Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Negative ion chemical ionization GCMS analysis of valproic acid and its metabolites Kassahun, Kelem 1987

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

Item Metadata

Download

Media
831-UBC_1987_A6_7 K38.pdf [ 6.4MB ]
Metadata
JSON: 831-1.0096961.json
JSON-LD: 831-1.0096961-ld.json
RDF/XML (Pretty): 831-1.0096961-rdf.xml
RDF/JSON: 831-1.0096961-rdf.json
Turtle: 831-1.0096961-turtle.txt
N-Triples: 831-1.0096961-rdf-ntriples.txt
Original Record: 831-1.0096961-source.json
Full Text
831-1.0096961-fulltext.txt
Citation
831-1.0096961.ris

Full Text

NEGATIVE ION CHEMICAL IONIZATION GCMS ANALYSIS OF VALPROIC ACID AND ITS METABOLITES by KELEM KASSAHUN B . S c . (Pharm.) Addis Ababa U n i v e r s i t y , 1981 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in THE FACULTY OF GRADUATE STUDIES Facul ty of Pharmaceutical Sciences ( D i v i s i o n of Pharmaceutical Chemistry) We accept t h i s thes is as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA August 1987 © KELEM KASSAHUN, 1987 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 1956 Main Mall Vancouver, Canada V6T 1Y3 DE-6(3/81) - 11 -ABSTRACT V a l p r o i c ac id (VPA) i s a major ant iconvulsant drug widely used in the treatment of absence s e i z u r e s . VPA i s extensive ly metabolized in humans. Several VPA metabol i tes possess ant iconvulsant a c t i v i t y and other metabol i tes are impl icated in rare but fa ta l cases of h e p a t o t o x i c i t y . A highly s e n s i t i v e and more s p e c i f i c ana ly t i ca l method was required to analyze the large number of VPA metabo l i tes , some of which are present at t race l e v e l s . The ob ject ive of th is study was to develop such a method and to make a prel iminary a p p l i c a t i o n of the method to the determination of t race VPA l e v e l s and to search for new VPA metabo l i tes . The s u i t a b i l i t y of analyzing halogenated der iva t ives of VPA and i t s metabol i tes by negative ion chemical i o n i z a t i o n (NICI) GCMS was evaluated for the desired s e n s i t i v i t y and s p e c i f i c i t y . An assay was thus developed for VPA in serum and s a l i v a based on NICI-GCMS of the pentafluorobenzyl (PFB) d e r i v a t i v e . The NIC I spectrum of the PFB ester of VPA was dominated by a s ing le fragment i o n , the m/z 143 ([M-181]") i o n . When the m/z 143 ion was monitored the lower l i m i t of detect ion was 2 ng/mL of VPA in serum or s a l i v a . Using [^Hsl-VPA as the in terna l standard, the i n t r a - and in te r -assay v a r i a t i o n s were less than 10 % at serum VPA concentrat ions of 10 to 800 ng/mL. L i n e a r i t y was observed over the concentrat ion range of 10 ng/mL to 25 ug/mL. The NICI assay was employed to quant i tate VPA in serum (total and free) and s a l i v a in f i v e healthy volunteers who took part in a drug i n t e r a c t i o n study between VPA and carbamazepine (CBZ). A total of 63 pai red s a l i v a and serum samples were analyzed by NICI-GCMS; 33 before the adminis t ra t ion of CBZ and 30 a f t e r CBZ. The % decrease in the average VPA concentrat ion a f te r CBZ was 27.91 ± 3.48, 36.85 ± 13.64, and 48.13 ± 7.70, for serum t o t a l , serum free and s a l i v a VPA, r e s p e c t i v e l y . There was a s i g n i f i c a n t reduction (p<0.025) in the average VPA concentrat ion in a l l three b i o l o g i c a l f l u i d s . The average s a l i v a to serum free VPA r a t i o was 18.92% ± 6.25 before CBZ and 16.37% ± 2.82 fo l lowing CBZ. The average s a l i v a to serum tota l VPA r a t i o was 2.43% ± 0.86 before CBZ and 1.67% ± 0.50 fo l lowing CBZ, i n d i c a t i n g that the s a l i v a to serum tota l VPA r a t i o was concentrat ion dependent. A strong c o r r e l a t i o n was found between s a l i v a and both serum free (r = 0.9035 ± 0.0784) and serum tota l VPA (0.9058 ± 0.0450) (af ter CBZ). The free f r a c t i o n of VPA did not increase a f te r CBZ administrat ion suggesting that the decrease in VPA concentrat ion a f te r CBZ was not re la ted to changes in the free f r a c t i o n of VPA. PFB der iva t i ve formation of VPA metabol i tes was f a c i l e and resu l ted in uniform d e r i v a t i z a t i o n of a l l metaboli tes s tud ied . In the NICI mass spectra most of the ion current was c a r r i e d by the [M-181]" fragment i o n , the only exception being that of 3-keto VPA. The base peak in the NICI spectrum of PFB der iva t i zed 3-keto VPA was [M-181-C02]". Isolated metabol i tes were i d e n t i f i e d with the help of twin ions (deuterated and undeuterated) in the mass spectra and by comparison of mass spectra and retent ion times with synthet ic reference compounds. Urine or serum metabol i tes were analyzed in one chromatographic run and SIM chromatograms obtained. Serum and urine con t ro ls showed no i n t e r f e r i n g peaks and the a n a l y t i c a l method appears su i tab le for a sens i t i ve assay of VPA metabol i tes . - iv -The N I C I method employing PFB d e r i v a t i v e s was s e n s i t i v e enough to detect VPA metabol i tes in s a l i v a . Seven metabol i tes were detected. The r a t i o of Z to E isomers of 2-ene VPA was much greater in s a l i v a than in serum (3.82 v s . 0 .458) , suggesting d i f f e rences in the t ransport or plasma prote in binding proper t ies of these two isomers. A new VPA metabo l i te , assigned the st ructure 4 ' - ke to -2 -ene VPA was detected in u r ine . The mass spectrum and retent ion time of th is new metabol i te matched that of one compound which was present in a synthet ic mixture conta in ing 4 ' -ke to -2 -ene VPA. Another new metaboli te which appears to be 2 - ( 2 ' - p r o p e n y l ) - g l u t a r i c acid was a lso detected in u r i n e . The synthesis of 4 ' -ke to -2 -ene VPA was attempted using two d i f f e r e n t synthet ic methods. The f i r s t method which involved the dehydrogenation of the 0-TMS d ia lky l ketene acetal of ethyl 2-propyl-4-oxopentanoate apparently resul ted in the formation of the p o s i t i o n a l isomer, 4-keto-2-ene VPA. The second synthet ic route was based on the dehydration of 4 -carboethoxy-2 -e thy leneth ioketa l -5 -hydroxyheptane and produced 4 ' -ke to -2 -ene VPA. However, i t was not p o s s i b l e to i s o l a t e s u f f i c i e n t product for NMR c h a r a c t e r i z a t i o n . - V -TABLE OF CONTENTS Page ABSTRACT i i LIST OF TABLES v i i i LIST OF FIGURES x LIST OF SCHEMES x i i i LIST OF ABBREVIATIONS xiv ACKNOWLEDGEMENT x v i i I. INTRODUCTION 1 A. Pharmacokinetics of VPA 2 B. Metabolism of VPA 3 C. Ant iconvulsant a c t i v i t y of VPA metabol i tes 5 D. T o x i c i t y of VPA and i t s metabol i tes 7 E . In teract ion between VPA and carbamazepine 10 F. Ana ly t i ca l methods for VPA and i t s metabol i tes 11 G. Negative ion chemical i on i za t ion 13 1. Ion forming react ions in NICI 14 2. Negative ion reagent gas systems 14 3. S e n s i t i v i t y of NICI 16 4. Factors which determine the s e n s i t i v i t y of sample detect ion in NICI 16 H. Object ives 17 II. EXPERIMENTAL 18 A. Chemicals and Mater ia ls 18 1. General 18 2. VPA metabol i tes and internal standards 19 B. Instrumentation 20 1. Gas chromatography mass spectrometry 20 2. GC-Electron Capture Detection 21 3. Other instruments 21 - v i -Page C. Human Study 21 D. Ana lys is of Samples 22 1. Serum and s a l i v a standards 22 2. Sample preparat ion 24 3. D e r i v a t i z a t i o n 25 4. Serum free l e v e l s 26 5. Comparison of the s e n s i t i v i t y of EI (t-BDMS) with NICI (PFB, bis-TFMB) 26 6. I d e n t i f i c a t i o n of VPA metabol i t e s 27 E. Chemical Synthesis 27 1. Attempted synthesis of 2 - ( 2 ' - o x o p r o p y l ) - 2 -pentenoic ac id (4 ' -keto-2-ene VPA) v ia ethyl 2-propyl-4-oxopentanoate 27 2. Synthesis of 2 - (2 1 -oxopropy l ) -2 -penteno ic ac id s t a r t i n g with a protected 4-oxopentanoic ac id 31 3. Synthesis of ethyl 2-propyl-3-oxopentanoate 35 4. Synthesis of ethyl 2-propyl-3-hydroxypentanoate 35 III. RESULTS AND DISCUSSION 36 A. NICI-GCMS Assay Development 36 1. D e r i v a t i z a t i o n 36 2. GC-Electron Capture Detect ion 37 3. Opt imizat ion of MS parameters for NICI 40 4. Comparison of the r e l a t i v e s e n s i t i v i t y of d e r i v a t i z e d VPA by EI and NICI 45 5. Quant i ta t ive ana lys is with the PFB d e r i v a t i v e 48 B. VPA l e v e l s in serum (free and to ta l ) and s a l i v a before and a f ter the adminis t ra t ion of CBZ 56 1. E f f e c t of CBZ on serum and s a l i v a l e v e l s of VPA 60 2. Serum f ree VPA l e v e l s 74 3. S a l i v a VPA l e v e l s 76 - v i i -Page C. I d e n t i f i c a t i o n of VPA metabol i tes using NICI-GCMS of t h e i r PFB d e r i v a t i v e s 93 1. Negative ion spectra of PFB d e r i v a t i z e d VPA metabol i tes 95 2. PFB as an e lec t ron capture NICI-GCMS d e r i v a t i v e for VPA metabol i tes 112 3. NICI (PFB) versus EI (t-BDMS) spectra of VPA metabol i tes 113 4. Selected ion chromatograms 114 5. VPA metabol i tes in s a l i v a 119 6. Detect ion of new VPA metabol i tes 123 D. Synthesis 131 1. Attempted synthesis o f 2 - ( 2 ' - o x o p r o p y l ) - 2 -pentenoic ac id (4 ' -keto-2-ene VPA) v ia ethyl 2-propyl-4-oxopentanoate 131 2. Synthesis of 4 ' -ke to -2 -ene VPA s t a r t i n g with a protected 4-oxopentanoic ac id 135 SUMMARY AND CONCLUSIONS 147 REFERENCES 150 APPENDIX 161 - v i i i -LIST OF TABLES Comparison of the r e l a t i v e s e n s i t i v i t i e s of three VPA d e r i v a t i v e s Comparison of [^Hsl-VPA and OA as internal standards Serum VPA l e v e l s Ug/mL) in two subjects on VPA steady-s ta te as measured by EI (t-BDMS) and NICI (PFB) Serum t o t a l , serum free and s a l i v a concentrat ions (ng/mL) of VPA and the i r r e l a t i o n s h i p to each other in volunteer W.T. before the administ rat ion of CBZ Serum t o t a l , serum free and s a l i v a concentrat ions Ug/mL) of VPA and the i r r e l a t i o n s h i p to each other in volunteer W.T. a f te r the administ rat ion of CBZ Serum t o t a l , serum free and s a l i v a concentrat ions (ug/mL) of VPA and the i r r e l a t i o n s h i p to each other in volunteer M.S. before the administ rat ion of CBZ Serum t o t a l , serum free and s a l i v a concentrat ions (ug/mL) of VPA and t h e i r r e l a t i o n s h i p to each other in volunteer M.S. a f te r the adminis t ra t ion of CBZ Serum t o t a l , serum free and s a l i v a concentrat ions Ug/mL) of VPA and t h e i r r e l a t i o n s h i p to each other in volunteer R.M. before the adminis t ra t ion of CBZ Serum t o t a l , serum free and s a l i v a concentrat ions Ug/mL) of VPA and the i r r e l a t i o n s h i p to each other in volunteer R.M. a f te r the administ rat ion of CBZ Serum t o t a l , serum free and s a l i v a concentrat ions Ug/mL) of VPA and t h e i r r e l a t i o n s h i p to each other in volunteer B.A. before the the adminis t ra t ion of CBZ Serum t o t a l , serum free and s a l i v a concentrat ions Ug/mL) of VPA and t h e i r r e l a t i o n s h i p to each other in volunteer B.A. a f t e r the administ rat ion of CBZ - ix -LIST OF TABLES (CONT'D) Table Page 8a Serum t o t a l , serum free and s a l i v a concentrat ions 69 (ug/mL) of VPA and t h e i r r e l a t i o n s h i p to each other in volunteer F . A . before the administrat ion of CBZ 8b Serum t o t a l , serum free and s a l i v a concentrat ions 70 (ug/mL) of VPA and t h e i r r e l a t i o n s h i p to each other in volunteer F . A . a f te r the administ rat ion of CBZ 9 Time-averaged ra t ios (6-8 samples) and c o r r e l a t i o n s 71 between serum t o t a l , serum free and s a l i v a VPA concentrat ions in f i v e volunteers before the administ rat ion of CBZ 10 Time-averaged r a t i o s (6-8 samples) and c o r r e l a t i o n s 72 between serum t o t a l , serum free and s a l i v a VPA concentrat ions in f i v e volunteers a f ter the administ rat ion of CBZ 11 Decrease (%) in average VPA concentrat ion 75 a f te r CBZ adminis t ra t ion 12 Ions (m/z) monitored in NICI mode for VPA and 116 [ 2 H6]-VPA metabol i tes X -LIST OF FIGURES F igure Page 1 Metabolic pathways of va lp ro ic ac id in human 6 2 Total ion current p lo t of the PFB esters of VPA (a) 38 and 0A(b) in the NICI mode 3 Typica l chromatograms of PFB d e r i v a t i v e s of VPA(a) 41 and 0A(b) obtained with GC-ECD 4 C a l i b r a t i o n curve for PFB d e r i v a t i z e d VPA in ethyl 42 acetate obtained using GC-ECD. 5 NICI mass spectra of the PFB (A) and bis-TFMB (B) 46 der iva t ives of VPA. 6 EI mass spectra of the PFB (A) and bis-TFMB (B) 47 der iva t ives of VPA. 7 SIM chromatograms of VPA, OA (m/z 143) and [ 2 He]-VPA 50 (m/z 149) from serum spiked with these substances. 8 C a l i b r a t i o n curve for serum total VPA. 52 9 C a l i b r a t i o n curve for serum free VPA. 53 10 C a l i b r a t i o n curve for s a l i v a VPA. 54 11 C a l i b r a t i o n curve for EI(t-BDMS) determination of 55 serum tota l VPA. 12 SIM chromatogram of PFB d e r i v a t i z e d VPA obtained 57 with 10 pg of VPA extracted from serum 13 Re la t ionsh ip between VPA concentrat ions (ug/mL) in 59 serum determined by EI (t-BDMS) and NICI (PFB) 14 S a l i v a concentrat ion- t ime p r o f i l e s for f i v e 77 volunteers at steady s ta te VPA 15 Concentrations time curves for seum t o t a l , 78 serum free and s a l i v a VPA in one volunteer (M.S.) 16 Concentrat ion-t ime curve for serum free VPA 79 before and a f te r CBZ adminis t ra t ion in one volunteer (W.T. ) . 17 Concentrat ion-t ime curve for s a l i v a VPA before and 80 a f t e r CBZ adminis t ra t ion in one volunteer (W.T. ) . 18 Re la t ionship between serum total and s a l i v a VPA 81 concentrat ions in one volunteer (W.T. ) . - x i -LIST OF FIGURES (CONT'D) F igure Page 19 Re la t ionsh ip between serum free and s a l i v a VPA 82 concentrat ions in one volunteer (W.T.) 20 The r e l a t i o n s h i p between serum total and s a l i v a VPA 83 concentrat ions in a l l f i v e volunteers 21 The r e l a t i o n s h i p between serum free and s a l i v a VPA 84 concentrat ions in a l l f i ve volunteers 22 A p l o t of free f r a c t i o n versus serum total concentrat ion 85 of VPA 23 Total ion current p l o t , in the NICI mode, of the 94 PFB der iva t i zed urine exract from a volunteer on VPA steady s t a t e , a lso given selected doses of [ ^ l - V P A 24 NICI mass spectrum of VPA ( 2 H 0 + z%) PFB esters 96 25 NICI mass spectrum of 3-ene VPA ( 2 H 0 + 2 He) 97 PFB esters 26 NICI mass spectrum of (E)-2-ene VPA ( 2 H Q + 2%) 98 PFB es te rs 27 NICI mass spectrum of 2 ,4 -d iene VPA ( 2 H 0 + ^ 5 ) 99 PFB es te rs 28 NICI mass spectrum of ( E , E ) - 2 , 3 ' - d i e n e VPA ( 2 H 0 + 100 2 H6) PFB esters 29 A) NICI mass spectrum of 4 ' -ke to -2 -ene VPA ( 2 H Q + 101 2 H3) PFB e s t e r s , B) EI (t-BDMS) mass spectrum of 4 ' -ke to -2 -ene VPA 30 NICI mass spectrum of 3-keto VPA ( 2 H Q + 2 H6) 102 PFB esters 31 NICI mass spectrum of 3-0H VPA ( 2 H 0 + 2%) 103 PFB esters 32 NICI mass spectrum of 4-keto VPA ( 2 H Q + ^ 3 ) 104 PFB esters 33 NICI mass spectrum of 4-0H VPA ( 2 H 0 + 2 H6) 105 PFB esters 34 NICI mass spectrum of 5-OH VPA ( 2 H 0 + 2 H 5 ) 106 PFB esters - x i i -LIST OF FIGURES (CONT'D) F igure Page 35 NICI mass spectrum of 2-PSA ( 2 H Q + % ) 107 PFB esters 36 NICI mass spectrum of 2-PGA ( 2 H 0 + % ) 108 PFB esters 3 7 A) Mass chromatograms at m/z 141 1 0 9 B) NICI mass spectrum of 4 -ene VPA PFB ester 3 8 EI(t-BDMS) mass spectrum of 3-OH VPA 115 3 9 SIM chromatograms of the PFB der iva t ives of VPA 117 and [ 2 H6l-VPA metabol i tes in a urine ext ract 40 SIM chromatograms of the PFB der iva t ives of serum 118 VPA and [ 2 He]-VPA metabol i tes 41 SIM chromatograms of the PFB der iva t ives of 120 VPA and [ 2 H6]-VPA metabol i tes in a s a l i v a ex t ract 42 NICI (A) and EI (B) mass spectra of VPA re la ted 124 material in urine that appears to be 2 - ( 2 ' - p r o p e n y l ) -g l u t a r i c ac id 4 3 EI mass spectra of A) 4 ' - k e t o - 2 - e n e VPA and B) 4 -keto 127 VPA extracted from urine without a l k a l i n e treatment 44 NICI mass spectra of A) 4 ' - k e t o - 2 - e n e VPA and B) 4 -keto 128 VPA extracted from urine without a l k a l i n e treatment 45 Mass chromatograms (m/z 2 1 3 ) of the t-BDMS 136 der iva t i ves of synthesized 2 - p r o p y l - 4 - o x o - 2 -pentenoic ac id ( 4 - ke to - 2 - ene VPA) 46 Mass spectra of the t-BDMS der iva t ives of the isomers 137 of 2 - p r o p y l - 4 - o x o - 2 - p e n t e n o i c acid 47 EI mass spectra of the isomers of the ethyl esters of 141 4 ' - k e t o - 2 - e n e VPA 4 8 Mass chromatograms at m/z 2 1 3 of t-BDMS d e r i v a t i z e d 143 synthet ic and nat ive 4 ' - k e t o - 2 - e n e VPA 4 9 Mass spectra of the t-BDMS d e r i v a t i v e s of synthet ic and 144 nat ive 4 ' - k e t o - 2 - e n e VPA 50 Mass chromatograms at m/z 155 of the PFB d e r i v a t i v e s 145 of synthet ic and i s o l a t e d 4 ' - k e t o - 2 - e n e VPA x i i i -LIST OF SCHEMES Scheme Page 1 O r i g i n of the m/z 113 anion in the NICI mass 111 spectrum of PFB d e r i v a t i z e d 3-keto VPA 2 Proposed fragmentation pathway for the t-BDMS 126 de r i va t i ve of a new VPA metaboli te assigned the s t ruc ture 4 ' -ke to -2 -ene VPA 3 Synthet ic route for 2 -propyl -4-oxopentanoic ac id 132 4 Attempted synthesis of 2 - (2 ' -oxopropy l ) -2 -penteno ic 134 ac id 5 Synthet ic route for 2 - (2 ' -oxopropy l ) -2 -penteno ic 139 ac id - x i v -VPA [ 2 H 6 ] -VPA 2 , 3 ' - d i e n e VPA 2 ,4 -d iene VPA 4 , 4 ' - d i e n e VPA 2- ene VPA 3- ene VPA 4- ene VPA 3- keto VPA 4- keto VPA 3- OH VPA 4- OH VPA 5- OH VPA 4 ' -ke to -2 -ene VPA 4-keto-2-ene VPA 3 ' -ke to -4 -ene VPA 3-keto-4-ene VPA 4 ' -0H-2-ene VPA 4 ' -0H-4-ene VPA bis-TFMB Bp CBZ CI CSF DDQ LIST OF ABBREVIATIONS Va lpro ic ac id (2-propylpentanoic acid) [ 2 H6] -Va lp ro ic ac id 2-(1 ' -propeny l ) -2 -pentenoic ac id 2 -propy l -2 ,4 -pentad ieno ic acid 2 - (2 ' -p ropeny l ) -4 -penteno ic ac id 2 -propyl -2 -pentenoic acid 2 -propyl -3 -pentenoic ac id 2 -propyl -4 -pentenoic acid 2-propyl -3-oxopentanoic ac id 2-propyl -4-oxopentanoic acid 2-propyl-3-hydroxypentanoic ac id 2-propyl -4-hydroxypentanoic acid 2-propyl -5-hydroxypentanoic ac id 2 - (2 ' -oxopropy l ) -2 -penteno ic acid 2 -propy l -4 -oxo-2 -pentenoic ac id 2 - (1 ' -oxopropy l ) -4 -penteno ic acid 2 -propy l -3 -oxo-4 -pentenoic ac id 2 - (2 ' -hydroxypropy l ) -2 -pentenoic acid 2-(2 ' -hydroxypropyl ) -4 -pentenoic ac id 3 ,5 -b i s ( t r i f luoromethy l )benzy l b o i l i n g point carbamazepine chemical i o n i z a t i o n cerebrospinal f l u i d 2 , 3 - d i c h l o r o - 5 , 6 - d i c y a n o - l , 4 - b e n z o q u i n o n e - XV -LIST OF ABBREVIATIONS (CONT'D) DMAP 4-dimethylaminopyridine E trans ECD e lec t ron capture detect ion EI e lec t ron impact eV e lec t ron v o l t s GC gas chromatography GCMS gas chromatography mass spectrometry I.D internal diameter IR in f ra red LDA l i th ium di isopropylamide L i t . 1i terature m mul t i pi et MHz megahertz MS mass spectrometer MSTFA N-methyl-N-tr imethyl s i l y l t r i f luoroacetamide m/z mass to charge r a t i o NICI negative ion chemical i o n i z a t i o n NMR nuclear magnetic resonance OA octanoic ac id PFB pentaf l uorobenzyl PFBB pentafl uorobenzyl bromide q quadruplet r c o r r e l a t i o n c o e f f i c i e n t s s i n g l e t SIM selected ion monitoring 6 chemical s h i f t - xv i -LIST OF ABBREVIATIONS (CONT'D) t t r i pi e t t-BDMS t e r t i a r y b u t y l d i m e t h y l s i l y l t-BDMSCl t e r t i a r y b u t y l d i m e t h y l s i l y l c h l o r i d e THF tetrahydrofuran TIC total ion current TMS t r i m e t h y l s i l y l Z c i s - x v i i -ACKNOWLEDGEMENT I am very great ful to Dr. F. Abbott for his exce l l en t superv is ion and support throughout the course of the study. Specia l thanks go to Mr. R. Burton f o r h is valuable ass is tance in the gas chromatography - mass spectrometry work. I am a lso thankful to members of my research committee, Dr. J . Axel son , Dr. H. B u r t , Dr. K. McErlane and Dr. J . Orr for the i r time and guidance. I would a lso l i k e to acknowledge the var ious ass is tances by my lab mates, R. Lee, G. S l a t t e r , S . Panesar and M. Lee. The typing serv ice provided by Mrs. S. Rodgers and Mrs. A . Vance i s . s i n c e r e l y acknowledged. The f i n a n c i a l support in the form of a Fe l lowship provided by the World Health Organizat ion i s g r e a t f u l l y acknowledged. - 1 -I. INTRODUCTION V a l p r o i c ac id (d i -n -propy l ace t i c a c i d , 2-propylpentanoic a c i d , VPA) i s a major ant iconvulsant drug now in use throughout the world. VPA has been known since 1881 but i t s ant iconvulsant proper t ies were demonstrated much l a t e r by Meunier et a l . (1963). S t r u c t u r a l l y VPA i s a simple branched f a t t y ac id and hence i t d i f f e r s from the usual a n t i e p i l e p t i c drugs in that i t lacks nitrogen and a r ing s t r u c t u r e . Its s t ruc tura l formula i s : CH 3 - CH 2 - CH 2 CH - COOH CH 3 - CH 2 - CH 2 VPA i s useful in a var ie ty of se izures inc lud ing primary general ized s e i z -ures of the p e t i t mal and myoclonic types, t o n i c - c l o n i c seizures and p a r t i a l s e i z u r e s . The prec ise mode of act ion of VPA remains uncer ta in , although i t has been suggested that VPA exerts i t s act ions through e f fec ts on gamma-aminobutyric a c i d . VPA i s extensive ly metabolized in humans and experimental animals. The metabolism of VPA i s very complex; so far 17 metabol i tes have been i d e n t i f i e d in humans. Several of these metabolites possess ant iconvulsant proper t ies and other metabol i tes are thought to be involved with rare but fa ta l hepatotox ic i ty associated with VPA therapy. In view of the ant iconvulsant a c t i v i t y and/or potent ia l t o x i c i t y of the metabol i tes , there i s a great deal of i n t e r e s t in studying VPA metab-o l i s m . Some of the VPA metabol i tes are found in minor quant i t i es and t race metabol i tes might not have been i d e n t i f i e d because of the lack of - 2 -s e n s i t i v i t y of cur ren t ly used ana ly t i ca l methods. Hence, there i s a need fo r the development of convenient , highly s e n s i t i v e and s p e c i f i c methods of a n a l y s i s . The purpose of th is work was to develop a highly s e n s i t i v e and s p e c i f i c method of ana lys is and apply the method to measure trace VPA l e v e l s and to search for new VPA metabol i tes that may be present at trace l e v e l s . A. Pharmacokinetics of VPA The pharmacokinetics of VPA have been extens ive ly studied both in humans and animals and have been reviewed by Gugler and von Unruh (1980), Schobben et a l . (1980), M o r s e l l i and Franco-Morsel l i (1980), and Rimmer and Richens (1985). A f te r oral admin is t ra t ion , VPA i s rap id ly and almost completely absorbed, peak plasma l e v e l s being at ta ined within one to four hours (Schobben et a l . , 1980). In terms of absolute b i o a v a i l a b i l i t y d i f f e r e n t formulat ions of the drug appear to be b ioequ iva len t . Therapeutic plasma l e v e l s are general ly 50 to 100 ug/mL with some pat ients requ i r ing plasma concentrat ions in excess of 100 u.g/mL (Brum' and Wi lder , 1979). There i s a s i g n i f i c a n t r e l a t i o n s h i p between the reduc-t i o n in the number of se izures and increas ing serum VPA l e v e l s (Gram et a l . , 1979). It has a lso been found that the r e l a t i o n s h i p between dose and plasma concentrat ion i s c u r v i l i n e a r , i . e . the plasma concentrat ion to dose r a t i o decreases with increas ing doses. Studies in rodents have shown that VPA i s d i s t r i b u t e d r a p i d l y , reaching the bra in in a few minutes (Vajda, 1983). The apparent volume of d i s t r i b u t i o n i s in the range of 0.1 to 0.4 L/kg (Gugler et a l . , 1977; - 3 -Perucca et a l . , 1978). Th is small volume of d i s t r i b u t i o n ind icates that VPA i s d i s t r i b u t e d only to the c i r c u l a t i o n and rap id ly exchangable extra-c e l l u l a r water. VPA does not appear to be bound to i n t r a c e l l u l a r proteins in bra in nor i s taken up s e l e c t i v e l y by the bra in of humans (Goldberg and Todoro f f , 1980). VPA i s highly bound to plasma prote ins (average 90%) (Gugler and Mue l l e r , 1978; Loscher , 1978). Par t l y because of the high plasma prote in b ind ing , VPA concentrat ion in the cerebrospinal f l u i d is 10% of that in plasma and the s a l i v a concentrat ion of VPA ranges from 0.4 to 6% of the plasma concent ra t ion . Prote in binding of VPA i s concentrat ion dependent and the free f r a c t i o n var ies two-fold within the therapeut ic range (Levy et a l . , 1986). The plasma clearance of VPA ranges from 5 to 10 mL/min (Klotz and Anton in , 1977; Gugler et a l . , 1977). In c h i l d r e n higher clearance values are found which may be explained by greater volumes of d i s t r i b u t i o n (Schobben et a l . , 1980). The plasma e l iminat ion h a l f - l i f e i s in the range of 9 to 18 hours in monotherapy and i s 6 to 12 hours when VPA i s administered with other a n t i e p i l e p t i c drugs (Levy et a l . , 1986). B. Metabolism of VPA Metabolism i s the major means for the e l iminat ion of VPA and renal excre t ion of the unchanged drug accounts for less than 5% of the adminis-tered dose (Gugler et a l . , 1977). In sp i te of i t s simple structure the metabolism of VPA i s very complex and complete e l u c i d a t i o n of i t s metabo-l i c pathways has proved e l u s i v e . The metabolism of VPA has been exten-s i v e l y studied in man and var ious animals and was i n i t i a l l y reviewed by Gugler and von Unruh (1980). More recen t ly , in th is laboratory , the - 4 -metabol i te 4-keto VPA was i d e n t i f i e d and 2 - p r o p y l s u c c i n i c ac id and 2-propylmalonic ac id were charac te r i zed as VPA metabol i tes using deuterated t racers and GCMS ana lys is (Acheampong et a l . , 1983). VPA undergoes g lucuron ida t ion , f J - , w - and ( oo- l ) -ox idat ion to produce a large number of metabol i tes (Loscher , 1981a; Granneman et a l . , 1984a). The human metabol ic pathways of VPA are summarized in F igure 1. In both man and r a t , B-oxidat ion and glucuronidat ion are the two primary pathways. In a s ingle dose study in man 15 to 20% of the administered dose was excreted as VPA glucuronide ( B i a l e r et a l . , 1985). B-Oxidat ion of VPA gives r i s e to 2-ene VPA, 3-OH VPA and 3-keto VPA with 2-ene VPA and 3-Keto VPA being the major metabol i tes in plasma (Nau and Loscher , 1984). 3-Keto VPA i s also a major ur inary metabol i te and i s considered to a r ise pr imar i ly as a r e s u l t of mitochondrial o x i d a t i o n . The w- ox idat ion pathway leads to 5-0H VPA, 2 - p r o p y l g l u t a r i c ac id and 2-propylmalonic a c i d . The ( O J - 1 ) - ox idat ion pathway r e s u l t s in the formation of 4-OH VPA, 4-keto VPA and 2 - p r o p y l s u c c i n i c a c i d . Products of a- and (o ) - l ) - o x i d a t i o n are found in minor quant i t i es in the serum of e p i l e p t i c pat ients (Abbott et a l . , 1986a). The OJ and (oj-1) pathways are cytochrome P-450 mediated (Pr icke t t and B a i l l i e , 1984). By administer ing 5-0H VPA to the r a t , 2-propyl gl u t a r i c acid was shown to be the terminal product of the w-ox idat ion pathway and s i m i l a r l y adminis t ra t ion of 4-0H VPA resu l ted in the production of (^ -1 ) metabol i tes (Granneman et a l . , 1984a). The unsaturated metaboli tes 3-ene VPA and 4-ene VPA were not observed a f te r treatment with 4-0H VPA and 5-0H VPA and administrat ion of these unsaturated metabol i tes produced n e g l i g i b l e amounts of hydroxy metabo l i tes . Thus 3-ene VPA and 4-ene VPA do not belong to the OJ and - 5 -pathways and are thought to o r ig ina te through dehydrogenation of VPA. These unsaturated metabol i tes are fur ther metabolized to produce dienes VPA. One of these diene metabol i tes i s a major serum metabolite and has been assigned the s t ructure ( E , E ) - 2 , 3 ' - d i e n e VPA by Acheampong and Abbott (1985). In addi t ion to the above known metabolic pathways, mul t ip le minor metabol ic pathways are thought to operate in the metabolism of VPA. For example, P r i c k e t t and B a i l l i e (1984) showed that incubat ion of VPA with ra t l i v e r microsomes led to the formation of 3 - ,4 - and 5-OH VPA. The 3-OH VPA was thought to be a product of B-oxidat ion but the above study suggests that 3-OH VPA can al so be formed by cytochrome P-450 dependent o x i d a t i o n . More recent ly Ret t ie et a l . (1987) have demonstrated that cytochrome P-450 ca ta lyzes the formation of 4-ene VPA from VPA. C . Ant iconvulsant a c t i v i t y of VPA metabol i tes VPA d isp lays a l a te onset of a n t i e p i l e p t i c e f f e c t s (Rowan et a l . , 1979; Henriksen and Johannessen, 1980) and a car ry -over e f f e c t a f te r drug admin is t ra t ion i s d iscont inued (Lockard and Levy, 1976). These observa-t i o n s suggest that ac t ive metabol i tes may be formed which accumulate in the b r a i n . In a study by Nau and Loscher (1982) of the pharmacokinetic and pharmacological proper t ies of VPA and 2-ene VPA in the mouse, i t was found that 2-ene VPA was c leared from the plasma and brain slower than the parent drug i n d i c a t i n g that 2-ene VPA may contr ibute to the ant iconvulsant e f f e c t of chronic VPA therapy. The ant iconvulsant a c t i v i t y of several VPA metabol i tes has been s tud ied using d i f f e r e n t animal models of ep i lepsy (Loscher , 1981b; Loscher and Nau, 1983; Keane et a l . , 1985; Loscher and Nau, 1985). In one of C H , - C H , - C H , C K J - C H J - C H J CHCOOGlu VALPROIC ACID OH C H , - C H - - C H , 6-OH VPA 1 HOOC-CH,-CH, CHCOOH ;CHCOOH CHj -CM 2 -CH 2 2 Z-PropylaluUrlc »ctd CHj—CM?—CH2 CHj—CH2—CH2 CHCOOH OH CH j -CH -CH 2 CHCOOH C H j - C H 2 - C H 2 A-OH VPA CHj—C—CH2 C H j - C H 2 - C H 2 4-Keto VPA CHCOOH C H 2 - C H - C H 2 CHCOOH C l i p CH-CH 2 CH 3 —CH ? -CH 2 4-ene VPA C H j - C M - C H ^ y 2 M ' - d l e n e VPA CHCOOH CH ? =CH-CH 2 CHCOOH C H ^ C H - C H ^ C H j - C H ^ CH^ C-COOH C H j - C H 2 - C H 2 f (t) 2.4-dle 3-ene VPA \ / ne VPA C H 3 _ C H 2 - C H ^ C H 3 - C H 2 - C H ^ 2-ene VPA • ,OH C H j - C H 2 — C H ^ C-COOH C H j - C ^ - C H ^ 3-OH VPA CH-COOH CHj—CH=CH^ ^ C - C O O H H 2(t) .3' ( t ) -d lene VPA CH — C H 2 — C  -COOH C H j - C H 2 - C H 2 - C H ^ COOH HOOC-CH, C H j - C H ^ C H ^ CHCOOH C H j - C H ^ C ' ^ C H J - C H 2 - C H 2 / CHCOOH 2-PronyliMlonlc «c ld 2-Propyltucclnlc «ctd 3-Keto VPA Figure 1. Metabolic pathways of v a l p r o i c a c i d in human (Abbott et a l . , 1986b). - 7 -these studies (Loscher , 1981b) the ant iconvulsant a c t i v i t y of VPA metabol i tes was determined by measuring t h e i r e f f e c t s on the thresholds f o r the maximal e l e c t r o c o n v u l s i on and the penty lenetet razole induced convuls ions in mice. Of the tested metabol i tes 2-ene VPA and 4-ene VPA were the most potent d i s p l a y i n g 50 to 90% of the potency of VPA. The other metabol i tes t e s t e d , i . e . 3-OH VPA, 3-keto VPA, 5-0H VPA, 2 - p r o p y l g l u t a r i c a c i d , 3-ene VPA and 4-OH VPA also gave r i s e to s i g n i f i c a n t threshold e l e v a t i o n s . Another study (Loscher and Nau, 1983) has shown that although several VPA metabol i tes were present in plasma of dogs and rats a f te r acute and long term treatment with VPA only 2-ene VPA was found in the bra in of both animals. The 2-ene VPA accumulated in some brain regions during chronic treatment and was found to be 1.3 times more potent than the parent drug when c a l c u l a t i o n was based upon whole bra in concentra-t i o n s . Furthermore, s tudies with the trans isomer of 2-ene VPA ind ica te a comparable ant iconvulsant p r o f i l e with that of the parent drug without the po ten t ia l for embryotoxici ty associa ted with the l a t t e r (Loscher et a l . , 1984). In a recent study Acheampong and Abbott (unpublished data) have shown that 2 , 3 ' - d i e n e VPA has an ant iconvulsant a c t i v i t y comparable to that of 2-ene VPA whereas 4-keto VPA was i n a c t i v e . The ant iconvulsant property of ac t ive VPA metabol i tes does not appear to be super ior to that of VPA. However, 2-ene VPA or 2 , 3 ' - d i e n e VPA could be an a l t e r n a t i v e to VPA i f i t can be shown that e i ther i s not associated with the hepato tox ic i ty caused by VPA therapy. D. T o x i c i t y of VPA and i t s metabol i tes The common adverse e f f e c t s with VPA therapy are nausea, vomit ing, g a s t r o i n t e s t i n a l d is turbance , thrombocytopenia and behavioral disturbance - 8 -(Schmidt, 1984). P a n c r e a t i t i s has a lso been re la ted to VPA therapy (Wyl l ie e t a l . , 1984). The most ser ious tox ic e f f e c t of VPA i s the hepatotox ic i ty which appears to be an i d i o s y n c r a t i c react ion in a small population of p a t i e n t s . The VPA-induced hepatic t o x i c i t y can assume two d i f f e r e n t forms (Gram and Bentsen, 1983). The f i r s t form is associated with an increase in l i v e r enzymes and appears to be dose r e l a t e d . The other form const i tu tes an i r r e v e r s i b l e l i v e r damage. Its frequency has been estimated as 1 in 20,000 pat ients (Jaevons, 1984). The c l i n i c a l symptoms of the l i v e r t o x i c i t y inc lude hepa toce l lu la r necros is and microves icu la r s t e a t o s i s . The l a t t e r i s s i m i l a r to that observed in Reye's syndrome and Jamaican vomiting s ick -ness (Gerber et a l . , 1979). The hepato tox ic i ty of VPA i s be l ieved to be associated with the mono and/or doubly unsaturated metaboli tes of VPA. Kochen et a l . (1984) noted an increased formation of diunsaturated metabol i tes in pat ients with s i d e - e f f e c t s as opposed to pat ients without s i d e - e f f e c t s . The 4 , 4 ' - d i e n e VPA which has never been observed before was detected along with 4-ene VPA in one pat ient who died from hepatic f a i l u r e (Kochen et a l . , 1983). Most f a t a l hepato tox ic i ty cases have been in mul t ip le drug therapy (Drei fuss and S a n t i l l i , 1986). Th is may be due to an increased formation of one or more of the tox ic metabol i tes . In r a t s , Granneman et a l . (1984b) found that phenobarbital coadminist rat ion caused s i g n i f i c a n t increases in the plasma l e v e l s of 4-ene VPA and 5-OH VPA. The most l i k e l y metabol i te to be involved with the l i v e r t o x i c i t y i s 4-ene VPA. The 4-ene VPA i s s t r u c t u r a l l y s i m i l a r to the metaboli te of - 9 -hypoglycin A that i s responsib le for Jamaican vomiting s i c k n e s s , and 4-pentenoic acid which produces a fa t ty l i v e r in the rat (Nau and Loscher , 1984). The mechanism by which 4-ene VPA may cause hepatotox ic i ty i s not known. It has been postulated that VPA and unsaturated metaboli tes cause t h e i r hepatotox ic i ty by i n h i b i t i o n of the B-oxidat ion pathway (Kesterson e t a l . , 1984). VPA causes a m i l d , t rans ien t i n h i b i t i o n of the B-oxidat ion pathway by sequestrat ion of CoA, while 4-ene VPA is thought to cause a prolonged and potent i n h i b i t i o n due to the formation of 4-ene VPA-CoA. The 4-ene VPA-CoA might be a potent i n h i b i t o r of a s p e c i f i c enzyme(s) in the B-oxidat ion system. Furthermore, mu l t ip le biochemical react ions may r e s u l t from the numerous VPA metabol i tes . High l e v e l s of unusual keto ac ids such as 3- and 4-keto VPA might a lso a f f e c t B-oxidat ion by i n t e r f e r i n g with the B -ke to -acy l th io lase enzyme (Kesterson et a l . , 1984). The metabo l i tes , 4-ene VPA and 4-OH VPA have been found to be tox ic in cu l tu red rat hepatocytes (Kingsley et a l . , 1983). In general terms VPA causes various metabolic disturbances because i t i n h i b i t s several enzymes involved in intermediary c e l l metabolism. Thurston et a l . (1985) have reported that a s ing le therapeut ic dose of VPA a f f e c t s the metabolism of carbohydrates, fa ts and amino ac ids in in fant mice. A l s o , a recent study (Turnbull et a l . , 1986) i n d i c a t e s that one gram of VPA given o r a l l y causes metabolic disturbances in normal humans. In a study that sought to address the mechanism of the t o x i c i t y of 4-ene VPA Rettenmeier et a l . have detected 3-0H-4-ene VPA as one of the metabol i tes of 4-ene VPA in the perfused rat l i v e r (1985) and in the Rhesus monkey (1986a). These authors postulate that the detect ion of - 10 -3-0H-4-ene VPA i s i n d i r e c t evidence for the formation of 3-keto-4-ene VPA from 4-ene VPA. They suggest that 4-ene VPA is metabo l ica l ly act iva ted to 3-keto-4-ene VPA which i s capable of a l k y l a t i n g mitochondrial p r o t e i n s . The 3-keto-4-ene VPA l i k e other a ,B-unsaturated carbonyl compounds i s h ighly reac t ive and can undergo Michael addi t ions to give covalent adducts of nucleophi les (Eder et a l . , 1982). Te ra togen ic i t y i s the second major tox ic e f f e c t of VPA (Brown et a l . , 1985). VPA i s teratogenic and embroyotoxic in r a b b i t s , rats and mice (Petrere et a l . , 1986). VPA crosses the placenta and can a f f e c t the fetus (Dickinson et a l . , 1979). Publ ished case reports of fe ta l malformation in e p i l e p t i c mothers on VPA descr ibe various malformations inc lud ing spina b i f i d a (Rimmer and Richens, 1985). The mechanism of VPA te ra togen ic i ty i s unknown but Brown et a l . (1985) have suggested that the biochemical mechanism of VPA t e r a t o g e n i c i t y d i f f e r s from that of the hepa to tox ic i t y . E . In teract ion between VPA and carbamazepine Carbamazepine (CBZ) drug i n t e r a c t i o n s have been recent ly reviewed by Baciewicz (1986). Because the biotransformat ion of CBZ i s inducib le as well as suscept ib le to i n h i b i t i o n , and CBZ i s prote in bound to a large extent (Levy and Koch, 1982), i n te rac t ions between VPA and CBZ are to be expected. To date there have been c o n f l i c t i n g reports of poss ib le i n t e r a c t i o n s between the two. Bowdle et a l . (1979) demonstrated a decrease in CBZ minimum steady-s ta te concentrat ion when VPA was administered concomitant ly with CBZ. In another study (Levy et a l . , 1984a), where VPA was given for one week to seven e p i l e p t i c pat ients rece iv ing chronic CBZ, i t was found that s teady-state CBZ l e v e l s were reduced by 3-59% in six - 11 -pat ien ts and unchanged in one pa t ien t . On the other hand several studies (P isan i et a l . , 1981; Brodie et a l . ,1983; McKauge et a l . , 1981) did not f i n d s i g n i f i c a n t d i f f e rences in plasma l e v e l s of CBZ when CBZ and VPA were administered concomitant ly . The e f f e c t of CBZ on VPA plasma l e v e l s has also been s tud ied . Bowdle et a l . (1979) observed that minimum steady-s ta te concentrat ions of VPA dec l ined and c learance increased when CBZ was given to normal volunteers on VPA s t e a d y - s t a t e . S i m i l a r l y , Hoffman et a l . (1981) found that the h a l f - l i f e of VPA was reduced from 15 to 6 . 9 hours and c learance increased from 8.0 to 13.7 mt/min when the two drugs were given together . In the present study the e f f e c t of CBZ on the prote in binding ( in vivo) of VPA w i l l be invest iga ted in f i v e healthy volunteers as a part of a general study of the e f f e c t of CBZ on VPA metabolism. F. A n a l y t i c a l methods for VPA and i t s metabol i tes For the ana lys is of VPA, separat ion of the drug from b i o l o g i c a l f l u i d s i s requ i red , usua l ly by a c i d i f y i n g and ext rac t ing the serum or ur ine sample with organic so lven ts . The most e f f i c i e n t solvent for the ex t rac t ion of VPA and i t s metabol i tes i s ethyl acetate (Abbott et a l . , 1986a). Many methods have been described in the l i t e r a t u r e for the ana lys is o f VPA in b i o l o g i c a l f l u i d s . These include high-performance l i q u i d chrom-atography (Sutheimer et a l . , 1979; A l r i c et a l . , 1981; Kl ine et a l . , 1982; Moody and A l l a n , 1983; Nakamura et a l . , 1984; Kushida and I s h i z a k i , 1985), enzyme immunoassay (H igg ins , 1983; Siegmund et a l . , 1981), gas chromatography, and gas chromatography - mass spectrometry. - 12 -GC determination of VPA has been by far the most common method. VPA has been assayed by GC under ivat ized (Loscher, 1977; Kwong et a l . , 1980; Freeman and Rawal, 1980; Berry and C l a r k e , 1978). The drug has also been quant i ta ted as the methyl ( C a l e n d r i l l o and Reynoso, 1980), butyl (Hulshoff and Roseboom, 1979), tr imethyl s i l y l (Loscher, 1981a), phenacyl (Gupta et a l . , 1 9 7 9 ) , hexaf luoroisopropyl (Nishioka and Kawai, 1983) or t -bu ty l d i m e t h y l s i l y l ( Abbott et a l . , 1982) d e r i v a t i v e s . The phenacyl es te r of VPA has a lso been analyzed by GC with e lec t ron capture detect ion (Chan, 1980). The GCMS methods include the quant i ta t ion of VPA as i t s methyl ester (von Unruh et a l . , 1980), i d e n t i f i c a t i o n of VPA metabol i tes using the methyl es te r and t -buty l d i m e t h y l s i l y l d e r i v a t i v e s (Acheampong et a l . , 1983), simultaneous ana lys is of VPA and eight of i t s metaboli tes using t r i m e t h y l s i l y l d e r i v a t i v e s (Nau et a l . , 1981) and quant i ta t ion of 2-3- and 4-ene VPA as t h e i r t r i m e t h y l s i l y l ester (Rettenmeier et a l . , 1986b). A chemical i o n i z a t i o n (CI) GCMS assay of the ethyl esters of VPA metabol i tes (Granneman et a l . , 1984a), a CI GCMS method for the determination of VPA (Balkon, 1979) and a d i r e c t i n s e r t i o n CI method (Schier et a l . , 1980) have a l s o been reported. The most complete GCMS assay i s that of Abbott et a l . (1986a) which enables the simultaneous determination of VPA and 12 metabol i tes in a s i n g l e chromatographic run. Th is assay i s based upon selected ion monitor ing of the e lec t ron impact i o n i z a t i o n of t -bu ty l dimethyl s i l y l d e r i v a t i v e s of the drug and i t s metabol i tes . The assay i s r e l a t i v e l y s e n s i t i v e and s p e c i f i c . However, the large number of VPA metabol i tes , - 13 -some of which are present at very low concentrat ions (e s pe c ia l l y those impl ica ted in h e p a t o t o x i c i t y ) , the poss ib le in ter fe rence of endogeneous f a t t y a c i d s , make the search for an even more s e n s i t i v e and s p e c i f i c method of ana lys is necessary. The above mentioned assay of Abbott et a l . has a lower l i m i t of detect ion of 0.1 ng/ml. The serum l e v e l s of some VPA metabol i tes (4-ene VPA, 3-ene VPA, 2 ,4-diene VPA, 5-0H VPA, 4-OH VPA) are f requent ly near the lower detect ion l i m i t s . In th is work a highly s e n s i t i v e and more s p e c i f i c ana lys is for VPA and a l l i t s metabol i tes was to be developed. The method was to be based on the technique of e lec t ron capture negative ion chemical i o n i z a t i o n GCMS. Because of t h i s i o n i z a t i o n technique, at l e a s t one order of magnitude increase in s e n s i t i v i t y over any current method was expected. The GCMS s p e c i f i c i t y would be fur ther enhanced because of the sof t i o n i z a t i o n nature of negative ion chemical i o n i z a t i o n . The need for a more s e n s i t i v e and s p e c i f i c a n a l y t i c a l method cannot be overemphasized. Such a method w i l l be valuable in the study of the metabolism of VPA metabol i tes in small animals in order to e luc ida te metabol ic pathways and in the detect ion of intermediary metabol i tes which may be responsib le for the hepatotox ic i ty of VPA therapy. G. Negative ion chemical i o n i z a t i o n In a number of l abora to r i es there has been a recent in te res t in using negative ion chemical i o n i z a t i o n (NICI) mass spectrometry coupled with GC for the ana lys is of c e r t a i n fa t ty a c i d s . This has been e s p e c i a l l y t rue fo r the determination of prostanoids as t h e i r pentafluorobenzyl d e r i v a t i v e s . The technique of NICI mass spectrometry i s a r e l a t i v e l y new - 14 -technique and has been used to solve s t ructura l and ana ly t i ca l problems only during the l a s t decade. 1. Ion forming react ions in NICI Under conventional EI c o n d i t i o n s , 70eV e l e c t r o n s , and source pressures in the range of 10" 5 to 10" 7 t o r r , formation of negative ions occurs by the ion pai r mechanism and i s dominated by low mass fragment ions (Hunt et a l . , 1976 ). Under CI c o n d i t i o n s , i . e . at source pressures of about 1 t o r r , negative ions can be produced in two ways (Dougherty, 1981; Watson, 1985): a . E lec t ron /molecu le react ions i Resonance e lec t ron capture AB + e" • AB" (<0.1eV) i i D i s s o c i a t i v e e lec t ron capture AB + e" »• A- + B" (0-15eV) i i i Ion-pa i r formation AB + e" *> A" + B + + e - (>10eV) b. Anion/molecule react ions AB + C" * ABC" or (AB-H)" + HC 2. Negative ion reagent gas systems a . Bronsted-base reagent systems These reagent substances play a s i m i l a r ro le as the reagent gases in p o s i t i v e CI and r e s u l t in ion/molecule r e a c t i o n s . The Bronsted-base - 15 -reagent systems include H", N H 2 " , O H - , 0 T , 0 2 T , C H 3 O - , F" and C T (Har r i son , 1983). They react e i ther by proton abst ract ion or adduct format ion. b. E lec t ron capture reagent systems The capture of e lec t rons by a molecule i s a resonance process which requi res e lec t rons of near-thermal energy (Har r i son , 1983). With a high source pressure the simplest type of process which leads to negative ion formation i s where the reagent gas acts only as a moderating gas to produce a high populat ion of thermal energy e lec t rons which are captured by sample molecules with some e lec t ron a f f i n i t y . The reagent gas can also act in the capaci ty of c o l l i s i o n a l s t a b i l i z a t i o n of the newly formed negative ions (Hass, 1980). C lasses of compounds that have i n t r i n s i c a l l y high NICI s e n s i t i v i t y are general ly o x i d i z i n g and a l k y l a t i n g agents (Dougherty, 1981). P o s i t i v e e l e c t r o n a f f i n i t i e s are observed for many halogenated compounds, quinones and n i t r o compounds (Howe et a l . , 1981). For molecules which lack e lect ron capture c a p a b i l i t y , d e r i v a t i z a t i o n of the molecule with a su i tab le d e r i v a t i z i n g agent that endows the molecule with p o s i t i v e e lec t ron a f f i n i t y i s p o s s i b l e . Der iva t ives such as pentafluorobenzaldehyde (to form a S c h i f f base with aromatic amines), pentaf luorobenzoyl ha l ide ( for phenols and amines) and t e t r a f l u o r o p h t h a l i c anhydride ( for amines) have been used for NICI mass spectrometry. Pentaf luorobenzyl bromide has been employed as the d e r i v a t i z i n g reagent for NICI GCMS ana lys is of fa t ty acids and prostag land ins . - 16 -3. S e n s i t i v i t y of NICI I t can be shown that the ion currents obtained in CI are general ly as intense as those observed in EI (Har r ison , 1983). Furthermore, the CI ion cur rent may be concentrated in a few i o n s . The s e n s i t i v i t y of CI systems i s dependent upon k, the rate constant of the CI r e a c t i o n . E f f i c i e n t CI reac t ions w i l l show bet ter or equal s e n s i t i v i t i e s to those of E I . I n e f f i c i e n t i o n i z a t i o n react ions i . e . those with smal ler rate constants w i l l have lower s e n s i t i v i t i e s . Compounds that are amenable to e lec t ron capture as opposed to i o n / molecule react ions in NICI can have high rate constants as a r e s u l t of the high mob i l i t y of the e l e c t r o n . Th is can r e s u l t in extraordinary s e n s i t i v i t y in tha t , when a molecule possesses both a p o s i t i v e e lect ron a f f i n i t y and large cross sect ion for e lect ron capture , the negative ion spectrum which depends upon e lec t ron capture can exh ib i t up to 100 times the s e n s i t i v i t y found with other i o n i z a t i o n techniques. 4. Factors which determine the s e n s i t i v i t y of sample detect ion in NICI. The formation of negative ions by e lec t ron capture i s st rongly depen-dent upon the e lec t ron a f f i n i t y of the ana ly te , the energy of the e l e c t r o n s e f f e c t i n g the i o n i z a t i o n and the degree of c o l l i s i o n of molecule ions with neutra ls (Dougherty, 1981). The s e n s i t i v i t y with which a sample can be detected, the re fo re , depends upon the extent to which newly formed and exc i ted anions can be s t a b i l i z e d by reagent gas molecules (Chapman, 1985). C o l l i s ional processes can a lso lead to e lec t ron e jec t ion and hence there w i l l be an optimal source pressure for maximum s e n s i t i v i t y . In a d d i t i o n , the r e l a t i v e importance of c o l l i s ional s t a b i l i z a t i o n to e lect ron detachment w i l l depend upon the internal energy of the reagent gas - 17 -molecules and hence s e n s i t i v i t y w i l l be strongly in f luenced by source temperature (Hass, 1980). Furthermore, e lec t ron absorbing impur i t ies such as halogenated solvents can deplete the thermal e lec t rons in the ion source and t h i s r e s u l t s in a drop in s e n s i t i v i t y . High s e n s i t i v i t i e s can not be maintained as the concentrat ion of substrate molecules increases s ince the number of thermal e lec t rons in the ion source i s f i n i t e . L inear response ranges must be determined because non- l inear response may s ta r t as low as 10 ng in some cases (Stout , 1984). H. Object ives 1. The main ob ject ive of th is study was to develop a highly s e n s i t i v e and s p e c i f i c method for the detect ion and quant i ta t ion of trace leve l VPA and i t s metabol i tes . The s u i t a b i l i t y of halogenated der iva t ives and negative ion chemical i o n i z a t i o n GCMS were to be evaluated for ach iev ing the des i red a n a l y t i c a l method. 2. The a n a l y t i c a l method was to be used to measure VPA in serum ( to ta l and free) and in s a l i v a in f i v e healthy volunteers who p a r t i c i p a t e d i n a drug i n t e r a c t i o n study between VPA and carbamazepine. The e f f e c t of carbamazepine on the free f r a c t i o n of VPA was to be determined and the u t i l i t y of measuring s a l i v a r y concentrat ion of VPA in a drug in te rac t ion study evaluated. 3. A search for new VPA metabol i tes was to be done using the twin ion technique and employing both e lec t ron impact ( t - b u t y l d i m e t h y l s i l y l d e r i v a t i v e s ) and negative ion chemical i o n i z a t i o n (pentafluorobenzyl d e r i v a t i v e s ) mass spectrometry. Potent ia l new VPA metabol i tes and known metabol i tes were to be synthesized (as required) fo r use as reference standards. - 18 -II. EXPERIMENTAL A . Chemicals and M a t e r i a l s . 1. General Chemicals were reagent grade and obtained from the fol lowing sources . a . A l d r i c h Chemical Co. (Milwaukee, Wisconsin) : 3 , 5 - B i s ( t r i f l u o r o m e t h y l ) b e n z y l bromide, Boron t r i f l u o r i d e e thera te , t -Buty ld imethyl s i l y l c h l o r i d e , n -Buty l l i th ium (1.6 M in hexane), Calcium hydr ide , 18-Crown-6, 2 , 3 - D i c h l o r o - 5 , 6 - d i c y a n o - l , 4 - b e n z o q u i n o n e , Di i s o p r o p y l -amine, D i isopropy le thy l amine, 4-Dimethylaminopyridine, 1 , 2 - E t h a n e d i t h i o l , Isopropylcyclohexylamine, Li thium aluminium hydr ide , Methanesulfonyl c h l o r i d e , Pentanoic a c i d , Potassium hydride (35% d ispers ion in mineral o i l ) , Propionyl c h l o r i d e , Sodium hydride (50% d ispers ion in mineral o i l ) , Tetrahydrofuran, T r ie thy l amine. b. A l f a Products (Danvers, Massachusetts): Pentaf luorobenzyl bromide c . BDH Chemicals (Toronto, Ontar io ) : Acetone, A c e t o n i t r i l e , Benzene, C i t r i c ac id (anhydrous), Ether (anhydrous), Hydrochlor ic a c i d , Potassium i o d i d e , Sodium hydroxide, Sodium su l fa te (anhydrous), S u l f u r i c a c i d . d . B r i t i s h Drug House (Poole , U .K . ) : Iodoethane, P y r i d i n e . - 1 9 -e. Caledon Laborator ies L t d . (Georgetown, Ontar io ) : Dichloromethane, E thano l , Ethyl acetate . f . Eastman Kodak Co. (Rochester , New York) : Ethyl acetoacetate , 4-0xopentanoic a c i d , Propionaldehyde. g . F i s h e r S c i e n t i f i c Co. (Fa i r lawn, New J e r s e y ) : Bromine, t - B u t a n o l , Cadmium carbonate. h. Ma l l inkrodt Chemicals (St . L o u i s , M i s s o u r i ) : Potassium carbonate (anhydrous), Sodium bicarbonate . i . Matheson Coleman and Be l l Co. (Norward, Ohio): C h l o r o t r i m e t h y l s i l a n e , Phosphorus t r ib romide , 2 , 4 , 6 - T r i -methy lpyr id ine . j . N ichols Chemical Company (Montreal , Canada): Mercuric c h l o r i d e , k. P ie rce Chemical Company (Rockford, I l l i n o i s ) : N - M e t h y l - N - t r i m e t h y l s i l y l t r i f l u o r o a c e t a m i d e . 2. VPA metabol i tes and in terna l standards D i -n -p ropy l ace t ic ac id (VPA) was obtained from ICN Biochemicals Inc. K+K Labs. (P la inv iew, New York) . The internal standard octanoic ac id (OA) was purchased from N u t r i t i o n a l Biochemicals Corporat ion (C leve land , Ohio) . The synthesis of the other in terna l standard used, [ ^ J - V P A has been reported (Acheampong et a l . , 1984). The VPA metabol i tes used as reference standards were obtained from th is labora-tory and t h e i r synthesis has been publ ished (Acheampong et a l . , 1983). These metabol i tes included 4-ene VPA, 3-ene VPA, 2-ene VPA, 4-OH VPA, 5-OH VPA, 2 - p r o p y l g l u t a r i c ac id and 2 - p r o p y l s u c c i n i c a c i d . The synthe-s i s of 2 ,4 -d iene VPA and 2 , 3 ' - d i e n e VPA w i l l be reported elsewhere. - 20 -B. Instrumentation 1. Gas Chromatography Mass Spectrometry a . C a p i l l a r y Column GCMS C a p i l l a r y column GCMS a n a l y s i s was performed on a Hewlett-Packard 5987A gas-chromatograph mass spectrometer with an RTE-6 data system. E lec t ron impact spectra were obtained at e lect ron energy of 70eV, and ion source pressure of 1.8 x I O - 6 T o r r . EI GCMS ana lys is of the t-BDMS d e r i v a t i v e s was done under the fo l lowing c o n d i t i o n s : OV-1701 bonded phase column, 25 m x 0.32 mm I.D. with a f i lm thickness of 0.25 u (Quadrex S c i e n t i f i c , New Haven, Connect icu t ) ; oven temperature, 50°C to 100°C at 3 0 ° / m i n , 100°C to 260°C at 8 ° / m i n ; source temperature, 240°C; open s p l i t i n t e r f a c e , 250°C; i n j e c t i o n port temperature, 240°C; helium flow ra te , 1 mL/min. Negative ion chemical i o n i z a t i o n spectra were recorded at 120-170eV depending upon the value of the source pressure at which the instrument was tuned. Source pressure was about 1 t o r r . Operating condi t ions were: oven temperature, 50°C to 140°C at 3 0 ° / m i n ; 140°C to 250°C at 5 ° / m i n ; reagent gases, methane, ammonia, argon-methane; source temperature, 2 0 0 ° C ; open s p l i t i n t e r f a c e , 250°C; i n j e c t i o n port temparture, 240°C; c a r r i e r gas, helium at a flow rate of 1 mL/min. One uL of sample was i n j e c t e d and the mode of i n j e c t i o n was s p l i t l e s s . b. Packed Column GCMS Intermediates and end products of the synthet ic react ions were monitored using a Hewlett-Packard 5700A gas chromatograph in ter faced to a Varian M a t - I l l mass spectrometer v ia a var iab le s l i t separator . Ion iza t ion energy was 70eV and source pressure 5 x 10~6 t o r r . - 21 -Scanning range was 15-750 mass uni ts with one scan taken every 5 seconds. Data was processed by an o n - l i n e Varian 620L computer system. Operating c o n d i t i o n s : column (1.8 m x 2 mm I.D) packed with 2% Dexsil 300 on 100/200 mesh Supelcoport (Supelco, I n c . , B e l l e f o n t e , Pennsy lvan ia ) . Temperature program: i n i t i a l 5 0 ° C , rate 8°C/min to 2 7 0 ° C , hold 5 min at 270°C . 2. GC - E lec t ron Capture Detect ion PFB d e r i v a t i v e s of v a l p r o i c and octanoic ac ids were analyzed by e lec t ron capture detect ion using a HP-5840A gas chromatograph, modified fo r c a p i l l a r y column use. Temperature program: 150°C (hold 4 min) then 20°C/min to 2 4 0 ° C , hold 3 min at 240°C . Argon/methane flow: 6 mL/min (column), 40 mL/min (de tec tor ) . Column: the same as in l a . 3 . Other Instruments The IR spectrum of ethyl 2-propyl-4-oxopentanoate was obtained as a neat l i q u i d f i l m on sodium ch lo r ide disks using a Unicam SP-1000 spectrophotometer. Proton NMR spectra were recorded on Bruker WP-80 and N i c o l e t Oxford-270 instruments at the NMR f a c i l i t y in the Department of Chemistry, U .B .C . NMR solvent was C D C I 3 and the internal standard te t ramethy ls i l ane . C. Human Study The b lood , urine and s a l i v a samples used fo r metabolite i d e n t i f i c a t i o n and quant i ta t ion of VPA were part of a drug in te rac t ion study between carbamazepine (CBZ) and VPA and were c o l l e c t e d as part of the M.Sc. graduate research of Sukhbinder Panesar. F ive healthy male volunteers par t ic ipa ted in the study. Volunteers received an average of - 22 -16.4 mg/kg/day of VPA in syrup form. The drug was administered in two equal doses, one at 8 a.m. and the other 8 p.m. On day 9 o f the study 100 mg CBZ twice d a i l y was added to the dosing regimen which was inc r -eased to 200 mg for the evening dose on day 16. One volunteer , FA, also rece ived s ix doses of 700 mg [ 2H6]-VPA on days 8 , 9, 25, 26 (twice) and 27. Blood was c o l l e c t e d in s t e r i l e , non-heparinized vacutainers p r io r to the morning dose of days 7 and 23 a f t e r an overnight fas t and at 0 . 5 , 1, 1 . 5 , 2, 2 . 5 , 3 , 5, 7, 9, 12, 24, 30, 36 and 48h a f te r the dose. The samples were allowed to c l o t and serum obtained a f te r c e n t r i f u g a t i o n . Urine samples were c o l l e c t e d in 2h b l o c k s , other convenient blocks and also overnight . Total urine volume was recorded and a homogenous a l iquo t saved. S a l i v a samples were c o l l e c t e d fo l lowing st imulat ion with 5% c i t r i c a c i d so lu t ion and were taken simultaneous to blood samples. Four mL of the c i t r i c ac id so lu t ion was held in the mouth for 2 min and spat out. The s a l i v a sample (3 to 5 mL) was then c o l l e c t e d a f te r 2 min. The pH of the s a l i v a samples was not measured immediately a f te r c o l l e c t i o n . However, during the c o l l e c t i o n of blank s a l i v a fo l lowing the same procedure the pH was measured and found to be constant at around 7.3 f o r a number of such samplings. A l l b i o l o g i c a l samples were stored at - 2 0 ° C unt i l analyzed. D. A n a l y s i s of Samples 1. Serum and S a l i v a Standards Stock so lu t ions of VPA (500 ug/mL), OA (250 ug/mL), and [ 2 H 6 ] -VPA - 23 -(24 Lig /mL) in water were made by d i l u t i n g concentrated so lu t ions of the above substances in methanol with d i s t i l l e d water. For serum tota l c a l i b r a t i o n curves drug free serum was added to an a l i q u o t of VPA s o l u t i o n to give e i t h e r 150, 30 or 10 ug/mL VPA in serum. For the NICI assay the appropriate volume was taken from e i t h e r the 30 or 10 ug/mL VPA in serum to provide 24, 18, 12, 9, 6 , 4, and 3 Lig/mL VPA concentra t ions in a f i n a l volume of 100 \xl of serum. For EI a n a l y s i s the standards were prepared by p ipe t t ing the required volume from the 150 ug/mL VPA standard to give 90, 75, 60, 45, 30, and 15 ug/mL VPA concentra t ions in a f ina l volume of 100 uL o f serum. S a l i v a standards were prepared by p i p e t t i n g the appropriate volumes from e i t h e r a 5 Lig/mL or 0.5 Lig/mL VPA standard in blank s a l i v a . The c a l i b r a t i o n points prepared were 3 , 1.5, 1, 0 .5 , 0 .25, 0 . 1 , and 0.05 Lig/mL contained a f i n a l volume of 1 mL of s a l i v a . To determine serum free VPA concent ra t ions , VPA so lu t ions of 25 and 2.5 ug/mL in water were prepared and a l iquots taken to provide 15, 10, 7 .5 , 5, 2 .5 , 1, and 0.5 ug/mL concentrat ions in water. F ina l volume was 100 l i L . For purposes of i n v e s t i g a t i n g the p r e c i s i o n of the NICI assay, standards of VPA over the concentrat ion range of 800 to 10 ng/mL in serum were assayed three times on the day of the preparat ion of standards and repeated on days 4 and 7. Then the standard devia t ions of the slopes and the c o e f f i c i e n t of va r ia t ion at each c a l i b r a t i o n point were c a l c u l a t e d . To determine the recovery of VPA in serum and s a l i v a , serum and s a l i v a samples spiked with the same amounts of VPA used for the standard curves were prepared. These concentrat ions were then determined using a standard curve of VPA prepared in water. - 24 -For the NICI assay the peaks at m/z 143 of the PFB ester of VPA and m/z 149 o f the PFB ester of [ 2 He]-VPA (or m/z 143 of the PFB e s t e r of OA) were monitored. The c a l i b r a t i o n curves were obtained by a p l o t of the area r a t i o of VPA to that of [ 2 H6l-VPA or OA versus the known concentrat ion of VPA in serum or s a l i v a . The concentrat ion of each serum or s a l i v a sample was obtained using l i n e a r regression a n a l y s i s . A new standard curve was prepared p r io r to the run of each batch of serum or s a l i v a samples. 2. Sample preparat ion To measure serum total VPA l e v e l s 100 uL o f serum sample was taken and d i l u t e d f i v e times with blank serum. Then 100 uL o f the d i lu ted sample was t rans fe r red in to a 3.5 mL screw cap septum v i a l . Internal standard (40 uL from a 24 ug/ml so lu t ion of [ 2 H6]-VPA) was added fol lowed by 3N NaOH to make the pH 12-13. The samples were subsequently heated for 1 hour at 6 0 ° C . A f t e r cool ing to room temperature the pH was brought down to 2 using 4N HC1 and the samples allowed to s i t at room temperature for 10 minutes. Then the samples were extracted with 500 uL o f ethyl acetate by gentle ro ta t ion for 20 minutes. To increase the recovery of the drug the ex t rac t ion step was repeated with another 500 uL o f ethyl acetate . The organic layer was then t rans fer red to a v i a l conta in ing anhydrous Na 2S04, vortexed and cent r i fuged at 2000 rpm for 20 minutes. The supernatant was t ransfer red to another v ia l and the volume reduced to about 200 uL under N 2 . F i n a l l y the sample was d e r i v a t i z e d to give e i ther PFB or t-BDMS d e r i v a t i v e s . S a l i v a samples were prepared by taking 1 mL of sample followed by the add i t ion of in terna l standard (30 uL of 12 ug/mL [ 2 He]-VPA s o l u t i o n ) . The pH of the samples was adjusted to about 2 and the - 25 -samples were extracted with 3 mL ethyl acetate and fo l lowing c e n t r i f u g a t i o n to break the emulsion, were t reated as for serum samples. For the determination of f ree serum VPA concent ra t ion , 1 mL of serum was cent r i fuged and 100 uL of the u l t r a f i l t r a t e taken. The u l t r a f i l t r a t e was then t reated exact ly as serum samples except that e x t r a c t i o n was done once and that 40 ul of 12 ug/mL [^Hsl-VPA s o l u t i o n was used as an in terna l standard. 3. D e r i v a t i z a t i o n To form the pentaf luorobenzyl (PFB) de r i va t i ve the concentrated serum or s a l i v a ex t rac t was t rans fe r red into a 1 mL con ica l react ion v i a l and 10 ul o f d i isopropy l ethyl amine (neat) was added followed by 10 uL o f 30% pentaf l uorobenzyl bromide (PFBB) so lu t ion in ethyl acetate . The sample was then heated in a heating block for 45 minutes at 4 0 ° C . Samples for EI a n a l y s i s were de r iva t i zed to give t-BDMS d e r i v a t i v e s by adding 50 uL o f t-BDMSCl in pyr id ine (containing 5% DMAP) and heating at 60°C for 4 hours . The TMS d e r i v a t i v e s of ur inary metabol i tes were prepared by adding 50 uL of MSTFA reagent and heating at 60°C for 30 minutes. For GC-ECD a n a l y s i s of the PFB ester of VPA using OA as the in te rna l standard the fo l lowing d e r i v a t i z a t i o n procedure was fo l lowed. Ten ul of each VPA (from 50 ng/mL to 10 iig/mL VPA) so lu t ion in ethyl acetate and 10 uL o f OA (400 ng/mL OA so lu t ion in ethyl acetate) and 60 ul of 0.5% PFBB so lu t ion in benzene (containing 1.5 mg/mL 18-crown-6) were t r a n s f e r r e d to a react ion v i a l and a few c r y s t a l s of potassium acetate added. The react ion v ia l was then allowed to s i t at room temperature for 1 hour, fo l lowing which 65 ul of the react ion mixture was pipetted into another react ion v ia l and the volume made to - 26 -200 uL with ethyl acetate . The solvent was then evaporated with N 2 to complete dryness and the residue reconst i tu ted with 500 ui of ethyl ace ta te . 4. Serum free l e v e l s Serum free l e v e l s of VPA were determined a f te r u l t r a f i l t r a t i o n . U l t r a f i l t r a t i o n - was c a r r i e d out with YMT u l t r a f i l t r a t i o n membranes in a MPS-1 m i c r o p a r t i t i o n system (Amicon C o r p . , Danvers, Massachusetts) . The cent r i fuge was a Beckman Model J 2 - 2 1 . A 45° angle rotor was used which was e q u i l i b r a t e d to 20°C before use. Cent r i fuga t ion was car r ied out for 20 minutes at 3500 rpm. 5. Comparison of the s e n s i t i v i t y of EI(t-BDMS) with NICI (PFB, bis-TFMB) To compare the s e n s i t i v i t y of the f luor ina ted der iva t ives (PFB, 3 ,5 -b i s ( t r i f luoromethy l )benzy l (bis-TFMB)) r e l a t i v e to that of the t-BDMS d e r i v a t i v e , a known amount of VPA was der iva t i zed for EI and NICI a n a l y s i s . The bis-TFMB d e r i v a t i v e of VPA was prepared in the same way as the PFB d e r i v a t i v e described above. Each d e r i v a t i v e was prepared (octanoic acid used as internal standard) in such a way that the f ina l concentrat ion in the react ion v ia l was 5 ug/mL. The t-BDMS der iva t i ve was run in the EI mode and the other two der iva t i ves under both EI and NICI c o n d i t i o n s . The samples were run on f i v e d i f f e r e n t days in two weeks t ime. The mean area r a t i o s were taken and the r a t i o of NICI to EI c a l c u l a t e d . - 27 -6 . I d e n t i f i c a t i o n of VPA metabol i tes For t h i s purpose urine samples and one serum sample from the volunteer on VPA steady state who had a lso been given s ix doses of C 2H6]-VPA were used. Two mL urine samples were se lec ted in such a way that the drug and metabol i tes were e i t h e r mainly l a b e l l e d , both l a b e l l e d and u n l a b e l l e d , or predominantly u n l a b e l l e d . Samples analyzed inc luded both before and a f te r adminis t ra t ion of CBZ. Af te r the usual work up, each urine sample was d e r i v a t i z e d to give the PFB, t-BDMS, TMS, and methyl es ter d e r i v a t i v e s . The re tent ion times of a l l d e r i v a t i v e s were recorded and mass spectra obtained in the NICI mode fo r PFB d e r i v a t i v e s , and EI mode for the other d e r i v a t i v e s . In addi t ion to the above samples, de r iva t i zed extracts were a lso prepared without a l k a l i n e treatment of the urine samples p r io r to e x t r a c t i o n . Metabol i te peaks were i d e n t i f i e d by i n j e c t i n g synthet ic standards. E . Chemical Synthesis 1. Attempted synthesis of 2 - (2 ' -oxopropy l ) -2 -penteno ic acid (4 ' -ke to -2 -ene VPA) v ia ethyl 2 -propyl -4-oxopentanoate . a . Synthesis of ethyl 2-bromopentanoate Pentanoic ac id (20.4 g , 0.2 mol) was placed in a 500 mL f lask with a re f lux condenser whose top end was connected to a gas absorption d e v i c e . Bromine (35 g , 0.22 mol) was added fol lowed by 1 mL of - 28 -phosphorus t r ib romide . The mixture was s t i r r e d and heated with an o i l bath at 70°C for 30 minutes and then at 100°C for 6 hours by which time a l l the bromine had reacted . The react ion mixture was subsequently d i s t i l l e d using a water pump in order to remove residual hydrogen bromide. The product was then d i s t i l l e d under reduced pressure . Bp 7 6 - 7 9 ° C / 0 . 0 3 mm. [ L i t . (Acheampong, Ph.D. thes is ) bp 1 0 2 ° - 1 0 5 ° C / 2 . 5 mm]. Mass spectrum:(MW=181) m/z 55 (100%), 138(87%), 140(85%), 27(78%), 41(63%) 29(61%), 43(34%), 94(32%), 101(25%). 2-Bromopentanoic ac id was converted to i t s ethyl es ter by r e f l u x i n g a mixture of 2-bromopentanoic ac id (76g, 0.42 mol ) , ethanol (80 mL,1.36 mol ) , benzene (150 mL), and concentrated s u l f u r i c ac id (1.7 mL) for 12 hours using a Dean-Stark water separat ion u n i t . A f te r washing the react ion mixture with saturated NaHC03 and water, pure ethyl 2-bromopentanoate was obtained by d i s t i l l a t i o n . Bp 4 2 - 4 5 ° C / 0 . 2 mm. [ L i t . (Acheampong, Ph.D. thes is ) bp 6 0 - 6 2 ° C / 3 . 0 mm]. Mass spectrum:(MW=209) m/z 29(100%), 55(89%), 166(22%), 168(20%), 101(12%) 129(10%), 140(8%), 138(6%). b. Synthesis of ethyl 2-propyl-4-oxopentanoate To anhydrous THF (100 mL) (dr ied with l i th ium aluminium hydride) i n a 250 mL f l a s k , sodium hydride (5.76 g (50% d i s p e r s i o n ) , 0.12 mol) was added fol lowed by dropwise addi t ion of ethyl acetoacetate (13 g , 0.1 mol) over a period of 30 minutes. A f te r s t i r r i n g for an addi t ional 15 minutes, ethyl 2-bromopentanoate (20.9 g , 0.1 mol) was added drop by drop and the so lu t ion ref luxed for 6 hours . D i s t i l l e d water (40 mL) - 29 -was then added and the r e s u l t i n g mixture f i l t e r e d under s u c t i o n . The organic layer was separated and the aqueous phase extracted three times with e ther . The combined organic layer was dr ied over anhydrous Na2S04 and the ether evaporated. The crude product was d i s t i l l e d to give ethyl 2 - p r o p y l - 3 - a c e t y l succ ina te . Bp 1 1 8 ° C / 0 . 2 mm. Mass spectrum: (MW=258) m/z 43(100%), 129(40%), 97(35%), 174(32%), 115(26%), 143(18%), 185(12%), 213(11%). To obtain the 2-propyl -4-oxopentanoic ac id a mixture of the acy lsucc ina te (9.9 g , 0.04 mol) and concentrated HC1 (40 mL, 0.4 mol) was heated under re f lux for 8 hours. The r e s u l t i n g mixture was extracted three times with ether and the ext ract dr ied over anhydrous Na2S04- The solvent was then removed and the crude product d i s t i l l e d to give 2-propyl -4-oxopentanoic ac id contaminated with a small amount of ethyl 2 -propyl -4-oxopentanoate . R e d i s t i l l a t i o n gave pure 2-propyl -4-oxopentanoic a c i d . Bp 1 1 5 ° C / 0 . 0 7 mm. [ L i t . (Acheampong et a l . , 1983) bp 1 3 3 - 1 3 6 ° C / 2 . 5 mm]. Mass spectrum: (MW=158) m/z 43(100%), 58(19%), 101(16%), 83(15%), 73(14%), 140(8%), 158(1%). The corresponding ethyl es ter of the above acid was made by r e f l u x i n g 18-crown-6 (0.3 g ) , ethyl iodide (1.56 g , 0.01 mol ) , 2 -propyl -4-oxopentanoic acid (0.79 g , 0.005 mol) and K 2 C O 3 (4 g) in THF for 6 hours . The mixture was f i l t e r e d and THF removed. The residue was then f rac t iona ted by d i s t i l l a t i o n and y i e l d e d pure ethyl 2 - p r o p y l -4-oxopentanoate which showed a s ing le peak upon GCMS a n a l y s i s . The e s t e r i f i c a t i o n react ion was also c a r r i e d out in a s i m i l a r manner to - 30 -that of ethyl 2-bromopentanoate. Bp 9 7 ° C / 1 . 7 mm. Mass spectrum: (MW=186) m/z 43(100%), 129(32%), 101(30%), 141(16%), 29(14%), 73(12%). IR Spectrum (neat f i l m ) : 2900 cm"l ( 0 -CH 2 CH 3 ) , 1735 cm" 1 (C=0). NKR Spectrum: 0 .9 ( t ,3H,CH 3 -CH2- ) ; 1 . 1 - 1 . 6 (m,4H,CH2-CH2) ; 1 . 2 ( t , 3 H , - C H 3 ) ; 2 . l ( s , 3 H , C H 3 - C 0 ) , 2 .3-3(m,3H,CH 2 -CH); 4 .2 (q ,2H,0CH 2 ) . c . Introduction of the double bond by dehydrogenation of the 0-TMS d i a l k y l ketene acetal of ethyl 2-propyl -4-oxopentano-ate with the keto funct ion protected with TMS. Li thium di isopropylamide (LDA) was prepared by dr ipping n - b u t y l l i t h i u m (13.5 mL, 0.022 mol ) , to di isopropylamine (2.95 mL, 0.022 mol) i n THF (25 mL) at 0°C over 20 min per iod . The mixture was then cooled in a dry ice acetone bath to -78°C and ethyl 2 - p r o p y l - 4 -oxopentanoate (1 .86 g ,0 . 0 1 mol) added dropwise and allowed to react for 60 min, fo l lowing which c h l o r o t r i m e t h y l s i l a n e (3.9 g , 0.036 mol) was added dropwise over a 10 min per iod . The temperature was allowed to a t t a i n 25°C and the mixture s t i r r e d for 60 min. The THF was d i s t i l l e d o f f and the residue reconst i tu ted with 5 mL of dry benzene. Then dichlorodicyanobenzoquinone (DDQ) (2.27 g , 0 . 01 mol) was d isso lved in benzene under N2 and 2 , 4 , 6 - t r i m e t h y l p y r i d i n e (1 mL) in benzene was added dropwise to the DDQ s o l u t i o n . A f te r 10 min th is mixture was added to the s i l y l ether under N2 and the mixture s t i r r e d for 2 hours. A f t e r d i l u t i n g with ether the react ion mixture was washed with 1 M NaOH and the aqueous phase extracted three times with ether . The combined organic layer was washed success ive ly with HC1, NaOH, and water and the ex t rac t dr ied over anhydrous Na2S04- GCMS ana lys is of the crude - 31 -product ind ica ted the presence of three components. One of these was the s t a r t i n g m a t e r i a l . The other two were: 1) TMS enol ether of ethyl 2 -propyl -4-oxopentanoate: Mass spectrum:(MW=258) m/z 73(100%), 75(65%), 185(55%), 130(44%), 115(42%) 97(20%), 45(18%), 213(10%), 143(15%), 215(8%), 243(5%), 258(3%). 2) TMS enol ether of the ethyl 2 -propyl -4 -oxo-2-pentenoate or ethyl 2 - (2 ' -oxopropy l ) -2 -pentenoate . Mass spectrum: (MW=256) m/z 73(100%), 75(40%), 95(28%), 43(25%), 137(18%), 109(12%), 183(9%), 167(7%), 213(10%), 227(5%), 256(4%), 241(3%). A small port ion of the react ion mixture was taken and made a l k a l i n e with NaOH and s t i r r e d for three days at room temperature in order to e f f e c t hydro lys is of both the TMS enol ether and the ethyl ester of the product . The mixture was then a c i d i f i e d and extracted with e ther . Fol lowing the evaporation of the ether a port ion of the residue was d e r i v a t i z e d to give the t-BDMS d e r i v a t i v e s which were analyzed by c a p i l l a r y GCMS. Two peaks were detected with mass m/z 213 (M-57) + and had s i m i l a r mass spec t ra . 2. Synthesis of 2 - (2 ' -oxopropy l ) -2 -penteno ic acid s t a r t i n g with a protected 4-oxopentanoic acid a . Pro tec t ion of 4-oxopentanoic acid through a d i t h i o ketal Commercially a v a i l a b l e 4-oxopentanoic ac id was p u r i f i e d by f r a c t i o n a l d i s t i l l a t i o n and was converted into i t s ethyl ester in a manner s i m i l a r to that descr ibed for 2-bromopentanoic a c i d . Ethyl 4-oxopentanoate (11.5 g , 0.08 mol) was d isso lved in dichloromethane (d r ied over calcium hydride) and 10 mL (0.12 mol) of 1 ,2 -e thanedi th io l was added fol lowed by 2 mL of boron t r i f l u o r i d e e therate . The solut ion was s t i r r e d at room temperature overnight and 100 mL of 5% sodium - 32 -hydroxide added. The organic layer was separated and washed with water and dr ied over MgSO^. Evaporation of the solvent and subsequent d i s t i l l a t i o n of the residue af forded 15g (85%) of ethyl 4 -e thy leneth ioketa lpentanoate . Bp 1 1 2 ° C / 0 . 5 mm. Mass spectrum: (Mw=220) m/z 119(100%), 59(20%), 175(17%), 29(15%), 61(14%), 87(13%), 115(13%), 220(12%), 87(10%), 205(2%). NMR Spectrum: 6 1 . 3 ( t , 3H , -CH 3 >; 1 .8(s ,3H ,CHg- ) ; 2.25(t,2H,_CH_ 2-CH 2-C0); 2 . 6 ( t , 2 H , C H 2 - C H 2 - C 0 ) ; 3 .3 -3 .4 (m,4H ,CH 2 -CH 2 -S ) ; 4 .7 ( q , 2H ,0CH 2 ) . b. Synthesis of 4 -carboethoxy-2 -e thy leneth ioketa l -5-hydroxyheptane To di isopropylamine (8.2 mL, 0.06 mol) in 100 mL THF at 0°C was added dropwise n -bu ty l l i th ium (38 mL, 0.06 mol) . The mixture was s t i r r e d for 15 min and cooled to - 7 8 ° C and ethyl 4 -e thy lene th ioke ta l -pentanoate (11 g, 0.05 mol) in THF was added dropwise and the mixture s t i r r e d for 60 min. Then propionaldehyde (4 mL, 0.055 mol) was added and the react ion mixture allowed to s t i r for 2.5 hours . The mixture was then quenched with 15% HC1 and extracted with e ther . The etheral e x t r a c t was washed with water and saturated NaHC03 and dr ied (Na2S04). A f t e r evaporation of the ether the residue was d i s t i l l e d under reduced pressure and afforded 6 g of the product ( thick ye l lowish l i q u i d ) which was shown to be homogeneous by GCMS. - 33 -Mass spectrum:(MW=278) m/z 119(100%), 29(35%), 59(30%), 185(27%), 43(25%), 111(22%), 61(21%), 139(20%), 120(15%), 121(14%), 220(10%), 159(8%), 186(7%), 205(4%), 233(2%), 278(1%). c . Synthesis of 4 -carboethoxy-2-ethy leneth ioketa l -4 -heptene Dehydration of 4-carboethoxy-2-ethylenethioketal -5-hydroxyheptane was c a r r i e d out using methanesulfonyl c h l o r i d e and potassium hydr ide. Hydroxy compound (4 g , 0.015 mol ) , t r ie thylamine (3 mL, 0.02 mol ) , and dichloromethane (40 mL) were cooled to 0 ° C . Methanesulfonyl ch lo r ide (1.6 mL, 0.02 mol) in dichloromethane was added dropwise and the mixture s t i r r e d for 60 min. The p r e c i p i t a t e formed was f i l t e r e d of f and the solvent removed by f l ash evaporat ion. The mesylate was then taken up in dry THF and potassium hydride (1.2 g , 0.03 mol) added at 0°C and the react ion mixture s t i r r e d for 12 hours at room temperature fo l lowing which unreacted potassium hydride was neut ra l i zed with t -butanol and water. The mixture was extracted with ether and dr ied (Na2S04). GCMS ana lys is of the crude product showed f i v e peaks. Two of these peaks in the TIC trace corresponded to the two geometric isomers of the desi red product with one of these isomers being the major component. The other three peaks were not i d e n t i f i e d . Mass spectrum: (Mw=260) Isomer with the retent ion time of 19.18 min: m/z 127(100%), 29(58%), 199(45%), 132(22%), 45(21%), 119(21%), 59(20%), 99(20%), 41(19%), 61(18%), 74(15%), 200(14%), 155(10%), 159(9%), 187(6%), 260(5%), 215(3%). - 34 -Mass spectrum: Isomer with the re tent ion time of 20.24 min: m/z 199(100%), 127(85%), 29(60%), 99(29%), 200(28%), 74(27%), 112(27%), 155(25%), 61(22%), 41(20%), 45(18%), 59(17%), 159(14%), 187(10%), 260(6%), 215(2%). d . Removal of the 1 ,3 -d i th io lane protect ing group Cleavage of the d i th io lane group was accomplished with mercuric c h l o r i d e in the presence of cadmium carbonate. A port ion of the product mixture from above (c) (2.5 g) was d isso lved in acetone (100 mL) and water (10 mL), mercuric ch lo r ide (2 g ) , and cadmium carbonate(2 g) were added and the mixture s t i r r e d at room temperature for 24 hours . At th is stage addi t iona l mercuric c h l o r i d e (0.8 g) and cadmium carbonate (0.8 g) were added and s t i r r i n g continued for another 72 hours . Then the mixture was f i l t e r e d and the acetone removed. The residue was d issolved in ether and the etheral so lu t ion washed success ive ly with water, 10% potassium i o d i d e , water and dr ied over anhydrous Na2S04- A f te r removal of the ether the residue was analyzed by GCMS. The TIC p lo t showed 9 peaks two of which (minor components) corresponded to the two isomers of the desi red product. Mass spectrum: (MW=184) Peak with the retent ion time of 9.06 min: m/z 111(100%), m/z 43(90%), 29(15%), 55(12%), 112(11%), 184(9%). Mass spectrum: Peak with the retent ion time of 9.65 min: m/z 111(100%), 43(65%), 55(20%), 29(19%), 184(10%), 112(9%). A port ion of the product from above was then t reated with a l k a l i and s t i r r e d for 3 days fo l lowing which i t was a c i d i f i e d and extracted with ether . The ether was evaporated and two port ions of the residue were taken and der iva t i zed to give PFB and t-BDMS d e r i v a t i v e s for NICI and EI a n a l y s i s , r e s p e c t i v e l y . The mass spectra and retent ion times were then compared to those from urine e x t r a c t . - 35 -3. Synthesis of ethyl 2-propyl-3-oxopentanoate To isopropylcyclohexylamine (16.9 g , 0.12 mol) in dry THF (80 mL), at 0 ° C , was added n-butyl 1 ithiurn (77 mL, 0.12 mol) dropwise and the mixture was s t i r r e d for 15 min. The mixture was cooled to -78°C and ethyl pentanoate (13 g , 0.1 mol) fol lowed by propionyl ch lo r ide (9.3 g , 0.1 mol) in THF were added and s t i r r i n g continued for 25 minutes. Then the react ion mixture was quenched with 15% HC1 and extracted with e ther . A f te r the usual work-up and f r a c t i o n a l d i s t i l l a t i o n pure ethyl 2-propyl-3-oxopentanoate was obtained. Bp 8 6 ° C / 0 . 1 mm. Mass spectrum: (MW=186) m/z 57(100%), 29(72%), 101(33%), 55(20%), 73(19%), 130(10%), 144(9%), 43(8%), 157(4%). 4. Synthesis of ethyl 2-propyl-3-hydroxypentanoate n-Butyl1ithiurn (77 mL, 0.12 mol) was added dropwise to di isopropylamine (12 mL, 0.12 mol) in dry THF at 0 ° C , over a 20 min p e r i o d . The mixture was allowed to s t i r for a fur ther 15 min and cooled to - 7 8 ° C and ethyl pentanoate (15.6 g , 0.12 mol) in 10 mL THF was added dropwise and the mixture s t i r r e d for 60 min. Then propionaldehyde (7.4 g , 0.1 mol) was added and the s t i r r i n g continued f o r 2 hours. The react ion mixture was neut ra l i zed with 15% HC1 and extracted with e ther . Fol lowing f l a s h evaporation of the ether the residue was d i s t i l l e d to y i e l d ethyl 2-propyl -3-hydroxypentanoate. BP 92°C /4 mm. [ L i t . (Acheampong et a l . , 1983) bp 7 0 - 7 2 ° C / 0 . 2 mm]. Mass Spectrum: (MW=188) m/z 101(100%), 73(85%), 55(65%), 29(45%), 57(38%), 130(32%), 113(30%), 84(20%), 41(15%), 159(12%), 143(10%). - 36 -III. RESULTS AND DISCUSSION A . NICI GCMS assay development A highly s e n s i t i v e assay has been developed for VPA in serum and s a l i v a based on the NICI-GCMS of the PFB e s t e r . The assay was employed to quant i ta te VPA in serum and s a l i v a in f i v e volunteers at VPA steady s t a t e . The assay i s a lso app l icab le to VPA metabo l i tes , although no quant i ta t ion of the metabol i tes was made in the present study. The chromatographic c h a r a c t e r i s t i c s of the PFB d e r i v a t i v e s of a l l VPA metabol i tes have been determined and se lected ion chromatograms obta ined. The assay development is discussed below from the point of view of de r i va t i ve format ion, i n i t i a l GC-ECD a n a l y s i s , opt imizat ion of mass spectrometer parameters, and s e l e c t i o n of a su i tab le internal standard. 1. D e r i v a t i z a t i o n D e r i v a t i z a t i o n with PFBB was c a r r i e d out by a modi f ica t ion of the method of Min et a l . (1980) for the d e r i v a t i z a t i o n of a tr imethyl prostaglandin E 2 ana log. Th is i s based upon the observat ion that the Cj carboxyl group of prostaglandins undergoes f a c i l e react ion with benzy l i c h a l i d e s . To form the PFB der iva t i ves of VPA and i t s metabol i tes samples were heated at 40°C for 45 minutes in the presence of 10 uL of 30% PFBB so lu t ion in ethyl acetate and 10 uL of d i i s o p r o p y l e t h y l amine and t h i s procedure was used throughout the study. Longer react ion times or higher temperatures produced a yel low gummy - 37 -substance. I n i t i a l l y t r i e t h y l amine was used, but was replaced with d i i s o p r o p y l e t h y l amine, a bulky amine, which minimizes quaternary ammonium s a l t formation with PFBB (Wickramasinghe et a l . , 1973). In most of the NICI work done with p ros tag land ins , solvent i s removed a f te r the ex t rac t ion step and the residue i s reconst i tu ted with a c e t o n i t r i l e p r i o r to d e r i v a t i z a t i o n . With VPA and i t s metabol i tes , such a procedure would not be des i rab le because of t h e i r low molecular weight compared to prostaglandins and hence complete solvent evaporat ion before d e r i v a t i z a t i o n was thought to r e s u l t in loss of sample. Because of the presence of the ex t rac t ing solvent (ethyl a c e t a t e ) , there was evidence of the e s t e r i f i c a t i o n of a small amount of hydrolyzed ethyl acetate ; however th is was not a problem. Removal of the excess PFBB p r i o r to i n j e c t i o n of de r i va t i zed sample in to t the GCMS did not prove poss ib le without loss of peak i n t e n s i t i e s (see below). Because of the excess reagent solvent was d iver ted and scanning commenced 4 minutes a f te r sample i n j e c t i o n to reduce the amount of reagent enter ing the mass spectrometer ion source. The d e r i v a t i z i n g procedure adopted appears to r e s u l t in optimum d e r i v a t i z a t i o n since no addi t iona l peaks were observed upon attempts to methyl ate the components of the react ion mixture a f te r the d e r i v a t i z a t i o n step with PFBB. A typ ica l tota l ion chromatogram trace f o r the PFB esters of VPA and OA in the negative ion mode i s shown in F igure 2. 2. GC-Elect ron Capture Detect ion (ECD) F igure 3 shows typ ica l GC-ECD chromatograms of the PFB der iva t ives of VPA and OA. Since GC with ECD i s in p r i n c i p l e s i m i l a r to e lect ron - 38 -a b T 2 i r~ \ 5 TIME (min) 6 ~ r 8 Figure 2. Total ion current p lo t o f the PFB esters o f VPA(a) and 0A(b) in the NICI mode. - 39 -capture NICI, i t was of i n t e r e s t to i n i t i a l l y evaluate the assay with GC. When the d e r i v a t i z a t i o n procedure out l ined above was used and 1 or 2 uL o f sample in jec ted into the GC, there was a high background due to the excess d e r i v a t i z i n g reagent. To remove most of the excess PFBB the fo l lowing procedures were evaluated. a . Treatment with a l k a l i A f t e r the d e r i v a t i z a t i o n s tep , solvent was removed under N 2 and the residue t reated with 400 uL of 3N NaOH and/or 400 uL of ION NaOH and extracted with ethyl acetate . T h i s , however, did not remove excess reagent because the product presumably formed i s pentaf luorobenzyl a lcohol which i s soluble both in water and most organic so lven ts . b. Evaporation of the react ion mixture at elevated temperature ( 4 0 - 6 0 ° C ) . The reagent peak was s i g n i f i c a n t l y reduced but so were analyte and in terna l standard peaks. PFBB i s not very v o l a t i l e r e l a t i v e to VPA-PFB and condi t ions required to evaporate excess reagent i n v a r i a b l y cause l o s s of the sample. c . S e l e c t i v e solvent ex t rac t ion To s e l e c t i v e l y remove excess PFBB, d i f f e r e n t solvents (ethyl ace ta te , hexane, heptane, iso-octane e t c . ) were employed. The react ion mixture was made b a s i c , extracted with one of the solvents and re -ex t rac ted with another solvent of d i f f e r i n g p o l a r i t y . It was found that both d e r i v a t i v e and the PFBB reagent were soluble in a l l so lvents . d . Column chromatographic separat ion S i l i c a gel and sephadex LH-20 short columns were used in an attempt to separate the der iva t ives from PFBB; but , again separation was not p o s s i b l e . - 40 -When a d i f f e r e n t d e r i v a t i z a t i o n procedure that used 0.5% PFBB s o l u t i o n in the presence of 18-crown-6 and potassium acetate was employed, i t was poss ib le to obtain the chromatograms shown in Figure 3 . An acceptable c a l i b r a t i o n curve for VPA in ethyl acetate using OA as in terna l standard was a lso obtained (Figure 4.) Th is procedure adopted from Rubio and Garland (1985) uses a l e s s e r amount of PFBB and i s supposedly advantageous in that phenols are not d e r i v a t i z e d because o f the use of the l ess basic potassium acetate . 3 . Opt imizat ion of MS parameters for NICI It i s known that under CI cond i t ions a number of fac tors in f luence the spectrum obtained and hence the s e n s i t i v i t y . During our i n i t i a l work with NICI, we experienced lack of reproducible s e n s i t i v i t y . The area counts obtained for the same concentrat ion of de r iva t i zed VPA on d i f f e r e n t runs and days var ied at times by more than a fac tor of 100. There was a lso f i lament sag and the NICI cond i t ions tended to destroy the e lec t ron m u l t i p l i e r . E f f o r t s to manipulate instrument var iab les (e lec t ron energy, source pressure , etc) in order to get reproducible s e n s i t i v i t i e s did not prove s u c c e s s f u l . The lack of r e p r o d u c i b i l i t y at that time can be explained by the f a c t that i t was not poss ib le to maintain reagent gas flow at a reasonably constant va lue . The other problems such as f i lament sag were probably due to the fact that the pressure gauge reading d id not r e f l e c t the actual pressure in the ion source . With the instrument in apparently good working order , and with the tune values remaining reasonably constant , the e f f e c t s of important - 41 -b T 1 1 r 1 3 5 7 TIME (min) Figure 3. Typica l chromatograms o f the PFB d e r i v a t i v e s o f VPA(a) and 0A(b) obtained with GC-ECD. - 42 -Figure 4. C a l i b r a t i o n curve for PFB d e r i v a t i z e d VPA in ethyl acetate obtained using GC-ECD. - 43 -mass spectrometer va r iab les on the s e n s i t i v i t y to VPA-PFB were determined. Source temperatures t r i e d were 150, 190 and 240°C . R e l a t i v e s e n s i t i v i t i e s of the VPA -PFB peak were 7 . 56 , 2 . 94 and 1.0 r e s p e c t i v e l y . The s e n s i t i v i t y thus appeared to increase as the source temperature decreased, but for routine ana lys is a compromise must be made between an optimum low temperature and the cons idera t ion of the ion source becoming excess ive ly contaminated at lower temperatures. Most of the ana lys is was, the re fo re , c a r r i e d out at a source temperature of 200°C . When varying the source pressure s e n s i t i v i t y was maximal at about the highest pressure to le ra ted by the instrument (1 t o r r ) . For pressures l ess than 1 t o r r , the s e n s i t i v i t y decreased dramat ica l ly u n t i l at low pressures i .e less than 0.3 t o r r there was no signal at a l l . The highest response for the VPA-PFB d e r i v a t i v e was obtained near the highest e lec t ron energy handled by the instrument (240eV). At e lec t ron energies l e s s than lOOeV the monitored i o n , m/z 143", was not detected. A l l the above instrumental parameters are i n t e r r e l a t e d . For example, s e n s i t i v i t y depends upon the extent to which newly formed anions are s t a b i l i z e d by reagent gas molecules, but s ince c o l l i s i o n can a l s o lead to e lec t ron detachment, s e n s i t i v i t y w i l l a lso depend upon the in te rna l energy of the reagent gas so that s e n s i t i v i t y w i l l be highly in f luenced by source temperature. The i o n i z i n g e lec t ron energy used a lso depends upon the value of the source pressure . Because i t i s - 44 -d i f f i c u l t to assess absolute enhancement of s e n s i t i v i t y , the e f f e c t of these interdependent parameters i s measured approximately by varying one parameter while maintaining the others at some optimal va lue . The e f f e c t of reagent gases on the s e n s i t i v i t y of the VPA-PFB was a l s o inves t iga ted using methane, ammonia, and argon-methane (95:5). Ammonia gave an apparently greater s e n s i t i v i t y , which i s s i m i l a r to that reported by Miyazaki et a l . (1984) for p ros tag land ins , but resu l ted in short f i lament l i f e . There was no d i f f e rence between methane and argon-methane with respect to sample response or background i o n s . No carbon conta in ing ions of higher masses ( ion/molecule adducts) were observed with methane and th is gas was used throughout a l l subsequent experiments. In sp i te of i t s high s e n s i t i v i t y the reported use of NICI i s l i m i t e d because in addi t ion to other reasons ( e . g . a v a i l a b l i l i t y of the inst rument ) , poor r e p r o d u c i b i l i t y of NICI mass spectra i s often mentioned as a r e s t r i c t i n g fac tor (Oehme et a l . , 1986). Furthermore, a drawback of most instruments with negative ion detect ion c a p a b i l i t y is that important parameters such as temperature and pressure can only be measured approximately. In our work with VPA we had d i f f i c u l t y gett ing reproducib le s e n s i t i v i t y and some changes in the mass spectrometer were required to improve the r e p r o d u c i b i l i t y . The reagent gas flow c o n t r o l l e r of the mass spectrometer was replaced with a simple needle v a l v e . That meant the source pressure gauge i n d i c a t o r gave c l o s e r values to the actual source pressure . A mod i f i ca t ion to the source heater was a lso made. These changes brought a dramatic improvement in r e p r o d u c i b i l i t y and, to a l e s s e r extent s e n s i t i v i t y . In our exper ience, - 45 -we found ion source pressure to be a c r i t i c a l parameter for obtaining reproducib le s e n s i t i v i t i e s . 4. Comparison of the r e l a t i v e s e n s i t i v i t y of der iva t i zed VPA by EI and NICI Once reproducible s e n s i t i v i t i e s were obta ined, i t was poss ib le to determine the r e l a t i v e s e n s i t i v i t y of two f luor ina ted der iva t ives to that of the t-BDMS d e r i v a t i v e . The NICI and EI spectra of the PFB and bis-TFMB d e r i v a t i v e s of VPA are shown in Figures 5 and 6. The f l u o r i n a t e d der iva t i ves were analyzed in both the EI and NICI modes whi le the t-BDMS d e r i v a t i v e was analyzed by EI and the m/z 201 (M-57) + ion monitored. In the NICI spectra of both f luor ina ted d e r i v a t i v e s , the base anion i s m/z 143 corresponding to the loss of the pentaf luorobenzyl and 3 , 5 - b i s ( t r i f l u o r o m e t h y l ) b e n z y l moie t ies . In the EI s p e c t r a , the base peaks are m/z 181 (PFB) and m/z 227 (bis-TFMB) which are the complementary p o s i t i v e ions to those of the NICI. Since these ions represent moiet ies introduced by d e r i v a t i z a t i o n , the EI spectra of these der iva t i ves lack s p e c i f i c i t y . A l l three der iva t i ves were prepared in a manner that the f ina l concentrat ion in the react ion v i a l was an amount of de r i va t i ve equivalent to 5 ng/uL of VPA. The r e s u l t obtained a f te r analyzing the samples i s shown in Table 1. The area counts are the means of f i v e determinat ions. In the EI mode, the s e n s i t i v i t y of a l l three d e r i v a t i v e s was s i m i l a r . In the NICI mode, the PFB d e r i v a t i v e was found to be 30-50 times more s e n s i t i v e than the t-BDMS d e r i v a t i v e by E I . The PFB d e r i v a t i v e (NICI mode) a lso proved to be about 5 times more s e n s i t i v e than the s i m i l a r f l u o r i n a t e d d e r i v a t i v e , bis-TFMB. - 46 -A) 100 _ 143 >-l— • — i CO B) 100 _ >-t— I — I CO •z. LU t— UJ LU or 50 C H j - C H 2 - C H2 v 4 - O f C H CH3-CH2- C H 2 14: 181 M/Z 250 350 C CH^—C C H^-^C^—C H2 ^ C H - C - O + C H 1 4 3 _ J 150 242 250 M/Z 350 450 Figure 5. NICI mass spectra of the PFB(A) AND bis-TFMB(B) de r iva t i ves o f VPA. - 47 -A) 100 _ 181 CH - C H - C H P| J ^ C H - C - 0 C H 3 - C H 2 - C H 2 " C H 2 - C 6 F 5 1 — 181 282 240 25 • i l I il 125 M / Z 225 325 B) 100 >-i — cn > 57 227 C H 3 - C H 2 - C H 2 C H 3 - C H 2 - C H 2 0 II , C H - C - O f C H 2 - C 6 H 3 ( C F 3 ) 2 -227 or 50 JLLl 285 328 Ml I I •• ••• I || I 355 150 M / Z 250 350 Figure 6. EI mass spectra of the PFB(A) and bis-TFMB(B) d e r i v a t i v e s of VPA. - 48 -5 . Quant i ta t ive ana lys is with the PFB d e r i v a t i v e a . Internal standard The in terna l standard used throughout the present study was [ 2 H5]-VPA. The use of an in terna l standard that gives a common ion (OA) was a lso i n v e s t i g a t e d . The r e s u l t of the ana lys is of serum spiked with seven d i f f e r e n t concentrat ions (10 to 800 ng/mL) of VPA with the same amount of e i t h e r [ 2H6]-VPA or OA as internal standard is shown in Table 2. The r e s u l t s are the means of three determinat ions. Figure 7 shows t y p i c a l se lec ted ion chromatograms of the PFB der iva t i ves of VPA, [ 2 H6]-VPA and OA extracted from serum. As seen from Table 2, the c a l i b r a t i o n curve with [ 2H6]-VPA as the in terna l standard is super ior to that with OA. The in tercepts are much c l o s e r to the o r i g i n and the in t ra -assay v a r i a b i l i t y i s lower in the case of [ 2 H63-VPA. In GCMS assays using SIM, both stable i s o t o p e - l a b e l l e d analogs and chemical ly re la ted compounds have been used as in terna l standards. Claeys et a l . (1977) evaluated the p r e c i s i o n of assays using e i ther internal standard and concluded that s tab le i s o t o p e - l a b e l l e d internal standards produced the lowest variance f a c t o r s due to sample manipulation and instrumental e r r o r s . On the other hand, Lee and M i l l a r d (1975) have argued that a substance is most accurate ly determined using an internal standard g iv ing a common ion because of the advantage of monitoring a s ing le ion and hence a gain in s e n s i t i v i t y and s t a b i l i t y . Under CI c o n d i t i o n s , though, because of the extreme dependence of the spectra on source pressure and temperature i t can be said that more accurate measurements can be made with stable i s o t o p e - l a b e l led internal standard since i t s physicochemical propert ies c l o s e l y approximate that of the analyte . Our resu l ts seem to support t h i s . - 49 -TABLE 1. Comparison of the relative sensi t iv i t ies of three VPA derivatives*. Der iva t ive Mode of Ion Area Re la t ive s e n s i t i v i t y i o n i z a t i o n monitored count NICI/EI PFB EI 181 8,714 47 NICI 143 412,000 bis-TFMB EI 227 8,228 13 NICI 143 103,500 t-BDMS EI 201 10,184 * Amount in jec ted in a l l cases was an amount of d e r i v a t i v e equivalent to 5 ng of VPA. TABLE 2. Comparison of [2Hg]-VPA and OA as internal standards* C 2H6]-VPA OA Intra-assay Vari at ion < 5% 8-15% r 2 0.9986 0.9960 y - i n t e r c e p t - 0.0027 0.1020 x - i n t e r c e p t 2.2483 - 12.0155 * Results obtained with seven d i f f e r e n t concentrat ions of VPA ranging from 10 to 800 ng/mL. The samples were assayed three times and values shown are means of the three determinat ions. - 50 -m/z 149 TIME (min) Figure 7. SIM chromatograms of VPA, OA (m/z 143) and [ H g ] -VPA (m/z 149) from serum spiked with these substances. - 51 -b. A n a l y t i c a l parameters The ex t rac t ion procedure employed was a modi f ica t ion of that of Abbott et a l . (1986a), modif ied to accommodate the small serum sample volume (100 uL) . For the ex t rac t ion of s a l i v a samples a s i m i l a r procedure was used, but samples were not t reated with a l k a l i p r i o r to a c i d i f i c a t i o n . The recovery of drug from serum and s a l i v a was greater than 95% compared to drug extracted from water. The c a l i b r a t i o n curves for VPA using [ 2 H6l-VPA as internal standard were obtained by monitoring the intense peaks at m/z 143 and 149 which are the base ions in the NICI spectra of VPA-PFB and [ 2 H6]-VPA-PFB, r e s p e c t i v e l y . The c o e f f i c i e n t of determinat ions, r 2 , were greater than 0.996 for both serum (free and to ta l ) and s a l i v a . The c a l i b r a t i o n curves for serum t o t a l , serum f r e e , and s a l i v a VPA are shown in Figures 8 , 9 and 10 r e s p e c t i v e l y . The c a l i b r a t i o n curve for the EI(t-BDMS) determination of serum tota l VPA i s given in F igure 11. The p r e c i s i o n of the NICI assay using PFB der iva t i ves was very good with an in t ra -and in te r - assay v a r i a t i o n of l e s s than 10% ( c o e f f i c i e n t of var ia t ion ) at 10-800 ng/mL of VPA in serum. The within assay and between assay standard devia t ions of the slope of the c a l i b r a t i o n curve were 0.000003 and 0.000021, r e s p e c t i v e l y (C.V.=0.299% and 1.92%). L i n e a r i t y was observed over the concentrat ion range of 10 ng/mL to 25 ug/mL. The lower l i m i t of detect ion was 2 ng/mL of VPA based on a 200 uL sample of serum. Figure 12 shows the SIM chromatogram of VPA obtained with 10 pg of VPA extracted from serum. This l i m i t of detect ion i s s i m i l a r to most of the reported values for PFB der iva t ives - 52 -2.5 -i 0 5 10 15 20 25 Concentration (ug/mL serum) Figure 8. C a l i b r a t i o n curve for serum to ta l VPA. Peak area r a t i o was obtained by monitor ing m/z 143 (VPA) and m/z 149 ( [ 2 H , ] - V P A ) . - 53 Figure 9. C a l i b r a t i o n curve for serum free VPA. Peak area r a t i o was obtained by monitor ing m/z 143 (VPA) and m/z 149 ( [ 2 H f i ] - V P A ) . - 54 -Figure 10. C a l i b r a t i o n curve for s a l i v a VPA. Peak area r a t i o was obtained by monitor ing m/z 143 (VPA) and m/z 149 ( [ Z H g ] - V P A ) . - 55 -Figure 11. C a l i b r a t i o n curve for EI(t-BDMS) determinat ion o f serum tota l VPA. Peak area r a t i o was obtained by monitor ing m/z 201 (VPA) and m/z 207 ( [ 2 H g ] - V P A ) . - 56 -o f prostanoids . In some c a s e s , with prostaglandins l i m i t s of detect ion as low as 200 fg have been reported with ammonia as the reagent gas (Miyazaki et a l . , 1984). The lower l i m i t of detect ion obtained with VPA-PFB i s not as good as the above example probably because of the f a c t that VPA i s a small molecule and hence with l ess surface area for e lec t ron capture . To assess how well the developed NICI assay compared with an EI (t-BDMS) assay rou t ine ly used for quant i ta t ion of VPA in our l a b o r a t o r y , serum samples of two volunteers were measured by both EI and NICI GCMS methods. These values are given in Table 3 , and Figure 13 shows the c o r r e l a t i o n of the serum values obtained by the two methods fo r one of the vo lunteers . As seen from Table 3 and Figure 13, agreement between EI and NICI methods for serum VPA i s e x c e l l e n t . Mean values were 28.4 vs 27.1 (B.A) and 40.3 vs 40.2 (F.A) and c o r r e l a t i o n s (r) were 0.98 and 0 .99 , r e s p e c t i v e l y . B. VPA l e v e l s in serum (free and to ta l ) and s a l i v a before and a f te r the adminis t ra t ion of CBZ A to ta l of 63 pa i red s a l i v a and serum samples were analyzed for VPA by NICI in SIM mode; 33 samples before CBZ and 30 a f t e r CBZ. The serum t o t a l , serum free and s a l i v a l e v e l s of VPA in the f i v e volunteers are presented in Tables 4 to 8 . The r e l a t i o n s h i p s between serum t o t a l , serum free and s a l i v a , as well as the c o r r e l a t i o n s between each other are a lso given in these t a b l e s . Tables 9 and 10 show the time-averaged ra t ios between serum free and serum t o t a l , s a l i v a and serum t o t a l , and s a l i v a and serum free - 57 -TIME (min) Figure 12. SIM chromatogram of PFB d e r i v a t i z e d VPA obtained with 10 pg of VPA extracted from serum (the second peak i s background from serum). - 58 -TABLE 3. Serum VPA levels (ug/mL) in two subjects on VPA steady-state as measured by EI (t-BDMS) and NICI (PFB). BA FA EI 40.31 40.56 42.50 25.93 23.20 27.97 25.03 33.80 31.78 27.61 26.05 21.85 16.94 14.35 NICI 42.36 42.44 43.78 20.56 22.27 25.90 21.07 31.05 30.36 26.65 23.60 19.01 16.40 13.26 EI 44.35 58.34 41.20 48.20 48.05 33.41 22.70 51.25 45.27 42.15 34.17 29.02 26.04 NICI 44.45 58.77 40.59 46.84 46.06 33.61 24.15 53.24 43.80 42.02 33.34 26.99 28.02 Mean: 28.42 C o r r e l a t i o n ( r ) : 27.053 0.9830 40.319 40.145 0.9913 - 59 -0.7710 0 10 20 30 40 50 60 El Figure 13. Re la t ionsh ip between VPA concentrat ions (ug/mL) in serum determined by EI(t-BDMS) and NICI (PFB). - 60 -whi le in Table 11 the % change in average VPA concentrat ion in a l l three b i o l o g i c a l samples a f te r CBZ adminis t ra t ion i s presented. Figure 14 i s the s a l i v a concentrat ion- t ime p r o f i l e for the f i v e vo lunteers . In F igure 15 the concentrat ion time curves for s a l i v a , f r e e , and total VPA f o r one volunteer are shown. F igures 16 and 17 are the concentrat ion time curves for serum free and s a l i v a VPA before and a f te r CBZ in one vo lunteer . Curves showing the degree of c o r r e l a t i o n between s a l i v a and both serum tota l and free VPA before and a f te r CBZ for one of the volunteers are given in Figures 18 and 19. The r e l a t i o n s h i p between serum tota l and s a l i v a and between serum free and s a l i v a for a l l the volunteers are shown in Figures 20 and 21. 1. E f f e c t of CBZ on serum and s a l i v a l e v e l s of VPA The % decrease in average VPA concentrat ion a f te r CBZ was 27.91 ± 3.48, 36.85 ± 13.64, and 48.13 ± 7.70 (Table 11) , for serum t o t a l , serum f r e e , and s a l i v a , r e s p e c t i v e l y . There was a s i g n i f i c a n t reduction (PO.025 ) of VPA concentrat ion in a l l three b i o l o g i c a l f l u i d s . The % decrease i s higher for serum free and s a l i v a compared to serum tota l f o r a l l volunteers except one (R.M. ) . The greater % decrease of serum f ree VPA and s a l i v a VPA a f te r CBZ i s because the free f r a c t i o n decl ined with decreasing tota l serum VPA concent ra t ions . The mean free f rac t ion f o r the f i v e subjects decreased from 0.1334 to 0.1072 a f t e r CBZ (Tables 9 and 10) . Because VPA i s a h ighly protein-bound drug, VPA i s l i k e l y to undergo drug i n t e r a c t i o n s at the prote in binding l e v e l . VPA i s known to d i s p l a c e phenytoin from i t s binding s i t e on serum albumin, and TABLE 4a. Serum to ta l , serum free and saliva concentrations (ug/mL) of VPA and their relationship to each other in volunteer W.T. before the administration of CBZ.  Time Serum Serum Sa l iva S a l i v a : to ta l S a l i v a : f ree Free: to ta l (h) total free r a t i o r a t i o r a t i o 1 52.95 6.75 0.891 0.017 0.132 0.127 2 54.57 5.29 1.020 0.019 0.193 0.097 3 47.50 4.60 1.114 0.023 0.242 0.097 5 41.30 4.36 0.756 0.018 0.173 0.106 7 38.50 3.15 0.472 0.012 0.150 0.082 9 32.47 2.45 0.323 0.010 0.132 0.076 12 25.80 1.88 0.345 0.013 0.184 0.073 Mean S.D. C .V . 0.0160 0.0042 26.25% 0.1723 0.0361 20.95% 0.0940 0.0175 18.62% C o r r e l a t i o n ( r ) : Between serum tota l and s a l i v a = 0.8945 : Between serum free and s a l i v a = 0.8254 : Between serum total and free = 0.9400 TABLE 4b. Serum to ta l , serum free and saliva concentrations (ug/mL) of VPA and their relationship to each other in volunteer W.T. after the administration of CBZ. Time (h) Serum tota l Serum free Sa l i va S a l i v a : to ta l r a t i o S a l i v a : f ree r a t i o Free: to ta l r a t i o 1 41.26 3.83 0.697 0.017 0.181 0.093 2 35.97 3.19 0.683 0.019 0.214 0.089 3 31.71 3.08 0.499 0.016 0.162 0.097 5 29.75 2.15 0.427 0.014 0.198 0.072 7 23.02 1.62 0.273 0.012 0.168 0.070 9 20.71 1.20 0.219 0.011 0.182 0.058 12 16.44 0.81 - - - 0.049 Mean S.D. C .V . 0.0148 0.0027 18.24% 0.1841 0.0175 9.50% 0.0754 0.0169 22.41% C o r r e l a t i o n ( r ) : Between serum total and s a l i v a = 0.9793 : Between serum free and tota l = 0.9615 : Between serum tota l and free = 0.9831 TABLE 5a. Serum tota l , serum free and saliva concentrations (ug/mL) of VPA and their relationship to each other in volunteer M.S. before the administration of CBZ. Time (h) Serum tota l Serum free S a l i v a S a l i v a : to ta l r a t i o S a l i v a : f ree r a t i o Free: to ta l r a t i o 1 92.63 8.33 1.533 0.017 0.184 0.090 2 63.60 9.48 1.020 0.016 0.108 0.149 3 65.31 9.25 1.274 0.020 0.138 0.142 5 50.74 6.48 1.093 0.022 0.169 0.128 7 51.06 8.74 1.029 0.020 0.118 0.171 9 51.57 7.94 - - - 0.154 24 21.48 1.75 0.343 0.016 0.196 0.081 Mean S.D. C .V . 0.0185 0.0023 12.43% 0.1522 0.0330 21.68% 0.1307 0.0311 23.78% Cor re la t ion ( r ) : Between serum tota l and s a l i v a = 0.9488 : Between serum free and s a l i v a = 0.8129 : Between serum total and free = 0.7463 TABLE 5b. Serum t o t a l , serum free and s a l i v a concentrat ions (ng /n l ) of VPA and t h e i r r e l a t i o n s h i p to each other 1n volunteer M.S. a f t e r the administrat ion of CBZ.  Time Serum Serum Sa l i va S a l i v a : tota l S a l i v a : f ree Free : to ta l (h) to ta l free r a t i o r a t i o r a t i o r o o O~7\M 0 0 7 07121 o.oss 1 57.60 6.81 - - - 0.118 2 50.38 3.64 0.670 0.013 0.184 0.072 3 49.69 3.35 0.472 0.010 0.140 0.067 5 41.26 2.31 - - - 0.056 7 38.04 2.01 0.612 0.016 0.304 0.053 9 29.64 - 0.411 0.014 12 26.03 1^94 0.355 0.014 0.182 0.075 Mean 0.0123 0.1870 0.0713 S.D. 0.0030 0.0628 0.0205 C.V . 24.24% 33.63% 28.83% C o r r e l a t i o n (r) : Between serum tota l and s a l i v a = 0.8493 : Between serum free and s a l i v a = 0.7500 : Between serum tota l and free = 0.8332 TABLE 6a. Serum to ta l , serum free and saliva concentrations (ug/mL) of VPA and their relationship to each other 1n volunteer R.M. before the administration of CBZ. Time (h) Serum tota l Serum free Sa l iva S a l i v a : to ta l r a t i o S a l i v a : f ree r a t i o Free: to ta l r a t i o 0 30.33 2.77 0.503 0.017 0.182 0.091 1 47.26 6.72 1.294 0.027 0.193 0.142 2 44.81 5.92 0.870 0.019 0.147 0.132 3 42.01 6.11 0.488 0.012 0.080 0.146 5 38.60 4.32 0.592 0.015 0.137 0.112 7 37.67 3.62 0.662 0.018 0.183 0.096 9 34.40 - - - - -12 29.47 2.89 _ _ _ 0.098 Mean S.D. C .V . 0.0180 0.0046 25 .66% 0.1536 0.0386 25.15% 0.1167 0.0213 18.30% Cor re la t ion (r) : Between serum total and s a l i v a = 0.8412 : Between serum free and s a l i v a = 0.6500 : Between serum total and free = 0.9556 Time (h) TABLE 6b. Serum to ta l , serum free and saliva concentrations (iig/mL) of VPA and their relationship to each other 1n volunteer R.M. after the administration of CBZ. Serum tota l Serum free Sa l i va S a l i v a : to ta l S a l i v a : free Free: to ta l r a t i o r a t i o r a t i o 0 1 2 3 5 7 9 12 22.79 42.43 39.35 30.56 27.61 24.10 18.11 6.69 6.82 3.10 2.56 1.66 1.20 0.175 0.590 0.388 0.757 0.430 0.268 0.185 0.008 0.014 0.010 0.014 0.010 0.010 0.089 0.111 0.139 0.105 0.154 0.157 0.101 0.092 0.069 0.066 0.0970 0.0328 33.835% Mean S.D. C .V . 0.0110 0.0022 20.32% 0.1196 0.0235 19.72% Cor re la t ion (r) : Between serum total and s a l i v a = 0.9011 : Between serum free and s a l i v a = 0.9549 : Between serum total and free = 0.9765 TABLE 7a. Serum to ta l , serum free and saliva concentrations (ug/mL) of VPA and their relationship to each other In volunteer B.A. before the administration of CBZ.  Time Serum Serum Sa l iva S a l i v a : to ta l S a l i v a : f ree Free: to ta l (h) tota l f ree r a t i o r a t i o r a t i o 1 42.36 5.40 - 0.128 - -2 42.44 5.38 1.064 0.025 0.198 0.127 3 43.78 5.49 1.522 0.035 0.277 0.125 5 20.56 - 0.611 0.029 - -7 22.27 3.39 1.270 0.057 0.375 0.152 9 25.90 2.82 1.233 0.047 0.437 0.109 12 21.07 2.76 0.773 0.035 0.280 0.131 Mean 0.0380 0.3134 0.1287 S.D. 0.0108 0.0834 0.0126 C.V. 28.63% 26.61% 9.79% Cor re la t ion (r) : Between serum tota l and s a l i v a = 0.6017 : Between serum free and s a l i v a = 0.4884 : Between serum tota l and free = 0.9733 TABLE 7b. Serum tota l , serum free and saliva concentrations (iig/mL) of VPA and their relationship to each other 1n volunteer B.A. after the administration of CBZ. Time (h) Serum tota l Serum free S a l i v a S a l i v a : to ta l r a t i o S a l i v a : f ree r a t i o Free: to ta l r a t i o 1 31.05 4.60 0.940 0.030 0.204 0.148 2 30.36 3.80 0.585 0.019 0.154 0.125 3 26.65 - 0.549 0.021 - -5 23.60 3.37 0.529 0.022 0.157 0.142 7 19.01 1.90 0.391 0.021 0.206 0.099 9 16.40 1.62 0.360 0.022 0.222 0.099 12 13.26 1.61 0.292 0.022 0.181 0.121 Mean S.D. c . v . 0.0224 0.0032 14.49% 0.1873 0.0254 13.56% 0.1223 0.0189 15.45% Cor re la t ion (r) : Between serum tota l and s a l i v a = 0.8727 : Between serum free and s a l i v a = 0.9354 : Between serum tota l and free = 0.9614 TABLE 8a. Serum tota l , serum free and saliva concentrations (pig/mL) of VPA and their relationship to each other in volunteer F.A. before the administration of CBZ. Time (h) Serum tota l Serum free Sa l i va S a l i v a : to ta l r a t i o S a l i v a : free r a t i o Free: to ta l r a t i o 0 44.45 8.21 0.889 0.020 0.108 0.185 1 64.66 13.71 2.963 0.046 0.216 0.212 2 56.58 13.11 2.432 0.043 0.186 0.231 3 58.77 11.04 3.230 0.055 0.293 0.188 5 40.59 9.18 1.218 0.030 0.133 0.226 7 46.84 7.99 1.209 0.026 0.151 0.171 9 46.06 7.88 0.483 0.011 0.061 0.171 12 33.61 6.45 0.559 0.017 0.087 0.192 Mean S .D . C . V . 0.0310 0.0145 46 .86% 0.1544 0.0706 45.74% 0.1979 0.0219 11.10% Cor re la t ion (r) : Between serum total and s a l i v a = 0.8933 : Between serum free and s a l i v a = 0.8818 : Between serum total and free = 0.9045 TABLE 8b. Serum t o t a l , serum free and s a l i v a concentrat ions (iig/mL) of VPA and t h e i r r e l a t i o n s h i p to each other 1n volunteer F . A . a f t e r the adminis t ra t ion of CBZ.  Time Serum Serum . Sa l iva S a l i v a : to ta l S a l i v a : free Free: to ta l (h) to ta l f ree r a t i o r a t i o r a t i o 0 24.15 4.44 0.353 0.015 0.080 0.183 1 53.24 8.73 - - - 0.163 2 43.80 7.17 0.998 0.023 0.139 0.163 3 42.02 7.33 1.294 0.031 0.177 0.174 5 33.34 4.86 0.773 0.023 0.159 0.145 7 26.09 4.61 0.572 0.022 0.124 0.176 9 28.02 5.44 - _ _ 0.194 Mean S.D. C .V . 0.0228 0.0051 22.26% 0.1405 0.0331 23.60% 0.1698 0.0149 8.76% Cor re la t ion (r) : Between serum tota l and s a l i v a = 0.9265 : Between serum free and s a l i v a = 0.9159 : Between serum tota l and free = 0.9654 TABLE 9. Time-averaged ratios (6-8 samples) and correlations between serum to ta l , serum free and saliva VPA concentrations in f ive volunteers before the administration of CBZ. Volun- S a l i v a : S a l i v a : Free: C o r r e l a t i o n (r) teer total r a t i o free r a t i o to ta l r a t i o T o t a l : S a l i v a Free: S a l i v a T o t a l : Free W.T. 0.0160 0.1723 0.0940 0.8945 0.8254 0.9400 M.S. 0.0185 0.1522 0.1307 0.9488 0.8129 0.7463 R.M. 0.0180 0.1536 0.1167 0.8412 0.6500 0 . 9556 F . A . 0.0310 0.1544 0.1970 0.8933 0.8818 0.9045 B.A. 0.0380 0.3134 0.1287 0.6017 0.4884 0.9733 Mean 0.0243 0.1892 0.1334 0.8945* 0.7925* 0.9039 S.D. 0.0086 0.0625 0.0343 C.V . 35.63% 33.06% 25.76% * excluding B . A . ' s values TABLE 10. Time-averaged ratios (6-8 samples) and correlations between serum to ta l , serum free and saliva VPA concentrations in f ive volunteers after the administration of CBZ. Volun- S a l i v a : S a l i v a : Free: C o r r e l a t i o n (r) teer total r a t i o free r a t i o tota l r a t i o T o t a l : S a l i v a Free: S a l i v a T o t a l : Free W.T. 0.0148 0.1841 0.0754 0 . 9793 0.9615 0.9831 M.S. 0.0123 0.1870 0.0713 0.8493 0.7500 0.8332 R.M. 0.0110 0.1195 0.0970 0.9011 0.9549 0.9765 F . A . 0.0228 0.1405 0.1698 0 . 9265 0.9159 0.9654 B.A . 0.0224 0.1873 0.1223 0.8727 0 . 9354 0.9614 Mean 0.0167 0.1637 0.1072 0.9058 0.9035 0.9439 S.D. 0.0050 0.0282 0.0362 C .V . 29.96% 17.27% 33.77% - 73 -s a l i c y l a t e in turn d isp laces VPA from i t s binding s i t e s in plasma (Levy and Koch, 1962). VPA i s a lso l i k e l y to compete with CBZ (75% bound) for prote in binding s i t e s . It has been shown in v i t r o that VPA reduces pro te in binding of CBZ, whereas addi t ion of CBZ to VPA did not change the serum binding of VPA (Mattson et a l . , 1982; Patel and Levy, 1979). Th is was explained by the greater binding a f f i n i t y of VPA compared to that of CBZ. Our in vivo resu l ts a lso confirm the above in v i t r o f i n d i n g . In the present study the free f r a c t i o n of VPA was ac tua l l y s l i g h t l y reduced (from 0.1334 to 0.1072) because of the lower serum to ta l VPA concentrat ions a f te r CBZ. Hence, in v i v o , CBZ does not d i s p l a c e VPA from plasma prote in binding s i t e s . In a study of the e f f e c t s of carbamazepine on VPA k i n e t i c s in normal s u b j e c t s , Bowdle et a l . (1979) found that the minimum steady-s ta te l e v e l s of VPA decreased s i g n i f i c a n t l y a f te r CBZ. Our r e s u l t s a lso showed a s i g n i f i c a n t decrease in VPA serum and s a l i v a concentrat ions a f te r two weeks on carbamazepine. An increase in the f ree f r a c t i o n of a drug usual ly r e s u l t s in increased clearance r a t e s , reduced s teady-s ta te tota l l e v e l s , and hence unchanged free concentra t ions (Koch-Weser and S e l l e r s , 1976). However, the decrease in VPA tota l l e v e l s observed cannot be due to an increase in the free f r a c t i o n since the free f r a c t i o n did not increase a f te r CBZ. Th is work was part of a general study on the e f f e c t of CBZ on VPA metabolism and was done s p e c i f i c a l l y to measure any e f f e c t of CBZ on the free f r a c t i o n of VPA. The e f f e c t of CBZ on VPA metabolism was f i r s t studied by Sukhbinder Panesar ( M . S c , 1987). It was found that VPA volume of d i s t r i b u t i o n did not change but that clearance and the e l iminat ion rate constant were increased s i g n i f i c a n t l y (p<0.05) - 74 -in the f i v e volunteers a f te r CBZ admin is t ra t ion . This suggests that CBZ induces the metabolism of VPA. 2. Serum free VPA l e v e l s The serum free l e v e l s of VPA were measured employing the technique of u l t r a f i l t r a t i o n . We chose u l t r a f i l t r a t i o n over equ i l ib r ium d i a l y s i s because of i t s s i m p l i c i t y and a previous report that free l eve ls of VPA obtained by u l t r a f i l t r a t i o n and equ i l ib r ium d i a l y s i s were found to be s t rongly cor re la ted (Levy et a l . , 1984b). There is no adsorpt ion of VPA to the f i l t r a t i o n membranes or f i l t r a t i o n device (Nau et a l . , 1984). The free VPA concentrat ions in th is study ranged from 13.71 to 0.81 ug/mL compared to tota l VPA concentrat ions of 92.63 to 13.26 Lig/mL. The free concentrat ions were highly cor re la ted with total concentrat ion over the concentrat ion range studied (Tables 4 to 8) ( e . g . , r=0.9439 ± 0.05, mean ± S . D . , a f te r CBZ, Table 10). The free f r a c t i o n s var ied from 0.05 to 0.23 and were more or less concentrat ion dependent. For a l l except one volunteer (B.A. ) the free f r a c t i o n s decreased with a decrease in tota l VPA concent ra t ion , i . e . a f te r CBZ. For B.A. there was no change in free f rac t ion poss ib ly because most of the tota l VPA concentrat ion values were less than 30 Lig/mL. The overa l l mean free f r a c t i o n was 0.1334 ± 0.0343 before CBZ and 0.1072 ± 0.0362 a f t e r CBZ (Tables 9 and 10). This concentrat ion dependence of free f r a c t i o n was evident even at VPA tota l concentra-t ions of less than 50 ug/ml. The free f r a c t i o n s obtained in th is study are in agreement with publ ished data for e p i l e p t i c pat ients (Otten et - 75 -TABLE 11. Decrease [%) 1n average VPA a d m i n i s t r a t i o n . concentrat ion a f t e r CBZ Volunteer Serum Total Serum Free S a l i v a W.T. 32.15 44.24 38.85 M.S. 28.89 54.49 47.21 R.M. 21.94 13.91 40.85 F . A . 29.97 40.13 55.56 B.A. 26.58 31.49 58.20 Mean 27.91 36.85 48.13 S .D . 3.481 13.649 7.701 - 76 -a l . , 1984; Garnett et a l . , 1983). The c o r r e l a t i o n between total VPA concentrat ion and free f r a c t i o n s was much lower than that of the total to f ree concentrat ion (0.336 v s . 0.787; c a l c u l a t i o n based on a l l values before and a f te r CBZ for a l l vo lun teers ) . The p lo t of total VPA concentrat ion versus free f r a c t i o n (Figure 22) appears to curve upwards showing the concentrat ion dependent binding of VPA. Because of the saturable nature of VPA plasma prote in b ind ing , the f ree f r a c t i o n i s not constant over the range of VPA concentrat ions used t h e r a p e u t i c a l l y as demonstrated in th is study and other studies (Levy et a l . , 1986). In a d d i t i o n , VPA binding is a f fected by free fa t ty acid concentrat ion and s a l i c y l a t e s , and can be modif ied by a number of d isease c o n d i t i o n s . For these reasons, total VPA l e v e l s do not r e f l e c t the free leve l which i s assumed to be the pharmacological ly ac t ive form. In sp i te of the strong c o r r e l a t i o n between total and free VPA, f ree concentrat ions cannot be accurate ly predicted from total VPA concentrat ions because of the high i n t e r - and i n t r a - s u b j e c t v a r i a b i l i t y . Hence, in s i t u a t i o n s where drug monitoring is requ i red , the monitoring of the free l e v e l s rather than the total i s l i k e l y to be more u s e f u l . 3. S a l i v a VPA l e v e l s The s a l i v a concentrat ions of VPA in the f i v e volunteers ranged from 0.175 to 3.23 ug/mL compared to serum tota l VPA concentrat ions of 13.26 to 92.63 ug/mL. The time-averaged s a l i v a to tota l and free VPA r a t i o s are given in Tables 4 to 8. The mean r a t i o s for the f i ve - 77 -v ° n „ 9 • • O • R.M. • B.A. v F .A. O • O • " " 1 1 1 1 1 0 2 4 6 8 10 12 Time, h Figure 14. Sa l i va concentra t ion- t ime p r o f i l e s for f i v e volunteers at steady s ta te VPA. - 78 -T i m e , h Figure 15. Concentrat ion- t ime curves for serum t o t a l , serum free and s a l i v a VPA in one volunteer ( M . S . ) . - 79 -1CH T i m e , h Figure 16. Concentrat ion- t ime curve for serum free VPA before and a f te r CBZ admin is t ra t ion in one volunteer (W.T . ) . - 80 -10-i 0 .1+ 0 Figure 17. 6 8 10 T i m e , h Concentrat ion- t ime curve for s a l i v a VPA before and a f t e r CBZ admin is t ra t ion in one volunteer (W.T . ) . Figure 18. Re la t ionship between serum tota l and s a l i v a VPA concentrat ions in one volunteer (W.T.) . 1.2 • Before CBZ O After CBZ S e r u m free V P A , u g / m L Figure 19. Re la t ionship between serum free and s a l i v a VPA concentrat ions in one volunteer (W.T. ) . Figure 20. The r e l a t i o n s h i p between serum to ta l and s a l i v a VPA concen-t ra t ions in a l l f i ve vo lunteers . 3.5-1 2.5-S e r u m free V P A , u g / m L Figure 2 1 . The r e l a t i o n s h i p between serum free and s a l i v a VPA concentra -t ions in a l l f i v e vo lunteers . - 85 -0 . 2 5 - 1 0 . 2 0 -O 0 . 1 5 -D . £ 0.10-^ 0 . 0 5 -• • • • • • • 0 . 0 0 - I— 50 i 10 I 2 0 i 3 0 i 4 0 i 6 0 Serum total VPA, ug/mL 7 0 Figure 22. A p lo t of free f r a c t i o n versus serum t o t a l concentrat ion of VPA (a l l values before and a f t e r CBZ fo r a l l f i v e vo lun teers ) . - 86 -volunteers were 0.0243 ± 0.0086, and 0.1892 ± 0.0625, for s a l i v a to to ta l and s a l i v a to f r e e , r e s p e c t i v e l y (Table 9) . These ra t ios were 0.0167 ± 0.0050 and 0.1637 ± 0.0282 a f t e r CBZ (Table 10). The time-averaged s a l i v a to serum free r a t i o s of four of the f i v e volunteers were very s i m i l a r (0.1723, 0.1522, 0.1536, 0.1544) as were the s a l i v a to tota l r a t i o s in three of the volunteers (0.0160, 0.0185, 0.0180) (Table 9 ) . The s a l i v a to free r a t i o s in four of the vo lunteers , a f ter CBZ (Table 10) were s i m i l a r and did not change markedly from the values obtained before CBZ. However, th is was not true of the s a l i v a to total r a t i o s . The s a l i v a to serum tota l VPA r a t i o s were in general agreement with publ ished data for healthy volunteers on mul t ip le doses (Gugler et a l . , 1977) but were higher than those of Abbott et a l . (1982) (s ing le -dose study) and Acheampong et a l . (1984) (mult ip le-dose study) . The s a l i v a to free r a t i o s were s i m i l a r to those reported by Abbott et a l . (1982) and Acheampong et a l . (1984) and were a lso comparable to values found for e p i l e p t i c pat ients (Gugler et a l . , 1980). The c o r r e l a t i o n s between s a l i v a VPA concentrat ions and both serum to ta l and free concentrat ions were very good in the majori ty of cases (Tables 9 and 10) . As seen in Figure 21 the overa l l c o r r e l a t i o n between serum free and s a l i v a VPA concentrat ion was also s t rong. a . A n a l y t i c a l aspects of s a l i v a VPA quant i ta t ion S a l i v a VPA concentrat ions have been measured by a number of a n a l y t i c a l techniques; GC (Blom and Guelen, 1977; Gugler et a l . , 1977; Fung and Ueda, 1982; Gugler et a l . , 1980); GCMS (Abbott et a l . , 1980; Acheampong et a l . , 1982); EMIT (Monaco et a l . , 1982). With the EMIT assay , no c o r r e l a t i o n was found between s a l i v a and plasma VPA and in - 87 -more than 60% of the cases s a l i v a l e v e l s were near zero . The c o r r e l a t i o n s found by Blom and Guelen were poor. Gugler et a l . (1977) a lso reported a high i n t r a - and i n t e r - s u b j e c t v a r i a b i l i t y of the s a l i v a to plasma r a t i o s . The VPA s a l i v a leve l work reported prev iously had shown that the s a l i v a concentrat ion of VPA is not equal to the unbound concentrat ion in serum, s a l i v a l e v e l s can often be e r r a t i c , a cons is ten t ra t io between s a l i v a and serum tota l or free i s not obta ined, and there is a high in te rsub jec t v a r i a b i l i t y . It has been suggested that because of i t s low pKa (4 .9 ) , VPA w i l l be mostly ion ized at plasma pH and hence l i t t l e of i t excreted into s a l i v a and that the s a l i v a concentrat ion of VPA w i l l be highly dependent on s a l i v a pH. In a d d i t i o n , the low s a l i v a concentrat ion may prove d i f f i c u l t to c o r r e c t l y measure by most a n a l y t i c a l techniques (Gugler et a l . , 1977). Although s a l i v a l e v e l s of VPA are not equal to the unbound concentrat ion in plasma, i t was of i n t e r e s t to f ind out whether a constant r e l a t i o n s h i p between the two e x i s t e d , so that the unbound concentrat ion could be estimated from s a l i v a r y l e v e l s . It was our in ten t ion to do the present study by incorporat ing two important f a c t o r s . We used a highly s e n s i t i v e and prec ise negative ion GCMS assay that can measure VPA in serum and s a l i v a accurate ly down to about 2 ng/mL. The other important fac tor was the use of reproducible and standardized methods of c o l l e c t i o n of s a l i v a samples. b. The e f f e c t of pH on the s a l i v a concentrat ion of VPA Since the secre t ion of r e l a t i v e l y strong a c i d i c and basic drugs in to s a l i v a is dependent upon the pH of s a l i v a and serum (Mucklow, - 88 -1982), i t i s genera l ly assumed that s a l i v a pH w i l l a f f e c t the s a l i v a concentrat ion of VPA. However, Abbott et a l . (1982) have observed h igh ly i n f l a t e d s a l i v a l e v e l s of VPA which could not be explained by pH e f f e c t s or f ree l e v e l s of the drug. S i m i l a r l y , Acheampong et a l . (1984) found d iscrepanc ies between the experimental ly obtained values and those c a l c u l a t e d using Mat in 's equation (Matin et a l . , 1974) for s a l i v a l e v e l s of VPA. It i s d i f f i c u l t to get an accurate measure of s a l i v a pH by conventional means because of the poss ib le d i f fe rence in pH of mixed s a l i v a compared to the pH of s a l i v a in contact with the e p i t h e l i a l c e l l s of the s a l i v a r y glands (Koup et a l . , 1975). For diazepam, i t has been found that although s a l i v a to plasma r a t i o s may show good c o r r e l a t i o n , s a l i v a r y diazepam l e v e l s are higher than the f ree diazepam concentrat ion in plasma (Gier et a l . , 1980). Since the pKa of diazepam (a weak base) i s 3 .3 , the concentrat ion of diazepam in s a l i v a i s not l i k e l y to be a f fec ted by s a l i v a r y pH. In the case of s a l i c y l a t e , a r e l a t i v e l y strong acid (pKa 3 ) , a high i n t r a - s u b j e c t v a r i a t i o n was found that was not re la ted to s a l i v a pH (Levy et a l . , 1980). It does not appear that the ro le of pH in the s a l i v a r y excret ion o f ind iv idua l drugs i s f u l l y understood. Against t h i s background, i t was f e l t that i t was more important to standardize s a l i v a c o l l e c t i o n by s t imula t ion with c i t r i c ac id and c o l l e c t i n g sample over a given time in te rva l rather than measuring whole mouth s a l i v a pH which would e i ther not be accurate (a f fec ted by c i t r i c acid) or otherwise would not r e f l e c t the pH in the s a l i v a r y g lands. In a study (Abbott , unpublished data) of the r e l a t i o n s h i p between s t imula t ion of s a l i v a flow with c i t r i c ac id and ensuing parot id s a l i v a r y pH, i t was found that the s a l i v a pH in s ix volunteers was reasonably constant fo l lowing s t imula t ion by 4 mL of 5% c i t r i c ac id - 89 -s o l u t i o n retained in the mouth for 2 minutes. The parot id s a l i v a pH plateaued between 2-3 minutes and was in the range of pH 7 .3 -7 .6 . This protocol was thus used for the c o l l e c t i o n of s a l i v a samples throughout the present study, in the hope to minimize i n t r a - and i n t e r - s u b j e c t v a r i a b i l i t y . c . The r e l a t i o n s h i p between s a l i v a and serum concentrat ions of VPA In Tables 4 to 8 i t can be seen that there are occasional s a l i v a VPA values which are d ispropor t ionate ly high and thus contr ibute to decreased c o r r e l a t i o n (both free and to ta l ) and high v a r i a b i l i t y in the s a l i v a to tota l or free r a t i o s . However, examination of Table 9 shows that in four of the f i v e volunteers the mean s a l i v a to free ra t ios are remarkably constant . The mean value for these four volunteers is 0.1581 with c o e f f i c i e n t of v a r i a t i o n of 5.2%. I n t e r e s t i n g l y , for four of the vo lunteers , s i m i l a r ra t ios were observed a f te r CBZ with a mean of 0.1747 and c o e f f i c i e n t of v a r i a t i o n of 11.3%, and for the three volunteers these values are s i m i l a r to those before CBZ. This good agreement between the s a l i v a to free ra t ios obtained two weeks apart , a t l e a s t in three of the volunteers could be a t t r ibu ted to the standardized sampling protocol and p r e c i s i o n of the assay. For one of the volunteers (B.A) the s a l i v a to free VPA r a t i o decreased from 0.3134 to 0 .1873-a f ter CBZ. This appears to cont rad ic t the f ind ing of Acheampong et a l . (1984) who reported that low s a l i v a to f ree ra t ios were observed at high serum concentrat ions of VPA. In f a c t , s a l i v a to free ra t ios should not change depending upon VPA serum concentrat ion since any changes in free VPA should bring a proport ional - 90 -change in s a l i v a l e v e l s . This i s cons is ten t with our resu l ts since there i s e s s e n t i a l l y no d i f fe rence between the s a l i v a to serum free r a t i o in a l l subjects before and a f ter CBZ except for B.A. whose ra t io decreases a f te r CBZ. The s a l i v a to serum tota l r a t i o did not show the good agreement exh ib i ted by the s a l i v a to free serum r a t i o before and a f te r CBZ. Af ter CBZ there was a s i g n i f i c a n t decrease in the s a l i v a to total r a t i o at l e a s t in four of the volunteers i n d i c a t i n g that the s a l i v a to total r a t i o was concentrat ion dependent, higher ra t ios being found at higher serum tota l concent ra t ions . Since serum tota l VPA concentrat ions decreased by 27.91S ± 3.48 (Mean ± S.D.) a f te r CBZ (Table 11) the % f ree decreases (36.85) and the re fo re , r e s u l t s in a corresponding % decrease in s a l i v a concentrat ions (48.13). The c o r r e l a t i o n between s a l i v a and serum tota l or free was bet ter for the samples a f te r CBZ (Table 9 and 10) . S i m i l a r l y , the i n t r a i n d i v i d u a l v a r i a b i l i t y was less a f t e r CBZ. There was no d i f fe rence between the v a r i a b i l i t y of s a l i v a to to ta l and s a l i v a to free r a t i o s . In summary, despi te i n t r a i n d i v i d u a l v a r i a b i l i t y the s a l i v a VPA to serum free r a t i o in three of the f i v e volunteers was reasonably constant over two d i f f e r e n t sampling per iods , two weeks apar t . For the s a l i v a VPA to tota l r a t ios such a r e l a t i o n s h i p was not found because of the decreased serum concentrat ion and accompanying decrease in free f r a c t i o n fo l lowing CBZ admin is t ra t ion . From t h i s work i t appears the time-averaged s a l i v a to free ra t io once determined could be used for assessing serum free VPA by measuring VPA in s a l i v a . The good c o r r e l a t i o n s found between s a l i v a and both serum tota l and free VPA concentrat ions suggest that measuring s a l i v a - 91 -VPA by t h i s method would be su i tab le for pharmacokinetic and drug i n t e r a c t i o n s t u d i e s . S a l i v a r y measurement of VPA could be useful for therapeut ic drug monitoring but i t i s u n l i k e l y to be of value using rout ine assay methods that do not have the s e n s i t i v i t y and accuracy of NICI. For example, Fung and Ueda (1982) were not able to detect any VPA in some of the i r s a l i v a samples using GC. d . The concept of s a l i v a r y clearance in the s a l i v a r y excret ion of drugs Zuidema and van Ginneken (1983a) and Kido (1982) have introduced the concept of s a l i v a r y c learance in an attempt to expla in some of the i r r e g u l a r i t i e s observed with s a l i v a r y l e v e l s of drugs. When a drug i s poor ly l i p o p h i l i c and hence with a low ext rac t ion r a t i o , permeation of the drug across the e p i t h e l i a l membrane i s ra te -determin ing. The s a l i v a c learance i s then independent of blood flow whereas s a l i v a flow remains dependent on the blood f low. The s a l i v a r y concentrat ion of the drug w i l l decrease with increas ing blood and s a l i v a flow since the drug i s not able to e q u i l i b r a t e between s a l i v a and blood at high flow because of i t s poor l i p i d s o l u b i l i t y . Such substances as urea and primidone, show flow dependent s a l i v a r y excret ion (Barte ls et a l . , 1979). Since drugs with good l i p o p h i l i c i t y have a high ex t rac t ion r a t i o , s a l i v a r y c learance w i l l be proport ional to blood flow and as a consequence s a l i v a flow and clearance w i l l change p r o p o r t i o n a l l y . In a study (Kido, 1982) with some ant iconvulsant drugs, where the k i n e t i c s of the drugs i n s a l i v a were studied a f te r s t imula t ion with p i l o c a r p i n e , s a l i v a r y l e v e l s of phenobarbital showed maximum values between 10 and 30 minutes. This observat ion could not be explained with Mat in 's equat ion. - 92 -With the add i t ion of the c learance parameter, however, the phenobarbital l e v e l s were bet ter expla ined. S a l i v a to plasma r a t i o s greater than one can be explained by e i t h e r ac t ive t ranspor t or i o n i z a t i o n in s a l i v a ( r e l a t i v e l y strong b a s e s ) . An a c t i v e t ranspor t mechanism appears to be true for p e n i c i l l i n (Zuidema and van Ginneken, 1983b). In the present study s a l i v a was obtained fo l lowing st imulat ion with c i t r i c a c i d . Whereas t h i s helps to e l iminate v a r i a t i o n s in res t ing s a l i v a pH between sampling per iods , a new var iab le can be introduced -s a l i v a flow ra te . This may par t l y expla in the e r r a t i c s a l i v a VPA l e v e l s observed. VPA i s poorly l i p o p h i l i c compared to other a n t i e p i l e p t i c drugs such as CBZ, phenytoin and phenobarbital (Goldberg and Todorof f , 1980). However, s ince the concentrat ion of VPA was found to be higher in the c i t r i c ac id st imulated s a l i v a compared to non-st imulated s a l i v a (Abbott , unpublished observa t ion ) , other fac tors such as f a c i l i t a t e d t ranspor t of VPA may be responsib le for the incons is ten t s a l i v a l e v e l s o f VPA. It has been s t ressed that l i t t l e VPA i s secreted into s a l i v a because of i t s low pKa. On the other hand the cerebrospinal f l u i d (CSF) to serum VPA concentrat ion r a t i o has been found to be 0.10 (Blom and Guelen, 1977), 0 .11 , (Monaco et a l . , 1982) and 0.08 (Kido, 1982). These concentra t ions in CSF are more or l ess equal to the unbound concentra-t i o n in serum and i t i s general ly accepted that the CSF concentrat ion r e f l e c t s the free drug in plasma. The normal pH of the CSF i s 7.32 (West, 1985). In the study by Abbott (unpublished data) i t was found that fo l lowing s t imula t ion of s a l i v a flow with c i t r i c ac id in s ix vo lunteers the parot id s a l i v a pH var ied very l i t t l e a f te r 2 minutes of the i n i t i a l s t imula t ion by c i t r i c ac id (7.46 ± 0 .14, n=20). Thus there - 93 i s l i t t l e d i f f e rence between st imulated s a l i v a and CSF pH whereas there i s almost a 5-10 f o l d d i f f e rence in the concentrat ion of VPA in the two t i s s u e s . Cornford et a l . (1985) have suggested that a f r a c t i o n of the VPA enter ing the c a p i l l a r i e s in the prote in bound form has the capaci ty to e q u i l i b r a t e with bra in because of enhanced drug d i s s o c i a t i o n from albumin in the brain c i r c u l a t i o n and that VPA i s a c t i v e l y transported out of the b r a i n . It has a lso been demonstrated in dogs that VPA is t ransported out of the CSF by the same anion e f f l u x mechanism that t ranspor ts y - aminobuty r ic ac id and probenecid out of the CSF (Loscher and F r e y , 1982). In sp i te of t h i s i t was found in humans that the concentrat ion of VPA in CSF was 7.6 to 25.0% of the tota l plasma concentrat ions in the range of 35.5 to 150.4 ug/mL (Vajda et a l . , 1981). This CSF concentrat ion apparently r e f l e c t s the free f r a c t i o n , in plasma, although free l e v e l s were not determined in t h i s study. From the above c o n s i d e r a t i o n s , i t i s conceivable that there i s an ac t ive t ranspor t of VPA out of s a l i v a which would expla in the low concentrat ion in s a l i v a r e l a t i v e to free serum and CSF concent ra t ions . C . I d e n t i f i c a t i o n of VPA metabol i tes using NICI-GCMS of t h e i r PFB d e r i v a t i v e s The to ta l ion current chromatogram p lo t of the PFB der iva t i zed ur ine ex t rac t from one volunteer at VPA steady state administered s e l e c t e d doses of [ 2H5]-VPA i s shown in Figure 23. Th is TIC p lot conta ins peaks for VPA and 14 VPA metabol i tes: 3-ene VPA, (Z)-2-ene VPA, (E)-2-ene VPA, ( E , Z ) - 2 , 3 ' - d i e n e VPA, 2,4- diene VPA, ( E , E ) - 2 , 3 ' - d i e n e VPA, a new VPA metabol i te (peak 10) , 3-keto VPA, 3-OH VPA diastereomers, 4-keto VPA, 4-OH VPA, 5-OH VPA, 2-PSA and 2-PGA. A l l 7 8 9 10 11 12 13 20 21 22 TIME (min) F igure 23. Total ion current p l o t , in the NICI mode, of the PFB d e r i v a t i z e d ur ine ext ract from a volunteer on VPA steady s t a t e , a lso given se lec ted doses of [ Hg]-VPA. Peak numbers correspond t o : 1= VPA, 2= 3-ene VPA, 3= (Z)-2-ene VPA, 4= (E) -2-ene VPA, 5= ( E , Z ) - [ 2 H 6 ] -2 ,3 ' -d iene VPA, 6= ( E , Z ) - 2 , 3 ' - d i e n e VPA, 7= 2 ,4-d iene VPA, 8= ( E , E ) - [ 2 H 6 J - 2 , 3 ' - d i e n e VPA, 9= ( E , E ) - 2 , 3 ' - d i e n e VPA, 10= 4 ' -ke to -2 -ene VPA, 11= 3-keto VPA, 12= 3-OH VPA, 13= 3-OH VPA, 14= 4-keto VPA, 15= 4-OH VPA, 16= 5-0H VPA, 17= 2-PSA, 18= 2-PGA. - 95 peaks of i n t e r e s t were su i tab ly resolved except those of 4-ene VPA and VPA. The 4-ene VPA peak i s swamped by the huge VPA peak. The existence o f the 4-ene VPA peak under the VPA peak was v e r i f i e d by obtaining the mass chromatograms at m/z 141. One of these peaks had ident ica l re tent ion time to that of in jec ted synthet ic 4-ene VPA (see l a t e r in F igure 37 ) . The i d e n t i f i c a t i o n of the i so la ted metabol i tes was f a c i l i t a t e d by the doublet fragment ions (deuterated and undeuterated) in the i r negative ion spectra and with the help of synthet ic reference compounds. The negative ion spectra of these ur inary metabol i tes along with the i r synthet ic standards are i l l u s t r a t e d in Figures 24 through 36. Since the urine sample analyzed contained mainly deuterated VPA and metabo l i tes , the i n t e n s i t i e s of l a b e l l e d and unlabel led ions were not equal in the spectra obtained for the drug and metabol i tes . 1. Negative ion spectra of PFB der iva t i zed VPA metabol i tes In the negative ion mass spectra of VPA and i t s metabol i tes almost a l l of the ion current i s c a r r i e d by a s ing le fragment an ion , [M-pentaf luorobenzyl (181)]~ an ion . This h ighly abundant, resonance s t a b i l i z e d carboxylate anion i s formed under d i s s o c i a t i v e e lec t ron capture by cleavage of the PFB-oxygen bond. Since excess energy from the i o n i z a t i o n process i s d iss ipa ted by bond c leavage, t h i s e lec t ron capture NICI technique i s a highly e f f i c i e n t soft i o n i z a t i o n process. In Figures 24 to 37 the negative ion spectra of VPA and i t s metabol i tes are presented. A l l the metabol i tes and the parent drug have the [M-181] - ion as t h e i r base peak, the only exception being 3-keto - 96 -A ) 100 t . d _1 L U CC 0 I. 50 149 143 150 CHj-CHj-CH^ 0 > 'CH—C—O-r-CHs-CeF CH^CH^ -143 ,2, "2 L6 r5 ( \ . 149) M / Z [ r 250 B) 100 X . L U a: 0 1 . 50 143 CH^—CH^-Cr^ CH—!!—04CH«-C.F CH^—CH^—C^ 143-n2 k 6 r 5 150 250 M / Z Figure 24. NICI mass spectrum of A) VPA-PFB ( H 0 and B) synthet ic VPA-PFB. - 97 -A) 100 _ 147 141 CH r CH5-CH, 0 / H - C - ° t C H 2 - C 6 F 5 CHj-CH=CH 141 — ( Z H 6 , 147) 50 150 M/Z 250 350 B) 100 _ 141 C H r C H r C H , 0 "3 ""2 "n2. CHj-CH = CH / C H - C — O + CH^-CgFj 141-50 150 M/Z ~ i 250 350 Figure 25. NICI mass spectrum of A) 3-ene VPA-PFB ( 2 H n + 2 H C ) (peak 2 , F ig 23) and B) syn the t ic 3-ene VPA-PFB. u 6 - 98 -A ) 100 147 L U cr 141 CH5-CH5-CH CHj-CH^-CH' —C—O-j-CHj- C6F5 141 ,2, I H 6 , 147) D J _ 50 150 M / Z 250 B) 'DO » _ 141 in z L U CH^—CH^—CHj CHg-CHj-CH 141 "2 u 6 ' 5 U J > L U 0 X _ 50 150 M / Z 250 350 Figure 26. NICI mass spectrum of A) (E)-2-ene VPA-PFB ( H 0 + H 6 ) (peak 4, F i g . 23) and B) syn the t ic (E)-2-ene VPA-PFB. - 99 -A) 100 _, >-i — C O en 50 144 139 CHr-CHr-CH, 0 3 2 2 N ii C - C - 0 + C H 5 - C t F c CHj=CH-CH " 2 l f 5 J L 150 M/Z 139 — ( 2 H 5 , 144) 250 350 B) 100 >-i — CD U J or 50 139 C H - C H 2 - C H 2 x 0 C - C - 0 CH^CH-CH 139 C H 2 - C 6 F 5 150 M/Z 250 350 Figure 27. NICI mass spectrum of A) 2 ,4 -d iene VPA-PFB ( 2 H 0 + 2 H 5 ) (peak 7, F i g . 23) and B) syn the t ic 2 ,4-d iene VPA-PFB. - 100 -A) 100 _, >-UJ cc 50 145 C-Hr-CH = CH 0 3 \ II 2 # C Hj-CH^-CH C - C - 0 + CHs-C,F 145-150 M/Z 250 2 "6 r 5 1 1 350 B) 100 _ >-C O L U CC 50 139 CHr-CH=CH 0 3 \ II C - C - 0 + CHj-CH^-CH 139-C H 2 - C 6 F 5 I I 150 250 M/Z 1 I 350 Figure 28. NICI mass spectrum of A) ( E , E ) - 2 , 3 ' - d i e n e VPA-PFB ( H f i) (peak 8, F ig .23) and B) synthet ic ( E , E ) - 2 , 3 ' - d i e n e VPA-PFB. - 101 -A ) 100 X _ , 158 > i — c r _j L U C C 0 X. 0 II CHj-C—CH 2 0 C H J - C H J - C H C— C — O - K H ^ C g F j 155 /2 ( H 3 , 158) 100 % _ to > cx _ i L l J CC 0 X. M / Z ~T 300 ~~1 400 111 ( M - 1 5 9 ) + 114 (M-159 + 2 H 3 ) + 213 ( M - 5 7 ) + 171 141 iee ,216 (M-57 + 2 H 3 ) + i r 270 ( M ) + J . 200 300 m/z 2 2 Figure 29. A) NICI mass spectrum o f 4 ' - ke to -2 -ene VPA-PFB ( Hn + H 3 ) (peak 10, F i g . 2 3 ) , B) EI(t-BDMS) mass spectrum of 4 ' -ke to -2 -ene VPA. - 102 -A) 100 X _ , z L U » — cr _j L U CC 0 t. 50 119 113 0 CH^-CH^-C . p CH5-CHj-CH 2 CH - f -C -0 -CH«-C r F, 158 r 6 s 113-1 ( 2Hg, 119) 163 i — r -150 M / Z 250 i 1— 350 B) 100 > i — cc _ J L U CC 0 I. 50 113 CH^CH^-C 0 2 \ , I II CH C H j - C H ^ ^ 4 c ' — O - C H p C g F j 113 150 250 M / Z 337 350 Figure 30. NICI mass spectrum of A) 3-keto VPA-PFB ( 2 H n + 2 H C ) fneak 11 F i g . 23) and B) syn the t ic 3-keto VPA-PFB. ° 6 - 103 -A ) too t . in UJ 0 t _ l _ 50 165 159 150 OH I CHj-CH^-CH^ CH—C-O+CHy-C^F, 159 ,2, TVs ( 'H 6 , 165) 250 350 M / Z B)l00 x . 159 i n z U J z OH CHT-CH5-CH 0 CH—C—0-159 -C H r C 6 F 5 cr _ i U J on 0 x_ 50 150 250 350 M / Z Figure 31 N I C I mass spectrum o f A) 3 - 0 H VPA-PFB ( HQ + H 6) (peaks 12 and 1 3 , F i g . 2 3 ) and B) syn the t ic 3 - 0 H VPA-PFB. - 104 -A ) 100 X . >-t— to CC 0 X. 160 157 I CHs-C-CH CHy-CH2-CH2 2 - - - c ' - o i CH-157 ,2, •CH^gFg ( H 3 . 160) i , a , L 50 150 250 M / Z 350 B ) I 0 0 X L U 0 X. 157 0 U C H T - C — C H , 0 3 2 \ II C H J - C H J - C H J C H — C — 0 + C H 5 5 - C , F 157-" T L 6 r 5 M / Z 50 150 350 2 2 Figure 32. NICI mass spectrum of A) 4-keto VPA-PFB ( Hg + H 3 ) (peak 14, F i g . 23) and B) syn the t ic 4-keto VPA-PFB. - 105 -A) 100 _ >-C D or 50 165 159 1 r OH CHr-CH-CH, 0 •» * \ II C H - C - 0 4 C H 5 - C , F CHj-CH^-C^ 150 M/Z "2 L 6 r 5 159 ( 2 H 6 , 165) 250 350 B) 100 _ >-C D L U L U 50 159 OH I C H r C H - C H , 0 3 2 v II C H - C - 0 + CH5-C c F c CHj- CH^-CH^ i r 159-"2 "6 5 150 M/Z 250 350 Figure 33. NICI mass spectrum o f A) 4-OH VPA-PFB ( 2 H n + 2 H f i ) (peak 15, F i g . 23) and B) syn the t ic 4-OH VPA-PFB. - 106 -A) 100 _, >-I— C O z L U I— z L U cn 50 164 159 OH I CH5-CH5-CH, 0 2 2 2 v v 11 , C H - C - O f CH=-C £F C CHj-CH^-CH^ 150 M/Z "2 ^6rS 159 ( 2 H 5 , 164) 250 350 B) '0L1 >-t— C O z LU 50 159 OH I CHj-CH^-CH, 0 C H j - C H ^ C H ^ „ H —C—0-l-CH=-C cF 159-2 v 6 ' 5 150 M/Z 250 350 Figure 34. NICI mass spectrum o f A) 5-0H VPA-PFB ( 2 H n + 2 Hc) (peak 16, F i g . 23) and B) syn the t ic 5-0H VPA-PFB. - 107 -A ) 100 _ 342 CO U J F 5C 6^H 2-0-C-CH 0 ^CH-C-0 CH3-CH2-CH2 339 ( 2H 3,342) C H 2 - C 6 F 5 LLJ > U J cc 144 334 100 J u 200 M/Z 300 400 B) 100 _, >-(— CO U J U J c r U J cc 100 339 F 5C 6-CH 2-0-C-CH^ jj ,CH-C-0 CH3-CH2-CH2 339 C H2 - V 5 ^ I I 200 300 M/Z 400 Figure 35. NICI mass spectrum o f A) 2-PSA-diPFB ( 2 H n + 2Ho) (peak 17, F i g . 23) and B) syn the t ic 2 -PSA-d iPFB. - 108 -A ) 100 »_, 150 250 356 35 F 5 C 6 - C H 2 - ° - ^ H 2 - C H 2 N 8 , / C H - d - 0 - [ c H R C 6 F 5 C H J - C H J - C H 2 350 M/Z 353 ,2 ( H 3 . 356) 450 - i — 1 B) 100 I 150 250 5 6 2 2 2 \ » C H - C — 0 + C H = - C , F C S CHj—CH^—CH^ 353-"2 V 5 350 M/Z 450 i — 1 Figure 36. NICI mass spectrum of A) 2-PGA-diPFB ( 2H_ + 2 H , ) (peak 18, F i g . 23) and B) syn the t ic 2-PGA-diPFB. - 109 -A) m/z 141 6.5 7.0 I— 7.5 — I — 8.0 8.5 TIME (min) B) too _ C O z. L U 50 141 | C H 3 - C H 2 - C H 2 |CH 2 = C H - C H 2 141-CH-C-O+CH - C , F C 2 6 5 150 M/Z 250 —1 350 Figure 37. A) mass chromatograms at m/z 141 from F i g . 23 (peak a = 4-ene VPA), B) NICI mass spectrum of syn the t i c 4-ene VPA-PFB. - 110 YPA. The negative ion spectrum of th is p a r t i c u l a r metaboli te is more complex than that of s i m i l a r compounds der iva t i zed with PFB. The base peak i s m/z 113 (m/z 119, 2 % ) , and there are two other major fragments; m/z 157 (M-181) and 337 (M-l) (Figure 30 ) . The fac t that the major peak i s m/z 113 rather than m/z 157 can be explained by the presence of a 3-keto group that f a c i l i t a t e s a rearrangement (decarboxylat ion) of the fragment ion to give the m/z 113 an ion . The mechanism of the formation of t h i s anion is given in Scheme 1. Under the cond i t ions of d e r i v a t i z a t i o n employed, there was no evidence of the d e r i v a t i z a t i o n of keto and hydroxy! groups. Hence, keto and hydroxyl metabol i tes of VPA y i e l d e d only mono-der ivat ives . This i s i n agreement with the report of S t r i f e and Murphy (1984) who noted the non- r e a c t i v i t y of the hydroxy funct ion of hydroxyeicosanoids towards PFBB. For 6 -oxo-prostagl andins i t was poss ib le to obtain d i - d e r i v a t i v e s v ia the 6-keto funct ion (using pentaf luorobenzyl -hydroxyl amine) and the carboxyl group i . e . PFB oxime and PFB ester (Waddell and B l a i r , 1983). Introduct ion of two e lec t ron captur ing groups into a molecule should increase detect ion s e n s i t i v i t y . In th is c a s e , however, the s e n s i t i v i t y was less than the mono-PFB der iva t i ve because of a fragmentation pathway that gave r i s e to l e s s abundant high mass ions of the d i - d e r i v a t i v e . The d ica rboxy l i c acid VPA metabo l i tes , 2-PSA and 2-PGA, gave di-PFB d e r i v a t i v e s with the usual base peak i . e . [M-181]" (Figures 35 and 36) . In the TIC t race (Figure 23) , 2-PGA appears to be the most prominent metaboli te based on i t s peak height and the leve l of 2-PSA also looks h igh . Since 2-PGA i s not the major ur inary metaboli te of VPA the enhanced s e n s i t i v i t y observed is because o f the presence of two PFB moie t i es . Despite the increased s e n s i t i v i t y - I l l -0 n C h v c h v c o 3 2 \ || C H - C — 0 - C H o C c F c / c o b 0 C H J - C H J - C ^ H C H - C - 0 ^ 0 I c h y c h y c CH + C0 o / 2 (m/z 113) Scheme 1. Or ig in of the m/z 113 anion in the NICI mass spectrum of PFB d e r i v a t i z e d 3-keto VPA. - 112 observed for the two d i c a r b o x y l i c a c i d s , 2-PGA and 2-PSA, a th i rd d i c a r b o x y l i c acid metabol i te , 2-PMA, was not detected. 2-Propylmalonic ac id (2-PMA) has been character i zed as a VPA metabol i te by Acheampong et a l . (1983). 2. PFB as an e lec t ron capture NICI GCMS d e r i v a t i v e for VPA metabol i t e s The PFB d e r i v a t i v e s of VPA and i t s metabol i tes have good chromatographic proper t ies g iv ing sharp peaks in a reasonable GC run t ime. As with other carboxy l ic compounds der iva t i zed with PFB (mainly prostanoids) the NICI spectra of the PFB der iva t i zed VPA metabol i tes i s t y p i f i e d by the [M-181] - an ion. This ion r e s u l t s from cleavage of the PFB-oxygen bond. The exact mechanism of t h i s process i s not known (Waddell and B l a i r , 1983). GCMS SIM i s a very s p e c i f i c technique because ana lys is i s c a r r i e d out on the bas is of two parameters; gas chromatographic retent ion time and the mass of the monitored i o n ( s ) . In a d d i t i o n , s ince fragmentation with PFB d e r i v a t i v e s i s almost always d i rec ted away from the i n t a c t analyte molecule (base peak i s the i n t a c t molecule l ess 1) , there is enhanced s p e c i f i c i t y . Because of the inherent s e n s i t i v i t y of e lec t ron capture NICI and since there is l i t t l e fragmentation the system is very s e n s i t i v e and hence ideal for SIM. The PFB d e r i v a t i v e i s a lso super ior to per f luoroa lky l der iva t ives because with the l a t t e r , the major i ty of the ion current i s ca r r i ed by small fragments from the d e r i v a t i z i n g moiety. For example, Stan and Reich (1980) have reported a detect ion l i m i t of l f g by NICI fo r the - 113 heptaf luorobutyrate d e r i v a t i v e s of hydroxy fa t ty ac id methyl e s t e r s . But the ions monitored were those a r i s i n g from the d e r i v a t i z i n g molecule and are l e s s s p e c i f i c . With bis-TFMB, however, an ident ica l fragmentation pattern to that of the PFB was obtained (Figure 5B) . In contrast to PFB and bis-TFMB, pentafluorobenzoyl acyl d e r i v a t i v e s of compounds conta in ing free hydroxy groups and/or secondary amines produce molecular anions with v i r t u a l l y no fragmentation (Ramesah and P i c k e t t , 1986). This i s thought to be due to the increased s t a b i l i t y of the PFB acyl d e r i v a t i v e compared to PFB es te r d e r i v a t i v e s . Less that 2% (Figure 5A) of the PFB ion (m/z 181), which i s the base peak in the EI spectrum of PFB der iva t i zed VPA, i s observed in the NICI spectrum. The PFB anion is poss ib ly not as stable as the resonance s t a b i l i z e d carboxylate anion ( S t r i f e and Murphy, 1984). The s t a b i l i t y of the benzyl cat ion in E I , which rearranges to the more stable t ropyl ium ion is well known and th is expla ins the s t r i k i n g abundance of the m/z 181 in the EI spectrum of VPA-PFB. However, the number of TT e lec t rons in a cyc lohepta t r ieny l anion ( tropyl ium anion) would not f i t the (4n+2) n e lec t ron ru le and i t would be antiaromatic (Morrison and Boyd, 1973) and hence only a small amount of the m/z 181 anion was observed in the NICI spectra of VPA and i t s metabol i tes . 3. NICI (PFB) versus EI (t-BDMS) spectra of VPA metabol i tes VPA metabol i tes have been assayed in our laboratory (Abbott et a l . , 1986a) as t h e i r t-BDMS d e r i v a t i v e s by monitoring the c h a r a c t e r i s t i c [M-57] + fragment ion which c o n s t i t u t e s the base peak f o r VPA and many of the unsaturated metabol i tes . Nevertheless, - 114 -the [M-57] + ion i s not the most intense ion fo r the polar metabo l i t es , 3-OH VPA, 3-keto VPA and 4-keto VPA, most of the ion current being c a r r i e d by fragment ions inc lud ing m/z 73 and m/z 75 from the d e r i v a t i z i n g moiety. Th is is i l l u s t r a t e d for 3-OH VPA in Figure 38. In addi t ion to the increased s e n s i t i v i t y , NICI ana lys is of PFB d e r i v a t i z e d metabol i tes appears to be super ior to the EI (t-BDMS) a n a l y s i s in many ways. The [M-181]" ion is the base anion for a l l metabol i tes except for 3-keto VPA and th is enables more s p e c i f i c and s e n s i t i v e detect ion for a l l VPA metabo l i tes . The keto and hydroxy metabol i tes can give e i ther mono or d i - d e r i v a t i v e s of t-BDMS depending upon the d e r i v a t i z a t i o n c o n d i t i o n s . With the t-BDMS reagent in p y r i d i n e , 3-OH VPA does not de r iva t i ze r e a d i l y and chromatographs p o o r l y . With PFB, 4-OH VPA d e r i v a t i z e s and gives NICI spectrum whereas with the t-BDMS method a de r i va t i ve of 4-OH VPA is not seen and 4-OH VPA i s analyzed as the under ivat ized y - l a c t o n e . F i n a l l y , PFB d e r i v a t i v e formation i s f a c i l e and the time i s short compared to the 4 hours required for t-BDMS d e r i v a t i v e formation. 4. Selected ion chromatograms Se lected ion chromatograms obtained by monitoring the appropriate ions for a l l VPA metabol i tes extracted from a urine sample are i l l u s t r a t e d in Figure 3 9. The deuterated analogs are a lso presented s ince the sample contained [^Hsl-VPA and i t s metabol i tes . The ions monitored are summarized in Table 12. S i m i l a r SIM chromatograms for serum VPA and [ 2H5]-VPA metabol i tes are presented in Figure 40. - 115 -75 25 JJLL OH I CH 3 -CH 2 -CH CH 3 -CH 2 -CH 2 0 CHn II I 3 -CH-C-0-Si-CH_ ' C(CH ) 147 115 129 1 8 7 217 ( M -57) + 199 159 i 1 1 1 1 m 1 245 261 125 M / Z 225 325 Figure 38. EI (t-BDMS) mass spectrum of 3-0H VPA. - 116 -TABLE 12. Ions (m/z) monitored in NICI mode for VPA and [2He]-VPA metabolites derivatized with PFB. COMPOUND UNDEUTERATED DEUTERATED 3-keto VPA 113 119 2 ,4-diene VPA 139 144 ( E , E ) - 2 , 3 ' - d i e n e VPA 139 145 (Z)-2-ene VPA 141 147 (E)-2-ene VPA 141 147 4-ene VPA 141 146 3-ene VPA 141 147 VPA 143 149 4 ' -ke to -2 -ene VPA 155 158* 4-keto VPA 157 160* 3-OH VPA 159 165 4-OH VPA 159 165 5-OH VPA 159 164 2-PSA 339 342** 2-PGA 353 356** * When a l k a l i i s used to hydrolyze conjugates, otherwise 161 and 163 * d i - d e r i v a t i v e s - 117 -M/Z 149 H/Z 139 10 11 12 13 TIME (min) 14 20 21 22 Figure 39. SIM chromatograms of the PFB d e r i v a t i v e s of VPA and [ H g ] -VPA metabol i tes in a ur ine e x t r a c t . Peaks: 1= 2 ,4 -d iene VPA, 2= ( E , E ) - 2 , 3 ' - d i e n e VPA, 3= 4-ene VPA, 4= 3-ene VPA, 5= (Z) -2-ene VPA, 6= (E)-2-ene VPA, 7= VPA, 8= [ 2 H 5 ] - 2 , 4 - d i e n e VPA, 9= ( E , E ) - [ 2 H 6 ] - 2 , 3 ' - d i e n e VPA, 10= [ 2 H g ] - 3 - e n e VPA, 11=(Z)-[ 2 H g ] - 2 - e n e VPA, 12= ( E ) - [ 2 H g ] - 2 - e n e VPA, 13= [ 2 H g ] - V P A , 14= 4 ' - ke to -2 -ene VPA, 15 and 18= u n i d e n t i f i e d peaks i n t e r -f e r i n g with 3-keto VPA peaks, 16= 4-keto VPA, 17= [ 2 H 3 ] -4 ' - ke to -2 -ene VPA, 19= 3-OH VPA, 20= 3-OH VPA, 21= 4-0H VPA, 22= 5-OH VPA, 23= [ 2 H 3 ] - 4 - k e t o VPA, 24= [ 2 H 5 ] - 5 - 0 H VPA, 25= [ 2 H g ] -3 -0H VPA, 26= [ 2 H 6 ] - 3 - 0 H VPA, 27= [ 2 H g ] - 4 -0H VPA, 28= 2-PSA, 29= [ 2 H 3 ] - 2 - P S A , 30= 2-PGA, 31= [ 2 H 3 ] - 2 - P G A - 118 -F igure 40. SIM chromatograms of the PFB d e r i v a t i v e s of serum VPA and [ H g ] -VPA metabo l i tes . Peaks VPA, 3= 2 ,4 -d iene VPA, 4 6 1= 3-keto VPA, 2--( E , E ) - 2 , 3 ' - d i e n e VPA, 5= 4-ene VPA, r H g ] - 3 - k e t o 3-ene VPA, 7= (Z)-2-ene VPA, 8= (E)-2-ene VPA, 9= VPA, 10= [ 2 H 5 ] - 2 , 4 - d i e n e VPA, 11= ( E , E ) - [ 2 H g ] - 2 , 3 ' - d i e n e VPA, 12= [ 2 Hg]-3-ene VPA, 13= (Z ) - [ 2 Hg] -2 -ene VPA, 14= ( E ) - [ 2 H g ] -2-ene VPA, 15= [ 2 H C ] - V P A , 16= 4 ' - ke to -2 -ene VPA, 17= 4-keto VPA, 18= r H 3 ] - 4 ' - k e t o - 2 - e n e VPA, 19= 3-OH VPA, 20= 3-OH VPA 21= 4-0H VPA, 22= 5-OH VPA, 23= [ 2 H 3 ] - 4 - k e t o VPA, 24= [ 2 H 5 ] -5-OH VPA, 25= [ 2 Hg]-3-0H VPA, 26= [ 2 H g ] - 3-OH VPA, 27= [ 2 H g ] -4-OH VPA, 28= 2-PSA, 29= [ 2 H 3 ] - 2 - P S A , 30= 2-PGA, 31= [ 2 H 3 ] -2-PGA. - 119 -Ur ine and serum cont ro ls showed no i n t e r f e r i n g peaks, however, the o r i g i n of the m/z 158 ion in the NICI mass spectrum of PFB der iva t i zed 3-keto VPA (Figure 30A) i s not c lea r since th is ion i s absent in the spectrum of synthet ic 3-keto VPA (Figure 30B). In summary GCMS-NICI using PFB d e r i v a t i v e s appears to be super ior , in terms of d e r i v a t i v e formation and s e n s i t i v i t y to other cur rent ly a v a i l a b l e methods for the ana lys is of VPA metabo l i tes . A l l metaboli tes gave mono-der ivat ives (d ica rboxy l i c acid metabol i tes formed d i - d e r i v a t i v e s ) and were analyzed simultaneously in one chromatographic run . The d e r i v a t i v e formation and chromatographic c h a r a c t e r i s t i c s have been def ined and SIM chromatograms obta ined. The ana lys is has not yet been used to quant i tate metabo l i tes , but i t i s only a matter of s e l e c t i n g appropriate internal standards and applying the method. 5. VPA Metabol i tes in S a l i v a Because of the high s e n s i t i v i t y of the NICI method with PFB d e r i v a t i v e s i t was poss ib le to detect and i d e n t i f y VPA metabol i tes in s a l i v a . The metabol i tes detected were 4-ene VPA, 3-ene VPA, (Z)-2-ene VPA, (E) -2-ene VPA, ( E , E ) - 2 , 3 ' - d i e n e VPA, 3-keto VPA and 4-keto VPA. The metabol i tes were f i r s t detected by SIM, but i t was also possib le to detect them under l i n e a r scan c o n d i t i o n s . The l a t t e r was performed on a s a l i v a sample that a lso contained metabol i tes of [ 2 H6]-VPA. The metabol i tes were p o s i t i v e l y i d e n t i f i e d with the help of t h e i r twin ions and by i n j e c t i n g standards fo r re tent ion time comparison. The SIM chromatograms of fourteen s a l i v a r y metabol i tes of VPA and [ 2H63-VPA are shown together in Figure 41. I n t e r e s t i n g l y , none of the more polar - 120 -Sjn/z 160 m/z 157 —I— 10 11 TIME (min) F igure 41. SIM chromatograms of the PFB d e r i v a t i v e s of VPA and [ H g ] -VPA metabol i tes in a s a l i v a e x t r a c t . Peaks: 1= 3-keto VPA, 2= [ 2 H g ] -3-keto VPA, 3= ( E , E ) - 2 , 3 ' - d i e n e VPA, 4= 4-ene VPA, 5= 3-ene VPA, 6= (Z)-2-ene VPA, 7= (E)-2-ene VPA, 8= VPA, 9= ( E , E ) - [ 2 H g ] -2 , 3 ' - d i e n e VPA, 10= [ 2 H 5 ] - 4 - e n e VPA, 11= [ 2 H g ] - 3 - e n e VPA, 12= ( Z ) - [ 2 H g ] - 2 - e n e VPA, 13= ( E ) - [ 2 H g ] - 2 - e n e VPA, 14= [ 2 H g ] - V P A , 15= 4-keto VPA, 16= [ 2 H 3 ] - 4 - k e t o VPA. - -121 hydroxy metabol i tes were detected in s a l i v a . The l e v e l s of a l l metabol i tes appear to be higher than that of (E)-2-ene VPA which i s the major VPA metabol i te in serum. This i s not s u r p r i s i n g since th is metabol i te i s bound to plasma prote ins in excess of 98% (Nau et a l . , 1 9 8 4 ) . What i s i n t r i g u i n g , however, i s the fac t that there is a higher leve l of (Z)-2-ene in s a l i v a than (E) -2 -ene . Under the experimental condi t ions employed for the ex t rac t ion of VPA metabol i tes inc lud ing the isomers of 2-ene VPA from serum, i t does not appear that there is a conversion of (E)-2-ene VPA to (Z)-2-ene VPA, and i t i s u n l i k e l y that there wi l l be an in terconvers ion during the ex t rac t ion of these compounds from s a l i v a . In a recent study of VPA metaboli te l e v e l s in the serum of ped ia t r i c pat ients by Abbott et al .(1986a) i t was found that the concentrat ion of (E)-2-ene was about 30 times that o f (Z) -2 -ene. In the present study the ra t io of Z to E isomers of 2-ene VPA was much greater in s a l i v a than in serum (3.82 v s . 0.458), suggesting d i f fe rences in the t ransport or plasma protein binding proper t ies of these two isomers. Albumin, the most abundant protein in plasma i s the most important drug binding protein (Sjoholm, 1984). Albumin has a broad binding s p e c i f i c i t y with compounds of widely d i f f e r e n t s t ructures inc lud ing f a t t y a c i d s , b i l i r u b i n and many drugs binding with high a f f i n i t y . For f a t t y a c i d s , binding appears to be dependent upon the s t ructure of the f a t t y ac id (Spector and F l e t c h e r , 1978). The strength of binding increases as the chain length of the fa t ty acid inc reases . Also for a given chain length the presence of a s ing le c i s double bond increases the strength of the binding (oleate > s t e a r a t e ) . This report i s not - 122 cons is ten t with our observat ion for the isomers of 2-ene VPA, but Spector and F l e t c h e r (1978) did not compare the binding of c i s oleate to that of i t s trans isomer. Drug metabol i tes are genera l ly more polar and less protein bound than the i r parent drugs (Drayer, 1984). This i s not , however, true with t rans-2-ene VPA which i s more protein bound and has a s im i la r e l imina t ion h a l f - l i f e to that of the parent drug (Loscher et a l . , 1984). Because of i t s high protein binding t rans-2-ene VPA i s l i k e l y to have a d i s p l a c i n g e f f e c t on VPA binding and may r e s u l t in drug metaboli te i n t e r a c t i o n . The binding of a c i d i c drugs to albumin or to other human plasma prote ins i s to some extent s t e r e o s e l e c t i v e with respect to enantiomers (Drayer, 1984; van Ginneken et a l . , 1983). To our knowledge, there is no report in the l i t e r a t u r e with respect to s t e r e o s e l e c t i v e protein binding of geometric isomers. There are some examples of s t e r e o s e l e c t i v e metabolism of trans and c i s isomers (Vermeulen and Breimer, 1983). In ra t ad ipocytes , c is -unsa tura ted fa t ty ac ids st imulated l ipogenes is whereas saturated or t rans-unsaturated fa t ty ac ids were i n e f f e c t i v e (Shechter and Hem's, 1984). From a s t ructura l point of view, c i s isomers of fa t ty acids have a bend at the double bond in contrast to trans isomers and saturated fa t ty acids which extend in a l i n e a r conformation (Morrison and Boyd, 1973). For th is reason, c i s isomers f i t with each other or other molecules very poor ly . Th is may expla in why c i s - 2 - e n e VPA can be l e s s protein bound than t rans-2 -ene VPA. The higher s a l i v a leve l of the c i s isomer of 2-ene VPA might as well be due to a s t e r e o s e l e c t i v e t ransport to s a l i v a . There may be a f a c i l i t a t e d transport of the c i s isomer into s a l i v a or converse ly , the - 123 trans isomer may be s t e r e o s e l e c t i v e l y t ransported out of s a l i v a . In order to del ineate the mechanism of the apparent higher s a l i v a leve l of the c i s isomer of 2-ene VPA compared to the trans isomer, more experiments are r e q u i r e d , inc lud ing in v i t r o and in v ivo protein b inding studies of the pure isomers. Should the serum prote in binding proper t ies prove to be markedly d i f f e r e n t , i t i s l i k e l y that the c i s isomer w i l l a lso have d i f f e r e n t pharmacokinetic and pharmacodynamic proper t ies from that of the trans isomer. 6. Detect ion of new VPA metabol i tes Four d e r i v a t i v e s , namely, TMS, methyl e s t e r , t-BDMS and PFB (NICI) were employed for the detect ion of new VPA metabo l i tes . The l a t t e r two d e r i v a t i v e s proved to be more useful because of t h e i r super ior s e n s i t i v i t y and typ ica l d iagnost ic fragment i o n s . New metabol i tes inc lud ing one which appears to be 2 - ( 2 ' - p r o p e n y l ) - g l u t a r i c acid and an unsaturated keto metaboli te were apparent. In Figure 42 are shown the EI and NICI spectra of a VPA re la ted compound in urine (absent in contro l urine) that appears to be 2 - ( 2 ' - p r o p e n y l ) - g l u t a r i c a c i d . The s t ruc tura l assignment fo r th is apparent new metabol i te i s made d i f f i c u l t by the fac t that the i n t e n s i t y of the deuterated ions in both the EI and NICI spectra are much l e s s than those of other metabol i tes . The 2 - (2 ' -propenyl ) - g l u t a r i c ac id has been charac ter i zed as a metabol i te of 4-ene VPA in the Rhesus monkey (Rettenmeier et a l . , 1986a). The mass spectra of the unsaturated keto metaboli te contains c h a r a c t e r i s t i c twin fragment ions that help reveal i t s i d e n t i t y with - 124 -A> 100 351 >-r— C O r— cx _ l L U F 5 C 6 - C H 2 - ° - C - C H 2 - C V 0 CH 2=CH-CH 2 351 ,2 ^CH-C-O- C H 2 - C 6 F 5 ( H2,353) 331 353 100 200 M/Z 300 400 B) 100 _ 343 (M-57)+ tBDMS-0-C-CH2-CH2, CH 2=CH-CH 2 /^ CH-C-O-tBDMS 73 LU 0£ 147 2,36 115 133 50 150 T 250 M/Z B45 385 T 350 I— 450 Figure 42. NICI (A) and EI (B) mass spectra of VPA re la ted mater ia l i ur ine that appears to be 2 - ( 2 ' - p r o p e n y l ) - g l u t a r i c a c i d . - 125 some c e r t a i n t y . Diagnost ic fragment ions and ion doublets in the mass spectra of the d e r i v a t i v e s of the new metaboli te were compared to those o f known metabol i tes to aid in i d e n t i f i c a t i o n . The EI and NICI mass spectra of these d e r i v a t i v e s are presented in Figure 29. In the EI mass spectrum m/z 213 ( 2 H3, 216) corresponds to [M-57 ] + , m/z 111 ( 2 H3,114) i s [M-159] + and m/z 270 i s the molecular i o n . The fragmentation pathway proposed for the t-BDMS d e r i v a t i v e of the new metabol i te i s shown in Scheme 2. In the NICI mass spectrum, m/z 155 ( 2 H3, 158) i s the [M-181]" anion. From these spectra i t was apparent that the new metaboli te was an unsaturated keto compound (the [M-57] + ion of the t-BDMS d e r i v a t i v e s of 3-keto and 4-keto VPA i s 215 and the [M-181] - ion of the PFB d e r i v a t i v e of 4-keto VPA i s 157). S ince the twin ions were separated by only three daltons instead of s i x , three of the deuterium atoms must have been l o s t from the metabo l i te . With compounds l i k e m e t h a d o n e ^ (Hsia et a l . , 1976 ) having a keto group a to C 2 H2, the deuterium atoms can r e a d i l y exchange for hydrogen in a l k a l i n e s o l u t i o n . In the work-up procedure, the ur ine sample from which the spectra in Figure 29 were obtained was treated with a l k a l i in order to hydrolyze conjugates. F igures 43 and 44 are the EI and NICI mass spectra r e s p e c t i v e l y , of the d e r i v a t i v e s of the new metaboli te and 4-keto VPA obtained without a l k a l i n e treatment of the urine sample. In both the EI and NICI spect ra of the new metaboli te and 4-keto VPA the twin ions are now separated by s ix mass units ( ( 1 1 1 , 117), (213, 219), (155, 161), (215, 221), (157, 163)). This ind ica tes that the keto group i s at p o s i t i o n 4 and hence the new metaboli te must be a 4-keto - 126 -+ C —CH-II 3 0 m/z 43 0 II C H j - C — C H 2 CH 3 -CH 2 -CH 0 II CHj- C —CH 2 C H ^ C H ^ C H 0 II , C - C - 0 - S i C ( C H - ) - ( C H _ ) 3 ' 3 ^ " 3 y 2 m/z 270 f > C - C - 0 - S i C ( C H 3 ) 3 ( C H 3 ) 2 • 0 - S i C ( C H 3 ) 3 ( C H 3 ) , 0 II C H = - C - C H 0 0 3 2 \ in C - C // C H ^ C H ^ C H - C 4 H g 0 II C H r C - C H 9 0 C - C - 0 - S 1 ( C H 3 ) 2 C H ^ C H ^ C H m/z 213 m/z 139 (not observed) -CO 0 C H 3 - C - C H C H ^ C H ^ C H m/z 111 (base peak) Scheme 2. Proposed fragmentation pathway f o r the t-BDMS d e r i v a t i v e of a new VPA metabol i te assigned the s t ruc ture 4 ' -ke to -2 -ene VPA. - 127 -A) 100 _ >-h -CO > t— cr _ i U J cn 43 25 75 117 111 98 0 fl CHs-C—CH, CHj-CH^CH CH, , C+-C—0—Si — CH, C(CH 3 ) 3 111-( 2 H 6 > 117) 149 219(M-57 + 2H 6) + 129 191 177 213 125 M/Z 225 270 (M+) 276 325 B) 100 _, >— I— C O L U L U cn 43 75 ,1 I I 25 0 I) C H j - C - C H 2 Q c \ II I 3 x C H - C - 0 - S i - C H 3 C H r C H r C H 2 C ( C H 3 ) 3 221 (M-57+2H6) 193 101 129 148 215 281 125 M/Z 225 325 Figure 43. EI mass spectra o f A) 4 ' - ke to -2 -ene VPA and B) 4-keto VPA ext racted from urine without a l k a l i n e treatment. - 128 -A) 100 _ >— C O > r— cr _ i L U CC *< 50 161 155 0 II C H 3 - C - C H 2 , C H 3 - C H 2 - C H l o t 155 ,2 ( H6,161) 150 M/Z 250 C H 2 - C 6 F 5 350 B ) 100 _ >-t— C O cc 50 163 157 0 II C H 3 - C - C H 2 . C H 3 - C H 2 - C H 2 0 II C H - C - O f C H _ - C , F c i. o _> 157-( H 6 , 163) 150 M/Z 250 350 Figure 44. NICI mass spectra o f A) 4 ' -ke to -2 -ene VPA and B) 4-keto VPA extracted from ur ine without a l k a l i n e treatment. - 129 -compound. The double bond cannot be at pos i t ion 4 ' because there is no evidence for the l o s s of a deuterium atom. There fore , the new metabol i te may be one of the fo l lowing: 4 ' -ke to -2 -ene VPA, 4-keto -2 -ene VPA or 4 ' - k e t o - 3 - e n e VPA. The l a t t e r p o s s i b i l i t y i s un l i ke ly s ince the new metaboli te would most l i k e l y be a d e r i v a t i v e of the major unsaturated metabol i tes in serum ( i . e . 2-ene VPA or 2 , 3 ' - d i e n e VPA). From the peak height in F igure 23 (peak 10) the new metabolite is a prominent peak and has been detected in a l l urine samples analyzed. I ts leve l in ur ine appears to be higher than 2 ,4 -d iene VPA. It was detected both before and af ter induct ion by CBZ. The new metaboli te i s re la ted to 3 ' - ke to -4 -ene VPA which was detected as a VPA metabol i te in the Rhesus monkey (Rettenmeier et a l . , 1986b), in that i t i s an unsaturated keto metaboli te of VPA. The new metaboli te was detected both by EI and NICI GCMS. The complimentary nature of the two i o n i z a t i o n methods in terms of the i d e n t i f i c a t i o n of VPA metabol i tes cannot be overemphasized. NICI with PFB in most cases gave the d iagnost ic and abundant [M-181]" ion which corresponds to the i n t a c t molecular anion l e s s one. Once the NICI s t r u c t u r a l information was obta ined, the t-BDMS d e r i v a t i v e under EI gave more fragment ions and the c h a r a c t e r i s t i c i o n , [M-57 ] + , which aided in the postu la t ion of a st ructure for the unknown compound. Thus, the two GCMS systems were h ighly complimentary to each other in i d e n t i f y i n g known VPA metabol i tes and the new VPA metabo l i te . The new VPA metabol i te appears to be 4 ' - ke to -2 -ene VPA although t h i s s t ructura l assignment should be considered tenta t ive (see sect ion D) . For the sake of convenience the new metaboli te has been referred - 130 to as 4 ' -ke to -2 -ene VPA in the tex t . The 4 ' - ke to -2 -ene VPA could be detected in ur ine under l i n e a r scan condi t ions but in serum by SIM o n l y . One poss ib le o r i g i n of 4 ' - ke to -2 -ene VPA i s ) -hydroxylat ion o f the saturated side chain of 2-ene VPA followed by dehydrogenation of i t s precursor 4 ' -0H-2-ene VPA, by the enzyme alcohol dehydrogenase in a manner analogous to that of the conversion of (co-l)-hydroxy fa t ty acids to (co-D-keto fa t ty acids (Bjorkhem, 1972). The poss ib le precursor i . e . 4 ' -0H-2-ene VPA, however was not detected in serum or u r ine . This p o s s i b l e metabolic o r i g i n of 4 ' - ke to -2 -ene VPA i s s i m i l a r to the report by Rettenmeier et a l . (1986a) where 4 ' -0H-4-ene VPA was apparently found to be a minor metaboli te of 4-ene VPA in the Rhesus monkey. The 4 ' - ke to -2 -ene VPA could also ar ise from the hydration of the 3 ' -doub le bond of 2 , 3 ' - d i e n e VPA followed by o x i d a t i o n . Addi t ion of water across a double bond was evident in one study of the metabolism of unsaturated VPA metabol i tes (Granneman et a l . , 1984a). In rats given 2 , 3 ' - d i e n e VPA, however, no evidence was seen for the formation of 4 ' - ke to -2 -ene VPA (Ron Lee , M.Sc. t h e s i s ) . Regardless, 4 ' - ke to -2 -ene VPA must be derived from e i ther a mono or d i -unsaturated VPA metabol i te and t h i s confirms that unsaturated VPA metabol i tes are l i k e l y to give r i s e to p o t e n t i a l l y tox ic ox idat ion products . The detect ion of 4 ' - ke to -2 -ene VPA a lso demonstrates the complex nature of VPA metabolism which involves a va r ie ty of enzymes and mul t ip le minor pathways such as dehydrogenation, hydra t ion , reduct ion and hydroxy!a t ion . - 131 D. Synthesis From the spectral data discussed above and metabolic c o n s i d e r a t i o n s , i t was f e l t that the new VPA metaboli te was 4 ' -ke to -2 -ene VPA and the synthesis of t h i s compound was attempted employing l i t e r a t u r e methods. 1. Attempted synthesis of 2 - (2 ' -oxopropy l ) -2 -penteno ic acid (4 ' -ke to -2 -ene VPA) via ethyl 2 -propyl -4-oxopentanoate . 2-Propyl -4-oxopentanoic a c i d , I, was synthesized from ethyl 2-bromopentanoate and e thy lacetoaceta te . 2-Bromopentanoic acid was prepared by bromination of pentanoic acid according to standard procedures. Ethyl 2-bromopentanoate and ethyl acetoacetate were then condensed in the presence of NaH (Scheme 3 ) . This synthet ic route was adopted from that of Acheampong et a l . (1983) which i s based on the method of Lawessen et a l . (1962) f o r the preparat ion of Y-keto a c i d s . The f ina l product was obtained by decarboxylat ion and hydro lys is of the acy lsucc ina te intermediate. D i s t i l l a t i o n of the crude product y ie lded 2-propyl -4-oxopentanoic acid contaminated with small amounts of i t s ethyl es te r . R e d i s t i l l a t i o n y i e l d e d pure 2-propyl -4-oxopentanoic a c i d . The acid was converted to i t s ethyl ester i n i t i a l l y by using ethyl iod ide in the presence of potassium carbonate and 18-crown-6 (Fedorynoki et a l . , 1978). However, because the keto funct ion was not a f fec ted by a c i d , ethyl alcohol in the presence of s u l f u r i c acid was used for the e s t e r i f i c a t i o n of most of the synthesized acid since th is procedure resul ted in a higher y i e l d . The iden t i t y of ethyl 2 -p ropy l -4 -oxopentanoate was es tab l ished by GCMS, IR and NMR (see appendix). - 132 -C H J - C H ^ - C H / J - C H ^ - C — O H B r 0 PBr, 0 II CH rCHr-CHr-CH-C-OH o c c i Br EtOH H + 0 II C H - t - C H ^ C H a - C H - C - O C - H , -3 2 2 I do Br 0 0 0 II M II C H r - C - C H ^ C - O C 0 H C + C H rCH^-CH^-CH-C-OC 0H c NaH THF A 0 C 0 0 C „ H c II I 2 5 CH=-C-CH^ 0 3 \ II C H - C - 0 C o H c / 2 5 CH^ — CH^-""CH2 0 II CHr-C-CH, Cone. HC1 A \ CH CH2~ CH2 0 II C H - C - O H (I) Br Scheme 3. Synthet ic route for 2 -propyl -4-oxopentanoic a c i d . - 133 -Int roduct ion of the double bond at the 2 or 2 1 pos i t ion was attempted by the ox idat ion of the t r i m e t h y l s i l y l enol ether using the hydride abst ract ing reagent DDQ according to Jung and Pan (1977). The synthet ic strategy was to use two equivalents of LDA and chlorot r imethyl si lane since both the keto and carboxyl carbonyls wi l l form the TMS enols (Scheme 4 ) . The presence of TMS contain ing compounds, a mono- and di-TMS d e r i v a t i v e s , was observed by GCMS before the react ion with DDQ. The mono -TMS enol was p r e f e r e n t i a l l y formed as ind ica ted by a 3:1 r a t i o of the mono-TMS enol to the di-TMS e n o l . Fol lowing the react ion with DDQ, GCMS ana lys is of the crude product ind ica ted the presence of three components in about equal q u a n t i t i e s . These were the s t a r t i n g m a t e r i a l , the TMS enol ether of ethyl 2 -propyl -4 -oxopentanoate , and the TMS enol ether of ethyl 2 -p ropy l -4-oxo-2-pentenoic acid (or the TMS enol ether of ethyl 2 - (2 ' -oxopropyl ) -2 -pentenoic a c i d ) . The y i e l d s of ox idat ion products of enol ethers are genera l ly moderate and oxidat ion is almost a l l of the time incomplete r e s u l t i n g in considerable s ta r t ing material remaining at the end of the react ion (Jung and Pan, 1977). When the enolate product of 4-keto VPA i s trapped as a s i l y l enol ether s t ructures A and B are p o s s i b l e . 0 - TMS 0 - TMS CH 2=C-CH 2R CH3-C=CHR A B The composit ion of the regio isomer ic enolate mixture of unsymmetrical ketones i s governed by k i n e t i c or thermodynamic factors (Carey and Sundberg, 1983). By appropriate se lec t ion of experimental condit ions 134 -0 It CH=-C-CH, CH^—CH — CH 2 0 II C H - C - O H EtI K 2 C 0 3 0 II 18-crown-6 THF CH=-C—CH, / CH^- C H ^ CH 2 8 'CH—C-OC„H 2"5 LDA THF •78° TMS CI 2 n 5 O-TMS l CH - C - C H „ 0—TMS ^ C = C - 0 C „ H CH^-- C H ^ CH 2 DDQ benzene 0—TMS I CHs=C—CH v CH^-" CH2 - CH 2 C —C —0C o H 2"5 I OH H + 0 II C H r C - C H ^ 3 ^ CH — C H ^ CH 2 0 II C - C - O H ( I D Scheme 4. Attempted synthesis o f 2 - (2 1 -oxopropy l ) - 2 -pen teno ic a c i d . - 135 under which an enol ate i s formed from a ketone i t i s poss ib le to e s t a b l i s h e i t h e r k i n e t i c or thermodynamic c o n t r o l . If the base is strong and s t e r i c a l l y bulky and i f aprot ic solvents are used, the major product formed w i l l be the product of k i n e t i c c o n t r o l . In the present reac t ion LDA and THF were used and hence, the dominant enol ether wi l l be A which i s the product of k ine t i c c o n t r o l . Under these condi t ions t h e r e f o r e , the f ina l product w i l l have the double bond at e i ther p o s i t i o n s 2 or 2 ' . The formation of the double bond on the same chain as the keto funct ion wi l l be favored because of con jugat ion . A por t ion of the product mixture was t reated with a l k a l i to e f f e c t h y d r o l y s i s of both the TMS enol ether and the e s t e r . Fol lowing a c i d i f i c a t i o n and ex t rac t ion the ext ract was der iva t i zed to give the t-BDMS e s t e r . The mass chromatograms at m/z 213 [M-57] + are shown in F igure 45 (two peaks) and the corresponding mass spectra in Figure 46. The mass spectra as well as the re tent ion times for the two products were d i f f e r e n t from that of the new metaboli te extracted from ur ine . Since the mass spectra were d i f f e r e n t from that of 4 ' - ke to -2 -ene VPA ( I I I ) , these compounds are poss ib ly the two geometric isomers of 2 -propy l -4 -oxo-2 -pentenoic acid ( I I ) . 2. Synthesis of 4 ' - ke to -2 -ene VPA s t a r t i n g with a protected 4-oxopentanoic acid The 4-oxopentanoic acid was converted to i t s ethyl ester and the keto funct ion protected by means of 1 , 3 - d i t h i o l ane (Scheme 5 ) according to Hatch et a l . (1978). The ethyl 4-ethylenethioketalpentanoate was - 136 -(b) 10.39 m/z 213 (a) 10.30 8 9 10 11 I 12 TIME (min) Figure 45. Mass chromatograms (m/z 213) of the t-BDMS d e r i v a t i v e s o f synthesized 2 -propy l -4 -oxo-2 -penteno ic ac id (4-keto-2-ene VPA). - 137 -A) 100 _ >-t— t—i CO z LU > CC _J LU CC 4 3 25 75 95 185 31 l i L l L 155 171 213 (M-57) 0 II CH -C-CHJ 3 CH3-CH2-CH2 0 f 3 C-C-O-Si-CH, I 3 C(CH3)3 255(M-15) 125 M/Z 225 325 B) 100 _ CO z LU LU cr _ J LU cc 4 3 75 . u l 25 213 ( M " 5 7 ) " 0 » CH CH3-C-CH. 0 C H 3 C^-C-O-Si-CH C H 3 - C H 2 - C H 2 j j 115 171 155 185 1" 125 M/Z 270(Mt) 225 325 Figure 46 Mass spectra o f A) peak a and B) peak b in Figure 45. - 138 obtained in good y i e l d (85%) and i t s i d e n t i t y confirmed by GCMS and NMR ( fo r NMR spectrum see appendix). The synthet ic route employed was an a ldol condensation of ethyl 4 -e thy leneth ioketa l pentanoate with propionaldehyde. The enolate formed with LDA was condensed with propionaldehyde to form the B-hydroxy ester which was then dehydrated with methane sul fonyl c h l o r i d e fol lowed by potassium hydr ide . This synthet ic route was adopted from the method for preparing 8-hydroxy, B 1 , y ' -unsaturated esters (Kende and Toder, 1982; Acheampong and Abbot t , 1985). The 4-carboethoxy-2-ethylenethioketa l -5 -hydroxyheptane was obtained in a moderate y i e l d and was shown to be homogeneous by GCMS. The hydroxy compound was then mesylated and treated with potassium hydr ide. Af ter the dehydration step, GCMS ana lys is of the crude product showed f i v e peaks. Two of these peaks (one of them was the major component) in the TIC p l o t corresponded to the two geometric isomers of the desi red product with the protect ing group ( for EI spect ra see appendix). The three compounds could not be i d e n t i f i e d from mass spectral data . Since i t was not poss ib le to i s o l a t e the product , the product mixture as such was t reated with mercuric c h l o r i d e in the presence of cadmium carbonate in order to c leave the 1 , 3 - d i t h i o l a n e protect ing group (Pappas and Nace, 1959). The TIC p lo t of the r e s u l t i n g product mixture showed nine peaks, two of which (minor components) were the two isomers of the ethyl ester of 4 ' - ke to -2 -ene VPA. Their mass spectra are shown in Figure 47. Even though Hg(II) s a l t s have been used for the cleavage of the 1 , 3 - d i t h i o l a n e d e r i v a t i v e s of a number of compounds, in the present case the deprotect ion step was i n e f f i c i e n t and resulted in the appearance of addi t ional compounds. The protect ing group i t s e l f - 139 -0 0 II II C H ^ C - C H j - C H ^ C — OH EtOH + 0 II H 0 II C H 3 - C - C H 2 - C H 2 - C - O C 2 H 5 HSCH2CH2SH ( C 2 H 5 ) 2 0 . B F 3 CH5-CH„ I 2 | 2 S S 0 \ / II CH 3 — C — C H ^ C H ^ C — O C 2 H 5 C H ^ C H ^ C H CH^-CH, l 2 | 2 S S \ / C H 3 — C — C H 2 LDA THF -78 C CH5-CH5-CH 3 2 1 OH 0 x 11 C H - C — OC 0 H C / 2 5 MSC1 E t 3 N C H 2 C 1 2 Scheme 5. Synthet ic route for 2-(2 ' - o x o p r o p y l)-2 - p e n t e n o i c a c i d . - 140 -Scheme 5. (Continued) CH5 -CH, I 2 I 2 S S \ / C H - — C — C H 0 0 C H - C - O C H CHr-CH^-CH 3 2 1 2"5 0S0 2 C H 3 KH THF C H - C H „ I 2 I 2 S S \ / C H — C - C H 9 0 3 2 S n C-C — 0C„H C H ^ C H ^ C H ^ 2n5 HgCl 2 CdC0„ H 20 Q II CHj-C — C H 3 0 II C H r C - C H 0 0 3 2 N || C - C - 0 C „ H CH^—CH^—CH 2"5 I OH !H + 0 II CHr-C-CH 0 0 II C—C-OH CH^-CH^-CH (I I I) - 141 A) 100 _ 111 >-I— cn LU LU 43 2 9 5 5 C H „ - C - C H „ II ,C fC-0 -G \H C H 3 - C H 2 - C H 111-1 2 5 184 (Mt) LL B) 25 100 _ >-i— t—) CO z LU I— z LU > I— c r 125 M / Z 111 225 LU 43 2 9 5 5 0 II C H 3 - C - C H 2 C H 3 - C H 2 - C H 111-1 C-0 - C " 2 H 5 184 (Mt) I , \ 11 I 25 125 M / Z 225 Figure 47. EI mass spectra of the isomers o f the ethyl es te r o f 4 ' - ke to -2-ene VPA A) Isomer with the shor ter re tent ion t ime. B) Isomer with the longer re tent ion t ime. - 142 might have caused a compl icat ion in the synthesis since three unknown compounds were observed a f ter the dehydration step with potassium hydr ide . As a r e s u l t , the crude product mixture was analyzed by GCMS. A f t e r hydro lys is of the e s t e r , a port ion of the product mixture was d e r i v a t i z e d to give the t-BDMS and PFB d e r i v a t i v e s . The mass chromatograms at m/z 213 [M-57] + are shown in Figure 48 along with that from the urine ex t rac t . F igure 4 9 shows the mass spectra of peaks a and b from Figure 48. The mass spectra of peaks a and b are i d e n t i c a l and the i r re tent ion times are almost the same (8.12 minutes versus 8.15 minutes) . There fore , the chromatographic and mass spectral data of the new metaboli te match one synthet ic product , which i s considered to be 4 ' -ke to -2 -ene VPA on the basis of intermediates that l e d to i t s formation and the synthet ic route used. In the EI mass spectra of 4 ' -ke to -2 -ene VPA the m/z 111 ion appears to be a c h a r a c t e r i s t i c i o n . The m/z 111 i s the base peak in both the t-BDMS (nat ive and synthet ic) (Figure 49) and ethyl (Figure 47) es te rs of 4 ' - ke to -2 -ene VPA. There are q u a l i t a t i v e as well as quan t i t a t i ve d i f f e rences in the mass spectra of 4 ' - ke to -2 -ene VPA and 2 -propy l -4 -oxo-2 -penteno ic acid (4-keto-2-ene VPA). In the mass spectrum of the former (Figures 47 and 49) m/z 111 i s the base peak whereas in the l a t t e r (Figure 46) t h i s ion is absent. In a d d i t i o n , the m/z 213 ion i s the base peak in the mass spectrum of 2 -propy l -4 -oxo-2 -pentenoic a c i d , but has an i n t e n s i t y of l e s s than 50% in that of 4 ' - ke to -2 -ene VPA. In the m/z 155 mass chromatogram obtained by NICI o f synthesized 4 ' - ke to -2 -ene VPA there are four peaks whose mass spectra are the same - 143 -A) 10.46 m/z 213 (a) 8.12 I 8 76 8.88 m/z 213 9.92 10.22 TIME (min) F igure 48. A) Mass chromatograms at m/z 213 of t-BDMS d e r i v a t i z e d syn-t h e t i c 4 ' - ke to -2 -ene VPA. B) Mass chromatograms at m/z 213 of t-BDMS d e r i v a t i z e d nat ive 4 ' - ke to -2 -ene VPA. - 144 -A) 100 _ >-r— co LU-L U > cc 43 25 111 75 103 0 II CHr -C -CH. V C H J - C H ^ C H " ? H3 ; C - ( - C - 0 - S i - C H 3 C(CH 3 ) 3 111-213(M-57) + 1 2 7 1 4 9 ,171185 2 5 5 270 (Mt) I I. 125 M/Z 225 325 B) 100 _ CO L U I— z »—» L U CC 43 25 73 1 1 1 0 II C H r C - C H 0 CHj-CH^CH 111—I C f C - 0 - S i - C H 3 C(CH 3 ) 3 213(M-57) + 171 127 149 185 _L 270(M-255 125 M/Z 225 —1 325 Figure 49. Mass spectra o f A) peak a and B) peak b in Figure 48. - 145 -TIME (min) Figure 50. A) Mass chromatograms at m/z 155 o f PFB d e r i v a t i z e d synthe t ic 4 ' - ke to -2 -ene VPA. B) Mass chromatogram at m/z 155 o f PFB d e r i v a t i z e d 4 ' - k e t o - 2 -ene VPA from a ur ine e x t r a c t . - 146 with m/z 155 as the base peak. The re tent ion time of one of these peaks matches that from the urine ext ract (Figure 50 ) . The two other peaks could be the isomers of 4 ' - ke to -3 -ene VPA s ince i t i s poss ib le that small amounts of t h i s pos i t iona l isomer may be formed as a r e s u l t of the removal of a y proton from the mesylate d e r i v a t i v e during the dehydrat ion step. It i s d i f f i c u l t , however, to say fo r cer ta in which o f the peaks are those of 4 ' - ke to -2 -ene VPA or 4 ' - k e t o - 3 - e n e VPA since i t i s the smaller peak that has i d e n t i c a l re tent ion time to that of the metabol i te extracted from ur ine . It i s a lso poss ib le that the two l a r g e r peaks in Figure 50A could be due to side products in the synthet ic product mixture. Because the only ion of s i g n i f i c a n c e in the mass spectra of the peaks in Figure 50 i s 155, an inference about these compounds can not be made on the basis of the i r NICI s p e c t r a . The synthet ic strategy fo r 4 ' - ke to -2 -ene VPA worked well except at the deprotect ion step. The keto group was protected with 1 ,3 -d i th io lane group because of the a v a i l a b i l i t y of the reagents at that t ime. From t h i s exper ience, i t appears that a 1 ,3-dioxolane would be a better pro tec t ing group because of i t s ease of removal. In summary, the new VPA metabol i te is most l i k e l y 4 ' - ke to -2 -ene VPA. This s t ructura l assignment must be considered tenta t ive however, u n t i l a successful synthesis provides s u f f i c i e n t product to obtain NMR data . - 147 SUMMARY AND CONCLUSIONS The PFB d e r i v a t i v e s of VPA and i t s metaboli tes produced intense [M-181]" ions ( [M-181-C0 2 ] - for 3-keto VPA), which are s u i t a b l e for SIM ana lys is as well as i d e n t i f i c a t i o n of metabo l i tes . Three VPA d e r i v a t i v e s (PFB, bis-TFMB and t-BDMS) were compared with respect to detect ion s e n s i t i v i t y . In the EI mode, a l l three d e r i v a t i v e s had s i m i l a r s e n s i t i v i t i e s . In NICI, the PFB d e r i v a t i v e was found to be 30-50 times more s e n s i t i v e than the t-BDMS d e r i v a t i v e by E I . The PFB d e r i v a t i v e (NICI mode) a lso proved to be about f i v e times as s e n s i t i v e as the s i m i l a r f l u o r i n a t e d d e r i v a t i v e , bis-TFMB. Comparison of a s tab le i so tope - labe l led internal standard ([ 2 H5]-VPA) and an internal standard that gives a common ion (OA) showed that the [ 2H6]-VPA was a super ior in terna l standard for the NICI assay of VPA. A highly s e n s i t i v e and prec ise NICI assay was developed that can quant i ta te VPA in serum and s a l i v a accurate ly down to 2 ng/mL. Serum VPA concentrat ions obtained by NICI were in exce l len t agreement with those obtained by an EI (t-BDMS) assay used for rout ine VPA quant i ta t ion in th is labora tory . - 148 Paired s a l i v a and serum samples were assayed for VPA using the NICI assay developed, in f i v e healthy volunteers both before and a f t e r CBZ admin is t ra t ion . In sp i te of the considerable i n t r a - s u b j e c t v a r i a b i l i t y , the time-averaged s a l i v a to serum free r a t i o s were remarkably s i m i l a r in three volunteers both before and a f t e r CBZ admin is t ra t ion . The time-averaged s a l i v a to serum total r a t i o s decreased a f te r CBZ i n d i c a t i n g that the s a l i v a to serum to ta l r a t i o was concentrat ion dependent. An ac t ive t ransport of VPA out of s a l i v a i s invoked to explain the lower concentrat ion of VPA in s a l i v a (18.92% ± 6 . 2 5 of serum free) compared to serum free and CSF concent ra t ions . The good c o r r e l a t i o n s found between s a l i v a and both serum tota l and free VPA concentrat ions suggest that measuring VPA in s a l i v a by the NICI method would be su i tab le for drug in te rac t ion and pharmacokinetic s t u d i e s . The condi t ions for de r i va t i ve formation and the chromatographic behavior of the PFB der iva t i ves of VPA metabol i tes have been determined and SIM chromatograms obtained. This ana ly t i ca l method appears to be super ior in terms of ease of de r i va t i ve formation and s e n s i t i v i t y to cur ren t ly ava i l ab le GCMS methods for the a n a l y s i s of VPA metabo l i tes . Seven VPA metabol i tes were i d e n t i f i e d in s a l i v a . There appears to be a s t e r e o s e l e c t i v e plasma prote in binding or t ransport of the geometric isomers of 2-ene VPA as ind icated by t h e i r s a l i v a l e v e l s r e l a t i v e to those in serum. - 149 -10. A new VPA metabol i te which from mass spectral and chromatographic data appears to be 4 ' -ke to -2 -ene VPA was detected in human ur ine . The detect ion of t h i s metabol i te , apparently a r i s i n g from the ox idat ion of 2 ,3 ' -d iene VPA or 2-ene VPA, confirms that unsaturated VPA metabol i tes are l i k e l y to give r i s e to p o t e n t i a l l y tox ic ox idat ion products . Another new metabol i te that appears to be 2 - ( 2 ' - p r o p e n y l ) - g l u t a r i c acid was also detected in u r ine . 11. The synthesis of 4 ' -ke to -2 -ene VPA was attempted using two d i f f e r e n t synthet ic routes . The synthet ic route in which 4-oxopentonoate protected through d i t h i o ketal was used as the s t a r t i n g material y i e l d e d the protected 4 ' -ke to -2 -ene VPA. The deprotect ion s tep , however, was i n e f f i c i e n t and consequently, i t was not poss ib le to i s o l a t e s u f f i c i e n t product for obtaining NMR data . - 150 REFERENCES F . S . Abbott , R. Burton, J . Orr , D. Wladichuk, S. Ferguson and T . H . Sun. V a l p r o i c ac id ana lys is in s a l i v a and serum using selected ion monitoring (e lec t ron ion iza t ion ) of the t e r t . - B u t y l d i m e t h y l s i l y l d e r i v a t i v e s . J . Chromatogr. , 227, 433 (1982). F . S . Abbott , J . Kassam, A. Acheampong, S. Ferguson, S. Panesar, R. Burton, K. F a r r e l l and J . Orr . C a p i l l a r y gas chromatography-mass spectrometry of v a l p r o i c ac id metabol i tes in serum and urine using t e r t . - B u t y l d i m e t h y l s i l y l d e r i v a t i v e s . J . Chromatogr. , 375, 285 (1986a). F . S . Abbott , J . Kassam, J . M . Orr and K. F a r r e l l . The e f f e c t of a s p i r i n on v a l p r o i c ac id metabolism. C l i n . Pharmacol. T h e r . , 40, 94 (1986b). A . A . Acheampong. Quant i ta t ive s t ructure -ant iconvul sant a c t i v i t y studies of v a l p r o i c ac id analogues. Ph.D. T h e s i s , Un ivers i ty of B r i t i s h Columbia, Vancouver (1985), p.55. A . Acheampong and F . S . Abbott . Synthesis and stereochemical determination of diunsaturated va lp ro ic ac id analogs inc lud ing i t s major diunsaturated metabol i te . J . l i p i d R e s . , 26, 1002 (1985). A. Acheampong, F. Abbott and R. Burton. I d e n t i f i c a t i o n of v a l p r o i c ac id metabol i tes in human serum and urine using hexadeuterated va lp ro ic acid and gas chromatographic mass spectrometric a n a l y s i s . Biomed. Mass Spectrom., 10, 586 (1983). — A . A . Acheampong, F . S . Abbott , J . M . Orr , S.M. Ferguson and R.W. Burton. Use of hexadeutrated va lp ro ic ac id and gas-chromatography-mass spectrometry to determine the pharmacokinetics of va lp ro ic a c i d . J . Pharm. S c i . , 73, 489 (1984). R. A l r i c , M. C o c i g l i o , J . P . Blayac and R. Puech. Performance evaluat ion of a reversed phase, high performance l i q u i d chromatographic assay of v a l p r o i c ac id invo lv ing a "solvent demixing" ext ract ion procedure and precolumn d e r i v a t i z a t i o n . J . Chromatogr. , 224, 289 (1981). A .M. Baciewicz . Carbamazepine drug i n t e r a c t i o n s . Ther. Drug M o n i t . , 8, 305 (1986). J . Balkon. Rapid CI-MS determination of va lpro ic a c i d ; adaptat ion of a gas chromatographic assay. J . Ana l . T o x i c o l . , 2» 78 (1979). H. B a r t e l s , E. Gunther and S . Wall i s . Flow-dependent s a l i v a r y primidone l e v e l s in e p i l e p t i c c h i l d r e n . E p i l e p s i a , 20_, 431 (1979). D . J . Berry and L . A . C la rke . Determination of va lpro ic ac id (d ipropy lace t ic acid) in plasma by g a s - l i q u i d chromatography. J . Chromatogr. , 156, 301 (1978). M. B i a l e r , Z. Hussien, I. Raz, 0. Abramsky, Y. Herishanu and F. Pachys. Pharmacokinetics of v a l p r o i c ac id in volunteers a f te r a s ing le dose study. Biopharm. Drug D i s p o s . , 6 , 33 (1985). - 151 I. Bjorkhem. On the ro le of alcohol dehydrogenase in to-oxidation of fa t ty a c i d s . Eur . J . Biochem., 30, 441 (1972). G . F . Blom and P . J . M . Guelen. The d i s t r i b u t i o n of a n t i - e p i l e p t i c drugs between serum, s a l i v a and cerebrospinal f l u i d . In: A n t i e p i l e p t i c drug moni tor ing , C. Gardner-Thorpe, D. Janz , H. Meinard i , C . E . Pippenger, e d s . , Pitman Med ica l , Tunbridge Wel ls , Kent (1977), p. 287. T . A . Bowdle, R.H. Levy and R .E . C u t l e r . E f f e c t s of carbamazepine on v a l p r o i c ac id k i n e t i c s in normal sub jec ts . C l i n . Pharmacol. T h e r . , 26_, 629 (1979). M . J . Brod ie , G. For res t and W.G. Rapeport. Carbamazepine 10,11-epoxide concentrat ions in e p i l e p t i c s on carbamazepine alone and in combination with other an t iconvu lsants . Br . J . C l i n . Pharmacol . , 16_, 747 (1983). N.A. Brown, P .B . Farmer and M. Coakley. Va lp ro ic ac id t e r a t o g e n i c i t y : demonstration that the biochemical mechanism d i f f e r s from that of valproate h e p a t o t o x i c i t y . Biochemical Soc ie ty T ransac t ions , 13^, 75 (1985). J . Brum" and B . J . Wi lder . Va lpro ic a c i d . Review of a new a n t i e p i l e p t i c drug. A r c h . N e u r o l . , 36 , 393 (1979). B. C a l e n d r i l l o and G. Reynoso. A micromethod for the on-column methylation of v a l p r o i c ac id by g a s - l i q u i d chromatography. J . Ana l . T o x i c o l . , 4 , 272 (1980). F . A . Carey and R . J . Sundberg. Advanced Organic Chemistry, part B, 2nd e d . , Plenum Press , New York (1983), p. 5-10. S . C . Chan. Monitoring serum va lp ro ic ac id by gas chromatography with e lec t ron capture de tec t ion . C l i n . Chem., 26_, 1528 (1980). J . R . Chapman. P r a c t i c a l Organic Mass Spectrometry. John Wiley and Sons, New York (1985), p.79-100. M. C l a e y s , S . P . Markey and W. Maenhaut. Variance ana lys is of er ror in se lec ted ion monitoring assays using var ious internal standards. A p r a c t i c a l study case. Biomed. Mass Spectrom., 4 , 122 (1977). E. M. Cornford , C P . Diep and W.M. Pardr idge. B lood-bra in b a r r i e r t ransport of va lp ro ic a c i d . J . Neurochem., 44_, 1541 (1985). R.G. D ick inson , R.C. Harland, R.K. Lynn, W.B. Smith and N. Gerber. Transmission of va lp ro ic ac id (Depakene) across the p lacenta: h a l f - l i f e of the drug in mother and baby. J . P e d i a t r . , 94, 832 (1979). R .C . Dougherty. Negative chemical i o n i z a t i o n mass spectrometry. A n a l . Chem., 53 , 625A (1981). D . E . Drayer. Prote in binding of drugs, drug enantiomers, ac t i ve drug metabol i tes and therapeut ic drug moni tor ing. In: Drug-protein b ind ing , M.M. Reidenberg, S. E r i l l , e d s . , Prager P u b l i s h e r s , New York (1984), p.332. F. E. Dre i fuss and N. S a n t i l l i . Va lpro ic ac id hepatic f a t a l i t i e s : Ana lys is o f United States cases . Neurology, 36_ (Suppl . 1 ) , 175 (1986). - 152 E. Eder, D. Henschler and T. Neudecker. Mutagenic proper t ies of a l l y l i c and a , B-unsaturated compounds: cons idera t ion of a l k y l a t i n g mechanisms. Xenob io t ica , 12, 831 (1982). M. Fedorynoki , K. Wojciechowski, Z. Matacz and M. Makosza. Sodium and potassium carbonates: E f f i c i e n t strong bases in s o l i d - l i q u i d two-phase systems. J . Org. Chem., 43, 4682 (1978). D . J . Freeman and N. Rawal. Ext ract ion of under ivat ized v a l p r o i c ac id from serum before gas-chromatography. C l i n . Chem., 26_, 6 7 4 (1980). K .C . Fung and K. Ueda. Sa l i va and serum v a l p r o i c ac id l e v e l s in e p i l e p t i c c h i l d r e n . J . P e d i a t r . , 100, 512 (1982). W.R. Garnet t , C . L . March, C. T a r a s i d i s and J . M . P e l l o c k . Assessment of total and free v a l p r o i c ac id l e v e l s in serum and plasma in c h i l d r e n . C h i l d Neurology S o c i e t y , 12th National Meeting, Oct . 13-15, 1983, 43P. N. Gerber, R.G. D ick inson, R.C. Harland, R.K. Lynn, D. Houghton, J . I . Antonias and J . C . Schimschock. Reye- l ike syndrome associated with va lpro ic ac id therapy. J . P e d i a t r . , 95, 142 (1979). J . J . G i e r , B . J . ' t Har t , P . F . Wilderink and F . A . Nelemans. Comparison of plasma and s a l i v a l e v e l s of diazepam. Br. J . C l i n . Pharmac , 10, 151 (1980). M.A. Goldberg and T . Todorof f . Brain binding of ant iconvulsants : Carbamazepine and va lp ro ic a c i d . Neurology, 3J3, 826 (1980). L. Gram and K.D. Bentsen. Hepatic t o x i c i t y of a n t i e p i l e p t i c drugs: A review. In: Current therapy in e p i l e p s y , M.V. L ivanainen, e d . , Acta Neurol Scand . , 68 , (Suppl . 97) , 81 (1983). L. Gram, H. F l a c h s , A. Wurtz-Jorgensen, J . Parnas and B. Andersen. Sodium va lproa te , serum level and c l i n i c a l e f f e c t in ep i lepsy : A c o n t r o l l e d study. E p i l e p s i a , 20, 303 (1979). G.R. Granneman, S . I . Wang, J . M . Machinist and J.W. Kesterson. Aspects of the metabolism of v a l p r o i c a c i d . Xenob io t ica , 14, 375 (1984a). G.R. Granneman, S . I . Wang, J.W. Kesterson and J . M . Mach in is t . The hepato tox ic i ty of v a l p r o i c ac id and i t s metaboli te in r a t s . II. Intermediary and va lp ro ic ac id metabolism. Hepatology, 4, 1153 (1984b). R. Gugler and G. Muel ler . Plasma prote in binding of va lp ro ic ac id in healthy subjects and in pat ients with renal d isease . Br . J . C l i n . Pharmacol . , 5, 441 (1978). - 153 R. Gugler , M. Eichelbaum, A. S c h e l l , W. Froscher , H. K i e f e r , H.U. Schulz and G. M u l l e r , In: A n t i e p i l e p t i c Therapy: Advances in drug moni tor ing, S . I . Johannessen, P . L . M o r s e l l i , C . E . Pippenger, A. Richens, D. Schmidt, H. Me inard i , e d s . , Raven P r e s s , New York (1980), p. 121. R. Gugler , A. S c h e l l , M. Eichelbaum, W. Froscher and H.U. Schu lz . D i s p o s i t i o n of va lp ro ic ac id in man. Europ. J . C l i n . Pharmacol . , 12, 125 (1977). R. Gugler and G . E . von Unruh. C l i n i c a l pharmacokinetics of va lp ro ic a c i d . C l i n . Pharmacokinet ics, 5 , 67 (1980). R.N. Gupta, F. Eng and M.L. Gupta. Gas-chromatographic ana lys is for va lpro ic ac id as phenacyl e s t e r . C l i n . Chem., 25_, 1303 (1979). A . G . Har r ison . Chemical Ion izat ion Mass Spectrometry. CRC, Boca Raton, F l o r i d a (1983), p.29-31, 63-80. J . R . Hass. Negative ion chemical i o n i z a t i o n mass spectrometry. In: Mass Spectrometry, part B, C. M e r r i t t , J r . , C .N. McEwen, e d s . , Marcel Dekker, I n c . , New York (1980), p.287-297. R.P. Hatch, J . Shringarpure and S.M. Weinreb. Studies on total synthesis of the o l ivomycins . J . Org. Chem., 43, 4172 (1978). O.Henriksen and S . I . Johannessen. C l i n i c a l observat ions of sodium valproate in c h i l d r e n : An evaluat ion of therapeut ic serum l e v e l s . In: A n t i e p i l e p t i c therapy: Advances in Drug Moni tor ing , S . I . Johannessen, P . L . M o r s e l l i , C . E . Pippenger, A. Richens, D. Schmidt, H. Meinard i , e d s . , Raven Press , New York (1980), p. 253. T . N . H igg ins . Enzyme immunoassay of va lp ro ic acid using the Abbott ABA-200 Analyzer . C l i n . Biochem., 16, 222 (1983). F. Hoffmann, G . E . von Unruh and B . C . J a n c i k . Va lp ro ic ac id d i s p o s i t i o n in e p i l e p t i c pat ients during combined a n t i e p i l e p t i c maintenance therapy. Eur . J . C l i n . Pharmacol . , 19 , 383 (1981). I. Howe, D. Wil l iams and R. Bowen. Mass Spectrometry P r i n c i p l e s and A p p l i c a t i o n s , 2nd e d . , McGraw-Hill I n c . , New York (1981), p.186-189. J . C . H s i a , J . C . L . Tarn, H.G. G i l e s , C C . Leung and H. Marcus. Markers for detect ion of supplementation in narcot ic programs-Deuterium l a b e l l e d methadone. Sc ience , 193, 498 (1976). A . Hulshoff and H. Roseboom. Determination of va lpro ic acid (d i -n -propy l ace t i c acid) in plasma by g a s - l i q u i d chromatography with pre-column b u t y l a t i o n . C l i n i c a Chimica A c t a , 93, 9 (1979). D.F. Hunt, G . C S t a f f o r d , J r . , F.W. Crow and J.W. Russe l . Pulsed p o s i t i v e negative ion chemical i o n i z a t i o n mass spectrometry. A n a l . Chem., 48, 2098 (1976). - 154 P.M. Jaevons. None-dose-related side e f f ec ts of va lproate . E p i l e p s i a , 25, (Suppl . 1 ) , S50 (1984). ~ M.E. Jung and Y . G . Pan. Oxidation of t r i a l k y l s i l y l enol ethers via hydride a b s t r a c t i o n : A new procedure for ketone to enone convers ion . J . Org. Chem. 42, 3951 (1977). P . E . Keane, J . Simiand and M. Morre. Comparison of the pharmacological and biochemical p r o f i l e s of va lp ro ic ac id (VPA) and i t s cerebral metabol i te (2-en-VPA) a f t e r oral administ rat ion in mice. Meth and F ind Exptl C l i n Pharmacol . , ]_, 83 (1985). A . S . Kende and B.H. Toder. Stereochemistry of deconjugative a l k y l a t i o n of es te r d ieno la tes . S t e r e o s p e c i f i c tota l synthesis of the 1 i t s e n o l i d e s . J . Org. Chem., 47, 163 (1982). J .W. Kesterson , G.R. Granneman and J . M . Mach in is t . The hepatotox ic i ty of v a l p r o i c ac id and i t s metabol i tes in r a t s . I. T o x i c o l o g i c , biochemical and h is topa tho log ic s t u d i e s . Hepatology, 4_, 1143 (1984). H. K ido. A study on cerebrospinal f l u i d and s a l i v a to serum concentrat ion r a t i o s of an t iconvu lsants -phenyto in , phenobarbi ta l , v a l p r o i c acid and primidone in e p i l e p t i c pa t i en ts . Se ish in Shinkeigaku Zasshi ( J p n . ) , 84, 661 (1982) (Eng. A b s t r . ) . — E.D . K i n g s l e y , P. Gray, K .G . Tolman and R. Tweedale. The t o x i c i t y of metabol i tes of sodium valproate in cu l tured hepatocytes. J . C l i n . Pharmacol . , 23_, 178 (1983). W.F. K l i n e , D.P. Enagonio, D . J . Reeder and W.E. May. L iqu id chromatographic determination of va lp ro ic ac id in human serum. J . L i q . Chromatogr. , 5, 1697 (1982). U. K lo tz and K.H. Antonin . Pharmacokinetics and b i o a v a i l a b i l i t y of sodium va lproa te . C l i n . Pharmacol. T h e r . , 21 ,^ 736 (1977). W. Kochen, A. Schneider and A. R i t z . Abnormal metabolism of va lp ro ic acid in fa ta l hepatic f a i l u r e . Eur. J . P e d i a t r . , 141, 30 (1983). W. Kochen, H.P. Sprunck, B.Tauscher and M. Klemens. F ive doubly unsaturated metabol i tes of v a l p r o i c ac id in urine and plasma of pat ients on v a l p r o i c a c i d therapy. J . C l i n . Chem. C l i n . Biochem., 22, 309 (1984). J . Koch-Weser and E.M. S e l l e r s . Binding of drugs to serum albumin. N. Eng. J . Med. , 294, 311 (1976 ). J . R . Koup, W.J . Jusko and A . L . Goldfarb . pH-Dependent secre t ion of procainamide into s a l i v a . J . Pharm. S c i . , 64, 2008 (1975). T . Kwong, A. Bacharach, T. Kankoski , F. M i r a b e l l a , R. Martinez and J . K e l l e r . A rapid g a s - l i q u i d chromatographic assay for v a l p r o i c ac id in serum. C l i n . Chem., 26, 1053 (1980). - 155 K. Kushida and T. I s h i z a k i . Concurrent determination of v a l p r o i c acid with other a n t i e p i l e p t i c drugs by high performance l i q u i d chromatography. J . Chromatogr. , 338, 131 (1985). S . Lawesson, M. Dahlen and C. F r i e s e l l . Preparat ion of 3 -subs t i tu ted l e v u l i n i c e s t e r s . Ac ta . Chem. S c a n d . , 16_, 1191 (1962). M.G. Lee and B . J . M i l l a r d . A comparison of unlabel led and l a b e l l e d internal standards for q u a n t i f i c a t i o n by s ing le and mul t ip le ion moni tor ing. Biomed. Mass Spectrom. , 2 , 78 (1975). G. Levy, J . A . Prockna l , R. Olufs and L.M. Pachman. Re la t ionship between s a l i v a s a l i c y l a t e concentrat ion and free or tota l concentrat ion in serum of c h i l d r e n with j u v e n i l e rheumatoid a r t h r i t i s . C l i n . Pharmacol. T h e r . , 27, 619 (1980). ~ ~ R.H. Levy, T . A . Morel and, P . L . M o r s e l l i , M. Guyot, A. Brachet-Liermain and P. Lo iseau . Carbamazepine/valproic acid in te rac t ions in man and rhesus monkey. E p i l e p s i a , 25_, 338 (1984a). R.H. Levy, P.N. F r i e l , I. Johno, L.M. L inth icum, L. C o l i n , K. Koch, V .A . Ra isys , A . J . Wilensky and N.R. Temkin. F i l t r a t i o n for free drug leve l moni tor ing: Carbamazepine and va lpro ic a c i d . Ther. Drug M o n i t . , 6 , 67 (1984b). R. Levy and K.Koch. Drug in te rac t ions with va lp ro ic a c i d . Drugs, 24, 543 (1982). ~ R.H. Levy, A . J . Wilensky and P.N. F r i e l . Other a n t i e p i l e p t i c drugs. In: Appl ied Pharmacokinetics. P r i n c i p l e s of therapeut ic drug moni tor ing, W.E. Evans, J . J . Schentag, W.J. Jusko, e d s . , Appl ied Therapeut ics Inc. Spokane, Washington (1986), p. 540. J . S . Lockard and R.H. Levy. Va lpro ic a c i d : Revers ib ly act ing drug E p i l e p s i a , 17, 477 (1976). W. Loscher . Concentrat ion of metaboli tes of va lpro ic ac id in plasma of e p i l e p t i c p a t i e n t s . E p i l e p s i a , 22, 169 (1981a). W. Loscher . Ant iconvulsant a c t i v i t y of the metaboli tes of va lp ro ic a c i d . Arch . Int. Pharmacodyn., 249, 158 (1981b). W. Loscher . Rapid determination of valproate sodium in serum by g a s - l i q u i d chromatography. E p i l e p s i a , 18, 225 (1977). W. Loscher . Serum prote in binding and pharmacokinetics of valproate in man, dog, rat and mouse. J . Pharmacol. Exp. T h e r . , 204, 255 (1978). W. Loscher and H.H. Frey . Transport of GABA at the blood - CSF i n t e r f a c e . J . Neurochem., 38, 1072 (1982). - 156 W. Loscher and H. Nau. D i s t r i b u t i o n of va lp ro ic ac id and i t s metaboli tes in var ious brain areas of dogs and rats a f te r acute and prolonged treatment. J . Pharmacol, and Exp. T h e r . , 226, 845 (1983). W. Loscher and H. Nau. Pharmcological evaluat ion of various metaboli tes and analogues of v a l p r o i c a c i d . Ant iconvulsant and tox ic potencies in mice. Neuropharmacology, 24, 427 (1985). W. Loscher , H. Nau, C. Marescaux and M. Vergnes. Comparative evaluat ion of ant iconvulsant and tox ic potencies of v a l p r o i c ac id and 2 -en -va lp ro ic ac id in d i f f e r e n t animal models of e p i l e p s y . Eur . J . Pharmacol . , ^ 9 , 211 (1984). S . B . Mat in, S .H . Wan and J . H . Karam. Pharmacokinetics of tolbutamide: P r e d i c t i o n by concentrat ion in s a l i v a . C l i n . Pharmacol. T h e r . , 16, 1052 (1974). ~ ~ G . F . Mattson, R.H. Mattson a n d J . A . Cramer. In teract ion between va lpro ic ac id and carbamazepine: An in v i t r o study of prote in b ind ing . Ther. Drug M o n i t . , 4 , 181 (1982). L . McKauge, J . H . Tyrer and M.J . Ead ie . Factors in f luenc ing simultaneous concentrat ions of carbamazepine and i t s epoxide in plasma. Ther. Drug M o n i t . , 3 , 63 (1981). G. Meunier, G. Car raz , Y. Meunier, P.Eymard and M. Aimart . Propr ie tes pharmacodynamique de V a c i d e n -d ipropy lacet ique . Therapie , 18_, 435 (1963). B.H. Min, J . Pao, W.A. Gar land, J . A . F . de S i l v a and M. Parsonnet. Determination of an ant isecre tory tr imethyl prostaglandin E 2 analog in human plasma by combined c a p i l l a r y column gas chromatography - negative chemical i o n i z a t i o n mass spectrometry. J . Chromatogr., 183, 411 (1980). H. Miyazak i , M. I s h i b a s h i , H. Takayama, K. Yamashita, I. Suwa and M. K a t o r i . Microdeterminat ion of prostaglandins and thromboxane B 2 by c a p i l l a r y gas chromatography and negative ion chemical i o n i z a t i o n mass spectrometry. J . Chromatogr. , 28 9 , 2 4 9 (1984). F . Monaco, S. P i redda , R. Mutani , C. Mastropaolo and M. Tondi . The free f r a c t i o n of v a l p r o i c ac id in t e a r s , s a l i v a , and cerebrospinal f l u i d . E p i l e p s i a , 23 , 23 (1982). J . P . Moody and S.M. A l l a n . Measurement of va lpro ic acid in serum as the 4-bromophenacyl es te r by high performance l i q u i d chromatography. C l i n i c a Chimica A c t a , 127, 263 (1983). R . J . Morrison and R.N. Boyd. Organic Chemistry, 3rd e d . , A l l y n and Bacon Inc. Boston (1973), p.1058, p.328. P . L . M o r s e l l i and R. Franco-Morsel 1 i . C l i n i c a l pharmacokinetics of a n t i e p i l e p t i c drugs in adu l ts . Pharmac. T h e r . , 10_, 65 (1980). - 157 J . C . Mucklow. The use of s a l i v a in therapeut ic drug moni tor ing. Ther. Drug M o n i t . , 4 , 2 2 9 (1982). M. Nakamura, K. Kondo, R. Nishioka and S . Kawai. Improved procedure for the high-performance l i q u i d chromatographic determination of v a l p r o i c acid in serum as i t s phenacyl e s t e r . J . Chromatogr., 310, 450 (1984). H. Nau, H. Helge and W. Luck. Va lpro ic ac id in the per inata l per iod: Decreased maternal serum prote in binding resu l ts in fe ta l accumulation and neonatal displacement of the drug and some metabol i tes . J . P e d i a t . , 104, 627 (1984). H. Nau and W. Loscher . Va lpro ic a c i d : Brain and plasma l e v e l s of the drug and i t s metabo l i tes , ant iconvulsant e f f e c t s andy -Aminobutyric ac id (GABA) metabolism in the mouse. J . Pharmacol. Exp. T h e r . , 220, 654 (1982). H. Nau and W. Loscher. Va lpro ic ac id metabol i tes: Pharmacological and t o x i c o l o g i c a l s t u d i e s . E p i l e p s i a , 25_, ( S u p p l . l ) , S14 (1984). H. Nau, W. Wi t t foht , H. Schafer , C. Jakobs, D. Rating and H. Helge. Va lpro ic ac id and several metabol i tes . Quant i ta t ive determination in serum, u r i n e , breast milk and t i s s u e by gas chromatography - mass spectrometry using se lec ted ion moni tor ing. J . Chromatogr., 2 26 , 69 (1981 ). R. Nishioka and S. Kawai. New method for the gas chromatographic determination of va lp ro ic ac id in serum. J . Chromatogr., 277, 356 (1983). M. Oehme, D. Stockl and H. Knoppel. Comparison of the r e p r o d u c i b i l i t y of negative ion chemical i o n i z a t i o n mass spectra obtained by d i f f e r e n t reagent gases on two commercial quadrupole instruments. Ana l . Chem., 58, 554 (1986). N. Otten, K. H a l l , J . Irvine-Meek, M. Leroux, D. Budnik and S . Sesh ia . Free v a l p r o i c a c i d : Steady-State pharmacokinetics in pat ients with i n t r a c t a b l e e p i l e p s y . Can. J . Neurol . S c i . , 1^, 457(1984). N. Pappas and H.R. Nace. Norster iods II. App l ica t ion of the Favorsk i i rearrangement to the preparat ion of A-Norpregnanes. J . Am. Chem. S o c , 8^, 4556 (1959). I .H. Patel and R.H. Levy. Va lpro ic acid binding to human serum albumin and determination of free f r a c t i o n in the presence of ant iconvulsants and free f a t t y a c i d s . E p i l e p s i a , 20, 85 (1979). E. Perucca, G. G a t t i , G.M. Fr igo and A. Crema. Pharmacokinetics of v a l p r o i c ac id a f te r oral and intravenous admin is t ra t ion . Br . J . C l i n . Pharmacol . , 5_, 313 (1978). J . A . P e t r e r e , J . A . Anderson, R. Sakowski, J . E . F i t z g e r a l d and F . A . de l a I g l e s i a . Teratogensis of calcium valproate in r a b b i t s . Tera to logy , 34_, 263 (1986). - 158 F. P i s a n i , G. Oteri and R. D i p e r r i . Va lpro ic ac id f a i l s to d isp lace phenobarbitone and carbamazepine from plasma prote in binding s i tes in e p i l e p t i c p a t i e n t s . Br . J . C l i n . Pharmacol . , 12_, 81 (1981). K . S . P r i c k e t t and T . A . B a i l l i e . Metabolism of va lpro ic acid by hepatic microsomal cytochrome P-450. Biochem. Biophys. Res. Comm., 122, 1166 (1984). C . S . Ramesha and W.C. P i c k e t t . Measurement of sub-picogram quant i t i es of p l a t e l e t a c t i v a t i n g f a c t o r (AGEPC) by gas chromatography/negative ion chemical i o n i z a t i o n mass spectrometry. Biomed. Env i ron . Mass Spectrom., 1^ 3_, 107 (1986). A.W. Rettenmeier, W.P. Gordon, K . S . P r i c k e t t , R.H. Levy and T . A . B a i l l i e . Biotransformat ion and pharmacokinetics in the rhesus monkey of 2 -n -propy l -4 -penteno ic a c i d , a tox ic metaboli te of va lp ro ic a c i d . Drug Metab. D i s p o s . , 14, 454 (1986a). A.W. Rettenmeier, W.P. Gordon, K .S . P r i c k e t t , R.H. Levy, J . S . Lockard, K . E . Thummel and T . A . B a i l l i e . Metabolic fate of va lp ro ic ac id in the Rhesus monkey, formation of a tox ic metabol i te , 2 -n -propy l -4 -penteno ic a c i d . Drug Metab. D i s p o s . , 14, 443 (1986b). A.W. Rettenmeier, K . S . P r i c k e t t , W.P. Gordon, S.M. Bjorge, S . L . Chang, R.H. Levy and T . A . B a i l l i e . Studies on the biotransformat ion in the perfused rat l i v e r of 2 -n -propy l -4 -penteno ic a c i d , a metabolite of the a n t i e p i l e p t i c drug v a l p r o i c a c i d . Evidence for the formation of chemical ly react ive in termediates . Drug Metab. D i s p o s . , 13_, 81 (1985). A . E . R e t t i e , A.W. Rettenmeier, W.N. Howald and T . A . B a i l l i e . Cytochrome P-450-cata lyzed formation of Af* -VPA, a tox ic metaboli te of va lp ro ic a c i d . Sc ience , 235 , 890 (1987). E . M. Rimmer and A. Richens. An update on sodium va lproa te . Pharmacotherapy, 5, 171 (1985). A . J . Rowan, C D . B i n n i e , C .A. W a r f i e l d , H.Meinardi and J . W . A . Me i je r . The delayed e f f e c t of sodium valproate on the photoconvulsive response in man. E p i l e p s i a , 20, 61 (1979). F . Rubio and W.A. Gar land. Determination of a c e t y l t r i m o p r o s t i l and i t s metabol i te t r imopros t i l in human or dog plasma by gas chromatography -negat ive - ion chemical i o n i z a t i o n mass spectrometry. J . Chromatogr., 339, 313 (1985). G. M. S c h i e r , I .E . Tho Gan, B. Hal pern and J . Kor th . Measurement of sodium valproate in serum by d i r e c t - i n s e r t i o n chemical ion izat ion/mass spectrometry. C l i n . Chem., 26_, 147 (1980). D. Schmidt. Adverse e f f e c t s of va lproate . E p i l e p s i a , 25_, (Suppl . 1) , S44 (1984). F . Schobben, E. van der K l e i j n and T . B . Vree. Therpeutic drug monitoring of v a l p r o i c a c i d . Ther. Drug Mon i t . , 2_, 61 (1980). - 159 -Y. Shechter and Y . I . Hem's. c is -Unsatura ted fa t ty acids induce both l i p o g e n s i s and calcium binding in ad ipocytes . Biochimic et Biophysica A c t a , 805 , 89 (1984). L . B . Siegmund, T. Annedore, V. Wolfgang, J . Karl and K. Maximi l ian . Evaluat ion of a new VPA enzyme immunoassay and comparison with a c a p i l l a r y GC method. C l i n . Chem., 2J_t 169 (1981). I. Sjoholm. The s p e c i f i c i t y of drug binding s i tes on human serum albumin. In: Drug-protein b ind ing , M.M. Reidenberg and S. E r i l l , e d s . , Praeger Publ ishers New York (1984), p. 36. A . A . Spector and J . E . F l e t c h e r . Transport of fa t ty ac id in the c i r c u l a t i o n . In: Disturbances in l i p i d and l i p o p r o t e i n metabolism, J . M . Die tschy, A.M. Got to , J r . , J . A . Ontko, e d s . , American Phys io log ica l S o c i e t y , Maryland (1978), p. 229. H . J . Stan and M.S. Re ich . Detection of hydroxy fa t ty ac ids in b i o l o g i c a l samples using c a p i l l a r y gas chromatography in combination with pos i t i ve and negative chemical i o n i z a t i o n mass spectrometry. L i p i d s , 1S_, 1044 (1980). S. Stout . App l i ca t ion of gas chromatography negative ion chemical i o n i z a t i o n mass spectrometry in confirmatory procedures for p e s t i c i d e res idues . Biomed. Mass Spectrom., 207 (1984). R . J . S t r i f e and R.C. Murphy. Preparat ion of pentaf luorobenzyl esters of arachidonic acid l ipoxygenase metabol i tes . Analys is by gas chromatography and negat ive- ion chemical i o n i z a t i o n mass spectrometry. J . Chromatogr., 305, 3 (1984). C. Sutheimer, D. F re t tho ld and I. Sunshine. The determination of va lpro ic ac id in plasma using reversed-phase l i q u i d chromatography. Chromatography Newslet ter , 7 (2 ), 1 (1979). J . H . Thurston, J . E . C a r r o l l , R .E . Hauhart and J . A . S c h i r o . A s ing le therapeut ic dose of valproate a f f e c t s l i v e r carbohydrate, f a t , adenylate, amino a c i d , coenzyme A, and c a r n i t i n e metabolism in in fant mice: Poss ib le c l i n i c a l s i g n i f i c a n c e . L i f e Sc iences , 36_, 1643 (1985). D. M. T u r n b u l l , D . J . D ick , L. Wi lson, H. Sherrat t and K. A l b e r t i . Valproate causes metabolic disturbance in normal man. J . Neurology, Neurosurgery, and P s y c h i a t r y , 49 , 405 (1986 ). F . J . E . Vajda. Va lp ro ic ac id in the treatment of status e p i l e p t i c u s . In: Advances in Neurology, A . V . Delgado-Escueta, C G . Waster la in , D.M. Treiman, R . J . P o r t e r , e d s . , Raven P r e s s , New York, V o l . 34, 1983, p. 519. F . J . E . Vajda, G.A. Donnan, J . P h i l i p s and P .F . B l a d i n . Human b r a i n , plasma and cerebrospinal f l u i d concentrat ion of sodium valproate a f te r 72 hours of therapy. Neurology (NY), 31_, 486 (1981). C A . M . Van Ginneken, J . F . R . de Miranda and A. J . Be ld . S t e r e o s e l e c t i v i t y and drug d i s p o s i t i o n . In: Stereochemistry and b i o l o g i c a l a c t i v i t y of drugs, E . J . A r i e n s , W. Soud i jn , P.B.M.W.M. Timmermans, e d s . , Blackwell S c i e n t i f i c P u b l i c a t i o n s (1983), p. 55. - 160 N . P . E . Vermeulen and D.D. Breimer. S t e r e o s e l e c t i v i t y in drug and xenobiot ic metabolism. In: Stereochemistry and b i o l o g i c a l a c t i v i t y of drugs, E . J . A r i e n s , W. Soudi jn , P.B.M.W.M. Timmermans, e d s . , Blackwell S c i e n t i f i c P u b l i c a t i o n s (1983), p.33. G . E . von Unruh, B.Ch. Jancik and F. Hoffmann. Determination of va lpro ic acid k i n e t i c s in pat ients during maintenance therapy using a tetradeuterated form of the drug. Biomed. Mass Spectrom, ]_, 164 (1980). K.A. Waddell and I.A. B l a i r . Combined c a p i l l a r y column gas chromatography negative ion chemical i o n i z a t i o n mass spectrometry of prostanoids . Biomed. Mass Spectrom., 10, 83 (1983). J . T . Watson. Introduction To Mass Spectrometry. Raven P r e s s , New York (1985), p.192-196. J . B . West (ed) . Phys io log ica l Basis of Medical P r a c t i c e , 11th e d . , Wil l iams and W i l k i n s , Balt imore (1985), p. 607. J . A . F . Wickramasinghe, W. Morozowich, W.E. Hamlin and S .R. Shawn. Detection o f prostaglandin F20t as pentafl uorobenzyl ester by e lec t ron capture GLC. J . Pharm. S c i . , 6 £ , 1428 ( 1973). E . W y l l i e , R. W y l l i e , R.P. Cruse, G. Erenberg and D. Rothner. P a n c r e a t i t i s associa ted with va lp ro ic ac id therapy. A . J . D . C . , 138, 912 (1984). J . Zuidema and C .A .M . van Ginneken. Clearance concept in s a l i v a r y drug e x c r e t i o n . Pharm. Acta H e l v . , 58_, 88 (1983a). J . Zuidema and C .A .M . van Ginneken. Clearance concept in s a l i v a r y drug e x c r e t i o n . Pharm. Acta H e l v . , 58, 136 (1983b). - 161 -APPENDIX NMR, IR and mass spectra of some of the synthesized compounds. IJIICROMETf. US i»m) 13 14 16 18 70 75 i L to 11 JO 2 4000 3600 JJCO 7U00 V i i J j i v j j V i J w J l i .OJ 1000 l-i(/0 i i . L O U l i N C V (CM ') IR spectrum of ethyl 2-propyl -4-oxopentanoate. 1000 y.jo 40U <T> H-NMR (80 MHz) spectrum of ethyl 2 -propyl -4 -oxopentanoate . - 164 -100 >-I— I—H CO L U L U 73 45 29 25 CH 0-TMS 2 = C - ™ 2 * : C H - C - 0 - C 2 H 5 C H 3 - C H 2 - C H 2 185 115 130 97 143 L i L _ l L_J ! 1 5 243258 (Mt ) 125 M / Z 225 325 EI spectrum of the TMS enol ether of ethyl 2-propyl -4-oxopentanoate. - 165 -100 _ 73 >-h -C O L U c r _ i L U 43 29 25 0-TMS I 0 CH = C - C H || CH - C H - C H ^ 95 137 109 183 213 241 256 (Mt) 125 M / Z 225 325 EI spectrum of the TMS enol ether of ethyl 2 -propy l -4 -oxo-2 -pentenoate . r | mi | il 111 in i| I I 111 II 11111 \f\ 3 SO i. « 0 J 50 pew 20 A —' I 1 ..J \ CTl J L _ 1 1 1 I 1 1 1 1 I 1 1 1 1 1 i i — i — i — I — i — i — t — i — i — i — t — i — i — i — i — t — i — i — i — [ — i — i — i — j — i — i — i — i — i — r 4 0 3 5 3 0 2 b 2 0 l b 1 0 0 b 0 0 PPM 1 H-NMR (300 MHz) spectrum of ethyl 4 -e thy leneth ioketa lpentanoate . - 167 -100 _ >-I— i—i cn L U i— L U > f— C E _ l L U 29 25 127 I 2 I 2 S S C H 3 — C — C H 2 0 \ H C — C - 0 - C „ H c 2 5 C H ^ C H ^ C H 199 215 260 ( M t ) 125 M / Z 225 325 EI mass spectrum of 4-carboethoxy-2-ethy1enethioketa1-4-heptene (isomer with the shorter re ten t ion t ime) . - 168 -100 _ 199 >-i— i—i cn L U L U 29 25 4 127 I 2 I 2 s s \ / C H , - C — C H 9 0 3 2v y C - C - 0 - C O H , # 2 5 C H ^ C H A - C H 260 (M-) 215 125 M / Z 1 225 325 EI mass spectrum of 4 -carboethoxy-2 -e thy leneth ioketa l -4 -heptene (isomer with the longer re ten t ion t ime) . 

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.831.1-0096961/manifest

Comment

Related Items