Validation-Lessons learned from practical experience

With regard to the problems encountered and the experience gained in validation studies conducted in the past, suggestions have been made concerning criteria for the selection of the tests and laboratories to be included in a validation study, the selection and distribution of test chemicals, and procedures for the handling, analysis and interpretation of the resulting data. In particular, tests should have been developed to the extent that detailed protocols and standard operating procedures have been produced and evaluated. The laboratories should be chosen on the basis of evidence of their appropriate experience, competence and ability to comply with good laboratory practice (GLP) requirements. The choice of test chemicals depends primarily on the goals of the validation study and on the availability of reliable in vivo toxicity data of high quality. A biostatistician should be involved in the initial design of the validation study as well as in the analysis of the resulting data. The quality of the in vivo and in vitro data must be ensured, prior to determining the reproducibility and predictivity of the alternative test.

Why, when and how in vitro tests should be accepted into regulatory toxicology

The use of in vitro techniques in toxicological research is widespread, but, up to now, relatively little progress has been made in applying the knowledge gained in regulatory toxicity testing. In vitro tests should be accepted into regulatory toxicology for at least five reasons: scientific, humanitarian, legislative, logistical and economic. In particular, in vitro tests have the potential to provide a mechanistic basis for toxicity testing, and they may permit the use of tissues from more-appropriate target species and individuals, including humans. The relevance and reliability of the in vitro test, with regard to its use for a particular purpose and with particular types of chemicals, should have been adequately demonstrated (i.e. it should have been validated) prior to regulatory acceptance. There are several obstacles to this, including whether validation should be based on comparisons between in vitro data and animal data, and whether the in vitro tests should be expected to provide regulators with the same kinds of predictions and classification criteria that they currently obtain from animal tests. Regulatory incorporation should be a permissive process, rather than a restrictive one. Any scientifically defensible in vitro test which has been properly validated and independently recommended, should be acceptable for the specific purposes for which its use would be appropriate.

Structural requirements for the direct and cytochrome P450-dependent reaction of cyclic alpha,beta-unsaturated carbonyl compounds with glutathione: a study with coumarin and related compounds

The interaction of glutathione (GSH) with coumarin, or one of a series of compounds related to coumarin, was assessed in the absence and presence of liver microsomes (direct reaction and indirect reaction, respectively) to determine the structural requirements for direct and mono-oxygenase-mediated reaction of cyclic alpha,beta-unsaturated carbonyls with GSH. Acrolein was used as a positive control for the direct reaction, and produced complete or nearly complete depletion of GSH under all assay conditions. 5,6-Dihydro-2H-pyran-2-one and 2-cyclohexen-1-one also produced substantial depletion of GSH in the direct reaction, which was not increased by the addition of liver microsomes. Coumarin, 2H-pyran-2-one and precocene I (a substituted pyran lacking the 2-one structure) were not substrates for the direct reaction but did cause depletion of GSH when incubated in the presence of rat or human liver microsomes. These depletions were dependent on a functioning mono-oxygenase system as judged by the effects of omission of cofactors, addition of competitive or inactivating inhibitors of cytochrome P450, and induction. Dihydrocoumarin, delta-valerolactone, cyclohexanone and 4H-pyran-4-one were not substrates for either the direct or indirect reaction. These findings are rationalized on the basis of a direct nucleophilic attack of GSH on the alpha,beta-centre of the alpha,beta-unsaturated carbonyl compounds, which is hindered by benzenoid resonance in coumarin and 2H-pyran-2-one, for which enzyme-mediated reaction with GSH, probably via a 3,4-epoxide, is the favoured mechanism.

Species differences in the metabolism and hepatotoxicity of coumarin

1. Investigations of coumarin metabolism and hepatotoxicity have been reviewed. 2. Species differences in coumarin hepatotoxicity appear to be metabolism-mediated. 3. The rat, in which it is markedly hepatotoxic, primarily metabolises coumarin via 3-hydroxylation and cleavage of the heterocyclic ring. 4. Coumarin is less toxic in the baboon, gerbil and certain strains of mice, which resemble man in their extensive formation of the 7-hydroxy metabolite. 5. Liver toxicity in patients receiving relatively high daily doses of coumarin is very rare. 6. Recent studies indicate that coumarin 3,4-epoxide is the metabolic intermediate responsible for hepatotoxicity in the rat.

Metabolism of coumarin by rat, gerbil and human liver microsomes

1. o-Hydroxyphenylacetaldehyde was the major metabolite of coumarin (1 mM) in rat, gerbil and human liver microsomes. 2. Treatment of rats with phenobarbitone (PB) or beta-naphthoflavone increased the o-hydroxyphenylacetaldehyde formed. 3-Hydroxycoumarin was the other main metabolite produced by rat liver microsomes. 3. Liver microsomal metabolism of coumarin in gerbil was extensive with 3-, 5-, 6-, 7- and 8-hydroxycoumarins, and 3,7- and 6,7-dihydroxycoumarins produced, in addition to o-hydroxyphenylacetaldehyde. The profile of the hydroxy metabolites was altered by in vivo treatment of gerbils with cytochrome P-450 inducers, but there was no increase of coumarin metabolism. 4. Coumarin was metabolized by human liver microsomes to o-hydroxyphenylacetaldehyde, 7-hydroxycoumarin, 3-hydroxycoumarin, and trace amounts of 5-, 6- and 8-hydroxycoumarins. 5. At low substrate concentrations (0-10 microM) hepatic microsomal metabolism of coumarin in gerbil resembled that in man, with 7-hydroxycoumarin being a major metabolite. However, the production of o-hydroxyphenylacetaldehyde was greater in gerbil than human liver microsomes. 6. At higher substrate concentrations (1 mM) metabolism of coumarin by liver microsomes from PB-treated gerbils most closely resembled that by human liver microsomes. 7. The gerbil would appear to be a more appropriate animal model than rat for studies to assess the toxicological hazard of coumarin for man.