The inositol phosphate products formed during the cleavage of phosphatidylinositol by phosphatidylinositol-specific phospholipase C from Bacillus cereus were analyzed by 31P NMR. 31P NMR spectroscopy can distinguish between the inositol phosphate species and phosphatidylinositol. Chemical shift values (with reference to phosphoric acid) observed are 0.41, 3.62, 4.45, and 16.30 ppm for phosphatidylinositol, myo-inositol 1-monophosphate, myo-inositol 2-monophosphate, and myo-inositol 1,2-cyclic monophosphate, respectively. It is shown that under a variety of experimental conditions this phospholipase C cleaves phosphatidylinositol via an intramolecular phosphotransfer reaction producing diacylglycerol and D-myo-inositol 1,2-cyclic monophosphate. We also report the new and unexpected observation that the phosphatidylinositol-specific phospholipase C from B. cereus is able to hydrolyze the inositol cyclic phosphate to form D-myo-inositol 1-monophosphate. The enzyme, therefore, possesses phosphotransferase and cyclic phosphodiesterase activities. The second reaction requires thousandfold higher enzyme concentrations to be observed by 31P NMR. This reaction was shown to be regiospecific in that only the 1-phosphate was produced and stereospecific in that only D-myo-inositol 1,2-cyclic monophosphate was hydrolyzed. Inhibition with a monoclonal antibody specific for the B. cereus phospholipase C showed that the cyclic phosphodiesterase activity is intrinsic to the bacterial enzyme. We propose a two-step mechanism for the phosphatidyl-inositol-specific phospholipase C from B. cereus involving sequential phosphotransferase and cyclic phosphodiesterase activities. This mechanism bears a resemblance to the well-known two-step mechanism of pancreatic ribonuclease, RNase A.

Scheme 5 a a (a) TIPDSCl2 (2.5 equiv), pyridine, room temperature; (b) 35, CSA, DMF, 100 °C.

2,4-Di-O-benzoyl-myo-inositol 1,3,5-orthoformate, on heating in the presence of a base, undergoes transesterification to give 2,4,6-tri-O-benzoyl-myo-inositol 1,3,5-orthoformate and 2-O-benzoyl-myo-inositol 1,3,5-orthoformate in the solid state. The same reaction can also be performed by microwave irradiation instead of heating. The crystal structure of the dibenzoate reveals that the screw-axis-related molecules have the hydroxyl and the carbonyl groups ideally oriented for the reaction and gives a close picture of how such a reaction proceeds in enzymes. The structure of the corresponding acetate, (()-2-O-benzoyl-4-O-acetylmyo-inositol 1,3,5-orthoformate, lacks this geometry and hence is unreactive in the solid state. Both the acetate and the benzoate undergo base-catalyzed transesterification in solution.

Racemic 2,4-di-O-acyl-myo-inosityl 1,3,5-orthoesters undergo transesterification catalyzed by sodium carbonate with varying ease of reaction in the solid state; reactions in solution and melt do not show such varied differences. An interesting crystal of a 1:1 molecular complex of highly reactive racemic 2,4-di-O-benzoyl-myo-inosityl 1,3,5-orthoformate and its orthoacetate analogue exhibited better reactivity than the latter component alone. Single-crystal X-ray structures of the reactants have been correlated with the observed differences in the acyl-transfer efficiencies in the solid state. Although each of the derivatives helically self-assembles around the crystallographic 2(1) axis linked through O-H...O hydrogen bonding, the pre-organization of the reactive groups (C=O [El] and OH [Nu]), C-H...O and the C-H...pi interactions are significantly more favourable for the reactive derivatives than the less reactive ones. Bond-length distributions also showed differences; the O-C bond of the axial benzoyl group, which gets cleaved during the reaction, is longer (1.345-1.361 A) relative to the chemically equivalent O-C bond of the equatorial benzoyl group (1.316-1.344 A) in the reactive derivatives. These bond-length differences are not significant in the less reactive derivatives. The overall molecular organization is different too; the strikingly discrete helices, which may be viewed as "reaction tunnels" and are held by interhelical interactions, are clearly evident in the reactive derivatives in comparison with the less reactive ones.

No abstract