Biosynthesis of phosphatidylinositol.

Interest in our laboratory in the inositol lipids derives from the presence of relatively high amounts of phosphatidylinositol phosphate (PhIP) and phosphatidylinositol diphosphate (PhIPP) in excitable tissue. It was with this relationship in mind that we embarked on a study of biosynthetic pathways of inositol lipids several years ago. A first step in the elucidation of biosynthetic pathways of phospholipids goes back to 1953, when Kornberg and Pricerl showed that a liver microsomal fraction catalyzed the acylation of glycerol-3-phosphate (G-3-P, L-a-glycerophosphate) to diacyl-G-3-P (phosphatidate) , a lipid that was not previously known to be a biosynthetic intermediate substance. Subsequent studies indicated that phosphorylcholine was a precursor in the synthesis of phosphatidylch~line.~~~ Kennedy and Weiss, in 1955, found that ATP stimulated the incorporation of phosphorylcholine into lipid.4 An important step forward was taken when they learned that the stimulation by ATP was due to the presence of trace amounts of cytidine nucleotide.5 From this experiment evolved the role of CDP-choline and CDP-ethanolamine in phospholipid synthesis. A DPNH-linked dehydrogenase was found that reduced DHAP produced in glycolysis to G-3-P,6 which, in turn, could be acylated by the microsomal enzyme1 and fatty acyl-CoA to phosphatidate. Phosphatidate phosphohydrolasei was shown to cleave phosphatidate to P, and l,Zdiglyceride, which then reacted with CDP-choline or CDP-ethanolamine to yield phosphatidylcholine and phosphatidylethanolamine, respectively. A pathway for phospholipid synthesis was thus proposed, beginning with the reduction of DHAP to G-3-P, acylation to phosphatidate, dephosphorylation to diglyceride, and, finally, to yield phosphatidylcholine or phosphatidylethanolamine by reaction with the appropriate cytidine diphosphate derivative.s By analogy, we began to look for a similar system that would synthesize phosphatidylinositol (PhI). We looked in vain for an inositol kinase and for the presence of CDP-inositol, using inositol labeled with tritium by the Wilzbach technique or by reduction of scyllo-inosose.9 Although there was no evidence for inositol phosphate or CDP-inositol formation, we observed that cytidine nucleotides did significantly stimulate inositol incorporation into lipid in incubations with particulate fractions from kidney.1° When inositol was omitted, radioactivity from labeled CMP accumulated, as a function of time, into a