Synthesis, evaluation, and metabolism of novel [6]-shogaol derivatives as potent Nrf2 activators

Article abstract of DOI:10.1016/j.freeradbiomed.2016.03.026

Oxidative stress is a central component of many chronic diseases. The Kelch-like ECH-associated protein 1 (Keap1)-nuclear factor erythroid 2 p45-related factor 2 (Nrf2) system is a major regulatory pathway of cytoprotective genes against oxidative and electrophilic stress. Activation of the Nrf2 pathway plays crucial roles in the chemopreventive effects of various inducers. In this study, we developed a novel class of potent Nrf2 activators derived from ginger compound, [6]-shogaol (6S), using the Tg[glutathione S-transferase pi 1 (gstp1):green fluorescent protein (GFP)] transgenic zebrafish model. Investigation of structure-activity relationships of 6S derivatives indicates that the combination of an α,β-unsaturated carbonyl entity and a catechol moiety in one compound enhances the Tg(gstp1:GFP) fluorescence signal in zebrafish embryos. Chemical reaction and in vivo metabolism studies of the four most potent 6S derivatives showed that both α,β-unsaturated carbonyl entity and catechol moiety act as major active groups for conjugation with the sulfhydryl groups of the cysteine residues. In addition, we further demonstrated that 6S derivatives increased the expression of Nrf2 downstream target, heme oxygenase-1, in both a dose- and time-dependent manner. These results suggest that α,β-unsaturated carbonyl entity and catechol moiety of 6S derivatives may react with the cysteine residues of Keap1, disrupting the Keap1-Nrf2 complex, thereby liberating and activating Nrf2. Our findings of natural product-derived Nrf2 activators lead to design options of potent Nrf2 activators for further optimization.

Full text of DOI:10.1016/j.freeradbiomed.2016.03.026

Author’s Accepted Manuscript
Synthesis, Evaluation, and Metabolism of Novel
[6]-Shogaol Derivatives as Potent Nrf2 Activators
Yingdong Zhu, Pei Wang, Yantao Zhao, Chun
Yang, Anderson Clark, TinChung Leung, Xiaoxin
Chen, Shengmin Sang
www.elsevier.com
PII:
S0891-5849(16)30005-3
http://dx.doi.org/10.1016/j.freeradbiomed.2016.03.026
FRB12803
DOI:
Reference:
To appear in:
Free Radical Biology and Medicine
Received date: 12 January 2016
Revised date: 4 March 2016
Accepted date: 24 March 2016
Cite this article as: Yingdong Zhu, Pei Wang, Yantao Zhao, Chun Yang,
Anderson Clark, TinChung Leung, Xiaoxin Chen and Shengmin Sang, Synthesis,
Evaluation, and Metabolism of Novel [6]-Shogaol Derivatives as Potent Nrf2
A c t i v a t o r s , Free
Radical
Biology
and
Medicine,
http://dx.doi.org/10.1016/j.freeradbiomed.2016.03.026
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 galley proof before it is published in its final citable 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.

Synthesis, Evaluation, and Metabolism of Novel [6]-Shogaol Derivatives
as Potent Nrf2 Activators
Yingdong Zhu a, Pei Wang a, Yantao Zhao a, Chun Yang a, b, Anderson Clark c,
TinChung Leung c, Xiaoxin Chen d, and Shengmin Sang 1,*
a Laboratory for Functional Foods and Human Health, Center for Excellence in Post-
Harvest Technologies, North Carolina Agricultural and Technical State University, North
Carolina Research Campus, 500 Laureate Way, Kannapolis, NC 28081, USA
b Department of Colorectal Surgery, General Hospital of Ningxia Medical University,
Yinchuan 750004, P.R. China
c Nutrition Research Program, Julius L. Chambers Biomedical/Biotechnology Research
Institute, North Carolina Central University, North Carolina Research Campus, 500
Laureate Way, Kannapolis, NC 28081, USA
d Cancer Research Program, Julius L. Chambers Biomedical/Biotechnology Research
Institute, North Carolina Center University, 700 George Street, Durham, NC 27707, USA
* Corresponding author. Fax: 704-250-5709, E-mail: ssang@ncat.edu or shengminsang@yahoo.com
(S. Sang)
Abbreviations: ARE, antioxidant response element; br, broad; 6G, [6]-gingerol; GFP, green
fluorescent protein; gstp 1, glutathione S-transferase pi 1; hpf, hours postfertilization; Keap1, Kelch-like
ECH-associated protein 1; LC/MS, liquid chromatography-mass spectrometry; Nrf2, nuclear factor
erythroid 2 p45-related factor 2; 6S, [6]-shogaol; SFN, sulforaphane; TLC, thin-layer chromatography.
1

Running Title: [6]-Shogaol Derivatives as Nrf2 Activators
ABSTRACT:
Oxidative stress is a central component of many chronic diseases. The Kelch-like ECH-
associated protein 1 (Keap1)-nuclear factor erythroid 2 p45-related factor 2 (Nrf2) system
is a major regulatory pathway of cytoprotective genes against oxidative and electrophilic
stress. Activation of the Nrf2 pathway plays crucial roles in the chemopreventive effects
of various inducers. In this study, we developed a novel class of potent Nrf2 activators
derived from ginger compound, [6]-shogaol (6S), using the Tg[glutathione S-transferase
pi 1 (gstp1):green fluorescent protein (GFP)] transgenic zebrafish model. Investigation
of structure-activity relationships of 6S derivatives indicates that the combination of an
α,β-unsaturated carbonyl entity and a catechol moiety in one compound enhances the
Tg(gstp1:GFP) fluorescence signal in zebrafish embryos. Chemical reaction and in vivo
metabolism studies of the four most potent 6S derivatives showed that both ,-
unsaturated carbonyl entity and catechol moiety act as major active groups for
conjugation with the sulfhydryl groups of the cysteine residues. In addition, we further
demonstrated that 6S derivatives increased the expression of Nrf2 downstream target,
heme oxygenase-1, in both a dose- and time-dependent manner. These results suggest
that ,-unsaturated carbonyl entity and catechol moiety of 6S derivatives may react with
the cysteine residues of Keap1, disrupting the Keap1-Nrf2 complex, thereby liberating
and activating Nrf2. Our findings of natural product-derived Nrf2 activators lead to
design options of potent Nrf2 activators for further optimization.
Keywords:
2

[6]-Shogaol derivatives; Metabolism; Nrf2 activators; Tg(gstp1:GFP) transgenic
zebrafish model
INTRODUCTION
Nuclear factor erythroid-2 related factor 2 (Nrf2), a transcription factor and basic
leucine zipper protein, regulates the expression of cytoprotective molecules that
counteract against oxidative and electrophilic stress.1 Under basal conditions, Nrf2 is
constantly targeted for Kelch-like ECH-associated protein 1 (Keap1)-mediated
ubiquitination and subsequent proteasomal degradation to maintain a low level of Nrf2
protein in the cells. Upon activation in response to oxidative and electrophilic stress, Nrf2
detaches from Keap1, migrates to the nucleus, and binds to the antioxidant response
element (ARE) sequence to activate transcription of cytoprotective and detoxifying
genes.2-4 The association of Keap1 with Nrf2 mainly relies on the cysteine residues in the
protein Keap1, and the sulfhydryl groups of the cysteine residues are probably the direct
sensors of inducers of the phase II system.5, 6 Therefore, Nrf2 is a potential target for
many chronic diseases, such as chronic kidney disease, asthma, neurodegenerative
diseases, and cancer.7-13 It is known that small molecules Nrf2 activators, such as
sulforaphane (SFN),14 curcumin,15 and chalcone derivatives,16 have been identified as
cancer chemopreventive agents. Screening for Nrf2 activators would lead to the
discovery of new pharmaceuticals for oxidative stress-associated chronic diseases.
The Keap1-Nrf2 system is conserved among vertebrates, including zebrafish.17-19
Zebrafish Nrf2 protein shares six highly conserved Neh domains with mammalian Nrf2
proteins, which are considered to play critical functions in Nrf2 regulation.20 High
sequence identities in each Neh domain between zebrafish and mouse Nrf2 imply a
3

common regulatory mechanism among vertebrate Nrf2s.17 This finding justifies zebrafish
as a powerful tool to analyze the molecular basis of the Nrf2-Keap1 system. Among the
known endogenous targets of zebrafish Nrf2, the pi-class glutathione S-transferase 1 gene
(gstp1) showed the strongest induction in both electrophile-treated and Nrf2-
overexpressing embryos, and is regarded as an important determinant for the protection
against insults of various electrophilic agents.21 This gene has also been observed in
mammals and humans.22, 23 A green fluorescent protein (GFP) reporter gene driven by the
gstp1 promoter was created in the Tg(gstp1:GFP) transgenic zebrafish line by Dr.
Kobayashi's group; the promoter containing an ARE-like sequence located 50 bp
upstream of the transcription initiation site was shown to be essential and sufficient for
Nrf2 transactivation.24 Obviously, Tg(gstp1:GFP) transgenic zebrafish provides a new
and attractive platform to screen novel Nrf2 activators in vivo for drug discovery. A
recent study has successfully identified nitro-fatty acids and cyclopentenone
prostaglandins as potent Nrf2 activators using the Tg(gstp1: GFP) transgenic zebrafish
model.25
In a recent study, we identified [6]-shogaol (6S), a major component of dry ginger, as
an activator of Nrf2 in colon epithelial cells in vitro and in vivo.26 Our data obtained with
gene expression profiling, Western blotting and immunostaining clearly demonstrated
that 6S induced nuclear translocation of Nrf2 and activated Nrf2 target genes in an Nrf2-
dependent manner.27 With an α,β-unsaturated carbonyl group in the alkyl tail, 6S is a
typical Michael acceptor. We hypothesized that 6S could activate Nrf2 via alkylation of
cysteine residues of Keap1 protein. This was confirmed by our observations that 6S
modified Keap1 at 17 cysteine residues and these cysteine residues were located in all 5
4

domains of Keap1 protein. Keap1 is a cysteine-rich protein possessing 27 cysteine
residues in the human protein. Alkylation of one or more of the cysteine residues of
Keap1 by xenobiotic electrophiles appears to be an important signaling mechanism for
the regulation of ARE activity through Nrf2.28 These findings prompt us to synthesize a
series of 6S derivatives and conduct a structure-activity relationship (SAR) study of 6S
derivatives responsible for Nrf2 activation using the gstp1 reporter transgenic zebrafish
model. We designed and synthesized a panel of 6S derivatives (Schemes 1 and 2) to
probe their SAR and Nrf2 activation on transgenic zebrafish embryos. We further
investigated the formation of cysteine conjugates of representative 6S derivatives in mice
and in zebrafish embryos to understand their underlying molecular mechanisms.
MATERIALS AND METHODS
Materials. Anhydrous chemical reactions were carried out in oven-dried glassware
under a nitrogen atmosphere unless otherwise noted. Reactions were monitored by
analytical thin-layer chromatography (TLC) on 250 μm silica gel plates (GF254) (Merck)
and visualized under UV light. The products were isolated and purified by either
preparative TLC on 1000 μm silica gel plates (GF254) (Sorbent Technologies, catalog no.
1617124) or column chromatography (CC) using silica gel (Sorbent Technologies,
1
catalog no. 3093M-25). H, 13C NMR, and two-dimensional (2-D) NMR spectra were
recorded on a Bruker AVANCE 400 MHz or 600 MHz spectrometer (Bruker, Inc.,
Silberstreifen, Rheinstetten, Germany) using TMS as an internal standard. Chemical
shifts (δ) are expressed in ppm. Coupling constants (J) are expressed in Hz, and
multiplicities are indicated by s (singlet), d (doublet), t (triplet), q (quartet), and br
5

13
(broad). The C NMR spectra are proton decoupled. Shogaols (1–3) and gingerols (11–
13) used in the present study were purified from ginger extract.29 6S metabolites, 4–10,
14–17 and 25 were obtained from [6]-shogaol as described in our previous study.30 Other
chemicals were purchased from Sigma-Aldrich (St. Louis, MO) and were used without
further purification. All compounds used were > 95% pure. Fetal bovine serum and
penicillin/streptomycin were purchased from Gemini Bio-Products (West Sacramento,
CA). Human normal colonic epithelial cells (CCD 841 CoN, ATCC® CRL-170TM)
were obtained from ATCC (Manassas, VA).
6

N
N
N
F3C
HO
CHO
O
F3C
HO
HO
ii)
N
i)
iii)
O
29
CF3
O
28
CF3
R1
30
31c
O
R1
OH O
R2
R3
CHO
R2
R3
(CH2)4
iv)
R4
R4
R1
R2
R3
R4
R1
R2
R3
R4
33a H
33b H
33c H
33d H
33e H
33f
33g CF3
H
OAc OCH3
32a H
32b H
32c H
32d H
32e H
H
OAc OCH3
OAc OAc OCH3
CF3 OAc OCH3
H
CF3
CF3
H
OAc OAc OCH3
CF3 OAc OCH3
H
CF3
CF3
H
OAc OAc
OAc OAc
H
H
H
H
H
H
H
H
H
H
H
H
F
32f
F
32g CF3
O
O
R1
R4
R1
HO
R2
R3
vi)
v)
(CH2)4
(CH2)4
R2
R1
R2
H
R3
R4
R1
18 H
R2
OCH3
34a H
34b H
34c H
34d H
22
23
OAc OCH3
OAc OAc OCH3
CF3 OAc OCH3
H
CF3
CF3
H
20 OH OCH3
21 CF3 OCH3
OAc OAc
26 H
OH
H
F
H
H
H
H
H
H
24 CF3
Scheme 1. Synthesis of aldehyde (31c) and [6]-dehydroshogaol derivatives (18, 20-24,
o
and 26). Regents and conditions: i) CH3I, K2CO3, acetone, 0 C–rt, 18 h, yield 100%; ii)
n-BuLi, THF, -78 oC-rt, 25 min; B(OMe)3, -78 oC–rt, 18 h; NH3 in MeOH; and then H2O2,
o
0 C–rt, 2 h, yield 50%; iii) hexamethylenetetramine, TFA, reflux, 3 h, yield 20%; iv)
o
trans-3-nonen-2-one, n-BuLi, DIPA, THF, -78 C, 35 min, yield 34–90%; v) PTSA,
toluene, reflux, 5 min, yield 36–86%; vi) LiOH, THF/MeOH/H2O, rt, 5 min, yield 59–
83%.
7

Synthesis of [6]-Shogaol Derivatives. Aryl-1,4-dien-3-ones, typical Michael
acceptors, could be synthesized by Horner-Emmons olefination of 2-oxo-3-
alkenylphosphonates with benzaldehydes in the presence of lithium bromide and
triethylamine,31 or by conjugate addition of vinyl iodide to styryl-activated enones under
n-BuLi.32, 33 In light of several disadvantages of these procedures including high reactive
species required, and the use of expensive or noncommercial available reagents, we
herein developed a straightforward and simple strategy to synthesize aryl-dec-1,4-dien-3-
ones (6S derivatives), using commercial aromatic aldehydes and low reactive species 3-
nonen-2-one via nucleophilic addition and subsequent dehydration. In short, the
appropriate substituted aromatic aldehydes (32a-g) were treated with lithium species of
3-nonen-2-one to form -hydroxyenones (33a-g), and then a rapid PTSA-catalyzed
dehydration was followed, furnishing the target [6]-dehydroshogaol derivatives (18, 20-
24 and 26) (Scheme 1).
R1
O
R1
O
R1 OH
O
CHO
(CH2)4
(CH2)4
ii)
(CH2)4
i)
R2
R2
R2
R3
R3
34h H
34i
R1 R2 R3
R3
R1
R1
R2
R3
R2
R3
32a
H
OAc OCH3
OAc OCH3
OAc OAc
33h
33i
H
H
OAc OCH3
OAc OAc
32d H OAc OAc
H
O
iii)
HO
R
19
27
R = OCH3
R = OH
Scheme 2. Synthesis of [6]-dehydroparadol derivatives (19 and 27). Regents and
conditions: i) 2-nonanone, n-BuLi, DIPA, THF, -78 oC, 35 min, yield 32–85%; ii) PTSA,
8

toluene, reflux, 5 min, yield 41–68%; iii) LiOH, THF/MeOH/H2O, rt, 5 min, yield 79–
80%.
Likewise, [6]-dehydroparadol derivatives (19 and 27) were successfully prepared
using aromatic aldehydes (32a and 32d) and low reactive lithium species of 2-nonanone
(Scheme 2). Non-commercially available aldehyde 31c was prepared from 3-
(trifluoromethyl)phenol following the steps as reported in the literature.34 Cysteinyl
conjugates 35 and 36 were readily made under Michael addition condition according to
the method for the preparation of cysteine conjugate of 6S (Scheme 3A).30 Conjugate 37
was successfully synthesized through IBX-mediated oxidation to the o-quinone followed
by nucleophilic addition of L-cystein (Scheme 3A).35 The synthesis details and
spectroscopic data for these compounds are given in the Supplemental Information and
1
13
their structures and purity were confirmed by H and C NMR spectroscopy, and LC-
ESI/MS. All final products are > 95% pure.
9

O
Cys
O
A)
i)
i)
35
18
HO
HO
HO
O
O
Cys =
H2N
O
Cys O
O
OH
27
36
S
HO
HO
OH
OH
OH
OH
O
O
Cys
HO
ii)
37
25
B)
O
O
O
pathway A
H2N
H2N
H2N
OH
OH
OH
H
H
S
O
O
S
O
S
1
5
3
5
5
(CH2)4
(CH2)4
(CH2)4
1'
1'
4'
HO
HO
35
3'
18
O
O
NH2
pathway B
O
O
H2N
O
H
OH
5'
Cys
HO
5'
3
(CH2)6
3
1'
(CH2)6
1'
S
IBX
5'
4'
5'
S
OH
4'
O
25
37
HO
3'
3'
OH
4'
3'
O
H
O
OH
3'
O
OH
Scheme 3. Synthesis of cysteine conjugates of [6]-shogaol derivatives (35-37) (A) and
their proposed reaction mechanism (B). Regents and conditions: i) L-cystein, NaHCO3
(cat.), MeOH/PBS (5:1), rt, 10 min, yield 14-39%; ii) L-cysteine, IBX, MeOH/PBS, -78
oC-rt, 1.5 h, yield 78%.
LC/MS Analysis. LC/MS analysis was carried out with a Thermo-Finnigan Spectra
System, which consisted of an Accela high speed MS pump, an Accela refrigerated
autosampler, and an LCQ Fleet ion trap mass detector (Thermo Electron, San Jose, CA,
USA) incorporated with an electrospray ionization (ESI) interface. A Gemini-NX C18
10

column (150 mm × 4.6 mm i.d., 5 μm, Phenomenex) was used for separation at a flow
rate of 0.3 mL/min. The column was eluted with 100% A (5% aqueous methanol with 0.2%
formic acid) for 1 min, followed by linear increases in B (95% aqueous methanol with
0.2% formic acid) to 55% from 1 to 4 min, to 85% from 4 to 25 min, then to 100% B
from 25 to 30 min, and then with 100% B from 30 to 35 min. The column was then re-
equilibrated with 100% A for 5 min. The LC eluent was introduced into the ESI interface.
The positive ion polarity mode was set for the ESI source with the voltage on the ESI
interface maintained at approximately 4.6 kV. Nitrogen gas was used as the sheath gas
o
and auxiliary gas. Optimized source parameters, including capillary temperature (260 C),
sheath gas flow rate (31 arbitrary units), auxiliary gas flow rate (16 units), tube lens (34
V), and capillary voltage 9 V, were tuned using authentic samples. The collision-induced
dissociation (CID) was conducted with an isolation width 2 Da and normalized collision
energy of 35 for MS2. The mass range was measured from 100 to 800 m/z. Data
acquisition was performed with Xcalibur version 2.0 (Thermo Electron, San Jose, CA,
USA).
Transgenic Zebrafish and Embryos. Zebrafish AB transgenic strains Tg(gstp1:GFP)
were maintained in a Pentair Aquatic Ecosystem fish housing unit with 14 h light/10 h
dark cycle.25 The fish embryos were maintained at 28.5 °C in 0.3X Danieau’s solution
(19.3 mM NaCl, 0.23 mM KCl, 0.13 mM MgSO4, 0.2 mM Ca(NO3)2, 1.7 mM HEPES,
pH 7.0) containing 30 g/ml phenylthiourea (PTU) to inhibit pigmentation.
Tg(gstp1:GFP) cross with AB wild-type strain fluorescence embryos were used for the
experiment. Zebrafish embryos were washed, dechorionated and anaesthetized before
11

observations, and fluorescence imaging for analysis. All zebrafish experiments were
approved by IACUC committee at the North Carolina Central University.
Chemical Treatments. Transgenic zebrafish embryos at 1 day-post-fertilization (dpf)
were dechorionated and placed in Petri dishes containing different chemicals diluted in
0.3X Danieau’s solution containing PTU. The chemical stock was dissolved in DMSO at
5 mM concentration. The control contains DMSO at the corresponding concentration of
0.1% which showed no effect on embryonic development and no effect on Tg(gstp1:GFP)
fluorescence activity.
Fluorescent Imaging. An Olympus MVX10 Fluorescence Macroscope (Olympus,
Center Valley, PA) equipped with a Hamamatsu C9300-221 high-speed digital CCD
camera (Hamamatsu City, Japan) was used for fluorescence microscopy. Tg(gstp1:GFP)
fluorescent embryos were anaesthetized in tricaine and imaged at 2 dpf (day-post-
fertilization). MetaMorph Basic software (Olympus, Center Valley, PA) was used for
image acquisition and analysis.
Quantification of Nrf2 Reporter Activity by Tg(gstp1: GFP) Fluoresence in
Zebrafish Model. The fluorescence intensity of the olfactory neural epithelia is
quantitated using MetaMorph Basic software. In brief, GFP expression in the olfactory
regions of Tg(gstp1:GFP) embryos treated with various chemicals was measured after
24-hour treatment from 1 dpf to 2 dpf. The GFP induction was detectable by fluorescence
signal at the area of the olfactory epithelia in the anterior of the head region, the intensity
was measured using an area of a circle of 60 pixel diameter of the fluorescence olfactory
epithelia subtracted by a background non-fluorescence area next to the zebrafish embryo.
12

Both the left and right olfactory epithelia were measured to give an average fluorescence
value. The values are an average of measurements from at least 10 embryos.
Measurement of reactive oxygen species (ROS). Zebrafish embryos (20/group) at
1dpf were dechorionated, and treated with 5 µM compounds 1, 18, 26, and 27 over 2, 4, 8,
and 24 h. The embryos at each time point were then used for ROS extraction and analysis.
The ROS assay employed the cell-permeable fluorogenic probe 2, 7-
dichlorodihydrofluorescindiacetate [DCFH-DA] (Sigma Aldrich, St. Louis, MO) to
measure the relative changes in O-2 and H2O2 levels in Zebrafish embryos after treatment.
All the Zebrafish embryos were homogenized in 400 μL ice-cooled PBS, and 100 μL
homogenate was added into the well of 96-well plate in triplicate (PBS was run as
background). Each well was added 5 μL 20 mM DCFH-DA stock freshly prepared, and
then the plate was incubated at 37 °C for 15, 30, and 60 min. The plate was immediately
placed in a Biotek microplate reader to measure fluorescence at wavelengths of 485
(excitation) and 528 (emission) at the given incubation time points. The protein
concentrations of homogenate were determined using Bradford method. The corrected
fluorescent values were calculated as the increase rate per minute per mg protein and
normalized to respective control for each time point and are presented as fold induction
(n = 3).
Western Blotting. Human normal colonic epithelial cells (CCD 841 CoN) were
grown in Eagle’s Minimum Essential Medium with 10% fetal bovine serum and 1%
penicillin/streptomycin, and maintained at 37 oC in a 100% humidified atmosphere of 5%
CO2 and 95% air. Fresh growth medium was added to the cells every two days until
13

confluent. Cells were planted in 145 × 20 mm flat-bottomed tissue culture dishes and
growth to 70-80% confluence and then treated with compound 27 for various doses and
time points. At the end of incubation period, cell lysates were prepared in ice-cold Cell
Lysis Buffer (Cell Signaling, Danvers, MA) with 1% protease inhibitor cocktail and 1%
phenylmethylsulfonyl fluoride (Sigma, St. Louis, MO). Protein concentrations were
measured using BCA Protein Assay Kit (Thermo Scientific, Rockford, IL, USA).
Aliquots containing 30 μg protein were loaded onto a 10-12% sodium dodecyl
sulfatepolyacrylamide gel, transblotted onto polyvinylidene difluoride (PVDF) membrane
(Bio-Rad Laboratories, Beverly, CA), blocked with Tris buffered saline with 1% Casein
with 0.1% Tween-20, and then incubated with each of the primary antibodies of HO-1
o
and β-actin overnight at 4 C (Cell Signaling, Beverly, MA). The membrane was then
incubated with horseradish peroxidase-conjugated donkey anti-rabbit IgG (Cell signaling,
Danvers, MA). The bound complexes were detected with SuperSignal Chemiluminescent
Substract (Thermo Scientific, Rockford, IL). The immunoblot bands were quantified by
densitometry analysis, and the ratio to β-actin was calculated and presented.
Biotransformation of Xenobiotics 18 and 25-27 in Zebrafish Embryos. Zebrafish
embryos were staged and maintained according to NCCU IACUC guidelines. In short,
fifty zebrafish embryos at the 8 hour-post-fertilization (hpf) stage were incubated at 28.5
oC in 0.3X Danieau’s solution (19.3 mM NaCl, 0.23 mM KCl, 0.13 mM MgSO4, 0.2 mM
Ca(NO3)2, 1.7 mM HEPES, pH 7.0) with or without 5 M 18 and 25-27, respectively. At
24 hpf, embryos were dechorionated manually and the chorions were carefully removed
one by one from the culture medium. The zebrafish embryos were harvested and kept at -
80 oC till analyze by LC/MS.
14

Zebrafish Embryo Sample Preparation. 200 L Sodium acetate buffer solution
(pH 5.0) was added to 50 zebrafish embryos. Samples were homogenized for 90 s by an
Omni Bead Ruptor Homogenizer (Kennesaw, GA). 10 L 10% ascorbic acid and 300 L
sodium acetate buffer were added to the homogenates. The mixture was incubated in the
o
presence of β-glucuronidase (250 U) and sulfatase (3 U) at 37 C for 45 min, and then
600 L MeOH with 1% acetic acid was added. The resulting mixture was vortexed for 30
s and then centrifuged at 17000g for 10 min. The supernatant was removed and
evaporated under a gentle stream of nitrogen. The residue was reconstituted in 150 L 90%
MeOH with 0.2% AA, and 10 L was analyzed directly by LC/MS.
Mouse Study. The Institutional Review Board approved the protocol for mouse study
through the Animal Care and Facilities Committee at North Carolina Central University.
Female C57BL/6J mice were purchased from the Jackson Laboratory (Bar Harbor, ME,
USA) and allowed to acclimate for at least 1 week prior to the start of the experiment.
The mice were housed 5 per cage and maintained in air-conditioned quarters with a room
temperature of 20  2 ºC, relative humidity of 50  10 %, and an alternating 12-h
light/dark cycle. Mice were fed Rodent Chow #5001 (LabDiet) and water, and were
allowed to eat and drink ad libitum. Synthetic compounds 18 and 27 in DMSO were
administered to mice by oral gavage (200 mg/kg), respectively. Urine samples were
collected in metabolic cages (5 mice per cage) in 24 h after administration. The samples
were stored at -80 ºC until analysis.
Mouse Urine Sample Preparation. Enzymatic deconjugation of mouse urine was
performed as described previously with slight modifications.36 In brief, triplicate samples
were prepared in the presence of β-glucuronidase (250 U) and sulfatase (3 U) for 1.0 h at
15

37 oC. After incubation, 500 L methanol containing 0.2% AA was added to the medium.
The resulting suspension was centrifuged at 17,000 g for 5 min, and 10 L of supernatant
was analyzed directly by LC/MS.
Statistical analysis. All results are presented as means ± standard deviation. An
unpaired t test was used to determine potential differences between each treatment and
control. Comparisons between all treatments and control were measured by one way
ANOVA with Dunnett’s test using the GraphPad Prism version 5.04. A p value of less
than 0.05 was considered statistically significant in all tests.
RESULTS AND DISCUSSION
Potency of 6S Derivatives to Activate Nrf2 in Transgenic Zebrafish Embryos.
We have previously identified Nrf2 as a molecular target of ginger compound 6S in colon
epithelial cells in vitro and in vivo.27 In the current study, we found that 6S (1) is even
more active than SFN (2.40 vs. 2.03 times higher than vehicle) (Figure 1), a well-known
38
potent activator of Nrf2.37,
Using quantitative microscopy, we showed GFP
fluorescence intensity at the olfactory sensory neural epithelia of 6S-treated zebrafish
embryos is stronger than those of SFN-treated individuals (Figure 2). We hypothesized
that the ,-unsaturated carbonyl entity in the side chain of 6S plays a pivotal role for
Nrf2 activation.
To test our hypothesis, we compared the activity of 6S with its derivatives with the
conversion of ketone group into hydroxyl group (4), the reduction of olefinic double bond
(5), or both (14), as well as their related derivatives (6-10 and 15-17). Our results clearly
indicate that the departure of the ,-unsaturated carbonyl entity attenuated Nrf2-induced
activities (Figure 1). 6S demonstrated stronger activity than compounds 4 (1.13 times
16

higher than vehicle) and 5 (1.27 times higher than vehicle), and compound 14 had no
effect at all (Figure 1). Furthermore, thiol conjugation (6-9) also diminished the activity
of 6S. 6G (11), the hydrated precursor of 6S, was also less potent than 6S (1.87 vs. 2.40
times higher than vehicle). We further determined how the side chain length affected the
activity of 6S by comparing the effects of 6-, 8-, and 10S (1-3) in Tg(gstp1: GFP)
transgenic zebrafish embryos. Our results indicate that their activities are in the order of
6S > 8S > 10S (2.40 vs. 1.67 vs. 1.53) (Figure 1). A similar result was obtained for 6-, 8-,
and 10G (11-13) (1.87 vs. 1.65 vs. 1.29) (Figure 1). This suggests that the lipophilicity of
the alkyl tails in 6S derivatives diminishes the potency. Taken together, our results
supported that a central core consisting of an ,-unsaturated carbonyl entity and a ten-
carbon units of alkyl tail in 6S derivatives would exert a potent Nrf2 activation and could
serve as core structures to further develop novel Nrf2 activators.
17

R2 R3
R
R1
R4
O
R2
R3
(CH2)m
(CH2)m
(CH2)4
HO
HO
R1 R2
R3 R4
OH OCH3
R1
R2 R3
m
4
4
4
4
4
4
4
6
8
4
4
4
4
1, R = O, m = 4
2, R = O, m = 6
3, R = O, m = 8
4, R = OH, m = 4
18, H
H
R1
5, OCH3
6, OCH3
7, OCH3
8, OCH3
9, OCH3
10, OCH3
11, OCH3
12, OCH3
13, OCH3
O
O
O
O
O
O
O
O
O
H
O
20, H OH OH OCH3
21, H CF3 OH OCH3
22, H CF3
Cys
Nac
SCH3
GSH
OCH3
OH
H
H
H
H
H
H
23, F
CF3
H
H
R1
O
24, CF3
26, H
OH OH
(CH2)4
O
OH
OH
R2
(CH2)4
14, OCH3 OH
H
R3
15, OCH3 OH Cys
16, OCH3 OH Nac
17, OCH3 OH SCH3
R1 R2 R3
19, H OH OCH3
27, H OH OH
HO
25
OH
Compound ID
Ratio/CTL
1.00
Compound ID
Ration/CTL
0.96 ± 0.21
1.08 ± 0.65
0.78 ± 0.12
0.72 ± 0.14
5.67 ± 0.59
3.29 ± 0.35
1.26 ± 0.14
2.18 ± 0.53
1.28 ± 0.10
1.63 ± 0.17
1.55 ± 0.67
4.32 ± 0.50
6.96 ± 0.48
9.19 ± 0.12
DMSO
14
15
16
17
18
19
20
21
22
23
24
25
26
27
SFN
1
2.03 ± 0.49
2.40 ± 0.54
1.67 ± 0.55
1.53 ± 0.53
1.13 ± 0.21
1.27 ± 0.29
1.06 ± 0.20
1.04 ± 0.05
1.16 ± 0.47
0.93 ± 0.21
0.96 ± 0.24
1.87 ± 0.51
1.65 ± 0.29
1.29 ± 0.40
2
3
4
5
6
7
8
9
10
11
12
13
Figure 1. Structures of compounds 1-27 and their effects on Tg(gstp1:GFP) fluorescence
signal in the Tg(gstp1:GFP) transgenic zebrafish embryos at a concentration of 5 M.
The induced activities are indicated by the ratios of GFP fluorescence intensities in 6S
derivatives treatment to control. The values are expressed as the mean ± SD (n = 3-7).
CLT, DMSO control; GFP, green fluorescent protein; SFN, sulforaphane.
18

Figure 2. Tg(gstp1:GFP) fluorescence signal at the olfactory sensory neural epithelia of
zebrafish embryos at 2 day after 24 hr treatment with compounds 1, 11, 18, and 25-27, at
a concentration of 5 M. Top left is the bright field image of control (CTL) at the ventral
view of head region, the rest are fluorescent images of GFP signal. The inducible
Tg(gstp1:GFP) activity at the olfactory sensory neural epithelia was indicated by the two
rectangles. GFP, green fluorescent protein; 6G, [6]-gingerol; 6S, [6]-shogaol; SFN,
sulforaphane.
Recent studies have revealed that conjugated enones such as chalcones can stimulate
expression of Nrf2-dependent genes such as HO-1, NQO-1, and GCLM, and are potent
19

Nrf2 activators in vitro and in vivo.16, 21, 39 We therefore designed and synthesized
conjugated 6S derivatives 18 and 19 (Schemes 1 and 2). We observed that both
compounds 18, with a 1-aryl-1,4-dien-3-one entity in the structure, and 19, with a 1-aryl-
1-en-3-one entity in the structure, were more potent than 6S (5.67 vs. 2.40 and 3.29 vs.
2.40), suggesting that conjugated unsaturated ketones in the side chains of 6S derivatives
greatly favor the activation of Nrf2.
Aside from the chemical modifications on the side chain, the electronic properties of
the substituents on the aromatic rings of 6S derivatives were investigated. Firstly, we
observed that an extra electron-donating group (-OH) at the C-5' position of the aromatic
ring in 18, corresponding to 20 (Scheme 1), diminished the activity of 18 (1.26 vs. 5.67)
(Figure 1). Similarly, the introduction of an electron-withdrawing group (-CF3) into the
C-5' position of the ring in 18, corresponding to 21 (Scheme 1), also weakened the
activity of 18, but had better activity than 20 (2.18 vs. 1.26). We also observed that the
removal of the hydroxyl and methoxyl groups from 21, corresponding to 22, further
decreased the efficacy of 21 (1.28 vs. 2.18). Adding another electron-withdrawing group
(-F) into the C-6' position of the ring in 22, corresponding to 23 (Scheme 1), or changing
the substitution position of -CF3, corresponding to 24 (Scheme 1), only slightly increased
the activity of 22 (1.63 vs. 1.28 and 1.55 vs. 1.28) (Figure 1).
Interestingly, we noticed that derivative 25, a metabolite of 6S,40 containing a
catechol moiety and an isolated ketone group in the structure, exerted a comparable
potency (4.32 times higher than vehicle) to conjugated enone 18 (5.67 times higher than
vehicle) (Figures 1 and 2). This indicates that a catechol moiety in the structure is as
essential as an ,-unsaturated carbonyl entity for Nrf2 activation. Inspired by the
20

observation from catechol 25, we combined a catechol moiety and an unsaturated
carbonyl entity in a molecule, eventually resulting in the synthesis of complexes of
catechol and conjugated enone 26 and 27 (Schemes 1 and 2). In the end, we observed that
both 26 and 27 showed enhanced Nrf2-induced activity (6.96 times higher than vehicle
for 26; and 9.19-fold higher than vehicle for 27), and are more active than conjugated
enone 18 (5.67 times higher than vehicle), 19 (3.29 times higher than vehicle), and
catechol 25 (4.32 times higher than vehicle) (Figures 1 and 2). Interestingly, the activity
of 27 is higher than the sum of 25 (4.32 times higher than vehicle) and 19 (3.29 times
higher than vehicle), and the activity of 26 (6.96 times higher than vehicle) is also higher
than 25 and 18. These findings suggest that the combination of a catechol moiety and a
conjugated enone improves the Nrf2-induced activity.
Taken together, we identified four 6S derivatives 18 and 25-27 as potent Nrf2
activators in transgenic zebrafish embryos for the first time. Investigation on SAR of 6S
derivatives demonstrated that 1) an ,-unsaturated carbonyl entity in the alkyl tail is
crucial for Nrf2 activation; 2) a conjugated unsaturated ketone in the structure enhances
the activity; 3) a catechol moiety in the structure plays a pivotal role in Nrf2-induced
activities; and 4) coexistence of a catechol moiety and a conjugated unsaturated ketone in
one molecule improves the activity.
Effects of 6S and Its Derivatives (18, 26, and 27) on Cellular ROS Levels. In order
to determine whether ROS plays a role in the activation of Nrf2 pathway, we measured
the cellular ROS levels in zebrafish embryos treated with test agents at 2, 4, 8, and 24 h.
Our results clearly indicated that these compounds did not induce oxidative stress (Figure
21

3), suggesting that ROS production is not one of the mechanism of actions of these
derivatives.
6S Derivative (27) Increased HO-1 Expression in CCD 841 CoN Cells in Both a
Dose- and Time-Dependent manner. To assess the role of 6S derivatives on the
induction of Nrf2 downstream genes, human normal colonic epithelial cells CCD 841
CoN were treated for 24 h with compound 27, the most active 6S derivative to activate
Nrf2 in transgenic zebrafish embryos, at different doses of 5, 10, and 20 μM. As shown in
Figure 4A, 27 significantly increased HO-1 protein expression, compared to untreated
control (p < 0.05 for all concentrations). These data clearly illustrate that the
representative 6S derivative 27 induces HO-1 expression in CCD 841 CoN cells in a
dose-dependent manner.
On the other hand, CCD 841 CoN cells were treated for 3, 6, 12, and 24 h with
compound 27 at a dose of 10 μM. As seen in Figure 4B, 27 significantly increased HO-1
expression from time point of 3 h to 24 h when compared to control (p < 0.05 for all time
points). These observations demonstrate that derivative 27 increases HO-1 expression in
CCD 841 CoN cells in a time-dependent manner.
22

NS
NS
NS
NS
Figure 3. Effects of ROS productions in zebrafish embryos by compounds 1, 18, 26 and
27. Zebrafish embryos at 1 dpf were treated with compounds 1, 18, 26, and 27 at 5 μM
over 2, 4, 8, and 24 h, and ROS levels in the embryos at each time point were measured
using a cell-permeable fluorogenic probe DCFH-DA. One way ANOVA following
Dunnett's test was utilized to determine the potential differences between all treatments
and control. NS: no significant difference. ROS, reactive oxygen spicies.
23

A)
Compound 27
***
0 μM
5 μM
10 μM
20 μM
HO-1
***
**
β-Actin
***
B)
Compound 27
3 h
***
0 h
6 h
12 h
24 h
***
HO-1
**
β-Actin
Figure 4. Effects of compound 27 on HO-1 expression in human normal colonic
epithelial cells CCD-841CoN. (A) Compound 27 increased HO-1 expression in a dose-
dependent manner. Cells were treated with 27 at concentrations of 0, 5, 10, and 20 μM for
24 h, resectively. (B) Compound 27 increased HO-1 expression in a time-dependent
manner. Cells were treated by 27 at 10 μM for 0, 3, 6, 12, and 24 h, respectively. β-Actin
was used as an internal control. The fold changes in HO-1 expression are shown on the
right at each row using densitometric analyses of the bands. Results are mean ± SD (n =
3). Bar, standard error; *, p < 0.05; **, p < 0.01; ***, p <0.001. all statistical tests are
unpaired Student's t test, two-tailed, compared to control (0 μM or 0 h).
Chemical Reactivity of Three Representative 6S Derivatives (18, 25, and 27) with
L-Cysteine. It has been reported that reaction of the reactive cystine residues of Keap1
24

with electrophiles results in the formation of intermolecular disulfide bridges, thus
covalently linking two monomers of Keap1, thereby liberating Nrf2.5 Natural Nrf2/ARE
activators such as xanthohumol, isoliquiritigenin, and SFN, all containing electrophilic
groups as Michael acceptor, have been reported to react with the cysteine residues of
human Keap1.6, 41 To understand the underlying mechanisms that the newly synthesized
6S derivatives activate Nrf2, we first investigated the chemical reactivity of three
representative Nrf2 activators (18, 25, and 27) with L-cysteine in vitro. Similar to the
reaction between 6S and L-cysteine,30 Michael addition reactions between enone 18 or 27
and L-cysteine occurred immediately under slightly basic condition, thereby in 10 min
giving rise to respective cysteine conjugates 35 and 36, both with a cysteine residue
attached to the side chain (Scheme 3A). Regarding catechol compound 25, no Michael
addition took place under the same conditions as above because of the absence of an
unsaturated carbonyl entity as a Michael acceptor in the structure. Under oxidative
environments (i.e. IBX), reaction between 25 and L-cysteine, however, efficiently
produced cysteine conjugate 37, a compound with a cysteine residue binding to the
aromatic ring (Scheme 3A), which is consistent with previous observations from the
reaction of plant catechol, piceatannol.35 In addition, treatment of 25 with L-cysteine by
tyrosinase from mushroom in PBS generated cysteine conjugate 37 as well (Data not
shown).
Thus, we perceived that 1) electrophilic alkenes of ,-unsaturated carbonyl entities
in conjugated enones act as Michael acceptors and undergo Michael addition by
nucleophilic sulfhydryl groups of cysteine residues under basic conditions, as indicated in
pathway A (Scheme 3B); and 2) high susceptibility of catechol moiety in the structure
25

under oxidative environments leads to the formation of o-quinones as reactive Michael
acceptors and subsequently Michael additions take place at ortho-position of the catechol
groups by the treatment of cysteine residues, as described in pathway B (Scheme 3B). As
a result, we identified that both ,-unsaturated carbonyl entities and catechol moieties in
molecules act as major active groups for the assaults of the sulfhydryl groups of the
cysteine residues.
Table 1. Cysteine conjugates found in mice and in zebrafish embryos after treatment of
xenobiotics 18 and 25-27, respectively.
No. RT (min) [M +
H]+
MS/MS
396/378, 275 [M  Cys + H]+ (B)
384/263 [M  Cys + H]+
22.0
21.2
22.8
24.0
396
384
386
384
35
36
386/265 [M  Cys + H]+ (B), 137
384/367 [M  NH2 + H]+ (B), 295 [M  C3H7NO2 + H]+,
277, 169
36a
37
384/367, 295 [M  C3H7NO2 + H]+ (B), 267, 167
386/368 [M  H2O + H]+ (B), 351, 297 [M  C3H7NO2 +
H]+, 279
25.2
25.1
384
386
37a
37b
382/261 [M  Cys + H]+
19.6
382
38
Cys, cysteine; RT, retention time.
Structural Confirmation of Cysteine Conjugates (35-37) Obtained from the
Corresponding Chemical Reaction. Similar to cysteine conjugate of 6S,30 compound 35
was indicated as the cysteine conjugate of dienone 18. The attachment of the cysteine
26

residue at C-5 position in the alkyl tail of 18 was established by HMBC correlations
between C-5 (δC 42.0) and HCys- (δH 3.22/2.89) and H-6 (δH 1.63) as well as H-4 (δH
3.36/2.98) (Figure 5). This is supported by observing m/z 275 [M – Cys + H]+ (lose of
cysteine moiety from m/z 396, pattern a) as the major product ion in its MS/MS spectrum
(Table 1 and Figure 5). Therefore, compound 35 was identified as 5-S-cysteinyl-[6]-
dehydroparadol (Figure 5). Likewise, the interpretation of NMR data identified 36 as 1-S-
cysteinyl-1-(3,4-dihydroxyphenyl)decan-3-one, which was further confirmed by
observing m/z 263 [M – Cys + H]+ (lose of cysteine moiety from m/z 384, pattern a) as
1
the major product ion in its MS/MS spectrum (Table 1). The H NMR and 13C NMR
spectra of conjugate 37 were similar to those of 36, indicating 37 is a cysteine conjugate
of 25. The major differences between 36 and 37 were 1) only two aromatic protons at δH
6.82 (1H, brs) and δH 6.66 (1H, brs) remained in 37, and 2) signals for thio-methine (-
CH-S-) disappeared in 37. This suggested that the cysteine residue in 37 is bound to the
aromatic ring rather the alkyl tail. The linkage of the cysteine residue at C-5' position of
the catechol moiety was accomplished by HMBC correlations between C-3' (δC 119.2)
and HCys- (δH 3.47/2.95) (Figure 5B). This was further supported by its MS/MS fragment
ions at m/z 295 originated from the cleavage of CCys-β-S bond (pattern b), and m/z 169
formed by the cleavage of CCys-β-S bond followed by the α-cleavage of carbonyl group
between C2-C3 bond (Figure 5C). Compound 37 was therefore identified as 3'-S-
cysteinyl-1-(3,4-dihydroxyphenyl)decan-3-one (Figure 5A).
27

pattern a
B)
A)
C)
O
3
Cys
(CH2)4
O
Cys
1
m/ z 275
O
5
(CH2)4
a
H2N
S
4'
OH
35,
R = OCH3
38, R = OH
HO
b
HO
3'
R
35
O
O
pattern b
O
Cys R
(CH2)4
m/ z 295
H2N
(CH2)4
OH
HO
O
35
b
O
36, R = O;
3
HO
S
36a,
R = OH
(CH2)6
OH
HO
R
O
m/ z 169
pattern b
1
3'
OH
Cys
37
H2N
(CH2)4
3
OH
O
4'
O
b
37, R = O;
m/ z 295
HO
H2N
S
5'
37b,
R = OH
OH
(CH2)6
OH
OH
b
OH
S
HO
37
(CH2)6
Cys
HO
OH
(CH2)4
HO
m/ z 267
37a
OH
HMBC
37a
m/z 167
OH
Figure 5. Cysteine conjugates found in mice and in zebrafish embryos. A) Structures of
all cysteine conjugates 35, 36, 36a, 37, 37a, 37b, and 38. B) Main HMBC correlations in
the structures of 35 and 37, respectively. C) Typical fragmentation patterns of 35, 37, and
37a in their respective MS/MS spectra.
Formation of Cysteine Conjugates of 18 and 25-27 in Mice and in Zebrafish
Embryos. We have proved that the major active 6S derivatives could bind to the reactive
cysteine residues to form related cysteine conjugates in vitro. Whether 6S derivatives
undergo the similar pathway in vivo remains unknown. In this section, 18 and 27 were
administrated to mice by oral gavage and 18 and 25-27 were given to zebrafish embryos
in incubation medium. We discovered seven cysteine conjugates, including three
conjugates (35-37) that have been above identified in vitro and four new analogues (36a,
37a, 37b, and 38), using LC-MS approaches (Figure 5). As shown in Figure 6, conjugate
35 was predominantly present in both mouse urine and zebrafish embryos after treatment
28

of 18 (Figures 6A and 6B). Besides conjugate 36 and its reduced form 36a, three
additional conjugates 37, 37a, and 37b, with cysteine residue binding to the aromatic ring,
were detected from 27-treated mouse urine (Figure 6C). Likewise, four cysteine
conjugates 36, 36a, 37, and 37b were found in 27-treated zebrafish embryos (Figure 6D).
As expected, conjugates 37 and 37b were exclusively identified in 25-treated zebrafish
embryos (Figure 6E). Also, three conjugates 37, 37b, and 38 were detected in 26-treated
zebrafish embryos (Figure 6F). These in vivo observations demonstrated that 1)
unsaturated carbonyl entities in the molecules lead to cysteine conjugation occurring on
the alkyl side chain, evidenced by the presence of conjugates 35 in 18-treated mice and
zebrafish, 36 and 36a in 27-treated mice and zebrafish, and 38 in 26-treated zebrafish; 2)
catechol moieties in the structures induce the binding of cysteine residues to the aromatic
rings, evidenced by the appearance of conjugates 37 and 37b in both 25- and 26-treated
zebrafish, and 37, 37a, and 37b in 27-treated mice and zebrafish; and 3) coexistence of
unsaturated carbonyl entities and catechol moieties in a molecule leads to the formation
of both conjugations, evidenced by the existence of conjugates 37, 37b, and 38 in 26-
treated zebrafish, and 36, 36a, 37, 37a, and 37b in 27-treated mice and zebrafish. These
in vivo results verified that both ,-unsaturated carbonyl entities and catechol moieties
in the molecules act as major active sites for the conjugation with the reactive cysteine
residues under physiological conditions.
We have reported that 6S could activate Nrf2 via modification of Keap1 at 17
cysteine residues and these cysteine residues were located in all 5 domains of Keap1
protein.26 In light of our findings above, we proposed that both ,-unsaturated carbonyl
entities and catechol moieties in 6S derivatives could react with the reactive sulfhydryl
29