Costunolide specifically binds and inhibits thioredoxin reductase 1 to induce apoptosis in colon cancer

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54

CAN13549_proof ■ 15 October 2017 ■ 1/13

Cancer Letters xxx (2017) 1e13

Contents lists available at ScienceDirect

Cancer Letters journal homepage: www.elsevier.com/locate/canlet

Original Article

Q4 Q3

Costunolide specifically binds and inhibits thioredoxin reductase 1 to induce apoptosis in colon cancer Weishan Zhuge a, b, Ruijie Chen a, Katanaev Vladimir c, Xidan Dong d, Khan Zia e, Xiangwei Sun f, Xuanxuan Dai a, Miao Bao a, Xian Shen b, f, **, Guang Liang a, * a

Chemical Biology Research Center, School of Pharmaceutical Sciences, Wenzhou Medical University, Wenzhou, Zhejiang 325035, China Department of Gastrointestinal Surgery, The First Affiliated Hospital of Wenzhou Medical University, Wenzhou, Zhejiang 325035, China School of Biomedicine, Far Eastern Federal University, Sukhanova Street 8, Vladivostok, 690922, Russian Federation d Department of Pathology, The First Affiliated Hospital of Wenzhou Medical University, Wenzhou, Zhejiang 325035, China e Department of Pathology and Laboratory Medicine, Western University, London, ON N6A5C1, Canada f Department of Gastrointestinal Surgery, The Second Affiliated Hospital of Wenzhou Medical University, Wenzhou, Zhejiang 325000, China b c

a r t i c l e i n f o

a b s t r a c t

Article history: Received 13 August 2017 Received in revised form 3 October 2017 Accepted 6 October 2017

Colon cancer is one of the leading causes of cancer-related deaths. A natural sesquiterpene lactone, costunolide (CTD), showed inhibition of cancer development. However, the underlying mechanisms are not known. Here, we have examined the therapeutic activity and novel mechanisms of the anti-cancer activities of CTD in colon cancer cells. Using SPR analysis and enzyme activity assay on recombinant TrxR1 protein, our results show that CTD directly binds and inhibits the activity of TrxR1, which caused enhanced generation of ROS and led to ROS-dependent endoplasmic reticulum stress and cell apoptosis in colon cancer cells. Overexpression of TrxR1 in HCT116 cells reversed CTD-induced cell apoptosis and ROS increase. CTD treatment of mice implanted with colon cancer cells showed tumor growth inhibition and reduced TrxR1 activity and ROS level. In addition, it was observed that TrxR1 was significantly upregulated in existing colon cancer gene database and clinically obtained colon cancer tissues. Our studies have uncovered the mechanism underlying the biological activity of CTD in colon cancer and suggest that targeting TrxR1 may prove to be beneficial as a treatment option. © 2017 Published by Elsevier B.V.

Keywords: Colon cancer Costunolide Thioredoxin/thioredoxin reductase 1 Oxidative stress Endoplasmic reticulum stress

Introduction Colon cancer is a leading cause of cancer-related deaths. Currently, colon cancer is managed through surgery and chemotherapy [1]. The overall survival of patients with advanced colon

Q1

Q2

Abbreviations: ER, endoplasmic reticulum; TrxR1, thioredoxin/thioredoxin reductase 1; EIF2, eukaryotic initiation factor 2; ATF-4, activating transcription factor 4; CHOP, CAAT/enhancer-binding protein homologous protein; PI, propidium iodide; DCFH-DA, 20,70-dichlorodihydrofluorescein diacetate; MDM-2, murine double minute 2; Cdc2, cyclin-dependent kinase 1 cell division cycle protein 2; PARP, poly ADP-ribose polymerase; NAC, N-acetyl cysteine; MTT, 3-(4,5- dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide; Ki-67, nuclear protein associated with cell proliferation; MDA, malondialdehyde. * Corresponding author. ** Corresponding author. Department of Gastrointestinal Surgery, The Second Affiliated Hospital of Wenzhou Medical University, Wenzhou, Zhejiang 325000, China. E-mail addresses: [email protected] (X. Shen), [email protected] (G. Liang).

cancer has improved over the past few decades. Response to current systemic chemotherapies can reach up to 50%. However, resistance develops in nearly all patients with colon cancer and limits the therapeutic efficacies of many anti-cancer agents [2]. These developments eventually lead to chemotherapy failure. Studies to date have identified a number of molecular derangements which serve as biomarkers. These include microsatellite instability, CpG island methylator phenotype, chromosomal instability, and BRAF and KRAS mutations [3,4]. Using these biomarkers and subtyping colon cancer can lead to marked differences in survival [5]. However, drug resistance remains an obstacle to successful chemotherapy [6]. Therefore, a novel approach to combatting colon cancer is needed. The use of various natural and synthetic drugs for colon cancer is gaining attention in recent years. We know that up to seventy percent of all cancers correlate with diet and almost 90% of colon cancer may be preventable through modifying diet [7]. One such natural compound that shows remarkable anti-cancer activities in a host of human cancers is costunolide (CTD). CTD is a naturally

https://doi.org/10.1016/j.canlet.2017.10.006 0304-3835/© 2017 Published by Elsevier B.V.

Please cite this article in press as: W. Zhuge, et al., Costunolide specifically binds and inhibits thioredoxin reductase 1 to induce apoptosis in colon cancer, Cancer Letters (2017), https://doi.org/10.1016/j.canlet.2017.10.006

55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119

CAN13549_proof ■ 15 October 2017 ■ 2/13

2

W. Zhuge et al. / Cancer Letters xxx (2017) 1e13

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 Please cite this article in press as: W. Zhuge, et al., Costunolide specifically binds and inhibits thioredoxin reductase 1 to induce apoptosis in colon cancer, Cancer Letters (2017), https://doi.org/10.1016/j.canlet.2017.10.006

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

CAN13549_proof ■ 15 October 2017 ■ 3/13

W. Zhuge et al. / Cancer Letters xxx (2017) 1e13

occurring sesquiterpene lactone. In two studies utilizing chemically-induced models of colon cancer through exposure of azoxymethane (AOM), CTD reduced the frequencies of aberrant crypt foci (ACF) [8,9]. This effect was associated with supressed ornithine decarboxylase activity of the colonic mucosal tissue [8]. Ornithine decarboxylase catalyzes the first and committed step in the synthesis of polyamines which are important for DNA double strand-break repair pathway and as antioxidants. A decrease in polyamines suggests that CTD may enhance oxidative stress and associated DNA damage in colon cancer. In fact, CTD has been reported to enhance the production of reactive oxygen species (ROS) in oesophageal cancer [10], lung adenocarcinoma [11], bladder cancer [12], breast cancer [13], and ovarian cancer [14], among others. Although this may seem counterintuitive as ROS is believed to enhance tumorigenesis [15,16], it is now being realized that compounds which enhance ROS unmask the deleterious damage, particularly in cancer cells [17]. What remains enigmatic is the mechanism by which CTD enhances ROS and oxidative damage. Recent studies have shown that thioredoxin/thioredoxin reductase (TrxR) system contributes to tumor cell resistance to oxidative stress-induced apoptosis [18]. This raises the question whether CTD and other ROS enhancing agents target TrxR system to generate ROS. TrxR is overexpressed in many human cancers and plays a role in regulating intracellular redox balance [19]. Of the two mammalian TrxR forms, TrxR1 is believed to be the major redox-regulator and regulates cell proliferation, angiogenesis, transcription, and DNA repair [20,21]. In this study, we have examined the therapeutic effect of CTD on colon cancer cells in culture and in human colon cancer xenografts. Our studies show that CTD reduces colon cancer viability through causing cell cycle arrest and inducing apoptosis. Negative growth signals in colon cancer were mediated through elaboration of ROS. We also found that CTD directly binds and inhibits TrxR1 to increase ROS levels. In human colon cancer biopsy specimens, TrxR1 expression correlated with the expression of apoptotic genes. Our studies have discovered a novel mechanism of the anti-cancer activities of CTD. Materials and methods Reagents Costunolide (CTD) was obtained from Aladdin (Shanghai, China). N-acetyl-Lcysteine (NAC) was from Sigma-Aldrich (St. Louis, MO, USA). Antibodies against cyclin B1 (sc-245), B-cell lymphoma 2 (Bcl2, sc-492), Bcl2-associated protein x (Bax, sc-493), murine double mutant 2 (MDM2, sc-965), p53 (sc-126), cell division cycle 2 (Cdc2, sc-54), GAPDH (sc-32233), Ki-67 (sc-7846), TrxR1 (sc-28321), and horseradish peroxidase (HRP)-conjugated secondary antibodies were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Antibodies against cleaved form of Poly (ADP-ribose) polymerase (cleaved-PARP, 5625S), phosphorylated eukaryotic initiation factor 2a (p-eIF2a, 3398S), eIF2a (9722S), activating transcription factor-4 (ATF4, 11815S), CCAAT/-enhancer-binding protein homologous protein (CHOP, 2895S), and cleaved caspase-3 (9661S) were obtained from Cell Signaling Technology (Danvers, MA). FITC Annexin V Apoptosis Detection Kit I and propidium iodide (PI) were purchased from BD Pharmingen (Franklin Lakes, NJ). Reactive oxygen species probe 20 ,70 -dichlorodihydrofluorescein diacetate (DCFH-DA) was purchased from Thermo Fisher (Carlsbad, CA, USA).

3

(Shanghai, China). Cells were cultured in McCoy's 5A medium (Gibco, Eggenstein, Germany) supplemented with 10% heat-inactivated fetal bovine serum (Gibco), 100 U mL1 penicillin, and 100 mg mL1 streptomycin. Cell viability, cell cycle, and cell apoptosis assays The routine MTT assay for cell viability was used as described in details in the supplementary file. In addition, PI staining was used for cell cycle analysis and Annexin V/PI for apoptosis detection in an FACS Calibur flow cytometer (BD Biosciences, CA). The methodology was described in details in the supplementary file. For apoptotic statistics, the apoptotic cells contain Annexin V/PI double-positive cells and Annexin V-positive/PI-negative cells. Colony formation assay The routine colony formation assay using HCT-116 cells was described in details in the supplementary file. Western blotting analysis Total proteins from cultured cells or tumor tissues were isolated and the concentrations were measured by the Bradford protein assay kit (Bio-Rad, Hercules, CA). The routine Western Blot assay was described in details in the supplementary file. Densitometric measurements were performed using Image J (National Institute of Health, MD). Measurement of reactive oxygen species generation in cultured cells Intracellular ROS contents were measured by flow cytometry utilizing dichlorodihydrofluorescein diacetate (DCFH-DA). Briefly, cells were plated at 5  105 density in 6-well plates and allowed to attach for 12 h. Cells were then exposed to CTD for 1.5 h. NAC pretreatments were carried out at 5 mM for 1 h. Following treatments, cells were stained with 10 mM DCFH-DA at 37  C for 30 min in the dark. Cells were collected and DCF fluorescence was analyzed by FACS Calibur flow cytometer (BD Biosciences, CA). Images were also captured using Nikon epifluorescence microscope equipped with a digital camera (Nikon, Japan). Electron microscopy HCT-116 cells were treated with vehicle control (DMSO) or 30 mM CTD for 12 h. NAC pretreatments were carried out at 5 mM for 1 h. Cells were collected and fixed in phosphate buffer (pH 7.4) containing 2.5% glutaraldehyde for 6 h at 4  C. Cells were post-fixed in 1% OsO4 at room temperature for 60 min, stained with 1% uranyl acetate, dehydrated through graded acetone solutions, and embedded in Epon. Areas containing cells were block-mounted and cut into 70 nm sections and examined with the electron microscope (H-7500, Hitachi, Ibaraki, Japan). Cell transfections for gene silencing or overexpression Gene silencing: HCT-116 cells were plated in 6-well plates at a density of 5  104 for 24 h. siRNA against ATF4 or non-targeting control were transfected at a final concentration of 50 pmol mL1 using lipofectamine 2000 reagent (Invitrogen, CA). Culture medium was replaced with fresh medium after 6 h and cells were incubated for an additional 24 h. Then, cells were treated with 30 mM CTD for 15 h and used for subsequent experiments. siRNA oligonucleotides were synthesized by GenePharma (Shanghai, China). The two distinct sequences for ATF4 siRNA were designed and used (No.1: sense 50 -GCCUAGGUCUCUUAGAUGATT-30 and antisense 50 -UCAUCUAAGAGACCUAGGCTT-3’; No.2: sense 50 -GCGUAGUUCGCUAAGGUGATT-30 and antisense 50 -UCACCUUAGCGAACUACGCTT-30 ). Gene overexpression: The recombinant plasmid vector TXNRD1 coding TrxR1 protein was obtained from Addgene (Plasmid #38863, Addgene, Cambridge, MA). The TXNRD1 plasmid was transfected into colon cancer cell line (HCT-116) using Lipofectamine 3000 reagent (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's protocol. After 24 h post transfection, TXNRD1 expression in HCT-116 cells was confirmed by Western blotting analysis. Thioredoxin reductase-1 activity assays

Colon cancer cell lines Human colon cancer cell lines HCT-116, HT-29, and SW620 were purchased from the Institute of Biochemistry and Cell Biology, Chinese Academy of Sciences

TrxR1 activity was determined at room temperature using SpectraMax M5 microplate reader (Molecular Devices, USA). NADPH-reduced TrxR1 (160 nM) was incubated with various concentrations of CTD for 2 h at room temperature in 96-

Fig. 1. CTD suppresses cell viability, and induces cell cycle arrest and apoptosis in human colon cancer cells. (A) Chemical structure of costunolide (CTD). (B, C) Effect of CTD on human colon cancer cell viability. HCT-116, SW620 and HT-29 cells were treated with increasing concentration of CTD for 24 h (B) or 48 h (C). Cell viability was determined by MTT assay and the IC50 values were calculated. (D) Cell cycle phase analysis following exposure of colon cancer cells to CTD. Cells were exposed to CTD at 10, 20 or 30 mM for 15 h. DMSO was used as vehicle control. Figure showing flow cytometry histograms of cells stained with PI. (E) Effect of CTD on colon cancer cell apoptosis as assessed by Annexin V/PI staining. Cells were exposed to CTD for 20 h. (F, G) quantification of flow cytometry data for G2/M phase arrest (F) and percent apoptotic cells (G) [*p < 0.05, **p < 0.01 and***p < 0.001 compared to DMSO]. (H) Effect of varying CTD concentrations on colon cancer cell colony formation. Cells were incubated with CTD for 5 h and allowed to grow for 9 days. Colonies were stained by crystal violet dye. (I) Western blot analysis of G2/M cell cycle-associated proteins and apoptosis-associated proteins in cells challenged with CTD for 12 h (cell cycle protein) and 15 h (apoptosis protein). GAPDH was used as loading control. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Please cite this article in press as: W. Zhuge, et al., Costunolide specifically binds and inhibits thioredoxin reductase 1 to induce apoptosis in colon cancer, Cancer Letters (2017), https://doi.org/10.1016/j.canlet.2017.10.006

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

CAN13549_proof ■ 15 October 2017 ■ 4/13

4

W. Zhuge et al. / Cancer Letters xxx (2017) 1e13

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 Please cite this article in press as: W. Zhuge, et al., Costunolide specifically binds and inhibits thioredoxin reductase 1 to induce apoptosis in colon cancer, Cancer Letters (2017), https://doi.org/10.1016/j.canlet.2017.10.006

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

CAN13549_proof ■ 15 October 2017 ■ 5/13

W. Zhuge et al. / Cancer Letters xxx (2017) 1e13 well plates. A mixture of TE buffer (50 mM Tris-HCl, pH 7.5, 1 mM EDTA, 50 mL) containing 5,50 -dithiobis(2-nitrobenzoate) (DTNB) and NADPH was added to achieve final concentrations of 2 mM and 200 mM, respectively. The linear increase in absorbance at 412 nm during the initial 3 min was recorded. The same amounts of DMSO (1%, v/v) were added to the control experiments and the activity was expressed as the percentage of the control. TrxR1 activity in cells and tissue samples was measured by end-point insulin reduction assay as described previously [22,23]. Total proteins isolated were measured by the Bradford assay. Briefly, 100 mg total proteins were incubated in a final reaction volume of 50 mL containing 100 mM Tris-HCl (pH 7.6), 0.3 mM insulin, 660 mM NADPH, 3 mM EDTA, and 15 mM E. coli-derived Trx (Sigma) for 1 h at 37  C. The reaction was terminated by adding 200 mL of 1 mM 2,4-dinitrochlorobenzene (DTNB). A blank sample, containing everything except Trx, was treated in the same manner. The absorbance at 412 nm was measured, and the blank value was subtracted from the corresponding absorbance value of the sample. The activity was expressed as the percentage of the control.

5

measured by using a Lipid Peroxidation MDA assay kit (Beyotime Institute of Biotechnology).

Statistical analysis Data and statistical analysis in this study comply with the recommendations on experimental design and analysis in pharmacology [29]. All experiments were assayed in quintuplicate (n ¼ 5) and statistical analysis was performed only when a minimum of n ¼ 5 independent samples were acquired. All the data are reported as mean ± SEM. Statistical analysis was performed with GraphPad Prism 6.0 software (GraphPad, San Diego, CA, USA). We used one-way ANOVA followed by Dunnett's post hoc test when comparing more than two groups of data and one-way ANOVA, non-parametric KruskaleWallis test followed by Dunn's post hoc test when comparing multiple independent groups. When comparing two groups, the unpaired Student's t-test was used. A p value < 0.05 was considered statistically significant.

Surface plasmon resonance analysis for CTD-TrxR1 interaction The binding affinity of CTD with recombinant human TrxR1 protein (Sigma) was determined using a ProteOn XPR36 Protein Interaction Array system (Bio-Rad) with a GLH sensor chip (ProteOn). The methodology was described in details in the supplementary file. Molecular docking of CTD to the TrxR1 structural model To probe the interaction mode between CTD and TrxR1, molecular docking study was carried out by using AutoDock (version 4.2.6) [24] and was described in details in the supplementary file. The crystal structure of the human TRXR1 (PDB code: 2ZZ0) was derived from Protein Data Bank as the receptor [25]. AutoDockTools version 1.5.6 and PyMol was used to analyze the docking results [24,26]. Patient samples This study was approved by the Institutional Research Human Ethical Committee of the Wenzhou Medical University for the use of clinical biopsy specimens and informed consent was obtained from the patients. A total of 20 colon cancer biopsy samples were obtained. Clinical diagnosed was performed at the Second Affiliated Hospital of Wenzhou Medical University during the period of 2016e2017. Colon cancer tissues and matched tumor-adjacent morphologically normal colon tissues were frozen and stored in liquid nitrogen until further use. In vivo xenografts Animal studies were performed in compliance with the ARRIVE guidelines [27,28]. Experimental procedures followed Wenzhou Medical University's Policy on the Care and Use of Laboratory Animals. Four-week-old athymic BALB/c nu/nu female mice (18e22 g; total n ¼ 18) were purchased from Vital River Laboratories (Beijing, China). Animals were housed at a constant room temperature with a 12 h: 12 h light/dark cycle and fed a standard rodent diet and water. Mice were divided into three experimental groups randomly. HCT-116 cells were resuspended in 0.1 mL PBS and injected subcutaneously into the right lower limb of each nude mouse at 8  106 cells per mouse. When xenograft tumors reached a volume of 100e150 mm3, mice were treated with 7.5 mg kg1 or 15 mg kg1 CTD by intraperitoneal (i.p.) injection once a day for 12 days. Control mice received vehicle only. Since we found that the tumors in mice started to ulcerate at 12th day, we ended the experiment and sacrifice the mice at day 12. Tumor volumes were determined by measuring length (l) and width (w) and calculating volume using V ¼ 0.5  l  w2 at indicated time points. At the end of the study period, mice were sacrificed and tumor specimens were harvested and weighed. H&E staining, and DHE/DCFH-DA immunofluorescence staining for ROS The routine methodology for tumor tissue staining was described in details in the supplementary file. Malondialdehyde (MDA) assay Tumor samples from nude mice were homogenized. The tissue lysates were then centrifuged at 12 000  g for 10 min at 4  C to collect the supernatant. Total protein content was determined by using the Bradford assay. MDA levels were

Results CTD effectively suppresses cell viability by inhibiting growth and inducing apoptosis in human colon cancer cells CTD has been reported to exhibit inhibitory activity in a number of tumor types. However, the effect of CTD on colon cancer is not fully known. We cultured three established colon cancer cells, HCT116, SW620 and HT-29, in increasing concentrations of CTD and measured viability. Exposure of these colon cancer cells to CTD reduced the viability in a dose-dependent manner at both 24 h and 48 h (Fig. 1B and C). At 24 h exposure period, we obtained IC50 values of 21.63, 26.37, and 30.34 mM for HCT-116, SW620, and HT29, respectively. Although the pattern remained similar, longer term exposure appeared to be more effective as can be seen by reduced IC50 values. As cell viability is the sum of both positive and negative growth signals, we assessed cell growth and apoptosis in this experimental platform. Based on the viability data, we selected 10, 20, and 30 mM CTD for these studies. Propidium iodide (PI) staining of cells revealed accumulation of cells in the G2/M cell cycle phase following exposure to CTD (Fig. 1D and F). Here again, CTD produced cell cycle arrest in a dose-dependent manner. We confirmed these results by measuring cell cycle-associated proteins. Specifically, we measured murine double minute (MDM2) [30], cell division cycle protein 2 (CDC2, also known as cyclindependent kinase 1) [31], and cyclin B [31] levels in colon cancer cells following exposure to CTD. Indeed, exposure of cells to CTD reduced the levels of MDM2, CDC2, and cyclin B1 (Fig. 1I). Next, we assessed the second negative growth signal in cells following CTD challenge. We stained cells with Annexin V/PI and show induction of apoptotic death in colon cancer cells (Fig. 1E and G). These results were also confirmed by western blotting. CTD increased the levels of p53 and Bcl2-associated protein x (Bax) while reducing Bcl2 (Fig. 1I). In addition, the cleaved forms of PARP and caspase-3 were increased in cells exposed to CTD. PARP is one of the best characterized substrate of active caspases [32]. We then utilized the colony formation assay which incorporate these growth signals and shows whether cancer cells are able to form colonies. As shown in Fig. 1H, CTD prevented colony formation at 20 and 30 mM levels. Collectively, our findings show that CTD reduces growth of colon cancer cells and induces apoptotic cell death.

Fig. 2. CTD induces ROS accumulation. (A) ROS levels in colon cancer cells were assessed by staining DCFH-DA at different times following exposure to CTD at 30 mM. DCF fluorescence was measured by flow cytometry. The data came from 5 independent experiments and showed as mean ± SEM. (B) Colon cancer cells with or without 5 mM NAC were exposed to different concentrations of CTD (10, 20, and 30 mM) for 1 h and intracellular ROS levels were measured by DCF fluorescence. (C) G2/M phase accumulation in cells exposed to CTD (15 h), with or without pretreatment with 5 mM NAC for 1 h [**p < 0.01, ***p < 0.001 compared to DMSO]. Representative histograms are shown in Supplementary Fig. S1. (D) Apoptosis induction in cells exposed to CTD (20 h), with or without pretreatment with 5 mM NAC for 1 h [***p < 0.001 compared to DMSO]. Representative histograms are shown in Supplementary Fig. S1. (E) Colony formation in cells exposed to 30 mM CTD (5 h), with or without pretreatment with 5 mM NAC for 1 h. (F) Western blot analysis of proteins in cells pretreated with 5 mM NAC prior to 30 mM CTD exposure. For cell cycle phase proteins, CTD exposure was carried out for 12 h and for apoptosis-related protein, exposure was for 15 h.

Please cite this article in press as: W. Zhuge, et al., Costunolide specifically binds and inhibits thioredoxin reductase 1 to induce apoptosis in colon cancer, Cancer Letters (2017), https://doi.org/10.1016/j.canlet.2017.10.006

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

CAN13549_proof ■ 15 October 2017 ■ 6/13

6

W. Zhuge et al. / Cancer Letters xxx (2017) 1e13

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 Please cite this article in press as: W. Zhuge, et al., Costunolide specifically binds and inhibits thioredoxin reductase 1 to induce apoptosis in colon cancer, Cancer Letters (2017), https://doi.org/10.1016/j.canlet.2017.10.006

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

CAN13549_proof ■ 15 October 2017 ■ 7/13

W. Zhuge et al. / Cancer Letters xxx (2017) 1e13

CTD induces ROS accumulation in human colon cancer cells Elevated ROS levels have been identified as mechanisms behind the activity of many chemotherapeutic drugs [17]. CTD has also been shown to induce ROS in cancer cells [10,12e14]. To test whether CTD inhibited colon cancer growth through elaboration of ROS, we treated the three colon cancer cells with 30 mM CTD for varying time periods. Cells were then stained with 20 ,70 -dichlorofluorescin diacetate (DCFH-DA). DCFH-DA is rapidly de-esterified inside cells and is subsequently oxidized to fluorescence DCF in the presence of ROS. Measurement of DCF, therefore, allows for quantitative detection of ROS. Our results show that CTD rapidly increases ROS levels in colon cancer cells (Fig. 2A) and this induction is concentration-dependent (Fig. 2B, Supplementary Fig. S1A). We next assessed whether this increase in DCF fluorescence, and therefore ROS, is suppressed when cells are treated with an antioxidant. We pretreated colon cancer cells with N-acetyl cysteine before exposing the cells to CTD and measured ROS levels. NAC is commonly used as a precursor of glutathione (GSH) and can also interact directly with ROS and nitrogen species as a scavenger [33]. As expected, NAC pretreatment decreased DCF signal indicating reduction in intracellular ROS levels (Supplementary Fig. S1B). We then determined whether the negative growth signals from CTD can be attenuated if ROS levels are decreased. Pretreatment of cells with NAC was able to normalize CTD-induced growth arrest (Fig. 2C, Supplementary Fig. S1C) as well as induction of apoptosis (Fig. 2D, Supplementary Fig. S1D). Normalization of tumor cell growth by NAC was reflected in enhanced colony formation (Fig. 2E). Furthermore, NAC prevented CTD-induced reductions in G2/M-phase proteins (Fig. 2F) and alterations of apoptosis-related proteins (Fig. 2F). These findings indicate that CTD inhibits colon cancer growth through the generation of ROS as suppressing ROS levels normalized CTD-induced cytotoxicity. Endoplasmic reticulum stress pathway contributes to CTD-mediated apoptosis Previously, we have shown that ROS enhancing agents mediate downstream damage in cancer cells though perturbations in the endoplasmic reticulum (ER) stress pathway [34e36]. ER stress response or unfolded protein response (UPR) induces protein kinase R (PKR)-like endoplasmic reticulum kinase (PERK)-mediated phosphorylation of eIF2a. Phosphorylated-eIF2a then blocks protein translation but allows translation of ATF4. ATF4 is a key transcription factor in the ER stress pathway and mediates the induction of the pro-death transcriptional regulator CHOP. Therefore, we tested whether CTD-mediated ROS activates the ER stress pathway in colon cancer cells and utilized western blotting to probe for key ER stress proteins. Our results show rapid induction of peIF2a, ATF4 and CHOP in HTC-116 cells following exposure to CTD (Fig. 3A). Analysis further showed that this induction is consistent in all three colon cancer cell lines (Fig. 3B). Moreover, CTDmediated ROS was clearly inducing ER stress in cells as pretreatment with NAC normalized the induction of p-eIF2a and ATF4, as well as the downstream CHOP (Fig. 3C). Electron microscopy showed that a short 5-h CTD challenge is sufficient to cause ER

7

swelling (Fig. 3D). This morphological alteration was not seen in cells pretreated with NAC, confirming the immunoblotting data. Electron microscopy also revealed swollen mitochondria with disrupted cristae in cells exposed to CTD (Supplementary Fig. S2). This finding is not surprising given the bidirectional relationship between ER stress pathway and mitochondrial dysfunction. To building on our findings and to confirm a causal role of ERstress pathway in colon cancer apoptosis, we altered ATF4 levels in cells and assessed the effect of CTD. As mentioned earlier, ATF4 is a key transcription factor in the ER-stress pathway and induces the essential CHOP protein to execute ER-stress. We knocked ATF-4 protein down using three siRNA sequences which target ATF-4 mRNA (Fig. 3E and F). If ATF4 was involved in playing a role in CTD-mediated cell death, then knocking it down would be expected to reduce the cytotoxicity associated with CTD. As expected based on our hypothesis, ATF4 knockdown by No.1 siRNA sequence reduced apoptotic cell death by CTD and normalized viability (Fig. 3G and H). To avoid the possible off-target effect of siRNA sequence, we used a distinct sequence (No.2) to silence ATF-4 gene. As shown in the Supplementary Fig. S3, similar results were observed. The No.2 siRNA was also able to silence ATF-4 expression and block CTD-induced cell death. CTD binds and inactivates TrxR1 in human colon cancer cells Overexpression of TrxR1 in human cancers [19], as well as recent studies showing that TrxR system contributes to tumor cell resistance to oxidative stress [18], prompted us to determine whether TrxR1 is a target of CTD. Supporting this notion is a recent study which showed that dehydrocostus lactone, with similar in vitro activity profile as CTD, inhibited the activity of TrxR1 in HeLa cells [37]. We first assessed whether CTD binds to TrxR1 and we utilized surface plasmon resonance (SPR) to achieve this. Our results show that CTD does bind recombinant human (rh) TrxR1 in a dose-dependent manner (Fig. 4A). These results indicate that TrxR1 may one of the direct targets of CTD. We carried out molecular simulation docking of CTD-TrxR1 complex using AutoDock. As shown in Fig. 4B, our molecular docking results showed that CTD fits into the C-terminal redox center (active site) of TrxR1. The bulky 10-membered carbocyclic ring part forms hydrophobic interactions with the surrounding residues. Additionally, the oxygen atom of lactone ring interacts with the key residues Gln-494 through the formation of hydrogen bonds. Therefore, this simulation suggests that CTD binds and blocks the active site of TrxR1. To test this empirically, we evaluated the direct inhibitory effects of rhTrxR1 protein activity by using the 5,50 -dithiobis(2-nitrobenzoate) (DTNB) assay. The result showed decreasing activity of TrxR1 in a CTD-dose-dependent manner (Fig. 4C). We confirmed these results by measuring TrxR1 enzyme activity in lysates prepared from HCT116 cells. Consistent with the cell-free assay, TrxR1 activity in HCT116 cell lysates was reduced with increasing concentrations of CTD (Fig. 4D). These results show that CTD directly targets TrxR1 and inhibits its activity. We next determined whether overexpression of TrxR1 which occurs in cancers, would dampen the cytotoxic effect of CTD. We performed transfections to increase TrxR1 expression in HCT-

Fig. 3. CTD induces apoptosis in colon cancer cells by ROS-dependent ER stress pathway. (A) Expression of ER-stress pathway in colon cancer cells as assessed by protein induction of phosphorylated eIF2a, ATF4, and CHOP. Cells were exposed to 30 mM CTD for different time periods. eIF2a and GAPDH served as controls. (B) Western blot analysis of ER-stress pathway associated proteins in cells exposed to various concentrations of CTD for 3 h (ATF-4 and p-EIF2a) or 8 h (CHOP). (C) Effect of NAC pretreatment on CTD-induced ER stress pathway proteins. NAC was used at 5 mM for 1 h before exposure to CTD. (D) Electron microscopy images of HTC-116 cells exposed to CTD [8000 and 25,000 shown]. Cells were exposed to 30 mM CTD for 5 h. (E) Western blot analysis of ATF4 protein following No.1 siRNA transfection in HTC-116 cells [Negative Control ¼ negative control siRNA transfected cells treated with vehicle, Veh þ CTD ¼ negative control siRNA transfected cells treated with CTD, siRNA ¼ ATF4 siRNA transfected cells treated with vehicle, siRNA þ CTD ¼ ATF4 siRNA transfected cells treated with CTD]. (F) Densitometric quantification for panel F [5 independent experiments; *p < 0.05]. (G) Effect of ATF4 knockdown on CTD-induced apoptosis as assessed by Annexin V/PI staining. HTC-116 cells were transfected with ATF4 siRNA for 24 h and then exposed to 30 mM CTD. (H) Assessment of cell viability in HTC-116 cells following knockdown of ATF4 and exposure to 30 mM CTD. Viability was assessed by MTT assay [***p < 0.001].

Please cite this article in press as: W. Zhuge, et al., Costunolide specifically binds and inhibits thioredoxin reductase 1 to induce apoptosis in colon cancer, Cancer Letters (2017), https://doi.org/10.1016/j.canlet.2017.10.006

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

CAN13549_proof ■ 15 October 2017 ■ 8/13

8

W. Zhuge et al. / Cancer Letters xxx (2017) 1e13

Fig. 4. CTD binds to and inactivates TrxR1. (A) Surface plasmon resonance (SPR) assay to determined the direct binding of CTD to rhTrxR1. rhTrxR1 was added to different concentration of CTD and binding affinities were measured. (B) Molecular interaction between CTD and TrxR1 was simulated by docking software. (C, D) TrxR1 enzyme activity inhibition by CTD as determined by DTNB assay utilizing rhTrxR1 (C) and in HTC-116 lysates (D) as determined by end-point insulin reduction assay. (E) Western blotting analysis of stable overexpression of TrxR1 protein in HCT-116 cells after TXNRD1 plasmid transfection [Control ¼ no transfection, control plasmid ¼ control vehicle vector, TrxR1 ¼ TXNRD1 plasmid transfection]. (F) Quantification of TrxR1 protein levels from panel E. (G) TrxR1 overexpressing cells and vector control transfected cells were exposed to CTD and the cell viability was measured by MTT assay. IC50 values are shown. (H) TrxR1 overexpressing cells and control cells were exposed to 30 mM CTD and the intracellular ROS levels were measured by DCF fluorescence.

Please cite this article in press as: W. Zhuge, et al., Costunolide specifically binds and inhibits thioredoxin reductase 1 to induce apoptosis in colon cancer, Cancer Letters (2017), https://doi.org/10.1016/j.canlet.2017.10.006

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

CAN13549_proof ■ 15 October 2017 ■ 9/13

W. Zhuge et al. / Cancer Letters xxx (2017) 1e13

116 cells (Fig. 4E and F). Overexpression TrxR1 upon CTD exposure reduced CTD-induced growth inhibition and ROS increasing (Fig. 4G and H). These findings demonstrate that CTD binds and inactivates TrxR1 in human colon cancer cells. Deficits in TrxR1, in turn, increase ROS and induce colon cancer cell apoptosis, indicating that the anti-cancer activity of CTD is, at least partly, mediated by targeting TrxR1. TrxR1 is up-regulated in colon cancer Above results indicate that TrxR1 may be a therapeutic target for colon cancer. TrxR1 overexpression has been reported in several malignancies and may be associated with aggressive tumor growth and poor survival. However, reports of its involvement in colon cancer are scarce. A recent study showed that half of the colorectal carcinomas specimens examined (5/10) had increased Trx mRNA levels [38]. In addition, TrxR activity were increased in colorectal tumors compared to normal mucosa [38]. We examined TrxR1 mRNA levels (probe ID: 7958174) in colon cancer samples (tumor, n ¼ 20) and normal colon tissues (normal, n ¼ 20) using the public gene expression omnibus (GEO) profile data set (GSE65480). Compared to normal colon specimens, TrxR1 expression was significantly higher in colon cancer specimens (Fig. 5A). We further collected 20 colon cancer specimens from the patients with colon cancer and confirmed TrxR1 levels in biopsy specimens by using both immunochemical staining (Fig. 5B) and western blotting (Fig. 5C). The patient information was listed in the Supplementary Table S1. Compared to patient-paired normal colon tissues (tumor adjacent), colon cancer specimens showed increased TrxR1 immunoreactivity (Fig. 5B and C). These results support the idea that TrxR1 is significantly increased in colon cancer tissues.

9

CTD decreases TrxR1 and inhibits the growth of HCT-116 cell xenografts Our last objective was to confirm TrxR1 targeting by CTD and the resulting antitumor activity in a HCT-116 xenograft model. We implanted HCT-116 cells in athymic nu/nu mice and treated the mice with 7.5 mg kg1 or 15 mg kg1 CTD for 12 days. Both of these tested doses decreased tumor volume and weight in mice (Fig. 6A and B). No reductions in body weights were noted in any of the experimental groups (Fig. 6C). Histological analysis of liver, kidney, and heart tissues did not reveal any pathological alterations upon CTD administration (Supplementary Fig. S4). Analysis of resected tumor specimens showed induction of ATF4 and CHOP proteins (Fig. 6D) which was clearly seen with 15 mg kg1 CTD treatment. In addition to this readout signifying ER stress pathway activation, CTD increased the levels of cleaved/active caspase-3 (Fig. 6D). Immunohistochemical staining for cleaved caspase-3 also revealed robust immunoreactivity in tumors from mice that received 15 mg kg1 CTD (Fig. 6E). As our in vitro studies showed that CTD inhibited TrxR1 and activated ROS, we assessed ROS level and TrxR1 activity in tumor specimens from mice. As direct measures of ROS levels in tumor specimens, we stained the tissue sections with dihydroethidium (DHE) and DCFH-DA. DHE forms a red fluorescent product upon reaction with ROS and intercalates with DNA. We show that CTD increases ethidium fluorescence indicating increased ROS levels in tumor specimens (Fig. 6F). Similar results were obtained with DCF fluorescence, confirming increased ROS generation in tumor cells following CTD treatment (Fig. 6F). To provide quantitative measure of ROS in tumor specimens, we assessed lipid peroxidation product malondialdehyde (MDA). MDA level is a useful quantitative

Fig. 5. TrxR1 is up-regulated in colon cancer. (A) TrxR1 mRNA (probe ID 7958174) levels in human colon cancer tissues and normal colon specimens [n ¼ 20; GEO set, GSE65480]. (B) Representative immunohistochemical staining for TrxR1 in colon cancer tissues (T) and adjacent normal colon tissues (N) paired from the patients. (C) Western blot analysis of TrxR1 protein levels is different colon cancer tissues (T) and patient-matched, adjacent normal colon tissues (N). GAPDH used as loading control.

Please cite this article in press as: W. Zhuge, et al., Costunolide specifically binds and inhibits thioredoxin reductase 1 to induce apoptosis in colon cancer, Cancer Letters (2017), https://doi.org/10.1016/j.canlet.2017.10.006

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

CAN13549_proof ■ 15 October 2017 ■ 10/13

10

W. Zhuge et al. / Cancer Letters xxx (2017) 1e13

indicator of cellular oxidative stress [39]. Our results show significantly increased MDA levels in tumor lysates prepared from mice treated with CTD (Fig. 6G). Then, Fig. 6H show that CTD decreases TrxR1 activity in colon cancer xenografts. Taken together, these results are in line with our in vitro findings and show that CTD inhibits tumor growth by targeting TrxR1, and triggering ROS production and induction of apoptosis.

Discussion In this study, CTD directly bound and inhibited TrxR1 leading to increased accumulation of ROS in colon cancer cells, which was involved in inducing ER-stress-mediated apoptosis. We also show that TrxR1 levels are elevated in human colon cancer specimens. Finally, CTD is able to inhibit in preclinical human

Fig. 6. CTD inhibits the growth of HCT-116 xenografts by inhibiting TrxR1 and inducing oxidative stress. (A) HCT-116 cells were implanted in nude mice. Mice were then treated with CTD for 12 days. Graph showing tumor volume at indicated time periods. (B) Weights of tumor tissues resected from mice at the end of the treatment period. (C) Body weights of mice receiving CTD treatment at various intervals. (D) Western blot analysis of ATF4, CHOP, and cleaved-caspase 3 levels in resected tumor tissues. GAPDH was used as loading control. (E) Representative staining images of cleaved-caspase 3 in tumor tissues. Reactivity is seen as brown. (F) Staining of tumor tissue sections for DHE (red) and DCFH-DA (green). Tissues were counterstained with DAPI (Blue). Increased fluorescence intensity is indicative of increased ROS levels. (G) Levels of MDA in tumor tissue lysates showing increased lipid peroxidation product levels upon treatment with CTD [***p < 0.001 compared to vehicle treated group]. (H) Activity of TrxR1 in tumor tissue lysates as determined by end-point insulin reduction assay. [*p < 0.05, **p < 0.01, and ***p < 0.001, compared to vehicle treated mice]. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Please cite this article in press as: W. Zhuge, et al., Costunolide specifically binds and inhibits thioredoxin reductase 1 to induce apoptosis in colon cancer, Cancer Letters (2017), https://doi.org/10.1016/j.canlet.2017.10.006

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

CAN13549_proof ■ 15 October 2017 ■ 11/13

W. Zhuge et al. / Cancer Letters xxx (2017) 1e13

colon cancer xenograft model. These salient findings are summarized in Fig. 7. The pathways utilized by CTD to cause apoptosis may be cancer cell type-specific. For example, Lee and colleagues found that CTD induces apoptosis in HL-60 leukemia cells by mitochondrial permeability transition and cytochrome C [40]. Similar mechanism was also shown in bladder cancer cells [12] and ovarian cancer cells [14]. In contrast, Choi and colleagues found that CTD activated Fas, caspase-8, caspase-3, and cleavage of PARP in breast cancer cells [13]. They also showed that CTD did not disrupt mitochondrial membrane potential or produce any changes in Bcl-2 and Bax proteins [13]. It should be noted that the mitochondrial intrinsic pathway, death receptor-mediated extrinsic pathway, and the ER stress pathway are not fully exclusive. The three pathways are linked in that molecules in one pathway can influence the other [41]. Glimpse of this are clearly evident in our study. We found that CTD caused apoptotic cell death through alteration of Bcl2/Bax, activation of caspase-3, and cleavage of PARP. We also found disruption of ER in colon cancer cells. It is interesting to note that mitochondrial swelling was also observed in cells exposed to CTD. Therefore, it appears that multiple apoptotic mechanisms may be elicited in colon cancer cells. The differential responses of colon cancer cells to CTD compared to other cancer cells are certainly interesting and warrants further studies. One of the significant findings of our study is the direct binding and inhibition of TrxR1 by CTD. We used SPR to show quick association and dissociation of CTD with TrxR1. Molecular docking simulation revealed that CTD binds the active site of TrxR1. This finding is consistent with dose-dependent inhibition of TrxR1 activity observed in our studies. TrxR isoenzymes catalyze the reduction of oxidized thioredoxin by NADPH, and regulate multiple redox-based signaling pathways [42,43]. Inhibition of the Trx/TrxR system causes dramatic imbalance in the formation and the removal of ROS [19]. TrxR is elevated in many human cancers and is

11

associated with aggressive tumor behavior [44]. Our studies also show that human colon cancer specimens exhibit elevated levels of TrxR1 compared to patient-matched, normal colon tissues. Using colon cancer xenografts, we also show that CTD inhibited TrxR1 activity. This end-point was associated with reduced tumor growth and induction of apoptosis, thus, strongly implicating TrxR1 in colon cancer growth and progression. Based on the critical roles it plays in cancer and other diseases, the TrxR system is certainly emerging as a target for anticancer drug development [45,46]. In fact, auranofin, a metal phosphine complex, has been used for the clinical treatment of rheumatoid arthritis [47]. Recently, the mechanism of auranofin was discovered as an inhibitor of TrxR1 [48]. We have also shown that auranofin inhibits the growth of gastric cancer cells [49]. Interestingly, this inhibition of gastric cancer growth was linked to elevated ROS levels. The downstream consequence of TrxR1 inhibition and the central mechanism of colon cancer cell inhibition by CTD was elevated ROS. A body of evidence supports the idea that moderate levels of ROS contribute to tumor development by promoting cell proliferation, angiogenesis, and metastasis [50]. However, forcing excessive ROS generation and oxidative stress is able to produce DNA damage and an abnormal stress response triggering cancer cell death [51]. Many effective pro-oxidant chemotherapeutic agents rely on inducing this additional oxidative stress in cancer cells. This strategy may also offer selectivity. As cancer cells exhibit increased basal levels of ROS, chemotherapeutic agents may be able to reach the deleterious threshold in cancer cells at a much lower dose. Furthermore, this strategy may also circumvent the issue of developing drug resistance as ROS-inducing agents may utilize multiple mechanisms and protein targets. For example, our studies show that CTD inhibits TrxR1 to induce ROS in cancer cells. There are also reports that CTD is able to reduce the levels of nuclear factor-erythroid 2 (NF-E2) p45-related factor-2 (Nrf2) [52] and deplete glutathione levels [53]. Nrf2 regulates the transcription of a plethora of cytoprotective genes including catalase, superoxide dismutase, heme oxygenase-1, NAD(P)H:quinone oxidoreductase 1, glutathione peroxidase-2 and glutathione S-transferase [54,55]. This raises the interesting question whether Nrf2 is another target of CTD or represents a downstream response to TrxR1-mediated redox alteration. These studies warrant further investigation. In summary, our studies show that CTD effectively inhibits colon cancer growth, in vitro and in vivo. We have discovered that TrxR1 is a target of CTD. CTD inhibited TrxR1 activity and induced apoptotic cell death through ROS elevation. ROS increases in colon cancer cells utilized multiple mechanisms of apoptotic death including the activation of the ERstress pathway, alteration of Bcl2/ Bax, and caspases. In addition, using human colon cancer biopsy specimens and analyzing gene databases, we show that TrxR1 is increased in colon cancer. Therefore, TrxR1 may very well be an excellent target to pursue for the development of therapies for colon cancer patients. Our findings also suggest that CTD may have therapeutic utility as it inhibits TrxR1.

Author contributions

Fig. 7. Schematic showing the proposed mechanism of CTD in colon cancer. Our studies show that CTD inhibits TrxR1 activity leading to increased ROS levels in colon cancer cells. Elevated ROS level perturbs ER function leading to the activation of the ER stress pathway/unfolded protein response. ER stress pathway contributes to ROSinduced cell death by activating caspases.

Weishan Zhuge, Ruijie Chen, Xidan Dong, and Miao Bao: conception and design, collection, analysis and interpretation of data, manuscript writing; Xiangwei Sun and Xuanxuan Dai: collection and interpretation of data, manuscript revision; Weishan Zhuge, Xian Shen, and Vladimir Katanaev: collection and analysis of data. Xian Shen, Weishan Zhuge, Khan Zia, and Guang Liang: conception and design, interpretation of data, manuscript revision.

Please cite this article in press as: W. Zhuge, et al., Costunolide specifically binds and inhibits thioredoxin reductase 1 to induce apoptosis in colon cancer, Cancer Letters (2017), https://doi.org/10.1016/j.canlet.2017.10.006

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

CAN13549_proof ■ 15 October 2017 ■ 12/13

12

W. Zhuge et al. / Cancer Letters xxx (2017) 1e13

Disclosure statement The authors disclose no potential conflicts of interest. Acknowledgements

[20]

[21] [22]

The work was supported by National Natural Science Foundation of China [81622043] and Zhejiang Province Natural Science Funding of China [LY17H160055].

[23] [24]

Appendix A. Supplementary data Supplementary data related to this article can be found at https://doi.org/10.1016/j.canlet.2017.10.006.

[25]

[26] [27]

Conflict of interest None. References [1] A. Stein, D. Atanackovic, C. Bokemeyer, Current standards and new trends in the primary treatment of colorectal cancer, Eur. J. Cancer 47 (Suppl 3) (2011) S312eS314. [2] N.A. Dallas, L. Xia, F. Fan, M.J. Gray, P. Gaur, G. van Buren 2nd, et al., Chemoresistant colorectal cancer cells, the cancer stem cell phenotype, and increased sensitivity to insulin-like growth factor-I receptor inhibition, Cancer Res. 69 (2009) 1951e1957. [3] L. Shen, M. Toyota, Y. Kondo, E. Lin, L. Zhang, Y. Guo, et al., Integrated genetic and epigenetic analysis identifies three different subclasses of colon cancer, Proc. Natl. Acad. Sci. U. S. A. 104 (2007) 18654e18659. [4] G.H. Kang, Four molecular subtypes of colorectal cancer and their precursor lesions, Arch. Pathol. Lab. Med. 135 (2011) 698e703. [5] A.I. Phipps, P.J. Limburg, J.A. Baron, A.N. Burnett-Hartman, D.J. Weisenberger, P.W. Laird, et al., Association between molecular subtypes of colorectal cancer and patient survival, Gastroenterology 148 (2015) 77e87 e72. [6] H. Thomas, H.M. Coley, Overcoming multidrug resistance in cancer: an update on the clinical strategy of inhibiting p-glycoprotein, Cancer Control 10 (2003) 159e165. [7] A. Bishayee, Y. Haskell, C. Do, K.S. Siveen, N. Mohandas, G. Sethi, et al., Potential benefits of edible berries in the management of aerodigestive and gastrointestinal tract cancers: preclinical and clinical evidence, Crit. Rev. Food Sci. Nutr. 56 (2016) 1753e1775. [8] T. Kawamori, T. Tanaka, A. Hara, J. Yamahara, H. Mori, Modifying effects of naturally occurring products on the development of colonic aberrant crypt foci induced by azoxymethane in F344 rats, Cancer Res. 55 (1995) 1277e1282. [9] H. Mori, T. Kawamori, T. Tanaka, M. Ohnishi, J. Yamahara, Chemopreventive effect of costunolide, a constituent of oriental medicine, on azoxymethaneinduced intestinal carcinogenesis in rats, Cancer Lett. 83 (1994) 171e175. [10] P. Hua, M. Sun, G. Zhang, Y. Zhang, G. Song, Z. Liu, et al., Costunolide induces apoptosis through generation of ROS and activation of P53 in human esophageal cancer Eca-109 cells, J. Biochem. Mol. Toxicol. 30 (2016) 462e469. [11] Z. Wang, X. Zhao, X. Gong, Costunolide induces lung adenocarcinoma cell line A549 cells apoptosis through ROS (reactive oxygen species)-mediated endoplasmic reticulum stress, Cell Biol. Int. 40 (2016) 289e297. [12] A. Rasul, R. Bao, M. Malhi, B. Zhao, I. Tsuji, J. Li, et al., Induction of apoptosis by costunolide in bladder cancer cells is mediated through ROS generation and mitochondrial dysfunction, Molecules 18 (2013) 1418e1433. [13] Y.K. Choi, H.S. Seo, H.S. Choi, H.S. Choi, S.R. Kim, Y.C. Shin, et al., Induction of Fas-mediated extrinsic apoptosis, p21WAF1-related G2/M cell cycle arrest and ROS generation by costunolide in estrogen receptor-negative breast cancer cells, MDA-MB-231, Mol. Cell Biochem. 363 (2012) 119e128. [14] Y.I. Yang, J.H. Kim, K.T. Lee, J.H. Choi, Costunolide induces apoptosis in platinum-resistant human ovarian cancer cells by generating reactive oxygen species, Gynecol. Oncol. 123 (2011) 588e596. [15] E.O. Hileman, J. Liu, M. Albitar, M.J. Keating, P. Huang, Intrinsic oxidative stress in cancer cells: a biochemical basis for therapeutic selectivity, Cancer Chemother. Pharmacol. 53 (2004) 209e219. [16] S. Toyokuni, K. Okamoto, J. Yo, H. Hiai, Persistent oxidative stress in cancer, FEBS Lett. 358 (1995) 1e3. [17] H.-R. Teppo, Y. Soini, P. Karihtala, Reactive oxygen species-mediated mechanisms of action of targeted cancer therapy, Oxid. Med. Cell. Longev. (2017) 11. [18] S.J. Kim, Y. Miyoshi, T. Taguchi, Y. Tamaki, H. Nakamura, J. Yodoi, et al., High thioredoxin expression is associated with resistance to docetaxel in primary breast cancer, Clin. Cancer Res. 11 (2005) 8425e8430. [19] M.P. Rigobello, V. Gandin, A. Folda, A.K. Rundlof, A.P. Fernandes, A. Bindoli, et al., Treatment of human cancer cells with selenite or tellurite in combination

[28]

[29]

[30]

[31] [32] [33] [34]

[35]

[36]

[37]

[38]

[39]

[40]

[41] [42]

[43] [44]

[45] [46] [47] [48]

[49]

with auranofin enhances cell death due to redox shift, Free Radic. Biol. Med. 47 (2009) 710e721. M.H. Yoo, X.M. Xu, B.A. Carlson, A.D. Patterson, V.N. Gladyshev, D.L. Hatfield, Targeting thioredoxin reductase 1 reduction in cancer cells inhibits selfsufficient growth and DNA replication, PLoS One 2 (2007), e1112. E.S. Arner, A. Holmgren, The thioredoxin system in cancer-introduction to a thematic volume of Seminars in Cancer Biology,, Semin. Cancer Biol. 16 (2006) 419. S.E. Eriksson, S. Prast-Nielsen, E. Flaberg, L. Szekely, E.S. Arner, High levels of thioredoxin reductase 1 modulate drug-specific cytotoxic efficacy, Free Radic. Biol. Med. 47 (2009) 1661e1671. A. Holmgren, M. Bjornstedt, Thioredoxin and thioredoxin reductase, Methods Enzymol. 252 (1995) 199e208. G.M. Morris, R. Huey, W. Lindstrom, M.F. Sanner, R.K. Belew, D.S. Goodsell, et al., AutoDock4 and AutoDockTools4: automated docking with selective receptor flexibility, J. Comput. Chem. 30 (2009) 2785e2791. Y.C. Lo, T.P. Ko, W.C. Su, T.L. Su, A.H. Wang, Terpyridine-platinum(II) complexes are effective inhibitors of mammalian topoisomerases and human thioredoxin reductase 1, J. Inorg. Biochem. 103 (2009) 1082e1092. W.L. DeLano, The PyMOL Molecular Graphics System, DeLano Scientific, San Carlos, CA, 2002. C. Kilkenny, W. Browne, I.C. Cuthill, M. Emerson, D.G. Altman, N.C.R.R.G.W. Group, Animal research: reporting in vivo experiments: the ARRIVE guidelines, Br. J. Pharmacol. 160 (2010) 1577e1579. J.C. McGrath, G.B. Drummond, E.M. McLachlan, C. Kilkenny, C.L. Wainwright, Guidelines for reporting experiments involving animals: the ARRIVE guidelines, Br. J. Pharmacol. 160 (2010) 1573e1576. M.J. Curtis, R.A. Bond, D. Spina, A. Ahluwalia, S.P. Alexander, M.A. Giembycz, et al., Experimental design and analysis and their reporting: new guidance for publication in BJP, Br. J. Pharmacol. 172 (2015) 3461e3471. L. Meng, T. Lin, R.Y. Tsai, Nucleoplasmic mobilization of nucleostemin stabilizes MDM2 and promotes G2-M progression and cell survival, J. Cell Sci. 121 (2008) 4037e4046. E.A. Nigg, Cyclin-dependent protein kinases: key regulators of the eukaryotic cell cycle, Bioessays 17 (1995) 471e480. G.M. Cohen, Caspases: the executioners of apoptosis, Biochem. J. 326 (Pt 1) (1997) 1e16. M. Zafarullah, W.Q. Li, J. Sylvester, M. Ahmad, Molecular mechanisms of Nacetylcysteine actions, Cell Mol. Life Sci. 60 (2003) 6e20. J. Zhang, Z. Feng, C. Wang, H. Zhou, W. Liu, K. Kanchana, et al., Curcumin derivative WZ35 efficiently suppresses colon cancer progression through inducing ROS production and ER stress-dependent apoptosis, Am. J. Cancer Res. 7 (2017) 275e288. C. Feng, Y. Xia, P. Zou, M. Shen, J. Hu, S. Ying, et al., Curcumin analog L48H37 induces apoptosis through ROS-mediated endoplasmic reticulum stress and STAT3 pathways in human lung cancer cells, Mol. Carcinog. 56 (2017) 1765e1777. Y. Hu, C. Zhao, H. Zheng, K. Lu, D. Shi, Z. Liu, et al., A novel STAT3 inhibitor HO3867 induces cell apoptosis by reactive oxygen species-dependent endoplasmic reticulum stress in human pancreatic cancer cells, Anticancer Drugs 28 (2017) 392e400. M. Yang, J. Zhang, Y. Li, X. Han, K. Gao, J. Fang, Bioassay-guided isolation of dehydrocostus lactone from Saussurea lappa: a new targeted cytosolic thioredoxin reductase anticancer agent, Arch. Biochem. Biophys. 607 (2016) 20e26. M. Berggren, A. Gallegos, J.R. Gasdaska, P.Y. Gasdaska, J. Warneke, G. Powis, Thioredoxin and thioredoxin reductase gene expression in human tumors and cell lines, and the effects of serum stimulation and hypoxia, Anticancer Res. 16 (1996) 3459e3466. K. Gong, W. Li, Shikonin, a Chinese plant-derived naphthoquinone, induces apoptosis in hepatocellular carcinoma cells through reactive oxygen species: a potential new treatment for hepatocellular carcinoma, Free Radic. Biol. Med. 51 (2011) 2259e2271. M.G. Lee, K.T. Lee, S.G. Chi, J.H. Park, Costunolide induces apoptosis by ROSmediated mitochondrial permeability transition and cytochrome C release, Biol. Pharm. Bull. 24 (2001) 303e306. S. Elmore, Apoptosis: a review of programmed cell death, Toxicol. Pathol. 35 (2007) 495e516. E.S. Arner, Focus on mammalian thioredoxin reductaseseimportant selenoproteins with versatile functions, Biochim. Biophys. Acta 1790 (2009) 495e526. A. Matsuzawa, Thioredoxin and redox signaling: roles of the thioredoxin system in control of cell fate, Arch. Biochem. Biophys. 617 (2017) 101e105. D.T. Lincoln, E.M. Ali Emadi, K.F. Tonissen, F.M. Clarke, The thioredoxinthioredoxin reductase system: over-expression in human cancer, Anticancer Res. 23 (2003) 2425e2433. P. Nguyen, R.T. Awwad, D.D. Smart, D.R. Spitz, D. Gius, Thioredoxin reductase as a novel molecular target for cancer therapy, Cancer Lett. 236 (2006) 164e174. S. Urig, K. Becker, On the potential of thioredoxin reductase inhibitors for cancer therapy, Semin. Cancer Biol. 16 (2006) 452e465. I.C. Shaw, Gold-based therapeutic agents, Chem. Rev. 99 (1999) 2589e2600. V. Gandin, A.P. Fernandes, M.P. Rigobello, B. Dani, F. Sorrentino, F. Tisato, et al., Cancer cell death induced by phosphine gold(I) compounds targeting thioredoxin reductase, Biochem. Pharmacol. 79 (2010) 90e101. P. Zou, M. Chen, J. Ji, W. Chen, X. Chen, S. Ying, et al., Auranofin induces apoptosis by ROS-mediated ER stress and mitochondrial dysfunction and

Please cite this article in press as: W. Zhuge, et al., Costunolide specifically binds and inhibits thioredoxin reductase 1 to induce apoptosis in colon cancer, Cancer Letters (2017), https://doi.org/10.1016/j.canlet.2017.10.006

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

1 2 3 4 5 6 7 8 9

CAN13549_proof ■ 15 October 2017 ■ 13/13

W. Zhuge et al. / Cancer Letters xxx (2017) 1e13 displayed synergistic lethality with piperlongumine in gastric cancer, Oncotarget 6 (2015) 36505e36521. [50] G. Waris, H. Ahsan, Reactive oxygen species: role in the development of cancer and various chronic conditions, J. Carcinog. 5 (2006) 14. [51] D. Trachootham, J. Alexandre, P. Huang, Targeting cancer cells by ROSmediated mechanisms: a radical therapeutic approach? Nat. Rev. Drug Discov. 8 (2009) 579e591. [52] F. Ahmad, D. Dixit, V. Sharma, A. Kumar, S.D. Joshi, C. Sarkar, et al., Nrf2-driven TERT regulates pentose phosphate pathway in glioblastoma, Cell Death Dis. 7 (2016), e2213.

13

[53] E. Butturini, E. Cavalieri, A.C. de Prati, E. Darra, A. Rigo, K. Shoji, et al., Two naturally occurring terpenes, dehydrocostuslactone and costunolide, decrease intracellular GSH content and inhibit STAT3 activation, PLoS One 6 (2011), e20174. [54] T.W. Kensler, N. Wakabayashi, S. Biswal, Cell survival responses to environmental stresses via the Keap1-Nrf2-ARE pathway, Annu. Rev. Pharmacol. Toxicol. 47 (2007) 89e116. [55] M. Kobayashi, M. Yamamoto, Nrf2-Keap1 regulation of cellular defense mechanisms against electrophiles and reactive oxygen species, Adv. Enzyme Regul. 46 (2006) 113e140.

Please cite this article in press as: W. Zhuge, et al., Costunolide specifically binds and inhibits thioredoxin reductase 1 to induce apoptosis in colon cancer, Cancer Letters (2017), https://doi.org/10.1016/j.canlet.2017.10.006

10 11 12 13 14 15 16 17 18