Structure-activity relationships of lysophosphatidic acid analogs

Abstract

The physiologic effects of lysophosphatidic acid (LPA) remain poorly understood. Our ignorance is due in part to lack of medicinal chemistry focussed on this pleiotropic lipid mediator. Beginning with commercially available phospholipids tested on whole cells or tissues and continuing with synthetic analogs tested at recombinant LPA receptors, the features of the LPA pharmacophore have become visible. An active LPA mimetic has a long aliphatic chain terminating in a phosphate monoester; bulky substitutions at the second carbon (relative to the phosphate) are tolerated poorly and a dissociable proton near the phosphate group seems required for optimal activity. These requirements are met by substituting ethanolamine for the glyceryl group in LPA. Substitutions at the second carbon of the N-acyl ethanolamide phosphoric acid (NAEPA) result in highly active agonists, including some receptor type selective compounds, if the substituent is small (e.g. methyl, methylene amino, methylene hydroxy). However, bulky hydrophobic substituents lead to compounds with decreased agonist, or even antagonist, activities. Examination of naturally occurring plant lipids led to the discovery of another LPA receptor antagonist, di-octyl glyceryl pyrophosphate. An unexplained result obtained in testing the LPA mimetics is the strong stereoselectivity exhibited by some responses (e.g. calcium mobilization) and the lack of stereoselectivity of other responses (e.g. platelet aggregation).