UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Mechanistic studies of valproic acid hepatotoxicity : identification and characterization of thiol conjugates Kassahun, Kelem 1991

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-UBC_1991_A1 K37.pdf [ 9.14MB ]
Metadata
JSON: 831-1.0100542.json
JSON-LD: 831-1.0100542-ld.json
RDF/XML (Pretty): 831-1.0100542-rdf.xml
RDF/JSON: 831-1.0100542-rdf.json
Turtle: 831-1.0100542-turtle.txt
N-Triples: 831-1.0100542-rdf-ntriples.txt
Original Record: 831-1.0100542-source.json
Full Text
831-1.0100542-fulltext.txt
Citation
831-1.0100542.ris

Full Text

MECHANISTIC STUDIES OF VALPROIC ACID HEPATOTOXICITY: IDENTIFICATION AND CHARACTERIZATION OF THIOL CONJUGATES by KELEM KASSAHUN B.Sc.(Pharm.), Addis Ababa Univers i ty , 1981 M.Sc , Univers i ty of B r i t i s h Columbia, 1987 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES Faculty of Pharmaceutical Sciences Div is ion of Pharmaceutical Chemistry We accept th i s thesis as conforming to the required standard Univers i ty of B r i t i s h Columbia Apr i l 1991 ©Kelem Kassahun, 1991 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department The University of British Columbia Vancouver, Canada DE-6 (2/88) 11 ABSTRACT The severe hepatotoxicity of the antiepileptic drug valproic acid (VPA) is believed to be mediated through chemically reactive metabolites. The monounsaturated metabolite 4-ene VPA is steatogenic in the rat, and in a similar fashion to the hepatotoxin 4-pentenoic acid, is thought to be oxidized by mitochondria to a highly reactive a,/J-unsaturated ketone, 3-keto-4-ene VPA. The tripeptide thiol, glutathione (GSH), is known to react with a variety of electrophilic compounds that have the potential to interact with cellular macromolecules. The identification and structural characterization of GSH conjugates provides a means of identifying short-lived unstable electrophiles and thus an insight into the mechanisms of toxicity. This thesis describes the synthesis and characterization of thiol conjugates of reactive metabolites derived from the in vivo metabolism of VPA, 4-ene VPA, (E)-2,4-diene VPA, and 4-pentenoic acid. A negative ion chemical ionization gas chromatographic/mass spectrometric (NICI/GC/MS) method for the determination of VPA and 14 of its metabolites in a single chromatographic run was developed. A combination of pentafluorobenzyl and trimethylsilyl derivatization resulted in the [M-181]" anion as the base peak for all the metabolites measured. When these ions were monitored sensitivities in the low picogram range were achieved. The VPA metabolite profile was determined in pediatric patients on VPA monotherapy and on combined therapy with either carbamazepine or clobazam. 4-Ene- and (E)-2,4-diene-VPA were found to be minor metabolites with serum levels below 1% that of VPA. In patients on combined therapy with carbamazepine, the w and u-l pathways of VPA metabolism were induced, while products of ^-oxidation were significantly decreased. Polytherapy had no significant effect on the serum levels of 4-ene- or (E)-2,4-diene-VPA. Rats were dosed intraperitoneally with 100 mg/kg of the sodium salts of VPA, 4-ene-, (E)-2,4-diene-VPA, 4-pentenoic or (E)-2,4-pentadienoic acids. Methylated bile extracts were analyzed by high pressure liquid chromatography and liquid chromatography/tandem mass spectrometry (LC/MS/MS) for GSH conjugates while urine samples were analyzed by GC/MS and LC/MS for N-acetylcysteine (NAC) conjugates and other metabolites. The GSH conjugate of (E)-2,4-diene VPA was detected in the bile of rats treated with 4-ene- and (E)-2,4-diene-VPA. The NAC conjugate was a major urinary metabolite of rats given (E)-2,4-diene VPA and was a prominent urinary metabolite of those animals given 4-ene VPA. The structures of these metabolites were confirmed by comparing GC/MS or LC/MS properties of the isolated metabolites to those of synthetic standards. The GSH and NAC conjugates of (E)-2,4-diene VPA were chemically synthesized and their structures established to be (E)-5-(glutathion-S-yl)-3-ene VPA and (E)-5-(N-acetylcystein-S-yl)-3-ene VPA by nuclear magnetic resonance spectroscopy and mass spectrometry. In contrast to the very slow reaction of the free acid of (E)-2,4-diene VPA with GSH, the methyl ester reacted rapidly with GSH to yield the adduct. In vivo it appears the diene forms an intermediate with enhanced electrophilic reactivity to GSH as indicated by the facile reaction of the diene with GSH in vivo (about 40% of the (E)-2,4-diene VPA administered to rats was excreted as the NAC conjugate in 24 hr). In rats treated with 4-pentenoic and/or (E)-2,4-pentadienoic acids the following conjugates were identified and characterized by synthesis: GSH and cysteine conjugates of 3-oxo-4-pentenoic acid, GSH and NAC conjugates of (E)-2,4-pentadienoic acid, and the NAC conjugate of acrylic acid. The results thus provided the first direct biochemical evidence for the in vivo formation of the metabolite of 4-pentenoic acid considered responsible for the irreversible inhibition of fatty acid metabolism. The results also revealed basic differences between the mitochondrial metabolism of 4-ene VPA and 4-pentenoic acid. The 3-keto-4-ene VPA and its GSH and NAC conjugates were synthesized in order to facilitate the in vivo identification of these compounds following the administration of VPA, 4-ene-, or (E)-2,4-diene-VPA to rats. However, neither the 3-keto-4-ene VPA nor its thiol derivatives were evident in any of the treatments. The NAC conjugate of (E)-2,4-diene VPA was also found to be a metabolite of VPA in patients. The level of the conjugate appeared to be higher in two patients who recovered from VPA-induced liver toxicity. The characterization of GSH and NAC (in humans and rats) conjugates of (E)-2,4-diene VPA suggests that VPA is metabolized to a chemically reactive intermediate that may contribute to the hepatotoxicity of the drug. TABLE OF CONTENTS Page ABSTRACT i i TABLE OF CONTENTS v LIST OF FIGURES vii LIST OF SCHEMES xi LIST OF TABLES xii LIST OF ABBREVIATIONS x i i i ACKNOWLEDGEMENTS xvi i I . INTRODUCTION 1 (1) Fatty acid metabolism 3 (2) Metabolism of VPA and 4-ene VPA 6 (3) Inhibition of fatty acid metabolism by analogs of 4-ene VPA 13 (4) Metabolic disturbances caused by VPA 16 (5) Teratogenicity of VPA 19 (6) The hepatotoxicity of VPA 20 (7) Glutathione conjugation and its toxicological significance 29 (8) Rationale and objectives 32 I I . EXPERIMENTAL (1) Chemicals and materials 35 (2) Animals and surgical equipment 37 (3) Instrumentation 38 (4) Chemical syntheses 40 vi (5) Animal experiments 57 (6) Human study 60 (7) Quantitation of metabolites 60 I I I . RESULTS (1) Chemical syntheses 65 (2) Development of an analytical method for profiling VPA metabolism in patients 98 (3) Metabolism studies involving VPA, 4-ene-and (E)-2,4-diene-VPA 111 (4) Identification and characterization of the thiol conjugates of reactive metabolites of 4-pentenoic acid 127 (5) Attempted synthesis of the CoA thio ester of (E)-2,4-diene VPA 140 IV. DISCUSSION (1) Chemistry 142 (2) Development of analytical methods 151 (3) Metabolism 156 (4) Significance of the findings of the metabolism studies to the mechanism of the hepatotoxicity of VPA 168 V . SUMMARY AND CONCLUSIONS 176 V I . REFERENCES 182 V I I . APPENDICES 198 V I 1 LIST OF FIGURES Figure Page IA The beta-oxidation pathway of VPA metabolism in humans 8 IB Microsomal oxidation, desaturation and glucuronidation pathways of VPA metabolism in humans 9 2 Structurally related liver toxin carboxylic acids 25 3 Mitochondrial metabolism of 4-pentenoic acid leading to the generation of chemically reactive intermediates 26 4 Metabolic pathway of VPA proposed to lead to the production of reactive and hepatotoxic metabolites 27 5 EI/GC/MS TIC plot and spectrum of synthesized methyl 3-keto-4-ene VPA 67 6 *H-NMR spectrum of synthetic GSH conjugate of methyl 3-keto-4-ene VPA 69 7 ^C-NMR spectrum of synthetic GSH conjugate of methyl 3-keto-4-ene VPA 70 8 Reconstructed FAB mass spectrum of the GSH conjugate of methyl 3-keto-4-ene VPA in 3-nitrobenzyl alcohol matrix 71 9 EI/TIC chromatogram of an HPLC purified synthetic sample of the dimethyl ester of the NAC conjugate of 3-keto-4-ene VPA showing the decomposition of the compound upon GC/MS analysis 72 10 PCI and EI/GC/MS spectra of the dimethyl ester of the NAC conjugate of 3-keto-4-ene VPA 73 11 Reconstructed FAB mass spectrum of the GSH conjugate of methyl (E)-2,4-diene VPA in 3-nitrobenzyl alcohol matrix 76 12 *H-NMR spectrum of the GSH conjugate of methyl (E)-2,4-diene VPA 78 13 ldC-NMR spectrum of the GSH conjugate of methyl (E)-2,4-diene VPA 79 14 EI mass spectra of synthetic and isolated NAC conjugate of (E)-2,4-diene VPA (dimethyl ester) 81 15 EI/GC/MS TIC chromatogram and spectrum of a synthetic sample of methyl 3-oxo-4-pentenoate 84 vi i i 16 LC/MS spectrum, in the positive ion mode, of synthesized GSH conjugate of methyl 3-oxo-4-pentenoate 85 17 *H-NMR spectrum of the GSH conjugate of methyl 3-oxo-4-pentenoate 86 18 ^C-NMR spectrum of the GSH conjugate of methyl 3-oxo-4-pentenoate 87 19 EI and PCI mass spectra of the NAC conjugate of 3-oxo-4-pentenoate, dimethyl ester 89 20 LC/MS/MS of the (M+H)+ of the dimethyl ester of the NAC conjugate of 3-oxo-4-pentenoate 90 21 LC/MS mass spectrum of the GSH conjugate of methyl (E)-2,4-pentadienoate 92 22 *H-NMR spectrum of the GSH conjugate of methyl (E)-2,4-pentadienoate 93 23 ^C-NMR spectrum of the GSH conjugate of methyl (E)-2,4-pentadienoate 94 24 LC/MS profile of a fully methylated sample of the GSH conjugate of (E)-2,4-pentadienoate 95 25 Daughter ion (parent, m/z 448) MS/MS spectra of peaks 1 (top) and 2 (bottom) in Fig. 24 corresponding to isomeric GSH conjugates of (E)-2,4-pentadienoate 96 26 Mass chromatogram (m/z 176) and spectrum of a crude synthetic sample of the NAC conjugate of (E)-2,4-pentadienoate (dimethyl ester) 97 27 NICI/GC/MS spectrum of the 0-TMS-PFB derivative of 3-keto VPA 101 28A Typical SIM chromatograms obtained from the serum extract of a patient to which internal standards have been added 102 28B Typical SIM chromatograms obtained from the serum extract of a patient to which internal standards have been added 103 29 A representative TIC chromatogram (EI mode) from the methylated urine extract of a rat dosed with (E)-2,4-diene VPA 112 30 A) EI/TIC chromatogram from the methylated urine extract of a rat given 4-ene VPA. B) HPLC profile of the methylated bile extract of rats dosed with 4-ene VPA 113 ix 31 HPLC profiles of the methylated bile extract of rats treated with (E)-2,4-diene VPA, a methylated synthetic sample of the GSH conjugate of (E)-2,4-diene VPA, and methylated control bile extract 115 32 LC/MS chromatogram (negative ion mode) obtained from the methylated bile extract of rats dosed with (E)-2,4-diene VPA 116 33 Daughter ion spectra obtained by argon CAD of the ( M + O O C C H 3 ) " anion of the negative ion LC/MS spectrum of methylated bile extract from rats dosed with (E)-2,4-diene VPA and authentic methylated GSH conjugate of (E)-2,4-diene VPA 117 34 Mass chromatogram of m/z 141 obtained from the PFB derivatized urine extract of rats treated with (E)-2,4-diene VPA showing the monounsaturated metabolites of (E)-2,4-diene VPA 118 35 Extracted ion (m/z 229) and mass spectrum from the PFB and TMS derivatized urine extract of rats given (E)-2,4-diene VPA showing the presence of 3-0H-4-ene VPA 120 36 Mass chromatogram (m/z 196, ([M-181]")) and spectrum of the PFB derivatized urine extract of rats treated with (E)-2,4-diene VPA, apparently indicating the presence of the glycine amide conjugate of (E)-2,4-diene VPA 121 37 EI/TIC chromatogram of the methylated urine extract of rats dosed with the GSH conjugate of methyl (E)-2,4-diene VPA 123 38 LC/MS extracted ion (m/z 348) and MS/MS spectrum obtained from the methylated urine extract of rats given the GSH conjugate of methyl 3-keto-4-ene VPA 124 39 A typical selected ion current chromatogram for the PFB derivatized NAC conjugate of (E)-2,4-diene VPA in the urine extract of a patient on VPA therapy and from the urine extract of a volunteer administered a 50:50 mixture of VPA and [ZHe]VPA 125 40 Selected ion current chromatograms ([M-181]") of the PFB derivatives of the analyte and internal standard extracted from control urine 128 41 A typical calibration curve for the quantitation of the NAC conjugate of (E)-2,4rdiene VPA obtained from the PFB derivative of the conjugate using the NAC conjugate of acrylic acid as the internal standard 129 42 EI/TIC chromatogram of the methylated urine extract of X rats dosed with (E)-2,4-pentadienoic acid showing the cysteine conjugate of 3-oxo-4-pentenoate and isomeric NAC conjugates of (E)-2,4-pentadienoate 130 43 LC/MS trace (extracted ion): A) from the methylated urine extract of rats given (E)-2,4-pentadienoic acid, B) synthetic cysteine conjugate of methyl 3-oxo-4-pentenoate 131 44 LC/MS spectra of the peaks in Fig. 43 132 45 EI/GC/MS spectra of synthetic and isolated cysteine conjugate of methyl 3-oxo-4-pentenoate 133 46 EI/GC/MS spectra of authentic and native NAC conjugate of (E)-2,4-pentadienoate (dimethyl ester) 135 47 Positive ion LC/MS chromatogram (extracted ion, m/z 478): A) methylated bile extract of rats dosed with (E)-2,4-pentadienoic acid, B) methylated authentic GSH conjugate of 3-oxo-4-pentenoate 136 48 Daughter ion spectra (parent, m/z 478) of A) Scans 102-106 (Fig. 47A) and B) Scans 102-106 (Fig. 47B) representing tetra-methylated GSH conjugate of 3-oxo-4-pentenoate 138 49 LC/MS extracted ion chromatogram (m/z 448): A) methylated bile extract of rats treated with (E)-2,4-pentadienoic acid, B) methylated synthetic GSH conjugate of (E)-2,4-pentadienoate 139 50 Daughter ion spectra (parent, m/z 448): A) Scans 111-115 (Fig. 49A), B) Scans 115-121 (Fig. 49B) 141 51 Scheme showing the possible regioisomeric GSH conjugates of methyl (E)-2,4-diene VPA as a result of 1,6- and 1,4-additions 148 52 Structures of the possible isomeric NAC conjugates of methyl (E)-2,4-pentadienoate 150 53 Proposed mitochondrial beta-oxidation of 4-ene VPA on the propenyl and propyl chains 164 LIST OF SCHEMES Scheme 1 Synthetic route for 3-keto-4-ene VPA methyl ester and its GSH and NAC conjugates 2 Outline of the synthetic route of the thiol conjugates of (E)-2,4-diene VPA and its methyl ester 3 Synthetic pathway of methyl 3-oxo-4-pentenoate and its thiol conjugates 4 Synthesis of methyl (E)-2,4-pentadienoate and its GSH and NAC adducts 5 Proposed mechanism for the biosynthesis of the GSH conjugate of (E)-2,4-diene VPA from 4-ene VPA XI 1 LIST OF TABLES Table Page 1 *H-NMR data of (E)-2,4-diene VPA methyl ester and its NAC and GSH conjugates 77 2 Ions monitored in the NICI mode for VPA and metabolites derivatized with PFB and TMS 100 3 Day to day precision and accuracy data for VPA and metabolites in serum 105-6 4 Serum concentrations (ng/ml) of VPA metabolites in 26 pediatric patients on VPA monotherapy 108 5 Mean concentrations normalized to dose (/xg/mL/mg/kg) of VPA and metabolites in the serum of patients 109 6 Urinary concentrations (ng/mg creatinine) of the conjugated and unconjugated fractions of VPA metabolites in patients on monotherapy 110 xi i i LIST OF ABBREVIATIONS API atmospheric pressure ionization ATP adenosine triphosphate bp boiling point bs broad singlet C-18 octadecylsilane CAD collision activated dissociation CBZ carbamazepine CI chemical ionization CoA coenzyme A thio ester CoASH coenzyme A CLBZ clobazam D deuterium d doublet (NMR), deuterium DBU l,8-diazabicyclo[5.4.0.]undec-7-ene S chemical shift dd doublet of doublets E trans EI electron impact FAB fast atom bombardment FAD flavine adenine dinucleotide GC/MS gas chromatography/mass spectrometry GSH glutathione GSSG glutathione disulfide hr hour(s) HMPA hexamethylphosphoramide HPLC high performance liquid chromatography Hz Hertz i .d . internal diameter 0 i .p. intraperitoneal i .v. intravenous J coupling constant kg kilogram LC liquid chromatography LC/MS liquid chromatography/mass spectrometry LDA lithium diisopropylamide m multiplet, meter M+ molecular ion mg milligram 2-MGA 2-methylglutaric acid min minutes mL mil l i l i ter mmoles millimoles MS mass spectrometry MS/MS tandem mass spectrometry MSTFA N-methyl-N-trimethylsilyl-trifluoroacetamide MW molecular weight m/z mass to charge ratio NAC N-acetyl-L-cysteine NAD nicotinic acid adenine dinucleotide NICI negative ion chemical ionization NMR nuclear magnetic resonance spectroscopy PCI positive chemical ionization PFBB pentafluorobenzyl bromide PFB pentafl uorobenzyl ppm parts per million psi pounds per square inch q quartet s singlet SIM selected ion monitoring t triplet t-BDMS tertiarybutyldimethylsilyl THF tetrahydrofuran TIC total ion current TLC thin layer chromatography TMS trimethylsilyl VPA valproic acid, valproate [2H]6VPA [^Hlsvalproic acid 2,3'-diene VPA 2-(l'-propenyl)-2-pentenoic acid 2,4-diene VPA 2-propyl-2,4-pentadienoic acid 2,4'-diene VPA 2-(2'-propenyl)-2-pentenoic acid 2-ene VPA 2-propyl-2-pentenoic acid [2H]32-ene VPA [2H]32-propyl-2-pentenoic acid 3-ene VPA 2-propyl-3-pentenoic acid 4-ene VPA 2-propyl-4-pentenoic acid 3-keto VPA 2-propyl-3-oxopentanoic acid [2H]33-keto VPA [^H]32-propyl-3-oxopentanoic acid 4-keto VPA 2-propyl-4-oxopentanoic acid 3-OH VPA 2-propyl-3-hydroxypentanoic acid 4-OH VPA 2-propyl-4-hydroxypentanoic acid 5-OH VPA 3-0H-4-ene VPA 3'-0H-4-ene VPA 3-keto-4-ene VPA 3'-keto-4-ene VPA 2-PGA 2-PSA Z 2-propyl-5-hydroxypentanoic acid 2-propyl-3-hydroxy-4-pentenoic acid 2-(1'-hydroxypropyl)-4-pentenoic ac 2-propyl-3-oxo-4-pentenoic acid 2-(l'-oxopropyl)-4-pentenoic acid 2-propylglutaric acid 2-propylsuccinic acid cis ACKNOWLEGDEMENTS I am grateful to my supervisor, Dr. Frank Abbott, for his guidance, support, and friendship throughout the course of the study. I am also thankful to members of my supervisory committee, Dr. James Axel son, Dr. Guenter Eigendorf, Dr. Keith McErlane, and Dr. James Orr for their time and guidance. Special thanks go to Mr. Roland Burton for his technical assistance and valuable suggestions. I would also like to acknowledge the various assistances by my colleagues, Dr. Ronald Lee, Dr. Abdul Mutlib, Anthony Borel, and Sue Panesar. 1 I . INTRODUCTION The antiepileptic properties of valproic acid (VPA) were discovered serendipitously (Meunier et al., 1963) nearly a century after the compound was first synthesized and used as an organic solvent. This branched-chain fatty acid is unique among antiepileptic drugs both in its structure and mechanism of action. VPA is very effective in a variety of seizure types and thus is regarded as an important new type of antiepileptic drug, and today enjoys widespread use throughout the world. VPA is a valuable drug for the treatment of primary generalized seizures (Gram and Bentsen, 1985; Wilder et a/ . , 1983) and is the drug of choice for children with absence seizures (Jeavons et al., 1977). Emerging indications for VPA are psychiatric ailments such as bipolar disorders (Penry and Dean, 1989). Initially VPA was considered a safe drug with only minor side effects, but with more widespread use it has been shown to cause fatal hepatotoxicity in man (Zimmerman and Ishak, 1982; Zafrani and Berthelot, 1982). The fatal hepatotoxicity is considered to be an idiosyncratic reaction primarily occurring in children under two years of age who were taking VPA along with other anticonvulsant drugs (Dreifuss et al., 1987). The mechanism of VPA hepatotoxicity is not understood, although the possible involvement of a reactive metabolite has been suggested (Zimmerman and Ishak, 1982). Because the monounsaturated metabolite 4-ene VPA is a structural analog of the fatty acid oxidation inhibitor 4-pentenoic acid, and is steatogenic in the rat, it has been proposed that 4-ene VPA contributes to the hepatotoxicity of VPA through metabolic activation to a reactive intermediate. The biotransformation of xenobiotics can lead to reactive intermediates and metabolites that are capable of interactions with cellular components. The formation of chemically reactive metabolites from drugs is well documented, and it is generally accepted that reactive metabolites mediate the toxic reactions to many drugs and other xenobiotics (Nelson, 1982). The tripeptide thiol, glutathione (GSH), is known to react with a variety of electrophilic compounds thus preventing to a certain extent their reaction with essential components of the cell (Chasseaud, 1976). The main objective of the present study was to determine whether chemically reactive metabolite(s) of VPA, which might contribute to the hepatotoxic response to the drug, are formed and can be trapped as their thiol conjugates in vivo. As a background to the rationale and objectives of the thesis, the following topics will be briefly described. The metabolism of VPA and/or 4-ene VPA appears to play a crucial role in the hepatotoxicity of VPA, and because VPA shares common metabolic pathways with fatty acids, fatty acid metabolism will be briefly discussed. This will be followed by a discussion of the metabolism of VPA and 4-ene VPA. Hypoglycin and 4-pentenoic acid are structurally related to 4-ene VPA and are known to be liver toxins, and hence the mechanism by which these analogs of 4-ene VPA cause hepatotoxicity will be presented. An outline of the metabolic disturbances caused by VPA and the teratogenicity of VPA will be presented because some of the consequences of the metabolic disturbances caused by VPA might be related to the hepatotoxicity of VPA. Then a literature survey of the incidence and pathological manifestations of VPA hepatotoxicity and a discussion of the proposed biochemical mechanisms will be presented. Finally, there will be a short discussion of glutathione conjugation and its toxicological significance as this would be relevant to the reactive metabolite hypothesis of VPA hepatototoxicity. 1. Fatty acid metabolism A. Mitochondrial fl-oxidation The sequence of reactions by which fatty acids are degraded is called ^-oxidation. A single oxidation cycle results in the shortening of the chain by two carbons through the splitting off of an acetyl CoA unit (Devlin, 1986; Stryer, 1988). The initial step in the oxidation of fatty acids is their conversion to CoA thio ester derivatives. This reaction is catalyzed by acyl CoA synthetase and occurs in the endoplasmic reticulum or outer mitochondrial membrane. Fatty acids that are present in the mitochondria can also be activated; however, the physiological significance of this mitochondrial process is unknown. Fatty acyl CoAs are formed outside of the mitochondria but are oxidized in the mitochondrial matrix. Since long-chain acyl CoAs do not readily cross the mitochondrial membrane, a transport mechanism is required. This is accomplished by forming acyl carnitine, between carnitine and the acyl CoA, which is shuttled across the inner mitochondrial membrane. Medium-chain (CS-CJO) acyl CoAs can permeate into the mitochondrial matrix, and hence their oxidation is not dependent on carnitine. Also, short-chain fatty acids can cross the inner mitochondrial membrane directly i .e. without forming CoA thio esters. In the first step of j3-oxidation, a double bond of the trans-configuration is formed between carbons 2 and 3 of the acyl CoA, and in the process the oxidizing agent, FAD, is reduced to FADH2- This reaction is catalyzed by several acyl dehydrogenases the specificity of which is dependent on chain length. The second step is hydration of the double bond to a hydroxyacyl CoA by the enzyme enoyl CoA hydratase. The reaction is stereospecific resulting in only the L-isomer. This step precedes the second oxidation reaction, which converts the hydroxy group into a keto group. The reaction is catalyzed by L-3-hydroxyacyl CoA dehydrogenase and the cofactor is NAD+. The final step of ^-oxidation is cleavage of the carbon-carbon bond of the acyl CoA by a molecule of CoASH to yield acetyl CoA and an acyl CoA shortened by two carbon atoms. This thiolytic cleavage is catalyzed by /J-ketoacyl CoA thiolase, an enzyme with a broad specificity with regard to chain length. Beta-oxidation of unsaturated fatty acids requires two additional enzymes; an isomerase which converts a cis- 3-ene double bond to a trans- 2-ene double bond, and an epimerase which catalyzes conversion of the D-3-hydroxyacyl CoA configuration to L-3-hydroxyacyl CoA. The enzyme, 2,4-dienoyl CoA reductase, which is present in the mitochondria as well as in peroxisomes, catalyzes the reduction of polyunsaturated fatty acids possessing a 2,4-dienoyl CoA structure (Osmundsen and Hovik, 1988). This results in the 3-ene analog which is 0-oxidized following isomerization to 2-enoyl CoA. The level of 2,4-dienoyl CoA reductase is known to be induced by clofibrate treatment of rats (Borrebaek et a7., 1980). B. Peroxisomal fl-oxidation It is now well known that in the cell the ^-oxidation system is also present in peroxisomes (Lazarow, 1987). However, the physiological significance of the peroxisomal ^-oxidation is considered to be relatively low with most of the oxidation of fatty acids occurring in the mitochondria (Devlin, 1986; Yamada et al., 1987). Peroxisomal /3-oxidation is involved in the metabolism of dicarboxylic acids, bile acid, and very long-chain fatty acids which are poorly oxidized by mitochondria (Yamada et al., 1987). Like the mitochondria, the peroxisomes also participate in the metabolism of xenobiotics (Yamada et al., 1984; Yamada et al., 1987; Suzuki et al., 1990). The mammalian peroxisomal fatty acid system differs from the mitochondrial )3-oxidation in three respects. First, carnitine is not required for the entry of fatty acyl CoAs into peroxisomes. Second, the init ial dehydrogenation step is accomplished by a cyanide-insensitive oxidase system and leads to the formation of H 2 O 2 , which is eventually eliminated by catalase. Third, the specificity of peroxisomal enzymes is for longer chain fatty acids. The observation that ^-oxidation in peroxisomes does not proceed below C s (Devlin, 1986) has led to the hypothesis that the function of peroxisomes is to shorten long-chain fatty acids for eventual mitochondrial /3-oxidation. C. Omega and Omeqa-1 oxidations Hydroxylation of the methyl carbon at the far end of the molecule from the carboxyl group (w-oxidation) or on the carbon next to the methyl end (w-l-oxidation) are minor pathways for fatty acid metabolism (Devlin, 1986; van Hoof et al., 1988). The reactions are catalyzed by the mixed-function oxidase system and require cytochrome P-450, molecular O2, and NADPH. By this metabolic pathway, monocarboxylic acids are converted to hydroxyacids in the endoplasmic reticulum. These hydroxyacids are subsequently converted to dicarboxylic acids by the actions of cytosolic alcohol- and aldehyde-dehydrogenases. The dicarboxylic acids are then excreted, or are subject to /f-oxidation from either end, depending on the tissue and its metabolic state. Organic aciduria (excretion of medium-chain dicarboxylic acids) is increased in conditions of excessive fatty acid degradation (eg. ketosis and diabetes mellitus) or defective mitochondrial fatty acid metabolism (eg. inherited metabolic diseases, hypoglycin poisoning, riboflavin deficiency)(van Hoof et al., 1988). 2. Metabolism of VPA and 4-ene VPA A. VPA Metabolism is the major route of elimination of VPA in both animals and humans with renal excretion of the unchanged drug accounting for less than 5% of the administered dose (Gugler et al., 1977). The metabolic pathways of VPA are similar to those of endogeneous fatty acids. In spite of this fact and its simple structure, the metabolism of VPA has proved to be very complex, and complete elucidation of its metabolic pathways has proved elusive. The metabolism of VPA has been extensively studied in man and various animals and has been reviewed by Gugler and von Unruh (1980) and more recently by Eadie et a7.(1988). These studies indicate four main metabolic pathways: glucuronidation, /?-oxidation, w-oxidation and (w-l)-oxidation. The human metabolic pathways of VPA are summarized in Figs. IA and IB. In both man and the rat, ^-oxidation and glucuronidation are the two primary pathways. Although VPA metabolism is qualitatively similar in humans and the rat, the level of /3-oxidation metabolites in rats given high doses of VPA is lower than in humans (Granneman et al., 1984a). Glucuronidation is quantitatively the most important route of VPA transformation, with 15-20% of the administered dose in man being excreted as VPA glucuronide (Bialer et al., 1985). Metabolism of VPA is dose-dependent; with increasing doses of VPA, glucuronidation is increased at the expense of /3-oxidation (Granneman et al., 1984b). In the rhesus monkey the principal pathways of VPA metabolism in decreasing quantitative importance were glucuronide formation, w-oxidation, /3-oxidation and (w-l)-oxidation (Rettenmeier et al., 1986a). Beta-oxidation of VPA produces trans- and cis-2-ene-, 3-0H- and 3-keto-VPA with 3-keto- and 2-ene-VPA being major metabolites in serum (Abbott et al., 1986). 3-Keto VPA is also a major urinary metabolite and is considered to arise through mitochondrial /3-oxidation. Thus, /3-oxidation of VPA is incomplete, and this has been attributed to its branched-chain structure (Rettenmeier et al., 1985). Although for many xenobiotics /3-oxidation results in a chain shortened by two carbons (Suzuki et al., 1990; Vyas et al., 1990; Tsaconas and Padieu, 1989; Grindel et a / . , 1979), for VPA such a chain-shortening has yet to be demonstrated. The involvement of 2-methyl-branched chain acyl CoA dehydrogenase, but not other acyl dehydrogenases, in the dehydrogenation of VPA to 2-ene VPA has recently been demonstrated (Ito et al., 1990). COOH (E)-3-ene VPA (E.E)-2,3'-diene VPA Figure IA. The beta-oxidation pathway of VPA metabolism in humans broken lines indicate likely metabolic routes not yet confirmed. COOH (E ) -2 ,4 -d iene VPA Figure IB. Microsomal oxidation, desaturation and glucuronidation pathways of VPA metabolism in humans. The w-oxidation pathway leads to the formation of 5-OH VPA which is further oxidized in the cytosol to give 2-PGA. The (w-1)-pathway results in the formation of 4-0H-, 4-keto-VPA and 2-PSA. Products of w-and (w-l)-oxidation of VPA are found in minor quantities in the serum of epileptic patients (Abbott et a 7 . , 1986). The w-and (w-1)-pathways are cytochrome P-450 mediated; Prickett and Baillie (1984) showed that incubation of VPA with rat liver microsomes led to the formation of 3-, 4- and 5-OH-VPA. This indicated that the metabolite 3-OH VPA can be formed by both cytochrome P-450 dependent direct hydroxylation and /J-oxidation, and thus may not be a good indicator of the extent of /*-oxidation of VPA. Indeed, it has been shown, using stable isotope techniques, that 3-OH VPA has this dual origin (Rettenmeier et a 7 . , 1987). Gamma- and delta-dehydrogenation of VPA gives rise to 3-ene- and 4-ene-VPA, respectively. The cellular site of the formation of 3-ene VPA has not been established although there is evidence from studies using stable isotopes that 3-ene VPA is formed reversibly from 2-ene VPA (Rettenmeier et a 7 . , 1987). It can also be speculated that 3-ene VPA and 2,4-diene VPA (see below) might have a product-precursor relationship since it is known that fatty acids having the 2,4-dienoyl structure are reduced by 2,4-dienoyl CoA reductase to the analog with the double bond at the 3-position (Osmundsen and Hovik, 1988). Rettie et a7.(1987) have demonstrated that cytochrome P-450 catalyzes the formation of 4-ene VPA from VPA. The proposed mechanism involves formation of a carbon centered free radical intermediate at C 4 , via hydrogen abstraction, which partitions between recombination (formation of 4-OH VPA) and elimination (formation of 4-ene VPA) pathways (Rettie et al., 1988). The monounsaturated metabolites, which also are excreted as glucuronide conjugates, are further metabolized to give dienes of VPA. The 2,4-diene VPA is formed from 4-ene VPA and is found as the minor diene metabolite in the serum of patients (Granneman et a7., 1984a; Rettenmeier et al., 1985). The major diunsaturated metabolite in the serum of subjects treated with VPA has been characterized as (E,E)-2,3'-diene VPA by Acheampong and Abbott (1985) and probably arises from 3-ene VPA (Rettenmeier et al., 1987). New VPA metabolites continue to be identified as exemplified by the recent identification of 4'-keto-2-ene VPA as a minor serum metabolite (Kassahun et al., 1989). Interest in VPA metabolism emanates from the possible contribution of metabolites to both the therapeutic and toxic effects of the drug. Thus, the unsaturated metabolites 2-ene VPA (Loscher, 1981a; Loscher and Nau, 1985) and 2,3'-diene VPA (Abbott and Acheampong, 1988) were found to have significant anticonvulsant activity in rodent models. The 4-ene VPA metabolite is thought to be involved in the hepatic damage of VPA (Zimmerman and Ishak, 1982; Rettenmeier et al., 1985). B. 4-ene VPA The 4-ene VPA metabolite has been implicated in the hepatotoxicity of VPA, and consequently its metabolism has been extensively studied (Granneman et al., 1984a; Rettenmeier et al., 1985, 1986b). Eight metabolites were identified by GC/MS in rat bile and perfusate from isolated rat liver. The metabolites were identified as (E)-2,4-diene VPA, 4,5-diOH VPA-7-lactone, 2-propyl-2-enylglutaric acid, 2-PGA, 3'-keto-4-ene-, 5'-0H-4-ene-, 5-OH- and 3-0H-4-ene-VPA. In the rhesus monkey, the fraction of the dose of 4-ene VPA recovered in urine in the form of metabolites and parent compound was much lower than that of VPA implying that 4-ene VPA or one or more of its metabolites tend to be incorporated into triglycerides (examples of xenobiotics that are incorporated into the adipose tissue are known (Fears, 1985)) or bind covalently to cellular components. In the rhesus monkey 19 metabolites were characterized indicating that the metabolism of 4-ene VPA is extremely complex and involves mitochondrial/ peroxisomal, microsomal and cytosolic enzymes. The major metabolites identified were (E)-2,4-diene VPA, 4-ene VPA glucuronide and 3'-keto-4-ene VPA. Beta-oxidation on the propenyl side chain would lead to (E)-2,4-diene VPA and 3-OH-4-ene VPA. The latter metabolite could also arise from cytochrome P-450 dependent hydroxylation (Prickett and Bail l ie , 1986). Thus far no information is available on the identity of the enzyme(s) involved in the formation of (E)-2,4-diene VPA. Mitochondrial/peroxisomal /J-oxidation has been implicated based on such factors as the stereospecificity of the product, i .e. the (E)-isomer. By analogy to the metabolism of VPA, the end product of the oxidation of 4-ene VPA should be the electrophilic a,^-unsaturated ketone, 3-keto-4-ene VPA. However, this compound was not detected in any of the studies of 4-ene VPA metabolism (Granneman et al., 1984a; Rettenmeier et al., 1985, 1986b). On the other hand, /J-oxidation on the saturated side chain of 4-ene VPA gave rise to the expected keto compound (3'-keto-4-ene VPA) and the hydroxy intermediate (3'-0H-4-ene VPA) but not the diene intermediate (2,4'-diene VPA) (Rettenmeier et al., 1986b). Like VPA, conjugation with glucuronic acid was a major route of 4-ene VPA metabolism. The glycine conjugate of (E)-2,4-diene VPA was identified in significant amounts indicating that, unlike VPA, glycine conjugation plays a significant role in the metabolism of 4-ene VPA. The minor pathways of 4-ene VPA metabolism in the rhesus monkey included hydroxylation at positions 3, 3', 4', and 5' to give a series of regioisomeric hydroxy metabolites and their derivatives. There was also evidence for the formation of an epoxide intermediate and/or its precursor free radical species during the oxidation of the double bond of 4-ene VPA by cytochrome P-450 (Prickett and Baill ie, 1986). This reaction was accompanied by alkyation of cytochrome P-450 and formation of 4,5-diOH VPA-7-lactone, detected both in microsomal incubations and the urine of monkeys treated with 4-ene VPA. From the available evidence both mitochondrial /3-oxidation and cytochrome P-450 mediated oxidation appear to be involved in the generation of chemically reactive intermediates of 4-ene VPA. 3. Inhibition of fatty acid metabolism by analogs of 4-ene VPA A. Hypoglvein Hypoglycin (L-(methylenecyclopropyl)alanine) is the toxic hypoglycemic principle of the unripe fruit of the Jamaican ackee tree (Sherratt, 1986). Consumption of this fruit causes widespread disturbances in carbohydrate and l ipid metabolism (Sherratt and Osmundsen, 1976), the c l inical , biochemical, and histopathological manifestations of which resemble those of Reye's syndrome and VPA hepatotoxicity. The main symptoms of hypoglycin poisoning are severe hypoglycemia, vomiting, depletion of liver glycogen, fat deposits in the liver, increased plasma free fatty acids, and massive dicarboxylic aciduria (Sherratt, 1986). This is often followed by death. Hypoglycin is metabolized to methylenecyclopropylpyruvic acid which is then oxidatively decarboxylated to methylenecyclopropylacetic acid (MCPA) in the mitochondria by the enzymes that catalyze metabolism of amino acids (Sherratt, 1986). The MCPA is activated to its CoA thio ester, part of which can be conjugated with glycine and eliminated from the body. The primary metabolic effect of hypoglycin, inhibition of fatty acid oxidation, is a consequence of the inactivation of several dehydrogenases by MCPA-CoA. The MCPA-CoA is considered to be a suicide, mechanism-based inhibitor of dehydrogenases (Wenz et al., 1981). The initial step of the dehydrogenation of acyl CoA esters by the general acyl CoA dehydrogenase enzyme is abstraction of an a-proton. This would give a transient a-carbanion which would be stabilized by the terminal double bond present in the molecule, but eventually lead to ring opening and covalent addition of the MCPA-CoA to the flavin electrophile of the enzyme. Hypoglycin toxicity is attenuated by the administration of glycine since the latter promotes excretion of MCPA as the glycine conjugate (Al-Bassam and Sherratt, 1981). Clofibrate pretreatment of rats protects rats from the toxic effects of hypoglycin (van Hoof et al., 1985) presumably because clofibrate induces mitochondrial ^-oxidation. B. 4-Pentertoic acid The hypoglycin analog, 4-pentenoic acid, is a hypoglycemic agent and strong inhibitor of mitochondrial fatty acid oxidation (Billington et a7., 1978a; Billington et aJ. , 1978b). In rats, 4-pentenoic acid produces most of the features of Reye's syndrome, i .e. fatty liver, hypoglycemia, abnormal mitochondria, elevation of serum transaminase, hyperammonemia, etc. In an effort to elucidate the mechanism by which 4-pentenoic acid inhibits fatty acid metabolism, the metabolism of 4-pentenoic acid has been studied by a number of investigators (Brendel et a7., 1969; Holland et a7., 1973; Hiltunen and Davis, 1981; Schulz, 1983). 4-Pentenoic acid is ini t ia l ly converted to its CoA thio ester and /3-oxidized to 2,4-pentadienoyl CoA, which is then channeled into two pathways. In the major pathway a key step is the NADPH-dependent reduction of 2,4-pentadienoyl CoA to 3-pentenoyl CoA, which is subsequently ^-oxidized Jo propionyl CoA and acetyl CoA. The minor pathway leads to the direct hydroxylation of 2,4-pentadienoyl CoA to give 3-hydroxy-4-pentenoyl CoA, which is then oxidized to 3-oxo-4-pentenoyl CoA. The formation of acryloyl CoA, which should be the end-product of this pathway, has been reported (Brendel et al., 1969). The inhibition of ^-oxidation by 4-pentenoic acid appears to involve the irreversible inactivation of 3-ketoacyl CoA thiolase by 3-oxo-4-pentenoyl CoA (Schulz, 1983; Schulz, 1987). It has been proposed that 3-oxo-4-pentenoyl CoA acts as a suicide inhibitor of the thiolase enzyme while being cleaved to acryloyl CoA and acetyl CoA. The inactivation is accompanied by Michael addition of the highly electrophilic a,/J-unsaturated ketone to a nucleophilic center on the thiolase enzyme. Studies of the inhibition of thiolase by 3-oxo-4-pentenoyl CoA were based on in situ formation of 3-oxo-4-pentenoyl CoA by incubating 2,4-pentadienoyl CoA with the necessary enzymes and co-factors. Thus, there is only indirect evidence for the formation of 3-oxo-4-pentenoic acid; an authentic standard had not been used in those studies. Although the thiolase enzyme is inhibited much more strongly by 3-oxo-4-pentenoyl CoA, 2,4-pentadienoyl CoA (without the cofactors required to convert it to 3-oxo-4-pentenoic acid) does inhibit this enzyme to some extent (Holland et al., 1973). 3-0xo-4-pentenoic acid is also an inhibitor of carnitine acetyltransferase (Zhong et al., 1985). There is also evidence that in vivo the site of inhibition of fatty acid /J-oxidation by 4-pentenoic acid could be at the level of the acyl dehydrogenase (Thayer, 1984). 4. Metabolic disturbances caused by VPA By virtue of its fatty acid structure, VPA interferes with several aspects of fatty acid metabolism. VPA inhibits /J-oxidation of long- and medium-chain fatty acids (Turnbull et al., 1983; Coude et al., 1983; Thurston et al., 1983a). Peroxisomal /J-oxidation is increased in the liver of rats and mice treated with VPA (Horie and Suga, 1985; Draye and Vamecq, 1987); this appears to be a compensatory mechanism for the inhibition of mitochondrial function. Administration of VPA results in an increased l ipid content of the liver (Jezequel et al., 1984; Olson et al., 1986), inhibition of ketogenesis, and decreased plasma levels of /3-hydroxybutyrate in young rats (Kesterson et al., 1984) and humans (Thurston et al., 1983b). Increased excretion of C6-Cin dicarboxylic acids is observed in the urine of patients given VPA (Mortensen et al., 1980a). All the above biochemical effects of VPA are consistent with inhibition of fatty acid ^-oxidation. VPA itself is thought to inhibit /J-oxidation by sequestering CoA due to the accumulation of valproyl CoA (Kesterson et al., 1984; Thurston et al., 1985). The inhibition of fatty acid synthesis in isolated hepatocytes by VPA may also be related to the accumulation of valproyl CoA (Becker and Harris, 1983). VPA therapy results in decreased liver and plasma carnitine levels (Bohles et al., 1982; Ohtani et al., 1982; Thurston et al., 1983b; Murphy et al., 1985). Valproylcarnitine has been identified as a minor metabolite in the urine of patients taking VPA (Millington et al., 1985). Sequestration of hepatic CoA by VPA may result in fatty acids forming excess acyl carnitines, thus producing a deficiency of carnitine (Eadie et al., 1988). Bohles et a7.(1982) found reduced serum carnitine levels in a fatal hepatotoxic case, but another incident of VPA-induced hepatic failure (Laub et al., 1986) was not associated with decreased carnitine levels, and carnitine administration did not reverse the toxicity. The level of glucose in the blood of young rats, mice, and starved rats is decreased by administration of VPA (Turnbull et al., 1983; Kesterson et al., 1984; Graf et al., 1985). Treating isolated rat hepatocytes with VPA and its metabolites (2-ene-, 4-ene-, 4-0H-, 5-0H-VPA and 2-PSA) causes concentration-dependent inhibition of gluconeogenesis (Turnbull et al., 1983; Rogiers et al., 1985). The mechanism by which VPA inhibits gluconeogenesis is not known, but it has been suggested that gluconeogenesis might be inhibited at the pyruvate carboxylase level. Acetyl CoA which induces pyruvate carboxylase is decreased after VPA administration (Cotariu and Zaidman, 1988). VPA also interferes with energy metabolism by inhibiting pyruvate oxidation and the activities of the enzymes pyruvate dehydrogenase and pyruvate decarboxylase (Turnbull et a7., 1983). In isolated liver mitochondria, oxidative phosphorylation is inhibited by VPA (Hayasaka et al., 1986). VPA therapy in humans is associated with hyperammonemia, which is mainly of renal origin, and is reversible in the absence of hepatic dysfunction (Coulter and Allen, 1981; Rawat et al., 1981; Warter et al., 1983; Zaccara et a7., 1987). A VPA-induced increase in plasma ammonia may be dose-related (Coulter and Allen, 1981) or non-dose-related (Zaret et al., 1982). One cause of the hyperammonemia may be the inhibition of carbamoyl-phosphate synthetase by the high acyl CoA to free CoA ratio caused by VPA (Turnbull et al., 1983); VPA is known to inhibit urea synthesis in healthy adults (Hjelm et al., 1986, 1987) and in isolated hepatocytes (Turnbull et al., 1983; Coude et al., 1983). VPA induces hyperglycinemia (Bartlett, 1977; Cherruau et a7., 1981), an effect attributed to the inhibition of the glycine cleavage system by both VPA and valproyl CoA (Kochi et al., 1979; Mortensen et al., 1980b; Martin-Gal 1ardo et a7., 1983). One possible mechanism of the antiepileptic activity of VPA is related to its ability to induce changes in the metabolism of y-aminobutyric acid (GABA). The drug significantly raises GABA levels in the whole brain of rodents (Horton et a7., 1977) in a dose-dependent manner (Miyazaki et al., 1988). GABA levels have been shown to be increased in the plasma of patients on VPA therapy (Loscher and Schmidt, 1980). The effect of VPA on GABA levels is related to its inhibitory action on GABA-transaminase (Loscher et a / . , 1981) and succinate semi aldehyde dehydrogenase (Sawaya et a / . , 1975), enzymes involved in the degradation of GABA. Thus, i t is clear from the preceding account that VPA interferes with several aspects of intermediary metabolism and energy production. However, the evidence available does not implicate the involvement of any single metabolic effect of the drug as the cause of the fatal hepatotoxicity of the drug. 5. Teratogenicity of VPA Genetic factors related to epilepsy, seizures during pregnancy, and the teratogenic potential of antiepileptic drugs can all contribute to a high rate of teratogenicity in epileptic patients (Tsuru et al., 1988). Next to hepatotoxicity, teratogenicity is the major toxic effect of VPA (Brown et al., 1985). VPA is teratogenic in the mouse, rat, rabbit, and monkey (Hauck and Nau, 1989). VPA crosses the placenta and can affect the fetus (Dickinson et al., 1979). Published case reports of fetal malformation in epileptic mothers on VPA include neural tube defects (spina bifida) and other defects collectively known as fetal valproate syndrome (Nau et al., 1981a; Tein and MacGregor, 1985; Jager-Roman et al., 1986). In a prospective study of 172 deliveries, the incidence of malformations in infants exposed to VPA polytherapy was higher than that of VPA monotherapy or monotherapy with any of the antiepileptic drugs. The incidence was also higher than combinations that did not include VPA (Kaneko et al., 1988). In an animal study where six VPA metabolites were tested, only 4-ene VPA was clearly teratogenic (Nau and Loscher, 1986). Since both VPA and 4-ene VPA have teratogenic activity (Brown et al., 1985; Hauck and Nau, 1989), it appears there are multiple biochemical mechanisms underlying the teratogenic activity of these compounds. In a comparative study' of the teratogenicity of VPA and the enantiomers of 4-ene VPA, it was found that the S-enantiomer was more potent than VPA while the R-enantiomer was less potent (Hauck and Nau, 1989). It is interesting to note at this point that the production of (E)-2,4-diene VPA was found to be higher in hepatocyte incubations containing (S)-4-ene VPA compared to that of its antipode (Porubek et al., 1988). Whether the biochemical mechanism of the teratogenicity is similar to the hepatotoxicity is not clear at present. However, since electrophilic intermediates are known to cause the teratogenicity of other agents (e.g. the arene oxide metabolite of thalidomide (Nelson, 1982)), i t would be of interest to determine whether reactive metabolites derived from VPA or 4-ene VPA are involved in the teratogenicity of these agents. In light of what is known about the metabolism of (S)-4-ene VPA, and the fact that it is more teratogenic than VPA, it could be speculated that (E)-2,4-diene VPA or its further metabolite might be involved in the teratogenicity of both 4-ene VPA and VPA. 6. The hepatotoxicity of VPA A. Incidence and pathological manifestations Many investigators classify VPA associated hepatotoxicity into two forms (Gram and Bentsen, 1983; Dreifuss and Langer, 1988; Cotariu and Zaidman, 1988). The first form is a reversible, dose-related condition in which laboratory markers of hepatotoxicity disappear upon withdrawal of VPA therapy while the second form is a non-dose-related idiosyncratic liver failure that results in death. However, there are a number of problems associated with such a classification. Several cases of seemingly irreversible hepatotoxicity were reversed (Jeavons, 1984; Scheffner et a / . , 1988), and thus it is not clear whether the two forms of hepatotoxicity are distinct diseases or merely manifestations of the degree of severity. In a recent case report (Kimura et a7., 1989), two patients who developed hepatotoxicity related to VPA therapy and who presented with Reye-like syndrome and coma, recovered after discontinuation of VPA. Necrosis, which is a frequent feature of VPA hepatotoxicity, was present in one of the patients. Microvesicular fat, a common histological finding of fatal cases, was the dominant histological feature in two subjects with non-fatal liver injury (Powell-Jackson et a7., 1984). In the non-fatal cases clinical and laboratory findings do not allow differentiation from those of fatal cases. The moderately frequent incidence of minor hepatic toxicity, the reversibility of some of the severe cases, and the rarity of fatal cases, suggests that VPA hepatotoxicity in general is common. The hepatotoxicity appears to become fatal in those patients with a decreased capacity to detoxify the drug and its metabolites either because of illness or an inherited metabolic disease. Thus, by definition, the hepatotoxicity becomes idiosyncratic. The fatal hepatotoxicity associated with VPA has been described in terms of a Reye-like syndrome and has similarities to Jamaican vomiting sickness (hypoglycin poisoning)(Gerber et al., 1979; Zimmerman and Ishak, 1982). Prior to 1987 more than 100 patients had died world-wide from VPA-induced liver failure (Scheffner et al., 1988). A majority of the cases of hepatotoxicity occurred in children, most of them developing their first symptoms of toxicity within six months of initiation of VPA treatment. Many of the affected children, 34 out of 67 cases reviewed by Jeavons (1984), had other conditions such as mental retardation, metabolic disorders, neurological diseases and genetic abnormalities. In almost all cases the symptoms that patients presented with were vomiting, nausea, abdominal pain and apathy. Of the 67 cases analyzed by Jeavons (1984) and for those for which data were available, biluribin levels were abnormal in 40 and normal in seven. The serum glutamic-oxaloacetic transaminase level was abnormal in 56 cases, normal in one and unknown in eight, while serum glutamic-pyruvic transaminase levels were abnormal in 40 and unknown in 23. There was also evidence of a blood coagulation disorder in almost all cases where there was information. For 20 patients having serum ammonia data, ammonia levels were high in 16 and normal in four. In the 23 cases of VPA-associated fatal liver failure analyzed by Zimmerman and Ishak (1982), the most frequent histological finding was microvesicular steatosis, which was present in 17 of 21 cases for which data were available. The review by Jeavons (1984) identified 22 cases of steatosis accompanied by necrosis, 8 steatosis only, 20 necrosis only, and four steatosis with cirrhosis. According to a review of United States cases (Dreifuss et al., 1987), the overall rate of fatal hepatotoxicity related to VPA was 1 in 10000. Risk factors for fatal hepatotoxicity were identified as patient age, co-medication with other antiepileptic drugs and the presence of other diseases. Patients at high risk (incidence of 1 in 500) were children under two years of age who were on polytherapy and who had other medical problems such as mental retardation, developmental delay, congenital abnormalities, and metabolic disorders. In a follow-up study (fatalities in 1985-1986) a four-fold decrease in hepatic fatalities associated with VPA therapy was reported (Dreifuss et a7., 1989; Dreifuss, 1989). On the basis of a German study, the incidence of VPA-related fatal liver failure was placed at 1 in 5000 (Scheffner et a7., 1988). These authors point out that based on the criteria set out by Dreifuss et a7. (1987), they were unable to single out any high risk group from a review of 16 fatal cases. Hence they argue that all patients taking VPA should be considered at risk for fatal hepatotoxicity. B. Biochemical mechanism The mechanism by which VPA causes hepatotoxicity is unknown although one of its metabolites, 4-ene VPA, has been heavily implicated (Gerber et al., 1979; Zimmerman and Ishak, 1982; Rettenmeier et al., 1985, 1986b), and various biochemical mechanisms have been proposed (Eadie et al., 1988). Because microvesicular steatosis, which is the most frequent lesion in VPA-related hepatotoxicity, is also the histological hallmark of Jamaican vomiting sickness caused by MCPA and of 4-pentenoic acid toxicity in the rat, it has been suggested that similar biochemical mechanisms underlie the liver toxicity caused by these three agents. Support for this comes from the close structural similarity between the VPA metabolite 4-ene VPA, 4-pentenoic acid, and MCPA (Fig. 2). As mentioned earlier 4-pentenoic acid is a potent inhibitor of fatty acid metabolism and is apparently metabolized to the reactive electrophilic species, 3-oxo-4-pentenoic acid (Fig. 3), which alkylates 3-ketoacyl CoA thiolase (Schulz, 1983). Because 4-ene VPA is the 2-propyl analog of 4-pentenoic acid, it has been postulated by Rettenmeier et a7.(1985) that 4-ene VPA may undergo a similar bioactivation (Fig. 4). This may explain the liver steatosis observed in patients with VPA-induced liver damage. It has indeed been shown that 4-ene VPA is a strong steatogen in rats (Kesterson et a7., 1984), is toxic to cultured hepatocytes in a dose-dependent fashion (Kingsley et al., 1983) and inhibits ^-oxidation of fatty acids in vitro (Bjorge and Bail l ie , 1985) and in vivo (Granneman et al., 1984c). More recently, the covalent binding of radiolabeled 4-ene VPA to proteins of rat liver has been documented (Porubek et al., 1989). The role of 4-ene VPA in VPA hepatotoxicity has recently been reviewed by Baillie (1988). A major metabolic pathway of 4-ene VPA in the rat and monkey is /J-oxidation leading to the formation of (E)-2,4-diene VPA (Rettenmeier et al., 1985, 1986b). It was hypothesized that (E)-2,4-diene VPA undergoes further /J-oxidation to give the ultimate toxic electrophilic species 3-keto-4-ene VPA (Baillie, 1988). However, in metabolic studies involving rats (Granneman et a7., 1984a; Rettenmeier et a7., 1985) and the rhesus monkey (Rettenmeier et al., 1986b), there was no evidence for the formation of 3-keto-4-ene VPA. This lack of detection was presumed to result from the unstable and highly reactive nature of this compound. COOH 4-ENE VALPROIC ACID COOH 4-PENTENOIC ACID M E T H Y L E N E C Y C L O -PROPYLACET IC ACID Figure 2. Structurally related liver toxin carboxylic acids. 4-pentenoic acid 'COOH "COOH (E)-2,4-pentadienoic acid CH2=CHCOOH + CH 3C0-S-CoA acrylic acid acetyl CoA COOH 3-OH-4-pentenoic acid COOH 3-oxo-4-pentenoic acid Figure 3. Mitochondrial metabolism of 4-pentenoic acid leading to the generation of chemically reactive intermediates. ro COOH MICROSOMAL OXIDATION 4-ene VPA \ /3-OXIDATION \ MITOCHONDRIA COOH (E)-2,4-diene VPA Figure 4. Metabolic pathway of VPA proposed to lead to the production of reactive and hepatotoxic metabolites. The finding that VPA hepatotoxicity is more prevalent in polytherapy than monotherapy (Dreifuss et al., 1987) is consistent with the view that a metabolite(s) of VPA produced in large quantity by the induced liver might be partly responsible for the hepatotoxicity. However, l i t t l e has been published on VPA metabolite profiles in patients who had liver damage coincident with VPA therapy. In one report of a child who died of VPA hepatotoxicity, the plasma and urine levels of 4-ene- and 2,4-diene-VPA were highly elevated (Kochen et al., 1983). In another case of fulminant hepatic failure (Kuhara et al., 1990), 2-PGA was detected at abnormally high levels whereas VPA glucuronide and 3- keto VPA were undetectable indicating the impairment of VPA glucuronidation and /J-oxidation. However, in the absence of additional information it is not clear whether the increased formation of certain metabolites is an effect or cause of the hepatotoxicity. Since the centrizonal necrosis of the liver observed in VPA-induced liver toxicity is not seen in other lesions associated with microvesicular steatosis (Jamaican vomiting sickness, Reye's syndrome, 4- pentenoic acid toxicity), some investigators (Zimmerman and Ishak, 1982) have been prompted to conclude that the necrosis part of the VPA lesion may be caused by a reactive metabolite different from that which causes microvesicular steatosis. Other mechanisms that do not involve formation of a chemically reactive metabolite, but that have been proposed to contribute to the toxicity are: carnitine deficiency (Coulter, 1984), interference of VPA with intermediary metabolism in a compromised liver, and depletion of CoA (Kesterson et al., 1984). 7. Glutathione conjugation and its toxicological significance The formation of chemically reactive metabolites from drugs and other xenobiotics is responsible for the initiation of many types of tissue lesions. Mitchell et a7.(1982) classified chemically reactive metabolites into electrophiles, organic free radicals, and those that generate active G"2 species such as O2". By binding covalently to proteins and nucleic acids, electrophilic intermediates can cause toxic effects, mutations, and cancer. There is a good correlation between covalent binding of electrophiles to tissue components and tissue necrosis (Gillette, 1974). Many of the agents that are chemically reactive or are metabolically converted to chemically reactive intermediates affect cellular levels of GSH (Boobis et al., 1989). This includes compounds that are substrates for GSH conjugation and those responsible for the oxidation of GSH to GSH disulfide (GSSG). GSH acts as a nucleophilic scavenger by converting electrophilic centers to S-linked conjugates and as a substrate in the GSH redox system that is involved in the destruction of hydroperoxides. GSH is an essential constituent of the ce l l , and failure to maintain adequate levels is detrimental to the survival of the ce l l . The protective role of GSH against chemically induced injury to tissues can be compromised by the depletion of GSH, which may lead to cell injury and death (Reed and Fariss, 1984). A. Glutathione GSH (L-7-glutamyl-L-cysteinylglycine) constitutes 90% of the non-protein thiol in the cell (Reed, 1985). The cellular concentration of GSH varies between 0.5 to 10 mM depending on the cell type (Kosower and Kosower, 1978). The GSH level in the liver is close to twice that of the kidneys and over threefold higher than in the lung. Most of the intracellular GSH is present in the reduced form with less than 5% present as GSSG. Chemically GSH is characterized by the -/-glutamyl linkage which makes it resistant to degradation by typical proteases. The cysteinyl residue provides a nucleophilic thiol group that reacts with electrophiles and metabolically produced oxidizing species. The high number of hydrophilic groups of GSH ensures high water solubility for its conjugates. Its relatively high molecular weight provides for the preferential biliary excretion of GSH adducts. GSH and its thiolate anion are soft and strong nucleophiles that react well with soft electrophiles but poorly with both weak and hard electrophiles (Ketterer et al., 1983). B. Glutathione S-transferases GSH S-transferases are a group of isoenzymes with overlapping substrate specificity and widespread distribution in nature. In mammals, GSH transferases are found in large quantities, representing 10% of the cytosolic protein of rat liver, which can reach greater than 20% following induction (Jakoby, 1985). Although the highest level of transferase is in the cytosol (0.2 to 0.5 mM), membrane-bound transferases have been detected in the microsomes and mitochondria (Ketterer et al., 1983). In the rat liver there are at least seven soluble isoenzymes with similar dimeric structure and in which the monomers fall into three differing molecular weight categories. In addition to catalysis, these enzymes serve storage and transport functions (Kaplowitz, 1980). The transferases have a low-affinity, broad-specificity site for lipophilic compounds and another site which is specific for GSH and its 0-alanine analog (Jakoby, 1985). Transferases catalyze the initial step of NAC adduct (mercapturic acid) synthesis in which a thioether bond is formed between GSH and the carbon atom of an electrophilic substrate (Habig et a7., 1974). The GSH conjugate is normally excreted into bile and is also metabolized by a series of reactions to the NAC conjugate which may be excreted in the bile or urine. The broad specificity of the transferases has perhaps evolved to accommodate the wide spectrum of structural types of toxic substances to which the organism is exposed. The conjugation reactions catalyzed by transferases fall into two basic categories: nucleophilic substitution and nucleophilic addition to a polarized double bond. Transferases also catalyze the reaction of GSH with substrates having heteroatom electrophilic centers. The basis of the catalytic mechanism of transferases appears to be to promote the formation of the thiolate anion (GS") from GSH (Jakoby and Habig, 1980). This seems reasonable since the rate of reaction of GS" with electrophiles is three orders of magnitude greater than that of GSH. The mechanism proposed for the catalytic activity is related to the proximity effect (Jakoby and Habig, 1980). The active site of the enzyme consists of a specific binding area for GSH (the effect of binding to which is the reduction of the pKa of GSH) and another electrophilic binding site, which is close to the GSH binding site and is characterized by a lipophilic spacious surrounding. Thus, when a sufficiently reactive electrophilic substrate is in proximity to the thiolate anion a reaction takes place. Formation of GSH conjugates can also occur nonenzymatically. However, in vivo, it is assumed that the contribution of the noncatalytic reaction is negligible owing to the slow formation of electrophiles by enzymes with slow rates of reaction such as cytochrome P-450 (Ketterer et al., 1983). 8. Rationale and objectives While there may be multiple mechanisms by which VPA causes hepatotoxicity, there is a strong biochemical basis to implicate the 4-ene VPA pathway as discussed in preceding sections. The putative reactive metabolite, 3-keto-4-ene VPA, that would arise from this pathway has not been identified in serum or urine, presumably because of its chemical instability. The 3-keto-4-ene VPA, akin to other a,/J-unsaturated carbonyl compounds, is highly reactive and can theoretically undergo Michael additions to give covalent adducts of nucleophiles. As alluded to in an earlier section, reactive intermediates produced in the body are usually detoxified by the formation of GSH conjugates. In view of the nature of the putative reactive metabolite in the case of VPA, it is possible that a GSH conjugate may be formed in vivo. Hence the logical derivative of 3-keto-4-ene VPA in urine will be the NAC conjugate, while it may be possible to detect the GSH conjugate in bile. The (E)-2,4-diene VPA metabolite is presumably a /J-oxidation product of 4-ene VPA and is a key intermediate in the metabolism of 4-ene VPA. A detailed study of the metabolism of this diene will not only enable the possible detection of the putative a,^-unsaturated ketone but also the detection of other metabolites that point to additional sources of toxicity. The metabolism of 4-pentenoic acid will be studied in a parallel fashion to that of 4-ene VPA with the aim of characterizing thiol conjugates of reactive metabolites of 4-pentenoic acid. The information that will be obtained from this study has a twofold significance. Firstly, i t may provide the answer to the question of whether the parallel drawn between the metabolism of 4-ene VPA and 4-pentenoic acid and hence the proposed mechanism of hepatotoxicity is realistic. Secondly, the detection of the thiol conjugates of the a,^-unsaturated keto metabolite of 4-pentenoic acid will provide the first direct biochemical evidence for the formation of the metabolite of 4-pentenoic acid responsible for the irreversible inhibition of /3-ketoacyl CoA thiolase, one of the enzymes in the fatty acid /3-oxidation pathway. Hence the present study will focus on the metabolism of 4-ene VPA, (E)-2,4-diene VPA, and 4-pentenoic acid and/or (E)-2,4-pentadienoic acid with particular emphasis on proving the formation of reactive metabolites of VPA and their thiol conjugates and devising methods for their detection in humans when applicable. Specific objectives 1. To study the metabolism of 4-ene- and (E)-2,4-diene-VPA in rats with particular emphasis on pathways generating reactive metabolites and to characterize any novel metabolites spectroscopically and by synthesis. 2. To chemically synthesize 3-keto-4-ene VPA and its NAC adduct and to investigate the formation of these metabolites in rats following the administration of VPA, 4-ene- and (E)-2,4-diene-VPA. 3. To synthesize the GSH conjugate of 3-keto-4-ene VPA both chemically and from in situ metabolism of (E)-2,4-diene- and 4-ene-VPA by rat liver mitochondrial fraction. 4. To study the metabolism of 4-pentenoic acid and/or (E)-2,4-pentadienoic acid with the objective of trapping the electrophilic 3-keto-4-pentenoic acid in vivo in the form of its thiol conjugates. 5. To develop a highly sensitive NICI assay for the quantitation of the known VPA metabolites, including 4-ene- and (E)-2,4-diene-VPA, in pediatric patients and to use the method for the detection of new metabolites. 6. To investigate the formation of NAC conjugates of reactive metabolites in patients on VPA therapy and to develop a method to quantitate these conjugates as a measure of exposure of patients to toxic metabolites of VPA. I I . EXPERIMENTAL 1. Chemicals and materials Chemicals were reagent grade and obtained from the following sources. A. Aldrich Chemical Co. (Milwaukee, Wisconsin) Acrolein, allyl bromide, Amberlite XAD-2 resin, p-benzoquinone, n-butyllithium (1.6 M in hexane), calcium hydride, chloroform-d, deuterium oxide, l,8-diazabicyclo[5.4.0]undec-7-ene (DBU), DiazaldR, diisopropylamine, diisopropylethyl amine, hexamethylphosphoramide (HMPA), lithium aluminium hydride, methanesulfonyl chloride, methanol-d4, 2-methylglutaric acid (2-MGA), palladium (II) chloride, pentanoic acid, (E)-2-pentenoic acid, 4-pentenoic acid, potassium hydride (35% dispersion in mineral o i l ) , propionaldehyde B. Alfa Products (Danvers. Massachusetts) Pentafluorobenzyl bromide (PFBB) C. J . T. Baker Inc. (Phillipsburg. New Jersey) Sil ica gel for flash chromatography, solid phase extraction cartridges (6 mL size) D. BDH Chemicals (Toronto. Ontario) Acetone, benzene, ether (anhydrous), hydrochloric acid, sodium dihydrogen orthophosphate, sodium hydroxide, sodium sulfate (anhydrous), petroleum ether, pyridine, tetrahydrofuran (THF) 36 E. Boehringer Mannheim (Montreal, Quebec) Adenosine triphosphate (ATP), bovine serum albumin F. Caledon Laboratories Ltd. (Georgetown, Ontario) All solvents from this supplier were of distilled-in-glass grade-acetonitrile, dichloromethane, ethyl acetate, methanol G. ICN Biochemicals Inc. (Plainview, New York) Di-n-propylacetic acid (VPA) H. Mallinckrodt Chemicals (St. Louis, Missouri) Acetic acid (glacial), sodium bicarbonate I. Matheson Ltd. (Edmonton, Alberta) Hydrogen chloride gas J . Pierce Chemical Co. (Rockford, Illinois) N-Methyl-N-trimethylsilyl-trifluoroacetamide (MSTFA), triethylamine K. Sigma Chemical Co. (St. Louis, Missouri) Coenzyme A (CoA), glutathione (GSH), N-acetyl-L-cysteine (NAC), urethane L. Supelco (Oakville, Ontario) 3% Dexsil 300 on 100/120 mesh Supelcoport M. Whatman Ltd. (Maidstone, Kent, England) Silica gel plates 37 N. Compounds synthesized in our laboratory 2-Propyl-2-pentenoic acid (E- and Z-isomers, 2-ene VPA), 2-propyl-3-pentenoic acid (3-ene VPA, stereochemistry not determined), 2-(l'-propenyl)-2-pentenoic acid (E,E-and E,Z-isomers, 2,3'-diene VPA), 2-propyl-3-oxopentanoic acid (3-keto VPA), 2-propyl-4-oxopentanoic acid (4-keto VPA), 2-propyl-3-hydroxypentanoic acid (3-OH VPA), 2-propyl-4-hydroxypentanoic acid as the -/-lactone (4-OH VPA), 2-propylglutaric acid (2-PGA), 2-propylsuccinic acid (2-PSA), di-([3,3,3-2H]propyl)acetic acid ([2H6]VPA), [3,5,5-2H]3-heptene-4-carboxylic acid ([2H3]2-ene VPA), [3',3',3'-2H]2-propyl-3-oxopentanoic acid ([2H3]3-keto VPA) (Acheampong et al., 1983; Acheampong et al., 1984; Abbott et al., 1986; Lee et al., 1989). 2. Animals and surgical equipment Animals and surgical equipment were purchased from the following suppliers. Animal Care Facility (University of British Columbia, Vancouver, B.C) Sprague Dawley male rats (250-350 g) Becton Dickinson (Rutherford, New Jersey) Yale needle 23G, 1", 19G 1.5"; tuberculin syringe 1 cc Clay Adams (Parsippany, New Jersey) Polyethylene tubing PE-10 Ethicon Sutures Ltd. (Peterborough, Ontario) 4-0 Silk 3. Instrumentation A. Capillary column 6C/MS A Hewlett-Packard 5987A quadrupole GC/MS was used in electron impact (EI), negative ion chemical ionization (NICI), and positive chemical ionization (PCI) modes with the following conditions and settings. EI analysis: column, DB-1701 (J and W Scientific, Folsom, California), 30 m x 0.32 mm x 0.25 /im film thickness; oven, 50-140°C at 30°/min, 140-260°C at 5°/min, hold 10 min; source, 240°C; injection port, 240"C; open split interface, 250°C. Electron energy was 80 eV and source pressure 10"6 torr. The carrier gas was helium at a flow rate of 1 mL/min with a back pressure of 10 psi. NICI analysis: column, DB-1, 30 m x 0.32 mm x 0.25 /im. Oven temperature was programmed from 200-250°C at 30°/min, hold 20 min, 250-260°C at 2°/min> hold 12 min, for conjugated metabolites and 110-140'C at 30°/min, 140-260°C at 5°/min for non-conjugated metabolites. Other temperatures were the same as for EI except for the source temperature which was set at 200°C. NICI spectra were recorded at an ionization energy of 120 eV and source pressure of ca. 1 torr using methane as the moderating gas. B. Packed column GC/MS Intermediates and end products of synthetic reactions were characterized using a Hewlett-Packard 5700A gas chromatograph interfaced to a Varian MAT-111 mass spectrometer via a variable s l i t separator. Data were processed using a Packard Bell computer (IBM AT clone) and a program written in our laboratory. Mass spectra were obtained from m/z 14 to 750 at 12 scans per minute. Total ion current (TIC) chromatograms were plotted from m/z 50 to 500. Ionization energy was 70 eV and source pressure 5 x 10"^  torr. Other operating conditions: column, 1.8 m x 2 mm i .d . packed with 3% Dexsil 300 on 100/200 mesh Supelcoport; oven temperature, init ial 50°C, rate 8°/min to 270°C, hold 10 min at 270°C. The injection port, transfer line and separator temperatures were held at 250°C. C. HPLC HPLC analysis was carried out on an HP-1050 series LC with a quaternary pump and variable wavelength UV detector fitted with either a Hypersil 0DS analytical column (4.6 x 200 mm) or a Whatman Parti si 1 10 0DS-2 preparative column (9 x 250 mm). The wavelength of detection was 210-220 nm. The mobile phase consisted of various proportions of 0.01 or 0.02 M NaH2P04 buffer (adjusted to pH 3.5 with H3PO4) and methanol. All analyses were performed isocratically at a flow rate of 1 mL/min (analytical column) or 3 mL/min (preparative column). D. LC/MS and LC/MS/MS LC/MS and LC/MS/MS were performed at Cantest Ltd. of Vancouver on a Sciex atmospheric pressure ionization (API) tandem mass spectrometer coupled to a heated nebulizer LC/MS interface. The LC system consisted of an Applied Bio System 140A pump and a 4.6 X 200 mm 0DS column connected to a Valco VE106C injector. The mobile phase consisted of acetonitrile and 10 mM ammonium acetate buffer, programmed 20 to 100% acetonitrile over a 10 min period. The nebulizing gas was N2 and vapor temperature 150°C. Collision activated dissociation (CAD) daughter ion spectra were obtained using argon as the collision gas (collision energy 20-60 eV) in the second quadrupole. Detection was both in the positive and negative ion modes. E. FAB/MS FAB spectra were obtained in the Department of Chemistry and were recorded on an AEI MS9 mass spectrometer with an in-house built saddle field gun and ion source. Ionization was effected using xenon gas at 1-1.5 mA current and 8 kV accelerating potential. Spectra were obtained at a scan rate of 10 sec/decade. Samples in methanolic solutions were applied to the 3-nitrobenzyl alcohol matrix. F. NMR spectroscopy All NMR spectra were performed in the NMR facil ity of the Chemistry Department. High field *H-NMR spectra were recorded on a Bruker WH-400 spectrometer using tetramethylsilane as the reference. 13C-NMR were obtained on either a Varian XL-300 or Bruker AC-200 spectrometer operating at 75 MHz, by broad-band decoupling on the proton frequency and by an attached proton test. Samples were dissolved in methanol-d4 and D2O in the case of conjugated metabolites and CDCI3 for all other compounds. 4. Chemical syntheses A. General Purification of synthetic products was carried out either by fractional distillation or medium pressure (flash) chromatography according to the method of St i l l et al. (1978). Up to 2 g of material were purified using a 2.5 X 20 cm si l ica gel column. The eluent composition was selected on the basis of an rf value of about 0.3 on TLC. The sample was dissolved in the less polar component of the mobile phase and eluted at 2 mL/min with approximately 5 mL fractions being collected. The syntheses described in this thesis that involve alkylation reactions are based on the general procedure of the alkylation of ester enolates described by R.J. Cregge et al. (1973). B. Synthesis of 3-keto-4-ene VPA This compound was synthesized as the methyl ester by oxidizing methyl 3-0H-4-ene VPA with tert-butyl hydroperoxide in the presence of chromium trioxide. The 3-0H-4-ene VPA was prepared using the following procedure. n-Butyllithium (69 mL, 0.11 mol) was added dropwise to diisopropylamine (15.5 mL, 0.11 mol) in THF (100 mL) at 0°C. The mixture was stirred for 20 min and cooled in a dry ice acetone bath to -78°C, and methyl pentanoate (11.6 g, 0.1 mol) in THF (10 mL) was added dropwise over a period of 30 min. After stirring the reaction mixture for another 30 min, acrolein (6.7 mL, 0.1 mol) in 10 mL THF was added dropwise and the reaction allowed to continue for 60 min before being quenched with 15% HCl solution. The mixture was extracted with ether, the etheral extract washed with NaHC03 and water, and dried over anhydrous Na2S04« The product (3-0H-4-ene VPA methyl ester) was purified by fractional distillation (bp 62"C/0.1 mm, yield 72%). GC/MS: (MW=172) m/z 87(100%), 55(82%), 116(32%), 130(3%), 141(2%), 155(2%). iH-NMR: 8 0.92(t,3H,CH 3-), 1.28-1.73(m,4H,-CH2-CH2-), 2.47-2.57(m,IH,-CH-), 2.60-2.73(bs,lH,-0H), 3 .70(s ,3H,-OCH3) , 4.31(t,lH,-CH-0), 5.15-5.35(m,2H,CH2=), 5.80-5.92(m,1H,=CH-). tert-Butyl hydroperoxide (9.6 mL, 70% solution, 0.07 mol) and methyl 3-OH-4-ene VPA (1.72 g, 0.01 mol) were successively added to a stirred solution of chromium trioxide (100 mg, 0.001 mol) in 20 mL dichloromethane (Muzart, 1987a, 1987b). After stirring for 6 hr at room temperature the mixture was treated with saturated NaHC03 solution and the organic layer dried (Na2S04). The product (3-keto-4-ene VPA methyl ester) was purified by flash chromatography using 10% ethyl acetate in pet. ether (yield 75%). GC/MS: (Mw=170) m/z 55(100%), 128(9%), 84(7%), 96(5%), 139(3%), 170(1%). iH-NMR: 6 0.94(t,3H,CH3-), 1.30-1.37(m,2H,CH3-CH2-), 1.82-1.95(m,2H,-CH 2 -), 3.71(t,lH,-CH-), 3.73(s,3H,-0CH3), 5.86-6.51(m,3H,CH2=CH-). C. Synthesis of the GSH conjugate of 3-keto-4-ene VPA methyl ester To a solution of GSH (921 mg, 3 mmol) in 5 mL of water was added a solution of 3-keto-4-ene VPA methyl ester (510 mg, 3 mmol) in 10 mL of methanol. The pH of the mixture was adjusted to 7.8 with 10 M NaOH solution and the mixture stirred at room temperature for 30 min at which time all substrate was converted to product as determined by HPLC. The product was purified by preparative HPLC. Repeated injections (200 /iL) of a methanol ic solution of the crude product (70 mg/mL) were made on the preparative column using a mobile phase of 0.02 M NaH2P04-methanol (60:40). The conjugate eluted at 4.02 min. FAB/MS diagnostic ions: m/z 171(M-GS)+, 478(M+H)+, 500(M+Na)+. 1H-NMR: 6 0.94(t,3H,CH3-), 1.27-1.36(m,2H,CH3-CH2-), 1.81(t,2H,-CH2-CH2-), 2.15-2.23(m,2H,Glu#3'), 2.55-2.62(m,2H,Glu7T'), 2.79(t,2H,-S-CH2-CH2-C0-), 2.83-2.9O(m,lH,Cys0), 2.94(t,2H,-S-CH2-CH2-CO-), 3.05-3.12(m,IH, Cys/8'), 3.71(t,lH,Glua), 3.75(s,3H,-0CH3), 3.80(s,2H,Glyaa'), 4.57(lH,dd ,Cysa). 13C-NMR: 6 14.2(CH3-), 21.5(CH3-CH2-), 26.3(-CH2-CH2-), 27.7(-S-CH2-CH2-CO-), 31.1(-S-CH2-CH2-C0-), 32.9(Glu0), 34.6(Cys0), 44.2(Glu-y), 48.2(Glya), 53.2(-0CH3), 54.3(Cysa), 55.3(Glua), 59.8(-CH-C00CH3, weak signal), 172.2(Cys-CO-), 172.5(G1u-CO-), 174.3(Gly-C00H), 175.3(Glu-COOH), 176(-C00CH3), 206.9(-CH2-CO-CH-). D. Synthesis of the methyl NAC conjugate of 3-keto-4-ene VPA methyl  ester This adduct was synthesized by refiuxing the substrate with the methyl ester of NAC according to the method of Delbressine et a/. (1981) for the synthesis of the methyl ester of the NAC conjugate of methyl crotonate. The methyl ester of NAC was synthesized by stirring NAC in excess methanol in the presence of thionyl chloride (van Bladeren et al., 1980). To a solution of methyl 3-keto-4-ene VPA (1.72 g, 10 mmol) in 50 mL p-dioxane, methyl NAC (1.77 g, 10 mmol) and triethylamine (2.8 mL, 20 mmol) were added and the mixture refluxed for 1 hr. The solvent was removed by flash evaporation and the residue purified by HPLC using 0.01 M NaH2P04-methanol (40:60) as the mobile phase with which the conjugate eluted at 6.7 min. GC/MS: PCI: m/z 348(100%)(M+H)+, 376(40%)(M+C2H5)+, 388(18%)(M+C3H5)+. iH-NMR: S 0.94(t,3H,CH3-), 1.27-1.36(m,2H,CH3-CH2-), 1.80(t,2H,-CH2-CH 2), 2.01(s,3H,-NHCOCH3), 2.78(t,2H,-S-CH2-CH2-CO-), 2.82-2.86(m,IH, Cys0), 2.89(t,2H,-S-CH2-CH2-C0-), 2.98-3.O4(lH,m,Cys0'), 3.73(s,3H,-OCH3), 3.75(s,3H,-S-CH2-CH-C00CH3), 4.63(lH,dd,Cysa). 13C-NMR: 6 14.2(CH3-), 21.6(CH3-CH2-), 22.4(-NHC0CH3),26.7(-CH2-CH2-), 31.2(-S-CH2-CH2-CO-), 34.6(-S-CH2-CH2-C0-), 43.0(Cys/J), 52.8(Cysa), 52.9(-0CH3), 53.7(-S-CH2-CH-C00CH3), 59.5(-C0-CH-C00CH3, weak signal), 171.9(-NHC0CH3), 172.5(-COOCH3), 173.2(-S-CH2-CH-C00CH3), 205.5(-CH2-CH2-CO-CH-). E. Synthesis of (E)-2,4-diene VPA Methyl 3-0H-4-ene VPA synthesized as described in section B above was converted to 2,4-diene VPA employing two synthetic approaches. The hydroxy compound (8.6 g, 0.05 mol), triethylamine(7.7 mL, 0.055 mol) and dichloromethane (100 mL) were cooled to 0°C and methanesulfonyl chloride (4.3 mL, 0.055 mol) was added dropwise and the mixture stirred for 45 min at room temperature (Lee et al., 1989). The precipitate formed was filtered off and following the removal of the solvent the residue was reconstituted with THF (80 mL), cooled to 0°C and potassium hydride (11.4 g, 0.1 mol) added. The reaction mixture was stirred at room temperature for 7 hr before being quenched cautiously at -78°C with water. The product was isolated by extraction with ether and purified by fractional distillation (bp 6VC/1 mm) to give 3.3 g (yield 42%) of a 3:1 mixture of the E and Z isomers of 2,4-diene VPA methyl ester. The product was further purified by flash chromatography (5% ethyl acetate in pet. ether) and afforded 94% of the E isomer as determined by NMR. To obtain the free acid, a portion of the ester was refluxed in two equivalents of aqueous NaOH at 60°C for 24 hr. The mixture was then treated with dilute HCl solution, extracted with ether and, following the removal of ether, was purified by column chromatography (5% methanol in chloroform). The purity of the diene, as judged by HPLC relative peak area, was 98%. In the second synthetic route the mesylate synthesized as described above was subjected to elimination using 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU). To a solution of the mesylate in THF (5.0 g, 0.02 mol), DBU (3.8 mL, 0.025 mol) was added and the mixture refluxed for 2 hr (Rettenmeier et a / . , 1989). The solution was then allowed to cool to room temperature, quenched with water and extracted with ether. The etheral layer was washed successively with dilute HCl and NaOH and dried (Na2S04). Following column purification as described above, the methyl ester of (E)-2,4-diene VPA was obtained in 76% yield. GC/MS: (MW=154)(methyl ester): m/z 95(100%), 154(83%), 93(63%), 79(59%), 59(54%), 123(49%), 67(45%), 139(25%), 111(22%). GC/MS: (Mw=140)(free acid): m/z 140(100%), 95(97%), 67(82%), 41(75%), 111(50%), 125(34%). ^H-NMR: (methyl ester): 8 0.92(t,3H,CH3-), 1.41-1.52(m,2H,CH3-CH2-), 2.40(t,2H,-CH 2-CH 2-), 3.76(s,3H,-OCH3), 5.44(dd,1H,CH2=), 5.56(dd, 1H,CH2=), 6.61-6.72(m,lH,=CH-), 7.17(d,IH,-CH=). *H-NMR: (free acid) 8 0.93(t,3H,CH3-), 1.45-1.57(m,2H,CH3-CH2-), 2.41(t,2H,-CH 2-CH 2-), 5.50(dd,lH,CH2=), 5.62(dd,1H,CH2=), 6.62-6.74(m, 1H,=CH-), 7.30(d,lH,-CH=). F. Synthesis of the GSH conjugate of (E)-2.4-diene VPA To a solution of (E)-2,4-diene VPA (140 mg, 1 mmol) in methanol (5 mL), GSH (614 mg, 2 mmol) in 0.1 M NaH2P04 (pH 7.4, 20 mL) was added and the mixture stirred at room temperature. The progress of the reaction was monitored by TLC using the solvent system of chloroform/methanol/acetic acid (7:3:1) for development and spraying with Young's reagent (Knight and Young, 1958). One hr after the commencement of the reaction a product could be detected on TLC and the stirring was stopped after another 7 hr. The reaction mixture was extracted with ether to remove unreacted starting material and the aqueous portion was transferred onto Sep-pak C-18 cartridges, washed with water and eluted with methanol/H20 (1:2). Following lyophilization of the eluate, the residue was analyzed by FAB/MS. FAB/MS diagnostic ions: m/z 448(M+H)+, 470(M+Na)+. G. Synthesis of the GSH conjugate of the methyl ester of (E)-2,4-diene  VPA This conjugate was synthesized essentially following the procedure described in section C, the main difference being that stirring was continued for 5 hr instead of 1. At that time 50% of the substrate was j converted to product as determined by HPLC. The crude product was purified by preparative HPLC using the solvent system 0.02 M NaH2P04-methanol (50:50) with which the conjugate eluted at 5.28 min. FAB/MS diagnostic ions: m/z 155(M-GS)+, 462(M+H)+, 484(M+Na)+. *H-NMR: 6 0.92(t,3H,CH3-), 1.24-1.36(m,2H,CH3-CH2-), 1.48-1.59(m,lH,-CH 2 -CH 2 -), 1.66-1.77(m,lH,-CH2-CH2-), 2.15-2.23(m,2H,Glu/?0'), 2.51-2.64 (m,2H,Glurr'), 2.72-2.81(m,lH,Cys/0, 2.96-3.03(m,lH,CysjS'), 3.10(q,lH,-CH=CH-CH-), 3.19(d,2H,-S-CH2-CH=CH-), 3.70(s,3H,-OCH3), 3.73(t,lH, Glua), 3.79(s,2H,Glya), 4.55(dd,lH,Cysa), 5.51-5.67(m,2H,-CH=CH-). 13C-NMR: S 14.1(CH3-), 21.1(CH3-CH2-), 27.7(-CH 2-CH 2-), 32.9(Glu£) , 34.1(Cys)9), 35.4(G1UT), 44.0(Glya), 49.8((-CH=CH-CH-), 52.7(Cysa), 54.0(Glua), 55.3(-0CH3), 129.7(-CH=CH-), 132.0(-CH=CH-), 172.4(Cys-C0-), 174.2(Glu-C0-), 175.3(Gly-C00H), 175.7(Glu-C00H), 176.9(-C00CH3). H. Synthesis of the dimethyl ester of the NAC conjugate of (E)-2,4-diene  VPA To a solution of methyl (E)-2,4-diene VPA (1.54 g, 10 mmol) in 50 mL p-dioxane, methyl NAC (1.77 g, 10 mmol) and triethylamine (5.6 mL, 40 mmol) were added and the mixture refluxed for 3 hr. Following the removal of the solvent the crude product was purified by preparative HPLC. With the mobile phase of 0.01 M NaH2P04-methanol (40:60) the conjugate eluted at 10.09 min. GC/MS: PCI: m/z 332(100%)(M+H)+, 360(24%)(M+C2H5)+, 372(12%)(M+C3H5)+. *H-NMR: 6 0.92(t,3H,CH3-), 1.27-1.38(m,2H,CH3-CH2-), 1.49-1.58(m,lH,-CH 2-CH 2-), 1.68-1.78(m,lH,-CH2-CH2-), 2.02(s,3H,-NHC0CH3), 2.70-2.79 (m,lH,Cys/3), 2.91-2.99(m,lH,Cys0'), 3.10(q,lH,-CH=CH-CH-), 3.17 (d,2H,-S-CH2-CH=CH-), 3.68(s,3H,-0CH3), 3.75(s,3H,-S-CH2-CH-C00CH3), 4.58(dd, lH,Cysa), 5.49-5.64(m,2H,-CH=CH-). 13C-NMR: S 14.1(CH3-), 21.3(CH3-CH2-), 22.4(-NHC0CH3), 32.5(-CH 2-CH 2-), 34.1(Cys/3), 35.6(-S-CH2-CH=CH-), 49.8(Cysa), 52.4(-0CH3), 53.0(-S-CH2-CH-C00CH3), 53.2(-CH=CH-CH-), 129.7(-CH=CH-), 132.5(-CH=CH-), 172.7(-NHC0CH3), 173.3(-C00CH3), 176.3(-S-CH2-CH-COOCH3). I. Synthesis of methvl 3-keto-4-pentenoate A solution of di isopropyl amine (12.5 mL, 0.09 mol) in THF was cooled to 0°C and n-butyllithium (56 mL, 0.09 mol) added in a dropwise fashion. The mixture was stirred for 20 min, cooled to -78°C and methyl acetate (7.1 mL, 0.08 mol) in THF was added dropwise. After stirring for 10 min acrolein (5.2 mL, 0.08 mol) was added and the mixture stirred for a further 60 min before being quenched with 15% HCl solution. Following extraction with ether, the extract was washed with NaHC03 and water. The ether was removed and the product purified by fractional distillation (bp 75°C/8 mm) to yield 5.2 g of ester (yield 50%). GC/MS: (MW=130) m/z 43(100%), 57(83%), 74(51%), 98(32%), 81(4%), 87(3%), 113(2%), 130(2%). 1HZNMR: 8 2.50-2.65(m,2H,-CH2-), 2.85-3.05(bs, 1H,-0H), 3.73(s,3H,-0CH3), 4.50-4.60(m,lH,-CH(OH)-), 5.15(dd,1H,CH2=), 5.35(dd,1H,CH2=), 5.85-5.95(m,lH,=CH). A 4.5 g portion (0.035 mol) was oxidized to the final product in a fashion similar to that described for the synthesis of methyl 3-keto-4-ene VPA in section B above. Pure methyl 3-keto-4-pentenoate was obtained in 41% yield by distillation under reduced pressure (70°C/23 mm). GC/MS: (MW=128) m/z 55(100%), 69(18%), 96(16%), 97(8%), 128(7%), 127(3%). JH-NMR: (keto and enol forms with a ratio of 2:1) (1). keto form; 8 3.63(s,2H,-CH2-C00CH3), 3.75(s,3H,-OCH3), 5.85-6.50(m,3H,CH2=CH-); enol form; 8 5.10(s,lH,=CHC00CH3). J . Synthesis of the thiol conjugates of methyl 3-keto-4-pentenoate Both GSH and NAC conjugates were synthesized in a manner similar to that described for the GSH conjugate of methyl 3-keto-4-ene VPA by mixing a methanolic solution of the substrate and aqueous solutions of the conjugating agents (pH of mixture=7.8) and stirring for 30 min. The NAC conjugate was also synthesized as the dimethyl ester by refiuxing methyl NAC with methyl 3-keto-4-pentenoate in dioxane. For the GSH conjugate the reaction mixture was lyophilized and characterized without further purification since it was deemed to be pure by HPLC. LC/MS: m/z 436(100%)(M+H)+, 309(51%)(GS+3H)+, 308(49%)(GS+2H)+, 404(12%)(M-0CH3)+. *H-NMR: 6 2.O7-2.2O(m,2H,Glu/S0'), 2.45-2.57(m,2H,Glury'), 2.75(d,2H,-S-CH 2-CH 2-CO-), 2.8O-2.89(m,lH,Cys0), 2.90-2.98(m,2H,-S-CH2-CH2-CO-), 3.07-3.13(m,lH,Cys0'), 3.68-3.83(m,6H,Glua,-0CH3,Glyaa'), 4.56(dd,lH, Cysa). 13C-NMR: 6 25.1(-S-CH2-CH2-CO-), 26.5(Glu0), 31.7(-S-CH2-CH2-CO-), 33.4(Cys/3), 43.6(Glya), 53.1(-0CH3), 53.3(Cysa), 54.4(Glua), 170.5(Cys-C0-), 172.2(Glu-CO-), 174.6(Gly-C00H), 175.2(Glu-C00H), 177.1(-COOCH3), 206.5(-CH2-CO-CH2-). The NAC conjugate of methyl 3-keto-4-pentenoate was purified by HPLC using 0.01 M NaH2P04-methanol (50:50) (retention time 3.72 min). GC/MS: (MW=305) (dimethyl ester) PCI: m/z 178(100%)(N-acetylcysteinyl moiety+2H)+, 248(48%)(M-NHC0CH3+H)+, 276(17%)(M-NHCOCH3+C2H5)+, 288(5%) (M-NHC0CH3+C3H5)+. 1H-NMR: (monomethyl ester) 6 2.00(s,3H,-NHCOCH3), 2.75(t,2H,-S-CH2-CH2-C0-), 2.78-2.87(m,lH,Cys0), 2.90(t,2H,-S-CH2-CH2-), 3.03-3.09(m,IH, Cys/3'), 3.73(s,3H,-0CH3), 4.40(dd, lH,Cysa). 13C-NMR: 6 22.8(-NHC0CH3), 26.5(-S-CH2-CH2-C0-), 35.8(-S-CH2-CH2-C0-), 42.9(Cys0), 52.8(Cysa), 55.6(-0CH3), 170.2(-NHC0Crl3), 172.9(-C00CH3), 177.2(-COOH), 204.1(-S-CH2-CH2-CO-). K. Synthesis of 2,4-pentadienoic acid, its methyl ester and the thiol  conjugates of the ester Acrolein (14.7 mL, 0.22 mol) was added to a solution of malonic acid (21.8 g, 0.2 mol) in pyridine (30 mL) containing 100 mg hydroquinone. The temperature was increased to 80°C and the mixture stirred for 3 hr. The reaction was quenched by pouring the mixture into ice cold water followed by dilute acid (Worley and Young, 1972). The product was isolated by ether extraction and purified by recrystallization from pet. ether (yield 57%, 100% E-isomer). GC/MS: (MW=98) m/z 98(100%), 97(72%), 53(71%), 42(47%), 70(35%), 81(21%). *H-NMR: S 5.55(dd, 1H,CH2=), 5.65(dd,1H,CH2=), 5.95(d,IH,J=17Hz,trans,-CH=CH-), 6.45-6.55(m,lH,CH2=CH-), 7.35-7.45(m,IH,-CH=CH-), 11.6-12.0 (bs,lH,-0H). The acid was converted to its methyl ester by refiuxing a mixture of acid (10 g, 0.1 mol), methanol (30 mL) and H2S04 (0.3 mL) for 2 hr. The mixture was treated with water and the product extracted with ether and dried (Na2S04). Pure ester was obtained by distillation at reduced pressure (37°C/10 mm, yield 42%). GC/MS: (MW=112) m/z 53(100%), 81(92%), 112(43%), 97(32%). 1H-NMR: 8 3.05(s,3H,-OCH3), 4.85(dd,1H,CH2=), 4.95(dd,1H,CH2=), 5.25(d,lH,J=17Hz,trans,-CH=CH-), 5.70-5.85(m,1H,CH2=CH-), 6.55-6.65 (m ,1H,-CH=CH-). To synthesize the GSH conjugate of the above ester, a methanolic solution of the ester (0.67 g, 6 mmol) and an aqueous solution of GSH (1.84 g, 6 mmol) (pH of mixture = 7.8) were stirred for 4 hr at which time 70% of the substrate was converted to product. Pure GSH conjugate was obtained by preparative HPLC (solvent system, 40% methanol, 60% 0.02 M NaH2P04; retention time 4.49 min). LC/MS: m/z 420(100%)(M+H)+, 147(25%)(ammoniated glutamyl residue), 129(20%)(glutamyl moiety). *H-NMR: 6 2.10-2.25(m,2H,Glu/3)3'), 2.50-2.65(m,2H,Glu77'), 2.70-2.8O(m,lH,Cys0), 2.95-3.O5(m,lH,Cys0'), 3.13(d,2H,-CH2-C00CH3), 3.17 (d,2H,-S-CH2-CH=CH-), 3.70(s,3H,-OCH3), 3.73(t,1H,G1 ua), 3.78 (s,2H, Glya), 4.55(dd,lH,Cysa), 5.52-5.62(m,lH,J=16Hz,-CH=CH-), 5.65-5.75 (m ,lH,J=16Hz,-CH=CH-). 13C-NMR: 8 27.7(Glu0), 32.8(Cys0), 34.1(Glu-y), 38.0(-CH2-C00CH3), 44.2 (Glya), 52.9(Cysa), 54.0(-0CH3), 55.3(Glua), 126.3(-CH=CH-), 131.1(-CH=CH-), 172.6(Cys-C0-), 174.4(Glu-C0-), 175.1(Gly-COOH), 175.4(Glu-C00H), 176.1(-C00CH3). The dimethyl ester of the NAC conjugate of (E)-2,4-pentadienoic acid was prepared by refiuxing methyl NAC with the substrate in a manner similar to that of the formation of the NAC conjugate of (E)-2,4-diene VPA. Pure conjugate was obtained by preparative HPLC using the solvent system of 0.01 M NaH2P04 a n d methanol (50:50) (retention time 10.55 min). GC/MS: (MW=289) PCI: m/z 290(100%)(M+H)+, 318(24%)(M+C2H5)+, 330(8%) (M+C3H5)4-. *H-NMR: 6 2.0(s,3H,-NHC0CH3), 2.68-2.78(m, lH,Cys/3), 2.90-3.00(m,IH ,Cys/S'), 3.12(d,2H,-CH2-C00CH3), 3.17(d,2H,-S-CH2-CH=CH-), 3.68(s,3H,-OCH3), 3.75(s,3H,-CH-C00CH3), 4.61(dd,lH,Cysa), 5.50-5.60 (m,lH,J=15.9 Hz,-CH=CH-), 5.65-5.75(m,lH,J=15.90Hz,-CH=CH-). 13C-NMR: 6 22.3(-NHC0CH3), 32.6(Cys0), 34.3(-CH2-C00CH3), 52.5(-CH-COOCH3), 52.9(Cysa), 53.3(-CH2-C00CH3), 126.6(-CH=CH-), 130.9(-CH=CH-), 172.7(-NHC0CH3), 173.2(-CH-C00CH3), 173.7(-CH2-C00CH3). L. Attempted synthesis of the CoA ester of (E)-2,4-diene VPA The anhydride of (E)-2,4-diene VPA was prepared by adding triethylamine (14 /iL, 100 /xmol) to a solution of (E)-2,4-diene VPA (14 mg, 100 /xmol) in THF (Goldman and Vagelos, 1961). After stirring the mixture for 15 min ethyl chloroformate (10 juL, 100 /imol) in THF was added and the reaction allowed to proceed for 30 min at 0°C. The precipitate formed was filtered off and the product analyzed by GC/MS. GC/MS: (MW=212) m/z 123(100%), 95(45%), 67(41%), 167(2%), 139(1%), 140(1%). CoASH (70 mg, 90 /xmol) was dissolved in 2 mL of THF-H20 mixture (7:3) and the pH adjusted to 8 with 1 M NaOH solution. The anhydride was dissolved in 2 mL THF and added dropwise to the CoASH solution under N 2 , while maintaining the pH of the reaction mixture at 8 by the addition of NaOH solution. After 30 min the pH of the mixture was adjusted to 5 with 1% HCIO4. Following lyophilization of the mixture, the product was analyzed by HPLC and LC/MS. M. Synthesis of 4'-keto-2-ene VPA This new VPA metabolite was prepared as the ethyl ester by oxidizing the terminal double bond of synthesized ethyl 2-(2'-propenyl)-2-pentenoate to the corresponding keto function. n-Butyllithium (28 mL, 0.045 mol) was added dropwise to diisopropylamine (6.2 mL, 0.045 mol) in THF (80 mL) at 0°C. The mixture was stirred for 15 min and cooled to -78°C, and ethyl 4-pentenoate (5.12 g, 0.04 mol) in THF was added dropwise over a period of 20 min. After stirring the reaction mixture for another 20 min, propionaldehyde (2.9 mL, 0.04 mol) was added and the reaction allowed to continue for 45 min before being quenched with 15% HCl solution. The product, ethyl 2-(l'-hydroxypropyl)-4-pentenoate, was isolated by extraction with ether and purified by fractional distil lation (bp 85°C/1.5 mm). The dehydration of the hydroxy ester was carried out using methanesulfonyl chloride and potassium hydride. Hydroxy ester (4.10 g, 0.022 mol), triethylamine (3.3 mL, 0.024 mol) and dichloromethane (60 mL) were cooled to 0°C and methanesulfonyl chloride (1.9 mL, 0.024 mol) was added dropwise and the mixture stirred for 60 min. The precipitate formed was filtered off and the solvent removed. The mesylate was then taken up in THF and potassium hydride (5.03 g, 0.044 mol) was added at 0°C and the reaction mixture stirred for 4 hr at room temperature. Following neutralization of the excess potassium hydride, the mixture was extracted with ether and pure ethyl 2-(2'-propenyl)-2-pentenoate was obtained by distil lation (bp 44°C/0.5 mm, yield 52%). The final product was obtained by palladium-catalyzed oxidation of the terminal double bond of ethyl 2-(2'-propenyl)-2-pentenoate (Rettenmeier et a7., 1986a; Tsuji, 1984). Palladium(II) chloride (21 mg, 0.12 mmol) and p-benzoquinone (0.71 g, 6.6 mmol) were added to a 7:1 mixture of dimethylformamide/water (16 mL) and then the diene (1.0 g, 6 mmol) was added over a period of 5 min. The reaction was allowed to proceed for 24 hr, at which time the reaction mixture was poured into cold 10% HCl (15 mL) and extracted with ether. The ether extract was washed with 10% NaOH and dried over Na2S04. Following the removal of the ether, the residue was distil led (bp 94°C/1 mm) to yield a mixture of E-and Z-isomers of ethyl 4'-keto-2-ene VPA (the E:Z ratio as determined by GC/MS was 95:5; yield 41%). The major isomer was assigned the E-configuration on the basis of stereoselectivity of the synthetic procedure (Lee et al., 1989). GC/MS: (MW=184) m/z 43(100%), 138(61%), 67(59%), 69(53%), 113(42%), 95(33%), 143(32%), 139(24%), 169(2%), 184(2%). *H-NMR: 5 1.07(t,3H,CH3-CH2-CH=), 1.29(t,3H,CH3-CH20-), 2.ll-2.25(m,2H, -CH2-CH=), 2.22(s,3H,CH3-C0-), 3.44(s,2H,-C0-CH2-), 4.29(q,2H,-0CH2-), 6.99(t,lH,J=6Hz,-CH=). N. Synthesis of 4'-keto-3-ene VPA This compound was synthesized in a manner similar to that described for 4'-keto-2-ene VPA by oxidizing the corresponding diene. To a solution of diisopropylamine (8.0 mL, 0.055 mol) at 0°C, n-butyllithium (34.0 mL, 0.055 mol) and HMPA (9.5 mL, 0.055 mol) were added successively and the mixture stirred for 20 min. The mixture was then cooled to -78°C and (E)-ethyl 2-pentenoate (6.4 g, 0.05 mol) added dropwise over a period of 20 min and stirring continued for 30 min. This was followed by the addition of allyl bromide (4.5 mL, 0.05 mol) and further stirring for 30 min. Finally the reaction mixture was quenched with 15% HCl solution and extracted with ether. Pure ethyl 2(2'-propenyl-3-pentenoate was obtained by distillation at reduced pressure (bp 60T/4 mm; yield 56%). The ratio of the Z:E isomers was determined by GC/MS to be 97:3, the major isomer assigned the Z-configuration on the basis of the known stereoselectivity of the synthetic route (Kende and Toder, 1982). A portion of the diene was oxidized to the keto function in the same way as that described for 4'-keto-2-ene VPA, to give 4'-keto-3-ene VPA (bp 98°C/2 mm; yield 38%). GC/MS: (MW=184) m/z 43(100%), 138(48%), 67(46%), 95(22%), 139(18%), 155(1%), 169(1%), 184(1%). *H-NMR: cS 1.23(d,3H,CH3-CH=), 1.26-1.28(t,3H,CH3-CH20-), 1.72(d,2H,-C0-CH 2 -), 2.17(s,3H,CH3-CO-), 3.00-3.07(m,IH,-CH=CH-CH-CH2-), 4.15(q,2H,-0CH 2-), 5.25-5.33(t,lH,J3)4=10Hz,CH3-CH=CH-,cis), 5.60-5.80(m,1H,CH3-CH=). 0. Synthesis of 4-ene VPA n-Butyllithium (94 mL, 0.15 mol), was added to a THF solution of diisopropylamine (21 mL, 0.15 mol) at 0°C and the mixture stirred for 20 min. The mixture was cooled to -78°C and HMPA (26 mL, 0.15 mol) and ethyl pentanoate (18.2 g, 0.14 mol) were added successively and stirring continued for 45 min. Allyl bromide (12 mL, 0.14 mol) was added and the reaction mixture stirred for 1.5 hr. Following the usual work-up and fractional distillation (bp 55'C/3 mm) 4-ene VPA ethyl ester was obtained in 61% yield. A portion of the ester was converted to the free acid by refiuxing the ester in two equivalents of aqueous NaOH. Distillation afforded pure 4-ene VPA (107°C/4.5 mm). GC/MS: (MW=142) m/z 57(100%), 56(83%), 85(64%), 41(47%), 100(7%), 142(6%), 113(3%). *H-NMR: 6 0.92(t,3H,CH3-), 1.28-1.70(m,4H,-CH2-CH2-) , 2.21-2.51(m,3H,-CH-CH2-CH=). 5.01-5.13(m,2H,CH2=), 5.72-5.85(m,lH,-CH=). P. Synthesis of 5-0H VPA This metabolite was synthesized by hydroboration of 4-ene VPA (Rettenmeier et al., 1985). To a stirred solution of ethyl 4-ene VPA (2.5 g, 0.015 mol)in THF (20 mL) at 0°C, borane (1 mL) was added dropwise and the mixture stirred for 30 min at room temperature. The mixture was cooled and excess hydride neutralized with water. Then 3 M NaOH (7 mL, 0.02 mol) and 30% H 2 O 2 (2.4 mL) were added dropwise at a rate that allowed for the temperature of the reaction mixture to remain below 350C. The mixture was then refluxed for 60 min. The product was isolated by ether extraction and purified by distillation (bp 88°C/0.1 mm; yield 32%). GC/MS: (MW=188) m/z 101(100%), 55(90%), 143(43%), 128(41%), 113(40%), 159(8%), 188(2%). Q. Synthesis of the dimethyl ester of the NAC conjugate of acrylic acid This adduct of acrylate was synthesized for use as an internal standard, following the same procedure for the synthesis of other NAC adducts described in preceding sections. GC/MS: (MW=263) m/z 43(100%), 172(51%), 204(47%), 55(46%), 88(43%), 190(12%). 5. Animal experiments A. In vivo metabolism studies A group of 3 rats was used for each treatment. For urine collections the animals were housed in stainless steel metabolic cages and were allowed free access to food and water. For the bile duct cannulation, each rat was anesthesized with urethane (1.1 g/kg) which was administered i .p . as an aqueous solution (0.3 g/mL). A midline incision was made and the portion of the bile duct at the junction with the duodenum located. A puncture was made into the duct and a cannula of PE-10 tubing inserted through the opening directed toward the liver. The cannula was ligated at the duodenal junction with silk thread and exteriorized through the lower abdominal wall with the help of a bore of an externally inserted 19G1 needle. After the surgical procedure, the incision was sutured and the animal placed in a metabolism cage. Control bile and urine were collected for 1 hr and following the administration of test compounds, bile and urine were collected for 8 and 24 hr, respectively. Samples were kept frozen at -20°C until analyzed. Each group of animals was given a 100 mg/kg i .p . dose of the sodium salts of the following compounds: VPA, 4-ene VPA, (E)-2,4-diene VPA (isomeric purity 94%), 4-pentenoic acid and (E)-2,4-pentadienoic acid (isomeric purity 100%). To prepare solutions of these compounds for administration to rats, 200 mg of each compound was weighed out and an equivalent amount of NaOH (3 M solution) added. When necessary, the pH of the solution was adjusted to 7.4 using 3 M HCl solution and the volume made up to 2 mL with water. To study the metabolic disposition of the GSH conjugates of (E)-2,4-diene- and 3-keto-4-ene-VPA, the conjugates (methyl ester form) were dissolved in normal saline and a dose of 40 mg/kg injected into a tail vein of each rat. B. Preparation of bile and urine for mass spectral analysis To identify GSH conjugates in bile, 40 mL of pooled bile (from each treatment group and control animals) was diluted with 40 mL of disti l led water and passed through a 2.5 x 20 cm Amberlite XAD-2 column. The bile was allowed to percolate slowly through the column packing. The column was then washed with 100 mL distil led water and eluted with 150 mL methanol. The eluate was evaporated to dryness and a portion of the residue dissolved in methanol. The solution was treated with methanolic hydrogen chloride gas, dried under the flow of N 2 , dissolved in methanol and treated with diazomethane generated from DiazaldR. The sample was then analyzed by HPLC, LC/MS and/or LC/MS/MS. An authentic sample of each conjugate (2-4 mg) was also methylated in the same way as the bile sample. For EI/GC/MS analysis, a 1 mL urine sample was taken from treated or control animals and acidified to pH 1.5-2.0 with 3 M HCl solution. The sample was saturated with 1 g NH4CI and extracted with ethyl acetate (3 mL) by gentle rotation of the phases for 30 min. The layers were separated by centrifugation, the organic portion dried over anhydrous Na2SC«4 and the volume reduced to 100 nl under N 2 . The concentrated extract was methylated with diazomethane. C. In vitro metabolism studies Animals were sacrificed by decapitation and the livers immediately removed. The liver was sheared with scissors and rinsed in 0.9% NaCl solution and homogenized in Tris-HCl (0.05 M, pH 7.5 containing 1.15% KCI) buffer with a Potter-Elvehjem tissue homogenizer. The homogenate was centrifuged at 600 g for 20 min and the pellet discarded. The supernatant was centrifuged at 7,500 g for 10 min, the pellet recovered and recentrifuged at 7,500 g for 10 min and the pellet (mitochondrial fraction) suspended in sucrose, homogenized and stored at -80°C until used. The supernatant from the 7,500 g spin was centifuged at 17,000 g for 10 min and the pellet discarded. The supernatant was further centrifuged at 100,000 g for 65 min and the resulting supernatant was saved as a source of cytosolic enzymes. The pellet was suspended in EDTA/KCI and centrifuged at 100,000 g for 70 min. The resulting pellet was resuspended in sucrose, homogenized, and saved as the microsomal fraction. Protein concentrations in all cellular fractions were determined by the method of Lowry et al. (1951) using bovine serum albumin as the standard. Biological synthesis of the CoA thio ester of (E)-2,4-diene VPA: Microsomal incubations of the substrate were carried out at 37°C for 0, 30, and 60 min in a total volume of 2 mL (Yamaoka et al., 1988). (E)-2,4-diene VPA (100 /xM) was incubated with the microsomal fraction (3 mg protein) suspended in 2.0 mL of 60 mM potassium phosphate buffer (pH=7.4) containing 1 mM CoASH, 5.0 mM ATP, and 5.0 mM MgCl2. After the indicated incubation times the reaction was stopped by adding 25 /xL acetic acid and 6 mL methanol. The mixture was centrifuged at 1000 g, the supernatant evaporated to dryness and the residue analyzed by LC/MS. 6. Human study Patients studied were regular visitors to the Seizure Clinic of British Columbia's Children's Hospital and included 26 receiving VPA only and 33 receiving VPA in combination with carbamazepine (CBZ) or clobazam (CLBZ). The ages of the patients ranged from 1.0 to 19.8 years. The mean VPA daily doses were 18.6 mg/kg for the monotherapy group and 32.9 mg/kg for the polytherapy group. VPA metabolite concentrations were determined in serum (trough) and overnight urine samples. One healthy male volunteer weighing 70 kg participated in the study in order to facilitate metabolite identification by the twin ion technique. The volunteer was given a 50:50 mixture of VPA and [2H6]VPA until steady state was reached i.e. 350 mg doses of each drug given twice daily for three days. 7. Quantitation of metabolites A. Quantitation of non-thiol metabolites Serum and urine standards: A reference standard was prepared by spiking VPA and the 14 metabolites mentioned below in either control serum or urine. Six standards including a blank sample were prepared for each run by taking aliquots of standard serum or urine and making up to a final volume of 250 /iL with control serum or urine. The concentrations for the serum standards making up the calibration curves were as follows: 24.8, 49.6, 74.4, 99.2 and 124 /ig/mL for VPA; 0.030, 0.300, 0.600, 0.900 and 1.20 itg/mL for 4-ene VPA; 0.394, 0.788, 1.18, 1.58 and 1.97 ng/ml for 3-ene VPA; 0.016, 0.162, 0.324, 0.486 and 0.648 /ig/mL for (Z)-2-ene VPA; 3.07, 6.16, 9.23, 12.3 and 15.4 ng/ml for (E)-2-ene VPA; 0.016, 0.160, 0.320, 0.480 and 0.640 /ig/mL for (E,Z)-2,3'-diene VPA; 3.10, 6.20, 9.30, 12.4 and 15.5 /xg/mL for (E,E)-2,3'-diene VPA; 0.204, 0.408, 0.612, 0.816 and 1.02 jig/mL for (E)-2,4-diene VPA; 0.380, 0.760, 1.14, 1.52 and 1.90 itg/mL for 3-OH VPA; 0.320, 0.640, 0.960, 1.28 and 1.60 /xg/mL for 5-OH VPA; 0.208, 0.416, 0.624, 0.832 and 1.04 /xg/ml_ for 4-keto VPA; 2.0, 4.0, 6.0, 8.0 and 10.0 /xg/mL for 3-keto VPA; 0.200, 0.400, 0.600, 0.800 and 1.0 /xg/mL for 2-PGA; and 0.020, 0.198, 0.396, 0.594 and 0.792 /xg/mL for 2-PSA. Stock solutions of the internal standards were prepared in disti l led water at a concentration of 50 /xg/mL for each of [2H3]2-ene VPA, [2H3].3-keto VPA and [2H6]VPA and 10 /xg/mL for 2-MGA. Equal volumes of each internal standard from the above solutions were mixed and a 100 xxL aliquot was added to each sample. The calibration curves were obtained by plotting the peak area ratio of VPA or metabolite to that of the internal standard versus the concentration of the compound by a computer program written in our laboratory and which also provided a reading of sample concentration. [2H6]VPA and [2H3]2 -ene VPA served as the internal standards for VPA and the unsaturated metabolites, respectively. [2H3]3-keto VPA was used to quantitate 3-keto VPA while 2-MGA served as the internal standard for the dicarboxylic acid metabolites. [2H3]2-ene VPA also served as the internal standard for the hydroxy metabolites and 4-keto VPA. Sample preparation: Depending on the amount of available serum, 50-250 /xL of serum sample were taken, internal standards added and the pH was adjusted to 2 with 3 M HCl solution. The sample was allowed to sit at room temperature for 10 min and extracted with 500 /xL of ethyl acetate by mechanical rotation of the phases for 20 min. To increase the recovery of the analytes the extraction step was repeated with another 500 /iL of ethyl acetate. After each extraction step the sample was centrifuged at 450 g for 10 min in order to break the emulsion formed. The combined organic layer was then transferred to a 3.5 mL screw cap vial containing anhydrous Na2S04, vortex mixed, and centrifuged at 450 g for 20 min. The supernatant was then transferred to another vial and the volume reduced to approximately 200 /iL under N 2 . To determine total urinary levels of VPA and metabolites, urine samples (50-250 /iL) were treated with 3 M NaOH (pH 12-13) and incubated at 60°C for 60 min before their pH was adjusted to 2 with 3 M HCl solution and extracted with ethyl acetate. For free urine concentration measurements, urine samples were treated as for serum except for a single extraction step. The concentration of metabolites as conjugates was detemined from the difference between total and free urine concentrations. To form the pentafluorobenzyl (PFB) derivative of the carboxyl groups the extract was transferred to a 1 mL conical reaction vial and 10 /iL of di isopropyl ethyl amine and 10 /iL of 30% PFBB solution in ethyl acetate were added and the sample heated in a heating block at 50°C for 60 min. The PFB esters were then treated with 50 /iL of MSTFA and heated at 60°C for 30 min to silylate metabolites containing the hydroxyl and 3-oxo moiety. A 1 /iL volume of the derivatized sample was injected into the GCMS system in the splitless mode using an automatic sample injector (Hewlett-Packard Model 7673A) under NICI conditions. B. Quantitation of the NAC conjugate of (E)-2,4-diene VPA Calibration curve: The dimethyl ester of the NAC conjugate of (E)-2,4-diene VPA was hydrolyzed by stirring the conjugate in two equivalents of NaOH at room temperature for 10 hr. An aliquot of the de-esterified conjugate was taken and made up to volume with water to obtain a stock solution of 5 mg/mL. 50 and 200 and /xL samples of the stock solution were then taken and the volume adjusted to 100 mL with control urine. An appropriate volume was taken from either of these standard solutions to yield standard concentrations of 0.05, 0.10, 0.50, 2.0, 4.0, and 6.0 ng/ml in a final volume of 1 mL. An internal standard, the NAC conjugate of acrylic acid, was prepared by hydrolyzing the methyl NAC derivative of methyl acrylate in aqueous NaOH in the same way as that of the dimethyl ester of (E)-2,4-diene VPA. An aliquot of the hydrolyzed sample was utilized to prepare a stock solution of 2.5 mg/mL, from which an amount was taken to give an internal standard concentration of 5 /xg/mL in water. Sample preparation: A 1 mL urine sample was taken from each of 28 pediatric patients on VPA therapy and 200 /xL of the internal standard solution added. The samples were acidified, saturated with NH4CI and extracted with ethyl acetate in a similar way as that described in an earlier section. To determine the amount of (E)-2,4-diene VPA dose recovered as the NAC conjugate, urine was obtained from three rats for 24 hr, the volume recorded, and a 200 /xL sample assayed for the NAC conjugate. To determine the extent of the reversal of the conjugation of (E)-2,4-diene VPA under alkali , samples were treated with 3 M NaOH to pH=12 and heated at 60°C for 60 min before the usual sample work-up. For derivative formation the concentrated extract was treated with 20 /iL of 20% PFBB in ethyl acetate and 10 /iL of di isopropyl ethyl amine and heated at 50°C for 60 min. A 1 /iL volume of the derivatized sample was injected into the GC/MS in NICI mode and selected ion monitoring (SIM) of the ions m/z 414 and 482. III. RESULTS 1. Chemical syntheses A. 3-Keto-4-ene VPA This putative metabolite of VPA was synthesized as the methyl ester by oxidizing methyl 3-0H-4-ene VPA according to Scheme 1. The product gave satisfactory GC/MS (Fig. 5) and ^-NMR (App. 1) data. Methyl 3-0H-4-ene VPA was first synthesized by alkylating the a-carbanion of methyl pentanoate, generated with lithium diisopropylamide (LDA) by acrolein, according to the general procedure for the alkylation of ester enolates (Cregge et al., 1973). An attempt to synthesize the t i t le compound by alkylating methyl pentanoate with acryloyl chloride in the presence of LDA resulted in a product showing addition of (i-Pr)2NH across the double bond of 3-keto-4-ene VPA. Similarly, an effort to synthesize the free acid form of 3-keto-4-ene VPA by oxidizing the free acid of 3-0H-4-ene VPA was not successful; the product obtained apparently being the decarboxylated 3-keto-4-ene VPA. A method previously reported for the synthesis of the ethyl ester of 3-keto-4-ene VPA (Rettenmeier et al., 1985) did not prove useful for preparing the free acid form of this /J-keto acid. B. GSH and NAC conjugates of methyl 3-keto-4-ene VPA The GSH adduct was synthesized by mixing a methanolic solution of the substrate with an aqueous solution of GSH and adjusting the pH of the mixture to 7.8 (Scheme 1). The reaction was instantaneous and the conversion quantitative; no substrate was detected 30 min after commencement of the reaction. Following purification by preparative 'COOMe (i-Pr)2NH, n-BuLi • THF. -78°C H2C=CHCHO COOMe (CH 3)3COOH Cr0 3 . CH 2CI 2 GSH MeOH/H 20 OH-COOMe 0 (C 2 H 5 ) 3 N Me-NAC A COOMe COOMe 5-(glutothion-S-yl)3-keto VPA methyl ester CH 2-CH-COOMe NHCOCH 3 5-(N-acetylcystein-S-yl)-3-keto VPA . dimethyl ester Scheme 1. Synthetic route for 3-keto-4-ene VPA methyl ester and its GSH and NAC conjugates. cn 67 A _ 2 9 f t 0 5 3 G5 8 2 9 6 1 4 5 153.87 1 8 1 i l • O i l 1. i i nl i • m i n • • i h n n r i i m 2 2 4 2 3 9 2 5 J 2 6 2 « l . . . . l . « . t l . i . . l i n i l i .,.,1 T o t a l I o n C h r o n a t o g r a n 5 5 2 7 18 S c a n U1Q3 4 1 C H o = C H - ( 55 C H - C I C 3 H 7 1 3 9 - X ' OCH3 0 II OH' I C H 2 = C H - C ^ C H = C - O C H 3 8 4 7 3 -I • 96 KJ9 • 1^. , I,, , 1 2 8 1 7 0 m/z Figure 5. EI/GC/MS TIC plot and spectrum of synthesized methyl 3-keto-4-ene VPA. HPLC, the structure of the adduct was determined to be 5-(glutathion-S-yl)-3-keto VPA by NMR and FAB/MS. The h-, 13C-NMR and FAB/MS spectra are shown in Figs. 6 , 7 and 8, respectively. The signals for the CH2=CH-protons in the *H-NMR spectrum of the substrate (App. 1) were replaced by signals for C H 2 - C H 2 - in the spectrum of the GSH conjugate (Fig. 6 ) . Because the methine proton (-COCH(COOCH3)C3H7) exchanged with deuterium of the NMR solvent (methanol-d4 and D2O) , there was no signal for this proton in the *H-NMR spectrum and there was only a weak signal for the -CH-(or -CD-) carbon in the 13C-NMR spectrum (Fig. 7 ) . Refiuxing the methyl ester of NAC with the methyl ester of 3-keto-4-ene VPA in the presence of two equivalents of (C2H5)3N afforded the NAC conjugate in a good yield. The NMR spectra (App. 2 and 3) were consistent with the structure 5-(N-acetylcystein-S-yl)-3-keto VPA. The NAC conjugate decomposed thermally in the GC/MS injector as determined by GC/MS and LC analysis of the same sample (Fig. 9). The decomposition of the conjugate was a reversal of the Michael addition (retro-Michael reaction) resulting in the substrate and methyl NAC. A similar decomposition has been reported for other mercapturates upon GC analysis (Onkenhout et a7., 1982). As the EI mass spectrum (Fig. 10A) did not furnish molecular weight information, for purposes of positive MS identification the conjugate was analyzed by PCI (Fig. 10B). C. GSH and NAC conjugates of (E)-2,4-diene VPA The synthetic pathway to these compounds is depicted in Scheme 2. The (E)-2,4-diene VPA was synthesized in a manner similar to a published procedure (Lee et al., 1989). However, another reagent (DBU) was also tried for the elimination of the mesylate. DBU elimination of the m c d e CH 2 CH 2 CH3 CH 2CH2C0CrJC00CH3 S-CH 2CHCONHCH 2COOH NH NHo 0=CCH 2 CH 2 CHCOOH 3 P P M Figure 6 . *H-NMR spectrum of synthetic GSH conjugate of methyl 3-keto-4-ene VPA. cn to 0 g h i C H 2 C H 2 C H 3 C H ? C H o C 0 C H C 0 0 C H 3 S - C H 0 C H C O N H C H 0 C O O H j ^ | | m -^ n NH fvJH2 0 = C C H 9 C H 9 C H C 0 0 H op ^ q ^ r s d CD3OD m ' I I ' ' I ' I ' I I I I I I I I I I I I I 1 , I I I I I I I I I I I I I I I I I I I I I I I I I •! I 1 I 1 I I I I I I I I I I I I I I 1 I I I I I I I 1 1 I I I 1 1 1 1 1 200 180 160 140 120 100 80 60 40 20 i 1 i I 1 i i 0 P P M Figure 7. 13C-NMR spectrum of synthetic GSH conjugate of methyl 3-keto-4-ene VPA. o ioo-/. _ 136 * 171 154 >-CD LU > CL" _ J LU or 107 12£ OS C 3 H 7 CH 2CH 2COCHCOOCH 3 f " " ¥|CH 2CHCONHCH 2COOH [0=CCH 2CH 2 xpHCOOH 1 3 0 \ 4 0 3 ^ 3 0 6 2 0 3 ' 203 403 308 246 274 * 289 331 387 415 478 (M + H) 462 500 (M + Na) + 100 200 300 m/z 400 500 Figure 8. Reconstructed FAB mass spectrum of the GSH conjugate of methyl 3-keto-4-ene VPA in 3-nitrobenzyl alcohol matrix. matrix ions 60090-55000-50000-45000-40000-35000->> 30000-4J •r-c 25000-= 20000-15000-10000-5000-0-4 .92 39 9? 7.75 i 6 78.87 r-j-. 10 21 .25 C 18.42 12 n r 1 14 1 i 1 i 16 18 T I ' Tv l 1 20 22 ' I ' I 24 26 28 TIME (min) Figure 9. EI/TIC chromatogram of an HPLC purified synthetic sample of the dimethyl ester of the NAC conjugate of 3-keto-4-ene VPA showing the decomposition of the compound upon GC/MS analysis. Peaks: a = methyl 3-keto-4-ene VPA, b = methyl NAC and c = methyl NAC conjugate of methyl 3-keto-4-ene VPA. The inset is the HPLC chromatogram obtained from the same sample. A 5 5 / gUllll Illlll 171 L C3H7 C H 2 C H 2 C O C H C O O C H 3 S - C H 2 C H C O O C H 3 N H C O C H 3 1 7 1 \ 1 3 9 8 8 / 7 3 / l i : \ •II. l l l i i l l , • 1 1 • 8 8 (nil 1 1 3 0 / III Ii 1 ' ' 1 2 0 11 1 ll ( M - C O O C H 3 ) 2 8 8 -1 1 0 -100 -90 >> a CD h-70 [ -60 cu > cu 1 • • • 1 • 1 6 0 1 7 7 1 9 6 2 2 4 III -I-J—1—1—1—1—1-2 0 0 2 3 8 / I 2 5 6 • 1 1 1 • 1 2 4 0 •30 2 0 1 0 •-0 2 8 0 m/z B 93H7 C H 2 C H 2 C O C H C O O C H 3 S - C H 2 C H C O O C H 3 f ! j H C O C H 3 173 (M + H) (M + C 2H 5) 41 1 7 1 144 \ (M + C 3 H 5 ) P98 H>8 >> F78 c cu +-> 3 8 ' F28 S 8 1 2 8 1 6 8 2 8 8 m/z 2 4 8 2 8 8 3 2 8 3 6 8 Figure 10. A) mass spectrum of peak c, Fig. 9, corresponding to the dimethyl ester of the NAC conjugate of 3-keto-4-ene VPA. B) PCI mass spectrum of the dimethyl ester of the NAC conjugate of 3-keto-4-ene VPA. 74 COOH COOH MeOH GSH 0.1 M NoH 2P0 4 pH 7.4 (7 h) (E)-5-(glutathion-S-yl)3-ene VPA (5%) COOMe I CH2-CH-COOMe I NHCOCH3 (E)-5-(N-acetylcystein-S-yl)3-ene VPA dimethyl ester Scheme 2. Outline of the synthetic route of the thiol conjugates of (E)-2,4-diene VPA and its methyl ester. mesylate resulted in a higher overall yield than did KH but gave a similar isomeric composition of the product. Column chromatography aided in the enrichment of the isomeric content of 2,4-diene VPA to 94% of the E isomer. Efforts to synthesize the GSH conjugate of the free acid of (E)-2,4-diene VPA in alcoholic aqueous solution (pH=7.8) were not successful. The use of a phosphate buffer (pH=7.4), however, resulted in the production of the desired product albeit in very low yields; 5% in 7 hr. On the other hand, the methyl ester of (E)-2,4-diene VPA readily reacted with GSH in methanolic aqueous solution (pH=7.8) with 50% of the substrate being converted to product within 5 hr. The product was purified by preparative HPLC and characterized by FAB/MS, *H- and 1 3 C -NMR. The FAB/MS spectrum showed structurally informative ions at m/z 484(M+Na)+, 462(M+H)+ and 155(M-GS)+ (Fig. 11). The most important structural evidence came from the NMR spectrum. The proton chemical shift data are presented in Table 1 and Fig. 12. Signals at 5.44, 5.56, 6.66 and 7.17 8 which were present for the substrate were absent in the GSH adduct while new signals at 3.10, and 5.51-5.67 8 appeared for the substrate portion of the adduct. The methylene protons (=C-CH.2CH2CH3(C00CH3)) which have a single resonance in the spectrum of 2,4-diene VPA (App. 4) are non-equivalent in the GSH adduct (Fig. 12) indicating that the methylene carbon might be adjacent to a chiral center in the latter. The *H- and 13C-NMR (Fig. 13) signals of the glutathionyl moiety of the conjugate were similar to literature values of GSH conjugates (Nocerini et al., 1985; te Koppele et al., 1986; Ishida et al., 1989; Nakagawa and Tayama, 1989). From these data plus the data of decoupling experiments the structure of the GSH conjugate of 1005! _ 136 LU LU > CC I LU CC OV. 176 C3H7 155L _ _CrJ2CH=CHCHCOOCH 3 T S - C H 2 C H C O N H C H 2 C O O H __N_H_j331 N H 2 6=C~CH2CH2CHCOOH 306 155 289 3D 8 333 374 462 (M + H) + 4 8 4 (M + Na) + 100 200 300 m/z Figure 11. Reconstructed FAB mass spectrum of the GSH conjugate of methyl (E)-2,4-diene VPA in 3-nitrobenzyl alcohol matrix. Some fragment ions have m/z values of 2 amu higher than shown on the structure as a result of double proton transfer. * matrix peaks —1 cn 77 7 8 9 C H 2 C H 2 C H 3 1 2 3 I C H 9 = C H C H = C C O O C H 3 * 4 5 6 J 7 8 9 C H 2 C H 2 C H 3 C H 2 C H = C H C H C O O C H 3 S - C H o C H C O O C H o 10^ I 1112 K T N H C O C H 3 14 1 5 J 7 8 9 C H 2 C H 2 C H 3 C H ? C H = C H C H C O O C H o I 4 5 6 J S - C H o C H C O N H C H o C O O H K T I 1112 13^14 NH I 0 = C C H 2 C H 2 C H C 0 0 H 15 16^ 17^ 18 19 TABLE 1 II III *H-NMR data of (E)-2,4-diene VPA methyl ester and i t s NAC (methyl ester) and GSH conjugates I 0.92 a(3H,t 6) 9 c 1.46 (2H,m) 8 2.40 (2H,t) 7d 3.76 (3H,s) 6 5.44 (lH.dd) l e 5.56 (lH.dd) l e 6.66 (lH.m) 2 7.17 (lH,d) 3 f II 0.92 (3H,t) 9 1.34 (2H,m) 8 1.54 (lH.m) 7 1.73 (lH.m) 7 2.02 (3H,s) 15 2.75 (lH.m) 10 2.94 (lH.ra) 10 3.10 (IH.q) 4 3.17 (2H,d) 1 3.68 (3H,s) 6 3.75 (3H,s) 13 4.58 (lH.dd?) 11 5.49- 5.64 (2H,m) 3,2 III 0.92 (3H,t) 9 1.31 (2H,m) 8 1.54 (lH.ra) 7 1.71 (lH.m) 7 2.17 (2H,m) 17 2.56 (2H,m) 16 2.76 (lH.ra) 10 3.01 (lH.ra) 10 3.10 (lH.q) 4 3.19 (2H,d) 1 3.70 (3H,s) 6 3.73 (lH.t) 18 3.79 (2H,s) 13 4.55 (lH,dd / ?) 11 5.51-5.67 (2H,m) 3,2 a Chemical s h i f t (ppm) E JBX=10 H Z , JAB=2 H Z ? J=4 and 8.0 Hz b M u l t i p l i c i t y JAX= 1 7 H Z , JAB=2 H Z h J=4.2 and 8.4 Hz c Assignment ^ J=12 Hz d Coupling constant, J=7 Hz 5 4 3 2 1 0 PPM Figure 12. H-NMR spectrum of the GSH conjugate of methyl (E)-2,4-diene CO 9 h i CH 2CH 2CH 3 CH2CH=CHCHCOOCH3 S-CH 2 C H C 0 N H C H 2 C 0 0 H j ^ I 1 m ^ n NH NH2 0 = C C H 2 C H 2 C H C 0 0 H op ' Q s ci ib m»in**i)»ni H C D 3 O D a,d m p q TMS i 180 160 1 — 140 120 100 PPM 80 60 40 20 t 0 Figure 13. 13C-NMR spectrum of the GSH conjugate of methyl (E)-2,4-diene VPA. 10 (E)-2,4-diene VPA was concluded to be 5-(glutathion-S-yl)-3-ene VPA. The 10C-NMR, shown in Fig. 13 also supported this structural assignment for the conjugate. The signals for the -CH=CH- carbons are present at 132.I and 129.8 5. The chemical shift for -CH-(C00CH.3)CH2CH2CH3 occurs at 49.0 6 as verified by the attached proton test. The methyl ester of (E)-2,4-diene VPA reacted rapidly with methyl NAC and yielded a single product when analyzed by LC and EI/GC/MS. The GC/MS spectrum of the conjugate is given in Fig. 14. Structurally informative ions are present at m/z 331 (M+), 300(M-OCH.3) + , 272(M-C00CH3)+, 176(SCH2CH(NHC0CH3)C00CH3)+, 155(M-(SCH2CH(NHC0CH3)C00CH3))+ and 144(CH2CH(NHC0CH3)C00CH3)+. The ]H-NMR and 13C-NMR (App. 5 and 6) data confirmed the position of the double bond to be at the same position as in the GSH conjugate i .e. at position 3. From these spectra and a series of homonuclear decoupling experiments the structure of the adduct was concluded to be 5-(N-acetylcystein-S-yl)-3-ene VPA. The quartet for -CH-(C00CH3)CH2CH2CH3 at 3.10 (Table 1) collapsed into a doublet and triplet respectively, upon decoupling of the signals at 1.73 and 5.55 6. Irradiation of the signals at 3.10 and 3.17 S resulted in the appearance of doublets on the downfield (3.10) and upfield (3.17) side of the 5.49-5.64 6 signal assigned to the -CH=CH- protons. The coupling constant of the vinylic protons was 16 Hz indicating a trans configuration for the double bond (Silverstein et a/ . , 1981). A similar coupling constant value was obtained for the vinylic protons in the GSH conjugate. 81 43 95 1.1 .Illlll] 80 134 144 II, CoH 3 n 7 155 [; CH2CH=CHCHCOOCH3 r~S-CHoCHCOOCHo 176 176 155 JL N H C O C H 3 L , I„|J 212 240 . 1 261 \ 272 / M 331 + = 300 pi 10 rioo >> •»-> ; *i— J80 ns - cu r-7 0 +-> ; c 1—< Z cu > r*0 at - 1 — r3 0 cu ;: ia 120 160 200 240 280 320 m/z B 95 176 71 U 134 144 117 illliul il. 155 p 3 H 7 CH2CH=CHCHCOOCH3 jf S-CH 2CHCOOCH 3 NHCOCH3 2 * 2 240 272 MM 331 JL. .,L.lJ.. ... .1. I. c-110 ^.00 r90 +J 300 r8a r7 0 ^ 0 :S0 r3 0 E20 r!0 0) c cu > 4-> cu 80 120 160 200 240 320 m/z Figure 14. EI mass spectra of A) authentic dimethyl ester of the NAC conjugate of (E)-2,4-diene VPA and B) peak marked with an asterisk, Fig. 29). D. Methyl 3-oxo-4-pentenoate This a,^-unsaturated keto ester is a known compound and has previously been synthesized as the ethyl ester by a synthetic route that involved several steps which were tedious and inefficient (Collins and Tomkins, 1977). In the present study the methyl ester was synthesized by an efficient, two-step synthetic route as shown in Scheme 3. The spectroscopic data of the compound was in agreement with that in the literature (Collins and Tomkins, 1977) with the *H-NMR showing the keto and enol forms of the compound in a ratio of 2:1 (App. 12). The GC/MS spectrum of the compound is shown in Fig. 15 and is consistent with the structure of the desired product. E. GSH and NAC conjugates of 3-oxo-4-pentenoate The 3-oxo-4-pentenoate reacted readily with GSH in alkaline aqueous/alcohol solution (Scheme 3). The structure of the product was determined to be 5-(glutathion-S-yl)-3-oxopentanoate using LC/MS (Fig. 16) and *H- and *3C-NMR (Figs. 17 and 18) spectroscopy. As was the case with the GSH conjugate of 3-keto-4-ene VPA, there was no signal for the methylene protons adjacent to the two carbonyl groups in the GSH conjugate of 3-oxo-4-pentenoate. The NAC conjugate was also synthesized in a manner similar to that of the GSH conjugate by stirring NAC and the substrate in aqueous methanolic solution. The NMR spectra of the purified product are shown in App. 13 and 14. For GC/MS analysis an attempt was made to methylate the free carboxyl group; however, upon treatment with methanolic HCl gas followed by CH2N2 the NAC conjugate was deacetylated to give the cysteine conjugate. The NAC conjugate was also synthesized as the CH 3COOMe (i-Pr) 2NH, n-BuLi • THF, - 7 8 ° C H 2 C=CHCHO (CH 5 ) 3 COOH^ COOMe ^ ~ K OH C r 0 3 . C H 2 C I 2 GSH M e O H / H 2 0 OH-COOMe NAC MeOH/H 20 OH-COOMe COOMe 5-(glutothion-S-yl)3-oxopentanoote C H 2 - C H - C O O H N H C O C H 3 5-(N-acetylcystein-S-yl)3-oxo-M e O H - H C l / pentanoate > ^ C H 2 N 2 f ^ j ^ C O O M e ? 0 le CH 2-CH-COOMe NH 2 5-(cystein-S-yl)3-oxopentanoate Scheme 3. Synthetic pathway of methyl 3-oxo-4-pentenoate and its thiol conjugates. CO H O T o t a l I o n C h r o n a t o g r a n Scan HJLJO m/z Figure 15. EI/GS/MS TIC chromatogram and spectrum of a synthetic sample of methyl 3-oxo-4-pentenoate. 85 100T 4 0 4 -n 3 0 6 75-CH 2 CH 2 COCH 2 COiOCH3 _| i S-CH 2 CHCONHCH 2 COOH rjJH fsjH2 0=CCH 2 CH 2 CHCOOH 309 •g SO-cr 25-HSCHoCHCONHCHoCOOH Z \ + NHo 179 o+" 150 404 200 250 300 436 , + (M + H) 350 m/z 400 450 500 550 Figure 16. LC/MS spectrum, in the positive ion mode, of synthesized GSH conjugate of methyl 3-oxo-4-pentenoate. 86 C H 2 C H 2 C O C H 2 C O O C H 3 Figure 17. JH-NMR spectrum of the GSH conjugate of methyl 3-oxo-4-pentenoate. a b o d e f C H 2 C H 2 C O C H 2 C O O C H 3 g h i J S - C H 2 C H C O N H C H 2 C O O H f^H fsjH2 0 = C C H 9 C H 9 C H C 0 0 H I m ^ n ^ o D | 1 1 ' 1 I 1 1 1 1 | ' ' I ' I 1 1 | 1 ' 1 1 I 1 1 1 1 | 1 1 1 ' M ' I I | I I I I | I I I I | I I M | I I I I | I I I 'I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I 1 I I I 1 I I 1 I I I I I i I I I I I I l '| I I [ I I , I I I I 240 220 200 180 160 140 120 100 80 60 40 20 0 PPM Figure 18. 13C-NMR spectrum of the GSH conjugate of methyl 3-oxo-4-pentenoate. co dimethyl ester starting with methyl NAC. Compared to other conjugates studied, the dimethyl ester of the NAC conjugate of 3-oxo-4-pentenoate underwent extensive MS fragmenation upon GC/MS analysis (Fig.19A). Even the PCI spectrum did not furnish molecular weight information (Fig. 19B). Satisfactory mass spectral identification was obtained only using LC/MS/MS (Fig. 20). F. GSH and NAC conjugates of (E)-2,4-pentadienoate Methyl 2,4-pentadienoate was synthesized stereospecifically (100% E isomer) following a literature method (Worley and Young, 1972). The ester reacted readily with GSH according to Scheme 4 with 70% of the substrate converted to product in 4 hr. The LC/MS, *H- and 13C-NMR data of the HPLC purified product are illustrated in Figs. 21, 22 and 23, respectively. From these spectroscopic data and a series of homonuclear decoupling experiments, the structure of the conjugate was established to be (E)-5-(glutathion-S-yl)-3-pentenoate. Thus, this analog of (E)-2,4-diene VPA reacted with GSH by the same mechanism as that of (E)-2,4-diene VPA. Unlike that of (E)-2,4-diene VPA, however, the reaction of (E)-2,4-pentadienoate produced two isomers detectable by LC/MS (Fig. 24) and having the same MS/MS mass spectra (Fig. 25). The dimethyl ester of the NAC conjugate of (E)-2,4-pentadienoate was prepared in a manner similar to that described for the NAC conjugate of (E)-2,4-diene VPA methyl ester. The mass chromatogram (m/z 176) obtained from a crude sample is illustrated in Fig. 26A. There are four isomeric peaks with the same mass spectra. Figure 26B shows the mass spectrum of the major peak and contains several ions which are informative of the structure- m/z 289(M+), 258(M-0CH3)+, 230(M-COOCH3)+, 89 6 0 II 1 C H 2 C H 2 C O C H 2 C O O C H 3 S-CH 2 CHCOOCH 3 (M-NHCOCH3-COOCH3) + N H C O C H 3 1 8 8 / 8 8 7 1 Illl. 1 1 7 1 1 3 ll ll 1 5 6 1 2 8 1 7 0 \ 1 7 7 •t—1—.—f—>—1—1—1—1—1—' 1 6 0 1 8 0 2 0 0 2 2 9 : r l 1 0 L 1 0 0 ^ 9 0 >1 +-> - •r~ C O c CD ~ -t-> C : 6 0 •—1 cu > •r— - +J ra ^ 4 0 1 — cu \ cc r 3 0 720 r l 0 -I—•—I— 6 0 8 8 1 0 0 -1—'—1— 1 —r 1 2 0 1 — 1 — ' — r 1 4 0 m/z 220 B 1 7 8 C H 2 C H 2 C 0 C H 2 C 0 0 C H 3 r " S ~ - C H 2 C H C 0 0 C H 3  1 7 6 N H C 0 C H 3 7 1 1 0 2 \ 1 4 4 II 1 1 2 / 1 3 6 \ 1 6 0 / (M-NHCOCH3 + H) + 110 100 H98 (M-NHCOCH3 + C 2 H 5 ) + 1 8 8 2 0 6 • 1 ' • 8 0 1 2 0 1 6 0 m/z 200 2 4 8 288 788 >> - 7 0 4-> •r— co _ c cu -t-> c t—* r 5 0 cu - > - •r-J 4 0 4-> CTJ 1 — \ CD al - 3 0 : r 2 0 J 1 0 '-0 2 8 0 Figure 19. EI (A) and PCI (B) mass spectra of the NAC conjugate of 3-oxo-4-pentenoate, dimethyl ester. 90 100! 75-in c j | 50 <D s a> CC 25-178 129 119 100 161 144 ill 169 186 232-129 246 I i I i L CH 2 CH 2 COJCH2]COOCH3 f S i f C H 2 C H C O O C H 3 1 7 6 " AHCOCH3 1 6 1 — J 1 — 1 4 4 306 232 246 i | < l L , H , . . ( . . . i | i (M + H) + 150 200 250 m/z 300 350 400 Figure 20. LC/MS/MS of the (M+H)+ of the dimethyl ester of the NAC conjugate of 3-oxo-4-pentenoate. 91 H 9 C = C H C H 0 MeOH C H 2 ( C O O H ) 2 • • H + , A GSH MeOH/H 2 C> ^COOMe OH-Me-NAC ( C 2 H 5 ) 5 N A COOMe s ( E )-5- (g lu ta th ion-S-y l )3-pentenoate COOMe C H 2 - C H - C O O M e I N H C O C H 3 ( E )-5-(N-ocety lcys te in-S-y l )3-pentenoate Scheme 4. Synthesis of methyl (E)-2,4-pentadienoate and its GSH and NAC adducts. 92 100! 420 (M + H) + 75J 1 50H C H 2 C H = C H C H 2 C O O C H 3 S - C H 2 C H C O N H C H 2 C O O H f^H fvJH2 0 = C C H 2 C H 2 C H C O O H 25) 147 129 100 150 291 256 274 200 250 300 m/z 376 402 § 442 308 357 I i S l 350 400 450 500 Figure 21. LC/MS mass spectrum of the GSH conjugate of methyl (E)-2,4-pentadienoate. 93 —i— . — | —- i— i—.—i—.—i—.—i— i— i—•—i—'—i—•—r—•—i—•—i—•—i—•—r—-•—i— r — i—•—i • i 1 i • r-S.8 S.G 5.4 5.2 5.0 4.8 4.6 4. 4:- 4.2 4;B 3.8 3.G 3.4 3.2 3.0 ,2.8 2.G 2.4 2. 2 2.0 PPM Figure 22. AH-NMR spectrum of the GSH conjugate of methyl (E)-2,4-pentadienoate. a b c d e f CH2CH=CHCH2COOCH3 g j S-CH2CHCONHCH2COOH ^ fjJH2 0=CCHoCH?CHCOOH I m ^ n ^ o p c b C D 3 O D m: l i 1 1 I ' ' ' ' I i ' ' 1 1 1 1 1 1 1 1 M 1 i | 1 1 1 1 1 1 1 11 1 1 1 n - p I 1 1 1 I I I I 1 I I I I I I I r-l f I I' M J T,.| I I I I I I I j M I I I I II I I I I I I 1 I I I I I I I I I I I M I M I I I 1 I I I I I I I I I 1 I M I I 2 4 0 2 2 0 200 180 160 140 120 100 80 6 0 4 0 20 PPM 0 Figure 23. 13C-NMR spectrum of the GSH conjugate of methyl (E)-2,4-pentadienoate. 95 Figure 24. LC/MS profile of a fully methylated sample of the GSH conjugate of (E)-2,4-pentadienoate. Peaks 1 and 2 were identified as isomeric GSH conjugates of (E)-2,4-pentadienoate. 75H £ 50H 25-j 100, 113 144 176 CH2CH = CHCH2COOCH3 2[ 8 S-CH2CHCONHCH2COOCH3 ljir7 ,NH2 T 0 = CCH9CH2!CHC00CH3 144 | I I 359—' 188 448 305 288 100 200 300 359 (M + H) + 400 m/z 176 75 c SO rx 25 144 113 188 CH 2 CH=CHCH 2 COOCH 3 S-CH 2 CHCONHCH 2 COOCH 3 NH |SIH2 0 = C C H 2 C H 2 C H C O O C H 3 (M + H) + \ . 4 4 9 305 100 200 m/z 288 359 300 —I— 400 Figure 25. Daughter ion (parent, m/z 448) MS/MS spectra of peaks 1 (top) and 2 (bottom) in Fig. 24 corresponding to isomeric GSH conjugates of (E)-2,4-pentadienoate. 97 cu 32900-1 28000-24000-20000 16000-12000-8000-4000" A m/z 176 • • • I • • • • I • • • • I • • • • • • • ' ) » • • • I • I r i " ' | i • • • • | 14.0 15.0 16.0 17.0 18.0 19.0 20.0 21.0 22.0 23.0 24.0 B TIME (min) 176 71 43 If i ..iJ 88 ll. 117 113 134 \ 80 I. Illi. 144 / 166 ll.. ll 230 258 —I —I I I I I [3 CH 2 CH=CHCH 2 5 C 0 !0CH3 J~~ S ^ C H 2 C H C O O C H 3 | N H C O C H 3 1—144 198 230 / \ 120 ' ' 1 160 1,1. i i 258 / M 1" 289 \ 110 100 3^0 co 70 £= cu +-> t60 ~ cu ^0 > ra b40 •— cu 01 F/30^ •20 -10 200 240 280 m/z Figure 26. A) Mass chromatogram of m/z 176 obtained from a crude synthetic sample of the NAC conjugate of (E)-2,4-pentadienoate (dimethyl ester) and B) EI/GC/MS spectrum of the peak labelled 4. 176(SCH2CH(NHCOCH3)COOCH3)+ and 113(M-(SCH2CH(NHC0CH3)C00CH3))+. From this spectrum and *H-NMR (App. 9, obtained from an HPLC purified sample) and 10C-NMR (App. 10) plus decoupling experiments the structure of the conjugate was determined to be 5-(N-acetylcystein-S-yl)-3-pentenoate. The coupling constant between the vinylic protons was 15.90 Hz indicating a trans configuration for the double bond. 2. Development of an analytical method for profiling VPA metabolism in  patients A sensitive and selective GC/MS assay was required for the simultaneous quantitation of VPA and 14 of its metabolites in order to assess the 4-ene VPA pathway relative to other VPA metabolic pathways in pediatric patients. In a previous communication (Kassahun et a7., 1989), the use of NICI/GC/MS for the identification of VPA metabolites as their pentafluorobenzyl (PFB) derivatives was described. In the present study the NICI methodology was used to determine the levels of the putative hepatotoxic metabolites of VPA along with the other known metabolites. A. Negative ion GC/MS properties of VPA metabolites The PFB derivatives of the metabolites and the parent drug have excellent negative ion response as reported earlier (Kassahun et a7., 1989). With the DB-1 column, however, the PFB derivatives of the hydroxy metabolites chromatographed poorly and sensitivity was compromised. Trimethylsilylation of the hydroxyl groups after PFB derivatization of the carboxyl moiety greatly improved both peak shape and sensitivity of detection. The PFB derivative of 3-keto VPA, unlike the other metabolites, fragmented to yield the [M-I8I-CO2]" (M-181 corresponds to the loss of the pentafluorobenzyl moiety) ion as the base peak (Kassahun et al., 1989). This fragmentation was blocked with TMS derivatization of the 3-oxo moiety, resulting in the [M-181]" ion as the major ion in the mass spectrum. Hence, SIM chromatograms were obtained by monitoring the [M-181]" anions which were the base peaks of all the metabolites studied i .e. both PFB esters and O-TMS-PFB derivatives. The metabolites assayed and the negative ions monitored are given in Table 2. The negative ion spectra obtained for VPA and its metabolites are typified by the example in Fig. 27, with the [M-181]" as almost the only ion in the mass spectrum. Thus, the combination of PFB and TMS derivatization produced suitable negative ions and chromatographic characteristics which allowed the development of a simultaneous assay of VPA and 14 metabolites. Typical SIM chromatograms representing all the metabolites assayed are shown in Figs. 28A and 28B. The use of the DB-1 column and PFB derivatization enabled the complete separation of 4-ene VPA from VPA as well as the separation of the monounsaturated metabolites as seen in Fig. 28A, m/z 141. PFB derivatization also resulted in baseline separation of the geometric and/or positional isomers of the diunsaturated metabolites (Fig. 28A, m/z 139). The putative toxic metabolite, (E)-2,4-diene VPA is clearly separated from (E,Z)-2,3'-diene VPA. When the t-BDMS and TMS derivatives were analyzed under similar conditions, these two metabolites overlapped and being isobaric were counted as one. In the case of 4-OH VPA only the open chain form was monitored since the 7-lactone would be transparent to NICI. Because the base diisopropylethyl amine was used for the derivatization reaction, it was assumed that most of the 4-OH VPA would be in the open chain form. Table 2. Ions monitored (m/z) in the NICI mode for VPA and metabolites derivatized with PFB and TMS COMPOUND [M-181]" (E,Z)-2,3'-Diene VPA 139 (E)-2,4-Diene VPA N 139 (E,E)-2,3'-Diene VPA 139 4-Ene VPA 141 3-Ene VPA 141 (Z)-2-Ene VPA 141 (E)-2-Ene VPA 141 VPA 143 4-Keto VPA 157 3-Keto VPAa 229 4-OH VPA a ' b 231 3-OH VPAa 231 5-OH VPAa 231 2-PSAc 339 2-PGAc 353 a O-TMS-PFB derivatives b two isomers c di-PFB derivatives 101 100% c n U J U J cr. OX .J L 100 229 200 229 M/Z 300 400 Figure 27. NICI/GC/MS spectrum of the O-TMS-PFB derivative of 3-keto VPA. m/z 139 m/z 141 m/z 143 m/z 144 m/z 149 m/z 157 ll L 13 TIME (min) Figure 28A. Typical SIM chromatograms obtained from the serum extract of a patient on VPA to which internal standards have been added. Peaks: m/z 139, 1 = (E,Z)-2,3'-diene VPA, 2 = (E)-2,4-diene VPA, 3 = (E,E)-2,3'-diene VPA; m/z 141, 1 = 4-ene VPA, 2 = 3-ene VPA„ 3 = (Z)-2-ene VPA, 4 = (E)-2-ene VPA; m/z 143, MPA; m/z 144, 1 = (Z)-[2H3]2-ene VPA, 2 = (E)-[%]2-ene VPA, m/z 149, [^H6]VPA; m/z 157, 1 = 4-keto VPA. 103 m/z 232 u m/z 325 11 13 15 17 19 21 1 1 m/z 339 18 20 22 m/z 353 20 22 24 TIME (min) Figure 28B. Typical SIM chromatograms obtained from the serum extract of a patient on VPA to which internal standards have been added (same run as in Fig. 28A). Peaks: m/z 229, 1 = 3-keto VPAj m/z 331, 1 & 3 = 4-OH VPA, 2 = 3-OH VPA, 4 = 5-0H VPA; m/z 232, 1 = [%]• 1 = 2-MGA; m/z 339, 1 = 2-PSA; m/z 353, 1 = 2-PGA. |3-keto VPA; m/z 325, Greater than 90% of the 3-keto VPA formed a diderivative i.e. 0-TMS-PFB. Formation of a small amount of 3-keto VPA monoderivative did not present a problem because a deuterated analog was used as the internal standard. Even though the two geometric isomers of the 0-TMS-PFB derivative of 3-keto VPA were formed, only one isomer, which accounted for 93% of the ion current, was monitored. B. Analytical parameters Calibration curve linearity was observed over the concentration ranges tested. Non-linear response was evident in the case of VPA when serum volumes of greater than 250 /xL were used. Correlation coefficients were typically greater than 0.99 for all standard curves. The method showed good precision with a day to day variation of less than 15% for all 15 analytes. The day to day precision and accuracy of the assay for individual metabolites is shown in Table 3. The precision of the assay for analytes having deuterated analogs as internal standards (VPA, 2-ene-, 3-keto-VPA) was especially high, although this could be attributable to the higher levels measured for these metabolites. For 4-0H VPA it was necessary to maintain the extraction pH fairly constant in order to obtain reproducible results. A deuterated internal standard for 4-OH VPA would eliminate this problem. Trace metabolites such as 4-ene- and (E,Z)-2,3'-diene-VPA were readily quantitated from 50 /xL of serum. The lower limit of detection of the metabolites .(signal to noise ratio > 3) varied between 1 to 8 ng/mL based on a 250 /LtL sample of serum. Table 3. Day to day precision and accuracy data for VPA and metabolites in serum (n = 5) Compound Spiked Found C V . (#g/mL) (/*gM) (%) 4-Ene VPA 0.30 0.31 6.2 1.20 1.24 1.4 3-Ene VPA 0.39 0.38 10.2 1.97 2.04 2.2 (Z)-2-Ene VPA 0.16 0.15 6.0 0.65 0.67 1.1 (E)-2-Ene VPA 3.07 3.06 2.7 15.39 15.52 0.8 (E,E)-2,3'-Diene VPA 3.10 3.20 6.2 15.50 16.26 1.8 (E,Z)-2,3'-Diene VPA 0.18 0.21 11.0 0.64 0.62 2.0 (E)-2,4-Diene VPA 0.20 0.19 12.3 1.02 1.08 3.7 VPA 24.88 27.70 2.0 124.40 125.50 1.5 4-OH VPA 0.38 0.37 11.8 1.90 2.01 1.3 3-OH VPA 0.19 0.22 6.1 0.98 0.98 3.6 Table 3. (continued). Day to day precision and accuracy data for VPA and metabolites in serum Compound Spiked Found CV (/xg/mL) (Aig/mL) (%) 5-OH VPA 0.80 0.95 13.8 1.60 1.50 6.1 4-Keto VPA 0.21 0.23 14.6 1.04 1.04 3.9 3-Keto VPA 2.00 2.5 2.8 10.00 9.72 1.3 2-PSA 0.19 0.17 15.6 0.60 0.58 8.5 2-PGA 0.20 0.20 15.3 1.00 1.05 2.4 107 C. VPA metabolite profiles in pediatric patients The assay was employed to determine the levels of VPA and 14 of its metabolites in the serum and urine of 59 pediatric patients. The serum concentrations for 26 patients on VPA monotherapy are shown in Table 4. The major metabolites in serum were (E)-2-ene-, (E,E)-2,3'-diene- and 3-keto-VPA. The putative toxic metabolite, 4-ene VPA, was present at low levels (mean = 0.20 itg/mL). The mean concentration of the diene metabolite ((E)-2,4-diene VPA) suspected to be a hepatotoxin was 0.65 /xg/mL. The concentrations of metabolites (normalized to dose) in the patient groups on combined therapy with other antiepileptic drugs is shown in Table 5. The serum levels of 4-keto- and 5-0H-VPA were significantly (P<0.05) higher in the VPA + CBZ group compared to the monotherapy group. Also, the serum concentrations of VPA and its oxidation products were lower in the VPA + CBZ group. CLBZ had no significant effect on the VPA metabolite profile. In urine the major metabolites were 3-keto-, 4-0H-VPA and 2-PGA (Table 6). The unsaturated metabolites were excreted mainly as conjugates, while the keto metabolites were predominantly excreted in the free form. The hydroxy metabolites and 2-PGA were found in urine in both the conjugated and free forms. The levels of 4-ene- and (E)-2,4-diene-VPA were determined in the urine of two patients who had developed VPA-induced hepatotoxicity and were found not to be different from the levels in other patients with comparable VPA urinary concentrations. Table 4. Serum concentrations (/xg/mL) of VPA metabolites in 26 pediatric patients on VPA monotherapy Compound Mean Range % of VPA 4-Ene VPA 0.20 0.04-0.40 0.2 3-Ene VPA 1.80 0.72-3.30 2.0 (Z)-2-Ene VPA 0.11 0.05-0.18 0.1 (E)-2-Ene VPA 6.44 2.51-10.30 7.2 (E,E)-2,3'-Diene VPA 6.31 2.47-13.50 7.1 (E,Z)-2,3'-Diene VPA 0.28 0.06-0.38 0.3 (E)-2,4-Diene VPA 0.65 0.15-1.16 0.7 VPA 89.2 39.8-134.0 100 4-OH VPA 0.71 0.14-1.64 0.8 3-OH VPA 0.97 0.10-2.27 1.1 5-0H VPA 0.45 0.13-0.96 0.5 4-Keto VPA 0.60 0.16-1.27 0.7 3-Keto VPA 3.63 0.78-8.24 4.1 2-PSA 0.02 Tr.-0.15 0.02 2-PGA 0.16 0.02-0.30 0.2 Table 5. Mean concentrations normalized to dose (/xg/mL/mg/kg) of VPA and metabolites in the serum of patients Compound VPA only n = 26 VPA + CBZ n = 21 VPA + CLBZ n = 12 4-Ene VPA 0.009 0.007 0.008 3-Ene VPA 0.109* 0.042 0.086 (Z)-2-Ene VPA 0.005* 0.002 0.005 (E)-2-Ene VPA 0.311* 0.136 0.274 (E,E)-2,3'-Diene VPA 0.304* 0.129 0.201 (E,Z)-2,3'-Diene VPA 0.012* 0.004 0.007 (E)-2,4-Diene VPA 0.035 0.029 0.029 VPA 4.88* 2.31 3.82 4-OH VPA 0.053 0.051 0.046 3-OH VPA 0.064 0.049 0.038 5-0H VPA 0.022 0.031* 0.028 4-Keto VPA 0.030 0.049* 0.026 3-Keto VPA 0.311* 0.126 0.102 2-PSA 0.005 0.006 0.004 2-PGA 0.010 0.007 0.006 * Significantly different at p<0.05 (VPA only vs. VPA + CBZ) Table 6. Urinary concentrations (/xg/mg creatinine) of the conjugated and unconjugated fractions of VPA metabolites in 9 pediatric patients on VPA monotherapy Compound Mean (Range)3 Mean (Range)b 4-Ene VPA 0.25 (0.02-1.53) 0.03 (Tr-0.17) 3-Ene VPA 0.21 (0.03-0.93) 0.05 (Tr-0.28) (Z)-2-Ene VPA 0.94 (0.10-2.77) 0.03 (Tr-0.22) (E)-2-Ene VPA 21.5 (1.0-54.3) 0.92 (Tr-5.73) (E,E)-2,3'-Diene VPA 17.2 (3.85-56.9) 1.12 (Tr-6.12) (E,Z)-2,3'-Diene VPA 1.23 (0.24-3.56) 0.16 (Tr-0.26) (E)-2,4-Diene VPA 3.68 (0.18-11.5) 0.52 (Tr-3.17) VPA 262 ( 15.6-1191) 15.3 (0.43-96) 4-OH VPA 17.2 (0.45-79.4) 29.7 (3.16-94.9) 3-OH VPA 10.6 (0.83-26.7) 10.1 (5.01-18.4) 5-0H VPA 5.21 (Tr-16.30) 12.6 (2.56-34.8) 4-Keto VPA 0.51 (Tr-3.60) 9.64 (1.09-28.1) 3-Keto VPA 7.93 (Tr-28.0) 160 (33.9-362) 2-PSA 0.23 (Tr-1.48) 1.57 (0.23-5.66) 2-PGA 6.45 (1.49-25.1) 32.0 (4.26-66.2) a Conjugated fraction b Unconjugated fraction 3. Metabolism studies involving VPA. 4-ene- and (E)-2.4-diene-VPA A. Animal studies Identification of the NAC conjugate of (E)-2,4-diene VPA in urine: When the urine extract from rats treated with (E)-2,4-diene VPA was methylated and analyzed by capillary GC/MS, a substantial part of the ion current was carried by the peak at retention time of 21.95 min. A representative chromatogram is shown in Fig. 29. The mass spectrum of this compound was identical to that of the synthetic NAC adduct of (E)-2,4-diene VPA (Fig. 14) and the compound had identical retention times with the synthetic standard when analyzed on two columns of different polarity. Using the same techniques the conjugate was also positively identified to be a prominent urinary metabolite of rats treated with 4-ene VPA (Fig. 30A). The conjugate could be detected in the urine extract of rats given a single dose of VPA but required the use of SIM because of the low levels of the conjugate. Of the (E)-2,4-diene VPA dose administered to rats, 40.2% + 8.7 (n=3) was excreted in urine as the NAC conjugate in 24 hr. The reversibility of the conjugation of (E)-2,4-diene VPA with NAC was investigated by heating the conjugate at 60°C in alkaline aqueous solution (pH=12). Under these conditions there was approximately 50% decomposition of the conjugate back to (E)-2,4-diene VPA and NAC. The GSH conjugate of (E)-2,4-diene VPA in rat bile: The bile extract was methylated with diazomethane before analysis by HPLC and the GSH conjugate of (E)-2,4-diene VPA was identified as the methyl ester. Methylation was necessary as efforts to synthesize a relatively pure conjugate of the free acid of (E)-2,4-diene VPA were not successful. The reaction of the free acid with GSH was very slow and purification of the 112 400000 4400(3(3 400000 360000 2 1 . 9 5 20 22 24 26 28 3 0 T I M E (min) Figure 29. A representative TIC chromatogram (EI mode) from the methylated urine extract of a rat dosed with (E)-2,4-diene VPA. The peak marked with an asterisk was identified as the dimethyl ester of the NAC conjugate of (E)-2,4-diene VPA. 2 2 0 0 0 2 6 0 0 0 1 8 0 6 0 1 6 0 6 0 1 4 0 0 0 •<- 12008H cr CD 1 0 0 0 0 -8 0 0 0 -6 0 0 0 -4 0 0 0 : 2008 - j 0-A 1 9 . 0 2 8 . 0 Jul/* f f \ Y Zl .0 2 2 . 0 2 3 . 0 TIME (min) 2 4 . 0 2 5 . 0 2 6 . 0 2 7 . 0 B ui z o D. (/) UJ CC T 0 T 4 TIME (min) Figure 30. A) EI/TIC chromatogram from the methylated urine extract of a rat given 4-ene VPA. The peak marked with an asterisk was characterized as the methylated NAC conjugate of (E)-2,4-diene VPA. B) HPLC profile of the methylated bile extract of rats dosed with 4-ene VPA. The peak marked with an asterisk was identified as the methylated GSH conjugate of (E)-2,4-diene VPA. conjugate by HPLC proved difficult because of co-eluting polar contaminants. The HPLC profile of the methylated bile extract from rats treated with (E)-2,4-diene VPA is shown in Fig. 31 together with a t r i -methylated sample of synthetic GSH conjugate of (E)-2,4-diene VPA. The peak marked with an asterisk had the same retention time as the authentic standard. The peak from the bile extract co-chromatographed with the synthetic standard but was absent in control bile. An identical peak was present in the methylated bile extract of animals administered 4-ene VPA (Fig. 30B). For spectral characterization these samples were analyzed by LC/MS/MS. Figure 32 is the LC/MS trace of the methylated bile extract from rats dosed with (E)-2,4-diene VPA. The mass spectra of the conjugate from synthetic and biological samples matched with each other and contained intense acetate attachment ions (m/z 548) of the methylated GSH conjugate of (E)-2,4-diene VPA (App. 15 and 16). The daughter ion (parent, m/z 548) spectra from the two sources were nearly identical as shown in Fig. 33. The spectra contained several ions which were structurally informative. Of particular significance was the fragment m/z 127 which was indicative of the position of the double bond. Non-thiol metabolites of (E)-2,4-diene VPA: The following monounsaturated acids were identified and characterized as metabolites of (E)-2,4-diene VPA in rat urine: 3-ene- (stereochemistry not determined), 4-ene- and (E)-2-ene-VPA (Fig. 34). On the basis of peak height the level of 3-ene VPA was higher than that of (E)-2-ene VPA and that of 4-ene VPA was intermediate between the two. 115 HI GO z o 0. CO LU CC VL_ ~r 0 B n r 4 8 —r 12 T" 0 4 8 I— 12 i 0 i 4 "T" 8 D 1^  0 4 8 I— 12 TIME (min) Figure 31. HPLC profile of the methylated bile extract of rats treated with (E)-2,4-diene VPA (A). The peak marked with an asterisk had the same retention time as the methylated synthetic sample of the GSH conjugate of (E)-2,4-diene VPA shown in (B) but was absent in the methylated control bile extract (D). Chromatogram C was obtained from a mixture of equivalent amounts of the conjugate in synthetic and bile samples. 116 Figure 32. LC/MS chromatogram (negative ion mode) obtained from the methylated bile extract of rats dosed with (E)-2,4-diene VPA. 117 75 50 25 (301-CH3-H2O)" 127 Vy 268 W-1 , ijl l^ lUllllihl.ltlljl., 128- H CoH - 127 ( 3n7 CH2CH=CHCHCOOCH3 r SiCHoCHCONHCHoCOOCHo 334 " 0=CCH2CH2CHCOOCH3 333 548 (M + OOCCHg)-456 100 ) t|Jhllll.J]J, I .l.l/l.ll.,^  600 C 3H 7 CH2CH=CHCHCOOCH3 S-CH2CHCONHCH2COOCH3 NH NH2 0=CCH2CH2CHCOOCH3 548 456 Figure 33. Daughter ion spectra obtained by argon CAD of the (M+OOCCH3)" anion of the negative ion LC/MS spectrum of A) methylated bile extract from rats dosed with (E)-2,4-diene VPA and B) authentic methylated GSH conjugate of (E)-2,4-diene VPA. 118 V00008-8 0 0 0 0 8 -7 0 0 0 0 0 -6 8 0 8 8 8 ->> £ 5 8 0 8 0 0 -<D +-> C •—• 4 0 0 0 0 0 -3 0 0 0 0 0 2 0 0 6 8 8 -1 0 0 0 0 0 -8 1 ' I ' 9 . 8 m/z 141 i 1 • i 1 • 1 8 . 8 | i"i—i—Y 11 . 8 1 2 . 8 TIME (min) 1 3 . 8 1 4 . 8 Figure 34. Mass chromatogram of m/z 141 obtained from the PFB derivatized urine extract of rats treated with (E)-2,4-diene VPA. Peaks 1, 2 and 3 were identified as 4-ene-, 3-ene- and (E)-2-ene-VPA, respectively. The 3-0H-4-ene VPA was identified as a urinary metabolite of (E)-2,4-diene VPA (Fig. 35) with the mass chromatogram of m/z 229 ([M-181]") showing a single major peak. A similarly derivatized synthetic sample of 3-0H-4-ene VPA had the same GC/MS properties. A synthetic sample of another regioisomer, 3-0H'-4-ene VPA, was available and gave two diastereomeric peaks upon GC/MS analysis. This isomer could not be detected in urine samples. Mass spectral evidence was obtained for the presence of the glycine amide conjugate of (E)-2,4-diene VPA in rat urine. The NICI mass spectrum of the PFB-derivatized urine extract (Fig. 36) contained the expected [M-181]" ion and another ion at [M-I8I-CO2]" which appeared to be diagnostic for the glycine amide metabolite. The presence of a heteroatom adjacent to the a-carbon seems to result in the decarboxylation of the carboxylate anion fragment as was also demonstrated for 3-keto VPA (Kassahun et al., 1989). 3-Keto-4-ene VPA and its thiol conjugates: The 3-keto-4-ene VPA and its thiol conjugates were sought in the biological fluids of rats treated with VPA, 4-ene-, and (E)-2,4-diene-VPA. The search for these compounds was made easier by the availability of synthetic standards. However, neither the 3-keto-4-ene VPA itself nor its thiol conjugates were evident in any of the treatment groups. Disposition of the GSH conjugates of 3-keto-4-ene VPA and (E)-2,4-diene VPA in rats: The GSH conjugates of 3-keto-4-ene VPA and (E)-2,4-diene VPA (methyl esters of the carboxyl groups of the substrates) were administered i .v . to rats in order to determine the type of S-l inked urinary metabolites of these conjugates. For both GSH conjugates the only detectable conjugated metabolites were the respective NAC 120 80000H 70000 60000 >, 50000 c/> S 40000-30000H 28880-18808-m/z 229 33 .58 • • • i • • •"" • i • • • • i • • • • l • • • • l • • • • l • • • • l 30.0 31.0 32.0 33.0 34.0 35.0 36.0 37.0 3 7 - 5 038.21 - i — i f i r | — f t i * i — i 38.8 TIME (min) 139 229 229 C H 2 C 6 F 5 0 -S i (CH 3 ) 3 173 148 168 180 r j T T T T j T T T T - p T 200 219 110 r100 9^0 ^0 >> P70 263 264 308 220 240 ' i1 1 1 1 i 1 co c CD ^ 0 a CD b58 > 0 b-30 •20 •10 260 280 300 m/z Figure 35. Extracted ion (m/z 229) from the PFB and TMS derivatized urine extract of rats given (E)-2,4-diene VPA (above) and the NICI mass spectrum of the peak with a retention time of 33.58 min (below). ca 121 320080-•89000-M0000-j=1680w8-+-> 120000-S0Q00-B I 11 \ m/z 196 40000-, I . I 8 12 16 20 24 28 32  TIME (min) 3 6 "40 4 t 3 £ 196 152 93H7 CH2=CH-CH=C-CO-NH-CH 2-i-COO-|-CH2-C 6F5 152—1 196—' 325 S09 261 100 200 1—i—1—1—n—1—J 1 1 388 379 y 474 ,-„S51 S9f 1/ / 671 / H 10 j-100 790 JQ0 : co -70 c cu 7 6 0 ^ cu is© +J ro : os 738 s$ 7 2 0 7 1 0 400 500 608 m/z Figure 36. A) mass chromatogram (m/z 196, ([M-181]")) from the PFB derivatized urine extract of rats treated with (E)-2,4-diene VPA, B) NICI mass spectrum of peak 1 which on a mass spectral basis appears to be the glycine amide conjugate of (E)-2,4-diene VPA. conjugates (Figs. 37 and 38). There was no evidence for the cysteine conjugates of these compounds. For the NAC conjugate of 3-keto-4-ene VPA GC/MS was not an appropriate method of analysis since the conjugate decomposed into the substrate and methyl NAC upon GC/MS analysis with no intact conjugate being observed. A satisfactory result was obtained only by LC/MS as shown in Fig. 38. The two peaks in Fig. 38 had the same MS/MS spectra and are assumed to be diastereomeric NAC conjugates of 3-keto-4-ene VPA. Although the monomethyl esters of the GSH conjugates were administered to rats, the respective NAC conjugates were recovered in the urine only as free acids. B. Human study Identification of the NAC conjugate of (E)-2,4-diene VPA in patients on VPA therapy: The NAC conjugate of (E)-2,4-diene VPA was found to be a urinary metabolite of VPA in patients. Figure 39, m/z 482, is a typical selected ion current chromatogram in the NICI mode obtained from the PFB derivatized urine extract of a patient. The ion monitored was the M-pentafluorobenzyl fragment anion of the di-PFB derivative of the conjugate. A full scan mass spectrum of the conjugate could also be obtained, in the NICI mode, from the derivatized urine extract of some of the patients. Further evidence of the formation of the conjugate in humans was obtained from the urine extract of a volunteer who was given a 50:50 mixture of VPA and [2H6]VPA to steady state. By monitoring m/z 482, a mass chromatogram similar to that seen for patient samples (Fig. 39 top) was obtained. When the ion at m/z 487 [M-181+2H5]~ was monitored (Fig. 39 bottom) a peak appeared at the expected retention time (25.82 min) 123 70008-60000-50009-40000-10000-14 .21 8 . 4 2 5 .11 cu 30000- 4 . 1 6 G 20000" ) 0 : | | 6 .72 18 .52 14 25 21 .75 12 16 20 T I M E (min) • < • » i " [ 1 1 1 1 1 1 1 1 1 1 1 1 1 • • 1 1 ' • 24 28 32 Figure 37. EI/TIC chromatogram of the methylated urine extract of rats dosed with the GSH conjugate of methyl (E)-2,4-diene VPA. The peak marked with an asterisk was identified to be the dimethyl ester of the NAC conjugate of (E)-2,4-diene VPA. 100 A 75 c 50 25 171 L_ r" 1 76 316—1 C3H7: C H 2 C H 2 C 0 C H C 0 i 0 C H 5 :|:: S 7 r C H 2 C H C O O C H 3 • 1 • 1 NHCOCH 203—11—144 c so (M + H) + Figure 38. A) LC/MS extracted ion (m/z 348) obtained from the methylated urine extract of rats given the GSH conjugate of methyl 3-keto-4-ene VPA, B) MS/MS of scans 198-218 which were characterized to represent the dimethyl ester of the NAC conjugate of 3-keto-4-ene VPA. 25.98 2 8 8 6 6 -laeee-leeee-1 4 8 8 6 -1 2 0 6 6 -1 6 0 0 6 -8 0 6 8 -6 8 8 8 -4 6 6 8 -2 e e e -m/z 482 23.48 A 6 | • n - . • , , , • , • ff ,', • ,\ . . . fpt ,'• , >-/>. p-r^-T-i-i-T-rrT-r, . / • , V. » • • . . • "l . . I .-7 , 2 8 . 8 2 1 . 0 2 2 . 0 2 3 . 0 2 4 . 8 2 5 . 0 2 6 . 0 2 7 . 0 2 8 . 0 2 9 . 8 3 6 . 0 3 1 . 0 3 2 . 8 3 3 . 0 29.92 TIME (min) 1 4886-j 1 2 0 8 6 -1 8 6 8 8 -30.67 8 6 6 6 -i/i c 6 6 8 8 - 1 4eee-2 6 6 8 -m/z 487 29.74 I 25.82 26.31 L U 2 1 ? 0 r 2 2 ' . 8 2 3 .G 1 2 4 . 0 2 5 . ' 0 26. ' . 0 2 7 ' . 0 2 8 . 0 2 9 . 6 3 8 . 8 3 1 . 0 3 2 . 8 3 3 . 0 3 4 . 0 3 5 . 6 TIME (min) SIM chromatogram of the PFB derivatized patient on VPA therapy. The peak marked to the di-PFB derivative of the NAC m/z 487: Peak marked with an asterisk is the pentadeuterated analog of the di-PFB derivative of the NAC conjugate of (E)-2,4-diene VPA obtained from the urine extract of a volunteer administered six doses of a 50:50 mixture of VPA and [ Hs]VPA. Figure 39. m/z 482: A typical urine extract of an epileptic with an asterisk corresponds conjugate of (E)-2,4-diene VPA i .e. slightly before the unlabelled metabolite (25.98 min). It is to be noted that the deuterated analogs of VPA metabolites elute slightly ahead of their unlabelled counterparts on capillary columns used for their analysis (Kassahun et al., 1989). In the bottom chromatogram of Fig. 39 the peaks with retention times of 26.31 and 30.67 min are background peaks also present in control urine. The peak with a retention time of 29.74 min appears to be an isomer of the conjugate (see below). From the ion chromatogram obtained from the urine extract of the volunteer the m/z 482 to 487 ratio was not unity even though a 50:50 mixture of VPA and [ 2Hs]VPA was administered. Instead the level, of the deuterated conjugate was 65% of the undeuterated analog. This small isotope effect was also observed in the metabolism of [2Hs]VPA to (E)-[2H5]2,4-diene VPA. As seen in Fig. 39 there was a second peak at m/z 482 (retention time 29.92 min), about equal in intensity to the peak identified as the NAC conjugate of (E)-2,4-diene VPA (retention time 25.98 min). This peak was absent in control urine and is thought to be an isomer of the conjugate. The synthetic sample of the NAC conjugate of (E)-2,4-diene VPA also contained this peak but at less than 10% of the intensity of the peak for the major isomer (App. 22). A corresponding isomer was also found in the urine extract of the subject administered labelled VPA (Fig. 39, m/z 487, 29.74 min). Quantitation of the NAC conjugate of (E)-2,4-diene VPA in human urine: The urinary concentration of the NAC conjugate of (E)-2,4-diene VPA was determined in 28 pediatric patients on VPA therapy. Quantitation of the conjugate was carried out using an NICI/GC/MS method developed for this purpose. Figure 40 shows the SIM chromatograms of the [M-181]" anions obtained from the internal standard and synthetic sample extracted from urine. A typical calibration curve is shown in Fig. 41. The assay had good precision with intra- and inter-day variation of less than 5 and 10%, respectively. The lower limit of detection (signal-to-noise ratio=3) was 10 pg injected on column. The mean concentration of the conjugate in the urine of these 28 patients was 1.38 /zg/ml with a range of trace to 5.74 itg/ml. The level of the conjugate when normalized to creatinine was 3-4 times higher in two patients who had developed VPA induced liver failure compared to those without this side effect. There was no difference in the urinary level of the (E)-2,4-diene VPA itself among the patients studied. 4. Identification and characterization of the thiol conjugates of  reactive metabolites of 4-pentenoic acid Urinary metabolites: Figure 42 illustrates a partial EI/TIC chromatogram obtained from the methylated urine extract of a rat dosed with 100 mg/kg of (E)-2,4-pentadienoic acid. The peak labelled 1 was identified as the cysteine conjugate of 3-oxo-4-pentenoic acid using various mass spectral techniques and by comparison with a synthetic standard. The retention times (GC/MS, LC/MS (Fig. 43)) as well as mass spectra (LC/MS (Fig. 44), PCI/GC/MS, EI/GC/MS (Fig. 45)) of the isolated metabolite were the same as those of the synthetic standard. The synthesized monomethyl NAC conjugate of 3-keto-4-pentenoate was converted to the dimethyl ester by treating the conjugate with methanolic HCl gas followed by CH2N2. This, however, resulted not only in the formation of the dimethyl ester but also caused deacetylation of 1600-1208-| 800-400-C 3 H 7 C H 2 - CH = CH - CH - COO - C H 2 - C 6 F 5 S-CH 2 -CH-NHCOCH 3 COO - C H 2 - C 6 F 5 m/z 482 1 i 1 i 1 i 12 14 i "f ri l i l i l1. •' 1 i l i l 16 > 11 T ' I 1 18 20 22 24 I I" ' I • I 26 28 30 1600->>1200-S 800-+-> c 400-10 12 m/z 414 JL C H 2 - C H 2 - COO - CH 2 - C 6 F 5 S-CH 2 -CH-NHCOCH 3 COO - C H 2 - C 6 F 5 1 i 1 i 1 i 1 i 1 i 1 i 1 i 1 i 1 i 1 i 1 i ' i 1 i 1 i 1 i 1 i 1 i 14 16 18 20 22 24 26 28 30 T IME (min) Figure 40. SIM chromatograms ([M-181]") of the PFB derivatives of the analyte (m/z 482, peak 1) and internal standard (m/z 414, peak 2) extracted from control urine. Concentration (ug/mL urine) Figure 41. A typical calibration curve for the quantitation of the NAC conjugate of (E)-2,4-diene VPA obtained from the PFB derivative of the conjugate using the NAC conjugate of acrylic acid as the internal standard. 130 80800-70000-60000-50000-S 40000-30000-20000-10000-16.0 17.8 18.0 19.0 20.0 21.0 22.0 23.0 TIME (min) Figure 42. EI/TIC chromatogram of the methylated urine extract of rats dosed with (EJ-^-pentadienoic acid. Peaks labelled were identified to be the cysteine conjugate of 3-oxo-4-pentenoate (peak 1) and isomeric NAC conjugates of (E)-2,4-pentadienoate (peaks 2, 3 and 4). the NAC conjugate was shown to be deacetylated upon treatment with methanolic HCl gas followed by CH2N2. Figure 43. LC/MS trace (extracted ion): from the methylated urine extract of rats given (E)-2,4-pentadienoic acid (top) and synthetic cysteine conjugate of methyl 3-oxo-4-pentenoate (bottom). 132 264 100! 754 197 (M + H) 204 ?32 I I I I CH 2 CH 2 COCH2 l CO!OCH3 S - C H 2 C H C O O C H 3 ' [Vi2 (M + NH4) + (t.i l l | f.m v .<jil jl .. JytLL, 350 400 450 100 204 232 •E SO C H 2 C H 2 C O C H 2 l C O ' i O C H 3 S - C H 2 C H C O O C H 3 AH2 350 400 450 Figure 44. LC/MS spectra: scans 231 (Fig. 43, bottom). -237 (Fig. 43, top) and scans 228-239 133 A 43 C H 2 C H 2 C O C H 2 C O O C H 3 S - C H 2 C H C O O C H 3 N H 2 S8 \ 5*5 .III 112 162 / il , 148 \ •I Ii ll i 144 172 204 198 / •1 18 •180 M + 79W 780 >-> -t-> - •r— CO j-70 c cu -t-> c 6^8 i—i cu T50 > - +-> rg r40 i— CU al r30 720 -10 X r i<-0 88 120 160 m/z 200 240 B 43 172 CH 2 CH 2 COCH 2 COOCH 3 S-CH 2 CHCOOCH 3 l\H2 74 \ 88 / • l 1 80 1 12 101 / 140 \ I I i 144 204 / 190 110 :108 790 l«0 >> h70 £ CU r +•> -60 c 263 -10 -50 cu cu cc ^0 <° 30 1-20 128 160 m/z 208 240 Figure 45. EI/GC/MS spectra of A) synthetic cysteine conjugate of methyl 3-oxo-4-pentenoate and B) peak 1, Fig. 42. the NAC conjugate to yield the cysteine conjugate. The identity of this S-linked compound was established using EI, PCI and LC/MS mass spectral techniques. Because the urine extract was also treated the same way (i.e. methanolic HCl gas followed by CH2N2) it appears that the cysteine conjugate was formed artif ic ial ly from the NAC conjugate in vitro. In Fig. 42 the peaks marked 2, 3 and 4 were identified as isomeric NAC conjugates of (E)-2,4-pentadienoic acid by comparing retention times and mass spectral data (Fig. 46) with an authentic standard. The major peak was assigned the structure (E)-5-(N-acetylcystein-S-yl)3-pentenoate on the basis of the spectroscopic data of the synthetic compound. Thus the main NAC conjugate of (E)-2,4-pentadienoic acid in the rat is (E)-5-(N-acetylcystein-S-yl)3-pentenoic acid. The same conjugate was identified as a urinary metabolite of rats treated with 4-pentenoic acid. The amount of the NAC conjugate of (E)-2,4-pentadienoic acid recovered in urine of rats treated with either (E)-2,4-pentadienoic acid or 4-pentenoic acid was much less than that of the NAC conjugate of (E)-2,4-diene VPA recovered after comparable doses of (E)-2,4-diene VPA or 4-ene VPA. The NAC conjugate of a third chemically reactive metabolite of 4-pentenoic acid, namely acrylic acid, was also identified and characterized in the urine of rats dosed with (E)-2,4-pentadienoic acid. The level of this conjugate, as judged by its peak height, was much smaller than that of the NAC conjugate of (E)-2,4-pentadienoic acid. Biliary metabolites: Bile from (E)-2,4-pentadienoic acid treated rats was purified by solid phase extraction, methylated and analyzed by LC and LC/MS/MS. Figure 47A is the LC/MS trace (extracted ion, m/z 478) of the methylated bile extract obtained in the positive ion mode using a 135 43 11 J L 71 ..ii 88 ll. 117 113 134 I. Illi, 144 166 176 .113 258-| C H 2 C H - C H C H 2 C O O C H 3 r " S - C H 2 C K C O O C H 3  1 ., N H C O C H 3 198 230 / \ r l 10 100 h90 I - O 0 I , I. n M 289 + -246 258 1 i >—> • i ' 280 70 :68 5^0 ^6 30 20 •10 00 c cu e cu > cu ' 1 I ' 80 120 I ' ' ' I 1 160 m/z 200 240 B 71 / 81 117 113 134 \ Illlll I 144 / 166 lln 176 • • l 1 160 C H 2 C H = C H C H 2 C O O C H 3 S - C H 2 C H C 0 0 C H 3 N H C 0 C H 3 198 230 / \ M 289 + -' I 1 200 258 / I 110 100 ^0 ^0 . c/l -70 c cu +-> t60 5 cu :50 -40 •— cu CC j-30 3^  20 10 •-0 • i 1 1 80 120 240 280 m/z Figure 46. EI/GC/MS spectra of A) authentic NAC conjugate of (E)-2,4-pentadienoate (dimethyl ester) and B) peak 4, Fig. 42. 136 75-S 50-25 50 3.0 100 6.1 150 92 Scan/Time (min) 200 12.3 250 15.4 Figure 47. Positive ion LC/MS chromatogram (extracted ion, m/z 478), A) from the methylated bile extract of rats dosed with (E)-2,4-pentadienoic acid, B) methylated authentic GSH conjugate of 3-oxo-4-pentenoate. mixture of CH3CN and NH4COOCH3 buffer as the mobile phase. The corresponding extracted ion chromatogram from the methylated synthetic standard is shown in Fig. 47B. The ion at m/z 487 represents (M+H)+ of the tetra-methylated GSH conjugate of 3-oxo-4-pentenoate. There were three major peaks at m/z 478 for both the synthetic and biological samples. The daughter ion spectra of the peaks from the bile extract were identical to those from the synthetic sample as shown in Fig. 48 for one pair of these peaks. The three peaks had the same MS/MS fragmentation patterns except for minor differences in the intensity of some of the fragment ions. Thus isomeric GSH conjugates of 3-oxo-4-pentenoate were present in the methylated bile extract of rats dosed with (E)-2,4-pentadienoic acid. In addition to the two free carboxyl groups, a third carbon of the GSH adduct of methyl 3-oxo-4-pentenoate was methylated upon treatment with diazomethane to give (M+Hj+ ion of m/z 478. The fragment ion, m/z 143 (Fig. 48), indicated that the additional methyl group was on the substrate portion of the conjugate. Because the methylene protons of the carbon between the two carbonyl groups are acidic, i t appears that this carbon was the site of the additional methylation. This would give a pair of diastereomers and would account for two of the three isomers observed. The keto function of the substrate moiety of the molecule may be enolized and consequently the hydroxyl group methylated. This enolized isomer may account for the third isomer observed. Figure 49A represents an LC/MS extracted ion chromatogram (m/z 448) from the methylated bile extract of rats given (E)-2,4-pentadienoic acid. In Fig. 49B the LC/MS trace (extracted ion, m/z 448) representing the (M+H)+ of the the methylated synthetic GSH conjugate of (E)-2,4-138 100 75 c 50 25 100 75 S. 50 25 143 111 101 446-f H 3 478 ^ CH 2 CH 2 COCHCO ! O C H 3 r : :~'~ 334 C H 2 C H C O N H C H 2 C O O C H 3 [59 $"2 0 = C C H 2 C H 2 p H C O O C H 3 •302 389-159 193 303 3 3 6 2)4 . 271 247 U.»li^ .A(^ AJ...JLl 357 389 100 200 300 400 m/z B 143 111 101 .,4 iJ,U4 v.-**, C H 3 C H 2 C H 2 C O C H C O O C H 3 S - C H 2 C H C O N H C H 2 C O O C H 3 0 = C C H 2 C H 2 C H C O O C H 3 159 303 193 11 214 229 271 247 336 100 200 300 400 m/z 446 (M + H) + 500 478 357 4 4 6 I 389 I ..I. J, JL. 500 Figure 48. Daughter ion spectra (parent, m/z 478) of A) Scans 102-106 (Fig. 47A) and B) Scans 102-106 (Fig. 47B) representing tetra-methylated GSH conjugate of 3-oxo-4-pentenoate. 139 Figure 49. LC/MS extracted ion chromatogram (m/z 448): A) methylated bile extract of rats treated with (E)-2,4-pentadienoic acid, B) methylated synthetic GSH conjugate of (E)-2,4-pentadienoate. pentadienoate is shown. The retention times of the major peaks from the two sources were the same with nearly identical MS/MS fragmentation patterns as shown in Fig. 50. Hence (E)-5-(glutathion-S-yl)3-pentenoic acid was the major GSH conjugate of (E)-2,4-pentadienoic acid in the rat. 5. Attempted synthesis of the CoA thio ester of (E)-2.4-diene VPA The chemical synthesis of the CoA ester of (E)-2,4-diene VPA was attempted by reacting the anhydride of the diene with CoASH under slightly alkaline conditions. The LC/MS spectrum of the synthetic sample did not show the pseudomolecular ion of the expected molecule. A possible explanation for the lack of mass spectral identification of the product is that the diene may have formed a di-CoA derivative (Mw=1657) which was probably susceptible to thermal degradation when the heated nebulizer LC/MS interface was used. In addition to the formation of the more conventional CoA ester derivative the diene can act as a Michael acceptor to form a conjugate of CoA where CoA is added across the double bond of the terminal carbon of (E)-2,4-diene VPA. The microsomal fraction of rat liver was used as a source of synthetase for the biological synthesis of the CoA ester of (E)-2,4-diene VPA. The LC/MS spectrum obtained from the incubation mixture contained a number of ions that could arise from the thio ester of (E)-2,4-diene VPA. However, confirmation of the structure was not possible due to the unavailability of an authentic standard. 100 176 •a so •£ 50-100 L _ C H 2 C H = C H C H 2 C O O C H 3 288 S - C H 2 C H C O N H C H 2 C O O C H 3 l O = C C H 2 C H 2 ; C H C O O C H 3 359-J 305 288 3S9 (M + H) + 300 400 500 m/z C H 2 C H = C H C H 2 C O O C H 3 S - C H 2 C H C O N H C H 2 C O O C H 3 r^ iH fvjH2 O = C C H 2 C H 2 C H C O O C H 3 305 288 359 100 200 300 400 m/z Figure 50. Daughter ion spectra (parent, m/z 448): A) Scans 111-(Fig. 49A), B) Scans 115-121 (Fig. 49B). IV. DISCUSSION 1. Chemistry Several compounds were synthesized in anticipation of proposed metabolic schemes while others were prepared as the need arose to characterize substances identified as potential metabolites by mass spectrometry. A. a.fl-Unsaturated keto metabolites and their thiol conjugates The putative metabolites of 4-ene VPA and 4-pentenoic acid, 3-keto-4-ene VPA and 3-oxo-4-pentenoic acid respectively, are characterized by two salient chemical properties: they are a,/}-unsaturated carbonyl compounds and are also /?-keto acids. These chemical features largely determined the synthetic routes that could be successfully used for their synthesis, and gave chemical instability to the compounds under conditions normally used for the analysis of VPA metabolites. Considerable difficulty was encountered in the synthesis of 3-keto-4-ene VPA. Several synthetic approaches were tried to obtain the free acid form of the compound in an amount sufficient for the desired experiments. However, partly because of the /J-keto acid nature of 3-keto-4-ene VPA the product obtained was mostly the decarboxylated analog (dienol). For this reason the methyl ester of 3-keto-4-ene VPA was synthesized since, for purposes of comparison of biological samples with that of a synthetic standard, the biological sample could easily be methylated using diazomethane. The ester was synthesized by alkylating the enolate of methyl pentanoate with acrolein and oxidizing the resulting hydroxy compound using tert-butyl hydroperoxide in the presence of a catalytic amount of chromium trioxide (Scheme 1). The use of tert-butyl hydroperoxide for the oxidation of alcohols to carbonyl compounds, as reported by Muzart (1987a, 1987b), obviates the need for large quantities of chromium (VI) reagents. This has advantages in that residual chromium would not interfere with final sample work-up and also that the oxidation would take place under milder conditions. An attempt to de-esterify a sample of 3-keto-4-ene VPA methyl ester under alkaline conditions was not successful, the ester being converted to an undetermined product(s). Because the strong UV absorbance observed for methyl 3-keto-4-ene VPA disappeared after the above sample manipulation, the product may be a result of addition of water across the double bond of 3-keto-4-ene VPA to give 5-hydroxy-3-keto VPA. Unfortunately GC/MS analysis of the product(s) did not furnish adequate information on their identity. The 3-oxo-4-pentenoate was also synthesized as the methyl ester using a procedure similar to that followed for 3-keto-4-ene VPA. Reported syntheses of 3-oxo-4-pentenoate have employed rather long and tedious synthetic routes (eg. Collins and Tomkins, 1977; Stork and Guthikonda, 1972; Pelletier et al., 1968). In the present study methyl 3-oxo-4-pentenoate was synthesized by an efficient two-step method (Scheme 3). Alkylation of the a-carbanion of methyl acetate with acrolein and oxidation of the resulting hydroxy compound afforded the 3-oxo-4-pentenoate in 41% yield. Synthesis of the GSH conjugate of methyl 3-keto-4-ene VPA was first attempted using triethylamine as a catalyst. NMR analysis of the resulting product showed that the conjugate had formed a triethylamine salt. On the other hand triethylamine was not a problem in the synthesis of the NAC conjugate since methyl NAC was used as one of the starting materials. The GSH conjugates of methyl 3-keto-4-ene VPA and 3-oxo-4-pentenoate were prepared by mixing a methanolic solution of the substrate with an aqueous solution of GSH and adjusting the pH of the mixture to 7.8. Under these conditions the reaction was instantaneous and almost quantitative. A facile reaction of these compounds with GSH and NAC is to be expected as the reaction of a,^-unsaturated carbonyl compounds with Michael donors such as thiols is a well known fact (Friedman et al., 1965). The GSH and NAC conjugates of methyl 3-keto-4-ene VPA and 3-oxo-4-pentenoate were fully characterized using various spectral techniques (Figs. 6-8, 10, 16-20). In each case the adduct was the result of conjugate addition where the thiolate ion of GSH and/or NAC attacked the terminal carbon of the chain bearing the keto function. The methine and methylene protons which are present between the two carbonyl groups of the substrate moieties of the thiol conjugates of methyl 3-keto-4-ene VPA and 3-oxo-4-peiitenoate exchanged with deuterium quite readily and thus gave no *H-NMR signals when D2O was used as the NMR solvent (Figs. 6 and 17). These protons were so acidic that the methylene carbon of methyl 3-oxo-4-pentenoate was methylated upon treatment with diazomethane (Fig. 48). For 3-keto-4-ene VPA such a propensity for methylation was not observed probably because of the steric hindrance of the propyl substituent at position 2. A. 2,4-Diunsaturated carboxvlic acid metabolites and their thiol  conjugates (E)-2,4-diene VPA and (E)-2,4-pentadienoic acid were synthesized using modifications of the methods of Lee et al. (1989) and Worley and Young (1972), respectively. When these acids were reacted with either GSH or NAC under conditions similar to that described in the preceding section for the a,/J-unsaturated keto metabolites, the rate of reaction was very slow and consequently l i t t l e product was obtained. The poor reactivity of the free acids appears to be due to the fact that at the pH (7.4) used for the synthesis the acids would be largely in the ionized form. Charge repulsion would preclude activation of the terminal double bond to Michael addition by GSH. On the contrary, the methyl ester derivatives of these diene metabolites reacted with GSH and NAC fairly rapidly resulting in good yields of the respective adducts. Compared to the rate of reaction of the a,^-unsaturated keto metabolites however, the reaction of the diene esters with GSH was slower as indicated by a 50% (in 5 hr) and 90% (in 30 min) yield for the GSH conjugates of methyl (E)-2,4-diene- and methyl 3-keto-4-ene-VPA, respectively. Methyl esters of (E)-2,4-diene VPA and (E)-2,4-pentadienoate belong to the category of compounds with a general structure of C=C-C=C-Z where Z is COOR, CHO, COR, CN, NO2, etc. Such systems can undergo Michael-type of reactions by a mechanism that involves either a 1,2-, 1,4-, or 1,6-addition (March, 1985). In the synthesis of the GSH conjugate of methyl (E)-2,4-diene VPA only one product was observed by LC/MS and NMR (Figs. 12 and 13) and was established to be the result of the addition of the thiolate ion at the terminal carbon with the remaining double bond appearing at position 3. In a 1,4-addition the thiolate ion would be attached to carbon 3 and the terminal double bond (position 4) would remain intact (Fig. 51). Methyl (E)-2,4-diene VPA or (E)-2,4-pentadienoate were not expected to undergo additions of the 1,2-type because a 1,2-addition is generally speaking not likely for compounds with structures similar to these diene esters (Oare and Heathcock, 1989). The major GSH conjugate of (E)-2,4-pentadienoate methyl ester had a similar structure with the thiolate attached at the terminal carbon and the double bond appearing at position 3. Thus, both diene analogs underwent nucleophilic addition reactions mainly via the 1,6-mechanism. The configuration of the double bonds of the major GSH conjugates of methyl (E)-2,4-diene VPA and (E)-2,4-pentadienoate was determined by NMR to be trans, i .e. the structures of the conjugates were (E)-5-(glutathion-S-yl)-3-ene VPA and (E)-5-(glutathion-S-yl)-3-pentenoate. This was established from the coupling constant of the vinylic protons which was 16 Hz. To the best of our knowledge there is very l i t t l e information in the literature regarding the stereochemistry of products arising from similar reactions. In the mammalian liver the enzyme 2,4-dienoyl CoA reductase catalyzes the NADPH-dependent addition of two hydrogen atoms to the double bonds of fatty acyl CoAs having a 2,4-dienoyl CoA structure (Osmundsen and Hovik, 1988). The double bond of the product appears at position 3 and in some cases was shown to have the trans configuration (Dommes and Kunau, 1984). Although the mechanism of the reaction of a diene ester with GSH i.e. nucleophilic addition is likely to be different from that of the reduction of a 2,4-dienoyl CoA fatty acid, the two reactions are similar in as much as that in the products of both reactions the remaining double bond appears at position 3. In view of the mechanism of the reaction of methyl (E)-2,4-diene VPA or methyl (E)-2,4-pentadienoate with GSH it is not clear why the reaction should be stereoselective for the trans isomer. A plausible explanation is that such an orientation would result in the more thermodynamically stable product, with the two bulky groups on either side of the double bond farther apart than they would be in the cis isomer. Methyl (E)-2,4-diene VPA is prochiral and forms a chiral center upon reaction with GSH. Hence the reaction of (E)-2,4-diene VPA with GSH can result in either diastereomers, geometric or regio-isomers. The structures of the regioisomers and the chiral centers in the conjugates are shown in Fig. 51. Similar isomers for the GSH conjugate of methyl (E)-2,4-pentadienoate will be observed i f both 1,4- and 1,6-addition mechanisms are followed. For the synthesized GSH conjugate of (E)-2,4-diene VPA only one isomer was detected by LC/MS whereas NICI/GC/MS of the NAC conjugate revealed two isomers with the minor isomer having a longer retention time (App. 22). The major isomer was characterized to be (E)-5-(N-acetylcystein-S-yl)-3-ene VPA. The minor isomer is not likely to be the (Z) or cis isomer of the above compound because cis isomers normally elute before their trans counterparts on the semi polar capillary columns used for the analysis of VPA metabolites. Therefore the minor isomer might be a diastereomer or regioisomer. The pattern of isomers arising from the metabolism studies will be discussed in the metabolism section. 148 93*7 G S - C H 9 - C H = C H - C H - C O O M e C 3 H 7 C H o = C H - C H - C H - C O O M e GS Figure 51. Scheme showing the possible regioisomeric GSH conjugates of methyl (E)-2,4-diene VPA as a result of 1,6- (A) and 1,4- (B) additions. The chiral centers that can be formed are marked with asterisks. The reaction of methyl (E)-2,4-pentadienoate with GSH produced two isomers as detected by LC/MS (Fig. 24) while the reaction with NAC resulted in four isomers which were detected by GC/MS (Fig. 26A). With a 1,6-addition mechanism the reaction of methyl (E)-2,4-pentadienoate with GSH is expected to produce at most two geometric isomers. The ^C-NMR spectrum (Fig. 23) of the GSH conjugate of methyl (E)-2,4-pentadienoate appears to indicate two such isomers with a pair of signals for the -CH=CH- carbons. When the nucleophlic addition follows a 1,4-mechanism (for structures see Fig. 52) a chiral center would be created at position 3 of the substrate moiety of the GSH or NAC adduct again resulting in a maximum of two isomers (Fig. 52). The four isomers of the NAC conjugate of methyl (E)-2,4-pentadienoate observed by GC/MS therefore may be the two geometric isomers of 5-(N-acetylcystein-S-yl)-3-pentenoate (peaks 3 and 4, Fig. 26A) and the two diastereomers of the conjugate with the structure of 3-(N-acetylcystein-S-yl)-4-pentenoate (peaks 1 and 2, Fig. 26A). Unfortunately the EI spectra of these compounds were the same and did not enable the discrimination between regioisomers. Definitive structural assignment of these compounds will have to await the preparative isolation of the compounds in sufficient quantities for NMR analysis. The apparent lack of formation of the analogous conjugate of (E)-2,4-diene VPA i.e. 3-(N-acetylcystein-S-yl)-4-ene VPA might be attributable to the steric interference of the propyl substituent at position 2, making 1,4-addition unlikely. Reversibility of the conjugation reaction with thiols and the in vitro stability of GSH and NAC conjugates is an important consideration when dealing with such conjugates. Toxicity of some GSH conjugates can H O O C - C H - C H 0 - S - C H 9 ^ / H H \ / H N H C O C H 3 / C = ? C v / \ \ C H 2 - C O O M e HOOC-CH-CH 2 -S-CH 2 ^CH 2 -COOMe , NHCOCH3 (E)-5-(N-acetylcystein-S-yl)3-pentenoate (Z)-5-(N-acetylcystein-S-yl)3-pentenoate CH2-COOMe H O O C - C H - C H 2 - S " r!iHCOCH3 CH2-COOMe -H H-CH=CH2 (3R,2,R)-3-(N-acetylcystein-S-yl)4-pentenoate S-CH 2-CH-COOH N H C O C H 3 CH=CH 2 (3S,2'R)-3-(N-acetylcystein-S-yl)4-pentenoate Figure 52. Structures of the possible isomeric NAC conjugates of methyl (E)-2,4-pentadienoate. Four isomers were detected by GC/MS for the synthetic sample while three isomers were identified in the urine extract of rats treated with (E)-2,4-pentadienoic acid. result from the reversal of the conjugation reaction (van Bladeren, 1988). An example of a reversible conjugation is that of the conjugates of N-(l-methyl-3,3-diphenylpropyl)isocyanate which were found to release the isocyanate under mild alkaline conditions (Mutlib et al., 1990). Knowledge of the factors influencing the stability of GSH and NAC conjugates is also important from the analytical point of view. Heating the NAC conjugate of (E)-2,4-diene VPA with alkali resulted in some reversibility of the conjugation reaction as evidenced by the amount of conjugate before and after alkaline treatment. The NAC conjugate of (E)-2,4-diene VPA was found to be quite stable under the GC conditions used for its analysis, with no trace of decomposition products of the NAC conjugate being detected. This is in contrast to the NAC conjugate of methyl 3-keto-4-ene VPA which readily underwent decomposition to NAC and methyl 3-keto-4-ene VPA in the GC inlet (Fig. 9). 2. Development of analytical methods A. Assay development for non-thiol metabolites and profiles in patients Despite the importance given to the metabolic profile of VPA due to the potential anticonvulsant activity and toxicity of some VPA metabolites, very l i t t l e has been published on VPA metabolite profiles in epileptic patients. In this study it was sought to obtain an assessment of the 4-ene VPA pathway relative to other metabolic pathways in epileptic patients on monotherapy and combined therapy with CBZ and CLBZ. Partly because of their large number, the wide concentration range at which they are found in biological fluids, and similarities to endogeneous compounds, the simultaneous quantitation of VPA metabolites is of considerable difficulty. Published GC/MS methods for the analysis of VPA metabolites are based on either t-BDMS (Abbott et a7., 1986) or TMS (Nau et al., 1981b; Tatsuhara et a7., 1987; Rettenmeier et a7. 1989) derivatives. Two methods which employed TMS derivatives (Nau et al., 1981b; Tatsuhara et al., 1987) did not include all the important metabolites, while in a third method (Rettenmeier et al., 1989) separation was achieved by using a much longer column and GC run time compared to the others. Although the previously reported assay from our laboratory (Abbott et al., 1986) was satisfactory for most metabolites, there were chromatographic problems with some of the hydroxy metabolites. In the present study a GC/MS method for the simultaneous quantitation of VPA and 14 of its metabolites was developed using PFB derivatives and negative ion detection. The assay made use of the inherent sensitivity and soft ionization nature of electron-capture NICI for a successful quantitation of this large array of metabolites. By a combination of PFB (carboxyl group) and TMS (hydroxyl and 3-oxo moiety) derivatization and the use of a non-polar capillary column, adequate chromatographic separation and sensitivity were achieved. The NICI assay developed was superior to published methods in many respects. Separation of 14 metabolites and the parent compound was achieved in a reasonable GC run time of 25 min with all compounds giving sharp peaks as shown in Fig. 28. VPA was completely resolved from 4-ene VPA as were the monounsaturated metabolites. PFB derivatization resulted in baseline separation of the geometric and regioisomeric diunsaturated metabolites. This is significant because with t-BDMS and TMS derivatives of the toxic metabolite (E)-2,4-diene- and (E,Z)-2,3'-diene-VPA overlap and having the same [M-57]+ or [M-15]+ fragment ions are counted as one. Other metabolites, such as 3-OH VPA, which chromatograph poorly as t-BDMS derivatives were analyzed quite readily by using TMS and PFB derivatives. The NICI method was highly sensitive for VPA being 30-50 times more sensitive than the t-BDMS derivative by EI. The lower limit of detection for metabolites varied between 1-8 ng/mL based on a signal to noise ratio of > 3. The assay was successfully employed for the metabolic profiling of VPA in pediatric patients and for tissue distribution studies of VPA metabolites in animals. The major serum metabolites in pediatric patients on VPA monotherapy were (E)-2-ene-, (E,E)-2,3'-diene-, and 3-keto-VPA as shown in Table 4. (E)-2-Ene- and 3-keto-VPA are /3-oxidation products (Loscher, 1981b), whereas the origin of (E,E)-2,3'-diene VPA is unknown although its concentration pattern suggests it might be a derivative of (E)-2-ene VPA probably via 3-ene VPA (Rettenmeier et al., 1987). The 4-ene VPA metabolite was found at low levels (mean = 0.20 /xg/mL) while (E)-2,4-diene VPA had a mean serum concentration of 0.65 /xg/mL. The serum concentration values obtained for most of the metabolites in this study compare favorably to those previously reported by Abbott et al. (1986) and Tatsuhara et al. (1987). Notable exceptions were the levels of 4-ene- and (E,E)-2,3'-diene-VPA which were found to be lower and higher, respectively, than in the previous study (Abbott et al., 1986). The reasons for these discrepancies appear to be the better resolution of 4-ene VPA from VPA and the availability of a synthetic standard for preparing the calibration curve of (E,E)-2,3'-diene VPA in the present assay. The serum levels of 4-keto- and 5-OH-VPA were significantly higher in the VPA + CBZ group compared to the monotherapy group (Table 5), suggesting the induction of w (5-OH VPA) and u-1 (4-keto VPA) pathways, consistent with the fact that CBZ induces VPA metabolism (Panesar et al., 1989). The serum concentrations of VPA and its /3-oxidation products were lower in the VPA + CBZ group. The decrease in the serum concentration of 2-ene VPA, when VPA was coadministered with CBZ, was also documented in a study in adult volunteers (Panesar, 1987). Polytherapy had no significant effect on the serum levels of 4-ene- and (E)-2,4-diene-VPA. Similar results were obtained in a recent report of VPA urine metabolite profiles (Levy et al., 1990) where CBZ had no significant effect on the urinary levels of 4-ene VPA. B. Assay of the NAC conjugate of (E)-2,4-diene VPA Once the presence of the NAC conjugate of (E)-2,4-diene VPA in the urine of patients taking VPA was established, it was necessary to develop a method for the quantitative determination of the conjugate in those patients. The measurement of NAC conjugates in urine allows for non-invasive monitoring of the exposure of subjects to toxic agents. Determination of NAC conjugates in a readily available biological fluid such as urine, also enables the estimation of the extent of GSH conjugation and the characterization of the chemistry of GSH conjugation without the need for surgical procedures in animals. The NAC conjugate of (E)-2,4-diene VPA was isolated from urine by acidifying urine to pH 1-2, adding NH4CI, and extracting with ethyl acetate. It was not possible to determine recovery because the authentic NAC conjugate was available as the dimethyl ester and had to be hydrolyzed and then extracted from aqueous solution. Using a similar extraction procedure, the recoveries (concentration range low /ig/mL) for benzyl and phenyl mercapturic acids were 94 and 70%, respectively (Laham and Potvin, 1987; Jongeneelen et al., 1987). For future quantitative work it will be important to preparatively isolate the free acid form of the NAC conjugate of (E)-2,4-diene VPA to use as the standard. The quantitation of NAC conjugates has been carried out using various analytical instruments such as UV spectrophotometry, HPLC, and GC (Vermeulen, 1989). The most selective and sensitive assay for the NAC conjugates of (Z-) and (E)-l,3-dichloropropene was GC/MS by SIM (Onkenhout et a / . , 1986). The NAC conjugate of (E)-2,4-diene VPA was found to be sufficiently thermostable for GC analysis. The assay was based on PFB derivatization and negative ion detection. The conjugate underwent efficient derivatization with PFBB to yield a di-PFB derivative with suitable GC/MS properties as shown in Fig. 40. The di-PFB derivative of the conjugate gave a sharp peak and its NICI mass spectrum (App. 17) was characterized by an intense ion at m/z 482 which corresponded to the loss of one PFB moiety with the remaining PFB retained on the carboxylate anion fragment. The dimethyl ester of the NAC conjugate of acrylic acid was synthesized and, after hydrolysis with alkali, was used as the internal standard. Being an NAC conjugate as well as a dicarboxylic acid, the internal standard had similar chromatographic and mass spectral properties to that of the analyte and was expected to behave in a similar way in solution. This enabled the development of an assay for the NAC conjugate of (E)-2,4-diene VPA in patient urine using 1 mL samples. The sensitivity of the assay was 2 ng/mL with a signal to noise ratio of >3. To our knowledge this is the first report of the use of PFB derivatization and negative ion detection for the analysis of thiol conjugates. The concentration of the conjugate in some of the patients studied was in the low ng/mL level. However, the sensitivity of the assay, which was mainly due to the type of derivative and use of NICI, was adequate for the analysis. The mean concentration of the conjugate in patients on VPA was found to be 1.38 /xg/mL (n=28) with a range of trace to 5.74 jug/mL. Using the NICI methodology, a comparison of the levels of the conjugate in VPA monotherapy and combined therapy with other antiepileptic drugs is currently underway in our laboratory. 3. Metabolism A. 4-Ene- and (E)-2.4-diene-VPA The serum levels of the toxic metabolites of VPA namely, 4-ene-and (E)-2,4-diene-VPA were found to be respectively, 0.2 and 0.7% that of VPA (Table 4) indicating that they are only minor metabolites in the serum of epileptic children. Although polytherapy is known to be a predisposing factor to VPA hepatotoxicity, the level of either 4-ene- or (E)-2,4-diene-VPA was not increased in combination therapy as shown in Table 5. In rats too, phenobarbital treatment did not greatly enhance the formation of 4-ene VPA from VPA (Granneman et al., 1984c). This does not necessarily mean that polytherapy with microsomal enzyme inducers fails to lead to an increased formation of 4-ene VPA. An alternate reasoning is that the increased formation of 4-ene VPA is not necessarily reflected by its serum levels; an increased formation clearance of 4-ene VPA was demonstrated after CBZ administration in a drug interaction study of CBZ and VPA (Panesar, 1987) and also in patients given VPA in combination with either CBZ or phenytoin (Levy et al., 1990). In a previous study (Tennison et al., 1988), the monitoring of 4-ene VPA serum levels was not found to predict risk of fatal hepatotoxicity. Despite the lack of a sufficient number of studies of metabolite levels from patients who died or recovered from VPA-related hepatotoxicity, the available evidence suggests that 4-ene VPA concentration in serum is not predictive of potential VPA hepatotoxicity. This raises the possibility that 4-ene VPA may be metabolized to secondary metabolites that are covalently bound to tissue macromolecules. The metabolism studies of 4-ene- and (E)-2,4-diene-VPA in the rat were therefore undertaken in part to address this apparent dissociation between the 4-ene VPA metabolite level in patients and VPA hepatotoxicity. As discussed in the introduction to 4-ene VPA and its hepatotoxic analogs, 4-pentenoic acid and hypoglycin are substrates for enzymes of the fatty acid /J-oxidation system. It is well known that the endoplasmic reticulum is the site for several Phase I and some Phase II routes of drug metabolism. However, for drugs with acyl and alkyl side chains /}-oxidation is also an important metabolic pathway (Yamada et al., 1987). For example, various hypolipidemic drugs are activated to their acyl CoA derivatives and are subsequently /J-oxidized to various metabolites (Aarsland et al., 1990). Xenobiotic CoA thio esters play a critical role in the metabolism, activity, and toxicity of a wide variety of xenobiotic carboxylic acids (Caldwell, 1984). Thus, i t is not surprising that 4-ene VPA is capable of activation to its CoA ester and subsequent /J-oxidation. Detection of thiol conjugates and other metabolites of (E)-2,4-diene VPA: In the present study it was found that 4-ene VPA was bioactivated to a metabolite that avidly reacted with GSH. The product of this interaction of reactive metabolite with GSH was determined to be (E)-5-(glutathion-S-yl)-3-ene VPA i.e. the GSH conjugate of (E)-2,4-diene VPA. The NAC conjugate was subsequently confirmed as a metabolite in the urine of patients administered therapeutic doses of VPA. The presence of these thiol conjugates in the respective species and biological fluids was established using various spectral techniques as presented in the results section. The interaction of (E)-2,4-diene VPA with GSH and NAC in vitro produced the respective adducts, the structures of which were determined using NMR and mass spectrometry. The site of attachment of GSH or NAC with the substrate molecule was unequivocally demonstrated to be at the terminal carbon resulting in a shift of the double bond to the 3 position. Several isomers, including geometric, diastereomeric, and regio-isomers could theoretically be formed from the interaction of GSH with (E)-2,4-diene VPA. Only one isomer was observed by HPLC for both the bile extract (Fig. 32) and synthetic standard. On the other hand, GC/MS of the NAC conjugate isolated from the urine of patients on VPA therapy appears to indicate a second isomer of the conjugate (Fig., 39 top). Whether this isomer is geometric, diastereomeric or positional to the identified NAC conjugate remains to be established. The detection of these thiol metabolites provided evidence that VPA is metabolized in vivo to a chemically reactive intermediate capable of reaction with cellular nucleophiles most probably in the liver. The capacity of (E)-2,4-diene VPA to form tissue bound residues is supported by a recent tissue distribution study in rats wherein (E)-2,4-diene VPA could not be detected in extracts of liver tissue homogenates even though it was present in serum (Lee et a7., 1991). In contrast, VPA and other metabolites were detectable in both tissues. The NAC conjugate of (E)-2,4-diene VPA was a major urinary metabolite of (E)-2,4-diene VPA as indicated by the nearly 40% of the dose of the diene recovered as the NAC conjugate. Other metabolites of (E)-2,4-diene VPA that were identified in urine were 3-0H-4-ene-, 3-ene-(stereochemistry not determined), 4-ene-, and (E)-2-ene-VPA. The metabolic origin of 3-ene VPA from (E)-2,4-diene VPA could be rationalized by invoking the action of the 2,4-dienoyl CoA reductase enzyme on (E)-2,4-diene VPA. The subsequent action of isomerase, the enzyme responsible for converting a 3-ene double bond to a trans 2-ene double bond, would then give (E)-2-ene VPA. (E)-2-ene VPA is expected to undergo further /3-oxidation to yield other VPA metabolites such as 3-0H-and 3-keto-VPA, but no effort was made to detect these metabolites. 3-Ene- and (E)-2-ene-VPA were also detected in urine following the administration of 4-ene VPA to monkeys (Rettenmeier et a7., 1986b). The detection of 4-ene VPA as a metabolite of (E)-2,4-diene VPA can be rationalized by invoking the 1,4-addition of hydrogens to (E)-2,4-diene VPA catalyzed by 2,4-dienoyl CoA reductase. The formation of a 1,4-addition product (4-ene VPA) in addition to the 1,6-addition product (3-ene VPA) may be due to the branched nature of the diene as opposed to endogeneous substrates of the enzyme. Branching may cause the enzyme to lose some of its regio-selectivity. Alternatively, 4-ene VPA may be formed by the action of 2-enoyl CoA reductase on (E)-2,4-diene VPA. This enzyme is known to act on fatty acids with a 2-enoyl CoA structure (Kunau and Dommes, 1978). The amount of 3-ene VPA recovered from urine after (E)-2,4-diene VPA administration to rats was not quantitated. However, the reduction of (E)-2,4-diene VPA to 3-ene VPA is not likely to be the major pathway of (E)-2,4-diene VPA metabolism as the diene was found to be largely excreted as the NAC conjugate. In contrast, the major metabolic pathway of 4-pentenoic acid was the reduction of its diene metabolite (E)-2,4-pentadienoic acid to 3-pentenoic acid and oxidation of the latter to acetyl CoA and propionyl CoA (Schulz, 1983). Thus, (E)-2,4-diene VPA tended to bind with GSH or accumulate to a greater extent than (E)-2,4-pentadienoic acid. The relatively inefficient reduction of (E)-2,4-diene VPA to 3-ene VPA might be related to the hepatotoxic properties of (E)-2,4-diene VPA. This is supported by the observation that inhibition of fatty acid oxidation by 4-pentenoic acid and hence the resulting toxicity is abolished by treating rats with clofibrate (Borrebaek et al., 1980; van Hoof et a7., 1985). By its strong inducing effect on 2,4-dienoyl CoA reductase, clofibrate treatment channels the metabolism of (E)-2,4-pentadienoyl CoA toward the 3-pentenoyl CoA pathway. In this connection it would be interesting to determine i f treatment of rats with clofibrate or glycine would decrease or abolish the toxicity of (E)-2,4-diene VPA. Hypoglycin toxicity is attenuated by the administration of glycine (Al-Bassam and Sherratt, 1981). Similarly, administration of glycine might stimulate the excretion of (E)-2,4-diene VPA as its glycine amide conjugate. Of course administration of NAC would be expected to decrease the toxicity of (E)-2,4-diene VPA by increasing GSH in the liver. GSH precursors such as NAC are effective against acetaminophen toxicity largely through provision of GSH for conjugation with the reactive metabolite of acetaminophen (Boobis et a/ . , 1989). 3-Keto-4-ene VPA and its thiol conjugates: The metabolism studies in rats involving 4-ene- and (E)-2,4-diene-VPA were conducted primarily to try and isolate the thiol derivatives of 3-keto-4-ene VPA. The experimental findings were somewhat surprising in that a major metabolite of (E)-2,4-diene- and 4-ene-VPA was in fact the GSH conjugate of (E)-2,4-diene VPA. The finding that (E)-2,4-diene VPA can be activated to an electrophilic intermediate capable of rapid reaction with GSH was not anticipated. Because intact 3-keto-4-ene VPA may not be sufficiently stable to survive in urine or under sample work-up conditions, i t was hoped that detection in the form of its thiol conjugates would be feasible. In vitro, 3-keto-4-ene VPA reacted readily with GSH and NAC and consequently synthetic samples were available to facilitate the search for these metabolites. However, neither the 3-keto-4-ene VPA nor its thiol conjugates were evident in any of the treatment groups. In patient urine the 3-keto-4-ene VPA was not detected even though the more sensitive NICI technique was used and the ions expected to arise from the PFB derivative of 3-keto-4-ene VPA were monitored. Other forms of 3-keto-4-ene VPA that were searched for by GC/MS but not detected included the decarboxylated derivative of the acid. It is generally accepted that NAC conjugates that appear in the urine are the major end products of the GSH pathway although cysteine conjugates, S-methyl derivatives, and sulphoxides can also be formed (Onkenhout, 1983). Metabolites derived from the GSH conjugate of methyl 3- keto-4-ene VPA were determined by administering the GSH conjugate to rats. The NAC conjugate was the only GSH-derived metabolite that could be identified in urine (Fig. 38). Hence i f the GSH conjugate of 3-keto-4- ene VPA was formed in vivo, the NAC conjugate should have been detected in urine. The init ial proposal to synthesize the GSH conjugate of 3-keto-4-ene VPA from in situ metabolism of 4-ene VPA in mitochondrial fractions of rat liver was not attempted as this would offer no mechanistic information in the absence of TA? vivo formation of 3-keto-4-ene VPA. One technique that might be useful to establish the in vivo formation of 3-keto-4-ene VPA would be to isolate and characterize the protein (enzyme) adduct of the compound after administration of a labelled analog of 4-ene- or (E)-2,4-diene-VPA to rats. There are a number of possible reasons why 3-keto-4-ene VPA or its thiol addition metabolites were not detected in the present study. The 3-keto-4-ene VPA and its thiol conjugates could have been formed, but may have been at levels below the detection limit of our analytical method which was based on EI/GC/MS of the dimethyl ester in the case of the NAC conjugate. It was not possible to develop an NICI method based on the PFB esters for a more sensitive detection of the NAC conjugate of 3-keto-4-ene VPA. Attempts to prepare the free acid form of the conjugate were unsuccessful. The detection of 3-0H-4-ene VPA, the immediate precursor to 3-keto-4-ene VPA, after (E)-2,4-diene VPA administration to rats is the best evidence for the in vivo formation of 3-keto-4-ene VPA. The 3-0H-4-ene VPA was also detected in previous metabolism studies of 4-ene VPA (Rettenmeier et al., 1985, 1986b). It should nevertheless be noted that 3- 0H-4-ene VPA could also be formed by a direct microsomal hydroxylation of 4-ene VPA. The latter was identified as a urinary metabolite of (E)-2,4-diene VPA as discussed below. As well, 3-0H-4-ene VPA having a substituent at position 2 and a terminal double bond may not be a good substrate for the /J-oxidation enzyme responsible for oxidation of hydroxy acyl CoAs to keto acyl CoAs. Another reason for the lack of detection of 3-keto-4-ene VPA or its derivatives in the metabolism studies that were undertaken is that 4- ene-/(E)-2,4-diene-VPA may not be metabolized at all to 3-keto-4-ene VPA. (E)-2,4-diene- and 3'-keto-4-ene-VPA, which is a product of /J-oxidation on the saturated side chain of 4-ene VPA (for metabolic scheme see Fig. 53), are major metabolites of 4-ene VPA. (E)-2,4'-diene VPA, an intermediate in the metabolism of 4-ene VPA to 3'-keto-4-ene VPA, was not detected as a metabolite of 4-ene VPA (Fig. 53) (Rettenmeier et al., 1985). In the present study a major urinary metabolite of (E)-2,4-diene VPA in the rat was the NAC conjugate of (E)-2,4-diene VPA with apparently l i t t l e /J-oxidation of the diene taking place. Taken together, these observations suggest that while 4-ene VPA undergoes /3-oxidation on the saturated side chain readily to give 3'-keto-4-ene VPA, the diene structure of (E)-2,4-diene VPA is somehow an impediment to /J-oxidation and hence there is less likelihood of the formation of 3-keto-4-ene VPA from (E)-2,4-diene VPA. This conclusion is supported by the relative ( E ) - 2 , 4 - d i e n e VPA ( E ) - 2 , 4 ' - d i e n e VPA \ \ 3-0H-4-ene VPA 3 « - 0 H - 4 - e n e VPA 3-keto-4-ene VPA Figure 53. Proposed mitochondrial beta-oxidation of 4-ene VPA on the propenyl (A) and propyl (B) chains (Rettenmeier et al., 1985). The compounds in brackets could theoretically be formed but were not detected in the study. amounts of the NAC conjugates of (E)-2,4-diene VPA and (E)-2,4-pentadienoic acid recovered after equivalent doses of the acids. The amount of the NAC conjugate of (E)-2,4-diene VPA recovered in urine was much higher than that of the NAC conjugate of (E)-2,4-pentadienoic acid based on a comparison of peak height and area. (E)-2,4-Pentadienoic acid is known to be efficiently /J-oxidized to give acetyl CoA and propionyl CoA (Schulz, 1983) and was metabolized to 3-oxo-4-pentenoic acid as found in the present study. These factors explain why a relatively small amount of the NAC conjugate of (E)-2,4-pentadienoic acid was recovered in urine. B. 4-Pentenoic and (E)-2,4-pentadienoic acids The metabolism studies involving 4-pentenoic and (E)-2,4-pentadienoic acids were carried out for two reasons. The first was to determine i f parallels can realistically be drawn between the metabolism of 4-ene VPA and 4-pentenoic acid and hence a similar proposed mechanism of hepatotoxicity. The second reason was to isolate and characterize spectroscopically the thiol conjugates of 3-oxo-4-pentenoic acid, the putative metabolite of 4-pentenoic acid considered to be responsible for the irreversible inhibition of the /J-ketoacyl CoA thiolase enzyme. The NAC adduct of (E)-2,4-pentadienoic acid was found to be a metabolite of both 4-pentenoic and (E)-2,4-pentadienoic acids (Figs. 42 and 46), demonstrating a similarity between 4-pentenoic acid and 4-ene VPA with respect to the metabolic fate of their diene derivatives. The putative a,^-unsaturated keto metabolite of 4-pentenoic acid, 3-oxo-4-pentenoic acid, was isolated and detected in the form of its thiol conjugates following the administration of (E)-2,4-pentadienoic acid to rats (Figs. 42-45, 47, 48). The thiol derivative of 3-oxo-4-pentenoic acid that was characterized in urine extracts was the cysteine conjugate (Fig. 42). Treatment of the urine extract with HCl gas and CH2N2 prior to GC/MS analysis not only resulted in the esterification of the NAC conjugate but also caused deacetylation to give the cysteine conjugate of 3-oxo-4-pentenoate. Similar treatment of an authentic sample of the NAC conjugate of 3-oxo-4-pentenoate also yielded the cysteine conjugate, thus verifying that the deacetylation of the NAC conjugate had occurred in vitro. The GSH conjugate of 3-oxo-4-pentenoate was detected in the bile of rats treated with (E)-2,4-pentadienoic acid. Three isomeric conjugates (Fig. 47) with similar MS/MS fragmentation patterns were identified. The bile extract was methylated with diazomethane prior to analysis and as a result the spectra indicated that in addition to the methylation of three carboxyl groups there was one additional methyl group in the molecule. The daughter ion spectra of the compounds indicated that the methyl group was attached to a carbon of the substrate moiety of the conjugate (Fig. 48). The site of this methylation is likely to be the methylene carbon between the two carbonyl groups. The protons of this methylene carbon were found to be highly acidic and readily exchanged with deuterium of D2O. Introduction of a methyl group on this a-carbon will render that carbon chiral and together with the chiral carbons of L-cysteine and L-glutamic acid will result in the formation of two diastereomers. These diastereomers probably account for two of the three isomers observed for the tetra-methylated GSH conjugate of 3-oxo-4-pentenoate. The third isomer detected can be accounted for by assuming enolization of the conjugate and methylation of the resulting hydroxyl group. The *H-NMR spectra of methyl 3-oxo-4-pentenoate did show both enol and keto forms of the compound as indicated in the experimental section and shown in App. 12. Enolization of the molecule can theoretically give rise to a pair of geometric isomers. But in this case only one isomer was clearly seen. This is not unusual since in the case of the 0-TMS ether of 3-keto VPA it was found that one of the geometric isomers accounted for 93% of the GC/MS ion current observed as mentioned in the results section. The unsaturated keto metabolite, 3-oxo-4-pentenoic acid, is generally believed to be the cause of the hepatotoxicity of 4-pentenoic acid (Schulz, 1983), although it has never been characterized in the studies conducted thus far. The results of the present study thus provide the first direct biochemical evidence for the formation of the metabolite of 4-pentenoic acid considered responsible for the irreversible inhibition of fatty acid metabolism. In comparing the metabolism of 4-ene VPA and 4-pentenoic acid or metabolites derived therefrom in rats the following differences can be noted: 1) Following equivalent doses of the two diene metabolites, the amount of NAC conjugate of (E)-2,4-pentadienoic acid recovered in urine was much less than that of the NAC conjugate of (E)-2,4-diene VPA. 2) Beta-oxidation of (E)-2,4-pentadienoic acid resulted in the formation of 3-oxo-4-pentenoic acid as detected in the form of its thiol derivatives whereas the corresponding metabolite from 4-ene VPA could not be detected. 3) In the case of 4-pentenoic and (E)-2,4-pentadienoic acids it was difficult to detect the parent compound in urine collections taken over 24 hr. In the urine of rats treated with 4-ene- or (E)-2,4-diene-VPA, the administered compounds were readily detectable. Thus the results suggest that there is a fundamental difference between the mitochondrial metabolism of 4-ene VPA and that of 4-pentenoic acid. 4. Significance of the findings of the metabolism studies to the  mechanism of the hepatotoxicity of VPA A novel metabolite of VPA, 5-(N-acetylcystein-S-yl)-3-ene VPA, has been identified in rat and human urine. 5-(Glutathion-S-yl)-3-ene VPA was detected in rat bile following the administration of either (E)-2,4-diene- or 4-ene-VPA. The identification and characterization of these thiol adducts marks the first real evidence for the formation of chemically reactive intermediates from VPA metabolism in vivo. Among the well established class of compounds that are substrates for GSH conjugation are the a,^-unsaturated carbonyl compounds (Chasseaud, 1979); although not entirely unique the formation of the GSH conjugate of (E)-2,4-diene VPA is unusual since it is a 1,6-addition rather than the 1,4-Michael addition commonly observed with a,/J-unsaturated carbonyl compounds. A high percentage of the (E)-2,4-diene VPA related material recovered from the urine of rats dosed with the diene was in the form of the NAC conjugate. Compared to the methyl ester of the diene which reacted readily with GSH to form the adduct, the in vitro reaction of the free acid of the diene with GSH was very slow. This result is similar to the finding that acrylic acid was less reactive than its esters when reacted with GSH in vitro (Sanders et al., 1988). The Michael reaction of GSH with a,/J-unsaturated acids is bound to be slow because the ionization of acids at physiological pH prevents activation of the double bond for nucleophilic attack. Because the in vitro reaction of the free acid of (E)-2,4-diene VPA with GSH was slow and yet urinary recovery of the NAC conjugate of the diene was high it is most likely that GSH reacted in vivo with an activated form of the diene, namely with the CoA ester form as shown in Scheme 5. This is a reasonable assumption based on the fact that (E)-2,4-diene VPA formed a glycine amide conjugate previously characterized by Rettenmeier et a/.(1986b) after 4-ene VPA administration to monkeys and also detected in the present study. Formation of a CoA thio ester intermediate is an obligatory step for glycine conjugation (Fears, 1985). The apparent facile reaction in vivo of (E)-2,4-diene VPA with GSH raises the question of the possible contribution of this diene to the hepatotoxicity of VPA. (E)-2,4-diene VPA is a potent hepatotoxin in the rat with one of its toxic manifestations being the induction of mitochondrial lesions (Kesterson et al., 1984). Since this diene is considered to be a mitochondrial oxidation product of 4-ene VPA CoA (Rettenmeier et al., 1985, 1986b) it appears that mitochondria are both the site of formation and the target for toxicity of (E)-2,4-diene VPA and/or (E)-2,4-diene VPA CoA. It is conceivable then that, by virtue of its electrophilic nature, (E)-2,4-diene VPA CoA may covalently bind to a nucleophilic site of a mitochondrial enzyme which would account for the potent inhibition of mitochondrial /J-oxidation of fatty acids observed for (E)-2,4-diene VPA in rats (Kesterson et al., 1984). 0 C H 2 = C H - C H 2 - C H - C - O H R 0 II C H 2 = C H - C H 2 - C H - C - S - C o A R R = C 3 H 7 0 ^ GS~ C H 2 ^ C H ^ C H ^ C ^ C - S - C o A R 0~ _ G S - C H 2 - C H = C H - C = C - S - C o A R 0 _ ll G S - C H 2 - C H = C H - C - C - S - C o A ~~ R H fl G S - C H o - C H = C H - C H - C - S - C o A ^ I R Scheme 5. Proposed mechanism for the biosynthesis of the GSH conjugate of (E)-2,4-diene VPA from 4-ene VPA. Before considering an alternate mechanism for the hepatotoxicity of (E)-2,4-diene VPA CoA, a brief description of the cellular compartmentalization of GSH seems relevant. In the liver GSH is compartmentalized to the cytoplasm (85-90%) and the mitochondrial matrix (10-15%) (Jocelyn, 1975; Reed, 1985). The presence of a discrete mitochondrial pool of GSH was first proposed by Vignais and Vignais (1973) and confirmed by other investigators (Jocelyn and Kamminga, 1974; Jocelyn, 1975; Meredith and Reed, 1982, 1983). Liver mitochondria contain a complete GSH redox system consisting of GSH, GSH reductase, GSH peroxidase, and an NADPH generating system (Reed, 1990). There is considerable experimental evidence to support the concept that mitochondrial GSH is metabolically distinct from that of the cytoplasm. The low mitochondrial rate of GSH turnover (tj/2 = 30 hr) compared to that of the cytoplasm (tj/2 = 2 hr) indicates the two pools are not free to equilibrate (Meredith and Reed, 1982). Moreover, the rate of incorporation of [^S]methionine or cysteine into the mitochondrial pool was found to be 15 times less than that of the cytoplasmic pool. In incubations containing the GSH depleting agent, diethylmaleate (185 jiM), mitochondrial levels were not changed after 60 min whereas cytopasmic levels of GSH were decreased to 40% of control values. Mitochondria depend largely on GSH peroxidase to detoxify peroxides because they do not contain catalase. Consequently GSH plays a protective role against mitochondrial cytotoxicity by providing reducing equivalents to GSH peroxidase (Meredith and Reed, 1982). Several reports indicated that cell injury occurred when cellular GSH levels fell below 10-15% of control values which is the amount of GSH associated with the mitochondria (Anundi et a/ . , 1979; Younes and Siegers, 1981; Mitchell et al., 1985; Casini et a/ . , 1985). It was shown that, although diethylmaleate depleted cytoplasmic GSH, the mitochondrial GSH level was unaltered and there was no resulting cellular injury. On the other hand, ethacrynic acid, a compound that can penetrate the mitochondrial membrane, produced marked cytotoxicity coincident with mitochondrial GSH depletion (Meredith and Reed, 1982, 1983). These findings indicated that not only is the GSH pool in the mitochondria metabolically distinct but also is the critical GSH pool influencing cytotoxicity. An alternate mechanism for the hepatotoxicity of the reactive (E)-2,4-diene VPA CoA ester could then be the localized depletion of GSH in mitochondria. Such a condition would lead to oxidative stress within the cell and result in subsequent cellular injury. The cellular pool of GSH in mitochondria is relatively small and when depleted can lead to significant cytotoxicity as discussed above. Thus a situation whereby (E)-2,4-diene VPA is formed in unusually high levels by susceptible individuals might explain the idiosyncratic nature of VPA-induced hepatotoxicity. Evidence in support of this theory was obtained from four patients described as having VPA-associated hepatotoxicity (Farrell and Abbott, 1991). All patients survived with three successfully treated with NAC. The lymphocytes from these patients demonstrated a significantly higher cell k i l l when incubated with VPA plus microsomes than did lymphocytes from controls. The addition of GSH to the incubation medium fully protected the cells. The overall evidence suggests that abnormalities in the GSH cell defence mechanisms may have played a role in the predisposition of these patients to VPA hepatotoxicity. The potential role of GSH in VPA-associated hepatotoxicity is supported by recent studies in the rat where single toxic doses of VPA were observed to significantly depress the GSH peroxidase and reductase enzymes responsible for GSH homeostasis (Cotariu et al., 1990). Thus a combination of enhanced production of one or more reactive electrophilic metabolites, and a deficiency in the ability to maintain a normal GSH level resulting either from genetic factors or a drug induced effect could well form the basis for the idiosyncratic nature of VPA hepatotoxicity. The observation that phenobarbital pretreatment of animals induces the formation of 4-ene VPA from VPA (Rettie et al., 1987) also sugggests an increased formation of the (E)-2,4-diene VPA in the induced liver. This of course would be true in the presence of an intact 0-oxidation system since (E)-2,4-diene VPA is considered to be a mitochondrial oxidation product of 4-ene VPA (Rettenmeier et al., 1985, 1986b). However, the expected increased formation of these metabolites is not always reflected by their serum or urine levels as discussed elsewhere in the thesis. This may be rationalized for example by further metabolism of (E)-2,4-diene VPA to its thiol conjugates. The data from the two patients who recovered from VPA-related hepatotoxicity tends to support this view. Our finding that (E)-2,4-diene VPA is activated in vivo to an electrophilic species that readily reacts with GSH and that the GSH conjugate of (E)-2,4-diene VPA is a metabolite of 4-ene VPA is consistent with the report of Porubek et al. (1989) of the covalent binding of 4-ene VPA to proteins of the liver. What needs to be established is whether polytherapy leads to increased urinary levels of the NAC conjugate of (E)-2,4-diene VPA in patients. The measurement of this metabolite may prove to be more indicative of exposure to a toxic species than that of a readily metabolized precursor such as 4-ene- or even (E)-2,4-diene-VPA itself. Like 4-ene VPA, the (E)-2,4-diene VPA is hepatotoxic in the rat (Kesterson et al., 1984; Granneman et al., 1984c) and was highly elevated in one reported hepatic failure case (Kochen et al., 1983). In the present study, the level of the NAC conjugate of (E)-2,4-diene VPA was found to be 3-4 times higher in two patients with hepatic failure compared to those without this side effect. Taken together, these observations point to the potential role of (E)-2,4-diene VPA in the toxicity of VPA and suggest that this diene should exhibit greater hepatotoxicity than 4-ene VPA. In a previous study (Kesterson et al., 1984) that compared the hepatotoxic properties of 4-ene- and (E)-2,4-diene-VPA in the rat, 4-ene VPA was in fact found to be the more steatogenic. This in vivo result might be explained by reasoning that (E)-2,4-diene VPA produced in situ from 4-ene VPA CoA would be present at the site of toxicity (mitochondria) in the activated form. Administered (E)-2,4-diene VPA on the other hand would be conjugated with GSH in the cytosol and made less available for intramitochondrial enzymes. Such a hypothesis needs further clarification. It has been postulated that 0-oxidation of (E)-2,4-diene VPA leads to formation of the ultimate hepatotoxin, the highly electrophilic 3-keto-4-ene VPA (Rettenmeier et al., 1985). This metabolite of VPA has yet to be detected (Rettenmeier et al., 1985, 1986b; Granneman et al., 1984a). In the present metabolic study of (E)-2,4-diene VPA neither the 3-keto-4-ene VPA itself nor its NAC conjugate could be demonstrated under our experimental conditions even though the thiol conjugates of 3-oxo-4-pentenoate were identified and characterized in rat urine and bile. In light of this finding, the contribution of (E)-2,4-diene VPA to the hepatotoxicity of VPA may be more significant than thought previously. However, the formation of 3-keto-4-ene VPA in vivo can not be completely ruled out since it may bind avidly to tissue macromolecules with thiol conjugates being below the level of detection. Future studies with (E)-2,4-diene VPA should include experiments on GSH depletion and resulting toxicity, the testing of the toxicity of (E)-2,4-diene VPA CoA using a suitable in vitro model, and the monitoring of the NAC conjugate in patients on VPA mono- and poly-therapy. It would also be important and mechanistically informative to determine i f any of the enzymes in the fatty acid oxidation pathway are inactivated (covalently modified) by the (E)-2,4-diene VPA CoA. V. SUMMARY AND CONCLUSIONS A possible mechanism for the hepatotoxicity of valproic acid was investigated based on the metabolism of this drug to reactive electrophilic intermediates. The formation of the GSH and NAC conjugates of reactive metabolites derived from VPA, 4-ene-, or (E)-2,4-diene-VPA was studied in the rat. Metabolism of the hepatotoxin 4-pentenoic acid which is an analog of 4-ene VPA was also studied. Because of the hepatotoxic potential of 4-ene- and (E)-2,4-diene-VPA in humans the levels of these metabolites were determined in the serum of epileptic patients. For this purpose a highly sensitive and selective NICI/GC/MS assay, which enabled the simultaneous quantitation of VPA and 14 of its metabolites, was developed. 4-Ene- and (E)-2,4-diene VPA were minor serum metabolites with average concentrations below 1 ng/ml. Polytherapy with CBZ and CBLZ had no significant effect on the levels of either 4-ene- or (E)-2,4-diene VPA. In the patient groups taking VPA in combination with CBZ there was an induction of w and w-1 pathways of VPA metabolism while the levels of the /J-oxidation products were significantly decreased. Several compounds were synthesized as metabolite standards. The putative metabolite of 4-ene VPA, 3-keto-4-ene VPA and its analog 3-oxo-4-pentenoic acid were synthesized as their methyl esters by an efficient two-step method. The GSH conjugates of these two a,^-unsaturated keto compounds were prepared in almost quantitative yields by adding methanolic solutions of the substrates to an alkaline aqueous solution of GSH. The GSH and NAC conjugates of methyl 3-keto-4-ene VPA and 3-oxo-4-pentenoate were fully characterized using NMR and mass spectrometry. In each case the adduct formed was the result of conjugate addition across the activated double bond. When the free acids of (E)-2,4-diene VPA and (E)-2,4-pentadienoic acid were reacted with either GSH or NAC under conditions similar to that described for the a,^-unsaturated keto metabolites, the rate of reaction was very slow and consequently l i t t l e product was obtained. The poor reactivity of the free acids appears to be due to the fact that at the pH (7.4) used for the synthesis the acids would be largely in the ionized form. This would prevent activation of the terminal double bond to Michael addition by GSH. In contrast, the methyl ester derivatives of these diene metabolites reacted relatively rapidly with GSH and NAC to give good yields of the respective adducts. Compared to the rate of reaction of the a,^-unsaturated keto metabolite standards, however, the reaction of the diene esters with GSH was slower as indicated by a 50% (in 5 hr) and 90% (in 30 min) yield for the GSH conjugates of methyl (E)-2,4-diene- and methyl 3-keto-4-ene-VPA, respectively. The structure of the GSH conjugate of methyl (E)-2,4-diene VPA was established to be (E)-5-(glutathion-S-yl)-3-ene VPA by NMR and mass spectrometry. The major GSH conjugate of (E)-2,4-pentadienoate had a similar structure with the thiolate attached to the terminal carbon and the double bond appearing at position 3. Thus, both diene analogs underwent nucleophilic addition reactions mainly via the 1,6-mechanism. The GSH conjugate of (E)-2,4-diene VPA was found as a biliary metabolite of rats dosed with 4-ene- and (E)-2,4-diene VPA. The NAC conjugate of (E)-2,4-diene VPA was a major metabolite in the urine of rats treated with these compounds. The NAC conjugate was subsequently confirmed as a metabolite of VPA in the urine of patients. The detection 178 of these thiol metabolites provided evidence that VPA is metabolized in vivo to a chemically reactive intermediate capable of reaction with cellular nucleophiles most probably in the liver. (E)-2,4-diene VPA is prochiral and consequently the interaction of the diene with GSH can produce diastereomers, geometric- or regio-isomers. Two isomers of the NAC conjugate of (E)-2,4-diene VPA were detected by NICI/GC/MS both for the synthetic standard and as metabolites in human urine. The major synthetic isomer and metabolite was spectroscopically characterized to be (E)-5-(N-acetylcystein-S-yl)-3-ene VPA. Based on retention time, the second isomer appears to be a diastereomer of the conjugate resulting from a 1,6-addition of GSH to the diene. The urinary concentration of (E)-5-(N-acetylcystein-S-yl)-3-ene VPA was determined in 28 patients on VPA therapy. Quantitation of the conjugate was carried out using an NICI/GC/MS method developed for this purpose. The mean concentration of the conjugate in the urine of these patients was 1.38 /xg/mL with a range from trace to 5.74 xxg/mL. The level of the conjugate when normalized to creatinine was found to be 3-4 times higher in two patients who had developed VPA-induced liver failure compare to those without this side effect. There was no difference in the urinary levels of the (E)-2,4-diene VPA itself among the patients studied. The metabolism studies in rats involving 4-ene- and (E)-2,4-diene-VPA were conducted primarily to isolate and characterize the thiol derivatives of 3-keto-4-ene VPA. This was based on the analogy of 4-ene VPA to 4-pentenoic acid which is metabolized to the ultimate hepatotoxin 3-oxo-4-pentenoic acid. The 3-keto-4-ene VPA or its thiol conjugates 179 could not be demonstrated as metabolites even though synthetic standards were available to facilitate their detection. Thus 3-keto-4-ene VPA either does not arise as a metabolite of 4-ene VPA, or reacts rapidly with cellular macromolecules so that thiol conjugates, i f formed, were below our limit of detection. The metabolism of 4-pentenoic acid was studied in order to examine the parallel drawn between the metabolism of 4-ene VPA and 4-pentenoic acid and to determine i f the proposed mechanism of hepatotoxicity was realistic. In rats dosed with 4-pentenoic and (E)-2,4-pentadienoic acids, the thiol conjugates of (E)-2,4-pentadienoic acid were detected, demonstrating a similarity between the metabolism of 4-pentenoic acid and 4-ene VPA. The putative a,/5-unsaturated keto metabolite of 4-pentenoic acid, 3-oxo-4-pentenoic acid, was isolated and detected in the form of its thiol conjugates following the administration of (E)-2,4-pentadienoic acid to rats. Three isomers of the GSH conjugate were detected by LC/MS/MS in methylated bile extract. The isomers appeared to arise through the enolization of the molecule and subsequent methylation of the hydroxyl group of the enol, and the introduction of a chiral center at the a-carbon of the substrate portion of the conjugate. These experiments provided the first direct biochemical evidence for the formation of the reactive metabolite of 4-pentenoic acid thought responsible for the irreversible inhibition of fatty acid metabolism. The results also indicated that there was a fundamental difference between the mitochondrial metabolism of 4-ene VPA and that of 4-pentenoic acid. In summary, a novel metabolite of VPA, the NAC conjugate of (E)-2,4-diene VPA has been identified in rat and human urine. The identification and characterization of this thiol adduct marks the first real evidence for the formation of chemically reactive intermediates from VPA metabolism in vivo. Although not entirely unique, the formation of the GSH conjugate of (E)-2,4-diene VPA is unusual since it is a 1,6-addition rather than the 1,4-Michael addition commonly observed with a,/J-unsaturated carbonyl compounds. Because the in vitro reaction of the free acid of (E)-2,4-diene VPA with GSH was slow and yet urinary recovery of the NAC conjugate of the diene in rats was high it was hypothesized that (E)-2,4-diene VPA was activated to nucleophilic addition by the formation of the CoA thio ester. The apparent facile reaction in vivo of (E)-2,4-diene VPA with GSH raises the question of the possible contribution of this diene to the hepatotoxicity of VPA. (E)-2,4-diene VPA is a potent hepatotoxin in the rat, with one of its toxic manifestations being the induction of mitochondrial lesions. Thus, i t appears that mitochondria are both the site of formation and the target for toxicity of (E)-2,4-diene VPA/(E)-2,4-diene VPA CoA. It is conceivable then that, by virtue of its electrophilic nature, (E)-2,4-diene VPA CoA may covalently bind to a nucleophilic site of a mitochondrial enzyme which would account for the reported potent inhibition of mitochondrial ^-oxidation of fatty acids by (E)-2,4-diene VPA in rats. An alternate mechanism for the hepatotoxicity of the reactive (E)-2,4-diene VPA CoA ester could be the localized depletion of GSH in mitochondria. Such a condition would lead to oxidative stress within the cell and result in subsequent cellular injury. The cellular pool of GSH in mitochondria is relatively small and when depleted can lead to significant cytotoxicity. Thus a situation whereby (E)-2,4-diene VPA is formed in unusually high levels by susceptible individuals, and a deficiency in the ability to maintain a normal GSH level, resulting either from genetic factors or a competition for detoxification when VPA is given in combination with other drugs, might explain the idiosyncratic nature of VPA-induced hepatotoxicity. V I . REFERENCES Aarsland A. , Berge R.K., Bremer J . and Aarsaether N. The hypolipidemic peroxisome-proliferating drug, bis (carboxymethylthio)-l.lO decane, a dicarboxylic metabolite of tiadenol, is activated to an acylcoenzyme A thioester. Biochim.Biophys.Acta 1033, 176-183 (1990). Abbott F.S. and Acheampong A.A. Quantitative structure-anticonvulsant activity relationships of valproic acid, related carboxylic acids and tetrazoles. Neuropharmacology 27, 287-294 (1988). Abbott F.S. , Kassam J . , Acheampong A., Ferguson S., Panesar S., Burton R., Farrell K. and Orr J . Capillary gas chromatography-mass spectrometry of valproic acid metabolites in serum and urine using tert.-butyldimethylsilyl derivatives. J . Chromatogr. 375, 285-298 (1986). Acheampong A. and Abbott F.S. Synthesis and stereochemical determination of diunsaturated valproic acid analogs including its major diunsaturated metabolite. J.Lipid Res. 26, 1002-1008 (1985). Acheampong A. , Abbott F. and Burton R. Identification of valproic acid metabolites in human serum and urine using hexadeuterated valproic acid and gas chromatographic mass spectrometric analysis. Biomed.Mass Spectrom. 10, 586-595 (1983). Acheampong A.A. , Abbott F.S. , Orr J .M. , Ferguson S.M. and Burton R.B. Use of hexadeutrated valproic acid and gas chromatography-mass spectrometry to determine the pharmacokinetics of valproic acid. J.Pharm.Sci. 73, 489-494 (1984). Al-Bassam S.S. and Sherratt H.S.A. The antagonism of the toxicity of hypoglycin by glycine. Biochem. Pharmacol. 30, 2817-2824 (1981). Anundi I. , Hogberg J . and Stead A.H. Glutathione depletion in isolated hepatocytes: Its relation to l ipid oxidation and cell damage. Acta Pharmacol.Toxicol. 45, 45-51 (1979). Baillie T.A. Metabolic activation of valproic acid and drug-mediated hepatotoxicity. Role of the terminal olefin, 2-n-propyl-4-pentenoic acid. Chem.Res.Toxicol. 1, 195-199 (1988). Bartlett K. Sodium valproate and hyperglycinaemia. Lancet 716 (1977). Becker CM. and Harris R.A. Influence of valproic acid on hepatic carbohydrate and lipid metabolism. Arch.Biochem.Biophys. 223, 381 -392 (1983). Bialer M., Hussien Z . , Raz I., Abramsky 0., Herishanu Y. and Pachys F. Pharmacokinetics of valproic acid in volunteers after a single dose study. Biopharm.Drug Dispos. 6, 33-42 (1985). Billington D., Osmundsen H. and Sherratt H.S.A. Mechanisms of the metabolic disturbances caused by hypoglycin and by pent-4-enoic acid in vitro studies. Biochem.Pharmacol. 27, 2879-2890 (1978a). Billington D., Osmundsen H. and Sherratt H.S.A. Mechanisms of the metabolic disturbances caused by hypoglycin and by pent-4-enoic acid in vivo studies. Biochem.Pharmacol. 27, 2891-2900 (1978b). Bjorge S.M. and Baillie T.A. Inhibition of medium-chain fatty acid B-oxidation in vitro by valproic acid and its unsaturated metabolite, 2-n-propyl-4-pentenoic acid. Biochem.Biophys.Res.Commun. 132, 242-252 (1985). Bohles H. , Richter K., Wagner-Thiessen E. and Schafer H. Decreased serum carnitine in valproate induced Reye syndrome. Eur.J.Pediatr. 139, 185-186 (1982). Boobis A.R., Fawthrop D.J. and Davies D.S. Mechanisms of cell death. 7"/ps 10, 275-280 (1989). Borrebaek B., Osmundsen H. and Bremer J . In vivo induction of 4-enoyl-CoA reductase by clofibrate in liver mitochondria and its effect on pent-4-enoate metabolism. Biochem.Biophys.Res.Commun. 93, 1173-1180 (1980). Brendel K., Corredor C P . and Bressler R. The effect of 4-pentenoic acid on fatty acid oxidation. Biochem.Biophys.Res.Commun. 34, 340-347 (1969). Brown N.A., Farmer P.B. and Coakley M. Valproic acid teratogenicity: demonstration that the biochemical mechanism differs from that of valproate hepatotoxicity. Biochem.Soc.Trans. 13, 75-77 (1985). Caldwell J . Xenobiotic acyl-coenzymes A: critical intermediates in the biochemical pharmacology and toxicology of carboxylic acids. Biochem.Soc.Trans. 12, 9-11 (1984). Casini A . F . , Pompella A. and Comporti M. Liver glutathione depletion induced by bromobenzene, iodobenzene, and diethylmaleate poisoning and its relation to l ipid peroxidation and necrosis. Amer.J.Pathol. 118, 225-237 (1985). Chasseaud L.F. Conjugation with glutathione and mercapturic acid excretion. In"Glutathione: Metabolism and function" (l.M.Arias and W.B.Jakoby, eds.), pp.77-114, Raven Press, New York, 1976. Chasseaud L.F. The role of glutathione and glutathione S-transferases in the metabolism of chemical carcinogens and other electrophilic agents. In "Advances in cancer research", vol.29 (G.Klein and S.Weinhouse, eds.), pp.175-274, Academic Press, New York, 1979. Cherruau B., Mangeot M., Demelier J . F . , Charpentier C , Pelletier C. and Lemonnier. Metabolic abnormalities observed in the rat after administration of sodium dipropyl acetate A. Toxicol. Lett. 8, 39-42 (1981). Collins D.J. and Tomkins CW. Acid-catalyzed isomerization of ethyl 5-hydroxy-7a-methyl-l-oxo-2,6,7,7a-tetrahydro-lH-indene-4-carboxylate. Aust.J.Chem. 30, 443-450 (1977). Cotariu D. and Zaidman J . L . Valproic acid and the liver. Clin.Chem. 34, 890-897 (1988). Cotariu D., Evans S., Zaidman J . L . and Marcus 0. Early changes in hepatic redox homeostasis following treatment with a single dose of valproic acid. Biochem. Pharmacol. 40, 589-593 (1990). Coude F.X. , Grimber G., Pelet A. and Benoit Y. Action of the antiepileptic drug, valproic acid, on fatty acid oxidation in isolated rat hepatocytes. Biochem.Biophys.Res.Commun. 115, 730-736 (1983). Coulter D.L. and Allen R.J. Hyperammonemia with valproic acid therapy. J.Pediatr. 99, 317-319 (1981). Coulter D.L. Carnitine deficiency: a possible mechanism for valproate hepatotoxicity. Lancet 689 (1984). Cregge R . J . , Herrmann J . L . , Lee C.S., Richman J . E . and Schlessinger. A convenient one flask procedure for ester alkylation R.H. Tet. Lett. 26, 2425-2428 (1973). Delbressine L . P . C , Seutter-Berlage F. and Seutter E. Identification of urinary mercapturic acids formed from acrylate, methacrylate and crotonate in the rat. Xenobiotica 11, 241-247 (1981). Devlin T.M. "Textbook of biochemistry with clinical correlations", 2nd edition, pp.356-381, John Wiley & Sons, New York, 1986. Dickinson R.G., Harland R . C , Lynn R.K., Smith W.B. and Gerber N. Transmission of valproic acid (Depakene) across the placenta: half-l i fe of the drug in mother and baby. J.Pediatr. 94, 832-835 (1979). Dommes V. and Kunau W.H. 2,4-Dienoyl coenzyme A reductases from bovine liver and Escherichia col i . J . Biol. Chem. 259, 1781-1788 (1984). Draye J.P. and Vamecq J . The inhibition by valproic acid of the mitochondrial oxidation of monocarboxylic acid and w-hydroxymonocarboxylic acids: Possible implications for the metabolism of gamma-aminobutyric acid. J . Biochem. 102, 235-242 (1987). Dreifuss F.E. and Langer D.H. Side effects of valproate. Am.J.Med. 84 (Suppl.IA), 34-41 (1988). Dreifuss F.E. Valproic acid hepatic fatalities: revised table. Neurology 39, 1558 (1989). Dreifuss F . E . , Langer D.H., Moline K.A. and Maxwell. Valproic acid hepatic fatalities. J . E . II US experience since 1984. Neurology 39, 201-207 (1989). Dreifuss F . E . , Santi H i N., Langer D.H., Sweeney K.P., Moline K.A. and Menander K.B. Valproic acid hepatic fatalities a retrospective review. Neurology 37, 379-385 (1987). Eadie M.J . , Hooper W.D. and Dickinson R.G. Valproate-associated hepatotoxicity and its biochemical mechanisms. Med.Toxicol. 3, 85-106 (1988). Farrell K. and Abbott F. The role of N-acetylcysteine in the treatment of valproate hepatotoxicity. In "Valproate associated hepatotoxicity", (J.K.Penry and R.H.Levy, eds.), Raven Press, in press. Fears R. Lipophilic xenobiotic conjugates: The pharmacological and toxicological consequences of the participation of drugs and other foreign compounds as substrates in l ipid biosynthesis. Prog.Lipid Res. 24, 177-195 (1985). Friedman M., Cavins J .F . and Wall J.S. Relative nucleophilic reactivities of amino groups and mercaptide ions in addition reactions with a,^-unsaturated compounds. J.Am.Chem.Soc. 87, 3672-3681 (1965). Gerber N., Dickinson R.G., Harland R . C , Lynn R.K., Houghton D., Antonias J . I . and Schimschock J .C. Reye-like syndrome associated with valproic acid therapy. J.Pediatr. 95, 142-144 (1979). Gillette J.R. Commentary. A perspective on the role of chemically reactive metabolites of foreign compounds in toxicity. I. Correlation of changes in covalent binding of reactivity metabolites with changes in the incidence and severity of toxicity. Biochem. Pharmacol. 23, 2785-2794 (1974). Goldman P. and Vagelos P.R. The specificity of triglyceride synthesis from diglycerides in chicken adipose tissue. J . Biol. Chem. 236, 2620-2623 (1961). Graf R., Gossrau R., Merker H .J . , Schwabe R., Stahlmann R. and Nau H. Enzyme cytochemistry combined with electron microscopy, pharmacokinetics, and clinical chemistry for the evaluation of the effects of steady-state valproic acid concentrations on the mouse. Histochemistry 83, 347-358 (1985). Gram L. and Bentsen K.D. Hepatic toxicity of antiepileptic drugs: A review. In"Current therapy in epilepsy" (M.V.Livanainen, ed.), Acta Neurol.Scand. 68(Suppl.97), 81-90 (1983). Gram L. and Bentsen K.D. Valproate. An updated review. Acta Neurol.Scand. 72, 129-139 (1985). Granneman G.R, Wang S.I., Machinist J.M. and Kesterson J.W. Aspects of the metabolism of valproic acid. Xenobiotica 14, 375-387 (1984a). Granneman G.R., Marriott T.B. , Wang S.I., Sennello L.T. and Hagen N.S. Aspects of the dose-dependent metabolism of valproic acid.In"Metabolism of antiepileptic drugs"(R.H.Levy, ed.), pp.97-104, Raven Press, New York,1984b. Granneman G.R., Wang S.I., Kesterson J.W. and Machinist J.M. The hepatotoxicity of valproic acid and its metabolites in rats. II. Intermediary and valproic acid metabolism. Hepatology 4, 1153-1158 (1984c). Grindel J .M. , Migdalof B.H. and Cressman W.A. The comparative metabolism and disposition of penfluridol-3H in the rat, rabbit, dog, and man. Drug Metab. Dispos. 7, 325-329 (1979). Gugler R. and von Unruh G.E. Clinical pharmacokinetics of valproic acid. Clin.Pharmacokin. 5, 67-83 (1980). Gugler R., Schell A. , Eichelbaum M., Froscher W. and Schulz H.U. Disposition of valproic acid in man. Eur.J.Clin.Pharmacol. 12, 125-132 (1977). Habig W.H., Pabst M.J. and Jakoby. Glutathione S-transferases W.B. The first enzymatic step in mercapturic acid formation. J. Biol. Chem. 249, 7130-7139 (1974). Hauck R.S. and Nau H. Asymmetric synthesis and enantioselective teratogenicity of 2-n-propyl-4-pentenoic acid (4-ene VPA), an active metabolite of the anticonvulsant drug, valproic acid. Toxicol. Lett. 49, 41-48 (1989). Hayasaka K., Takahashi J . , Kobayashi Y., Jinuma K., Narisawa K. and Taka K. Effects of valproate on biogenesis and function of liver mitochondria. Neurology 36, 351-356 (1986). Hiltunen J.K. and Davies. Metabolism of pent-4-enoate in rat heart E .J . Reduction of the double bond. Biochem. J. 194, 427-432 (1981). Hjelm M., Oberholzer V., Seakins J . , Thomas S. and Kay J.D.S. Valproate inhibition of urea synthesis. Lancet 923 (1987). Hjelm M., Oberholzer V., Seakins J . , Thomas S. and Kay J.D.S. Valproate-induced inhibition of urea synthesis and hyperammonemia in healthy subjects. Lancet 859 (1986). Holland P . C , Senior A.E. and Sherratt H.S.A. Biochemical effects of the hypoglycaemic compound pent-4-enoic acid and related non-hypoglycaemic fatty acids. Effects of their coenzyme A esters on enzymes of fatty acid oxidation. Biochem. J. 136, 173-184 (1973). Horie S. and Suga T. Enhancement of peroxisomal /3-oxidation in the liver of rats and mice treated with valproic acid. Biochem. Pharmacol. 34, 1357-1362 (1985). Horton R.W., Anlezark C M . , Sawaya M.C.B. and Meldrum B.S. Monoamine and GABA metabolism and the anticonvulsant action of di-n-propylacetate and ethanolamine-O-sulphate. Eur.J.Pharmac. 41, 387-397 (1977). Ishida T . , Kumagai Y., Ikeda Y., Ito K., Yano M., Toki S., Mihashi K., Fujioka T . , Iwase Y. and Hachiyama S. (8S)-(Glutathion-S-ylJdihydromorphinone, a novel metabolite of morphine from guinea pig bile. Drug Metab. Dispos. 17, 77-81 (1989). Ito M., Ikeda Y. , Arnez J . G . , Finocchiaro G. and Tanaka K. The enzymatic basis for the metabolism and inhibitory effects of valproic acid: dehydrogenation of valproyl-CoA by methyl-branched-chain acyl-CoA dehydrogenase. Biochim.Biophys.Acta 1034, 213-218 (1990). Jager-Roman E . , Deichl A. , Jakob S., Hartmann A.M., Koch S., Rating D., Steldinger R., Nau H. and Helge H. Fetal growth, major malformations and minor anomalies in infants born to women receiving valproic acid. J.Pediatr. 108, 997-1004 (1986). Jakoby W.B. and Habig W.H. Glutathione transferases. In"Enzymatic basis of detoxication, Vol.II" (W.B.Jakoby, ed.),pp.63-94, Academic Press, New York, 1980. Jakoby W.B. Glutathione transferases: An overview. In"Methods in enzymology, vol.113" (A.Meister, ed.), pp.495-499, Academic Press, New York, 1985. Jeavons P.M. Non-dose-related side effects of valproate. Epilepsia 25(Suppl.l), S50-S55 (1984). Jeavons P.M., Clark J . E . and Maheshwari M.C. Treatment of generalized epilepsies of childhood and adolescence with sodium valproate (Epilim). Develop.Med.Child.Neurol. 19, 9-25 (1977). Jezequel A.M., Bonazzi P., Novelli G., Venturini C. and Orlandi F. Early structural and functional changes in liver of rats treated with a single dose of valproic acid. Hepatology 4, 1159-1166 (1984). Jocelyn P.C. Some properties of mitochondrial glutathione. Biochim.Biophys.Acta 369, 427-436 (1975). Jocelyn P.C. and Kamminga A. The non protein thiol of liver mitochondria. Biochim.Biophys.Acta 343, 356-362 (1974). 188 Jongeneelen F . J . , Dirven H.A.A.M., Leijkkers CM. and Henderson P.T. S-phenyl-N-acetylcysteine in urine of rats and workers after exposure to benzene. J.Ana.Toxicol. 11, 100-104 (1987). Kaneko S., Otani K., Fukushima Y., Ogawa Y., Nomura Y., Ono T. , Nakane Y., Teranishi T. and Goto M. Teratogenicity of antiepileptic drugs: Analysis of possible risk factors. Epilepsia 29, 459-467 (1988). Kaplowitz N. Physiological significance of glutathione S-transferases. Gastrointest.Liver Physiol. 2, G439-G444 (1980). Kassahun K., Burton R. and Abbott F.S. Negative ion chemical ionization gas chromatography/mass spectrometry of valproic acid metabolites. Biomed. Environ. Mass Spectrom. 18, 918-926 (1989). Kende A.S. and Toder. Stereochemistry of deconjugative alkylation of ester dienolates B.H. Stereospecific total synthesis of the litsenolides. J.Org.Chem. 47, 163-167 (1982). Kesterson J.W., Granneman G.R. and Machinist J . The hepatotoxicity of valproic acid and its metabolites in rats. I. Toxicologic, biochemical and histopathologic studies. Hepatology 4, 1143-1152 (1984). Ketterer B., Coles B. and Meyer D.J. The role of glutathione in detoxication. Environ.Health Perspect. 49, 59-69 (1983). Kimura A. , Yoshida I. , Yamashita F . , Kuriya N., Yamamoto M. and Nagayama K. The occurrence of intramitochondrial Ca2+ granules in valproate-induced liver injury. J.Pediatr.Gastroenterol.Nutr. 8, 13-18 (1989). Kingsley E . , Gray P., Tolman K.G. and Tweedale R. The toxicity of metabolites of sodium valproate in cultured hepatocytes. J.Clin.Pharmacol. 23, 178-185 (1983). Knight R.H. and Young. Biochemical studies of toxic agents L. II The occurrence of premercapturic acids. Biochem. J. 70, 111-119 (1958). Kochen W., Schneider A. and Ritz A. Abnormal metabolism of valproic acid in fatal hepatic failure. Eur.J.Pediatr. 141, 30-35 (1983). Kochi H. , Hayasaka K., Hiraga K. and Kikuchi G. Reduction of the level of the glycine cleavage system in the rat liver resulting from administration of dipropyl acetic acid; an experimental approach to hyperglycinemia. Arch.Biochem.Biophys. 198, 589-597 (1979). Kosower N.S. and Kosower E.M. The glutathione status of cells. In"International review of cytology", V.54, pp.109-160, Academic Press, New York, 1978. Kuhara T . , Inoue Y. , Matsumoto M., Shinka T. , Matsumoto I., Kawahara N. and Sakura N. Markedly increased w-oxidation of valproate in fulminant hepatic failure. Epilepsia 31, 214-217 (1990). Kunau W.H. and Dommes. Degradation of unsaturated fatty acids P. Identification of intermediates in the degradation of cis-4-decenoyl-CoA by extracts of beef-liver mitochondria. Eur. J. Biochem. 91, 533-544 (1978). Laham S. and Potvin M. Biological conversion of benzaldehyde to benzylmercaptiiric acid in the Sprague-Dawley rat. Drug Chem.Toxicol. 10, 209-225 (1987). Laub M . C , Paetzke-Brunner I. and Jaeger G. Serum carnitine during valproic acid therapy. Epilepsia 27, 559-562 (1986). Lazarow P.B. The role of peroxisomes in mammalian cellular metabolism. J.Inher.Metab.Dis. 10 (Suppl. 1), 11-22 (1987). Lee R.D. Pharmacokinetics, tissue distribution, and pharmacodynamics of valproic acid and its unsaturated metabolites in rats. Ph.D. thesis, 1991. Lee R.D., Kassahun K. and Abbott F.S. Stereoselective synthesis of the diunsaturated metabolites of valproic acid. J.Pharm.Sci. 78, 667-671 (1989). Levy R.H., Rettenmeier A.W., Anderson C D . , Wilensky A . J . , Friel P.N., Baillie T.A. , Acheampong A. , Tor J . , Guyot M. and Loiseau P. Effects of polytherapy with phenytoin, carbamazepine, and stiripentol on formation of 4-ene valproate, a hepatotoxic metabolite of valproic acid. Clin.Pharmacol.Ther. 48, 225-235 (1990). Loscher W. and Nau H. Pharmacological evaluation of various metabolites and analogues of valproic acid. Anticonvulsant and toxic potencies in mice. Neuropharmacology 24, 427-435 (1985). Loscher W. and Schmidt D. Increase of human plasma GABA by sodium valproate. Epilepsia 21, 611-615 (1980). Loscher W. Anticonvulsant activity of the metabolites of valproic acid. Arch.Int.Pharmacodyn. 249, 158-163 (1981a). Loscher W. Concentrations of metabolites of valproic acid in plasma of epileptic patients. Epilepsia 22, 169-178 (1981b) Loscher W., Bohme G., Schafer H. and Kochen W. Effect of metabolites of valproic acid on the metabolism of GABA in brain and brain nerve endings. Neuropharmacology 20, 1187-1192 (1981). Lowry O.H., Rosebrough N.J . , Lewis-Farr A. and Randall R.J. Protein measurement with folin phenol reagent. J . Biol. Chem. 193, 265-275 (1951). March J . "Advanced organic chemistry", 3rd edition, pp.664-666, John Wiley & Sons, New York, 1985. Martin-Gallardo A. , Rodriguez P., Lopez M., Benavides J . and Ugarte M. Effect of dipropyl acetate on the glycine cleavage enzyme and glycine levels: a possible experimental approach to non-ketotic hyperglycinemia. Biochem. Pharmacol. 16, 2877-2882 (1983). Meredith M.J. and Reed D.J. Depletion in vitro of mitochondrial glutathione in rat hepatocytes and enhancement of l ipid peroxidation by adriamycin and l,3-bis(2-chloroethyl)-1-nitrosourea (BCNU). Biochem. Pharmacol. 32, 1383-1388 (1983). Meredith M.J. and Reed D.J. Status of the mitochondrial pool of the glutathione in the isolated hepatocyte. J . Biol. Chem. 257, 3747-3753 (1982). Meunier G., Carraz G., Meunier Y., Eymard P. and Aimart M. Proprietes pharmacodynamique de l'acide n-dipropylacetique. Therapie 18, 435-438 (1963). Millington D.S., Bohan T.P. , Roe C P . , Yergey A.L. and Liberato D.J. Valproylcarnitine: a novel drug metabolite identified by fast atom bombardment and thermospray liquid chromatography-mass spectrometry. Clin.Chim.Acta 145, 69-76 (1985). Mitchell D.B., Acosta D. and Bruckner J.V. Role of glutathione depletion in the cytotoxicity of acetaminophen in a primary culture system of rat hepatocytes. Toxicology 37, 127-146 (1985). Mitchell J .R. , Hughes H. , Lauterburg B.H. and Smith C V . Chemical nature of reactive intermediates as determinant of toxicologic responses. Drug Metab.Rev. 13, 539-553 (1982). Miyazaki C , Natsuyama K., Ichikawa M. and Gato S. Effect of VPA on cerebral amino acids: mechanism of GABA elevation and possible causal relation of VPA-induced encephalopathy and glutamine level. Chem.Pharm.Bull. 36, 3589-3594 (1988). Mortensen P.B., Gregersen N., Kolvraa S. and Christensen E. The occurrence of C6-C10-dicarboxylic acids in urine from patients and rats treated with dipropyl acetate. Biochem.Med. 24, 153-161 (1980a). Mortensen P.B., Kolvraa S. and Christensen E. Inhibition of the glycine cleavage system: hyperglycinemia and hyperglycinuria caused by valproic acid. Epilepsia 21, 563-569 (1980b). Murphy J . V . , Marqurdt K.M. and Shug A.L. Valproic acid associated abnormalities of carnitine metabolism. Lancet 820-821 (1985). Mutlib A . E . , Talaat R.E. , Slatter J.G. and Abbott F.S. Formation and reversibility of S-linked conjugates of N-(l-methyl-3,3-diphenyl propyl)isocyanate, an in vivo metabolite of N-(l-methyl-3,3-diphenyl propyljformamide, in rats. Drug Metab. Dispos. 18, 1038-1045 (1990). Muzart J . Chromium(VI)oxide-70% tert-butyl hydroperoxide, a simple catalytic system for oxidation of alcohols to carbonyl compounds. Tet.Lett. 28, 2133-2134 (1987a). Muzart J . Practical chromium(VI)oxide-catalyzed benzylic oxidations using 70% tert-butyl hydroperoxide. Tet.Lett. 28, 2131-2132 (1987b). Nakagawa Y. and Tayama S. Formation of ortho-phenylphenol glutathione conjugates in the rat liver. Xenobiotica 19, 499-507 (1989). Nau H. and Loscher W. Pharmacologic evaluation of various metabolites and analogs of valproic acid: teratogenic potencies in mice. Fund.Appl.Toxicol. 6, 669-676 (1986),. Nau H. , Rating D.,Koch S., Hauser I. and Helge H. Valproic acid and its metabolites: placental transfer, neonatal pharmacokinetics, transfer via mother's milk and clinical status in neonates of epileptic mothers. J.Pharmacol.Exp.Ther. 219, 768-776 (1981a). Nau H. , Wittfoht W., Schafer H., Jakobs C , Rating D.and Helge H. Valproic acid and several metabolites. Quantitative determination in serum, urine, breast milk and tissue by gas chromatography-mass spectrometry using selected ion monitoring. J.Chromatogr. 226, 69-78 (1981b). Nelson S.D. Metabolic activation and drug toxicity. J. Med. Chem. 25, 753-765 (1982). Nocerini M.R., Yost G.S., Carlson J .R. , Liberato D.J. and Breeze R.G. Structure of the glutathione adduct of activated 3-methylindole indicates that an imine methide is the electrophilic intermediate. Drug Metab. Dispos. 13, 690-694 (1985). Oare D.A. and Heathcock C H . Stereochemistry of the base-promoted Michael addition reaction. In "Topics in stereochemistry" (E.L.Eliel and S.H.Wilen, eds.), pp.227-407, John Wiley & Sons, New York, 1989. Ohtani Y. , Endo F. and Matsuda I. Carnitine deficiency and hyperammonemia associated with valproic acid therapy. J.Pediatr. 101, 782-785 (1982). Olson M.J . , Handler J.A. and Thurman R.G. Mechanism of zone-specific hepatic steatosis caused by valproate: inhibition of ketogenesis in periportal regions of the liver lobule. MoT.Pharmacol. 30, 520-525 (1986). Onkenhout W., Guijt G .J . , De Long H.J. and Vermeulen. Capillary gas chromatographic analysis of mercapturic acids N.P.E. Decomposition of mercapturic acids according to a "retro-Michael" reaction. J . Chromatogr. 243, 362-367 (1982). Onkenhout W., Mulder P.P.J . , Boogaard P . J . , Buijs W. and Vermeulen N.P.E. Identification and quantitative determination of mercapturic acids formed from Z- and E-l,3-dichloropropene by the rat, using gas chromatography with three different detection techniques. Arch.Toxicol. 59, 235-241 (1986). Onkenhout W., Vermeulen N.P.E., Luijten W.C.M.M. and De Jong. Electron impact mass spectrometric analysis of mercapturic acid methyl esters H.J. Fragmentation of mercapturic acid methyl esters according to retro-Michael reactions. Biomed.Mass Spectrom. 10, 614-619 (1983). Osmundsen H. and Hovik R. 0-Oxidation of polyunsaturated fatty acids. Biochem.Soc.Trans. 16, 420-422 (1988). Panesar S.K. The effect of carbamazepine on valproic acid metabolism. M.Sc. thesis, 1987. Panesar S.K., Orr J .M. , Farrell K., Burton R.W., Kassahun K. and Abbott F.S. The effect of carbamazepine on valproic acid disposition in adult volunteers. Br.J.clin.Pharmac. 27, 323-328 (1989). Pelletier S.W., Chappell R.L. and Prabhakar S. A stereoselective synthesis of racemic andrographolide lactone. J.Am.Chem.Soc. 90, 2889-2893 (1968). Penry J.K. and Dean J .C. The scope and use of valproate in epilepsy. J.Clin.Psychiatry 50(Suppl.), 17-22 (1989). Porubek D . J . , Barnes H., Theodore L . J . and Baillie T.A. Enantioselective synthesis and preliminary metabolic studies of the optical isomers of 2-n-propyl-4-pentenoic acid, a hepatotoxic metabolite of valproic acid. Chem.Res.Toxicol. 1, 343-348 (1988). Porubek D .J . , Gri11o M.P. and Baillie T.A. The covalent binding to protein of valproic acid and its hepatotoxic metabolite, 2-n-propyl-4-pentenoic acid, in rats and in isolated rat hepatocytes. Drug Metab. Dispos. 17, 123-130 (1989). Powell-Jackson P.R., Tredger J.M. and Williams R. Hepatotoxicity to sodium valproate: a review. Gut 25, 673-681 (1984). Prickett K.S. and Baillie T.A. Metabolism of unsaturated derivatives of valproic acid in rat liver microsomes and destruction of cytochrome P-450. Drug Metab. Dispos. 14, 221-229 (1986). Prickett K.S. and Baillie T.A. Metabolism of valproic acid by hepatic microsomal cytochrome P-450. Biochem.Biophys.Res.Commun. 122, 1166-1173 (1984). Rawat S., Borkowski W.J. and Swick H.M. Valproic acid and secondary hyperammonemia. Neurology 31, 1173-1174 (1981). Reed D.J. and Fariss M.W. Glutathione depletion and susceptibility. Pharmacol. Rev. 36, 25S-33S (1984). Reed D.J. Cellular defence mechanisms against reactive metabolites. In"Bioactivation of foreign compounds" (M.W.Anders, ed.), pp.71-108, Academic Press, New York, 1985. Reed D.J. Glutathione: Toxicological implications. Annu.Rev.Pharmacol .Toxicol. 30, 603-631 (1990). Rettenmeier A.W., Gordon W.P, Prickett K.S, Levy R.H. and Baillie T.A. Biotransformation and pharmacokinetics in the rhesus monkey of 2-n-propyl -4-pentenoic acid, a toxic metabolite of valproic acid. Drug Metab.Dispos. 14, 454-463 (1986b). Rettenmeier A.W., Gordon W.P., Barnes H. and Baillie T.A. Studies on the metabolic fate of valproic acid in the rat using stable isotope techniques. Xenobiotica 17, 1147-1157 (1987). Rettenmeier A.W., Gordon W.P., Prickett K.S. , Levy R.H., Lockard J .S . , Thummel K.E. and Baillie T.A. Metabolic fate of valproic acid in the rhesus monkey, formation of a toxic metabolite, 2-n-propyl-4-pentenoic acid. Drug Metab. Dispos. 14, 443-453 (1986a). Rettenmeier A.W., Howald W.N., Levy R.H., Witek D .J . , Gordon W.P., Porubek D.J. and Baillie T.A. Quantitative metabolic profiling of valproic acid in humans using automated gas chromatography/mass spectrometric techniques. Biomed. Environ. Mass Spectrom. 18, 192-199 (1989). Rettenmeier A.W., Prickett K.S., Gordon W.P., Bjorge S.M., Chang S.L. , Levy R.H. and Baillie T.A. Studies on the biotransformation in the perfused rat liver of 2-n-propyl-4-pentenoic acid, a metabolite of the antiepileptic drug valproic acid. Evidence for the formation of chemically reactive intermediates. Drug Metab. Dispos. 13, 81-96 (1985). Rettie A . E . , Boberg M., Rettenmeier A.W. and Baillie T.A. Cytochrome P-450-catalyzed desaturation of valproic acid in vitro. J. Biol. Chem. 263, 13733-13738 (1988). Rettie A . E . , Rettenmeier A.W., Howald W.N. and Baillie T.A. Cytochrome P-450 catalyzed formation of 4-ene VPA, a toxic metabolite of valproic acid. Science 235, 890-893 (1987). Rogiers V. , Vandenberghe Y. and Vercruysse A. Inhibition of gluconeogenesis by sodium valproate and its metabolites in isolated rat hepatocytes. Xenobiotica 15, 759-766 (1985). Sanders J .M. , Burka L.T. and Matthews H.B. Metabolism and disposition of n-butyl acrylate in male Fischer rats. Drug Metab. Dispos. 16, 429-434 (1988). Sawaya M.C.B., Horton R.W. and Meldrum B.S. Effects of anticonvulsant drugs on the cerebral enzymes metabolizing GABA. Epilepsia 16, 649-655 (1975). Scheffner D., Konig St., Rauterberg-Ruland I., Kochen W., Hofmann W.J. and Unkelbach St. Fatal liver failure in 16 children with valproate therapy. Epilepsia 29, 530-542 (1988). Schulz H. Inhibitors of fatty acid oxidation. Life Sci. 40, 1443-1449 (1987). Schulz H. Metabolism of 4-pentenoic acid and inhibition of thiolase by metabolites of 4-pentenoic acid. Biochemistry 22, 1827-1832 (1983). Sherratt H.S.A. and Osmundsen H. On the mechanisms of some pharmacological actions of the hypoglycaemic toxins hypoglycin and pent-4-enoic acid. A way out of the present confusion. Biochem. Pharmacol. 25, 743-750 (1976). Sherratt H.S.A. Hypoglycin, the famous toxin of the unripe Jamaican ackee fruit. Tips 7, 186-191 (1986). Silverstein R.M., Bassler G.C. and Morrill T.C. "Spectrometric identification of organic compounds", 4th edition, p.235, John Wiley & Sons, New York, 1981. St i l l W.C, Kahn M. and Mitra A. Rapid chromatographic technique for preparative separations with moderate resolution. J.Org.Chem. 43, 2923-2925 (1978). Stork G. and Guthikonda R.N. Stereoselective total synthesis of (± ) -yohimbine, (±)-psi-yohimbine and (±)-;8-yohimbine. J.Am.Chem.Soc. 94, 5109-5110 (1972). Stryer L. "Biochemistry", 3rd edition, pp.469-493, W.H.Freeman & Company, New York, 1988. Suzuki H. , Mori K., Yamada J . and Suga T. Contributions of peroxisomal beta-oxidation system to the chain shortening of N-(alpha-methylbenzyl)azelaamic acid. Biochem. Pharmacol. 39, 1975-1981 (1990). Tatsuhara T . , Muro H. , Matsuda Y. and Imai Y. Determination of valproic acid and its metabolites by gas chromatography-mass spectrometry with selected ion monitoring. J . Chromatogr. 399, 183-195 (1987). te Koppele J .M. , van der Mark E . J . , Boerrigter J .C.O. , Brussee J . , van der Gen A. , van der Greef J . and Mulder. a-Bromoisovalerylurea as model substrate for studies on pharmacokinetics of glutathione conjugation in the rat. I G.J. (Bio-)synthesis, analysis and identification of diastereomeric glutathione conjugates and mercapturates. J . Pharmacol. Exp. Ther. 239, 898-904 (1986). Tein I . and MacGregor D.L. Possible valproate teratogenicity. Arch.Neurol. 42, 291-293 (1985). Tennison M.B., Miles M.V., Pollack G.M., Thorn M.D. and Dupuis R.E. Valproate metabolites and hepatotoxicity in an epileptic population. Epilepsia 29, 543-547 (1988). Thayer. Inhibition of mitochondrial fatty acid oxidation in pentenoic acid-induced fatty liver W.S. A possible model for Reye's syndrome. Biochem. Pharmacol. 33, 1187-1194 (1984). Thurston J . H . , Carroll J . E . , Dodson W.E., Hauhart R.E. and Tasch V. Chronic valproate administration reduces fasting ketonaemia in children. Neurology 33, 1348-1350 (1983b). Thurston J . H . , Carroll J . E . , Hauhart R.E. and Schiro T.A. A single therapeutic dose of valproate affects liver carbohydrate, fat, adenylate, amino acid, coenzyme A, and carnitine metabolism in infant mice: possible clinical significance. Life Sci. 36, 1643-1651 (1985). Thurston J . H . , Carroll J . E . , Norris B . J . , Hauhart R.E. and Schiro J.A. Acute in vivo and in vitro inhibition of palmitic acid and pyruvate oxidation by valproate and valproyl-coenzyme A in liver of infant mice. Ann.Neurol.14, 384-385 (1983a). Tsaconas C. and Padieu P. Gas chromatographic-mass spectrometric assessment of the pharmacokinetics of amineptine and its main metabolite in volunteers with liver impairment. J . Chromatogr. 487, 313-329 (1989). Tsuji J . Synthetic applications of the palladium-catalyzed oxidation of olefins to ketons. Synthesis 5, 369-384 (1984). Tsuru N., Maeda T. and Tsuruoka T. Three cases of delivery under sodium valproate- placental transfer, milk transfer and probable teratogenicity of sodium valproate. Jpn.J.Psychiatr. 42, 89-96 (1988). Turnbull D.M., Bone A . J . , Bartlett K., Koundakjian P.P. and Sherratt H.S.A. The effects of valproate on intermediary metabolism in isolated rat hepatocytes and intact rats. Biochem. Pharmacol. 32, 1887-1892 (1983). van Bladeren P.J. Formation of toxic metabolites from drugs and other xenobiotics by glutathione conjugation. 7"/ps 9, 295-299 (1988). van Bladeren P . J . , Buys W., Breimer D.D. and van der Gen A. The synthesis of mercapturic acids and their esters. Eur.J.Med.Chem. 15, 495-497 (1980). van Hoof F . , Hue L . , Vamecq J . and Sherratt H.S.A. Protection of rats by clofibrate against the hypoglycaemic and toxic effects of hypoglycin and pent-4-enoate. An ultrastructural and biochemical study. Biochem. J . 229, 387-397 (1985). van Hoof F . , Vamecq J . , Draye J.P. and Veitch K. The catabolism of medium- and long-chain dicarboxylic acids. Biochem.Soc.Trans. 16, 423-424 (1988). Vermeulen N.P.E. Analysis of mercapturic acids as a tool in biotransformation, biomonitoring and toxicological studies. Tips 10, 177-181 (1989). Vignais P.M. and Vignais P.V. Fuscin, an inhibitor of mitochondrial SH-dependent transport-linked functions. Biochim.Biophys.Acta 325, 357-374 (1973). Vyas K.P., Kari P.H., Pitzenberger S.M., Halpin R.A., Ramjit H.G., Arison B., Murphy J .S . , Hoffman W.F., Schwartz M.S., Ulm E.H. and Duggan. Biotransformation of lovastatin 1 D.E. Structure elucidation of in vitro and in vivo metabolites in the rat and mouse. Drug Metab. Dispos. 18, 203-211 (1990). Warter J .M. , Brandt C.H., Marescaux C.H., Rumbach L . , Micheletti G., Chabrier G., Krieger J . and Imler M. The renal origin of sodium valproate-induced hyperammonemia in fasting humans. Neurology 33, 1136-1140 (1983). Wenz A. , Thorpe C. and Ghisla S. Inactivation of general acyl-CoA dehydrogenase from pig kidney by a metabolite of hypoglycin A. J . Biol. Chem. 256, 9809-9812 (1981). Wilder B . J . , Ramsay R.E. , Murphy J . V . , Karas B . J . , Marquardt K. and Hammond E .J . Comparison of valproic acid and phenytoin in newly diagnosed tonic-clonic seizures. Neurology 33, 1474-1476 (1983). Worley R. and Young R.N. Anionic polymerization and copolymerization of 1-cyanobutadiene. Eur.Polymer J. 8, 1355-1359 (1972). Yamada J . , Horie S., Watanabe T. and Suga T. Participation of peroxisomal /3-oxidation system in the chain-shortening of a xenobiotic acyl compound. Biochem.Biophys.Res.Commun. 125, 123-128 (1984). Yamada J . , Ogawa S., Horie S., Watanabe T. and Suga T. Participation of peroxisomes in the metabolism of xenobiotic acyl compounds: comparison between peroxisomal and mitochondrial /3-oxidation of w-phenyl fatty acids in rat liver. Biochim.Biophys.Acta 921, 292-301 (1987). Yamaoka A. , Sumimoto H., Isobe R. and Minakami S. Formation of leukotriene B4-coenzyme A ester by rat liver microsomes. Biochem.Biophys.Res.Commun. 154, 1248-1252 (1988). Younes M. and Siegers C P . Mechanistic aspects of enhanced lipid peroxidation following glutathione depletion in vivo. Chem.Biol.Interact. 34, 257-266 (1981). Zaccara C , Campostrini R., Paganini M., Messori A. , Valenza T. , Arnetoli G. and Zappoli R. Long-term treatment with sodium valproate: monitoring of venous ammonia concentrations and adverse effects. Ther.Drug Monitor. 9, 34-40 (1987). Zafrani E.S. and Berthelot P. Sodium valproate in the induction of unusual hepatotoxicity. Hepatology 2, 648-649 (1982). Zaret B.S., Beckner R.B., Marini A.M., Wagle W. and Passarelli C. Sodium valproate-induced hyperammonemia without clinical hepatic dysfunction. Neurology 32, 206-208 (1982). Zhong J . , Fong J .C. and Schulz H. Inhibition of carnitine acetyltransferase by metabolites of 4-pentenoic acid. Arch.Biochem.Biophys. 240, 524-529 (1985). Zimmerman H.J. and Ishak K.G. Valproate-induced hepatic injury: analysis of 23 fatal cases. Hepatology 2, 591-597 (1982). 198 VII. APPENDICES J L J U I i I I i i | i i i i i i i i i | i i i i i i I i i i i | i i i i | i i i i | i i i i | i i 160 140 i i | i i i i | i i i i | i i i i | i i 120 100 1 I 1 1 1 1 | 2 2 0 2 0 0 180 i i | i i i i | i i i 80 r - p - r r 60 i i i i | i i 40 i i i I i M i j i i 20 0 PPM Appendix 3. 1,3C-NMR spectrum of the dimethyl ester of the NAC conjugate of 3-keto-4-ene VPA. o Appendix 5. *H-NMR spectrum of the methyl NAC conjugate of methyl (EV 2,4-diene VPA. S I i i i i i i i i i | i i i i | i i i i ) i i i i | i i i i [ i i i i | i i i i j i i i i | i i i i | i i i i | i i i i | i i i i | i i i i | i i i i | i i i i | i i i 200 180 160 140 120 100 . 8 0 60 40 1 I 1 ' 1 ' 1 ' 20 , , , , , , 0 PPM Appendix 6. 13C-NMR spectrum of the NAC conjugate of (E)-2,4-diene VPA, dimethyl ester. ro o < i ' 1 ' 1 i ' ' 1 1 i 1 1 1 1 i i i i i i i •! i •- • * i i i i i i '' i * 12.5 12 .2 11.5 11 . B IB.5 I B .a 9 . 5 19.B B .5 8.B 7 .5 / ' I I I I I I ' I I • I I I 6 .5 6 .0 S .S PPM Appendix 7. ^-NMR spectrum of (E)-2,4-pentadienoic acid. o r f J 6.1 B.2 8.B 7.8 7.6 7.4 7.2 7.B 6. 6 6. 6 6. 4 6. 2 6 I y 1 1 ' 1—" 1 1 1 1 1 . 1 , r — S.K 5.6 5.4 5.2 S.B 4. 8 4. 6 4.« I — ' — I 1—T 4.2 4. PPM 3. 6 3.6 3.* - 1 — 1 — 1 — ' — r Appendix 8.^ H-NMR spectrum of methyl (E)-2,4-pentadienoate. V I 1 1 1 1 M • i • I ' • I I I I I I I I I I II I I I II I I M I I I I I I I I 1 I I ) I I I 1J-I I I I | " II I I I I I | I I I I I I I I I | I I I I | I I I I I I I I I | I I I I I 2 2 0 o n n 1 Q r i • " : n ••An - ~ ~ - - - ' 2 0 0 180 160 140 120 100 i i i i I i i i i I M i i I i r 80 6 0 40 20 0 PPM Appendix 10. 13C-NMR spectrum of (E)-5-(N-acetylcystein-S-yl )3-pentenoate, dimethyl ester. ro o 0  ~' i 1 i 1 i 1 i ' i ' i rr _i 1 i—'—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—|—i—|—,—|—,—|—,—|—,—|—,—.— I 6 .8 6.6 6.1 6.2 6.0 S.S "5.6 5 .4 5.2 5.2 4.8 4.6 4 .4 a. 2 4.0 3.8 3.6 3 . 4 3 2 3 0 2 8 2 6 2 4 PPM Appendix 11. AH-NMR spectrum of methyl 3-hydroxy-4-pentenoate. ro o I t PPM Appendix 12. H-NMR spectrum of methyl 3-oxo-4-pentenoate. ro i—• o 211 —1—i—.—i—i—|—•—.—i—i—|—i—i—i—i—|—i—i—i—i—|—i—i—i—i—|—.—,—i—i—|—i—j—i—i—|—i—i—i—i—r— 6.0 5.5 S.0 4.S 4.0 3.5 3.0 2.5 2.0 PPM Appendix 13. XH-NMR spectrum of 5-(N-acetylcystein-S-yl)3-oxopentanoate, mono-methyl ester. | i i i i | i i i i | n i i | m ~ r p i i i | i i i i | i~i-ri~i i i i i | u I I | i i M | i i n -pn- r rp r 240 220 200 180 160 140 120 U J n - n - [ - r r r n • i ITI j I i i i | i i i i | i I i i | I i i i | i i i i | i i I i | i l I i j i I M | i i i I j 100 80 60 40 20 0 PPM Appendix 14. 1JC-NMR spectrum of 5-(N-acetyl.cystein-S-yl)3-oxopentanoate, mono-methyl ester. 10O 302 75 c 50 25 268 0-I 200 334 320 548 474 250 300 350 400 m/z 524 488 450 500 550 600 Appendix 15. LC/MS mass spectrum (negative ion mode) of the t r i -methylated GSH conjugate of (E)-2,4-diene VPA. The ion at- m/z 548 is the (M+OOCCH3)" anion of the conjugate. ro. 1—1 co 268 100-, 283 75-c 50-CD 25-255 200 302 548 334 320 311 561 403 474 524 i x illliili-250 300 350 400 m/z 450 500 550 600 Appendix 16. LC/MS mass spectrum (negative ion mode) of a bile component (methylated) from rats dosed with (E)-2,4-diene VPA that was identified as the GSH conjugate of (E)-2,4-diene VPA. The ion at m/z 548 is the (M+OOCCH3)" anion of the fully methylated conjugate. 215 4 3 2 -iee -90 139 \ _ t L 162 106 158 194 ^ 3 JL/ 2 8 6 2 5 7 / 3 2 2 44S 3 5 9 3 9 3 4 2 5 / /• f * s r7e ^68 ^58 -48 - 3 8 Be 2 5 8 3 0 8 350 488 4 5 0 see m/z >> c CL) CU > +-> ro CU Appendix 17. NICI/GC/MS mass spectrum of the PFB derivative of NAC conjugate of (E)-2,4-diene VPA. The ion m/z 482 is the [M-181]" anion of the di-PFB derivative of the conjugate. 216 9 7 i 1 1 1 0 0 1 6 2 / • I .. 3 3 9 3 8 3 2 1 6 2 6 0 3 1 1 y y 4 4 0 4 7 7 5 6 9 5 4 5 6 8 8 1 10 1 0 0 ^90 ^ 0 7 0 : 6 0 J58 ^ 0 •30 •20 He 200 3 9 0 4 0 0 / b 4 5 / t>41 \ . / / / / \ : 5 0 0 600 m/z 7 0 0 c cu > +-> cu a: Appendix 18. NICI/GC/MS mass spectrum of the PFB derivatized NAC conjugate of (E)-2,4-pentadienoic acid. The ion at m/z 440 is the [M-181]" anion of the di-PFB derivative of the conjugate. 217 55 87 102 115 144 / 204 \ 172 264 22232 80 1 - 1—J I—r~ 160 • • ' I ' 240 266 V r l 10 100 ^90 ~-Q<d •70 ;60 J50 ^40 292 r 3 0  r 20 HL0 120 200 m/z 280 Appendix 19. PCI/GC/MS mass spectrum of the methylated cysteine conjugate of 3-oxo-4-pentenoate. The ion at m/z 264 is the [M+H]+ ion of the conjugate. 218 71 y las 128 196 / .1 234 285 -i—•—r 80 120 160 -i—•—i—'—i—1—r 200 240 280 ->—i—r 320 360 m/z 414 r l 10 J100 9^0 ^0 >> +-> 70 •*"" c cu F-60 4-> ;50 ^ 0 400 cu > cu F-30 6c r20 J10 Appendix 20. NICI/GC/MS mass spectrum of the PFB derivatized NAC conjugate of acrylic acid. The peak at m/z 414 is the [M-181]" of the di-PFB derivative of the conjugate. 219 IA C CU 1 4 0 0 0 1 2 6 0 8 -1 0 0 0 8 -8 0 0 0 -6 0 0 0 -•—1 4000H 2 0 0 0 " 24 .67 >2 . 5 5 14 1 6 ^ ^ ^ ^ ^ I ' I 1 I 1 I ' I 1 I 1 T • I ' I ' I ' I ' I 1 I ' 1 ' I ' I ' I • I . 18 2 8 2 2 2 4 2 6 2 8 3 0 3 2 3 4 3 6 3 8 T IME (min) 4 1 4 2 9 7 1 5 1 1 9 7 2 4 7 2 6 1 / \ l l . t j 3 7 5 \ 4 5 2 i • • • • i • • • • i 2 0 0 3 0 0 5 6 0 \ 6 4 4 4 9 7 I ,| 5 8 9 \ 2 4 . 7 1 rain. r l 1 0 J 1 0 0 ^90 ^ 0 •70 ^ 0 ^50 'r4Q •30 E-28 10 6 5 8 i • • > • i 4 0 0 * 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 — IA c cu +-> c cu > cu at 5 8 0 6 0 0 m/z Appendix 21. NICI/GC/MS extracted ion (m/z 414) chromatogram and spectrum from rat urine treated with (E)-2,4-pentadienoic acid. The peak at 24.67 min was identified as the NAC conjugate of acrylic acid. 44866-4eeee-36066-1 32060^ 28660J c zteee-+-> c 20303-12686^ 8888^ 4888- 2 > 14 A .22 26 28 32 34 TIME (min) Appendix 22. TIC chromatogram, in the NICI mode, of a PFB derivatized synthetic sample of (E)-2,4-diene VPA. Peaks 1 and 2 are isomeric NAC conjugates of (E)-2,4-diene VPA. Appendix 23. *H-NMR spectrum of 4'-keto-2-ene VPA ethyl ester. 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0100542/manifest

Comment

Related Items