 QQ).1.SLY.SLx) CALL PLOT< [LLL=NUM( KK] 216 A [INITIALIZE plotter at TERCEPT ---——-‘-'-——- PLOT horizontal arrow pointing at the] IIJITIALTZE plotter , TERCEPT valge three inches above t0p of Y axis on the graph [ENCODE TERCEPT, DECODE AS NUM(KK), IC<=1,6| PLOT NUANKK), KK=1,6 to right of the above ‘ arrow as done for the run number plot after LLL computed GO TO statement above I y , GO TO K: abo 217 type of alternative is eXpressed over the arrow leading from this to the subsequent choice operation. For instance, an IF statement has two exiting arrows. and over one may read "less than or equal to" and over the other "more than.” In all cases, the path of calculations and other Operations follows the path of the arrows from one opera- titui to another. Re-enteries into the program are described where they occur in as much detail as Space allows. The operations performed on the data after it has been entered into the computer are described below: 1) A least square analysis is performed. and the best straight line through all of the input data points is calculated. 2) Each point is then tested by a modified chi- square analysis to determine if it is more than 15 per cent from the I value calculated by the computer for the X value of that point on the line calculated in 1 above. 3) If all points are within 15 per cent of the line. the exclusion limit is progressively lowered to 11. 7 and 3 per cent. Once a point is excluded (see h), all remaining points are retested at all exclusion limits again following recalculation of the new best straight line by least squares. b) If at any time a point is further from the calcu- lated line than the exclusion limit, this point is dropped from the array. and a new least squares EN" itu, . r“ ‘v‘q'v'W—‘r- _l 5.2.3192. I— I 218 line is calculated for the remaining points, which are then retested as in 2 and 3 above. A new line is calculated each time a point is excluded. and the final line acceptable has all points remaining good to at least 3 per cent. 5) For each line that is calculated, both the Slope and intercept on the Y axis are calculated and printed out. In addition. each point that is excluded is printed out along with the exCluSion limit at that point in the execution. 6) The computer plots the final (best) straight line calculated, including all the points but indicat- ing by means of lines in the boxes which points have been excluded. The Y axis is divided into units of 1000, and the X axis is divided into units of one by small blips. An arrow is drawn pointing to the Y intercept, and its numerical value is printed adjacent to this arrow. The words DEAL and the run number are printed just below the X axis. Three typical sets of results are presented on the following pages. Each set of data consists of both printed and plotted results. The data presented was provided as follows: 1) L.C. 6 - pyruvate kinase subunit molecular weight extrapolated to zero concentration. by Dr. M. S. Kayne e ~ {g l. 11‘“ _ “the... I.» 219: L. c. RUN 000006 ULTRACENTRIFUGE RUN 698L050 ON 3/12/65 3600 RUN ON 2/20/67 PROGRAM DATED 2/24/67 mLEEULAR WEIGHT VS. CONCENTRATION RAN DATA POINT X Y '7 1 1.00 68574 133 2 1.00 61177 he... 4 1.50 61280 {- 5 1-50 53109 6 1 .75 55097 E“; .s' 7 1.75 57886 3 8 2.00 57982 '9 9 2.25 55919 10 2.50 53552 11 2.75 56335 12 3.00 56825 13 3.25 65030 14 3.50 54227 15 3-75 53702 16 4.00 48120 17 4.25 56312 18 4.50 51715 19 4.75 58736 20 5.00 53336 21 5.00 44007 CALCULATIONS 1 SLOPE = .2287.08 INTERCEPT -...- 62848.04 POINT 13 EXCLUDED WITH LIMIT = 15 PER CENT 2 SLOPE = -2391.60 INTERCEPT = 62665.15 POINT 1 EXCLUDED WITH LIMIT = 11 PER CENT 3 SLOPE = 4904.69 INTERCEPT 2: 60793.19 POINT 19 EXCLUDED WITH LIEIT = 11 PER CENT 4 SLOPE = -2351. 51 INTEICEPT = 61677.04 POINT 21 EXCLUDED WITH LIMIT = 11 PER CENT 5 SLOPE = -1809.39 INTERCEPT = 60549.97 POINT 5 EXCLUDED WITH LIMIT 2: 7 PER CENT 6 SLOPE: -2072.75 INTERCEPT = 61581.97 POINT 16 EXCLUDED WITH LIMIT = 7 PER CENT 220 7 SLOPE = -1767.41 INTERCET = 61097.19 POINT 2 EXCLUDED WITH LIMIT = 3 PER CENT 8 SLOPE = -1 581+. 50 INTEFCEPT = 601445.93 POINT 4 EXCLUDED WITH LIMIT = 3 PER CENT 9 SLOPE = -1306.27 INTERCEPT = ‘ 59380.30 POINT 6 EXCLUDED wITM LIMIT = 3 PER CENT 10 SLOPE = -1’+77.08 INTERCEPT 2: 60066.27 POINT 10 EXCLUDED WITH LIHIT = 3 PER CENT 11 SLOPE = -1589.85 INTERCEPT = 60671.36 POINT 17 EXCLUDED WITH LIMIT = 3 PER CENT 12 SLOPE = -1819.58 INTERCEPT = 61115.05 ALL POINTS NW GOOD TO WITHIN 3 PER CENT 221 DEAL RUN# E 222 L. c. RUN 000007 ULTRACENTRIFUGE RUN 5013mm ON 2/10/65 3600 RUN ON 2/20/67 PROGRAM DATED 2/24/67 MOLECULAR WEIGHT VS. CONCENTRATION RAW DATA POINT X I 1 2. 50 61945 2 3.00 55522 3 3-50 53973 4 4.00 51588 5 it~50 48797 6 5.00 51678 7 6.00 42926 8 7.00 40710 9 7450 37180 10 8.00 36297 11 8 . 50 34862 12 9.00 34653 13 10.00 31237 CALCULATIONS 1 SLOPE = -3872.41 INTERCEPT = 68104.03 POINT 1 EXCLUDED WITH LIMIT :2 3 PER CENT 2 SLOPE = -3646.14 INTERCEPT‘ = 66377447 POINT 6 EXCLUDED WITH LIMIT = 7 PER CENT 3 SLOPE = -3557416 INTERCEPT = 65482.15 POINT 9 EXCLUDED WITH LIMIT = 3 PER CENT 4 SLOPE = -3 524.14 INTERCEPT = 65434.76 POINT 7 EXCLUDED WITH LIMIT = '3 PER CENT POINT 13 EXCLUD- WITH LIMIT = 3 PER CENT 6 SLOPE = -3623.72 INTERCEPT = 66066.06 POINT 12 EXCLUDED WITH LIMIT = 3 PER CENT 7 SLOPE = 4759.24 INTERCEPT = 66639.97 ALL POINTS NW GOOD TO WITHIN 3 PER CENT d, FQLJN# DEAL L. c. RON 000011 ULTRACENTRIPUGE RUN 722RCLO ON 5/ 6/65 3600 RUN ON 2/20/67 224' MR WEIGHT VS. CONCENTRATION PROGRAM DATED 2/24/67 RAN'DATA POINT x Y 1 1.50 40787 2 1.70 41238 3 2.00 40533 4 3.00 39517 5 3.00 35422 6 3.20 34788 7 3.20 35066 8 3.50 33686 9 3.70 32459 10 4.00 31278 11 4.20 31283 12 4.50 29678 13 4. 50 29228 14 4.70 28665 16 5.00 27639 CALCULATIONS 1 SLOPE = .4135.40 INTERCEPT = 48355.47 POINT 4 EXCLUDED WITH LINIT = 7 PER CENT 2 SLOPE = .4023.01 INTERCEPT -.- 47717.50 ALL POINTS NW GOOD TO WITHIN 3 PER CENT 225 DEAL RUN-1* 1! 226 2) L.C. 7 - phosphorylase E’subunit molecular weight extrapolated to zero concentration, by Dr. W. C. Deal. 3) L.C. 11 - pyruvate kinase subunit (different condi- tions) molecular weight extrapolated to zero con- centration, by Dr. M. S. Kayne. To determine the value of the computer program, it is of interest to Judge the difference in the results produced by drawing the best estimated line and having the computer draw the best calculated line. It can be seen that for L.C. 11, the results would be nearly equal. However, for L.C. 7, the best estimated line extrapolates to approximately 713000, whereas the value determined to be correct to within three per cent was 66,640, and this difference amounts to an error of approximately 7 per cent. For L.C. 6, the data scatters so badly that extrapolation by eye could give values ranging from 57,000 to 68,000, but the computer calculated the best straight line through ten of the points to extrapolate to 61,115. For both of these last two sets of data, the com- puter program proved to be an invaluable aid in determining the best extrapolated molecular weight, even from badly scattered data. It should be noted in closing that the computer is JJDuted to straight line extrapolations, which may not always be the best data-fitting curve. The human element must determine this, but even in these cases, the accurate plot produced by the computer provides a representation of the data to simplify visual estimation of the best curve.