Inhibition of peroxide removal systems and ascorbate-induced cytotoxicity in pancreatic cancer

Transcription

1 University of Iowa Iowa Research Online Theses and Dissertations Spring 2016 Inhibition of peroxide removal systems and ascorbate-induced cytotoxicity in pancreatic cancer Hannah Van Beek University of Iowa Copyright 2016 Hannah Van Beek This thesis is available at Iowa Research Online: Recommended Citation Van Beek, Hannah. "Inhibition of peroxide removal systems and ascorbate-induced cytotoxicity in pancreatic cancer." MS (Master of Science) thesis, University of Iowa, Follow this and additional works at: Part of the Pharmacology Commons

2 INHIBITION OF PEROXIDE REMOVAL SYSTEMS AND ASCORBATE-INDUCED CYTOTOXICITY IN PANCREATIC CANCER by Hannah Van Beek A thesis submitted in partial fulfillment of the requirements for the Master of Science degree in Pharmacology in the Graduate College of The University of Iowa May 2016 Thesis Supervisors: Professor Joseph J. Cullen, M.D. Professor Dawn E. Quelle, Ph.D.

3 Graduate College The University of Iowa Iowa City, Iowa CERTIFICATE OF APPROVAL MASTER S THESIS This is to certify that the Master s thesis of Hannah Marie Van Beek has been approved by the Examining Committee for the thesis requirement for the Master of Science degree in Pharmacology at the May 2016 graduation. Thesis Committee: Joseph J. Cullen, Thesis Supervisor Dawn E. Quelle, Thesis Supervisor John G. Koland

4 Abstract: Compared to normal cells, cancer cells tend to have higher concentrations of reactive oxygen species (ROS) such as hydrogen peroxide (H 2O 2) due to an accelerated cellular metabolism. The high ROS content leaves cancer cells increasingly susceptible to oxidative stress-induced cell death. This susceptibility can be manipulated in selective cancer therapy by further increasing production of ROS or inhibiting peroxide removal systems or a combination of the two. Pharmacological ascorbate (high-dose intravenous ascorbate) has been shown to sensitize pancreatic cancer to ionizing radiation (IR) by increasing production of ROS such as H 2O 2. Glutathione reductase (GR) and thioredoxin reductase (TrxR) are both important enzymes in peroxide removal systems. GR and TrxR function to recycle key electron donors in the cellular removal of H 2O 2. We hypothesized that inhibiting the peroxide removal systems via inhibition of GR and TrxR would enhance ascorbate-induced cytotoxicity in pancreatic cancer cells. Inhibition of TrxR activity enhanced ascorbate-induced cytotoxicity in MIA PaCa-2 pancreatic cancer cells. Additionally, knockdown of GR protein expression in combination with pharmacological ascorbate treatment increased MIA PaCa-2 pancreatic cancer cell sensitivity to IR. In MIA PaCa-2 and 403 F1 patient-derived pancreatic cancer cells, inhibition of both TrxR and GR activity combined with pharmacological ascorbate enhanced radiosensitivity. However, this effect was not seen in 339 patient-derived pancreatic cancer cells treated with the same dose of ascorbate. In conclusion, inhibition of TrxR activity, GR activity, or both enhances radiosensitivity and ascorbate-induced cytotoxicity in some, but not all, pancreatic cancer cell lines. Treatments combining ascorbate with inhibition of H 2O 2 removal may be an effective strategy for treatment of pancreatic adenocarcinoma. ii

5 Public Abstract: Pancreatic adenocarcinoma is the fourth leading cause of cancer deaths in the United States. Surgical resection of the primary tumor is currently the only treatment with curative potential. Adjuvant treatments have thus far not been shown to significantly improve long-term survival. Compared to normal cells, cancer cells have increased reactive oxygen species (ROS) levels because of their accelerated metabolism. The high ROS content of cancer cells leaves them even more susceptible to oxidative stress-induced cell death. This particular trait can be manipulated in selective cancer therapy. Glutathione reductase (GR) and thioredoxin reductase (TrxR) are important enzymes in peroxide removal systems that function to recycle key electron donors in the cellular removal of hydrogen peroxide (H 2O 2). Intravenous (pharmacological) ascorbate produces high plasma concentrations that are cytotoxic to tumor cells. Ascorbate auto-oxidation generates H 2O 2. Likewise, radiation therapy is known to increase levels of ROS, including H 2O 2. We hypothesized that inhibiting the peroxide removal systems via inhibition of GR and TrxR will enhance ascorbate-induced cytotoxicity in pancreatic cancer cells. Here, we show that inhibition of TrxR activity, GR activity, or both enhances radiosensitivity and ascorbate-induced cytotoxicity in MIA PaCa-2 and 403 F1 pancreatic cancer cell lines, but not in 339 pancreatic cancer cells. Therapies combining ascorbate with inhibition of H 2O 2 removal have the potential to be an effective strategy in treatment of pancreatic adenocarcinoma. iii

6 Table of Contents: List of Figures:... v Introduction:... 1 Methods:... 4 Cell Culture... 4 Thioredoxin Reductase Activity... 4 Clonogenic Assay... 5 Knockdown of Glutathione Reductase Expression... 5 Western Blotting... 5 Glutathione Reductase Activity... 6 Statistical Analysis... 6 Results:... 6 Inhibition of Thioredoxin Reductase... 6 Inhibition of Glutathione Peroxidase System... 7 Inhibition of Hydrogen Peroxide Removal Systems... 8 Discussion:...15 References:...17 iv

7 List of Figures: Figure 1. Thioredoxin and glutathione peroxidase (GPx) antioxidant enzyme systems and inhibitors. 3 Figure 2. Inhibition of TrxR enhances ascorbate-induced cytotoxicity in MIA PaCa-2 cells. 9 Figure 3. GSR2, but not scrambled or GSR4 vectors, decreased GR protein expression and enzymatic activity. 10 Figure 4. Knockdown of GR protein expression and ascorbate enhanced radiosensitivity in MIA PaCa-2 cells. 11 Figure 5. 2-DG and ascorbate enhanced radiosensitivity in MIA PaCa-2 and 403 F1 pancreatic cancer cells. 13 v

8 Introduction: In the United States, pancreatic cancer accounts for 3% of all cancers and 7% of cancer deaths [1]. Pancreatic adenocarcinoma of ductal epithelial origin, comprises the majority of pancreatic cancer cases, and is the fourth leading cause of cancer deaths. The incidence of pancreatic cancer has been increasing slightly over the past decade [1]. Surgical resection of the tumor is currently the only potentially curative treatment. However, less than 10% of patients are eligible for surgical resection, and median survival is only 12 to 18 months after resection. Additionally, fewer than 5% of patients survive 5 years following resection. Radiation and chemotherapy are common adjuvant therapies in the treatment of pancreatic cancer. Nevertheless, these therapies have failed to significantly improve long-term survival. Less than 10% of patients benefit from radiation while less than 20% respond to chemotherapy. Very few patients survive beyond 5 years, and most will ultimately die of metastatic recurrence [2]. Almost all cancers, including pancreatic cancer have elevated levels of reactive oxygen species (ROS), promoting tumor development and progression [3]. High levels of ROS in cancer cells are often the result of increased metabolic activity as well as mitochondrial dysfunction [4]. Normally, aerobic generation of energy occurs when molecular oxygen is reduced to water via the transfer of four electrons to oxygen. In principle, this reaction continues without production of intermediates. However, this reaction also generates partially reduced oxygen species. - Superoxide (O 2 ) may account for as much as two percent of total oxygen consumption. In turn, hydrogen peroxide (H 2O 2) and oxygen are produced by the both the spontaneous dismutation of superoxide and the superoxide dismutase - catalyzed reaction. In normal aerobic cells, ROS exist in balance with biochemical antioxidants, and disruption of this balance can result in oxidative stress [5]. Dividing cells require rapid ATP generation, increased synthesis of macromolecular building blocks, and tighter control of cellular redox status. Cancer cell metabolism is altered to balance the need for energy with the need for macromolecules and redox balance. This need for increased ATP production results in increased generation of ROS and H 2O 2. The amplified ROS content of cancer cells necessitates greater control of antioxidant systems in order to avoid oxidative stress [6]. Many cancer therapies have been designed to exploit the elevated levels of ROS in cancer cells by increasing cellular ROS levels to overwhelm antioxidant systems and induce cellular 1

9 damage and tumor cell apoptosis. This can occur through random damaging functions of ROS or a ROS-mediated induction of apoptosis. These results are often achieved through chemotherapy or radiation therapy. Because cancer cells have a higher basal level of ROS, they are more dependent upon antioxidant systems as compared to normal cells with lower basal ROS levels. This characteristic allows selective toxicity of some agents towards cancer cells compared to normal cells [4]. Damage caused by ROS is prevented and repaired by an antioxidant network (Figure 1). The enzyme catalase can remove H 2O 2 present in high concentrations. In contrast, the selenium-dependent enzyme, Glutathione peroxidase (GPx) has high affinity for H 2O 2, and can remove it even when H 2O 2 is present at low concentrations. Glutathione (GSH) provides the electrons for GPx removal of H 2O 2 [7]. For every molecule of H 2O 2 removed, two molecules of glutathione donate an electron and are oxidized to glutathione disulfide (GSSG). Glutathione disulfide is known to be toxic to cells [8]. Glutathione disulfide reductase (GR) is responsible for recycling GSSG to maintain GSH levels. GR removes two electrons from NADPH, donating them to GSSG, regenerating cellular stores of GSH for further H 2O 2 removal [9]. The GSH system is one of two major systems responsible for removal of H 2O 2. The other major peroxide removal system is the thioredoxin system. In this system, peroxiredoxin reduces H 2O 2 to H 2O at the expense of thioredoxin [10]. Thioredoxin contains a catalytic site in which two neighboring cysteines are cycled between the active reduced dithiol and oxidized disulfide forms [4]. Thioredoxin reductase (TrxR) is then responsible for regeneration of the reduced dithiol form of thioredoxin, donating electrons obtained from NADPH [10]. The antioxidant enzymes can be targeted either directly or indirectly via modulation of NADPH availability. NADPH is an important electron donor in the reduction of both GSSG and thioredoxin. Energy from the conversion of glucose-6-phosphate into ribulose-5-phosphate during glycolysis is utilized during the oxidative phase of the pentose phosphate cycle to reduce two molecules of NADP + to NADPH. 2-deoxyglucose (2-DG) is a glucose analog in which the 2- hydroxyl group is replaced by a hydrogen atom. This effectively prevents the molecule from undergoing glycolysis and allowing it to competitively inhibit the production of glucose-6- phosphate [11]. Glucose transporters facilitate the uptake of 2-DG. Many types of cancer cells tend to have higher glucose uptake than normal cells, and therefore have higher uptake of 2-DG [12]. For this reason, 2-DG has been investigated as a potential anti-cancer agent. Additionally, by inhibiting glycolysis, 2-DG indirectly prevents NADP + reduction to NADPH. Therefore, it may 2

10 have an effect on the ability of GR and TrxR to reduce their substrates, GSSG and thioredoxin, respectively [13]. Auranofin is an orally available organic gold compound traditionally used to treat rheumatoid arthritis. It is known to interact with selenol-containing residues, and inhibit thioredoxin reductase [14]. This inhibition of H 2O 2 removal is the rationale behind recent interest in auranofin use as an anticancer therapy. Figure 1. Thioredoxin and glutathione peroxidase (GPx) antioxidant enzyme systems and inhibitors. GSH = glutathione; GSSG = glutathione disulfide; GR = glutathione disulfide reductase; GPx = glutathione peroxidase; Trx-(SH) 2 = reduced thioredoxin; Trx-S 2 = oxidized thioredoxin; TR = thioredoxin reductase; Prx = peroxiredoxin; G-6-P = glucose-6-phosphate; 6-P-G = 6-phosphoglucono-δ-lactone; Inhibitors of the pathway are: Auranofin; Viral vectors; 2DG = 2-deoxy-D-glucose. Ascorbate (also known as Vitamin C and ascorbic acid) has been established as an effective reducing agent. It will readily undergo two one-electron oxidations, forming both ascorbate radical (Asc - ) and dehydroascorbic acid (DHA). At physiologic concentrations, ascorbate anion (AscH - ) functions primarily as a reducing agent and antioxidant, and readily donate an electron to potentially damaging free radicals such as hydroxyl radicals (HO ). Ascorbate radical is relatively unreactive and readily dismutes to ascorbate and DHA [10, 15]. These antioxidant properties of ascorbate are arguably its most well-known properties. A normal, healthy human will typically have a plasma ascorbate concentration of μm. These are the levels at which ascorbate functions as an antioxidant [10]. At pharmacological 3

11 (high) doses, ascorbate acts as a pro-oxidant, readily oxidizing in the presence of catalytic metals. It can act as a one-electron reducing agent, and is oxidized to ascorbate radical while reducing ferric (Fe 3+ ) to ferrous (Fe 2+.- ) iron. Ferrous iron can react with O 2 and reduce it to O 2 which then dismutes to H 2O 2 and O 2 [10]. High-dose intravenous ascorbate is well tolerated when administered to patients. Ascorbate concentrations of 25 mm were obtained in patients who received ascorbate (1.5 g/kg). As of July 2015, the mean survival of subjects completing at least 2 cycles (8 weeks) of therapy is now 15 ± 2 months compared to the historical controls of 6 months [16]. Pharmacological ascorbate has been since shown to sensitize several cancer types to radiation therapy by increased production of ROS such as H 2O 2, overwhelming antioxidant systems [17]. Ascorbate is well-tolerated and seems to lack the normal cell toxicity of other cancer treatments, making it a promising adjuvant for cancer therapy. One working model for the development of cancer therapeutics is the combination of inhibited peroxide removal and pro-oxidant pharmacological agents. This combination may sufficiently overwhelm antioxidants over the toxicity threshold [4]. We hypothesized that inhibition of peroxide-removal systems would enhance ascorbate-induced cytotoxicity in pancreatic cancer cells. Methods: Cell Culture The pancreatic cancer cell line MiaPaCa-2 (ATCC) were cultured in DMEM (Gibco, USA) supplemented with 10% FBS and antibiotics. The patient-derived pancreatic cancer cell lines, 339 and 403 F1 (Medical College of Wisconsin surgical oncology tissue bank), were cultured in F-12 media (Gibco, USA) supplemented with 10% FBS, insulin, EGF, hydrocortisone, bovine pituitary extract and antibiotics. Thioredoxin Reductase Activity Cells were collected by scraping and thioredoxin reductase activity measured using a colorimetric Thioredoxin Reductase Assay Kit (Sigma-Aldrich, St. Louis, MO) as per manufacturer s directions. In the assay, TrxR catalyzes the reduction of 5,5 -dithiobis(2- nitrobenzoic) acid (DTNB) with NAPDH to 5-thio-2-nitrobenzoic acid (TNB). TNB produces a strong yellow color that is measured at 412 nm. Because other enzymes can also reduce 4

12 DTNB, a TrxR specific inhibitor is utilized to measure DTNB reduction specifically due to TrxR activity. The assay mixtures (1 ml) contained 100 mm potassium phosphate, ph 7.0, 10 mm EDTA, 0.24 mm NADPH, 3 M DTNB, with and without TrxR inhibitor. The change in absorbance at 412 nm was measured over 1 min at 25 C. A unit of TrxR was defined as the amount of TrxR which will cause an increase in A 412 of 1.0 per minute per ml at ph 7.0 at 25 C. Activity was defined as units per mg protein in the sample. Clonogenic Assay Cells (1.2 x 10 5 ) were seeded in 60 mm culture dishes and treated 48 h later. After treatment, cells were trypsinized and seeded into 6-well plates at 300 cell/well or 600 cell/well in 4 ml growth media. Colonies formed between 7 and 14 days at 37 C were fixed with 70% ethanol and stained with Coomasie blue. Colonies with more than 50 cells were counted. Knockdown of Glutathione Reductase Expression The following shrna sequences targeting the glutathione reductase mrna were designed by the University of Iowa Viral Vector core and were cloned downstream of the mu6 promoter: GSR2 shrna: sense 5 -GGCCATGCAGCCTTCACGAG-3 and antisense 5 -TTCGTGAAGGCTGCATGGCC-3 GSR4 shrna: sense 5 -GCTTAGGAATAACCAGCGAT-3 and antisense 5 -ATCGCTGGTTATTCCTAAGC-3 The mu6shrna and a CMVeGFP cassette were then packaged into a feline immunodeficiency virus (FIV) lentiviral vector to enable stable transduction. MIA PaCa-2 cells (3 x 10 5 ) were seeded in 100 mm culture dishes. 48 h later, cells were transduced with 5 MOI viral vector in 1x Polybrene (4 μg/ml) in serum-free DMEM. After 6 hours, transduction media was removed and replaced with complete media. Cells received either scrambled viral vector (Scrambled) or one of two glutathione reductase-targeted shrna expressing vectors (GSR2 and GSR4). At least 2 days following transduction, flow cytometry was used to strictly select cells expressing the GFP reporter. Following flow cytometry, cells were allowed to grow for 2 weeks before further analysis. Western Blotting Cells were collected by scraping; protein concentrations were determined using a Bio-Rad DC Bradford Protein Assay (Bio-Rad Laboratories, Hercules, CA). Protein (8 μg) was electrophoresed in a Bio-Rad 4-20% Ready gel at 100 V. The proteins were electrotransferred 5

13 onto a PVDF membrane (Millipore, Billerica, CA) and blocked with 5% nonfat dry milk in 0.1% Tween-PBS (TPBS) for 60 min. The membranes were incubated with a rabbit polyclonal antiglutathione reductase antibody (1:1000, Abcam, Cambridge, MA) at 4 C overnight. Membranes were washed 5 times with 0.1 % TPBS and incubated with secondary antibodies conjugated with horseradish peroxidase (1:25 000, Millipore, Temecula, CA). Blotting with GAPDH antibody (1:10 000, Millipore, Temecula, CA) was used as a loading control. After washing with TPBS, membranes were stained with Super Signal West Pico Chemiluminescent Substrate (Pierce Biotechnology, Rockford, IL) and exposed to Classic Blue Autoradiography Film (Molecular Technologies, St Louis, MO). Adjusted density was determined by densitometry using ImageJ analysis software (NIH). Glutathione Reductase Activity Cells were collected by scraping and glutathione reductase activity measured using a colorimetric Glutathione Reductase Assay Kit (Sigma-Aldrich, St. Louis, MO) as per manufacturer s directions. GR activity is measured by the decrease in absorbance at 340 nm caused by the oxidation of NADPH when glutathione is reduced in the presence of GR. The assay mixtures (1 ml) contained mm potassium phosphate, ph 7.5, mm EDTA, 0.1 mm NADPH, and 1 mm oxidized glutathione. The change in absorbance at 340 nm was measured over 100 sec at 25 C. A unit was defined as the amount of GR which will cause a decrease in A 340 of 1.0 per minute per ml at ph 7.0 at 25 C. Activity was defined as units per mg protein in the sample. Statistical Analysis A single factor ANOVA followed by the Tukey post hoc test was used to determine statistical differences between means for multiple comparisons. A Student s t-test was used to determine statistical differences between means for two comparisons. All means were calculated from one or two independent experiments in triplicate, with error bars representing the SD. All data are expressed as mean ± SD. Results: Inhibition of Thioredoxin Reductase Auranofin (AF) has been shown to interact with selenol-containing residues and inhibit TrxR [14]. MIA PaCa-2 pancreatic cancer cells treated with AF (1 μm) for 3 h showed that TrxR 6

14 activity was significantly inhibited compared to vehicle-treated controls (Figure 2A). After 24 h AF treatment, TrxR activity was not significantly different from activity following 3 h AF treatment (n = 3, P < 0.05). Pharmacological ascorbate has been shown to induce cytotoxicity in pancreatic cancer cells via its oxidation and eventual production of H 2O 2 [2]. Therefore, inhibition of TrxR and resultant impairment of the thioredoxin peroxide removal system should enhance ascorbate-induced cytotoxicity in pancreatic cancer cells. When MIA PaCa-2 cells were treated with AF (1 μm) for 3 or 24 h and Asc (2 mm) for 1 h, the surviving fraction was significantly decreased compared to cells treated with AF alone (Figure 2B, n = 6, P < 0.01). Inhibition of Glutathione Peroxidase System Glutathione reductase (GR) activity is required for efficient peroxide removal by the glutathione peroxidase system. By removing two electrons from NADPH and donating them to GSSG, GR regenerates cellular stores of GSH. These stores provide electrons for the removal of H 2O 2 by glutathione peroxidase [9]. To study the specific contributions of glutathione reductase inhibition to ascorbate-induced cytotoxicity, MIA PaCa-2 cells were transduced with a lentiviral vector encoding GR-targeted shrna (GSR2 and GSR4) or a scrambled control vector (Scrambled). To determine the efficiency of GR knockdown, Western blot analysis and GR enzyme activity assays were performed. Cells transduced with GSR2 vector demonstrated decreased GR protein compared to non-transduced and Scrambled-transduced cells (Figure 3A). GR enzyme activity assays demonstrated a significant decrease in GR activity in GSR2-transduced cells compared to nontransduced controls and both Scrambled and GSR4-transduced cells (Figure 3B, n = 3, P < 0.05). To investigate the effect of GR knockdown on MIA PaCa-2 cell response to pharmacological ascorbate treatment, clonogenic survival assays were performed. Cells transduced with Scrambled, GSR2, and GSR4 vectors showed similar clonogenic survival when treated with 0 1 mm concentrations of ascorbate for 1 h. When treated with 2 mm ascorbate, both GSR2 and GSR4 transduced cells showed significant decreases in clonogenic survival compared to Scrambled vector transduced cells (Figure 4A, n = 3, P < 0.05). Previous studies from the laboratory have shown that pharmacologic ascorbate treatment enhances ionizing radiation (IR)- induced cytotoxicity in pancreatic cancer cells [17]. To determine whether inhibition of GR enhances ascorbate-induced radiosensitization, clonogenic 7

15 survival assays were performed combining GR knockdown with pharmacological ascorbate and IR treatments. Neither scrambled nor GSR4-transduced MIA PaCa-2 cells showed significant differences between radiation alone and radiation with ascorbate (0.5 mm) treatments (Figure 4B, 4D, two independent experiments, each in triplicate, P < 0.05). However, GSR2-transduced cells showed a significant decrease in surviving fraction when treated with both radiation (4 Gy) and ascorbate (0.5 mm) for 1 h (Figure 4C, two independent experiments, each in triplicate, P < 0.05). Inhibition of Hydrogen Peroxide Removal Systems 2-Deoxy-Glucose (2-DG) is a glucose analog that competitively inhibits glucose metabolism, creating a chemically induced state of glucose deprivation, resulting in a decreased level of NADPH required for GR and TrxR activity, inhibiting hydroperoxide detoxification [11, 12]. MIA PaCa-2 cells treated with 2-DG (25 mm) for 2 h and ascorbate (0.5 mm) and IR (2 Gy, and 4 Gy) for 1 h showed significantly decreased clonogenic survival compared to cells treated with IR alone (Figure 5A; two independent experiments, each in triplicate; 2 Gy, P < 0.01; 4 Gy, P < 0.05). Similarly, 403 F1 patient-derived pancreatic cancer cells showed significant decreases in the surviving fraction of cells treated with 2-DG (25 mm, 2 h), ascorbate (0.5 mm, 1 h) and IR (2-4 Gy, 1 h) compared to cells treated only with IR (Figure 5B, n = 3, P < 0.05). 339 patientderived cells showed no significant difference between combination 2-DG (25 mm, 2 h), ascorbate (0.5 mm, 1 h) and IR (2-4 Gy, 1 h) therapy and IR alone (Figure 5C, n = 3, P < 0.05). 8

16 Figure 2A Figure 2B Figure 2. Inhibition of TrxR enhances ascorbate-induced cytotoxicity in MIA PaCa-2 cells. A) AF (1 μm) treatment for 3 h significantly decreased TrxR activity in MIA PaCa-2 cells, *P < 0.05 vs. 0 h, means ± SD, n = 3. B) Combination of AF (1 μm) for 3 or 24 h and Asc (2mM) for 1 h significantly decreased MIA PaCa-2 clonogenic cell survival, #P < 0.01 vs. AF, means ± SD, n = 3. 9

17 Figure 3A Control Scrambled GSR2 GSR4 GR (58 kda) GAPDH (37 kda) Adjusted Density Figure 3B Figure 3. GSR2, but not scrambled or GSR4 vectors, decreased GR protein expression and enzymatic activity. A) Western blot demonstrates knockdown of GR protein expression in MIA PaCa-2 cells by GSR2 viral transduction compared to control and scrambled. B) GSR2 viral transduction of MIA PaCa-2 cells significantly decreased GR enzymatic activity, *P < 0.05 vs. control, means ± SD, n = 3. 10

18 Figure 4A Figure 4B Figure 4. Knockdown of GR protein expression and ascorbate enhanced radiosensitivity in MIA PaCa-2 cells. A) Scrambled, GSR2, and GSR4 transduced MIA PaCa-2 cells showed similar clonogenic survival when treated with 0-1 mm concentrations Asc for 1 h, *P < 0.05, means ± SD, n = 3. B) Ascorbate (0.5 mm) for 1 h did not significantly change clonogenic survival in scrambled-transduced MIA PaCa-2 cells treated with IR (0-4 Gy), *P < 0.05 vs. IR, means ± SD, two independent experiments, each in triplicate. 11

19 Figure 4C Figure 4D Figure 4 - continued. Knockdown of GR protein expression and ascorbate enhanced radiosensitivity in MIA PaCa-2 cells. C) GSR2 vector transduction significantly decreased clonogenic survival of MIA PaCa-2 cells treated with ascorbate (0.5 mm) and IR (4 Gy) for 1 h, *P < 0.05 vs. IR, means ± SD, two independent experiments, each in triplicate. D) Ascorbate (0.5 mm) for 1 h did not significantly change clonogenic survival in GSR4-trasduced MIA PaCa- 2 cells treated with IR (0-4 Gy), *P < 0.05 vs. IR, means ± SD, two independent experiments, each in triplicate. 12

20 5A 5B Figure 5. 2-DG and ascorbate enhanced radiosensitivity in MIA PaCa-2 and 403 F1 pancreatic cancer cells. A and B) MIA PaCa-2 and 403 F1 cells treated with 2- DG (25 mm) for 2 h, ascorbate (0.5 mm) for 1 h and IR (2-4 Gy) for 1 h showed significantly decreased clonogenic survival compared to cells receiving IR alone, #P < 0.01 vs. IR, *P < 0.05 vs. IR, means ± SD, MIA PaCa-2: two independent experiments, each in triplicate, 403 F1: n = 3. 13

21 Figure 5C Figure 5 - continued. 2-DG and ascorbate enhanced radiosensitivity in MIA PaCa-2 and 403 F1 pancreatic cancer cells. C) 2-DG (25 mm) for 2 h, ascorbate (0.5 mm) for 1 h and IR (2-4 Gy) for 1 h did not significantly change 339 patient-derived pancreatic cancer cell clonogenic survival compared to IR alone, *P < 0.05 vs. IR, means ± SD, n = 3. 14

22 Discussion: Previous studies have demonstrated pharmacological ascorbate-induced cytotoxicity in pancreatic cancer cells by an H 2O 2-mediated mechanism [2, 10]. For this reason, we investigated inhibition of peroxide removal systems as a potential adjuvant to pharmacological ascorbate in treatment of pancreatic cancer. Inhibition of TrxR by AF significantly decreased clonogenic cell survival when combined with pharmacological ascorbate treatment in pancreatic cancer cells compared to AF alone. This supported the hypothesis that inhibition of the thioredoxin peroxide removal system can enhance ascorbate-induced cytotoxicity in pancreatic cancer cells. The glutathione peroxidase system requires GR activity to efficiently remove peroxide [9]. To study the effects of decreased GR activity, we developed GSR2 and GSR4 lentiviral vectors encoding a GR-targeted shrna to stably knockdown GR expression in MIA PaCa-2 pancreatic cancer cells. These vectors enabled us to decrease GR activity without additional off-target effects. The GSR2 lentiviral vector knocked down GR protein expression and significantly decreased GR activity in MIA PaCa-2 cells. Additionally, we showed that knockdown of GR in GSR2-transduced cells did not enhance ascorbate-induced cytotoxicity when treated with low dose (0-1mM) ascorbate alone. However, GSR2-transduced cells showed significant decreases in clonogenic survival when treated with both radiation and ascorbate. Conversely, GSR4 and scrambled vector transduced cells treated with ascorbate did not exhibit enhanced sensitivity to radiation compared to radiation alone. Hence, knockdown of GR protein expression and ascorbate increased radiosensitivity in MIA PaCa-2 pancreatic cancer cells. The glucose analog, 2-DG competitively inhibits glucose metabolism, resulting in decreased NADPH. Because NADPH is required for GR and TrxR activity, this effectivity inhibits H 2O 2 metabolism by both the thioredoxin and glutathione peroxidase systems [11, 12]. In both MIA PaCa-2 and 403 F1 patient-derived pancreatic cancer cells, combination of 2-DG, ascorbate, and radiation therapy significantly decreased clonogenic survival compared to radiation alone. 339 patient-derived pancreatic cancer cells showed no difference in clonogenic survival between combination therapy and radiation alone. These data show that 2-DG and ascorbate increased radiosensitivity in some, but not all pancreatic cancer cell types. Unpublished data (Claire Doskey, personal communication, unreferenced; Table 1) shows a higher rate constant for H 2O 2 removal in 339 cells than in 403 F1 and MIA PaCa-2 cells. In addition, 339 cells have been shown to have more active catalase molecules per cell than both 403 F1 and MIA PaCa-2 15

23 cells. Taken together, these data may explain why 339 cells appear to be more resistant to the cytotoxic effects of the combination 2-DG, ascorbate and radiation treatment. MIA PaCa F1 339 Obtained from ATCC Med College Wisc Tissue Bank Disease Carcinoma Adenocarcinoma Adenocarcinoma Kras status Mutated p53 status Mutated Smad4 status Negative Positive Rate constant for H 2O 2 removal Active catalase units/cell Table 1. Characteristics of MIA PaCa-2, 403 F1, and 339 pancreatic cancer cells. Rate constant for H 2O 2 removal and active catalase molecules per cell based on unpublished data from Claire Doskey, personal communication, unreferenced. In summary, inhibition of TrxR activity, GR activity, or both can sensitive some pancreatic cancer cell lines to pro-oxidants such as pharmacological ascorbate and ionizing radiation. While 2-DG has not yet been approved for use in humans in the United States, pharmacological ascorbate and auranofin are immediately available pharmacological agents. Treatments that combine ascorbate with inhibition of H 2O 2 removal have the potential to be an effective strategy for treatment of pancreatic adenocarcinoma. 16

24 References: [1] American Cancer Society. (2014) Cancer Facts & Figures American Cancer Society [2] Du, J., Martin, S.M., Levine, M., Wagner, B.A., Buettner, G.R., Wang, S., Taghiyev, A.F., Du, C., Knudson, C.M., and Cullen, J.J. (2010) Mechanisms of ascorbate-induced cytotoxicity in pancreatic cancer. Clin Cancer Res. 16, [3] Du J., Nelson, E.S., Simons, A.L., Olney, K.E., Moser, J.C., Schrock, H.E., Wagner, B.A., Buettner, G.R., Smith, B.L., Teoh, M.L.T., Tsao, M.S.T., and Cullen, J.J. (2013) Regulation of pancreatic cancer growth by superoxide. Mol Carcinog. 52, [4] Liou, G.-Y., and Storz, P. (2010) Reactive oxygen species in cancer. Free Radic. Res. 44, [5] Liu, Y., Fiskum, G., and Schubert, D. (2002) Generation of reactive oxygen species by the mitochondrial electron transport chain. J. Neurochem. 5, [6] Waris, G., and Ahsan, H. (2006) Reactive oxygen species: Role in the development of cancer and various chronic conditions. J Carcinog 5. [7] Kohen, R., and Nyska, A. (2002) Oxidation of biological systems: Oxidative stress phenomena, antioxidants, redox reactions, and methods for their quantification. Tox. Path. 30, [8] Filomeni, G., Rotilio, G., and Ciriolo, M.R. (2003) Glutathione disulfide induces apoptosis in U937 cells by a redox-mediated p38 MAP kinase pathway. FASEB J. 17, [9] Andreyev, A.Yu., Kushnareva, Yu.E., and Starkov, A.A. (2005) Mitochondrial metabolism of reactive oxygen species. Biochem. 70, [10] Du, J., Cullen, J.J., and Buettner, G.R. (2012) Ascorbic acid: Chemistry, biology and the treatment of cancer. Biochimica et Biophysica Acta (BBA) - Reviews on Cancer.1826, [11] Wick, A.N., Drury, D.R., Nakada, H.I., and Wolfe, J.B. (1957). Localization of the primary metabolic block produced by 2-deoxyglucose. J. Biol. Chem. 224, [12] Coleman, M.C.. Asbury, C., Daniels, D.H., Du, J., Aykin-Burns, N., Smith, B.J., Li, L., Spitz, D.R., Cullen, J.J. (2008) Inhibition of glucose metabolism in pancreatic cancer induces cytotoxicity via metabolic oxidative stress. Free Radic Biol Med. 44, [13] Pelicano, H., Martin, D.S., Xu, R.H., and Huang, P. (2006) Glycolysis inhibition for anticancer treatment. Oncogene. 25, [14] Talbot, S., Nelson, R., and Self, W.T. (2008) Arsenic trioxide and auranofin inhibit selenoprotein synthesis: Implications for chemotherapy for acute promyelocytic leukaemia. Br. J. Pharmacol. 154, [15] Parrow, N.L., Leshin, J.A., and Levine, M. (2013) Parenteral ascorbate as a cancer therapeutic: A reassessment based on pharmacokinetics. Antiox. and Redox Signal. 19, [16] Welsh, J.L., Wagner, B.A., van't Erve, T.J., Zehr, P.S., Berg, D.J., Halfdanarson, T.R., Yee, N.S., Bodeker, K.L., Du, J., Roberts, L.J. 2nd, Drisko, J., Levine, M., Buettner, G.R., Cullen, J.J. (2013) Pharmacological ascorbate with gemcitabine for the control of metastatic and node-positive pancreatic cancer (PACMAN): results from a phase I clinical trial. Cancer Chemother Pharmacol. 71, [17] Du, J., Cieslak, J.A. 3rd, Welsh, J.L., Sibenaller, Z.A., Allen, B.G., Wagner, B.A., Kalen, A.L., Doskey, C.M., Strother, R.K., Button, A.M., Mott, S.L., Smith, B., Tsai, S., Mezhir, J., Goswami, P.C., Spitz, D.R., Buettner, G.R., Cullen, J.J. (2015) Pharmacological ascorbate radiosensitizes pancreatic cancer. Cancer Res. 75,