Multiple pathways involved in the biosynthesis of anandamide

Abstract

Endocannabinoids, including anandamide (arachidonoyl ethanolamide) have been implicated in the regulation of a growing number of physiological and pathological processes. Anandamide can be generated from its membrane phospholipid precursor N-arachidonoyl phosphatidylethanolamine (NAPE) through hydrolysis by a phospholipase D (NAPE-PLD). Recent evidence indicates, however, the existence of two additional, parallel pathways. One involves the sequential deacylation of NAPE by α,β-hydrolase 4 (Abhd4) and the subsequent cleavage of glycerophosphate to yield anandamide, and the other one proceeds through phospholipase C-mediated hydrolysis of NAPE to yield phosphoanandamide, which is then dephosphorylated by phosphatases, including the tyrosine phosphatase PTPN22 and the inositol 5′ phosphatase SHIP1. Conversion of synthetic NAPE to AEA by brain homogenates from wild-type and NAPE-PLD^−/− mice can proceed through both the PLC/phosphatase and Abdh4 pathways, with the former being dominant at shorter (<10 min) and the latter at longer (60 min) incubations. In macrophages, the endotoxin-induced synthesis of anandamide proceeds uniquely through the phospholipase C/phosphatase pathway.

Introduction

N-Acylethanolamines were first identified as endogenous lipids in the hypoxic myocardium (Epps et al., 1979) where their biosynthesis was found to occur in two steps. First, a transacylase N-acylates membrane phosphatidyl ethanolamine (PE) (Natarajan et al., 1981), which is then hydrolyzed by a phospholipase D (PLD) to yield N-acylethanolamines and phosphatidic acid (Schmid et al., 1983). A low abundance member of this class of lipids, N-arachidonoyl ethanolamide (AEA, also called anandamide) was later identified as the first endogenous cannabinoid ligand (Devane et al., 1992). Since then, endocannabinoids (ECs) acting at specific cannabinoid receptors in the brain and periphery have been implicated in the regulation of a growing number of physiological functions and pathological processes (Pacher et al., 2006). Compounds that selectively inhibit EC actions at cannabinoid receptors or enhance them by blocking EC degradation show great therapeutic potential (Pacher et al., 2006). Inhibitors of enzymes that metabolize ECs were found to potentiate some but not other EC actions (Kathuria et al., 2003), which suggests that blocking EC biosynthesis may have a pharmacological profile different from that achieved by blocking cannabinoid receptors, and thus may be used therapeutically to avoid some of the unwanted side effects of the latter.

Anandamide is generated from the membrane phospholipid precursor N-arachidonoyl PE (NAPE) through a two-step process similar to that described above (Di Marzo et al., 1994). The calcium-dependent transacylase that catalyzes the transfer of arachidonic acid (AA) from the sn-1 position of phosphatidyl choline to the nitrogen atom of PE to generate NAPE has not yet been identified. However, a NAPE-specific PLD that hydrolyzes NAPE to yield anandamide has been cloned (Okamoto et al., 2004) and the purified protein characterized (Wang et al., 2006). When overexpressed in cells, NAPE-PLD selectively reduces NAPE and increases anandamide levels (Okamoto et al., 2005). Unexpectedly, no significant reduction in anandamide levels was found in the brain of NAPE-PLD knockout mice (Leung et al., 2006), or in macrophages following siRNA knockdown of NAPE-PLD (Liu et al., 2006). Furthermore, tissues from NAPE-PLD^−/− mice contained an enzymatic activity capable of converting NAPE to AEA in a calcium-independent manner (Leung et al., 2006), suggesting the existence of additional, parallel biosynthetic pathways. There has been evidence since the 1980s for enzymatic activities that can generate N-acylethanolamines from NAPE through sequential deacylations yielding lyso-NAPE and glycerophospho-ethanolamines, respectively (Natarajan et al., 1984). Accordingly, a recently identified group IB secretory phospholipase A₂ can convert NAPE to 2-lyso-NAPE, which is then metabolized to anandamide through a calcium-independent mechanism sensitive to inhibition by methyl arachidonoyl fluorophosphate (MAFP) (Sun et al., 2004). However, the restricted tissue expression of this PLA₂ suggested the existence of additional enzymes involved in generating lyso-NAPE. One such enzyme is the recently identified α,β-hydrolase 4 (Abhd4), which can act on either NAPE or lyso-NAPE to generate the glycerophospho-arachidonoyl ethanolamide (GpAEA), which is then acted on by a metal-dependent, phosphodiesterase to yield AEA (Simon and Cravatt, 2006). Another alternative pathway identified in the RAW264.7 mouse macrophage cell line involves the hydrolysis of NAPE by a phospholipase C to yield phosphoanandamide (pAEA) which is then dephosphorylated by phosphatases, including the putative tyrosine phosphatase PTPN22 (Liu et al., 2006). This latter pathway was found to be responsible for the endotoxin (LPS)-induced increase in AEA biosynthesis in macrophages (Liu et al., 2003, Liu et al., 2006). In the present study we have analyzed the relative importance of the NAPE-PLD, Abhd4 and PLC/phosphatase pathways in AEA biosynthesis in RAW264.7 cells and in mouse brain.