Helicobacter pylori urease belongs to a family of highly conserved urea-hydrolyzing enzymes. A common feature of these enzymes is the presence of two Lewis acid nickel ions and a reactive cysteine residue in the active site. The H+/K(+)-ATPase inhibitor omeprazole is a prodrug of a sulfenamide which covalently modifies cysteine residues on the luminal side of the H+/K(+)-ATPase of gastric parietal cells. Omeprazole and eight analogues were selected based on their chemical, electronic, and kinetic properties, and each was incubated with viable H. pylori in phosphate-buffered saline at pH 7.4 for 30 min, after which 100 mM urea was added and the amount of ammonia formed analyzed after a further 10 min. Inhibition between 0% and 100% at a 0.1 mM concentration was observed for the different analogues and could be expressed as a function of the pKa-value of the pyridine, the pKa-value of the benzimidazole, the overall lipophilicity, and, most importantly, the rate of sulfenamide formation, in a quantitative structure-activity relationship. The inhibition was potentiated by a lower pH (favoring the formation of the sulfenamide) but abolished in the presence of beta-mercaptoethanol (a scavenger of the sulfenamide). Structural analogues incapable of yielding the sulfenamide did not inhibit ammonia production. Treatment of Helicobacter felis-infected mice with 230 mumol/kg flurofamide b.i.d. for 4 weeks, known to potently inhibit urease activity in vivo, as a means of eradicating the infection, was tested and compared with the effect of 125 mumol/kg omeprazole b.i.d. for 4 weeks. Neither treatment proved efficacious.

A relationship between the structure of 21 2-[[(2-pyridyl)methyl]thio]-1H-benzimidazoles (6) and their anti Helicobacter pylori activity expressed as minimum bactericidal concentration (MBC) values is described. Observed MBCs ranged from 256 to 1 microg/mL. The structure-activity relationship (SAR) showed that larger and more lipophilic compounds, especially compounds with such substituents in the 4-position of the pyridyl moiety, generally had lower MBC values. Four new compounds that were predicted to be potent by the established SAR model were synthesized and tested. One such compound, i.e., 2-[[(4-[(cyclopropylmethyl)oxy]-3-methyl-2-pyridyl)methyl]thio]-1H-be nzimidazole (18), was tested for in vivo efficacy in a mouse Helicobacter felismodel (125 micromol/kg bid given orally for 4 days, n = 4). Unfortunately, antibacterial activity could not be clearly demonstrated in this model. Instead a potent acid secretion inhibition was observed. This finding was attributed to the methylthio compound being oxidized to the corresponding methyl sulfinyl derivative, i.e., a proton pump inhibitor, in vivo. Although the antibacterial activity had the potential of decreasing H. felis cell counts in vivo the proton pump inhibitory effect became dominant and actually promoted H. felis cell growth. Hence, we conclude that the antibacterial utility of the 2-[[(2-pyridyl)methyl]thio]-1H-benzimidazoles (6) as a compound class is compromised by their propensity to become proton pump inhibitors upon metabolic oxidation in vivo.