A series of 4-amino-2,2-diarylbutyronitriles (3) prepared for testing as inhibitors of gastrointestinal propulsive activity did not show any enhancement over such existing agents as diphenoxylate and loperamide. However, conversion of the nitrile group to a 2-methyl-1,3,4-oxadiazol-5-yl function led to compounds 5g and 5j, statistically equipotent to diphenoxylate and loperamide in the mouse and showing a very low order of analgesic activity. Structural modifications determined that the best separation of antipropulsive and analgesic effects was obtained when the amino group was bicyclic and the oxadiazole ring had a 2-methyl substituent. The most potent compounds were and analogues of diphenoxylate and loperamide where the oxadiazole ring was present, but these compounds had marked analgesic activity.

An efficient large-scale process to prepare the HIV protease inhibitor urea intermediate, N-[3(S)-[bis(phenylmethyl)amino-2(R)-hydroxy-4-phenylbutyl]-N ′-(1,1-dimethylethyl)-N-(2methylpropyl)urea was developed. The protected alcohol, β-(S)-[bis(phenylmethyl)amino]benzenepropanol, was obtained in 95% yield in one step by the benzylation of L-phenylalaninol with benzyl bromide under aqueous conditions. Oxidation of protected alcohol with sulfur trioxide pyridine complex in DMSO at 15 °C gave the corresponding aldehyde in quantitative yield. The dimethyl sulfide byproduct was easily removed by nitrogen sparging and treatment of the effluent gas stream with bleach solution. Diastereoselective reaction of the chiral amino aldehyde with (chloromethyl)lithium at -35 °C followed by warming to room temperature gave the desired epoxide stereoselectively in good yield. A DOE (statistical design of experiment) study indicated that the reaction concentration and halogen reagent were important factors for this reaction. To simplify the operations and to increase the productivity of epoxide, a continuous process was developed. Regioselective ring opening of epoxides with isobutylamine followed by reaction of the resulting amine with tert-butyl isocyanate in isopropyl alcohol gave the urea N-[3(S)-[bis(phenylmethyl)amino]-2(R)-hydroxy-4-phenylbutyl]-N ′-(1,1-dimethylethyl)-N-(2-methylpropyl)urea, in good yield. The process improvements for the crystallization of urea are also discussed.

A series of substituted 2-[(2-benzimidazolylsulfinyl)methyl]anilines were synthesized as potential inhibitors of the acid secretory enzyme H+/K+ ATPase. Substitutions on the aniline nitrogen atom resulted in potent enzyme inhibition in vitro but weak activity in gastric fistula dogs. Electron-donating substituents on the aniline ring enhanced in vitro and in vivo potency relative to the unsubstituted analogue. The potency showed a correlation to the calculated pKa of the aniline nitrogen atom. Substitutions on the aniline and benzimidazole rings did not further enhance potency. Di- and trisubstituted aniline derivatives were potent inhibitors of the enzyme system. The preferred combination of substituents were a methoxy group on the benzimidazole ring and a single alkyl group on the aniline ring. One such compound, 76, was an effective inhibitor of acid secretion in the dog and was selected for further pharmacological study.

No abstract

Diisopyramide phosphate has been recognized as a new antiarrhythmic agent.1 Our screening program for antiarrhythmic activity has generated a large amount of biological data on compounds having structural variations. We consider the data appropriate for analysis of the structureactivity relationship employing the quantum statistical concepts.*J In doing this, we gain insight into the antiarrhythmic action, in the hope of predicting the drug potency before a potential antiarrhythmic agent is synthesized.The compounds shown in Table I have the structural variations near the carbonyl group and show a wide vari-ance in the ir absorption peak of the carbonyl group. The wave numbers ranging from as low as 1622 cm-' (compound 10) up to as high as 1735 cm-' (compound 12) represent the absorption of the fundamental vibrational transitions ( I = 0 -1 = 1, and 1 is the vibrational quantum number). This difference in wave numbers measures a contribution to the variation in the drug potency.The partition coefficients of the compounds shown in Table I were experimentally determined in an octanol-pH 7.4 phosphate buffer system* and found to be substantially influenced by the chemical structures. However, the partition coefficients alone did not correlate significantly with the drug potency. At this point, the lipophilicity and C=O absorption frequency are only useful correlative variables, and this study does not imply that these variables are the only variables of importance in drug-transfer and drug-receptor interaction.Theoretical Method. The ventricular arrhythmia induced in the unanesthetized dog by ligation of the anterior descending coronary artery was reduced by intravenous treatment with the compounds in Table I. If antiarrhythmic potency (AP) is defined as mean maximal reduction in extrasystoles per mean effective dose, eq 2 of part l2 must be slightly modified as mean maximal reduction in extrasystoles --where C is the molar concentration and A is the proportionality constant, which has the same unit as d(response)/ Cdt. Substituting eq 5 of part 12 and eq 18 of part 23 in eq 1, taking logarithms on both sides, and collecting constants in one term, we obtain Table I. Data for Observed and Calculated Antiarrhythmic Potency R Compd R 7? Ir, cm-1, Ln A P a YO-0