Article pubs.acs.org/JAFC
Activation of the Kelch-like ECH-Associated Protein 1 (Keap1)/NF-E2Related Factor 2 (Nrf2) Pathway through Covalent Modification of the 2‑Alkenal Group of Aliphatic Electrophiles in Coriandrum sativum L. Yumi Abiko,† Mai Mizokawa,‡ and Yoshito Kumagai*,†,‡ †
Faculty of Medicine and ‡Masters Program in Environmental Sciences, Graduate School of Environmental Sciences, University of Tsukuba, 1-1-1 Tennodai, Tsukuba, Ibaraki 305-8575, Japan S Supporting Information *
ABSTRACT: Phytochemicals able to activate the transcription factor NF-E2-related factor 2 (Nrf2) were isolated from an extract of Coriandrum sativum L. (C. sativum) leaves by preparative octadecyl silica column chromatography. Ultraperformance liquid chromatography and liquid chromatography−tandem mass spectrometry analysis of the isolated components after derivatization with 2-diphenylacetyl-1,3-inandione-1-hydrazone and experiments with HepG2 cells revealed that (E)-2-alkenals with different carbon numbers play a role in Nrf2 activation in these cells. Such Nrf2 activation appears to be attributable to S-alkylation of Kelch-like ECH-associated protein 1 (Keap1), the negative regulator for Nrf2, as determined by a biotin-PEAC5-maleimide assay. Interestingly, (E)-2-butenal caused Keap1 modification and Nrf2 activation, whereas butanal did not. These results suggest that (E)-2-alkenals with an α,β-unsaturated aldehyde moiety, which is a common substituent in phytochemicals isolated from C. sativum leaves, activate the Keap1/Nrf2 pathway associated with cellular protection. KEYWORDS: Coriandrum sativum, 2-alkenal, Keap1/Nrf2, electrophile, α,β-unsaturated aldehyde
■
the cytoplasm.20 When the reactive thiol groups of Kelch-like ECH-associated protein 1 (Keap1), which is the negative regulator of Nrf2, are covalently modified by electrophiles, Nrf2 translocates into the nucleus and interacts with small Maf proteins, and then the complex binds to the antioxidant responsive element (ARE) on DNA.20,21 Although it has been reported that CSLEs can activate Nrf2,8 determination of the phytochemicals in the plant responsible for Nrf2 activation remains to be elucidated. To address this issue, we isolated and identified Nrf2 activators from a CSLE. We report for the first time direct evidence that Nrf2 activation mediated by CSLE is mainly due to (E)-2-alkenals with different carbon numbers because these phytoelectrophiles are able to modify Keap1.
INTRODUCTION The leaves of Coriandrum sativum L. (C. sativum), also known as cilantro, Chinese parsley, or paxi, are not only consumed as a spice to add flavor to food but also traditionally used as a medicine for gastrointestinal disorders, pain, inflammation, and oxidative stress.1,2 Although a variety of chemicals have been identified in C. sativum leaves,3−5 the exact nature of the bioactive substances in the leaves that provide relief from such disorders is poorly understood. Lipopolysaccharide-induced inflammation has been suppressed, presumably by rutin in the leaves of C. sativum.6 A C. sativum leaf extract (CSLE) has been shown to have an antioxidative effect in vitro, and the extract containing fatty acids (C16−18), including linoleic and linolenic acids, protected HaCaT cells against H2O2-induced oxidative damage by inhibiting reactive oxygen species production.7,8 However, the particular phytochemicals in the extract that are responsible for such bioactivity are unknown. The volatile oil obtained from n-pentane extraction of C. sativum leaves contains fatty alcohols and a series of aliphatic electrophiles containing an (E)-2-alkenal, which make up approximately 70% of the oil from C. sativum leaves.9 Aliphatic electrophiles with an α,β-unsaturated aldehyde moiety are of interest, as these compounds are found to be easily bound to protein nucleophiles, resulting in activation or inactivation of electrophilic signal transduction pathways.10−15 Transcriptional factor NF-E2-related factor 2 (Nrf2), which is known to be activated by electrophilic signaling, regulates phase II xenobiotic detoxification enzymes, phase III transporters, and antioxidant proteins, which are known to have a cytoprotective role against oxidative stress and environmental pollutants, such as heavy metals, arsenic, and quinones.16−20 Under normal conditions, Nrf2 rapidly undergoes degradation by ubiquitin-proteasomes in © 2014 American Chemical Society
■
MATERIALS AND METHODS
Materials. C. sativum fresh leaves were purchased from Mekon Foods (Chiba, Japan). (E)-2-Decenal (93.5% purity determined by GC) and (E)-2-dodecenal (96.1% purity determined by GC) were purchased from Tokyo Chemical Industry (Tokyo, Japan). Glutathione (GSH) was obtained from Wako Pure Chemical Industries Ltd. (Osaka, Japan). Tris(2-carboxyethyl)phosphine (TCEP) was purchased from Nacalai Tesque Inc. (Kyoto, Japan). 2-Diphenylacetyl-1,3-inandione-1-hydrazone (DAIH), anti-Nrf2 antibodies, anti-Keap1 antibodies, antiglutamate−cysteine ligase, modifier subunit (GCLM), and antiglutamate−cysteine ligase, catalytic subunit (GCLC), antibodies were purchased from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Anti-heme oxygenase-1 (HO-1), horseradish peroxidase (HRP)-linked anti-rabbit IgG, and biotin−PEAC5-maleimide (BPM) were purchased from Stressgen (Victoria, BC, Canada), Cell Signaling Technology Received: Revised: Accepted: Published: 10936
July 1, 2014 October 7, 2014 October 12, 2014 October 13, 2014 dx.doi.org/10.1021/jf5030592 | J. Agric. Food Chem. 2014, 62, 10936−10944
Journal of Agricultural and Food Chemistry
Article
Figure 1. CSLE-mediated activation of Nrf2 in HepG2 cells and modification of purified Keap1. (A) Cells were exposed to CSLE (10, 25, or 50 μM) for 1, 3, or 6 h, and then total cell proteins were subjected to Western blot analysis with the indicated antibodies. (B) CSLE was incubated with GSH (10 mM) for 30 min at room temperature. The cells were treated with GSH-incubated CSLE (50 μg/mL) for 1 h, and total cell proteins were subjected to Western blot analysis with the indicated antibodies.
Figure 2. HPLC analysis of CSLE (A) and fractions separated by Ultra Pack ODS-S-50B column chromatography (B). Detection at 230 nm. Inc., Osaka, Japan); elution was with 25% acetonitrile in water at a flow rate of 5 mL/min. Phytochemicals with retention times of 2−60 min (fraction I, 10 mg), 60−100 min (fraction II, 10 mg), 100−150 min (fraction III, 10 mg), 150−220 min (fraction IV, 20 mg), and 220−300 min (fraction V, 10 mg) were collected, evaporated in vacuo, and freeze-dried. Phytochemicals in each fraction were monitored using analytical highperformance liquid chromatography (HPLC) [YMC-Pack, ODS-AM, 250 × 4.6 mm, i.d., 5 μm (YMC CO Ltd., Kyoto, Japan); mobile phase, acetonitrile/water, 25:75 (v/v); flow rate, 1 mL/min; detection, 230 nm]. Detection of the Covalent Modification of Keap1 by Electrophiles. The covalent modification of Keap1 by electrophiles was detected using a BPM-labeling assay as previously described.22 Recombinant murine Keap1 was expressed as a C-terminal His-tagged fusion protein from BL21-CodonPlus (DE3)-RIPL cells (Stratagene, La Jolla, CA, USA) and purified using Ni-NTA agarose (Qiagen, Valencia, CA, USA).23 Samples, including Keap1−decenal and −dodecenal adducts, were incubated with 25 μM BPM for 30 min at 37 °C and then subjected to Western blot analysis with an HRP-linked antibiotin antibody (Cell Signaling Technology). Detection of 2-Alkenals Using DAIH Labeling Analysis. Each sample (17 μL) was incubated with 85 μL of a derivatization reagent
(Beverly, MA, USA), and Invitrogen (Carlsbad, CA, USA), respectively. All other reagents were of the highest grade available. Cell Culture. HepG2 cells were obtained from RIKEN Cell Bank (Tsukuba, Japan). The cells were cultured in minimal essential medium (MEM; Wako) containing 10% fetal bovine serum, antibiotics (100 U/mL penicillin and 100 μg/mL streptomycin), and 2 mM L-alanyl-L-glutamine (Invitrogen) in an incubator supplemented with 5% CO2 at 37 °C. In extracting total protein, 8 × 105 cells were seeded in 35 mm dishes. The cells were preincubated in serum-free medium for 12 h before treatment with compounds. CSLE Extraction. C. sativum leaves (78 g) were frozen in liquid nitrogen and transferred to a 1 L flask containing 80 mL of distilled deionized water and 160 mL of n-hexane. The flask was refluxed in an oil bath at 80 °C for 2 h, and the n-hexane layer was collected using a separating funnel. The n-hexane layer was passed over anhydrous sodium sulfate to remove any trace of water and then treated with argon gas. The n-hexane layer was evaporated in vacuo to remove n-hexane, and the residues were redissolved in acetonitrile or dimethyl sulfoxide (DMSO). Samples were stored at −80 °C prior to use. Separation of CSLE. The CSLE (70 mg) in DMSO was separated into five fractions by preparative column chromatography using an Ultra Pack ODS-S-50B column (300 × 26 mm i.d., 50 μm; Yamazen Science 10937
dx.doi.org/10.1021/jf5030592 | J. Agric. Food Chem. 2014, 62, 10936−10944
Journal of Agricultural and Food Chemistry
Article
Figure 3. Nrf2 activation during exposure of HepG2 cells to the CSLE fractions. The cells were exposed to each fraction (5, 20, or 100 μM) for 1 h, and then the cell lysates were subjected to Western blot analysis with the indicated antibodies. The bands were quantified using ImageJ software, and the density of each band was normalized to that of actin. Each value is the mean ± SE of three determinations. (∗) p < 0.05 and (∗∗) p < 0.01 versus control.
Figure 4. Identification of (E)-2-alkenals in CSLE and fractions by UPLC-MS/MS following reaction with DAIH: (A) reaction of (E)-2-alkenal with DAIH leading to the dehydrated reaction product [predicted MS numbers of the fragments are m/z 323.13 derived from DAIH and m/z 25.2 + 14n derived from the (E)-2-alkenal residue of the (E)-2-alkenal-DAIH adduct]; (B) observed MS number of the reaction products with DAIH and fragments derived from the 2-alkenal−DAIH adduct by UPLC-MS/MS analysis. consisting of 0.06% DAIH, 140 μM HCl, and 93% acetonitrile for 15 min on ice. A 10 μL sample of the reaction mixture was analyzed by ultraperformance liquid chromatography (UPLC) and liquid chromatography−tandem mass spectrometry (LC-MS/MS). Briefly, the reaction products were characterized using an Acquity UPLC system (Waters Co., Milford, MA, USA) equipped with an Acquity BEH C18 column (2.1 mm × 50 mm i.d., 1.7 μm) held at 35 °C in the direct injection mode. Mobile phases A [water containing 0.1% (v/v) formic acid] and B [acetonitrile containing 0.1% (v/v) formic acid] were linearly mixed using a gradient program, and the instrument was calibrated immediately prior to each series of experiments. The flow rate was 0.3 μL/min, and the mobile phase composition was as follows: 15% B for 10 min; linear increase over 20 min to 95% B; maintained at 95% B for 2 min before a linear return to 15% B over 3 min. The total running time, including conditioning the column to the initial conditions, was 35 min. The eluted chemical compounds were then transferred to the electrospray source of the Synapt High Definition MS system (Waters Co.). The system control and analysis of the mass spectra were
performed using Waters MassLynx software ver. 4.1. Electrospray ionization was used with a capillary voltage of 2.8 kV and a sampling cone voltage of 35 V. Low (6 eV) or elevated (stepped from 15−60 eV) collision energy was used to generate either intact precursor ions (low energy) or product ions (elevated energy). The source temperature was 80 °C, and the detector was operated in the negative ion mode. Data were collected from m/z 50−1000. All analyses were acquired with an independent reference; leucine enkephalin [M − H]− ion as lock mass (m/z 554.2615) was infused via the LockSpray ion source to ensure accuracy and reproducibility. Data were analyzed with MassLynx version 4.1 software and MassFragment version 1.1 software. Luciferase Assay. Transient transfection of cDNAs was performed using Lipofectamin 2000 (Invitrogen) according to the manufacturer’s instructions. Briefly, HepG2 cells were seeded at 0.5 × 105 cells/cm2 in 12-well plates. After the cells were incubated overnight, 1.6 μg of ARE-Luciferase cDNA and 0.16 μg of pRL-TL cDNA or Lipofectamin 2000 reagent were mixed with OPTI-MEM (Invitrogen) in separate tubes and incubated for 20 min at room temperature to allow the 10938
dx.doi.org/10.1021/jf5030592 | J. Agric. Food Chem. 2014, 62, 10936−10944
Journal of Agricultural and Food Chemistry
Article
formation of complexes. These solutions were then added to the cells and incubated for 12 h, followed by a further 12 h of incubation in serum-free medium. The cDNA-transfected cells were then exposed to the samples, washed with phosphate-buffered saline (PBS), and lysed with Passive Lysis Buffer (Promega, Madison, WI, USA). Luciferase activity was measured using a Dual-Luciferase Reporter Assay System (Promega) according to the manufacturer’s instructions. Western Blot Analysis. HepG2 cells were exposed to CSLE, butanal (99.85% purity, Sigma-Aldrich, St. Louis, MO, USA), (E)-2butenal (99% purity, Nacalai), (E)-2-decenal, or (E)-2-dodecenal, washed with ice-cold PBS, and then collected by scraping into a 2% sodium dodecyl sulfate (SDS) solution. After the cells were lysed at 95 °C for 20 min, protein concentrations were determined using the BCA Protein Assay (Piece, Rockford, IL, USA). Each sample was mixed with a half-volume of SDS−polyacrylamide gel electrophoresis (PAGE) loading buffer (62.5 mM Tris-HCl, pH 6.8; 6% SDS; 24% glycerol; 50 mM TCEP; and 0.015% bromophenol blue) and incubated at 95 °C for 5 min. The cellular proteins were then separated using SDS-PAGE and transferred to polyvinylidene difluoride membranes (Bio-Rad Laboratories, Hercules, CA, USA). The membranes were blocked with 5% skim milk in TTBS (20 mM Tris-HCl, pH 7.5; 150 mM NaCl; and 0.1% Tween 20) and then incubated with primary antibody in TTBS. To detect immunoreactive proteins, HRP-conjugated anti-rabbit IgG and an enhanced system (Chemi-Lumi One L; Nacalai, Kyoto, Japan) were used. Representative blots were from three independent experiments. Determination of Electrophilicity Index. The lowest unoccupied molecular orbital energy (ELUMO) and highest occupied molecular orbital energy (EHOMO) were determined using Gaussian 09 (ver. 8.0) software (Conflex, San Diego, CA, USA). The ground state equilibrium geometries of each structure were calculated with density-functional RB3LYP 6-31G(d). The global hardness (η) was calculated as η ≈ (ELUMO − EHOMO)/2, softness (σ) was calculated as σ = 1/η, and the electronic chemical potential (μ) was calculated as μ ≈ (EHOMO + ELUMO)/2. The electrophilicity index (ω) was calculated as ω = μ2/2η.24,25 Statistical Analysis. Statistical significance was assessed using Student’s t test or Dunnett’s post hoc test after a one-way analysis of variance with Microsoft Excel for Mac 2011 version 14.4.3; p < 0.05 was considered significant.
■
ion mode of the CSLE components after derivatization with DAIH (see Figure 4A).26 Authentic (E)-2-decenal and (E)-2-dodecenal were converted into DAIH derivatives with retention times of 16.61 min (m/z 489.3) and 20.95 min (m/z 517.3), respectively (Figure 4 and Figure S3 in the Supporting Information), whereas no such peaks and corresponding molecular masses were seen without DAIH (Figure S3A). Using the DAIH derivative method, we identified (E)-2-alkenal-related chemicals, including (E)-2-decenal (compound 2 in fraction II), (E)-2-undecenal (compound 4 in fraction III), and (E)-2-dodecenal (compound 5 in fraction IV) (Figures 4 and 5). Compound 3 was not an (E)-2-alkenal because
RESULTS
Activation of Nrf2 by CSLE. Exposure of HepG2 cells to CSLE resulted in activation of Nrf2 and up-regulation of downstream proteins, such as GCLM and HO-1 (Figure 1A); however, CSLE at a concentration of 100 μg/mL repressed the Nrf2 activation and thus GCLM and HO-1 gene expression. Pretreatment with GSH markedly diminished CSLE-mediated activation of Nrf2 in the cells (Figure 1B), suggesting electrophiles contributed to the activation. Consistent with this, the thiol groups of Keap1, the negative regulator for Nrf2, were modified by CSLE in a concentration-dependent manner as determined by a BPM assay (Supporting Information Figure S1). Separation of Nrf2 Activators in CSLE. As shown in Figure 2A, the CSLE contained eight major compounds detected at 230 nm; compounds 2 and 5 were identical to authentic (E)-2-decenal and (E)-2-dodecenal, respectively (Figure 2 and Figure S2 in the Supporting Information). The CSLE components were separated into five fractions by Ultra Pack ODS-S-50B column chromatography (Figure 2B), and then the Nrf2 activation potential of each fraction was evaluated in HepG2 cells. Unexpectedly, fractions I−V all activated Nrf2 to a similar extent (Figure 3), suggesting that the observed Nrf2 activation might be due to electrophiles with (E)-2-alkenal substituents, as various (E)2-alkenals have been observed previously in a CSLE.3,9 Analysis of (E)-2-Alkenals in CSLE. To investigate whether or not the components in each fraction contained (E)-2-alkenal moieties, we performed UPLC-MS/MS analysis in the negative
Figure 5. UPLC and LC-MS/MS spectra of reaction products after reaction of each fraction with DAIH: (A−D) UPLC/MS chromatograms of derivatives of 2-alkenal with DAIH (left) monitored at m/z 323.1, respectively. The mass spectra were obtained from derivatives with retention times of 16.7 min (A), 19.1 min (B), 21.0 min (C), 22.6 min (C) and 22.4 min (D), corresponding to the DAIH adducts.
this molecule could not be converted to a DAIH derivative (data not shown). Fraction V contained (E)-2-tridecenal, although we did not examine whether (E)-2-tridecenal is compound 6, 7, or 8 (Figures 4 and 5). (E)-2-Hexenal, (E)-2-tetradecenal, (E)-2pentadecenal, (E)-2-hexadecenal, and (E)-2-heptadecenal were also detected in the CSLE by DAIH derivatization (Figure 4 and Figure S4 in the Supporting Information). These results suggest that there are a variety of aliphatic electrophiles with an (E)-2-alkenal group in the CSLE, which is in agreement with previous studies.3,9,16 Activation of Nrf2 Is Coupled to Covalent Modification of Keap1 by (E)-2-Decenal and (E)-2-Dodecenal. Exposure of HepG2 cells to (E)-2-decenal and (E)-2-dodecenal under 10939
dx.doi.org/10.1021/jf5030592 | J. Agric. Food Chem. 2014, 62, 10936−10944
Journal of Agricultural and Food Chemistry
Article
Figure 6. (E)-2-Decenal- or (E)-2-dodecenal-mediated activation of Nrf2 and up-regulation of downstream gene products. (A, B) The cells were exposed to (E)-2-decenal (A) or (E)-2-dodecenal (B) for 3 h, and total cell proteins were subjected to Western blot analysis with the indicated antibodies. (C) ARE-luciferase and pRL-TK cDNAs were transfected into HepG2 cells. After exposure of the cells to (E)-2-decenal or (E)-2-dodecenal for 6 h, luciferase activity was measured. (D, E) The cells were exposed to 25 μM (E)-2-decenal (D) or (E)-2-dodecenal (E) for 6, 12, or 24 h, and total cell proteins were subjected to Western blot analysis with the indicated antibodies. The bands were quantified using ImageJ software, and the density of each band was normalized to that of actin. Each value is the mean ± SE of three determinations. (∗) p < 0.05 and (∗∗) p < 0.01 versus control.
Figure 7. Modification of cellular Keap1 and purified Keap1 by (E)-2-decenal and (E)-2-dodecenal. (A) Recombinant murine Keap1 was incubated with (E)-2-decenal and (E)-2-dodecenal in 20 mM Tris-HCl (pH 8.5) at 37 °C for 20 min and allowed to react with 25 μM BPM (30 min at 37 °C). The reaction mixture was subjected to Western blot analysis with avidin−horseradish peroxidase conjugate. The bands were quantified using ImageJ software, and the band densities were normalized to that of Keap1. Each value is the mean ± SE of three determinations. (∗) p < 0.05 and (∗∗) p < 0.01 versus control. (B, C) HepG2 cells were incubated with (E)-2-decenal (B) and (E)-2-dodecenal (C) for 30 min and then lysed with RIPA buffer. Cell lysates were incubated with BPM (50 μM) for 30 min and then immunoprecipitated with avidin−agarose. The precipitated proteins were subjected to Western blot analysis with an anti-Keap1 antibody. The bands were quantified using ImageJ software. Each value is the mean ± SE of three determinations. (∗) p < 0.05 and (∗∗) p < 0.01 versus control.
nontoxic conditions (
