Can Small Chemical Modifications of Natural Pan-inhibitors Modulate the Biological Selectivity? the Case of Curcumin Prenylated Derivatives Acting as HDAC or mPGES-1 Inhibitors

Article abstract of DOI:10.1021/acs.jnatprod.5b00700

Curcumin, or diferuloylmethane, a polyphenolic molecule isolated from the rhizome of Curcuma longa, is reported to modulate multiple molecular targets involved in cancer and inflammatory processes. On the basis of its pan-inhibitory characteristics, here we show that simple chemical modifications of the curcumin scaffold can regulate its biological selectivity. In particular, the curcumin scaffold was modified with three types of substituents at positions C-1, C-8, and/or C-8′ [C₅ (isopentenyl, 5-8), C₁₀ (geranyl, 9-12), and C₁₅ (farnesyl, 13, 14)] in order to make these molecules more selective than the parent compound toward two specific targets: histone deacetylase (HDAC) and microsomal prostaglandin E₂ synthase-1 (mPGES-1). From combined in silico and in vitro analyses, three selective inhibitors by proper substitution at position 8 were revealed. Compound 13 has improved HDAC inhibitory activity and selectivity with respect to the parent compound, while 5 and 9 block the mPGES-1 enzyme. We hypothesize about the covalent interaction of curcumin, 5, and 9 with the mPGES-1 binding site.

Full text of DOI:10.1021/acs.jnatprod.5b00700

Article
pubs.acs.org/jnp
Can Small Chemical Modifications of Natural Pan-inhibitors
Modulate the Biological Selectivity? The Case of Curcumin
Prenylated Derivatives Acting as HDAC or mPGES‑1 Inhibitors
Mehrdad Iranshahi,†,‡ Maria Giovanna Chini,§,‡ Milena Masullo,§ Amirhossein Sahebkar,† Azita Javidnia,⊥
Mahsa Chitsazian Yazdi,† Carlo Pergola,∥ Andreas Koeberle,∥ Oliver Werz,∥ Cosimo Pizza,§
Stefania Terracciano,§ Sonia Piacente,*,§ and Giuseppe Bifulco*,§
†Biotechnology Research Center and School of Pharmacy, Mashhad University of Medical Sciences, Vakil Abad Boulevard, Opposite
Mellat Park, 91775-1365 Mashhad, Iran
§Department of Pharmacy, University of Salerno, Via Giovanni Paolo II 132, 84084 Fisciano (SA), Italy
⊥Department of Medicinal Chemistry, Faculty of Pharmacy, Tehran University of Medical Sciences, Enghlab Street, 14155-6559
Tehran, Iran
∥Institute of Pharmacy, Friedrich Schiller University Jena, Philosophenweg 14, 07743 Jena, Germany
S
* Supporting Information
ABSTRACT: Curcumin, or diferuloylmethane, a polyphe-
nolic molecule isolated from the rhizome of Curcuma longa, is
reported to modulate multiple molecular targets involved in
cancer and inflammatory processes. On the basis of its pan-
inhibitory characteristics, here we show that simple chemical
modifications of the curcumin scaffold can regulate its
biological selectivity. In particular, the curcumin scaffold was
modified with three types of substituents at positions C-1, C-8,
and/or C-8′ [C5 (isopentenyl, 5−8), C10 (geranyl, 9−12), and
C
15 (farnesyl, 13, 14)] in order to make these molecules more
selective than the parent compound toward two specific
targets: histone deacetylase (HDAC) and microsomal
prostaglandin E2 synthase-1 (mPGES-1). From combined in silico and in vitro analyses, three selective inhibitors by proper
substitution at position 8 were revealed. Compound 13 has improved HDAC inhibitory activity and selectivity with respect to the
parent compound, while 5 and 9 block the mPGES-1 enzyme. We hypothesize about the covalent interaction of curcumin, 5, and
9 with the mPGES-1 binding site.
he isolation and characterization of active natural products
(NPs) have produced a multitude of derivatives useful
two targets involved in cancer and inflammatory processes:
histone deacetylase (HDAC) and microsomal prostaglandin E2
synthase-1 (mPGES-1).
T
both in therapy and in the study of physiological mechanisms.
In clinical applications, in fact, analogues and/or derivatives of
NPs have often been used in order to modulate the
pharmacokinetic or pharmacodynamic properties of lead
compounds. Some representative examples involve the alkaloid
family; for example, nalorphine (1), with an allyl group as R1,
antagonizes the effects of morphine (2, from Papaver
somniferum), while ethylmorphine (3), with an ethyl group as
R2, causes a complete drop of analgesic activity (Figure 1),
showing that small and simple modifications of a parent
compound can influence its pharmacological activity.
Several examples of chemical modifications of the curcumin
scaffold have been reported in order to modify the selectivity,
potency, and the bioavailability1,2 of this pan-inhibitor. Recent
results3,4 have shown that the prenyl groups at C-9 are able to
modulate the biological selectivity of curcumin (4, Figure 2).
Here we describe the case of this polyphenolic molecule and its
prenylated congeners (5−14, Figure 2) that selectively inhibit
Curcumin (4), a polyphenolic molecule isolated from the
rhizome of Curcuma longa (Zingiberaceae), has a profile of a
safe drug, and it has been used in the treatment of
inflammatory, cancer, neurodegenerative, cardiovascular, and
metabolic diseases, modulating multiple pathways. In particular,
curcumin directly inhibits some biological targets (e.g.,
transcription factors, protein kinases) or indirectly up-regulates
(e.g., JNK, p53, DN5) or down-regulates (TNF, EGFR)5
numerous others crucial for several inflammatory and cancer
pathologies5−8 by noncovalent hydrophobic and hydrogen-
bonding interactions.9,10 Owing to its multitude of actions,6,10
curcumin has several clinical applications, and for these reasons
this natural product and its derivatives are currently involved in
Received: May 5, 2015
© XXXX American Chemical Society and
American Society of Pharmacognosy
A
DOI: 10.1021/acs.jnatprod.5b00700
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
Figure 1. Examples of morphine analogues whose activity is influenced by small chemical substitutions on the scaffold.
The first class of enzyme is involved in the epigenetic
modifications24,25 linked to cancer development and pro-
gression. The impact of curcumin (4, Figure 2) on these
epigenetic mechanisms26,27 has been evaluated and has revealed
its ability to inhibit HDAC in a more effective manner with
respect to the well-known HDACi, such as valproic acid and
sodium butyrate,2 appearing as a new member of HDAC
inhibitors (HDACi).13−18,28−31 In particular, when combined
with HDACi, curcumin suppresses tumor progression proteins
and cell migration in vitro and blocks tumor growth and
metastasis in vivo.32
On the other hand, microsomal prostaglandin E2 synthase-1
enzyme, belonging to the membrane-associated proteins
involved in the eicosanoid and glutathione metabolism
(MAPEG) family, has emerged as an attractive target for the
development of novel anti-inflammatory and anticancer
drugs.33 This enzyme is responsible, via the arachidonic acid
cascade, for the conversion of COX-derived unstable peroxide
34
PGH2 into PGE2
without affecting the constitutive
prostaglandins implicated in physiological functions. It is
overexpressed in several inflammatory disorders,35−39 and for
this reason, the identification of new mPGES-1 inhibitors is
essential for the development of safer drugs devoid of classical
NSAID side effects. Closely related to the inflammatory
diseases, mPGES-1 is involved in the pathogenesis of different
cancer forms and in induction of angiogenesis.40 However,
despite several inhibitors being identified,41−43 only a few of
them exhibit in vivo anticancer and anti-inflammatory44−46
properties.
The recent resolution of the human X-ray structures of
mPGES-1 from 201347 to date48 and of the high-resolution X-
ray structure of human mPGES-1 in lipidic mesophase in
201449 in complex with a potent inhibitor have represented
attractive starting points for the rational design of new mPGES-
1 inhibitors.
Collectively these considerations prompted an evaluation of
the influence of simple prenyl groups50,51 at C-8 and/or C-8′
and C-1 on modulation of biological selectivity of the parent
compound on HDACs and mPGES-1 of some synthetically
accessible curcumin derivatives (5−14). In particular, we have
performed a computational study of the affinities of 4−14,
identifying 5 and 9 as selective mPGES-1 inhibitors, and 13,
which inhibits HDACs and also shows an increment of potency
with respect to curcumin. Interestingly, because the α,β-
unsaturated β-diketo moiety of curcumin covalently interacts
with protein thiols9,10,52,53 through a Michael54 reaction and
mPGES-1 has a glutathione (GSH) cofactor in the catalytic
binding site, we have envisaged a covalent interaction between
4, 5, and 9 with the GSH cofactor.
Figure 2. Chemical structure of curcumin (4) and its derivatives (5−
14).
many clinical trials.2 Thus, it would be crucial to clarify, at the
molecular level, the selectivity of curcumin toward a specific
target, in order to develop new and more active and selective
analogues. To prove this hypothesis, we investigated two
targets directly inhibited by curcumin11,12 and which were
exhaustively studied: HDAC13−18 and mPGES-1.19−23
B
DOI: 10.1021/acs.jnatprod.5b00700
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
Figure 3. Glide GScore (dark bars at the left of the dotted black line) and number of key interactions (light bars at the right of the dotted black line)
of 4−11, 13, and 14 for HDACs. For the calculated activity the interactions with the following amino acids were considered discriminant: (a)
Leu276, Phe144, Tyr209, the π−π interaction with His183 and Phe210, and the coordination (mono- or bidentate mode) with Zn2+ for HDAC2;
(b) Phe152, Tyr306, His143, His180, the π−π interaction with Trp141, and the coordination (mono- or bidentate mode) with Zn2+ for HDAC8; (c)
Tyr170, Phe168, His159, His158, the π−π interaction with Phe227, and the coordination (mono- or bidentate mode) with Zn2+ for HDAC4; (d)
Phe140, Arg126, Gly301, the π−π interaction with Phe200 and His171, and the coordination of Zn2+ for HDAC6; (e) Phe738, His669, Pro542,
Glu543, the π−π interaction with His709, and the coordination of Zn2+ for HDAC7.
Figure 4. Three-dimensional model of cocrystallized molecules (GSH analogue and β-octyl glucoside (dark blue sticks), PDB code: 4AL1) in the
receptor binding site of mPGES-1. In panel B the crucial amino acids of the mPGES-1 receptor are depicted as sticks colored by chain (A, red; B,
green). The π-stacking interaction is represented by a cyan dotted line.
ered. For the HDACs, the coordination of the Zn2+ ion and the
RESULTS AND DISCUSSION
■
interactions with the receptor counterparts have been
Docking Studies. A small set of readily synthesizable
considered essential for the simulated inhibition,14,65,66 and,
curcumin analogues, 5−14 (Figure 2), were designed gradually
from the analysis of the computational data (Figure 3), 4,
together with the molecules with substitution at C-8 (5, 9, and
substituting the curcumin scaffold at C-8, C-8′ (R1 and R2), and
C-1 (R3 and R4) with simple prenylated moieties [C5
13), showed the most interesting results. In particular, they are
(isopentenyl), C10 (geranyl), and C15 (farnesyl)]. These
able to chelate the zinc ion in a bidentate (for HDAC2) and
compounds were used as suitable case studies to test the
monodentate fashion (HDAC8, HDAC4, HDAC6, HDAC7)
influence of simple chemical modifications of the curcumin
scaffold on the regulation of its biological selectivity. In order to
prove the hypothesis of this work, computational studies were
and to arrange the R1 substituent in the cavities of the HDAC’s
surface. On the other hand, the compounds with R1 and R2
substitution groups (6, 10, 14) retain weak calculated affinities
for HDACs (Figure 3), while 7, 8, and 11 (R3 and R4 at C-1)
performed by molecular docking (Glide software version 6.1,
standard and extra precision level)55−58 on these targets, to
are able to interact only with HDAC4 due to the different shape
disclose the structural basis for modulation of the bioactivity of
4.
of the residues at the rim of the catalytic channel. None of the
compounds are able to chelate the zinc ion of HDAC1 and
HDAC3, probably because of the dimension of the catalytic
channel14,59 and the shape of the protein surface,14 which
hinder the access of the zinc-chelating moiety. Finally, full
substitution with geranyl moieties at R1−R4 of 12 leads to
complete inactivity.
Concerning the mPGES-1 receptor, the 3D model of
curcumin (4) and its analogues (5−14) with the X-ray
structure of the target47 are reported for the first time. As
For computational studies, the X-ray and homology model
structures14 of HDACs of class I (HDAC1, 2, 3, and 8)59−62
and class II (HDAC4, 6, and 7)14,63,64 and the recent
experimental structure of mPGES-147 were used. From a
qualitative point of view, the ability of compounds to bind the
targets, expressed by a Glide score (kcal/mol) and by the
number of key interactions established with the ligand binding
sites of HDACs and mPGES-1,14,20,22,23,47−49,66 was consid-
C
DOI: 10.1021/acs.jnatprod.5b00700
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
Figure 5. Glide GScore and number of key interactions (Tyr130chain A, Ser127chain A, Asp49chain B, Gln134chain A, His53chain B, Phe44chain B
Thr131chain A, Tyr28chain B) of 4−10, 13, and 14 for mPGES-1.
,
Figure 6. Three-dimensional models of the interactions of curcumin (4, yellow sticks) in the mPGES-1 binding site. In panel B the crucial amino
acids of the mPGES-1 receptor are depicted as sticks colored by chain (A, red; B, green). The π-stacking interaction is represented by a cyan dotted
line.
Scheme 1. General Synthetic Procedure and Chemical Structures of Synthesized Curcumin Derivatives
reported by Geschwindner et al.,47 the mPGES-1 active site is
divisible into the cofactor (GSH) and substrate (PGH2)
binding site, and it includes the N-terminal (helices II and IV)
and C-terminal (helix I) parts and an adjacent monomer
cytoplasmic domain. The pattern of binding is well represented
by the cocrystallized structure of mPGES-1 with the GSH
analogue [L-γ-glutamyl-S-(2-biphenyl-4-yl-2-oxoethyl)-L-cystei-
nylglycine] and a β-octyl glucoside moiety (Figure 4),
disclosing several key interactions for a rational design of a
substrate site inhibitor.
key interactions with the receptor counterpart (Figure 5), were
performed:47−49 (a) π−π with Tyr130chain A, indicative of a
good accommodation within the GSH binding site;20 (b)
effective contacts (van der Waals and polar interactions) with
Ser127chain A, a key residue involved in PGH2 recognition, and
with Asp49chain B, Gln134chain A, His53chain B, Phe44chain B
,
Thr131chain A, and Tyr28chain B, belonging to the binding
groove.20,22,23
Thus, curcumin interacts in the enzyme active site with the
same set of residues observed for the GSH analogue and β-octyl
glucoside (Figure 6), strongly supporting the significance of the
docking results. The two phenyl rings of curcumin (Figure 6)
Docking calculations between curcumin (4) and analogues
5−14 and mPGES-1, in order to verify the presence of some
D
DOI: 10.1021/acs.jnatprod.5b00700
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
Table 1. Individual and Total Yields of Prenylated Curcumin Analogues (5−14)
molar ratio
5
6
7
8
9
10
11
12
13
14
total yield
curcumin:1-bromo-3-methyl-2-butene:DBU
1:1:1
24
19
10
24
34
60
84
1:3:3
16
37
1
2:24:21
38
curcumin:geranyl bromide:DBU
1:1:1
15
15
38
36
53
78
2:24:21
8
19
curcumin:farnesyl bromide:DBU
1:1:1
21
5
26
interact with grooves A (Gln134Chain A, Tyr130Chain A, and
Ile32Chain B) and B (Leu39Chain B, Phe44Chain B, His53Chain B, and
Asp49Chain B) on the receptor molecular surface, and it
establishes further contacts with Gly35Chain B, Arg38Chain B, and
the GSH cofactor of chain A.
Comparing the results for the two targets, 5, 9, and 13
represent the most active derivatives of 4, showing C-8 as the
important determinant for the modulation of the predicted
biological selectivity and activity.
In order to verify the above in silico results, the synthesis and
biological evaluation of compounds 5−14 were next pursued.
Synthesis. Compounds 5−14 were synthesized by
dissolution of curcumin in acetone and subsequent addition
of 1-bromo-3-methyl-2-butene (2 mmol, 240 μL) and DBU
(1,8-diazabicyclo[5.4.0]undec-7-ene).67 By using different
molar ratios (excess of 1-bromo-3-methyl-2-butene and
DBU) and temperature (reflux) compounds 5−14 were
obtained and successively purified by column chromatography
on silica gel and preparative TLC (Scheme 1). Precumins I−IV
(5−8) are characterized by the presence of one prenyl moiety
at C-8 (precumin I), two prenyl moieties at C-8 and C-8′
(precumin II), three prenyl moieties at positions C-1, C-8, and
C-8′ (precumin III), and four prenyl moieties, two at C-1 and
the other two at C-8 and C-8′ (precumin IV).
Table 2. In Vitro HDAC Inhibitory Activity of Tested
Compounds 4−14
compound
IC50
curcumin (4)
187 μM
5
5.62 mM
6
347 μM
a
7
a
8
9
495 μM
188 μM
217 μM
786 μM
84.2 μM
122 μM
1.42 nM
10
11
12
13
14
trichostatin A
a
Empty cells indicate no inhibition or absence of compound activity
that could not be fit to an IC50 curve.
respect to curcumin (187 μM). Compounds 10, 14, and
curcumin showed comparable activity (Table 2), and precumin
I (5) displayed weaker activity (IC50 = 5.62 mM). Precumins
III (7) and IV (8) were ineffective for HDAC inhibition (data
not shown).
In order to synthesize geranyl derivatives (9−12) of
curcumin, named gercumins I−IV, curcumin was dissolved in
acetone and reacted with different molar ratios of geranyl
bromide in the presence of DBU. The resulting compounds,
purified by column chromatography on silica gel, showed the
same geranyl pattern of substitution with respect to 5−8.
Similarly, for the synthesis of farcumins I (13) and II (14),
curcumin, dissolved in acetone, was treated with farnesyl
bromide in the presence of DBU. The purified compounds
were curcumin derivatives characterized by a farnesyl moiety at
C-8 (farcumin I) and at C-8 and C-8′ (farcumin II). The
percentage yields of compounds 5−14 are reported in Table 1.
The structures of compounds 5−14 have been confirmed by
1D (1H and 13C) and 2D NMR (HSQC, HMBC, and COSY)
experiments along with ESIMS analysis.
HDAC Biological Results. To test HDAC inhibitory
activity, curcumin (4) and compounds 5−14 have been
screened in vitro by using a fluorescence-based assay in HeLa
cell nuclear extracts (see Experimental Section) in 10-dose IC50
mode with 2-fold serial dilution starting at 3.333 mM.
Trichostatin A (TSA), a well-established HDAC inhibitor,
was used as control compound in a 10-dose IC50 with 3-fold
serial dilution starting at 10 μM. The IC50 values for tested
compounds are reported in Table 2.
mPGES-1 Biological Results. In order to study com-
pounds 5−14 for inhibition of mPGES-1, a cell-free activity
assay was used, where the substrate PGH2 is converted by
mPGES-1 in microsomes from IL-1β-activated A549 cells to
PGE2.68 Curcumin (4) potently inhibited mPGES-1 activity
with an IC50 value of 0.22 μM, which is in agreement with a
previous finding.12 The monoisopentenylated 5 and the
monogeranylated 9 showed potent inhibition of mPGES-1,
although the IC50 values (0.93 and 1.02 μM, respectively) are
comparable to 4. Similarly, derivatives carrying two to four
isopentenyl (6−8) or geranyl (10−12) residues as well as one
or two farnesyl moieties (13, 14) were bioactive, but the
potency was further impaired and the IC50 values were
determined in the range 2−8 μM.
Selectivity Profile. The aforementioned results have
permitted the creation of a basic SAR profile of the most
promising molecules with respect to parent compound 4
(IC50‑for‑mPGES‑1 = 0.22 μM, IC50‑for‑HDACs = 187 μM), which
show remarkable inhibition in the low μM range for mPGES-1
(5, IC50 = 0.93 μM, and 9, IC50 = 1.02 μM), and in vitro
interesting HDAC (13, IC50‑for‑HDACs = 84.2 μM) inhibitory
activity. From the computational analysis, C-8 of the curcumin
scaffold (Figures 3 and 5) was identified as a modification point
able to modulate the biological activity of the parent pan-
inhibitor compound, and this result is consistent with biological
data, which have confirmed the hypothesis, and that in
Among the screened compounds, farcumin I (13) was found
to be the most active, with an IC50 value of 84.2 μM with
E
DOI: 10.1021/acs.jnatprod.5b00700
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
Figure 7. SAR summary profile for 5−14.
Figure 8. Three-dimensional models of 4 (yellow), 5 (light blue), 9 (light pink), and 13 (fuchsia) in complex with HDAC2 (class I) (panel A) and
HDAC6 (class II) (panel B).
particular the length of the chain at C-8 mainly influences the
bioselectivity of curcumin analogues (Figure 7).
13. The in silico analysis of compounds 6−8, 10, 11, and 14,
i.e., curcumin analogues with three substitutions (7 and 11) at
the other possible positions (Figure 2), revealed that, even if
they are able to weakly interact with HDAC4, they have
reduced biological activity due to the steric hindrance
responsible for reducing the optimal contacts with the other
HDAC isoforms. Only compound 14, possessing two farnesyl
groups as R1 and R2, acts as a weak inhibitor, proving the alkyl
chain with 15 atoms as the preferred substituent on the
curcumin skeleton for the HDAC inhibition.
The 3D models of curcumin and of the most promising
candidates 5, 9, and 13 in complex with one representative
receptor for each class of HDACs (HDAC2 and HDAC6)
(Figures 8−10) were illustrated to identify the main features of
new potential curcumin-based HDACi and to summarize the
critical points for the modulation of the biological activity on
HDACs shown by these compounds.
Concerning HDAC2, the contacts with Leu276, Phe144, and
Tyr209, the π−π interaction with His183 and Phe210, and the
coordination (mono- or bidentate mode) with the metal atom
were considered as key interactions. As shown above (Table 2),
the substitution at C-8 of curcumin (4) with increasing alkyl
group size, yielding 5, 9, and 13, influences the inhibitory
activity on HDAC. In fact, the better biological profile observed
for compound 13 is due to the larger farnesyl moiety, which is
able to establish peculiar hydrophobic contacts with the
macromolecule surface (namely, with Gln31, Glu103, Gly30,
and Pro106) with respect to 4, 5, and 9 (Figure 9).
As observed for HDAC2 by computational analysis, the gain
derived from the contacts of the R1 alkyl chain of 13 with
HDAC6 relating to the other compounds seems to be the
driving force of the target−ligand complexes. In addition to the
key contacts with Phe140, Arg126, and Gly301, the π−π
interaction with Phe200 and His171, and the coordination of
Zn2+, 13 is also interacting with Asp169, Asp267, His130, and
Phe86 (Figure 10).
From the analysis of binding modes of 5−14 in comparison
with 4 in the mPGES-1 active site, the molecules with one
substitution at R1 (Figure 11) maintain a similar pattern of
contacts, in particular the hydrophobic contacts with groove B
and the key π−π interaction with Tyr130ChainB
In addition, 5, 9, and 13 increase their interactions in groove
A, making further interactions with Val29Chain B and Ile25Chain B
.
,
and, Leu142Chain A is involved in hydrophobic contact with 9
and 13. On the other hand, the increased chain length of the
substituent at C-8 of 9 and 13 permits interaction with
Cys137Chain A and Ile33Chain B, respectively. Moreover, the
presence of the R2, R3, and R4 alkyl chains hampers the
interactions with grooves A and B and produces the loss of the
key contacts with the mPGES-1 binding site, namely, with
Arg126Chain A, Arg38Chain B, and Phe44Chain B for 6 and with
Asp49Chain B for 6 and 8 (see Figure 5 and Supporting
Information). Comparing the binding mode of 5 with respect
to 6−8, which shows the prenylated moieties at R1, R2, R3, and
R4, respectively, it is evident how the alkyl chain R2 causes the
shifting of aromatic rings in grooves A and B (Figure S39,
Supporting Information), and the substituents at R3 and R4,
On the other hand, compounds with multiple prenyl, geranyl,
and farnesyl substitutions (6−8, 10, 11, and 14) show a
decrease of calculated and experimental potency with respect to
F
DOI: 10.1021/acs.jnatprod.5b00700
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
Figure 9. 2D interaction diagrams of 4, 5, 9, and 13 with the HDAC2 binding site. The hydrophobic residues are displayed in green, the polar
residues are displayed in light blue, the positively charged residues are in violet, and the negatively charged residues are in red. H-bonds are indicated
as pink dotted arrows (side chain) and continuous (backbone) pink arrows, and π−π interactions are depicted as continuous green arrows.
protruding toward the solvent/membrane region, interact in an
unfavorable manner.
Moreover, as anticipated above, on the bases of the biological
data, the possibility that curcumin, 5, and 9 establish a covalent
interaction with the GSH cofactor in the mPGES-1 catalytic site
was also evaluated. In particular, the covalent docking module69
at R2, R3, and R4 could restrain the formation of the covalent
complexes.
In summary, by means of the docking studies, the qualitative
predictions have proved that the different size of substituent
groups at C-8 can modulate the biological selectivity of
curcumin. In fact, these results confirm that the topology and
the size of alkyl groups are the most important determinants for
biological selectivity and activity of these curcumin scaffold-
based inhibitors. In particular, the increased C-8 chain-length of
13 allows favorable interactions with the HDAC receptor
surfaces, while it causes a great structural difference with respect
to the dimension of the natural PGH2 substrate of mPGES-1.
On the other hand, the dimension of the alkyl chains of 5 and 9
represents the maximum length allowed to retain a biological
activity on this MAPEG member receptor, and it is compatible
with the covalent hypothesis inhibition.
of the Schrodinger suite was used to obtain a binding mode
̈
between the polyphenolic molecules and the enzyme,
simulating a Michael addition, where the SH of GSH attacks
the α,β-unsaturated β-diketo moiety.9,10,52−54 Curcumin (4)
and its congeners (5 and 9) are able to bind the GSH (Figure
12), keeping the fundamental interactions with the mPGES-1
binding site, maintaining quite similar binding modes with
respect to the noncovalent ones (see above).
In all three models the nonsubstituted aromatic portion
interacts with the amino acids of groove B (Ans46Chain B
,
Asp49Chain B, Phe 44Chain B, etc.) in the same manner, while the
other part of the molecules in the compound-GSH-mPGES-1
complex interact with groove A, like the GSH analogue
cocrystallized with the enzyme (Figure 4). In particular, 8-OH
of 4 forms a hydrogen bond with Gln134Chain A (Figure 12a),
while the 8′-OH of 5 (Figure 12b) and 9 (Figure 12c) interacts
with Asp49Chain B. Moreover, the phenyl ring substituted with
prenyl moiety (Figure 12b) of 5 retains a π−π interaction with
Tyr130Chain A, while 9 (Figure 12c) establishs a cation−π
EXPERIMENTAL SECTION
■
General Experimental Procedures. Melting points were
determined on a Sanyo Gallenkamp apparatus. NMR experiments
were performed on a Bruker DRX-600 spectrometer (Bruker
BioSpinGmBH, Rheinstetten, Germany) equipped with a Bruker 5
mm TCI CryoProbe at 300 K. All 2D NMR spectra were acquired in
CDCl3 (99.95%, Sigma-Aldrich), and standard pulse sequences and
phase cycling were used for DQF-COSY, HSQC, and HMBC spectra.
The NMR data were processed using Xwinnmr software. Exact masses
were measured by an AB SCIEX Voyager DE mass spectrometer
equipped with a 337 nm laser and delay extraction and operated in
positive-ion reflector mode. Samples were analyzed by MALDI-TOF
mass spectrometry. A mixture of analyte solution and α-cyano-4-
hydroxycinnamic acid (Sigma) was applied to the metallic sample plate
interaction with Arg126Chain A
.
The alkyl groups at C-8 show hydrophobic interactions with
Ile25Chain B, Val29Chain B, and Leu142Chain B, demonstrating from
the structural point of view the possibility of covalent
inhibition; in fact, further substitutions with prenylated moieties
G
DOI: 10.1021/acs.jnatprod.5b00700
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
Figure 10. 2D interaction diagrams of 4, 5, 9, and 13 with the HDAC6 binding site. The hydrophobic residues are displayed in green, the polar
residues are displayed in light blue, the positively charged residues are in violet, and the negatively charged residues are in red. H-bonds are indicated
as pink dotted arrows (side chain) and continuous (backbone) pink arrows. π−π and cation−π interactions are depicted as continuous green and red
arrows, respectively.
Figure 11. Three-dimensional models of 4 (yellow), 5 (light blue), 9 (light pink), and 13 (fuchsia) with mPGES-1 binding site. In panel B the
crucial amino acids of the mPGES-1 receptor are depicted as sticks colored by chain (A, red; B, green). The π-stacking interaction and hydrogen
bonds are represented by cyan and yellow dotted lines, respectively.
and dried. Mass calibration was performed with the ions from
adrenocorticotropic hormone (ACTH) fragment 18−39 human at
2465.1989 Da and α-cyano-4-hydroxycinnamic acid at 190.0504 Da as
internal standard. Column chromatography was conducted with silica
gel (Mesh 230−400, 0.040−0.063 mm, Merck).
Curcumin Preparation. Curcumin 65−70% (product code #C
1386) was purchased from Sigma-Aldrich. Curcumin of 65−70%
purity was further purified70 by silica gel column chromatography,
using CH2Cl2 as eluent. The isolated curcumin was crystallized from
MeOH/H2O (9:1), as orange needles. Physicochemical characteristics
H
DOI: 10.1021/acs.jnatprod.5b00700
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
Figure 12. Three-dimensional models of covalent complexes between 4 (yellow sticks), 5 (light blue sticks), 9 (light pink sticks), and cofactor GSH
(stick colored by atom type) of mPGES-1. The crucial amino acids of the mPGES-1 receptor are depicted as sticks colored by chain (A, green; B,
red; C, blue). The π-stacking interaction, cation−π, and hydrogen bonds are represented by cyan, green, and yellow dotted lines, respectively.
1
Table 3. H NMR and 13C NMR Data for Precumins I−IV (5−8) (600 MHz, δ ppm, in CDCl3)
5
6
7
8
position
δH (J in Hz)
5.82, s
δC
102.4
δH (J in Hz)
δC
δH (J in Hz)
δC
δH (J in Hz)
δC
1
5.83, s
102.2
184.3
122.9
141.5
129.3
110.8
150.8
151.7
113.6
123.5
56.9
111.9
184.3
125.6
141.8
129.6
110.7
150.4
151.8
113.6
123.5
56.9
70.0
2, 2′
3, 3′
4, 4′
5, 5′
6, 6′
7, 7′
8, 8′
9, 9′
10, 10′
OCH3
184.5, 184.4
123.1, 123.0
141.8, 141.7
129.3, 129.0
111.0, 111.1
148.4, 149.0
151.3, 146.1
113.6, 115.8
123.8, 123.6
56.9
198.5
120.8
145.0
128.4
111.1
150.8
152.2
113.5
124.4
57.2
6.48, 6.51, d (16.0)
7.58, 7.61, d (16.0)
6.52, d (16.0)
7.63, d (16.0)
6.88, d (15.5)
7.67, d (15.5)
6.69, d (15.5)
7.68, d (15.5)
7.06, 7.08, d (2.0)
7.09, d (2.0)
7.08, d (1.5)
7.02, d (2.0)
6.87, 6.93, d (8.0)
6.90, d (8.5)
7.13, dd (8.5, 2.0)
3.93, s
6.86, d (8.5)
7.12, dd (8.5, 1.5)
3.91, s
6.84, d (8.5)
7.12, dd (8.5, 2.0)
3.91, s
7.12, 7.11, dd (8.0, 2.0)
3.93, s
prenyl moiety at C-8
4.65, d (7.0)
prenyl moieties at C-8,8′
prenyl moieties at C-8,8′
prenyl moieties at C-8,8′
1″
2″
3″
4″
5″
66.9
4.65, d (7.0)
66.7
4.62, d (7.0)
66.7
4.62, d (7.0)
66.7
5.53, brt (7.0)
120.4
139.4
19.4
5.53, brt (7.0)
120.4
139.6
19.4
5.52, brt (7.0)
120.4
139.6
19.4
5.50, brt (7.0)
120.4
139.6
19.4
1.76, s
1.80, s
1.76, s
1.80, s
1.75, s,
1.76, s,
27.1
27.1
1.78, s
27.1
1.79, s
27.1
prenyl moiety at C-1
3.28 (2H), d (6.7)
5.16, t (6.7)
prenyl moieties at C-1
2.79 (2H), d (6.7)
4.88, t (6.7)
1‴
2‴
3‴
4‴
5‴
27.1
30.2
119.8
133.0
19.4,
27.4
119.3
136.5
19.2,
27.4
1.75, s
1.85, s
1.60, s
1.68, s
were in agreement with literature data.70 TLC experiments showed
that curcumin was completely free of demethoxycurcumin and
bisdemethoxycurcumin.
as eluent. The fractions (80 mL each) were analyzed by TLC (using a
mixture of petroleum ether/acetone/MeOH, 2.5:1:0.1). Precumin II
(6) was further purified by preparative TLC using a petroleum ether/
acetone/MeOH (2.5:1:0.1 v/v) system as eluent. For the synthesis of
precumin III (7), a similar procedure was applied using curcumin (2
mmol, 736 mg), 1-bromo-3-methyl-2-butene (6 mmol, 720 μL), and
DBU (6 mmol, 900 μL), followed by purification by silica gel column
chromatography (55 cm × 5 cm). Precumin III (7) was further
purified by preparative TLC using a petroleum ether/acetone/MeOH
(2.5:1:0.1 v/v) system as eluent. For the synthesis of precumin IV (8),
curcumin (2 mmol, 736 mg) was dissolved in acetone (50 mL). Then,
1-bromo-3-methyl-2-butene (24 mmol, 2880 μL) and DBU (21 mmol,
Curcumin Analogues with C5 (Prenyl) Side Chain(s):
Precumins I−IV (5−8). For the synthesis of precumins I (5) and
II (6), curcumin (2 mmol, 736 mg) was dissolved in acetone (50 mL).
Then, 1-bromo-3-methyl-2-butene (2 mmol, 240 μL) and DBU (2
mmol, 300 μL) were added, and the mixture was refluxed for 2 h at 37
°C. Precumins I and II were isolated by decantation of the reaction
mixture with distilled H2O (75 mL) and EtOAc (75 mL), followed by
separation and solvent evaporation of the organic phase. The products
were purified by silica gel column chromatography using a 40 cm × 5
cm column and a petroleum ether/acetone/MeOH (5:1:0.1) mixture
I
DOI: 10.1021/acs.jnatprod.5b00700
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
1
Table 4. H NMR and 13C NMR Data for Gercumins I−IV (9−12) (600 MHz, δ ppm, in CDCl3)
9
10
11
12
δH (J in Hz)
5.82, s
δC
101.1
δH (J in Hz)
δC
δH (J in Hz)
δC
δH (J in Hz)
δC
1
5.85, s
101.4
182.6
122.6
140.9
127.4
110.2
149.8
150.3
113.0
123.0
56.1
111.9
184.3
125.6
141.8
129.6
110.8
150.4
151.8
113.6
123.5
55.9
68.3
2, 2′
3, 3′
4, 4′
5, 5′
6, 6′
7, 7′
8, 8′
9, 9′
10, 10′
OCH3
183.4, 183.3
122.1, 122.2
140.7, 140.5
127.9, 128.9
109.7, 110.0
149.5, 146.8
150.8, 146.7
112.7, 115.2
122.5, 122.8
56.1
197.4
119.9
143.9
127.1
110.3
149.6
150.9
112.7
123.5
56.1
6.51, 6.48, d (16.0)
7.61, 7.59, d (16.0)
6.52, d (16.0)
7.63, d (16.0)
6.88, d (15.5)
7.67, d (15.5)
6.65 d (16.0)
7.66 d (16.0)
7.06, 7.08, d (1.8)
7.11, d (1.8)
7.06, d (1.5)
6.99 d (1.8)
6.88, 6.93, d (8.3)
6.90, d (8.0)
7.13, dd (8.0, 1.8)
3.95 s
6.87, d (8.5)
7.11, dd (8.5, 1.5)
3.91, s
6.82 d (8.3)
7.08 dd (8.3, 1.8)
3.86 s
7.12, 7.11, dd (8.3, 1.8)
3.93, 3.95, s
geranyl moiety at C-8
4.67, d (6.3)
geranyl moieties at C-8,8′
geranyl moieties at C-8,8′
geranyl moieties at C-8,8′
1″
2″
3″
4″
5″
6″
7″
8″
9″
10″
66.4
4.69, d (6.3)
66.3
4.66, d (6.6)
66.2
4.63 d (7.0)
66.0
5.51, t (6.3)
119.2
141.4
39.6
5.53, t (6.3)
119.5
141.4
39.6
5.50, d (6.6)
119.3
141.2
39.5
5.46 t (7.0)
119.2
141.3
39.7
2.08, (2H) m
2.12, (2H) m
5.09, t (6.8)
2.10, (2H) m
2.14, (2H) m
5.11, t (6.8)
2.06, (2H) m
2.11, (2H) m
5.08, t (6.8)
2.04 (2H) m
2.09 (2H) m
5.03 t (6.8)
26.4
26.4
26.2
26.4
123.8
131.7
26.1
123.8
131.4
26.2
123.9
131.8
26.0
124.1
131.3
26.0
1.68, s
1.61, s
1.75, s
1.70, s
1.63, s
1.77, s
1.66, s
1.63 s
18.0
18.3
1.59, s
17.7
1.62, s
16.6
17.0
17.3
1.74, s
16.9
1.70, s
16.9
geranyl moiety at C-1
3.25, (2H) d (6.5)
5.16, t (6.5)
geranyl moieties at C-1
2.77, (2H) d (7.0)
4.89, t (7.0)
1‴
2‴
3‴
4‴
5‴
6‴
7‴
8‴
9‴
10‴
26.1
29.1
118.9
141.8
39.9
118.3
138.3
40.2
2.06, (2H) m
2.11, (2H) m
5.08, (6.8)
1.95, (2H) m
1.99, (2H) m
5.04, t (6.8)
26.6
26.5
124.2
131.8
26.1
124.1
131.4
26.0
1.55, s
1.56, s
1.58, s
1.63, s
1.56, s
1.54, s
17.6
17.7
17.9
17.7
3150 μL) were added, and the mixture was refluxed for 3.5 h and then
sonicated for 1 h at 30 °C.
After isolation and solvent evaporation, the products were purified
by silica gel column chromatography (35 cm × 3 cm). Eleven fractions
(each 50 mL) were collected and analyzed by TLC using petroleum
ether/EtOAc (different ratios).
was dissolved in acetone (50 mL) and treated with farnesyl bromide
(purity: 95%; 2 mmol, 532 μL) in the presence of DBU (2 mmol, 300
μL). The mixture was refluxed for 2 h at 37 °C. Isolation and
purification steps were performed as described for precumins. After the
reaction was completed (confirmed by TLC), the products were
isolated by decantation of the reaction mixture with distilled H2O (75
mL) and EtOAc (75 mL). After separation and solvent evaporation of
the organic phase, the products were purified by silica gel column
chromatography, using petroleum ether/acetone/MeOH (5:1:0.1 v/v)
as eluent.
Precumin I (5): viscous, brownish liquid; yield is reported in Table
1; IR (KBr) νma1x 3360 (br OH), 2925 (CH), 1685 (CO), 1665
(CC) cm−1; H NMR (600 MHz, CDCl3) and 13C NMR (150
MHz, CDCl3) data are reported in Table 3; ESIMS m/z 437.2 [M +
H]+; HR-MALDI-TOFMS [M + H]+ m/z 437.1968 (calcd for
C26H29O6, 437.1964).
Curcumin Analogues with C10 (Geranyl) Side Chain(s):
Gercumins I−IV (9−12). Curcumin (2 mmol, 736 mg) was dissolved
in acetone (50 mL) and treated with geranyl bromide (2 mmol, 400
μL) in the presence of DBU (2 mmol, 300 μL). The mixture was
refluxed for 2.5 h at 37 °C. After the reaction was completed
(confirmed by TLC), the products were isolated by decantation of the
reaction mixture with distilled H2O (75 mL) and EtOAc (75 mL).
After separation and solvent evaporation of the organic phase, the
products were purified by silica gel column chromatography using a
mixture of petroleum ether/acetone/MeOH (5:1:0.1 v/v) as eluent.
The fractions (80 mL each) were analyzed by TLC [petroleum ether/
acetone/MeOH (2.5:1:0.1 v/v)]. Synthesis of gercumins III (11) and
IV (12) was carried out using the above procedure with the following
molar ratios: curcumin (1 mmol, 368 mg), geranyl bromide (12 mmol,
2400 μL), and DBU (10.5 mmol, 1575 μL). Isolation and purification
of products was performed by decantation with H2O (60 mL) and
EtOAc (60 mL), followed by silica gel column (60 cm × 5 cm)
chromatography using a mixture of petroleum ether/acetone/MeOH
(10:1:2.5 v/v) as eluent. The fractions (30 mL each) were analyzed by
TLC [petroleum ether/acetone/MeOH (5:1:2.5 v/v)].
Precumin II (6): orange crystals; yield is reported in Table 1; mp
1
120−121 °C; 3440 (br OH), 2934 (CH), 1650 (CC); H NMR
(600 MHz, CDCl3) and 13C NMR (150 MHz, CDCl3) data are
reported in Table 3; ESIMS m/z 505.3 [M + H]+; HR-MALDI-
TOFMS [M + H]+ m/z 505.2595 (calcd for C31H37O6, 505.2590).
Precumin III (7): orange crystals; yield is reported in Table 1; mp
1
143−144 °C; 3448 (br OH), 2934 (CH), 1640 (CC); H NMR
(600 MHz, CDCl3) and 13C NMR (150 MHz, CDCl3) data are
reported in Table 3; ESIMS m/z 573.3 [M + H]+; HR-MALDI-
TOFMS [M + H]+ m/z 573.3218 (calcd for C36H45O6, 573.3216).
Precumin IV (8): pale yellow crystals; yield is reported in Table 1;
Curcumin Analogues with C15 (Farnesyl) Side Chain:
Farcumins I and II (13 and 14). Curcumin (2 mmol, 736 mg)
1
mp 58−60 °C; 3455 (br OH), 2945 (CH), 1648 (CC); H NMR
J
DOI: 10.1021/acs.jnatprod.5b00700
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
(600 MHz, CDCl3) and 13C NMR (150 MHz, CDCl3) data are
reported in Table 3; ESIMS m/z 641.4 [M + H]+; HR-MALDI-
TOFMS [M + H]+ m/z 641.3847 (calcd for C41H53O6, 641.3842).
Gercumin I (9): viscous, brownish liquid; yield is reported in Table
5; ESIMS m/z 777.5 [M + H]+; HR-MALDI-TOFMS [M + H]+ m/z
777.5098 (calcd for C51H69O6, 777.5094).
In Vitro Biological Assays. Inhibition of HDAC activity in HeLa
nuclear extracts by curcumin (4) and 5−14 was measured using the
HDAC fluorimetric assay on nuclear extract substrate BML-KI104
Fluor de Lys by Reaction Biology Corp. (Malvern, PA, USA). HDAC
activities were evaluated by testing the compounds in 10-dose IC50
mode with 2-fold serial dilution starting at 3.333 mM. HDAC control
compound trichostatin A was tested in a 10-dose IC50 with 3-fold serial
dilution starting at 10 μM. Determination of fluorescence of tested
compounds was performed to ensure that the samples gave no
interference in the assay. All compounds exhibited no fluorescent
background except compounds 8, 11, and 12, which showed a weak
fluorescent background at 3.33 and 1.11 mM.
Preparations of A549 cells and determination of mPGES-1 activity
were performed as described by Koeberle et al.68 In brief, cells were
treated with 2 ng/mL IL-1β for 72 h at 37 °C, 5% CO2, harvested,
sonicated, and homogenized (homogenization buffer: 0.1 M K3PO4
buffer, pH 7.4, 1 mM PMSF, 60 μg/mL STI, 1 μg/mL leupeptin, 2.5
mM glutathione, and 250 mM sucrose). The homogenate was
centrifuged at 10000g for 10 min and 174000g for 1 h at 4 °C, the
resulting pellet (microsomal fraction) was resuspended in 1 mL of
homogenization buffer, and the total protein concentration was
determined. Microsomal membranes were diluted in K3PO4 buffer
(0.1 M, pH 7.4) containing 2.5 mM glutathione. Test compounds or
vehicle was added, and after 15 min at 4 °C reaction (100 μL total
volume) was initiated by addition of 20 μM PGH2. After 1 min at 4
°C, the reaction was terminated using stop solution (100 μL; 40 mM
FeCl2, 80 mM citric acid, and 10 μM 11β-PGE2 as internal standard),
followed by solid-phase extraction and analysis of PGE2 by HPLC.
Computational Details. The chemical structures of investigated
compounds were built with Maestro71 (version 9.6) Build Panel, then
processed with LigPrep72 at a pH of 7.4 1.0, and finally minimized
1
1; 3445 (br OH), 2938 (CH), 1650 (CC); H NMR (600 MHz,
CDCl3) and 13C NMR (150 MHz, CDCl3) data are reported in Table
4; ESIMS m/z 505.3 [M + H]+; HR-MALDI-TOFMS [M + H]+ m/z
505.2595 (calcd for C31H37O6, 505.2590).
Gercumin II (10): yellow crystals; yield is reported in Table 1; mp
62−64 °C; 3440 (br OH), 2934 (CH), 1650 (CC); 1H NMR (600
MHz, CDCl3) and 13C NMR (150 MHz, CDCl3) data are reported in
Table 4; ESIMS m/z 641.4 [M + H]+; HR-MALDI-TOFMS [M + H]+
m/z 641.3845 (calcd for C41H53O6, 641.3842).
Gercumin III (11): viscous, brownish liquid; yield is reported in
1
Table 1; 3450 (br OH), 2940 (CH), 1645 (CC); H NMR (600
MHz, CDCl3) and 13C NMR (150 MHz, CDCl3) data are reported in
Table 4; ESIMS m/z 777.5 [M + H]+; HR-MALDI-TOFMS [M + H]+
m/z 777.5097 (calcd for C51H69O6, 777.5094).
Gercumin IV (12): viscous, brownish liquid; yield is reported in
1
Table 1; 3455 (br OH), 2936 (CH), 1655 (CC); H NMR (600
MHz, CDCl3) and 13C NMR (150 MHz, CDCl3) data are reported in
Table 4; ESIMS m/z 913.6 [M + H]+; HR-MALDI-TOFMS [M + H]+
m/z 912.6270 (calcd for C61H84O6, 912.6268).
Farcumin I (13): viscous, brownish liquid; yield is reported in Table
1
1; 3452 (br OH), 2936 (CH), 1640 (CC); H NMR (600 MHz,
CDCl3) and 13C NMR (150 MHz, CDCl3) data are reported in Table
5; ESIMS m/z 573.3 [M + H]+; HR-MALDI-TOFMS [M + H]+ m/z
573.3219 (calcd for C36H45O6, 573.3216).
Farcumin II (14): viscous, brownish liquid; yield is reported in Table
1
1; 3453 (br OH), 2938 (CH), 1648 (CC); H NMR (600 MHz,
CDCl3) and 13C NMR (150 MHz, CDCl3) data are reported in Table
Table 5. 1H NMR and 13C NMR Data for Farcumins I and II
(13 and 14) (600 MHz, δ ppm, in CDCl3)
using OPLS 2005 force field (Schrodinger Suite 2013).
̈
Protein 3D models were prepared using the Schrodinger Protein
̈
Preparation Wizard (Maestro version 9.6).71 Calculations were
performed using the Glide software package (standard precision
13
14
(SP) and extra precision (XP) modes, version 6.1, Schrodinger
̈
package)55−58 in order to determine the binding mode of compounds
4−14 into the HDAC and mPGES-1 cavities. The receptor grids were
generated focused on sites of pharmacological interest for each protein
(Table S1 in the Supporting Information). The sampling step was set
to expanded sampling mode (4 times), keeping 10 000 ligand poses for
the initial phase of docking, followed by 800 ligand poses selected for
energy minimization. A maximum number of 15 (SP mode) and 15
(XP mode) output structures were saved for each ligand, with a scaling
factor of 0.8 related to van der Waals radii with a partial charge cutoff
of 0.15. A postdocking optimization of the obtained docking outputs
for each SP and XP mode was performed, accounting for a maximum
of five poses based on a 0.5 kcal/mol rejection cutoff for the obtained
minimized poses.
position
δH (J in Hz)
5.83 s
δC
102.2
184.3, 184.2
δH (J in Hz)
δC
1
5.85, s
102.9
184.8
124.9
141.5
129.3
111.0
150.8
151.7
113.6
123.5
57.1
2, 2′
3, 3′
4, 4′
5, 5′
6, 6′
7, 7′
8, 8′
9, 9′
6.52, 6.48, d (16.0) 122.7, 122.9 6.51, d (15.7)
7.63, 7.60 d (16.0) 141.5, 141.3 7.63, d (15.7)
129.5, 129.0
7.07, 7.10, d (1.5)
110.7, 109.8 7.12, d (1.5)
150.5, 148.1
151.5, 146.8
6.89, 6.95, d (8.3)
113.5, 115.8 6.90, d (8.3)
123.4, 123.6 7.14, dd (8.3, 1.5)
10, 10′ 7.13, 7.15, d (8.3)
OCH3
3.95, 3.96, s
56.9
3.96, s
farnesyl moiety at C-8
farnesyl moieties at C-8,8′
The covalent docking between 4, 5, 9, and the cofactor GSH of
mPGES-1 was done using the covalent docking69 protocol of the
1″
4.69 (2H), d (6.3) 66.9
4.70 (2H), d (6.3) 66.9
2″
3″
4″
5″
6″
7″
8″
9″
10″
11″
12″
13″
14″
15″
5.53, t (6.3)
120.2
142.7
40.7
5.54, t (6.3)
120.2
142.7
40.8
Schrodinger suite. In particular, the Michael addition was chosen as
̈
reaction type, between the α,β-unsaturated β-diketo moiety (identified
as the ligand functional group after the Ligfilter step)73 and the
cofactor GSH (reactive receptor residue). Covalent docking experi-
ments were performed generating a receptor grid focused on the
mPGES-1 binding site, considering the residues close to the cofactor
GSH as the centroid of the docking box (7.80 (x), −9.26 (y), −8.16
(z)), with inner box and outer box dimensions of 14 × 14 × 14 Å and
52 × 52 × 52 Å, respectively. For scoring the poses, the option of the
calculate affinity score using Glide was used, where a noncovalent
binding affinity score using the results of Glide docking and a score-in-
place calculation using the final docked pose is calculated. The average
of the two GlideScore values is used for the affinity. Docking results
and illustrations of the 3D and 2D models were generated using
Maestro version 9.6.71
2.11 (2H), m
2.16 (2H), m
5.12, t (6.8)
2.11 (2H), m
2.16 (2H), m
5.12, t (6.8)
27.5
27.4
124.9
136.9
40.8
124.9
136.9
40.8
2.00 (2H), m
2.08 (2H), m
5.12, t (6.8)
1.98 (2H), m
2.08 (2H), m
5.12, t (6.8)
28.0
28.0
124.9
132.9
18.1
124.9
132.9
18.1
1.63, s
1.71, s
1.63, s
1.78, s
1.63, s
1.71, s
1.63, s
1.78, s
27.1
27.1
18.1
18.1
18.1
18.1
K
DOI: 10.1021/acs.jnatprod.5b00700
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
R.; Bifulco, G.; De Riccardis, F. J. Am. Chem. Soc. 2007, 129, 3007−
3012.
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.jnat-
prod.5b00700.
■
S
(19) Chini, M. G.; De Simone, R.; Bruno, I.; Riccio, R.; Dehm, F.;
Weinigel, C.; Barz, D.; Werz, O.; Bifulco, G. Eur. J. Med. Chem. 2012,
54, 311−323.
(20) Lauro, G.; Strocchia, M.; Terracciano, S.; Bruno, I.; Fischer, K.;
Pergola, C.; Werz, O.; Riccio, R.; Bifulco, G. Eur. J. Med. Chem. 2014,
80, 407−415.
(21) De Simone, R.; Chini, M. G.; Bruno, I.; Riccio, R.; Mueller, D.;
Werz, O.; Bifulco, G. J. Med. Chem. 2011, 54, 1565−1575.
(22) Chini, M. G.; Ferroni, C.; Cantone, V.; Dambruoso, P.; Varchi,
G.; Pepe, A.; Fischer, K.; Pergola, C.; Werz, O.; Bruno, I.; Riccio, R.;
Bifulco, G. MedChemComm 2015, 6, 75−79.
(23) Terracciano, S.; Lauro, G.; Strocchia, M.; Fischer, K.; Werz, O.;
Riccio, R.; Bruno, I.; Bifulco, G. ACS Med. Chem. Lett. 2015, 6, 187−
191.
NMR spectra of 5−14, computational details (PDB
codes, grid details), and 2D interactions diagram of 4−14
with HDAC and mPGES-1 binding sites (PDF)
AUTHOR INFORMATION
Corresponding Authors
*For S. Piacente, Phone: +39-089-969763. Fax: +39-089-
969602. E-mail: piacente@unisa.it.
*For G. Bifulco, Phone: +39-089-969741. Fax: +39-089-
969602. E-mail: bifulco@unisa.it.
■
(24) Kristensen, L. S.; Nielsen, H. M.; Hansen, L. L. Eur. J.
Pharmacol. 2009, 625, 131−142.
Author Contributions
(25) Ropero, S.; Esteller, M. Mol. Oncol. 2007, 1, 19−25.
(26) Reuter, S.; Gupta, S. C.; Park, B.; Goel, A.; Aggarwal, B. B. Genes
Nutr. 2011, 6, 93−108.
‡M. Iranshahi and M. G. Chini contributed equally to this work.
Notes
The authors declare no competing financial interest.
(27) Teiten, M.-H.; Dicato, M.; Diederich, M. Mol. Nutr. Food Res.
2013, 57, 1619−1629.
(28) Chen, Y.; Shu, W.; Chen, W.; Wu, Q.; Liu, H.; Cui, G. Basic
Clin. Pharmacol. Toxicol. 2007, 101, 427−433.
ACKNOWLEDGMENTS
■
Financial support by Associazione Italiana Ricerca sul Cancro
(AIRC) Grant IG 2012−IG_12777−Giuseppe Bifulco is
gratefully acknowledged.
(29) Liu, H. L.; Chen, Y.; Cui, C. H.; Zhou, J. F. Acta Pharmacol. Sin.
2005, 26, 603−609.
(30) Wu, Q.; Chen, Y.; Li, X. J. Huazhong Univ. Sci. Technol., Med. Sci.
2006, 26, 199−210.
REFERENCES
(31) Shu, L.; Khor, T. O.; Lee, J.-H.; Boyanapalli, S. S. S.; Huang, Y.;
Wu, T.-Y.; Saw, C. L. L.; Cheung, K.-L.; Kong, A.-N. T. AAPS J. 2011,
13, 606−614.
■
(1) Anand, P.; 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.
(2) Fan, X.; Zhang, C.; Liu, D.-b.; Yan, J.; Liang, H.-p. Curr. Pharm.
Des. 2013, 19, 2011−2031.
(32) Lin, K.-T.; Wang, Y.-W.; Chen, C.-T.; Ho, C.-M.; Su, W.-H.;
Jou, Y.-S. Clin. Cancer Res. 2012, 18, 4691−4701.
(33) Chang, H.-H.; Meuillet, E. J. Future Med. Chem. 2011, 3, 1909−
1934.
(3) Minassi, A.; Sanchez-Duffhues, G.; Antonio Collado, J.; Munoz,
E.; Appendino, G. J. Nat. Prod. 2013, 76, 1105−1112.
(4) Koeberle, A.; Munoz, E.; Appendino, G. B.; Minassi, A.; Pace, S.;
Rossi, A.; Weinigel, C.; Barz, D.; Sautebin, L.; Caprioglio, D.; Collado,
J. A.; Werz, O. J. Med. Chem. 2014, 57, 5638−5648.
(5) Aggarwal, B. B.; Sung, B. Trends Pharmacol. Sci. 2009, 30, 85−94.
(6) Teiten, M.-H.; Eifes, S.; Dicato, M.; Diederich, M. Toxins 2010, 2,
128−162.
(34) Funk, C. D. Science 2001, 294, 1871−1875.
(35) Claveau, D.; Sirinyan, M.; Guay, J.; Gordon, R.; Chan, C. C.;
Bureau, Y.; Riendeau, D.; Mancini, J. A. J. Immunol. 2003, 170, 4738−
4744.
(36) Kojima, F.; Naraba, H.; Miyamoto, S.; Beppu, M.; Aoki, H.;
Kawai, S. Arthritis Res. Ther. 2004, 6, R355−R365.
(37) Masuko-Hongo, K.; Berenbaum, F.; Humbert, L.; Salvat, C.;
Goldring, M. B.; Thirion, S. Arthritis Rheum. 2004, 50, 2829−2838.
(38) Li, X. F.; Afif, H.; Cheng, S.; Martel-Pelletier, J.; Pelletier, J. P.;
Ranger, P.; Fahmi, H. J. Rheumatol. 2005, 32, 887−895.
(39) Subbaramaiah, K.; Yoshimatsu, K.; Scherl, E.; Das, K. M.;
Glazier, K. D.; Golijanin, D.; Soslow, R. A.; Tanabe, T.; Naraba, H.;
Dannenberg, A. J. J. Biol. Chem. 2004, 279, 12647−12658.
(40) Nakanishi, M.; Montrose, D. C.; Clark, P.; Nambiar, P. R.;
Belinsky, G. S.; Claffey, K. P.; Xu, D.; Rosenberg, D. W. Cancer Res.
2008, 68, 3251−3259.
(41) Wiegard, A.; Hanekamp, W.; Griessbach, K.; Fabian, J.; Lehr, M.
Eur. J. Med. Chem. 2012, 48, 153−163.
(42) Roersch, F.; Wobst, I.; Zettl, H.; Schubert-Zsilavecz, M.;
Groesch, S.; Geisslinger, G.; Schneider, G.; Proschak, E. J. Med. Chem.
2010, 53, 911−915.
(43) Shiro, T.; Takahashi, H.; Kakiguchi, K.; Inoue, Y.; Masuda, K.;
Nagata, H.; Tobe, M. Bioorg. Med. Chem. Lett. 2012, 22, 285−288.
(44) Bauer, J.; Koeberle, A.; Dehm, F.; Pollastro, F.; Appendino, G.;
Northoff, H.; Rossi, A.; Sautebin, L.; Werz, O. Biochem. Pharmacol.
2011, 81, 259−268.
(45) Koeberle, A.; Rossi, A.; Zettl, H.; Pergola, C.; Dehm, F.; Bauer,
J.; Greiner, C.; Reckel, S.; Hoernig, C.; Northoff, H.; Bernhard, F.;
Doetsch, V.; Sautebin, L.; Schubert-Zsilavecz, M.; Werz, O. J.
Pharmacol. Exp. Ther. 2010, 332, 840−848.
(46) Norberg, J. K.; Sells, E.; Chang, H.-H.; Alla, S. R.; Zhang, S.;
Meuillet, E. J. Pharm. Pat. Anal. 2013, 2, 265−288.
(7) Coussens, L. M.; Werb, Z. Nature 2002, 420, 860−867.
(8) Wang, S.-L.; Li, Y.; Wen, Y.; Chen, Y.-F.; Na, L.-X.; Li, S.-T.; Sun,
C.-H. Biomed. Environ. Sci. 2009, 22, 32−39.
(9) 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.
(10) Priyadarsini, K. I. Curr. Pharm. Des. 2013, 19, 2093−2100.
(11) Bora-Tatar, G.; Dayangac-Erden, D.; Demir, A. S.; Dalkara, S.;
Yelekci, K.; Erdem-Yurter, H. Bioorg. Med. Chem. 2009, 17, 5219−
5228.
(12) Koeberle, A.; Northoff, H.; Werz, O. Mol. Cancer Ther. 2009, 8,
2348−2355.
(13) Chini, M. G.; Terracciano, S.; Riccio, R.; Bifulco, G.; Ciao, R.;
Gaeta, C.; Troisi, F.; Neri, P. Org. Lett. 2010, 12, 5382−5385.
(14) Di Micco, S.; Chini, M. G.; Terracciano, S.; Bruno, I.; Riccio, R.;
Bifulco, G. Bioorg. Med. Chem. 2013, 21, 3795−3807.
(15) Terracciano, S.; Chini, M. G.; Riccio, R.; Bruno, I.; Bifulco, G.
ChemMedChem 2012, 7, 694−702.
(16) Terracciano, S.; Chini, M. G.; Bifulco, G.; D’Amico, E.;
Marzocco, S.; Riccio, R.; Bruno, I. Tetrahedron 2010, 66, 2520−2528.
(17) Grolla, A. A.; Podesta, V.; Chini, M. G.; Di Micco, S.; Vallario,
A.; Genazzani, A. A.; Canonico, P. L.; Bifulco, G.; Tron, G. C.; Sorba,
G.; Pirali, T. J. Med. Chem. 2009, 52, 2776−2785.
(18) Maulucci, N.; Chini, M. G.; Di Micco, S.; Izzo, I.; Cafaro, E.;
Russo, A.; Gallinari, P.; Paolini, C.; Nardi, M. C.; Casapullo, A.; Riccio,
L
DOI: 10.1021/acs.jnatprod.5b00700
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
(47) Sjogren, T.; Nord, J.; Ek, M.; Johansson, P.; Liu, G.;
Geschwindner, S. Proc. Natl. Acad. Sci. U. S. A. 2013, 110, 3806−3811.
(48) Luz, J. G.; Antonysamy, S.; Kuklish, S. L.; Condon, B.; Lee, M.
R.; Allison, D.; Yu, X. P.; Chandrasekhar, S.; Backer, R.; Zhang, A.;
Russell, M.; Chang, S. S.; Harvey, A.; Sloan, A. V.; Fisher, M. J. J. Med.
Chem. 2015, 58, 4727−4737.
(73) Schrodinger Software Release 2013-3; Schrodinger LLC, New
York, NY, 2013.
̈
(49) Li, D.; Howe, N.; Dukkipati, A.; Shah, S. T. A.; Bax, B. D.; Edge,
C.; Bridges, A.; Hardwicke, P.; Singh, O. M. P.; Giblin, G.; Pautsch, A.;
Pfau, R.; Schnapp, G.; Wang, M.; Olieric, V.; Caffrey, M. Cryst. Growth
Des. 2014, 14, 2034−2047.
(50) Genovese, S.; Curini, M.; Epifano, F. Phytochemistry 2009, 70,
1082−1091.
(51) Epifano, F.; Genovese, S.; Menghini, L.; Curini, M.
Phytochemistry 2007, 68, 939−953.
(52) Marcu, M. G.; Jung, Y.-J.; Lee, S.; Chung, E.-J.; Lee, M.-J.;
Trepel, J.; Neckers, L. Med. Chem. 2006, 2, 169−174.
(53) Liu, Z.; Xie, Z.; Jones, W.; Pavlovicz, R. E.; Liu, S.; Yu, J.; Li, P.-
k.; Lin, J.; Fuchs, J. R.; Marcucci, G.; Li, C.; Chan, K. K. Bioorg. Med.
Chem. Lett. 2009, 19, 706−709.
(54) 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.
(55) Glide Version 6.1; Schrodinger, LLC: New York, NY, 2013.
̈
(56) Halgren, T. A.; Murphy, R. B.; Friesner, R. A.; Beard, H. S.;
Frye, L. L.; Pollard, W. T.; Banks, J. L. J. Med. Chem. 2004, 47, 1750−
1759.
(57) Friesner, R. A.; Banks, J. L.; Murphy, R. B.; Halgren, T. A.;
Klicic, J. J.; Mainz, D. T.; Repasky, M. P.; Knoll, E. H.; Shelley, M.;
Perry, J. K.; Shaw, D. E.; Francis, P.; Shenkin, P. S. J. Med. Chem. 2004,
47, 1739−1749.
(58) Friesner, R. A.; Murphy, R. B.; Repasky, M. P.; Frye, L. L.;
Greenwood, J. R.; Halgren, T. A.; Sanschagrin, P. C.; Mainz, D. T. J.
Med. Chem. 2006, 49, 6177−6196.
(59) Millard, C. J.; Watson, P. J.; Celardo, I.; Gordiyenko, Y.;
Cowley, S. M.; Robinson, C. V.; Fairall, L.; Schwabe, J. W. R. Mol. Cell
2013, 51, 57−67.
(60) Bressi, J. C.; Jennings, A. J.; Skene, R.; Wu, Y.; Melkus, R.; De
Jong, R.; O’Connell, S.; Grimshaw, C. E.; Navre, M.; Gangloff, A. R.
Bioorg. Med. Chem. Lett. 2010, 20, 3142−3145.
(61) Watson, P. J.; Fairall, L.; Santos, G. M.; Schwabe, J. W. R. Nature
2012, 481, 335−340.
(62) Somoza, J. R.; Skene, R. J.; Katz, B. A.; Mol, C.; Ho, J. D.;
Jennings, A. J.; Luong, C.; Arvai, A.; Buggy, J. J.; Chi, E.; Tang, J.; Sang,
B. C.; Verner, E.; Wynands, R.; Leahy, E. M.; Dougan, D. R.; Snell, G.;
Navre, M.; Knuth, M. W.; Swanson, R. V.; McRee, D. E.; Tari, L. W.
Structure 2004, 12, 1325−1334.
(63) Bottomley, M. J.; Lo Surdo, P.; Di Giovine, P.; Cirillo, A.;
Scarpelli, R.; Ferrigno, F.; Jones, P.; Neddermann, P.; De Francesco,
R.; Steinkuhler, C.; Gallinari, P.; Carfi, A. J. Biol. Chem. 2008, 283,
26694−26704.
(64) Schuetz, A.; Min, J.; Allali-Hassani, A.; Schapira, M.; Shuen, M.;
Loppnau, P.; Mazitschek, R.; Kwiatkowski, N. P.; Lewis, T. A.;
Maglathin, R. L.; McLean, T. H.; Bochkarev, A.; Plotnikov, A. N.;
Vedadi, M.; Arrowsmith, C. H. J. Biol. Chem. 2008, 283, 11355−11363.
(65) Bieliauskas, A. V.; Pflum, M. K. H. Chem. Soc. Rev. 2008, 37,
1402−1413.
(66) Thaler, F.; Mercurio, C. ChemMedChem 2014, 9, 523−536.
(67) Askari, M.; Sahebkar, A.; Iranshahi, M. Iran J. Basic Med. Sci.
2009, 12, 63−69.
(68) Koeberle, A.; Siemoneit, U.; Buehring, U.; Northoff, H.; Laufer,
S.; Albrecht, W.; Werz, O. J. Pharmacol. Exp. Ther. 2008, 326, 975−
982.
(69) Covalent Docking, protocol 2013-3. Glide Version 6.1, Prime
Version 3.4; Schrodinger, LLC: New York, NY, 2013.
̈
(70) Peret-Almeida, L.; Cherubino, A. P. F.; Alves, R. J.; Dufosse, L.;
Gloria, M. B. A. Food Res. Int. 2005, 38, 1039−1044.
(71) Maestro, Version 9.6; Schrodinger, LLC: New York, NY, 2013.
̈
(72) LigPrep, Version 2.8; Schrodinger, LLC: New York, NY, 2013.
̈
M
DOI: 10.1021/acs.jnatprod.5b00700
J. Nat. Prod. XXXX, XXX, XXX−XXX