Polymorphism at the ADH2 and ADH3 loci of alcohol dehydrogenase (ADH) has been shown to have an effect on the predisposition to alcoholism in Asian individuals. However, the results are not conclusive for white individuals. We have analyzed the ADH genotype of 876 white individuals from Spain (n = 251), France (n = 160), Germany (n = 184), Sweden (n = 88), and Poland (n = 193). Peripheral blood samples from healthy controls and groups of patients with viral cirrhosis and alcohol-induced cirrhosis, as well as alcoholics with no liver disease, were collected on filter paper. Genotyping of the ADH2 and ADH3 loci was performed using polymerase chain reaction-restriction fragment length polymorphism methods on white cell DNA. In healthy controls, ADH2*2 frequencies ranged from 0% (France) to 5.4% (Spain), whereas ADH3*1 frequencies ranged from 47.6% (Germany) to 62.5% (Sweden). Statistically significant differences were not found, however, between controls from different countries, nor between patients with alcoholism and/or liver disease. When all individuals were grouped in nonalcoholics (n = 451) and alcoholics (n = 425), ADH2*2 frequency was higher in nonalcoholics (3.8%) than in alcoholics (1.3%) (P = .0016), whereas the ADH3 alleles did not show differences. Linkage disequilibrium was found between ADH2 and ADH3, resulting in an association of the alleles ADH2*2 and ADH3*1, both coding for the most active enzymatic forms. In conclusion, the ADH2*2 allele decreases the risk for alcoholism, whereas the ADH2*2 and ADH3*1 alleles are found to be associated in the European population.

Aldo–keto reductases (AKRs) are NAD(P)H-dependent oxidoreductases that catalyse the reduction of a variety of carbonyl compounds, such as carbohydrates, aliphatic and aromatic aldehydes and steroids. We have studied the retinal reductase activity of human aldose reductase (AR), human small-intestine (HSI) AR and pig aldehyde reductase. Human AR and HSI AR were very efficient in the reduction of all-trans-, 9-cis- and 13-cis-retinal (kcat/Km=1100–10300 mM−1·min−1), constituting the first cytosolic NADP(H)-dependent retinal reductases described in humans. Aldehyde reductase showed no activity with these retinal isomers. Glucose was a poor inhibitor (Ki=80 mM) of retinal reductase activity of human AR, whereas tolrestat, a classical AKR inhibitor used pharmacologically to treat diabetes, inhibited retinal reduction by human AR and HSI AR. All-trans-retinoic acid failed to inhibit both enzymes. In this paper we present the AKRs as an emergent superfamily of retinal-active enzymes, putatively involved in the regulation of retinoid biological activity through the assimilation of retinoids from β-carotene and the control of retinal bioavailability.