Parthenolide Induces Oxidative Stress and Impedes Cell Migration by SuppressingWnt Pathway and Epithelial-Mesenchymal-Transition (EMT) in HCT-116 Metastatic Colorectal Cancer Cells

Colorectal cancer (CRC) is a vital cause of cancer morbidity and mortality. 50% of CRC patients suffer from an aggressive metastatic disease which ultimately fallout in death. Inmetastatic cancer, tumour cellsmigrate, invade, and inally colonise to the distant organ by degrading their attachments with the extracellular matrix. Parthenolide (PTL) is a secondary metabolite of feverfew (Tanacetum parthenium) plant. It shows its cytotoxic effect towards cancer cells via different cellular signalling pathways like inhibition of NF-κB, STAT3, MAPK, JNK pathways, activation of p53 etc. In the present study, we have assessed anti-cancer and anti-metastatic potential of PTL against human HCT-116 metastatic colorectal cancer cells. Analysis of cellular oxidative status (GSH/GSSG) of PTL treated HCT-116 cells showed a signi icant decrease (p<0.05) in GSH level while GSSG level was increased signi icantly (p<0.05) on PTL treatment. PTL also increased the amount of intracellular reactive oxygen species. The qRT-PCR analysis revealed that PTL down-regulates cfos, c-jun and N-cadherin expression and up-regulates E-cadherin expression indicating inhibition of cell migration and metastasis by EMT pathway. PTL inhibited the MMP-9 expression in a dose-dependent fashion and inhibited cancer cell migration by regulating Wnt/β-catenin signalling through the upregulation of DKK-1 protein expression indicating PTL has a promising anticancer potential against HCT-116 metastatic colorectal carcinoma cells.

Parthenolide, HCT-116, Epithelial-Mesenchymal Transition, Wnt Signalling, Colorectal Cancer, Metastasis A Colorectal cancer (CRC) is a vital cause of cancer morbidity and mortality. 50% of CRC patients suffer from an aggressive metastatic disease which ultimately fallout in death. In metastatic cancer, tumour cells migrate, invade, and inally colonise to the distant organ by degrading their attachments with the extracellular matrix. Parthenolide (PTL) is a secondary metabolite of feverfew (Tanacetum parthenium) plant. It shows its cytotoxic effect towards cancer cells via different cellular signalling pathways like inhibition of NF-κB, STAT3, MAPK, JNK pathways, activation of p53 etc. In the present study, we have assessed anti-cancer and anti-metastatic potential of PTL against human HCT-116 metastatic colorectal cancer cells. Analysis of cellular oxidative status (GSH/GSSG) of PTL treated HCT-116 cells showed a signi icant decrease (p<0.05) in GSH level while GSSG level was increased signi icantly (p<0.05) on PTL treatment. PTL also increased the amount of intracellular reactive oxygen species. The qRT-PCR analysis revealed that PTL down-regulates cfos, c-jun and N-cadherin expression and up-regulates E-cadherin expression indicating inhibition of cell migration and metastasis by EMT pathway. PTL inhibited the MMP-9 expression in a dose-dependent fashion and inhibited cancer cell migration by regulating Wnt/β-catenin signalling through the upregulation of DKK-1 protein expression indicating PTL has a promising anticancer potential against HCT-116 metastatic colorectal carcinoma cells.

INTRODUCTION
Colorectal cancer (CRC) causes a global crisis in terms of morbidity and mortality. It is the 3 rd most fatal and 4 th most recurrently identi ied cancer worldwide. 50% of CRC patients suffer from the metastatic hepatic disease over their lifetime, which ultimately fallout in death for more than two-thirds of these patients (Bray et al., 2018). Since there has been no remarkable development for metastatic CRC management, healing rate has remained near to the ground over the decades. Aggressive local invasion and metastasis have made colon cancer challenging to deal. Therefore, illuminating the mechanism of invasive metastasis of CRC and novel drug discovery that target colorectal cancer and its metastasis-related diseases, is urgently needed. Metastasis is a complex process during which cancer cells aid their migration, invasion, and inal colonisation in distant sites by degrading their anchorage with the extracellular matrix (ECM). Matrix metal-loproteinase -2/-9 (MMP-2/-9) perform an essential role in the degradation of ECM and involvement in cancer metastasis (Gonzalez-Avila et al., 2019). Epithelial-mesenchymal transition (EMT), a process, which plays a vital role in cancer progression and metastasis. EMT ease the metastasis by progressing the epithelial cancer cells to impair their cell-cell and cell-ECM adhesion characteristics and to acquire migratory and invasive characteristics (Lin and Wu, 2020).
Sesquiterpene lactones are secondary metabolites isolated from plants of Asteraceae family, Tanacetum parthenium and have been used to treat certain ailments traditionally. Parthenolide (PTL) is one of the active sesquiterpene lactones of this plant. PTL contains two functional groups, one epoxide group and one α-methylene-γ-lactone ring, which readily react at the nucleophilic region of biomolecules (Dey et al., 2016). Parthenolide has shown their cytotoxicity or induces apoptosis in different cancer cells. PTL shows its anti-cancer activity via different cellular signalling pathways viz inhibition of NFκB, STAT3, MAPK, JNK pathways, activation of p53, suppression of nucleic acid synthesis pathway, depletion of intercellular thiols, induction of oxidative stress, promoting mitochondrial dysfunction, interference of cellular calcium homeostasis, cell cycle arrest at G2/M phase, depletion of HDAC1 and inhibits tubulin carboxypeptidase activity (Zhang et al., 2004;Siyuan et al., 2004). Literature data showed that PTL selectively targets cancer cells. It has been reported that PTL speci ically targets stem cells of acute myelogenous leukaemia (AML) and their progenitor cells without damaging normal hematopoietic cells (Guzman et al., 2005;Baranello et al., 2015). However, how Parthenolide affects cancer metastasis is not well understood. In this present study, we have assessed the antimigratory/anti-metastatic potential of PTL against human HCT-116 colorectal cancer cells.
The antibodies against DKK-1, MMP-9, GAPDH, anti-Rb secondary with HRP and anti-Mouse secondary with HRP conjugate, ECL reagent were purchased from Abcam, USA and Bio-Rad, USA.

Preparation of Drug
A primary stock solution of Parthenolide was prepared using absolute ethanol as a solvent and kept at -20°C. The inal drug concentrations were prepared at the time of treatment by diluting the stock solution with media.

Cell lines maintenance and culture
The HCT-116 colon cancer cell line was a gift from Dr Sanjay Ghosh, Professor, University of Calcutta (India). The cell line was maintained in DMEM (high glucose) complete media added with 10% heatinactivated FBS, along with L-glutamine (2 mM), penicillin (100 U/ml), streptomycin (100 µg/ml) in 5% CO 2 and 95% humidi ied conditions maintaining 37 • C. Cells were cultured to the exponential phase until the number of cells grown up to 1.0 x 10 6 cells/ml.

Fluorescence microscopic study using Acridine Orange and Ethidium Bromide
To determine cell death, we used Acridine Orange (AO) and Ethidium Bromide (EtBr).
In each well, 2x10 4 HCT-116 cells were seeded in a 6well plate and kept at 37 • C for 24 hrs CO 2 incubator. PTL treatment was done at 5µg/ml dose and 10µg/ml dose for 48 hours. After the treatment, cells were trypsinised and washed with PBS. 10µl of the cells were then pipetted and placed on a clean glass slide and mixed with 10µl each of acridine orange (50µg/ml) and ethidium bromide (50µg/ml) (Gomes et al., 2011).
The stained cells were observed using the EVOS ® FL Cell Imaging System (Life Technologies, USA) at 400X magni ication.

Intracellular ROS measurement by Fluorescence microscopy
The quantity of reactive oxygen species generation was assessed using H 2 DCFDA staining as per the method by Chattopadhyay et al. (2014). In brief, experimental cells (2 x10 5 cells/mL) were treated with PTL for 48 h. There was a positive control set where experimental cells were treated with H 2 O 2 (100 mM) for 30 min before estimation. After 48 h of treatment, cells were washed, followed by 30 min incubation with H 2 DCFDA (1 mg/ml ) at 37 • C. Next, the cells were washed thrice with the media. Cells were observed under the EVOS ® FL Cell Imaging System (Life Technologies, USA). All measurements were done in triplicate.

Reduced Glutathione (GSH) measurement
Reduced GSH of the cell lysates were estimated as stated by Halder et al. (2018). To the cell lysate, 25% of trichloroacetic acid was mixed with and centrifuged at 2000 RPM for 15 min, and the protein sediments were collected. The supernatants were aspirated, and they were diluted up to 1 ml using 0.2 M sodium phosphate buffer (pH 8.0). Then, 2 ml of 0.6 mM DTNB (Ellman's reagent) was added to each of the vials. The experimental set up was incubated at 37 • C for 10 min. After the reaction of GSH with DTNB, the absorbance was observed at 405 nm. Readings were taken using a standard curve prepared with different doses of standard GSH and expressed as µg of GSH per mg of protein. All measurements were done in triplicate.

Oxidised Glutathione (GSSG) measurement
The oxidised glutathione (GSSG) levels of the cell lysates were evaluated on the production of derivatised intermediate after the reaction of GSH with 2-vinyl pyridine as per Mahapatra et al. (2009). In brief, 0.5 ml sample and 2 µL, 2-vinyl pyridine reagent was added and kept at 37 • C for 1 h. 4% sulfosalicylic acid (SSA) was added to this for deproteinisation and centrifuged for 1 min at 1,000×g. The supernatants were collected, and GSSG concentrations were determined after reacting them with DTNB at 412 nm and compared with standard GSSG curve. Respective GSSG concentrations were calculated and expressed as µg of GSSG per mg of protein.
All measurements were done in triplicate.

Wound Healing Assay
On an 80% con luent six-well plate, straight scratches were made across the diameter of the well using P100 pipette tip. Cells, that detached were washed using PBS, and fresh serum-free medium were added with different doses of PTL.
Cell migration towards the wound areas was observed after 48 h. The ability of the cells to heal the wounds were assessed by measuring the ultimate areas of healed wounds. Images were captured in Olympus IX71 microscope at 100x magni ication. Each scratch area was measured thrice using ImageJ software.

Analysis of Gene Expression by q-RT-PCR
Total cellular RNA was isolated using 'RNeasy ® Mini Kit (250)' (Qiagen, Netherlands) and RNA was measured using Eppendorf BioSpectrometer ® . Then an equal volume of each experimental samples was reverse-transcribed to form cDNA using 'High-Capacity cDNA Reverse Transcription Kit' (Invitrogen TM , USA) following the manufacturer's instruction. The cDNAs were then used for the qPCR analysis of gene expression using GAPDH as housekeeping control. The list of primers used for the qPCR reactions was mentioned in Table 1, and ampli ication genes were performed in CFX96 Touch Real-Time PCR Detection System (Bio-Rad, USA) using SsoAdvanced TM Universal SYBR ® Green Supermix (Bio-Rad Laboratories, USA)for 30 cycles. The threshold cycle (CT) readings were determined using the equation of ∆CT = CT (target) − CT (endogenous control) and fold changes were measured as 2 −∆(∆CT ) . The results were plotted on the graph for the presentation.

Western blot analysis
Western Blot study was done by our previously standardised lab protocol (Sarkar et al., 2017). Experimental cells were seeded for treatment on 60 mm plates (Himedia, India), washed with Phosphate Buffer Saline before lysis with RIPA buffer consisting of protease inhibitor cocktail. The cell lysates were then centrifuged for 30 minutes at 10000 rpm at 4°C and supernatants were collected.Protein concentrations were estimated using Bradford reagent taking bovine serum albumin as standard, and samples were diluted at 1:1 ratio using sample buffer (consisting of 4% SDS, 0.5 M Tris-HCl (pH 6.8), 20% glycerol and 0.002% bromophenol blue). Samples containing 50 µg of total protein (solubilised in sample buffer) were resolved SDS-PAGE gel and then electro transblotted to a Polyvinylidene di luoride (PVDF) membrane. Then the PVDF membrane was blocked with 5% non-fat dried milk (Himedia, India). Membranes were incubated with the DKK1 and MMP-9 antibodies. After washing, membranes were incubated with HRP-conjugated secondary antibody and developed with ECL reagent and observed under ChemiDoc XRS+ System (Bio-Rad, USA).

Statistical Analysis
The data were represented as mean ± SEM, n=3. One-way ANOVA test (using a statistical package, Origin 6.1, Northampton, MA 01060 USA) was done to compare among the means of control and treated groups. P < 0.05 was considered as signi icance limit.

RESULTS
Cytotoxicity of PTL against HCT-116 cells was tested by MTT assay after 48 h treatment. The percentage of HCT-116 cancer cell death signi icantly reduced upon treatment of PTL. The 50% inhibitory concentration (IC 50 ) value of PTL in HCT-116 cells was determined using different doses (1, 5, 10, 50, 100 µg/ml) of the drug for 48 h. Respective cell death percentages were plotted in a graph using Statistica software. The IC 50 value of PTL in HCT-116 cells was noted as 12.19 µg/ml. After 48 h, morphologies of cells were altered drastically. PTL induced dose-dependent cell death which was observed at 48 h under phase-contrast microscopy (Magni ication 200x). ( Figure 1A, Figure 1B)

Fluorescence microscopic analysis
The luorescence images after PTL treatment observed with EtBr-AO staining. These typical luorescent dye stains in such a way that the healthy cells with undamaged DNA give green luorescence, late apoptotic/necrotic cells having fragmented DNA emit orange/red-coloured luorescence. Our experimental result shows that PTL treatment decreased the number of viable cells and an increased number of apoptotic/necrotic cell population. The signi icant population of HCT-116 cells stained with orange colour indicates cellular apoptosis in the PTL treated groups in a dose (5µg/ml and 10µg/ml) dependent manner. ( Figure 1C)

Determination of Cellular Reactive Oxygen Species
The luorescence intensity of Dichloro-dihydroluorescein diacetate (DCFH 2 -DA) represents ROS generation. In luorescence microscopic image analysis, it was observed that in HCT-116 cells, DCFH 2 -DA luorescence intensity was elevated signi icantly (p<0.05) by PTL treatment. Results presented as the mean of three experiments. Values were represented Mean ±SEM (p<0.05 signi icant). (Figure 2A, Figure 2B)

Measurement of Cellular GSH and GSSG
Reduced Glutathione (GSH) level was estimated using cell lysate of different groups. The result shows a 33.33% decrease in GSH level in 5µg/ml PTL treated group while comparing with the control. In 10µg/ml PTL treatment group, the level of GSH was decreased by 58.3% compared to that of control. PTL treatment increased GSSG level significantly (p<0.05) in HCT-116 cell lysate. It has been observed that PTL (5µg/ml and 10µg/ml) treatment upsurge GSSG level by 75% and 87.5% respectively while comparing with the control. ( Figure 2C, Figure 2D)

Wound Healing Assay
PTL at concentrations of 5µg/ml and 10µg/ml promote dose-dependent inhibition of migratory properties of HCT-116 cells. The gap did not achieve full wound closure for both doses. PTL treatments after 48 h with a dose of 5µg/ml and 10µg/ml demonstrated that the percentage of recovery was decreased with respect to control. The cell migration assay was performed by wound closure by the migrated cells initial (0 h) and inal (48 h) wound area of the scratches, and they were measured by ImageJ (NIH) software. In 5µg/ml of PTL treatment in migrating cells, 30% of the wound healing was inhibited as compared with the control. But in 10µg/ml treatment setup, the wound area was increased by 40% as compared to the initial wound created at zero hour of the experiment due to extensive cell death. (Figure 3)

qRT-PCR analysis of the expression of transcription factors and genes associated with EMT
Treatment with PTL (10µg/ml) for 48 h in HCT-116 cells showed a down-regulated expression of c-fos, cjun and N-cadherin whereas it induces up-regulated expression of E-cadherin. The igure showed expression of different genes in the RNA level when the results of PTL treated HCT-116 cells were compared to that of their respective controls. Results are represented as levels of mRNA expressions equalised with GAPDH as the control. Our data showed that PTL (10µg/ml) decreased 40% of c-jun expression and 50% of c-fos expression in HCT-116 cells (Figure 4A, Figure 4B). The same dose of the treatment showed 50% up-regulated E-cadherin expression ( Figure 4C) and 30% down regulation of N-cadherin expression ( Figure 4D).

Western Immunoblot analysis of DKK-1 and MMP-9
The effect of PTL on DKK-1 and MMP-9 expression in HCT-116 colon cancer cells were determined quantitatively using Western immunoblot analysis. The data were further analysed by Image-J software. We measured the expression levels of DKK-1, a key negative regulator of the Wnt signalling pathway, after treatment with varying concentrations of PTL. Here, we observed that PTL treatment at 5µg/ml dose showed 1.2 fold expression of   DKK-1 whereas 10µg/ml PTL treatment showed 2.0 fold DKK-1 expression as compared with the control after 48 h of treatment. Next, we measured MMP-9 expressions. MMP-9 is a vital biomarker of invasion and metastatic properties of cancer. Data showed that 5µg/ml and 10µg/ml of PTL treatments lower MMP-9 expression by 20% and 30% respectively as compared with the control indicating the antimetastatic effect of PTL in a dose-dependent fashion. (Figure 5)

DISCUSSION
Metastasis is the most complicated stage of neoplasm being the major lethal cause due to cancer. Epithelial to mesenchymal transition is the process through which tumour cells migrates and metastasise into the circulation and distant organs. Colorectal cancer has become the most common gastrointestinal malignancy in the recent decade. Therefore, search for a potent anti-cancer drug which can potentially prevent proliferation and migration of cancer cells and downregulate EMT markers is going on. Parthenolide is a key component among sesquiterpene lactones present in Feverfew. PTL has an epoxide functional group and an α-methylene-γ-lactone ring were having nucleophilic reaction ability with biological molecules, especially having cysteine thiol (-SH) groups by 'Michael addition reaction' (Dey et al., 2016). It has been shown to inhibit growth or induce apoptosis in several tumour cell lines (Guzman et al., 2005;Carlisi et al., 2016). In the present study, we have examined the therapeutic ef icacy of PTL against migration and metastatic properties of HCT-116 colorectal cancer cells in vitro. This study showed that PTL signi icantly killed the HCT-116 cells in a dose-dependent fashion. Collectively, 5µg/ml and 10µg/ml dose of PTL with 48 h incubation period were selected. Being free radicals, ROS molecules are highly reactive and show a crucial role in cell signalling regulation, causing oxidative cell damage and ultimately cell death. In normal physiological homeostasis, the amount of ROS remains lower during metabolism, which is effectively quenched by many antioxidant enzymes of the glutathione system. Cellular ROS is generated via different enzymatic relations, viz. the mitochondrial respiratory chain reaction, membrane-bound superoxide generating enzyme, NADPH oxidase etc. (Carlisi et al., 2016).  From the luorescence microscopic image analysis, it has been observed that in HCT-116 cells, DCFH 2 -DA luorescence intensity was signi icantly elevated (p<0.05) by PTL treatment. Increased level of ROS excites the release of several in lammatory molecules, including TNF-α. TNF-α stimulates NF-κB and JNK, which eventually increases apoptotic and necrotic cell death (Shen and Pervaiz, 2006). To observe the cell death pattern due to PTL treatment, we stained the cancer cells with EtBr and AO and the luorescence images after PTL treatment were observed. These typical staining stains that the healthy cells with undamaged DNA and give green luorescence, and late apoptotic/necrotic cells' having fragmented DNA emits orange/red-coloured luorescence. In our experiment, it is evident that PTL treatment decreases the number of viable cells and increases the apoptotic cell population in a dosedependent manner. Glutathione, an important cellular antioxidant, protects cells from cellular peroxides and different free radicals. This study showed that GSH levels in HCT-116 cancer cell line were signi icantly decreased (p<0.05) when treated with PTL. Conversely GSSG level in cancer cells significantly increased (p<0.5) when treated with PTL. Therefore, it can be said that PTL treatment can alter the cellular redox balance and directs cells towards oxidative damage.
Most of the cancer patients died due to metastasis. Tumour cell migration and invasion to the circulation from the surrounding tissue through epithelial to mesenchymal transition is considered as an initial step of the metastasis. To examine the effect of PTL in colon cancer migration, an in vitro wound healing study was carried out. We have measured the percentage of inhibition of wound healing which was calculated using ImageJ software analysis. After treatment with 5µg/ml PTL for 48 h, we observed that the wound closure rate decreased by 30%.
Additionally, 10µg/ml PTL treatment increased the wound area by 39% due to extensive cell death. To further con irm we performed gene expression analysis by the real-time qPCR method of some genes that are essential modulator of epithelial to mesenchymal transition. c-fos is a proto-oncogene that is expressed in many cancers. c-fos overexpression promotes cancer cell growth and angiogenesis. c-fos has a DNA binding domain which remains in a dimer form with c-jun gene product and ultimately forms the transcription factor activating protein 1 (AP-1). Since c-fos is a member of the AP-1 family, it is associated primarily with signal transduction, cell proliferation and cellular differentiation (Milde-Langosch, 2005). In our experimental result, we found a signi icant decrease in c-fos and c-jun gene expression level, indicating the antiproliferative and anti-angiogenic potential of PTL.
Additionally, we have studied gene expression of two important markers of EMT pathway, E-cadherin and N-cadherin. E-cadherin is an adhesion protein which represents the epithelial nature of cells, whereas N-cadherin is a marker of mesenchymal phenotype. E-cadherin downregulation and Ncadherin upregulation switch on the EMT pathway.
Our experimental data revealed that PTL could signi icantly up-regulate E-cadherin gene expression The scienti ic report suggested that expression of E-cadherin modulates β-catenin translocation, thereby inhibiting the Wnt pathway (Loh et al., 2019). So, in this study PTL induced overexpression of E-cadherin may also take part in suppressing Wnt pathway in HCT-116 colon cancer cells. Wnt pathway is one of the major pathways that promote metastasis. In colon cancer cells, it was reported that the Wnt/β-catenin pathway is hyperactivated and nuclear translocation and accumulation of βcatenin promotes metastasis (Tenbaum et al., 2012).
Hence, the pharmacological intervention, which blocks canonical Wnt pathway holds promising indication to control metastatic migration of colon cancer cells. DICKKOPF-1 (DKK-1) is an antagonist of the Wnt pathway, and therefore, we have analysed the effect of PTL treatment on DKK-1 protein expression in HCT-116 cells.
Our present study, the up-regulated expression of DKK-1 by PTL treatment in HCT-116 cells, indicates inhibition of Wnt/β-catenin pathwaymediated EMT in this colon cancer cell. The mesenchymal phenotype marker N-cadherin form synergistic crosstalk with ibroblast growth factor receptor via the extracellular domains and aid the ERK 1/2 pathway activation and MMP-9 expression (Suyama et al., 2002).
Generally increased MMP-9 expression and activation is one of the hallmarks of tumour progression, including angiogenesis, invasion and metastasis. In our study, PTL repressed the MMP-9 protein expression in a dose-dependent fashion. Present data suggest that PTL treatment inhibits cell migration/invasion by regulating EMT markers (E-cadherin and N-cadherin) and MMP-9. (Figure 6)

CONCLUSIONS
In conclusion, this can be demonstrated that PTL inhibits cell proliferation and induces HCT-116 cell death through several pathways including oxidative stress, thiol depletion, DNA damage and protooncogene downregulation. Our indings also provide the shreds of evidence that PTL inhibits cell migration through the modulation of Wnt signalling, EMT pathway and MMP. These indings support that, PTL has a promising anti-cancer potential against metastatic HCT-116 colorectal carcinoma cells.

ACKNOWLEDGEMENT
Authors are obliged to the Department of Biotechnology, Government of West Bengal, India for granting research fund to Dr Biplab Giri. We express our sincere thanks to Professor Sanjay Ghosh, Department of Biochemistry, University of Calcutta, for his generous gift of providing us with the HCT-116 cells.