ABSTRACT: We substituted neutral amino acids for some positively charged residues (R42, K107, K146, R148 and K229) that line the active site of aldolase A in an effort to determine binding sites for inositol 1, 4,5-trisphosphate. In addition, D33 (involved in carbon-carbon bond cleavage) was mutated. K229A and D33S aldolases showed almost no catalytic activity, but Ins(1,4,5)P(3) binding was similar to that determined with the use of wild-type aldolase A. R42A, K107A, K146R and R148A had markedly decreased affinities for Ins(1,4,5)P(3) binding, increased EC(50) values for Fru(1,6)P(2)-evoked release of bound Ins(1,4,5)P(3) and increased K(i) values for Ins(1,4, 5)P(3)-evoked inhibition of aldolase activity. K146Q (positive charge removal) had essentially no catalytic activity and could not bind Ins(1,4,5)P(3). Computer-simulated docking of Ins(1,4,5)P(3) in the aldolase A structure was consistent with electrostatic binding of Ins(1,4,5)P(3) to K107, K146, R148, R42, R303 and backbone nitrogens, as has been reported for Fru(1,6)P(2) binding. Results indicate that Ins(1,4,5)P(3) binding occurs at the active site and is not dependent on having a catalytically active enzyme; they also suggest that there is competition between Ins(1,4,5)P(3) and Fru(1, 6)P(2) for binding. Although Ins(1,4,5)P(3) binding to aldolase involved electrostatic interactions, the aldolase A Ins(1,4, 5)P(3)-binding domain did not show other similarities to pleckstrin homology domains or phosphotyrosine-binding domains known to bind Ins(1,4,5)P(3) in other proteins.
