Phosphatidylinositol (PI) holds a potential of becoming an important dietary supplement due to its effects on lipid metabolism in animals and humans manifested as a decrease of the blood cholesterol and lipids, and relief of the metabolic syndrome. To establish an efficient, enzymatic system for PI production from phosphatidylcholine and myo-inositol as an alcohol acceptor, our previous study started with the wild-type Streptomyces antibioticus phospholipase D (SaPLD) as a template for generation of PI-synthesizing variants by saturation mutagenesis targeting positions involved in acceptor accommodation, W187, Y191, and Y385. The isolated variants generated PI as a mixture of positional isomers, among which only 1-PI exists in nature. Thus, the current study has focused to improve positional specificity of W187N/Y191Y/Y385R SaPLD (NYR) which generates PI as a mixture of 1-PI and 3-PI in the ratio of 76/24, by subjecting four residues of its acceptor-binding site to saturation mutagenesis. Subsequent screening pointed at NYR-186T and NYR-186L as the most improved variants producing PI with a ratio of 1-/3-PI=93/7 and 87/13, respectively, at 37 degrees C. Lowering the reaction temperature further improved the specificity of both variants to 1-/3-PI>97/3 at 20 degrees C with no change in total PI yield. Structure model analyses imply that G186T and G186L mutations increased rigidity of the acceptor-binding site, thus limiting the possible orientations of myo-inositol. The two newly isolated PLDs are promising for future application in large-scale 1-PI production. Biotechnol. Bioeng. 2016;113: 62-71. (c) 2015 Wiley Periodicals, Inc.