Search

The synthesis and binding affinities to the digitalis Na(+),K(+)-ATPase receptor of a series of 3 beta,14 beta-dihydroxy-5 beta-androstane and 3 beta-hydroxy-14-oxoseco-D-5 beta-androstane derivatives bearing a 17 alpha-(aminoalkoxy)imino chain are reported; some derivatives were also studied for their inotropic activity. Our recently proposed model of interaction of molecules with the digitalis receptor was used to design these compounds. On that basis, the possibility to design novel potent inhibitors of Na(+),K(+)-ATPase without being constrained by the stereochemistry of the classical digitalis skeleton in the D-ring region was predicted. The binding affinities of the most potent compounds in the two series, (EZ)-17 alpha-[2-[(2-aminoethoxy)imino]ethyl]-5 beta-androstane-3 beta,14 beta-diol (6f) and (EZ)-3 beta-hydroxy-17 alpha-[2-[(2-aminoethoxy)imino]ethyl]-14,15-seco-5 beta-androstan-14-one (24c) are higher than that of the potent natural compound digitoxigenin, despite the unusual alpha-exit of the substituent in position 17 of 6f or the disruption of the D-ring in 24c. These results further support the validity of our recently proposed model of binding at the digitalis receptor. Results of the inotropic tests on guinea pig atrium deserve further investigation on the pharmacological profile of these derivatives.

A series of digitalis-like compounds with a 17-aminoalkoxyiminoalkyl or -alkenyl substituent was synthesized and evaluated for inhibition of Na(+),K(+)-ATPase and for inotropic activity. The highest inhibition was found with compounds having the substituent in configuration 17beta and the amino group at a distance of 6 or 7 bonds from C(17) of the digitoxigenin skeleton. The presence of the oxime function strengthens the interaction with the receptor, more if alpha,beta-unsaturated, thus mimicking the electronic situation of the unsaturated lactone in natural digitalis compounds. The most active compounds showed Na(+),K(+)-ATPase inhibitory potencies (IC(50)) 17-25 times higher than the standards digitoxigenin and digoxin and 3-11 times higher inotropic potencies (EC(50)) in isolated guinea pig left atria. These features are supported by a molecular model suggesting the possible interactions of the groups described above with particular amino acid residues in the H1-H2 domains of Na(+),K(+)-ATPase. Some interactions are the classical ones already described in the literature; a new, very strong interaction of the basic group with the Cys138 was found and adds new possibilities to design compounds interacting with this region of the receptor. The most interesting compounds were also studied in vivo in the anesthetized guinea pig for evaluating their inotropic effect versus the lethal dose. Compounds 9 and 12 showed a slightly higher safety ratio than digoxin and deserve further evaluation.

A series of 17 beta-(hydrazonomethyl)-5 beta-androstane-3, beta,14 beta-diol derivatives was synthesized and evaluated in the displacement of [3H]ouabain binding from Na+,K(+)-ATPase. The data were explored with multiple linear regression and partial least-squares to find possible quantitatives structure-activity relationships. Good correlations were found between binding to the receptor and van der Waals volumes or molar refractivities of the 17 beta-hydrazonomethyl substituents and pKa values of the compounds. Equivalent results were obtained using the proton affinity (calculated using MOPAC) of the hydrazone residues instead of experimental pKa. As basicity or related electronic factors of the substituents explain a significant portion of the observed changes in the activity, an ion-pair interaction between a carboxylate residue of the enzyme and the protonated 17 beta-hydrazonomethyl group, as postulated by Thomas, plays an important role in the interaction of the ligand to the Na+,K(+)-ATPase receptor.