Metabolism of dexamethasone in the human kidney: Nicotinamide adenine dinucleotide-dependent 11β-reduction

Abstract

Recently, two distinct isoenzymes of 11β-hydroxysteroid-dehydrogenase (11β-HSD) have been cloned and characterized in several species: The isoenzyme 11β-HSD-I is widely distributed, bidirectional, prefers NADP(H) and has a low substrate affinity. The isoenzyme 11β-HSD-II seems to exclusively oxidize physiological glucocorticoid, uses NAD as cosubstrate, has high substrate affinity, and is only found in mineralocorticoid target tissues and the placenta. Synthetic steroids fluorinated in position 9, however, are rapidly reduced by human kidney cortex slices. We attempted to find out which isoenzyme is responsible for this unexpected reductase activity. We studied the 11β-HSD activity towards cortisol (F)/cortisone let and dexamethasone (D)/11-dehydro-dexamethasone (DH-D) in microsomes prepared from human kidney cortex. For the reaction E to F (not for DH-D to D!), glucose-6-phosphate and glucose-6-phosphate-dehydrogenase had to be added as a NADH/NADPH-regenerating system. Oxidation off to E: NAD was the exclusively used cosubstrate; the affinity [Michael's constant (K(m)) for F = 25.5 nmol/L] and the maximum velocity (V(max) = 22.9 nmol/mg/min) were high. Reduction of E to F: Without the NADH/NADPH-regenerating system, this reaction was very slow. With this system, the K(m) value for E was in the nanomolar range (80.6 nmol/L) and the V(max) value was very low (0.88 nmol/mg/min). The reaction was clearly NADH-preferring. For the steroid pair F/E, the quotient V(max) (oxidation)/V(max)/V(reduction) (26) demonstrates an equilibrium far on the 11-keto side. Oxidation of D to DH-D: With NAD as the only used cosubstrate, the kinetic analysis is corn patible with the existence of two different NAD-dependent isoenzymes: K(m) for D = 327 nmol/L, V(max) = 53.5 nmol/mg/min and K(m) for D = 81.2 nmol/L; V(max) = 20.4 nmol/mg/min. Reduction of DH-D to D: The maximum velocity was higher than that of all other reactions tested: V(max) =226.0 nmol/mg/min. The reaction was exclusively NADH-dependent; the K(m) value for DH-D was 68.4 nmol/L. For D/DH-D, the ratio V(max)(oxidation)/V(max)(reduction) was 0.24, demonstrating a shift to reductase activity with the reaction equilibrium far on the 11- hydroxy side. The reaction F to E was inhibited by E, DH-D, and D in a concentration-dependent manner. In conclusion, the cosubstrate dependence, the K(m) value of the oxidation of F and the product inhibition are in good correspondence with data for the cloned human 11β-HSD-II. The NADH-dependent 11β-reduction of E and especially of DH-D are inconsistent with the dogma of an unidirectional 11β-HSD-II. The preference of D for the reductase reaction in human kidney slices is probably caused by the fluor atom in position 9, is catalyzed by 11β HSD-II, and leads to an activation of 11-DH-D to D in the human kidney.