Marliolide inhibits skin carcinogenesis by activating NRF2/ARE to induce heme oxygenase-1

Article abstract of DOI:10.1016/j.ejmech.2018.02.068

Heme oxygenase-1 (HO-1) catalyzes the enzymatic degradation of heme to produce three anti-oxidant molecules: carbon monoxide (CO), ferrous ion (Fe²⁺), and biliverdin. Induction of HO-1 is currently considered as a feasible strategy to treat oxidative stress-related diseases. In the present study, we identified marliolide as a novel inducer of HO-1 in human normal keratinocyte HaCaT cells. Mechanism-based studies demonstrated that the induction of HO-1 by marliolide occurred through activation of NRF2/ARE via direct binding of marliolide to KEAP1. Structure-activity relationship revealed chemical moieties of marliolide critical for induction of HO-1, which renders a support for Michael reaction as a potential mechanism of action. Finally, we observed that marliolide significantly inhibited the papilloma formation in DMBA/TPA–induced mouse skin carcinogenesis model and this event was closely associated with lowering the formation of 8-OH-G and 4-HNE in vivo. Together, our study provides the first evidence that marliolide might be effective against oxidative stress-related skin disorders.

Full text of DOI:10.1016/j.ejmech.2018.02.068

Accepted Manuscript
Marliolide inhibits skin carcinogenesis by activating NRF2/ARE to induce heme
Oxygenase-1
June Lee, Karabasappa Mailar, Ok-Kyung Yoo, Won Jun Choi, Young-Sam Keum
PII:
S0223-5234(18)30205-8
10.1016/j.ejmech.2018.02.068
EJMECH 10245
DOI:
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 25 August 2017
Revised Date: 14 February 2018
Accepted Date: 21 February 2018
Please cite this article as: J. Lee, K. Mailar, O.-K. Yoo, W.J. Choi, Y.-S. Keum, Marliolide inhibits skin
carcinogenesis by activating NRF2/ARE to induce heme Oxygenase-1, European Journal of Medicinal
Chemistry (2018), doi: 10.1016/j.ejmech.2018.02.068.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT
100
Acetone
DMBA/TPA
80
DMBA/TPA + Marliolide
ꢃ
ꢀ
2
+
2
2
ARE
60
40
20
0
2
2
2
NRF2/HO-1
6
.($3ꢁ
6
.($3ꢁ
+
2
+2
+
2
***
ROS
***
ꢁ
ꢂ
**
8
ꢁ
ꢂ
ꢁꢂ
+6 .($3ꢁ
*
Week
1
2
3
4
5
6
7
9
10

ACCEPTED MANUSCRIPT
Marliolide Inhibits Skin Carcinogenesis by Activating NRF2/ARE
to Induce Heme Oxygenase-1
June Lee, Karabasappa Mailar, Ok-Kyung Yoo, Won Jun Choi* and Young-Sam Keum*
College of Pharmacy and Integrated Research Institute for Drug Development, Dongguk
University, 32 Dongguk-ro, Goyang, Gyeonggi-do 10326, Korea
Corresponding authors
Won Jun Choi, Ph.D.
Phone: +82-31-961-5219. Fax: +82-31-961-5206. Email: mp89@dongguk.edu
Young-Sam Keum, Ph.D.
Phone: +82-31-961-5215. Fax: +82-31-961-5206. Email: keum03@dongguk.edu

ACCEPTED MANUSCRIPT
ABSTRACT
Heme oxygenase-1 (HO-1) catalyzes the enzymatic degradation of heme to produce
2
+
three anti-oxidant molecules: carbon monoxide (CO), ferrous ion (Fe ), and biliverdin.
Induction of HO-1 is currently considered as a feasible strategy to treat oxidative stress-
related diseases. In the present study, we identified marliolide as a novel inducer of HO-1 in
human normal keratinocyte HaCaT cells. Mechanism-based studies demonstrated that the
induction of HO-1 by marliolide occurred through activation of NRF2/ARE via direct
binding of marliolide to KEAP1. Structure-activity relationship revealed chemical moieties of
marliolide critical for induction of HO-1, which renders a support for Michael reaction as a
potential mechanism of action. Finally, we observed that marliolide significantly inhibited the
papilloma formation in DMBA/TPA–induced mouse skin carcinogenesis model and this
event was closely associated with lowering the formation of 8-OH-G and 4-HNE in vivo.
Together, our study provides the first evidence that marliolide might be effective against
oxidative stress-related skin disorders.

ACCEPTED MANUSCRIPT
1
. Introduction
Oxidative stress is closely associated with aberrant exposure of reactive oxygen
-
species (ROS, also known as oxygenated free radical molecules) such as superoxide (O ),
2
.
hydrogen peroxide (H O ), and hydroxyl radical ( OH) [1]. Although an adequate amount of
2
2
ROS is necessary for maintaining proper redox homeostasis, a high amount of ROS is
detrimental because it provokes direct oxidative damages on cellular macromolecules [2]. In
order to deal with excessive oxidative stress, aerobic organisms possess a number of phase II
cytoprotective enzymes including heme oxygenase-1 (HO-1), whose transcription is
controlled by NF-E2-related factor 2 (NRF2) [3]. Under basal condition, NRF2 is constantly
poly-ubiquitinated by Cullin-3 (Cul3)/Kelch-like ECH-associated protein 1 (KEAP1) in the
cytoplasm. Upon exposure to pro-oxidants or electrophiles, NRF2 is released from
Cul3/KEAP1, translocates into the nucleus, and activates HO-1 transcription by binding to
the cis-acting element of promoter, termed antioxidant response element (ARE).
HO-1 is the first and rate-limiting enzyme in the heme degradation pathway to
2
+
generate three anti-oxidant products: carbon monoxide (CO), ferrous ion (Fe ), and
biliverdin [4]. CO protects lungs and hearts against endotoxin, ischemia/reperfusion injury,
cardiac xenograft rejection, and asthma by modulating cytokine production, cell proliferation,
and apoptosis [5]. Biliverdin and its reduced product bilirubin are efficient scavengers of
ROS [6]. Ferrous iron induces ferritin expression and contributes to iron sequestration [7].
Collectively, these imply that all enzymatic products of HO-1 possess beneficial activities.
Supporting this idea, HO-1 knock-out mice spontaneously develop inflammatory diseases
and are highly susceptible to experimental sepsis [8]. Phenotypical alterations of individuals
with genetic HO-1 deficiency are also similar to those observed in HO-1 knock-out mice [9].
Marliolide (Fig. 1A) is a natural product having γ-lactone ring with α,β-unsaturated
carbonyl group, which was firstly isolated from the leaves of Mollinedia marliae and the bark
of Cinnamomum cambodianum [10, 11]. However, there exist no literatures demonstrating
the pharmacological activity and related biochemical mechanisms of marliolide. In the
present study, we demonstrate that marliolide strongly induced HO-1 in human normal
keratinocyte HaCaT cells. We also demonstrate that marliolide could form a stable adduct
with KEAP1 and activate NRF2/ARE to induce HO-1. Finally, we provide evidence that
marliolide protects mice against oxidative damages in vivo, thereby exhibiting significant
skin chemopreventive activity.

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT
2
2
. Results
.1. Marliolide Induce HO-1 by NRF2-dependent ARE activation
Induction of phase II cytoprotective enzymes, including HO-1 has been recognized
as a feasible strategy to suppress carcinogenesis and other chronic diseases caused by
oxidants and electrophiles [12]. In addition, recent studies demonstrate that HO-1 can
suppress immune-mediated inflammatory disorders by limiting innate or adaptive immune
responses [13] or by promoting polarization of macrophages towards an anti-inflammatory
M2 phenotype [14]. Therefore, it is not surprising to observe that numerous HO-1 chemical
inducers were effective for treatment of chronic inflammatory disorders [15]. In accordance
with this, we have attempted to find new HO-1 inducers and found that marliolide strongly
induced HO-1 (Fig. 1B) without perturbing cellular integrity in human keratinocyte HaCaT
cells: marliolide did not affect the cell-cycle pattern (Fig. 1C) and the viability (Fig. 1D), and
failed to induce apoptosis (Fig. 1E).
To examine whether marliolide could activate ARE-dependent gene expression, we
subcloned DNA oligonucleotides containing 3x ARE sequence into pGreenFire dual reporter
plasmid and established stable HaCaT-ARE-GFP-luciferase cells by lentiviral transduction
and subsequent selection with puromycin (Fig. 2A). To examine whether HaCaT-ARE-GFP-
luciferase cells are responsive to ARE inducer, we exposed sulforaphane, a prototypical
inducer of ARE, to HaCaT-ARE-GFP luciferase cells and found that it significantly increased
ARE luciferase activity in HaCaT-ARE-GFP-luciferase cells (Fig. 2B). Likewise, marliolide
increased ARE-dependent GFP expression (Fig. 2C) and the luciferase activity of HaCaT-
ARE-GFP-luciferase cells in dose- (Fig. 2D, Left Panel) and time-dependent manner (Fig. 2D,
Right Panel).
Because NRF2 is a leucine-zipper transcription factor responsible for ARE-
dependent gene activation [16], we exposed marliolide to HaCaT cells and measured NRF2
level. Our results showed that marliolide increased NRF2 level in dose- (Fig. 3A, Upper
Panel) and time-dependent manner (Fig. 3A, Lower Panel). Marliolide also caused
translocation of NRF2 into the nucleus (Fig. 3B). To examine whether the induction of HO-1
by marliolide was dependent on Nrf2 genotype, we have prepared Nrf2 (+/+) and Nrf2 (-/-)
mouse embryonic fibroblasts (MEFs) (Fig. 3C, Upper Panel) and exposed them to marliolide.
As a result, the induction of HO-1 by marliolide was attenuated in Nrf2 (-/-) MEFs, compared
to that in Nrf2 (+/+) MEFs (Fig. 3C, Lower Panel). The induction of HO-1 mRNA level by

ACCEPTED MANUSCRIPT
marliolide was also significantly attenuated in Nrf2 (-/-) MEFs compared that in Nrf2 (+/+)
MEFs (Fig. 3D). These results illustrate that marliolide induction of HO-1 occurs through
NRF2/ARE activation.
2
.2. Chemistry
We have recently reported the asymmetric total synthesis of marliolide 2 [17].
Synthesis of marliolide 2 and its (Z)-isomer 3 is described in Fig. 4A. With marliolide 2 in
hand, we prepared saturated marliolide 4 by catalytic hydrogenation of 2 using H Pd/C (5%)
2
,
in presence of acetic acid [18]. Interestingly, along with the hydrogenated product 4 (73%),
we isolated and identified compound 5 (10%) in minor proportion, which is formed by the
hydrogenation of alkene and coincident dehydroxylation at β-position (Fig. 4B). During our
attempt to synthesize β-keto-γ-butyrolactone by oxidation of secondary hydroxyl group in
marliolide, we obtained compound 6 from unexpected rearrangement. Hence, to check out
the effects of β-hydroxyl group, we turned our attention to synthesizing compound 8 without
affecting other functionalities of 2. First, hydrogenation of internal olefinic bond in 6 using
H Pd/C (5%) in presence of acetic acid afforded compound 7, followed by elimination of
2
,
secondary alcohol produced the desired compound 8 (Fig. 4B). Further, compound 9 was
prepared by reacting 2 with isovaleryl chloride in presence of DMAP. To investigate the
effect of chirality of γ-methyl in marliolide, we utilized our strategy to synthesize epi-
marlioide 11 by way of the known sequential reactions (Fig. 4C). Condensation of the key
intermediate lactone epi-1 with myristyl aldehyde, selective protection of resulting
diastereomeric diol with TBS, and mesylation of remaining alcohol followed by elimination
afforded 10 in 16% yield for three steps. Finally, treatment of 10 with HCl produced the
desired epimer of marliolide, 11 [17]. A series of marliolide derivatives 22-31 with different
chain length at α-position were also prepared according to the previously reported general
methods (Fig. 4D).
2
.3. Structure Activity Relationship of Marliolide
It is generally accepted that NRF2/ARE-dependent gene activation by chemicals
largely occurs via three ways: (1) direct conjugation and subsequent inactivation of KEAP1
by Michael reaction [19], (2) direct inhibition between NRF2 and KEAP1 interface [20], and
(3) activation of intracellular signaling pathways, leading to NRF2 transactivation [21]. A
close examination of marliolide structure shows that it contains α,β–unsaturated lactone

ACCEPTED MANUSCRIPT
moiety belonging to Michael acceptor class. This intrigued us to examine whether marliolide
induction of HO-1 could be attributable to its binding to KEAP1. After synthesizing
biotinylated marliolide (please refer to Supporting information), we incubated biotin and
biotinylated marliolide (Fig. 5A, Left Panel) with HaCaT cell lysates and performed an
immunoprecipitation with streaptavidin-agarose bead followed by Western blotting with
KEAP1 and NRF2 antibodies. As a result, we observed that biotinylated marliolide could
selectively bind to KEAP1, but not NRF2 (Fig. 5A, Right Panel).
After setting up a hypothetic mechanism of HO-1 induction by marliolide: marliolide
binds to KEAP1 through Michael reaction (Fig. 5B), we synthesized a number of marliolide
derivatives and examined their effects on HO-1 level in HaCaT cells (Fig. 5C). As expected,
marliolide (2) caused a strong time-dependent induction of HO-1. 3 also induced HO-1,
although its induction kinetic was retarded compared with 2. However, 4 failed to induce
HO-1, providing a convincing clue that marliolide induction of HO-1 might occur via
Michael reaction. This hypothesis could be further examined by the use of another marliolide
derivative, in which the carbonyl group of the γ-butyrolactone ring was reduced into the
hydroxyl group. In spite of repetitive efforts, however, we failed to synthesize this derivative
due to its intrinsic instability: this might be due to hemi-ketal formation and subsequent
opening of γ-butyrolactol ring (Data not shown). In addition, we observed that both 5 and 8
failed to induce HO-1. The failure of 5 to induce HO-1 was somewhat unexpected in that the
hydroxyl group of γ-butyrolactone ring did not seem to participate in Michael reaction.
Although the exact reason is unclear, the observation that 9 failed to induce HO-1 renders a
strong support that hydroxyl group in γ-butyrolactone ring is critical for the induction of HO-
1
by marliolide. Finally, we observed that 11 induced HO-1, although its induction kinetic
was slower than 2. This indicates that configuration of hydroxyl group does not significantly
interfere with the ability of marliolide to induce HO-1.
We conducted a second set of SAR study to ask whether varying the aliphatic chain
length of marliolide could affect the degree of HO-1 induction (Fig. 5D). We observed that
both 22 and 23 failed to induce HO-1. When compared with 2, 24 exerted a weaker and 25
exhibited a retarded induction of HO-1. In contrast, 26, 27, 28, and 29 seemed to exhibit
equivalent induction of HO-1 compared with 2, while 30 and 31 exerted a weaker or delayed
induction of HO-1 compared with 2. To quantitatively compare the induction level of HO-1
mRNA, we exposed 2, 26, 27, 28, and 29 to HaCaT cells and conducted real-time RT-PCR

ACCEPTED MANUSCRIPT
assay at different time points with HO-1 specific primers. Compared with 2, 29 exhibited a
stronger induction of HO-1 mRNA at 4 h post-treatment, 26, 27, 28, and 29 exhibited a
stronger induction of HO-1 mRNA at 8 h post-treatment, and 26 and 27 exhibited a stronger
induction of HO-1 mRNA at 12 h post-treatment (Fig. 5D). This result suggests that 26, 27,
2
8, and 29 can exhibit a stronger induction of HO-1 mRNA than marliolide in vitro,
depending on the exposure time. Finally, our MTT assay result showed that all marliolide
derivatives used in the present study were not cytotoxic to HaCaT cells (Fig. 5E).
2
.4. Marliolide Attenuates DMBA/TPA-induced Papilloma Formation in Hairless Mice.
Because marliolide strongly induced NRF2/ARE-dependent HO-1 in HaCaT cells,
we examined whether marliolide could induce NRF2 and phase II cytoprotective enzymes in
vivo. To address this issue, we have topically applied marliolide and 29, which induced HO-1
mRNA comparably or stronger than marliolide (Fig. 5D) in the back of hairless mice and
conducted Western blotting using mouse skin lysates. Our results showed that both marliolide
and 29 increased the expression of NRF2 and phase II cytoprotective enzymes (HO-1 and
GCLC) in mouse skin (Fig. 6A). Because carcinogenesis requires multiple genetic mutations
that can be facilitated by exposure to oxidative stress [22], we assumed that marliolide would
exhibit anti-oxidant activities through induction of NRF2/HO-1 to inhibit carcinogenesis in
vivo. We adopted DMBA/TPA two-stage mouse skin carcinogenesis model to address this
question. During the course of study, topical application of marliolide significantly
suppressed the number (Fig. 6B, Upper Panel) and incidence (Fig. 6B, Lower Panel) of
DMBA/TPA-induced papilloma in the back of hairless mice. At autopsy, we observed that
marliolide inhibited the growth of DMBA/TPA-induced papilloma (Fig. 6C) and suppressed
the in vivo formation of 8-hydroxyguanosine (8-OH-G) and 4-hydroxynoneal (4-HNE) (Fig.
6
D), both of which are well-known oxidative stress markers. These results suggest that
marliolide exhibits significant anti-oxidant and anti-carcinogenic effects in vivo.

ACCEPTED MANUSCRIPT
3
. Conclusion
Previous studies have demonstrated that induction of NRF2/HO-1 can alleviate
diverse oxidative stress-related disorders such as cancer [23], diabetes [24], cardiovascular
diseases [25, 26]. In addition, natural compounds that possess the ability to induce HO-1 and
NRF2 can inhibit development of inflammatory diseases [27, 28]. Therefore, exploring new
NRF2/HO-1 activators for treatment of pro-inflammatory diseases aroused significant
commercial interests in the last decade. One example is an oral formulation of dimethyl
fumarate (DMF), which was developed for patients with recurrent multiple sclerosis (MS)
and finally approved by the Food and Drug Administration in 2013 [29]. In line with this idea,
we found that marliolide was not cytotoxic (Fig. 1) and strongly induced HO-1 by eliciting
NRF2-dependent ARE activation in HaCaT cells (Fig. 2 and 3). Our results also showed that
marliolide induced HO-1 by acting as Michael acceptor for KEAP1, but not NRF2 (Fig. 5A).
In addition, SAR study revealed chemical moieties of marliolide responsible for the induction
of HO-1 (Fig. 5C). It is known that KEAP1 contains 25 cysteine residues in mouse and 27
cysteine residues in human [30]. In particular, some cysteine residues of KEAP1 are located
adjacent to basic amino acid residues. Therefore, KEAP1 seems to contain reactive cysteine
residues with low pK values, which might serve as excellent targets for Michael reaction.
a
However, evidence is still lacking on which cysteine residue(s) of KEAP1 form a direct
adduct with marliolide, although several cysteine residues of KEAP1 (Cys 151, Cys 273 and
Cys 288, for example) were previously proposed as preferential sites for Michael reaction
[
31].
Our results show that marliolide and 29 are efficient inducers of NRF2 and its
downstream targets in vivo (Fig. 6A). Using DMBA/TPA two-stage mouse skin
carcinogenesis model, we observed that marliolide significantly inhibited the incidence of
papilloma in hairless mice (Fig. 6B). The inhibition of papilloma formation by marliolide
could be ascribed to its ability to induce NRF2-dependent phase II cytoprotective enzymes
(Fig. 6A) and lower the degree of oxidative stress in vivo (Fig. 6D). Moreover, we note that
2
6, 27, 28, and 29 elicited a stronger induction of HO-1 mRNA than marliolide (Fig. 5D) at
certain time points without apparent cytotoxicity in HaCaT cells (Fig. 5E). This raises a
possibility that 26, 27, 28, and 29 could be further harnessed as novel pharmacological leads
of NRF2/HO-1 inducers. Because 26 [(-)-licunolide B] [32], 28 [(-)-lincomolide B] [33] and
3
0 [(-)-isodihydromahubanolide B] [34] were found in particular species of fungus or plants,

ACCEPTED MANUSCRIPT
it is possible to speculate that 27 and 29 might exist somewhere in nature. However, the exact
roles of these marliolide derivatives are currently unclear.

ACCEPTED MANUSCRIPT
4
4
. Experimental Section
.1. Cell Culture, Chemicals, Plasmids, and Antibodies
DMEM, RPMI, FBS, and penicillin/streptomycin (Pen/Strep) were purchased from
WELGENE (Daegu, Korea). HaCaT cells were purchased from Korean Cell line Bank (Seoul,
Korea). HaCaT cells were grown in RPMI media supplemented with 10% FBS and
Penicillin/Streptomycin (100 U/ml). Sulforaphane, DMBA, TPA, etoposide, and antibodies
against NRF2, KEAP1, total actin, 8-OH-G were purchased from Santa Cruz biotechnology
(
Santa Cruz, CA, USA). Polybrene was purchased from Merck-Millipore (Merck-Millipore
Korea, Daejeon, Korea). pGreenfire reporter vector was purchased from System Bioscience
Palo Alto, CA, USA). Lentiviral helper plasmids (pMD2.G and psPAX.2) were acquired
(
from Addgene (Cambridge, MA, USA). 3x Tandem ARE DNA oligonucleotides were bought
from Macrogen (Seoul, Korea), annealed, and subcloned into pGreenfire vector using a
double-digestion with Cla1 and Spe1 restriction enzymes. HO-1 antibody was purchased
from Enzo Life Science (Farmingdale, NY, USA). Antibodies against GCLC, Cleaved-
Caspase-3 and Cleaved-PARP were purchased from Cell Signaling Technology (Beverly,
MA, USA). 4-HNE antibody was purchased from Abcam (Cambridge, MA, USA).
Marliolide and all derivatives were dissolved in DMSO (vehicle) and used in all in vitro
experiments at a dilution ratio of 1/1000.
4
.2. Measurement of Cell-Cycle by Fluorescence-activated Cell Sorting (FACS).
6
After treatment of marliolide, an equal number of HaCaT cells (1x10 /group) was
dispensed into glass tubes and fixed with a solution containing 1x PBS and 70% EtOH at a
ratio of 3:7. After washing with 1x PBS, cells were mixed with propidium iodide (20 µg/ml)
and the changes in cell-cycle were measured by FACS Calibur flow cytometer (Becton
Dickinson, San Jose, CA, USA).
4
.3. Immunofluorescence (IF) Assay
After treatment of marliolide, confluent HaCaT cells were grown on a slice glass and
incubated with blocking serum (1% BSA) for 30 min. After washing with 1x PBS, HaCaT
cells were fixed with paraformaldehyde and hybridized with primary antibodies overnight at
4
°C. These slides were washed with 1x PBS and probed with fluorescein isothiocyanate
(FITC)-conjugated rabbit secondary antibodies (Jackson-ImmunoResearch, West Grove, PA,

ACCEPTED MANUSCRIPT
USA). The fluorescent images were obtained with C2 confocal microscope (Nikon Korea,
Seoul, Korea).
4
.4. Generation of HaCaT-GFP-luciferase Cells by Lentiviral Transduction
Using JETPEI reagent (Polyplus-Transfection, New York, NY, USA), 293T
packaging cells were transfected with 3 µg pGreenFire-ARE plasmid together with lentiviral
helper vectors (3 µg pMD2.G and 3 µg psPAX.2). After 48 h transfection, viral supernatant
was collected, filtered, and transduced into confluent HaCaT cells, which were further
selected with puromycin (1 µg/ml) for 48 h.
4
.5. Monitoring GFP and Luciferase Activity in HaCaT-ARE-GFP-luciferase Cells
Established HaCaT-ARE-GFP-luciferase cells were plated in 24-well culture plates
5
at a density of 2x10 cells/well. After treatment of marliolide, cellular GFP expression was
observed by fluorescent microscopy. For the luciferase assay, HaCaT-ARE-GFP-luciferase
cells were washed with 1x PBS and lyzed with luciferase lysis buffer [0.1 M potassium
phosphate buffer at pH 7.8, 1 % Triton X-100, 1 mM DTT, 2 mM EDTA]. The luciferase
activities were measured by GLOMAX Multi-system. (Promega, Madison, WI, USA) and
normalized by protein concentration.
4
.6. Western Blot Analysis
After treatment of marliolide and derivatives, cultured cells were collected and
washed with ice-cold 1x PBS buffer. Cells were collected by centrifugation and resuspended
with RIPA lysis buffer [50 mM Tris-HCl at pH 8.0, 150 mM NaCl, 1% NP-40, 0.5%
deoxycholic acid, 0.1% sodium dodecyl sulfate (SDS), 1 mM Na VO , 1 mM dithiothreitol
3
4
(DTT), 1 mM phenylmethylsulfonyl fluoride (PMSF)] on ice for 30 min. Cell lysates were
collected and protein concentration was measured by BCA Protein Assay Kits (Thermo-
Fisher Scientific, Waltham, MA, USA). Equal amounts of cell lysates were resolved by SDS-
PAGE and transferred to PVDF membranes. After 1 h incubation in blocking buffer (5% skin
milk in 1x PBST), the membrane was hybridized with appropriated primary antibodies
overnight at 4℃. After washing with 1x PBST three times, the membranes were hybridized
with HRP-conjugated secondary antibodies (Thermo Fischer scientific, Waltham, MA, USA).
After washing with 1x PBST three times, the membranes were finally visualized by ECL
detection system. Total actin blot was used as a control to illustrate an equal loading of

ACCEPTED MANUSCRIPT
samples.
4
.7. Real-time Reverse Transcription-Polymerase Chain Reaction (RT-PCR)
After treatment of marliolide and derivatives, HaCaT cells were collected and total
RNA was extracted by Hybrid-R RNA extraction kit (GeneAll, Seoul, Korea). Total RNA
was subject to cDNA synthesis using PrimeScript RT-PCR kit (TAKARA Korea, Seoul,
Korea). Real-time RT-PCR analysis was performed using SYBR mix on CFX384 Real-time
system (BioRad, Hercules, CA, USA). Amplification protocol comprises following PCR
cycles: a single cycle of 5 min at 95℃, 40 cycles of 10 sec at 95℃, 10 sec at 59℃, and 20
sec 72℃, and a final cycle of 10 sec at 95℃. Real-time PCR primers used in the present
study are as follows: HO-1 PCR primers [5’-ATGCCCCAGGATTTGTCAGA-3’ (Forward)
and 5’-ACCTGGCCCTTCTGAAAGTT-3’ (Reverse)] and GAPDH PCR primers [5’-CAC
AGTCCATGCCATCACTG-3’ (Forward) and 5’-GTCCACCACTGACACGTTG-3’
(Reverse)]. The relative mRNA level of HO-1 was normalized by that of GAPDH.
4
.8. MTT assay.
HaCaT cells were seeded in 96 well plate (30,000 cells/well) and exposed to
marliolide or its derivatives for 24 h. After washing with 1x PBS, cells were incubated with
00 µl MTT solution (5 mg/ml) for 4 h. After a lysis with 100 µl DMSO, the resulting
absorbance was measured by spectrophotometer at the wavelength of 560 nm.
1
4
.9. Immunoprecipitation of KEAP1 by Biotinylated Marliolide
Confluent HaCaT cells were lyzed with RIPA buffer. HaCaT lysates (800 µg) were
then mixed with biotin (10 nmole) and biotinylated marliolide (10 nmole) followed by
immunoprecipitation with streptavidin-agarose bead (Cell Signaling Technology, Waltham,
MA, USA) overnight at 4°C. After washing beads with ice-cold 1x PBS buffer, Western blot
analysis was conducted using KEAP1 and NRF2 antibodies.
4
.10. DMBA/TPA-induced Two-Stage Mouse Skin Carcinogenesis Model
Six-week male hairless mice were purchased from Daehan Biolink (Eumseong,
Korea). Animals were housed in sterile filter-capped microisolator cages and provided with
water and diet ad libitum. After a week acclimation, individual mice were topically applied
with acetone (vehicle) or DMBA (2 µmole) once in the first week. Starting from the second

ACCEPTED MANUSCRIPT
week, the back of hairless mice were applied with TPA (10 nmole) alone or in combination
with marliolide (10 µmole) three times per week. During the course of experiment, the
number of papilloma was manually counted at the end of every week. At sacrifice, mouse
skin was immersed 30% formalin and embedded in paraffin block. After preparation of slides,
hematoxylin/eosin (H/E) staining was conducted and the images were captured by phase-
contrast microscopy. For immunofluorescence (IF) assay, the slides were hybridized with
primary 8-OH-G and 4-HNE antibodies overnight at 4℃ and probed with FITC-conjugated
secondary antibodies (Jackson Immunoresearch, West grove, PA, USA). The fluorescent
image were obtained with C2 confocal microscope (Nikon Korea, Seoul, Korea). The animal
experiment was carried out under an Institutional Animal Care and Use Committee-approved
Protocol (IACUC-2015-066) from Dongguk University (Seoul, Korea).
4
.11. Statistics
Statistical analysis was conducted using Student’s t-test. Asterisks indicate a
statistical significance with *P<0.05, **P<0.01 and *** p<0.001.
4
.12. General Chemical Procedures
Except where noted, all the materials were purchased from commercial suppliers and
used without further purification. All reactions were routinely carried out under an inert
atmosphere of dried nitrogen, in hot-oven dried glassware. Proton nuclear magnetic
1
resonance ( H-NMR) Spectra (CDCl , DMSO-d ) were recorded on a Varian (400 MHz)
3
6
spectrometer (Varian Medical Systems, Inc., Palo Alto, CA, USA). The 1H-NMR data are
reported as peak multiplicities: s for singlet, d for doublet, dd for doublet of doublets, t for
1
3
triplet, q for quartet, br s for broad singlet and m for multiplet. C-NMR spectra (CDCl ,
3
DMSO-d ) were recorded on Varian (100 MHz) spectrometer. The chemical shifts are
6
reported as parts per million (ppm), downfield from tetramethylsilane (TMS) internal
standard relative to the solvent peak with coupling constants in hertz (Hz). Optical rotations
were determined on Jasco P-2000 polarimeter in appropriate solvent. Infrared spectra were
recorded on FT-IR (NICOLET-iS5). Melting points were measured on Thermoscientific-9200.
Elemental analyses (C, H) were used to determine purity of all synthesized compounds, and
the results were within ± 0.4% of the calculated values, confirming > 95% purity (EA1110
CHNS-0, CE Instrument, Italy). Reactions were monitored with TLC (Merck precoated
6
0F254 plates). Spots were detected by viewing under a UV light, colorizing with charring

ACCEPTED MANUSCRIPT
after dipping in anisaldehyde solution or basic KMnO solution. Column chromatography
4
was performed on silica gel 60 (230-400 mesh Kieselgel 60). The mass spectra were recorded
using LRMS (electron ionization MS) obtained on a Shimadzu-2020 or using HRMS
(electrospray ionization MS) obtained on a G2 QTOF mass spectrometer.
4
.13. Synthetic Procedures of Marliolide and Its Derivatives
E,4S,5S)-Dihydro-4-hydroxy-5-methyl-3-tetradecylidenefuran-2(3H)-one,
2). Marliolide 2 was obtained as off-white solid in accordance with reported procedure [17]
(
marliolide
(
o
20
1
starting from
D-ribose: mp 72-74 C; [α] = -92.0 (c=1.0, CHCl ); H NMR (400 MHz,
D 3
CDCl ) δ 6.97-6.93 (t, J = 8.0 Hz, 1H), 4.83-4.81 (t, J = 5.2 Hz, 1H), 4.56-4.50 (m, 1H),
3
2
.45-2.34 (m, 2H), 1.65-1.59 (m, 1H), 1.56-1.50 (m, 2H), 1.45 (d, J = 6.8 Hz, 3H), 1.39-1.25
13
(
bs, 20H), 0.88 (t, J = 6.4 Hz, 3H); C NMR (100 MHz, CDCl ) δ 169.9, 147.8, 130.5, 78.7,
3
6
7.8, 31.9, 29.9-29.4, 28.4, 22.7, 14.1, 14.0; IR (neat) 3383, 2920, 2846, 1745, 1692, 1212,
-1
+
1
050 cm ; HRMS (ESI) m/z calcd for C H O [M+H] : 311.2583, found: 311.2583; Anal.
1
9
34
3
calculated for C H O : C, 73.50; H, 11.04, found: C, 73.48; H, 11.03.
1
9
34
3
(Z,4S,5S)-Dihydro-4-hydroxy-5-methyl-3-tetradecylidenefuran-2(3H)-one
(3).
Compound 3 was obtained as off-white solid in accordance with reported procedure [17]: mp
o
20
1
5
5-57 C; [α] = -50.1 (c=0.6, CHCl ); H NMR (400 MHz, CDCl ) δ 6.58-6.54 (t, J = 6.4
D 3 3
Hz, 1H), 4.66-4.64 (t, J = 5.6 Hz, 1H), 4.56-4.52 (m, 1H), 2.80-2.68 (m, 2H), 1.70 (d, J = 6.0
Hz, 1H), 1.50-1.42 (m, 2H), 1.39 (d, J = 6.4 Hz, 3H), 1.39-1.25 (bs, 20H), 0.88 (t, J = 6.4 Hz,
1
3
3
H); C NMR (100 MHz, CDCl ) δ 168.9, 149.9, 129.2, 77.9, 71.4, 31.9, 29.7-29.4, 29.3,
3
-1
2
8.8, 27.9, 22.6, 14.1, 14.1; IR (neat) 3383, 2920, 2846, 1745, 1692, 1212, 1050 cm ; LCMS
+
(
ESI) m/z calcd for C H O [M+H] : 311.26, found: 311.35; Anal. calculated for C H O :
1
9
34
3
19 34
3
C, 73.50; H, 11.04, found: C, 73.35; H, 10.01.
(
3R,4S,5R)-3-Dihydro-4,5-dihydroxy-3-tetradecylfuran-2(3H)-one (4) and (3R,5R)-3-
Dihydro-5-hydroxy-3-tetradecylfuran-2(3H)-one (5). To a stirred solution of 2 (30 mg,
.096 mmol) in ethanol (1 ml), were added 5% Pd/C (5 mg) and AcOH, and the suspension
0
was stirred overnight under an atmosphere of hydrogen at room temperature. The resulting
suspension was filtered through a pad of celite and the filtrate was evaporated. The crude
residue was purified by silica gel column chromatography (hexanes/EtOAc=9:1 v/v) to afford
o
20
1
4
(22 mg, 73%) as off-white solid: mp 91-93 C; [α] = -33.3 (c=1.0, CHCl ); H NMR (400
D 3

ACCEPTED MANUSCRIPT
MHz, CDCl ) δ 4.48-4.42 (m, 1H), 4.31 (t, J = 3.8 Hz, 1H), 2.54 (m, 1H), 2.32 (bs, 1H),
3
1
.85-1.76 (m, 1H), 1.7-1.58 (m, 1H), 1.44 (d, J = 6.4 Hz, 3H), 1.44-1.39 (bm, 2H), 1.3-1.25
1
3
(
bs, 22H), 0.88 (t, J = 6.6 Hz, 3H); C NMR (100 MHz, CDCl ) δ 178.1, 79.1, 71.1, 47.6,
3
-1
2
9.3-29.7, 31.9, 27.6, 23.3, 22.7, 14.1, 13.7; IR (neat) 2913, 2851, 1731, 1588, 1203 cm ;
+
HRMS (ESI) m/z calcd for C H O [M+H] : 313.2743, found: 313.2758; Anal. calculated
1
9
36
3
for C H O : C, 73.03; H, 11.61, found: C, 73.38; H, 11.73. Unintended minor product,
1
9
36
3
o
20
compound 5 (3 mg, 10.4%) was obtained as off-white solid: mp 58-60 C; [α] D= -11.9
1
(
c=0.28, CHCl ); H NMR (400 MHz, CDCl ) δ 4.46 (m, 1H), 2.63-2.55 (m, 1H), 2.5-2.43
3
3
(m, 1H), 1.95-1.89 (bm, 1H), 1.46 (m, 1H), 1.42 (d, J = 6.0 Hz, 3H), 1.31-1.25 (bs, 25H),
1
3
0
.88 (t, J = 6.6 Hz, 3H); C NMR (100 MHz, CDCl ) δ 179.0, 75.0, 41.5, 37, 31.9, 30.3,
3
-1
2
9.3-29.6, 27.4, 22.6, 21.0, 14.1; IR (neat) 2918, 2846, 1757, 1466, 1206 cm ; HRMS (ESI)
+
m/z calcd for C H O [M+H] : 297.2794, found: 297.2637; Anal. calculated for C H O :
1
9
36
2
19 36
2
C, 76.97; H, 12.24, found: C, 76.89; H, 12.23.
(S)- 3-(1-Hydroxyotetradecyl)-5-methylfuran-2(5H)-one (6). To a solution of 2 (15 mg,
0
.05 mmol) in acetone (1 ml), Jones reagent (0.5 ml (prepared from 35 mg of Cr O + 0.03
2
3
o
o
ml of conc. H SO + 0.5 ml of water at 0 C)) was added at 0 C. Reaction mixture was
2
4
slowly warmed to room temperature and stirred for 3 h. The reaction mixture was quenched
with sat. NaHCO solution and extracted with ethyl acetate. Combined organic layer was
3
dried over MgSO , filtered and evaporated. The crude residue was purified by silica gel
4
column chromatography (hexanes/EtOAc=9:1 v/v) to give 6 (10 mg, 66%) as off-white solid:
1
H NMR (400 MHz CDCl ) δ 7.19 (m, 1H), 5.09-5.04 (m, 1H), 4.48 (m, 1H), 1.81-1.63 (m,
3
2
H), 1.45 (d, J=6.8 Hz, 3H), 1.25 (bs, 21H), 0.88 (t, J=8.6 Hz, 3H).
(5S)-Dihydro-3-(1-hydroxyotetradecyl)-5-methylfuran-2(3H)-one (7). To
a
stirred
solution of 6 (10 mg, 0.03 mmol) in ethanol (1 ml), were added catalytic amount of AcOH
and 5% Pd/C (2 mg) and the flask was sealed with a hydrogen balloon. The suspension was
stirred at room temperature for 1 h. The resulting suspension was filtered through a pad of
celite and the filtrate was evaporated. The crude residue was purified by silica gel column
1
chromatography (hexanes/EtOAc=9:1 v/v) to afford 7 (10 mg, 99%) as off-white solid: H
NMR (400 MHz, CDCl ) δ 4.49-4.41 (m, 1H), 4.19 (m, 1H), 2.78 (m, 1H), 2.61 (m, 1H),
3
2
.43 (m, 1H), 2.04-1.95 (bm, 2H), 1.45 (m, 8H), 1.25 (bs, 42H), 0.89 (m, 6H).

ACCEPTED MANUSCRIPT
(E,S)-Dihydro-5-methyl-3-tetradecylidenefuran-2(3H)-one (8). To a solution of 7 (10 mg,
0
.03 mmol) in methylene chloride (1.0 ml), triethylamine (0.06 ml, 0.48 mmol) and
o
methanelsulfonyl chloride (0.01 ml, 0.16 mmol) were added at 0 C and the reaction mixture
was stirred at the same temperature for 3 h. Excess TEA (0.06 ml, 0.48 mmol) was added and
o
the reaction mixture was slowly warmed to 45 C and maintained overnight. Reaction
completion was monitored by TLC. Reaction mixture was diluted with methylene chloride
and washed with sat. NaHCO solution. Organic layer was dried over MgSO , filtered and
3
4
evaporated. The crude residue was purified by silica gel column chromatography
1
(
hexanes/EtOAc=9.8:0.2 v/v) to afford 8 (1.6 mg, 17%, E-isomer) as white solid: H NMR
(
400 MHz, CDCl ) δ 6.7-6.75 (m, 1H), 4.69-4.64 (m, 1H), 3.02 (m, 1H), 2.43-2.37 (m, 1H),
3
2
3
.18-2.12 (m, 2H), 1.48 (m, 2H), 1.42 (d, J = 6.4 Hz, 3H), 1.25 (bs, 20H), 0.89 (t, J = 6.8 Hz,
H).
(E,2S,3S)-Tetrahydro-2-methyl-5-oxo-4-tetradecylidenefuran-3-yl
3-methylbutanoate
(9). To a stirred solution of 2 (25 mg, 0.08 mmol) in methylene chloride (2 ml), were added
DMAP (49 mg, 0.41 mmol) and isovaleryl chloride (73 mg, 0.61 mmol). The reaction
mixture was stirred at room temperature overnight. The reaction mixture was quenched with
sat. NaHCO solution and extracted in methylene chloride (3 times). Combined organic layer
3
was washed with brine solution, dried over MgSO , filtered and evaporated. The crude
4
residue was purified by silica gel column chromatography (hexanes/EtOAc=9:1 v/v) to afford
20
1
9
(25 mg, 79%) as a syrupy solid: [α] = -73.5 (c=0.88, CHCl ); H NMR (400 MHz, CDCl )
D 3 3
δ 6.98 (dt, J = 8.0, 1.4 Hz, 1H), 6.06 (d, J = 4.8 Hz, 1H), 4.69-4.63 (m, 1H), 2.29-2.22 (m,
H), 2.25 (d, J = 7.2 Hz, 2H), 2.15-2.08 (m, 1H), 1.48 (m, 2H), 1.35 (d, J = 6.0 Hz, 3H), 1.25
2
13
(
bs, 20 H), 0.97 (s, 3H), 0.96 (s, 3H), 0.88 (t, J = 6.8 Hz, 3H); C NMR (100 MHz, CDCl ) δ
3
1
72.1, 169.2, 148.8, 126.7, 68.2, 43, 31.9, 30.2, 29.6-29.3, 28.2, 25.6, 22.6, 22.3, 14.2, 14.1;
-1
IR (neat) 2926, 2853, 1767, 1737, 1680, 1455, 1289, 1112, 989, 759 cm ; HRMS (ESI) m/z
+
calcd for C H O [M+H] : 395.3161, found: 395.3169; Anal. calculated for C H O : C,
2
4
42
4
24 42
4
7
3.05; H, 10.73, found: C, 73.58; H, 10.63.
(E,4R,5S)-Dihydro-4--(tert-butyldimethysilyl)-5-methyl-3-tetradecylidenefuran-2(3H)-
one (10). To a stirred solution of lithium diisopropylamide (20.37 mmol, 2.0 M in heptane) in
o
THF (2 ml) at -78 C, was added a solution of epi-1 (500 mg, 4.31 mmol) in THF (5 ml)
dropwise via syringe under N atmosphere. After 2 h of stirring, a solution of myristyl
2

ACCEPTED MANUSCRIPT
aldehyde (1.1 g, 5.17 mmol) in dry THF (5 ml) and HMPA (5 ml) was added dropwise via
o
o
syringe at -78 C and the reaction mass was allowed to warm to -30 C over a period of 3 h
and the same temperature was maintained for 3 h. The reaction mixture was quenched with
3
0 ml of sat. NH Cl and extracted with EtOAc (50ml x 3). Combined organic layer washed
4
with sat. brine, dried over MgSO , filtered and evaporated. The crude residue was purified by
4
silica gel column chromatography (EtOAc/CH Cl =0-20% EtOAc) gave diastereomeric
2
2
mixture product diol (384 mg). The purified product was taken for next step without further
analysis. To a stirred solution of above product (1.17 mmol) in DMF (4 ml) at RT, imidazole
(
2.34 mmol) and tert-butyldimethylchlorosilane (1.4 mmol) were added under N atmosphere
2
and the reaction mixture was stirred for overnight. Reaction mixture was quenched with sat.
NaHCO and extracted with EtOAc. Organic layer was dried over MgSO , filtered and
3
4
evaporated. The crude residue was purified by silica gel column chromatography
hexanes/EtOAc=9:1v/v) gave TBS protected diastereomeric mixture product (311 mg). The
(
purified product was taken for next step without further analysis. To a stirred solution of TBS
protected product (0.702 mmol) in CH Cl (4.0 ml), triethylamine (10.6 mmol) and
2
2
o
methanelsulfonyl chloride (3.44 mmol) were added at 0 C and the reaction mixture was
stirred at the same temperature for 3 h. Excess TEA (10.6 mmol) was added and the reaction
o
mixture was slowly warmed to 45 C and the temperature was maintained for overnight.
Reaction completion was monitored by TLC. Reaction mixture was diluted with CH Cl and
2
2
washed with sat. NaHCO . Organic layer was dried over MgSO , filtered and evaporated.
3
4
The crude residue was purified by silica gel column chromatography
hexanes/EtOAc=9.8:0.2 v/v) to afforded 10 (214 mg, 16% for three steps) as a white solid:
(
o
1
mp 52-54 C; H NMR (400 MHz, CDCl ) δ 6.95-6.91 (t, J = 6.6 Hz, 1.2, 1H), 4.51 (s, 1H),
3
4
.43-4.38 (q, J = 6.6 Hz,1H), 2.34-2.23 (m, 2H), 1.51-1.45 (m, 2H), 1.3 (d, J = 6.8 Hz, 3H),
.26 (bs, 20H), 0.88 (bm, 12H), 0.13 (d, J = 6.8 Hz, 6H).
1
(E,4R,5S)-Dihydro-4-hydroxy-5-methyl-3-tetradecylidenefuran-2(3H)-one (11). To a
solution of 10 (150 mg, 0.35 mmol) in methanol (5 ml), 1M HCl (0.3 ml) was added and the
o
reaction mixture was heated to 60 C and maintained overnight. The reaction completion was
monitored by TLC. Reaction mixture was cooled to room temperature, evaporated in high
vacuo. The crude residue dissolved in ethyl acetate, washed with sat. NaHCO solution.
3
Organic layer was dried over MgSO , filtered and evaporated. The crude residue was purified
4
by silica gel column chromatography (hexanes/EtOAc=9:1 v/v) to give 11 (77 mg, 70%) as

ACCEPTED MANUSCRIPT
o
20
1
off-white solid: mp 48-50 C; [α] = 42.6 (c=0.1, CHCl ); H NMR (400 MHz, CDCl ) δ 7.0
D
3
3
(t, J = 8.0, 1H), 4.55 (d, J = 6.4 Hz, 1H), 4.52-4.47 (m, 1H), 2.8-2.2 (m, 2H), 1.92 (bs, 1H),
1
3
1
.54-1.48 (m, 2H), 1.35 (d, J = 6.4 Hz, 3H), 1.3-1.26 (bs, 20H), 0.88 (t, J = 6.6 Hz, 3H); C
NMR (100 MHz, CDCl ) δ 169.4, 148.6, 129.3, 82.4, 72.2, 31.9, 29.7-29.3, 28.4, 22.6, 19.7,
3
-1
1
4.1; IR (neat) 3365, 2919, 2846, 1723, 1681, 1463, 1341, 1265, 1047 cm ; HRMS (ESI) m/z
+
calcd for C H O [M+H] : 311.2586, found: 311.2573; Anal. calculated for C H O : C,
1
9
34
3
19 34
3
7
3.50; H, 11.04, found: C, 73.48; H, 11.01.
General procedure for the preparation of compounds 12-21
o
To a stirred solution of lithium diisopropylamide (4.5 eq., 2.0 M in heptane) in THF at -78 C,
was added a solution of 1 (1eq.) in THF dropwise via syringe under N atmosphere. After 2 h
2
of stirring, a solution of proper aldehyde (1.2eq.) in dry THF and HMPA (1:1, v/v) was added
o
o
dropwise via syringe at -78 C and the reaction mass was allowed to warm to -30 C over a
period of 3 h and the same temperature was maintained for 3 h. The reaction mixture was
quenched with sat. NH Cl and extracted with EtOAc (three times). Combined organic layer
4
washed with brine, dried over MgSO , filtered and evaporated. The crude residue was
4
purified by silica gel column chromatography (EtOAc/CH Cl =0-20% EtOAc) gave
2
2
diastereomeric mixture product. The purified product was taken for next step without further
analysis. To a stirred solution of above product (1eq.) in DMF at RT, imidazole (2.0 eq.) and
tert-Butyldimethylchlorosilane (1.2 eq.) were added under N atmosphere and the reaction
2
mixture was stirred overnight. Reaction mixture was quenched with sat. NaHCO and
3
extracted with EtOAc. Organic layer was dried over MgSO , filtered and evaporated. The
4
crude residue was purified by silica gel column chromatography (hexanes/EtOAc=9:1v/v)
gave TBS protected diastereomeric mixture product. The purified product was taken for next
step without further analysis. To a stirred solution of TBS protected product (1 eq.) in CH Cl ,
2
2
o
triethylamine (15 eq.) and methanelsulfonyl chloride (5 eq.) were added at 0 C and the
reaction mixture was stirred at the same temperature for 3 h. Excess TEA (15 eq.) was added
o
and the reaction mixture was slowly warmed to 45 C and the temperature was maintained
overnight. Reaction completion was monitored by TLC. Reaction mixture was diluted with
CH Cl and washed with sat. NaHCO . Organic layer was dried over MgSO , filtered and
2
2
3
4
evaporated. The crude residue was purified by silica gel column chromatography
hexanes/EtOAc=9.8:0.2 v/v).
(

ACCEPTED MANUSCRIPT
(E,4S,5S)-Dihydro-4-(tert-butyldimethysilyl)-5-methyl-3-butylidenefuran-2(3H)-one (12)
o
1
Waxy solid, yield 13% (for 3 steps): mp 32-34 C; H NMR (400 MHz, CDCl ): δ 6.8 (t, J =
3
7
.8 Hz, 1H), 4.8 (d, J = 4.8 Hz, 1H), 4.45 (m, 1H), 2.38-2.25 (m, 2H), 1.57-1.48 (m, 2H),
.37 (d, J = 6.4 Hz, 3H), 0.98 (t, J = 7.2 Hz, 3H), 0.89 (bs, 9H), 0.11 (s, 6H).
1
(E,4S,5S)-Dihydro-4-(tert-butyldimethysilyl)-5-methyl-3-hexylidenefuran-2(3H)-one (13)
1
Colorless oil, yield 14% (for 3 steps): H NMR (400 MHz, CDCl ) δ 6.67 (t, J = 7.2 Hz, 1H),
3
4
.7 (d, J = 4.8 Hz, 1H), 4.35 (m, 1H), 2.28-2.14 (m, 2H), 1.43-1.32 (m, 2H), 1.26 (d, J = 6 Hz,
3
H), 1.22-1.17 (bm, 4H), 0.78 (bs,12H), 0.00 (s, 6 H).
(E,4S,5S)-Dihydro-4-(tert-butyldimethysilyl)-5-methyl-3-octaylidenefuran-2(3H)-one (14)
1
Colorless oil, yield: 14% (for 3 steps): H NMR (400 MHz, CDCl ) δ 6.8-6.76 (t, J = 7.6Hz,
3
1
H), 4.83 (d, J = 4.4 Hz, 1H), 4.45-4.4 (m, 1H), 2.37-2.25 (m, 2H), 1.53-1.41 (m, 2H), 1.37
(d, J = 6.4 Hz, 3H), 1.32-1.27 (bm, 8H), 0.89 (bs, 9H), 0.88 (t, J = 7.6 Hz, 3H), 0.11 (s, 6H).
(E,4S,5S)-Dihydro-4-(tert-butyldimethysilyl)-5-methyl-3-nonylidenefuran-2(3H)-one (15)
1
Colorless oil, yield 12% (for 3 steps): H NMR (400 MHz, CDCl ) δ 6.8-6.76 (t, J = 7.8Hz,
3
1
H), 4.84 (d, J = 4.8 Hz, 1H), 4.45-4.12 (m, 1H), 2.39-2.25 (m, 2H), 1.53-1.41 (m, 2H), 1.37
(d, J = 6.4 Hz, 3H), 1.27 (bs, 10H), 0.89 (bs, 9H), 0.88 (t, J = 8 Hz, 3H), 0.11 (s, 6H).
(E,4S,5S)-3-Decylidene-dihydro-4-(tert-butyldimethysilyl)-5-methylfuran-2(3H)-one (16)
1
Colorless oil, yield 13% (for 3 steps): H NMR (400 MHz, CDCl ) δ 6.8-6.76 (dt, J = 1.4,
3
6.8Hz, 1H), 4.84 (d, J = 4.8 Hz, 1H), 4.47-4.41 (m, 1H), 2.39-2.25 (m, 2H), 1.53-1.43 (m,
2H), 1.37 (d, J = 6.8 Hz, 3H), 1.32-1.26 (bs, 12H), 0.89 (bs, 9H), 0.88 (t, J = 7.2 Hz, 3H),
0.11 (s, 6H).
(E,4S,5S)-Dihydro-4-(tert-butyldimethysilyl)-5-methyl-3-undecylidenefuran-2(3H)-one
(17)
1
Colorless oil, yield 13% (for 3 steps): H NMR (400 MHz, CDCl ) δ 6.8-6.64 (t, J = 7.8Hz,
3
1
H), 4.72 (d, J = 4.8 Hz, 1H), 4.35-4.29 (m, 1H), 2.28-2.14 (m, 2H), 1.42-1.31 (m, 2H), 1.26
(d, J = 6 Hz, 3H), 1.15 (bs, 14H), 0.78 (bs, 9H), 0.77 (t, J = 6.8 Hz, 3H), 0.0 (s, 6H).
(E,4S,5S)-3-Dodecylidene-dihydro-4-(tert-butyldimethysilyl)-5-methylfuran-2(3H)-one

ACCEPTED MANUSCRIPT
(18)
1
Colorless oil, yield 13% (for 3 steps): H NMR (400 MHz, CDCl ) δ 6.8-6.76 (t, J = 7.6 Hz,
3
1
H), 4.84 (d, J = 4.4 Hz, 1H), 4.47-4.41 (m, 1H), 2.39-2.25 (m, 2H), 1.53-1.41 (m, 2H), 1.37
(d, J = 6.4 Hz, 3H), 1.26 (bs, 16H),0.89 (bs, 9H), 0.88 (t, J = 7.6 Hz, 3H), 0.11 (s, 6H).
(
(
E,4S,5S)-Dihydro-4-(tert-butyldimethysilyl)-5-methyl-3-tridecylidenefuran-2(3H)-one
19)
1
Colorless oil, yield 14% (for 3 steps): H NMR (400 MHz, CDCl ) δ 6.8-6.76 (dt, J = 1.4, 7.6
3
Hz, 1H), 4.83 (d, J = 4.4 Hz, 1H), 4.46-4.4 (m, 1H), 2.37-2.25 (m, 2H), 1.53-1.42 (m, 2H),
1
.37 (d, J = 6.4 Hz, 3H), 1.26 (bs, 18 H), 0.89 (bs, 9H), 0.88 (t, J = 7.2 Hz, 3H), 0.11 (s, 6H).
(
E,Z,4S,5S)-Dihydro-4-(tert-butyldimethysilyl)-5-methyl-3-hexadecylidenefuran-2(3H)-
one (20)
Colorless oil, yield 13% (for 3 steps): H NMR (400 MHz, CDCl ) δ 6.8 (t, J = 7.8 Hz, 1H),
1
3
4
.84 (d, J = 4.4 Hz, 1H), 4.46 (m, 1H), 2.39-2.25 (m, 2H), 1.53-1.41 (m, 2H), 1.37 (d, J = 6.8
Hz, 3H), 1.29 (bs, 24H), 0.89 (bs, 9H), 0.88 (t, J = 7.2 Hz, 3H), 0.11 (s, 6H).
(E,4S,5S)-Dihydro-4-(tert-butyldimethysilyl)-5-methyl-3-octadecylidenefuran-2(3H)-one
(21)
1
Colorless oil, yield 13% (for 3 steps): H NMR (400 MHz, CDCl ) δ 6.8-6.75 (dt, J = 1.6,
3
7
Hz, 1H), 4.83 (d, J = 4.4 Hz, 1H), 4.45-4.4 (m, 1H), 2.37-2.27 (m, 2H), 1.51-1.4 (m, 2H),
.37 (d, J = 6.8 Hz, 3H), 1.25 (bs, 28 H), 0.9 (bs, 9H), 0.88 (t, J = 7.2 Hz, 3H), 0.11 (s, 6H).
1
General procedure for the preparation of compounds 22-31
To a stirred solution of above TBS protected product (0.2 mmol) in methanol (4 ml), 1M aq.
o
HCl (0.2 ml) was added, and the reaction mixture was slowly heated to 60 C and maintained
overnight. The reaction completion was monitored by TLC. Reaction mixture was cooled to
room temperature, volatiles were evaporated. The crude residue was dissolved in EtOAc,
washed with sat. NaHCO . Organic layer was dried over MgSO , filtered and evaporated.
3
4
The crude residue was purified by silica gel column chromatography (hexanes/EtOAc=4:1
v/v).

ACCEPTED MANUSCRIPT
(E,4S,5S)-3-Butylidene-dihydro-4-hydroxy-5-methylfuran-2(3H)-one (22). Compound 22
o
20
1
was obtained as waxy white solid (65%): mp 44-46 C; [α] = -174.7 (c=1.2, CHCl ); H
D
3
NMR (400 MHz, CDCl ) δ 6.9 (dt, J = 6.4, 1.4 Hz, 1H), 4.8 (t, J = 5.8 Hz, 1H), 4.57-4.5 (m,
3
1
0
3
H), 2.44-2.33 (m, 2H), 1.9 (d, J = 7.2 Hz, 1H) 1.62-1.53 (m, 2H), 1.45 (d, J = 6.8 Hz, 3H),
13
.98 (t, J = 7.2 Hz, 3H); C NMR (100 MHz, CDCl ) δ 170.0, 147.5, 130.7, 78.7, 67.7,
3
-1
1.7, 21.7, 13.9, 13.8; IR (neat): 3415, 2960, 1731, 1674, 1196, 1035, 987 cm ; HRMS (ESI)
+
m/z calcd for C H O [M+H] : 171.1021, found: 171.1017; Anal. calculated for C H O : C,
9
14
3
9
14
3
6
3.51; H, 8.29, found: C, 63.25; H, 8.41.
(E,4S,5S)-3-Hexylidene-dihydro-4-hydroxy-5-methylfuran-2(3H)-one (23). Compound 23
2
0
1
was obtained as waxy white solid (75%): [α] = -115.3 (c=2.4, CHCl ); H NMR (400 MHz,
D
3
CDCl ) δ 6.95 (t, J = 7.8, 1H), 4.8 (t, J = 5.2 Hz, 1H), 4.57-4.5 (m, 1H), 2.45-2.32 (m, 2H),
3
2
3
1
.2 (bs, 1H) 1.55-1.5 (m, 2H), 1.47 (d, J = 6.8 Hz, 3H), 1.39-1.29 (m, 4H), 0.91 (t, J = 6.8 Hz,
13
H); C NMR (100 MHz, CDCl ) δ 170.2, 147.8, 130.4, 78.9, 67.7, 31.4, 29.8, 28.1, 22.4,
3
-1
3.9; IR (neat): 3408, 2928, 1728, 1674, 1200, 1034, 984 cm ; HRMS (ESI) m/z calcd for
+
C H O [M+H] : 199.1334, found: 199.1341; Anal. calculated for C H O : C, 66.64; H,
11
18
3
11 18
3
9
.15, found: C, 66.95; H, 9.17.
(E,4S,5S)-Dihydro-4-hydroxy-5-methyl-3-octaylidenefuran-2(3H)-one (24). Compound
o
25
1
2
4 was obtained as a syrupy solid (66%): mp 42-44 C; [α] = -123.1 (c=0.092, CHCl ); H
D 3
NMR (400 MHz, CDCl ) δ 6.93-6.89 (dt, J = 0.8, 8.2 Hz, 1H), 4.87 (bs, 1H), 4.56-4.52 (m,
3
1
H), 2.54 (bs, 1H), 2.44-2.23 (m, 2H), 1.54-1.49 (m, 2H), 1.46 (d, J = 6.4 Hz, 3H), 1.35-1.27
1
3
(
bm, 8H), 0.88 (t, J = 6.8 Hz, 3H); C NMR (100 MHz, CDCl ) δ 170.4, 147.9, 130.4, 79.1,
3
6
9
7.6, 31.7, 29.8-29, 28.4, 22.6, 14, 13.9; IR (neat) 3391, 2925, 2852, 1722, 1676, 1205, 1037,
-1
+
82 cm ; HRMS (ESI) m/z calcd for C H O [M+H] : 227.1647, found: 227.1647; Anal.
1
3
22
3
calculated for C H O : C, 68.99; H, 9.80, found: C, 69.32; H, 9.91.
1
3
22
3
(E,4S,5S)-Dihydro-4-hydroxy-5-methyl-3-nonylidenefuran-2(3H)-one (25). Compound
o
25
2
5 was obtained as low melting solid (72%): mp 45-47 C; [α] D= -110.2 (c=0.104, CHCl3);
1
H NMR (400 MHz, CDCl ) δ 6.93-6.89 (dt, J = 1.2, 8.4 Hz, 1H), 4.82 (t, J = 4.8 Hz, 1H),
3
4
.57-4.5 (m, 1H), 2.45-2.34 (m, 2H), 1.82 (bs, 1H), 1.56-1.48 (m, 2H), 1.47 (d, J = 6.8 Hz,
1
3
3
1
H), 1.39-1.27 (bm, 10H), 0.88 (t, J = 6.8 Hz, 3H); C NMR (100 MHz, CDCl ) δ 170,
3
47.8, 130.4, 78.7, 67.7, 31.8, 29.8-29.1, 28.4, 22.6, 14, 13.9; IR (neat) 3381, 2921, 2844,

ACCEPTED MANUSCRIPT
-
1
+
1
2
6
722, 1688, 1456, 1208, 1039, 980 cm ; HRMS (ESI) m/z calcd for C H O [M+H] :
14 24 3
41.1804, found: 241.1804; Anal. calculated for C H O : C, 69.96; H, 10.07, found: C,
1
4
24
3
9.56; H, 9.95.
(E,4S,5S)-3-Decylidene-dihydro-4-hydroxy-5-methylfuran-2(3H)-one, (-)-licunolide B
o
25
(
26). Compound 26 was obtained as off-white solid (77%): mp 51-53 C; [α] D= -112.9
1
(
c=0.096, CHCl : H NMR (400 MHz, CDCl ) δ 6.94-6.89 (dt, J = 1.2, 8.4 Hz, 1H), 4.82 (t, J
3
3
=
5.2 Hz, 1H), 4.56-4.51 (m, 1H), 2.47-2.23 (m, 2H), 2.36 (bs, 1H), 1.53-1.49 (m, 2H), 1.46
1
3
(
d, J = 6.8 Hz, 3H), 1.33-1.26 (m, 12H), 0.88 (t, J = 7.2 Hz, 3H); C NMR (100 MHz, CDCl )
3
δ 170.3, 147.9, 130.4, 79, 67.6, 31.8, 29.8-29.2, 28.4, 22.6, 14.1, 13.9; IR (neat) 3382, 2921,
-1
+
2
848, 1726, 1694, 1457, 1203, 1028, 980 cm ; HRMS (ESI) m/z calcd for C H O [M+H] :
15 26 3
2
55.196, found: 255.196; Anal. calculated for C H O : C, 70.83; H, 10.30, found: C, 70.56;
15 26 3
H, 10.18.
(E,4S,5S)-Dihydro-4-hydroxy-5-methyl-3-undecylidenefuran-2(3H)-one (27). Compound
o
20
1
2
7 was obtained as off-white solid (70%): mp 58-60 C; [α] = -111.5 (c=0.072, CHCl ); H
D 3
NMR (400 MHz, CDCl ) δ 6.95-6.91 (dt, J = 8.0, 1.4 Hz, 1H), 4.81 (t, J = 5.4 Hz, 1H), 4.56-
3
4
.5 (m, 1H), 2.43-2.34 (m, 2H), 2.3 (d, J = 7.2 Hz, 1H), 1.53-1.48 (m, 2 H), 1.46 (d, J = 6.8
1
3
Hz, 3H), 1.37-1.26 (bs, 14H), 0.88 (t, J = 6.8 Hz, 3H); C NMR (100 MHz, CDCl ) δ 170.1,
3
1
1
2
7
47.8, 130.4, 78.8, 67.7, 31.9, 29.3-29.8, 28.4, 22.7, 14.1, 13.9; IR (neat) 3381, 2920, 2846,
-1
+
723, 1688, 1457, 1262, 1026, 982 cm ; HRMS (ESI) m/z calcd for C H O [M+H] :
16
28
3
69.2117, found: 269.2118; Anal. calculated for C H O : C, 71.60; H, 10.52, found: C,
1
6
28
3
1.89; H, 10.63.
(E,4S,5S)-3-Dodecylidene-dihydro-4-hydroxy-5-methylfuran-2(3H)-one, (-)-lincomolide
o
25
B (28). Compound 28 was obtained as off-white solid (68%): mp 63-65 C; [α] D= -90.2
1
(
c=0.084, CHCl : H NMR (400 MHz, CDCl ) δ 6.93-6.89 (dt, J = 0.8, 8.2 Hz, 1H), 4.87 (bs,
3
3
1
H), 4.56-4.52 (m, 1H), 2.45 (bs, 1H), 2.45-2.36 (m, 2H), 1.53-1.49 (m, 2H), 1.46 (d, J = 6.8
1
3
Hz, 3H), 1.26 (bs, 16H), 0.88 (t, J = 6.6 Hz, 3H); C NMR (100 MHz, CDCl ) δ 170.3, 147.9,
3
1
1
30.4, 79, 67.6, 31.9, 29.8-29.3, 28.4, 22.6, 14.1, 13.9; IR (neat) 3381, 2920, 2847, 1721,
-1
+
692, 1460, 1215, 1031, 983 cm ; HRMS (ESI) m/z calcd for C H O [M+H] : 283.2273,
1
7
30
3
found: 283.2273; Anal. calculated for C H O : C, 72.30; H, 10.71, found: C, 72.78; H,
1
7
30
3
1
1.84.

ACCEPTED MANUSCRIPT
(E,4S,5S)-Dihydro-4-hydroxy-5-methyl-3-tridecylidenefuran-2(3H)-one (29). Compound
o
25
1
2
9 was obtained as off-white solid (78%): mp 67-68 C; [α] = -109.7 (c=0.084, CHCl ); H
D 3
NMR (400 MHz, CDCl ) δ 6.93-6.89 (dt, J = 1.2, 8.4 Hz, 1H), 4.81 (bs, 1H), 4.56-4.5 (m,
3
1
1
7
1
2
H), 2.45-2.36 (m, 2H), 2.54 (d, J = 5.6 Hz, 1H), 1.55-1.48 (m, 2H), 1.46 (d, J = 6.8 Hz, 3H),
1
3
.25 (bs, 18H), 0.88 (t, J = 6.8 Hz, 3H); C NMR (100 MHz, CDCl ) δ 170.2, 147.8, 130.4,
3
8.9, 67.7, 31.9, 29.8-29.3, 28.4, 22.6, 14.1, 13.9; IR (neat) 3382, 2917, 2846, 1723, 1691,
-1
+
454, 1208, 1033, 987 cm ; HRMS (ESI) m/z calcd for C H O [M+H] : 297.243, found:
1
8
32
3
97.243; Anal. calculated for C H O : C, 72.93; H, 10.88, found: C, 72.74; H, 10.66.
1
8
32
3
(E,4S,5S)-3-Hexadecylidene-dihydro-4-hydroxy-5-methylfuran-2(3H)-one,
(-)-
isodihydromahubanolide B (30). Compound 30 was obtained as off-white solid (74%): mp
o
20
1
7
8-80 C; [α] = -79.8; H NMR (400 MHz, CDCl ) δ 6.98 (t, J = 7.8 Hz, 1H), 4.84 (t, J =
D 3
5
.6 Hz, 1H), 4.56-4.5 (m, 1H), 2.45-2.32 (m, 2H), 1.65 (bs, 1H) 1.54-1.48 (m, 2H), 1.47 (d,
1
3
J = 6.8 Hz, 3H), 1.25 (bs, 24H), 0.89 (t, J = 6.8 Hz, 3H); C NMR (100 MHz, CDCl ) δ
3
1
2
3
7
69.9, 147.8, 130.4, 78.7, 67.7, 31.9, 29.8-29.3, 28.4, 22.7, 14.1, 13.9; IR (neat): 3880, 2918,
-1
+
847, 1724, 1690, 1457, 1265, 1023, 987 cm ; HRMS (ESI) m/z calcd for C H O [M+H] :
21
38
3
39.2899, found: 339.2876; Anal. calculated for C H O : C, 74.51; H, 11.31, found: C,
2
1
38
3
4.38; H, 11.33.
(E,4S,5S)-Dihydro-4-hydroxy-5-methyl-3-octadecylidenefuran-2(3H)-one
(31).
o
20
Compound 31 was obtained as off-white solid (72%): mp 83-85 C; [α] D= -64.8 (c=0.08,
1
CHCl ); H NMR (400 MHz, CDCl ) δ 6.95-6.91 (dt, J = 8.0, 1.4 Hz, 1H), 4.81 (t, J = 5.8 Hz,
3
3
1
H), 4.56-4.50 (m, 1H), 2.45-2.34 (m, 2H), 2.0 (d, J = 6.8 Hz, 1H), 1.55-1.49 (m, 2H), 1.48
1
3
(
d, J = 5.6 Hz, 3H), 1.3-1.25 (bs, 28H), 0.88 (t, J = 6.8 Hz, 3H); C NMR (100 MHz, CDCl )
3
δ 170.1, 147.8, 130.4, 78.8, 67.7, 31.9, 29.3-29.8, 28.4, 22.7, 14.1, 13.9; IR (neat) 3381, 2913,
-
1
+
2
3
7
846, 1723, 1689, 1455, 1213, 1026, 987 cm ; HRMS (ESI) m/z calcd for C H O [M+H] :
23 42 3
67.3212, found: 367.3214; Anal. calculated for C H O : C, 75.36; H, 11.55, found: C,
2
3
42
3
5.09; H, 11.43.

ACCEPTED MANUSCRIPT
Author Contributions
The manuscript was written through contributions by all authors. All authors
approved the final version of manuscript.
Acknowledgement
We declare competing financial interest in the utility of marliolide derivatives. This
research was supported by the National Research Foundation of Korea (NRF) grant funded
by the Korea Ministry of Science (NRF-2017R1D1A1B03036116), the Bio & Medical
Technology Development Program of the National Research Foundation of Korea (NRF)
funded by the Ministry of Science, ICT & Future Planning (No. 2012M3A9C1053532), and
the GRRC program of Gyeonggi province (GRRC-DONGGUK2016-B04).
Keywords
HO-1, Heme oxygenase-1; NRF2, NF-E2-related factor 2; KEAP1, Kelch-like ECH-
associated protein 1; ARE, antioxidant response element; GCLC, glutamate-cysteine ligase
modifier catalytic unit; MEF, mouse embryonic fibroblast; DMBA, 12-
Dimethylbenz[a]anthracene; TPA, 12-O-Tetradecanoylphorbol-13-acetate; 8-OH-G, 8-
hydroxyguanosine; 4-HNE, 4-hydroxynonenal.
Appendix A. Supplementary data
Supplementary data related to this article can be found at.

ACCEPTED MANUSCRIPT
References
[
[
1] H. Sies, C. Berndt, D.P. Jones, Oxidative Stress, Annu Rev Biochem, 86 (2017) 715-748.
2] M. Schieber, N.S. Chandel, ROS function in redox signaling and oxidative stress, Curr
Biol, 24 (2014) R453-462.
3] T. Suzuki, M. Yamamoto, Stress-sensing mechanisms and the physiological roles of the
Keap1-Nrf2 system during cellular stress, J Biol Chem, 292 (2017) 16817-16824.
4] N.G. Abraham, J.M. Junge, G.S. Drummond, Translational Significance of Heme
Oxygenase in Obesity and Metabolic Syndrome, Trends Pharmacol Sci, 37 (2016) 17-36.
5] S.W. Ryter, J. Alam, A.M. Choi, Heme oxygenase-1/carbon monoxide: from basic
science to therapeutic applications, Physiol Rev, 86 (2006) 583-650.
6] T. Jansen, M. Hortmann, M. Oelze, B. Opitz, S. Steven, R. Schell, M. Knorr, S. Karbach,
[
[
[
[
S. Schuhmacher, P. Wenzel, T. Munzel, A. Daiber, Conversion of biliverdin to bilirubin by
biliverdin reductase contributes to endothelial cell protection by heme oxygenase-1-evidence
for direct and indirect antioxidant actions of bilirubin, J Mol Cell Cardiol, 49 (2010) 186-195.
[
7] R. Gozzelino, M.P. Soares, Coupling heme and iron metabolism via ferritin H chain,
Antioxid Redox Signal, 20 (2014) 1754-1769.
8] K.D. Poss, S. Tonegawa, Reduced stress defense in heme oxygenase 1-deficient cells,
Proc Natl Acad Sci U S A, 94 (1997) 10925-10930.
9] A. Yachie, Y. Niida, T. Wada, N. Igarashi, H. Kaneda, T. Toma, K. Ohta, Y. Kasahara, S.
Koizumi, Oxidative stress causes enhanced endothelial cell injury in human heme oxygenase-
deficiency, J Clin Invest, 103 (1999) 129-135.
10] Y. Zhao, Y.-W. Guo, W. Zhang, Rotundifolides A and B, Two New Enol-Derived
Butenolactones from the Bark of Litsea rotundifolia var. oblongifolia, Helv Chim Acta, 88
[
[
1
[
(2005) 349-353.
[
11] B.M. Claros, A.J. da Silva, M.L. Vasconcellos, A.P. de Brito, G.G. Leitao, Chemical
constituents of two Mollinedia species, Phytochemistry, 55 (2000) 859-862.
12] R. Motterlini, R. Foresti, Heme oxygenase-1 as a target for drug discovery, Antioxid
Redox Signal, 20 (2014) 1810-1826.
13] M.P. Soares, I. Marguti, A. Cunha, R. Larsen, Immunoregulatory effects of HO-1: how
does it work?, Curr Opin Pharmacol, 9 (2009) 482-489.
14] Y. Naito, T. Takagi, Y. Higashimura, Heme oxygenase-1 and anti-inflammatory M2
macrophages, Arch Biochem Biophys, 564 (2014) 83-88.
[
[
[

ACCEPTED MANUSCRIPT
[
15] M.L. Ferrandiz, I. Devesa, Inducers of heme oxygenase-1, Curr Pharm Des, 14 (2008)
4
73-486.
16] Q. Ma, X. He, Molecular basis of electrophilic and oxidative defense: promises and
perils of Nrf2, Pharmacol Rev, 64 (2012) 1055-1081.
17] K. Mailar, W.J. Choi, The first asymmetric synthesis of marliolide from readily
accessible carbohydrate as chiral template, Carbohydr Res, 432 (2016) 31-35.
18] S.I. Lee, J.H. Jang, G.S. Hwang, D.H. Ryu, Asymmetric synthesis of alpha-alkylidene-
[
[
[
beta-hydroxy-gamma-butyrolactones via enantioselective tandem Michael-aldol reaction, J
Org Chem, 78 (2013) 770-775.
[
19] S. Magesh, Y. Chen, L. Hu, Small molecule modulators of Keap1-Nrf2-ARE pathway as
potential preventive and therapeutic agents, Med Res Rev, 32 (2012) 687-726.
20] D.A. Abed, M. Goldstein, H. Albanyan, H. Jin, L. Hu, Discovery of direct inhibitors of
[
Keap1-Nrf2 protein-protein interaction as potential therapeutic and preventive agents, Acta
Pharm Sin B, 5 (2015) 285-299.
[
21] Y.S. Keum, B.Y. Choi, Molecular and chemical regulation of the Keap1-Nrf2 signaling
pathway, Molecules, 19 (2014) 10074-10089.
[
(
[
(
[
22] J.K. Kundu, Y.J. Surh, Inflammation: gearing the journey to cancer, Mutat Res, 659
2008) 15-30.
23] L.Y. Chau, Heme oxygenase-1: emerging target of cancer therapy, J Biomed Sci, 22
2015) 22.
24] V. Pittala, V. Sorrenti, G. Romeo, A. Acquaviva, C. Di Giacomo, L. Salerno,
Therapeutic Potential of Caffeic Acid Phenethyl Ester (CAPE) in Diabetes, Curr Med Chem,
(2016).
[
25] J.F. Ndisang, Synergistic Interaction Between Heme Oxygenase (HO) and Nuclear-
Factor E2- Related Factor-2 (Nrf2) against Oxidative Stress in Cardiovascular Related
Diseases, Curr Pharm Des, 23 (2017) 1465-1470.
[
26] Y. Cheng, J. Rong, Therapeutic Potential of Heme Oxygenase-1/carbon Monoxide
System Against Ischemia-Reperfusion Injury, Curr Pharm Des, 23 (2017) 3884-3898.
27] Y.L. Xie, J.G. Chu, X.M. Jian, J.Z. Dong, L.P. Wang, G.X. Li, N.B. Yang, Curcumin
[
attenuates lipopolysaccharide/d-galactosamine-induced acute liver injury by activating Nrf2
nuclear translocation and inhibiting NF-kB activation, Biomed Pharmacother, 91 (2017) 70-
7
7.
[
28] V. Pittala, L. Vanella, L. Salerno, C. Di Giacomo, R. Acquaviva, M. Raffaele, G. Romeo,

ACCEPTED MANUSCRIPT
M.N. Modica, O. Prezzavento, V. Sorrenti, Novel Caffeic Acid Phenethyl Ester (Cape)
Analogues as Inducers of Heme Oxygenase-1, Curr Pharm Des, 23 (2017) 2657-2664.
[
29] R.J. Fox, D.H. Miller, J.T. Phillips, M. Hutchinson, E. Havrdova, M. Kita, M. Yang, K.
Raghupathi, M. Novas, M.T. Sweetser, V. Viglietta, K.T. Dawson, C.S. Investigators,
Placebo-controlled phase 3 study of oral BG-12 or glatiramer in multiple sclerosis, The New
England journal of medicine, 367 (2012) 1087-1097.
[
30] W. Li, A.N. Kong, Molecular mechanisms of Nrf2-mediated antioxidant response, Mol
Carcinog, 48 (2009) 91-104.
31] E. Kansanen, S.M. Kuosmanen, H. Leinonen, A.L. Levonen, The Keap1-Nrf2 pathway:
Mechanisms of activation and dysregulation in cancer, Redox Biol, 1 (2013) 45-49.
32] H. Tanaka, Y. Takaya, J. Toyoda, T. Yasuda, M. Sato, J. Murata, H. Murata, K.
[
[
Kaburagi, O. Iida, K. Sugimura, E. Sakai, Two new butanolides from the roots of litsea
acuminate, Phytochemistry Letters, 11 (2015) 32-36.
[
33] I.L. Tsai, C.H. Hung, C.Y. Duh, J.H. Chen, W.Y. Lin, I.S. Chen, Cytotoxic butanolides
from the stem bark of Formosan Lindera communis, Planta Med, 67 (2001) 865-867.
34] V. Martinez, M. Yoshida, O. Gottlieb, ω-ethyl, ω-ethenyl and ω-ethynyl-α-lactones from
Clinostemon mahuba., Phytochemistry, 20 (1981) 459-464.
[