2-Arachidonoylglycerol (2-AG) is a monoacylglycerol (MAG) molecule containing an esterified arachidonic acid chain at the sn-2 position of the glycerol backbone (Figure 1). Together with structurally similar N-arachidonoylethanolamine (anandamide), 2-AG has been extensively studied as an endogenous ligand of cannabinoid receptors (CB1 and CB2) in the brain and other mammalian tissues and plays several physiological functions in both central and peripheral organs [1]. During either acute or chronic pathological conditions, the tissue and plasma levels of 2-AG, as well as of the other endocannabinoids (EC) or N-acylethanolamides (NAEs), are altered. The modulation of circulating or tissue EC and NAEs levels produces changes in cannabinoid receptor activation that acts to re-establish the homeostasis of other mediators, such as γ-aminobutyric acid and glutamate, and to restore the cell to its original steady-state condition [2]. Due to the prominent role of these signaling molecules, the literature reports numerous biochemical studies on the EC and NAEs levels associated with pathological conditions [3,4,5,6,7] or on the modification of the endocannabinoid system activity in cellular [8,9,10] or animal [11,12] pathological models. Moreover, the literature reports different LC or GC-MS methods that quantify EC and NAEs to study and understand their role in pathological processes and homeostasis restoration [3,7,13]. The importance of having pure and stable analytical standards for EC and NAEs is clear.

Chemical synthesis of 2-AG would enable ready access to this ligand for extensive study to elucidate the functions of this major neurochemical signaling molecule in biological and biochemical studies. However, the synthesis, isolation, storage, and testing of 2-AG are all compromised by the instability of the compound to both air oxidation, affording olefin rearranged hydro-peroxides, and ester rearrangement, affording 1-arachidonoyl glycerol (1-AG). Indeed, the four double bonds present in the arachidonoyl chain render 2-AG susceptible to oxidation processes prompted by drastic reaction conditions. Moreover, the main issue in the synthesis of pure 2-AG is the rapid migration of the arachidonoyl group from the secondary to the primary hydroxyl group of the glycerol, resulting in the formation of more stable 1-AG. The 1-MAG isomeric conformation is so thermodynamically stable that, at equilibrium, the acyl migration reaches a 1:9 2-MAG:1-MAG ratio [14,15]. This rearrangement to 1-AG, or 1-acyl glycerol derivatives in general, occurs under different reaction conditions, including temperature, solvent polarity, water activity, fatty acid chain length, acid/base impurities, time, and even in storage conditions at room temperature in various biologically compatible solvents. These variables must be considered during purification, analysis, and subsequent reactions using 2-MAG as an intermediate. Furthermore, the separation of these isomers is ineffective on a preparative scale, leaving mixtures that could compromise the interpretation of biological study results. Commercially available analytical standards indeed are often sold as organic solvent solutions, principally acetonitrile solutions, with a purity of up to 98% and 9:1 ratio 2-AG:1-AG. Very low amounts of pure 2-AG are also commercially available, but even in this case the presence of 1-AG is around 10% and the stability of the standard is guarantee only for 2–3 months.

To date, 2-AG is synthesized by either chemical [16,17,18,19] or chemoenzymatic methods [20,21,22,23]. The chemical methods described for the synthesis of 2-AG are based on the same procedure: acylation of suitable 1,3-protected glycerol, or glycidol, precursors with an activated arachidonic acid derivative, followed by deprotection and separation of the isomeric arachidonoylglycerols. Han and co-workers [24] use the triisopropylsilyl protecting group (TIPS) for the 1- and 3-hydroxyl functions of glycerol that can be removed by prolonged treatment with a solution of tetra-n-butylammoniofluoride (TBAF), buffered with acetic acid, at low temperatures. Prolonged treatment required for protecting group cleavage leads to 2 to 1 partial AG isomerization and increases the chance of arachidonic chain auto-oxidation. In the two other approaches 1,3-benzylidene-glycerol is used as substrate and, after the introduction of the arachidonoyl part, the acid-labile benzylidene group of the 1,3-benzylidene-2-acylglycerols is separated by reaction with boric acid; the resulting mixture of the 1,3- and 1,2-phenylboronate esters are hydrolyzed by washing with water to produce 2-monoglycerides [18,25]. However, the protecting groups’ cleavage with boric acid proved difficult to complete and undergoes increasing isomerization to 1-AG after the longer reaction time required. The use of phenylboronic acid instead of boric acid for the cleavage of benzylidene acetal gave better results, leading to a cleavage and to the formation of 13% of 1-arachidonoylglycerol 2,3-phenylboronate ester [26], that being the least subject to rearrangement in the final deblocking phase (<1% as determined by HPLC or ¹H NMR analysis) leading to a total yield of 65%. Another synthesis of 2-AG used, as starting material, glycidyl arachidonate obtained from glycidol and commercially available arachidonic acid [16]. The transformation of glycidyl arachidonate into the 2-AG derivative is promoted by trifluoroacetic anhydride. 1,3-Dibenzyloxy-2-propanol was chosen as 1,3 protected glycerol, because it is possible to remove the protecting groups in a clean and quantitative manner using B-chlorocatecholborane as a selective reagent [17].

B-Chlorocatecholborane was also used for acetal cleavage of cis-arachidonoylbenzylidene glycerol [19] that demonstrated the mildness of the reagent on an especially sensitive transformation while also affording an improved quality of product of current interest. Nevertheless, the silyl [24] and benzyl [17,19] protected glycerol approaches experienced at least 5% rearrangement, demonstrating the sensitivity of this molecule to deblocking conditions and identifying it as a probe of the mildness of reaction conditions.

The enzymatic methods described for the synthesis of 2-AG are based on a two-step enzymatic process [20,21,22,23]. 1,3-Dibutyrylglycerol is synthesized by selective esterification of glycerol, using vinyl butyrate in the presence of immobilized Candida antarctica lipase B and anhydrous dichloromethane as solvent. The resulting butyryl groups act to protect these sites for the subsequent arachidonate ester placement at the free sn-2 position to form 2-arachidonoyl-1,3 dibutyrylglycerol. Next, the butyryl groups are selectively removed in a lipase-mediated transesterification reaction to yield 2-AG. Even if the use of immobilized Candida antarctica lipase B allows milder reaction conditions, these methods are time consuming, lead to byproduct formation, require the following purification step after every reaction step, and show isomerization of around 10%. Using immobilized enzymes as biocatalysts is a well-known technique that allows enzyme reuse and often improves enzyme stability, biocompatibility, and resistance to environmental changes [27,28]. Moreover, they are more easily removed from the reaction mixture leading to handy synthetic procedures. However, the various enzyme immobilization methods, such as physical adsorption, entrapment or encapsulation, chemical cross-linking, and covalent binding, can differently affect enzyme stability, chemical or conformational structures or enzyme active sites surrounding solvent, and enzyme selectivity and specificity. [29]. Furthermore, interfacial activation of the lipase, the fixation of the lipase in its open (and active) form, is also modulated by the different immobilization methods and supports. Thus, having a “universal” optimal biocatalyst is far from straightforward. Nevertheless, having different immobilized biocatalysts with the same enzyme allows forproducing a large library of biocatalysts from a specific lipase [30].

Searching for an alternative and efficient methodology that would bypass these problems during 2-AG synthesis, we have developed an enzyme-catalyzed efficient and highly regioselective synthesis that enables an easy release of 2-AG without rearrangement to 1-AG from a precursor that exhibits resistance to oxidation and ester rearrangement during storage.