The thia-Michael reactivity of zerumbone and related cross-conjugated dienones: Disentangling stoichiometry, regiochemistry, and addition mode with an NMR-spectroscopy-based cysteamine assay

Article abstract of DOI:10.1002/ejoc.201500237

The cross-conjugated and electrophilic dienone system of the humulane sesquiterpene zerumbone (1a) was modified by E/Z photochemical isomerization and/or by removal of homoconjugation with the isolated endocyclic double bond of the medium-sized ring. The site (C-6/C-9), mode (transient or irreversible), stoichiometry (single or twofold), and comparative rates of thiol addition were evaluated using an NMR-spectroscopy-based cysteamine assay. Dramatic effects were seen, and this highlights the subtleties of the reaction and the limitations of our predictive power in this field. For biological endpoints sensitive to thiol trapping, a substantial separation between Michael reactivity and biological activity was found for 1a and its analogues. This supports the view that shape complementarity plays a critical role in the covalent binding of Michael acceptors to their macromolecular target(s). Using the NMR-spectroscopy-based cysteamine assay, the thia-Michael reactivity of the cross-conjugated dienone zerumbone and its photoisomers and epoxides was investigated, and marked differences were found. For biological endpoints sensitive to Michael acceptors, a substantial separation between reactivity and biological activity was observed.

Full text of DOI:10.1002/ejoc.201500237

Job/Unit: O50237
/KAP1
Date: 05-05-15 12:22:36
Pages: 7
FULL PAPER
DOI: 10.1002/ejoc.201500237
The Thia-Michael Reactivity of Zerumbone and Related Cross-Conjugated
Dienones: Disentangling Stoichiometry, Regiochemistry, and Addition Mode
with an NMR-Spectroscopy-Based Cysteamine Assay
Giovanni Appendino,*[a] Alberto Minassi,[a] Juan A. Collado,[b] Federica Pollastro,[a]
Giuseppina Chianese,[c] Orazio Taglialatela-Scafati,[c] Mehdi Ayyari,[d] Victor Garcia,[b] and
Eduardo Muñoz[b]
Dedicated to Professor Iwao Ojima on occasion of his 70th birthday
Keywords: Natural products / Michael addition / Enones / Thiols / Medium-ring compounds / Biological activity /
Structure-activity relationships
The cross-conjugated and electrophilic dienone system of the
humulane sesquiterpene zerumbone (1a) was modified by
seen, and this highlights the subtleties of the reaction and the
limitations of our predictive power in this field. For biological
E/Z photochemical isomerization and/or by removal of homo- endpoints sensitive to thiol trapping, a substantial separation
conjugation with the isolated endocyclic double bond of the
medium-sized ring. The site (C-6/C-9), mode (transient or
irreversible), stoichiometry (single or twofold), and compara-
tive rates of thiol addition were evaluated using an NMR-
spectroscopy-based cysteamine assay. Dramatic effects were
between Michael reactivity and biological activity was found
for 1a and its analogues. This supports the view that shape
complementarity plays a critical role in the covalent binding
of Michael acceptors to their macromolecular target(s).
Zingiber zerumbet Smith), an oriental spice and medicinal
plant,[3] and it has been used as a starting material for the
semisynthesis of valuable natural products, as well as in
diversity-oriented synthesis.[4]
Introduction
Because of concerns over nonspecific toxicity and lack
of selectivity, the Michael acceptor motif is rarely intro-
duced into drug leads by design.[1] Paradoxically, our diet
is rife with Michael acceptors, and food plants provide
countless leads to investigate the biological role of Michael
reactivity in molecules that are substantially devoid of tox-
icity at dietary dosages.[2] The sesquiterpene zerumbone
(1a) represents an interesting compound to investigate the
chemical and biological subtleties of the Michael reaction.
Zerumbone is easily available in high yield (up to Ͼ4% on
a dry-weight basis) from lampuyang (shampoo ginger,
[a] Dipartimento di Scienze del Farmaco,
via Bovio 6, 28100 Novara, Italy
E-mail: appendino@pharm.unipmn.it
https://www.pharm.unipmn.it/it
[b] IMIBIC, Reina Sofía University Hospital, University of
Córdoba, Department of Cell Biology, Physiology, and
Immunology,
Avda Menéndez Pidal s/n, 14004 Córdoba, Spain
[c] Dipartimento di Farmacia, Università di Napoli Federico II,
Via Montesano 49, 80131 Napoli, Italy
[d] Department of Phytochemistry, Medicinal Plants and Drugs
Research Institute, Shahid Beheshti University,
G. C., Evin 1983963113, Tehran, Iran
The covalent interaction of zerumbone with proteins in-
volved in cell proliferation and defence against oxidative/
electrophilic stress makes it a potential candidate for cancer
Supporting information for this article is available on the
WWW under http://dx.doi.org/10.1002/ejoc.201500237.
Eur. J. Org. Chem. 0000, 0–0
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1

Job/Unit: O50237
/KAP1
Date: 05-05-15 12:22:36
Pages: 7
G. Appendino et al.
FULL PAPER
prevention and treatment.[5] The Michael reactivity of to have a value of ca. 12 Hz in 1b and 1c, consistent with a
zerumbone is actually surprising, since cross-conjugated Z configuration, and ca. 16 Hz in 1d and zerumbone (1a),
cyclohexadienones lack thiol-trapping properties.[6] This pointing to an E configuration. On the other hand, the
suggests a specific contribution from the medium-sized-ring allylic methyl group at C-7 (Me-13) showed a ROESY
environment. Furthermore, the mode of the thiol addition cross-peak with 6-H in 1c and 1d, while in zerumbone and
to zerumbone is also puzzling. In situ NMR spectroscopy 9-isozerumbone (1b) this methyl group was correlated with
experiments have shown that the addition to the disubsti- the C-5 methylene group. Thus, the cross-conjugated trisub-
tuted Δ9 double bond is transient, reversing upon a solvent stituted double bond (Δ6) was E in 1a and 1b, and Z in 1c
switch from DMSO to CDCl3, whereas the trisubstituted and 1d. In a similar way, the other allylic methyl group (Me-
Δ6 double bond reacts in an irreversible way.[6] All Michael 12) showed a ROESY cross-peak with the C-1 methylene
reactions are fundamentally reversible and susceptible to group in zerumbone (1a), 1b, and 1c, which shows an E
transthiolation,[7] but zerumbone adds to the growing configuration, whereas in 1d this ROESY cross-peak was
number of compounds where this reversibility trait is replaced by one with 2-H, diagnostic of a Z configuration.
1
brought to the extreme, resulting in thiol addition products Based on comparison of H NMR spectroscopic data, 2,6-
that are transient, resist isolation, and can be made to un- diisozerumbone (1d) might correspond to a photoisomer
dergo retro-Michael reaction by a simple solvent switch or wrongly reported as 6-isozerumbone (1e).[10] Taken to-
by heating.[8] Remarkably, removal of this transient electro- gether, these results show that irradiation can result in the
philictraitleadstoacompletelossofbiologicalactivity,which isomerization not only of the two conjugated double bonds,
suggests that fleeting Michael reactivity can also have bio- but also of the nonconjugated double bond of zerumbone,
logical consequences.[8]
presumably because of an antenna effect, where excitation
To shed light on this behaviour, and to provide clues of of the dienone chromophore is transmitted to a nearby
general value for the biological relevance of thiol trapping olefin site, as documented, for instance, in the formation of
by electrophilic biologically active agents, we have evaluated lumi-paclitaxel from the irradiation of paclitaxel.[11] At full
the relationship between structure and Michael reactivity in conversion, the ratio of photoisomers 1b/1c/1d was ca.
a series of analogues of zerumbone where the cross-conju- 7:2:1.
gated dienone system is modified configurationally and
Incomplete degassing and/or oxygen leaking during the
electronically by photochemical isomerization and/or re- irradiation was associated with the formation of variable
moval of transannular interaction (homoconjugation) with amounts of epoxides rac-2a[12] and rac-2b, resulting in re-
the isolated endocyclic double bond of the medium-sized moval of homoconjugation between the cross-conjugated
ring.[9] The cysteamine assay, based on the reaction of thiol- dienone system and the isolated endocyclic double bond.
trapping compounds with cysteamine (3) in DMSO and Epoxidation of the isolated double bond had dramatic
phase-switch to CDCl3,[6] was used to obtain information effects on the chemical shift of the terminal carbon atoms
on the stoichiometry and regiochemistry of addition, and of the dienone system, especially in 9-isozerumbone (1b),
on the relative reactivities of the analogues.
with a downfield shift of ca. 9 ppm for the C-6 signal, and
of ca. 4 ppm for the C-10 signal, suggesting a significant
through-space orbital interaction, a typical feature of me-
dium-sized cyclic polyolefins.[9] The yield of the epoxides
was increased considerably when the irradiation was carried
Results and Discussion
The configurational isomers of zerumbone (i.e., 1b–1d) out under a flow of air; however, this did not result in com-
were obtained by photochemical isomerization of the natu- plete conversion. Despite the presumably radical mecha-
ral product. The photochemistry of zerumbone has pre- nism of this oxidation, the configurations of the epoxides
viously been investigated under rather unusual conditions, were the same as those of the products obtained by treating
namely combining irradiation and thermal isomeriza- zerumbone (1a) and 9-isozerumbone (1b) with peracids.
tion.[10] Under these conditions, the 9Z isomer of zerum- The regioselectivity of the epoxidation, with attack
bone [9-isozerumbone (1b)] was formed along with a mix- occurring at the non-enone double bond, is as expected
ture of products resulting from both thermal and photo- from consideration of electron density and polarization, but
chemical pericyclic processes. Irradiation of zerumbone at it is nevertheless in contrast to the outcome of the electro-
room temperature prevented the formation of the “ther- philic addition of bromine to zerumbone, which, surpris-
mal” photoproducts and cleanly gave 9-isozerumbone (1b) ingly, occurred chemoselectively at the trisubstituted conju-
as the major reaction product (42%), accompanied by gated dienone double bond.[13] These contrasting observa-
minor amounts of two doubly isomerized analogues, tions are presumably related to the lack of planarity of the
namely 6,9-diisozerumbone (1c) and 2,6-diisozerumbone dienone system of zerumbone. In the solid state, torsion
(1d). The configurational assignment of the photoiso- angles with the carbonyl oxygen atom of ca. 45° and 35°
merized products was based on evaluation of the coupling were observed for the disubstituted and trisubstituted dien-
constant J9,10 for the disubstituted (Δ9) double bond, and one double bonds, respectively, resulting in a skew arrange-
on inspection of the ROESY spectra for the two trisubsti- ment that makes zerumbone chiral and, in principle,
tuted double bonds (Δ2 and Δ6). The coupling constant be- resolvable.[14] Pyramidalization and alkyl substitution raise
tween the vicinal olefin protons 9-H and 10-H was found the energy of an olefin HOMO.[15] This effect dominates
2
www.eurjoc.org
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 0000, 0–0

Job/Unit: O50237
/KAP1
Date: 05-05-15 12:22:36
Pages: 7
Reactivity of Zerumbone and Related Dienones
the reaction with a soft electrophile like bromine, but with
harder electrophiles like the distal oxygen atom of peracids,
the more electron-rich nonconjugated double bond is pre-
ferred.
The Michael reactivity of the natural product and its an-
alogues was then investigated with the cysteamine assay.[6]
Zerumbone (1a) and its derivatives are potentially bifunc-
tional thiol-trapping agents, and the assay was therefore
carried out by first treating solutions of the substrates in
DMSO with ca. 1.5 equiv. of cysteamine to identify the
most reactive site, and next with an excess (4 equiv.) of this
reagent to evaluate the overall reactivity pattern. The re-
versibility was evaluated by using a solvent shift, diluting
the DMSO solution of the adducts to 1:10 with CDCl3. In
reactions with nitrogen and oxygen nucleophiles, the trisub-
stituted dienone double bond of zerumbone was reported
to be more reactive than the disubstituted double bond,[4]
and a similar trend was observed in the NMR-spec-
troscopy-based cysteamine assay. An excess of the reagent
was found to be necessary for the disappearance of the AB
system assigned to the Δ9 double bond. Remarkably, the
Michael addition at the Δ9 disubstituted dienone bond was
reversed by the solvent switch, while the one at the other
Scheme 1. Stoichiometry, regiochemistry and nature (irreversible/
transient) of the Michael addition to zerumbone (1a), its stereoiso-
mers 1b–1d, and epoxides 2a and 2b. A plain line indicates an
irreversible addition, and a dotted line a transient addition.
In comparative experiments in which equimolar amounts
dienone bond (Δ6) was not.[6] It is not clear how a different of zerumbone (1a) and its 9Z isomer 1b were treated with
degree of pyramidalization and the methyl substitution con- a substoichiometric amount of cysteamine, only the 9Z iso-
tribute to this profile of reactivity, but transient thiol ad- mer reacted, which shows that isomerization increases the
dition has also been reported in a strained and partially reactivity of the disubstituted dienone double bond. Fur-
pyramidalized macrocyclic linear dienone.[16]
ther competition experiments in which equimolecular pairs
Doubly isomerized 6Z,9Z dienone 1c showed the same of isomers were treated with a substoichiometric amount of
reactivity as the natural product (Scheme 1), but marked cysteamine established the reactivity sequence 9Z Ͼ 6E Ͼ
differences in reactivity were observed for the other isomers. 6Z Ͼ 9E. In similar competition experiments, the epoxide
Thus, 9Z isomer 1b and 2Z,6Z isomer 1d behaved as mono- 2a of zerumbone and the epoxide 2b of 9-isozerumbone re-
dentate electrophiles, exclusively giving irreversible acted more slowly than their parent compounds, which
monoaddition to the disubstituted Δ9 double bond, even suggests that homoconjugation facilitates the interaction
with a large excess (6 equiv.) of cysteamine. These observa- with thiols, as would be expected for a lowering of the
tions show that isomerization can reverse the reactivity LUMO energy. Taken together, the results of the cyste-
profile of the dienone system in terms of stoichiometry, re- amine assay showed that zerumbone and its analogues can
giochemistry, and mode of addition. The gem-dimethyl sub- react with thiols in different ways. Zerumbone (1a) and its
stitution at the allylic carbon atom of the disubstituted 6Z,9Z isomer 1c behaved as bifunctional electrophiles,
double bond (C-11) could explain the lower reactivity of while the other isomers 1d and 1d behaved as monofunc-
this dienone bond compared to the trisubstituted one in tional and irreversible electrophiles. Remarkably, the
zerumbone (1a) and its 6Z,9Z isomer 1c. However, mono- Michael reactivity of the 9Z double bond was completely
isomerization of the dienone system as in 1b and 1d reverses removed by epoxidation of the nonconjugated double bond.
this trend, making the disubstituted double bond more re-
Since Michael reactivity is the hallmark of the biological
active than the trisubstituted one, possibly because of an profile of zerumbone,[5] we were interested in investigating
increased polarization that overrides the steric effect of the whether, and to what extent, these differences in reactivity
allylic gem-disubstitution. Remarkably, epoxidation had no translated to differences in biological activity. Thus, the
effect on the Michael reactivity of zerumbone (1a) but activity of the natural product and its derivatives were
dramatically affected that of its 9Z isomer 1b, quenching investigated in a series of validated biochemical (NF-κB
the thiophilicity of the disubstituted double bond, and lead- inhibition, Nrf2 activation), and phenotypical [cytotoxicity,
ing to the exclusive and irreversible reaction of the trisubsti- ROS (reactive oxygen species) generation] endpoints of
tuted dienone bond, even with a large excess (6 equiv.) of zerumbone.[5,17] Consistently with previous reports,[5,17]
cysteamine. These observations highlight the difficulty of zerumbone inhibited NF-κB, activated Nrf2, induced ROS,
predicting the site and stoichiometry of thiol addition in and was toxic to cancer cells. This pleiotropic trait of the
complex electrophilic molecules[7] and undoubtedly provide natural product was, however, lost in its isomers and deriva-
food for thought for computational studies aimed at pro- tives. Thus, epoxides 2a and 2b were unable to induce ROS
viding a mechanistic basis to rationalize this baffling accumulation or inhibit NF-κB, but were still capable to
behaviour.
activate the Nrf2 pathway, especially compound 2b. On the
Eur. J. Org. Chem. 0000, 0–0
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
3

Job/Unit: O50237
/KAP1
Date: 05-05-15 12:22:36
Pages: 7
G. Appendino et al.
FULL PAPER
other hand, the 9Z isomer 1b, which was significantly less
pro-oxidant than 1a, was cytotoxic, but did not show NF-
κB inhibitory activity, despite its higher Michael reactivity
than that of the natural product (see the Supporting Infor-
mation for detailed biological activity data). These observa-
tions show that reactivity with thiols and biological activity
of these cross-conjugated electrophilic dienones are signifi-
cantly dissociated. Shape complementarity is therefore criti-
cal to biological activity; presumably, this works by posi-
tioning the thiol-trapping site in the correct orientation to
interact with the reactive cysteine residue(s) of the target
molecule(s).
Experimental Section
General Remarks: 1H (500 MHz) and 13C (125 MHz) NMR spectra
were measured with a Varian INOVA spectrometer. Chemical shifts
Conclusions
were referenced to the residual solvent signal (CDCl3: δH
=
7.26 ppm; δC = 77.0 ppm). Homonuclear 1H connectivities were de-
termined with COSY experiments. One-bond heteronuclear 1H,13C
connectivities were determined with HSQC experiments. Through-
space 1H proximities were determined with ROESY experiments
with a mixing time of 250 ms. Two- and three-bond 1H,13C connec-
tivities were determined with gradient 2D HMBC experiments op-
timized for 2,3J = 9 Hz. Low- and high-resolution electrospray
(ESI) mass spectra were obtained with an LTQ OrbitrapXL
(Thermo Scientific) mass spectrometer. Silica gel 60 (70–230 mesh)
was purchased from Macherey–Nagel. Reactions were monitored
Zerumbone is undoubtedly a rather peculiar electrophilic
reagent. Its reactivity profile, as well as its stability com-
pared to that of its deoxy derivative humulene (4),[18] might
be due to the existence of a troponoid motif resulting from
a homoconjugated interaction between the dienone system
and the nonconjugated double bond (Figure 1). Neverthe-
less, we believe that the conclusions of this study provide
clues of general relevance. Thus, the puzzling changes ob-
served upon isomerization of the dienone system or re-
moval of homoconjugation highlight how difficult it is to by TLC on Merck 60 F254 (0.25 mm) plates, which were visualized
by UV inspection and/or staining with H2SO4 (5% in ethanol) and
heating. The organic phases were dried with Na2SO4 before con-
centration. Zerumbone was isolated from the essential oil of sham-
poo ginger by direct crystallization,[21] and epoxyzerumbone (2a)
was prepared according to a literature procedure.[12] The purity
(95% or higher) of all of the final products that were evaluated for
reactivity and biological activity was assessed by analytical HPLC
using isocratic acetonitrile/1% formic acid in water (80:20) as elu-
ent, with UV detection at 240 nm with a JASCO Herculite appara-
tus equipped with a Phenomenex Luna C18 column.
predict reactivity in multifunctional electrophilic agents,
while the lack of correlation between biological activity and
thiophilicity exemplifies the difficulty of translating test-
tube data into the complexity of cell-based experiments.
Photochemical Isomerization of Zerumbone (1a): A solution of
zerumbone (1a) (1.00 g, 4.5 mmol) in ethanol (100 mL) was de-
gassed by bubbling nitrogen through it for 10 min and then irradi-
ated in a water-cooled immersion-well photochemical reactor by
using a 125 W high-pressure mercury lamp. The course of the reac-
tion was monitored by TLC on silica gel (petroleum ether/EtOAc,
9:1), by checking the disappearance of zerumbone (Rf = 0.36) and
the formation of the isomerized products [Rf(1b) = 0.59; Rf(1c) =
Rf(1d) = 0.42; Rf(2a) = Rf(2b) = 0.16]. After 2 h, the reaction was
complete, and the solution was concentrated. The semisolid residue
was purified by gravity column chromatography on silica gel (petro-
leum ether/EtOAc, 98:2 to 80:20) to give 1b (410 mg, 42%), a mix-
ture of 1c and 1d (180 mg, 18%), and a mixture of epoxides 2a and
2b (72 mg, 7%). The mixture of 1c and 1d was resolved by prepara-
tive HPLC on silica gel (petroleum ether/EtOAc, 95:5) to give 1c
(84 mg) and 1d (39 mg). Pure samples of epoxides 2a and 2b were
obtained by epoxidation of zerumbone (1a) and 9-isozerumbone
(1b) rather than by purification of the mixture of these two com-
pounds obtained from the irradiation of 1a.
Figure 1. Homoconjugated interaction between the crossed di-
enone system and the nonconjugated double bond of zerumbone,
as expressed by troponoid structure B.
Due to a common reactivity-based mechanism of action,
compounds like acrolein (5), a carcinogenic pollutant,
curcumin (6), a dietary phenolic, and sunitinib (7), an anti-
cancer selective kinase inhibitor, belong, basically, to the
same group of thiol-trapping agents. Specificity of action is
bound to a “lock-and-dock” strategy,[19] namely the exqui-
site coupling of a general decrease of reactivity and the pos-
sibility of involvement in multicenter engagements capable
to accelerate the addition reaction. In this context, the iden-
tification of stoichiometry of addition, reactivity center(s)
and ranking within a series of analogues are of critical rele-
vance, and we have shown how the cysteamine assay, a
quick and unsophisticated test, can provide this infor-
mation.[20]
(9Z)-Isozerumbone (1b): Colourless crystals, m.p. 62–63 °C. 1H
NMR (CDCl3, 500 MHz): δ = 5.86 (d, J = 12.3 Hz, 1 H, 9-H), 5.78
(m, 1 H, 6-H), 5.34 (d, J = 12.3 Hz, 1 H, 10-H), 5.00 (t, J = 8.8 Hz,
1 H, 2-H), 2.27 (m, 2 H, 5-H2), 1.96 (m, 1 H, 1a-H), 1.94 (m, 2 H,
4
www.eurjoc.org
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 0000, 0–0

Job/Unit: O50237
/KAP1
Date: 05-05-15 12:22:36
Pages: 7
Reactivity of Zerumbone and Related Dienones
4-H2), 1.90 (m, 1 H, 1b-H), 1.92 (s, 3 H, 13-H3), 1.40 (s, 3 H, 12- Cytotoxicity Assays: SK-S-NH cells were seeded at a density of
H3), 1.07 (s, 6 H, 14-H3, 15-H3) ppm. 13C NMR (CDCl3, 104 cells/well in 96-well plates (200 μL), and were incubated with
125 MHz): δ = 201.6 (C-8), 151.6 (C-6), 143.6 (C-10), 137.5 (C-7), increasing concentrations of the compounds for 24 h. After that,
134.2 (C-3), 127.3 (C-9), 126.2 (C-2), 41.7 (C-11), 40.8 (C-1), 38.6 MTT [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium brom-
(C-4), 33.5 (C-5), 29.0 (C-14), 25.1 (C-15), 10.7 (C-13), 15.6 (C-12) ide); 5 mg/mL; 100 μL] from a mixed solution of MTT and DMEM
ppm. MS (ESI+): m/z = 241 [M + Na]+. HRMS (ESI+): calcd. for
C15H22ONa 241.1568; found 241.1570.
(Dulbecco’s modified Eagle medium) (1:2) was added to each well,
and cells were incubated at 37 °C in the dark for 4 h. Then the
reaction was stopped, the supernatant was removed, DMSO
(100 μL) was added to each well, and the wells were incubated for
10 min with gentle shaking. Finally, the absorbance was measured
at 550 nm by using a TriStar LB 941 instrument (Berthold Technol-
ogies). The absorbance of untreated cells was taken as 100%
viability, and the percentage of cellular viability was calculated.
(6Z,9Z)-Diisozerumbone (1c): Colourless powder, m.p. 75–77 °C.
1H NMR (CDCl3, 500 MHz): δ = 6.60 (m, 1 H, 6-H), 5.99 (d, J =
12.2 Hz, 1 H, 9-H), 5.60 (d, J = 12.2 Hz, 1 H, 10-H), 5.18 (t, J =
8.1 Hz, 1 H, 2-H), 2.41 (m, 2 H, 5-H2), 2.30 (m, 2 H, 4-H2), 2.05–
1.95 (m, 2 H, 1-H2), 1.75 (s, 3 H, 13-H3), 1.53 (s, 3 H, 12-H3),
1.05 (s, 3 H, 14-H3), 0.83 (s, 3 H, 15-H3) ppm. 13C NMR (CDCl3,
125 MHz): δ = 204.8 (C-8), 144.2 (C-10), 137.5 (C-7), 136.2 (C-3),
131.5 (C-6), 126.5 (C-9), 125.0 (C-2), 43.1 (C-11), 41.6 (C-1), 38.2
(C-4), 33.5 (C-5), 29.0 (C-14), 25.1 (C-15), 21.4 (C-13), 19.7 (C-12)
ppm. MS (ESI+): m/z = 241 [M + Na]+. HRMS (ESI+): calcd. for
C15H22ONa 241.1568; found 241.1572.
NF-κB and Nrf2 Transcriptional Assays: Anti-NF-κB (NF =
nuclear factor) activity was investigated in 5.1 cells, a validated “in
vitro” model for the study of TNFα-induced NF-κB activation
(TNF = tumor necrosis factor).[22] The 5.1 cells were stimulated
with TNFα (20 ng/mL) in the presence or absence of the com-
pounds for 6 h. For the activation of the antioxidant response ele-
ment (ARE), SK-N-SH cells were transiently transfected with the
pGL3-ARE-Luc plasmid, which contains the Nqo1 antioxidant
response element fused to the luciferase reporter gene. After 24 h,
the cells were stimulated with the compounds for 6 h. After stimu-
lation, the cells were washed twice in PBS (phosphate-buffered
saline) and lysed in tris-phosphate pH = 7.8 (25 mm), MgCl2
(8 mm), DTT (1 mm), Triton X-100 (1%), and glycerol (7%) at
room temperature in a horizontal shaker for 15 min. After centri-
fugation, the supernatants were used to measure luciferase activity
by using an Autolumat LB 9510 instrument (Berthold) following
the instructions of the luciferase assay kit (Promega, Madison, WI,
USA). The protein concentration in the cell extracts was measured
by the Bradford method (BioRad, Hercules, CA, USA). The RLU/
μg (RLU = relative light units) was calculated, and the results were
expressed as percentage of inhibition of NF-κB activity induced by
TNFα (100% activation). For Nrf2 activation, the specific ARE-
Luc induction was expressed as induction relative to the control
(untreated cells).
(2Z,6Z)-Diisozerumbone (1d): Colourless powder, m.p. 56–57 °C.
1H NMR (CDCl3, 500 MHz): δ = 6.34 (d, J = 16.2 Hz, 1 H, 10-
H), 5.88 (d, J = 16.2 Hz, 1 H, 9-H), 5.53 (m, 1 H, 6-H), 5.13 (t, J
= 8.1 Hz, 1 H, 2-H), 2.31 (m, 2 H, 4-H2), 2.18 (m, 2 H, 5-H2),
2.05–1.91 (m, 2 H, 1-H2), 1.91 (s, 3 H, 13-H3), 1.69 (3 s, 3 H, 12-
H3), 1.15 (s, 3 H, 14-H3), 1.15 (s, 3 H, 15-H3) ppm. 13C NMR
(CDCl3, 125 MHz): δ = 203.8 (C-8), 161.7 (C-10), 137.5 (C-7),
137.0 (C-3), 131.5 (C-6), 127.5 (C-9), 123.8 (C-2), 42.9 (C-11), 40.1
(C-1), 37.0 (C-4), 33.5 (C-5), 29.3 (C-14), 24.2 (C-15), 21.4 (C-13),
20.7 (C-12) ppm. MS (ESI+): m/z = 241 [M + Na]+. HRMS (ESI+):
calcd. for C15H22ONa 241.1568; found 241.1565.
Epoxidation of (9Z)-Zerumbone (1b): m-Chloroperbenzoic acid
(MCPBA; 80%; 84 mg, 0.8 mmol, 2 equiv.) was added to a solution
of 1b (85 mg, 0.4 mmol) in CH2Cl2 (2 mL). The mixture was stirred
at room temp. for 15 min, then it was diluted with CH2Cl2 (10 mL),
and worked up by concentration. The residue was purified by grav-
ity column chromatography on neutral alumina (10 g; petroleum
ether/EtOAc, 8:2) to give 2b (80 mg, 93%).
(9Z)-Isozerumbone Oxide (2b): Colourless foam. 1H NMR (CDCl3,
500 MHz): δ = 6.04 (d, J = 11.8 Hz, 1 H, 10-H), 5.98 (br. t, 1 H,
6-H), 5.43 (d, J = 11.8 Hz, 9-H), 2.87 (dd, J = 7.9, 3.7 Hz, 1 H, 2-
H), 2.85 (m, 1 H, 4a-H), 2.20 (m, 2 H, 5-H2), 1.91 (s, 3 H, 13-H3),
1.82 (dd, J = 13.2, 3.7 Hz, 1 H, 1a-H), 1.25 (s, 3 H, 3 H, 12-H3),
1.18 (m, 2 H, 4b-H, 1b-H), 1.10 (s, 6 H, 14-H3, 15-H3) ppm. 13C
NMR (CDCl3, 125 MHz): δ = 202.8 (C-8), 159.6 (C-6), 147.9 (C-
10), 139.5 (C-7), 128.4 (C-9), 62.9 (C-2), 61.5 (C-3), 42.7 (C-4), 38.2
(C-1), 36.1 (C-5), 29.8 (C-14), 24.8 (C-11), 24.0 (C-15), 15.6 (C-13),
12.2 (C-12) ppm. MS (ESI+): m/z = 257 [M + Na]+. HRMS (ESI+):
calcd. for C15H22O2Na 257.3238; found 257.3244.
ROS Determination: To evaluate the induction of ROS, SK-N-SH
cells were treated with increasing concentrations of compounds for
12 h. Then the cells were incubated in PBS with DiOC6(3) (green
fluorescence; 20 nm; Molecular Probes, Eugene, OR, USA) at 37 °C
for 20 min, followed by analysis with a FACS Canto II flow cyto-
meter (Beckton Dickinson, NJ, USA).
Statistical Analysis: Data are expressed as meanϮSD. Differences
were analysed by the Student’s t test. P Ͻ 0.05 was considered
significant.
Cysteamine Assay: In a standard 5 mm NMR tube (Armar Chemi-
cals), the substrate (ca. 5 mg) was dissolved in [D6]DMSO
(500 μL), and the NMR spectrum was recorded. Cysteamine
(1.5 equiv.) was then added, and the spectrum was recorded again
5 min after the addition. A second aliquot (4 equiv.) of cysteamine
was added, and the spectrum was registered again. An aliquot
(25 μL) of the solution was then transferred into a second NMR
tube containing CDCl3 (500 μL), and a new spectrum was re-
corded. A positive assay was evidenced by the disappearance of the
dienone signals of the substrate, and the reversibility of the Michael
addition by the reappearance of some of them after dilution to 1:20
with CDCl3. Comparative reactivity experiments were carried out
by treating roughly equimolar amounts of two substrates with a
substoichiometric amount of cysteamine (0.8–1.0 equiv. based on a
single reactive site in only one of the two substrates).
Acknowledgments
We thank Dr. Daniel Joulain for a generous gift of zerumbone,
and Mr. Andry Leeman (Mensa Group, Jakarta, Indonesia) for
providing fresh zerumbet roots. We are grateful to Ministero
dell’Università e Ricerca Scientifica e Tecnologica (MURST), Italy
for financial support (Project 2009RMW3Z5: “Metodologie sintet-
iche per la generazione di diversità molecolare di rilevanza bio-
logica”). Mass and NMR spectra were recorded at the “Centro di
Servizio Interdipartimentale di Analisi Strumentale”, Università di
Napoli “Federico II”.
[1] J. Singh, R. C. Petter, T. A. Baillie, A. Whitty, Nat. Rev. Drug
Discovery 2011, 10, 307–17.
Eur. J. Org. Chem. 0000, 0–0
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
5

Job/Unit: O50237
/KAP1
Date: 05-05-15 12:22:36
Pages: 7
G. Appendino et al.
FULL PAPER
[2]
[3]
[13] T. Kitayama, T. Okamoto, R. K. Hill, Y. Kawai, S. Takahashi,
S. Yonemori, Y. Yamamoto, K. Ohe, S. Uemura, S. Sawada, J.
Org. Chem. 1999, 64, 2667–2672.
[14] S. R. Hall, S. Nimgirawath, C. L. Raston, A. Sittatrakul, S.
Thadamiti, H. Thirasasana, A. H. White, Aust. J. Chem. 1981,
34, 2243–2247.
[15] A. Nicolaides in Pyramidalized Alkenes in Strained Hydro-
carbons (Ed.: H. Daidziuk), Wiley-VCH, Weinheim, 2009, pp.
113–115.
[16] P. Gupta, U. Sharma, T. C. Schulz, E. S. Sherrer, A. McLean,
A. J. Robins, L. M. West, Org. Lett. 2011, 13, 3920–3923.
[17] T. Kitayama, Biosci. Biotechnol., Biochem. 2011, 75, 199–207.
M. Birringer, Pharm. Res. 2011, 28, 2680–2694.
N. J. Yob, S. M. Jofrry, M. M. R. M. M. Affandi, L. K. Teh,
M. Z. Salleh, Z. A. Zakaria, Evidence-Based Complementary
Altern. Med. 2011, 1–12.
For a review on the chemistry of zerumbone, see: T. Kitayama,
Biosci. Biotechnol. Biochem. 2011, 75, 199–207.
R. Prasannan, K. A. Kalesh, M. K. Shanmugam, A. Nachi-
yappoa, L. Ramachandra, A. Nguyen, A. P. Kumar, M. Laksh-
amana, K. S. Ahn, G. Sethi, Biochem. Pharmacol. 2012, 84,
1268–1276.
[4]
[5]
[6]
C. Avonto, O. Taglialatela-Scafati, F. Pollastro, A. Minassi, V.
Di Marzo, L. De Petrocellis, G. Appendino, Angew. Chem. Int. [18] For instance, the nonconjugated double bond of humulene is
Ed. 2011, 50, 467–471; Angew. Chem. 2011, 123, 487–491.
rapidly oxidized in air, whereas, in the absence of light, no reac-
tion takes place with zerumbone.
[7] M. H. Johansson, Mini-Rev. Med. Chem. 2012, 12, 1330–1344.
[8] M. B. Sporn, K. T. Liby, M. M. Yore, L. Fu, J. M. Lopchuk,
G. W. Gribble, J. Nat. Prod. 2011, 74, 537–545.
[9] a) L. T. Scott, Pure Appl. Chem. 1986, 58, 105–110; b) G. Ap-
pendino, M. Calleri, G. Chiari, J. Chem. Soc. Perkin Trans. 2
1986, 205–209.
[19] L. Guterman, Chem. Eng. News 2011, 89, 19–26.
[20] For previous studies on the Michael reactivity of curcumin
with thiols (including cysteamine), see: N. Al-Rifai, H.
Ruecker, S. Amslinger, Chem. Eur. J. 2013, 19, 15384–15395.
In experiments to compare their reactivity toward thiols, curcu-
min and zerumbone showed comparable rates of addition (see
Supporting Information).
[10] H. N. Subba Rao, N. P. Damodaran, S. Dev, Tetrahedron Lett.
1967, 8, 227–233.
[21] M. R. Sulaiman, E. K. Perimal, M. N. Akhtar, A. S. Mo-
hamad, M. H. Khalid, N. A. Tasrip, F. Mokhtar, Z. A. Zaka-
ria, N. H. Lajis, D. A. Israf, Fitoterapia 2010, 81, 855–858.
[22] E. Munoz, M. V. Blazquez, C. Ortiz, C. Gomez-Diaz, P. Navas,
Biochem. J. 1997, 325, 23–28.
[11] S.-H. Chen, C. M. Combs, S. E. Hill, V. Farina, T. W. Doyle,
Tetrahedron Lett. 1992, 33, 7679–7680.
[12] H. W. D. Matthes, B. Luu, G. Ourisson, Tetrahedron 1982, 38,
3129–3135.
Received: February 18, 2015
Published Online: ᭿
6
www.eurjoc.org
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 0000, 0–0

Job/Unit: O50237
/KAP1
Date: 05-05-15 12:22:36
Pages: 7
Reactivity of Zerumbone and Related Dienones
Natural Products
Using the NMR-spectroscopy-based cyste-
amine assay, the thia-Michael reactivity of
the cross-conjugated dienone zerumbone
and its photoisomers and epoxides was in-
vestigated, and marked differences were
found. For biological endpoints sensitive to
Michael acceptors, a substantial separation
between reactivity and biological activity
was observed.
G. Appendino,* A. Minassi, J. A. Collado,
F. Pollastro, G. Chianese,
O. Taglialatela-Scafati, M. Ayyari,
V. Garcia, E. Muñoz ......................... 1–7
The Thia-Michael Reactivity of Zerum-
bone and Related Cross-Conjugated Dien-
ones: Disentangling Stoichiometry, Regio-
chemistry, and Addition Mode with an
NMR-Spectroscopy-Based Cysteamine As-
say
Keywords: Natural products / Michael ad-
dition / Enones / Thiols / Medium-ring
compounds / Biological activity /
Structure-activity relationships
Eur. J. Org. Chem. 0000, 0–0
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
7