The Deorphanization of TRPV1 and the Emergence of Octadecadienoids as a New Class of Lipid Transmitters

It has been roughly a half-century since the Hungarian scientist Nicholas Jancsó postulated the existence of a “pain receptor” through which capsaicin, the active lipidic principal in hot chile peppers, exerts its “stimulatory and desensitizing” actions (1). Little more than a decade ago, Caterina et al. reported on the molecular identification of that pain receptor as TRPV1 (first termed vanilloid receptor 1 or VR1) (2), and ever since this discovery, a search has been underway to ascertain the nature of what has turned out to be an elusive class of endogenous TRPV1 agonists. Although much work has focused on various eicosanoids as possible “endovanilloids” (Table 1), Patwardhan and colleagues have now added new and highly intriguing molecules as candidate ligands. In two recently published papers, they show that oxidized metabolites of linoleic acid (OLAMs) satisfy a number of critical criteria to be classified not merely as endogenous activators of this hotly pursued and heat-responsive channel but as peripheral neurochemical conduits of heat itself (3, 4). Thus, these scientists have at once potentially deorphanized TRPV1, both in the CNS and in the periphery, and heralded the emergence of “octadecadienoids,” metabolic products of linoleic acid, as a new class of algogenic (i.e., pain causing) substances. These findings raise critical questions regarding the precise role(s) of TRPV1 in nociceptive transduction and about the importance of fatty acids in the modulation of pain and related functions. Moreover, they point to new and fascinating lines of further inquiry, particularly with regard to the potential development of novel therapeutic strategies to treat pain. In the present Viewpoint, we address several of the more salient aspects of the present works as well as speculate on their implications [see also (5, 6)].

In the first article, Patwardhan et al. show that depolarization of neurons in rat spinal cord slices with high extracellular potassium leads to the release of at least one OLAM, 9-hydroxyocta-decadienoic acid (9-HODE), and that perfusates from these spinal cord slices are capable of activating native or recombinant TRPV1 (3). The exogenous administration of 9-HODE or of other oxidative metabolites of linoleic acid—such as 13-HODE, 9-oxoODE, or 13-oxoODE—also activated TRPV1, albeit less potently, suggesting the existence of a novel lipid family of endovanilloid substances. Most intriguingly perhaps are the demonstrations that the intrathecal application of exogenous OLAMs produced mechanical allodynia and that immunoneutralization of 9- and 13-HODE attenuated the production of mechanical allodynia caused by the intraplantar injection of complete Freund’s adjuvant (CFA). At the outset, one must be impressed by the simple, yet elegant “active fraction” approach taken by the authors, reminiscent of the early days of Guillemin and Schally and their contentious search for brain hormones.

The blockade of increased intracellular calcium in trigeminal ganglion neurons to depolarized spinal cord eluates by AMG9810, a purported competitive antagonist of TRPV1, suggests that the agonist effect of the OLAMs at TRPV1 may be orthosteric (i.e., with respect to the site that binds the canonical agonist ligand capsaicin). However, subsequent experimentation has cast doubt on such a conclusion, as discussed below. Insofar as perfusates from spinal cord neuronal slices exposed to a non-specific depolarizing stimulus (i.e., high potassium concentration) were able to ectopically activate TRPV1, it will be important to clearly define the in situ conditions under which such processes may occur and the neurochemical mediators thereof. That an intraplantar injection of CFA appeared able to produce the elaboration of spinal OLAMs does indeed suggest that there is a natural neurophysiologic correlate to the reported ex vivo observations, although relatively high concentrations of exogenous OLAMs were employed in these studies. Thus, it will be important to determine the concentration-response relationships potentially linking OLAM activation of TRPV1 to mechanical allodynia. Additionally, the difficult question of whether sufficient concentrations of OLAMs may be produced in the spinal cord to activate TRPV1 under physiologic conditions (e.g., after depolarization or treatment with CFA) must be addressed.

The present data, however, strongly support the notion that the production of these 18-carbon lipid metabolites occurs at a critical site of TRPV1 function, the dorsal spinal cord, whereas their production in the periphery, the other major site of TRPV1 activation, is addressed by these investigators in their follow-up publication (4). With respect to the creative immunoneutralization approach, it would be important to know the effect of either antibody (i.e., 9-HODE-specific or 13-HODE-specific) alone or why they must be given in combination, notwithstanding the claims of redundancy. Future studies could be aimed at identifying which endogenous spinal mediators are released by OLAM-stimulated TRPV1 and which second-order neurons and other downstream events are activated as a consequence. Finally, it will be important to determine the specific enzymatic and metabolic pathways (e.g., lipoxygenases, free radicals, etc.) that lead to the formation of these metabolites under physiological and pathophysiological conditions, as this will likely inform therapeutic intervention strategies.

In the second article, this same group turns its attention to the periphery and to the activation of thermo-sensitive afferents by heat (4). Specifically, they show that the same OLAMs described above are generated in mouse or rat skin biopsies exposed to noxious heat. Again, pharmacologic approaches were used to imply that these thermally generated OLAMs are algogenic (in wild type but not TRPV1-knockout mice) via TRPV1 activation, either by blocking these effects with nordihydroguaiaretic acid (NDGA), which inhibits the formation of OLAMs by lipoxygenase or free radicals, or by the TRPV1 antagonist iodo-resiniferatoxin (I-RTX). Of note is the observation that NDGA blocked nearly 50% of the heat-evoked current through TRPV1 and blocked only 30% of calcitonin gene-related peptide (CGRP) release in rat trigeminal ganglion neurons, whereas I-RTX blocked 100% of CGRP release (4). These findings may suggest that there are non-OLAM mediators of TRPV1 activation liberated by heat, that heat itself is at least in part a directly acting “agonist” or, possibly, that OLAMs are already pre-formed and in some sort of heat-releasable pool. It is also reasonable to consider what the effect of heat may be on the other arm of the linoleic acid metabolic pathway leading to the production of γ-linolenic acid—that is, does heat have any effect on 6-desaturase? Of course, the precise mechanisms by which heat should “selectively” activate lipoxygenases or increase reactive species would likely also be an important area for further study. This second series of investigations did, however, provide more substantive information about the range of OLAM concentrations required to activate TRPV1, which may be useful in future interpretations of studies in which the endogenous levels of these mediators are quantified. The temperature-response relationship for their production was described for both 9-HODE and 13-HODE, which matched well with the temperature range known to activate TRPV1.

Although the data cannot help but convince researchers that metabolic oxidation products of linoleic acid are strong candidates as endogenous activators of TRPV1, an issue that remains to be addressed is the time course for the production of OLAMs in response to heat, particularly in the context of primary afferent activation. The authors are credited for having at least attempted to demonstrate that the in vitro kinetics of TRPV1 activation to heat and to OLAMs are similar. That said, this demonstration does not persuade one that the latency from the time of application of noxious heat to the living tissue to the time at which the nociceptive behavior becomes manifest can be temporally accounted for, given the intervening requirements for the generation and liberation of endogenous OLAMs, their activation of neuronal TRPV1, and the propagation of that signal centrally. Indeed, the studies utilizing skin biopsy superfusates employed a twenty-minute collection period. On the other hand, is it possible that the latency imposed by this metabolic conversion and indirect activation may underlie, at least in part, the phenomenon of secondary pain, the delayed response to noxious stimulation thought to be carried by the slower, unmyelinated C-fibers (7).

As mentioned above, an existing controversy with regard to TRPV1 physiology and the therapeutic potential of TRPV1 antagonists is whether and under what circumstances (e.g., animal pain models and human pain conditions) TRPV1 is capable of mediating mechanical hypersensitivity. The multifarious results of experiments designed to answer these questions directly using a variety of chemotypically distinct TRPV1 antagonists raises new questions in light of the present findings. Perhaps the most interesting puzzle revealed by the second study relates to the biophysical and pharmacologic nature of OLAM agonism at TRPV1. Whereas the first article demonstrated antagonism by AMG9810 (a competitive TRPV1 antagonist that also potentiates proton-induced activation of TRPV1), the second article suggests that the OLAMs may act allosterically, because their effects were not blocked by AMG8562, an antagonist that blocks the activation of TRPV1 by capsaicin but not by heat. Indeed, the canonical TRPV1 agonist capsaicin has been routinely if not universally used as an agonist of recombinant TRPV1 in high-throughput screens for antagonists by countless pharmaceutical companies, and often, though not always, these antagonists have been shown to block responses to heat and acid (H⁺) in secondary screens. If OLAMs turn out to be orthosteric agonists with respect to capsaicin, then it is reasonable that most of the reported antagonists should block OLAM-mediated activation as well. If OLAMs turn out not to be orthosteric agonists, then the highly variable reports in the literature regarding the ability of these antagonists to ameliorate mechanical hypersensitivity in various models may indeed relate to their variable ability to block endogenous TRPV1 agonists, such as the OLAMs. The intriguing possibility that there exists a differential interaction of OLAMs with TRPV1 in the CNS and the periphery will require further study.

Given the apparent ability of OLAMs to activate TRPV1 both spinally and in the skin, one may wonder whether this class of compounds activates TRPV1 in other central or peripheral locations, such as the brain or the gut. If so, under what conditions are they produced (i.e., is heat the only stimulus for their production) and what are their effects (e.g., are they thermoregulatory or anxiogenic)? The question also remains whether exposure to OLAMs results in a desensitization of TRPV1 or whether these compounds provide a feed-forward system to delay desensitization.

Based on these two studies (3, 4), OLAMs are added to the list of lipids that are thought to have a direct stimulatory action on TRPV1 [for review, see (8)]. These lipids include anandamide (9, 10), the lipoxygenase products, 12(S)- and 15(S)-hydroperoxyeicosatetraenoic acid (HPETE) (11), N-arachidonoyl-dopamine (12), and N-oleoyl-dopamine (13), and they have different relative potencies for activating TRPV1 (Table 1). Furthermore, some controversy remains whether they are actual endogenous TRPV1 agonists or whether they modulate excitability indirectly. Major issues for such classification of all these lipids remain: are they synthesized in response to noxious stimuli at appropriate sites? What is the time course of their synthesis? Are sufficient concentrations produced to activate TRPV1? All are key criteria for ultimately implicating any one as a true endovanilloid. Assuming that these lipids are endogenous activators, it also will be important to ascertain whether they work in concert with other modulators or independently to activate TRPV1.

The discovery that aspirin blocks cyclooxygenase garnered Sir John Vane the Nobel Prize in 1982 and energized a prolific spate of research from numerous laboratories, showing among other things that prostaglandins are profound sensitizers of nociceptive afferent neurons and producers of hyperalgesia. Since then, researchers in the pain field have focused their work on various metabolic products of arachidonic acid, the eicosanoids, as the source of a number of classes of endogenous compounds that modulate nociception, most recently including the endocannabinoids. Whereas it is well established that the eicosanoids can act via G protein–coupled receptors to produce their effects, more recently, it has been appreciated that eicosanoids and other endogenous lipids can act as direct activators of the cationic channel TRPV1. Together, the two studies from the Hargreaves group (3, 4) extend such an appreciation by adding OLAMs to this list of versatile lipid transmitter substances, thereby providing pivotal results that could have far reaching implications for scientists and clinicians studying mechanisms of acute and chronic pain. Additionally, their studies broaden the horizon on the potential role that these and other lipids play in modulating additional targets [e.g., OLAMs acting at the peroxisome proliferator-activated receptor γ (PPARγ)] and/or regulating neuronal function more generally. Although we know much about the metabolic products of archadonic acid, little is known of the other fatty acids that are ubiquitously found in membranes throughout the body or the enzymes that liberate them, such as fatty acid amide hydrolase and monoacylglycerol lipase, to name but a few. In a similar regard, one of the more promising avenues inspired by the present work will be an emboldened search for the identities of the presumably lipidic endogenous mediators of other TRP channel family members, especially the thermoTRPs, including TRPV3 and TRPM8. These and many other intriguing questions raised by these studies will generate much new, exciting research in the future.

• Copyright © 2010

Michael R. Vasko, PhD, is Paul Stark Professor of Pharmacology and Chair of The Department of Pharmacology and Toxicology at Indiana University School of Medicine. His major research area involves determining the cellular mechanisms mediating peripheral sensitization and the enhanced release of neurotransmitters from nociceptive sensory neurons that occurs during tissue trauma and inflammation. He serves on the Editorial Advisory Boards of Molecular Interventions and The Journal of Pain. Address correspondence to MRV. E-mail vaskom{at}iupui.edu; fax 317-274-1560.

Christopher M. Flores, PhD, obtained his doctoral degree in Pharmacology from Georgetown University and has devoted the majority of his career to elucidating the cellular and molecular mechanisms of pain and analgesia. He ran a federally funded academic laboratory as Associate Professor of Pharmacology and Endodontics at the University of Texas Health Science Center at San Antonio until 2002, and, more recently, led the search for novel analgesic drugs as Head of Pain Discovery for Johnson & Johnson Pharmaceutical Research and Development, LLC. Currently, Chris is Senior Director of External Innovation in the Neuroscience Therapeutic Area at Johnson & Johnson, where he is building new partnerships in the quest to solve diseases of the nervous system. Address correspondence to CMF. E-mail cflores2{at}its.jnj.com; fax 609-730-2069.