(-)-Tarchonanthuslactone: Design of New Analogues, Evaluation of their Antiproliferative Activity on Cancer Cell Lines, and Preliminary Mechanistic Studies

Article abstract of DOI:10.1002/cmdc.201500246

Natural products containing the α,β-unsaturated δ-lactone skeleton have been shown to possess a variety of biological activities. The natural product (-)-tarchonanthuslactone (1) possessing this privileged scaffold is a popular synthetic target, but its biological activity remains underexplored. Herein, the total syntheses of dihydropyran-2-ones modeled on the structure of 1 were undertaken. These compounds were obtained in overall yields of 17-21 % based on the Keck asymmetric allylation reaction and were evaluated in vitro against eight different cultured human tumor cell lines. We further conducted initial investigation into the mechanism of action of selected analogues. Dihydropyran-2-one 8 [(S,E)-(6-oxo-3,6-dihydro-2H-pyran-2-yl)methyl 3-(3,4-dihydroxyphenyl)acrylate], a simplified analogue of (-)-tarchonanthuslactone (1) bearing an additional electrophilic site and a catechol system, was the most cytotoxic and selective compound against six of the eight cancer cell lines analyzed, including the pancreatic cancer cell line. Preliminary studies on the mechanism of action of compound 8 on pancreatic cancer demonstrated that apoptotic cell death takes place mediated by an increase in the level of reactive oxygen species. It appears as though compound 8, possessing two Michael acceptors and a catechol system, may be a promising scaffold for the selective killing of cancer cells, and thus, it deserves further investigation to determine its potential for cancer therapy. Fighting the big C: We describe the synthesis of a new family of analogues based on the scaffold of the natural product (-)-tarchonanthuslactone; these compounds were evaluated in vitro against tumor cell lines. We further conducted an initial investigation into the mechanism of action, including the inhibition of phosphatases and glutathione-S-transferase and the production of reactive oxygen species.

Full text of DOI:10.1002/cmdc.201500246

DOI: 10.1002/cmdc.201500246
Full Papers
(À)-Tarchonanthuslactone: Design of New Analogues,
Evaluation of their Antiproliferative Activity on Cancer Cell
Lines, and Preliminary Mechanistic Studies
Luiz Fernando Toneto Novaes,[a] Carolina Martins Avila,[a] Karin Juliane Pelizzaro-Rocha,[b]
DØbora Barbosa Vendramini-Costa,[c] Marina Pereira Dias,[b] Daniela Barreto Barbosa Trivella,[d]
Jo¼o Ernesto de Carvalho,[c] Carmen Veríssima Ferreira-Halder,[b] and Ronaldo Aloise Pilli*[a]
Natural products containing the a,b-unsaturated d-lactone
skeleton have been shown to possess a variety of biological
activities. The natural product (À)-tarchonanthuslactone (1)
possessing this privileged scaffold is a popular synthetic target,
but its biological activity remains underexplored. Herein, the
total syntheses of dihydropyran-2-ones modeled on the struc-
ture of 1 were undertaken. These compounds were obtained
in overall yields of 17–21% based on the Keck asymmetric ally-
lation reaction and were evaluated in vitro against eight differ-
ent cultured human tumor cell lines. We further conducted ini-
tial investigation into the mechanism of action of selected ana-
analogue of (À)-tarchonanthuslactone (1) bearing an addition-
al electrophilic site and a catechol system, was the most cyto-
toxic and selective compound against six of the eight cancer
cell lines analyzed, including the pancreatic cancer cell line.
Preliminary studies on the mechanism of action of compound
8 on pancreatic cancer demonstrated that apoptotic cell death
takes place mediated by an increase in the level of reactive
oxygen species. It appears as though compound 8, possessing
two Michael acceptors and a catechol system, may be a promis-
ing scaffold for the selective killing of cancer cells, and thus, it
deserves further investigation to determine its potential for
cancer therapy.
logues. Dihydropyran-2-one
8
[(S,E)-(6-oxo-3,6-dihydro-2H-
pyran-2-yl)methyl 3-(3,4-dihydroxyphenyl)acrylate], a simplified
Introduction
Natural products have been and continue to be rich sources
for drug discovery with high impact especially in the develop-
ment of new anticancer agents. The great structural diversity
of compounds provided by nature associated with their medic-
inal significance has served as inspiration for the design of
new lead compounds.[1–3] In this respect, natural products dis-
playing a,b-unsaturated d-lactone moieties have attracted in-
creasing interest from synthetic and medicinal chemists.[4,5]
The a,b-unsaturated d-lactone unit is a privileged scaffold
widely distributed among natural products; it possesses
a large range of biological activities, including cytotoxic activi-
ty,[6,7] antileishmanial activity,[8,9] anti-inflammatory activity,[10,11]
Figure 1. Chemical structures of representative natural products displaying
the a,b-unsaturated d-lactone scaffold: (À)-tarchonanthuslactone (1), gonio-
thalamin (2), cryptocaria diacetate (3), and cryptomoscatone D2 (4).
[a] L. F. Toneto Novaes, Dr. C. Martins Avila, Prof. Dr. R. A. Pilli
Institute of Chemistry, University of Campinas (Unicamp)
C.P. 6154, 13084-971, Campinas, SP (Brazil)
and tubulin binding properties.[12] The ability of the conjugated
double bond to act as a Michael acceptor for biological nucle-
ophiles features prominently in these molecules (Figure 1).[13–15]
Tarchonanthuslactone (1) is a natural a,b-unsaturated d-lac-
tone that was first isolated by Bohlmann and Suwita in 1979
from Tarchonanthus trilobus.[16] Tarchonanthus is a small genus
occurring in the tropics and in southern Africa used in folk
medicine: T. trilobus var. galpinii, also known as broad-leaved
camphor bush, is used to induce vomiting and to increase
libido, whereas T. camphorata has been used for diabetes by
traditional health practitioners in Africa. Its cytotoxic and anti-
inflammatory activities have also been reported.[17]
E-mail: pilli@iqm.unicamp.br
[b] Dr. K. J. Pelizzaro-Rocha, M. Pereira Dias, Prof. Dr. C. V. Ferreira-Halder
Department of Biochemistry, Biology Institute
University of Campinas (Unicamp), 13083-862, Campinas, SP (Brazil)
[c] Dr. D. B. Vendramini-Costa, Prof. Dr. J. Ernesto de Carvalho
Division of Pharmacology and Toxicology—CPQBA
University of Campinas (Unicamp), 13083-970, Campinas, SP (Brazil)
[d] Dr. D. B. Barbosa Trivella
Brazilian Biosciences National Laboratory, National Center for Research in
Energy and Material, 13083-970, Campinas, SP (Brazil)
Supporting Information for this article is available on the WWW under
http://dx.doi.org/10.1002/cmdc.201500246.
ChemMedChem 2015, 10, 1687 – 1699
1687
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
Although tarchonanthuslactone (1) has become a popular
synthetic target since its isolation,[18–31] the lack of studies on
its biological activity is rather surprising.[32] Owing to our inter-
est in investigating the cytotoxic profile of a,b-unsaturated d-
lactone inspired natural products, we devised a synthetic ap-
proach to (À)-tarchonanthuslactone (1) and its analogues. The
synthetic strategy is based on the Keck catalytic asymmetric al-
lylation of the aldehyde derived from (R)-polyhydroxybutyrate,
a biorenewable starting material, followed by the construction
of the dihydropyran-2-one moiety by a ring-closing metathesis
reaction. A similar synthetic route was employed to access the
analogues (Figure 2).
Results and Discussion
Chemistry
Initially, reductive depolymerization of commercially available
poly[(R)-3-hydroxybutyrate] (22) was performed upon treat-
ment of 22 with LiAlH4 in THF (94% yield), according to the
conditions described by Seebach and Züger.[37] Protection of
the hydroxy groups of (R)-1,3-butanediol (23) with tert-butyldi-
methylsilyl (TBS) groups and selective deprotection of the pri-
mary alcohol by using HF·pyridine (HF·py) afforded mono-TBS
derivative 25 in 54% overall yield. Swern oxidation of alcohol
25 furnished desired aldehyde 26 in 86% yield. Catalytic asym-
metric allylation of aldehyde 26 under Keck’s conditions[38] pro-
vided homoallylic alcohol 27 in 44% yield with a diastereomeric
ratio (12:1) in favor of the syn stereoisomer (aldehyde 26 was
recovered in 28% yield). The configuration of the major isomer
was established by comparison of its specific optical rotation
and NMR spectroscopy data with those reported for ent-26.[39]
The construction of a,b-unsaturated d-lactone 30 was ach-
ieved by esterification of 27 with acryloyl chloride, followed by
ring-closing metathesis by using Grubbs first-generation cata-
lyst[40] and deprotection of the secondary alcohol with HF·py
(60% yield, three steps). With alcohol 30 in hand, (À)-tarcho-
nanthuslactone (1) and corresponding analogue 5 were pre-
pared by a sequence of esterification with the corresponding
protected acids mediated by 1-ethyl-3-(3-dimethylaminopro-
pyl)carbodiimide hydrochloride (EDC·HCl), followed by depro-
tection of the catechol with tetra-n-butylammonium fluoride
(TBAF) and benzoic acid in THF (1, 58% yield; 5, 50% yield)
(Scheme 1). NMR spectroscopic data and the specific rotation
of synthetic (À)-tarchonanthuslactone (1) matched those de-
scribed for the natural product. The total synthesis of (À)-
tarchonanthuslactone (1) was achieved in 10 steps in 7% over-
all yield.[32]
Herein, we describe the design and synthesis of analogues
of (À)-tarchonanthuslactone (1) to evaluate key structural fea-
tures related to its bioactivity regarding cancer cells. First, we
investigated the effects of introducing a second electrophilic
site in 1 by following an approach similar to that conducted
previously with the natural product piperlongumine.[33]
A
second reactive electrophilic site was introduced in (À)-tarcho-
nanthuslactone analogue 5, which thus generated a new a,b-
unsaturated system. Further, the role of this second electro-
philic site was also evaluated in compounds 6, 8, 10, and 12.
Additionally, we explored the impact of the stereogenic center
at C7 by using analogues 6–13, which were also developed by
targeting molecular simplification of the natural product. The
R1 and R2 substituents in the aromatic ring of the simplified an-
alogues were replaced by 3,4-dihydroxy substituents (see com-
pounds 8 and 9), which are present in (À)-tarchonanthuslac-
tone (1) and in biologically active natural products such as caf-
feic acid and its derivatives,[34,35] and by the 3,4-dimethoxy
group (see compounds 10 and 11) and the 3-methoxy-4-hy-
droxy group (see compounds 12 and 13) which is present in
antineoplastic agents such as combretastatin A4.[36] Finally, we
removed the a,b-unsaturated d-lactone moiety and examined
the biological profile of methyl esters 14–21 as a proof of con-
cept of the importance of the lactone moiety (Figure 2).
A similar synthetic approach was employed to access simpli-
fied analogues 6–13 (Scheme 2). This time, Keck allylation of al-
dehyde 35 afforded homoallylic alcohol 36 in a yield higher
than that observed for its superior homologue 27 (Scheme 1):
whereas alcohol 27 was obtained in 44% yield accompanied
Figure 2. Design concept for analogues of (À)-tarchonanthuslactone (compounds 5–21).
ChemMedChem 2015, 10, 1687 – 1699
www.chemmedchem.org
1688
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
The syntheses of protected
carboxylic acids 31 and 32,[41]
derived from ferulic acid,[41,42]
and methyl esters 14–21[42,43]
were performed by using an ap-
proach similar to that described
in the literature.
Biology
Effect of (À)-tarchonanthus-
lactone and its analogues in
cancer cell lines
To assess the differences in sen-
sitivity displayed by each cell
line to the same compound, (À)-
tarchonanthuslactone (1), ana-
logues 5–13, and methyl esters
14–21 were evaluated in vitro
against eight different cultured
human tumor cell lines: glioma
(U-251), breast (MCF-7), ovary ex-
pressing the multidrug resist-
ance phenotype (NCI/ADR-RES),
kidney (786-0), lung non-small
cells (NCI-H460), prostate (PC-3),
colon (HT-29), and pancreas
(Panc-1). Additionally, the com-
pounds were assayed in vitro
against spontaneously trans-
formed keratinocytes from histo-
logically normal skin (HaCat
cells). We further investigated
the effects of the natural prod-
uct and selected analogues in
candidate target enzymes.
Scheme 1. Synthesis of (À)-tarchonanthuslactone (1) and analogue 5. Reagents and conditions: a) LiAlH4, THF,
reflux then RT, 94%; b) TBSCl, imidazole, CH2Cl2, RT, 98%; c) HF·py, pyridine, THF, RT, 55%; d) Swern oxidation,
À788C, 86%; e) 1. (S)-BINOL, Ti(OiPr)4, TFA, CH2Cl2, 4Š molecular sieves, reflux; 2. Bu3SnAllyl, À78 to À308C, 44%,
dr=12:1; f) acryloyl chloride, Et3N, CH2Cl2, 08C, 75%; g) Grubbs first-generation catalyst, CH2Cl2, reflux, 80%;
h) HF·py, pyridine, THF, RT, quant.; i) 1. 31 or 32, EDC·HCl, DMAP, CH2Cl2, RT; 2. TBAF, PhCO2H, THF, 08C, 58% (1)
and 50% (5).
Cell growth was determined
spectrophotometrically by the
sulforhodamine B (SRB) assay,
and the analyses were based on
the U. S. National Cancer Insti-
tute (NCI) 60 tumor cell line anti-
cancer drug screen (NCI60).[44]
One of the advantages of the
SRB assay is the possibility to
measure the cell population
density at time zero (the time at
which drugs are added, t0), as
this allows for the calculation of
the cellular responses for total
Scheme 2. Synthesis of simplified analogues of (À)-tarchonanthuslactone (1). Reagents and conditions: a) TBDPSCl,
CH2Cl2, RT, quant.; b) OsO4, NaIO4, 2,6-lutidine, H2O, 1,4-dioxane, RT, 72%; c) 1. (S)-BINOL, Ti(OiPr)4, TFA, CH2Cl2, 4 Š
molecular sieves, reflux; 2. Bu3SnAllyl, À78 to À208C, 87%, er=19:1; d) acryloyl chloride, Et3N, CH2Cl2, 08C, 63%;
e) Grubbs first-generation catalyst, CH2Cl2, reflux, 68%; f) TBAF, PhCO2H, THF, 08C to RT 90%; g) 1. carboxylic acid,
EDC·HCl, DMAP, CH2Cl2, RT; for 10–13: 2. TBAF, PhCO2H, THF, 08C.
by recovery of aldehyde 26 (28%), aldehyde 35 afforded corre-
sponding alcohol 36 in 87% yield under the same reaction
conditions. The increased reactivity may be ascribed to the ac-
tivated carbonyl moiety resulting from the vicinal tert-butyldi-
phenylsilyloxy (OTBDPS) group in 35, as this activating effect is
attenuated by the b-silyloxy group in 26.
growth inhibition (TGI). Moreover, this method displays practi-
cal advantages for large-scale screening.[45] The TGI values
were calculated from t=t0, for which the amount of cells at
the end of drug incubation (t), 48 h of treatment, was equal to
the amount at the beginning (t0). The compounds were as-
sayed at concentrations between 0.16 and 250 mgmLÀ1, and
ChemMedChem 2015, 10, 1687 – 1699
www.chemmedchem.org
1689
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
doxorubicin (DOX), employed as a positive control, at a concen-
tration of 0.025–25 mgmLÀ1. The results of the cancer cell line
screen are summarized in Table 1.
in which the parent compounds were active. In summary, two
moieties are important for the antiproliferative activity of (À)-
tarchonanthuslactone derivatives: 1) the presence of the
second Michael acceptor (present in compounds 5, 6, 8, 10,
and 12) and 2) the presence of the catechol system (found in
compounds 8 and 9). These two molecular features, along
with structural simplification of the natural product, are com-
bined in compound 8, and thus, 8 is a promising scaffold for
the selective killing of cancer cells.
All the compounds shown in Figure 2 were evaluated
against a panel of the eight cancer cell lines listed above, but
only compounds 1 and 5–13 displayed any cytotoxic effect
toward U-251 (glioma), MCF-7 (breast), 786-0 (kidney), PC-3
(prostate), and Panc-1 (pancreas), and for that reason, the data
for ovary cells expressing the multidrug resistance phenotype
(NCI/ADR-RES), lung non-small cells (NCI-H460), and colon (HT-
29) are not shown in Table 1. To start, all the dihydropyran-2-
ones investigated were much less toxic to human keratino-
cytes (HaCat) than doxorubicin. Compound 1 exhibited anti-
proliferative properties against three cell lines assayed (i.e.,
glioma, kidney and pancreas). Selectivity of 1 to cancer cell
lines was reasonable relative to transformed keratinocytes
(HaCat, TGI>100 mm).
The introduction of a second electrophilic site led to a dis-
crete increase in the cytotoxic effect of compound 5 relative to
that observed for tarchonanthuslactone (1), particularly for the
U-251 and Panc-1 cell lines. The cytotoxic effects of compound
5 to the highly aggressive pancreas cell line Panc-1 (TGI=
23.8 mm) are more pronounced than the cytotoxic effects of
compound 1 (TGI=81.8 mm). Analogue 8 displayed a better
profile for breast (MCF-7) and prostate (PC-3) relative to that
shown by analogue 5. Notably, only analogue 8 displayed any
cytotoxic effect against the breast cancer cell line (MCF-7). The
beneficial effect of an additional double bond with electrophil-
ic capacity was also evident upon comparing the profiles of
the pairs of compounds 6 /7, 8/9, 10/11, and 12/13.
Catechol derivatives have metal-chelating properties and
can act as reducing agents.[46,47] Additionally, catechol can un-
dergo transformation into quinone intermediates that are able
to react with cellular nucleophiles, such as the sulfhydryl
groups of cysteines, to form catechol–protein conjugates.[48]
Furthermore, quinones can drive the formation of reactive
oxygen species (ROS), which are capable of oxidizing several
biological molecules.[49] On the other hand, natural products
with a catechol motif, for example, quercetin, possess anticanc-
er, antioxidant, and anti-inflammatory properties.[50]
Methyl esters 14–21 did not display significant activity
(TGI>100 mm) against the human cancer cell lines assayed.
These results indicate that the a,b-unsaturated d-lactone scaf-
fold imparts a significant role in determining the cytotoxicity
of (À)-tarchonanthuslactone analogues, as speculated previ-
ously.
Additionally, all compounds assayed have high TGI values
(TGI>100 mm) for the human keratinocytes cell line (HaCat),
which reveals selectivity to cancer cells.
(À)-Tarchonanthuslactone analogues induce apoptotic cell
In general, simplified analogues lacking a C8ÀC9 olefin (i.e.,
compounds 7, 9, 11, and 13) were less cytotoxic in all sensitive
cancer cell lines than the corresponding analogues that carry
the second electrophilic site (i.e., compounds 6, 8, 10 and 12).
Among the simplified analogues, compound 8 stands out as
the only analogue that is effective against the breast cancer
cell line (MCF-7, TG1=27.4 mm). Furthermore, 8 retained its po-
tency or showed enhanced potency against the other cell lines
death in pancreatic cancer cells
Recently, our research group has been interested in discover-
ing new natural or synthetic compounds that display therapeu-
tic potential against pancreatic cancer and in understanding
their molecular mechanisms of action.[51] Pancreatic cancer is
highly aggressive and has poor prognosis. This is mainly due
to the rapid development of chemoresistance to drug therapy,
Table 1. Total growth inhibition (TGI) values for (À)-tarchonanthuslactone (1), analogues 5–13, and doxorubicin (DOX).[a]
Compd
TGI [mm]
U-251
MCF-7
>100
>100
>100
>100
786-0
PC-3
>100
>100
58.1Æ17.5
>100
Panc-1
HaCat
>100
1
5
6
7
8
58.9Æ13.9
20.1Æ3.8
19.0Æ2.5
42.1Æ9.2
30.1Æ8.7
41.3Æ11.8
81.8Æ0.4
23.8Æ7.9
85.4Æ0.7
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
17.6Æ5.0
28.9Æ8.1
17.7Æ10.6
56.6Æ8.7
36.0Æ6.0
46.7Æ6.6
0.18Æ0.1
27.4Æ12.2
31.7Æ6.1
92.0Æ10.8
33.1Æ3.3
30.1Æ6.1
24.6Æ1.0
67.1Æ5.0
25.1Æ0.8
93.2Æ22.7
23.4Æ1.7
95.6Æ3.0
28.8Æ5.3
9
>100
>100
>100
>100
>100
69.4Æ8.9
48.3Æ9.9
10
11
12
13
DOX
>100
>100
38.9Æ24.1
95.0Æ4.9
1.9Æ0.4
33.1Æ16.6
>100
12.8Æ1.0
6.6Æ1.2
34.6Æ1.4
[a] Concentration that elicits TGI was determined from nonlinear regression analysis by using ORIGIN 8.0 (OriginLab Corporation). DOX was the positive
control. Tested compounds were not effective (TGI>100 mm) against NCI/ADR-RES, NCI-H460, and HT-29 cell lines. Test compounds were evaluated against
glioma (U-251), breast (MCF-7), kidney (786-0), prostate (PC-3), and pancreas (Panc-1) cancer cell lines and transformed keratinocytes (HaCat). Results repre-
sent the average of two independent experiments performed in triplicate. Values are the meanÆSEM.
ChemMedChem 2015, 10, 1687 – 1699
www.chemmedchem.org
1690
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
such as resistance to the chemotherapeutic agent gemcitabine,
the standard drug for pancreatic cancer disease.[52] In this con-
text, we selected the human pancreatic carcinoma cell line
(Panc-1) for additional studies among the cancer cell lines sen-
sitized by the synthetic compounds developed herein (Table 1).
The Panc-1 cell line presents higher chemoresistance to drug
therapy than others pancreatic cell lines available for anticanc-
er screening, such as BxPC-3, AsPC-1, and Mia-PaCa-2.[53]
After the initial screening, we selected the (À)-tarchonan-
thuslactone analogues to evaluate their ability to induce apop-
totic cell death. The experiments were performed by using
double staining for annexin V and 7-aminoactinomycin D (7-
AAD) and flow cytometry in Panc-1 cells. Two representative
compounds were selected: simplified analogue 8, which pos-
sesses two Michael acceptor functionalities and the catechol
moiety, and corresponding analogue 9, in which the C8=C9
double bond is removed.
As observed in Figure 3b, compound 8 increased significant-
ly the level of reactive oxygen species in Panc-1 cells, even at
low concentration (25 mm). On the other hand, compound 9,
which lacks the C8ÀC9 olefin, was unable to elicit the same
effect, as it only slightly changed the ROS levels at high con-
centrations (100 mm). Our results point to a positive correlation
between the rate of apoptosis and the generation of ROS in-
duced by analogue 8.
Insight from cancer biology suggests that increasing the
ROS levels may be a strategy to selectively target cancer
cells.[57–60] Previous reports have suggested that cancer cells
may be particularly sensitive to ROS-modulating small mole-
cules, especially to electrophilic small molecules.[33,58]
ROS function as signaling molecules, and they are produced
by normal cells during normal metabolic activities and are
cleared by the endogenous antioxidant systems. In cancer
cells, antioxidant systems are paradoxically increased, and it
has been shown that these cells are strictly dependent on the
antioxidant machinery.[58,61] Cancer cells are in constant oxida-
tive stress, which is derived from their high metabolic rate.
There are at least two possible mechanisms by which small
molecules can contribute to the increase in ROS levels and to
the selective killing of cancer cells: 1) ROS generation by the
an exogenous small molecule and 2) inhibition of the antioxi-
dant system of the cell.[58,59]
Both compounds 8 and 9 were able to induce apoptotic cell
death at concentrations greater than 50 mm (Figure 3a). In ad-
dition, our results clearly show that analogue 8 is more potent
in inducing apoptosis than analogue 9 (100 mm, p<0.001).
Currently, apoptotic cell death is the desired goal of many
cancer treatments, and therefore, apoptotic inducers and regu-
lators are considered to be of significant potential for cancer
therapy.[54] Apoptotic cell death can result from pro-apoptotic
signals which are dispatched by damaged DNA or from re-
sponse to oxidative stress.[55] ROS have been considered as cy-
totoxic products of cellular metabolism, and the overproduc-
tion of ROS in cells may enhance cell death by apoptosis.[56]
For this reason, we next examined whether the analogues of
(À)-tarchonanthuslactone (1) had the ability to generate en-
dogenous ROS in Panc-1 cells. Cells treated with analogues 8
and 9 were stained with the ROS-specific 2’-7’-dichlorodihydro-
fluorescein diacetate (DCFH-DA) dye and were analyzed by
flow cytometry. Inside the cell, nonfluorescent DCFH-DA is hy-
drolyzed to the polar derivative 2’-7’-dichlorodihydrofluores-
cein (DCFH), which is oxidized in the presence of H2O2 to fluo-
rescent 2’,7’-dichlorofluorescein (DCF).
As mentioned above, the catechol system is known to gen-
erate radicals through its oxidation to quinones, and the latter
are excellent radical stabilizers and ROS generators. Ubiqui-
none, in the electron-transport chain, and several natural prod-
ucts with antiproliferative activity (e.g., caulibugulones[62]) are
good examples of quinones that are ROS generators. ROS can
oxidase several biological molecules, including DNA, proteins,
and lipids, which causes cell damage and influences signaling
pathways. Owing to the increased ROS levels and the extreme
dependence of cancer cells to the stress machinery of ROS, the
selective killing of cancer cells was observed upon perturbing
the ROS balance in these cells.
Figure 3. Measurement of apoptosis and intracellular ROS production in Panc-1 cells. Cells were treated with analogues 8 and 9 at 25, 50, and 100 mm for
24 h and then stained with a) annexin-V and 7-AAD to evaluate apoptosis and b) DCFH-DA to evaluate ROS levels. Both cells were analyzed by flow cytometry
as described in the Experimental Section. Data are represented as meanÆSEM of three independent experiments performed in duplicate. b: p<0.01 versus
control, a: p<0.001 versus control, and *: p<0.001 versus compound 9 (100 mm). Statistical analysis was assessed with ANOVA followed by Tukey test.
ChemMedChem 2015, 10, 1687 – 1699
www.chemmedchem.org
1691
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
Our results indicate the relationship between ROS genera-
tion and Panc-1 apoptotic cell death. Therefore, we investigat-
ed possible molecular mechanisms that would lead to this
phenomenon.
Protein tyrosine phosphatase inhibition and ROS generation
by selected analogues of 1
First, we investigated the inhibition of protein tyrosine phos-
phatase (PTPase) by selected analogues of tarchonanthuslac-
tone (1). PTPases are important enzymes that play central roles
in cell signaling. It has been reported that several of these en-
zymes are overexpressed in cancer cells.[63] Furthermore, PTPas-
es are known to be negatively regulated by ROS.[64–66] In partic-
ular, some cancer cell lines overexpress the low molecular
weight protein tyrosine phosphatase (LMW-PTP), and this has
been correlated to the aggressiveness of these tumors.[67]
Therefore, we investigated the effects of the analogues of (À)-
tarchonanthuslactone (1) in the activity of three representative
members of PTPases, LMW-PTP, CDC-25B, and PTP-1B, all in-
volved in cancer, and their correlation with the generation of
ROS by the selected compounds (Figure 4).
We evaluated the inhibition of PTPases by selected ana-
logues of 1 by using a standard PTPase assay. Experiments
were performed with the purified enzymes and p-nitrophenyl
phosphate as the substrate. Two pH values were selected for
measuring the PTPase activity, according to their optimal pH
ranges for catalytic activity. Results show that 1 and its ana-
logues that bear the catechol system are micromolar inhibitors
of CDC-25B and PTP-1B (Figure 4a).
Figure 4. a) Protein tyrosine phosphatase inhibition by selected analogues
of 1; data are represented as meanÆSD of experiments performed in tripli-
cate. b) Generation of ROS by the compounds in the enzyme assays; data
are represented as meanÆSEM of experiments performed in duplicate. The
PTPases LMW-PTP, PTP-1B, and CDC-25B were analyzed at pH 5.5 (for PTP-1B
and LMW-PTP) and pH 8.2 (for CDC-25B and PTP-1B). The inhibitor concen-
tration was 100 mm.
ROS generation was measured directly from the PTPase
assays. For this, we used the FOX method,[68,69] which uses the
oxidation of iron as a sensor for detecting ROS. Indeed, we ob-
served that the compounds containing the catechol system
were able to generate ROS in an alkaline medium (pH 8.2) (Fig-
ure 4b).
cule antioxidant that plays a central role in the antioxidant
system of cells. We thus measured the interaction of reduced
glutathione (GSH) with analogue 8 by 1H NMR spectroscopy
(Figure 5). A solution of analogue 8 (1 mgmLÀ1) in D2O con-
taining the surfactant trisaminol was treated with GSH
(7 equiv) and NMR spectra were recorded every 5 min. We ob-
served that GSH quickly performed conjugate addition to the
dihydropyranone, as indicated by the complete disappearance
of the signals of hydrogen atoms at C2 and C3 in the first-re-
corded spectrum after the addition of GSH (t=5 min).
PTPases display an activated cysteine residue in their catalyt-
ic sites.[70] It has been reported that oxidation of this cysteine
residue can inactivate PTPases in a reversible or irreversible
fashion.[62,66] The strong correlation between the generation of
ROS and the inhibition of PTPase strongly suggests that
enzyme oxidation, probably induced by the generation of ROS
by compounds with the catechol system, takes place. We ob-
served in the cell assays that the presence of the catechol
system in these analogues increased cell sensitivity to the com-
pounds. Possibly, the generation of ROS by these compounds
can induce enzyme oxidation in the cells, which therefore con-
tributes to their potency.
Cell-free analysis of (À)-tarchonanthuslactone analogues
Stress-related enzymes such as catalase (CAT), glutathione re-
ductase (GR), superoxide dismutase (SOD), and glutathione-S-
transferase (GST) play central roles in the antioxidant and sig-
naling systems of cells and are considered attractive targets for
cancer chemotherapy.[58] From those, GST has been reported to
act directly and indirectly in the stress response to ROS[71,72]
and, further, to be one of the enzyme targets for small mole-
cules that selectively kill cancer cells.[57] We, therefore, evaluat-
ed the effects of selected (À)-tarchonanthuslactone analogues
in inhibiting GST in a cell-free system.
Antioxidant machinery depletion by selected analogues of 1
Aiming at investigating the influence of the compounds on
the antioxidant machinery of cells, we also analyzed the influ-
ence of selected derivatives of 1 on the enzymatic antioxidant
systems of cells. In the context of the inhibition of the antioxi-
dant systems of cells, there are reports of small molecules in-
terfering with the tripeptide glutathione, a known small-mole-
A GST inhibition assay was performed by using purified re-
combinant S. japonicum GST. First, the compounds were as-
ChemMedChem 2015, 10, 1687 – 1699
www.chemmedchem.org
1692
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
Figure 5. a) 1H NMR spectrum (D2O, trisaminol, 500 MHz) of analogue 8;
b) 1H NMR spectrum (D2O, trisaminol, 500 MHz) of analogue 8 treated with
GSH after 5 min.
sayed at a single concentration (i.e., 100 mm) to select potential
GST inhibitors. Subsequently, with the best inhibitors identified
concentration–response curves were generated by using the
concentration range of 1 to 100 mm (Figure 6).
This assay allowed us to find GST inhibitors such as 13 with
an IC50 value of 11 mm and 9 with a IC50 of 18 mm (Table 2). De-
spite the low micromolar inhibition of these analogues, there
is no correlation between GST inhibition and cytotoxic activity.
Therefore, we suggest that GST is not the main enzyme target
of 1 and its analogues. Other studies to evaluate potential pro-
tein targets for the studied compounds are in progress.
Figure 6. In vitro glutathione-S-transferase inhibition assay with selected
compounds. a) GST inhibition screening with compounds at 100 mm; b) GST
concentration–response curves with selected inhibitors; data are represent-
ed as meanÆSD of experiments performed in triplicate.
induced cancer cell lines. Pancreatic cancer is one of the dead-
liest of the solid malignancies. Growth inhibition and induction
of apoptosis constitute the major mechanisms of action of
most chemotherapeutics during cancer. Unfortunately, pancre-
atic cancer is inherently resistant to apoptosis in all conven-
tional cancer therapeutic agents, which poses a great chal-
lenge to clinicians for its treatment. We therefore focused on
this cancer type, and a preliminary study of the mechanism of
action of compound 8 in pancreatic cancer cells demonstrated
that apoptotic cell death mediated by reactive oxygen species
takes place.
Conclusions
The natural product (À)-tarchonanthuslactone (1), simplified
analogues 5–13, and methyl esters 14–21 were evaluated in vi-
tro against eight different cultured human tumor cell lines. We
also conducted an initial investigation into the mechanism of
action of selected analogues. Compound 8, which bears an ad-
ditional electrophilic site relative to 1 and a catechol system,
was the most cytotoxic and selective analogue of 1 evaluated.
Compound 8 elicited increased sensitivity to six of the eight
cancer cell lines analyzed, including pancreatic and hormone-
Experimental Section
Table 2. IC50 values for in vitro GST inhibition.[a]
Chemistry
Starting materials and reagents were obtained from commercial
sources and were used as received unless otherwise specified. Di-
chloromethane was treated with calcium hydride and was distilled
before use. Tetrahydrofuran was treated with metallic sodium and
benzophenone and was distilled before use. Anhydrous reactions
were performed with continuous stirring under an atmosphere of
dry nitrogen. Progress of the reactions was monitored by thin-layer
chromatography (TLC) analysis (Merck, silica gel 60 F254 on alumi-
num plates). Melting points were recorded with an Electrother-
mal 9100 apparatus. 1H NMR and 13C NMR spectra were recorded
Compd
IC50 [mm]
1
6
7
8
9
>100
68Æ30
>100
ꢀ100
18Æ4
11 Æ3
13
[a] Data are the meanÆSEM of at least two independent experiments
performed in triplicate.
ChemMedChem 2015, 10, 1687 – 1699
www.chemmedchem.org
1693
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
with Bruker 250, 400, 500, and 600 spectrometers, and the chemi-
cal shifts (d) are reported in parts per million (ppm) relative to sig-
nals of the deuterated solvent as the internal standards (CDCl3
dH =7.26 ppm, dC =77.00 ppm; [D6]acetone dH =2.05 ppm, dC =
29.92 ppm; [D4]methanol dH =3.31 ppm, dC =49.15 ppm). Mass
spectra were recorded with a Waters Xevo Q-Tof apparatus operat-
ing in electrospray ionization (ESI) mode. Fourier-transformed infra-
red (FTIR) spectra were recorded with a Thermo Scientific Nicolet
iS5. The specific rotations were measured at 258C with a PerkinElm-
er 341 polarimeter and a sodium lamp. The purities of the target
compounds were determined by HPLC with a Waters Alliance ap-
paratus with a C18 column (5 mm, 4.6 mm”150 mm), eluting with
a gradient of acetonitrile and water. The preparation of tarchonan-
thuslactone (1) and dehydro analogue 5 is described in Ref. [32].
(S)-(6-Oxo-3,6-dihydro-2H-pyran-2-yl)methyl 3-(3,4-dimethoxy-
phenyl)propanoate (11). Yield: 83 mg (84%). Colorless oil; Rf =0.11
1
(SiO2, hexanes/EtOAc 60:40); [a]2D5 =À23 (c=1.0 in MeOH); H NMR
(600 MHz, CDCl3): d=2.21–2.34 (m, 2H), 2.64 (t, J=7.6 Hz, 2H),
2.87 (t, J=7.6 Hz, 2H), 3.81 (s, 3H), 3.83 (s, 3H), 4.24 (m, 2H), 4.57
(dq, J=11.5, 4.7 Hz, 1H), 5.98 (ddd, J=9.8, 2.5, 0.9 Hz, 1H), 6.68–
6.72 (m, 2H), 6.74–6.77 (m, 1H), 6.83 ppm (ddd, J=9.6, 5.8, 2.6 Hz,
1H); 13C NMR (151 MHz, CDCl3): d=25.7, 30.4, 35.8, 55.8, 55.9, 64.5,
75.1, 111.3, 111.7, 120.2, 121.2, 132.7, 144.5, 147.5, 148.9, 163.2,
172.5 ppm; IR (NaCl): n˜ =1027, 1156, 1260, 1516, 1732, 2836,
2938 cmÀ1; HRMS: m/z: calcd for C17H20O6 +H+: 321.1338 [M+H]+;
found: 321.1396.
General procedure for esterification and deprotection to pro-
duce 8, 9, 12, and 13: A solution of acid (0.63 mmol, 2 equiv) and
alcohol 39 (40 mg, 0.31 mmol, 1 equiv) in anhydrous CH2Cl2 (2 mL)
was added to a solution of EDC·HCl (119 mg, 0.63 mmol, 2 equiv)
and DMAP (38 mg, 0.31 mmol, 1 equiv) in anhydrous CH2Cl2 (9 mL)
at 258C, and the mixture was stirred at 258C for 6 h. Upon comple-
tion, the mixture was diluted with EtOAc (90 mL) and extracted
with an aqueous solution of 0.5m HCl (40 mL). The organic phase
was washed with an aqueous saturated solution of NaHCO3
(30 mL), dried (MgSO4), and concentrated. The crude material was
dissolved in anhydrous THF (20 mL) and benzoic acid (76 mg,
0.62 mmol, 2 equiv for 8 and 9; 38 mg, 0.31 mmol, 1 equiv for 12
and 13) and a solution of 1m TBAF in THF (0.62 mL, 0.62 mmol,
2 equiv for 8 and 9; 0.31 mL, 0.31 mmol, 1 equiv for 12 and 13)
were added at 08C. The mixture was stirred at this temperature for
1 h, and then an aqueous solution of 1m HCl (30 mL) was added,
and the aqueous phase was extracted with EtOAc (2”60 mL). The
organic phases were grouped, washed with brine (30 mL), and
dried (MgSO4). The product was purified by column chromatogra-
phy (SiO2, hexanes/EtOAc 30:70 for 8 and 9; hexanes/EtOAc 60:40
for 12 and 13) to afford the ester.
General procedure for esterification to produce 6, 7, 10, and 11:
A solution of the acid (0.63 mmol, 2 equiv) and alcohol 39 (40 mg,
0.31 mmol, 1 equiv) in anhydrous CH2Cl2 (2 mL) was added to a so-
lution of EDC·HCl (119 mg, 0.63 mmol, 2 equiv) and 4-dimethylami-
nopyridine (DMAP; 38 mg, 0.31 mmol, 1 equiv) in anhydrous
CH2Cl2 (9 mL) at 258C, and the mixture was stirred at 258C for 6 h.
Upon completion, the mixture was diluted with EtOAc (90 mL) and
extracted with an aqueous solution of 0.5m HCl (40 mL). The or-
ganic phase was washed with an aqueous solution of saturated
NaHCO3 (30 mL), dried (MgSO4), and concentrated. The product
was purified by column chromatography (SiO2, hexanes/EtOAc
60:40) to afford the corresponding ester.
(S)-(6-Oxo-3,6-dihydro-2H-pyran-2-yl)methyl cinnamate (6). Yield:
70 mg (87%). White solid; Rf =0.31 (SiO2, hexanes/EtOAc 60:40);
mp: 106–1098C; [a]2D0 =À39 (c=1.0 in CHCl3); 1H NMR (500 MHz,
CDCl3): d=2.38–2.55 (m, 2H), 4.41 (d, J=4.7 Hz, 2H), 4.58 (dq, J=
11.4, 4.6 Hz, 1H), 6.02–6.07 (m, 1H), 6.45 (d, J=16.0 Hz, 1H), 6.90
(ddd, J=9.8, 5.9, 2.6 Hz, 1H), 7.35–7.39 (m, 3H), 7.49–7.53 (m, 2H),
7.71 ppm (d, J=16.0 Hz, 1H); 13C NMR (126 MHz, CDCl3): d=25.9,
64.7, 75.4, 117.1, 121.4, 128.2 (2C), 129.0 (2C), 130.6, 134.1, 144.6,
145.9, 163.3, 166.5 ppm; IR (KBr): n˜ =1168, 1244, 1635, 1716,
2919 cmÀ1; HRMS: m/z: calcd for C15H14O4 +H+: 259.0970 [M+H]+;
found: 259.1020.
(S,E)-(6-Oxo-3,6-dihydro-2H-pyran-2-yl)methyl 3-(3,4-dihydroxy-
phenyl)acrylate (8). Yield: 65 mg (72%). Brown oil; Rf =0.44 (SiO2,
hexanes/EtOAc 30:70); [a]2D5 =À28 (c=1.0 in MeOH); 1H NMR
(600 MHz, [D4]methanol): d=2.42–2.53 (m, 2H), 4.37–4.40 (m, 2H),
4.77 (dq, J=10.3, 4.9 Hz, 1H), 4.85 (brs, 2H), 5.98–6.02 (m, 1H),
6.29 (d, J=15.9 Hz, 1H), 6.78 (d, J=8.2 Hz, 1H), 6.96 (dd, J=8.2,
2,1 Hz, 1H), 7.03–7.07 (m, 2H), 7.58 ppm (d, J=15.9 Hz, 1H);
13C NMR (151 MHz, [D4]methanol): d=26.6, 65.7, 77.3, 114.4, 115.2,
116.5, 121.2, 123.1, 127.6, 146.8, 147.6, 147.8, 149.7, 166.0,
168.6 ppm; IR (NaCl): n˜ =1161, 1256, 1600, 1632, 1699, 2958,
3401 cmÀ1 (broad); HRMS: m/z: calcd for C15H14O6 +H+: 291.0869
[M+H]+; found: 291.0971.
(S)-(6-Oxo-3,6-dihydro-2H-pyran-2-yl)methyl
3-phenylpropa-
noate (7). Yield: 71 mg (88%). Colorless oil; Rf =0.31 (SiO2, hex-
1
anes/EtOAc 60:40); [a]2D5 =À66 (c=1.0 in CHCl3); H NMR (500 MHz,
CDCl3): d=2.22–2.36 (m, 2H), 2.69 (t, J=7.6 Hz, 2H), 2.96 (t, J=
7.6 Hz, 2H), 4.26 (d, J=4.6 Hz, 2H), 4.58 (dq, J=11.1, 4.7 Hz, 1H),
6.00 (ddd, J=9.8, 2.5, 1.1 Hz, 1H), 6.85 (ddd, J=9.8, 5.8, 2.8 Hz,
1H), 7.17–7.22 (m, 3H), 7.26–7.30 ppm (m, 2H); 13C NMR (126 MHz,
CDCl3): d=25.5, 30.6, 35.4, 64.4, 75.0, 120.9, 126.2, 128.2 (2C), 128.4
(2C), 140.0, 144.6, 163.1, 172.3 ppm; IR (NaCl): n˜ =1081, 1104, 1161,
1246, 1389, 1454, 1497, 1604, 1732, 2951, 3028, 3062 cmÀ1; HRMS:
m/z: calcd for C15H16O4 +H+: 261.1127 [M+H]+; found: 261.1176.
(S)-(6-Oxo-3,6-dihydro-2H-pyran-2-yl)methyl
3-(3,4-dihydroxy-
phenyl)propanoate (9). Yield: 67 mg (74%). Pale-yellow oil; Rf =
0.44 (SiO2, hexanes/EtOAc 30:70); [a]2D5 =À22 (c=1.0 in MeOH);
1H NMR (600 MHz, [D4]methanol): d=2.27–2.31 (m, 2H), 2.61 (t, J=
7.3 Hz, 2H), 2.77 (t, J=7.4 Hz, 2H), 4.21 (dd, J=12.1, 3.6 Hz, 1H),
4.26 (dd, J=12.2, 5.5 Hz, 1H), 4.60 (tdd, J=8.1, 5.3, 3.6 Hz, 1H),
4.83 (brs, 2H), 5.95 (dt, J=9.8, 1.9 Hz, 1H), 6.52 (dd, J=8.1, 2.1 Hz,
1H), 6.64 (d, J=2.1 Hz, 1H), 6.66 (d, J=7.9 Hz, 1H), 6.96 ppm (dt,
J=9.4, 4.3 Hz, 1H); 13C NMR (151 MHz, [D4]methanol): d=26.4,
31.3, 36.9, 65.5, 77.0, 116.4, 116.5, 120.6, 120.9, 133.3, 144.6, 146.1,
147.9, 166.0, 174.2 ppm; IR (NaCl): n˜ =1083, 1114, 1260, 1446, 1520,
1604, 1716, 2958, 3315 cmÀ1 (broad); HRMS: m/z: calcd for
C15H16O6 +H+: 293.1025 [M+H]+; found: 293.1104.
(S,E)-(6-Oxo-3,6-dihydro-2H-pyran-2-yl)methyl 3-(3,4-dimethoxy-
phenyl)acrylate (10). Yield: 83 mg (84%). Colorless oil; Rf =0.14
1
(SiO2, hexanes/EtOAc 60:40); [a]2D5 =À26 (c=1.0 in MeOH); H NMR
(500 MHz, CDCl3): d=2.37–2.53 (m, 2H), 3.87 (s, 3H), 3.87 (s, 3H),
4.38 (d, J=4.6 Hz, 2H), 4.71 (dq, J=11.4, 4.6 Hz, 1H), 6.00–6.04 (m,
1H), 6.30 (d, J=15.9 Hz, 1H), 6.83 (d, J=8.2 Hz, 1H), 6.89 (ddd, J=
9.8, 5.8, 2.6 Hz, 1H), 7.01 (d, J=1.7 Hz, 1H), 7.71 (dd, J=8.4, 1.8 Hz,
1H), 7.62 ppm (d, J=15.9 Hz, 1H); 13C NMR (126 MHz, CDCl3): d=
25.9, 55.9, 56.0, 64.5, 75.4, 109.7, 111.1, 114.7, 121.3, 122.9, 127.1,
144.6, 145.8, 149.2, 151.4, 163.3, 166.7 ppm; IR (NaCl): n˜ =1140,
1159, 1260, 1513, 1716, 2937 cmÀ1; HRMS: m/z: calcd for C17H18O6 +
H+: 319.1182 [M+H]+; found: 319.1221.
(S,E)-(6-Oxo-3,6-dihydro-2H-pyran-2-yl)methyl
3-(4-hydroxy-3-
methoxyphenyl)acrylate (12). Yield: 68 mg (72%). Colorless oil;
Rf =0.11 (SiO2, hexanes/EtOAc 60:40); [a]2D5 =À29 (c=1.0 in MeOH);
ChemMedChem 2015, 10, 1687 – 1699
www.chemmedchem.org
1694
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
1H NMR (500 MHz, [D6]acetone): d=2.52–2.56 (m, 2H), 3.92 (s, 3H),
4.39 (d, J=4.9 Hz, 2H), 4.74–4.81 (m, 1H), 5.97 (dt, J=9.8, 1.8 Hz,
1H), 6.44 (d, J=16.0 Hz, 1H), 6.88 (d, J=8.1 Hz, 1H), 7.05 (ddd, J=
9.8, 4.4, 4.1 Hz, 1H), 7.15 (dd, J=8.1, 1.8 Hz, 1H), 7.36 (d, J=1.8 Hz,
1H), 7.65 ppm (d, J=15.9 Hz, 1H); 13C NMR (126 MHz, [D6]acetone):
d=26.3, 56.3, 65.3, 76.3, 111.2, 111.3, 115.1, 116.0, 121.3, 124.2,
127.3, 146.5, 148.7, 150.1, 163.6, 167.2 ppm; IR (NaCl): n˜ =1159,
1515, 1711, 3400 cmÀ1 (broad); HRMS: m/z: calcd for C16H16O6 +H+:
305.1025 [M+H]+; found: 305.1090.
pletion, the mixture was diluted with EtOAc (15 mL), extracted
with water (10 mL), and successively washed with an aqueous solu-
tion of 1m HCl (10 mL) and brine (10 mL). Then, the organic phase
was dried (MgSO4) and concentrated. The solid was dried under re-
duced pressure (1.0 kPa) at 608C for 4 h to obtain 32 as a pale-
yellow solid (1.078 g, 95%): Rf =0.40 (SiO2, hexanes/EtOAc 75:25);
1
mp: 157–1608C; H NMR (250 MHz, CDCl3): d=0.22 (s, 6H), 0.23 (s,
6H), 0.99 (s, 9H), 1.00 (s, 9H), 6.24 (d, J=15.8 Hz, 1H), 6.81–6.87
(m, 1H), 7.01–7.08 (m, 2H), 7.67 ppm (d, J=16.0 Hz, 1H); 13C NMR
(62.9 MHz, CDCl3): d=À4.0 (2C),À3.9 (2C), 18.6, 18.6, 26.0 (6C),
115.0, 120.8, 121.3, 122.9, 127.8, 147.2, 147.4, 150.1, 173.2 ppm.
(S)-(6-Oxo-3,6-dihydro-2H-pyran-2-yl)methyl 3-(4-hydroxy-3-me-
thoxyphenyl)propanoate (13). Yield: 73 mg (77%). Colorless oil;
Rf =0.11 (SiO2, hexanes/EtOAc 60:40); [a]2D5 =À22 (c=1.0 in MeOH);
1H NMR (600 MHz, CDCl3): d=2.21–2.33 (m, 2H), 2.64 (t, J=7.6 Hz,
2H), 2.86 (t, J=7.6 Hz, 2H), 3.83 (s, 3H), 4.22–4.27 (m, 2H), 4.57
(dq, J=11.3, 4.7 Hz, 1H), 5.69 (brs, 1H), 5.99 (ddd, J=9.7, 2.6,
1.1 Hz, 1H), 6.65 (dd, J=8.0, 2.0 Hz, 1H), 6.68 (d, J=1.9 Hz, 1H),
6.79 (d, J=8.0 Hz, 1H), 6.84 ppm (ddd, J=9.7, 5.7, 2.7 Hz, 1H);
13C NMR (151 MHz, CDCl3): d=25.7, 30.6, 36.0, 55.9, 64.5, 75.2,
111.0, 114.4, 120.8, 121.1, 132.1, 144.2, 144.7, 146.6, 163.4,
172.6 ppm; IR (NaCl): n˜ =1033, 1082, 1153, 1237, 1515, 1720, 2934,
3424 cmÀ1 (broad); HRMS: m/z: calcd for C16H18O6 +H+: 307.1182
[M+H]+; found: 307.1219.
(E)-3-{4-[(tert-Butyldimethylsilyl)oxy]-3-methoxyphenyl}acrylic
acid. Freshly distilled diisopropylethylamine (5.38 mL, 30.9 mmol,
3 equiv) and tert-butyldimethylsilyl chloride (3.88 g, 25.8 mmol,
2.5 equiv) were added to a mixture of ferulic acid (2000 mg,
10.3 mmol, 1 equiv) in anhydrous CH2Cl2 (18 mL) at 258C, and the
mixture was stirred for 14 h. Then, the mixture was diluted with
EtOAc (25 mL), extracted with water (10 mL), and successively
washed with an aqueous solution of 1m HCl (2”15 mL) and brine
(15 mL); the organic phase was dried (MgSO4) and concentrated to
obtain a yellow oil. This oil was dissolved in THF (10 mL) and solid
K2CO3 (800 mg) and water (1 mL) were added. The mixture was
stirred for 2 h. Upon completion, the mixture was diluted with
EtOAc (20 mL), extracted with water (15 mL), and successively
washed with an aqueous solution of 1m HCl (15 mL) and brine
(15 mL). The organic phase was dried (MgSO4) and concentrated.
The solid was dried under reduced pressure (1.0 kPa) at 608C for
4 h to obtain the acid as a pale-yellow solid (2.90 g, 95%): Rf =0.46
(SiO2, hexanes/EtOAc 60:40); mp: 188–1908C; 1H NMR (250 MHz,
CDCl3): d=0.14 (s, 6H), 0.98 (s, 9H), 3.72 (s, 3H), 6.34 (d, J=
15.3 Hz, 1H), 6.65–7.05 (m, 3H), 7.65 ppm (d, J=15.3 Hz, 1H);
13C NMR (62.9 MHz, [D4]methanol): d=À4.3 (2C), 19.4, 26.3 (3C),
56.1, 112.2, 119.9, 122.1, 123.0, 130.5, 144.9, 148.3, 152.6,
173.1 ppm.
General procedure for hydrogenation to produce saturated
acids: Pd/C (5% w/w, 40 mg) was added to a solution of acid
(0.92 mmol) in EtOAc (5 mL). The heterogeneous mixture was
stirred under a hydrogen atmosphere for 3 h. The mixture was fil-
tered through Celite, and the solvent was removed under reduced
pressure.
3-{3,4-Bis[(tert-butyldimethylsilyl)oxy]phenyl}propanoic
acid
(31). Yield: 373 mg (99%). Pale-yellow solid; Rf =0.54 (SiO2, hex-
anes/EtOAc 75:25); mp: 88–898C; 1H NMR (250 MHz, CDCl3): d=
0.18 (s, 12H), 0.98 (s, 18H), 2.62 (t, J=7.5 Hz, 2H), 2.84 (t, J=
7.8 Hz, 2H), 6.60–6.78 ppm (m, 3H); 13C NMR (62.9 MHz, CDCl3): d=
À3.95 (4C), 18.6 (2C), 26.1 (6C), 30.1, 36.1, 121.2 (2C), 121.3, 133.4,
145.4, 146.8, 179.7 ppm.
(E)-3-(3,4-Dimethoxyphenyl)acrylic acid. Dimethyl sulfate (4.6 mL,
48.6 mmol) was added to
a mixture of ferulic acid (2.00 g,
10.3 mmol, 1 equiv) and potassium carbonate (13.0 g, 94 mmol,
9.1 equiv) in acetone (50 mL). The mixture was heated at reflux
overnight, filtered through a short column of silica gel (EtOAc as
eluent), and the solvent was evaporated under reduced pressure.
The residue was dissolved in methanol (35 mL) and an aqueous so-
lution of sodium hydroxide 10% w/v (35 mL). This mixture was
heated at reflux for 3 h. The mixture was neutralized with an aque-
ous solution of 6m HCl at 08C. The mixture was filtered, and the
solid was washed with cold water. The solid was dried under re-
duced pressure (1.0 kPa) at 708C to produce the acid as a white
solid (1.98 g, 92%): Rf =0.46 (SiO2, hexanes/EtOAc 30:70); mp: 178–
1808C; 1H NMR (250 MHz, CDCl3): d=3.93 (s, 6H), 6.33 (d, J=
16.0 Hz, 1H), 6.88 (d, J=8.2 Hz, 1H), 7.08 (d, J=1.4 Hz, 1H), 7.14
(dd, J=8.2, 1.3 Hz, 1H), 7.73 ppm (d, J=16.0 Hz, 1H); 13C NMR
(62.9 MHz, CDCl3): d=56.1 (2C), 110.0, 111.2, 115.1, 123.3, 127.2,
147.1, 149.4, 151.7, 172.7 ppm.
3-{4-[(tert-Butyldimethylsilyl)oxy]-3-methoxyphenyl}propanoic
acid. Yield: 282 mg (99%). Pale-yellow solid; Rf =0.50 (SiO2, hex-
anes/EtOAc 60:40); mp: 39–438C; 1H NMR (250 MHz, CDCl3): d=
0.13 (s, 6H), 0.98 (s, 9H), 1.89 (brs, 2H), 2.86 (brs, 2H), 3.74 (s, 3H),
6.58–6.77 ppm (m, 3H); 13C NMR (62.9 MHz, [D4]methanol): d=
À4.3 (2C), 19.4, 26.4 (3C), 32.2, 38.2, 56.0, 113.7, 121.6, 121.8, 136.4,
144.5, 152.2, 178.5 ppm.
3-(3,4-Dimethoxyphenyl)propanoic acid. Yield: 191 mg (99%).
White solid; Rf =0.41 (SiO2, hexanes/EtOAc 30:70); mp: 89–938C;
1H NMR (250 MHz, CDCl3): d=2.66 (t, J=7.9 Hz, 2H), 2.91 (t, J=
7.9 Hz, 2H), 3.85 (s, 3H), 3.86 (s, 3H), 6.71–6.82 ppm (m, 3H);
13C NMR (62.9 MHz, CDCl3): d=30.3, 36.0, 55.9, 56.0, 111.5, 111.8,
120.2, 132.9, 147.7, 149.0, 179.1 ppm.
(E)-3-{3,4-bis[(tert-Butyldimethylsilyl)oxy]phenyl}acrylic acid (32).
Freshly distilled diisopropylethylamine (2.91 mL, 16.7 mmol,
5.0 equiv) and tert-butyldimethylsilyl chloride (2.09 g, 13.9 mmol,
5.0 equiv) were added to a suspension of caffeic acid (500 mg,
2.78 mmol, 1 equiv) in anhydrous CH2Cl2 (3.4 mL) at 258C; the mix-
ture became a solution, which was stirred at 258C for 14 h. The
mixture was diluted with EtOAc (15 mL), extracted with water
(5 mL), and successively washed with an aqueous solution of 1m
HCl (2”10 mL) and brine (10 mL). Then, the organic phase was
dried (MgSO4) and concentrated to obtain a yellow oil. This oil was
dissolved in THF (4 mL) and solid K2CO3 (400 mg) and water
(0.7 mL) were added. The mixture was stirred for 2 h. Upon com-
(Allyloxy)(tert-butyl)diphenylsilane (34). Imidazole (7.46 g,
109 mmol, 1.2 equiv) and tert-butyldiphenylsilyl chloride (30 mL,
115 mmol, 1.25 equiv) were added to a solution of allyl alcohol 33
(5.30 g, 91.2 mmol, 1.0 equiv) dissolved in anhydrous CH2Cl2
(182 mL) 08C. The temperature was increased to RT, and the mix-
ture was stirred for 8 h; then, the solvent was removed under re-
duced pressure. The residue was poured into water (50 mL) and
extracted with Et2O (4”50 mL). The organic phases were com-
bined, dried (MgSO4), filtered, and concentrated under reduced
pressure. The product was purified by column chromatography
ChemMedChem 2015, 10, 1687 – 1699
www.chemmedchem.org
1695
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
(SiO2, hexanes/EtOAc 95:5) to afford 34 as a colorless oil (27.0 g,
quantitative yield): Rf =0.77 (SiO2, hexanes/EtOAc 95:5); 1H NMR
(250 MHz, CDCl3): d=1.08 (s, 9H), 4.20–4.24 (m, 2H), 5.12 (dd, J=
10.4, 1.7 Hz, 1H), 5.38 (dd, J=17.1, 1.8 Hz, 1H), 5.86–6.02 (m, 1H),
7.35–7.47 (m, 6H), 7.67–7.73 ppm (m, 4H); 13C NMR (62.9 MHz,
CDCl3): d=19.3, 26.8 (3C), 64.6, 113.9, 127.6 (4C), 129.6 (2C), 133.7
(2C), 135.5 (4C), 137.0 ppm.
phase was extracted with Et2O (3”10 mL). The organic phases
were grouped, dried (MgSO4), filtered, and concentrated. The prod-
uct was purified by column chromatography (SiO2, hexanes/EtOAc
90:10) to afford 37 as a colorless oil (407 mg, 63%): Rf =0.71 (SiO2,
hexanes/EtOAc 80:20); [a]2D5 =À5.7 (c=1.0 in CHCl3); Lit.[74] ent-
compound [a]D = +8.3 (c=0.986 in CHCl3); 1H NMR (250 MHz,
CDCl3): d=1.04 (s, 9H), 2.35–2.60 (m, 2H), 3.74 (d, J=4.7 Hz, 2H),
5.00–5.19 (m, 3H), 5.65–5.89 (m, 2H), 6.04–6.19 (m, 1H), 6.35–6.46
(m, 1H), 7.33–7.48 (m, 6H), 7.62–7.72 ppm (m, 4H); 13C NMR
(62.9 MHz; CDCl3): d=19.4, 26.9 (3C), 35.2, 64.5, 73.8, 118.1, 127.8
(5C), 128.9, 129.8 (2C), 130.7, 133.5, 133.5, 135.7 (2C), 135.7 (2C),
165.7 ppm.
2-[(tert-Butyldiphenylsilyl)oxy]acetaldehyde (35). 2,6-Lutidine
(10.7 mL, 9.84 g, 2.2 equiv), a solution of 2.5% w/w osmium tetrox-
ide in tBuOH (235 mg of OsO4, 0.02 equiv), and sodium periodate
(39.5 g, 184 mmol, 4.4 equiv) were added to a solution of alkene
34 (12.5 g, 42.2 mmol, 1 equiv) in a mixture of water/1,4-dioxane
(1:3 v/v, 500 mL). The mixture was stirred at RT for 24 h. Upon com-
pletion, the mixture was diluted with water (150 mL) and EtOAc
(300 mL) and then filtered. The phases were separated, and the
aqueous phase was extracted with EtOAc (2”300 mL). The organic
phases were combined, washed with an aqueous solution of 1m
HCl (2”100 mL) and a saturated aqueous solution of NaHCO3
(100 mL), and dried (MgSO4). The solvent was removed under re-
duced pressure, and the product was purified by column chroma-
tography (SiO2, hexanes/EtOAc 95:5) to afford 35 as a brown oil
(9.09 g, 72%): Rf =0.34 (SiO2, hexanes/EtOAc 90:10); 1H NMR
(250 MHz, CDCl3): d=1.14 (s, 9H), 4.24 (s, 2H), 7.36–7.51 (m, 6H),
7.66–7.77 (m, 4H), 9.74 ppm (s, 1H); 13C NMR (62.9 MHz, CDCl3):
d=19.4, 26.8 (3C), 70.1, 128.1 (4C), 130.2 (2C), 132.6 (2C), 135.6
(4C), 201.8 ppm.
(S)-6-{[(tert-Butyldiphenylsilyl)oxy]methyl}-5,6-dihydro-2H-pyran-
2-one (38). Grubbs first-generation catalyst (112 mg, 0.14 mmol,
0.1 equiv) was added to a solution of acrylate 37 (535 mg,
1.4 mmol, 1.0 equiv) in CH2Cl2 (140 mL) at 40–458C. The mixture
was heated at reflux for 6 h. After this time, DMSO (0.5 mL,
7.1 mmol, 5 equiv) was added at RT. The mixture was stirred for
12 h. Then, water (15 mL) was added. The phases were separated,
and the organic material was extracted with water (5”15 mL). The
organic phase was dried (MgSO4) and concentrated. The product
was purified by column chromatography (SiO2, hexanes/EtOAc
70:30) to afford 38 as a brown oil (339 mg, 68%): Rf =0.37 (SiO2,
hexanes/EtOAc 70:30); [a]2D5 =À38 (c=1.5 in CHCl3). Lit.[75] ent-com-
pound [a]D = +34.2 (c=1.5 in CHCl3); 1H NMR (250 MHz, CDCl3):
d=1.07 (s, 9H), 2.35–2.68 (m, 2H), 3.84 (d, J=5.0 Hz, 2H), 4.52 (dq,
J=10.6, 4.9 Hz, 1H), 5.97–6.05 (m, 1H), 6.88 (ddd, J=9.8, 5.5,
2.8 Hz, 1H), 7.35–7.55 (m, 6H), 7.62–7.70 ppm (m, 4H); 13C NMR
(62.9 MHz, CDCl3): d=19.3, 26.0, 26.9 (3C), 64.9, 77.7, 121.3, 127.9
(4C), 130.0 (2C),132.9, 133.1, 135.7 (2C), 135.7 (2C), 145.0,
163.9 ppm.
(S)-1-[(tert-Butyldiphenylsilyl)oxy]pent-4-en-2-ol (36).
A 50 mL
round-bottomed flask containing a magnetic stir bar was charged
with powdered activated 4 Š molecular sieves (2.5 g). After the ad-
dition of the molecular sieves, the flask was flame dried under flow
of N2. After this, anhydrous CH2Cl2 (6.4 mL), (S)-(À)-1,1’-bi-2-naph-
thol [(S)-BINOL; 189 mg, 0.66 mmol, 0.2 equiv], 1m trifluoroacetic
acid (TFA) in anhydrous CH2Cl2 (10 mL), and Ti(OiPr)4 (99 mL) were
added. The mixture was heated at reflux for 1 h, at which point
the color of the solution was observed to change dark red to
brown. After this period, the oil bath was removed, the tempera-
ture was allowed to reach RT, and aldehyde 35 (994 mg,
3.33 mmol, 1.0 equiv) was added. The mixture was stirred for
15 min. The flask was placed in a bath at À788C and allyltributyl-
stannane (1.6 mL, 4.98 mmol, 1.50 equiv) was added slowly to the
mixture. The mixture was stirred at À208C for 4 days. After this
period, brine (20 mL) was added to the mixture, and the tempera-
ture was allowed to reach RT. After 1 h, the mixture was filtered.
The aqueous phase was extracted with CH2Cl2 (4”20 mL). The or-
ganic phases were grouped, dried (MgSO4), filtered, and concen-
trated under reduced pressure. The product was purified by
column chromatography (SiO2, hexanes/EtOAc 90:10) to afford 36
as a colorless oil (988 mg, 87%): Rf =0.61 (SiO2, hexanes/EtOAc
80:20); [a]2D5 =À3 (c=1.0 in CHCl3). Lit.[73] ent-compound [a]D = +3
(c=0.986 in CHCl3); 1H NMR (250 MHz, CDCl3): d=1.07 (s, 9H),
2.15–2.29 (m, 3H), 3.55 (dd, J=10.1, 7.0 Hz, 1H), 3.67 (dd, J=10.1,
3.8 Hz, 1H), 3.73–3.85 (m, 1H), 5.02–5.14 (m, 2H), 5.80 (ddt, J=
17.1, 10.1, 7.0 Hz, 1H), 7.34–7.48 (m, 6H), 7.61–7.76 ppm (m, 4H);
13C NMR (62.9 MHz, CDCl3): d=19.4, 27.0 (3C), 37.7, 67.4, 71.4,
117.6, 127.9 (4C), 129.9 (2C), 133.3 (2C), 134.4, 135.6 ppm (4C).
(S)-6-(Hydroxymethyl)-5,6-dihydro-2H-pyran-2-one (39). A solu-
tion of 1m TBAF in THF (1.14 mL, 1.14 mmol, 1.05 equiv) was
added dropwise to a solution of lactone 38 (400 mg, 1.09 mmol,
1.0 equiv) and benzoic acid (139 mg, 1.14 mmol, 1.05 equiv) in an-
hydrous THF (55 mL) at 08C. The mixture was stirred for 4 h at RT,
and then a saturated aqueous solution of NaHCO3 (250 mL) was
added. The mixture was extracted with EtOAc (7”200 mL), and the
organic phases were grouped, dried (MgSO4), filtered, and concen-
trated. The product was purified by column chromatography (SiO2,
EtOAc) to afford 39 as a colorless oil (126 mg, 90%): Rf =0.34 (SiO2,
EtOAc); [a]2D5 =À159 (c=1.0 in CHCl3). Lit.[14] ent-compound [a]D =
+160 (c=0.85 in CHCl3); 1H NMR (400 MHz, CDCl3): d=2.20 (brs,
1H), 2.26–2.36 (m, 1H), 2.55–2.66 (m, 1H), 3.74 (dd, J=12.3, 4.8 Hz,
1H), 3.88 (dd, J=12.3, 3.3 Hz, 1H), 4.55 (ddd, J=12.3 Hz, 8.3,
4.0 Hz, 1H), 6.03 (ddd, J=9.8, 2.8, 0.8 Hz, 1H), 6.93 ppm (ddd, J=
9.5, 6.3, 2.3 Hz, 1H); 13C NMR (62.9 MHz, CDCl3): d=25.4, 64.0, 78.5,
121.1, 145.4, 163.9 ppm.
Biology
In vitro antiproliferative assay: Human tumor cell lines U-251
(glioma), MCF-7 (breast), NCI-H460 (lung, non-small cells), HT-29
(colon), PC-3 (prostate), 786-0 (kidney), and NCI-ADR/RES (ovarian
expressing multiple drugs resistance phenotype) were obtained
from the National Cancer Institute at Frederick, MA, USA. Human
pancreatic cancer cells (Panc-1) were purchased from the Rio de Ja-
neiro Cell Bank (Rio de Janeiro, RJ, Brazil). The non-tumor cell line
HaCat (human keratinocytes) was donated by Prof. Dr. Ricardo
Della Coletta, FOP/UNICAMP.
(S)-1-[(tert-Butyldiphenylsilyl)oxy]pent-4-en-2-yl acrylate (37).
Freshly distilled triethylamine (0.46 mL, 3.28 mmol, 2.0 equiv) was
added to a solution of alcohol 36 (560 mg, 1.64 mmol, 1.0 equiv) in
anhydrous CH2Cl2 (8.2 mL) at 08C, and this was followed by the
slow addition of acryloyl chloride (0.20 mL, 2.46 mmol, 1.5 equiv).
The mixture was stirred for 1.5 h at RT. After this period, the sol-
vent was removed and brine (10 mL) was added. The aqueous
With the exception of Panc-1 cells, stock cultures were grown in
RPMI 1640 (GIBCO BRL) medium supplemented with 5% fetal
ChemMedChem 2015, 10, 1687 – 1699
www.chemmedchem.org
1696
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
bovine serum (FBS, GIBCO), 100 UmLÀ1 penicillin, and 100 mgmLÀ1
streptomycin at 378C with 5% CO2. Panc-1 cells were cultured in
DMEM (Nutricell, Brazil) containing 100 UmLÀ1 penicillin,
100 mgmLÀ1 streptomycin, and 10% FBS (GIBCO, Brazil) under the
same conditions of temperature and atmosphere.
sence of the enzyme by using the same experimental conditions
to account for possible non-enzymatic product formation. Stock
solutions of the tested compounds were prepared in DMSO, which
resulted in final DMSO concentration of 5% in the assay. A control
group containing DMSO but not the inhibitors was conducted in
parallel to represent 100% enzyme activity and was used to nor-
malize the data. Initially, a single concentration (100 mm) of the in-
hibitors was used to verify potential GST inhibitors in the series.
Subsequently, concentration–response curves were conducted
with the best inhibitors identified by using the concentration
range of 1–100 mm. Enzyme initial velocities were calculated in the
control and test groups. Data was normalized against the DMSO
controls, plotted as percent of remaining enzyme activity versus
the log of inhibitor concentration. Experimental curves were fitted
by using the logistic four-parameter equation in the GraphPad
software version 5 (GraphPad Prism, San Diego, CA, USA). Experi-
ments were conducted in triplicate.
Cells in 96-well plates (100 mL cellswellÀ1) were exposed to tarcho-
nanthuslactone and its analogues at concentrations 0.25, 2.5, 25,
and 250 mgmLÀ1 in DMSO/RPMI or DMEM at 378C, 5% CO2, in air
for 48 h. Doxorubicin was used as a positive control (0.025, 0.25,
2.5, and 25 mgmLÀ1). The final DMSO concentration did not affect
cell viability (0.1%). Afterward, cells were fixed with 50% trichloro-
acetic acid, and cell proliferation was determined by spectrophoto-
metric quantification (540 nm) of cellular protein content by using
the sulforhodamine B assay. The TGI (concentration that produces
total growth inhibition or cytostatic effect) was determined
through nonlinear regression analysis by using the concentration–
response curve for each cell line (Table 1) in the ORIGIN 8.0 soft-
ware (OriginLab Corporation).
Protein phosphatase assays: The DNA sequence coding the human
protein phosphatases PTP-1B, LMW-PTP, and CDC-25B were cloned
into pET28a+ expression vectors and were expressed in E coli BL-
21DE3 with a 6-His N-terminal tag. The full-length sequence of
LMW-PTP (UniProt code 24666) and the catalytic domains of CDC-
25B (UniProt code P30305, isoform 3, residues 391-580) and PTP-1B
(UniProt code P18031, residues 1-298) were selected. Expressed
proteins were purified by Ni-affinity chromatography, followed by
size-exclusion purification by using Superdex 75 resin (GE Life Sci-
ences). Final buffers contained 100 mm Tris (pH 7.5), 150 mm NaCl,
1 mm dithiothreitol, and 10% glycerol. Samples were concentrated
to 5–10 mgmLÀ1 and were stored in liquid nitrogen. Protein con-
centration was evaluated by using the method of Bradford.[76] Pro-
tein purity was checked by PAGE-SDS electrophoresis,[77] and enzy-
matic activity was confirmed with the para-nitrophenol phosphate
(pNPP) activity assay.
Measurement of intracellular ROS production: Intracellular ROS
levels were measured by flow cytometry in cells loaded with the
redox-sensitive dye DCFH-DA (Sigma–Aldrich, St. Louis, MO, USA).
Briefly, Panc-1 cells were washed with HBSS (Hank’s buffered salt
solution) medium and were incubated in the dark for 30 min at
378C with 10 mm DCFH-DA. Then, Panc-1 cells were treated with
25, 50, and 100 mm of compounds 8 and 9. After 1 h of treatment,
cells were harvested and resuspended in HBSS medium. Fluores-
cence was recorded with the FL-1 channel of a Guava Easycyte
Mini flow cytometer (Guava Technologies, Hayward, CA). The data
were analyzed with the software CytoSoft 4.1, Guava Express Pro
program.
Measurement of apoptosis: Phosphatidylserine externalization was
analyzed with the Guava Nexin Assay Kit (Guava Technologies, Hay-
ward, CA) in accordance with the manufacturer’s instructions.
Panc-1 cells were treated with 25, 50, and 100 mm of compounds 8
and 9 for 24 h. Then, cells were harvested and resuspended at
a density of 1”105 cells in 100 mL of phosphate buffer saline (PBS).
Binding buffer (100 mL) containing annexin-V and 7-AAD was
added on the cells, which were then incubated in the dark for
20 min at room temperature. Thereafter, the cells were analyzed by
flow cytometry (Guava Easycyte Mini, Guava Technologies, Hay-
ward, CA).
IC50 measurements were conducted by using the PTPase classic
substrate pNPP, as previously reported.[78–80] Assays were performed
in 96-well plates by using an automated 8-channel pipet (Explorer
Eppendorf). Inhibitor stocks were prepared in DMSO, and enzymes
were incubated with the tested compounds in the concentration
range of 0.1–100 mm for 15 min at 308C in reaction buffer. The final
DMSO concentration in each well was 5% in a total volume of
90 mL. Reaction buffers were adjusted to the ideal pH range for
each PTPase and contained 100 mm sodium acetate (pH 5.5) and
0.005% Triton X-100 for LMW-PTP and PTP-1B and 100 mm bis-tris-
propane (pH 8.2) and 0.005% Triton X-100 for CDC-25B. After the
inhibitor incubation time, a 10 times stock solution containing the
substrate was added. The final substrate concentration was in the
range of the Km value presented to each enzyme. The enzymatic
reaction occurred over 10 min at 378C and was stopped by the ad-
dition of 1n NaOH (100 mL). Absorbance at l=405 nm, relative to
pNP reaction product production, was measured in each well by
using a M2e plate reader (Molecular Devices). Absorbance was nor-
malized to the control group, which was measured under the
same experimental conditions, however without the inhibitor (re-
lated to 100% enzyme activity), and was plotted in a graph of in-
hibitor concentration versus remaining enzyme activity. The experi-
mental data were fit by using the logistic four-parameter equation
in the GraphPad software version 5 (GraphPad Prism). At least two
independent experiments conducted in triplicate were used for
average and standard deviation calculation of the reported IC50
values. Blank samples (in the absence of the enzyme) were also
measured in each experiment to account for possible interference
of the inhibitor in the absorbance measurement.
Glutathione-S-transferase assay: Recombinant S. japonicum GST was
produced in E. coli BL21DE3 by using the commercially available
PGex vector (GE Life Sciences). Protein was purified by GSH-agar-
ose (GE Life Sciences) affinity chromatography and eluted in
a buffer containing 20 mm Tris (pH 7.5), 150 mm NaCl, and 5 mm
GSH. Protein purity and ample quality were monitored by SDS-
PAGE and dynamic light scattering. Protein concentration was mea-
sured by absorbance in a NanoDrop instrument by using the GST
extinction coefficient of 42860mÀ1 cmÀ1 at l=280 nm. The purified
protein was dialyzed in 50 mm Tris (pH 7.0) and 50 mm NaCl and
concentrated to 1 mgmLÀ1
.
Enzyme activity was measured in the presence and absence of the
test compounds by using GSH (0.7 mm) and 1-chloro-2,4-dinitro-
benzene (CDNB, 2.5 mm) as substrates. The reaction buffer con-
tained 100 mm Tris (pH 7.0), 0.01% Triton X-100, and 30 nm re-
combinant GST. The enzyme was incubated with the tested com-
pounds for 30 min at 208C prior to the addition of the substrates.
Formation of the GSH–CDNB conjugate by the enzyme was moni-
tored for 5 min at l=340 nm, 258C, by using an Envision (Perki-
nElmer) plate reader. Blank experiments were conducted in the ab-
ChemMedChem 2015, 10, 1687 – 1699
www.chemmedchem.org
1697
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
[18] T. Nakata, H. Noriaki, I. Katsumi, O. Takeshi, Tetrahedron Lett. 1987, 28,
5661–5664.
[19] Y. Mori, M. Suzuki, J. Chem. Soc. Perkin Trans. 1 1990, 1809–1812.
[20] G. SolladiØ, L. Gressot-Kempf, Tetrahedron: Asymmetry 1996, 7, 2371–
2379.
[21] M. V. R. Reddy, A. J. Yucel, P. V. Ramachandran, J. Org. Chem. 2001, 66,
2512–2514.
[22] S. D. Garaas, T. J. Hunter, G. A. O’Doherty, J. Org. Chem. 2002, 67, 2682–
2685.
[23] D. Enders, D. Steinbusch, Eur. J. Org. Chem. 2003, 4450–4454.
[24] S. Baktharaman, S. Selvakumar, V. K. Singh, Tetrahedron Lett. 2005, 46,
7527–7529.
[25] P. Gupta, S. V. Naidu, P. Kumar, Tetrahedron Lett. 2005, 46, 6571–6573.
[26] G. Sabitha, K. Sudhakar, N. M. Reddy, M. Rajkumar, J. S. Yadav, Tetrahe-
dron Lett. 2005, 46, 6567–6570.
[27] M. S. Scott, C. A. Luckhurst, D. J. Dixon, Org. Lett. 2005, 7, 5813–5816.
[28] S. George, A. Sudalai, Tetrahedron: Asymmetry 2007, 18, 975–981.
[29] J. Yadav, P. Reddy, M. Gupta, C. Chary, Synthesis 2007, 3639–3646.
[30] J. S. Yadav, N. N. Kumar, M. S. Reddy, A. R. Prasad, Tetrahedron 2007, 63,
2689–2694.
[31] H.-Y. Lee, V. Sampath, Y. Yoon, Synlett 2009, 0249–0252.
[32] G. F. P. de Souza, L. F. T. Novaes, C. M. Avila, L. F. R. Nascimento, L. A. Vel-
loso, R. A. Pilli, Molecules 2015, 20, 5038–5049.
[33] D. J. Adams, M. Dai, G. Pellegrino, B. K. Wagner, A. M. Stern, A. F. Shamji,
Proc. Natl. Acad. Sci. USA 2012, 109, 15115–15120.
[34] M. F. Tolba, A. Esmat, A. M. Al-Abd, S. S. Azab, A. E. Khalifa, H. A. Mosli,
IUBMB Life 2013, 65, 716–729.
[35] J.-Y. Lee, H.-J. Choi, T.-W. Chung, C.-H. Kim, H.-S. Jeong, K.-T. Ha, J. Nat.
Prod. 2013, 76, 1399–1405.
[36] R. Granata, L. Locati, L. Licitra, Curr. Opin. Oncol. 2013, 25, 224–228.
[37] D. Seebach, M. Züger, Helv. Chim. Acta 1982, 65, 495–503.
[38] G. E. Keck, K. H. Tarbet, L. S. Geraci, J. Am. Chem. Soc. 1993, 115, 8467–
8468.
[39] P. Kumar, P. Gupta, S. V. Naidu, Chem. Eur. J. 2006, 12, 1397–1402.
[40] G. C. Vougioukalakis, R. H. Grubbs, Chem. Rev. 2010, 110, 1746–1787.
[41] S. A. Snyder, F. Kontes, J. Am. Chem. Soc. 2009, 131, 1745–1752.
[42] F. Bisogno, L. Mascoti, C. Sanchez, F. Garibotto, F. Giannini, M. Kurina-
Sanz, J. Agric. Food Chem. 2007, 55, 10635–10640.
Reactive oxygen species detection in the enzymatic assays (FOX
method): Detection of ROS in the protein phosphatase assays was
performed by using the method of FOX (Fe2+/xylenol orange), ac-
cording to the protocol used by Ogusucu et al.,[69] adapted to mi-
croplates. The FOX reactant (100 mL) two times concentrated was
added to the same volume of the phosphatase assays, which con-
tained the phosphatase reaction buffer (0.5 mm Fe2+SO4, 0.05 mm
H2SO4, 0.2 mm xylenol orange, and 20 mm glucose) and the test
compounds at 100 mm (in the absence of the phosphatase en-
zymes). The reaction was left for 15 min, and the absorbance at
l=560 nm was measured by using a M2e microplate reader (Mo-
lecular Devices). Absorbance values were normalized to the values
found in the DMSO controls and are expressed as the relative in-
crease in ROS production. Experiments were conducted in dupli-
cate.
Acknowledgements
The authors thank the Fundażo de Amparo à Pesquisa do
Estado de S¼o Paulo (FAPESP) (research grants 2009/51602-5,
2010/17544-5, 2010/08673-6, 2012/09254-2, 2012/24385-6, and
2013/08896-3) and the Institute of Chemistry/UNICAMP for finan-
cial and academic support, and the Brazilian National Biosci-
ences Laboratory for use of the Laboratory of Spectroscopy and
Calorimetry (LEC). The authors also acknowledge Prof. Afon-
so C. D. Bainy and Jaco Mattos from the Federal University of
Santa Catarina (Brazil) for their support in conducting initial
analysis of the antioxidant enzymes.
Keywords: antitumor agents · cancer · natural products ·
reactive oxygen species · tarchonanthuslactone
[43] B. Narasimhan, D. Belsare, D. Pharande, V. Mourya, A. Dhake, Eur. J. Med.
Chem. 2004, 39, 827–834.
[44] A. Monks, D. Scudiero, P. Skehan, R. Shoemaker, K. Paull, D. Vistica, J.
Natl. Cancer Inst. 1991, 83, 757–766.
[1] J. Szychowski, J.-F. Truchon, Y. L. Bennani, J. Med. Chem. 2014, 57, 9292–
9308.
[45] L. V. Rubinstein, R. H. Shoemaker, K. D. Paull, R. M. Simon, S. Tosini, P.
Skehan, J. Natl. Cancer Inst. 1990, 82, 1113–1118.
[2] J. Hong, Chem. Eur. J. 2014, 20, 10204–10212.
[3] M. S. Butler, A. A. B. Robertson, M. A. Cooper, Nat. Prod. Rep. 2014, 31,
1612–1661.
[4] J. A. Marco, M. Carda, J. Murga, E. Falomir, Tetrahedron 2007, 63, 2929–
2958.
[5] V. Boucard, G. Broustal, J. M. Campagne, Eur. J. Org. Chem. 2007, 225–
236.
[6] A. de Fatima, L. Modolo, L. Conegero, R. Pilli, C. Ferreira, L. Kohn, Curr.
Med. Chem. 2006, 13, 3371–3384.
[7] ff. de Fµtima, L. K. Kohn, J. E. de Carvalho, R. A. Pilli, Bioorg. Med. Chem.
2006, 14, 622–631.
[8] M. CastaÇo, W. Cardona, W. QuiÇones, S. Robledo, F. Echeverri, Molecules
2009, 14, 2491–2500.
[46] J. Liu, A. Mori, Arch. Biochem. Biophys. 1993, 302, 118–127.
[47] R. C. Barcelos, J. C. Pastre, V. Caixeta, D. B. Vendramini-Costa, J. E. de Car-
valho, R. A. Pilli, Bioorg. Med. Chem. 2012, 20, 3635–3651.
[48] T. G. Hastings, M. J. Zigmond, J. Neurochem. 1994, 63, 1126–1132.
[49] M. Brisson, C. Foster, P. Wipf, B. Joo, R. J. Tomko, Jr., T. Nguyen, J. S.
Lazo, Mol. Pharmacol. 2007, 71, 184–192.
[50] F. Puoci, C. Morelli, G. Cirillo, M. Curcio, O. I. Parisi, P. Maris, D. Sisci, N.
Picci, Anticancer Res. 2012, 32, 2843–2847.
[51] K. J. Pelizzaro-Rocha, M. B. de Jesus, R. R. Ruela-de-Sousa, C. V. Naka-
mura, F. S. Reis, A. de Fµtima, Biochim. Biophys. Acta Mol. Cell Res. 2013,
1833, 2856–2865.
[52] A. N. Shah, J. M. Summy, J. Zhang, S. I. Park, N. U. Parikh, G. E. Gallick,
Ann. Surg. Oncol. 2007, 14, 3629–3637.
[9] M. J. Balunas, M. F. Grosso, F. A. Villa, N. Engene, K. L. McPhail, K. Tidge-
well, Org. Lett. 2012, 14, 3878–3881.
[53] W. Huanwen, L. Zhiyong, S. Xiaohua, R. Xinyu, W. Kai, L. Tonghua, Mol.
Cancer 2009, 8, 125.
[10] S.-M. Lee, W.-G. Lee, Y.-C. Kim, Y.-C. Kim, H. Ko, Bioorg. Med. Chem. Lett.
2011, 21, 5726–5729.
[54] T. H. Ward, J. Cummings, E. Dean, A. Greystoke, J. M. Hou, A. Backen, Br.
J. Cancer 2008, 99, 841–846.
[11] T. A. Johnson, J. Sohn, A. E. Ward, T. L. Cohen, N. D. Lorig-Roach, H.
Chen, Bioorg. Med. Chem. 2013, 21, 4358–4364.
[12] T. Usui, H. Watanabe, H. Nakayama, Y. Tada, N. Kanoh, M. Kondoh, T.
Asao, K. Takio, H. Watanabe, K. Nishikawa, T. Kitahara, H. Osada, Chem.
Biol. 2004, 11, 799–806.
[13] I. M. Serafimova, M. A. Pufall, S. Krishnan, K. Duda, M. S. Cohen, R. L. Ma-
glathlin, Nat. Chem. Biol. 2012, 8, 471–476.
[14] D. L. Boger, M. Hikota, B. M. Lewis, J Org Chem. 1997, 62, 1748–1753.
[15] L. Bialy, H. Waldmann, Chem. Eur. J. 2004, 10, 2759–2780.
[16] F. Bohlmann, A. Suwita, Phytochemistry 1979, 18, 677–678.
[17] M. van Huyssteen, P. J. Milne, E. E. Campbell, M. van de Venter, Afr. J.
Tradit., Complementary Altern. Med. 2011, 8, 150–158.
[55] L. Galluzzi, I. Vitale, J. M. Abrams, E. S. Alnemri, E. H. Baehrecke, M. V. Bla-
gosklonny, Cell Death Differ. 2012, 19, 107–120.
[56] D. Trachootham, J. Alexandre, P. Huang, Nat. Rev. Drug Discovery 2009,
8, 579–591.
[57] L. Raj, T. Ide, A. U. Gurkar, M. Foley, M. Schenone, X. Li, Nature 2011,
475, 231–234.
[58] G. T. Wondrak, Antioxid. Redox Signaling 2009, 11, 3013–3069.
[59] J. Luo, N. L. Solimini, S. J. Elledge, Cell 2009, 136, 823–837.
[60] K. D. Tew, D. M. Townsend, Curr. Opin. Chem. Biol. 2011, 15, 156–161.
[61] L. A. Sena, N. S. Chandel, Mol. Cell 2012, 48, 158–167.
ChemMedChem 2015, 10, 1687 – 1699
www.chemmedchem.org
1698
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
[62] M. Brisson, T. Nguyen, P. Wipf, B. Joo, B. W. Day, J. S. Skoko, E. M.
Schreiber, C. Foster, P. Bansal, J. S. Lazo, Mol. Pharmacol. 2005, 68,
1810–1820.
[63] L. Bialy, H. Waldmann, Angew. Chem. Int. Ed. 2005, 44, 3814–3839;
Angew. Chem. 2005, 117, 3880–3906.
[64] P. Chiarugi, P. Cirri, Trends Biochem. Sci. 2003, 28, 509–514.
[65] P. Chiarugi, G. Pani, E. Giannoni, L. Taddei, R. Colavitti, G. Raugei, J. Cell
Biol. 2003, 161, 933–944.
[66] A. Salmeen, J. N. Andersen, M. P. Myers, T.-C. Meng, J. A. Hinks, N. K.
Tonks, Nature 2003, 423, 769–773.
[72] K. D. Tew, D. M. Townsend, Drug Metab. Rev. 2011, 43, 179–193.
[73] Y. Díaz, F. Bravo, S. Castillón, J. Org. Chem. 1999, 64, 6508–6511.
[74] A. Kamal, M. Balakrishna, P. V. Reddy, S. Faazil, Tetrahedron: Asymmetry
2010, 21, 2517–2523.
[75] T. V. Hansen, Tetrahedron: Asymmetry 2002, 13, 547–550.
[76] M. M. Bradford, Anal. Biochem. 1976, 72, 248–254.
[77] U. K. Laemmli, Nature 1970, 227, 680–685.
[78] Z.-Y. Zhang, J. C. Clemens, H. L. Schubert, J. A. Stuckey, M. W. Fischer,
D. M. Hume, M. A. Saper, J. E. Dixon, J. Biol. Chem. 1992, 267, 23759–
23766.
[67] P. A. Ferreira, R. R. Ruela-de-Sousa, K. C. S. Queiroz, A. C. S. Souza, R.
Milani, R. A. Pilli, PLoS ONE 2012, 7, e44312.
[79] E. B. Gottlin, X. Xu, D. M. Epstein, S. P. Burke, J. W. Eckstein, D. P. Ballou,
J. Biol. Chem. 1996, 271, 27445–27449.
[68] R. Ogusucu, D. Rettori, D. C. Munhoz, L. E. S. Netto, O. Augusto, Free
Radical Biol. Med. 2007, 42, 326–334.
[80] F. Liu, D. E. Hill, J. Chernoff, J. Biol. Chem. 1996, 271, 31290–31295.
[69] R. Ogusucu, D. Rettori, L. E. S. Netto, O. Augusto, J. Biol. Chem. 2009,
284, 5546–5556.
[70] A. J. Barr, E. Ugochukwu, W. H. Lee, O. N. F. King, P. Filippakopoulos, I.
Alfano, Cell 2009, 136, 352–363.
Received: June 3, 2015
Revised: August 3, 2015
[71] K. D. Tew, Y. Manevich, C. Grek, Y. Xiong, J. Uys, D. M. Townsend, Free
Radical Biol. Med. 2011, 51, 299–313.
Published online on August 25, 2015
ChemMedChem 2015, 10, 1687 – 1699
www.chemmedchem.org
1699
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim