Dissecting the pharmacophore of curcumin. Which structural element is critical for which action?

Article abstract of DOI:10.1021/np400148e

The dietary phenolic curcumin (1a) is the archetypal network pharmacological agent, but is characterized by an ill-defined pharmacophore. Nevertheless, structure-activity studies of 1a have mainly focused on a single biological end-point and on a single structural element, the aliphatic bis-enoyl moiety. The comparative investigation of more than one end-point of curcumin and the modification of its aromatic region have been largely overlooked. To address these issues, we have investigated the effect of aromatic C-prenylation in the three archetypal structural types of curcuminoids, namely, curcumin itself (1a), its truncated analogue 2a (C₅-curcumin), and (as the reduced isoamyl version) the tetrahydro derivative 3a, comparatively evaluating reactivity with thiols and activity in biochemical (inhibition of NF-κB, HIV-1-Tat transactivation, Nrf2 activation) and phenotypic (anti-HIV action) assays sensitive, to a various extent, to thia-Michael addition. Prenylation, a validated maneuver for bioactivity modulation in plant phenolics, had no effect on Michael reactivity, but was detrimental for all biological end-points investigated, dissecting thiol trapping from activity, while hydrogenation attenuated, but did not completely abrogate, the activity of 1a. The C ₅-curcuminoid 2a outperformed the natural product in all end-points investigated and was identified as a novel high-potency anti-HIV lead in a cellular model of HIV infection. Taken together, these observations show that Michael reactivity is a critical element of the curcumin pharmacophore, but also reveal a surprising sensitivity of bioactivity to C-prenylation of the vanillyl moiety.

Full text of DOI:10.1021/np400148e

Article
pubs.acs.org/jnp
Dissecting the Pharmacophore of Curcumin. Which Structural
Element Is Critical for Which Action?
Alberto Minassi,*,† Gonzalo Sanchez-Duffhues,‡ Juan Antonio Collado,‡ Eduardo Munoz,‡
́
̃
and Giovanni Appendino*,†
†Dipartimento di Scienze del Farmaco, Universita
‡Instituto Maimon
ides de Investigacion Biomedica de Cor
Inmunología, Universidad de Cordoba, C/ Maria Virgen y Madre s/n, 14004 Cor
̀
del Piemonte Orientale, Via Bovio 6, 28100, Novara, Italy
doba (IMIBIC), Departamento de Biología Celular, Fisiología e
doba, Spain
́
́
́
́
́
́
S
* Supporting Information
ABSTRACT: The dietary phenolic curcumin (1a) is the archetypal network pharmacological agent, but is characterized by an
ill-defined pharmacophore. Nevertheless, structure−activity studies of 1a have mainly focused on a single biological end-point
and on a single structural element, the aliphatic bis-enoyl moiety. The comparative investigation of more than one end-point of
curcumin and the modification of its aromatic region have been largely overlooked. To address these issues, we have investigated
the effect of aromatic C-prenylation in the three archetypal structural types of curcuminoids, namely, curcumin itself (1a), its
truncated analogue 2a (C5-curcumin), and (as the reduced isoamyl version) the tetrahydro derivative 3a, comparatively
evaluating reactivity with thiols and activity in biochemical (inhibition of NF-κB, HIV-1-Tat transactivation, Nrf2 activation) and
phenotypic (anti-HIV action) assays sensitive, to a various extent, to thia-Michael addition. Prenylation, a validated maneuver for
bioactivity modulation in plant phenolics, had no effect on Michael reactivity, but was detrimental for all biological end-points
investigated, dissecting thiol trapping from activity, while hydrogenation attenuated, but did not completely abrogate, the activity
of 1a. The C5-curcuminoid 2a outperformed the natural product in all end-points investigated and was identified as a novel high-
potency anti-HIV lead in a cellular model of HIV infection. Taken together, these observations show that Michael reactivity is a
critical element of the curcumin pharmacophore, but also reveal a surprising sensitivity of bioactivity to C-prenylation of the
vanillyl moiety.
spectrum anti-inflammatory profile is of interest for the
prevention/treatment of pathological conditions sustained or
triggered by chronic inflammation (cancer, diabetes, Alz-
heimer’s disease),1,3 but curcumin has also been reported to
interact with over one hundred further biological targets,4
including some of great relevance for HIV research, such as the
proteins integrase and protease, as well as Tat-mediated
transactivation of the HIV-LTR.5
A systematic study of the biological profile of curcumin (1a)
is impossible under the reductionist mantra of one target−one
disease−one drug, but several of its actions might be related,
sharing a common mechanistic basis. Thus, the NF-κB pathway
ith almost 3000 preclinical investigations and potential
Wbeneficial effects that even include genetic diseases, the
diarylheptanoid curcumin (1a) is one of the most thoroughly
investigated and most promising dietary natural products.1 Due
to a dismally low oral bioavailability2 and to the lack of financial
incentives in the development of a nonpatentable natural
product, the clinical documentation on curcumin is inconsistent
with its pharmacological potential, and most of its beneficial
effects in humans lack validation, being only suggested by
epidemiological correlations, supported by studies on animal
models, or simply extrapolated from in vitro findings.3 At a
molecular level, curcumin is believed to act as a master switch
of inflammation by inhibiting pro-inflammatory transcription
factors (NF-κB, STAT-3) and the function and expression of
inflammatory enzymes (COXs and LOs).1,3 This broad-
Received: February 20, 2013
© XXXX American Chemical Society and
American Society of Pharmacognosy
A
dx.doi.org/10.1021/np400148e | J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
might be involved not only in the anti-inflammatory action of
curcumin but also in its antiviral action, since this compound
reverses Tat-induced dissociation of HDAC1 from LTR by
preventing the binding of p65/NFκB to the LTR promoter.5
From a medicinal chemistry standpoint, the pharmacophore of
curcumin is a unique blend of Michael acceptor, metal-
chelating, and antioxidant properties, and its dissection has so
far been elusive.6 Most structure−activity studies on curcumin
have aimed at the modification of its diacryloylmethane moiety,
which can be simplified to a dienone to afford the so-called C5-
curcuminoids, replaced by a 1,3-distyryl-substituted hetero-
cyclic core, or alkylenated (alkylated) at its central methylene
carbon with substantial retention, and sometimes even
improvement, of activity against a single specific biochemical
or phenotypic end-point.6 Conversely, less attention has been
devoted to the two aromatic moieties of the natural product,
and most studies in the area refer to simplified C5-curcuminoids
or to analogues of the natural product that bear additional
oxygenated functions.6 Surprisingly, the introduction of carbon
substituents on the aromatic moieties has received scarce
attention. Nature seems to have followed suit, since none of the
over three hundred diarylheptanoids reported so far from
various Zingiberaceous plants is C-prenylated,7 despite the
relevance of this group for the biological profile of phenolics
(flavonoids, stilbenoids), where it can induce novel bioactivity,
modulate a pre-existing one, and markedly improve bioavail-
ability.8 Apparently, Nature seems to have overlooked an
important possibility to generate biologically interesting
chemical diversity. Intrigued by these considerations, we have
investigated the prenylation of curcumin (1a) and its C5-
analogue (2a), evaluating the effects of this maneuver on the
reactivity with thiols, an important feature of the curcuminoid
pharmacophore,3,9 and on the activity toward three thiol-
trapping-sensitive biochemical targets of curcuminoids (NF-κB,
Tat-induced activation of the HIV-LTR promoter, Nrf2),3 as
well as on the phenotypic translation of the NF-κB and Tat
inhibition into HIV inhibition. Tetrahydrocurcumin (3a) and
its dihydroprenyl (=isoamyl) analogue (3b) were also prepared
and evaluated as nonelectrophilic versions of curcumin (1a)
and its diprenyl derivative (1b), respectively.
prenylation could, in principle, improve absorption8 while
leaving the biological profile largely unaffected. Owing to the
inherent propensity of phenols to undergo O- rather than C-
prenylation, and the high reactivity of the enolized 1,3-
dicarbonyl system of curcumin with electrophiles,10 the direct
prenylation of the natural product was not possible. 5′,5″-
Diprenylcurcumin (1b) was therefore prepared by condensing
the boric complex of 2,4-pentandione (acetylacetone) with 5-
prenylvanillin (6). By stabilizing its enol tautomer, formation of
a boric complex prevents alkylation of the central methylene of
acetylacetone, quenching the methylene-centered Knoevenagel
reactivity at the expense of the methyl-centered aldol-type
reactivity (Pabon reaction).12 5-Prenylvanillin (6) could, in
principle, be obtained from vanillin by α,α-dimethyl prop-
argylation, semireduction, and Claisen rearrangement (Scheme
1).11 Unexpectedly, this seemingly straightforward strategy
could not be implemented without a redox detour, since the
semireduction of the triple bond of 5 could not be affected at
the aldehyde stage, but only with the corresponding benzylic
alcohol, which next had to be reoxidized (Scheme 1).
Apparently, the palladium catalyst was inactive in the presence
of the aldehyde carbonyl, a somewhat surprising observation.
For the synthesis of the corresponding C5-curcuminoids (2b),
acetylacetone was replaced by acetone, a change that, for
unclear reasons, required protection as a Mem (methoxyethox-
ymethyl) ether of the phenolic hydroxyl group of 5-
prenylvanillin (6) to effect the bidirectional crotonic con-
densation. The protecting group could then be removed with
13
the THF complex of SnCl4 without affecting the acid-
sensitive prenyl group. The saturated analogues 3a and 3b were
prepared from the corresponding curcuminoids (1a and 2a) by
catalytic hydrogenation, an uneventful reaction.
Curcumin (1a), C5-curcumin (2a), and their prenyl
derivatives (1b and 2b, respectively) all gave a positive assay
for thiol trapping, reacting instantaneously and irreversibly with
cysteamine in DMSO to afford the corresponding 1,7-bis
Michael adduct.14 The positivity of the assay was evidenced by
the disappearance of the AB olefin system in the spectra, and
the irreversible nature of the addition by the failure of the AB
olefin system to reappear upon dilution with CDCl3.14
Competitive experiments were carried out by reacting
equimolecular binary mixtures of these curcuminoids with
substoichiometric amounts of cysteamine. Under these
conditions, no difference in reactivity was observed between
curcumin (1a) and its prenyl derivative (1b) or between the
monocarbonyl analogue of the natural product 2a and its prenyl
derivative 2b. Conversely, 2a showed a higher reactivity than
curcumin in the comparative cysteamine assay. Thus, when an
equimolecular mixture of 1a and 2a was treated with half of the
stoichiometric amount of cysteamine required for the formation
of the bis-adducts, a ca. 4:1 mixture of the bis-Michael adduct of
2a and that of curcumin was obtained (Figure 1). As expected,
no reaction occurred with the tetrahydro derivative 3a or its
isoamyl derivative 3b and cysteamine.
Having assessed the reactivity of the curcuminoids 1a−3b
with cysteamine, their NF-κB inhibitory activity was inves-
tigated in a stably transfected cell lines 5.1, a lymphoid T cell
line in which the HIV-1 LTR is activated by TNF-α through an
NF-κB-dependent mechanism.15 The HIV-1 LTR promoter
contains two κB sites that are critically required to respond to
TNFα, and it is well known that deletion of the two κB sites in
the LTR promoter abolishes completely the response to TNFα.
Thus, 5.1 cells represent an excellent cellular model for the
RESULTS AND DISCUSSION
■
Curcumin (1a) is the classic network pharmacological agent,
since its cellular mechanism of action is not dependent on the
targeting of a single molecular target, but is, apparently, the
result of a biochemical synergism that involves the weak
targeting of different proteins within related signaling net-
works.1,4 This biochemical pleiotropism is paralleled by the
vagueness of the curcumin pharmacophore.6 Aromatic C-
B
dx.doi.org/10.1021/np400148e | J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
Scheme 1. Synthesis of 5,5′-Diprenylcurcumin (1b) and 5,5′-Diprenyl-C5-curcumin (2b)
investigated in HeLa Tat-Luc cells in which the HIV-1 LTR
is directly activated by the HIV-1 Tat protein. In accordance
with the results observed in the NF-κB assays, prenylation
abolished the anti-Tat activity of curcumin (1a) and was
detrimental in all cases, while the most potent compound was
the C5-analogue 2a (IC50 = 4.7 μM) (Figure 4). These results
provided a rationale for investigating the HIV-1 inhibitory
activity of curcuminoids. The antiviral activity was evaluated in
Jurkat cells infected with the pNL4-3 HIV-1 clone pseudotyped
with the VSV envelope, bypassing the natural mode of HIV-1
entry into these cells and supporting a robust HIV-1
replication. Upon integration into host chromosomes, this
recombinant virus expresses the firefly luciferase gene, and
consequently luciferase activity in infected cells correlates with
the rate of viral replication. High luciferase activity levels were
detected 12 h after cellular infection with the VSV-pseudotyped
HIV-1 clone, and pretreatment of the cells 30 min prior to
infection with increasing doses of 1a and 2a revealed that the
latter outperforms curcumin by 4−5-fold (Figure 5). In our
experiments, the time-points used to study the activities of the
compounds were 6 h (Figures 2 and 6) and 12 h (Figures 4 and
5), respectively, and only marginal cytotoxicity was found with
compound 2a at these time-points.
Figure 1. Comparative thiol trapping assay of an equimolecular
mixture of curcumin (1a) and C5-curcumin (2a) treated with two
molar equivalents of cysteamine in DMSO-d6. The arrows indicate the
signals of the two equivalent enone β-protons (H-1 and H-7) of C5
curcumin (2a, lower field doublet, δ 7.66) and curcumin (1a, higher
field doublet, δ 7.55). After the addition of two equivalents of
cysteamine, the ratio between the two signals became 4:1, showing the
preferential reaction of C5-curcumin (2a) compared to curcumin (1a).
Finally, the set of curcuminoids was evaluated for another
bioactivity sensitive to thiol trapping, namely, the activation of
antioxidant response element (ARE)-mediated transcriptional
induction.17 This activity was investigated in a luciferase
reporter assay in SK-N-SH cells. In this cellular model,
curcumin (1a) was a weak activator of the Nrf2 pathway, and
this activity was clearly enhanced in its C5 analogue (2a), in
accordance with all the previous structure−reactivity and
structure−activity observations. Interestingly, it was found
that low doses of 2a activated ARE, while higher concentrations
were less potent and even reduced the basal level of ARE-Luc
activity in SK-N-SH cells. Hormesic models of activity have
been desscribed for curcumin also in other cellular models,18
and the activity of 2a at 1 μM concentration may be
physiologically relevant. Saturated curcumin (3a) did not
activate the ARE pathway, and prenylated curcumin (1b) and
the C5 analogue (2b) were also inactive in this assay. In
nonstimulated cells, NF-κB is normally sequestered in the
cytoplasm by a family of inhibitory proteins, which bind NF-κB,
inhibit its DNA-binding, and prevent its nuclear accumulation,
namely, IκB proteins.19 Specific extracellular stimuli such as
cytokines and phorbol esters lead to the rapid phosphorylation,
ubiquitination, and ultimately proteolytic degradation of IκBα,
which frees NF-κB to translocate to the nucleus, where it
screening of anti-NF-κB compounds. Curcumin (1a) inhibited
NF-κB with an IC50 of 16.4 μM, a value consistent with those
reported previously for this compound (Figure 2).1 This
activity was completely lost for its prenylated derivative (1b)
and was markedly attenuated in compounds 2b, 3a, and 3b.
Conversely, and in accordance with previous findings,6 activity
was higher for the C5 analogue 2a, which showed an IC50 value
of 4.92 μM. The effect of compound 2a, the most potent
compound toward cysteamine and the most powerful inhibitor
of NF-κB transcriptional activity, was investigated on IκBα
degradation in Jurkat cells. Stimulation with PMA (phorbol
miristate acetate) induced a rapid phosphorylation and
degradation of the cytoplasmic IκBα protein, coupled with
phosphorylation of the MAPKs JNK 1+2 and NF-κB subunit
p65.16 All these activities were inhibited by 2a in a
concentration-dependent manner (Figure 3). PMA-induced
ERK 1+2 activation by phosphorylation was not affected by
preincubation with 2a, indicating the selectivity of action on
NF-κB and JNK inhibition (Figure 3). These observations
validated the mechanism of NF-κB inhibition by curcuminoids
in a functional assay.
To evaluate the functionality of these observations, inhibition
of Tat-induced HIV LTR transactivation by 1a−3b was
C
dx.doi.org/10.1021/np400148e | J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
Figure 2. Effects of curcumin (1a) and derivatives on TNFα-induced HIV-1-LTR activation. 5.1 cells were preincubated with the compounds at the
indicated doses and stimulated with TNFα (2 ng/mL) for 6 h. The luciferase activity was measured, and results are presented as the percentage of
inhibition compared to TNFα alone. Five independent experiments were performed, and data are presented as mean SD; *p < 0.05, **p < 0.01,
and ***p < 0.005 in Student’s t test.
stimulation of NF-κB transactivation.19 Several reports have
shown that curcumin inhibits NF-κB by targeting the IKC
complex and also the MAPK JNK pathway.21,22 Taken
together, these observations suggest that the NF-κB inhibitory
mechanism of curcumin (1a) and its C-5 analogue (2a) is
remarkably similar, despite the truncation of the pharmaco-
phore. Owing to the lower potency,21 details on the action of
the tetrahydrocurcuminoids could not be elucidated in this
system.
The downstream biochemical implications of the NF-κB
inhibition data were investigated by evaluating the HIV-1 Tat-
(transactivator or transcription) inhibitory activity of com-
pounds 1a,b−3a,b. Tat is a protein essential for virus
replication that interacts with the transactivation response
element (TAR) RNA stem-loop and recruits the positive
transcriptional elongation factor (p-TEFb), which, in turn,
phosphorylates the C-terminal domain of the RNA polymerase
II and activates transcription elongation from the HIV long
terminal repeat promoter (LTR).23 Curcumin (1a) inhibits
HIV-1 Tat-mediated transactivation of the of the HIV-LTR
seemingly by a mechanism dependent on the HDAC1-NF-κB
pathway, since this substance reverses Tat-induced dissociation
of HDAC1 from LTR, preventing the binding of p65/NF-κB to
LTR promoters stimulated in Tat-stimulated cells.24 It remains
to be investigated if compound 2a inhibits Tat-induced HIV-1
LTR activation by the same mechanims described for curcumin
Figure 3. C5-Curcumin (2a) inhibits NF-κB and JNK signaling
pathways. Jurkat cells were incubated with 2a at the indicated
(1a).
The inhibition of HIV-gene regulation by targeting both Tat
concentrations and treated with PMA for 30 min, and total protein was
and NF-κB signaling pathways cripples crucial stages in the
extracted. The detection of IκBα phosphorylation and degradation,
cycle of viral replication and has therefore great clinical
total and phosphorylated p65, total and phosphorylated ERK, and
potential to prevent drug resistance, the scourge of AIDS
treatment. Curcumin has also been reported to target the viral
proteins integrase and protease with IC50 values of 40 and 100
μM, respectively.25 To investigate the phenotypical translation
of the biochemical assays, the anti-HIV activities of curcumin
(1a) and 2a were evaluated comparatively in a model of HIV-1
infection. Jurkat cells infected with the pNL4-3 HIV-1 clone
pseudotyped with the VSV envelope were employed. As shown
total and phosphorylated JNK was performed by Western blots. The
results are representative of three independent experiments.
regulates gene transcription. IκBα degradation is triggered by
its phosphorylation from the IKK complex (IKC) at Ser-32 and
Ser-36 and is considered the major step in NF-κB regulation.20
In addition to this pathway, a second level of NF-κB activation
involves the phosphorylation of the p65 and the subsequent
D
dx.doi.org/10.1021/np400148e | J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
Figure 4. Effects of curcumin (1a) and its derivatives 1b, 2a,b, and 3a,b on Tat-induced HIV-1-LTR activation. HeLa Tat-Luc cells were incubated
with either DRB (50 μM) or the test compounds at the indicated doses, and after 12 h luciferase activity was measured. Results are presented as the
percentage of activation relative to untreated cells. Four independent experiments were performed, and data are presented as mean SD; *p < 0.05,
**p < 0.01, and ***p < 0.005 in Student’s t test.
Figure 5. Effects of curcumin (1a) and C5-curcumin (2a) on HIV-1 replication. Jurkat cells (106/mL) were pretreated with the compounds for 30
min and then infected with VSV-pseudotyped-pNL4-3.Luc.R−E− (200 ng p24) for 12 h. Luciferase activity in cell extracts was determined, and
results are represented as percentage of activation SD compared to nontreated infected cells (100% activation). Three independent experiments
were performed, and data are presented as mean SD; *p < 0.05, **p < 0.01, and ***p < 0.005 in Student’s t test.
Figure 6. Effects of curcumin (1a) and derivatives on ARE-Luc activation. SK-N-SH cells were transfected with the ARE-Luc plasmid and then
stimulated with increasing concentrations of the compounds for 6 h. The luciferase activity was measured, and results are presented as the fold
induction over untreated cells. Three independent experiments were performed, and data are presented as mean SD; *p < 0.05, **p < 0.01, and
***p < 0.005 in Student’s t test.
E
dx.doi.org/10.1021/np400148e | J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
catalyst (ca. 80 mg) under a hydrogen atmosphere. The suspension
was filtered on a Celite pad and washed with CH2Cl2. The combined
organic layers were evaporated, and the crude product was dissolved in
toluene (10 mL) and next stirred at room temperature with activated
MnO2 (5 g) for 3 h. The suspension was worked up by filtration on
Celite, and the filtrate was evaporated, dissolved in N,N-diethylaniline
(20 mL), and heated with stirring at 165 °C (oil bath temperature) for
1.5 h. After cooling to RT, the reaction was quenched by the addition
of 2 N H2SO4 and extraction with EtOAc. After drying and
evaporation, the residue was purified over silica gel (petroleum
ether−EtOAc, 8:2, as eluant) to afford 6 as a pale yellow solid (967
mg; 75% over the four steps): mp 66 °C; IR (KBr) νmax 3141, 3001,
2912, 2852, 2728, 1666, 1592, 1464, 1310, 1149, 945, 851, 698 cm−1;
1H NMR (CDCl3, 300 MHz) δ 9.79 (1H, s), 7.28 (2H, bs), 6.26 (1H,
s), 5.31 (1H, t, J = 7.35 Hz), 3.93 (3H, s), 3.38 (1H, d, J = 7.35 Hz),
1.75 (3H, s), 1.72 (3H, s); 13C NMR (CDCl3, 75 MHz) δ 191.4,
149.5, 147.0, 133.7, 129.0, 127.8, 127.66, 121.3, 107.0, 56.3, 27.8, 25.8,
17.9; HREIMS m/z 220.1084 (calcd for C13H16O3, 220.1099)
in Figure 4 and in accordance with the results of the
biochemical assays, the truncated analogue 2a outperformed
curcumin (1a) by 4−5-fold, while prenylcurcumin was devoid
of any significant action.
Taken together, our observations show that a close
relationship exists between reactivity and bioactivity in
curcumin, C5-curcumin, and tetrahydrocurcumin, the arche-
typal curcuminoids from the C7-, C5-, and tetrahydro series. C5-
Curcumin (2a), the most electrophilic compound and most
potent ligand for the transcription factors, was also validated in
a phenotypic assay for anti-IV activity. Remarkably, insertion of
an ortho-prenyl group on the phenyl moiety was detrimental for
activity but not relevant for thiol trapping, suggesting that a
specific curcuminoid recognition site, unable to accommodate
relatively bulky aromatic alkyl substituents such as a prenyl
group, should exist on these targets.26 The discovery that a
single point mutation unrelated to the pharmacophoric
signature of curcumin can abolish its biological action against
specific targets is a remarkable observation for a group of
substances otherwise dominated by an overall modest
sensitivity to chemical modification and somewhat flat
structure−activity relationships.6,26 Therefore, prenylcurcumin
(1b) may well serve as a negative control agent for studies
aimed at unraveling the biological profile of curcumin (1a),
dissecting actions mediated by thiol trapping from those
differently modulated. The observation that prenylation was
detrimental in compounds from all three series also suggests
that the pharmacophore of curcuminoids is complex and not
exclusively reactivity-based.
5′,5″-Diprenylcurcumin (1b). To a stirred solution of 5-
prenylvanillin (5, 1 g, 4.53 mmol) in dry DMF (1 mL) were added
sequentially B2O3 (317 mg, 2.13 mmol, 0.47 molar equiv), 2,4-
pentandione (232 μL, 2.26 mmol, 0.5 molar equiv), and B(OCH3)3
(504 μL, 0.73 mmol, 3.3 molar equiv). The reaction was stirred at 90
°C, and n-butylamine (100 μL) was added over a period of 2.5 h. After
1 h, the reaction was cooled to 60 °C, and 5% AcOH (10 mL) was
added. The mixture was stirred at 60 °C for 3 h, during which a
precipitate was formed. After filtration and washing with water, the
crude product was purified over silica gel (petroleum ether−EtOAc,
8:2, as eluant) to afford 5′,5″-diprenylcurcumin (1b) as an orange
solid (510 mg; 45%): mp 145 °C; IR (KBr) νmax 3515, 2998, 2968,
2913, 2854, 1624, 1598, 1492, 1301, 1265, 1132, 1078, 1067, 967, 840,
1
550 cm−1; H NMR (CDCl3, 300 MHz) δ 7.55 (2H, d, J = 16 Hz),
6.98 (2H, d, J = 1.8 Hz), 6.92 (2H, d, J = 1.8 Hz), 6.44 (2H, d, J = 16
Hz), 5.92 (1H, s), 5.79 (1H, s), 5.30 (2H, t, J = 7.0 Hz), 3.90 (6H, s),
3.34 (4H, d, J = 7.0 Hz), 1.75 (6H, s), 1.70 (6H, s); 13C NMR
(CDCl3,75 MHz) δ 183.4, 146.6, 145.7, 141.0, 133.4, 127.9, 126.9,
123.6, 121.7, 121.6, 107.4, 101.0, 56.2, 28.0, 25.9, 17.9; HREIMS m/z
504−2523 (calcd for C31H36O6, 504.2512).
EXPERIMENTAL SECTION
■
General Experimental Procedures. IR spectra were taken on a
FT-IR Thermo Nicolet equipment. 1H (300 MHz) and 13C (75 MHz)
NMR spectra were measured on a JEOL Eclipse 300 instrument.
Chemical shifts were referenced to the residual solvent signal (CDCl3:
δH = 7.26, δC = 77.0). High-resolution ESIMS were obtained on a
LTQ OrbitrapXL (Thermo Scientific) mass spectrometer. Silica gel 60
(70−230 mesh) and RP-18 used for gravity column chromatography
(GCC) were purchased from Macherey-Nagel. Reactions were
monitored by TLC on Merck 60 F254 (0.25 mm) plates, which
were visualized by UV inspection and/or staining with 5% H2SO4 in
ethanol and heating. Organic phases were dried with Na2SO4 before
evaporation. Synthetic curcumin (1a), prepared using the Pabon
reaction,12 was used in all biological assays, while tetrahydrocurcumin
(3a) was obstained by catalytic hydrogenation of synthetic curcumin
on Pd/C.27 Chemical reagents and solvents were from Aldrich. All
final compounds were at least 95% pure as evaluated by HPLC.28
3-Methoxy-4-(2-methylbut-3-in-2-yloxy)benzaldehyde (5). To a
stirred solution of vanillin (1 g, 6.56 mmol) in dry DMF (12 mL) were
added sequentially K2CO3 (1.78 g, 13.13 mmol, 2 molar equiv), KI
(1.85 g, 11.15 mmol, 1.7 molar equiv), CuI (25 mg, 0.13 mmol, 0.02
molar equiv), and 3-chloro-3-methyl-1-butine (1.7 mL, 15.08 mmol,
2.3 molar equiv). The reaction was stirred at 65 °C, with its progress
monitored by TLC (petroleum ether−EtOAc, 8:2, as eluant, Rf =
0.33). After 1.5 h, the reaction was quenched by the addition of water
and extraction with ether. The combined organic layers were dried,
filtered, and evaporated, and the residue purified over silica gel
(petroleum ether−EtOAc, 9:1, as eluant) to afford 529 (1.29 g, 89%) as
a pale yellow solid.
5′,5″-Diprenyl C5-Curcumin (2b). To a stirred solution of 5-
prenylvanillin (200 mg, 0.9 mmol) in dry CH2Cl2 (2.5 mL) were
added sequentially TEA (252 μL, 1.8 mmol, 2 molar equiv) and
MEM-Cl (132 μL, 1.17 mmol, 1.3 molar equiv). The reaction was
stirred at room temperature overnight and was next quenched by the
addition of 2 N H2SO4 and extraction with CH2Cl2. The combined
organic layers were dried, filtered, and evaporated. The crude product
dissolved in absolute EtOH (2.5 mL) was stirred for 3 h at room
temperature in the presence of acetone (916 μL; 0.45 mmol, 0.5
equiv) and 20% NaOH (264 μL). The reaction was finally quenched
by the addition of 2 N H2SO4 and extraction with EtOAc, and the
combined organic phases were dried, filtered, and evaporated. The
crude residue was purified over silica gel (petroleum ether−EtOAc,
7:3, as eluant) to afford 100 mg (17% over two steps) of Mem-
protected 2b, which was next dissolved in THF (2.5 mL). After
cooling (0 °C), SnCl4 (500 μL) was added, and the reaction was
stirred at room temperature overnight. The reaction was quenched by
the addition of saturated NaHCO3 and extraction with EtOAc. The
combined organic layers were dried, filtered, and evaporated, and the
crude product was purified over silica gel (petroleum ether−EtOAc,
7:3, as eluant) to afford 3b as a brownish powder (60 mg, overall 9%
from 5): mp 120 °C; IR (KBr) νmax 3513, 3461, 3394, 2965, 2929,
2855, 1732, 1591, 1492, 1428, 1273, 1154, 1078, 984, 938, 847 cm−1;
1H NMR (CDCl3, 300 MHz) δ 7.64 (2H, d, J = 15.9 Hz), 7.03 (2H,
bs), 7.00 (2H, bs), 6.95 (2H, d, J = 15.9 Hz), 5.32 (2H, t, J = 1.2 Hz),
3.93 (6H, s), 3.36 (4H, d, J = 7.3 Hz), 1.76 (6H, s), 1.73 (6H, s); 13C
NMR (CDCl3, 75 MHz) δ 189.0, 146.4, 146.0, 143.6, 133.5, 127.9,
126.7, 124.1, 123.2, 121.7, 107.4, 56.2, 28.0, 25.9, 17.9; HREIMS m/z
462.2415 (calcd for C291H34O5, 462.2406).
5-Prenylvanillin (6). To a stirred solution of 5 (1.29 g, 5.88 mmol)
in EtOH (11 mL) and CH2Cl2 (1.1 mL) was added NaBH4 (199 mg,
5.29 mmol, 0.9 mol.ar equiv). The reaction was stirred at room
temperature for 1 h and then quenched by the addition of 2 N H2SO4
and extraction with CH2Cl2. The combined organic layers were dried,
filtered, and evaporated. The residue was dissolved in MeOH (10 mL)
and stirred overnight at room temperature in the presence of Lindlar
5′,5″-Diisoamyltetrahydrocurcumin (3b). To a stirred solution of
prenylcurcumin (1b, 450 mg, 0.89 mmol) in EtOAc (4 mL) and
MeOH (4 mL) was added Pd/C (10%, cat.). The reaction was stirred
F
dx.doi.org/10.1021/np400148e | J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
overnight at room temperature under an atmosphere of hydrogen and
then worked up by filtration on Celite. The organic phases were
evaporated, and the residue was purified over silica gel (petroleum
ether−EtOAc, 9:1, as eluant) to afford 3b as a yellowish solid (125 mg;
28%): mp 92 °C; IR (KBr) νmax 3536, 3476, 2947, 2924, 2863, 1702,
1601, 1499, 1464, 1437, 1295, 1230, 1153, 1107, 1068, 915, 837, 795,
and immunodetection of specific proteins was carried out with primary
antibodies using an ECL system (GE Healthcare). The steady-state
levels for β-tubulin, total p65, pan-ERK 1+2, and pan-JNK 1+2 served
as a loading control for IκBα (phospho and total), phospho-p65,
phospho-ERK 1+2, and phospho JNK 1+2, respectively.
Statistics. Statistical analyses and IC50 calculations were performed
using GraphPad Prism (Graph Pad Software, San Diego, CA, USA).
Sample population means were compared against control population
means in an unpaired two-tailed Student’s t test. A p value < 0.05 was
considered statisitically significant.
1
567 cm−1; H NMR (CDCl3, 300 MHz) δ 6.54 (2H, bs), 6.52 (2H,
bs), 5.50 (1H, s), 3.83 (6H, s), 2.89 (4H, m), 2.56 (4H, m), 1.53
(10H, m), 0.92 (12H, d, J = 6.7 Hz); 13C NMR (CDCl3, 75 MHz) δ
193.4, 146.2, 141.8, 131.6, 128.8, 121.8, 108.3, 56.0, 40.7, 39.1, 31.6,
28.0, 27.6, 22.7; HREIMS m/z 512.6766 (calcd for C31H44O6,
512.6775).
ASSOCIATED CONTENT
■
Cell Lines and Reagents. Jurkat cells were grown at 37 °C and
5% CO2 in supplemented RPMI 1640 containing 10% heat-inactivated
fetal bovine serum, 2 mM glutamine, penicillin (50 U/mL), and
streptomycin (50 μg/mL). SK-N-SH, Hela-Tat-Luc, and 293T cells
were grown in supplemented DMEM medium. The 5.1 clone line is a
Jurkat-derived clone stably transfected with a plasmid containing the
luciferase gene driven by the HIV-LTR promoter and was maintained
in complete RPMI medium supplemented with G418 (200 μg/mL).15
The Hela-Tat-Luc is a stably transfected cell line with the plasmids
pLTR-Luc and pcDNA3-Tat and was described previously.15 The
antibodies anti-IκBα (C-21, sc-371), anti-NF-κB-p65 (F-6, sc-8008),
and antiphospho-ERK1+2 (sc-7383) were acquired from Santa Cruz
Biotechnology (San Diego, CA, USA). The antibodies against
antiphospho-p65 (Ser 536), antiphospho-IκBα (Ser32/36), anti-JNK,
and antiphospho-JNK 1+2 (9255S) were from Cell Signaling
(Danvers, MA, USA). The mAb anti-tubulin and anti-ERK 1+2
(M5670) were purchased from Sigma Co. (St. Louis, MO, USA).
Luciferase Assays. For the anti-NF-κB activity 5.1 cells were
stimulated with TNFα (20 ng/mL) in the presence or the absence of
the compounds for 6 h. For the activation of the antioxidant response
element, SK-N-SH cells were transiently transfected with the pGL3-
ARE-Luc plasmid and 24 h later stimulated with the compounds for 6
h. The cells were washed twice in PBS and lysed in 25 mM Tris-
phosphate pH 7.8, 8 mM MgCl2, 1 mM DTT, 1% Triton X-100, and
7% glycerol during 15 min at RT in a horizontal shaker. After
centrifugation, the supernatant was used to measure luciferase activity
using an Autolumat LB 9510 (Berthold) following the instructions of
the luciferase assay kit (Promega, Madison, WI, USA).
VSV-Pseudotyped HIV-1 Infection Assa. High-titer VSV-
pseudotyped recombinant virus stocks were produced in 293T cells
by co-transfection of pNL4-3.Luc.R−E− together with either the
pcDNA3-VSV or the epNL3 plasmid encoding the vesicular stomatitis
virus G-protein using the calcium phosphate transfection system as
previously described.30 Jurkat cells (106/mL) were plated on a 24-well
plate and were pretreated with the compounds for 30 min. After
pretreatment, cells were inoculated with virus stocks (200 ng of p24),
and 24 h later cells were washed twice in PBS and lysed in 25 mM
Tris-phosphate pH 7.8, 8 mM MgCl2, 1 mM DTT, 1% Triton X-100,
and 7% glycerol for 15 min at RT. Then, the lysates were spun down
and the supernatant was used to measure luciferase activity using an
Autolumat LB 9510 (Berthold, Bad Wildbad, Germany) following the
instructions of the luciferase assay kit (Promega). The results are
represented as the percent of activation (considering the infected and
untreated cells 100% activation) or RLU. Results represent the mean
of four different experiments.
S
* Supporting Information
This material is available free of charge via the Internet at
http://pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Author
*Tel: +39 0321 375849 (A.M.), +39 0321 373744 (G.A.). Fax:
+ 39 0321 375621 (A.M.), +39 0321 375621 (G.A.). E-mail:
minassi@pharm.unipmn.it (A.M.), appendino@pharm.unipmn.
it (G.A.).
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
G.A. and A.M. are grateful to MURST (Italy) for financial
support (project 2009RMW3Z5: Metodologie Sintetiche per la
̀
Generazione di Diversita Molecolare di Rilevanza Biologica).
G.S.-D., J.A.C., and E.M. were supported by MICINN grant
SAF2010-19292 and by ISCIII-RETIC RD06/006.
REFERENCES
■
(1) Goel, A.; Kunnumakkara, A. J.; Aggarwal, B. B. Biochem.
Pharmacol. 2008, 75, 787−809.
(2) Anand, P.; Kunnumakkara, A. B.; Newman, R. A.; Aggarwal, B. B.
Mol. Pharmacol. 2007, 4, 807−818.
(3) Aggarwal, B. B.; Sung, B. Trends Pharmacol. Sci. 2009, 30, 85−94.
(4) (a) Zhou, H.; Beevers, C. S.; Huang, S. Curr. Drug Targets 2011,
12, 332−347. (b) Gupta, S. C.; Prasad, S.; Kim, J. H.; Patchva, S.;
Webb, L. J.; Priyadarsini, I. K.; Aggarwal, B. B. Nat. Prod. Rep. 2011,
28, 1937−1955.
(5) Zhang, H. S.; Ruan, Z.; Sang, W. W. J. Cell. Physiol. 2011, 226,
3385−3391.
(6) Preetha, A.; Thomas, S. G.; Kunnumakkara, A. B.; Sundaram, C.;
Harikumar, K. B.; Sung, B.; Tharakan, S. T.; Misra, K.; Priyadarsini, I.
K.; Rajasekharan, K. N.; Aggarwal, B. B. Biochem. Pharmacol. 2008, 76,
1590−1611.
(7) Lv, H.; She, G. Nat. Prod. Commun. 2010, 5, 1687−1708. This
review summarizes the occurrence of 307 structurally different
diarylheptanoids in plants.
(8) Minassi, A.; Giana, A.; Ech-Chahad, A.; Appendino, G. A. Org.
Lett. 2008, 10, 2267−2270 and references therein.
(9) (a) Marcu, M. G.; Jung, Y. J.; Lee, S.; Chung, E. J.; Lee, M. J.;
Trepel, J.; Neckers, L. Med. Chem. 2006, 2, 169−174. (b) Wahlang, B.;
Pawar, Y. B.; Bansal, A. K. Eur. J. Pharm. Biopharm. 2011, 77, 275−
281.
Western Blot. Jurkat cells (106 cells/mL) were preincubated with
the indicated compounds for 30 min and then stimulated with PMA
(50 ng/mL) for 15 min. Cells were then washed with PBS and
proteins extracted in 50 μL of lysis buffer (20 mM Hepes pH 8.0, 10
mM KCl, 0.15 mM EGTA, 0.15 mM EDTA, 0.5 mM Na3VO4, 5 mM
NaF, 1 mM DTT, leupeptin 1 μg/mL, pepstatin 0.5 μg/mL, aprotinin
0.5 μg/mL, and 1 mM PMSF) containing 0.5% NP-40. Protein
concentration was determined by the Bradford assay (Bio-Rad,
Richmond, CA, USA), and 30 μg of proteins was boiled in Laemmli
buffer and electrophoresed in 10% SDS/polyacrylamide gels.
Separated proteins were transferred to nitrocellulose membranes
(0.5 A at 100 V; 4 °C) for 1 h. Blots were blocked in TBS solution
containing 0.1% Tween 20 and 5% nonfat dry milk overnight at 4 °C,
(10) Ferrari, E.; Pignedoli, F.; Imbriano, C.; Marverti, G.; Basile, V.;
Venturi, E.; Saladini, M. J. Med. Chem. 2011, 54, 8066−8077.
(11) (a) Nicolaou, K. C.; Pfefferkorn, J. A.; Roecker, A. J.; Cao, G.-
Q.; Barluenga, S.; Mitchell, H. J. J. Am. Chem. Soc. 2000, 122, 9939−
9953. (b) Nicolaou, K. C.; Pfefferkorn, J. A.; Mitchell, H. J.; Roecker,
A. J.; Barluenga, S.; Cao, G.-Q.; Affleck, R. L.; Lillig, J. E. J. Am. Chem.
Soc. 2000, 122, 9954−9967.
(12) Pabon, H. J. Rec. Trav. Chim. 1964, 83, 379−386.
G
dx.doi.org/10.1021/np400148e | J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
(13) Appendino, G.; Cravotto, G.; Palmisano, G.; Annunziata, R.;
Szallasi, A. J. Med. Chem. 1996, 39, 3123−3131.
(14) Avonto, C.; Taglialatela-Scafati, O.; Pollastro, F.; Minassi, A.; Di
Marzo, V.; De Petrocellis, L.; Appendino, G. Angew. Chem., Int. Ed.
2011, 50, 467−471.
́
(15) Marquez, N.; Sancho, R.; Macho, A.; Moure, A.; Masip, I.;
Messeguer, A.; Munoz, E. Biochem. Pharmacol. 2006, 71, 596−604.
̃
(16) San
́
chez-Duffhues, G.; Calzado, M. A.; de Vinuesa, A. G.;
Appendino, G.; Fiebich, B. L.; Loock, U.; Lefarth-Risse, A.; Krohn, K.;
Munoz, E. Biochem. Pharmacol. 2009, 77, 1401−1409.
̃
(17) Jungand, K. A.; Kwak, M. K. Molecules 2010, 15, 7266−7291.
(18) (a) Demirovic, D.; Rattan, S. I. Biogerontology 2011, 12, 437−44.
(b) Biringer, M. Pharm. Res. 2011, 28, 2680−2694.
(19) Schmitz, M. L.; Mattioli, I.; Buss, H.; Kracht, M. ChemBioChem
2004, 5, 1348−1358.
(20) Karin, M.; Ben-Neriah, Y. Annu. Rev. Immunol. 2000, 18, 621−
663.
(21) A previous study reporting only marginal, if any, activity of
hydrogenated curcuminoids as NF-κB inhibitors was based on an
indirect evaluation of this action and seems to have underestimated the
activity of these compounds. Sandur, S. K.; Pandey, M. K.; Sung, B.;
Ahn, K. S.; Murakami, A.; Sethi, G. Carcinogenesis 2007, 28, 1765−
1773.
(22) Chen, Y. R.; Tan, T. H. Oncogene 1998, 17, 173−178.
(23) Johri, M. K.; Mishra, R.; Chhatbar, C.; Unni, S. K.; Singh, S. K.
Expert Opin. Biol. Ther. 2011, 11, 269−283.
(24) Zhang, H. S.; Ruan, Z.; Sang, W. W. J. Cell Physiol. 2011, 226,
3385−3391.
(25) Mazumder, A.; Raghavan, K.; Weinstein, J.; Kohn, K. W.;
Pommier, Y. Biochem. Pharmacol. 1995, 49, 1165−1170.
(26) The redundancy of targets and the lack of a consensus binding
motif have even led to the suggestion that curcumin does not directly
interact with many of its protein end-points, but rather inserts into
lipid bilayers, modifying the energetic costs of their deformation and,
indirectly, modulating the functioning of membrane-bound proteins.
Barry, J.; Fritz, M.; Brender, J. R.; Smith, P. E. S.; Lee, D.-K.;
Ramamoorthy, A. J. Am. Chem. Soc. 2009, 131, 4490−4499.
(27) Changtam, C.; de Koning, H. P.; Ibrahim, H.; Sajid, M. S.;
Gould, M. K.; Suksamrarn, A. Eur. J. Med. Chem. 2010, 45, 4446−
4457.
(28) Cuomo, J.; Appendino, G.; Adam, S.; Dern, A. S.; Schenier, E.;
McKinnon, T. P.; Brown, M. J.; Togni, S.; Dixon, B. M. J. Nat. Prod.
2011, 74, 664−669.
(29) Babu, K. S.; Raju, B. C.; Praveen, B.; Kishore, K. H.; Murty, U.
S.; Rao, J. M. Heterocycl. Commun. 2003, 9, 519−526.
́
(30) Sancho, R.; Marquez, N.; Gomez-Gonzalo, M.; Calzado, M. A.;
Bettoni, G.; Coiras, M. T.; Alcami, J.; Lopez-Cabrera, M.; Appendino,
G.; Munoz, E. J. Biol. Chem. 2004, 279, 37349−37359.
̃
H
dx.doi.org/10.1021/np400148e | J. Nat. Prod. XXXX, XXX, XXX−XXX