ION CHANNELS, RECEPTORS AND TRANSPORTERS

Umbellulone modulates TRP channels

Jian Zhong & Alberto Minassi & Jean Prenen &

Orazio Taglialatela-Scafati & Giovanni Appendino &

Bernd Nilius

Received: 10 October 2011 /Accepted: 11 October 2011 /Published online: 29 October 2011# Springer-Verlag 2011

Abstract Inhalation of umbellulone (UMB), the offensiveprinciple of the so-called “headache tree” (California baylaurel, Umbellularia californica Nutt.), causes a painfulcold sensation. We therefore studied the action of UMB andsome derivatives devoid of thiol-trapping properties on the“cold” transient receptor potential cation channels TRPA1and TRPM8. UMB activated TRPA1 in a dose-dependentmanner that was attenuated by cysteine-to-serine isostericmutation in TRPA1 (C622S), while channel block was

observed at higher concentration. However, although activa-tion by mustard oil was completely prevented in thesemutants, UMB still retained activating properties, indicatingthat it acts on TRPA1 only as a partial electrophilic agonist.UMB also activated TRPM8, but to a lower extent thanTRPA1. Removing Michael acceptor properties of UMB(reduction or nucleophilic trapping) was detrimental for theactivation of TRPA1, but increased the blocking potency. Thiswas, however, attenuated by acetylation of the hydroxylatedanalogs. All UMB derivatives, except the acetylated deriva-tives, were also TRPM8 activators. They acted, however, in abimodal manner, inhibiting the channel more potently thanUMB, and with tetrahydro-UMB being the most potentTRPM8 activator. In conclusion, UMB is a bimodal activatorof TRPA1 and a weak activator of TRPM8. Non-electrophilicderivatives of UMB are better TRPM8 activators than thenatural product and also potent blockers of this channel aswell as of TRPA1. The lack of effects of the acetylated UMBderivatives suggests that steric hindrance may prevent accessto the recognition site for the bicyclic monoterpene pharma-cophore on TRPA1 and TRPM8.

Keywords TRP channels . TRPA1 . TRPM8 . Headachetree . Umbellulone . Umbellulone derivatives .

Chemosensation . Electrophiles . Mustard oil . Menthol

Introduction

Umbellularia californica Nutt., a tree indigenous tosouthwestern Oregon and Northern California, has longbeen known to emit offensive vapors that can triggerviolent headache crises in sensitive persons [26]. Theseproperties, anecdotally documented in the scientific litera-ture since the second half of the nineteenth century [14],

Jian Zhong and Alberto Minassi contributed equally to this work.Giovanni Appendino and Bernd Nilius share senior last authorship.

Electronic supplementary material The online version of this article(doi:10.1007/s00424-011-1043-1) contains supplementary material,which is available to authorized users.

J. Zhong : J. Prenen : B. Nilius (*)Department of Molecular Cell Biology, Laboratory Ion ChannelResearch, Campus Gasthuisberg, Katholieke Universiteit Leuven,Herestraat 49,3000 Leuven, Belgiume-mail: bernd.nilius@med.kuleuven.be

J. ZhongCenter for Hypertension and Metabolic Diseases, Department ofHypertension and Endocrinology, Daping Hospital, Third MilitaryMedical University, Chongqing Institute of Hypertension,Chongqing 400042, China

A. Minassi :G. AppendinoDipartimento di Scienze Chimiche, Alimentari,Farmaceutiche e Farmacologiche,Via Bovio 9,28100 Novara, Italy

O. Taglialatela-ScafatiDipartimento di Chimica delle Sostanze Naturali,Università di Napoli Federico II,Via Montesano 49,80131 Naples, Italy

Pflugers Arch - Eur J Physiol (2011) 462:861–870DOI 10.1007/s00424-011-1043-1

have gained the plant the infamous nickname of “headachetree,” and have long remained unexplained. In the wake ofa recent report of medically documented cluster headache-like attacks triggered by the inhalation of U. californica [4–6], the monoterpene umbellulone (UMB, 1) was identifiedas the offensive principle from this plant [22]. Umbellulone,a compound readily absorbed by nasal route [22], activatesthe transient receptor potential (TRP) cation channelTRPA1 and, in animal experiments, can cause meningealvasodilation by liberation of calcitonin gene-related proteinfrom trigeminal nerve endings in the dura mater [22]. Theseeffects are abolished by TRPA1 antagonists, and are notobserved in TRPA1-nil animals, suggesting that activationof TRPA1 is critical for the migraine-inducing action ofumbellulone [22].

TRPA1 is activated by various electrophilic naturalproducts [24, 25] that covalently modify cysteine residuesof the channel [20]. While the classic activators of TRPA1form isolable adducts with thiol groups, those fromumbellulone could be observed in solution but not isolatedfrom the reaction medium, due to a rapid reversal of theaddition reaction during work up [3, 22]. This behavior isvery rare but not unprecedented within Michael acceptors,and compounds showing it are referred to as “reversible”thiol sinks [3]. Since thiol addition underlies the interactionof TRPA1 with its classic modulators, it was interesting toinvestigate in more detail the interaction of umbellulonewith this ion channel, evidencing a possible biologicaltranslation of the reversibility of its thiol-trapping trait. Acomplementary approach that combines point mutation of thelead structure and its target protein was used, and theinvestigation was extended also to TRPM8 due to thestructural similarity of umbellulone with menthol, the arche-typal ligand of this coolant receptor [7], and to the intenselycold sensation evoked by the inhalation of umbellulone.

Methods

Isolation of umbellulone Umbellulone was obtained from thedried leaves of U. californica as detailed previously [3, 22].

Cysteamine assay for Michael acceptor behavior Thesubstrate (0.1 mmol) was treated with cysteamine(0.20 mmol) in DMSO-d6 (4 mL/mmol), and the 1HNMR spectrum was registered 5 min after the addition(Spectrum A). An aliquot (25 μL) of the solution was nexttransferred into a NMR tube containing CDCl3 (0.5 mL),and a new spectrum was recorded (Spectrum B). A positiveassay was evidenced by specific changes in Spectrum Acompared to the spectrum of the pure product, and thereversibility by its reappearance upon dilution 1:20 withCDCl3 (Spectrum B).

Chemical modification of umbellulone Compounds 3a,b,4a, and 5 were prepared from umbellulone as described inthe literature [10, 28]. Acetyl tetrahydroumbellulone [4b]was prepared by acetylation of tetrahydroumbellulone, andits spectroscopic characterization will be reported in afuture comprehensive paper on the chemistry of umbellulone.

Reagents Umbellulone was dissolved in 100% DMSO atthe final concentration of 1 M; further dilution was inaqueous buffer. If not otherwise indicated, all reagents werefrom Sigma-Aldrich.

Cell culture Stable transfected mouse TRPA1-CHO werekindly obtained from Dr. Patapoutian (Scripps, La Jolla).For measurements on TRPM8, a stable cell line establishedin our laboratory was used, TRPM8-HEK293.

Electrophysiology Ionic currents were recorded in thewhole cell with an EPC-10 (HEKA Elektronik, Lambrecht,Germany) using ruptured patches sampled at 5 kHz andfiltered at 2.9 kHz as previously described configurations ofthe patch-clamp technique [23]. Between 50% and 80% ofthe series resistance was compensated. We used in thisstudy 200-ms voltage ramps from a holding of 0 mV from−100 to +100 mV applied every 2 s. The followingsolutions were used for whole-cell measurements in thebath (millimolar): 150 NaCl, 5 MgCl2, 10 (4-(2-hydrox-yethyl)-1-piperazineethanesulfonic acid) (HEPES), 10 glu-cose, and NaOH to pH 7.4. Pipette solution contained(millimolar): 100 K-Aspartate; 40 KCl; 2 MgCl2; 4 Na2-ATP, 5 EGTA, and 1.2 CaCl2 for buffering at 50 nM freeCa2+, 10 HEPES, and KOH to pH 7.2.

Fig. 1 Umbellulone and derivatives. The chemical structures ofumbellulone (UMB,[1]), its non-electrophilic derivates [2–5], and itsnaturally occurring analogs [6–8]. [2]=Dihydro-UMB; [3a]=β-Umbellulol; [3b]=Acetyl β-umbellulol; [4a]=Tetrahydro-UMB;[4b]=Acetyl tetrahydro-UMB; [5] Malonildihydro-UMB; [6]=Piperitone; [7]=Verbenone; [8]=3-Carene

862 Pflugers Arch - Eur J Physiol (2011) 462:861–870

Statistical analyses Statistical analysis and graphical pre-sentations were performed using the Origin ver. 7.0software (OriginLab, Northampton, MA, USA). Data arepresented as means±SEM. We used the one-way analysis ofvariance followed by the post hoc Bonferroni’s test forcomparisons of multiple groups and the unpaired two-tailedStudent’s t test for comparisons between two groups. P<0.05 was considered statistically significant.

Results

Michael acceptor properties

Figure 1 gives an overview on the various derivatives ofUMB tested in this electrophysiological approach. While

umbellulone gave a positive cysteamine assay for a reversibleMichael acceptor, all the other compounds were unreactive inthe test and did not show any thiol-trapping property.

Umbellulone is a bimodal activator of TRPA1

UMB is a reversible activator of TRPA1 [22], allowingrepetitive activation of TRPA1, in contrast to the activationof TRPA1 with mustard oil (MO; allyl isothiocyanate). Thereversible mode of activation allows activation by MO afterstimulation with UMB (Fig. 2a, c). In addition to this novelmode of activation, UMB can also block TRPA1 channelspre-activated by MO (Fig. 2a–d). This block probablycauses an off-response (increase of currents throughTRPA1) after washout of UMB (Fig. 2a–d). To quantifythis block, we measured MO-activated currents as differentialcurrent before and during MO application (ΔIMO) and during

Fig. 2 Umbellulone is a bimodalactivator of TRPA1. a and cRepresentative time courses ofwhole-cell currents throughTRPA1 measured at membranepotential of −80 mV (blackcircles) and +80 mV (red circles)during application of umbellu-lone (UMB) at concentrations of300 μM (a) and 1 mM (c), and100 μM mustard oil (MO). Notethe rapid rebound activation afterwashout of 1 mM UMB and thefast, reversible current inhibitionof UMB on TRPA1 pre-activatedby 100 μM MO. b and dCurrent–voltage (I–V) relation-ships obtained with voltageramps from −100 to +100 mV atthe times indicated in (a) and (c),respectively. e Dose–responsecurves for the inhibition on MO-induced currents by UMB. TheIC50 value was 408±52 μM

Pflugers Arch - Eur J Physiol (2011) 462:861–870 863

MO application and co-application of UMB (ΔIUMB).Inhibition was defined by

Inh %ð Þ ¼ ΔIMO �ΔIUMB

ΔIMO� 100

Block was fast, reversible, and voltage independent (Fig. 2a,c and see current voltage relationships in Fig. 2b, d). An IC50

of 408±52 μM was calculated. The fast/reversible block andthe off-response let us hypothesize that UMB may act ablocker probably of the MO dilated pore [18, 25].

Umbellulone is only a partially electrophilic TRPA1activator

To investigate the mechanism of TRPA1 activation byUMB, we studied its current activation through the mutants

TRPA1 channel in which the most reactive cysteine [20],C622, was exchanged by a serine, C622S (Fig. 3). UMBactivates TRPA1 transiently, and subsequent application ofMO also causes channel activation (Fig. 3a), in sharpcontrast with what was observed with MO, that is unable toactivate the channel after a first MO challenge. Incidentally,current–voltage (IV) curves after MO show much lessoutward rectification than those from UMB activation(Fig. 3b). Mutants (C622S) TRPA1 were still sensitive toactivation by UMB, albeit less sensitive than the wild-typechannels, whereas MO activation in these mutants wasstrikingly diminished (Fig. 3c, d). UMB still activated afraction of mutant TRPA1 channels, where the secondaryMO application was nearly ineffective (Fig. 3e). The smallsecondary current activation by MO in wild-type channelsprobably results from a partial cross-desensitization. Thesedata point to a dual model of activation: partly electrophilic

Fig. 3 Comparison of effects ofumbellulone on whole-cellcurrents through wild-type andC622S mutant TRPA1. Solu-tions with an extracellularCa2+ concentration of 1.5 mMwere used in the experimentsto accelerate TRPA1 inactiva-tion. a Representative timecourse of whole-cell currentsactivated by 1 mM UMB,followed by 50 μM MO throughwild-type TRPA1 measured atmembrane potential of −80 mV(black circles) and +80 mV(red circles). UMB activatedthe TRPA1 transiently, andsubsequent application of MOalso caused channel activation.b I–V relations measured at thetimes indicated in (a). c Repre-sentative time course of whole-cell currents elicited by 1 mMUMB, followed by 50 μM MOthrough C622S mutant TRPA1measured at membrane potentialof −80 mV (black circles)and +80 mV (red circles).C622S mutant TRPA1 was lessactivated than the wild-typechannels by both UMB and MO,especially MO activation wasstrikingly diminished. d I–Vrelations measured at the timesindicated in (c). f Comparisonof average values for currentsactivated by 1 mM UMB and50 μM MO through wild-typeor C622S mutant TRPA1

864 Pflugers Arch - Eur J Physiol (2011) 462:861–870

but also significantly non-electrophilic. Interestingly, theUMB block was maintained in the C622S mutants as slowand small currents through TRPA1 are still activated byMO. The inhibition of the small remaining current activatedby 100 μM MO was 82±15% (n=3, data not shown). Thisblock is, obviously, not linked to the covalent modificationof the reactive cysteines.

Modulation of TRPA1 activation by derivativesof umbellulone

Dihydroumbellulone ([2], Dihydro-UMB) could slowlyincrease currents through TRPA1, and could be administeredrepetitively (Fig. 4a). In comparison to the MO activation ofTRPA1, the one from Dihydro-UMB was smaller (approxi-mately 25% of the outward MO-activated current), and theoutward currents were more activated than the inward

currents (see IV curves in Fig. 4b, c). When applied duringadministration of MO, Dihydro-UMB potently inhibited thesecurrents (Fig. 4a, c). Activation of TRPA1 by Dihydro-UMB(EC50 of approximately 22 μM, Fig. 4d) was also observed inthe double mutants C622S/C642S (Fig. 4d). Half maximalblock of the MO-activated current through TRPA1 had anIC50 of 339±83 μM (n=7, Fig. 4e). Dihydro-UMB had asimilar blocking potency than in the wild-type channel. Thus,for 1 mM Dihydro-UMB, currents activated in the C622S/C642S mutants were 78±18% (n=4), not significantlydifferent than those from the wild-type block (data notshown).

β-Umbellulol [3a] did not show any TRPA1 activationproperties, but was still an efficient blocker of MO-activatedTRPA1 channels (Fig. 5a). Acetyl β-umbellulol [3b] was alsounable to activate TRPA1, but showed a reduced blockingpotency compared to 3a (Fig.5b). Tetrahydroumbellulone

Fig. 4 Modulation of TRPA1activation by dihydroumbellu-lone [2]. a Representative timecourse of whole-cell currentsmeasured at membrane potentialof −80 mV (black circles) and+80 mV (red circles) duringapplication of dihydroumbellu-lone (Dihydro-UMB) and MO atindicated concentrations. Notethe small and slowly increasedcurrent elicited by Dihydro-UMB and the potent inhibitionof Dihydro-UMB on TRPA1pre-activated by 100 μM MO. band c I–V relations measured atthe times indicated in (a). dComparison of dose–responsecurves of Dihydro-UMB onwild-type and C622/642Smutant TRPA1. e Dose–re-sponse curves for the inhibitionon MO-induced current byDihydro-UMB. The IC50 valuewas 339±83 μM

Pflugers Arch - Eur J Physiol (2011) 462:861–870 865

[4a], while lacking any activating properties, could alsoeffectively inhibited MO-activated currents through TRPA1(Fig. 5c). Acetylation had a different effect on 3a and 4asince tetrahydroumbellulone acetate [4a] was still an efficientTRPA1 blocker (Fig. 5d). In conclusion, all reducedderivatives of UMB lacked significant TRPA1-activatingproperties except dihydro-UMB, which showed modestresidual TRPA1-activating properties. The inhibitory potencywas dihydroumbellulone [2]>tetrahydroumbellulone [4a]~β-umbellulol [3a]~acetyl tetrahydroumbellulone [4b] >>>acetyl β-umbellulol [3b] (calculates IC50 values in micromo-lar 339±83, 380±29, 420±42, 490±55 and >1 mM, respec-tively, n between 3 and 7). The application of highconcentrations of these compounds (1 mM) was mostlyaccompanied by an off-response, which might indicate thatsome small activating properties are maintained but strikingly

overlap with the inhibitory effects. The behavior of malonildi-hydroumbellulone [5] was similar to that of β-umbellulol [3a]and tetrahydroumbellulone [4a], but was characterized by amuch more accentuated off-response already at 300 μM,causing a complete block of the 100 μM MO response(Suppl. 1).

Modulation of TRPM8 activation by umbelluloneand its derivatives

UMB could activate TRPM8 very quickly and is reversible(Fig. 6a). The shape of the IV curves was not changed, andthe activated channel showed a striking outward rectifica-tion (Fig. 5b, c). The activating efficiency for 1 mM UMBwas approximately 30% of the activating effects of 100 μMmenthol, and was therefore modest (Fig. 6d). Umbellulone

Fig. 5 Modulation of TRPA1activation by derivatives ofumbellulone. Representativetime courses of whole-cellcurrents measured at membranepotential of −80 mV (blackcircles) and +80 mV (redcircles), and corresponding I–Vrelations during application ofβ-umbellulol (a), acetyl β-umbellulol (b), tetrahydro-UMB(c), acetyl tetrahydro-UMB (d),and MO at indicated concentra-tions. All these compounds hadno TRPA1 activation potency,but had an efficient block effecton MO pre-activated TRPA1currents, except for acetylβ-umbellulol with a reducedblocking potency (b)

866 Pflugers Arch - Eur J Physiol (2011) 462:861–870

was only a modest (<10%) inhibitor of menthol-activatedTRPM8 currents (Fig. 6a, e). Dihydroumbellulone [2] hadonly marginal activating properties, but could, nevertheless,still inhibit outward currents through TRPM8 (data notshown, 41±8% inhibition of TRPM8 current activated by100 μM menthol, n=3).

β-Umbellulol [3a] was a robust activator of TRPM8 andalso showed potent inhibition properties (Fig. 7a), but bothproperties were substantially lost in its acetyl derivative[3b] (Fig. 7b). Also, tetrahydroumbellulone [4a] was arobust TRPM8 activator, albeit with only a modest inhibitoryactivity (Fig. 6c). As with β-umbellulol, acetylation wasdetrimental for both activities (Fig. 7d).

Figure 8 summarize these activation data. All derivativesof UMB are activators of TRPM8 except the acetylderivatives [3b] and [4b] and the malonate adduct 5. The

potency of β-umbellulol [3a] and tetrahydroumbellulone[4a] was significantly higher than that of UMB (p>0.05),while the malonate adduct was substantially inactive, actingas a sort of negative reference for these properties (Fig. 8a).The TRPM8 inhibitory properties are summarized inFig. 8b. All compounds showed a certain degree ofinhibitory properties, peaking in the malonate adduct [5].Interestingly, this compound was unable to activate TRPM8,and, with an almost complete block at 1 mM, behaved as atrue inhibitor (Fig. 8b and Supplementary figure).

Discussion

The olfactory properties of monoterpenoids have attracted aremarkable attention, generating an enormous amount of

Fig. 6 Modulation of TRPM8activation by umbellulone. aRepresentative time course ofwhole-cell currents throughTRPM8 measured at membranepotential of −80 mV (blackcircles) and +80 mV (redcircles) during application ofUMB and menthol at indicatedconcentrations. Note 1 mMUMB activates TRPM8 veryfast and is reversible. b and cI–V relations measured at thetimes indicated in (a). The acti-vated channel showed a strikingoutward rectification. d Com-parison of average values ofincreased outward and inwardcurrents induced by UMB withdifferent concentrations andmenthol. e Bars showed a smallpotency of UMB to inhibitTRPM8 current pre-activated bymenthol. Inhibitory effects areless than 10%

Pflugers Arch - Eur J Physiol (2011) 462:861–870 867

structure–activity data and paving the way for the molecularstudy of olfaction [9, 21]. Conversely, the study of theirsensory properties has long been confined to the coolant andpleasant effect of menthol, incidentally dissected from theolfactory minty note of this compound [7, 21]. Our interestin this field was piqued by the surprising observation thatinhalation of the offensive monoterpene ketone umbellulonecauses, as a prelude to painful autonomic symptoms, anintensely cold and unpleasant sensation in the ipsilateralnostril [6, 22]. Since TRPA1 and TRPM8 have beenidentified as cold sensors [19], the offensive sensoryexperience caused by the inhalation of umbellulone couldbe due to the stimulation of nasal trigeminal nerve endingsexpressing these channels. We have previously reported thatumbellulone can indeed activate TRPA1 [22], and provide

new evidence that its activation mode shows differencescompared to the one of MO, the classic TRPA1 activator,and that umbelllulone can also activate TRPM8.

Since TRPA1 is activated by Michael acceptors, theidentification of UMB as a TRPA1 activator is surprisingsince β,β-disubstituion is expected to totally preventconjugate addition of nucleophiles to enones and UMBdoes not form, indeed, any isolable adduct with thiols. Onthe other hand, UMB forms stable Michael adducts withcarbon nucleophiles [10], and was shown to trap thiolgroups in a click (very fast and quantitative) fashion,elusive, however, since the adducts reverted upon attemptedisolation [3]. Since the classic activators of TRPA1 like MOform stable and isolable thiol adducts, it was interesting tocompare the activation mode of TRPA1 by UMB and MO.

Fig. 7 Modulation of TRPM8activation by derivatives ofumbellulone. Representativetime courses of whole-cellcurrents measured at membranepotential of −80 mV (blackcircles) and +80 mV (redcircles), and corresponding I–Vrelations during application ofβ-umbellulol (a), acetylβ-umbellulol (b), tetrahydro-UMB (c), acetyl tetrahydro-UMB (d), and menthol atindicated concentrations

868 Pflugers Arch - Eur J Physiol (2011) 462:861–870

Several interesting features emerged in this study. The firstone is that UMB is a mechanistically hybrid activator,apparently combining covalent interaction at reactive ankyrinthiols of TRPA1 with noncovalent interaction with a secondsite. Thus, in sharp contrast with MO, UMB retainedsignificant activating potency also for mutant forms of TRPA1where the active cysteine residues have undergone isostericsubstitution to serine. Furthermore, UMB is a bimodal ligandthat activates TRPA1 at low concentration, but that paradoxi-cally inhibits it at higher concentrations, as already describedfor menthol and nicotine, two noncovalent ligands [17, 29].The nature of the very fast block and fast unblock of TRPA1by UMB is not known, but the kinetics suggests that thiscompound could act as a pore blocker, as probably do alsoits analogs [2]-[5]. Interestingly, chemical modification coulddissect the two actions, as exemplified by the malonyl adduct[5], that behaves as a pure antagonist, and represents aninteresting new chemotype of TRPA1 antagonist. Reductionof the enone system of UMB led to compounds lackingelectrophilic properties, as evidenced by a negative cyste-amine assay, but still showing residual TRPA1-activatingcapacity, in accordance with the suggestion that UMB has alsoa noncovalent-activating site on this ion channel. Conversely,the inhibitory potency increased significantly, and a completedissection between activation and inhibition could beachieved with the malonyl adduct [5]. Taken together, theseobservations suggest that UMB is a privileged structure forbinding to TRPA1. The related monoterpenoids piperitone[6] and verbenone [7] did not show any significant affinityfor TRPA1, while its deoxygenated analog 3-carene [8]was a good activator, suggesting that the cyclopropane

ring is a critical element for the monoterpenoid TRPA1pharmacophore.

When assayed for TRPM8 activation, UMB showed onlymodest potency, but the activity increased in its reducedanalogs, in accordance with the recent identification ofderivatives of umbellulol as potent coolant agents [28]. Thereduced analogs are devoid of the offensive trait of thenatural product that seems entirely related to activation ofTRPA1. The potent TRPM8-activating properties of [3a] and[4a] are somewhat surprising since the presence of thecyclopropane ring leads to a substantial flattening of thecyclohexane ring compared to menthol. Acetylation of thereduced umbellulone derivatives was, in general, detrimentalfor affinity to both TRPA1 and TRPM8, a somewhatunexpected finding since short-chain menthol esters fullyretain the activity of the natural product on TRPM8 [27].

The identification of neural targets for volatile terpenoidsis interesting since these compounds, as well as other plantvolatiles, can have a surprisingly good brain penetration byinhalation, where they can potentially trigger pharmacologicalresponses [12, 13]. The blurred literature on α-thujone, awrongly identified cannabinoid ligand but actually a GABA-A modulator [16], and the controversies surrounding thescientific bases of aromatherapy [15] might have substan-tially relegated the investigation on the neurological rele-vance of plant volatiles to the realm of olfaction, where theirpreeminence is unquestionable [21]. The study of TRPs andcannabinoid receptors [11], and the current clinical interestfor essential oil constituents like thymoquinone [1] or menthol[8], are now bringing volatile compounds to the forefront ofthe biomedical research.

Fig. 8 Comparison of theeffects of umbellulone and itsderivatives on TRPM8 (a, b)Bar graphs showed calibratedactivation effects (a) and inhib-itory properties (b) of the UMBderivatives on TRPM8. Allderivatives of UMB are activa-tors of TRPM8 except theiracetylated form and malonylumbellulone. Malonyl umbellu-lone was the most potentTRPM8 inhibitor, with analmost complete block at 1 mM

Pflugers Arch - Eur J Physiol (2011) 462:861–870 869

Acknowledgments The work was supported by grants from theBelgian Federal Government (IUAP P6/28. BN), the ResearchFoundation-Flanders (F.W.O.) (G.0565.07 and G.0686.09, BN), andthe Research Council of the KU Leuven (GOA 2009/07 and EF/95/010, BN). JZ was supported by grants from the Natural ScienceFoundation of China (no. 30900619) and the National Basic ResearchProgram of China (no. 2012CB517806).

References

1. Akhondina J, Kjanifar H, Raoofziaee M, Moayedpour A, ToosiMB, Khajedaluee M (2001) The effect of thymoquinone onintractable pediatric seizures (pilot study). Epilepsy Res 93:39–43

2. Appendino G, Minassi A, Pagani A, Ech-Chahad A (2008) Therole of natural products in the ligand deorphanization of TRPchannels. Curr Pharm Des 14:2–17

3. Avonto C, Taglialatela-Scafati O, Pollastro F, Minassi A, DiMarzo V, De Petrocellis L, Appendino G (2011) An NMRspectroscopic method to identify and classify thiol-trappingagents: revival of Michael acceptors for drug discovery? AngewChem Int Ed Engl 50:467–471

4. Benemei S, Appendino G, Geppetti P (2009) Pleasant naturalscent with unpleasant effects: cluster headache-like attackstriggered by Umbellularia californica. Cephalalgia 30:744–746

5. Benemei S, Nicoletti P, Capone JG, De Cesaris F, Geppetti P(2009) Migraine. Handb Exp Pharmacol 194: 75–89

6. Benemei S, Nicoletti P, Capone JG, Geppetti P (2009) CGRPreceptors in the control of pain and inflammation. Curr OpinPharmacol 9:9–14

7. Boedding MB, Wissenbach U, Flockerzi V (2001) Characterisationof TRPM8 as a pharmacophore receptor. Cell Calcium 42:618–628

8. Borhani Haghighi A, Motazedian S, Rezaiim R, Mohammadim F,Salarian L, Pourmokhtarim M, Khodaei S, Vossoughi M, Miri R(2010) Cutaneous application of menthol 10% solution as anabortive treatment of migraine without aura: a randomised,double-blind, placebo-controlled crossed-over study. Int J ClinPract 64:451–458

9. De Maria S, Nqai J (2010) The cell biology of smell. J Cell Biol191:443–452

10. Eastman RH (1954) Conjugative effects of cyclopropane rings. I.Synthesis and properties of 1-methyl-4-isopropyltricyclo[4,1,01,6,02,4]heptanon-5. J Am Chem Soc 76:4115–4117

11. Gertsch J, Leonti M, Raduner S, Racz I, Chen JZ, Xie XQ,Altmann KH, Karsak M, Zimmer A (2008) Beta-caryophyllene isa dietary cannabinoid. Proc Natl Acad Sci U S A 195:9099–10104

12. Guo J, Duan J, Shang E, Tang Y, Qian D (2009) Determination ofligustilide in rat brain after nasal administration of essential oilfrom Rhizoma chuanxiong. Fitoterapia 80:168–172

13. Guo J, Shang E, Duang J, Tang Y, Qian D (2011) Determinationof ligustilide in the brain of freely moving rats using microdialysiscoupled with ultra performance liquid chromatography/massspectrometry. Fitoterapia 82:441–445

14. Heamy JP (1875) Oreodaphne californica. Am J Pharm 47:105–110

15. Herz RS (2001) Aromatherapy facts and fictions: a scientificanalysis of olfactory effects on mood, physiology and behavior.Int J Neurosci 119:263–290

16. Hoeld KM, Sirisoma NS, Ikeda T, Narahashi T, Casida JE (2001)Alfa-thujone (the active component of absinthe): gamma-aminobutyric acida type A receptor modulation and metabolicdetoxification. Proc Natl Acad Sci U S A 97:3826–3831

17. Karashima Y, Damann N, Prenen J, Talavera K, Segal A,Voets T, Nilius B (2007) Bimodal action of menthol on thetransient receptor potential channel TRPA1. J Neurosci27:9874–9884

18. Karashima Y, Prenen J, Talavera K, Janssens A, Voets T, Nilius B(2010) Agonist-induced changes in Ca(2+) permeation throughthe nociceptor cation channel TRPA1. Biophys J 98:773–783

19. Karashima Y, Talavera K, Everaerts W, Janssens A, Kwan KY,Vennekens R, Nilius B, Voets T (2009) TRPA1 acts as a cold sensorin vitro and in vivo. Proc Natl Acad Sci U S A 106:1273–1278

20. Macpherson LJ, Dubin AE, Evans MJ, Marr F, Schultz PG,Cravatt BF, Patapoutian A (2007) Noxious compounds activateTRPA1 ion channels through covalent modification of cysteines.Nature 445:541–545

21. Maffei ME, Gertsch J, Appendino G (2011) Plant volatiles:production, function and pharmacology. Nat Prod Res 28:1359–1380

22. Nassini R, Materazzi S, Vriens J, Prenen J, Benemei S, De SienaG, la Marca G, Andrè E, Preti D, Avonto C, Sadofsky L, DiMarzo V, De Petrocellis L, Dussor G, Porreca F, Taglialatela-Scafati O, Appendino G, Nilius B, Geppetti P (2011) The‘headache tree’, via UMBELLULONE and TRPA1, activates thetrigeminovascular system. Brain. doi:10.1093/brain/awr272

23. Nilius B, Mahieu F, Prenen J, Janssens A, Owsianik G, VennekensR, Voets T (2006) The Ca2+-activated cation channel TRPM4 isregulated by phosphatidylinositol 4,5-biphosphate. EMBO J25:467–478, England

24. Nilius B, Owsianik G, Voets T, Peters JA (2007) Transientreceptor potential channels in disease. Physiol Rev 87:165–217

25. Nilius B, Prenen J, Owsianik G (2011) Irritating channels: thecase of TRPA1. J Physiol 589:1543–1549

26. Peattie DC (1953) A natural history of western trees. HoughtonMifflin, Boston

27. Sherkheli MA, Vogt-Eisele AK, Bura D, Beltràn Màarques LR,Gisselmann G, Hatt H (2010) Characterization of selectiveTRPM8 ligands and their structure activity response (S.A.R.)relationships. J Pharm Pharm Sci 13:242–253

28. Starkenmann C, Cayeuz J, Brauchli R, Mayenzet F (2011)Hemisynthesis of dihidroumbellulols from umbellulone: newcooling compounds. J Agric Food Chem 59:677–683

29. Talavera K, Gees M, Karashima Y, Meseguer VM, VanoirbeekJA, Damann N, Everaerts W, Benoit M, Janssens A, Vennekens R,Viana F, Nemery B, Nilius B, Voets T (2009) Nicotine activatesthe chemosensory cation channel TRPA1. Nat Neurosci 12:1293–1299

870 Pflugers Arch - Eur J Physiol (2011) 462:861–870