Structure-Pungency Relationships and TRP Channel Activation of Drimane Sesquiterpenes in Tasmanian Pepper (Tasmannia lanceolata)

Article abstract of DOI:10.1021/acs.jafc.7b02356

Sensory-guided fractionation of extracts of Tasmanian pepper berries revealed 20 drimane sesquiterpens, among which polygodial, warburganal, and 1β-acetoxy-9-deoxy-isomuzigadial exhibited the lowest pungency threshold concentrations on the tongue surface (0.6-2.8 nmol/cm²) and elicited a dose-dependent calcium influx into mTRPA1 expressing CHO cells with the lowest EC₅₀ values (4.5 ± 1.0 to 16.7 ± 7.5 μmol/L) and a good correlation to oral pungency thresholds (R² = 0.986, linear regression). Calcium imaging assays demonstrated these chemosensates to induce a calcium influx into cultured trigeminal neurons prepared from wildtype (TRPA1^+/+) mice, whereas no calcium influx was observed in neurons from TRPA1 knockout mice (TRPA1^-/-), thus confirming the α,β-unsaturated 1,4-dialdehyde structure to be the required structural motif for a low oral puncency thresholds and activation of the Transient Receptor Potential Channel A1 (TRPA1). Time-resolved NMR experiments confirmed the pungency mediating mechanism for electrophilic drimane sesquiterpene dialdehydes to be different from that found for other electrophilic pungent agents like isothiocyanates, which have been shown to undergo a covalent binding with cysteine residues in TRPA1. Instead, the high-impact chemosensates polygodial, warburganal, and 1β-acetoxy-9-deoxy-isomuzigadial showed immediate reactivity with the ?-amino group of lysine side chains to give pyrrole-type conjugates, thus showing evidence for TRPA1 activation by covalent lysine modification.

Full text of DOI:10.1021/acs.jafc.7b02356

Subscriber access provided by NEW YORK UNIV
Article
Structure-Pungency Relationships and TRP Channel Activation of
Drimane Sesquiterpenes in Tasmanian Pepper (Tasmannia lanceolata)
Klaus Mathie, Johanna Lainer, Stefan Spreng, Corinna Dawid,
David A. Andersson, Stuart Bevan, and Thomas Hofmann
J. Agric. Food Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jafc.7b02356 • Publication Date (Web): 28 Jun 2017
Downloaded from http://pubs.acs.org on June 30, 2017
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of Agricultural and Food Chemistry is published by the American Chemical
Society. 1155 Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 48
Journal of Agricultural and Food Chemistry
1
2
3
StructureꢀPungency Relationships and TRP Channel
Activation of Drimane Sesquiterpenes in Tasmanian
Pepper (Tasmannia lanceolata)
4
5
Klaus Mathie1, Johanna Lainer1, Stefan Spreng1, Corinna Dawid1, David
A. Andersson2, Stuart Bevan2, and Thomas Hofmann1*
6
7
8
9
1 Chair of Food Chemistry and Molecular Sensory Science, Technische Universität
München, LiseꢀMeitnerꢀStraße 34, Dꢀ85350 Freising, Germany, and
2 Kings’s College London, Wolfron Centre for AgeꢀRelated Diseases, London SE1
1UL, United Kingdom
10
11
12
13
14
15
16
17
18
19
20
21
22
*
23
24
25
26
Author to whom correspondence should be addressed
PHONE
FAX
+49ꢀ8161/71ꢀ2902
+49ꢀ8161/71ꢀ2949
EꢀMAIL
thomas.hofmann@tum.de
27
28
29
1
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 2 of 48
1
ABSTRACT
2
3
Sensoryꢀguided fractionation of extracts of Tasmanian pepper berries revealed 20
drimane sesquiterpens, amongst which polygodial, warburganal, and 1βꢀacetoxyꢀ9ꢀ
deoxyꢀisomuzigadial exhibited the lowest pungency threshold concentrations on the
tongue surface (0.6 ꢀ 2.8 nmol/cm2) and elicited a doseꢀdependent calcium influx into
mTRPA1 expressing CHO cells with the lowest EC50 values (4.5 ± 1.0 ꢀ 16.7 ± 7.5
ꢁmol/L) and a good correlation to oral pungency thresholds (R2 = 0.986, linear
regression). Calcium imaging assays demonstrated these chemosensates to induce
a calcium influx into cultured trigeminal neurons prepared from wildtype (TRPA1+/+)
mice, whereas no calcium influx was observed in neurons from TRPA1 knockout
mice (TRPA1ꢀ/ꢀ), thus confirming the α,βꢀunsaturated 1,4ꢀdialdehyde structure to be
the required structural motif for a low oral puncency thresholds and activation of the
Transient Receptor Potential Channel A1 (TRPA1). Timeꢀresolved NMR experiments
confirmed the pungency mediating mechanism for electrophilic drimane
sesquiterpene dialdehydes to be different from that found for other electrophilic
pungent agents like isothiocyanates, which have been shown to undergo a covalent
binding with cysteine residues in TRPA1. Instead, the highꢀimpact chemosensates
polygodial, warburganal, and 1βꢀacetoxyꢀ9ꢀdeoxyꢀisomuzigadial showed immediate
reactivity with the εꢀamino group of lysine side chains to give pyrroleꢀtype conjugates,
thus showing evidence for TRPA1 activation by covalent lysine modification.
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
KEYWORDS: pungency, drimane, TRPA1, TRPV1, taste dilution analysis, halfꢀ
tongue test, Tasmanian pepper, Tasmannia lanceolata
2
ACS Paragon Plus Environment

Page 3 of 48
Journal of Agricultural and Food Chemistry
1
INTRODUCTION
2
3
4
5
6
7
8
9
Tasmanian Mountain Pepper (Tasmannia lanceolata) is a spice plant from the family
of Winteraceae native to Tasmania and parts of Australia.1 Often recognised as
“bushfood”, Tasmanian Pepper is part of a range of edible indigenous plants which
have gained increasing popularity in modern cuisine.2,3 Next their use as medicinal
herb,4,5 the lanceolate leaves as well as the aromatic berries of the large shrub are
highly appreciated for their typical pungency and their characteristic aromatic and
freshꢀspicy odor,5 which makes Tasmanian mountain pepper an appreciated
ingredient for manufacturing of wasabi paste and chewing gum6.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
More than 50 years ago, the drimane sesquiterpene polygodial, 1 (Figure 1),
featuring an α,βꢀunsaturated 1,4ꢀdialdehyde structure, was isolated from Tasmanian
pepper leaves7 and was proposed as the key pungent principle of the berries.8 Over
the last decades, polygodial had been found in a range of botanicals, such as, e.g.
water pepper (Polygonum hydropiper),9 as well as in animal species,10,11 and is
widely discussed for its antifeedant12 properties and its role as a predator
deterrent.13,14 Just as wide as the occurrence of drimane aldehydes in nature, are
their biological activities, ranging from pungent taste12 to antimicrobial,15 antifungal,16
antibiotic,17 antifeedant,18 antinociceptive,19 and antihyperalgesic properties.20
Several studies proposed the biological activity to be directly related to its pungent
activity induced by the reactive α,βꢀunsaturated 1,4ꢀdialdehyde moiety.12,21,22
Although chemical reactions of polygodial with nucleophiles such as, e.g.
amino and mercapto groups,23ꢀ28 have been suggested, defined reaction products
could not yet be identified. More recent studies with pungent compounds from
mustard and garlic revealed that isothiocyanates undergo a reversible covalent
binding with the Transient Receptor Potential Channel A1 (TRPA1),29,30 which
3
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 4 of 48
1
2
together with TRPV1 is a key receptor channel protein involved in the perception of
pain, temperature and pungency.31 Next to these two polymodic, nonꢀselective cation
channels, only recently TRPV1 and TRPA1 agonists, such as, e.g. piperine,
capsaicin, 6ꢀgingerol and polygodial, were found to possess a marked effect on twoꢀ
pore domain (KCNK, K2P) potassium channels,32 which have been shown to “fineꢀ
tune” the cellular response to stimuli that active TRP channels.33 Today it is well
accepted that the reaction of pungent agonists, such as, e.g. isothiocyanates,
sulfides, thiols and small reactive unsaturated aldehydes like acrolein, with cysteine
residues plays a major role in TRPA1 activation. However, the unaltered channel
activation by unsaturated dialdehydes in TRPA1 mutants in which some key cysteine
residues were replaced,31,34 suggests a fundamentally different TRPA1 activation
mechanism by polygodial. Despite all recent efforts, the mechanism by which TRPA1
is activated by drimane sesquiterpenes as well as knowledge on structureꢀactivity
relationships within this class of components remain rather obscure.
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
Therefore, the objectives were to locate and identify the most intense pungent
phytochemicals by application of a sensoryꢀguided fractionation approach on
Tasmanian mountain pepper, and to study structureꢀactivity relationships by means
of human sensory tests, TRP channel activation studies, and NMRꢀspectroscopic
experiments on reactions between drimane sesquiterpenes and nucleophilic amino
acid residues as candidate TRPA1 targets.
23 MATERIALS AND METHODS
24
25
26
Chemicals and Materials. Unless stated otherwise, all chemicals were obtained
commercially from SigmaꢀAldrich (Steinheim, Germany) and were of p.a. grade. Nꢀ
4
ACS Paragon Plus Environment

Page 5 of 48
Journal of Agricultural and Food Chemistry
1
2
3
4
5
6
7
8
9
Acetylꢀ
L
ꢀlysine, formic acid, diethyl ether, ethanol absolute, and aluminium oxide 90
(basic, 0.063ꢀ0.200 mm, activity 1, for column chromatography) were obtained from
Merck (Darmstadt, Germany), deuterium oxide and chloroform (D100, 0.03% TMS)
from EurisoꢀTop (GifꢀSurꢀYvette, France), the solvents methanol, acetonitrile, ethyl
acetate, and hexane used for highꢀperformance liquid chromatography (HPLC) were
HPLC grade and from J.T. Baker (Deventer, Netherlands), NꢀacetylꢀLꢀarginine and
polygodial from Santa Cruz Biotechnology (Heidelberg, Germany), nerve growth
factor from Promega (Southampton, UK), penicillinꢀstreptomycin and laminin from
Invitrogen (UK), collagenase type IV and DNAse I from Worthington Biochem (US),
fetal bovine serum (FBS), MEM glutamax, hygromycin B, trypsine from Gibco Life
Technologies (UK), and FURA2ꢀAM from Molecular Probes (UK). Membrane filter
disks for filtration of HPLC samples (0.45 ꢁm) were purchased from Satorius AG
(Goettingen, Germany). Water used for chromatography was purified by means of a
MilliꢀQ Advantage A10 water purification system (Millipore, France), and bottled
water (Evian) was used for sensory analyses. Filter paper (Rundfilter Original 1,
Melitta, Minden, Germany) was used as carrier for sensory analyses using halfꢀ
tongue tests. Tasmanian pepper berries were purchased from the Australian retail (A
Taste Of The Bush; Matcham, New South Wales, Australia).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
Solvent Extraction of Tasmanian Pepper Berries. Tasmanian pepper
berries were ground using a laboratory blender and a portion (350 g) was extracted
four times with ethanol (1.8 L) at room temperature upon ultrasonification. After
filtration through a Buchner funnel (filter 1291, 90 mm, 84 g/m2, Sartorius, Germany),
ethanol was removed in vacuum at 40°C, followed by freezeꢀdrying. The dried extract
was kept at ꢀ20°C until used.
Sensory Analyses. General Conditions and Panel Training. 12 assessors
(age 22ꢀ38 years), who had given informed consent to participate the sensory tests
5
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 6 of 48
1
2
and had no history of known taste disorders, were recruited from the Chair of Food
Chemistry and Molecular Sensory Science (Freising, Germany). All participants were
trained in sensory experiments using reference compounds.35 Sensory assessments
were conducted at room temperature in a sensory panel room, equipped with
individual booths.
3
4
5
6
Modified Half-Tongue Test. Sensory test were performed by means of a
modified halfꢀtongue test according to literature.35 Briefly, test samples were
prepared by dissolving extracts, fractions or purified compounds in ethanol and
applying aliquots (20 µL) onto filter paper rectangles (1×2 cm). After solvent removal
in a nitrogen stream at 38 °C, the participants were presented with pairs of test
samples and control filters without additive, both encoded with three digit numbers
and in a randomised order. Pairs of sample and control filters were presented in
ascending order of application levels. The participants were asked to place both
filters of a pair onto their tongue at the same time, one onto each side. After a
stimulation time of 30 s, they were asked to select the sample perceived as more
pungent, as well as recording any other sensations perceived. Participants had
breaks of 5 min between each assessment, in which they were given water as a
palate cleanser.
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
Taste Dilution Analysis (TDA). An aliquot (4.3 g) of the pepper extract was
separated by RPꢀMPLC to give 10 fractions, which were separated from solvent in
vacuum at 40°C, followed by freezeꢀdrying. The residues were then dissolved in
equal volumes of ethanol (2 mL) in order to maintain the “natural” concentration
ratios between the fractions and diluted sequentially 1:1 with ethanol. Aliquots (20
µL) of each dilution were applied onto filter paper rectangles (1 × 2 cm) and halfꢀ
tongue tests were conducted as described above. The highest dilution at which a
sensory difference between the filter loaded with a pepper fraction and the control
6
ACS Paragon Plus Environment

Page 7 of 48
Journal of Agricultural and Food Chemistry
1
2
3
4
5
6
7
8
9
filter could be detected was defined as taste dilution (TD) factor36. The TDꢀfactors
determined by the participants were averaged.
Recognition Threshold Concentration. Threshold concentrations of the purified
compounds were determined in duplicate using the halfꢀtongue test as detailed
above. Individual recognition thresholds were determined by calculation of the
geometric mean of the two lowest correctly identified concentrations. The highest
concentration at which compounds were tested was 350 nmol on a filter vehicle.
Fractionation of Pepper Extracts by Medium Pressure Liquid
Chromatography (MPLC). Ethanol extracts of Tasmanian pepper berries, as well as
a hexane/ethyl acetate partition thereof, were fractionated by means of reversed
phase (RP) MPLC on 150 mm x 40 mm polypropylene cartridges using 25ꢀ40 µm
LichroPrep RP18 bulk material (Merck, Darmstadt, Germany) as stationary phase
and a gradient of 0.1% aqueous formic acid (solvent A) and methanol (solvent B) as
mobile phase (flow rate: 50 mL/min): 2 min / 20% B, 5 min / 40% B, 20 min / 80% B,
26 min / 80% B, 28 min / 100% B, 40 min / 100% B. Chromatography was monitored
using UV detection at 230 nm. The MPLC system (Büchi, Flawil, Switzerland)
consisted of a pump manager Cꢀ615 with three pump modules Cꢀ605, 20 mL
samples loop and 6ꢀwayꢀinjection valve, Cꢀ660 fraction collector, Cꢀ635 UV detector
and was controlled by the software SepacoreControl. Collected fractions were
separated from solvent in vacuum and freezeꢀdried twice prior to TDA and chemical
analysis, respectively.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
Preparation of Epipolygodial (2). Epipolygodial was obtained by conversion of
polygodial according to a literature protocol with some modifcations.35 Polygodial (36
mg) were dissolved in diethyl ether (8 mL), mixed with basic aluminium oxide (8 g;
activity 1) and stirred for 30 min. After filtration, the solvent was separated in vacuum
and epipolygodial isolated by preparative RPꢀHPLC on a 250 x 21.2 mm, 5 ꢁm,
7
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 8 of 48
1
2
Varian Pursuit C18 column with a mixture (50/50, v/v) of 0.1% aqueous formic acid
and acetonitrile as mobile phase (flow rate: 21.2 mL/min) and UV detection at 230
nm. Structure confirmation was performed by oneꢀ and twoꢀdimensional nuclear
resonance spectroscopy (1D/2DꢀNMR) and liquid mass spectrometry (LCꢀTOFꢀMS).
Spectroscopic data on compound 2, Figure 1, can be found as Supporting
Information.
3
4
5
6
7
Isolation and Identification of Pungent Phytochemicals. Most pungent
compounds in fractions judged with a high TDꢀfactor were enriched from the
ethanolic pepper extract by means of liquidꢀliquidꢀextraction using 0.1% aqueous
formic acid and hexane/ethyl acetate (8/2, v/v). The aqueous fraction was discarded
and the solvent of the organic fraction separated in vacuum, followed by freezeꢀ
drying. Aliquots were subjected to preparative MPLC for further fractionation under
the same conditions as detailed above for the taste dilution analysis (TDA). Individual
compounds were isolated from the MPLC fractions by means of preparative RPꢀ
HPLC on a 250 x 21.2 mm, 5 ꢁm, Varian Pursuit C18 column as stationary phase
using the following gradient of 0.1% aqueous formic acid (solvent A) and acetonitrile
(solvent B) as mobile phase (flow rate: 21.2 mL/min): 0 min / 30% B, 12 min / 40% B,
17 min / 40%B, 30 min / 45% B, 35 min / 55% B, 38 min / 55% B, 40 min / 100% B,
43 min / 100% B. Chromatography was monitored by means of an UV detector
(λ=230 nm) and an evaporative light scattering detector (ELSD) using a Sedex LTꢀ
ELSD Model 80 (Sedere, Alfortville, France). The HPLC system (Jasco, Grossꢀ
Umstadt, Germany) consisted of a PUꢀ2087 Plus pump, a DGꢀ2080ꢀ53 degasser, a
MDꢀ2010 Plus diode array detector, and run with the Chrompass 1.9 software. Reꢀ
chromatography was conducted, if needed, to obtain sufficient purities (>98%, HPLCꢀ
ELSD) for spectroscopic structure determination. Spectroscopic data on compounds
1 and 3 ꢀ 20, Figure 1, can be found as Supporting Information.
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
8
ACS Paragon Plus Environment

Page 9 of 48
Journal of Agricultural and Food Chemistry
1
2
3
4
5
6
7
8
9
TRP Channel Activation Experiments. Cell culture. Untransfected CHO
cells, CHO cells expressing mouse TRPA1 or rat TRPV1 were grown in MEMꢀ
medium supplemented with penicillin (100 U/mL), streptomycin (100 µg/mL), Lꢀ
glutamine (2 mM), FBS (10%) and for TRPA1ꢀtransfected cells additionally with
Hygromycin B (200 µg/mL).38 TRPA1 expression was tetracyclineꢀinduced (1 µg/mL)
12ꢀ24 h before calcium imaging measurements.39 Trigeminal ganglion (TG) neurons
were dissected from adult male or female mice and cultured using methods
described previously for Dorsal Root Ganglion neurons40.
Imaging of Intracellular Calcium Levels: Calcium imaging assays were
conducted as described previously.38,41,42 Briefly, Chinese Hamster Ovary (CHO)
cells and TG neurons were loaded with 2 µM FURAꢀ2AM in the presence of 1mM
probenecid for approximately 1h. The dye loading and subsequent experiments were
performed in a physiological saline solution containing NaCl (140 mmol/L), KCl (5
mmol/L), glucose (10 mmol/L), Hepes (10 mmol/L), CaCl2 (2 mmol/L), and MgCl2 (1
mmol/L) buffered to pH 7.4 with NaOH. Changes in intracellular calcium levels in
mTRPA1 or rTRPV1 expressing CHO cells were determined at 25°C using an
automated fluorometric plate reader (Flexstation 3, Molecular Devices). For calcium
imaging of neurons compounds were applied to cells by local continuous
microperfusion of solution through a fine tube placed very close to the cells being
studied. Experiments were conducted at 22±1 °C. Images of a group of cells were
captured every 2 s using 340 and 380 nm excitation wavelengths with emission
measured at 520 nm with a microscopeꢀbased imaging system (PTI). Analyses of
emissionꢀintensity ratios at 340 nm/380 nm excitation (R, in individual cells) were
performed using the ImageMaster suite of software.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
UPLC/Time-of-Flight Mass Spectrometry (UPLC/TOF-MS). Highꢀresolution
mass spectra of the target substances were measured on a SYNAPT G2S HDMS
9
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 10 of 48
1
2
(Waters UK Ltd., Manchester, UK) in the positive ESI and resolution modus with the
following parameters: capillary voltage +2.5 kV, sampling cone 30, extraction cone
4.0, source temperature 150 °C, desolvation temperature 450 °C, cone gas 30 L/h
and desolvation gas 850 L/h. The samples were introduced into the instrument via an
Acquity UPLC core system (Waters, Milford, MA, USA) consisting of a binary solvent
manager, a sample manager and a column oven. For chromatography, a 2 x
150 mm, 1.7 µm, BEH C18 column (Waters) was used as stationary phase with a
flow rate of 0.3 mL/min at a temperature of 40 °C and following gradient of
acetonitrile (solvent A) and aqueous formic acid (0.1% in water, pH 2.5; solvent B):
0 min / 50% B, 7 min / 100% B, 9 min / 100% B, 10 min / 50% B. The instrument was
calibrated over a m/z range of 100 to 1200 using a solution of sodium formate (0.5
mM) in a 2ꢀpropanol/water mixture (9/1, v/v). All data were lock mass corrected using
leucine enkephaline as the reference (m/z 556.2771, [M+H]+). Data acquisition and
interpretation were performed by using MassLynx software (version 4.1; Waters).
Nuclear Magnetic Resonance Spectroscopy (NMR). Oneꢀ and twoꢀ
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
1
dimensional H and 13C NMR spectra were acquired on a 400 MHz DRX and a 500
MHz Avance III Cryoprobe spectrometer (Bruker, Rheinstetten, Germany),
respectively. Chemical shifts were measured either by using tetramethylsilane (TMS)
as the internal standard or from residual chloroform signals (CDCl3, 0.03 % TMS,
Eurisoꢀtop, GifꢀsurꢀYvette, France). For structural elucidation and NMR signal
assignment 2DꢀNMR experiments, like COSYꢀ, HMQCꢀ, and HMBCꢀspectroscopy
were carried out using the pulse sequences taken from the Bruker software library.
Data processing was performed by using Topspin 1.3 software (Bruker, Rheinstetten,
Germany) and evaluated by using MestReNova 7.0.1 (Mestrelab Research, Spain).
To monitor the reaction of pungent aldehydes with amino acid residues, aliquots
(0.2 mL) of individual solutions of polygodial, warburganal, 1βꢀacetoxyꢀ9ꢀdeoxyꢀ
10
ACS Paragon Plus Environment

Page 11 of 48
Journal of Agricultural and Food Chemistry
1
isomuzigadial and changweikangaldehyde (2.5 mg/mL each) in deuterated methanol
2
3
were mixed with aliquots (0.4 mL) of stock solutions of the Nꢀacetylated amino acids
NαꢀacetylꢀLꢀlysine (3.38 mg/mL), NαꢀacetylꢀLꢀarginine (1.83 mg/mL), and NαꢀacetylꢀLꢀ
4
cysteine (2.5 mg/mL) in potassium hydrogenphospate buffer (20.4 mg KH2PO4
1
5
mg/mL in deuterium oxide, pH 7), and placed in 5 mm NMR tubes. H NMR spectra
6
(500 MHz, relaxation time: 1 s) were recorded immediately after mixing and in regular
time intervals thereafter whilst keeping the instrumental parameters constant.
Reactivity was evaluated by integrating compound specific signals.
7
8
9
10
11 RESULTS AND DISCUSSION
12
13
14
15
16
17
18
19
20
21
22
23
Tasmanian pepper berries were ground and exhaustively extracted with ethanol
to afford a strongly pungent powder after separating the solvent in vacuum. To locate
the pungent phytochemicals in the extract, a taste dilution analysis36 was conducted
in the following.
Taste Dilution Analysis (TDA). The ethanol extract of ground pepper berries
was separated into 10 fractions by means of RPꢀMPLCꢀUV (Figure 2) which, after
solvent separation, were dissolved in equal volumes of ethanol and, then, 1:1
dilutions were applied onto filter paper rectangles (1x2 cm) as vehicles for sensory
assessment in ascending order of concentration using the halfꢀtongueꢀtest. Medium
to nonꢀpolar fractions were perceived as intensely pungent with fraction F9 showing
the highest taste dilution (TD)ꢀfactor of 4096, followed by fractions F7, F8 and F10
24 judged with TDꢀfactors of 256. As the taste impact of the polar fractions F1 to F6 was
25
negligible, the target fractions F7ꢀ10 were enriched by liquidꢀliquid extraction in the
11
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 12 of 48
1
2
following and used for natural product isolation and chemosensory characterisation,
respectively.
3
Isolation and Identification of Pungent Phytochemicals. A total of 19 target
compounds were isolated from the most pungent tasting MPLC fractions F7 to F10
by means of preparative RPꢀHPLC. The major constituent polygodial (1, Figure 1)
could be unequivocally confirmed in fraction F9 by comparison of chromatographic
(RPꢀHPLC) and spectrometric data (NMR, LCꢀTOFꢀMS, UV) with those determined
for a reference standard.9,43 In addition, 13 compounds (3-13, 19-20) were isolated
from fraction F9, compound 16 was purified from fraction F8, and four compounds
(14-15, 17-18) from fraction F7. As fraction F10 contained mainly lipids and some
residual amounts of the same phytochemicals as found in fraction F9 (data not
shown), this fraction was not further used for natural product isolation. 13C NMR
experiments and accurate mass spectral data revealed that 10 of these additional 18
phytochemicals consist of 15 carbon atoms, thus suggesting sesquiterpene
substructures as found for polygodial, while the 13, 14 or 17 carbon atoms detected
in the remaining 8 phytochemicals indicate a truncated or a modified sesquiterpene
skeleton. 1D/2DꢀNMR analyses confirmed that, with the exception of compound 10,
all phytochemicals share a polygodialꢀtype, unsaturated trimethylꢀtransꢀdecaline
carbon backbone differing in the decoration of ring B with functional groups. Overall,
four groups of compounds, namely aldehydes, acids, lactones and alcohols, were
determined in their chemical structures in the following.
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
The four phytochemicals 3 to 6 showed characteristic NMR signals of
aldehydes and were identified by mass spectrometric and NMR spectroscopic
analysis. Compound 3 showed a pseudomolecular ion of m/z = 249.1419, thus
indicating an elemental composition of C15H22O3 and differing by only one oxygen
atom from polygodial (1). Two NMR signals with typical chemical shifts for aldehydes
12
ACS Paragon Plus Environment

Page 13 of 48
Journal of Agricultural and Food Chemistry
1
(9.73 ppm/HꢀC(11), 9.41 ppm/HꢀC(12)) and an olefinic proton HꢀC(7) at 7.27 ppm
were identified. Instead of the proton HꢀC(9) in 1, an oxygenꢀbound proton signal HOꢀ
C(9) has been assigned for the resonance signal detected at 4.09 ppm. In line with
this, the aldehyde proton HꢀC(11) appeared as singlet rather than as a doublet.
Therefore, this phytochemical could be unequivocally identified as the C9ꢀ
hydroxylated polygodial derivative warburganal (3, Figure 1). To the best of our
knowldege, compound 3 has not yet been reported in Tasmanian Pepper, although it
was found in other Warburgia species42,43 as well as water pepper (Polygonum
hydropiper).46
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
Compound 4 showed the typical NMR signal pattern of unsaturated drimane
dialdehydes, however, with modifications at the Aꢀring when compared to 1 indicated
by two additional quaternary, olefinic carbons resonating at 124.2 and 125.1 ppm and
coupling with two methyl groups H3ꢀC(13)/H3ꢀC(14) at 1.61/15.4 ppm and 1.63/19.0
ppm respectively as well as a methine proton HꢀC(1) shifted to the lower field at 4.89
ppm. A molecular formula of C17H22O4 (m/z = 291.1615 Da) showing a mass loss of
60 Da in the mass spectrum indicated an acetyl cleavage which, along with HMBC
couplings, identified the compound as 1βꢀacetoxyꢀ9ꢀdeoxyꢀisomuzigadial (4, Figure
1). While 9ꢀdeoxyisomuzigadial was isolated previously from Canella winterana,47 the
acetylated derivative 4 has yet not been reported in literature.
1
Compound 5 also showed an aldehyde substructure with a characteristic H
NMR chemical shift of 9.48 ppm, which showed HMBC connectivity to the carbon
atoms C(7), C(8) and C(9), and could be assigned to proton HꢀC(11). The molecular
formular C14H22O2 suggested a norsesquiterpene backbone as confirmed by 13C
NMR. Lowꢀfield shifted signals of a methine group (HꢀC(9)/3.85 ppm, C(9)/70.3 ppm)
and ROESY couplings between HꢀC(9) and the methyl group H3ꢀC(14) revealed the
identification of the previously not reported hydroxylated drimane derivative 5 (Figure
13
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 14 of 48
1
2
1). Consistent with the structurally related changweikang acid A, reported as natural
product in Polygonum hydropiper and Daphniphyllum calcynicum,48 compound 5 was
coined changweikang aldehyde.
3
1
4
Compound 6, showing a monoaldehyde motif by its characteristic H NMR
5
resonance signal at 9.33 ppm, revealed a sequential mass loss of 18 and 28 Da from
the molecular ion m/z 251 in LCꢀMS analysis, thus indicating the presence of a
carboxylic acid as additional functional group. In comparison to 1, NMR analysis
showed one of the aldehyde moieties to be oxidised to the carboxylic acid. In
particular, the absence of a COSY coupling to another proton leads to the
assignment of the aldehyde at C(12) and confirmed the structure as polygonic acid
(6, Figure 1), which has been reported earlier in water pepper Polygonum
hydropiper.49
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
Next to the aldehydes 1 and 3ꢀ6, NMR and LCꢀMS experiments led to the
identification of the isomeric hydroxylated carboxylic acids 7 ꢀ 9 as well as the
carboxylic acid 10 showing only a C13 carbon backbone. High resolution MS and 2Dꢀ
NMR experiments demonstrated compounds 7ꢀ9 to be the drimane
norsesquiterpenes polypiperic acid (7), epipolypiperic acid (8) and changweikang
acid A (9) sharing the same decaline backbone and differing only in the positions and
stereochemistry of their hydroxy and carboxy groups that were assigned based on
specific coupling patterns in HMBC and ROESY experiments. MS analysis of
compound 10 revealed a molecular formula of C13H20O2 and showed a cleavage of
water and CO2 fragments, thus indicating the presence of a carboxylic acid function.
This was confirmed by 13CꢀNMR showing a resonance at the low field of 170.34 ppm
for C(8). NMR also revealed strong similarities with other isolated drimanes and 2Dꢀ
NMR experiments confirmed that the compound is an indene derivative with ring A
being the identical to the other drimanes and ring B being an unsaturated 5ꢀ
14
ACS Paragon Plus Environment

Page 15 of 48
Journal of Agricultural and Food Chemistry
1
membered ring. HMBC couplings from the olefinic signal HꢀC(2) at 6.96 ppm to the
carboxylic group and all carbon signals of ring B as well as from the olefinic carbon
signal C(3) at 146.6 ppm into both rings A and B led to the identification of 10 as
3a,4,5,6,7,7aꢀHexahydroꢀ3a,7,7ꢀtrimethylꢀ1Hꢀindenꢀ3ꢀcarboxylic acid (Figure 1). To
the best of our knowledge, the carboxylic acids 7ꢀ10 are reported in Tasmanian
Pepper for the first time. Changweikang acid A (9) was previously identified in the
Chinese medicine Changweikang,48 polypiperic acid (7) was previously reported as
the corresponding methylester in Drimys granadensis50 and Polygonum hydropiper,51
whereas phytochemicals 8 and 10 have not been described in literature.
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
Moreover, nine drimane sesquiterpenes (11ꢀ19) were identified to share a
lactone structure. Compound 11, matching the molecular mass of polygodial (234
Da), was identified as the lactone Cinnamolide (11, Figure 1) with the characteristic
NMR resonance signal of 170.3 ppm for the carbonyl group C(12). HMBC
correlations of the carbonyl carbon to the olefinic proton HꢀC(7) at 6.87 ppm and to a
methylene group H2ꢀC(11) at 4.03 / 4.37 ppm led to the unequivocal identification of
the lactone structure at ring B. Although cinnamolide was first identified in
Cinnamosma fragrans52 and later in other species,44,53 this is the first report on 11 as
a phytochemical in Tasmanian pepper. Compounds 12ꢀ17 showed molecular masses
of 248, 250 and 266 Da respectively and the mass differences of 14, 16, and 32 amu
compared to cinnamolide (11) indicated ketoꢀ, monoꢀ and dihydroxy derivatives,
respectively. This was confirmed by 2DꢀNMR experiments which led to the
identification of 7ꢀketoconfertifolin (12), dendocarbin A (13), dendocarbin L (14),
dendocarbin M (15), ugandenial A (16) and fuegin (17), the structures of which are
displayed in Figure 1. The NMR signal pattern of further additional compounds (18,
19) was rather similar to the other constituents but appeared to consist of two
additional carbon atoms with chemical shifts as expected for acetyl moieties. This
15
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 16 of 48
1
2
was susbtantiated by 2DꢀNMR and the cleavage of 60 amu fragments from the
molecular ions and, thus, confirmed the target compounds as 3ꢀβꢀacetoxyꢀ7ꢀ
ketoconfertifolin (18) and 3ꢀβꢀacetoxydrimenin (19), respectively. None of the
lactones 11 – 19 have been reported earlier in Tasmanian pepper, although
compounds 11 – 17 and 19 have been reported as phytochemicals in other
botanical.13,54ꢀ57 3ꢀβꢀAcetoxyꢀ7ꢀketoconfertifolin (18) has to the best of our knowledge
not yet been reported in literature. Finally, the UVꢀinactive compound 20 could be
detected by evaporative light scattering detection in fraction F9 and was identified as
the 11,12ꢀdihydroxy derivative of polygodial, named drimendiol (20, Figure 1), by
comparison with literature data.43 Although drimendiol was reported earlier in Drimys
winteri,56 this is the first report on its occurrence in Tasmanian pepper.
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
Human Sensory Activity. Compounds 1, 3ꢀ7, 10, 11, 13ꢀ18, and 20 could be
isolated in sufficient quantities and purities (>98% HPLCꢀELSD) to determine
orosensory recognition threshold concentrations by means of an halfꢀtongue test
using filterꢀpaper vehicles as reported recently.35 In addition, epipolygodial (2) was
synthetically prepared from polygodial (1) and used for sensory studies. The lowest
orosensory thresholds of 0.6, 1.9, and 2.8 nm/cm2 were found for polygodial (1),
warburganal (3), followed by 1βꢀacetoxyꢀ9ꢀdeoxyꢀisomuzigadial (4), whereas
epipolygodial (2) and 3βꢀacetoxyꢀ7ꢀketoconfertifolin (18) showed 14 and 25 times
higher threshold concentrations when compared to 1 (Table 1). The data found for 1ꢀ
3 were in good agreement with the threshold concentration of 0.4 (1), 21.4 (2) and
2.0 nmol/tongue (3) reported earlier,59 and are in a similar order of magnitude as
found for piperine (3 nmol/cm2),35 the main pungent principle in black pepper.
Comparatively, high threshold levels of 30ꢀ70 nmol/cm2 were found for changweikang
aldehyde (5) and polygonic acid (6), which both have only one aldehyde function, as
well as for the hydroxylactones fuegin (17), cinnamolide (11), ugendenial A (16), and
16
ACS Paragon Plus Environment

Page 17 of 48
Journal of Agricultural and Food Chemistry
1
dendocarbins A, L and M (13, 14, 15). Drimendiol (20) and 3a,4,5,6,7,7aꢀhexahydroꢀ
3a,7,7ꢀtrimethylꢀ1Hꢀindenꢀ3ꢀcarboxylic acid (10) did not induce any pungent
sensation up to a maximum test concentration of 175 nm/cm2. These findings clearly
demonstrated the α,βꢀunsaturated 1,4ꢀdialdehyde structure to be the required
structural motif for a low orosensory threshold for pungency, thus confirming earlier
proposals.24,34,61,62 However, some of the isolated phytochemicals reported here,
such as, e.g. changweikang aldehyde (5) and polygonic acid (6), were also found to
induce a clear pungent orosensation, although at somewhat higher concentration
levels. In order to gain some insight into the mechanisms as to how the pungency of
these compounds is mediated, cellꢀbased TRPA1 and TRPV1 receptor studies were
conducted in the following as these ion channels, amongst other functions, are
known to play a crucial role for the perception of temperature, pain and pungency.63
Activation of Transient Receptor Potential Channels A1 (TRPA1) and V1
(TRPV1). While the vanilloid receptor TRPV1 is reported to be activated by pungent
compounds like capsaicin from chillies64 and piperine from black pepper,65 a wide
range of stimuli from noxious cold via environmental irritants up to diverse natural
products like polygodial are known agonists of the ion channel TRPA1.30,34,39,66ꢀ69
Therefore, selected phytochemicals isolated from Tasmanian pepper were screened
in cellꢀbased calcium influx experiments for their ability to activate TRPA1 and
TRPV1, respectively. In a second set of experiments, intracellular calcium imaging
experiments were performed using murine trigeminal neurons.
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
Using tetracycline induced mTRPA1 expressing CHO cells37 loaded with FURAꢀ
2AM as a calcium fluorescent dye, TRPA1 activation was investigated by FURAꢀ2
based calcium multiꢀwell plate assays as described earlier.42 Among all tested
compounds, the lowꢀthreshold dialdehydes polygodial (1), epipolygodial (2),
warburganal (3), and 1βꢀacetoxyꢀ9ꢀdeoxyꢀisomuzigadial (4), as well as the
17
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 18 of 48
1
2
monoaldehyde changweikang aldehyde (5) elicited a significant calcium influx into
the cells in a doseꢀdependent fashion (Figure 3). Polygonic acid (6), which exhibits
one aldehyde moiety additionally to its carboxylic function, and to a lesser extent the
hydroxylactones dendocarbin A (13), dendocarbins L/M (14/15) and ugandenial A
(16), evoked a calcium influx at the highest concentrations tested. Intringuingly, all
remaining compounds (7, 10, 11, 17, 18) did not show any significant calcium influx
at the tested concentration range to indicate a TRPA1 mediated mechanism for
pungency. The calcium influx observed for drimendiol (20) at a concentration of 200
µmol/L was identified as being unspecific and not related to TRPA1 activation as
control measurements with nonꢀtransfected CHO cells also showed a calcium
response (data not shown).
3
4
5
6
7
8
9
10
11
12
13
EC50 values for the TRPA1 response were calculated for the most potent agonists
from the log(concentration) response relationships fitted with a logistic function (
18
ACS Paragon Plus Environment

Page 19 of 48
Journal of Agricultural and Food Chemistry
1
Table 2). The compounds with a dialdehyde structure showed EC50 values
between 4.5 and 34.5 µM, which are significantly lower than the EC50 value
determined for the monoaldehyde changweikang aldehyde (340 µmol/L). The EC50
values of the dialdehydes are highly correlated with their oral recognition thresholds
(R2 = 0.986, linear regression). This suggests that the perception of pungency of the
dialdehydes is directly mediated via the TRPA1. The approximate EC50 value for
changweikang aldehyde is outside of this linear correlation and suggests that the
activation mechanism of the monoaldehyde is different.
2
3
4
5
6
7
8
9
The TRPV1 activation by the isolated drimanes was investigated in the same
fashion as the TRPA1 using a calcium imaging with rTRPV1 transfected CHO cells.
The majority of compounds showed no activation at the tested concentration ranges,
but the dialdehydes polygodial (1), warburganal (3) and 1βꢀacetoxyꢀ9ꢀdeoxyꢀ
isomuzigadial (4) showed some activation (data not shown). However, the
amplitudes of the responses were negligible and responses were only observed at
the highest concentrations tested, except for polygodial (1), which was the only test
compound to cause a significant calcium influx. Nevertheless, this was only observed
at concentrations approximately ten times higher than required for TRPA1 activation.
Therefore, it can be concluded that the TRPV1 activation has a minor contribution to
the perception of pungency for the investigated drimanes. However, it is possible that
the TRPV1 activation contributes a small component to the overall perceived
pungency in vivo.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
In a second set of experiments, the compounds showing TRPA1 activation in
the CHO model system were further studied using calcium imaging assays with
trigeminal neurons prepared from wildtype (TRPA1+/+) and TRPA1 knockout mice
(TRPA1ꢀ/ꢀ), respectively. Calcium influx into the cultured neurons loaded with FURA2ꢀ
26 AM was monitored by fluorescence microscopy. All dialdehydes 1ꢀ4 triggered
19
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 20 of 48
1
2
calcium influx into neurons prepared from Wildꢀtype mice (Figure 4 A, C, E, G),
whereas no calcium influx was observed in neurons from TRPA1ꢀ/ꢀ mice (Figure 4 B,
D, F, H). To further characterize the subpopulation of neurons activated by
dialdehydes 1ꢀ4 in trigeminal neurons from TRPA1+/+ mice, the same neurons were
challenged with allyl isothiocyanate (AITC; 50 ꢁmol/L) and capsaicin (1 ꢁmol/L) to
selectively stimulate and visualize TRPA1 and TRPV1 expressing neurons,
respectively, as well as potassium chloride (50 mmol/L) to activate all neurons.70 As
expected, responses to capsaicin and potassium chloride remained unchanged,
while the population of neurons activated by the aldehydes 1ꢀ4 was identical to that
activated by the TRPA1 agonist AITC. Microperfusion with dialdehydes 1ꢀ4 triggered
immediate calcium influx during the application of the compounds to neurons from
TRPA1+/+ mice as shown as an example for 1βꢀacetoxyꢀ9ꢀdeoxyꢀisomuzigadial (4) in
Figure 5 A, while no activation was found in neurons from TRPA1ꢀ/ꢀ mice (Figure 5 B).
3
4
5
6
7
8
9
10
11
12
13
14 Again, responses to capsaicin and potassium chloride remained unchanged, while
15
16
17
18
19
20
21
22
23
24
25
26
the neuron activated by dialdehyde 4 was also activated by AITC. These data
unequivocally confirm that the dialdehydes not only activate heterologously
expressed TRPA1 channels, but also native ion channels in murine trigeminal
neurons.
Also changweikang aldehyde (5) activated the same neuron population in
Wildtype neurons but, interestingly, evoked a different and interesting response.
Whilst microperfusion with dialdehydes 1ꢀ4 triggered calcium influx during the
application of the compounds, no calcium influx was observed during perfusion of
neurons from TRPA1+/+ mice with changweikang aldehyde (5) (Figure 5 C). Instead,
a rapid and significant calcium influx was observed when the neurons were washed
with buffer solution after perfusion of changweikang aldehyde at concentrations ≥240
µmol/L. This effect was repeatable and calcium levels decreased during subsequent
20
ACS Paragon Plus Environment

Page 21 of 48
Journal of Agricultural and Food Chemistry
1
administration of changweikang aldehyde (5). When the aldehyde was applied at 120
µmol/L, no effect was observed, neither during perfusion, nor upon washout. In
comparison, neurons from TRPA1ꢀ/ꢀ mice did not respond to 5, neither during
administration, nor during washout of 5 (Figure 5 D). These data for the first time
suggest changweikang aldehyde (5) to have a bimodal activating and inhibitory
function on the TRPA1 channel due to two distinct ligand interaction sites at the ion
channel with the activating site having a stronger binding affinity for 5 than the
inhibitory site. Such bimodal effects on the TRPA1 have been observed for other
natural compounds such as menthol,71 cinnamaldehyde and camphor.72 The
phenomenon of having both TRPA1 inhibiting and activating properties may well
explain why not all drimane sesquiterpenes, which were perceived as pungent in the
sensory studies, evoked significant TRPA1 or TRPV1 mediated in vitro responses.
Model Studies on the Binding of Drimane Sesquiterpenes to Nucleophilic
Amino Acid Residues of TRPA1. Previous studies attributed the TRPA1ꢀmediated
pungency of electrophilic pungent compounds, such as, e.g. isothiocyanates like
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16 AITC and disulfides like diallyldisulfide, to a reversible covalent binding to cysteine
residues in a complex reactive cysteine profile of the TRPA1 receptor protein,29,30,74
17
18 Amongst others, cysteine C621 had been suggested to be most critical for
19
20
21
22
23
24
25
electrophileꢀbinding and activation of TRPA1 but also that lysine residues, in
particular lysine K620, play a critical role for receptor activation.75 In comparison,
dialdeyhde sesquiterpenes are suggested to follow a different activation mechanism
than isothiocyanates and sulfides as polygodial retained full activity on the triple
cysteineꢀlysine TRPA1ꢀmutant.34 Although chemical reactions of polygodial with
61, 73
nucleophiles such as, e.g. amino and mercapto groups,23ꢀ28,
proposed, defined reaction products could not yet be identified.
have been
21
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 22 of 48
1
2
To study potential reactions between proposed nucleophilic amino acid side
chains of TRPA1 and drimane dialdehydes, timeꢀresolved NMRꢀspectroscopic
experiments were conducted on binary mixtures of polygodial and Nαꢀacetylꢀcysteine,
Nαꢀacetylꢀlysine, and Nαꢀacetylꢀarginine, respectively. The signal of the aldehyde
proton HꢀC(11) of polygodial was integrated relative to the TMS signal at each time
point. No reaction was observed between polygodial and Nαꢀacetylꢀcysteine or Nαꢀ
acetylꢀarginine (Figure 6, A), respectively, thus contradicting the hypothesized
reaction of polygodial with cysteine residues of TRPA1.12,22,27 In comparison, a rapid
degradation of polygodial was observed when incubated in the presence of Nαꢀ
acetylꢀlysine (Figure 6, A), e.g. the integrals of both aldehyde protons HꢀC(11) and Hꢀ
C(12) as well as the olefinic proton HꢀC(7) were decreased to 50% within the first
seven minutes of incubation (Figure 7). After 60 min, not even traces of polygodial
were detectable anymore (Figure 6, A), while a series of new proton resonance
signals corresponding to three reaction products were observed (Figure 7).
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
The first reaction product showed a proton singlet at 8.71 ppm for HꢀC(12), a
multiplet at 7.27 ppm for HꢀC(7), and a doublet at 5.75 ppm for HꢀC(11) and was
assigned as the cationic pyrrolinium ion 21 (Figure 7) by comparing the NMR data
23
18
19
20
21
22
23
24
25
26
with those reported for the reaction product of 1,4ꢀdialdehydes and methylamine.
The intermediate 21, formed upon nucleophilic reaction of polygodial and the lysine
side chain, was found to be formed very fast and, after running through a maximum
after 3 min, disappeared again 40 min after incubation (Figure 6, B). With increasing
incubation time and depletion of 21, two additional reaction products were formed
(Figure 6, B) and were identified as the two isomeric pyrroles 22 and 23 showing
characteristic resonance signals at 6.41/6.61 ppm and 6.67/6.45 ppm for the pairs of
aromatic pyrrole protons of 22 and 23, respectively (Figure 7). In addition, HꢀC(7) of
22 resonated at 4.42 ppm as a multiplet with small coupling constants of 1.9 and 4.1
22
ACS Paragon Plus Environment

Page 23 of 48
Journal of Agricultural and Food Chemistry
1
Hz, while the multiplet of HꢀC(7) of 23, resonating at 4.54 ppm, showed coupling
3
3
2
constants of 7.4 ( JHꢀC(6b)/HꢀC(7b)) und 9.4 Hz ( JHꢀC(6a)/HꢀC(7b)), thus indicating a 180°
3
4
angle between the vicinal protons HꢀC(7b) and HꢀC(6a) in this isomer (Figure 7).
To study the influence of the drimane structure on the reactivity with the εꢀamino
group of lysine, Nαꢀacetylꢀlysine was incubated with epipolygodial (2), warburganal
(3), 1βꢀacetoxyꢀ9ꢀdeoxyꢀisomuzigadial (4), and changweikang aldehyde (5),
respectively, followed by timeꢀresolved 1H NMRꢀspectrocopy (Figure 8). Both,
warburganal (3) and 1βꢀacetoxyꢀ9ꢀdeoxyꢀisomuzigadial (4) reacted quickly with Nαꢀ
acetylꢀlysine similarly to polygodial with somewhat slower reaction observed for 3,
most likely due to the additional hydroxylation at C9 (Figure , A). These data are well
in line with the lowest pungency threshold concentrations found for 1, 3 and 4 within
the very narrow range of 0.6 to 2.8 nmol/cm2. In comparison, the dialdehyde
epipolygodial (2) and the monoaldehyde changweikangaldehyde (5), showing
significantly higher pungency threshold concentrations of 8.6 and 27.0 nmol/cm2,
respectively, showed a much slower reaction with Nαꢀacetylꢀlysine, e.g. 20% of 2 and
5 had been reacted after 300 and 5500 min of incubation, respectively (Error!
Reference source not found. B and C). Based on these data, it may be concluded
that, due to the long reaction times, a reaction with lysine residues of the TRPA1
receptor is unlikely to be relevant for the pungent properties of epipolygodial (2) and
changweikang aldehyde (5). However, there is the possibility that the reaction is
catalyzed in vivo by stabilizing intermediate reaction products by cooperativity with
other amino acid side chains in TRPA1.
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
In summary, among the 20 drimane sesquiterpens isolated from Tasmanian
pepper berries, polygodial (1), warburganal (3), and 1βꢀacetoxyꢀ9ꢀdeoxyꢀ
isomuzigadial (4) exhibited the lowest pungency threshold concentrations within the
very narrow range of 0.6 to 2.8 nmol/cm2, elicited a doseꢀdependent calcium influx
23
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 24 of 48
1
2
into mTRPA1 expressing CHO cells with EC50 values between 4.5 ± 1.0 and 16.7 ±
7.5 ꢁmol/L (correlation to oral pungency thresholds: R2 = 0.986, linear regression)
and into cultured trigeminal neurons prepared from wildtype (TRPA1+/+) mice,
whereas no calcium influx was observed in neurons from TRPA1 knockout mice
(TRPA1ꢀ/ꢀ). These findings clearly confirmed the α,βꢀunsaturated 1,4ꢀdialdehyde
3
4
5
6
structure to be the required structural motif for a low orosensory threshold for
24,34,61,62
7
pungency,
and to activate TRPA1.12,21,22,31,34 Timeꢀresolved NMR
8
experiments proposed the pungency mediating mechanism for electrophilic drimane
sesquiterpene dialdehydes to be different from that found for other electrophilic
pungent agents like isothiocyanates, which have been shown to undergo a covalent
binding with cysteine residues in TRPA1.29,30 Instead, the highꢀimpact
chemosensates polygodial (1), warburganal (3), and 1βꢀacetoxyꢀ9ꢀdeoxyꢀ
isomuzigadial (4) showed immediate reactivity with the εꢀamino group of lysine side
chains to give pyrroleꢀtype conjugates (22, 23), thus showing evidence for TRPA1
activation by covalent lysine modification.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
ACKNOWLEDGEMENT
This work was supported in part by a grant from the TUM Graduate School. We
thank Jack Bircher and Clive Gentry for their support with cell culture and preparation
of neurons.
Supporting Information Available
Spectroscopic data on compounds 1 ꢀ 20. This material is available free of charge
via the Internet at http://pubs.acs.org.
24
ACS Paragon Plus Environment

Page 25 of 48
Journal of Agricultural and Food Chemistry
1
2
25
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 26 of 48
1
2
3
4
LITERATURE CITED
(1) Southwell, I. A.; Brophy, J. J. Differentiation within the Australian Tasmannia by
essential oil comparison, Phytochemistry. 1992, 31,3073–3081.
5
6
(2) Ahmed, A. K.; Johnson, K. A. TURNER REVIEW No. 3. Horticultural
development of Australian native edible plants, Aust. J. Bot. 2000, 48, 417.
7
8
9
(3) Konczak, I.; Zabaras, D.; Dunstan, M.; Aguas, P. Antioxidant capacity and
phenolic compounds in commercially grown native Australian herbs and spices,
Food Chem. 2010, 122, 260–266.
10
11
12
(4) Netzel, M.; Netzel, G.; Tian, Q.; Schwartz, S.; Konczak, I. Sources of
antioxidant activity in Australian native fruits. Identification and quantification of
anthocyanins, J. Agric. Food Chem. 2006, 54, 9820–9826.
13
14
15
(5) Netzel, M.; Netzel, G.; Tian, Q.; Schwartz, S.; Konczak, I. Native Australian
fruits ꢀ a novel source of antioxidants for food, Innov. Food Sci. Emerg. Technol.
2007, 8, 339–346.
16
17
18
(6) Menary, R. C. Mountain pepper extract, Tasmannia lanceolata. Quality
stabilisation and registration: a report for the Rural Industries Research and
Development Corporation; RIRDC: Barton, A.C.T, 2003.
19
20
(7) Loder, J. W. Occurence of the sesquiterpenes polygodial and guaiol in the
leaves of Drimys lanceolata (Poir.) Baill., Aust. J. Chem. 1962, 15, 389–390.
21
22
23
(8) Dragar, V. A.; Garland, S. M.; Menary, R. C. Investigation of the variation in
chemical composition of Tasmannia lanceolata solvent extracts, J. Agric. Food
Chem. 1998, 46, 3210–3213.
24
25
(9) Barnes, C. S.; Loder, J. W. The structure of polygodial: A new sesquiterpene
dialdehyde from Polygonum hydropiper L., Aust. J. Chem. 1962, 15, 322–327.
26
ACS Paragon Plus Environment

Page 27 of 48
Journal of Agricultural and Food Chemistry
1
(10) Jonassohn, M. Sesquiterpenoid unsaturated dialdehydes. Structural properties
that affect reactivity and bioactivity; PhD Thesis, Lund, Sweden, 1996.
2
3
4
(11) Jansen, B.; de Groot, A. The occurrence and biological activity of drimane
sesquiterpenoids, 1990.
5
6
(12) Kubo, I.; Ganjian, I. Insect antifeedant terpenes, hot tasting to humans,
Experientia. 1981, 37, 1063–1064.
7
8
9
(13) Sakio, Y.; Hirano, Y. J.; Hayashi, M.; Komiyama, K.; Ishibashi, M. Dendocarbins
AꢀN, new drimane sesquiterpenes from the nudibranch Dendrodoris
carbunculosa, J. Nat. Prod. 2001, 64, 726–731.
10
11
12
(14) Cimino, G.; de Rosa, S.; de Stefano, S.; Sodano, G.; Villani, G. Dorid
Nudibranch elaborates its own chemical defense, Science. 1983, 219, 1237–
1238.
13
14
15
(15) Anke, H.; Sterner, O. Comparison of the antimicrobial and cytotoxic activities of
twenty unsaturated sesquiterpene dialdehydes from plants and Mmshrooms,
Planta Med. 1991, 57, 344–346.
16
17
18
(16) Brennan, N. J.; Larsen, L.; Lorimer, S. D.; Perry, N. B.; Chapin, E. L.; Werk, T.
L.; Henry, M. J.; Hahn, D. R. Fungicidal sesquiterpene dialdehyde cinnamates
from Pseudowintera axillaris, J. Agric. Food Chem. 2006, 54, 468–473.
19
20
21
(17) McCallion, R. F.; Cole, A. L. J.; Walker, J. R. L.; Blunt, J. W.; Munro, M. H. G.
Antibiotic substances from New Zealand plants. II. Polygodial, an antiꢀcandida
agent from Pseudowintera colorata, JMPR. 1982, 44, 134–138.
22
23
24
(18) Gerard, P. J.; Perry, N. B.; Ruf, L. D.; Foster, L. M. Antifeedant and insecticidal
activity of compounds from Pseudowintera colorata (Winteraceae) on the
webbing clothes moth, Tineola bisselliella (Lepidoptera: Tineidae) and the
27
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 28 of 48
1
2
Australian carpet beetle, Anthrenocerus australis (Coleoptera: Dermestidae),
Bull. Entomol. Res. 1993, 83, 547–552.
3
4
5
6
(19) Malheiros, A.; CechinelꢀFilho, V.; Schmitt, C. B.; Santos, A. R. S.; Scheidt, C.;
Calixto, J. B.; Monache, F. D.; Yunes, R. A. A sesquiterpene drimane with
antinociceptive activity from Drimys winteri bark, Phytochem. 2001, 57, 103–
107.
7
8
(20) Mendes, G. L.; Santos, A. R. S.; Campos, M. M.; Tratsk, K. S.; Yunes, R. A.;
CechinelꢀFilho, V.; Calixto, J. B. Antiꢀhyperalgesic properties of the extract and
of the main sesquiterpene polygodial isolated from the barks of Drimys winteri
(Winteraceae), Life Sci. 1998, 63, 369–381.
9
10
11
12
13
(21) Ying, B.ꢀP.; Peiser, G. D.; Ji, Y.ꢀY.; Mathias, K. M.; Karasina, F.; Hwang, Y.ꢀS.
Structureꢀactivity relationships of phytotoxic sesquiterpenoids from Canella
winterana, J. Agric. Food Chem. 1995, 43, 826–829.
14
15
16
(22) Taniguchi, M.; Adachi, T.; Oi, S.; Kimura, A.; Katsumura, S.; Isoe, S.; Kubo, I.
Structureꢀactivity relationship of the Warburgia sesquiterpene dialdehydes,
Agric. Biol. Chem. 1984, 48, 73–78.
17
18
19
(23) Cimino, G.; Spinella, A.; Sodano, G. Identification of an intermediate in the
reaction between polygodial and methylamine in biomimetic conditions,
Tetrahedr. Lett. 1984, 25, 4151–4152.
20
21
22
23
(24) Cimino, G.; Sodano, G.; Spinella, A. Correlation of the reactivity of 1,4ꢀ
dialdehydes with methylamine in biomimetic conditions to their hot
taste: Covalent binding to primary amines as a molecular mechanism in hot
taste receptors, Tetrahedron. 1987, 43, 5401–5410.
28
ACS Paragon Plus Environment

Page 29 of 48
Journal of Agricultural and Food Chemistry
1
(25) D'Ischia, M.; Prota, G.; Sodano, G. Reaction of polygodial with primary amines:
An alternative explanation to the antifeedant activity, Tetrahedr. Lett. 1982, 23,
3295–3298.
2
3
4
5
6
(26) Gustafsson, J.; Jonassohn, M.; Kahnberg, P.; Anke, H.; Sterner, O. The
reactivity of the antibiotic sesquiterpene isovelleral towards primary amines,
Nat. Prod. Lett. 1997, 9, 253–258.
7
8
9
(27) Taniguchi, M.; Adachi, T.; Haraguchi, H.; Oi, S.; Kubo, I. Physiological activity of
warburganal and its reactivity with sulfhydryl groups, J. Biochem. 1983, 94,
149–154.
10
11
12
13
(28) Fritz, G. L.; Mills, G. D.; Warthen, D.; Waters, R. M. ReimerꢀTiemann adducts
as potential insect antifeedant agents. Reviewing the structureꢀactivity
relationship theory of the antifeedant, warburganal, J. Chem. Ecology. 1989, 15,
2607–2623.
14
15
(29) Hinman, A.; Chuang, H.ꢀh.; Bautista, D. M.; Julius, D. TRP channel activation by
reversible covalent modification, PNAS. 2006, 103, 19564–19568.
16
17
18
(30) Macpherson, L. J.; Dubin, A. E.; Evans, M. J.; Marr, F.; Schultz, P. G.; Cravatt,
B. F.; Patapoutian, A. Noxious compounds activate TRPA1 ion channels
through covalent modification of cysteines, Nature. 2007, 445, 541–545.
19
20
21
(31) Baraldi, P. G.; Preti, D.; Materazzi, S.; Geppetti, P. Transient Receptor Potential
Ankyrin 1 (TRPA1) Channel as emerging target for novel analgesics and antiꢀ
inflammatory agents, J. Med. Chem. 2010, 53, 5085–5107.
22
23
24
25
(32) Beltrán, L. R.; Dawid, C.; Beltrán, M.; Gisselmann, G.; Degenhardt, K.; Mathie,
K.; Hofmann, T.; Hatt, H. The pungent substances piperine, capsaicin, 6ꢀ
gingerol and polygodial inhibit the human twoꢀpore domain potassium channels
TASKꢀ1, TASKꢀ3 and TRESK. Front. Pharm. 2013, 4(141), 1–11.
29
ACS Paragon Plus Environment