Fluorinated analogues as mechanistic probes in valproic acid hepatotoxicity: comparative metabolic and pharmacokinetic studies

• Main Author: Tang, Wei
• Format: Others
• Language: English
• Published: 2009
• Online Access: http://hdl.handle.net/2429/6204

A serious drawback in the use of the anticonvulsant agent valproic acid (VPA) is the drug associated liver toxicity characterized by microvesicular steatosis frequently accompanied by necrosis. The main objective of this thesis was to test the hypothesis that the hepatotoxicity of VPA is due to the formation of reactive toxic metabolites. Firstly, metabolic activation of VPA was investigated by detection and characterization of drug-related thiol conjugates. Combined LC / MS / MS and NMR evidence clearly identified 5-GS-3-ene VPA-glucuronide I in the bile of rats dosed with (E)-2,4-diene VPA which is suspected to play a key role in VPA hepatotoxicity. Sufficient on-line LC / MS / MS data were obtained to indicate the presence of the NACglucuronide di-conjugate of (E)-2,4-diene VPA in both rat bile and urine. The amount of biliary 5-GS-3-ene VPA-glucuronide I was 7-fold greater than 5-GS-3-ene VPA, the sum of the two metabolites accounting for 6.6% of the dose. Incubation of 2,4-diene VPA-glucuronide with GSH in the presence of glutathione S-transferase (GST) enzyme led to the formation of the GSH-glucuronide di-conjugate. To the best of our knowledge, this is the first recorded instance in which glucuronide formation activates a drug to further conjugate with GSH via a Michael addition reaction. In other experiments, LC / MS / MS analysis of bile samples collected from rats dosed with 4-ene VPA, an analogue of the known hepatotoxicant 4-pentenoic acid (4PA), indicated the presence of the GSH, cysteinylglycine, cysteine and NAC conjugates of 4,5-epoxy VPA and (E)-2,4-diene VPA , respectively. Quantitatively, the biliary thiol conjugates accounted for 5% of the dose. This observation is novel for 4- ene VPA metabolism in terms of the degradation of GSH conjugates possibly occurring within the liver as opposed to an inter-organ process which involves the kidney. The GSH - and NAC-glucuronide di-conjugates of (E)-2,4-diene VPA were also identified as metabolites with 5-GS-3-ene VPA-glucuronide I representing 0.03% of the 4-ene VPA dose. Taken together, these data clearly indicate that reactive metabolites of VPA can react with hepatic GSH via several different metabolic pathways, the subsequent depletion of GSH having potential toxic consequences. Additionally, (E)-2,4-diene VPA, in its esterified forms, was demonstrated to be capable of alkylating reduced oxytocin at the free cysteine residues, implicating a direct modification of critical proteins by the diene metabolite of VPA . The role of GST in the conjugation of GSH with (E)-2,4-diene VPA was investigated using rat liver subcellular fractions as the source of GST enzymes. The GST mediated conjugation of GSH with (E)-2,4-diene VPA A/-acetylcysteamine thioester, a structural mimic of the corresponding CoA thioester, resulted in two isomeric products via either 5,6- or 1,6-addition, in agreement with in vivo observations. Only the 1,6-addition product was found for the spontaneous reaction of GSH with the unsaturated thioester (control). Quantitatively, GSH conjugates formed in the presence of the cytosol and sonic disrupted mitoplasts were 23- and 2-fold that of control, respectively. No reaction could be detected upon a mix of GSH with the free acid form of (E)-2,4-diene VPA. The results indicate that GST enzymes enhance the addition of GSH to (E)-2,4-diene VPA with the esterified diene being essential for the reaction. To further examine the metabolic activation hypothesis, oc-fluoro-4-ene VPA which was expected to be inert to p-oxidative metabolism was synthesized and its effect on rat liver studied in comparison with 4-ene VPA . Following treatment of rats for 5 days, 4-ene VPA, but not a-fluoro-4-ene VPA, induced severe hepatic microvesicular steatosis (> 8 5% affected hepatocytes) and alterations in mitochondria. Similar results were obtained when 4-pentenoic acid and 2,2-difluoro-4-pentenoic acid were compared. The p-oxidation product of 4-ene VPA, namely (E)-2,4-diene VPA , and the ^-acetylcysteine (NAC) conjugate of the diene could not be detected in rats administered oc-fluoro-4-ene VPA . In a separate acute study, mitochondrial GSH was determined to remain unchanged in rats treated with cc-fluoro-4-ene VPA but was reduced to 68% of control in those administered 4-ene VPA . These data are consistent with results derived from metabolic studies, suggesting that formation of a reactive intermediate is a key step in the events leading to 4-ene VPA, and possibly VPA, induced liver injury with depletion of mitochondrial GSH as one of the causative factors. A subsequent investigation was carried out to compare 4-ene VPA and a-fluoro- 4-ene VPA for their pharmacokinetic and protein binding properties. The serum concentration-time profiles of 4-ene VPA and a-fluoro-4-ene VPA were observed to resemble one another during the initial 200 min within which differences were apparent for the drug effects on mitochondrial GSH . The major phase II metabolites were the L-glutamine conjugate for cc-fluoro-4-ene VPA and the glucuronide ester for 4-ene VPA . The toxic metabolite (E)-2,4-diene VPA and its NAC conjugate were again detected only in 4-ene VPA treated rats. Despite differences in metabolism, the disposition to rat liver, the serum peak and free concentrations were comparable for 4-ene VPA and ocfluoro- 4-ene VPA. Thus, the apparent distinction between the two drugs in producing liver toxicity in rats is unlikely to be associated with pharmacokinetic differences. Finally, because of the apparent nonhepatotoxic property of a-fluoro-4-ene VPA , a-fluoro VPA was evaluated for anticonvulsant activity in mice. The ED50 of the drug was determined to be 1.7 mmol/kg with the peak activity occurring at 45 - 60 min following the dose, in contrast to 10 min for VPA . Subsequent kinetic studies revealed that the brain uptake of a-fluoro VPA was slower, the peak brain concentration arriving 45 min later than in the serum, whereas the peak brain level of VPA coincided with the peak serum level occurring within 15 min of the dose. On the other hand, a-fluoro VPA appeared to persist in the general circulation, resulting in its apparent slow elimination from the brain. a-Fluoro VPA was demonstrated to have anticonvulsant activity in the pentamethylenetetrazole seizure test in mice and to be capable of increasing brain synaptosomal GABA, although the connection between these two events remains to be clarified. These results suggest that a-fluoro VPA has potential as a new anticonvulsant drug. === Pharmaceutical Sciences, Faculty of === Graduate