ABSTRACT
Tridecaptin A1 is a linear antimicrobial lipopeptide comprised of 13 amino acids, including three diaminobutyric acid (Dab) residues. It displays potent activity against Gram-negative bacteria, including multidrug-resistant strains. Using solid-phase peptide synthesis, we performed an alanine scan of a fully active analogue, octyl-tridecaptin A1 , to determine key residues responsible for activity. The synthetic analogues were tested against ten organisms, both Gram-positive and Gram-negative bacteria. Modification of D-Dab8 abolished activity, and marked decreases were observed with substitution of D-allo-Ile12 and D-Trp5. Circular dichroism showed that octyl-tridecaptin A1 adopts a secondary structure in the presence of model phospholipid membranes, which was weakened by D-Dab8-D-Ala, D-allo-Ile12-D-Ala, and D-Trp5-D-Ala substitutions. The antimicrobial activity of the analogues is directly correlated to their ability to adopt a stable secondary structure in a membrane environment.
Subject(s)
Anti-Bacterial Agents/pharmacology , Bacteria/drug effects , Lipopeptides/pharmacology , Phospholipids/chemistry , Anti-Bacterial Agents/chemical synthesis , Anti-Bacterial Agents/chemistry , Dose-Response Relationship, Drug , Lipopeptides/chemical synthesis , Lipopeptides/chemistry , Microbial Sensitivity Tests , Molecular Conformation , Structure-Activity RelationshipABSTRACT
A facile two-step stereospecific synthesis of the protein farnesyltransferase inhibitor chaetomellic acid A (1) and its analogues was developed. Addition of organocuprates derived from Grignard reagents (e.g. tetradecylmagnesium chloride and CuBr.Me(2)S) to dimethyl acetylenedicarboxylate (DMAD) in tetrahydrofuran containing hexamethylphosphoramide was followed by capture of the resulting copper enolates with a variety of electrophiles (e.g. methyl iodide) to give dimethyl cis-butenedioate derivatives 4-11. Hydrolysis with lithium hydroxide generated the corresponding lithium carboxylates, which readily closed to 2,3-disubstituted maleic anhydrides 17-20 upon acid treatment. Compound 16, an analogue wherein the tetradecyl group of 1 is replaced by a farnesyl moiety, is 7-fold more potent than 1 as an inhibitor of protein farnesyltransferase from yeast and displays a 100:1 selectivity for this enzyme relative to yeast protein geranylgeranyltransferase. In contrast, analogue 15, which contains a geranylgeranyl side chain, shows ca. 10:1 selectivity for the latter enzyme.
