This dissertation is available at Iowa Research Online:

Transcription

1 University of Iowa Iowa Research Online Theses and Dissertations Spring 2016 Toward absolute quantitation in cell culture: expression of dose of xenobiotics and capacity of cells to remove hydrogen peroxide with implications to pharmacological ascorbate in cancer therapy Claire Marie Doskey University of Iowa Copyright 2016 Claire Marie Doskey This dissertation is available at Iowa Research Online: Recommended Citation Doskey, Claire Marie. "Toward absolute quantitation in cell culture: expression of dose of xenobiotics and capacity of cells to remove hydrogen peroxide with implications to pharmacological ascorbate in cancer therapy." PhD (Doctor of Philosophy) thesis, University of Iowa, Follow this and additional works at: Part of the Toxicology Commons

2 TOWARD ABSOLUTE QUANTITATION IN CELL CULTURE: EXPRESSION OF DOSE OF XENOBIOTICS AND CAPACITY OF CELLS TO REMOVE HYDROGEN PEROXIDE WITH IMPLICATIONS TO PHARMACOLOGICAL ASCORBATE IN CANCER THERAPY by Claire Marie Doskey A thesis submitted in partial fulfillment of the requirements for the Doctor of Philosophy degree in Human Toxicology in the Graduate College of The University of Iowa May 2016 Thesis Supervisor: Professor Garry R. Buettner

3 Graduate College The University of Iowa Iowa City, Iowa CERTIFICATE OF APPROVAL This is to certify that the Ph.D. thesis of PH.D. THESIS Claire Marie Doskey has been approved by the Examining Committee for the thesis requirement for the Doctor of Philosophy degree in Human Toxicology at the May 2016 graduation. Thesis Committee: Garry R. Buettner, Thesis Supervisor Joseph J. Cullen Jonathan A. Doorn Prabhat C. Goswami Gabriele Ludewig

4 ACKNOWLEDGEMENTS I would like to first and foremost thank Dr. Garry Buettner, who has been both an amazing thesis advisor and an excellent mentor to me. His vast knowledge on the different aspects of science still amazes me and having his guidance throughout my research projects has greatly added to my research experience here. I would like to thank all of the members of my Ph.D. committee, Dr. Joseph Cullen, Dr. Jonathan Doorn, Dr. Prabhat Goswami, and Dr. Gabriele Ludewig. Their support and insight have added greatly to my research. I am very grateful for the time they have taken in helping me over the past few years. My time in the Buettner lab has been great and I want to thank all of the lab members over the years that have made collaboration an enjoyable and enriching experience. A major contribution to this has been working with Brett Wagner. Brett s knowledge of scientific literature and his expertise on methods in the lab has helped me greatly throughout the years. Sharing an office for the past few years has led to many laughs and great times. I also want to give a special thanks to Joost van t Erve, who was a great mentor and friend to me during my rotation and first year in the Buettner lab. It was great working with him and he is still very helpful when I need some extra advice on research. Thank you to all the current and former members of the Buettner and Cullen labs that have collaborated with me and contributed to a great work environment: Juan Du, Codey Buranasudja, Justin Wilkes, Adrienne Klinger, Cam Cushing, Jordan Witmer, Malvika Rawal, and Hannah van Beek. ii

5 I would also like to thank my friends in Iowa City that have made my time outside of the lab very enjoyable. A special thanks to the trivia team that make Tuesday my favorite day of the week! I would like to acknowledge both the Interdisciplinary Graduate Program in Human Toxicology and Free Radical and Radiation Biology Program, especially Patricia Ramstad Buettner, David Purdy, Jennifer DeWitte, and Laura Hefley. They were all incredibly helpful and answered so many questions that I have had over my time here. Finally, I want to give a special thanks to my family who have been supportive and encouraging throughout my life and especially my time here: my husband, Bob, my parents Mary and Paul, my siblings Adam, Anna, and Luke. You have all been an inspiration to me and your support means the world to me. iii

6 ABSTRACT Basic health science research, which includes cell culture, typically underpins clinical and toxicological research. The results are used to predict biological effects of xenobiotics, e.g. environmental toxins and drugs, in humans. A goal of this research program was to apply aspects of quantitative redox biology to three separate, but interrelated projects that address improvements that can be made to evidence-based biological and toxicological research. This includes using absolute quantitation: to improve the specification of dose of xenobiotics added to cell culture systems; to determine absolute differences between the antioxidant capacity of tumor and normal cells; and to predict the implications of this new knowledge on the use of pharmacological ascorbate as an adjuvant in cancer therapy. Dose is a central parameter in determining the biological consequences of a xenobiotic; however, the dose of a xenobiotic at which these consequences are observed is dependent not only on biological variables, but also the physical aspects of cell culture experiments (i.e. cell number, medium volume). This is often overlooked due to the unrecognized ambiguity in the dominant metric used to express dose, i.e. initial concentration of xenobiotic. We hypothesized that specifying the dose of xenobiotics absolutely (as moles of xenobiotic per cell; mol cell -1 ) will reduce this ambiguity and provide additional information that is difficult to discern when traditional dosing metrics (initial concentration) are used. We investigated the use of mol cell -1 as an informative dosing metric using two model compounds: 1,4-benzoquinone and oligomycin A. When the dose of these two compounds was specified as mole cell -1, the toxicity observed was independent of the physical conditions used (i.e. number of cells, volume of medium). iv

7 This makes it a scalable dosing metric that reduces ambiguity between experiments having different physical conditions; allows direct comparison between different cell types; addresses the important issue of repeatability of experimental results, and could increase the translatability of information gained from in vitro experiments. We utilized quantitative methods to explore the absolute differences in the ability of tumor vs. normal cells to remove H 2 O 2 and how this impacts the use of pharmacological ascorbate as an adjuvant in cancer therapy. Ascorbate (AscH -, vitamin C) functions as a versatile reducing agent. At pharmacological doses (P-AscH -, plasma levels 20 mm), achievable through IV delivery, the oxidation of ascorbate can produce a high flux of H 2 O 2 in tumors. We hypothesized that the increased sensitivity of tumor cells to P-AscH - compared to normal cells (i.e. non-transformed) is due to their lower capacity to remove H 2 O 2. The rate constants (k cell ) for removal of H 2 O 2 revealed a differential in the capacity of cells to remove H 2 O 2, with the average k cell for normal cells (N = 10) being twice that of tumor cells (N = 15). The ED 50 of P-AscH - correlated directly with the capacity of cells to remove H 2 O 2. Quantitation made it possible to make comparisons across very different cell lines on an absolute basis. These results indicate that the capacity of cells to remove H 2 O 2 varies widely and in vivo measurement of this may predict which tumors may respond best to P-AscH - therapy. By designing experiments that begin with a quantitative dosing metric and utilize quantitation to produce absolute information from the results of experiments, we can better leverage data. We propose that this will lead to better predictions from such experiments. These enhancements to in vitro cell culture studies will increase the success v

8 in translation of data from in vitro experiments to in vivo animal studies and ultimately impact the success of extrapolation of basic science research to human clinical studies. vi

9 PUBLIC ABSTRACT Many years of basic health science research are typically needed before data can be extrapolated to humans, be it testing for the potential toxicity of environmental pollutants or the development of new drugs. Unfortunately, there have been a large number of failures in long-term studies, with the pre-clinical/basic science data not reflecting the biological consequences in humans. In vitro cell culture studies are major contributors to these preliminary studies. Much of the data derived from these experiments consists of relative comparisons across different treatment groups. We hypothesized that using absolute quantitation could improve the repeatability of experimental results, allow direct comparisons between different cell types, and increase the translatability of information gained from in vitro experiments. Here, we use absolute quantitation to: improve the specification of dose in cell culture model systems; determine absolute differences between the antioxidant capacity of tumor and normal cells; and explore what implications this has for the use of pharmacological ascorbate (high-dose vitamin C) as an adjuvant in cancer therapy. By designing experiments that begin with a quantitative metric for the dose of a xenobiotic (i.e. toxicant, pharmaceutical agent, biochemical tool) and utilizing approaches that yield absolute quantitative data from experiments, we can better leverage data to improve human health. We propose that these quantitative data will lead to better predictions, increasing the success in translation of data derived from in vitro experiments to in vivo animal studies and ultimately impact the success of extrapolation of basic science research to improve human health. vii

10 TABLE OF CONTENTS LIST OF TABLES... xi LIST OF FIGURES... xii LIST OF ABBREVIATIONS... xiv CHAPTER I: THEME, BACKGROUND, AND SIGNIFICANCE... 1 Theme... 1 Moles of a Substance per Cell is a Highly Informative Dosing Metric in Cell Culture. 2 Introduction... 2 Dosing Metrics Used in Cell Culture Studies ,4-Benzoquinone: Chemistry and Toxicity... 5 Oligomycin A: Inhibition of ATP... 6 Target Theory... 7 Differences in the Capacities of Tumor and Normal Cells to Remove H 2 O Introduction... 7 Oxidants... 8 Xenobiotics can Generate Oxidants Oxidative Stress Antioxidant Enzyme Systems that Remove H 2 O Antioxidant Enzymes in Cancer vs. Normal Implications for Sensitivity of Tumor Cells to Pharmacological Ascorbate in Cancer Therapy Introduction Pharmacological Ascorbate in Cancer Therapy Chemistry of Ascorbate Role of H 2 O 2 in Cytotoxicity of Ascorbate Pancreatic Cancer Significance CHAPTER 2: MATERIALS AND METHODS Materials Cell Lines Methods of Cell Counting Exposure to 1,4-Benzoquinone Exposure to Oligomycin A Exposure to Pharmacological Ascorbate Intracellular ATP Assay Clonogenic Survival Assay Trypan Blue Staining Exposure of Cells for GSH and GSSG Determination GSH and GSSG Determination with HPLC-BDD viii

11 H 2 O 2 Removal Assay: Determination of Rate Constant by which Cancer Cells Remove H 2 O Measurement of Catalase Activity Inhibition of Catalase with 3-Amino-1,2,4-Triazole Transduction with Adenovirus Catalase Measurement of Ascorbate Oxidation in Cell Culture Medium with Clark Electrode Oxygen Monitor Western Blot for Catalase Statistics CHAPTER 3: MOLES OF A SUBSTANCE PER CELL IS A HIGHLY I NFORMATIVE DOSING METRIC Introduction Results Comparison of Dosing Metrics, 1,4-Benzoquinone as an Example Comparison of Dosing Metrics, Oligomycin A Sequential Addition vs. Bolus Addition Causality, Census of Agent and Reaction Targets Intracellular Volume Affects the Apparent Toxicity of 1,4-BQ Discussion Different Experimental Platforms can Result in a Wide Range of Exposures The Case of H 2 O 2 and mol cell -1 or mol cell -1 s -1 as a Metric for Exposure Target Theory Bolus vs. Sequential Additions Limitations Recommendations Conclusions Future Directions CHAPTER 4: DIFFERENCES IN THE CAPACITIES OF TUMOR AND NORMAL CELLS TO REMOVE H 2 O Introduction Results Normal Cells have Higher Capacities for the Removal of H 2 O 2 in Comparison to Tumor Cells Catalase Activity Varies Across Tumor Cell Lines and Plays a Major Role in the Removal of Extracellular H 2 O Manipulation of Catalase Activity Causes Significant Changes in the Capacity of Cells to Remove H 2 O Physical Parameters of Cells Effect the Capacity of Cells to Remove H 2 O Discussion Conclusions Future Directions ix

12 CHAPTER 5: IMPLICATIONS FOR SENSITIVITY OF CELLS TO PHARMACOLOGIC ASCORBATE IN CANCER THERAPY Introduction Results Pharmacological Ascorbate is Oxidized in Cell Culture Medium The Dose of Pharmacological Ascorbate is Best Specified on a per Cell Basis The Differential Sensitivity to Ascorbate Across Pancreatic Cancer Cell Lines Correlates with the Capacity at which they Remove H 2 O Modulation of Catalase Activity in the Same Cell Line Mimics Results Seen in Different Cell Lines Inhibition of Catalase Sensitizes PANC-1 cells to Pharmacological Ascorbate ATP Correlates with Clonogenic Survival Following P-AscH - Treatment Discussion Conclusions Future Directions CHAPTER 6: SUMMARY Moles of a Substance per Cell is a Highly Informative Dosing Metric Differences in the Capacity of Tumor and Normal Cells to Remove H 2 O Implications for the Sensitivity of Cells to Pharmacological Ascorbate in Cancer Therapy APPENDIX A: COMBINATION OF ASCORBATE AND BLEOMYCIN Material and Methods Results and Discussion Future Directions APPENDIX B: BASAL CELL LINE DATA REFERENCES x

13 LIST OF TABLES Table 3.1 Table 4.1 Table 4.2 Table 4.3 Table B.1 Physical and biological parameters of cell lines used...64 Rate constants (k cell ) for H 2 O 2 removal by tumor cells...87 Rate constants (k cell ) for H 2 O 2 removal by normal cells...88 Intracellular volumes (pl) of cell lines...94 Basal cell line information xi

14 LIST OF FIGURES Figure 1.1 Information reported in methods sections of papers for cell culture studies is often insufficient to repeat exact exposures Figure 1.2 Quinone/semiquinone/hydroquinone triad of 1,4-benzoquinone Figure 1.3 Products formed upon the Michael addition reaction of thiols and primary amines to 1,4-benzoquinone Figure 1.4 Antioxidant networks of cells Figure 1.5 Chemistry of Ascorbate Figure 3.1 Cells can be counted consistently with three different methods Figure 3.2 Figure 3.3 Figure 3.4 Figure 3.5 Figure 3.6 Figure 3.7 Figure 3.8 Dose of 1,4-BQ expressed as mol cell -1 allows direct comparisons between different experimental conditions and more accurately reports toxicity than initial concentration in medium Dose specified as mol cell -1 more accurately reports toxicity of 1,4-BQ in cell culture experiments than initial concentration in medium Clonogenic survival correlates directly with intracellular ATP concentration following exposure to 1,4-BQ Expressing dose as mol cell -1 yields more information and can be helpful when using biochemical tools in cell culture experiments: ATP per cell decreases with increasing dose of oligomycin A on a per cell basis A single bolus addition or sequential additions of 1,4-BQ can provide different toxicities based on the endpoint measured Glutathione is not depleted with 1:1 stoichiometry upon exposure of MIA PaCa-2 cells to 1,4-BQ ED 50 of 1,4-BQ correlates directly with intracellular volume and mass of protein per cell for C6, MB231, A549, MIA PaCa-2, and HepG2 cell lines Figure 3.9 Depiction of target theory and exposure to 1,4-BQ Figure 3.10 Exposure to xenobiotics when specified as mol cell -1 varies greatly when using different experimental platforms for cell culture Figure 4.1 Normal cells have a more robust capacity to remove extracellular H 2 O 2 than tumor cells Figure 4.2 Catalase activity varies across cancer cell lines and correlates with the rate constant of H 2 O 2 removal (k cell ) xii

15 Figure 4.3 Figure 4.4 Figure 4.5 Inhibition of catalase has a major effect on the H 2 O 2 removal capacity of HepG2 cells Increasing catalase activity in MIA PaCa-2 cells directly increases their capacity to remove H 2 O Rate constants of H 2 O 2 removal correlate (k cell ) with intracellular volume Figure 5.1 Ascorbate is oxidized in DMEM generating a flux of H 2 O Figure 5.2 Dose of ascorbate is better specified on a per cell basis (pmol cell -1 ) than as initial concentration in the medium (mm) Figure 5.3 Figure 5.4 Figure 5.5 Figure 5.6 Figure 5.7 Figure A.1 The capacity to remove H 2 O 2 varies across different pancreatic cancer cell lines Sensitivity to ascorbate varies across pancreatic cancer cell lines and correlates with the capacity to remove extracellular H 2 O 2 (k cell ) Transduction of MIA PaCa-2 cells with adenovirus catalase at increasing MOIs (1-25 MOI) increases resistance to ascorbate as seen by ED 50 for clonogenic survival Inhibition of catalase with 3-amino-1,2,4-triazole sensitizes PANC-1 cells to ascorbate parallel to the decrease in k cell Clonogenic survival correlates closely with ATP per cell following 1 h exposure to P-AscH Clonogenic survival of MIA PaCa-2 cells following exposure to low doses of bleomycin Figure A.2 Bleomycin enhances the cytotoxicity of P-AscH - to MIA PaCa-2 cells Figure A.3 Bleomycin enhances the cytotoxicity of P-AscH - in MB231 cells Figure A.4 Bleomycin enhances the cytotoxicity of P-AscH - in A549 cells Figure A.5 Intracellular [ATP] does not change after 96 h exposure to bleomycin and does not correlate with the decrease in clonogenic survival xiii

16 LIST OF ABBREVIATIONS 1,4-BQ 3-AT AscH - ATP CAT Cys DMSO EC 50 ED 50 FBS GPx GR GSH GSSG H 2 O 2 HOCl HRP MOI MPO NADPH NIH OCR OSCP PARP PBS PCB PE P-AscH - 1,4-Benzoquinone 3-Amino-1,2,4-triazole Ascorbate Adenosine 5 triphosphate Catalase Cysteine Dimethyl Sulfoxide Effective Concentration 50 % clonogenic survival Effective Dose 50 % clonogenic survival Fetal Bovine Serum Glutathione Peroxidase Glutathione Reductase Glutathione Glutathione Disulfide Hydrogen Peroxide Hypochlorous Acid Horseradish Peroxidase Multiplicity of Infection Myeloperoxidase Nicotinamide adenine dinucleotide phosphate oxidase National Institutes of Health Oxygen Consumption Rate Oligomycin Sensitivity Conferring Protein Poly (ADP-ribose) polymerase Phosphate Buffered Saline Polychlorinated Biphenyl Plating Efficiency Pharmacological Ascorbate xiv

17 phpa Prx QRB RNS ROS RSS SOD Trx XOR para-hydroxyphenylacetic acid Peroxiredoxins Quantitative Redox Biology Reactive Nitrogen Species Reactive Oxygen Species Reactive Sulfur Species Superoxide Dismutase Thioredoxin Xanthine Oxidoreductase xv

18 CHAPTER I: THEME, BACKGROUND, AND SIGNIFICANCE Theme The research program for this thesis is centered on applying concepts of quantitative redox biology to three separate, but interrelated projects. The overall goal was to achieve absolute quantitation in cell culture experiments in order to improve cell culture models as an initial step in the extrapolation to biological consequences of xenobiotics to humans. This includes utilizing absolute quantitation: (1) to introduce a new specification of dose for xenobiotics in cell culture systems, i.e. mole of a substance per cell; (2) to determine absolute differences between the antioxidant capacity of tumor and normal cells; and (3) to predict the implications of this new knowledge on the use of pharmacological ascorbate as an adjuvant in cancer therapy. We explore this with the central hypothesis that by using absolute quantitation we can make improvements to the reproducibility of cell culture studies that form the basis of evidence-based biological and toxicological research. By designing experiments that begin with a quantitative dosing metric and utilizing quantitation to produce absolute information from the results of experiments, we can better leverage data. Designing experiments that begin with a quantitative dosing metric has the potential to increase the reproducibility of experiments, as well as scale them appropriately across different experimental platforms without changing the absolute amount of compound that cells are exposed to. Utilizing quantitation as mole of a substance per cell in reporting the results of experiments allows for direct and absolute 1

19 comparisons to be made across different cell lines, which are difficult to discern when relative comparisons are made. In addition, direct comparison of experimental results from laboratories around the world can be made on an absolute basis. We propose that these improvements will lead to better predictions from such experiments. These enhancements to in vitro cell culture studies will increase the success in translation of data from in vitro experiments to in vivo animal studies and ultimately impact the success of extrapolation of basic science research to human health studies. Moles of a Substance per Cell is a Highly Informative Dosing Metric in Cell Culture Introduction When assessing toxicity or efficacy of xenobiotics (i.e. toxicants, pharmaceutical agents, and biochemical tools), dose is key. Currently, using animal models to test the toxicity of xenobiotics is considered by many to be the best measure of risk of that compound when humans are exposed. However, the U.S. National Institutes of Health (NIH) is moving to limit the use of animals in testing for toxicity. The NIH is attempting to spread the 3Rs approach, in which alternative methods to whole animal testing should involve: reduction, replacement, and refinement [1]. In order to be successful in this goal it will be exceedingly important to look closely at how dose is expressed in in vitro experiments. This will allow us to gain as much information as possible with the goal of more accurately predicting the biological consequences in humans. Reproducibility in scientific research has also become a major focus among the scientific community. This issue encompasses several aspects of laboratory science including: statistical considerations, laboratory standards, practices, and reporting [2, 3]. 2

20 This is largely a concern for pre-clinical research because basic research (i.e. both in vitro cell culture models and in vivo animal studies) is depended upon for translation into clinical research. The biological consequences upon exposure of cells in culture to a xenobiotic are not only dependent on the cell type, but are also determined by the physical aspects of experiments. For many xenobiotics (i.e. toxicants, pharmaceutical agents, and biochemical tools) specifying dose as moles per cell in addition to extracellular concentration improves the repeatability of experiments by providing an absolute basis for experimental design and will therefore increase the value of predictions derived. The overall goal of this research is to more quantitatively express dose in cell culture studies. We hypothesize that by using absolute quantitation of dose in cell culture experiments, these methods will reduce the ambiguity between different experiments. Here we discuss the current dosing metrics used in cell culture studies, the chemistries of 1,4- benzoquinone and oligomycin A that make them good candidates for a new dosing metric, and target theory. Dosing Metrics Used in Cell Culture Studies Cell culture models can vary vastly from one another due to the numerous different conditions used to carry out experiments. Some such aspects that can make a significant impact on cell culture include: the medium formulation used, fetal bovine serum added to the medium (i.e. amount and lot number), plastic of vessel used (i.e. surface area and type), and depth of medium (i.e. changes in oxygen tension). Dose is the central parameter in assessing the biological consequences in in vitro experiments. The 3

21 physical aspects of in vitro experiments (i.e. cell culture) that determine the absolute amount of compound cells are exposed to include: the volume of medium used, the amount of compound added to the medium, and the number of cells treated. The current metric used to express dose in cell culture studies is nominal concentration (initial concentration of compound in the medium), which includes the amount of compound (mass or moles) per volume of medium (ml or L) the compound is suspended in [4]. This metric itself does not however contain information that makes the number of cells that are exposed or the volume of exposure medium readily available and often this information is not found within the Methods section of literature. Figure 1.1 shows a survey of the literature (N =10) that was completed for the Methods section of papers that contained pharmacological ascorbate (P-AscH - ) in cell culture cytotoxicity studies. Within the Methods section of these papers, the number of cells seeded at the beginning of the experiment was often found. However, the number of cells at the time of treatment with P-AscH - was often not reported. Precise volumes of medium used at the time of exposure were also often not available in the Methods section. It is sometimes possible to approximate this information from the cytotoxicity assay used based on the culture vessel that the experiment was carried out in. Missing information on these experimental conditions has a significant impact on deducing the actual exposure of cells to a xenobiotic, making it difficult to repeat experiments using the same experimental conditions. These omissions contribute to some of the reproducibility issues that are currently a concern in pre-clinical and basic science research [5]. Figure 1.1 demonstrates the difficulty in finding an absolute amount (mole) of compound that any number of cells is being treated with from these published 4

22 findings, further demonstrating that there is missing information about the details of experiments that could be provided. This motivated us to think of dosing metrics that may include absolute quantitation and account for the physical aspects of exposure in cell culture experiments. Moles of xenobiotic per cell provides a dosing metric in which pertinent information is included within the metric to repeat the experiment or scale an experiment to different platforms that require differing number of cells and volumes of medium. 1,4-Benzoquinone: Chemistry and Toxicity 1,4-Benzoquinone (1,4-BQ) is a common metabolite of benzene. It is among the simplest of the quinones in terms of structure; it is biologically and environmentally relevant, as it has been implicated as one of the major players in the carcinogenicity of benzene [6, 7]. Quinones can play both vital and detrimental roles in the cell. Some quinones result from the metabolism of parent compounds in the body (i.e. 1,4-BQ), which redox cycle between quinone, semiquinone, and hydroquinone species, generating superoxide anions (O 2 - ) and other reactive oxygen species (e.g. H 2 O 2 and hydroxyl radical, HO ) that can have potentially deleterious effects intracellularly (Figure 1.2) [6]. Other quinones that generate semiquinone radicals, such as ubiquinones, play vital roles in the cell, as their ability to redox cycle makes them critical components of the electrontransport chain [8]. While the redox cycling between quinones and semiquinone radicals results in one possible mechanism of toxicity, another mechanism of toxicity is the ability for quinones to covalently bind to biological nucleophiles. Quinones can be highly reactive, 5

23 as they are oxidants and electrophiles that readily form adducts with biological nucleophiles, including proteins, glutathione, and nucleic acids via Michael addition (Figure 1.3) [9,10]. Figure 1.3 shows the covalent binding of both a protein and glutathione to 1,4-BQ at the thiol and amine groups. Not only does this covalent bonding potentially disturb the macromolecular/cellular integrity and function, but adducted quinones can continue to redox cycle. This increases residence times of these harmful compounds in cells and generates more semiquinone radicals and subsequent reactive oxygen species [6, 11]. Because 1,4-BQ readily forms covalent bonds with amine- and thiol-containing compounds causing irreversible chemical changes to those biomolecules, we hypothesized that using mol cell -1 to express dose in cell culture would be more appropriate for 1,4-BQ than nominal concentration. Oligomycin A: Inhibition of ATP Unlike 1,4-BQ, which covalently binds to its target molecule causing a permanent, irreversible change, oligomycin A forms a tight complex with its target. Oligomycin A is an antibiotic that is isolated from Streptomyces diastatochromogenes. It inhibits the mitochondrial H + -ATP synthase, complex V of the electron transport chain. It does this by non-covalently binding to the oligomycin sensitivity-conferring protein (OSCP); at high concentrations it can also inhibit the Na + +K + -ATPase [12]. Oligomycin A serves as a widely used biochemical tool for studying mitochondrial respiration and oxidative phosphorylation. It forms a tight complex with OSCP with a V max = 0.77 µmol 6

24 min -1 mg -1, K m = 120 µm, and K i = 11 nm [13]. The dissociation constant of oligomycin A with the Na + -K + ATPase is K d 10-6 M [12, 13, 14]. Because oligomycin A forms a tight complex with its target protein, we hypothesized that mol cell -1 would also serve as a better metric for expressing dose than initial concentration. Target Theory Exposure has been thoroughly considered in the field of radiation biology. The concept of target theory has been suggested in this field to describe the relationship between radiation dose and an observed effect. As first described by Lea in 1946, target theory suggests that there are a certain number of sensitive points or targets within cells and when enough of those targets are damaged the cell will undergo some kind of biological effect or endpoint [15]. One of the goals of this study is to determine whether target theory also applies to toxicology when assessing direct-acting toxins that form covalent bonds with their target molecules. Differences in the Capacities of Tumor and Normal Cells to Remove H 2 O 2 Introduction The antioxidant capacity of different cells is often indicative of the metabolism and function of that cell (or tissue). Since reactive oxygen species are constantly being formed during normal metabolism, there must exist several antioxidant defenses to remove them. Reactive oxygen species (ROS) are largely implicated in human disease, 7

25 whether this is in causation or consequence. They are even utilized in the treatment of disease (i.e. cancer). Along with the endogenous production of ROS, many xenobiotics (exogenous sources; i.e. pharmaceutical agents, toxicants, environmental pollutants) are capable of generating ROS in the body as primary sources and during their metabolism. Similarly, these oxidative species must be scavenged biologically by antioxidants and/or antioxidant enzyme defenses to prevent oxidative damage that may ensue under oxidative stress. Differences in the antioxidant capacity of cells may therefore indicate which cells are better equipped to deal with an oxidative insult from a xenobiotic (i.e. toxicants or chemotherapeutics) vs. cells that will undergo excessive oxidative damage upon exposure. The overall goal of this research was to quantitatively determine the capacity of cells (tumor and normal) to remove H 2 O 2. We hypothesized that normal cells would have a higher capacity to remove H 2 O 2 than tumor cells, but there would be a large range of capacities across different cell types that would reflect their ability to deal with a H 2 O 2 generating agent (i.e. P-AscH - ) that we explore further in Chapter 5. Here we review the different types of oxidants, their effects, sources, and the antioxidant defenses responsible for their removal, with a particular focus on H 2 O 2. Oxidants Reactive oxygen species (ROS) are produced during the normal metabolic processes cells undergo, but can also be produced exogenously by xenobiotics upon exposure [16]. Both enzymatic and non-enzymatic processes are responsible for the 8

26 endogenous production of these species. Among the most noted enzymes responsible, are nicotinamide adenine dinucleotide phosphate (NADPH) oxidase, xanthine oxidoreductase (XOR), and myeloperoxidase (MPO) [17]. Some of the most biologically relevant oxidants include: superoxide (O 2 - ), hydrogen peroxide (H 2 O 2 ), hydroxyl radical (HO ), hypochlorous acid (HOCl), peroxyl, hydroperoxyl, and alkoxyl radicals [16,17]. Superoxide, hydroxyl, peroxyl, hydroperoxyl, and alkoxyl radicals are all free radicals, which is any species that contains one or more unpaired electrons [18]. H 2 O 2 and HOCl are non-radical oxidizing agents that can be easily converted into radical species [18]. H 2 O 2 is generated via several biological processes in the cell and is considered to be a major redox-signaling molecule, through its oxidation of thiol proteins [19]. It is a by-product of cellular respiration and an end-product of metabolic reactions, such as peroxisomal oxidation pathways. One of the prevalent reactions generating H 2 O 2 is the dismutation of O 2 - by superoxide dismutase (SOD) to yield H 2 O 2 and O 2. H 2 O 2 is a strong 2-e - oxidant (E = 1.32 V at ph 7.0) and a weak 1-e - oxidant (E = 0.32 V at ph 7.0) [19]. H 2 O 2 readily crosses cellular membranes, but has a high activation energy barrier that makes it unreactive to most biomolecules. The biomolecules that H 2 O 2 will react most favorably with include: transition metal centered proteins, selenoproteins, and selected thiol proteins [19]. While H 2 O 2 itself is not a very reactive oxidant, it can be activated in the presence of catalytic metal ions (i.e. Fe 2+ and Cu 1+ ) to generate hydroxyl radical, HO, which is the most oxidizing of the oxygen species. Hydroxyl radical can damage cellular components and DNA, as explained in further detail below. 9

27 Xenobiotics can Generate Oxidants There are several xenobiotics that are capable of generating oxidants at levels that can overwhelm the antioxidant defense of cells. This includes both environmental pollutants (i.e quinone-containing compounds (e.g. 1,4-BQ), cigarette smoke, radon, ozone (O 3 ), nitric oxide (NO x ), sulfur dioxide (SO 2 ), car exhaust, X-rays, and UV light) and chemotherapeutics (i.e. P-AscH -, histone deacetylase inhibitors, proteasome inhibitors, and redox cycling agents (e.g. doxorubicin)) that are often utilized for this toxic ability in cancer therapy [20, 21]. As an example, polychlorinated biphenyls (PCBs) can be oxidized to hydroquinones and quinones [10]. These species can be converted to their corresponding semiquinone free radicals. These radicals can enter a futile redox cycle resulting in the formation of reactive oxygen species, namely O 2 - and H 2 O 2 [10]. This is demonstrated in Figure 1.2. This quinone moiety is present in many environmental pollutants (metabolite of benzene, 1,4-BQ) as well as chemotherapeutics (i.e. doxorubicin). High levels of reactive oxygen species lead to a more oxidized redox environment in the cell, resulting in cellular damage and possibly death. There exist however, several antioxidants and antioxidant enzyme networks in the cell to metabolize them and are responsible for maintaining the reducing environment of the cell. Oxidative Stress Under normal conditions, there is a well-maintained balance of reactive oxygen, nitrogen, and sulfur species (ROS/RNS/RSS) and the antioxidants and enzyme systems 10

28 needed to metabolize them. The reducing environment of the cell that is responsible for helping prevent free radical damage is maintained by a multitude of antioxidant enzymes and vitamins, including: superoxide dismutase (SOD), catalase, glutathione peroxidase (GPx), peroxiredoxins (Prx), glutathione (GSH), ascorbate (vitamin C), alpha-tocopherol (vitamin E), and thioredoxin (Trx) [17, 18]. Oxidative stress is defined as a disturbance in this balance between the production of reactive oxygen species (or free radicals) and the antioxidant defenses that detoxify them, or in other words an alteration in the redox state of the cell [18]. This can lead to toxic effects in cells because oxidizing species that are not detoxified can cause oxidative damage to biomolecules (i.e. lipids, proteins, and DNA) leading to a multitude of downstream effects (i.e. inflammation or tissue injury) [21]. In particular, hydroxyl radicals can initiate lipid peroxidation, leading to damage of cellular membranes as well as lipoproteins [21]. Oxidation of proteins can lead to loss of or change in the protein structure that ultimately results in loss of function. This can affect enzyme activity when enzymes are the proteins damaged [21]. DNA is a biomolecule that undergoes oxidative damage at a high rate from normal metabolic processes. It has been estimated that the DNA in each cell is exposed to 10 4 oxidative hits per day, which causes 20 lesions [22]. Many of these lesions will go on to be repaired; however, a small percentage may remain unrepaired and accumulate over the lifespan. All of these damaging effects can have further downstream consequences, as well. Oxidative damage to the DNA is particularly deleterious and has been implicated in causing mutations that may ultimately result in cancer [21]. Oxidative damage to lipids and proteins has been suggested to cause downstream effects leading to disease-states as 11

29 well. Damage to lipids has been suggested to contribute to atherosclerosis, and protein oxidative damage causative of cataract formation [21]. Antioxidant Enzyme Systems that Remove H 2 O 2 As mentioned above, while H 2 O 2 is not very reactive with biomolecules, it can accumulate to relatively high concentrations in the cell due to its stability and can be converted to more reactive species inside the cell (i.e. HO ) [19]. The removal of H 2 O 2 by antioxidant enzymes is therefore very important. A schematic of the network of antioxidant enzymes is shown in Figure 1.4. The three main antioxidant enzymes responsible for the removal of H 2 O 2 include: catalase, GPx, and Prx (Figure 1.4). Catalase Catalase has been well-studied for its role in protecting cells from the toxic effects of H 2 O 2. Catalase is found in most aerobic living organisms. It is largely localized to the peroxisomes of mammalian cells [23]. Catalase is highest in erythrocytes, liver, and kidney [24]. It catalyzes the decomposition of H 2 O 2 into water and oxygen (reaction 1). When high fluxes of H 2 O 2 are present intracellularly, catalase is the enzyme that is primarily responsible for its removal [25]. 2 H 2 O 2 2 H 2 O + O 2 (1) Catalase is a tetramer of four identical polypeptide chains, each one with its own active site [23]. Each of the four identical subunits contains 527 amino acid residues, one heme group, namely iron (III) protoporphyrin IX, and a tightly bound NADPH molecule [23]. While not critical to the activity of catalase, the tightly bound NADPH protects catalase against inactivation by H 2 O 2 [26, 27]. 12

30 Catalase has been implicated in different oxidative stress-related disease states [23]. This includes catalase protein and activity levels, as well as polymorphisms of the CAT gene (catalase gene) [23]. In particular, polymorphisms of catalase have been implicated in several cancers, including breast cancer and hepatocellular carcinoma [28, 29, 30]. On a molecular level, the amounts of catalase protein are dependent on regulation of the CAT gene expression, which is regulated in part by peroxisome proliferator-activated receptor γ (PPARγ), tumor necrosis factor α (TNF-α), p53 protein and hypermethylation of CpG islands in the catalase promoter [23]. Regulatory mechanisms of catalase occur at the transcriptional, post-transcriptional and posttranslational levels [31]. Some of the mechanisms of catalase regulation in regards to cancer will be discussed below in the section Antioxidant Enzymes in Cancer vs. Normal Cells. Glutathione Peroxidases There are at least four different forms of GPx found in different compartments throughout the cell including: the nucleus, cytoplasm, and mitochondria [32]. Contrary to catalase, cytosolic GPx (GPx1) is responsible for removing low fluxes of H 2 O 2 (reaction 2). Its removal of H 2 O 2 requires GSH recycling by glutathione reductase (GR), which is dependent on NADPH production [25]. 2 GSH + H 2 O 2 GSSG + 2 H 2 O (2) 13

31 Peroxiredoxins Another antioxidant enzyme system responsible for the removal of H 2 O 2 is the peroxiredoxins (Prx). They are a family of thiol proteins that have high specificity for peroxides. These enzymes protect cellular components by removing low levels of peroxide that are produced during normal metabolism. Although they are slower at metabolizing H 2 O 2 than GPx1, it has been suggested that Prxs dispose of most H 2 O 2 generated intracellulary at low concentrations [23]. There are six Prx family members that are distributed in the cytoplasm, mitochondria, as well as other cellular compartments. Two different types of Prxs exist, 1-Cys and 2-Cys. Four of the six Prxs are 2-Cys (e.g. Prx 1,2,3,4). The 2-Cys Prxs function via an initial oxidation of a highly reactive (peroxidative) cysteine (Cys) that will condense with a second cysteine on an adjacent subunit [33]. This results in a disulfide-linked dimer, which is reduced by thioredoxin reductase with reducing equivalents from thioredoxin (Trx) and NADPH [33]. Prx will undergo reversible oxidations (reaction 3 and 4) to disulfides [34]. At high concentrations of H 2 O 2 the Cys in the active site can be over-oxidized (i.e. hyperoxidized) to cysteine sulfinic acid (Cys-SO 2 H) and even further to the cysteine sulfonic acid (Cys-SO 3 H) [35, 36]. This over-oxidation when encountered with high concentrations of H 2 O 2 and need for recycling limits the ability of Prx to function at high H 2 O 2 concentrations. Prx(reduced) + H 2 O 2 Prx(oxidized) + 2H 2 O (3) Prx (oxidized) + Trx(reduced) Prx(reduced) + Trx(oxidized) (4) 14

32 Kinetic Modeling of Peroxide Removal Systems Makino et al. constructed a metabolic model to describe how H 2 O 2 is eliminated in mammalian cells [25,37]. Catalase, GPx and Prx are all involved in the elimination of intra- and extracellular H 2 O 2. The contribution of Prx to the removal of high fluxes of H 2 O 2 is thought to be small [37]. They showed two reactions involved in the removal of H 2 O 2 that are very kinetically different [25]. The reaction attributed to catalase was first order, while the one attributed to GPx was Michaelis-Menten-like [25]. They further showed that the plot of concentration dependence of H 2 O 2 removal rate was biphasic and that at concentrations greater than 50 µm, there was a linear dependence [25]. They were able to estimate that GPx eliminated % of H 2 O 2 at physiological concentrations (less than 10 µm), however at higher concentrations the concentration of first-order removal (i.e. catalase) increases [25]. Antioxidant Enzymes in Cancer vs. Normal Biochemical studies of various different normal tissues have shown that the endogenous levels of antioxidant enzymes vary greatly across different tissue types [38]. It has been postulated that this reflects the different metabolism across different organs [39]. The intrinsic levels of antioxidant enzymes are low in a majority of cancer cell types as compared to non-transformed cells [38, 39]. Studies have shown that all but one human cancer cell type, a human renal adenocarcinoma, showed low levels of both catalase and GPx [39]. This suggests that the vast majority of cancer cells may lack the machinery to efficiently detoxify H 2 O 2. While in general the levels of catalase were low in cancer cells, catalase activity also varied greatly across different cancer cell lines [38]. 15

33 This may or may not correspond to a differential in capacity to remove H 2 O 2 and sensitivity to H 2 O 2 -producing agents. There have been several regulatory mechanisms explored to explain why catalase levels may be lower in cancers. It has been previously observed that down-regulation of catalase can be caused by hypermethylation of a 5 -cytosine-phosphate-guanine-3 (CpG) island in the catalase promoter during prolonged exposure to reactive oxygen species [40, 41]. This has been noted in hepatocellular carcinoma cells (HCC), showing a downregulation of catalase expression in HCC cells [40]. Down-regulation of antioxidant enzymes is recognized as being involved in the neoplastic transformation of cells [42]. Cancer cells, which have a higher proliferative capacity, showed higher H 2 O 2 levels and decreased catalase activities in comparison to normal cells [42]. One reason given for the higher levels of catalase in normal cells in comparison to tumor cells is that hydrogen peroxide functions as a second messenger of mitogenic signaling processes, which stimulates proliferation [23]. Contrarily, catalase suppresses growth factor dependent activation of mitogen-activated protein kinase (MAPK) and proliferation of cells [23]. Implications for Sensitivity of Tumor Cells to Pharmacological Ascorbate in Cancer Therapy Introduction Ascorbate (AscH - ) functions as a versatile reducing agent in biology. When used at low, physiological concentrations obtained through normal, healthy nutrition it exhibits antioxidant properties; however, when used at pharmacological doses (P-AscH - ) 16

34 achievable through IV delivery it can readily oxidize and deliver a high flux of H 2 O 2. This unique property of P-AscH - is currently being investigated as an adjuvant treatment in cancer therapy. Several in vitro and in vivo studies have shown a differential toxicity of P-AscH - between cancer cells and normal cells of the same tissue type. Additionally studies have implicated the H 2 O 2 produced from the oxidation of P-AscH - as the mediating factor in its toxicity to cancer cells. The initiative of adding P-AscH - therapy as an adjuvant to many standard of care protocols in cancer therapy has increased over the last two decades. Laboratories at The University of Iowa are participating largely in this initiative with in vitro, in vivo and clinical studies [43, 44, 45, 46]. This research aims to add to this initiative by exploring the mechanisms of P-AscH - in cancer therapy through the quantitative determination of the capacity of cells to remove extracellular H 2 O 2. The overall goal of this research is to determine whether the differential sensitivity of P-AscH - across different pancreatic cancer cells may be due to differences in cell s ability to remove H 2 O 2 (as determined in Chapter 4). We hypothesize that there will be a differential sensitivity to P-AscH - across different pancreatic cancer cells that is related to their ability to remove H 2 O 2. Pharmacological Ascorbate in Cancer Therapy In 1974, Ewan Cameron and Linus Pauling hypothesized that the use of ascorbate in cancer therapy could inhibit tumor growth [47]. In initial trials, cancer patients were given intravenous ascorbate (10 g/day) for 10 days, followed by oral ascorbate (10 g/day) indefinitely [47]. While these studies were uncontrolled, the outcomes were positive, 17

35 indicating decreased tumor growth and an increase in patient survival following the treatment [47]. This study was followed up by researchers at the Mayo Clinic, who did two controlled double-blind clinical trials in which only oral ascorbate (10 g/day) was given and contrary to the preliminary trial by Cameron, no difference was observed between the groups of patients [47, 48]. The different results from these clinical trials were proposed to be due to pharmacokinetics of ascorbate, i.e. differences between the oral and intravenous administration routes [49]. While the uptake of ascorbate from the intestinal tract is a tightly regulated process, pharmacokinetics of ascorbate indicate that when it is administered intravenously this can be bypassed and result in significant elevation of ascorbate in the plasma [47, 50]. There have been several Phase 1 clinical trials dedicated to researching the safety of P-AscH - in cancer therapy [43, 44, 45, 46]. However, the mechanism of H 2 O 2 -mediated cytotoxicity by P-AscH - as well as the selectively of it towards cancer cells, remains unclear. Chemistry of Ascorbate Ascorbic acid (AscH 2 ; vitamin C) can be both an antioxidant and a pro-oxidant. It is classically known for its antioxidant properties, but has recently been investigated for utilization in cancer therapy due to its pro-oxidant properties. The concentration of vitamin C obtained from normal nutrition in the plasma of healthy humans ranges from 40 to 80 µm [51, 52, 53]. At concentrations such as these, vitamin C plays the role of an endogenous antioxidant and cellular reducing agent. 18

36 At physiological ph (ph 7.0), the ascorbate monoanion (AscH - ) is the dominant form of vitamin C. Ascorbate is an electron donor that is able to donate one or two electrons. This property of AscH - allows it to serve as an excellent reducing agent and donor antioxidant. Ascorbate will undergo two one-electron oxidations, forming ascorbate radical (Asc - ) and dehydroascorbic acid (DHA). Asc - is not very reactive and will dismute to AscH - and DHA (reaction 5). It can also be enzymatically reduced back to ascorbate via NADH and NADPH-dependent reductases [50]. Ascorbate can readily donate an electron to oxidizing radicals (i.e. HO ), detoxifying the species and leading to the production of Asc -, which can then be recycled [54]. These processes are shown in Figure Asc - + H + AscH + DHA (5) AscH - can also exhibit pro-oxidant properties and whether the effects of AscH - observed will be antioxidant or pro-oxidant is very much dependent on the concentration of AscH - [50]. Ascorbate readily autoxidizes via ascorbate dianion (Asc 2- ) producing Asc - and O 2 - (reaction 6 and 8). This process is ph-dependent. The oxidation of AscH - is enhanced in the presence of catalytic metals (i.e. Fe 3+ ) (reaction 7 and 8). In short, AscH - can reduce ferric (Fe 3+ ) to ferrous (Fe 2+ ) iron (reaction 7). Fe 2+ can then react with O 2, reducing it to O 2 - (reaction 8). Superoxide radical readily dismutes nonenzymatically or via SOD to form H 2 O 2 (reaction 9). At the high plasma concentrations (> 20 mm) that are achievable through intravenous injection, the oxidation of AscH - will result in the production of a high flux of H 2 O 2 [55]. Asc 2 + O 2 Asc - + O 2 - (6) 19

37 AscH + Fe 3+ Asc - + Fe 2+ (7) Fe 2+ + O 2 Fe 3+ + O 2 - (8) O O H + H 2 O 2 + O 2 (9) Role of H 2 O 2 in Cytotoxicity of Ascorbate It has been well established that the cytotoxicity of pharmacological ascorbate (P- AscH - ) to cancer cells observed during in vitro studies is largely due to its generation of H 2 O 2 upon the oxidation of P-AscH - in the medium [55, 56]. While H 2 O 2 itself is not a very reactive oxidant, it can be activated in the presence of catalytic metal ions (i.e. Fe 3+ and Cu 2+ ) to HO, which is the most oxidizing of the oxygen species (Reactions 10 and 11). Hydroxyl radical can damage cellular components and DNA. Fe 3+ + O 2 - Fe 2+ + O 2 (10) Fe 2+ + H 2 O 2 Fe 3+ + HO (11) Ascorbate delivered at pharmacological concentrations has shown selective toxicity to several different tumor cell types [55]. While this selective cytotoxicity has been observed to be dependent on the generation of H 2 O 2, the mechanism by which this occurs is still under investigation. Several mechanisms for how the H 2 O 2 generated by P- AscH - elicits its cytotoxicity to tumor cells have been hypothesized and researched, including: the depletion of ATP leading to cell death to tumor cells. There have been three proposed mechanisms for how this may occur [57]. (1) H 2 O 2 can damage DNA, which will activate poly(adp-ribose) polymerase (PARP). PARP catabolizes NAD +, depleting substrate for NADH and ultimately ATP synthesis [58, 59]. (2) H 2 O 2 can oxidize glutathione to glutathione disulfide (GSSG). Reducing GSSG back to glutathione 20

38 will utilize NADPH provided by glucose from the pentose shunt. This decrease in glucose available for glycolysis or NADH production will decrease ATP levels [60, 61]. (3) H 2 O 2 could damage the mitochondria directly, decreasing ATP synthesis [62, 63, 64]. Questions still remain about why P-AscH - is eliciting these responses in tumor cells while at the same conditions there is no toxicity observed in normal cells. This research aims to determine whether the differential sensitivity of P-AscH - may be due to differences in the cell s ability to remove H 2 O 2 by looking at a particular model, pancreatic cancer, in which P-AscH - has been investigated for use as an adjuvant in cancer therapy [44, 46, 56, 65, 66, 57, 67]. Pancreatic Cancer Pancreatic cancer accounts for about 3 % of all new cancer cases reported each year (SEER Stat Fact Sheet) *. While the prevalence of pancreatic cancer is lower than some of the other types of cancer, the median survival following discovery of the disease is very low, months with surgical resection and 6 months without [68]. The percent of people surviving five years after being diagnosed with pancreatic cancer is only 7.2 % (SEER Stat Fact Sheet) *. Pancreatic cancer is the 8 th most common cancer, but the 4 th leading cause of cancer death. This is likely due to the difficulties in diagnosing pancreatic cancer. It is often not discovered until it is in the later stages, making it inoperable and it often metastasizing. Surgical resection of the tumor has the most potential for curative treatment of pancreatic cancer; unfortunately, surgery is an option for only about 20 % of patients *

39 with pancreatic cancer due to this late diagnosis. Whether or not surgery is performed, treatment of pancreatic cancer often includes combination therapies of radiation therapy and chemotherapy. In the recent past, the standard of care has consisted of gemcitabine (Gemzar), and possible addition of erlotinib (Tarceva; targeted therapy). Recently, the standard of care has been FOLFIRINOX, which consists of fluorouracil (5-FU), leucovorin (Wellcovorin), irinotecan (Camptosar), and oxaliplatin (Eloxatin) (cancer.net). Due to the multitude of side-effects resulting from this combination treatment, it can only be offered to patients that are otherwise healthy and these options have not significantly improved the median survival of patients. This has fueled the search to find more effective treatments for pancreatic cancer. There have been two phase I clinical trials at The University of Iowa to include P- AscH - as an adjuvant to the standard of care for patients with pancreatic cancer [44, 46]. This is therefore a very clinically- relevant disease model to further study the mechanism of P-AscH - as an adjuvant in cancer therapy. Significance With such a heavy reliance on in vitro models (i.e. cell culture studies) to guide clinical and translational research, it is imperative that these model systems be reproducible, informative, and allow for valuable predictions to be derived from experiments in which they are used. Increasing the quantitation in these studies can allow us to do this within our research labs, as well as across different research labs in different parts of the world. 22

40 Since dose is a central parameter in determining the biological consequences of xenobiotics (i.e. toxicants, pharmaceutical agents, and biochemical tools), we predict that using a dosing metric that includes quantitation of the absolute amount of compound used per number of cells exposed (mol cell -1 ) will improve the repeatability of the experiments performed. Using this dosing metric can vastly improve experimental design and has the potential to save time, money and resources. We also predicted that by utilizing quantitative methods, we would be able to determine absolute differences in the capacity of H 2 O 2 removal between several cell lines that are vastly different (tumor vs. normal, >10 different tissue origins). Quantitation allows direct and absolute comparisons to be made across different cell lines, which are difficult to discern when relative comparisons are made. Discerning differences in the capacities of cells to remove H 2 O 2 can have broad implications in predicting which cells might be most sensitive to a H 2 O 2 -generating xenobiotic, whether this be a toxicant or a pharmaceutical agent. We applied this theory to investigate whether different pancreatic cancer cells that demonstrated a wide-range of H 2 O 2 removal capacities, would show a differential sensitivity to P-AscH -. This application can provide additional evidence that H 2 O 2 is involved in the mechanism of P-AscH - toxicity to cancer cells and that catalase activity is critical in removing this H 2 O 2. These results indicate that catalase activity (or overall H 2 O 2 - removal capacity) could be a good marker for which tumors will respond to P- AscH - therapy. This information can also be used in finding combination therapies that may increase the efficacy of treatment for those tumors with higher catalase activities. 23

41 Overall, increasing quantitation in the biological and biochemical assays we perform as part of basic science research can improve them greatly. These improvements include: enhancing the reproducibility of cell culture studies, increasing the information obtained from these types of experiments, and increasing the value of the predictions we derive from the data obtained. These are all current initiatives of NIH, with the goal that improvement to these aspects of basic science in vitro results may lead to more translatability to animal models and have more meaningful connections to human health concerns. 24

42 % of papers reviewed Dose as concentra5on # cells seeded # of cells treated Exposure medium volume mol/cell could be calculated Figure 1.1: Information reported in methods sections of papers for cell culture studies is often insufficient to repeat exact exposures. A survey of the literature was completed for the methods section of papers that contained P-AscH - in cell culture cytotoxicity studies (N = 10). The number of cells seeded was found in the methods section of 90 % of the papers surveyed. The number of cells treated was only reported in 20 % of the studies along with precise exposure volumes of medium. With this missing information, it is difficult to determine the absolute amount (mole) of compound per cell used. 25

43 O 1 e - reduction O O O 2 O 2 2 e - reduction + Comproportionation (reaction 3) 1 e - reduction O OH O 2 O 2 OH Figure 1.2: Quinone/semiquinone/hydroquinone triad of 1,4-benzoquinone. The 1,4-benzosemiquinone radical is formed by the one-electron reduction of the quinone, which can be further reduced to a hydroquinone. Hydroquinones can autoxidize to yield semiquinone radicals as well. In this process superoxide anion (O 2 - ) can result from the reaction with molecular oxygen (O 2 ) [69]. O 2 - can dismute enzymatically via SOD and non-enzymatically to form H 2 O 2. When both quinone and hydroquinone are present, semiquinone radical will also be present via comproportionation [10]. The forms of the species shown are those that would dominate at ph 7. 26

44 OH SG + GS - OH O OH + Protein (RS - ) S-Protein O + Protein (RNH 2 ) OH OH NH-Protein OH Figure 1.3: Products formed upon the Michael addition reaction of thiols and primary amines to 1,4-benzoquinone. 1,4-Benzoquinone covalently bonds with thiols at k = 1 x 10 6 M -1 s -1 in ph 7 buffer and with amines at k = 1 x 10-2 M -1 s -1 in ph 7.4 buffer [9, 10]. Adapted from [70]. 27

45 Figure 1.4. Antioxidant networks of cells. Shown here are the three nodes of the hydrogen peroxide removal system in the cell. Superoxide dismutase (SOD) dismutes superoxide (O 2 - ) into H 2 O 2 and O 2. This H 2 O 2 or H 2 O 2 from other sources can be removed by glutathione peroxidase (GPx), peroxiredoxin (Prx), or catalase (CAT). (node 1) GPx is a selenoprotein that utilizes reducing equivalents from glutathione (GSH) to reduce H 2 O 2 to 2 H 2 O. Glutathione reductase (GR) reduces glutathione disulfide (GSSG) back to GSH, to maintain the function of GPx. (node 2) Prx utilizes reducing equivalents from thioredoxin (Trx) to remove H 2 O 2. (node 3) CAT reduces H 2 O 2 to H 2 O and O 2, but does not require reducing equivalents in the process. Figure from reference [71]. 28

46 Figure 1.5. Chemistry of ascorbate. Ascorbic Acid (AscH 2 ), has two ionizable hydroxyl groups and is water soluble. It has two different pk a s, pk 1 = 4.2 and pk 2 = The ascorbate monoanion (AscH - ) is the dominant form at ph 7.4. Ascorbate undergoes two consecutive 1-e - oxidations to ascorbate radical (Asc - ) and dehydroascorbic acid (DHA). The ascorbate radical will readily dismute to ascorbate and DHA. Figure from reference [50]. 29

47 CHAPTER 2: MATERIALS AND METHODS Materials 1,4-Benzoquinone, dimethyl sulfoxide, oligomycin A, adenosine 5 -triphosphate (ATP) disodium salt hydrate, 1-octane-sulfonic acid, sodium phosphate, acetonitrile, GSH, glutathione disulfide (GSSG), diethylenetriaminepentaacetic acid (DETAPAC), horseradish peroxidase (HRP), para-hydroxyphenylacetic acid (phpa), and 3-amino- 1,2,4-triazole (3-AT) were obtained from Sigma Aldrich (St. Louis, MO, USA). Ascorbic acid was obtained from Avantor Performance Materials, Inc. (Center Valley, PA, USA). Adenovirus catalase was obtained from the University of Iowa Viral Vector Core Facility (Iowa City, IA, USA). Cell Lines MIA PaCa-2, C6, HepG2, MDA-MB231, and A549 cells were purchased from American Type Culture Collection (Manassas, VA). Origin and bio-physical parameters of each cell are listed in Table 3.1. All cells were cultured in Dulbecco s modified eagle medium (DMEM) with high glucose from Invitrogen (Grand Island, NY), supplemented with 10 % fetal bovine serum (FBS) and Penicillin Streptomycin (80 units ml -1 ) at 37 C, 5 % CO 2. Sufficient medium was prepared to complete an experiment, including all replicates. All media preparations for a set of experiments contained FBS from the same lot number to minimize variation between experiments. Two patient-derived cell lines, 339 and 403, were obtained from the Medical College of Wisconsin (Milwaukee, WI). A375, Cal27, FaDu, H292, H1299, U87, U118, H6c7, melanocytes, normal human fibroblasts (12 and 46 years old), normal human astrocytes, HBePC, red blood cells and 30

48 FHs74int cells were donated from neighboring labs and were only used in experiments to determine their rate constant for H 2 O 2 removal. 339 and 403 cells were cultured in DMEM nutrient mixture F-12 (Ham) medium from Invitrogen (Grand Island, NY), supplemented with 6 % FBS, Penicillin Streptomycin (80 units ml -1 ), 0.1 % epidermal growth factor (EGF) human recombinant, 0.4 % Bovine Pituitary Extract, 4 % hydrocortisone, % insulin human recombinant and GlutaMAX I at 37 C, 5 % CO 2. Sufficient medium was prepared to complete an experiment, including all replicates. All media preparations for a set of experiments contained FBS from the same lot number to minimize variation between experiments. The volumes of MIA PaCa-2, C6, HepG2, MDA-MB231, and A549 cells were determined in triplicate using both a Moxi TM Z Mini Automated Cell Counter (ORFLO ) and a Z2 TM Coulter Counter (Beckman Coulter, Inc.). Cells were trypsinized, collected in PBS, and centrifuged to pellet. Cell pellets were then suspended in ISOTON II Diluent (Beckman Coulter, Inc.) and the cell size (intracellular volume) was immediately determined using both cell sizing methods. Protein mass per cell was measured in each cell line (i.e. MIA PaCa-2, C6, HepG2, MDA-MB231, and A549 cells) using the SDS-Lowry protein assay, n = 3 for each cell line. Cells were plated in triplicate 60 mm x 15 mm culture dishes at equal densities (500,000 cells per dish). Cells were allowed to adhere and grow 48 h after plating. Cells were then trypsinized, collected in PBS, centrifuged (193 g) and suspended in a small volume of PBS (200 µl). A portion of the cell pellet was used for counting with the hemocytometer, so the total number of the cells in each pellet was known. Cell pellets were kept in -80 C until time of analysis. Prior to protein mass determination, cell 31

49 pellets were thawed and sonicated for 1 min to lyse cells. Protein mass per cell was measured using the SDS-Lowry and cell counts allowed for the determination of protein mass per cell in each sample. Albumin from bovine serum (Sigma Aldrich; Cohn Fraction V, Sigma-A2153) was used as a standard [72]. Methods of Cell Counting MIA PaCa-2 cells were seeded in three T25 cell culture flasks. Cells were allowed to adhere and grow 48 h after plating. Cells were then trypsinized, collected in PBS, centrifuged (193 g) and suspended in 1.00 ml of PBS. Cells were counted with three different methods three times each: (1) hemocytometer, (2) Moxi TM Z Mini Automated Cell Counter (ORFLO ), and (3) Z2 TM Coulter Counter (Beckman Coulter, Inc.). Exposure to 1,4-Benzoquinone MIA PaCa-2, C6, HepG2, MDA-MB231, and A549 cells were seeded into multiple 25 cm 2 or 75 cm 2 culture flasks at equal density and allowed to grow until 70 % confluent. One of the flasks was used strictly for calculating the initial dose in units of mol cell -1. To achieve this, prior to exposure to 1,4-BQ, cells were counted in this flask with a hemocytometer; this number of total cells, which were present immediately prior to exposure, was used to calculate the initial (applied) dose in units of mol cell -1. Exposure media were prepared by addition of 1,4-BQ stock solution in DMSO to fresh culture media and vigorous mixing. All cell lines were cultured in and exposed in DMEM high glucose with 10 % FBS and Penicillin Streptomycin (80 units ml - ). Growth medium was exchanged with the exposure medium containing 1,4-BQ/DMSO or DMSO alone 32

50 (vehicle control). Exposures to 1,4-BQ ranged from 0 to 2000 femtomol cell -1 (femto = ; abbreviation = fmol cell -1 ), i.e. 0 to 320 µm under the physical conditions of these experiments. Cells were then incubated for 4 h at 37 C, 5 % CO 2. For control experiments, 2 % DMSO (0.28 M) was added to fresh culture medium so that the percentage DMSO was equivalent to that of cells exposed to 1,4-BQ. For a single experimental protocol where media volumes were varied within the experiment, the percentage of DMSO in fresh culture medium varied from % (0.035 to 0.51 M); control experiments were done using the highest and lowest concentrations of DMSO; no toxicity was observed. Exposure to Oligomycin A MIA PaCa-2 cells were seeded at varying cell densities (25,000 cells - 400,000 cells) in duplicate 6-well culture plates and allowed to adhere and grow for 48 h before exposure to oligomycin A. Prior to exposure to oligomycin A, the number of cells per well of the duplicate 6-well plate was determined with a hemocytometer to ascertain the number of cells in each well for the varying cell densities; this number of total cells, present immediately prior to exposure, was used to calculate the initial dose in units of mol cell -1. Exposure media were prepared by addition of oligomycin A stock solution in DMSO to fresh culture media. Growth medium was exchanged with the exposure medium containing oligomycin A (2 µm) in DMSO or DMSO alone (vehicle control). All controls and treatments had 0.3 % DMSO (0.039 M) in the medium. Exposures to oligomycin A ranged from 0 to 87 fmol cell -1 (0 µm in control treatments and 2 µm for 33

51 oligomycin treatments), depending on the cell density. Cells were exposed to oligomycin A for 1 h prior to measurement of intracellular ATP. Exposure to Pharmacological Ascorbate MIA PaCa-2, AsPC-1, PANC-1, 339, and 403 cells were seeded into multiple 60 mm 2 culture dishes at 250,000 cells per dish and were cultured for 48 h at 37 C, 5 % CO 2. One dish was used strictly for calculating the initial dose in units of mol cell -1. To achieve this, prior to exposure to ascorbate, cells were counted in this dish with a hemocytometer; this number of total cells, which were present immediately prior to exposure, was used to calculate the initial dose in units of mol cell -1. Growth medium was exchanged with DMEM high glucose medium with 10 % FBS and Penicillin Streptomycin (80 units ml -1 ) for all exposures to ascorbate. It is important that the medium formulation is consistent across ascorbate exposure, as subtle changes in the exposure medium can result in different rates of ascorbate oxidation and therefore differences in the flux of H 2 O 2 the cells are exposed to. For these studies, all cells were exposed in DMEM high glucose medium with 10 % FBS and Penicillin Streptomycin (80 units ml -1 ). After the replacement of growth medium with fresh DMEM high glucose complete medium (3.0 ml), ascorbate was added to medium to achieve exposures of picomoles cell -1 (pico = ; abbreviation = pmol cell -1 ), i.e. 0-8 mm. For control experiments, medium was replaced with fresh DMEM high glucose medium, but cells were untreated. Cells were then incubated for 1 h at 37 C, 5 % CO 2. 34

52 Intracellular ATP Assay A cell suspension (100 µl, 50,000 cells) was added to each well in an opaquewalled, 96-well plate. To this, 100 µl of reagent from an ATP kit (Promega, CellTiterGlo) was added to lyse the cells and initiate the luminescence reaction. After 10 min, luminescence was measured on a microplate reader. ATP standard curves with concentrations between µm were generated for each experiment. The ATP concentration was determined from the corresponding standard curve and converted to an intracellular concentration using the cell number and cell volume, as done for intracellular GSH concentrations [73, 74]. Clonogenic Survival Assay To assess the cytotoxicity of exposure to 1,4-BQ and P-AscH -, cells were plated for a clonogenic assay following the 4-h exposure to 1,4-BQ and 1-h exposure to P- AscH -. The exposure medium was removed, cells trypsinized and counted with a hemocytometer and plated at a cell density of 200 and 400 cells in 3.0 ml of medium in 60 mm 2 dishes, with the exception of HepG2 cells, which were plated at 2000 and 5000 cells per dish. Plates were incubated for 6-14 days at 37 C, 5 % CO 2. After the growth period, cells were fixed with 70 % ethanol and stained with Coomassie Blue. Colonies were counted as a grouping of 50 or more cells. The plating efficiency and surviving fraction were determined [75]; plating efficiency (PE) = (colonies counted / cells plated) x 100; survival fraction = (PE of treated sample / PE of control) x 100. From plots of clonogenic survival fraction vs. dose of 1,4-BQ, the Effective Dose 50 (ED 50, 50 % clonogenic survival) was determined. 35

53 Trypan Blue Staining As a measure of cell viability the trypan blue exclusion assay was employed [76]. Trypan blue (10 µl of 0.4 %) was added to 10 µl of a cell suspension from each exposure. Using 10 µl of this cell suspension, a hemocytometer was used to count stained and unstained cells; cell viability = (unstained cells / total cells) x 100. Exposure of Cells for GSH and GSSG Determination To assess the effect of exposure to 1,4-BQ on intracellular levels of GSH and GSSG, MIA PaCa-2 cells were cultured to 70 % confluence in 125 cm 2 flasks. Prior to exposure, the growth media were exchanged with exposure media as described above. Cells were exposed to 1,4-BQ at increasing mol cell -1 (0 to 500 fmol cell -1 ) for a period of 30 min at 37 ºC, 5 % CO 2. Also, additional experiments with a bolus mol cell -1 of 1,4-BQ (6.1 fmol cell -1 ) for up to 24 h were performed with cells harvested at 0, 10, 60, 400, and 1400 min after exposure. All cells in a flask, with the exception of a minute quantity for cell counting, were used to allow for the measurement of GSSG. GSH and GSSG Determination with HPLC-BDD Cells were removed from culture dishes with trypsin/edta. Samples were centrifuged and supernatant removed. Cell pellets were resuspended in 5 % perchloric acid (PCA) containing 100 µm diethylenetriaminepentaacetic acid (DTPA, alias DETAPAC). The processed samples were stored at -80 C until the day of analysis. Prior to analysis samples were thawed and centrifuged to remove the precipitated proteins. 36

54 To determine the amount of GSH and GSSG in cultured cells, HPLC with electrochemical detection (ESA CoulArray with a temperature-controlled (4 C) autosampler) was used following the protocol outlined by Park et al. [77]. The method is based on an electrochemical detection (ECD) system using a boron-doped diamond disc (BDD) electrode (Model 5040, ESA Biosciences, Chelmsford, MA, USA). Samples were loaded into auto-sampler vials with a 100 µl glass insert. For analysis a sample was loaded on the column and eluted isocratically with 97 % 25 mm sodium phosphate, ph 2.65, 1.4 mm 1-octane-sulfonic acid / 3 % acetonitrile for 60 min. With each set of samples, four standards containing GSH ( pmol on column) and GSSG ( pmol on column) were included; these standard curves were used to quantify analytes for that particular sample set. Quantitation was performed by integrating the GSH and GSSG peaks in the BDD electrode channel with ESA Coularray for Windows version The GSH amount (mol) was divided by the number of cells associated with the sample (counted prior to resuspension in perchloric acid/dtpa using a hemocytometer) giving a value of mol cell -1. The mol cell -1 value was then divided by the intracellular volume resulting in an average intracellular concentration of mol L -1 for GSH and GSSG [74]. H 2 O 2 Removal Assay: Determination of Rate Constant by which Cancer Cells Remove H 2 O 2 The rate constant (k cell ) of extracellular H 2 O 2 removal was determined in each of the different pancreatic cancer cell lines using the 96-well plate reader assay developed 37

55 and described by Wagner et al. [78]. Cells were seeded in rows E-G of a 96-well plate at a density of 15,000 cells per well. Cells are then incubated for 48 h prior to the assay at 37 C, 5 % CO 2. Briefly, extracellular H 2 O 2 (10 µm) is added to wells of a 96-well plate containing a known number of cells at different time points. The cells will remove this extracellular H 2 O 2 over time. The system is then quenched at a predetermined time and the concentration of extracellular H 2 O 2 remaining in the wells is determined. The quenching solution contains horseradish peroxidase (HRP) that reacts with the remaining H 2 O 2 in the wells. The activated HRP then oxidizes para-hydroxyphenylacetic acid (phpa) resulting in the formation of the fluorescent phpa dimer, providing the readout of the amount of the amount of H 2 O 2 remaining in the wells. With this method an observed rate constant for the removal of extracellular H 2 O 2 is determined on a per cell basis. The capacity of cells to remove extracellular H 2 O 2 can be determined by taking into account the information obtained: number of cells in the well, amount of extracellular H 2 O 2 remaining, total volume of medium, and the different times of exposure of cells to extracellular H 2 O 2. Measurement of Catalase Activity Catalase activity was measured in MIA PaCa-2, AsPC-1, PANC-1, 403, 339, HepG2, A549, and MB231 cell lysates using a spectrophotometric-based assay [79]. Briefly, cells were harvested at a density of x 10 6 cells in 200 µl PBS. Cells were counted with the hemacytometer, so a known number of cells were used in the assay. After completing cell lysis through sonication, the cell lysate is diluted in 50 mm phosphate buffer (ph 7.0) and 30 mm H 2 O 2 is added to the cell lysate in the cuvette to 38

56 yield a final concentration of 10 mm H 2 O 2 in the cuvette. The decomposition of H 2 O 2 was followed by the decrease in absorbance at 240 nm over time. Absorbance was measured every 10 s for 2 min. Active catalase molecules per cell were calculated from k, which is the slope of the ln(absorbance due to H 2 O 2 ) vs. time (seconds). We assume that catalase is fully released and dispersed into the suspension of the lysate. This number and the experimental k from the assay were used to determine the number of catalase monomers per cell. The rate constant k = 1.7 x 10 7 M -1 s -1 for the catalytic rate constant is the rate constant per monomer [79]. Catalase is a tetramer, thus the number of tetramers per cell will be ¼ of the monomer count. Inhibition of Catalase with 3-Amino-1,2,4-Triazole Catalase was inhibited using 3-amino-1,2,4-triazole. Cells were treated with 20 mm 3-AT for 1 h at 37 C, 5 % CO 2. After the 1 h incubation, cells are washed three times with PBS to remove extracellular 3-AT prior to being used for experiments described herein. Transduction with Adenovirus Catalase MIA PaCa-2 cells were plated 48 h prior to transduction. Complete DMEM medium was removed and cells were washed 2 times with serum-free DMEM medium. Cells were then transduced with adenovirus catalase (1 x pfu) for 24 hours at desired MOI (i.e. 1, 5, 10, 25, 50, and 100 for experiments herein) in serum-free DMEM medium. After 24 h, adenovirus catalase was removed and cells were washed with 39

57 complete DMEM medium prior to replacement with complete DMEM medium for 24 h incubation prior to being used for the experiments described herein. Measurement of Ascorbate Oxidation in Cell Culture Medium with Clark Electrode Oxygen Monitor The rate of oxygen consumption (OCR, d[o 2 ]/dt) upon addition of ascorbate to DMEM cell culture medium complete with 10 % FBS and Penicillin Streptomycin (80 units ml -1 ) was determined using a Clark electrode oxygen monitor (YSI Inc.) that is connected to an ESA Biostat microelectrode system (ESA Products, Dionex Corp.). The OCR represents the rate of H 2 O 2 production. The accumulation of H 2 O 2 is determined with this system through the addition of catalase (bovine liver, Sigma C-1345). Western Blot for Catalase Immunoreactive protein corresponding to catalase was identified and quantified from total cell protein in the different pancreatic cancer cell lines (MIA PaCa-2, AsPC-1, PANC-1, 403, and 339) by the specific reaction of the immobilized protein with its antibody, as described in [57]. Cellular protein was harvested from 100 mm 2 dishes when cells were 70 % confluent. Cells were washed three times with PBS (ph 7.0), scraped from the dishes and collected in 100 µl cell lysis buffer (with protease inhibitors). The suspension was then centrifuged for 5 min at 500 g at 4 C to pellet. The protein concentration in the supernatant was determined using the BCA protein assay kit according to manufacturer s instructions. 40

58 Protein (30 µg) was electrophoresed in a 4 % to 20 % Bio-Rad pre-cast gel. The proteins were then electrotransferred to Immobilon Transfer Membranes (Millipore). After blocking in 5 % nonfat milk for 1 h, the membranes were treated with anti-catalase antibody (1 in 1,000 dilution, Santa Cruz Biotechnology). Horseradish peroxidaseconjugated goat anti-rabbit (1 in 25,000 dilution) secondary antibody from Chemicon International. The washed blots were then treated with SuperSignal West Pico Chemiluminescent Substrate (Thermo Scientific) and exposed to Classic Blue Autoradiography Film (MIDSCI). Statistics Results are expressed as the mean ± SEM. Statistical analyses were performed using One-way analysis of variance and where appropriate the unpaired student s t test; p values less than 0.05 were considered statistically significant. Calculations were performed using IBM SPSS Statistics for Windows, Version 20.0 (IBM Corp., Armonk, NY). Statistical analysis was done using GraphPad Prism 6.04 software (GraphPad Software, San Diego, CA). Statistical significance was determined using two-tailed unpaired t-test with Tukey post-test; p < 0.05 was considered to be statistically significant. All experiments had at least n = 3 biological replicates (indicated in each figure caption) and error bars indicate standard error of the mean. 41

59 CHAPTER 3: MOLES OF A SUBSTANCE PER CELL IS A HIGHLY INFORMATIVE DOSING METRIC Introduction In the testing of xenobiotics, medicines, and natural products for biochemical and biological responses, the use of laboratory animals is regarded as the best model for providing information to predict effects in humans. The U.S. National Institutes of Health (NIH), as well as other research institutions worldwide, are seeking to minimize the use of animals in this 21 st century by encouraging the development, validation, and implementation of non-animal based studies (NIH Revitalization Act of 1993 SEC.404C ). To succeed, it is important to gain the maximum information possible from in vitro experiments with the goal to accurately predict biological effects in humans. A critical element in the foundation of scientific research is reproducibility. This problem encompasses a wide array of issues ranging from statistical considerations, to laboratory standards, practices, and reporting ** [80, and references therein]. Here we examine the topic of how to specify dose or exposure to a xenobiotic in cell culture experiments with the goal to address an aspect of the problem of reproducibility in science. This matter may also result in more successful translation of information from cell culture studies to whole organisms, thereby addressing the 3R s, replacement, reduction and refinement, for the use of animals in research [81]. Doskey CM *, van t Erve TJ *, Wagner BA, Buettner GR. (2015) Moles of a substance per cell is a highly informative dosing metric in cell culture. PLoS One. 10(7): e PMID: PMCID:PMC ( * These authors contributed equally to this work) NIH Revitalization Act of as accessed ** Principles and Guidelines for Reporting Preclinical Research at as accessed

60 When assessing the biological consequences of xenobiotics in in vitro experiments, dose is a central parameter [82, 83]. Groothuis et al. have reviewed some of the major issues with dose and reproducibility of cell culture experiments and the translation of in vitro observations to in vivo models [4]. This instructive review examines various dose-metrics, including nominal concentration, total concentration, freely available concentration, as well as various dose-metrics for xenobiotics associated with cells. The most common dosing metric in cell culture experiments is the initial concentration, i.e. nominal concentration (e.g. mol L -1, g L -1 ; see [4].), of a compound added to the culture medium [84, 85, 86]. Using the nominal concentration of a xenobiotic as a measure of exposure can be unexpectedly problematic by yielding ambiguous information on the true exposure of cells to xenobiotics in cell culture experiments and provide limited mechanistic insights [87, 88, 89]. Exposure is highly dependent on the actual experimental conditions, e.g. volume of the medium used and total moles or mass of xenobiotic. This can lead to large variations in experimental results from unrecognized differences in the actual exposure due to changes in the physical conditions (e.g. volume of medium and number of cells used) of experiments. This is especially important with the introduction of high-throughput screening techniques [90, 91]. In these techniques, low volumes of media coupled with low cell numbers in multiwell plates result in many changes in physical parameters compared to traditional cell culture vessels, e.g. cell culture dishes and flasks. The objective of this research is to re-evaluate how dose is considered and reported with the ultimate goal to reduce some of the ambiguity introduced by common dosing metrics. Here we examine the value of a complementary but different dosing- 43

61 metric, i.e. moles of xenobiotic applied normalized to the number of cells present at the time of exposure (i.e. counted cells). We hypothesize that for many xenobiotics, using moles per cell (mol cell -1 ) as a dosing metric will reduce the dependency of experimental results on physical parameters. Although this dosing metric does not directly indicate how much compound is in, or associated with, the cell, it provides the advantage of being readily implementable with every assay by counting the number of cells prior to exposure in a representative cell culture vessel (e.g. cell culture well, dish, or flask) and taking into account the volume of medium over the cells. Since the number of cells is the denominator of this metric, it is critical to have an accurate count of the cells treated or measured in the assay. To achieve this there are several methods currently used in research laboratories to count cells (e.g. hemocytometer, automated cell counters and mini-automated cell counters). Consistency in cell counts was achieved within each of the three counting methods and between the three methods (Figure 3.1). The hemocytometer microscope-based method gave consistently higher counts than the automated cell counters (Figure 3.1). This is possibly due to being able to distinguish and count individual cells that may cluster that would be indistinguishable by the automated cell counters and subject to elimination due to gating. Dose as mol cell -1 can also provide additional information not easily obtained by other dosing metrics. Many xenobiotics bring about irreversible changes in critical covalent bonds (targets) in cells; some form very stable complexes with cellular targets. Typically, a certain fraction of these targets must be transformed to bring about biochemical and biological consequences. For these agents, the actual information needed to unravel detailed mechanisms is: how many targets are there in a cell and how many 44

62 molecules (moles) of xenobiotic per cell are required to transform a critical number of targets that will result in a biochemical and biological response? We hypothesize that using mol cell -1 as the dosing metric in cell culture experiments will give researchers this type of information, thereby expanding the information that can be extracted from the data obtained from cell culture experiments. To assess these hypotheses, the effects of the electrophile 1,4-benzoquinone (1,4- BQ) and oligomycin A, which forms a highly stable complex with mitochondrial complex V, were investigated in a variety of cell lines. Using these compounds, mol cell -1 was investigated as a metric for: quantitative causality; the consequences of sequential vs. bolus additions; and the influence of intracellular volume on biological responses. We anticipate that by specifying dose as mol cell -1, reproducibility of results from in vitro experiments across different laboratories will be improved; additional information will be available from cell culture data; and translation of this information to whole organisms will be more successful. Results To test our hypotheses that expressing dose as mol cell -1 will: yield more information; lead to improved experimental design; and better predict biochemical and biological responses, 1,4-benzoquinone (1,4-BQ) and oligomycin A were used as model xenobiotics. 1,4-Benzoquinone is an electrophilic quinone that readily forms covalent bonds with amine- and thiol-containing biomolecules (Figure 1.3) while oligomycin A inhibits mitochondrial ATP production by forming a quite stable complex with 45

63 mitochondrial complex V [92]. Because these two xenobiotics have quite different chemistry they can serve as representative candidates to address our hypotheses. Comparison of Dosing Metrics, 1,4-Benzoquinone as an Example To evaluate the differences between expressing exposure as mol cell -1 compared to the initial concentration, A549 cells were exposed to 1,4-BQ for 4 h under two physically different experimental conditions, Figure 3.2. Toxicity was assessed using clonogenic assays (reproductive cell death). The physical conditions that were varied between the two experimental set-ups were the volume of medium and number of cells. Between the two experimental set-ups there was an 8-fold difference in the number of cells used, but only a 0.75-fold difference in the volume of medium used, yielding two very different exposures to 1,4-BQ. When exposure is expressed as the initial, i.e. nominal concentration of 1,4-BQ in the medium, the percent survival vs. dose for the two conditions were statistically very different; EC 50 (Effective Concentration yielding 50 % survival) values are 14 and 62 µm, Figure 3.2A. However, when dose is expressed as fmol cell -1, the values of ED 50 (Effective Dose 50 % survival) for the two experimental conditions are essentially the same, 160 and 170 fmol cell -1, Figure 3.2B. The ambiguity present in expressing dose in terms of initial concentration compared to mol cell -1 was further explored in the experiments of Figure 3.3A-3.3D. In the experiments of Figure 3.3A-B the initial concentration of 1,4-BQ was constant for all exposures (16.7 µm); however, mol cell -1 was varied for each experiment by exposing an identical number of MIA PaCa-2 cells in varying volumes of medium (5.0 ml to 30.0 ml or 10.0 ml to 80.0 ml). Clonogenic survival varied substantially as the volume of 46

64 medium varied, even though the initial concentration in medium was identical throughout all exposures, Figure 3.3A. In contrast, there is clear dose-dependence when dose is specified in mol cell -1, Figure 3.3B. In this experiment, mol cell -1 provides an unambiguous dosing metric that more accurately reflects the exposure as opposed to expressing dose as the initial added concentration of 1,4-BQ. Next, the number of MB231 cells exposed was varied, but in all treatments identical volumes of medium (10.0 ml medium) and identical concentrations of 1,4-BQ (16.7 µm) were used. Again, clonogenic survival was different in each experiment, even though the dose in terms of initial concentration in the medium was identical in all exposures, Figure 3.3C. In contrast, expressing dose in mol cell -1 rather than the initial concentration throughout these experiments better reflects the cellular exposure and clearly shows an anticipated dose-dependence, Figure 3.3D. ATP content per cell was used as an alternative endpoint after exposure of MIA PaCa-2 cells to 1,4-BQ, Figure 3.3E-F. As with the biological endpoint of clonogenic survival, ATP content varied substantially as the number of cells exposed varies, even though the initial concentration of 1,4-BQ in medium was identical throughout all exposures, Figure 3.3E. However, there is the anticipated dose-dependence when dose is specified in mol cell -1, Figure 3.3F. In these experiments, mol cell -1 of 1,4-BQ provides an unambiguous dosing metric that better reflects the exposure to a xenobiotic and relates to both biochemical and biological consequences as opposed to expressing dose as the initial added concentration of 1,4-BQ. It has been demonstrated that decreased cellular ATP can correlate with decreases in cell viability and increases in pro-apoptotic markers, such as caspases [93]. In Figure 47

65 3.3, we see that changes in ATP content per cell followed closely with results of the clonogenic assays, i.e. lower cellular ATP after exposure to 1,4-BQ correlates with poorer survival. A plot of clonogenic survival vs. intracellular ATP concentration produces a remarkable linear relationship between a biochemical parameter and a biological consequence, Figure 3.4. This is as might be expected because each measure is a function of the exposure to 1,4-BQ as expressed in units of mol cell -1. That each measure is a quantitative function of applied dose as mol cell -1 opens a new window to examine mechanisms of toxicity. Comparison of Dosing Metrics, Oligomycin A Oligomycin A is an antibiotic isolated from Streptomyces diastatochromogenes; it is an inhibitor of mitochondrial H + - ATP synthase, complex V of the electron transfer system. Oligomycin A is widely used as a biochemical tool for studying mitochondrial respiration and oxidative phosphorylation. It inhibits the H + -ATP synthase by binding to the oligomycin sensitivity-conferring protein (OSCP); at high concentrations oligomycin can also inhibit the Na + +K + -ATPase [92]. Unlike 1,4-BQ, oligomycin A does not covalently bind to its target, rather it forms a tight complex with OSCP and/or the Na + +K + -ATPase. It is a non-competitive inhibitor in its interaction with OSCP with a V max = 0.77 µmol min -1 mg -1, K m = 120 µm and K i = 11 nm [94]. Oligomycin A has a dissociation constant K d 10-6 M with the Na + -K + ATPase [92, 94, 95]. Because oligomycin A forms a tight complex, we hypothesized that mol cell -1 would be a better metric for expressing dose than initial concentration. In Figure 3.5A and B the number of MIA PaCa-2 cells exposed to oligomycin A was varied (0.09 x 10 6 to 1.1 x 10 6 cells), 48

66 but in all treatments cells were exposed to identical concentrations of oligomycin A (2 µm in 3.0 ml of medium). This achieves different amounts (moles) of oligomycin A on a per cell basis (0-87 fmol cell -1 ). ATP content was measured following exposure; the resulting steady-state intracellular ATP concentration was different after each treatment (i.e. different numbers of cells were exposed), even though the dose in terms of initial concentration in the medium was identical for all exposures, Figure 3.5A and B. The expected dose-response was achieved when dose is expressed in terms of mol cell -1. These data demonstrate that the effects of xenobiotics that are popular biochemical tools could provide different results under various experimental conditions. This potential ambiguity can be overcome by using mol cell -1 as a standardized and reliable dosing metric when specifying how these tools were used. Sequential Addition vs. Bolus Addition Another challenge in dosing is the comparison between bolus vs. sequential addition of the xenobiotic. To examine the usefulness of mol cell -1 in this setting, several smaller doses of 1,4-BQ were added sequentially throughout a 4-h exposure time and compared to a single bolus addition producing the same total amount of 1,4-BQ. MIA PaCa-2 cells were exposed to sequential additions or a single bolus addition of 1,4-BQ for a total cumulative dose of 600 fmol cell -1, then clonogenic survival and trypan blue staining assays were performed, Figure 3.6. Identical toxicity was observed between a single bolus dose of 1,4-BQ (600 fmol cell -1 ) versus 12 sequential doses (50 fmol cell -1 per addition), Figure 3.6A. However, the trypan blue assay results showed statistically significant differences in cytotoxicity between the two dosing methods, Figure 3.6B. 49

67 Thus, reproductive cell death is affected by the total dose of 1,4-BQ, whereas membrane integrity has an element of time and concentration because several smaller exposures over time appear to be less damaging than a single bolus dose; ultimate lethality was the same as seen by clonogenic survival pointing to the importance of the need to use an appropriate assay to provide the best information. Causality, Census of Agent and Reaction Targets For exposure-science to move ahead in this 21 st century, it is imperative to create quantifiable causal relationships between agent and biological target. Here, the use of mol cell -1 as a dosing metric to aid in this process was investigated. 1,4-BQ is known to react with many nucleophilic moieties in proteins and small molecules, intracellularly as well as extracellularly. However, the main target(s) for toxicity has(have) yet to be identified. Glutathione was investigated as a potential target of 1,4-BQ, which can be directly related to the observed toxicity. A dose-dependent depletion of intracellular GSH in MIA PaCa-2 cells upon exposure was observed, Figure 3.7A. As expected, only a small accumulation of GSSG was observed over 24 h, Figure 3.7A. The dose (mol cell -1 ) of 1,4-BQ required to achieve 90 % depletion of GSH was about 100 times the total amount of GSH in a cell. There is no 1:1 relationship between the amount of added 1,4-BQ and GSH depletion; this indicates the prevalence of multiple reactions of 1,4-BQ with targets other than GSH, Figure 3.7B. The kinetics of the depletion of intracellular GSH was rapid and no recovery was observed in the 24 h following exposure, Figure 3.7B. The dose of 1,4-BQ per cell required to observe significant toxicity is many orders of magnitude greater than the amount of intracellular 50

68 GSH. A consideration not addressed here is the amount of 1,4-BQ lost to other compartments, e.g. to the cell culture medium, cell culture vessel, evaporative losses, and other avenues, i.e. the 1,4-BQ that is not associated with cells. These potential routes for loss of xenobiotics in cell culture is the central theme of [4]. The reactivity/association of 1,4-BQ with the non-cellular components of experiments would greatly diminish the amount actually associated with cells, consistent with our observation. However, mol cell -1 provides considerably more information than nominal concentration and is a better starting point to investigate and account for the lost 1,4-BQ. Intracellular Volume Affects the Apparent Toxicity of 1,4-BQ Because the presumed principal mode of toxicity for 1,4-BQ is to form stable covalent bonds with cellular components, it is hypothesized from target theory that larger cells with more intracellular components may be more resistant to exposures of 1,4-BQ than smaller cells. To test this hypothesis and the usefulness of mol cell -1 in establishing this relationship, the toxicity of 1,4-BQ to five different cell lines with different intracellular volumes (Table 3.1) was determined, Figure 3.8A-D. A strong linear correlation between ED 50 (clonogenic assay) of 1,4-BQ with measured intracellular volume was observed (R 2 = 0.74), Figure 3.8A. As expected, the mass of protein per cell directly correlates with the intracellular volume (R 2 = 0.76) [96], Figure 3.8B. A very strong linear correlation between ED 50 of 1,4-BQ with the measured mass of protein per cell of the different cell lines was observed (R 2 = 0.96), Figure 3.8C. This is consistent with there being more targets (i.e. proteins) in larger cells than smaller cells that must be modified by 1,4-BQ to produce reproductive cell death, Figure

69 A standard approach for normalization of a biochemical or biological consequence, here ED 50, is to normalize the results to protein mass as determined with a protein assay. In the experiments of Figure 3.8 when ED 50 is normalized to mass of protein (here pg of protein), all cells appear to be equally susceptible to 1,4-BQ, Figure 3.8D. Normalization to mass of protein would be considered a biochemical normalization, whereas normalization to per cell would be a biophysical normalization. As demonstrated here, these two different normalizations can provide very different but complementary information. Discussion Cell culture is a widely used research tool in the life sciences. It is used to study the basic biology of health and disease and is essential in characterizing the biochemical and biological effects of a wide range of xenobiotics, including studies on the efficacy and toxicity of new therapeutics for the treatment of disease. Here, we propose a simple, straightforward, easy to implement, and cost-effective strategy to reconsider how dose and exposure are used and reported in typical cell culture experiments. This approach can greatly expand the quantity and quality of the information obtained in many settings. In this work, we provide examples of how mol cell -1 is a dosing metric that is less affected by physical parameters, such as volume of medium and number of cells, because it incorporates more information about the physical set-up within the metric. In doing so, it yields more consistent dose-response information that is not easily obtained by using other dosing metrics. This information can help to understand both biological (e.g., clonogenic survival) and biochemical (e.g., ATP content) endpoints. We anticipate that 52

70 this approach may also be useful for experiments with non-mammalian models of toxicity e.g. yeast, bacteria, and perhaps even C. elegans, zebra fish, and drosophila. However, if the ratio of extracellular (or extra-organism) volume containing the xenobiotic to intracellular (or intra-organism) volume is extremely large, then mol cell -1 or mol organism -1 will be inappropriate as the amount of substance per cell/organism could be far too great for the organism to make any significant change in the steady-state level of available xenobiotic, or availability is diffusion-limited. Then different aspects of exposure would need to be considered and be more appropriate. The concept of mol cell -1 extends not only to toxicant or drug, but also to agents used as biochemical tools to determine the fundamental mechanisms of biochemical and biological pathways. Oligomycin A is one such agent used to study mitochondrial function. As an example, oligomycin A is part of the mitochondrial stress test kit that Seahorse Bioscience provides for use with their XF analyzers to profile mitochondrial function. Here we show that the same initial concentration of oligomycin A has a different effect on the intracellular ATP concentration when varying numbers of cells are present. However, when applied dose is specified as mol cell -1 this caveat does not apply. Different Experimental Platforms can Result in a Wide Range of Exposures We demonstrate that mol cell -1 can reconcile what appear to be very different results from what would seem to be the same experiment, but performed under different physical configurations. Nowhere is this more important than in the state-of-the-art highthroughput screening of new drugs and toxicants. There is a wide range of experimental platforms (traditional culture vessels with robotic operation to low volume, low cell 53

71 number multi well plates) available for these types of studies. If not controlled for, these distinct platforms can provide different experimental results under seemingly similar conditions, Figure Note that for platforms commonly used in wet bioscience laboratories, dose per cell varies over a range of 60-fold under typical exposure conditions and an identical initial concentration. If platforms used for high-throughputscreening are included, then dose per cell varies over a fold range. When designing experiments using different platforms, exposure as mol cell -1 will reduce costs, improve reproducibility, and allow direct comparison between experiments. The Case of H 2 O 2 and mol cell -1 or mol cell -1 s -1 as a Metric for Exposure The concept of toxicity as a function of cell density has been noted before; an example is hydrogen peroxide (H 2 O 2 ) [97, 98, 99, and references therein]. Hydrogen peroxide can bring about both reversible and irreversible changes in covalent bonds. As a cellular toxin, the mechanism of toxicity for H 2 O 2 involves oxidation of proteins, lipids, DNA, as well as the intracellular redox buffer [100, 101, 102]. In different settings these mechanisms have different degrees of importance towards the overall toxicity of H 2 O 2. Recently Gülden et al. described the toxicity of bolus additions of H 2 O 2 to cells in culture on the basis of per cell exposure, i.e. µmol H 2 O 2 per 10 7 cells [98]. Their work was derived from earlier work by Spitz et al. [97] who concluded that the primary descriptor of toxicity of H 2 O 2 is nmol H 2 O 2 /mg cell protein for a specific cell line. As expected, when different physical setups of experiments were investigated they observed that the initial extracellular concentration of H 2 O 2 (molarity) did not necessarily correlate with the observed toxicity. However, their data could be easily interpreted when dose was 54

72 specified as nmol H 2 O 2 /mg cell protein; for a given cell line, this is proportional to mol cell -1, Figure 3.8B and Table 3.1. This concept applies not only to direct additions of H 2 O 2 to cells in culture but also to sources that generate a flux of H 2 O 2, e.g. the glucose/glucose oxidase system [103, 104] or the oxidation of ascorbate in cell culture [56]. A central consideration in designing successful experiments is knowledge on the flux (mol cell -1 s -1 ) of H 2 O 2 in a set of experiments. In all such experiments it is important to determine (or verify) the flux of H 2 O 2 from sources of H 2 O 2 on a mol cell -1 s -1 basis. It is only with this approach that repeatability can be achieved and the best understanding of experimental data can be obtained. With this approach the transition from observational biology to quantitative biology is possible [78, 96, 105]. The realization by a very few researchers over 20 years ago that mol cell -1 (nmol H 2 O 2 /mg cell protein) of H 2 O 2 is more accurate for specifying dose for exposure to H 2 O 2 remains largely unappreciated. Literally hundreds of scientific papers each year employ bolus addition of H 2 O 2 to the medium of cells in culture reporting only the initial concentration of H 2 O 2 in the medium. The vast majority of reports use initial molarity to specify dose, often without information on number of cells or volume of medium, i.e. cell density (cells L -1 ) limiting the information available in the data. Different physical setups of experiments will result in vastly different lifetimes of H 2 O 2 [78] and most probably different results. We demonstrate here that expression of dose as mol cell -1 has application beyond exposures to H 2 O 2. Dose of a xenobiotic as mol cell -1 can be easily implemented by making note of physical parameters used in the experiment, such as volume of medium 55

73 and number of cells. Utilizing this information in a dosing metric when setting up experiments allows for increased consistency in the larger body of information available in experimental data. Target Theory Target theory is a concept highly applicable when interpreting results from experiments using mol cell -1 as the dosing metric. As first described by Lea in 1946 [15], target theory assumes there to be a certain number of sensitive points or targets within cells. Damage to a certain percentage of those targets will cause the cell to undergo a response. The experiments with bolus vs. sequential additions of 1,4-BQ provide an excellent example of hits to a target accumulating to a critical percentage leading to biological effects, Figures 3.6 and 3.9. Here we assume that the damage to critical targets is not reversible. Our experimental results with the various cell lines varying cell volumes and thus a varying number of targets for 1,4-BQ, Figures 3.8 and 3.9 is consistent with different cell lines having a distinct number of targets. A certain fraction of these targets need to be hit to initiate an effect. This results in a different threshold for the amount of 1,4-BQ required amongst the cell lines before a specific toxicological endpoint is observed. The observation of a strong linear correlation of toxicity, ED 50, vs. cell size or protein mass cell -1 is consistent with the principles of target theory, assuming the number of targets susceptible to change by 1,4-BQ is proportional to cell volume or protein mass cell -1, Figure 3.8. This is in a way parallel to how administration of many drugs or exposure to toxicants is considered, i.e. the mass of the patient or area, e.g. per m 2. 56

74 The cell lines tested have quite different tissue origins (Table 3.1) and characteristics that contribute to the consequences that ensue upon exposure to a xenobiotic. However, the results of Figure 3.8 indicate that elements of target theory can be used to understand some aspects of these effects. This concept also underpins mol cell -1 as a unique approach to compare the availability of targets for a xenobiotic in various cell lines. Bolus vs. Sequential Additions Because of the rapid reaction of 1,4-BQ with thiols and amines, a bolus addition could have different apparent cellular toxicities, depending on the particular assay employed. We found that giving a single bolus dose of 1,4-BQ or 12 sequential smaller doses, which yielded the same total dose, produced the same toxicity, as seen by clonogenic survival, Figure 3.6. Parallel observations using activation of signaling pathways have been made with reactive lipid-derived electrophiles attesting the generality of this mechanism [106]. However, changes in membrane-integrity upon exposure to 1,4-BQ, as seen by the trypan blue assay, showed significant differences in apparent toxicity between the two methods of exposure. These observations indicate to the importance of time in addition to dose in understanding mechanisms of toxicity. The time frame for the 12 sequential additions, 4 h, was much less than the doubling time for MIA PaCa-2 cells, 24 h. From this experiment the short time and lower amount of 1,4-BQ allowed time for repair of the membrane damage as seen by trypan blue exclusion. However, the damage as seen by the clonogenic assay was apparently cumulative. One can speculate that if the time frame 57

75 for the sequential addition was on the order of, or greater than, the cell doubling time, then less toxicity would have been observed. The choice of assay to assess toxicity depends on the goals of the experiment and the information sought. Here we see that clonogenic survival is dependent on total dose, but membrane integrity is not associated with total dose under these experimental conditions. These apparent differences in toxicity are revealed by using mol cell -1 as the dosing metric providing the best information to understand the mechanism of lethality. Limitations In this work mol cell -1 is the nominal dose of a xenobiotic applied in the medium normalized to the number of cells present at time of exposure. This is not intended to imply that it is the dose associated with the cell, i.e. the cell burden or internal concentration. Mol cell -1 may not provide an advantage in all experimental settings. The biochemical and biological effects of xenobiotics that bind reversibly to cellular targets, such as hydrogen bonding, are governed by the equilibrium constant for the binding and, of course, mass action. For xenobiotics that act via equilibrium reactions, especially those with relatively large dissociation constants, the traditional dosing metric of initial concentration in the medium may be a useful dose metric. However for tight complexes, i.e. small dissociation constants, mol cell -1 may be a very informative metric. Here we propose that mol cell -1 can be especially valuable with agents that bring about irreversible changes in cells. As examples to test our hypothesis, we examined the classic electrophile 1,4-BQ, which makes irreversible covalent bonds with cellular components, and oligomycin A, which forms a tight complex with mitochondrial H

76 ATP synthase [92, 94, 95]. 1,4-Benzoquinone as an electrophilic species reacts readily with thiol- and amine-containing species forming covalent bonds, Figure 1.3 [9, 107, 108]; biological effects can ensue [109, 110, 111]. Mol per cell is a better approach than nominal concentration (i.e. initial concentration) to specify dose for these two quite different xenobiotics. We foresee that this approach to specify dose will be valuable all across cell culture for a wide range xenobiotics, be they toxicants, drugs, or standard biochemical tools. However, neither nominal concentration nor mol cell -1 addresses the amount lost due to chemical reactions with or binding to medium components or even the cell culture vessel itself as well as other possible routes, as addressed thoroughly by Groothuis et al. [4]. The only additional laboratory effort to arrive at mol cell -1 is to count the number of cells in an experiment. This requires little additional effort and resources. However, if the cell number changes significantly during an experiment, then additional considerations will be needed as the amount of xenobiotic per cell may not be easily estimated. Recommendations The vast majority of biochemical assays require some sort of normalization. For example enzyme assays are most often reported in some appropriate units that are normalized to amount of protein; examples are the assays for glutathione peroxidase activity [112], catalase activity [79], and superoxide dismutase activity [113]. These types of results from cell culture experiments can also be normalized to the number of cells [78, 96]. Normalizing to protein content is considered to be a biochemical normalization, 59

77 whereas normalizing to number of cells is a biophysical normalization. These two approaches provide different, but complementary information. Nominal extracellular concentration is not a normalized parameter. However, specifying dose or exposure in mol cell -1 (i.e. (nominal moles (or mass) of xenobiotic)/(number of cells)) offers a scalable parameter that can be used to design experiments and help interpret a wide variety of experimental results. Clearly this parameter will be an upper limit to the number of moles of xenobiotic, or downstream products, associated with a cell; some fraction will not be in/on the cells, but rather may be associated with media or experimental apparatus or lost through other mechanisms [4]. We recommend that when reporting on experiments that employ cell culture, researchers should specify the dose of xenobiotics both in traditional initial concentration units (nominal concentration) as well as mol cell -1 or mass cell -1 (a normalized nominal amount) to provide maximum information. Employing this dosing metric can be accomplished with little additional effort or resources. We also recommend cell number be determined prior to treatment in a representative flask (or well in a plate), as opposed to after treatment. This insures that all cells that are present at the time of exposure are accounted for and includes those cells that may undergo cell death during exposure, which would not be taken into account if cells are counted following the exposure. The method and time point of counting cells should always be clearly reported to reduce ambiguity and allow for replication of experiments. 60

78 Conclusions In this study, some of the advantages and limitations of mol cell -1 vs. initial molarity of the agent in the medium were examined. We conclude: 1. Dose expressed as mol cell -1 allows direct comparisons between different experimental conditions and can more accurately report the actual exposure compared to nominal (initial) concentration in the medium, Figures Dose/exposure as mol cell -1 can provide insightful information when exposures in cell culture are not from a single bolus addition, Figure The biological effects of xenobiotics that make irreversible covalent bonds depend on cell size and protein content, which is in line with concepts of target theory, Figures Mol cell -1 provides a scalable metric of dose that allows for successful transition between different cell culture platforms; i.e. all the information needed is included in this metric to scale experiments to different platforms that require differing number of cells and volumes of medium, Figure Data presented here make a compelling case to include mol cell -1 along with traditional nominal concentration units when specifying dose in cell culture studies. Future Directions It will be important to expand these studies to explore more compounds for which it may be appropriate to express dose as mol cell -1. It is likely that there will be several compounds in which mol cell -1 may not provide a distinct advantage. We predict that this 61

79 will likely be true for compounds that have low binding affinities for their target molecules or form readily reversible bonds with their target compounds. While it will be important to identify these compounds, until then we strongly suggest using both dosing metrics (i.e. initial concentration and mol cell -1 ). Including initial concentration of compound in the medium will be helpful in that it is a traditionally used dosing metric recognized by all and can be referred to in considering whether the doses used are physiologically relevant to those that can be achieved in vivo. Moles per cell as a dosing metric provides the advantage of containing all the information needed to replicate the experiment and also scale it to a different experimental platform. It will prove vital in setting up repeatable experiments with ease and a useful tool within and across different laboratories. Reporting both these dosing metrics can therefore provide complementary, but different information. Thus far, these studies have only been carried out in 2D cell culture systems. With the introduction of 3D cell culture systems and engineered tissue culture systems designed to better mimic in vivo environments, it will be important to test the utility of mol cell -1 as a dosing metric under such conditions. It will be interesting and informative to determine the influence of the physical parameters of these cell culture systems. The findings of toxicity of 1,4-BQ to cells of different sizes being consistent with target theory will be another very interesting observation to explore further. Many differences exist between cells other than intracellular volume and protein content (i.e. metabolic and energy demand, doubling time, antioxidant enzyme content) that may affect the cells susceptibility to toxins. It would be informative to adjust the intracellular volume of a single cell line and test the susceptibility to 1,4-benzoquinone. Hypotonic 62

80 and hypertonic solutions can be utilized to achieve this. Synchronization of cells in different phases of the cell cycle can also be done to achieve the change in cell size, although there may also be changes in the antioxidant content of the cells at different phases of the cell cycle as well. In addition, discovering other xenobiotics that may be consistent with target theory could also provide insight into their mechanisms of toxicity and which cells may be most sensitive to that toxicant. 63

81 Table 3.1 Physical and biological parameters of cell lines used Cell Line a Type Doubling time (h) Literature Intracellular Volume (pl) b Measured Intracellular Volume (pl) Protein mass per cell (pg) c C6 (CCL-107) Rat glioma 22 [114] 1.08 [115] 1.04 (0.12) d,e 0.95 (0.07) f,e 141 (26) e MDA-MB231 (HTB-26) A549 (CCL-185) MIA PaCa-2 (CRL-1420) Human mammary adenocarcinoma Adenocarcinoma alveolar epithelial Human pancreatic carcinoma 34 g 1.53 [96] 2.37 (0.01) d,e 2.21 (0.14) f,e 415 (16) 22 h 1.76 [115] 2.38 (0.01) d,e 2.69 (0.03) f,e 659 (15) 24 g 2.03 [96] 2.61 (0.02) d,e 2.10 (0.02) f,e 757 (59) HepG2 (HB-8065) Human hepatoma cells 42 g 2.54 [116] 2.83 (0.09) d,e 2.96 (0.07) f,e 1192 (186) a Name (ATCC #) b Literature values for intracellular volume of cell lines. c Protein mass per cell was measured in trypsinized cells using the SDS-Lowry protein assay, n = 3 for each cell line. Albumin from bovine serum (Sigma Aldrich; Cohn Fraction V, Sigma-A2153) was used as a standard [72]. d As measured with Moxi TM Z Mini Automated Cell Counter (ORFLO Technologies), n = 3 for each cell line. Presented is the mean of triplicate biological samples. e Standard error of the mean of three triplicate biological samples. f As measured with Z2 TM Coulter Counter, n = 3 for each cell line. Presented is the mean of triplicate biological samples. g As measured in triplicate under the cell culture conditions reported here. h ATCC as accessed on

82 2.5E+06 Number of Cells in Flask 2.0E E E E E+00 Figure 3.1. Cells can be counted consistently with three different methods. MIA PaCa-2 cells were grown in a 25-cm 2 tissue culture flask in DMEM medium until they were 70 % confluent. The cells were then trypsinized and counted using a hemocytometer, Moxi Z Mini Automated Cell Counter, and Z2 TM Coulter Counter Cell and Particle Counter (Beckman-Coulter, Miami, FL). Within each method the total number of cells in the flask from the cell-counts were very consistent (C.V. = +/- 3 %, n = 3). The variation across the three different methods was larger (C.V. = +/- 9 %), but there were no significant differences between the three methods of cell counting. Error bars are standard deviation of the mean, n = 3. 65

83 A % Clonogenic Survival x 10 6 cells 15 ml medium 1.55 x 10 6 cells 20 ml medium A ,4-BQ Initial Concentration (µm) B % Clonogenic Survival A ,4-BQ Dose (fmol cell -1 ) Figure 3.2. Dose of 1,4-BQ expressed as mol cell -1 allows direct comparisons between different experimental conditions and more accurately reports toxicity than initial concentration in medium. Clonogenic survival of A549 cells was observed after a 4-h exposure to 1,4-BQ using two different experimental conditions: black square, 13 x 10 6 cells exposed in 15.0 ml of medium; gray diamond, 1.55 x 10 6 cells exposed in 20.0 ml medium. The doses are expressed in: (A) Initial concentration in medium (µm), note that EC 50 depends on the physical setup of the experiment; (B) Mol cell -1 basis (here fmol cell -1 ); note that ED 50 is independent of the physical setup of the experiment. The two clonogenic survival curves are representative experiments with each being the median of six plates. Error bars represent the standard error of the median; many error bars are smaller than the symbol. Experiments were performed in collaboration with Thomas Joost van t Erve, PhD. 66

84 A Clonogenic Survival (%) Control MIA PaCa-2 5 ml 10 ml 20 ml 40 ml 30 ml 60 ml 80 ml ,4-BQ Initial Concentration (µm) B Clonogenic Survival (%) ml 10 ml 10 ml 20 ml MIA PaCa-2 20 ml 30 ml 40 ml 60 ml 80 ml ,4-BQ Dose (fmol cell -1 ) C 120 MB231 D 120 MB231 Clonogenic Survival (%) x 10 6 cells 3.3 x 10 6 cells 1.7 x 10 6 cells 0.84 x 10 6 cells ,4-BQ Initial Concentration (µm) Clonogenic Survival (%) x 10 6 cells 3.3 x 10 6 cells 1.7 x 10 6 cells 0.84 x 10 6 cells ,4-BQ Dose (fmol cell -1 ) 67

85 E Intracellular [ATP] (mm) MIA PaCa x 10 6 cells 0.85 x 10 6 cells 0.69 x 10 6 cells 0.27 x 10 6 cells 0.07 x 10 6 cells ,4-BQ Initial Concentration (µm) F Intracellular [ATP] (mm) MIA PaCa x 10 6 cells 0.69 x 10 6 cells 0.27 x 10 6 cells 0.07 x 10 6 cells ,4-BQ Dose (fmol cell -1 ) Figure 3.3. Dose specified as mol cell -1 more accurately reports toxicity of 1,4-BQ in cell culture experiments than initial concentration in medium. (A) These data show two different physical setups for the experiments, black circle (1.7 x 10 6 cells) and gray diamond (3.3 x 10 6 cells). Clonogenic survival of MIA PaCa-2 cells was measured in triplicate after a 4-h exposure to 16.7 µm 1,4-BQ in different volumes of medium, yielding a range of doses on a mol per cell basis. No dose dependence is observed when the dose of 1,4-BQ is expressed as the initial concentration in the culture medium. Error bars represent the standard deviation of the median of triplicate measures. (B) The data of panel A are transformed to express the dose of 1,4-BQ in units of fmol cell -1. A far better delineation of the dose-dependent toxicity is observed in each of the two sets of experiments. Error bars represent the standard deviation of the median of triplicate measures. (C) Using a varying number MDA-MB231 cells, clonogenic survival varied considerably after a 4-h exposure to 16.7 µm 1,4-BQ in 10.0 ml of medium in T- 25 culture flasks (each point has an n = 2 for biological replicates, n =3 within each replicate). Error bars represent the standard deviation of the mean. (D) As with MIA PaCa-2 cells, when the dose of 1,4-BQ is expressed in units of fmol cell -1, a much more informative delineation of the dose-dependent toxicity in MDA-MB231 cells is observed. Error bars represent the standard deviation of the mean. (E) Using a varying number of MIA PaCa-2 cells, intracellular ATP concentration of MIA PaCa-2 cells varied considerably after a 4-h exposure to 8.35 µm 1,4-BQ in 5.0 ml of medium in T-25 culture flasks (each point has a n = 2 for biological replicates, n = 2 with in each replicate). Error bars represent the standard deviation of the mean; many standard deviations are smaller than symbol. (F) When the dose of 1,4-BQ is expressed in units of fmol cell -1, a much more informative delineation of the dose-dependent effect on intracellular ATP concentration in MIA PaCa-2 cells is observed. Error bars represent the standard deviation of the mean. Experiments for A and B were performed in collaboration with Thomas Joost van t Erve, PhD. 68

86 % Clonogenic Survival R² = 0.99 MIA PaCa ATP (mm) Figure 3.4. Clonogenic survival correlates directly with intracellular ATP concentration following exposure to 1,4-BQ. Clonogenic survival of MIA PaCa-2 cells following 4-h exposure to ( fmol cell -1 ) 1,4-BQ was plotted against intracellular ATP concentration of MIA PaCa-2 cells also following 4-h exposure to ( fmol cell -1 ) 1,4-BQ. Clonogenic survival directly correlates with intracellular ATP concentration following 4-h exposure to 1,4-BQ. Clonogenic survival is presented as the mean of n = 3 biological replicates with error bars representing the standard error of the mean. Intracellular ATP is presented as the mean of n = 2 biological replicates with error bars representing the standard error of the mean. Some error bars are smaller than the symbols. 69

87 A 7 MIA PaCa-2 Intracellular ATP (mm) x 10 6 cells 1.1 x 10 6 cells 0.69 x 10 6 cells 0.34 x 10 6 cells x 10 6 cells 0 B 7 Intracellular ATP (mm) Oligomycin initial concentration (µm) 1.1 x 10 6 cells 1.1 x 10 6 cells 0.69 x 10 6 cells 0.34 x 10 6 cells MIA PaCa x 10 6 cells Oligomycin Dose (fmol cell -1 ) Figure 3.5. Expressing dose as mol cell -1 yields more information and can be helpful when using biochemical tools in cell culture experiments: ATP per cell decreases with increasing dose of oligomycin A on a per cell basis. The levels of ATP in MIA PaCa-2 cells were measured immediately after a 1-h exposure to oligomycin A. (A) ATP levels were measured following a 1-h exposure of MIA PaCa- 2 cells at varying cell densities to 2 µm oligomycin A in 3.0 ml medium. Doses of oligomycin A are expressed in initial concentration of oligomycin A in the medium (µm) (n = 4, error bars are standard deviation of the mean). (B) Doses of oligomycin A are expressed in mol cell -1 (fmol cell -1 ) (n = 4, error bars are standard deviation of the mean). 70

88 A 120 Bolus Addition Sequential Addition % Clonogenic Survival Clonogenic survival ,4-BQ Dose (fmol cell -1 ) B 120 Cell Viability (%) Bolus Addition Sequential Addition Trypan Blue ,4-BQ Dose (fmol cell -1 ) Figure 3.6. A single bolus addition or sequential additions of 1,4-BQ can provide different toxicities based on the endpoint measured. (A) Clonogenic survival of MIA PaCa-2 cells was evaluated after a bolus addition of 600 fmol cell -1 of 1,4-BQ or incremental additions of 1,4-BQ every 20 min over the 4-h exposure period (12 separate but equal additions) to yield a total dose of 600 fmol cell -1. Controls represent additions of DMSO only to the culture media using protocols parallel to additions of 1,4-BQ. Clonogenic survival was the difference in the clonogenic survival between the same two protocols. Each is different from the controls (p < 0.05). (B) Cell viability as indicated with trypan blue staining produced quite different results using a bolus dose of 1,4-BQ vs. sequential addition (n = 3, error bars are standard deviation of the mean). A single bolus addition produces a significant difference between control and bolus addition (p <0.05), whereas the sequential addition is the same as the control (p > 0.05). 71

89 A Concentration(mM) B Concentration (mm) GSH MIA PaCa-2 = 1.3 mm = 0.12 fmol cell -1 GSH GSSG ,4-BQ dose (fmol cell -1 ) 1,4-BQ dose = 6.1 fmol cell -1 GSH GSSG Time (min) Figure 3.7. Glutathione is not depleted with 1:1 stoichiometry upon exposure of MIA PaCa-2 cells to 1,4-BQ. (A) Intracellular concentration of GSH and GSSG in MIA PaCa-2 cells after a 30-min exposure to different doses of 1,4-BQ. The basal level of GSH in MIA PaCa-2 cell is 0.12 fmol cell -1 or 1.3 mm, assuming a uniform intracellular distribution and an intracellular volume of 2.03 pl cell -1, Table 3.1. To deplete 90 % of the basal intracellular volume of 1,4-BQ (10 fmol cell -1 ) is required. Additional depletion up to 99 % requires ~ 400 fold excess of 1,4-BQ. This implies that 1,4-BQ most likely reacts with extracellular targets; the fraction that enters the cells reacts with the multitude of targets available in the intracellular space. GSSG is not significantly produced (statistically) in these experiments. (B) Intracellular concentration of GSH and GSSG up to 24 h after 72

90 exposure to a bolus of 1,4-BQ. Following exposure of MIA PaCa-2 cells to 1,4-BQ (6.1 fmol cell -1 ), an immediate 31 % depletion in GSH levels is observed. There is no recovery of GSH for at least 24 h after exposure. There is little generation of GSSG, indicating negligible generation of H 2 O 2. On a mol cell -1 basis only 1 out of 100 of the molecules of 1,4-BQ that were present at the start of the exposure reacted with GSH; This demonstrates that GSH is not the exclusive target for 1,4-BQ. Shown here are typical experimental results, n = 3. The coefficient of variation in the GSH and GSSG measurements is 13 %. Experiments were performed in collaboration with Thomas Joost van t Erve. 73

91 A ED 50 (fmol cell -1 ) R² = 0.74 MIA PaCa-2 C6 A549 MB Intracellular Volume (pl) B 1600 HepG2 R² = 0.76 Protein Mass per cell (pg) 1200 HepG2 800 MIA PaCa-2 A MB231 C Intracellular Volume (pl) C ED 50 (fmol cell -1 ) R² = 0.96 HepG2 MIA PaCa-2 MB231 A549 C Protein mass per cell (pg) D 0.5 ED 50 (fmol / pg protein) Figure 3.8. ED 50 of 1,4-BQ correlates directly with intracellular volume and mass of protein per cell for C6, MB231, A549, MIA PaCa-2, and HepG2 cell lines. (A) The dose of 1,4-BQ (mol cell -1 ) at which 50 % clonogenic survival was observed for each cell type is plotted vs, the measured intracellular volume (Table 3.1). The correlation coefficient R 2 is Each cell line has n = 2 for biological replicates, n = 3 within each replicate. The measured intracellular volume represented is the mean of the two different methods of measuring intracellular volume. Each cell line was measured (n = 3) using a Z2 Coulter Counter and a Moxi Z Mini Automated Cell Counter in ISOTON II Diluent (Beckman Coulter, Inc). Error bars represent that standard error of the mean; some error bars are smaller than the symbol. (B) Mass of protein per cell directly correlates with intracellular volume (R 2 = 0.76). Protein content was measured in the five cell lines used (C6, MDA-MB231, A549, MIA PaCa-2, and HepG2) by the SDS-Lowry protein assay. Some of the uncertainties in the protein mass per cell are smaller than symbols. Error bars represent the standard error of the mean. Each protein measurement has n = 3 for biological replicates, n = 3 within each replicate. (C) The ED 50 of 1,4-BQ for C6, MB231, A549, MIA PaCa-2, and HepG2 cells (mol cell -1 ) is plotted vs. measured protein 74

92 mass (pg) per cell. The correlation coefficient R 2 is Error bars represent the standard error of the mean. (D) The ED 50 of 1,4-benzoquinone is expressed as fmol of 1,4-benzoquinone per pg protein for a cell. Each protein measurement has n = 3 for biological replicates, n = 3 within each replicate. Error bars represent the propagation of error as determined from the standard error of the means for both protein measurements and ED 50 of 1,4-BQ. When dose of 1,4-BQ is expressed as fmol pg -1 protein, there was no statistical difference in the ED 50 of 1,4-BQ observed across the different cell lines. ANOVA showed p > 0.05 for all comparisons. Experiments were performed in collaboration with Thomas Joost van t Erve, PhD. 75

93 Figure 3.9. Depiction of target theory and exposure to 1,4-BQ. The ten quinone moieties shown in each scenario represents the same mol cell -1. Here we assume that there are a certain number of sensitive/ reactive targets within cells and the number of targets is proportional to intracellular volume. Damage to some fraction of those targets will produce a biological effect. Because larger cells have a greater number of targets, more 1,4-BQ will be required to produce the same biological effect as observed with smaller cells. 76

94 Figure Exposure to xenobiotics when specified as mol cell -1 varies greatly when using different experimental platforms for cell culture. Hypothetical cell culture experiments are shown where general recommendations, such as from Invitrogen (Life Technologies, Grand Island, NY) are followed for: seeding density, estimated number of cells at confluence, and media volume when using different cell culture vessels. In all hypothetical experiments, cells were exposed to an initial concentration of 1 M xenobiotic in the medium. However, this leads to a wide range of doses when converted to mol cell -1. The size of each data point (circle) is proportional to the dose in units of nmol cell -1 ; the larger the area of the circle, the higher the exposure. Note that for platforms commonly used in wet bioscience laboratories, dose per cell varies over a range 60-fold; if platforms used for high-throughput-screening are included (1536- and 3456-well plates) then dose per cell varies over a fold range, despite the initial molar concentration of xenobiotic being the same for each experimental platform. Figure made in collaboration with Thomas Joost van t Erve, PhD. 77

95 CHAPTER 4: DIFFERENCES IN THE CAPACITIES OF TUMOR AND NORMAL CELLS TO REMOVE H 2 O 2 Introduction There are several xenobiotics (i.e. environmental pollutants, chemotherapeutics) that are capable of generating oxidants at levels that can overcome the antioxidant defense of cells. This can lead to toxic effects in cells because oxidizing species that are not detoxified can cause oxidative damage to biomolecules (i.e. lipids, proteins, and DNA) leading to a multitude of downstream effects (i.e. inflammation, tissue injury, DNA damage). One such oxidizing species is hydrogen peroxide (H 2 O 2 ). While H 2 O 2 has a relatively low reactivity, it can accumulate to high concentrations in the cell due to its stability and be converted to more reactive species (e.g. hydroxyl radical, HO ). The removal of H 2 O 2 by antioxidant enzymes is therefore very important. The three main antioxidant enzymes responsible for the removal of H 2 O 2 include: catalase, glutathione peroxidase (GPx), and the peroxiredoxins (Prx) (Figure 1.4). All catalyze the decomposition of H 2 O 2, however both GPx and Prx are reliant on the recycling of reducing equivalents. Kinetic modeling and experimental data have indicated that catalase is considered to be the major enzyme that detoxifies high concentrations of H 2 O 2. GPx and Prx are both very critical in removing low fluxes of H 2 O 2, but at higher than physiological concentration, such as those that could result from an oxidative insult by a xenobiotic, first-order removal by catalase becomes dominant [25, 37]. We utilized quantitative methods to explore absolute differences in the ability of cells to remove H 2 O 2. We hypothesized that there would be a wide range of antioxidant capacities across different tumor and normal cells; however, normal cells would be better 78

96 able to remove H 2 O 2 compared to tumor cells. We further predicted that the rate constant at which cells remove H 2 O 2 would be correlated with their catalase activity. Results Normal Cells have Higher Capacities for the Removal of H 2 O 2 in Comparison to Tumor Cells To investigate differences in the capacity of tumor and normal cells to remove H 2 O 2 the rate constant at which they removed extracellular H 2 O 2 from the medium was determined. When the rate constants (k cell ) by which cancer cells and normal cells of a variety of tissue types (i.e. skin, breast, pancreas, lung, brain, tongue, pharynx, liver, and intestine) remove extracellular H 2 O 2 were determined, results showed that both cancer cells (Table 4.1) and normal cells (Table 4.2) have a wide range of capacities to remove extracellular H 2 O 2 (Figure 4.1). On average normal cells (k cell = 5.5 x s -1 cell -1 L) have higher rate constants of removal of extracellular H 2 O 2 in comparison to cancer cells (k cell = 3.1 x s -1 cell -1 L) (Figure 4.1). Among a total 25 different cell lines measured, there was a 10-fold difference in the cell line with the lowest rate constant of H 2 O 2 removal (A375; 0.65 x s -1 cell -1 L) and the cell line with the highest rate constant of H 2 O 2 removal (Normal Human Astrocytes; 7.3 x s -1 cell -1 L). A wide range of rate constants for removal of extracellular H 2 O 2 was observed in the five different pancreatic cancer cell lines (Table 4.1), demonstrating that even within the same tissue types the k cell for H 2 O 2 removal varied. There was a five-fold difference in the rate constant for H 2 O 2 removal between 79

97 the MIA PaCa-2 cells (k cell = 1.1 x s -1 cell -1 L) and both the PANC-1 (5.1 x s -1 cell -1 L) and 339 cells (5.4 x s -1 cell -1 L) (Table 4.1). Catalase Activity Varies Across Tumor Cell Lines and Plays a Major Role in the Removal of Extracellular H 2 O 2 Since catalase is thought to be the primary enzyme involved in removing high fluxes of H 2 O 2, catalase activity was measured in eight tumor cell lines and its contribution to the k cell of H 2 O 2 removal was investigated. In parallel to k cell of H 2 O 2 removal, we observed that cancer cells of varying tissue origins had a wide range of catalase activity (Figure 4.2A). This variance in active catalase molecules per cell was also observed across cell lines of the same tissue type and was exemplified again in the pancreatic cancer cell lines observed (Figure 4.2A). As with the five-fold difference in the rate constant for H 2 O 2 removal, there was a five-fold difference in the catalase activity of MIA PaCa-2 (16,000 active catalase molecules per cell) and 339 cells (87,000 active catalase molecules per cell) (Figure 4.2A). As expected, the number of active catalase molecules per cell strongly correlated with the rate constants at which these cell lines remove extracellular H 2 O 2 (Figure 4.2B). Since catalase is thought to be the major contributing enzyme to the removal of high concentrations of H 2 O 2 it is not surprising that there is a strong correlation (R 2 = 0.88) between these two parameters in the cell lines (Figure 4.2B). The rate constant of H 2 O 2 removal being a function of the number of active catalase molecules per cell for the different tumor cells measured points to the major role it has in H 2 O 2 removal. 80

98 Manipulation of Catalase Activity Causes Significant Changes in the Capacity of Cells to Remove H 2 O 2 To further investigate the roles of the different enzymes involved in the removal of H 2 O 2, HepG2 cells, which have both a high basal catalase activity and a high rate constant for removal were utilized. Buthionine sulfoximine (BSO) inhibits gammaglutamyl-cysteine synthetase and will decrease the amount of glutathione in the cell. Glutathione is needed for the activity of GPx, and its recycling is a limiting factor in its activity at high fluxes of H 2 O 2. When glutathione synthesis was inhibited with BSO, there was no change in the k cell of H 2 O 2 removal. When catalase was inhibited using 3- amino-1,2,4-triazole (3-AT) in HepG2 cells, which have a high basal level of catalase activity, there was a near 4-fold decrease in the rate constant at which these cells remove extracellular H 2 O 2 (Figure 4.3A). These results both suggest and support the important role that catalase has in the removal of high concentrations of extracellular H 2 O 2. The number of active catalase molecules per cell calculated from the measurement of catalase activity in HepG2 cells following inhibition of catalase with 3-AT decreased 5-fold (Figure 4.3B). This decrease in catalase activity mirrors the decrease in the rate constant for extracellular H 2 O 2 removal (Figure 4.3A and B). Conversely, MIA PaCa-2 cells, which have a very low basal H 2 O 2 removal capacity and a markedly low catalase activity, were transduced with adenovirus catalase at varying virus loads (multiplicities of infection; MOI). Following transduction, the rate constant at which MIA PaCa-2 cells remove H 2 O 2 increased 1.5- to 80-fold with increasing MOI of adenovirus catalase (0-100 MOI) (Figure 4.4A). This increase in the rate constant for H 2 O 2 removal was directly correlated (R 2 = 0.91) with the resulting 81

99 active catalase molecules per cell that resulted post-transduction with adenovirus catalase (Figure 4.4B and C). Physical Parameters of Cells Effect the Capacity of Cells to Remove H 2 O 2 Cell size (i.e. intracellular volume) varies across different cell lines of both the same and different tissue types (Table 4.3). To determine whether the k cell of removal of H 2 O 2 might be a function of the size of the cell, the intracellular volume was measured and plotted against the rate constant for H 2 O 2 removal (Figure 4.5). There was a positive linear correlation between these two measures (Figure 4.5). The largest cells had the highest rate constant of removal of H 2 O 2 (PANC-1; 3.0 pl; 5.1 x s -1 cell -1 L), whereas the smallest cells had the lowest (A375; 1.59 pl; 0.65 x s -1 cell -1 L) (Table 4.3; Figure 4.5). Discussion Utilizing absolute quantitation, we measured the capacities of tumor and normal cells to remove hydrogen peroxide. Making comparisons across different cell lines is challenging; however, absolute quantitation of the parameters measured allowed for comparisons to be made across a variety of cell types with different cell sizes (i.e. intracellular volume). We observed the rate constants for removal of extracellular H 2 O 2 (k cell ) are on average 2-fold higher in normal cells than in cancer cells. Across the different tissue types as well as within the same tissue type (i.e. pancreatic cancer), there was a wide range of rate constants for removal of H 2 O 2 (k cell ). Previous biochemical studies of 82

100 various different normal tissues have shown that the endogenous levels of antioxidant enzymes vary greatly across the different tissue types [38]. It has been postulated that this reflects the different metabolism across different organs [39]. Our findings indicate that this may have implications in the sensitivity of these tissues to xenobiotics that generate H 2 O 2. Re-affirming the range of H 2 O 2 removal capacities, measurement of catalase activity of tumor cell lines of varying tissue origins also revealed a differential in the capacity of cells to remove H 2 O 2. Catalase activity strongly correlated with the k cell for removal of H 2 O 2. These findings may indicate that the catalase activity of a cell or tissue could be a good indicator of that tissue s ability to remove high fluxes of H 2 O 2. When different peroxide removal enzymes were inhibited and the k cell was measured, inhibition of catalase revealed a major impact. Inhibition of glutathione synthesis with buthionine sulfoximine (BSO), which will affect GPx activity, did not significantly alter the rate constant for removal of H 2 O 2. This provides additional evidence that catalase is a principal contributor in the removal of high concentrations of H 2 O 2. Conversely, when cells were transduced with adenovirus catalase to increase catalase activity there was an increase in the capacity of MIA PaCa-2 cells to remove H 2 O 2 that was MOI-dependent (0-100 MOI). The k cell for H 2 O 2 removal was a function of the catalase activity of the transduced MIA PaCa-2 cells. This also suggests the strong role of catalase in the removal of high concentrations of H 2 O 2 and shows modulation of catalase activity has a large effect on the capacity to remove H 2 O 2 within the same cell line in which all other parameters are held constant. 83

101 Conclusions In this study, the differences in the capacity of tumor vs. normal cells to remove H 2 O 2 were observed. We conclude that: 1. The rate constants (k cell ) for removal of extracellular H 2 O 2 are 2-fold higher in normal cells than in cancer cells, Table ; Figure The catalase activity of tumor cell lines of varying tissue origin revealed a differential in the capacity of the cells to remove H 2 O 2 that strongly correlated with the k cell of H 2 O 2 removal for each cell line, Figure Catalase plays a major role in removing high fluxes of H 2 O 2. a. Inhibition of catalase resulted in the most significant decrease in k cell of H 2 O 2 removal of the enzyme systems manipulated, Figure 4.3. b. Transduction of cells with adenovirus catalase resulted in a 1.5- to 80- fold increase in the k cell for removal that was strongly correlated with the catalase activity post-transduction, Figure 4.4. Data presented here show key differences between the capacities of tumor and normal cells to remove H 2 O 2, with catalase having a major role in the removal of high fluxes of H 2 O 2. This may have implications in the sensitivity of these tissues to xenobiotics that generate H 2 O 2, particularly in regards to chemotherapeutics that are redox-active. Future Directions These results establish that there is an absolute difference between the capacities of different cell types to remove H 2 O 2, with catalase being a major contributor to this 84

102 removal in vitro. A critical next step is determining if these same differences between the H 2 O 2 removal capacity of tumor and normal cells, as well as the catalase activity also occurs in vivo. A logical next step in these studies is to measure the catalase activity of adjacent tumor and normal tissue and determine the rate constant for H 2 O 2 removal to insure these findings hold true in a physiological environment. In particular, the question arises whether there is a range of catalase activities across different tumors? If so, this can have broad implications about how to go about treating those tumors differently. The results in this chapter provide strong additional evidence that catalase is the major enzyme involved in metabolizing H 2 O 2. It would be informative to do a stepwise inhibition of the other enzymes and co-factors involved in peroxide-removal networks of the cell. Utilizing absolute quantitation to do so would provide valuable information about the contributions of each of these enzymes in removing H 2 O 2. Catalase is largely localized to the peroxisomes of cells. The wide-range of catalase activities across the different pancreatic cancer cells (in particular) brings up the question of whether there are more peroxisomes in the cells with higher catalase activity. To answer this question, the peroxisomes can be stained and counted. Likewise, catalase can be stained and the amount of it packaged into the peroxisomes quantified. This may provide insight into why these differences in catalase activity and k cell for H 2 O 2 removal occur across different cell types. The finding that k cell of H 2 O 2 removal correlates with cell size would be an interesting finding to pursue further. While this relationship can be seen across different cell types, there are other factors that may be different across the cell lines and confound this. Modifying the size of the same cell type and then measuring k cell of H 2 O 2 removal to 85

103 confirm size-dependence would be a good follow-up study. While it is expected that modifying the cell size will impact the k cell of H 2 O 2 removal due to its effect on diffusion, it would be valuable to know the contribution of this that is reflected in the k cell. 86

104 Table 4.1: Rate constants (k cell ) for H 2 O 2 removal by tumor cells Cell Line Disease k cell (x10-12 s -1 cell -1 L) (SEM) MIA PaCa-2 Pancreatic 1.1 (0.1) Cancer AsPC (0.9) PANC (1.1) (0.7) (0.3) A375 Melanoma 0.65 (0.21) Cal27 Head & Neck 2.3 (0.6) Cancer FaDu 2.3 HepG2 Liver Cancer 4.2 (0.6) MB231 Breast Cancer 1.0 H292 Lung Cancer 3.0 (0.4) H (0.4) A (0.3) U87 Glioblastoma 4.8 (0.6) U (0.3) 87

105 Table 4.2: Rate constants (k cell ) for H 2 O 2 removal by normal cells Cell Line Tissue k cell (x10-12 s -1 cell -1 L) (SEM) H6c7 Pancreas 3.7 (0.4) Melanocytes Skin 6.3 (1.3) NHF (12 y) 5.9 NHF (46 y) 4.7 (0.8) NHA (#1) Brain 6.8 (0.7) NHA (#2) 4.4 (0.3) NHA (#3) 7.3 (0.2) HBePC Lung 6.7 (0.6) Red Blood Cells Blood 4.0 FHs74int Intestinal

106 8 k cell (x s -1 cell -1 L Tumor Normal Figure 4.1. Normal cells have a more robust capacity to remove extracellular H 2 O 2 than tumor cells. The rate constants, k cell, at which 15 tumor cell lines (listed in Table 4.1) and 10 normal cell lines (listed in Table 4.2) remove H 2 O 2 were measured. There was a wide range of capacities for removal of H 2 O 2 across all cell types. On average, normal cells (k cell = 5.5 x s -1 cell -1 L) had a 2-fold higher rate constant for removal H 2 O 2 than tumor cells (k cell = 3.1 x s -1 cell -1 L) (p < 0.05). 89

107 A Active Catalase Molecules per Cell MIA PaCa-2 MB231 A549 AsPC-1 HepG2 403 PANC B. Rate Constant of H 2 O 2 removal (k cell ) (x s -1 cell -1 L) A549 MIA PaCa-2 y = 2.6ln(x) R² = 0.9 MB231 HepG2 AsPC-1 PANC Active Catalase Molecules per Cell Figure 4.2. Catalase activity varies across cancer cell lines and correlates with the rate constant of H 2 O 2 removal (k cell ). (A) Catalase activity for cell lines of different tissue origins (i.e. pancreas, breast, lung, and liver) were determined and used to calculate the effective number of fully active catalase molecules per cell. This number varied 5-fold across the different cancer cell lines: from 16,400 molecules per cell (MIA PaCa-2) to 87,000 molecules per cell (339) (n = 3-9, error bars are standard error of the mean). (B) There is a strong correlation between the rate constant at which these cell lines remove extracellular H 2 O 2 and the number of active catalase molecules per cell (R 2 = 0.88). 90

108 A. k cell / s -1 cell -1 L HepG2 100 um BSO 20 mm 3-AT B Active Catalase Molecules per cell HepG2 HepG2 with 3-AT Figure 4.3. Inhibition of catalase has a major effect on the H 2 O 2 removal capacity of HepG2 cells. (A) Treatment of HepG2 cells with 100 µm buthionine sulfoximine (BSO) 24 h prior to the H 2 O 2 -removal assay to inhibit glutathione synthesis did not result in any change in the rate constant by which these cells remove H 2 O 2. However, treatment of HepG2 cells with 20 mm 3-AT for 1 h to inhibit catalase resulted in a four-fold decrease in the rate constant by which HepG2 cells remove extracellular H 2 O 2 (n = 4, error bars are standard error of the mean). (B) Catalase activity decreased 5-fold upon treatment with 20 mm 3- AT (n= 3, error bars are standard error of the mean). 91

109 A. k cell (x s -1 cell -1 L MIA PaCa-2 1 MOI 5 MOI 10 MOI 25 MOI 50 MOI 100 MOI B. Active catalase molecules per cell MIA PaCa-2 1 MOI 5 MOI 10 MOI 25 MOI 50 MOI 100 MOI C. k cell / s -1 cell -1 L MOI R² = MOI 3 5 MOI 2 1 MOI 1 0 MOI Active Catalase Molecules per Cell (x 10 5 ) Figure 4.4. Increasing catalase activity in MIA PaCa-2 cells directly increases their capacity to remove H 2 O 2 (A) Transduction of MIA PaCa-2 cells with adenovirus catalase (1-100 MOI) resulted in 1.5- to 80-fold increases in the rate constant by which these cells remove H 2 O 2 (n=4, error bars are standard error of the mean). (B) Catalase activity after transduction with MOI adenovirus catalase increased 1.5- to 2,500-fold from basal catalase activity of MIA PaCa-2 cells (n = 2). (C) There is a direct correlation between the number of active 92

110 catalase molecules per cell and the rate constant for removal of H 2 O 2 following transduction of MIA PaCa-2 cells with adenovirus catalase (R 2 = 0.91). 93

111 Table 4.3 Intracellular volumes (pl) of cell lines Cell Line MIA PaCa-2 AsPC-1 PANC-1 H6c7 A375 HepG2 MB231 A549 Intracellular volume (pl) a 2.36 (0.02) b 2.4 c 3.0 c 2.87 c 1.59 c 2.9 (0.09) b 2.29 (0.01) b 2.53 (0.01) b a Intracellular volume was measured using the Moxi TM Z Mini Automated Cell Counter (ORFLO Technologies), n = 3 for each cell line. Presented is the mean of triplicate biological samples. b Standard error of the mean of three triplicate biological samples. c n = 1 for each cell line k cell / s - 1 cell - 1 L R² = Intracellular Volume (pl) Figure 4.5. Rate constants of H 2 O 2 removal correlate (k cell ) with intracellular volume. The intracellular volume of 8 different cell lines was measured using the Moxi TM Z mini automated cell counter (ORFLO Technologies) (n = 3; mean shown). This measured intracellular volume was plotted against the rate constant of H 2 O 2 removal (k cell ) for each cell line. There was a direct correlation between the k cell and intracellular volume of each cell line (R 2 = 0.77). 94

112 CHAPTER 5: IMPLICATIONS FOR SENSITIVITY OF CELLS TO PHARMACOLOGIC ASCORBATE IN CANCER THERAPY Introduction Ascorbate (AscH - ) functions as a versatile reducing agent in biology. When used at low, physiological concentrations it exhibits antioxidant properties; however when used at pharmacological doses (P-AscH - ) achievable through intravenous injection, it can readily oxidize and deliver a high flux of hydrogen peroxide (H 2 O 2 ) [55, 117, 57, 56]. This unique property of P-AscH - is currently being investigated as an adjuvant treatment in cancer therapy. Several in vitro and in vivo studies have shown a differential toxicity of P-AscH - across different cancer types and between cancer cells and normal cells of the same tissue type [55, 56, 57, 67, 118, 119, 120, 121, 122, 123, 124, 125]. These studies have implicated the H 2 O 2 produced from the oxidation of P-AscH - as the principal mediating factor in its cytotoxicity to cancer cells. The differential sensitivity of cancer cells of different tissue types to P-AscH -, as well as their increased sensitivity over normal cells may be due to differences in their ability to remove H 2 O 2 and their endogenous levels of antioxidant enzymes to detoxify this H 2 O 2 (Chapter 4). Previous biochemical studies, as well as our results in Chapter 4, demonstrated that the intrinsic levels of antioxidant enzymes are low in a majority of cancer cell types as compared to non-transformed cells [38, 39]. In general, the levels of catalase were found to be low in cancer cells, but catalase activity also varied greatly across different cancer cell lines [38]. This suggests that the vast majority of cancer cells may lack the machinery needed to detoxify H 2 O 2 ; our use of absolute quantitation to determine the rate constant for removal of H 2 O 2 supports this. This may correspond to the differential in the sensitivity to H 2 O 2 -producing agents (i.e. P-AscH - ) and could be utilized to predict which 95

113 tumors may respond best to these types of therapies, as well as be manipulated to increase the efficacy. The wide-range of capacities at which different tumor cell types remove H 2 O 2 as well as their decreased capacity to remove H 2 O 2 in comparison to normal cells discussed in Chapter 4 led us to investigate whether the sensitivity of tumor cells to pharmacological ascorbate might be related to their capacity to remove H 2 O 2. We hypothesize that that across different tumor cell types there will be a differential sensitivity to P-AscH - that is correlated with their individual capacities to remove H 2 O 2. We investigated this in five different pancreatic cancer cell lines (including two patientderived cell lines). Results Pharmacological Ascorbate is Oxidized in Cell Culture Medium Oxidation of P-AscH - in both in vitro and in vivo settings generates a flux of H 2 O 2 [56]. The rate of oxygen consumption (OCR, -d[o 2 ]/dt) upon addition of ascorbate to DMEM cell culture medium was measured using a Clark electrode oxygen monitor (Figure 5.1). Addition of P-AscH - to DMEM cell culture medium complete with 10 % FBS to a final concentration of 6 mm resulted in an oxygen consumption rate of 48 nmol L -1 s -1, which represents the rate of H 2 O 2 production (Figure 5.1). Addition of catalase indicated an accumulation of 18 µm H 2 O 2 in the medium over the course of an experiment (Figure 5.1). In a typical experimental setting in which 125,000 cells were treated with 6 mm ascorbate in 3.0 ml of DMEM medium, this would result in the cells being exposed to a 1.2 pmol cell -1 s -1 flux of H 2 O 2. 96

114 The Dose of Pharmacological Ascorbate is Best Specified on a per Cell Basis In Chapter 3, we demonstrated that the specification of dose of xenobiotics (i.e. 1,4-benzoquinone and oligomycin A) in cell culture studies as moles of xenobiotic per number of cells in the experiment used yields more consistent results and reduces ambiguity across different physical experimental set-ups [126]. This is particularly true of xenobiotics that make irreversible changes to their target biomolecules. As shown in Figure 5.1, oxidation of P-AscH - in in vitro (i.e. in cell culture medium) settings generates a flux of H 2 O 2 (Figure 5.1). The toxicity of H 2 O 2 results in both irreversible and reversible changes to biomolecules and has been shown to also be cell density dependent [97, 98, 99]. Dose of P-AscH - used in cell culture studies is currently reported in terms of its concentration in the medium. Figure 5.2 demonstrates that specifying dose as moles P-AscH - per number of cells exposed, yields more consistent results and reduces ambiguity. When P-AscH - is specified as moles per cell a clear dose-response is observed in the cytotoxicity of P-AscH - to MIA PaCa-2 cells (Figure 5.2B), whereas expression of dose as the initial concentration in the medium produces ambiguous results when different physical set-ups (i.e. number of cells exposed) are used (Figure 5.2A). The Differential Sensitivity to Ascorbate Across Pancreatic Cancer Cell Lines Correlates with the Capacity at which they Remove H 2 O 2 Previous studies have indicated that there is a range of cancer cell sensitivity to ascorbate concentrations in vitro across different tissue types [117, 118]. We have found this to be true within the same tissue type as well. MIA PaCa-2, AsPC-1, 403, 339, and PANC-1 cells had a differential sensitivity to P-AscH - as measured by the dose that was 97

115 effective in killing 50% of the cells in vitro (ED 50 ) (Figure 5.4A). PANC-1 cells had an ED 50 of P-AscH - that was two times greater than MIA PaCa-2 cells, showing that MIA PaCa-2 cells were significantly more sensitive to P-AscH - than PANC-1 cells (Figure 5.4A). The five different pancreatic cancer cells examined have very different capacities to remove extracellular H 2 O 2, as observed by the rate constant at which they remove extracellular H 2 O 2 as well as the catalase activity of the cell lines (Table 4.1, Figure 5.3A and B). A western blot for catalase protein in the five pancreatic cancer cell lines revealed higher catalase protein levels in PANC-1 cells than MIA PaCa-2, AsPC-1, 403, and 339 cells (Figure 5.3C). While this is consistent with PANC-1 cells having among the highest catalase activities and k cell of H 2 O 2 removal, it did not reveal the differences among the other cell lines that we were able to detect with biochemical assay for catalase activity and rate constant of H 2 O 2 removal. For example, the 339 cells have a similar high capacity to remove H 2 O 2 as the PANC-a cells, but did not have higher catalase protein levels in comparison to the MIA PaCa-2, AsPC-1, and 403 cells (Figure 5.3C and D). The ED 50 of P-AscH directly correlated with the rate constant at which the cells remove extracellular H 2 O 2 (R 2 = 0.69, Figure 5.4B). MIA PaCa-2 cells were most sensitive to P-AscH - and had the lowest capacity to remove extracellular H 2 O 2, whereas PANC-1 cells were the least sensitive to P-AscH - and had the highest capacity of removal (Figure 5.4B). These results, showing strong correlations between the capacity of cells to remove extracellular H 2 O 2 and ED 50 of P-AscH -, support the important role of the H 2 O 2 removal system in the resulting toxicity observed from P-AscH -. This supports that P- AscH - may be more effective in cells that have a lower capacity to remove H 2 O 2. 98

116 Modulation of Catalase Activity in the Same Cell Line Mimics Results Seen in Different Cell Lines Across different pancreatic cancer cell lines we observed a strong correlation between the capacity for these cells to remove H 2 O 2 and their sensitivity to ascorbate. This was explored further within a single cell line, MIA PaCa-2 cells, following transduction with adenovirus catalase at varying MOIs (0-25 MOI) (Figure 5.5). As expected, we saw a shift in the dose-response curve following treatment with ascorbate that was MOI-dependent (Figure 5.5A). The dose of ascorbate that decreased clonogenic survival by 50 % (ED 50 ) very strongly correlated with the catalase activity resulting from the transduction of varying MOIs of adenovirus catalase (R 2 = 0.94) (Figure 5.5B). This is consistent with the results obtained across the different pancreatic cancer cell lines that have a wide-range of basal catalase activities (Figure 5.4B). Inhibition of Catalase Sensitizes PANC-1 cells to Pharmacological Ascorbate Catalase varies across tumor cell lines and plays a major role in the removal of H 2 O 2 generated at doses comparable to those capable of being generated by P-AscH - (Figure 4.2 and 4.3). The H 2 O 2 removal capacity of pancreatic cancer cell lines correlated with the ED 50 of P-AscH - in cell culture, with PANC-1 cells being the most resistant to P-AscH - and having the most robust capacity to remove extracellular H 2 O 2 (Figure 5.4B). We investigated whether inhibiting catalase activity in this cell line that has a robust capacity to deal with a large insult of H 2 O 2 would sensitize these cells to P- AscH -. When catalase was inhibited with 3-AT in PANC-1 cells prior to treatment with P-AscH -, the cells were sensitized to P-AscH - (Figure 5.6A). The dose of P-AscH - 99

117 needed to decrease clonogenic survival by 50 % was 1.5-fold less when cells were pretreated with 3-AT (Figure 5.6A). Pretreatment with 3-AT resulted in a 1.5-fold reduction in the rate constant at which PANC-1 cells remove H 2 O 2 (Figure 5.6B) and a 2-fold decrease in catalase activity (Figure 5.6C). The strong correlation between catalase activity and sensitivity to P-AscH -, as well as the effect of 3-AT inhibition of catalase on the rate constant of H 2 O 2 removal emphasize the role of catalase in the removal of H 2 O 2 at high concentrations, such as those achievable by P-AscH -. ATP Correlates with Clonogenic Survival Following P-AscH - Treatment The treatment of cancer cells with P-AscH - causes a dose-dependent decrease in ATP content per cell immediately after treatment (Figure 5.2). This decrease in ATP content per cell was investigated further in the different pancreatic cancer cell lines. Figure 5.7 shows the correlation of ATP content per cell immediately after a 1-h exposure of AsPC-1 cells to P-AscH - strongly correlates with clonogenic survival (R 2 = 0.83). Although not shown, this occurs in all pancreatic cell lines immediately following exposure to P-AscH -. Discussion Pharmacological ascorbate oxidizes in cell culture medium to generate a measurable flux of H 2 O 2. Several in vitro studies have indicated H 2 O 2 as the major cytotoxicity-mediating species of P-AscH - in cancer therapy [118]. There is a wide-range of capacities at which different tissue types remove H 2 O 2. We were able to quantitatively 100

118 determine such capacities for 10 different normal tissue cell types and 15 different cancer cell lines (Chapter 4). On average, the normal cells examined removed H 2 O 2 with a rate constant that was 2-fold higher than the cancer cell lines tested. We observed a large range in these rate constants of H 2 O 2 both across different tissue types and within different cell lines of the same tissue origin. In particular, there was a wide-range of k cell for removal of H 2 O 2 across the different pancreatic cancer cell lines (5-fold). P-AscH - has been studied thoroughly in the pancreatic cancer model in vitro, in vivo, and in clinical trials [44, 46, 56, 57, 67]. Utilizing the quantitative dosing metric (mol cell -1 ) established in Chapter 3 we were able to compare the absolute dose that was lethal to 50 % of the cells (ED 50 ) across five different pancreatic cancer cell lines, without ambiguity resulting from the physical conditions at which the experiments were carried out. We observed the k cell for removal of H 2 O 2 across the pancreatic cancer cell lines directly correlated with their sensitivity to P-AscH - (as measured by the ED 50 ). Our data support previous studies findings that catalase is the major contributor to the removal of high fluxes of H 2 O 2 in tumor cells. We observed that both increasing and decreasing the catalase activity had a significant effect on the rate constant of H 2 O 2 removal. We further investigated if similar manipulation of basal catalase activity would affect the cells sensitivity to P-AscH -. Increasing the catalase activity within the same cell line (MIA PaCa-2) increased resistance to P-AscH -. Many differences exist between different cell lines of both the same and different tissue origin; this result supports the contribution of catalase activity in protecting cells from P-AscH - and limits the other confounding factors that may be 101

119 present across the different cell lines. We saw this effect at MOIs that are 4 times less than MOIs that are used during typical viral transductions (25 MOI vs. 100 MOI). Transducing MIA PaCa-2 cells with 25 MOI increased the catalase activity 50-fold, whereas 100 MOI increased the catalase by 2,500-fold. It may be advantageous to transduce with lower MOIs than is the current practice. This may limit some of the offtarget effects of the virus on cells. For example, plating efficiency decreases with increasing MOI of adenovirus catalase. This can potentially confound results. Decreasing catalase activity increased sensitivity to P-AscH -. This suggests that a pharmacological inhibitor of catalase activity in tumor cells may be an effective combination therapy to increase the efficacy of P-AscH. - In these studies, we used 3-AT to inhibit catalase. While this catalase inhibitor is not currently utilized in the clinic or in vivo because it is not specific to tumor cells, there are other natural products that are potential catalase inhibitors currently being investigated. This includes: salicylic acid, anthocyanidins, methyldopa, and neutralizing antibodies [127, 128]. Thus, advances in targeting along with these types of reagents might offer an advantage. Furthermore, these results provide additional evidence that H 2 O 2 is involved in the mechanism of P-AscH - toxicity to cancer cells. The strong correlation between the capacity of different pancreatic cancer cells to remove H 2 O 2 and their sensitivity to P- AscH - suggests that in vivo measurement of catalase activity in tumors may predict which cancers will respond best to P-AscH - therapy. This information can also be used in finding combination therapies that may increase the efficacy of treatment for those tumors with higher catalase activities. For example, manganoporphyrins increase the flux of H 2 O 2 generated from P-AscH - when used in combination with P-AscH - [56]. They 102

120 have been shown to be synergistic with P-AscH - in in vitro and in vivo animal studies [56]. For tumor cells that have an increased capacity to remove H 2 O 2, combinations including agents (i.e. manganoporphyrins) that increase the flux of H 2 O 2 may be necessary. The correlation of decreased ATP content with clonogenic survival following exposure to P-AscH - suggests that in vivo measurement of ATP in tumors during and after P-AscH - therapy may indicate response to the treatment. Conclusions In this study, we observed that the differential sensitivity to P-AscH - across pancreatic cancer cells was strongly correlated with their individual capacities to remove extracellular H 2 O 2. We conclude that: 1. Pharmacological ascorbate is oxidized in cell culture medium to generate a flux of H 2 O 2, Figure Consistent with findings in Chapter 3, the dose of P-AscH - is also best expressed on a per cell basis, Figure Consistent with findings in Chapter 4, the catalase activity of different pancreatic cancer cell lines revealed a differential in the capacity of the cells to remove H 2 O 2, Figure The ED 50 of pharmacological ascorbate correlated with the capacity of pancreatic cancer cells to remove extracellular H 2 O 2, Figure Increasing catalase activity in MIA PaCa-2 cells (low basal catalase) increases resistance to exposure to ascorbate as seen by ED 50 for clonogenic survival, 103

121 while inhibition of catalase sensitizes PANC-1 cells (high basal catalase) to ascorbate parallel to the decrease in k cell, Figure This provides additional evidence that H 2 O 2 is involved in the mechanism of P-AscH - toxicity to cancer cells and that catalase activity is critical in removing this H 2 O 2. These results indicate that an in vivo measurement of catalase activity in tumors may predict which cancers will respond to P-AscH - therapy. This information can also be used in finding combination therapies that may increase the efficacy of treatment for those tumors with higher catalase activities. Future Directions We have shown that across different pancreatic cancer cell lines there are differences in catalase activity, as well as the overall capacities of these cell lines to remove H 2 O 2. These cell lines are all maintained in cell culture conditions. It would be a valuable next step to measure catalase activity in tumor biopsies to see if these differences persist across various tumors in vivo. This would be crucial in determining if it would be possible to assess whether measuring catalase activity levels in tumors may be a marker for which tumors may respond best to P-AscH -. We have shown in vitro that increasing catalase activity within the same tumor cell line, increased resistance to P-AscH -. Replicating these results in vivo would be vital in understanding if this information will be translational. To achieve this, stablytransduced cancer cells with different catalase activities could be injected into mice to form tumors with varying catalase activities. Then the mice can be treated with P-AscH - 104

122 and tumor growth can be tracked to determine if P-AscH - inhibits tumor growth more effectively in mice where lower catalase activity tumor cells were implanted. As previously mentioned, 3-AT is not currently used in vivo. Studying safe methods to selectively inhibit catalase in tumor cells in vivo could provide a potential combination therapy with P-AscH - that could increase its effectiveness in cancer therapy. Some pharmacological methods of inhibiting catalase in vivo include salicylic acid, anthocyanidins, methyldopa, and neutralizing antibodies. These compounds in combination with P-AscH - may work synergistically to inhibit tumor growth, but further studies would have to be done to evaluate the effectiveness as well as safety of these compounds with P-AscH -. The selectivity of these agents to tumor cells is not widelystudied and further investigation of this will be very important to its use in combination with P-AscH

123 DMEM (10 % FBS) AscH - (6 mm) [O 2 ] / umol L d[o 2 ]/dt = 48 nmol / L s -1 Catalase (500 U/mL) [H 2 O 2 ] = 18 μmol/l Time / s Figure 5.1. Ascorbate is oxidized in DMEM generating a flux of H 2 O 2. The rate of oxygen consumption of AscH - (6 mm) in DMEM with 10 % FBS is 48 nmol L -1 s -1. Addition of catalase leads to a return of oxygen, which indicates that 18 µmol L -1 of H 2 O 2 accumulated in the medium. 106

124 A. Intracellular [ATP] (mm) ,000 cells 473,000 cells 365,000 cells 165,000 cells 80,000 cells 45,000 cells Ascorbate (mm) B. Intracellular [ATP] (mm) ,000 cells 473,000 cells 365,000 cells 165,000 cells 80,000 cells Ascorbate (pmol cell -1 ) 45,000 cells Figure 5.2. Dose of ascorbate is better specified on a per cell basis (pmol cell -1 ) than as initial concentration in the medium (mm). MIA PaCa-2 cells at varying cell densities (45, ,000 cells) were treated with 5 mm ascorbate in 3.0 ml of medium for 1 h; ATP was measured immediately after. Dose of ascorbate is expressed as: (A) initial concentration of ascorbate in the medium; and (B) absolute amount of ascorbate (pmol) per cell. 107

125 A k cell / s -1 cell -1 L B Active Catalase Molecules per Cell (x10 5 ) C 60 kda Catalase 37 kda GAPDH D Ratio of Net Catalase: Net Loading Control AsPC-1 PANC-1 MIA PaCa-2 Figure 5.3. The capacity to remove H 2 O 2 varies across different pancreatic cancer cell lines. The H 2 O 2 removal capacity of five different pancreatic cancer cell lines was measured in Chapter 4. (A) The rate constant of H 2 O 2 removal (k cell ) varied 5-fold across the five cell lines. MIA PaCa-2 cells removed H 2 O 2 the slowest with k cell = 1.1 x s -1 cell -1 L. PANC-1 and 339 cells removed H 2 O 2 significantly faster with k cell = 5.1 x s -1 cell -1 L and 5.4 x s -1 cell -1 L, respectively. (B) The catalase activity of the different pancreatic cancer cell lines was measured and then used to calculate the number of active catalase molecules per cell. Similar to the rate constants for H 2 O 2 removal, the MIA 108