Toward an anandamide transporter

Almost every step in the elucidation of the structure and mechanism of action of the endocannabinoid system has been plagued by controversy. Although the psychoactive component of marijuana was identified in the mid-1960s (1), it took another two decades to show that it does not act by an unspecific mechanism, but via receptors, now known as the endocannabinoid receptors CB₁ and CB₂ (2). The identification of anandamide (AEA) in 1992 (3) as an endogenous cannabinoid ligand was originally challenged, because its level in mouse brain was found to be negligible. Several thousand publications later, it is now a well established transmitter. For more than a decade, a heated controversy, on paper and at meetings, has been going on over the transport of anandamide into the cell, a central process in its metabolism (Fig. 1). Is it a facilitated transport with the help of a transporter, presumably a protein, or does anandamide passively diffuse through the plasma membrane in a process that is not protein-mediated (4)? In this issue of PNAS, Moore et al. (5) report the identification of a high-affinity, saturable, anandamide binding site that is distinct from fatty acid amide hydrolase (FAAH), the enzyme that hydrolyzes anandamide (5). The authors bring forward evidence that this high-affinity binding site is, in fact, an anandamide transporter. The possibility that alternative transport mechanisms, be they diffusional, endocytic, or FAAH-mediated, are working in parallel to the facilitated transport has, however, not yet been excluded.

Until now, the biosynthesis and breakdown of anandamide were well understood, but its mechanism of uptake leading to termination of signaling was still ambiguous. In 1993, it was shown that anandamide was readily taken up by cells in culture and broken down into arachidonic acid and ethanolamine by an amidase, the FAAH (6). By breaking down anandamide, FAAH maintains an inward gradient that drives anandamide influx. Anandamide is unique among neurotransmitters/modulators in that its uptake is primarily under the control of a single enzyme (7). In addition, there appears to be an intracellular “sink,” composed, perhaps, of membranes and/or proteins that accumulate anandamide in cells in culture.

There is general agreement that anandamide uptake occurs by an ATP-independent process. However, the precise uptake mechanism at the plasma membrane remains unknown. More than 100 papers have been published predicated upon the supposition that a protein carrier mediates anandamide cellular uptake. More recently, a handful of papers have proposed that the uptake of anandamide occurs by simple diffusion. It has also been suggested that endocytosis via a caveola/lipid raft mediates anandamide transport (for reviews, see refs. 4, 8, and 9).

Moore and colleagues from Eli Lilly Laboratories (5) now report that LY2318912, a tetrazole-based urea derivative, is a potent inhibitor of anandamide uptake. It was used to identify a high-affinity, saturable, anandamide transporter binding site that is distinct from FAAH. LY2318912 differs in structure from the known inhibitors of FAAH, the most potent of which are aryl carbamates, which acylate a catalytic serine site (10), ketoheterocyclics that act by forming reversible hemiketals also with an active serine site (11) and, more recently, ketooxazolopyridines, again acting on serine hydrolases (12). Moore et al. (5) found that iodine-labeled LY2318912 high-affinity binding sites are present in RBL-2H3, wild-type HeLa (FAAH^–/–), and FAAH-transfected P2 membranes. The binding was consistent with a plasma membrane protein-binding site, being temperature-sensitive, rapid, and fully displaceable by anandamide. Although experiments to determine whether LY2318912 inhibits FAAH in FAAH-transfected HeLa or RBL cells are not presented, apparently it does not act by inhibition of this enzyme. Neither the binding affinity (K _d) nor the B _max changes significantly between FAAH-negative and FAAH-expressing cell types, and the presence of binding in FAAH^–/– cells points in the same direction. Further experiments show that LY2318912 not only competitively inhibits the transporter's binding site but also is not a substrate of the transporter. The authors argue that “If the described binding site was an anandamide transporter, then administration of selective binding site inhibitors would elevate endocannabinoid levels and generate a relevant physiological and behavioral response.” Indeed, administration of the labeled derivative led to a dose-dependent elevation of anandamide concentrations in rat cerebellum. They also tested the effect of LY2183240 in a pharmacologically relevant assay. They found that, dose-dependently, it attenuated formalin-induced paw-licking pain behavior in a model of persistent pain. These findings are consistent with the assumption that the levels of anandamide are elevated. These advances are exciting. We should look forward to the cloning and expression of the transport protein.

This provocative study should also be a catalyst for further experimentation in vitro on anandamide transport because it raises many interesting questions. For example, are these compounds effective at time points of <1 min? Shorter times isolate effects at the plasma membrane from those downstream membranes in the cell. Are these compounds FAAH inhibitors in the RBL-2H3 cell? Potent FAAH inhibitors should yield the same anandamide uptake profile with the very long incubations used in this study. If FAAH is inhibited with a highly selective and potent reagent (13, 14), what is the profile of anandamide uptake in the presence of the Lilly compounds? Do these transport inhibitors bind the CB₂ receptor in the immune-derived RBL-2H3 cells?

As mentioned above, Moore et al. (5) show that administration of LY2183240 to the rat raises anandamide levels in the brain and is effective in the treatment of pain without affecting motor function, suggesting that inhibitors of anandamide transport potentially will be very useful therapeutically. These findings are exciting because it has not been possible to separate nociceptive from motor and psychotropic effects with CB₁ agonists, and they raise additional topics for further experiments. Are the effects of LY2183240 in the rat additive when coadministered with FAAH inhibitors? What are the effects of the Lilly compound in wild-type and FAAH null mice? In other words, is it possible that an “off-target” reaction for the Lilly compound is indeed FAAH? Relief of pain without side effects of hypomotility, hypothermia, and catalepsy by activation of the CB₂ receptor has recently been shown in FAAH knockout mice (either genetically or chemically) (15). Is the Lilly compound acting via CB₂ in the animal studies? It is well known now that CB₂ agonists may also induce analgesia without the unwanted side effects of CB₁ activation (16).

Drugs are expected to be “magic bullets,” which act on a specific syndrome through a single, specific site, be it a receptor, an enzyme, or some other biological entity. Unfortunately, drugs very seldom are specific. This lack of specificity is of course true for exogenously administered cannabinoids. Tetrahydrocannabinol (THC), the psychoactive marijuana constituent, causes a myriad of effects, many of which represent undesirable side effects in a therapeutic situation. Because THC binds to both CB₁ and CB₂ cannabinoid receptors, which are involved in a very large number of physiological processes, this absence of selectivity is not surprising. Cannabinoids, be they endogenous, plant, or synthetic, have been found to affect almost every physiological system investigated, the nervous, the cardiovascular, the reproductive, and the respiratory systems, to name a few. And these cannabinoids have been found to be promising drugs in a long list of medical conditions. (For a recent review of cannabinoids in medicine, see ref. 17.) However, administration of cannabinoids for a very specific condition may affect too many physiological systems. But, as indicated above, there seems to be a way out of this problem. Anandamide is present in extremely low basal concentrations. Apparently it is formed mostly when the need arises in specific tissues. It activates the cannabinoid receptors near its site of synthesis and is then inactivated by transport into the cell and hydrolysed there. FAAH and transporter inhibitors, alone or together, enhance the level of anandamide around the specific site needed, without necessarily causing side effects. Anandamide may, of course, move to other sites by a variety of mechanisms, but such a process will presumably be negligible. This line of thought has led a large number of investigators, in academia and in the pharmaceutical companies, to look for novel FAAH and transporter inhibitors. The identification of a high-affinity binding site involved in the transport of anandamide by the Eli Lilly group represents a major step in understanding endocannabinoid metabolism and will present drug researchers with a very valuable tool.

Cannabinoids have been found to affect almost every physiological system investigated.