Ursodeoxycholic acid prompts glycolytic dominance, reductive stress and epithelial-to-mesenchymal transition in ovarian cancer cells through NRF2 activation

doi:10.21203/rs.3.rs-5190306/v1

Numerous secreted bacterial metabolites were identified with bioactivity in various neoplasias, including ovarian cancer. One such metabolite is ursodeoxycholic acid (UDCA), a secondary bile acid. Hereby, we assessed the bioactivity of UDCA in cell models of ovarian cancer, by applying UDCA in concentrations corresponding to the serum reference concentrations of UDCA (300 nM). UDCA induced epithelial-to-mesenchymal transition (EMT), increased the flux of glycolysis and reduced the naturally occurring oxidative stress in ovarian cancer cells. These changes were dependent on the activation of NRF2. The tumoral overexpression of UDCA-induced genes in humans correlated with worse survival. These results point out that bacterial metabolites may have opposite effects in different neoplasias and raise the possibility that UDCA-containing remedies on the long run may support cancer progression in ovarian cancer patients.

ursodeoxycholic acid

ovarian cancer

oxidative stress

EMT

Warburg metabolism

proliferation

NRF2

microbiome

Ovarian cancer is the second most common and most lethal gynecological malignancy after endometrial cancer [1, 2]. Ovarian cancer is not a single homogenous disease, but can be classified into two major types, type I and II with differences in the mechanism of carcinogenesis, tissue of origin and clinical prognosis [3, 4]. Type I tumors (30%) are low-grade, indolent tumors, while type II tumors (70%) are aggressive high-grade cancers that almost always present as advanced stage with high fatality [3, 5–8].

Dysbiosis is a pathological accommodation of the microbiome in compartments of an organism. Dysbiosis accompanies numerous diseases, among them neoplasias [9]. Dysbiosis accompanies ovarian cancer affecting, among others, the gut microbiome [10]. The healthy microbiome counteracts neoplastic cells through multi-pronged pathways in which the production of cytostatic metabolites play a prominent role [10] that was evidenced in ovarian cancer [11]. Cytostatic bacterial metabolites include compounds with various chemical structures, including bile acids.

Primary bile acids are synthetized in the liver [12]. Primary bile acids are secreted to the gastrointestinal tract to facilitate the emulsification of dietary fats and, hence, promote their digestion and uptake [12]. A fraction of primary bile acids are converted by intestinal bacteria to secondary bile acids through dihydroxylation and deconjugation [13]. Secondary bile acids are then absorbed from the gastrointestinal tract to the portal circulation to be subsequently cleared by the liver. Nevertheless, a fraction of the (secondary) bile acids remain in the systemic circulation and can elicit systemic effects similar to hormones [14]. Importantly, the ovary has a system for the production of bile acids [15, 16]. Mean follicular total bile acid level is around 10 µM [16], twice as the normal serum reference concentration.

Ursodeoxycholic acid (UDCA) is a secondary bile acid, its reference concentration in the human serum (conjugated + deconjugated UDCA) is between 100–300 µM [17–21]. Bile acids can be found, besides the systemic circulation, in the ascites of ovarian cancer patients [22]. A deregulation of bile acid biosynthesis was observed in ovarian cancer patients [23–26]. Furthermore, multiple bile acid receptors were shown to be involved in the regulation of hallmarks of ovarian cancer (reviewed in [12]) pointing towards a possible involvement of bile acids in regulating the progression of ovarian cancer.

In this study we assessed the effects of UDCA, a secondary bile acid that is a registered medication and in previous studies had beneficial antineoplastic effects in models of ovarian cancer (reviewed in [12]).

Ursodeoxycholic acid induces EMT and cellular diapedesis

Bile acids can influence epithelial-to-mesenchymal transition (EMT) in cell models of other cancers [21, 27, 28]. Therefore, we assessed whether UDCA has similar properties in the cell model of ovarian cancer. UDCA induced the expression of pro-EMT factors as the mRNA of Transcription factor 7-like 2 (TCF7L2) [29] and protein expression of β-catenin [30] and Snail [31] (Fig. 1A-B) suggesting that UDCA induces EMT. In line with these expressional changes, UDCA reduced the adherence of cells to each other and to the surface as shown in impedance measurements (Fig. 1C). Furthermore, UDCA supported chemotaxis and diapedesis of ovarian cancer cells in the Boyden chamber (Fig. 1D). However, UDCA did not influence the behavior of cells in an in vitro model of wound healing (Fig. 1E) and did not impact on the rate of cell proliferation when applied either in the reference concentration range (Fig. 1F) or in concentrations in the therapeutic range (Fig. 1G).

EMT can support drug resistance of cancer cells, among them, ovarian cancer cells [32]. As the therapy of ovarian cancer is built on platinum-based drugs [11], we assessed the sensitivity of A2780 cells and A2780 cisplatin resistant cells to cisplatin with and without UDCA treatment in a short timeframe (4 hours, MTT assay; Fig. 2A) and in a long timeframe (48 hours, SRB assay; Fig. 2B, C) (similar to [33]). Cisplatin had no effect in the short timeframe neither in cisplatin-sensitive nor in cisplatin-resistant cells (Fig. 2A), in contrast to the long treatment scheme (Fig. 2B, C). However, UDCA treatment did not affect cisplatin-induced cytostasis (Fig. 2B, C).

Ursodeoxycholic acid induces NRF2 expression and reduces oxidative stress

Bile acids in other cancer models impacted on cancer cell proliferation through mediating the redox status of cancer cells (e.g. [21, 34-36]). Changes to the redox status of cells upon bile acid treatment depends on changes to the expression and cellular localization of nuclear factor erythroid 2–related factor 2 (NRF2/NFE2L2) [37, 38].

UDCA induced the mRNA (Fig. 3A) and protein expression of NRF2 (Fig. 3B) without influencing the expression of Kelch-like ECH-associated protein 1 (KEAP1) (Fig. 3C) that is an intrinsic inhibitor of NRF2. These suggest the induction of NRF2 and in line with that, we observed a reduced immunosignal of 4-hydroxy-nonenal (4HNE), an oxidative stress marker [39] (Fig. 3D) and reduced hydroethidine signal, an indication of lower superoxide production [40] (Fig. 3E). UDCA treatment yielded 4HNE signals throughout the whole molecular weight range of the lysate (Fig. 3D), in contrast to our previous observations, where rather well-defined protein targets were identified.

Ursodeoxycholic acid induces glycolytic flux in A2780 cells

Cancer cells are characterized by changes to metabolism [41-43]. Bile acids [12, 14, 21] and, more specifically, UDCA [44-46] can induce oxidative metabolism raising the possibility of UDCA-elicited metabolic changes in A2780 cells. Similar, NRF2 activation and the subsequent change to cellular redox homeostasis can also induce changes to intermediary metabolism [37, 47-53].

UDCA induced the ECAR value in A2780 cells (Fig. 4A) suggesting increases in glycolytic flux. The effects of UDCA on oxidative phosphorylation (assessed through the OCR value) was very variable, there were experiments where UDCA induced and others where UDCA inhibited cellular oxygen consumption) (data not shown), hence changes to OCR was not considered. The induction of glycolysis was dependent on the activation of NRF2, as a pharmacological inhibitor of NRF2, ML385 [54] inhibited the induction of ECAR (Fig. 4A).

Finally, we assessed whether the induction of ECAR by UDCA can be generalized to other ovarian cancer cell lines. The treatment of ID8 murine ovarian cancer cells with UDCA led to an induction of ECAR (Fig. 4B), similar to the observations on A2780 cells.

In addition we assessed the activity of the AMP-activated kinase (AMPK) and the mechanistic target of rapamycin (mTOR) system in response to UDCA. The phosphorylation of p70 S6 kinase (S6K) at Thr389 was used as a proxy to assess the activity of mTOR complex 1 (mTORC1) and the phosphorylation of Akt kinase at Ser473 was used as a proxy to assess the activity of mTOR complex 2 (mTORC2). AMPK activity was judged by assessing an activating autophosphorylation site of AMPK (at Thr172). UDCA treatment did not influence the level of phosphorylation of S6K (Fig. 5A), Akt (Fig. 5B) and AMPK (Fig. 5C) suggesting that the activity of the AMPK/mTOR system is not influenced by UDCA treatment.

The induction of ursodeoxycholic acid-dependent genes associate with worse clinical outcome

In the study we identified a set of genes that were induced by UDCA either that level of mRNA/protein expression or at the level of activity (TCF7L2, CTNNB1, SNAI1 and NFE2L2). We assessed the composite effect of the 4 gene signature on patient survival using the GEPIA2 database. The tumoral overexpression of the 4 gene signature correlated with markedly worse survival with a hazard ratio of 1.3 (Fig. 6).

In this study we showed that UDCA, when administered in concentrations corresponding to the human serum reference concentration, promotes cancer progression. In previous studies on a plethora of different neoplasias UDCA exerted beneficial, antineoplastic effects (reviewed in [12]) by inducing apoptosis or supporting chemosensitization [55-57]. In this regard, our observations are novel and surprising.

UDCA exerted multi-pronged effects by impacting on EMT, cell adhesion, diapedesis, glycolytic dominance and oxidative stress. We observed the complex effects of bile acids in cell models of other neoplasia [21, 27] similar to previous studies on ovarian cancer [55-57]. The only contradiction among this study and previous observations is the lack of cytotoxicity or cytostasis in our data contrasted to prior art [56, 57], however, this is likely a concentration-dependent difference. Bile acids in high concentration can prove toxic, as bile acids can harm biomembranes, however, in submicromolar concentrations this is not likely (e.g. [21]).

Serum primary and secondary bile acid levels, including glycoursodeoxycholic acid decrease in ovarian cancer patients [23, 24]. Furthermore, bile acid biosynthesis [58-60], similar to other metabolites [61-63], decreases with age and age is a risk factor for ovarian cancer. These observations make internal ursodeoxycholic overproduction in patients unlikely, however, UDCA supplementation as medication may increase the risk of progression in patients with ovarian cancer.

We identified NRF2 overexpression as a mechanism of UDCA-induced changes. NRF2 overexpression [64, 65] supports cancer progression [37, 42], metastasis formation, induce EMT [66, 67] and glycolytic flux [37, 47-53]. Reduction of 4HNE signal upon UDCA treatment points out a reduction in naturally occurring oxidative stress that can be considered reductive stress [68]. Of note, reductive stress rearranges cellular intermediary metabolism and supports the Warburg-type metabolism [53] that we observed in UDCA-treated cells hereby. Warburg rearrangement occurs in ovarian cancer cells [69-73], and changes to cellular metabolism are associated with metastatic capacity [72-81] suggesting similar properties for UDCA.

To our surprise, we have not observed the induction of chemoresistance upon UDCA treatment neither in cisplatin resistant, nor in cisplatin sensitive A2780 cells, despite the widespread literature reports on resistance against numerous chemotherapy agents upon NRF2 induction in multiple cancers, including ovarian cancer [55, 57, 82-85].

Hereby, we provided evidence that UDCA treatment in a cellular models of ovarian cancer induces features of cancer progression and metastasis formation. High tumoral expression of the UDCA-induced genes correlate in humans with shorter survival suggesting that increases in UDCA concentration in the environment of ovarian cancer cells support cancer progression. These observations are in conflict with the prior observations on the widespread antineoplastic effects UDCA in multiple cancers [12], on a broader scale that points out that bacterial metabolites have cancer-specific actions and may have opposing features in different neoplasias. Furthermore, our results point towards a cautious administration of UDCA-containing drugs in ovarian cancer patients.

Chemicals

All chemicals were purchased from Sigma-Aldrich (St. Louis, MO, USA) unless stated otherwise. The sources of key chemicals are assembled in Table 1. UDCA was purchased from Sigma-Aldrich (St. Louis, MI, USA) and was dissolved in dimethyl sulfoxide (DMSO) to achieve the final concentration of 100 mM. Cisplatin was dissolved in 1xPBS at the concentration of 3 mM, ML385 was dissolved in DMSO in the concentration of 50 mM.

Table 1. The source of key chemicals

Compound	Company	Ref. No.
UDCA	Sigma-Aldrich 3050 Spruce Street, Saint Louis, MO 63103, USA	U5127 CAS 128-13-2
cisplatin	Sigma-Aldrich 3050 Spruce Street, Saint Louis, MO 63103, USA	232120 CAS 15663-27-1
ML385	Selleck Chemicals 14408 W Sylvanfield Drive, Houston, TX 77014 USA	S8790 CAS 846557-71-9

Cell lines

A2780 cells were cultured in RPMI 1640 medium supplemented with 10% fetal calf serum, 2 mM L-glutamine, 1% penicillin-streptomycin.

ID8 cells were cultured in high glucose DMEM (4.5 g/L glucose) medium supplemented with 4% fetal calf serum, 2 mM L-glutamine, 1% penicillin-streptomycin, 1% ITS supplement (I3146).

Cisplatin resistant A2780 cells were grown in RPMI 1640 medium supplemented with 10% fetal calf serum, 2 mM glutamine, 1% penicillin-streptomycin. Cisplatin resistant cells underwent selection (1 µM cisplatin) once a week for3 days before plating for assay.

Cell lines were regularly checked for mycoplasma contamination.

Methylthiazolyldiphenyl-tetrazolium bromide (MTT) reduction assay

MTT reduction assay was measured as previously described [33]. The assay determines the activity of mitochondrial complex I and can be used to detect rapid toxicity and the induction of apoptosis [86]. Briefly, cells were plated in 96 well plates the day before the assay. A2780 cells or cisplatin resistant A2780 cells were plated in 96 well plates for MTT assays with a number of 10⁴ cells/well. On the next day, cells were treated with cisplatin in itself or in a combination with 300 nM UDCA for 4 hours. Cisplatin was applied between the concentrations of 100 µM and 0.0017 µM with one-third dilution. Vehicle controls were DMSO for UDCA and 1xPBS for cisplatin.

Sulforhodamine B assay

The SRB assay measures acid-precipitable protein onto cells that is proportional to cell number, hence can be used to assess cell proliferation or long-term cytostasis [87]. SRB proliferation assay was measured as previously described [33].

For Figure 1., A2780 cells (10³ cells/well) were plated to 96 well plates and were treated with UDCA on the next day for 48 hours. For concentration range of serum reference, UDCA was applied between 1000 nM and 86 nM with four-fifths dilution. For therapeutic concentration range UDCA was applied between 15 µM and 0.4 µM with two-thirds dilution.

For Figure 2., A2780 cells (2x10³cells/well) and cisplatin resistant A2780 cells (6x10³ cells/well) were plated in 96 well plates. On the next day, cells were treated with cisplatin in itself or in a combination with 300 nM UDCA for 48 hours. Cisplatin was applied between the concentrations of 100 µM and 0.0017 µM with one-third dilution. Vehicle controls were DMSO for UDCA and 1xPBS for cisplatin.

Impedance measurements

Electric cell-substrate impedance sensing (ECIS Z-Theta, Applied BioPhysics) was applied as described previously [88]. A2780 cells (2x10⁴ or 5x10⁴) were plated into 8W10E ECIS plates (3-4 well/condition). This plate is recommended to monitor cell-cell connections. Impedance measurement was initiated upon plating and was continued onwards. Impedance was measured at multiple frequencies between 62.5 Hz and 64 kHz in 120 or 150 seconds (as a function of cell numbers) time intervals in real time. After reaching confluency (plateau in capacitance and resistance), half of the medium was replaced by 600 nM UDCA (yielding a final concentration of 300 nM) or DMSO as vehicle control. The impedance value corresponding to the plateau (i.e. confluence) was regarded as a baseline value. Impedance measurement was resumed for 24 hours. The average of the resistance values at 4 kHz ±30 minutes before and after the peak resistance was regarded as a peak value. The difference between the peak and the baseline values were expressed as fold change compared to the DMSO-treated samples and were plotted. All plates contained a well containing only medium as a background control for background correction.

Western blot

Cells were lysed in RIPA buffer (50 mM Tris, 150 mM NaCl, 0.1% SDS, 1% TritonX 100, 0.5% sodium deoxycolate, 1 mM EDTA, 1 mM Na₃VO₄, 1 mM NaF, 1 mM PMSF, protease inhibitor coctail). Protein concentrations were determined using a BCA protein assay kit (Pierce Biotechnologies, Rockford, IL, USA). Protein extracts were loaded onto 10% SDS polyacrylamide gels. After electrophoresis, proteins were transferred to nitrocellulose membranes and blocked using 5% BSA for 1 h at room temperature. Then, membranes were incubated with primary antibodies for overnight at 4 °C. Membranes were washed with Tris-buffered saline containing 0.1% Tween (TBS-T) followed by a 1 h incubation with IgG HRP conjugated secondary antibody (Cell Signaling Technology, Inc. Beverly, MA, 1:2000). Membranes were washed with TBS-T and visualization was performed by SuperSignal West Pico Solutions (Thermo Fisher Scientific). b-actin was used as a loading control. Blots were quantified by densitometry using Image Lab 6.1 software.

Antibodies used in this study are listed in Table 2.

Table 2. The antibodies used in the study

Abbreviations: mAb – monoclonal antibody, pAb – polyclonal antibody

Name	Species	Ig class	Dilution	Vendor and Cat No.	RRID
Name	Species	Ig class	Dilution	Vendor and Cat No.	RRID
Actin	Rabbit pAb		1:30000	Sigma-Aldrich A2066	AB 476693
Akt	Rabbit pAb		1:1000	Cell Signaling Technology #9272	AB 329827
Phospho-Akt (Ser473) (D9E) XP®	Rabbit mAb	IgG	1:2000	Cell Signaling Technology #4060	AB 2315049
AMPKα (D63G4)	Rabbit mAb		1:1000	Cell Signaling Technology #5832	AB 10624867
Phospho-AMPKα (Thr172) (40H9)	Rabbit mAb	IgG	1:1000	Cell Signaling Technology #2535	AB 331250
KEAP1 (D6B12)	Rabbit mAb	IgG	1:1000	Cell Signaling Technology #8047	AB 10860776
Nrf2	Rabbit pAb	IgG	1:1000	Abcam ab31163	AB 881705
Nrf2	Rabbit pAb	IgG	1:1000	Novus Biologicals NBP1-32822	AB 10003994
p70 S6 Kinase (49D7)	Rabbit mAb	IgG	1:1000	Cell Signaling Technology #2708	AB_390722
Phospho-p70 S6 Kinase (Thr389)	Rabbit pAb		1:1000	Cell Signaling Technology #9205	AB 330944
Snail (C15D3)	Rabbit mAb	IgG	1:1000	Cell Signaling Technology #3879	AB 2255011
4 Hydroxynonenal	Rabbit pAb	IgG	1:1000	Abcam ab46545	AB_722490
β-Catenin (D10A8) XP®	Rabbit mAb	IgG	1:1000	Cell Signaling Technology #8480	AB_11127855
Anti-rabbit IgG HRP-linked	Goat	IgG	1:10000	Cell Signaling Technology #7074	AB_2099233

Total RNA preparation and reverse transcription-coupled quantitative PCR (RT-qPCR)

Total RNA from A2780 cells was prepared using TRIzol reagent (cat. no. TR118, Molecular Research Center Inc., Cincinnati, OH, USA) according to the manufacturer’s instructions. DNase treatment of total RNA samples was performed using 2U of DNaseI (cat. no. AM2222, Thermo Fisher Scientific) per 10 µg RNA. Two μg of total RNA were used for reverse transcription using High-Capacity cDNA Reverse Transcription Kit (cat. no. 4368813, Thermo Fisher Scientific) according to the manufacturer’s instructions. The reverse transcription reaction mixtures were supplemented with RNase inhibitor (cat. no. N8080119, Thermo Fisher Scientific). The cDNA was cleaned with the nucleospin Gel and PCR clean-up (Macherey-Nagel Gmbh & Co. KG, Düren, Germany) according to the manufacturer’s instructions.

The qPCR reactions were performed with qPCRBIO SyGreen Lo-ROX Supermix (PCR Biosystems Ltd., London, UK) using 20 ng diluted cDNA with the appropriate primers at 500 nM final concentration in 10 μl total volume. Amplification reactions were performed using a Light-Cycler 480 system (Roche Applied Science, Basel, Switzerland). Cycling conditions are given in Table 3. Primers are listed in Table 4.

Table 3. Cycling conditions of qPCR reactions

Step	Temperature	Time	Cycles
Polymerase activation	95°C	10 min	1
Denaturation	95°C	10 s	50
Annealing/Extension	62°C	20 s	50
Melting curve analysis	95°C	5 s	1
	65°C	1 min
	65°C to 97°C (Continuous)	ramp rate 0.11°C/s
	40°C	30 s

Table 4 The specifics of the primer sets used in the study

Reporting is made in accordance with the QIME guideline [89, 90]

Gene name	Primers (5’-3’)	Length (bp)	Specificity check	Alignments (Sequence ID) Query coverage 100% Percent identity 100%	Primer location	Exon intron boundary	Calibration curve	Efficiency
NRF2	FW: TGAGCAAGTTTGGGAGGAGCT REV: ACTGGTTGGGGTCTTCTGTGG	245	BLAST	NM_001313903.2	FW 484-504 REV 728-708	YES	Slope: -3.428 YIntercept: 28.11 Error=0.0735	1.958
				NM_001313902.2	FW 613-633 REV 857-837
				NM_006164.5	FW 703-723 REV 947-927
				NM_001313904.1	FW 1165-1185 REV 1409-1389
				NM_001313901.1	FW 1202-1222 REV 1446-1426
				NM_001313900.1	FW 1110-1130 REV 1354-1334
				NM_001145413.3	FW 1215-1235 REV 1459-1439
				NM_001145412.3	FW 1236-1256 REV 1480-1460
TCF7L2	FW: ACGTACAGCAATGAACACTTCAC REV: GGCGATAGTGGGTAATACGG	128	BLAST	NM_001198530.2	FW 952-974 REV 1079-1060	YES	Slope: -3.699 YIntercept: 29.42 Error = 0.204	1.863
				NM_001146283.2	FW 1195-1217 REV 1322-1303
				NM_001349870.2	FW 778-800 REV 905-886
				NM_001146285.2 NM_001198528.2 NM_001146286.2 NM_001198527.2 NM_001198526.2 NM_001146284.2 NM_001198525.2 NM_001198529.2 NM_030756.5	FW 1054-1076 REV 1181-1162
				NM_001367943.1 NM_001146274.2 NM_001363501.2 NM_001198531.2	FW 1123-1145 REV 1250-1231
				NM_001349871.1	FW 310-332 REV 437-418
PPIA	FW: GTCTCCTTTGAGCTGTTTGCAGAC REV: CTTGCCACCAGTGCCATTATG	171	BLAST	NM_021130.5	FW 102-125 REV 272-252	YES	Slope: -3,485 YIntercept: 25.54 Error= 0.0366	1.936
RPLP0	FW: CCATTGAAATCCTGAGTGATGTG REV: GTCGAACACCTGCTGGATGAC	131	BLAST	NM_053275.4	FW 589-611 REV 719-699	YES	Slope: -3.552 YIntercept: 30.65 Error= 0.0548	1.912
RPLP0	FW: CCATTGAAATCCTGAGTGATGTG REV: GTCGAACACCTGCTGGATGAC	131	BLAST	NM_001002.4	FW 529-551 REV 659-639	YES	Slope: -3.552 YIntercept: 30.65 Error= 0.0548	1.912

Boyden chamber experiments

Cell invasion assay was determined using Corning BioCoat Matrigel Invasion Chambers (Corning, NY, USA) with 8.0 µm PET membrane in 24-well plates (cat.no.# 354480). In the control setup similar 8 µm PET membranes were used that were devoid of the matrigel layers. Cells (20.000 cells/well) were added to the top of insert in upper chamber in serum free medium and were seeded overnight. Then, cells were treated with LCA (0.03 µM) in 0.5 ml serum free medium. Simultaneously medium (0.75 ml) containing 10 % FBS, LCA (0.03 µM) and 100 ng/ml stromal cell-derived factor 1α (SDF1α) (Sigma-Aldrich, cat.no.# SRP4388) was added to the lower chamber as chemoattractant. After 48 h, non-invaded cells on the top of the filters were wiped off with cotton swabs. The lower surface of membranes containing invaded cells were washed in PBS and fixed with 100% methanol and stained with DAPI. The invading cells were counted with Opera Phoenix High Content Screening System (Perkin-Elmer Waltham, MA, USA) and pictures were analyzed using the Harmony 4.6 Software. The following calculations were performed:

% Invasion = (Mean of cells invading through Matrigel insert membrane / Mean of cells invading through Control insert membrane) *100

Invasion index = % Invasion of the Treated cell / % Invasion of the Control (non-treated) cell

Seahorse measurements

A2780 (6x10³) or ID8 (5x10³) cells were plated into Seahorse XF96 cell culture microplates (Agilent Technologies, Inc., Santa Clara, CAL, USA) and were treated with 300 nM UDCA or DMSO (as vehicle). In a subset of experiments UDCA-treated cells were treated with ML385 (1µM) for 48 hours. At the end of treatment medium was replaced by buffer-free DMEM supplemented with glucose (1 g/L) and the plate was incubated in a CO₂-free incubator for at least 1 hour and then the plate was subjected to Seahorse measurement using the Seahorse XF96 instrument. Five minute measurements were made that was repeated 5 times, interspersed by 30 seconds of mixing. After the assay cell numbers were determined by SRB assay. The stable ECAR readouts were averaged for each well, the first measurement point was omitted. The ECAR values were normalized for the SRB absorbance.

Database screening

We assessed the survival data in GEPIA2 [91] and database.

Statistical analysis

Statistical analysis was performed using 8.0.1 version of Graphpad Prism. Values were tested for normal distribution using the test indicated in the figure legends. When necessary, values were log normalized. The following statistical test, post hoc test and the level of significance is indicated in the figure captions. Nonlinear regression was performed using the built-in “[Inhibitor] vs. response – Variable slope (four parameters), least square fit” utility of Graphpad that yielded IC50 and Hill slope values.

Acknowledgements

The authors acknowledge the technical help of Ms. Kitti Barta.

The study was financed by grants from NKFIH (K142141, FK128387, FK146852, TKP2021-EGA-19, TKP2021-EGA-20) and the Hungarian Academy of Sciences (NKM2022-30). Project no. TKP2021-EGA-19 and TKP2021-EGA-20 were implemented with the support provided from the National Research, Development and Innovation Fund of Hungary, financed under the TKP2021-EGA funding scheme. EM and AS are funded by the Bolyai Fellowship of the Hungarian Academy of Sciences. Supported by the University of Debrecen Program for Scientific Publication. This project has received funding from the HUN-REN Hungarian Research Network.

Conflict of interest

Dr. Bai is a shareholder and CEO of Holobiont Diagnostics LTD. Involved in the development of microbiome-based diagnostic tools. Other authors declare no conflict of interest.

Data availability

Primary data is available at figshare.com (https://figshare.com/s/1f7e89f1c6e4cd82749d; DOI: 10.6084/m9.figshare.26825533)

Torre, L.A., et al., Ovarian cancer statistics, 2018. CA Cancer J Clin., 2018. 68(4): p. 284-296.
Colombo, N., et al., ESMO-ESGO consensus conference recommendations on ovarian cancer: pathology and molecular biology, early and advanced stages, borderline tumours and recurrent disease. Ann Oncol., 2019. 30(5): p. 672-705.
Kurman, R.J., Origin and molecular pathogenesis of ovarian high-grade serous carcinoma. Ann Oncol, 2013. 24 Suppl 10: p. x16-21.
Koshiyama, M., N. Matsumura, and I. Konishi, Recent concepts of ovarian carcinogenesis: type I and type II. Biomed Res Int, 2014. 2014: p. 934261.
Shih Ie, M. and R.J. Kurman, Ovarian tumorigenesis: a proposed model based on morphological and molecular genetic analysis. Am J Pathol, 2004. 164(5): p. 1511-8.
Wiegand, K.C., et al., ARID1A mutations in endometriosis-associated ovarian carcinomas. N Engl J Med, 2010. 363(16): p. 1532-43.
Carlson, J.W., et al., Serous tubal intraepithelial carcinoma: its potential role in primary peritoneal serous carcinoma and serous cancer prevention. J Clin Oncol, 2008. 26(25): p. 4160-5.
Kindelberger, D.W., et al., Intraepithelial carcinoma of the fimbria and pelvic serous carcinoma: Evidence for a causal relationship. Am J Surg Pathol, 2007. 31(2): p. 161-9.
Thomas, R.M. and C. Jobin, The Microbiome and Cancer: Is the 'Oncobiome' Mirage Real? Trends in Cancer, 2015. 1(1): p. 24-35.
Mikó, E., et al., Guideline for designing microbiome studies in neoplastic diseases. Geroscience, 2024.
Sipos, A., et al., The role of the microbiome in ovarian cancer: mechanistic insights into oncobiosis and to bacterial metabolite signaling. Molecular Medicine, 2021. 27(1): p. 33.
Režen, T., et al., The role of bile acids in carcinogenesis. Cell Mol Life Sci, 2022. 79(5): p. 243.
Ridlon, J.M. and J.S. Bajaj, The human gut sterolbiome: bile acid-microbiome endocrine aspects and therapeutics. Acta Pharm Sin B., 2015. 5(2): p. 99-105.
Watanabe, M., et al., Bile acids induce energy expenditure by promoting intracellular thyroid hormone activation. Nature., 2006. 439(7075): p. 484-489.
Smith, L.P., et al., The bile acid synthesis pathway is present and functional in the human ovary. PLoS One., 2009. 4(10): p. e7333.
Nagy, R.A., et al., The Origin of Follicular Bile Acids in the Human Ovary. Am J Pathol, 2019. 189(10): p. 2036-2045.
Trottier, J., et al., Profiling circulating and urinary bile acids in patients with biliary obstruction before and after biliary stenting. PLoS One, 2011. 6(7): p. e22094.
García-Cañaveras, J.C., et al., Targeted profiling of circulating and hepatic bile acids in human, mouse, and rat using a UPLC-MRM-MS-validated method. Jouranl of Lipid Research, 2012. 53(10): p. 2231-2241.
Ma, Z., et al., Serum metabolome and targeted bile acid profiling reveals potential novel biomarkers for drug-induced liver injury. Medicine (Baltimore), 2019. 98(31): p. e16717.
Sun, Z., et al., Distinct Bile Acid Profiles in Patients With Chronic Hepatitis B Virus Infection Reveal Metabolic Interplay Between Host, Virus and Gut Microbiome. Frontiers in Medicine, 2021. 8: p. 708495.
Miko, E., et al., Lithocholic acid, a bacterial metabolite reduces breast cancer cell proliferation and aggressiveness. Biochimica et Biophysica Acta - Bioenergetics, 2018. 1859(9): p. 958-974.
Hedenborg, G., et al., Bile constituents in ascitic fluid. Scand J Clin Lab Invest, 1988. 48(6): p. 543-52.
Ke, C., et al., Large-scale profiling of metabolic dysregulation in ovarian cancer. International Journal of Cancer, 2015. 136(3): p. 516-26.
Fan, L., et al., Use of Plasma Metabolomics to Identify Diagnostic Biomarkers for Early Stage Epithelial Ovarian Cancer. Journal of Cancer, 2016. 7(10): p. 1265-72.
Zhou, M., et al., Rapid mass spectrometric metabolic profiling of blood sera detects ovarian cancer with high accuracy. Cancer Epidemiology, Biomarkers & Prevention, 2010. 19(9): p. 2262-71.
Guan, W., et al., Ovarian cancer detection from metabolomic liquid chromatography/mass spectrometry data by support vector machines. BMC Bioinformatics, 2009. 10: p. 259.
Schwarcz, S., et al., The pro- and antineoplastic effects of deoxycholic acid in pancreatic adenocarcinoma cell models. Mol Biol Rep, 2023. 50(6): p. 5273-5282.
Schwarcz, S., et al., The bacterial metabolite, lithocholic acid, has antineoplastic effects in pancreatic adenocarcinoma. Cell Death Discov, 2024. 10(1): p. 248.
del Bosque-Plata, L., E.P. Hernández-Cortés, and C. Gragnoli, The broad pathogenetic role of TCF7L2 in human diseases beyond type 2 diabetes. Journal of Cellular Physiology, 2022. 237(1): p. 301-312.
Xue, W., et al., Wnt/β-catenin-driven EMT regulation in human cancers. Cellular and Molecular Life Sciences, 2024. 81(1): p. 79.
Suzuki, T., et al., Role of epithelial-mesenchymal transition factor SNAI1 and its targets in ovarian cancer aggressiveness. J Cancer Metastasis Treat, 2023. 9.
Shibue, T. and R.A. Weinberg, EMT, CSCs, and drug resistance: the mechanistic link and clinical implications. Nat Rev Clin Oncol, 2017. 11(10): p. 44.
Kacsir, I., et al., Half Sandwich-Type Osmium, Ruthenium, Iridium and Rhodium Complexes with Bidentate Glycosyl Heterocyclic Ligands Induce Cytostasis in Platinum-Resistant Ovarian Cancer Cells and Bacteriostasis in Gram-Positive Multiresistant Bacteria. Frontiers in Chemistry, 2023. 11: p. 1086267.
Kovács, P., et al., Lithocholic acid, a metabolite of the microbiome, increases oxidative stress in breast cancer. Cancers (Basel), 2019. 11(9): p. 1255.
Goldberg, A.A., et al., Lithocholic bile acid selectively kills neuroblastoma cells, while sparing normal neuronal cells. Oncotarget., 2011. 2(10): p. 761-782.
Goldberg, A.A., et al., Bile acids induce apoptosis selectively in androgen-dependent and -independent prostate cancer cells. PeerJ., 2013. 1: p. e122.
Smolková, K., et al., NRF2 in regulating cancer metabolism. Antioxidants and Redox Signaling, 2020. 33(13): p. 966-997.
Moi, P., et al., Isolation of NF-E2-related factor 2 (Nrf2), a NF-E2-like basic leucine zipper transcriptional activator that binds to the tandem NF-E2/AP1 repeat of the beta-globin locus control region. Proc Natl Acad Sci U S A, 1994. 91(21): p. 9926-30.
Zhang, H. and H.J. Forman, 4-hydroxynonenal-mediated signaling and aging. Free Radic Biol Med, 2017. 111: p. 219-225.
Benov, L., L. Sztejnberg, and I. Fridovich, Critical evaluation of the use of hydroethidine as a measure of superoxide anion radical. Free Radical Biology and Medicine, 1998. 25(7): p. 826-831.
Hanahan, D. and R.A. Weinberg, Hallmarks of cancer: the next generation. Cell, 2011. 144(5): p. 646-74.
Vander Heiden, M.G., L.C. Cantley, and C.B. Thompson, Understanding the Warburg Effect: The Metabolic Requirements of Cell Proliferation. Science, 2009. 324(5930): p. 1029-1033.
Warburg, O., The Metabolism of Carcinoma Cells. The Journal of Cancer Research, 1925. 9(1): p. 148-163.
Xavier, J.M., et al., Tauroursodeoxycholic acid increases neural stem cell pool and neuronal conversion by regulating mitochondria-cell cycle retrograde signaling. Cell Cycle, 2014. 13(22): p. 3576-89.
Fonseca, I., et al., Tauroursodeoxycholic Acid Protects Against Mitochondrial Dysfunction and Cell Death via Mitophagy in Human Neuroblastoma Cells. Molecular Neurobiology, 2017. 54(8): p. 6107-6119.
Yu, H., et al., Apoptosis induced by ursodeoxycholic acid in human melanoma cells through the mitochondrial pathway. Oncology Reports, 2019. 41(1): p. 213-223.
Pusapati, R.V., et al., mTORC1-Dependent Metabolic Reprogramming Underlies Escape from Glycolysis Addiction in Cancer Cells. Cancer Cell, 2016. 29(4): p. 548-562.
Shi, D., et al., Glycolysis inhibition via mTOR suppression is a key step in cardamonin-induced autophagy in SKOV3 cells. BMC Complement Altern Med, 2018. 18(1): p. 317.
Szwed, A., E. Kim, and E. Jacinto, Regulation and metabolic functions of mTORC1 and mTORC2. Physiological Reviews, 2021. 101(3): p. 1371-1426.
Chang, C.W., et al., ROS-independent ER stress-mediated NRF2 activation promotes warburg effect to maintain stemness-associated properties of cancer-initiating cells. Cell Death Dis, 2018. 9(2): p. 194.
Bollong, M.J., et al., A metabolite-derived protein modification integrates glycolysis with KEAP1-NRF2 signalling. Nature., 2018. 562(7728): p. 600-604.
Lee, S., et al., Impairment of HIF-1α-mediated metabolic adaption by NRF2-silencing in breast cancer cells. Redox Biology, 2019. 24: p. 101210-101210.
Hamanaka, R.B. and N.S. Chandel, Warburg Effect and Redox Balance. Science, 2011. 334(6060): p. 1219-1220.
Singh, A., et al., Small Molecule Inhibitor of NRF2 Selectively Intervenes Therapeutic Resistance in KEAP1-Deficient NSCLC Tumors. ACS Chem Biol, 2016. 11(11): p. 3214-3225.
Jin, Q., et al., Bile acids upregulate BRCA1 and downregulate estrogen receptor 1 gene expression in ovarian cancer cells. European Journal of Cancer Prevention, 2018. 27(6): p. 553-556.
Horowitz, N.S., et al., Novel cytotoxic agents from an unexpected source: bile acids and ovarian tumor apoptosis. Gynecologic Oncology, 2007. 107(2): p. 344-349.
Schuldes, H., et al., Reversal of multidrug resistance and increase in plasma membrane fluidity in CHO cells with R-verapamil and bile salts. European Journal of Cancer, 2001. 37(5): p. 660-7.
Ma, J., et al., Gut microbiota remodeling reverses aging-associated inflammation and dysregulation of systemic bile acid homeostasis in mice sex-specifically. Gut Microbes, 2020. 11(5): p. 1450-1474.
MahmoudianDehkordi, S., et al., Altered bile acid profile associates with cognitive impairment in Alzheimer's disease-An emerging role for gut microbiome. Alzheimers & Dementia, 2019. 15(1): p. 76-92.
Bertolotti, M., et al., Age-related changes in bile acid synthesis and hepatic nuclear receptor expression. Eur J Clin Invest, 2007. 37(6): p. 501-8.
Janssens, G.E., et al., A metabolomic signature of decelerated physiological aging in human plasma. GeroScience, 2023. 45(6): p. 3147-3164.
Jasbi, P., et al., Targeted metabolomics reveals plasma biomarkers and metabolic alterations of the aging process in healthy young and older adults. GeroScience, 2023.
Tian, Q., et al., Plasma metabolomic signatures of dual decline in memory and gait in older adults. GeroScience, 2023.
Konstantinopoulos, P.A., et al., Keap1 mutations and Nrf2 pathway activation in epithelial ovarian cancer. Cancer Res, 2011. 71(15): p. 5081-9.
Cho, H.-y., et al., Expression Patterns of Nrf2 and Keap1 in Ovarian Cancer Cells and their Prognostic Role in Disease Recurrence and Patient Survival. International Journal of Gynecologic Cancer, 2017. 27(3): p. 412-419.
Karimi Roshan, M., et al., Role of AKT and mTOR signaling pathways in the induction of epithelial-mesenchymal transition (EMT) process. Biochimie, 2019. 165: p. 229-234.
Helleman, J., et al., Pathway analysis of gene lists associated with platinum-based chemotherapy resistance in ovarian cancer: the big picture. Gynecol Oncol., 2010. 117(2): p. 170-6. .
Zhang, L. and K.D. Tew, Chapter Ten - Reductive stress in cancer, in Advances in Cancer Research, K.D. Tew and P.B. Fisher, Editors. 2021, Academic Press. p. 383-413.
Kellenberger, L.D., et al., The role of dysregulated glucose metabolism in epithelial ovarian cancer. J Oncol, 2010. 2010: p. 514310.
Mandai, M., et al., Ovarian clear cell carcinoma meets metabolism; HNF-1β confers survival benefits through the Warburg effect and ROS reduction. Oncotarget, 2015. 6(31): p. 30704-14.
Nantasupha, C., et al., Metabolic reprogramming in epithelial ovarian cancer. Am J Transl Res, 2021. 13(9): p. 9950-9973.
Tondo-Steele, K. and K. McLean, The "Sweet Spot" of Targeting Tumor Metabolism in Ovarian Cancers. Cancers (Basel), 2022. 14(19).
Szeőcs, D., et al., Cell-free ascites from ovarian cancer patients induces Warburg metabolism and cell proliferation through TGFβ-ERK signaling. GeroScience, 2024.
Rickard, B.P., et al., Malignant Ascites in Ovarian Cancer: Cellular, Acellular, and Biophysical Determinants of Molecular Characteristics and Therapy Response. Cancers (Basel), 2021. 13(17).
Ashraf, R. and S. Kumar, Mfn2-mediated mitochondrial fusion promotes autophagy and suppresses ovarian cancer progression by reducing ROS through AMPK/mTOR/ERK signaling. Cell Mol Life Sci, 2022. 79(11): p. 573.
Bandopadhyay, S., et al., SIRT6 promotes mitochondrial fission and subsequent cellular invasion in ovarian cancer. FEBS Open Bio, 2022. 12(9): p. 1657-1676.
Achreja, A., et al., Metabolic collateral lethal target identification reveals MTHFD2 paralogue dependency in ovarian cancer. Nat Metab, 2022. 4(9): p. 1119-1137.
Hagen, J.T., et al., Intrinsic adaptations in OXPHOS power output and reduced tumorigenicity characterize doxorubicin resistant ovarian cancer cells. Biochim Biophys Acta Bioenerg, 2022. 1863(8): p. 148915.
Wu, Y., et al., Targeting oxidative phosphorylation as an approach for the treatment of ovarian cancer. Front Oncol, 2022. 12: p. 971479.
Tyagi, K., S. Mandal, and A. Roy, Recent advancements in therapeutic targeting of the Warburg effect in refractory ovarian cancer: A promise towards disease remission. Biochim Biophys Acta Rev Cancer, 2021. 1876(1): p. 188563.
Ding, Y., et al., Molecular characteristics and tumorigenicity of ascites-derived tumor cells: mitochondrial oxidative phosphorylation as a novel therapy target in ovarian cancer. Mol Oncol, 2021. 15(12): p. 3578-3595.
Pascual, M.J., et al., Enhanced efficiency of the placental barrier to cisplatin through binding to glycocholic acid. Anticancer Research, 2001. 21(4a): p. 2703-7.
Li, D., et al., Targeting Nrf2 may reverse the drug resistance in ovarian cancer. Cancer Cell International, 2021. 21(1): p. 116.
Bao, L.J., et al., Nrf2 induces cisplatin resistance through activation of autophagy in ovarian carcinoma. Int J Clin Exp Pathol, 2014. 7(4): p. 1502-13.
Shim, G.S., et al., Acquisition of doxorubicin resistance in ovarian carcinoma cells accompanies activation of the NRF2 pathway. Free Radic Biol Med, 2009. 47(11): p. 1619-31.
Virag, L., A.L. Salzman, and C. Szabo, Poly(ADP-ribose) synthetase activation mediates mitochondrial injury during oxidant-induced cell death. Journal of Immunology, 1998. 161(7): p. 3753-3759.
Skehan, P., et al., New colorimetric cytotoxicity assay for anticancer-drug screening. Journal of the National Cancer Institute, 1990. 82(13): p. 1107-12.
Horváth, D., et al., Myosin phosphatase accelerates cutaneous wound healing by regulating migration and differentiation of epidermal keratinocytes via Akt signaling pathway in human and murine skin. Biochim Biophys Acta Mol Basis Dis, 2018. 1864(10): p. 3268-3280.
Bustin, S.A., et al., The MIQE guidelines: minimum information for publication of quantitative real-time PCR experiments. Clinical Chemistry, 2009. 55(4): p. 611-622.
Bustin, S.A., et al., Primer sequence disclosure: a clarification of the MIQE guidelines. Clinical Chemistry, 2011. 57(6): p. 919-21. .
Tang, Z., et al., GEPIA2: an enhanced web server for large-scale expression profiling and interactive analysis. Nucleic Acids Res, 2019. 47(W1): p. W556-w560.

There is a conflict of interest Dr. Bai is a shareholder and CEO of Holobiont Diagnostics LTD. Involved in the development of microbiome-based diagnostic tools. Other authors declare no conflict of interest.

SupplementaryWesternblots.pdf

Ursodeoxycholic acid prompts glycolytic dominance, reductive stress and epithelial-to-mesenchymal transition in ovarian cancer cells through NRF2 activation

Status:

Version 1

Abstract

Figures

Introduction

Results

Ursodeoxycholic acid induces EMT and cellular diapedesis

Ursodeoxycholic acid induces NRF2 expression and reduces oxidative stress

Ursodeoxycholic acid induces glycolytic flux in A2780 cells

The induction of ursodeoxycholic acid-dependent genes associate with worse clinical outcome

Discussion

Materials and Methods

Chemicals

Cell lines

Declarations

Acknowledgements

References

Additional Declarations

Supplementary Files

Status:

Version 1