Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Hypoxia and rapamycin induced changes to the cell cycle of multicellular spheroids and human tumour xenografts… Wong, Michelle 2006

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_2006-200209.pdf [ 13.77MB ]
Metadata
JSON: 831-1.0092948.json
JSON-LD: 831-1.0092948-ld.json
RDF/XML (Pretty): 831-1.0092948-rdf.xml
RDF/JSON: 831-1.0092948-rdf.json
Turtle: 831-1.0092948-turtle.txt
N-Triples: 831-1.0092948-rdf-ntriples.txt
Original Record: 831-1.0092948-source.json
Full Text
831-1.0092948-fulltext.txt
Citation
831-1.0092948.ris

Full Text

HYPOXIA AND RAPAMYCIN INDUCED CHANGES TO THE CELL CYCLE OF MULTICELLULAR SPHEROIDS AND HUMAN TUMOUR XENOGRAFTS LEADING TO POTENTIAL THERAPEUTIC ADVANTAGE by MICHELLE WONG B.Sc, The University of British Columbia, 1999 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY In THE FACULTY OF GRADUATE STUDIES (Experimental Medicine) THE UNIVERSITY OF BRITISH COLUMBIA October 2006 © Michelle Wong, 2006 ABSTRACT The cell cycle is a tightly regulated process that is functioning optimally when all factors contribute at the precise time and level that is required. However, in tumours some of these pathways are malfunctioning and to study this, we are utilizing spheroids as our in vitro tumour model system. In this thesis, we hypothesize that spheroids once adequately characterized will provide an invaluable means to assess the potential for using cytostatic agents to minimize accelerated repopulation and target radioresistant cells during radiation therapy of multicellular systems. The general methods involved for the majority of the work included immunoblotting, flow cytometry, as well as clonogenic assays. We have studied differences in the cell cycle time of subpopulations within each spheroid, where cells near the necrotic center showed a prolonged cell cycle time. Cycle time required by the peripheral cell layers was more rapid suggesting that cell kinetic variations may likely be a result of the local microenvironmental conditions such as hypoxia. Cyclin B1 and D levels were measured in spheroid populations that were subjected to a range of anoxic to aerobic conditions where cyclin levels decreased with decreasing concentrations of oxygen. Interestingly, these observations illustrate that measuring cyclin levels can provide a quick and convenient index for proliferation rate. The cytostatic agent, rapamycin, was analyzed for its influence on cell cycle progression and its affect on cyclin levels. Using human WiDr tumour xenografts, we found that the combination of rapamycin treatment with radiation had additive therapeutic benefits. Accelerated repopulation has been established as a limitation in radiation therapy of some cancers. Therapeutic advances have led to using accelerated fractionation regimens in an attempt to counteract the tumours' inherent ability to grow in the face of continued therapy. Since accelerated repopulation of irradiated tumours may be i i associated with the recruitment of quiescent cells into the cell cycle, rapamycin can potentially be used to control those cells. By incorporating rapamycin and concurrently administering radiation, we found a synergistic effect where the cell cycle time of the solid xenograft tumours were even slower than with either treatment alone. 111 TABLE OF CONTENTS ABSTRACT ii TABLE OF CONTENTS iv LIST OF FIGURES viii LIST OF SYMBOLS AND ABBREVIATIONS x ACKNOWLEDGEMENTS xiii CHAPTER I INTRODUCTION 1 1.1 TUMOUR MICROENVIRONMENT 2 1.2 SPHEROIDS: AN IN VITRO TUMOUR MODEL 4 1.2.1 Structure and growth 4 1.2.2 Physiology 7 1.3 FLUORESCENCE ACTIVATED CELL SORTING AND FLOW CYTOMETRY OVERVIEW 9 1.3.1 Hoechst 33342 9 1.3.2 Cell sorting techniques 12 1.3.2.1 Volume sort 12 1.3.2.2 Depth sort 14 1.3.3 Flow cytometry analysis 14 1.4 TUMOUR HYPOXIA 17 1.4.1 Classification of chronic versus transient hypoxia 17 1.4.2 Methods of hypoxia detection 17 1.4.2.1 Direct methods 19 1.4.2.2 Indirect methods 19 1.4.2.3 Spectroscopic methods 20 1.4.2.4 Bioreductive chemical probes 20 1.4.2.5 Intrinsic markers 21 1.4.3 Hypoxia in Spheroids 22 1.5 CELL CYCLE 23 1.5.1 Phases and Checkpoints 23 1.5.2 Cyclins and Cyclin Dependent Kinases 23 1.5.3 Cyclin Dependent Kinase Inhibitors 29 1.6 RADIOSENSITIVITY 32 1.6.1 Hypoxia and Radiation Therapy 32 1.6.2 Cyclin Alterations 33 1.7 RAPAMYCIN OVERVIEW: A CYTOSTATIC AGENT 37 1.8 RESEARCH OBJECTIVES 40 1.8.1 Specific Aims 40 1.8.2 Thesis Overview 41 iv CHAPTER II CELL CYCLE KINETIC STUDIES OF SPECIFIC REGIONS IN A SPHEROID 43 2.1 INTRODUCTION 44 2.2 MATERIALS AND METHODS 45 2.2.1 Spheroids 45 2.2.2 Reagents 47 2.2.3 Dual labelling with thymidine analogues 47 2.2.4 Cell cycle time kinetics 51 2.2.5 Fluorescence activated cell sorting 52 2.2.6 Flow cytometry with antibodies 53 2.3 RESULTS 55 2.3.1 Cell cycle times in spheroids vary by region 55 2.3.2 Labelling index of BrdUrd and IdUrd 61 2.4 DISCUSSION 64 CHAPTER III HYPOXIA INFLUENCED CHANGES IN THE LEVELS OF THE MITOTIC CYCLIN B1 WITHIN REGIONS OF A SPHEROID 68 3.1 INTRODUCTION 69 3.2 MATERIALS AND METHODS 72 3.2.1 Spheroids and reagents 72 3.2.2 Regulation of oxygen concentration 72 3.2.3 Trypsinization of spheroids 72 3.2.4 Western blotting 74 3.2.5 Fluorescence activated cell sorting 76 3.2.6 Flow cytometry with antibodies 76 3.2.7 Cryostat sectioning of spheroids 76 3.3 RESULTS 78 3.3.1 Cyclin B1 protein levels decrease with depth within a spheroid 78 3.3.2 Hypoxic conditions affect the levels of Cyclin B1 observed in spheroid subpopulations 80 3.3.3 Cell cycle distribution of the cells within the subpopulations of the spheroid 80 3.3.4 Cryosections of WiDr Spheroids for spatial determination of proteins under hypoxic treatment 84 3.4 DISCUSSION 87 CHAPTER IV OXYGEN DEPENDENCE OF CYCLIN D LEVELS IN SPHEROIDS COMPARED TO SINGLE CELLS 94 4.1 INTRODUCTION 95 4.2 MATERIALS AND METHODS 96 4.2.1 Spheroids, single cells, and reagents 96 4.2.2 Regulation of oxygen concentration 96 4.2.3 Single cell timepoints 96 4.2.4 Western Blotting .: 97 4.2.5 Fluorescence activated cell sorting 98 4.2.6 Flow cytometry with antibodies 98 4.2.7 Cryostat sectioning of spheroids 98 4.3 RESULTS 99 4.3.1 Cyclin D protein levels decline with increasing depth into a spheroid 99 4.3.2 Hypoxic conditions affect the levels of Cyclin D observed in spheroid subpopulations 99 4.3.3 Cell cycle distribution of the cells within the subpopulations of the spheroid 104 4.3.4 Cryosections of the WiDr Spheroids for spatial determination of proteins under hypoxic treatment 104 4.3.5 Hypoxic conditions minimally affect the levels of Cyclin D observed in single cell suspensions 105 4.4 DISCUSSION 109 CHAPTER V EFFECTS OF RAPAMYCIN IN VITRO 117 5.1 INTRODUCTION 118 5.2 MATERIALS AND METHODS 123 5.2.1 Spheroids and reagents 123 5.2.2 Toxicity, dose response, and time course of rapamycin on spheroids 123 5.2.3 Fluorescence activated cell sorting 124 5.2.4 Flow cytometry with antibodies 124 5.2.5 Cytospin slides of cells treated with rapamycin 124 5.2.6 Levels of cyclin D FACS analyzed prior to rapamycin treatment 125 5.3 RESULTS 127 5.3.1 Rapamycin toxicity in cells 127 5.3.2 Influence of rapamycin on Cyclin D levels in spheroids 127 5.3.3 Duration of rapamycin contact influences Cyclin D activity 130 5.3.4 Cytospin slides of single cells in contact with rapamycin 132 5.3.5 Determining the order of rapamycin effects 132 5.4 DISCUSSION 139 CHAPTER VI RAPAMYCIN UPTAKE IN VIVO AND USES FOR \ COMBINATION THERAPY WITH RADIATION 144 6.1 INTRODUCTION 145 6.2 MATERIALS AND METHODS 148 6.2.1 Tumours 148 6.2.2 Reagents 148 v i 6.2.3 Fluorescence activated cell sorting 149 6.2.4 Flow cytometry with antibodies 149 6.2.5 Rapamycin with lododeoxyuridine and Bromodeoxyuridine 150 6.2.6 Rapamycin in combination with radiotherapy 150 6.3 RESULTS 151 6.3.1 Rapamycin inhibits cell proliferation in vivo 151 6.3.2 Rapamycin sensitizes WiDr xenografts to Radiation 154 6.4 DISCUSSION 157 CHAPTER VII SUMMARY AND FUTURE DIRECTIONS 164 7.1 SUMMARY AND SIGNIFICANCE OF RESULTS 165 7.2 FUTURE DIRECTIONS 173 REFERENCES 175 v i i LIST OF FIGURES Figure 1.1 Spheroids 5 Figure 1.2 Hoechst 33342 molecular structure 10 Figure 1.3 Fluorescence activated cell sorting 11 Figure 1.4 , Volume and Depth Sorts 13 Figure 1.5 Flow Cytometry 15 Figure 1.6 Chronic diffusion-limited hypoxia 18 Figure 1.7 Diagrammatic view of the cell cycle 24 Figure 1.8 Schematic diagram of cyclin-Cdk interaction 25 Figure 1.9 Oscillations in cyclin levels during the cell cycle 27 Figure 1.10 Progression through the cell cycle 28 Figure 1.11 Two families of cyclin dependent kinase inhibitors 30 Figure 1.12 Molecular structure of rapamycin (C5 1H7 9N01 3) 38 Figure 2.1 Molecular structures of BrdUrd and IdUrd 46 Figure 2.2 Detection of incorporated BrdUrd (or IdUrd) 48 Figure 2.3 Single labelling with BrdUrd or IdUrd 49 Figure 2.4 Dual labelling with BrdUrd and IdUrd 50 Figure 2.5 Forward scatter and Time of flight flow cytometry dot plots 56 Figure 2.6 Single colour dot plots and DNA histograms 57 Figure 2.7 Dot plots and DNA content histograms of DAPI only samples 58 Figure 2.8 Cell cycle time kinetics 60 Figure 2.9 Labelling index for thymidine analogues 62 Figure 3.1 The gas apparatus used for controlling external oxygen concentrations 73 Figure 3.2 Western Blotting 75 Figure 3.3 Western Blots for Cyclin B1 and Cdk1 79 Figure 3.4 Flow cytometry graphs of Cyclin B1 on FACS sorted spheroids .... 81 Figure 3.5 Flow cytometry dot plots of cyclin B1 82 Figure 3.6 Percentage of cells in the S-phase of the cell cycle 83 Figure 3.7 Cryosections of WiDr spheroids 86 Figure 4.1 Western Blots for Cyclin D and Cdk4 100 Figure 4.2 Flow cytometry analysis of Cyclin D levels in spheroids 102 Figure 4.3 Flow cytometry dot plots of Cyclin D using WiDr spheroids 103 Figure 4.4 Cryosections of WiDr spheroids 106 Figure 4.5 Time course of the Cyclin D levels of single cells 107 Figure 5.1 mTOR protein 119 Figure 5.2 Toxicity of Rapamycin on human tumour cells 128 Figure 5.3 Cyclin D levels with varying doses of Rapamycin treatment 129 Figure 5.4 Levels of Cyclin D after treatment with 0.2ug/mL Rapamycin 131 Figure 5.5 Cytospin slides of WiDr cells treated with rapamycin 133 Figure 5.6 Rapamycin treatment on cells after FACS analysis using 21% oxygen 135 Figure 5.7 Rapamycin treatments on cells after FACS analysis using 0% and 21% oxygen 137 Figure 6.1 Labelling indices of WiDr xenografts after treatment with either 1 mg/kg or 3 mg/kg rapamycin 152 Figure 6.2 Potential doubling times (Tpot) of WiDr xenograft tumours 153 Figure 6.3 Labelling indices of WiDr xenografts after treatment with radiation and rapamycin 155 Figure 6.4 Potential doubling times (Tpot) of WiDr xenograft tumours after treatment with doses of radiation and rapamycin 156 ix LIST OF SYMBOLS AND ABBREVIATIONS 4E-BP1 eukaryotic initiation factor 4E binding protein 1 BrdUrd 5-bromo-2-deoxyuridine CAK Cdk-activating kinase cdc cell division cycle Cdk cyclin dependent kinase CKI cyclin dependent kinase inhibitor C0 2 carbon dioxide DAPI 4', 6-diamidino-2-phenylindole dihydrochloride hydrate DMSO dimethyl sulfoxide DNA deoxyribonucleic acid ECM extracellular matrix EGF epidermal growth factor FGi mean DNA content of cells in the G rphase FG2 mean DNA content of cells in the G2-phase FLU mean DNA content of labelled undivided cells Fs mean DNA content of cells in the S-phase FACS fluorescence activated cell sorting FBS fetal bovine serum FKBP12 FK506 binding protein Go quiescent cell stage of the cell cycle Gi first Gap phase in the cell cycle G 2 second Gap phase in the cell cycle HCI hydrochloric acid HPV human papilloma virus IdUrd 5-iodo-2-deoxyuridine i.p. intraperitoneal i.v. intravascular INK4 inhibitor of Cdk4 Kip kinase inhibitory proteins LI labelling index M mitosis phase in the cell cycle MEM minimal essential medium mRNA messenger RNA mTORC mammalian TOR complex mTOR mammalian target of rapamycin p70 S 6 K p70 S6 kinase PBS phosphate buffer saline PHAS-1 eukaryotic initiation factor 4E binding protein 1 PI3K phosphatidylinositol 3-kinase PIKK PI3K-related kinase family R point restriction point in the cell cycle Rb retinoblastoma protein RIPA radioimmunoprecipitation assay buffer RM relative movement RNA ribonucleic acid S synthesis phase in the cell cycle S6K S6 kinase protein SAGE serial analysis of gene expression SCID severe combined immunodeficient SDS sodium dodecyl sulphate t time T c cell cycle time Tpot potential doubling time T s time of S-phase TBS tris buffered saline TBS-T TBS containing Tween 20 TGF-p transforming growth factor-p Thr threonine TORC1 TOR complex 1 TORC2 TOR complex 2 UV ultraviolet VEGF vascular endothelial growth factor V0 initial volume of a tumour V(t) volume of a tumour at time t ACKNOWLEDGEMENTS My years spent at the BC Cancer Research Centre have enlightened my experience not only in scientific endeavours but also in the myriad of individuals I have had the pleasure to become acquainted with and whom have helped me enormously through this entire journey. First and foremost, I would like to thank my supervisor, Dr. Ralph Durand, for his endless patience and for giving me the opportunity to pursue my studies in his lab. Even as your administration obligations and responsibilities only got larger, I've never felt like I couldn't go and find you in your office for some guidance or ask you to solve any computer problem that the department may be facing at that moment. It still amazes me how you can pull out the circuitry map for ARTI and figure out exactly where the problems originated. I would also like to thank all the members of my supervisory committee, Dr. Peggy Olive, Dr. Gerry Krystal, and Dr. Michel Roberge. Their advice is thoroughly appreciated and they have helped guide me through the years with their suggestions and comments. They were very personable and I would be able to seek their opinions on my project or even on life itself. Under the Department of Experimental Medicine at UBC, I have been fortunate enough to be part of the program under the directorship spanning over two directors, Dr. Norman Wong and Dr. Vince Duronio. Both Dr. Wong and Dr. Duronio have given me sage advice from the early stages of my studies, to the encouragement in these final stages. My experience in ExMed would have not been the same without the help of our graduate secretary, Mr. Patrick Carew, who kept us informed of every upcoming award deadline as well as answering my menial questions. To Dr. Lloyd Skarsgard, thank you for giving me the opportunity to work with you on the anti-proton project and to learn first hand, your extraordinary forethought in experimental xi i i design and laboratory practices. Dr. Andrew Minchinton and Dr. Aly Karsan, senior principal investigators in our Medical Biophysics department, thank you for your continual helpful advice, especially during departmental meetings. In addition, thank you for giving me permission to use pieces of your equipment, from the cryostat to the phosphoimager. To Denise McDougal and Nancy LePard, thank you for your help with running all my samples on the FACS machine and for being so accommodating to my schedules, even when I was extremely demanding with getting more time on the machines. To Darrell Trendall, the first person I met when I started as a summer student, thank you for showing me how to do animal work and for keeping me updated on hockey related news. Sometimes it may seem like our lab shares too much information with one another, but your "coffee crisp" incident will be one story I will never forget. To Angie Nicol, thank you for not only teaching me all my tissue culture techniques, but also for being my confidant in personal matters of my life. You are truly one of the sweetest women I know, and thank you for allowing me to grow up with your boys and be their Auntie Michelle. To Dr. Arusha Oloumi, a former student in Dr. Peggy Olive's lab, thank you for teaching me molecular biology techniques and proving that you possess super powers by completing your degree while being a wonderful mother. To Susan McPhail, thank you for being the person I go to whenever have a question of any sort. You know an infinite number of techniques as well as the location of every piece of equipment in the department. In my dark thesis writing days, you even offered to read over my work (a job no one wants!) just because you genuinely wanted to help. To Wil Cottingham, our departmental administrative co-ordinator, thank you for your bottomless help with everything from awards and funding, to driving me home after our conference in Banff. xiv To Kevin Bennewith, my former grad student partner, thank you for all the laughs and tears. You are a truly great friend and I know you will be a wonderful father. Thank you for all the memories, our ice cream days, beach days, drinking days/nights, your all-girls stag in Seattle, our first Victoria conference, Frisbee and hearts tournaments, and of course, thank you for teaching me everything I know about hockey. To all the other members of the Medical Biophysics department at the BCCRC, you've helped keep me sane with our social outings, summer and Christmas parties, casual musings and those random Friday afternoons. To all the friends I have met as a grad student: Jenn, Jason, Arek, Nancy, Graeme, Chad, Sheela, Maisie, Steve, all the members of the GrasPods, and many more, thank you for making this experience not just bearable, but thoroughly enjoyable. I will cherish all the great times we have already shared and wish only the best for all of you. To all of my non-research friends, thank you for being so patient with my irregular time schedules and for constantly enduring any unintentional scientific jargon that seems to creep into my everyday speech. In particular, I'd like to thank Grace, my girlfriend since elementary school, who has always encouraged me and helped steer me towards the right path when things seemed uncertain. And to my family, you have been so wonderful to me. To my sister Connie, your never ending thirst for knowledge has always inspired me to be the best that I could be, thank you for always listening to me talk about everything and anything and even though you aren't physically close by, I know you are right behind me in support. Mom and Dad, thank you for everything you have sacrificed to give me all that I have today. You have always been so considerate of our careers, never wanting to interrupt, and only encouraged us to do whatever we desired to do. Connie and I owe all of our successes to the both of you. XV C H A P T E R 1: I N T R O D U C T I O N 1.1 TUMOUR MICROENVIRONMENT For normal cells to proliferate and differentiate, they rely heavily on the signals in the local microenvironment that may act as positive stimulators or inhibitory regulators. This situation stands true for the tumour microenvironment as well, where malignancies are found when host tissues help to induce, select and expand the tumourgenic cells (Loitta and Kohn, 2001). A few years back, the tumour microenvironment came more to the forefront of research focus primarily due to four advancements: more knowledge of molecular pathways, user friendly technological advances, the investigation into RNA profiles as well as DNA genetics, and the acceptance of cytokines, enzymes and the extracellular matrix (ECM) as integral parts for the molecular cross-talk of proteins (Witz, 2002). Tumourgenesis could be influenced by the microenvironment either by further increasing genetic instability, non-specifically inducing signalling cascades that were tightly regulated in normal cells, or by exerting selective external pressures which under normal situations, would have created the opposite effect on cells (Witz, 2002). Thus, malignant tumour cells could potentially recruit vasculature and stroma through production and secretion of stimulatory growth factors and cytokines. The locally activated host microenvironment in turn modifies the proliferative and invasive behaviour of tumour cells. Under those observations, it would appear that tumour cells would always be superior in growth to normal cells. However, tumour cell growth is unpredictable since they also have irregular vascular patterns as well as malformed tumour blood vessels. Thus, many regions of tumours have poor vascular perfusion and deprivations in nutrients and oxygen (Vaupel et al., 1989). This creates a population of tumour cells that are heterogeneous, whereas normal tissues are more structured and regulated in the cellular environment. 2 In normal tissues, the blood vessels are in some semblance of order where they are properly formed and functional, as well as evenly spaced for optimal diffusion of nutrients and oxygen to the surrounding cellular network. Blood vessels serve as the route of transport for both the nutrients and oxygen, thus the proximity of blood vessels and the blood flow rate to the tissues both contribute to the concentration available to the tumour vasculature. In tumour tissues, the malformed blood vessels and irregular vasculature create an unmet demand for both nutrients and oxygen in the newly formed tumour cells. Many areas within the tumour are found to be under-nourished and hypoxic which in turn, over long periods of time, create necrotic regions. Matters are further complicated when we try to define the degree of hypoxia and over which time frame of relevancy. It is easier to generalize the outcome of cellular environments in the extreme circumstance either when normalcy is maintained or if necrosis prevails. In any tumour system, these extreme situations are never exactly precise resulting in heterogeneous populations of cells. Most experimental work studying tumour vasculature had been done using animal systems, based primarily on ethics and convenience for technical manipulations. As live species, the observations model those anticipated for human patients, however, with any in vivo study there are intrinsic differences across species as well as internal complexities that could have an effect on the results. Superior in complexity to monolayers, spheroids (Section 1.2) are structurally similar to corded tumours in vivo due to the three-dimensional nature of the model, however, as an in vitro model, environmental factors can be controlled and manipulated (Sutherland et al., 1971; Sutherland and Durand, 1976). 3 1.2 SPHEROIDS: AN IN VITRO TUMOUR MODEL 1.2.1 Structure and Growth Spheroids are three-dimensional multicellular aggregates that form from single cell suspensions through proliferation and aggregation maintained through constant circulation within a spinner flask (Figure ^A). Depending on the generation time of the cell line, spheroids may grow to a diameter of 400-500 um within a span of 7 to 10 days for the V79-191b Chinese hamster lung fibroblasts, or approximately 20 days of incubation for the human lines, SiHa and WiDr, a cervical squamous cell carcinoma and a colon adenocarcinoma respectively. At this diameter of 400-500 um, we consider spheroids to be of optimal size where they are large enough to create a self-contained gradient from the microenvironment (Figure 16) yet not so overgrown as to create an enlarged necrotic centre encased in a tightly packed, thin layer of viable cells. Although three-dimensional culture systems such as multicellular spheroids are considered to mimic the in vivo environment of tumours, much of the biological complexity of the in vivo situation is lost. A major advantage of establishing three-dimensional, spherical aggregates from permanent cell lines is that basic mechanisms of cell growth, proliferation, and differentiation can be studied in a reproducible format with an internal environment dictated by the metabolism and adaptive responses of the cells. Thus, aggregates possess a three-dimensionally organized complex network displaying cell-to-cell and cell-to-matrix interactions in conjunction with a well defined morphological and physiological geometry affecting internal criteria such as cell shape, size, distribution, or enzymatic activity. There are two primary techniques used for cultivation of tumour spheroids as previously described in the literature (Mueller-Klieser, 1987; Lund-Johansen et al., 1992). Spinner flask cultures are most widely used by groups including our own mostly because large numbers of spheroids may be generated at the same time reaching 4 A Figure 1.1: Spheroids. (A) Scanning electron micrograph of a human cervical squamous cell (CaSki) spheroid. (S) Histological cross-section of a spheroid depicting a necrotic core encased in a large rim of viable cells. 5 diameters even as large as 1-2 mm. There are other approaches to avoid effects caused by stagnant medium, such as roller bottles or roller tubes, but in practice, these two are rarely utilized. However, for monitoring and manipulating single spheroids at particular growth stages, cultivation of aggregates on a stationary, nonadherent surface such as agar or agarose in 96-well plates is normally the method of choice. These agar or liquid overlay cultures are kept on a gyratory shaker following a defined initiation time to provide dispersed nutrient supply. With all methods, spherical aggregates of tumour cells form spontaneously in the absence of attachment to another substrate either directly in the flasks or during an initiation interval in agar-coated culture dishes. In addition to morphology, multicellular tumour spheroids also closely resemble in vivo solid tumours in their growth dynamics. Monolayer cultures grow exponentially if there aren't any spatial restrictions on Petri dishes, while nodular solid tumours are characterized by an early exponential phase followed by a period of reduced growth (Laird, 1964; McCredie et al., 1965). That first exponential phase is not dependent on the lack of availability of external factors while growth in the second phase does depend on the size of the tumour and the nutritional restrictions imposed on cells located in the interior of the tumour. Multicellular tumour spheroids are similarly characterized by early exponential growth followed by a period of slower growth (Sutherland, 1988). Mathematical models have been postulated to correlate the growth curves of solid tumour growth and those of multicellular spheroids. The growth phase can be separated into three distinct phases, the geometric, the linear and finally, the plateau phase (Conger and Ziskin; 1983). The geometric phase corresponds to the period of time when spheroids are beginning to aggregate and the proliferation of small spheroids, while the linear and plateau phases represent the stages when spheroids are developing the quiescent inner region and the gradual formation of the necrotic centre. Several groups have used the exponential-Gompertzian growth equation to test several tumour 6 cell lines grown as monolayers, multicellular spheroids, as well as solid tumours (Demicheli et al., 1989; Demicheli et al., 1991). Their results revealed that monolayer cultures do not adequately fit the model but both in vivo tumours and multicellular spheroids did fit the model very well, strongly suggesting that they resemble one another in their growth characteristics. The Gompertzian growth model can be represented mathematically as follows: V(f) = V0exp{ a/0 [1-exp (-#)]} where V(t) is the volume of the tumour at time t, VQ is the initial volume and a and B axe positive parameters (Chignola et al., 1999). The parameter a is the growth constant of the limiting exponential and is equal to the loge 2 divided by the doubling time. The doubling time of the Gompertz curve increases continuously and exponentially with time. The constant B determines the rate of increase, thus it signifies the slope of the curves (Steel, 1997). Analysis of tumour growth with the Gompertz equation provides not only a comparison for the outcome of anti-tumour therapies (Lloyd, 1975; Bassukas, 1993; Chignola et al., 1995), but also allows prediction of tumour progression in vivo even in the early stages of tumour growth and development (Norton et al., 1976). Understanding the growth of tumours may have profound implications for tumour therapy in experimental and clinical studies. 1.2.2 Physiology Spheroids grown from established tumour cell lines or, in some rare cases, directly from primary tumour biopsies, show growth patterns similar to those of tumours in vivo (Sutherland and Durand, 1976; Sutherland, 1986). As growth progresses, the percentage of cells that are proliferating decreases, while the proportion of quiescent 7 (non-proliferating) cells increases. Due to the three-dimensional nature of spheroid structure, cells are exposed to steep gradients of oxygen, glucose and other growth factors as well as the accumulation of toxic metabolic wastes, resulting in cell death and necrosis in the centre of the spheroids (Bauer et al., 1982). Beyond a critical size (approximately 500 um in diameter), most spheroids from permanent cell lines develop this necrotic core surrounded by the viable rim of cells consisting of both proliferating cells on the periphery and quiescent yet intact and viable cells between the core and the peripheral layer (Sutherland, 1988). The distance from the periphery of the spheroid to the onset of the necrotic centre may vary from 50 to 300 um, depending on a variety of factors including cell type, cell packing densities, nutrient consumption rates, and the concentration of these nutrients in the culture medium (Tannock and Kopelyan, 1986; Luk and Sutherland, 1987; Mueller-Klieser, 1987). However, the concentric arrangement of heterogeneous cell populations in spheroids as well as their growth pattern clearly mimics early stages of solid tumour growth in vivo, where there are tumour microregions with high proliferative activity close to the blood vessels, quiescent cells as intermediates, and necrotic areas at further distances. Scanning electron micrographs of the outer surface of a multicellular spheroid were generally found to be smooth in appearance where it was difficult to distinguish individual cells (Santini and Rainaldi, 1999). However for HT29 cells, Santini and Rainaldi found that even though the surface was smooth, the spheroids were covered by a layer of thick filaments which appeared to be closely associated with the surface itself. Of particular interest, they noted that there was high similarity in appearance between the multicellular spheroid and in vivo solid tumours from the same tissue of origin (Rosai, 1996; Santini and Rainaldi, 1999). 8 1.3 FLOURESCENCE ACTIVATED CELL SORTING AND FLOW CYTOMETRY OVERVIEW 1.3.1 Hoechst 33342 The blue fluorescent dye, Hoechst 33342 (Figure 1.2), belongs in the family of Hoechst dyes that differ slightly only in the wavelengths of their excitation and emission spectra. Hoechst dyes are nucleic acid stains that have multiple applications, including sensitive detection of DNA and cell number (Mocharla et al., 1987; Adams and Storrie, 1981), as well as uses for chromosome sorting (Lebo, 1982). The fluorescence of these dyes is very sensitive to both the DNA conformation and the state of chromatin. Consequently, Hoechst dyes are useful stains for the flow cytometric determination of DNA damage (Kubbies, 1990; Crissman et al., 1988) by monitoring the emission spectral shifts of the dyes (Ellwart and Dormer, 1990). Hoechst 33342 is a vital DNA stain that binds preferentially to A-T base pairs in the minor groove of DNA. The cells require no permeabilization for labelling, but do require physiologic conditions for optimal delivery to the cell nucleus. The binding of Hoechst 33342 to cellular DNA provides us with a non-biased fluorescence attachment to cells where those cells exposed to a higher concentration of Hoechst 33342 would stain more, versus those cells that have a lower concentration of the fluorescent dye. Thus, for use with the spheroid model, those cells on the periphery of the spheroid are first exposed to the dye, and as the Hoechst 33342 diffuses into the spheroid, the concentration of available Hoechst gradually reduces resulting in the inner cells staining with a much lower fluorescent intensity. With this tool, we are able to separate different subpopulations of cells from the model to determine differences between regions of the same tumour model system. The quantitative approach to do this involves using fluorescence activated cell sorting (Figure 1.3), which will be described in more detail in the next section. 9 Figure 1.2: Hoechst 33342 molecular structure. The ethoxy group (-OCH 2CH 3) is the position of the R group where different R groups results in Hoechst dyes at slightly different excitation and emission spectra, and with different capacities to cross the viable-cell membrane. 10 o Figure 1.3: Fluorescence activated cell sorting. Simple diagrammatic representation of FACS where fluorescently tagged single cells are separated out from mixed samples for pure populations of only one cell type. 11 1.3.2 Cell Sorting Techniques 1.3.2.1 Volume Sort Cells can be selected for sorting based on their fluorescence and cell size determined by light scatter. The saline stream, containing the cells, begins as a continuous stream as it is projected from the nozzle (Figure 1.3). The laser beam hits the stream just after it is projected from the nozzle and the cells interrogated at this point are collectively grouped as data referred to as the point of analysis. High frequency vibration of the nozzle causes the stream to eventually break up into droplets and allows deposition of selected droplets into test tubes. This is accomplished using an electrical charge that is applied to the stream and momentarily held until the droplet breaks off. If a cell fits the programmed sort criteria, that droplet retains the charge and is deflected by a positively and negatively charged pair of plates that pulls droplets into downstream test tube containers positioned on either side of the stream. Cells may be sorted into test tubes to collect large numbers of cells, or small, accurately counted numbers. Even individual cells may be directly deposited into microtitre plates. In the in vitro spheroid model, we utilize FACS to separate out cells with varying fluorescence within each spheroid creating multiple subpopulations. The number of subpopulations is determined by each individual scientist with a minimum of two fractions. Hoechst 33342 is the fluorescent dye that is generally used as described in Section 1.4.1. The fluorescent intensity increases as cells are closer to the periphery of the spheroid and this gradient of fluorescence is the basis for the sorting technique. Two methods of sorting spheroids have been used: the volume sort and the depth sort. Most groups sort spheroids based on volume, where there are equal numbers of cells in each fraction. Thus, with the spherical geometry of spheroids, to have equal numbers of cells in each layer (or subpopulation); each layer would have a different thickness (Figure 12 Figure 1.4: Volume and Depth Sorts. Each layer in the volume sort have equal cells per fraction shown diagrammatically in 3-dimension (A) and as a cross-section (S). Depth sorts have equal depth per fraction, in 3-dimension (C) and as a cross-section (D). 13 *\AA and 6). The advantage of the volume sort is the control of equal cell numbers in each sample and becomes especially important when performing survival assays where each sample should contain equal cell numbers prior to plating for accurate comparisons of colony growth. 1.3.2.1 Depth Sort Alternatively to the volume sort, we have chosen to use the depth sort method to separate out populations in the spheroids. To calculate the area of each spheroid, we measured spheroid diameters with a micrometer eyepiece ruler of a microscope. More automated methods are also available based on a microscope-image analyser that captures the area of each spheroid to estimate the entire spheroid volume. Figure 1.4C and D show two diagrammatic views of the depth sort where each subpopulation had equal depth into the spheroid. An analogy to this method can be described as the layers of an onion where each layer represents one fraction or subpopulation. The advantage of using the depth sort can be visualized when comparing Figure 1.4S and D, where both diagrams separate out the spheroid into six subpopulations, however, using the depth sort, we get more information from the inner layers of the spheroid since we are able to separate this area into 2-3 fractions as opposed to mostly one larger fraction from the volume sort. For our purposes in this thesis, all FACS sorting with spheroids was performed using the depth sort method. 1.3.3 Flow Cytometry Analysis Flow cytometry is the measurement of characteristics of single cells suspended in a flowing stream of saline. A focussed beam of laser light is positioned in line with the moving cells and light is scattered from the cell in both the forward and side direction. This information is picked up by detectors and stored for further analysis. The amount of forward scatter light is closely correlated with the size of the cells; whereas the amount 14 A 10 1 10 J' 10 3 10 \ LOG B44-FITC ~> i A O ~~I CM -g 1 • C O O ~= C O co -a i o ~a * ~H o o - | C5 o _J1 TE O __I r" •* i O _ i 7^ fljiirii;n|iiMjiiii|ii^ i>i|..ii|iai|iiiip 10 40 70 100 140 180 220 DMA Content -> Figure 1.5: Flow Cytometry. (A) Two colour flow cytometry analysis and (6) flow cytometry detection of cyclin B1 that identified G 2 phase cells. 15 of side scatter indicates nuclear shape and granularity. Cells that are labelled with a fluorescent marker, either using antibodies or a fluorescent dye such as Hoechst 33342, can be sorted depending on the intensity of this fluorescence. The laser beam excites the fluorochrome bound to the cells and as the fluorochrome molecule drops back from its excited state, fluorescent light is emitted which is then detected by sensors (Figure 1.5A). The quantity of fluorescent light emitted can be correlated with the expression of the cellular marker in question, such as cyclin B1 (Chapter 3) or cyclin D (Chapter 4). Figure 1.5 shows two flow cytometry histograms using either two fluorochromes {A) as markers for the two thymidine analogues, or one fluorochrome to label the expression of cyclin B1 (B). The use of FACS is a powerful tool where specific areas can be separated as well as analyzed for other cellular markers to determine regional differentiation in molecular levels. 16 1.4 TUMOUR HYPOXIA 1.4.1 Classification of Chronic versus Transient Hypoxia Blood perfusion in tumour vasculature is often markedly inferior to that of most normal tissues, leading to regions of hypoxia or poorly oxygenated viable cells (Figure 1.6). Cellular respiration depletes oxygen as it diffuses from blood vessels through packed layers of cells, with chronic hypoxia occurring at distances of 150-200 um from capillaries, where a majority of the oxygen has already been utilised (Thomlinson and Gray, 1955; Awwad et al., 1986). Areas of chronic hypoxia are continuously depleted of oxygen due to their distance from the blood vessels, where the limiting factor is the lack of oxygen availability. More recently, emerging evidence had been obtained for transient hypoxia in tumours, where cells become acutely hypoxic for short periods of time, followed by reoxygenation as the region regains oxygen from its surrounding blood vessels. Many groups have investigated the cause of transient hypoxia, leading to several explanations, all of which we believe to be a factor in the process. For instance, transient hypoxia may be a result of poor oxygen perfusion in tumour cell vasculature, limiting the distance of oxygen diffusion (Brown, 1979); or it may also result from unstable pressure of tumour blood flow, which would ultimately result in fluctuations in oxygen delivery to tumour cells (Dewhirst et al., 1996; Braun et al., 1999). By definition, the main difference between chronic and transient hypoxia depends on the duration of hypoxia exposure; the time frame is continuous and indefinite in chronic regions, but would be sporadic and temporary without any definitive or constant pattern in transient regions of hypoxic tumour cells. 1.4.2 Methods of Hypoxia Detection A simple clinical test for tumour hypoxia offers the promise of optimisation of treatment for individual patients based on the oxygenation status of their tumours and would 17 Figure 1.6: Chronic diffusion-limited hypoxia. Necrosis begins to occur at a radial distance of approximately 150 um from the blood vessels where a cross-section of a tumour cord is shown spanning out from the tumour blood vessel. As the cells get increasingly farther away (darker cells), oxygen is consumed limiting the amount of oxygen reaching these cells. 18 enable oncologists to analyze clinical trials using protocols that target hypoxic tumour cells. Many methods have been tested for measuring oxygen in tumours and tissues including direct, indirect and spectroscopic measurements, use of chemical probes, and intrinsic markers. 1.4.2.1 Direct Methods Direct measurements of oxygen have been made in human tumours and clearly show that many tumours are more hypoxic than their surrounding normal tissues. The "gold standard" for this method involves the use of the oxygen electrode using a probe with a diameter of 300 um. This probe is still large compared to the diameter of a tumour cell and is comparable to the size of a tumour cord. Thus an oxygen electrode may be simultaneously sampling a volume containing both well oxygenated and hypoxic cells. Practical use is still mainly confined to superficial lesions, or to lesions that can be exposed by surgery. Gatenby showed that measurements made with the oxygen electrode could predict the response of lymph node metastases to radiotherapy, with significant differences in the pre-treatment oxygenation status of responding and non-responding tumours (Gatenby et al., 1988). Oxygenation did not correlate with tumour stage, histological grade or tumour size and therefore was unable to be predicted by these parameters. In addition, oxygen electrode studies have shown benefits to the oxygenation status of tumours when patients breathe carbogen (Falk eta/., 1991). 1.4.2.2 Indirect methods As poor vascularisation of tumours is thought to imply poor delivery of oxygen leading to the presence of chronically hypoxic cells, measurements of the density of tumour vasculature could be indicative of the amount of hypoxia in tumours (Mueller-Klieser et al., 1987). Such measurements can be undertaken wherever a biopsy sample can be obtained from a human tumour, to be sectioned for histological analysis. Several studies with cervical carcinoma show a correlation of vascular density with local tumour control 19 and survival after radiotherapy (Siracka et al., 1988; Revesz et al., 1989). Although vascular density does not provide direct evidence of tumour oxygenation status and omits the possibility of transiently hypoxic cells, the predictive value of these measurements for survival after radiotherapy suggests that chronically hypoxic cells are an important component of the total hypoxic fraction that leads to radioresistance. 1.4.2.3 Spectroscopic methods Oxygen related changes in the spectroscopic properties of pyridine nucleotides or in those of cytochromes and other proteins have been extensively investigated as indicators of the oxygenation status of tissues in vivo (Chance et al., 1988). Near infrared spectroscopic measurements of haemoglobin deoxygenation can provide non-invasive information on the oxygenation of tissue but the reporter molecule is localized in the vasculature and would not be informative about the status of the cells which are hypoxic because of their distance from blood vessels. These redox-related changes are reversible by reoxygenation on exposure to air and are therefore not generally applicable to biopsy material (Nioka and Chance, 2005). This application in vivo is limited by the ability of light of the appropriate wavelength to penetrate tissues but this problem may be resolved by use of fast time-resolved spectroscopy (Torricelli et al., 2004; Swartling et al., 2005). An alternative to spectroscopic methods of measuring oxygen involves its quenching of long-lived excited states of molecules which are responsible for fluorescence. Quenching of delayed fluorescence from cells stained with acridine orange has the potential to measure the oxygen tensions in the immediate vicinity of DNA (Houba-Herin era/., 1984; Radu era/., 2004) 1.4.2.4 Bioreductive chemical probes The bioreductive metabolism of nitroaromatic compounds such as misonidazole involves the generation of an initial free radical which is so reactive towards oxygen that further 20 metabolism is effectively inhibited in oxygenated cells. In hypoxia, reactive metabolites are generated that bind to cellular macromolecules, thus labelling hypoxic cells even if they are subsequently reoxygenated (Urtasun et al. 1986; Cline et al., 1990). Although many types of nitroaromatic compounds have shown promise as probes for hypoxic cells in in vitro test systems, few have worked successfully in vivo, and almost all of those tested have been 2-nitroimidazoles (Franko and Chapman, 1982; Urtasun era/.,1986). Numerous isotopically-labelled 2-nitroimidazoles have been proposed as hypoxia markers. 3H-misonidazole has been administered to small numbers of patients with treatment-resistant metastatic melanomas, sarcomas, small cell lung cancers and squamous cell carcinomas, and localized regions of high labelling were observed in sections cut from biopsies of some of their tumours, especially those from melanomas and small cell lung cancers (Urtasun et al., 1986; Casciari et al., 1995; Jensen et al., 2000). However, the rather large amounts of radioactivity required and subsequent prolonged autoradiography of tumour sections make this technique very laborious for routine clinical use. 2-nitroimidazoles with immunohistochemically detectable side-chains have been investigated as potential hypoxia markers and using such compounds, hypoxia-specific staining in experimental tumours has been demonstrated (Hodgkiss et al., 1997; Webster era/., 1998; Rijken era/., 2000). 1.4.2.5 Intrinsic Markers An alternative and potentially superior strategy for the measurement of hypoxia is to utilize changes in expression of oxygen regulated proteins, or their mRNAs as markers for hypoxic cells. For example, this can be demonstrated by the induction of the erythropoietin gene expression by either anaemia or by the hypoxia caused by breathing rarefied air at high altitudes (Winter et al., 2005). Erythropoietin mRNA levels increase several hundred-fold in rodent liver and kidney cells, but the tissue specific nature of the erythropoietin expression probably precludes its use as a general marker of tumour 21 hypoxia (Gopfert et al., 1997; Kochling er al., 1998). Other examples of oxygen regulated proteins would be glycolytic and respiratory enzymes, stress proteins, and the vascular endothelial growth factor (VEGF). 1.4.3 Hypoxia in Spheroids With all methods of detecting hypoxia as described in the previous section, different timescales are involved with transient intermittent hypoxia and chronic diffusion-limited hypoxia and ideally, markers will be identified that are informative on the timescale of both types of hypoxia (Chaplin et al., 1986). Defining which type of hypoxia is found in a sample is complicated and often difficult in in vivo systems, because the most likely situation would be that both types are present at one point in time, either together or separately. However, with the in vitro spheroid model, hypoxic gradients are created via limitations of the available oxygen from the nutrient media, thus this would resemble the characteristics of chronic diffusion-limited hypoxia in solid tumours (Raleigh et al., 1987). With this aspect controlled, it is much easier to manipulate microenvironmental factors and to examine any hypoxia driven effects. 22 1.5 CELL CYCLE 1.5.1 Phases and checkpoints Cells in the process of proliferation may be categorized into distinct phases. When mitotic cells are dividing, they differ in many ways from cells in interphase. For example, cells undergoing the process of division are observed to be undergoing morphologically and biochemically defined periods of cell division: prophase, metaphase, anaphase and telophase. Likewise, as shown in Figure 1.7, interphase can be further sub-divided into three stages, namely, a Gi or pre-DNA synthesis period, an S or DNA synthesis period, and a G 2 or post-DNA synthesis period, prior to the M phase or mitosis. Cells undergoing DNA synthesis can be distinguished from other cells by nuclear labelling with DNA precursors, such as tritiated thymidine or thymidine analogues. It has been demonstrated that thymidine analogues, such as bromodeoxyuridine and iododeoxyuridine, are incorporated into a discrete population of cells (Laird, 1964; Sutherland eta/., 1971) which indicates that DNA synthesis is not continuously occurring, but rather is confined to a specific phase of the cell cycle, the synthesis phase (S-phase). The gaps between the S phase and mitosis, are called the Gi and G 2 phases and the duration of these varies depend on the cell type. However, most cells in normal tissues of adults are in a quiescent of G 0 state. Cells are receptive to stimuli to initiate proliferation in both the G 0 and the Gi phases of the cell cycle. The "restriction point", or R point, has been defined in cultured cells as the point in phase after which a cell is committed to enter DNA synthesis, as diagrammed in Figure 1.7. 1.5.2 Cyclins and cyclin dependent kinases The cell cycle is governed by a family of cyclin-dependent kinases (Cdks) whose activity is regulated by the binding of positive effectors, the cyclins (Sherr, 1994; Morgan, 1995), and by both activating and inactivating phosphorylation events. Cyclin binding to its 23 Source: Modified from Tannock and Hill, 1998 Figure 1.7: Diagrammatic view of the cell cycle. The restriction point (R point) represents a point in time in the phase beyond which cells are committed to proceed into the S phase. More simplistically, cells can be categorized either in interphase or undergoing mitosis. Interphase lasts much longer in duration and it is composed of the phase, the S phase, as well as the G 2 phase of the cell cycle. 24 C y c l i n binding Source: Adapted from Tannock and Hill, 1998 Figure 1.8: Schematic diagram of cyclin-Cdk interaction. 25 Cdk is required for kinase activation, but two different families of Cdk inhibitory molecules provide an additional level of control. The cyclin-dependent kinases regulate a series of biochemical pathways, or checkpoints, that integrate mitogenic and growth-inhibitory signals, monitor chromosome integrity, and co-ordinate the orderly sequence of cell cycle transitions (Hartwell, 1992). In mammalian cells, the family of Cdks, designated Cdk1 to 7, are conserved in size, ranging from 32 to 40 kDa, and share sequence homology. They are small serine/threonine kinases that are expressed at constant levels throughout the cell cycle and are catalytically inactive unless they are bound to cyclins. Phosphorylation of a conserved threonine, located in the catalytic cleft of the kinase (such as Thr161 for Cdk1 and Thr160 for Cdk2) is required for full activation, and this is catalyzed by the Cdk-activating kinase, or CAK (Solomon, 1993; Morgan, 1995). Cdk activation can be inhibited by phosphorylation of conserved inhibitory sites at Thr14 and Tyr15 by wee-1 kinases. Full Cdk activation requires dephosphorylation of these inhibitory sites by phosphatases of the cdc25 family as diagrammed in Figure 1.8. In general, cyclin levels oscillate during the cell cycle and cyclin mRNA and protein expression peak at the time of maximum kinase activation, contributing to discrete bursts of kinase activity at specific cell cycle transitions. Figure 1.9 illustrates this oscillating pattern of the various cyclins within the cell cycle (Sherr, 1996). The family of mammalian cyclins includes cyclins A to H, which all share a conserved sequence of about 100 amino acids, whose mutation disrupts both kinase binding and activation. Passage through G1 into S phase is regulated by the activities of cyclin D, E and cyclin A-associated kinases as shown in Figure 1.10 (Sherr, 1994; Reed, 1997). D-type cyclins associate with Cdks 2, 4, 5, and 6 and play an important role in re-entry of cells into the proliferative cycle from quiescence. A primary role for cyclin D-associated kinases appears to be the phosphorylation of the retinoblastoma protein, pRb. 26 Figure 1.9: Oscillations in various cyclin levels during the cell cycle. p27, a member of the Cip/Kip family of Cdk inhibitors, act to oppose the actions of Cdk2. They have been shown to have high levels in quiescent cells, however as cells enter the cell cycle and accumulate cyclin D dependent kinases, the Cip/Kip proteins are sequestered in complexes with cyclin D dependent Cdks resulting in the observed decrease in p27 as shown above. 27 Source: Adapted from Tannock and Hill, 1998 Figure 1.10: Progression through the cell cycle. This progression is governed by a series of cyclin dependent kinases whose activities are positively regulated by cyclins and negatively regulated by Cdk inhibitors. 28 Phosphorylation of pRb in G1 phase is required to allow progression from G1 to S phase. Cyclin E is associated with Cdk2 and, in most cells, cyclin E/Cdk2 activity peaks after the peak of cyclin D/Cdk2 activation. Overexpression of both cyclin D and E can accelerate the transition from G1 to S phase (Pardo et al., 1996). Mitotic or B-type cyclins, when associated with Cdk1, can control entry into or exit from mitosis. B-type cyclin levels increase during S phase. Phosphorylation of the inhibitory Thr14 and Tyr15 sites (see Figure 1.8) on Cdk1 keeps the kinase inactive until the G2/M transition. Dephosphorylation at these sites by cdc25 phosphatase, and CAK activation, triggers Cdk1 activation, which is essential for mitosis to occur (Lenormand et al., 1999). 1.5.3 Cyclin dependent kinase inhibitors Recent work has identified two families of Cdk inhibitory proteins (Sherr, 1995). The kinase inhibitory proteins (Kip) family members include p21, p27, and p57. p21 was identified both as a novel protein bound to Cdk2 and as the product of a gene whose transcription is activated by the negative growth regulatory p53. All Kip family members can bind to and inhibit a wide range of Cdk's, including Cdk1 through 6. Kip molecules bind more efficiently to cyclin/Cdk complexes than to either cyclin or Cdk alone (Hall et al., 1995). The number of Kip molecules available to bind Cdk proteins regulates kinase inhibition. When the ratio of Kip:Cdk molecules are low, the Kips may facilitate Cdk/cyclin assembly and kinase activation (LaBaer et al., 1997). However, at a certain critical point, an excess of Kip is reached and/or there are conformational changes so that the effect of the Kip binding becomes inhibitory. The second group of Cdk inhibitors is the INK4 (inhibitor of Cdk4) family (Reed et al., 1994; Sherr, 1995). They share a highly conserved motif that plays a role 29 cyclin D/cdk4-6 cyclin E/cdk2 inactive kinase active kinase inactive kinase inactive kinase active kinase inactive kinase Source: Adapted from Tannock and Hill, 1998 Figure 1.11: Two families of cyclin dependent kinase inhibitors. A schematic diagram showing mechanisms of inhibition of Cdks with cyclin D or cyclin E when bound by the Kip and INK4 families of Cdk inhibitors. 30 in protein/protein interactions. The INK4 proteins can inhibit specifically the cyclin D-dependent kinases, Cdk4 and Cdk6. The members of the INK4 family include p15, p16, p18, and p19. The INK4 proteins can displace cyclin D1 from Cdk4 complexes (Figure 1.11). It has been proposed that Cdk inhibitors, p15 and p27, co-operate to induce G1 arrest following TGF-p treatment (Reynisdottir et al., 1995). Upregulation of the p15 protein and increased binding of p15 to its targets, Cdk4 and Cdk6, occurs concomitantly with displacement of cyclin D1 and p27 from Cdk4 and an increase or stabilization of the association of p27 with cyclin E/Cdk2 complexes (Reynisdottir er al., 1995; Sandhu et al., 1997). The Cdk inhibitors also act to arrest the cell cycle when cells undergo DNA damage. It has been demonstrated that there is an induction of the Kip protein p21, mediated by p53, in the cellular response to irradiation (Petrocelli et al., 1996). The p53 protein levels increase dramatically following DNA damage and activate transcription of p21. The rise in p21 results in the increased binding of this inhibitor to target Cdks and aid in kinase inhibition. Thus, mammalian cells appear to have evolved a mechanism to coordinate DNA repair with cell cycle arrest in response to DNA damage. 31 1.6 RADIOSENSITIVITY 1.6.1 Hypoxia and radiation therapy The response of tumours to radiotherapy is heterogeneous, even with one tumour type or stage, and is dependent on many factors. These may include the intrinsic radiosensitivity and rate of proliferation of the tumour cells and the extent to which these cells are hypoxic (Bush et al., 1978). Hypoxic cells are known to be approximately three times more radioresistant than well oxygenated cells and their presence in tumours has been suggested to contribute to failures of local control in radiotherapy (Gray et al., 1953; Gatenby et al., 1988; Olive et al., 1997). Ionizing radiation may interfere with cell cycle processes through two important mechanisms: either by changing the pattern of cell proliferation or by killing the cells. Since the pattern of cell proliferation depends, in part, on the characteristics of the cell cycle, radiation effects involving the compartments of a proliferative cell system provide information on underlying mechanisms of radiation injury at the cellular level. More importantly, the generation and death of cells subsequent to these processes characterize the cell population structure (Dahlberg, 1999). The response of a proliferating cell population to ionizing radiation depends to a large degree on the distribution of the cells as a function of their position in the cell cycle. The radiation responses in mammalian cells, such as modification of DNA synthesis, mitotic and division delay and the production of chromosome aberrations affecting the proliferative capacity, depend in part on the cell cycle phase at the time of irradiation. Radiobiology has provided the methods for studying the mechanisms by which precise doses of ionizing radiation will interact with the cells of the tumour. Ideally, the application of ionizing radiation under clinical conditions in the treatment of malignant disease will result in eradication of the tumour cells and cure the patient. Clinical neoplasms that respond to radiotherapy by a rapid reduction in size and disappearance 32 within days or weeks are considered to be radiosensitive. Different tumour cell types or different growth conditions can render a population of cells as either more radioresistant or more radiosensitive. For example, a population of tumour cells growing in hypoxic conditions tends to be more radioresistant, whereas, cells subjected to low glucose conditions are more radiosensitive. Maebayashi et al. have done some studies that compare the radiosensitivity of a rat yolk sac tumour cell line with a normal or mutated p53 gene. They observed that a loss of p53 function by radiation-induced mutation of p53, decreased the radiosensitivity in this cell line. This observation was also seen in the human pancreatic adenocarcinoma cell lines, where cell lines expressing a mutant p53 gene were more radioresistant (Ng ef al., 1999). 1.6.2 Cyclin Alterations Ionizing radiation has been shown to induce perturbations in the cell cycle of eukaryotic cells. After irradiation, cells undergo a division delay which is reflected by increased time spent in the G 2 portion of the cell cycle (Elkind et al., 1963; Bedford et al., 1977). It is thought that this G2-phase delay contributes to the ability of the cells to survive irradiation. As mentioned in earlier sections, different cyclins are required for the progression of cells through the cell cycle. Cyclin D is required for the passage of cells through the GT phase and into the start of the S phase, while cyclin E aids cells through the transition between the G r S phases (Figure 1.10). The role of cyclin A involves regulating the movement of cells through the S phase and both cyclins A and B are involved in the transition from S-G 2 phase and in the advancement through the G 2 phase (Maity era/., 1994). Overexpression of cyclin D1, a Gi cell cycle regulator, is often found in many different tumour types, such as breast carcinoma and squamous cell carcinoma of the head and 33 neck. Coco Martin et al. (1999) studied the effect of cyclin D1 on radiosensitivity in a breast tumour cell line, MCF7, which contained a transfected cyclin D1 gene construct. This group demonstrated that MCF7 cells overexpressing cyclin D1 were more sensitive to ionizing radiation than the non-overexpressing counterparts. Expression of cyclin D1 in wild-type cells became slightly decreased after ionizing radiation, but this was not shown in the cyclin D1 overexpressing cells. Expression of cyclin E did not vary after radiation; however, cyclin A did show some decreased levels 24-36 hours post-irradiation. Irradiation with a dose of 5Gy resulted in a rapid increase of p53 and p21 in the cyclin D1-overexpressing cells. Total levels of the cyclin-dependent kinase inhibitor p21 markedly increased upon radiation, with a more pronounced effect in the cyclin D1-overexpressing cells. As a result, levels of p21 increased after irradiation, and this inhibitor associated with not only Cdk4 but also with Cdk2 and cyclin B1 in the cyclin D1-overexpressing cells (Coco Martin et al., 1999). After ionizing radiation, the overexpression of cyclin D1 reduced cell survival, and it also enhanced apoptosis. The sustained higher levels of p53 and p21 most likely cause the enhanced apoptosis after irradiation. After radiation, the higher levels of p53 subsequently induce expression of p21. These proteins were shown to have increased levels even in control MCF7 cells (Hain et al., 1996) but both protein levels remained higher after radiation in cyclin D1-overexpressing cells. p53 is an upstream regulator of p21, which is a cyclin-dependent kinase inhibitor. Increased levels of p53 and p21 can initiate a G2-M block, facilitate repair of DNA breaks and regulate exit from the G 2 checkpoint (Schwartz et al., 1997; Dulic et al., 1998). Increased activity of p21 leads also to the hypophosphorylation of pRb. Dephosphorylation of pRb prevents E2F transcriptional activity in cyclin D1-overexpressing cells, leading to a d - S block. Thus, ionizing radiation induces a transient G r S block in MCF7 cells and a more pronounced G2-M block (Nagasawa et al., 1998). The elevated cyclin D1 level may also be 34 responsible for the slightly faster release from the radiation induced G^S arrest in cyclin D1 overexpressing cells as compared with control cells. Many different groups have come to similar conclusions about the hypothesized mechanism of the effects of radiation on cyclin development and progression. However, there has been reported a difference of findings in regards to the levels of cyclin A production after irradiation. Although Coco Martin eta/. (1999) may have concluded that they saw decreased levels of cyclin A after ionizing radiation, Bernhard et al. (1994) observed the opposite findings in regards to cyclin A levels. This group found that cyclin A mRNA and protein levels rise at the expected time and continue to rise to levels greater than those found in the control cells. This contrasts with cyclin B mRNA and protein levels, which remain depressed after irradiation in the S phase. After the cells begin to exit from the block and enter the phase, the cyclin A mRNA and protein levels fall. These findings are consistent with the hypothesis formulated by Luca et al. that cyclin B is responsible for triggering the destruction of cyclin A (Luca et al., 1991). As well, other groups have found similar results in other human cells lines but have also found that at higher doses, 20Gy, cyclin A mRNA levels may also be depressed. When cells undergo DNA damage, this may induce the cells to undergo a Gi cell cycle arrest. The reduced levels of cyclin D may initiate this arrest, where cyclin D normally aids cells in the progression from Gi to S phases. The observation that many different tumour types often overexpress cyclin D1 may be correlated with the inability to induce GT arrest. The levels of cyclin B were reported to be decreased after ionizing radiation, and similar to cyclin D, this reduced level may be involved in cell cycle arrest; this time G 2 arrest. Currently, there are differing viewpoints in regard to the levels of cyclin A after ionizing radiation. Cyclin A is involved with the progression of cells through the S phase as well as the G 2 phase (along with cyclin B). Although both increased and decreased levels of 35 cyclin A were observed by two independent research groups, both findings are ambiguous. The group claiming increased levels of cyclin A actually state that after an initial increase, the levels of cyclin A decrease after the cells exit from the "block" (Bernhard et al., 1994). The actual time frame is not apparent in the reported findings and thus, this observation may actually be similar to those claiming lowered levels of cyclin A. They state that decreased levels were observed "24-36 hours after radiation" (Coco Martin et al., 1999) and this may be the time frame noted for cells to exit from the aforementioned "block". 36 1.7 RAPAMYCIN OVERVIEW: A CYTOSTATIC AGENT The cellular mechanisms affected by rapamycin have been studied in detail over the past few years (Hidalgo and Rowinsky, 2000; Blume-Jensen and Hunter, 2001). Rapamycin (Figure 1.12) binds with a 12 kDa FK506 binding protein (FKBP12) and inhibits the aptly named mammalian target of rapamycin, mTOR (also referred to as FRAP in humans, or RAFT-1 in rats). mTOR normally responds to growth factor / growth factor receptor interactions through the phosphatidylinositol 2-kinase / protein kinase B (PI3K / Akt) pathway (Downward, 1998; Scott er a/., 1998; Nave et al., 1999). mTOR is a member of the PI3K-related kinase family (PIKK) that is involved in many critical regulatory cell functions including cell cycle progression (Sarkaria et al., 1998). mTOR has serine/threonine kinase activity (Schmelzle and Hall, 2000) and aids in the mitogen activated induction of p70 S6 kinase (p70S6K). p70 S 6 K phosphorylates the 40S ribosomal protein subunit S6, thereby enhancing the translation of ribosomal proteins and other proteins involved in cell cycle progression. mTOR also phosphorylates the eukaryotic initiation factor 4E binding protein 1, 4E-BP1 (also referred to as PHAS-1). Unphosphorylated 4E-BP1, as occurs in quiescent cells inhibits the initiation of translation by binding to elF-4E and preventing its association with the elF-4F complex. The elF-4F complex normally increases the translation of mRNAs with regulatory elements in the 5'-untranslated region such as cyclin D1 (Sonenberg and Gingras, 1998). Thus inhibition of mTOR with rapamycin inhibits the action of both p70 S 6 K and 4E-BP1 to reduce the translation of proteins essential for cell cycle progression. Additional mechanisms have also been elucidated which may contribute to the ability of rapamycin to limit progression from G: to S phase. Rapamycin blocks the elimination of the cyclin dependent kinase (Cdk) inhibitor p27K,p1 and facilitates the formation of Cdk/cyclin and p27K i p 1 complexes (Nourse et al., 1994; Luo et al., 1996). The p27Kip1 thus sequesters Cdk/cyclin complexes that are necessary for cell cycle progression. In 37 Figure 1.12 Molecular structure of rapamycin (C51H79N013). 38 addition, rapamycin has been shown to upregulate p27 K i p l mRNA and protein in exponentially growing cells, although there is limited correlation to rapamycin sensitivity in some exponentially growing cell lines (Kawamata et al., 1998). Rapamycin also increases the turnover of cyclin D1 at the mRNA and protein level (Hashemolhosseini et al., 1998) which, when combined with its effect on the elF-4F complex, results in a deficiency of cyclin D1 in the Cdk4/cyclin D1 complexes. Cdk4/cyclin D1 complexes are necessary for phosphorylation of the retinoblastoma protein, pRb (Morice et al., 1993; Nourse et al. 1994), which is a well-known requirement for progression of cells from G1 to S phase. With this initial understanding of the mechanism of action of rapamycin, we are interested in the use of this agent to inhibit cell cycle progression of human tumour cells grown as multicellular spheroids. The hypothesis is that spheroids once adequately characterized will provide an invaluable means to assess the potential for using cytostatic agents to minimize accelerated repopulation and target radioresistant cells during radiation therapy of multicellular systems. 39 1.8 RESEARCH OBJECTIVES The cell cycle of tumour cells is a multi-faceted process with many key players, but in particular, tumour hypoxia has a rapid impact on cell cycle regulatory proteins, as well as changing the effectiveness of therapeutic agents. Radioresistance of hypoxic cells is an important limiting factor in tumour response to radiotherapy but can be targeted by the administration of cell cycle inhibitors such as rapamycin to manipulate intracellular targets that control cell cycle progression. 1.8.1 Specific aims I. To determine the differences in the cell cycle kinetics of various subpopulations (or regions) within spheroids grown from rodent and human cell lines. II. To identify hypoxia-induced differences in protein expression of the cyclin B1 and cyclin D cell cycle regulatory proteins when utilizing tumour cells grown as in vitro multicellular spheroids. III. To use a rapid and convenient index of proliferation by measuring the effects on cyclin B and cyclin D levels during short exposures of hypoxia. IV. To assess the utility of using cell cycle inhibiting agents such as rapamycin to control tumour cell cycle progression in spheroids. V. To determine the potential for combining a cytostatic agent with radiotherapy in order to target radioresistant hypoxic cells and minimize tumour regrowth during therapy. 40 1.8.2 Thesis overview The following chapters in this thesis will address each of the Specific Aims listed above. Chapter 2 will focus on the cell cycle kinetics of spheroids and the methods used to determine these values. Two thymidine analogues were used as tools to identify and quantify the proportion of cells that were in the S-phase of the cell cycle, in addition to following these cells as they continue to progress through the cycle. The main objective was to determine any differences in the cell cycle kinetics when comparing adjacent regions within the same model system, since theoretically, as a controlled steady state in vitro model, many complications found in in vitro animal models could be avoided. Chapters 3 and 4 will answer the questions posed in Specific Aims 2 and 3. Spheroids were subjected to a range of oxygenation states and the effect of this modification was determined by measuring the levels of two cell cycle regulators, namely cyclin B and cyclin D. The exposure time to hypoxia was quite short; however, this short exposure was sufficient to identify changes in the cyclin levels and thus, this procedure could be used as an easy and quick index for proliferation. The potential clinical applicability and relevance will also be discussed. Chapter 5 will introduce the use of the cytostatic agent, rapamycin, with our multicellular spheroid model and will determine the effects of this cell cycle agent as an inhibitor for tumour cell progression. Preliminary ground work to establish the toxicity and dose response of this drug with our model was first initiated prior to more detailed studies on the effects of this agent targeting against our cell cycle regulators of interest, cyclin B and cyclin D. Chapter 6 will investigate the therapeutic advantages of using rapamycin against human xenograft tumours in in vivo animal models and determine the potential for combining a cytostatic agent with a radiotherapy regimen in order to target radioresistant hypoxic cells and minimize tumour regrowth during therapy. Chapter 7 will include a thesis 41 summary and will present suggestions for further work, with particular reference to the potential clinical application of the data, methods, and concepts discussed in this thesis. 42 C H A P T E R 2 : C E L L C Y C L E K I N E T I C S T U D I E S O F S P E C I F I C R E G I O N S IN A S P H E R O I D 43 2.1 INTRODUCTION The most common and general definition for cancer can be stated as the uncontrolled growth of cells. There are many factors that could lead to this change in the kinetics of cellular growth; for example, these factors may include nanoscopic factors like point mutations to the more macroscopic change of gene expression and chromosome abnormalities (Hanahan and Weinberg, 2000; Blume-Jensen and Hunter, 2001). For most of our studies, we utilize the spheroid system as a model for tumour growth. Not only is this system easier to use primarily due to the in vitro nature of the system, but it also allows users to have a population of tumourgenic cells which were all derived from identical single cells. With a cellular model that should have similar characteristics throughout the entire system, we discovered that there were indeed differences dependent on the region of the cell within the spheroid, and in addition, the cell kinetics of each subregion also proved to have differing values. To measure the kinetics of the cells in spheroids, we used an indirect procedure that immunofluorescently dual-stains cells for flow cytometric measurements of two thymidine analogues, iododeoxyuridine and bromodeoxyuridine (Begg et a/., 1985; White et al., 1994). The presence of IdUrd and BrdUrd (Figure 2.1), and the amount of DNA were measured by three-colour flow cytometry making it possible to define different subpopulations within the S phase and also enabling the measurement of the progression of the cells through the cell cycle during the time following labelling (Bakker era/., 1991; Durand, 1997). 44 2.2 MATERIALS AND METHODS 2.2.1 Spheroids Chinese hamster V79-171b lung fibroblasts were maintained in exponential monolayer growth on 100 x 20mm plastic Petri dishes with subcultivation twice weekly in Eagle's MEM containing 10% fetal bovine serum (FBS; Hyclone Laboratories). The V79-171b clone of the V79 lung fibroblasts were originally obtained by our laboratories from Dr. Robert Sutherland. V79 cells have a high plating efficiency with a doubling time of 12 to 14 hours. There are mutations in the coding region of p53, therefore V79 cells do not have a functional p53 protein (Chaung et al., 1997). SiHa human cervical carcinoma cells, and WiDr human colon carcinoma cell lines were obtained from American Type Culture Collection and maintained in Eagle's MEM containing 10% FBS (Kennedy et al., 1997). SiHa cells were originally established from fragments of a primary tissue sample obtained from a Japanese patient. This cell line is reported to contain an integrated human papillomavirus type 16 genome (HPV-16) with one to two copies per cell. The presence of the E6 and E7 oncoproteins results in the inhibition of the normal regulatory functions of pRB and p53 proteins. Although WiDr cells were originally deposited with the ATCC as a colon adenocarcinoma line established from a 78 year old female, DNA fingerprinting has shown WiDr cells to be a derivative of the HT-29 line. Tumourgenic in nude mice, WiDr cells express p53 antigen with a G to A mutation changing the amino acid from Arg to His at position 273 in addition to expressing epidermal growth factor (EGF). To initiate spheroid growth, asynchronous cells were removed from the plastic Petri dishes by trypsinization using 0.1% trypsin (20 ml of 2.5% trypsin [Gibco BRL, Burlington, ON] in solution with 500 ml of sodium citrate buffer). Bellco glass spinner culture vessels (Vineland, NJ) were then inoculated with approximately 4 x 106 cells in 200 ml of medium containing 10% serum and stirred at 180 revolutions per minute; the gas phase included 5% C0 2 in air. During the growth of the spheroids, the medium was 45 A Figure 2.1 Molecular structures of BrdUrd and IdUrd. (A) 5-Bromo-2-deoxyuridine (B) 5-lodo-2-deoxyuridine 46 first changed on day 3 and then daily thereafter, and spheroids were used for experiments when they reached a size of 450-500 um in diameter (Sutherland et al., 1971; Sutherland and Durand, 1976). Experiments on the V79 spheroids commenced approximately 7 to 10 days after they were inoculated in the spinner culture flasks, and the SiHa and WiDr spheroids were used around 16 to 20 days after inoculation. 2.2.2 Reagents For the cell cycle time kinetic studies, we utilized two thymidine analogues, namely 5-lodo-2-deoxyuridine (IdUrd, Sigma, St. Louis, MO) and 5-Bromo-2-deoxyuridine (BrdUrd, Sigma, St. Louis, MO). IdUrd has a molecular weight of 354.10 and a molecular formula of C 9 H H I N 2 O 5 . BrdUrd differs in replacing the iodo- (iodine) group with the bromo-(bromine) group resulting in a slightly lower molecular weight of 307.10 and a molecular formula of CgHnBrNbOs (Figure 2.1). 2.2.3 Dual labelling with thymidine analogues To measure cell kinetics, the spheroids were double labelled using iododeoxyuridine and bromodeoxyuridine (Figure 2.2). IdUrd and BrdUrd are both thymidine analogues that are incorporated into the cells during the S phase of the cell cycle (Evans et al., 2000). The spheroids were first pulse-labelled with IdUrd at a concentration of 5 uM for 10 minutes. This label was subsequently removed and the spheroids were washed three times with warm culture medium containing 10% serum. After incubation in a water bath for 12 hours to 18 hours, the spheroids were then exposed to BrdUrd at a concentration of 5 uM for 10 minutes. Once again, the label was removed and the spheroids washed with the warm 10% serum-containing medium. The spheroids were incubated in the water 47 e Figure 2.2 Detection of incorporated BrdUrd (or IdUrd). (A) DNA double-strand helix supercoils unwind for synthesis of new strands of DNA and thymidine analogues are incorporated only at this stage of the cell cycle. (B) Fluorescently tagged antibodies detect the presence of BrdUrd and IdUrd. 48 B G1 G2 100-1 50 4 Add BrdUrd or IdUrd Figure 2.3 Single labelling with BrdUrd or IdUrd. (A) Cells in S-phase labelled after incorporation; (B-F) position of previously labelled cells as they progress through the cell cycle; (G) graph of percentage of labelled cells in mitosis where cells in positions (a-f) correspond to the positions in parts {A-F) above. The positions labelled (g, h, i, j) correspond to the same position as (c, d, e, f), but just one cell cycle behind. The position marked (b2) is the half-way point between (b - c), and the time of the S phase would be indicated as Ts. The T c is the cell cycle time from one cell cycle peak to another. 49 16-28 hrs . 4 hr Figure 2.4 Dual labelling with BrdUrd and IdUrd. The green arrow represents contact of the first label (e.g. BrdUrd) and after an interval of time; those cells would have progressed through the cell cycle. At this time, the second label contacts the cells (red arrow), and some of the initially labelled cells might be labelled as well. If the cell was present in the S-phase during both labels, the cells would fluoresce as yellow, or a mixture of the red and green fluorescent tags. 50 bath for 4 hours, after which they were harvested and trypsinized. The cells were then fixed in 70% ethanol and stored at - 2 0 ° C for at least half an hour. 2.2.4 Cell cycle time kinetics The incorporation of radioactive thymidine or thymidine analogues (like BrdUrd and IdUrd) into cellular DNA only occurs during the S-phase of the cell cycle, when the DNA helix has unwound for synthesis. Thus to determine the cell cycle time of cells, we utilize a dual-label procedure where cell kinetic measurements can be calculated using the values for the label uptake as well as the duration of the interval time between administration of the labels. In addition, with the use of flow cytometry, both BrdUrd and IdUrd can be detected using fluorescent antibodies, thus the labelling index of these labels can also be determined (Figure 2.3 and Figure 2.4). The calculation for T c was a derivation of a series of equations, first with the calculation of the relative movement (RM), which was the movement of S-phase cells in relation to the positions of G1 and G 2 phases using the following equation: FLU _ FGI RM = (1) FG2 - FQI where F G i and F G 2 were the mean DNA content of cells in G i and G 2 phases respectively, and FLU was the mean DNA content of labelled undivided cells. For the calculation of RM, the measurements of the DNA content were based on the labelling patterns of the second thymidine analogue, BrdUrd. From equation (1), the amount of time cells are in the S-phase of the cell cycle (Ts) was calculated as follows: 0.5 T s = x t (2) R M - 0 . 5 51 where t was the sampling time of the second label, BrdUrd, which for all of our experiments were 4 hours. Before calculating the T c using the interval time between the two thymidine analogue labels, a few more parameters need to be calculated such as the number of channels that the DNA content shifted from the mean DNA content of cells in the S phase (Fs) and the DNA content value at the end of the S-phase, again based on the labelling patterns of BrdUrd. Shift = (End of S-phase) - F s (3) Also, the difference in channels between the mean DNA content of double labelled cells in the S phase (i.e. F s from cells labelled with both IdUrd and BrdUrd) and F s from equation (3) was calculated: Difference = F s - (Fs of double labelled cells) (4) Thus, by using the labelling of two thymidine analogues as an index of the movement of the cells through the cell cycle to obtain values for equations (1) to (4), the cell cycle time (Tc) was calculated using equation (5): T c = interval time - Difference (5) (Shift / Ts) or if displayed according to the equation numbers: (4) T c = interval time (3)/Ts 2.2.5 Fluorescence activated cell sorting To obtain six cell subpopulations within each spheroid, spheroids were incubated with 2 JJM Hoechst 33342 (Sigma, St. Louis, MO) for 20 minutes, washed first with warm 52 MEM and disaggregated by trypsinization and repeated pipetting. This resulted in monodispersed, Hoechst-stained cells resuspended in complete medium (Durand, 1982; Durand and Olive, 1982). Fluorescence-activated cell sorting was carried out with a Becton Dickinson (Sunnyvale, CA) dual laser FACS 440 cell sorter. Hoechst 33342 intensity was measured using excitation at 350-360 nm (40 mW power) with emission monitored with a 449 ± 10 nm band pass filter (Hartwell, 1992). 2.2.6 Flow cytometry with antibodies After cell sorting, all spheroid subpopulations were fixed in 66% ethanol and incubated at -20°C for at least Va hour. The samples were subsequently washed once with PBS and again with cold buffer A [4% FBS, 0.1% triton X, and PBS], centrifuging in between at 1230 rpm for 10 minutes each time. After aspirating buffer A from each sample, they were immersed in 1ml of an acid solution [2N HCI and 0.5% triton X] for approximately 30 minutes at room temperature. Cells were neutralized by washing off the acid with ice cold MEM three times, again with centrifugation at 1230 rpm for 10 minutes between each step. After all three washes, samples were resuspended in buffer A and were allowed to sit in buffer A for at least 10 minutes before the addition of the primary antibody. The antibodies chosen were Br3 (anti-BrdUrd, Caltag Laboratories, San Francisco, CA) at a 1:100 dilution (100 u.l) and FITC-conjugated B-44 (anti-ldUrd and anti-BrdUrd, Becton Dickenson Immunocytometry Systems, San Jose, CA) at a 1:10 dilution (100 u.l). An IgG rCy3 goat anti-mouse antibody (Caltag Laboratories, San Francisco, CA) at a 1:50 dilution (100 LII) was used as the secondary antibody against the primary Br3 antibody. Lastly, cell DNA was stained with DAPI (4', 6-diamidino-2-phenylindole dihydrochloride hydrate, Sigma Chemical Company, St. Louis, MO) at 1 ug/ml (Ljungkvist et al., 2006). Flow cytometry was performed using established 53 methods as described previously in the literature (Begg et al., 1985; Clevenger et al., 1985; Olive and Durand, 1994; Ljungkvist et al., 2006). List mode files were collected using an Epics Elite flow cytometer and subsequently reprocessed for analysis. Doublet correction and gating were used to select the cell populations of interest with the "Winlist" software package and univariate DNA histograms were analyzed with the "Modfit" package (both from Verity Software House, Topsham, ME). Depending on the progression of the IdUrd-labelled cells at the time of analysis, and the advancement of the BrdUrd-labelled cells, the cell cycle time (Tc) can be calculated (Sham and Durand, 1998). Labelling indices (LI) were calculated for the BrdUrd-labelled cells as well as the IdUrd-labelled cells (Durand, 1997). 54 2.3 RESULTS 2.3.1 Cell cycle times in spheroids vary by region As three-dimensional, multicellular aggregates, spheroids have different growth kinetics compared to cells grown as monolayers in vitro as well as tumours grown in vivo. Cell cycle time measurements within each region of the spheroid were performed using the dual labelling method with two thymidine analogues, BrdUrd and IdUrd. To analyze these data, each FACS sample of a subpopulation in a spheroid was first analyzed by a flow cytometer to generate dual colour flow dot plots, one colour for each thymidine label (as shown previously in Figure 1.5). To generate these dual colour flow diagrams, two control histograms analyzed first include the forward scatter plots and the plots for the time of flight (Figure 2.5). These two dot plots are used to distinguish and gate out those cells that are higher in ploidy number, those that are dead, as well as any other small fragments of debris (Figure 2.5). As shown in Figure 2.5, cells are gated into region 1 (R1) and region 2 (R2) initially and all subsequent dot plots would be analyzed using only those cells in R1 and R2. In addition, when dual colour dot plots were analyzed, another important control we used included the analysis of the single colour plots, to clearly distinguish which population of cells are labelled with both BrdUrd and IdUrd. As a comparison to the dual colour plots (Figure 1.5), plots displaying either the green fluorescence only or red fluorescence only served not only as controls but an indicator of the location of the cells (Figure 2.6). The DNA histograms for each respective dot plot were also analyzed for a normal distribution of cells through the cell cycle. With increasing debris or 'unhealthy' cells, this would be reflected in the DNA histograms as wider Gi peaks and uneven distributions. The DNA content of cells were identified using DAPI and thus to control for cross-colour contamination, dot plots of DAPI only samples were analyzed for any green or red fluorescence (Figure 2.7). 55 A B Figure 2.5 Forward scatter and Time of flight flow cytometry dot plots. IdUrd only labelled cells (A and B) and BrdUrd only labelled cells (C and D). Plots are shown separating cells using forward scatter (A and C) or time of flight (Sand D). 56 A B Figure 2.6 Single colour dot plots and DNA histograms. IdUrd only labelled cells (A and B) and BrdUrd only labelled cells (C and D). DNA histograms for each population of cells labelled with IdUrd only (B) and BrdUrd only (D). 57 A B Figure 2.7 Dot plots and DNA content histograms of DAPI only samples. Cells were gated for forward scatter (A), time of flight (B), and checked for normal cell cycle distribution on the DNA histogram (C). When plotted with both colours, the entire population of cells were negative for both (D). 58 The Chinese hamster V79-171b spheroids have an approximate average cell cycle time (Tc) of 16 hours whereas both the human tumor lines (SiHa and WiDr) show an average T c value of approximately 24 hours (Figure 2.8). Fractions that are 'bright' contain cells stained with more Hoechst dye. Thus brighter fractions are found on the outer layers of the spheroid and 'dim' fractions are closer to the center of the spheroid. A prolonged cell cycle time is observed for cells in subpopulations close to the necrotic center (dim fractions). To calculate the T c of a cell, we utilize the dual labelling method of two thymidine analogues (IdUrd and BrdUrd), and one parameter of the calculation involves the interval time between these two thymidine analogues. Each thymidine analogue labelled all the cells in the S-phase at the time of the pulse with an interval time in between the two thymidine analogues at approximately the duration of one cell cycle. The T c calculations were based on the rationale that under 'perfect' situations, the cells that have labelled with the first label would have progressed through one cell cycle and should be in the S-phase once again when the second label was introduced to the cells. Under the caveat that the chosen interval time is exactly the T c for every cell and the cells had no interruption in its progress, then all cells in the S-phase would be dual-labelled, and no cells would contain with only one label. In reality, the chosen interval times were not exact, nor does every cell in the spheroid system have a similar cell cycle time. Thus, results revealed a proportion of cells that were dual-labelled, and another two fractions that were labelled with one label or the other. The T c measurements are based on the labelling index as well as the amount of time it takes the cells to complete the S-phase of the cell cycle. As shown in Figure 2.8, for each cell type, we have done this experiment using various interval times to test for variability and we observed that regardless of the interval time, the results for T c were similar. In the V79 experiments, the interval times between the two thymidine analogues pulses were 16, 18 and 20 hours. Since SiHa and WiDr cells have a T c longer than 20 hours, the pulsing interval 59 V79 Spheroids SiHa Spheroids WiDr Spheroids 0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7 Sorted Fractions (bright to dim) Figure 2.8 Cell cycle time kinetics. Within different regions of spheroids, the T c value varies as graphed in this figure from single determinations. Graphs are displayed according to depth into spheroids, where bright cells (i.e. fraction 1) was near the periphery and dim cells (i.e. fraction 6) was near the centre of the spheroid. The fraction marked as 0 was the unsorted samples to show the average value of T c when the cells were taken from the spheroids as a whole. 60 for these two human tumor lines was increased to 22, 24, 26 and 28 hours. Although the interval time is pertinent in T c calculations, each cell line shows similar T c patterns regardless of the interval time. Thus, this difference in growth patterns does not vary with changes in the interval time, and are more likely to be a consequence of the microenvironment or other cell cycle controls that activate or inhibit cellular proliferation either prior to entry into the S-phase or entry into mitosis. 2.3.2 Labelling index of BrdUrd and IdUrd As a control for the T c experiments, we wanted to confirm the accuracy of the calculations by analyzing the total percentage of cells from an entire population that labelled for the IdUrd and BrdUrd thymidine analogues (Figure 2.9). Spheroids were first pulsed with IdUrd before they were pulsed with BrdUrd. All three cell lines appeared to have an approximately 5-50% labelling index of IdUrd (Figure 2.9A). However, it can be observed that the proliferating cells incorporate much more IdUrd than the quiescent cells. After the IdUrd pulse, there were intervals of 16 to 28 hours before the introduction of the second label, BrdUrd. Similarly with BrdUrd, the three cell lines appeared to have a 5-50% labelling index (Figure 2.9B). Once again, it can be observed that the external, proliferating cells incorporate significantly more BrdUrd than the internal, mostly quiescent cells. Both the IdUrd and the BrdUrd graphs indicate that there is an increase in the number of labelled cells in the proliferating (bright) fraction when compared to the internal (dim) subpopulations. These figures displaying the IdUrd and the BrdUrd labelling (Figure 2.9) are important because they indicate that there is a consistent pattern of BrdUrd and IdUrd labelling. If the labelling fluctuated, then that would indicate that there are different percentages of the cell populations labelled in each scenario. This would prevent an accurate 61 A V 7 9 Spheroids SiHa Spheroids WiDr Spheroids 0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7 Sorted Fractions (bright to dim) B V 7 9 Spheroids SiHa Spheroids WiDr Spheroids Sorted Fractions (bright to dim) Figure 2.9 Labelling index for thymidine analogues. (A) IdUrd and (B) BrdUrd 62 comparison between each subpopulation. Although both the IdUrd and the BrdUrd labelling were found in the range of 5 to 50%, the labelling percentage of IdUrd was seen to be slightly greater than BrdUrd. This occurrence can be accounted for by the fact that IdUrd was the initial pulse. If the interval time between the two labels was exactly the time for one cell cycle, then those cells would have replicated, therefore, a proportion of the IdUrd labelled cells may have already replicated. 63 2.4 DISCUSSION Any tumour sample would inadvertently include some contamination of various normal tissue cells such that biopsy samples are usually (if not always) heterogeneous. Any attempts at kinetic measurements may be compromised by the presence of the different cell types, thus it is necessary to take extra precautions to ensure that the kinetic values are of the tumour and not from a heterogeneous population. Using the in vitro spheroid model system, each spheroid was comprised of identical immortalized single cells from the same cell lineage, that have clustered together in addition to multiplying from single cells to form large multicellular aggregates. This model system has been utilized for several decades, however, measurements of cell cycle time through the spheroid have not been previously reported. Even though the cells in the spheroid were all derived from the same parental line, Figure 2.8 outlines the cell cycle kinetic differences between six subpopulations within a spheroid. Cells nearest to the periphery of the spheroid were found to have the shortest cell cycle time of approximately 14 hours for the V 7 9 cells and 21 hours for both SiHa and WiDr spheroids. Conversely, those cycling cells closer to the centre of the spheroid had a much slower cell cycle time averaging 18 and 28 hours for the V 7 9 and for the human lines respectively. The results described in this chapter of pulse-labelling spheroids with BrdUrd and IdUrd in combination, show that cell cycle kinetic studies were feasible on a tumour model where we also had the ability to separate out regions of cells based on cellular depth into each spheroid. For this reason, the present application of this methodology gave rise to information that those cells closer to the periphery had the shortest Tc, and this value gradually increased with increasing depth into the spheroids. This phenomenon would be understandable in tumours since the population of cells furthest from blood vessels would also be devoid of important factors such as nutrients, oxygen and growth factors. Tumourgenic cells would respond according to the environmental factors surrounding 64 them and the three dimensional nature of the spheroid model also gives rise to similar characteristics. The difference in cell cycle time between regions of the spheroid was significant given that the cells were in vitro controlled cultured cells and placed in the three dimensional spatial orientation only for the duration of spheroidal growth. This gave rise to evidence for the influence of the microenvironment and how it could be the trigger point for further delays in the growth signalling processes. Pulse labelling of cells with thymidine analogues to determine the proliferative fraction of a population has become a common practice reported in many disciplines of study ranging from the cell cycle, to tumour perfusion studies, and BrdUrd has even been utilised in stem cell research (Luk et al., 2005). Used in combination, two thymidine analogues give the added information of predicting both the potential doubling time (Tpot) and T c of a given population of cells. This measurement becomes increasingly important during clinical trials using radiation therapy where treatment schedules for patients are still difficult to optimize for the general public when the speed of cellular proliferation would contribute greatly to the effectiveness of the treatment. However, when retrospective studies were performed using data from patients with head and neck cancers (Haustermans and Fowler, 2001), the authors found that the only pre-treatment kinetic parameter with a correlation with local control of treatment were the values for labelling index (LI), not Tpot. The major limitations for Tp o t calculations with tumour samples (especially in cross laboratory studies), are sample variability, intra- and intertumour variability, inter-laboratory variability as well as the heterogeneity of tumour samples contaminated with normal cells. These limitations were factors that were controlled in our system, thus the values we were able to obtain for LI and T c were accurate for our model system, though not used as a predictive value. It does provide information on the proliferative capacity of a particular cell line which could then be extrapolated to the cells in vivo which as stated 65 earlier, may be harder to test reliably. There may be a therapeutic advantage to use this cell kinetic technique as an initial test of the proliferative status of a solid tumour. In clinical protocols, most patients receive a standardized treatment of radiation therapy, mostly in addition to a series of chemotherapy as well. While this is a blanket approach to target a majority of the solid tumours, the reactions and resulting outcomes from the therapies range drastically for each patient. For example, we have come to understand that hypoxic tumour cells are highly radioresistant, resulting in poor therapeutic outcomes when using radiation therapy on tumours with a high hypoxic fraction. With the understanding that at times, the best therapeutic approach for a given patient may be to quickly initiate therapy, either with radiation or a cocktail of drugs, in hopes of delaying and preventing further growth especially in cancers in late stages of development. Although this may be the safe approach, it is often not the most effective since some solid tumours would inherently be more sensitive or resistant to certain therapies. Even though this knowledge has been established for years, it is often difficult to merge the two branches of medicine, the clinical and the academic arms, to develop a more individualistic approach to therapy for maximal positive effect. Durand and Aquino-Parsons (2004) looked at the response and evolution of solid tumours that were undergoing chemoradiotherapy. By taking sequential biopsies from the same patients as they underwent therapy, they were able to follow the response of cervical tumour cells. They concluded that accelerated repopulation began very rapidly after the onset of therapy and suggested that there was a connection between the growth potential of cells and the in situ response to therapy (Durand and Aquino-Parsons, 2004). Without a more individualistic approach, patients may be given treatments that have minimal advantage for therapy and more importantly, chemotherapy or radiation would stimulate the tumour cells and initiate accelerated repopulation creating the potential for secondary tumours in the patients. 66 Our studies also indicated the importance of the microenvironment on growth control and the advantageous use of proliferation studies to better understand solid tumours prior to treatment. The subsequent chapters in this thesis further investigate how the control processes were affected when we manipulated the microenvironment, mainly using hypoxic conditions, as well as manipulating the cell cycle regulation using drugs that hinder the process of cell proliferation. 67 C H A P T E R 3: H Y P O X I A I N F L U E N C E D C H A N G E S IN T H E L E V E L S O F T H E M I T O T I C C Y C L I N B1 W I T H I N R E G I O N S O F A S P H E R O I D 68 3.1 INTRODUCTION In normal tissues that undergo cell renewal, there is a balance between cell proliferation, growth arrest and differentiation, and loss of mature cells by programmed cell death, or apoptosis. Tumours grow because the homeostatic control mechanisms that maintain the appropriate numbers of cells in normal tissues are defective, leading to an imbalance between cell proliferation and cell death, and an expansion of the cell population. The development of autoradiography and its use with tritiated thymidine in the 1950s and 1960s, and the more recent application of fluorescence tagged analogues with flow cytometry, have allowed a detailed analysis of tumour growth in terms of the kinetics of proliferation of its constituent cells. The proliferative rate of tumour cells varies widely between tumours; nonproliferating cells are common, and there is often a high rate of cell death. The rate of cell proliferation in tumours may be an important factor in determining prognosis, or response to radiation or chemotherapy. Several tissues, including the bone marrow and gastrointestinal tissues contain cells with high rates of proliferation, and damage to these cells may be dose limiting for chemotherapy. An understanding of the molecular events that regulate the cell proliferation cycle and the mechanisms whereby malignant cells escape from cell cycle controls, from apoptotic cell death, and from the limitations of a finite life span are key to an understanding of the differences between growth of the normal cells and that of their malignant counterparts. These concepts are also important for understanding the interaction of drugs and radiation with tissues and are likely to provide leads for the development of new therapeutic strategies. The loss of regulatory control of the cell cycle, leading to unrestrained cell proliferation, is a hallmark of cancer. Cyclins, the regulatory subunits of cyclin-dependent kinases (Cdks), control the passage of proliferating cells through important checkpoints in the cell cycle (Dirks et al., 1997; Hall et al., 1996; Sherr, 1995). Cyclin binding to its Cdk is 69 required for kinase activation, and these Cdks regulate a series of biochemical pathways, or checkpoints, that integrate mitogenic and growth-inhibitory signals, monitor chromosome integrity, and co-ordinate the orderly sequence of cell cycle transitions (Hartwell, 1992). In general, cyclin levels oscillate throughout the cell cycle and the cyclin mRNA and protein expressions peak at the time of maximum kinase activation, contributing to discrete bursts of kinase activity at specific cell cycle transitions. The B-type cyclins are part of a family of mammalian mitotic cyclins which all share a conserved sequence of approximately 100 amino acids, where a mutation in this sequence would disrupt both kinase binding and activation. When associated with Cdk1, they can control both the entry and the exit from mitosis through the phosphorylation sites on Cdk1. Phosphorylation of the inhibitory Thr14 and Tyr15 sites keeps the kinase inactive until the G2/M phase transition and in contrast, dephosphorylation of these sites by Cdc25 phosphatase, and CAK activation, triggers Cdk1 activation, which is essential for mitosis to occur (Lenormand et al., 1999). Multicellular spheroids are composed of cells growing in a 3-dimensional structure simulating the growth and microenvironmental conditions of tumours. Spheroids can be used in a variety of research areas as an in vitro tumour model system of intermediate complexity between the standard monolayer or suspension cultures and tumours in vivo (Sutherland et al., 1981). The physical attributes of spheroids are advantageous, where not only are culture conditions controlled, but mathematical models have also been generated to correlate growth rate and diffusion rates in attempts to understand a macroscopic system within mesoscopic foundations (Delsanto et al., 2005). There has been previous work published using spheroids as an in vitro tumour model, using growth conditions to form spheroids of optimal size (Durand, 1993; Santini et al., 1999; Oloumi et al., 2000). In fact, spheroids have not only been used to test the effects of drugs currently used in the clinic (Durand and LePard, 1997; Durand and LePard, 2000) but 70 also as models for the effects of radiotherapy on solid tumours (Durand and Olive, 1982). One characteristic found in both tumours and large spheroids is the hypoxic nature of the cells near the central necrotic region. This sub-optimal growth condition would presumably change the growth kinetics of cells within the same spheroid. It is therefore important to elucidate the changes in the cell cycle under microenvironmental stresses, and determine the mechanisms behind these differences. In the previous chapter, experiments were designed to compare the actual cell cycle time of cells located in various regions within the spheroid using the established technique of perfusion-based cell sorting (Durand, 1986; Durand and LePard, 1997). In the present chapter, to answer questions regarding the connection of cell proliferation with the microenvironment, we further investigated the levels of an important cell cycle regulator, cyclin B1, in each of these spheroid subpopulations. We have determined that cells near the necrotic centre of the spheroid have a longer cell cycle time and the levels of cyclin B1 were lower in these regions. The role of hypoxia was directly observed by varying the external level of oxygen. Three cells lines were used to evaluate the generality of these observations. 71 3.2 MATERIALS AND METHODS 3.2.1 Spheroids and Reagents The preparation of spheroids was similar to the methods described in section 2.2.1. 3.2.2 Regulation of oxygen concentration The spheroids had controlled atmospheric conditions where gas was continuously flowing into the spheroid flasks over a 4 hour time period. The gas was supplied from pre-mixed gas tanks (Praxair, Danbury, CT) where gas tank regulators controlled the pressure of the gas, and hoses led directly from the tank to a gassing apparatus board (Figure 3.1). This board served as a central location where the use of one tank could simultaneously gas up to 12 flasks, or it had the versatility to allow up to 6 gas tanks to be used at the same time. The board had 12 'arms', each equipped with a flow meter, which were the connectors from the board to the spinner flasks. These flow meters controlled the flow rate of the gas into each spinner at 80 cc/min. The end of each hose had an adaptor that contained a luer lock to secure the hose to the needle that protruded out of a rubber stopper. The rubber stoppers also had another needle as a gas release valve for the spinner flasks. 3.2.3 Trypsinization of spheroids Sequential trypsinization is a method that generates relatively pure populations of cells from discrete regions in a spheroid (Freyer, 1988). The outer cell layer of spheroids was removed by agitating them with cold 0.1% trypsin in PBS for five minutes. The outer cells that dissociated from the spheroid were gently removed from the remaining spheroid, and these outer cells were transferred to cold medium. The process was repeated with the remaining cores and the cores were trypsinized to completion with this population of cells as the inner cells. The sequential method of trypsinization was used 72 A B Figure 3.1 The gas apparatus used for controlling external oxygen concentrations. (A) Angled view of the apparatus showing the gas tank bank, the control board mounted on the wall, as well as the water baths where the spinner flasks are immersed in water for constant temperature. (S) Front view of the apparatus showing more detail of the control board, containing 1 2 individual gauges that regulate the gas pressure as well as give flexibility for multiple ports with one gas tank. 73 to generate the outer and inner regions of cells for the Western blot analysis, however, in all other instances; spheroids were agitated for 10 minutes in warm 0.1% trypsin in sodium citrate buffer. 3.2.4 Western Blotting Whole cell lysates of spheroid subpopulations were prepared by lysing 5.0 X 106 cells in radioimmunoprecipitation assay (RIPA) buffer [0.15M NaCl, 1% sodium deoxycholate, 0.1% SDS, 1% Triton 100, 50 mM Tris-HCI (pH 7.4), 10//g/ml protease inhibitor mixture (Sigma), and 0.1 mM Na3V04]. Equal amounts of protein (BioRad Protein Assay) were denatured in sample loading buffer [10% 2-mercaptoethanol, 2% SDS, 30% glycerol, 0.025% bromophenol blue, 50 mM Tris-HCI (pH 6.8)] and loaded into the wells of a 10% SDS-polyacrylamide gel. After electrophoresis, gels were blotted onto a polyvinylidene fluoride membrane and analyzed for cyclin B1 using a mouse monoclonal antibody (Upstate Biotechnology, Lake Placid, NY) that was incubated with the membrane overnight at 4°C (Figure 3.2). To remove protein binding, membranes were placed in stripping buffer [1X TBS, 0.07% 2-mercaptoethanol, 0.2% SDS] for 30 minutes at 50°C, and rinsed with TBST prior to re-probing with another primary antibody. For Cdk1, we used a rabbit monoclonal antibody (Upstate Biotechnology, Lake Placid, NY) incubated with the membrane overnight at 4°C. Secondary goat anti-mouse antibodies (Calbiochem, San Diego, CA) or goat anti-rabbit antibodies (Calbiochem, San Diego, CA) conjugated to horseradish peroxidase (1:8000) were incubated with blots for 1 hour, and bands were detected using the enhanced chemilluminescence detection system (Amersham, Buckinghamshire, England). In addition, all blots were also probed for the levels of actin (Oncogene Research, Boston, MA) to confirm equal protein loading in each sample lane. 74 target nttro cellulose add HRP P _ k IS remove unbound HRP • a-0 I detect HRP Source: Adapted from (http://web.mit.edu/esqbio/www/rdria/rdna.html): Southerns, Northerns, Westerns, & Cloning: "Molecular Searching" Techniques, Barry White, MIT, 1995 Figure 3.2 Western Blotting. General overview of the process for Western Blotting from the beginning stages of a lysate mixture of proteins to antibody detection and chemiluminescence using horse radish peroxidase (HRP) as a probe. 75 3.2.5 Fluorescence activated cell sorting The methods for fluorescence activated cell sorting were similar to the methods described in section 2.2.5. 3.2.6 Flow cytometry with antibodies The methods for flow cytometry were similar to the methods described in section 2.2.6. The cells were allowed to sit in buffer A for at least 10 minutes before the addition of the primary antibody. Cyclin B1 monoclonal antibody (PharMingen, San Diego, CA) was diluted 1:100 in buffer A [4% FBS, 0.1% triton X, and PBS], 100>l of diluted antibody was added to each sample for an overnight incubation at 4°C. Samples were then topped up with buffer A and centrifuged at 1230 rpm for 10 minutes. The secondary antibody used was a goat anti-mouse Alexa 488 (Molecular Probes, Eugene, OR) at a 1:200 dilution in buffer A, 100//I of diluted antibody was added to each sample for 2 hours at 4°C. Washed samples were then treated to DAPI (4', 6-diamidino-2-phenylindole dihydrochloride hydrate, Sigma Chemical Company, St. Louis, MO), a DNA stain, at a concentration of 1 ug/ml (Ljungkvist et al., 2006). DAPI was added to each sample prior to sorting, and the samples were held on ice while in queue. 3.2.7 Cryotome sectioning of spheroids For immunohistochemical analysis of the cyclin B1 expression under various oxygen concentrations, spheroids were first placed in a water-jacketed spinner flask with one of the arms plugged with a rubber stopper equipped with two needles, one to attach to the gas tank connector hose, and the other as a gas release valve. The spheroids were gassed for a period of 4 hours and subjected to 2 /vM Hoechst 33342 for 20 minutes prior to the completion of the 4 hour time frame, after which they were removed from the flask and placed promptly in ice to be washed with cold PBS. After aspirating as much 76 of the PBS as possible, the spheroids were gently mixed with a tissue-embedding compound (Tissue-Tek® O.C.T. Compound, Bayer, Berkley, CA), and then frozen inside the cryostat on pre-chilled freezing discs at -20°C. The frozen blocks were sectioned using a cryotome into 10pm thick slices and placed on glass slides for immunohistochemistry analysis. The cyclin B1 antibodies used were the same as those used in the Western Blot analysis, and under a fluorescence microscope, detection of both the Hoechst and the antibody against cyclin B1 was visualized using different wavelengths of light to excite the different fluorescent colours. The colours shown in the images were false-colour digital images overlaid on top of one another to show both layers of detection of the two fluorescent tags. 77 3.3 RESULTS 3.3.1 Cyclin B1 protein levels decrease with depth within a spheroid For the western blots shown in Figure 3.3A lysates were from cells of V79-171b spheroids that were partially trypsinized to create two samples; an outer cell and an inner core fraction. We have also used unsorted cells (i.e. whole spheroids) as a comparison. The lysates were each probed for cyclin B1 and Cdk1. The bottom series of bands indicate the relative amounts of actin found in each sample and serves as our loading control. For the human tumor cell lines SiHa and WiDr (Figure 3.38 and C respectively), lysates were similarly partially trypsinized and probed for cyclin B1. Once again, levels of actin were determined as well as a control. The results indicate that for both the Chinese hamster and human cell lines, there are higher cyclin B1 protein levels in the outer fractions compared to the inner fractions and the spheroid as a whole. For the V79-171b cells, the blot was also subjected to the Cdk1 antibody probe after the blot was stripped of the cyclin B1 bound antibodies. Cdk1 is the active binding partner for B type cyclins and this protein also had a similar pattern of increasing protein levels when comparing the different regions within a spheroid. In addition, these spheroids were subjected to varying degrees of oxygen concentrations for 4 hours, as depicted by the percentages of oxygen shown at the top. The levels of the protein also increased as the growth conditions allowed greater amounts of oxygen, with the highest level of cyclin B1 protein found in spheroids subjected to air (21% oxygen) and levels decreased proportionally to decreasing amounts of oxygen. Although the spheroids were only gassed at the various oxygen concentrations for 4 hours, this result is consistent in all three cell lines. 78 A 0 % 5 % A I R uns ou t in uns out in uns o u t in Cyclin B l Cdkl B 0 % 5 % A I R uns ou t in uns o u t in uns out in Cyclin B l Actin 0 % 5 % A I R uns ou t in uns out in uns o u t in « • • . - Cyclin B l Actin Figure 3.3 Western Blots for Cyclin B1 and Cdk1. Lysates of spheroids that were partially trypsinized to create an outer cell fraction (out) and an inner core fraction (in). Whole spheroids were also trypsinized fully to create the unsorted (uns) sample. Cell lines tested include V79-191b (A), SiHa (S), and WiDr (C). 79 3.3.2 Hypoxic conditions affect the levels of Cyclin B1 observed in spheroid subpopulations Western analysis gives a subjective indication of the expression of protein levels; however, we also wanted to validate the results quantitatively using fluorescence activated cell sorting (FACS) and flow cytometry. Using FACS, we separated out six subpopulations of cells where each fraction represents a shell of cells of equal thickness. In the flow cytometry analysis, the cells were probed with antibodies against cyclin B1 with each curve from cells subjected to various oxygen mixtures ranging from 0% oxygen to 95% oxygen (Figure 3.4). Cells on the periphery of the spheroid had higher levels of cyclin B1 and these levels gradually decreased as cells approach the necrotic centre. In addition, the ambient levels of oxygen to which the spheroids were subjected had an effect on the levels of cyclin B1. Higher oxygen concentrations resulted in higher levels of cyclin B1 protein. In the cases where the spheroids were subjected to more hypoxic environments, the changes in protein levels within the spheroid did not change as drastically as the changes in more aerobic conditions. Paired Student's T-tests were performed comparing the data from the anoxic conditions (0% oxygen) to either air (21% oxygen) or hyperoxic conditions (95% oxygen). For V79 spheroids, when comparing 0% oxygen and air, these curves had a p value of 0.0446, while comparison of 0% and 95% oxygen had a p value of 0.0135. The p values for SiHa cells were 0.0106 and 0.0107 respectively, and WiDr cells also showed statistical significance with p values of 0.011 and 0.0049 respectively. To easily visualize this using the flow cytometry dot plots, we show an example plot using WiDr spheroids subjected to 0% oxygen (Figure 3.5A and B) or 21% oxygen (Figure 3.5C and D). Most notably, 0% oxygen resulted in very minor changes in cyclin levels even when comparing the cells in the periphery compared to those cells closest to the centre. Each series of experiments were done in triplicate for verification and validation. 80 A Cyclin B1 Levels for V 7 9 Spheroids 300 -i 250 -> o 200 -_ l EQ c 150 -o >. U c n 100 -a S 50 -0 -e 8\ • 0*/. Oxygen 53 1% Oxygen • 2% Oxygen V 5% Oxygen 10V* Oxygen • AIR © 95% Oxygen v 1 2 3 4 5 e Sort Fraction (Bright to Dim) Cyclin B1 Levels for SiHa Spheroids Cyclin B1 Levels for WiDr Spheroids 01 S 200 ca £ 150 o re o E I o • 0% Oxygen 1% Oxygen A 2% Oxygen V 5% Oxygen 10% Oxygen • AIR 95% Oxygen G 5 . I 4. x 5- - A 250 225 200 J 175 01 _ l T- 150 ffl £ 125 o O 1 0 0 c S 75 s 50 25 0 0 0% Oxygen 1% Oxygen A 2% Oxygen Y 5% Oxygen <> 10% Oxygen • AIR 0 95% Oxygen 1 1= 1 2 3 4 5 6 Sort Fraction (Bright to Dim) 1 2 3 4 5 6 7 Sort Fraction (Bright to Dim) Figure 3.4 Flow cytometry graphs of Cyclin B1 on FACS sorted spheroids. Using transient exposures to a range of oxygen concentrations, spheroids of various cell lines [{A) V79-171b, (8) SiHa, and (C) WiDr] were monitored for the changes in cyclin B1 levels in different regions of the spheroid. 81 A B Figure 3.5 Flow cytometry dot plots of cyclin B1 . Spheroids were subjected to 0% oxygen and were a sample from the interior of the spheroid (A) and from the periphery (B). In comparison, the spheroids subjected to 21% oxygen show higher cyclin B1 protein on the peripheral cells (D), compared to the inside (C) and to both samples exposed to 0% oxygen. 82 55 0 1 2 3 4 5 6 7 Sort Fraction (Bright to Dim) Figure 3.6 Percentage of cells in the S-phase of the cell cycle. There were progressively more cells in the synthesis phase of the cell cycle for those cells that were from the more peripheral regions of the spheroid. This trend was constant in all three cell lines. 83 3.3.3 Cell cycle distribution of the cells within the subpopulations of the spheroid To obtain an estimate of the proliferative status of the cells in the spheroids, we measured the percentage of cells undergoing the S-phase of the cell cycle in each cell subpopulation for all cell lines. Under the caveat that the position of the cell within the cell cycle is not only a close estimate of their proliferative status but would also indicate the proliferative trends of the cells, we gated the cells and tabulated the percentages of the total number of cells that were in the S-phase of the cell cycle at any one particular time. The results shown in Figure 3.6 were an average value of three independent studies for each cell line. Spheroids were exposed to Hoechst 33342 to create a florescence gradient via diffusion through the cells in the spheroids. After the spheroids were disaggregated by trypsinization into single cells, then they were run through the cell sorter for FACS analysis. In cells located near the surface of the spheroid, WiDr and SiHa cells had approximately 46% in S-phase and these values decreased to about 4% for those cells closest to the centre of the spheroid. The percentage of V7 9 cells with S-phase DNA content decreased from 42% to approximately 10% with increasing depth into the spheroid. These results indicate that the outer regions of the cells were actively proliferating for all three cell lines, while those cells deeper inside the spheroid had a lower S-phase fraction and were almost entirely quiescent. 3.3.4 Cryosections of WiDr Spheroids for spatial determination of Cyclin B1 under hypoxic treatment Immunohistochemistry provided visual confirmation of cyclin B1 distributions in WiDr cells under various oxygen concentrations. Intact spheroids were also subjected to the fluorescent Hoechst 33342 DNA-binding dye to visualize the decreasing fluorescent gradient with depth into the spheroid. After sectioning, the slides were counterstained 84 with a fluorescently labelled antibody staining for the presence of the cyclin B1 protein. In Figure 3.7, panel A shows the Hoechst fluorescence alone; panels B to F show spheroids that were subjected to 0%, 2%, 5%, 21% (air), and 95% oxygen for four hours prior to freezing in the mounting compound as described in "Materials and Methods". Hoechst 33342 binding is independent of the microenvironment unlike the levels of cyclin B1. Note the consistency with Figures 3.3 and 3.4. 85 Figure 3.7 Cryosections of WiDr spheroids. WiDr spheroids were subjected to hypoxic conditions, 0% and 2% (S and C respectively) as well as higher concentrations of oxygen, 5%, 21%, and 95% (D-F) prior to sectioning. Hoechst (blue) was added to observe the diffusion properties and sections were also probed against antibodies for cyclin B1 (red). The images displayed are an overlay of the two colours at a magnification of 100X. As a control, Hoechst alone (A) was also imaged with both fluorescent filters. 86 3.4 DISCUSSION There are many known mechanisms and control systems that are involved in the progression of cells throughout the cell cycle. However, in tumours, some of these pathways are disrupted or malfunctioning, resulting in abnormal control. In addition, it appears that cells within the same tumour are influenced by epigenetic mechanisms (e.g. the microenvironment). Spheroids are three-dimensional in nature and they have been shown to mimic the growth properties of tumours in vivo (Durand, 1976; Sutherland, 1988; Mueller-Klieser, 1997). In multicellular spheroid systems, oxygen and nutrients are supplied by the suspension growth medium and as a result, concentrations of these nutrients become increasingly less towards the interior of the spheroid. Consequently, we found more proliferating cells in the outer rim and quiescent cells mostly in the centre. Once a spheroid grows beyond a certain size, a necrotic core develops, and over time, more necrosis would be observed creating increasingly thinner outer shells of viable cells. There are many elements of the microenvironment that can affect tumour growth, including oxygen availability, tumour blood flow, pH and interstitial pressure. In spheroids, most of the growth factors are introduced through the growth medium. Therefore, the peripheral layers of the spheroid will be more accessible to all the nutrients, resulting in a shorter cell cycle time, while cells near the centre of the spheroids would have a reduced availability of nutrients and a prolonged cell cycle time and reduced growth fraction. However, in addition to the microenvironment, the differences seen in the cell cycle time for the six subpopulations may in fact be affected by other molecular controls that are activating or inhibiting cellular proliferation. Thus the interaction of the microenvironment and molecular controls may have significant implications for tumour growth, and for therapy. We have documented differences in the cell cycle time of different subpopulations within each type of spheroid, and have attempted to elucidate some causes of these 87 differences. While the spheroid cell subpopulations near the necrotic centre showed a prolonged cell cycle time, the peripheral cell layers had a more rapid cycle time suggesting that cell kinetic variations may be related to the local microenvironment. Most notably, cells near the interior of the spheroid would be subjected to hypoxic conditions. Recognizing the limitations of quantitatively measuring the levels of oxygen in each subpopulation, we instead chose to manipulate the external level of oxygenation. As expected, the outer areas of the spheroid had much greater amounts of the cyclin B1 protein compared to the inner regions (Figure 3.2). Also, the intensity of the band for the entire spheroid was lower than the outer fraction and higher than that found for the inner regions. This intermediate level of protein found for the whole spheroid can be explained by the fact that we had loaded the same amount of total protein in each lane of the SDS-polyacrylamide gel. As shown in our actin control, we had similar levels of protein loading in all lanes. Since the unsorted samples were a mix of all cells from the entire spheroid, there would be a smaller percentage of cells from the periphery when compared to the outer region sample. Thus, the trend where we saw intermediate cyclin B1 protein intensity for the unsorted sample further illustrates that there is a decreasing amount of cyclin B1 protein with increasing depth into the spheroid. For the V79, SiHa, and WiDr spheroids, not only did we find differential expression of cyclin B1 with respect to their location of the spheroid, but we also deduced the importance of oxygen, even for short time exposures. After as little as 4 hours of reduced oxygenation, cyclin B1 levels gradually decreased with decreasing concentrations of oxygen (Figure 3.3). Interestingly, these observations illustrate that measuring cyclin levels can provide a quick and convenient surrogate index for proliferation. When spheroids were transiently (4 hours) exposed to differing levels of oxygen, cyclin B1 levels changed markedly (Figure 3.3 and 3.4). For 21% and 95% oxygen, those cells near the periphery of the spheroid had a 2- to 3-fold increase in the levels of cyclin B1 88 when comparing those cells on the periphery with those near the centre of the spheroid. Even though the change in oxygen from 21% to 95% is a large range, the cyclin B1 levels did not significantly differ. We attribute this observation to the possibility that 21% oxygen is sufficient levels of oxygen for cells and further saturation of oxygen creates no definitive benefit, at least in the short term of up to 4 hours. Spheroids subjected to 2% to 10% oxygen showed a similar trend, where peripheral cells had higher cyclin B1 protein levels compared to their innermost cells; however, these levels were less than those found at higher oxygen concentrations in the same similar spheroid region. Comparing sort fractions 1 with 6, there is a maximal 2-fold increase in cyclin B1 levels. This trend dramatically decreased when the oxygen levels dropped to 1% and 0% oxygen. For the spheroid flask gassed at 1% oxygen, the relative position of a cell within a spheroid still created at most, a 30% higher cyclin B1 level in the peripheral cells. Even more drastic were the levels observed when spheroids were subjected to 0% oxygen, where there appears to be no change in the cyclin B1 protein level in any area within the spheroid. As a comparison between the V7 9 and the human cell lines, flow cytometry indicated that V7 9 signals were in general, at an elevated level, however, the trends seen were similar. The comparison of the cyclin B1 levels with respect to the location in the spheroid suggested that cells of similar origin grown in tissue culture could foster such different expressions. However, this gave rise to more evidence that there were signals in the spheroid microenvironment that would regulate the expression of cyclin proteins. Cyclin proteins highly associate with Cdks, such as Cdk1 for cyclin B1, as well as cyclin-dependent kinase inhibitors (Cki). One such Cki that inhibits the active kinase activity of cycin B1 with Cdk1 at the G 2 phase checkpoint would be the effects of p 2 1 wafi /ci P i p 2 1 wafi /ci P i h g s b e e p c i t e d t o b e j n d u c e d b y a v a r i e t y o f signals, including decreased pH (Ohtsubo et al., 1997), and hypoxia (Graeber et al., 1994). With the knowledge that hypoxia had effects on p2i w a f 1 / c i p 1 > there would be potential effects on 89 cyclin B1 indirectly or hypoxia itself could be a modulator of cyclin activation. The hypoxia studies we performed using fractionated spheroids were done with a range of oxygenation states, from extreme hypoxia (0%) up to 95% oxygen (Figure 3.4). The baseline for the studies was 21%, which is the amount of oxygen found in air. Throughout the entire spheroid, the differences in cyclin B1 for spheroids subjected to 21% and 95% oxygen were not significant, however, lowering the oxygenation caused dramatic changes in cyclin B1 levels. Under hypoxic conditions, each subpopulation of fractionated spheroids showed much lower intensities of the cyclin protein compared to the baseline. For us, the most intriguing result from our study of the effects of the external oxygenation states on cyclin levels was from the extreme hypoxic curve where the external oxygen was at 0%. Under these conditions, the once highly proliferating cells on the peripheral rim lost a majority, if not all, of their increased expression of the cyclin B1 protein, which we speculated to signify the loss and/or delay of cellular proliferation. This phenomenon was confirmed visually in our fluorescent images of spheroid sections (Figure 3.7), where we imaged Hoechst 33342 dye staining as well as fluorescent antibodies against cyclin B1. With similar growth environments and nutrient availability, the only difference between each panel was the oxygen mixture provided to the spheroids for 4 hours. Each panel showed the Hoechst 33342 diffusion gradient as it contacted the peripheral cells and entered the spheroid. This binding of Hoechst did not change in each image; however, the cyclin B1 was quite different. Panels D and E indicated those spheroids subjected to either 21% or 95% oxygen, and in both cases, there was cyclin B1 detected throughout the spheroid, with higher intensities near the outside of the spheroid. With more hypoxic gas mixtures, the levels of cyclin B1 were reduced, as indicated by dimmer fluorescence, and at 0% oxygen (8), there was hardly any cyclin B1 detected at all. 90 The DNA content analyses detailed how the regional location of a cell within a spheroid was a component in determining the fraction of cells in S-phase. These studies revealed a lower percentage of cells in S-phase with increasing proximity to the necrotic centre. The decline in the S-phase fraction could be a result of Gi-phase arrest, however, our results suggest that there were also molecular signals to arrest cells in the S- to G 2-phases of the cell cycle. This would have an influence on the cells by inhibiting them from entering the next cell cycle phase, and decreasing the numbers of cells proliferating in each round of the cell cycle. Since we have chosen to use three cell lines of hamster and human origin, the consistent patterns observed indicated that the molecular alterations were highly conserved across species. It had been previously documented that there were possible substances released by the necrotic centre of a spheroid that non-specifically induced arrest in all phases of the cell cycle (Freyer, 1988). This may partially explain the S- and G2-phase arrest; however, we also showed alterations in the G2-phase regulatory proteins, namely the cyclin B1 and Cdk1 proteins, which also would contribute to this effect. As an in vitro model system, spheroids have the benefit of cell culture reliability; however, their three-dimensional physical properties give them a complexity similar to that found in tumours. Thus, we believe that there is much information to gain in determining the signals in the spheroid environment that can regulate the cell cycle of tumourgenic cells. We have initiated these studies to not only investigate the changes in protein levels of our spheroid model, but we have also illustrated the importance of oxygen, and how even short periods of hypoxia can induce significant changes in the levels of important cell cycle proteins, and ultimately affect the rate of proliferation. There are still many unknown factors regarding the control mechanisms of the cell cycle, the interplay and cross-talk of the proteins, and the microenvironmental factors that control these processes. The microenvironment of a tumour constitutes a wide branch 91 of specific factors, such as hypoxia, nutrient availability, acidic pH levels, as well as cell-matrix interactions. We have undertaken the challenge to factor out the effects of transiently exposing the spheroids to hypoxia and observed dramatic changes in cyclin levels. The importance of oxygen to cyclin B1 protein levels revealed a consistent pattern that could also be manipulated as a diagnostic tool for hypoxia as well as a marker for the regulation of cell proliferation. It has been widely documented that hypoxic tumours pose clinical problems for treatment, mainly due to the radioresistance in these solid tumours (Lord et al., 1993; Olive and Durand, 1994). Thus the strong correlation we observed with cyclin B1 levels and hypoxia could potentially be used as a preliminary, non-invasive method of detecting tumour hypoxia. This would also be convenient for tumour biopsies or in vitro samples that were already taken without the administration of a hypoxia marker such as pimonidazole, CCI-103F or EF5 (Kennedy et al., 1997; Evans et al., 2000; Ljungkvist et al., 2006) where the levels of cyclin B1 protein might be an alternative to garner more information. Importantly, the results from these studies must be analyzed under the caveat that lowered cyclin B1 levels do not necessarily mean the cells were hypoxic, since our results can only suggest that hypoxic conditions give rise to lowered cyclin B1 levels. However, the reverse statement can not be substantiated with the data presented thus far. In other words, if cyclin levels in tumour cells were similar to that of the normal cells, then the tumour sample would very unlikely to be hypoxic, and conventional methods of radiotherapy may be prescribed without hesitation about radioresistance. If the cyclin levels measured in the tumour cells were lower than those of the normal cells, there is a chance that the tumour was hypoxic and at this point, additional and more specific methods of hypoxia detection would be utilized. Detecting the amount of cyclin B1 in a tumour sample would be a very quick and easy technique to perform, and more importantly, it would have the ability to divide patients further into two subgroups based on hypoxic severity. Levels of cyclin B1 92 protein suggests the possibility of hypoxic states and at this point, either more testing could be performed or even a therapy regimen more personalized and adapted to hypoxic tumours. In this chapter, not only have we shown that cyclin B1 levels vary dependent on the location of those cells within the spheroid, but we have also suggested that low cyclin B1 protein levels might be useful to confirm hypoxia. This relationship also led to the idea of implementing the use of this information as a predictive assay for tumour hypoxia, for the ultimate goal of creating more individualized regimens of cancer treatment for maximal effectiveness on every patient. 93 C H A P T E R 4 : O X Y G E N D E P E N D E N C E O F C Y C L I N D L E V E L S IN S P H E R O I D S C O M P A R E D T O S I N G L E C E L L S 94 4.1 INTRODUCTION Deregulation of cell division is one of the most common abnormalities found in cancer. The development and progression of cancer is dependent on various cellular genetic and epigenetic events, especially alterations of the cell cycle machinery at various checkpoints (Sherr, 1994; Hartwell, 1992). This precise control of the cell cycle is regulated by the activity of cyclins, cyclin-dependent kinases (Cdks), and cyclin-dependent kinase inhibitors (Dirks and Rutka, 1997; Hall and Peters, 1996; Sherr, 1995). Cyclin binding to its Cdk is required for kinase activation, and these Cdks regulate a series of biochemical pathways, or checkpoints, that integrate mitogenic and growth-inhibitory signals, monitor chromosome integrity and co-ordinate the orderly sequence of cell cycle transitions (Hartwell, 1992). In general, cyclin levels oscillate throughout the cell cycle and the cyclin mRNA and protein expressions peak at the time of maximum kinase activation, contributing to discrete bursts of kinase activity at specific cell cycle transitions (Paschoa et al., 2005). We speculated that different regions within a tumour would have different activity levels of cyclin D. Furthermore, we hypothesized that this change was not entirely dependent on the spatial location of a cell, but also a consequence of external factors such as hypoxia. Using multicellular spheroids as an in vitro tumour model system, we show herein the importance of the control of the cyclin D protein in the cell cycle, and how the levels change under conditions of hypoxia. In addition, we employ and examine the effects of rapamycin, an inhibitor of the mammalian target of rapamycin (mTOR) protein kinase that has recently been tested to be a new agent for anticancer therapy. 95 4.2 MATERIALS AND METHODS 4.2.1 Spheroids, single cells, and reagents The preparation of spheroids was similar to the methods described in section 2.2.1. Single cells from monolayers were also utilized for some studies as a comparison to the hypoxia treatment to the spheroids. 4.2.2 Regulation of oxygen concentration The regulation of the external oxygen concentration was similar to the methods described in section 3.2.2. As usual, 2 /vM Hoechst 33342 (Sigma, St. Louis, MO) was added to the media for 20 minutes; spheroids were then disaggregated and resuspended in complete medium for FACS. After cell sorting, all spheroid subpopulations were fixed in 66% ethanol and incubated at -20°C for at least VT. hour. The samples were subsequently washed 2 times with cold buffer A [4% FBS, 0.1% triton X100, and PBS] and spun at 1230 rpm for 10 minutes each time. After the final spin, the cells sat in buffer A for at least 10 minutes before the addition of the primary antibody. Cyclin D monoclonal antibody (Upstate Biotechnology, Lake Placid, NY) was diluted 1:100 in buffer A, 100 /J\ of diluted antibody was added to each sample for an overnight incubation at 4°C. Samples were then topped up with buffer A and centrifuged at 1230 rpm for 10 minutes. The secondary antibody used was a goat anti-mouse Alexa 488 (Molecular Probes, Eugene, OR) at a 1:200 dilution in buffer A, 100 /J\ of diluted antibody was added to each sample for 2 hours. Washed samples were then treated to DAPI also at a concentration of 1 ug/ml (Oloumi et al., 2000). 4.2.3 Single cell timepoints Experimental water-jacketed spinner flasks were filled with 40 ml of MEM with 10% FBS and placed in a 37°C water bath on top of magnetic stir plates. Each media filled 96 experimental flask was pre-gassed for 1 hour with various oxygen mixtures ranging from mixtures of 1% oxygen, 21% (air), and 95% oxygen. All of these various oxygen gas tanks were analyzed to also contain 5% C0 2 and balance N2. After pre-gassing to equilibrate the media, single cells from monolayer cultures were trypsinized and neutralized in MEM with 10% FBS and this cell suspension was added to the spinner flasks to a total volume of 50 ml. The single cells were continuously spinning using the magnetic stir bar inside the flask and they were gassed for several timepoints, from V2 hour to 4 hours. After each time point, the cells were removed from the flasks and centrifuged into a pellet at 1230 rpm for 10 minutes. Each sample was aspirated and fixed with ice-cold 70% ethanol. The samples were placed in a -20°C freezer for a minimum of 1/4 hour. Antibodies used to probe for cyclin D were the same as those described in the above sections. 4.2.4 Western Blotting Western blotting reagents and methods were similar to that previously described in section 3.2.4. However, after electrophoresis, gels were blotted onto a polyvinylidene fluoride membrane and analyzed for cyclin D using a mouse monoclonal antibody (Upstate Biotechnology, Lake Placid, NY) that was incubated with the membrane overnight at 4°C. To remove protein binding, membranes were placed in stripping buffer [1X TBS, 0.07% 2-mercaptoethanol, 0.2% SDS] for 30 minutes at 50°C, and rinsed with TBST prior to re-probing with another primary antibody. For Cdk4, we used a rabbit monoclonal antibody (Upstate Biotechnology, Lake Placid, NY) incubated with the membrane overnight at 4°C . Secondary goat anti-mouse antibodies (Calbiochem, San Diego, CA) or goat anti-rabbit antibodies (Calbiochem, San Diego, CA) conjugated to horseradish peroxidase (1:8000) were incubated with blots for 1 hour, and bands were detected using the enhanced chemiluminescence detection system (Amersham, 97 Buckinghamshire, England). In addition, all blots were also probed for the levels of actin (Oncogene Research, Boston, MA) to confirm equal protein loading in each sample lane. 4.2.5 Fluorescence activated cell sorting These methods were described previously in section 2.2.5. 4.2.6 Flow cytometry with antibodies Flow cytometry detection for all antibodies used in this chapter was performed using a three-laser Coulter Epics Elite ESP flow cytometer (Hialeah, FL). 4.2.7 Cryostat sectioning of spheroids Cryostat sectioning of spheroids has been described in section 3.2.7. The expression of cyclin D under various oxygen concentrations was visually determined. 98 4.3 RESULTS 4.3.1 Cyclin D protein levels decline with increasing depth into a spheroid For the western blots shown in Figure 4.1, these lysates were V79-171b spheroids that were partially trypsinized to create an outer cell and an inner core fraction. We have also used unsorted cells (i.e. whole spheroids). The lysates were each probed for cyclin D and Cdk4. The bottom series of bands indicate the relative amounts of actin found in each sample and serves as our loading control. For the V79-171P cells, the blot was also subjected to the Cdk4 antibody probe after the blot was stripped and re-blotted for the Cdk4 protein. Cdk4 is the active binding partner for Gi cyclins and this protein is also observed to have a similar pattern of increasing protein levels when comparing the different regions within a spheroid. In addition, these spheroids were subjected to varying degrees of oxygen concentrations for 4 hours, as depicted by the percentages of oxygen shown at the top. The levels of the protein also increased with increasing oxygen concentration with the highest level of cyclin D protein found in spheroids subjected to air (i.e. 21% oxygen). Although the spheroids were only gassed at the various oxygen concentrations for 4 hours, this result is quite dramatic. These western blots were done in replicate for reproducibility and to ensure accuracy. 4.3.2 Hypoxic conditions affect the levels of Cyclin D observed in spheroid subpopulations Western blotting analysis gives a macroscopic view of the expression of protein levels; however, we also wanted to validate the results quantitatively using FACS and flow cytometry. Using FACS, we were able to separate out six subpopulations of cells with a depth sort where each fraction represents a shell of cells of equal distance progressing into the spheroid along the spherical axis. To separate cells according to the depth of a 99 0% 5% AIR uns out in uns out in uns out in * Y * * """"^ | Cyclin D mmm mmm* g p g i Cdk4 i d ^ M M ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ . ^ • • ^ F ^ ^ ^ w ^ ^ ^ ^ ^ ^ ^ v ^ M ^ p V * > Aciin Figure 4.1 Western Blots for Cyclin D and Cdk4. Protein levels of Cyclin D and CKD4 in V79-171b spheroids that were partially trypsinized and subjected to varying ranges of oxygen. 100 cell within a spheroid, we utilized the diffusion gradient of Hoechst 33342, where cells that bind highly with Hoechst 33342 fluoresce brightly especially when compared to those cells near the centre of the spheroid. In Figure 4.2, the graphs are organized according to sort fraction where fraction 1 is the brightest (closest to the periphery of the spheroid) and fraction 6 is the dimmest (closest to the centre of the spheroid). In the flow cytometry analysis, the cells were probed with antibodies against cyclin D in each of the six subpopulations created using cell sorting techniques. The graphs are arranged with the sorted subpopulations from fractions 1 to 6 (bright to dim cells), and each line shows the population of cells subjected to various oxygen mixtures ranging from 0% oxygen to 95% oxygen. The levels of cyclin D were corrected for cellular auto fluorescence and non-specific antibody binding. Our results indicate that cells on the periphery of the spheroid have higher levels of cyclin D and these levels gradually decrease as cells approach the necrotic centre. In addition, the ambient levels of oxygen that the spheroids experienced had an effect on the levels of cyclin D. Higher oxygen concentrations produced higher levels of cyclin D protein. In the cases where the spheroids were subjected to more hypoxic environments, the changes in protein levels within the spheroid did not change as drastically as the changes in more aerobic conditions. Paired Student's T-tests were performed comparing the 0% oxygen curve to 21% oxygen, as well as comparing 0% to 95% oxygen. For the V79 spheroids, the p value between 0% and 21% was extremely low at a value of 7.49 x 10"6 and the p value between 0% and 95% oxygen was 1.83 x 10"5. The p values for SiHa were 0.0447 and 0.0213 respectively and for WiDr cells, the p values were 1.21 x 10*3 and 5.55 x 10"4 respectively. All comparisons in all three cell lines were thus statistically significant with p values less than 0.05. Examples of the raw data (flow cytometry dot plots) are shown in Figure 4.3 where the levels of cyclin D protein were visibly different between 101 V 7 9 S i H a WiDr « 100 H 120 100 80 60 40 -i 20 9 0% Oxygen it 1*. Oxygen A 2% Oxygen 5% Oxygen 10% Oxygen • AIR 0 95% Oxygen X 120 100 80 60 H 40 20 H • 0% Oxygen 1 1%Oxygen 2% Oxygen • 5% Oxygen 10% Oxygen • AIR 95% Oxygen * i 0 1 Sort Fraction (Bright to Dim) Figure 4.2 Flow cytometry analysis of Cyclin D levels in spheroids. Fluorescent antibodies probed against cyclin D were analyzed in samples from different regions of a spheroid, as well as samples that were transiently exposed to a range of oxygen concentrations. 102 B 10 20 30 40 50 60 70 80 90 100 DNA Content --> 30 40 5C 60 70 30 90 10C DNA Content --> 120 10 20 30 40 50 60 70 30 30 100 DNA Content --> 10 70 3C 40 50 60 70 90 90 '0C DNA Content --> 11 " i " 120 Figure 4.3 Flow cytometry dot plots of cyclin D using WiDr spheroids. WiDr spheroids were subjected to either 0% oxygen (A and B) or 21% oxygen (Cand D) and probed with the Alexa 488, a fluorescent secondary antibody for the primary cyclin D antibody. Plots were shown for the fraction of cells closest to the interior of the spheroid (A and Q and for those on the periphery of the spheroid (B and D). 103 spheroids grown in higher concentrations of oxygen. Most notably, 0% oxygen resulted in very minor changes in cyclin levels even when comparing the cells in the periphery compared to those cells closest to the centre. Each series of experiments was done in triplicate for verification and validation. 4.3.3 Cell cycle distribution of the cells within the subpopulations of the spheroid To obtain an estimate of the proliferative status of the cells in the spheroids, we calculated the percentage of cells undergoing the S-phase of the cell cycle in each cell subpopulation for all three of our cell lines. This procedure was similar to that described in section 3.3.3. The resulting data were also very similar to the graph detailed in Figure 3.6 and due to the close similarities, we have opted to not show the additional figure. 4.3.4 Cryosections of the WiDr Spheroids for spatial determination of Cyclin D under hypoxic treatment Immunohistochemistry was used to determine visually where the proteins and cells were distributed within an entire spheroid. Only WiDr cells were used to demonstrate the patterns shown and how various oxygen concentrations would alter the cyclin D protein patterns (Figure 4.4). Spheroids were subjected to the fluorescent Hoechst 33342 DNA-binding dye to visualize the decreasing fluorescent gradient with more depth into the spheroid. After sectioning, the slides were counterstained with a fluorescently labelled antibody staining for the presence of the cyclin D protein. Panel A serves as our control of just imaging the sections with the Hoechst fluorescence to show that a fluorescent red counterstain of the cyclin D protein does not mar the images for Hoechst staining. 104 Panels B to F show sections of spheroids that were subjected to 0%, 2%, 5%, 21% (air), and 95% oxygen for four hours prior to freezing in the mounting compound as described in "Materials and Methods". For each panel, 10 cryosections were made and placed on slides. Each slide contained a section from frozen block of the OCT compound which consisted of a population of spheroids. The images presented here were indicative of the majority of the sections. These images show how Hoechst 33342 binds to DNA regardless of the surrounding microenvironment, however, the levels of cyclin D detected by the antibody revealed changes. For panel 6, spheroids were treated with 0% oxygen, and the cyclin D proteins were barely noticeable anywhere on the section. On very close inspection, there is one very small spot of red staining in the upper right hand side of the spheroid. When the oxygen levels were increased to 2% and 5% (Figure 4.4C and D respectively), cyclin D proteins were visualized on the spheroid sections. Where 2% oxygen levels resulted in cyclin D protein found exclusively in the peripheral cells, an increase to 5% oxygen resulted in more cyclin D protein detection deeper into the spheroid. This protein level and increased intensity continued for the spheroids subjected to 21% and 95% oxygen (Figure 4.4E and F). Even though the difference in the oxygen levels of 21% and 95% is a large range, the sections indicate that there was a similar distribution of cyclin D under both conditions and the additional oxygen at this point didn't seem to further increase the levels, and might even have resulted in a slight decrease. 4.3.5 Hypoxic conditions minimally affect the levels of Cyclin D observed in single cell suspensions As shown in Figure 4.2, cyclin D levels respond rapidly to oxygen deprivation whilst 105 Figure 4.4 Cryosections of WiDr spheroids. WiDr spheroids were subjected to hypoxic conditions, 0% and 2% (S and C respectively) as well as higher concentrations of oxygen, 5%, 21%, and 95% (D-F) prior to sectioning. Hoechst (blue) was added to observe the diffusion properties and sections were also probed with antibodies for cyclin D (red). The images displayed are an overlay of the two colours using a magnification of 100X. As a control, Hoechst alone (A) was also imaged with both fluorescent filters. 106 Figure 4.5 Time course of oxygen modification of the cyclin D levels of single cells. 107 the oxygen dependence of single cell suspensions does not have such a strong relationship. Figure 4.5 displays the cyclin D levels observed in each cell line over a course of five timepoints. Each timepoint represents the length of time that a sample of single cells was gassed with various oxygen concentrations. The five timepoints were taken at Vz hour, 1, 2, 3, and 4 hours. Additionally, the oxygen mixtures spanned from hypoxic to oxic gas mixtures, from 1% oxygen, to 21% (air), and 95% oxygen. We observed that over a course of four hours, the levels of cyclin D remained relatively similar, with some decrease at the later timepoints. However, the patterns differed from the results shown with the spheroids, in that at most, the cyclin D levels in single cells decreased only 10% in hypoxic conditions of 1% oxygen. We expected that the oxygen concentration would have little or no effect on the cells for the 21% and 95% oxygen, and for the 1% oxygen, we would expect to see lowered levels of cyclin D produced, especially at the longer timepoints. This was not the case, however, and this stimulated interest in the possibility that cyclins are candidates as a reversible marker of proliferation. Over a course of 4 hours, single cells that were exposed to 1%, 21%, or 95% oxygen didn't have any significant changes in cyclin D levels. This trend was observed in all three cell lines. 108 4.4 DISCUSSION The proliferation rate of a tumour is a determining factor for conventional therapeutic strategies since more rapidly growing tumours have been shown to have an increased sensitivity to many anticancer agents (Arguello et al., 1998). Additionally, one major problem in chemotherapy treatments involves the development of drug resistance, and this resistance is not uncommon in solid tumours with low proliferation rates. Antitumour agents, such as taxol, highly depend on cell proliferation and would not be effective on non-proliferating cells. Those quiescent cells would survive treatment and possibly repopulate the tumor mass (Drewinko et al., 1981). Understanding the mechanisms underlying tumour cell proliferation has been a long standing challenge and in this study, our objective was to determine critical cell cycle machinery proteins and how the microenvironment could influence them utilizing multicellular spheroids as our model. Multicellular spheroids are useful in vitro models with similar complexity to in vivo solid tumours. Their three-dimensional structure and well-defined geometry make them ideal candidates for our studies that involved the differentiation of regional populations to mark differences in the levels of cell cycle proteins, namely cyclin D. Cyclins, the regulatory subunits of cyclin-dependent kinases (Cdks), control the passage of proliferating cells through checkpoints in the cell cycle (Dirks and Rutka, 1997; Hall and Peters, 1996; Sherr, 1995). Cyclin D is under the G i cyclin classification, which is involved in promoting the progression of cells from the Gi to S-phase of the cell cycle. For active kinase activity, cyclin D must be bound to its respective Cdk, Cdk4. Consistent with our previous studies, the results presented here support the use of spheroids for multi-fractionated subpopulations where we used the advantage of a controlled environment for specific manipulations. Our studies, however, clearly suggest the critical influence of oxygen on cyclin D levels, and ultimately, on cell proliferation. Surprisingly, we demonstrated how this effect seemed to be unique to the spheroids 109 because oxygen deprivation studies with single cells revealed very little change in cyclin D levels for the same cell lines. The role of hypoxia in cancer has been widely studied revealing that hypoxic tumours were highly radioresistant creating fundamental problems for therapy (Ljungkvist et al., 2006). It has been well recognized that patients should be pre-screened for hypoxia in an effort to pre-distinguish those tumours that would be radioresistant and allow oncologists to pursue other avenues of treatment at earlier stages of tumour progression. Although the result of radioresistance was established, differing reports on the mechanisms have surfaced suggesting that hypoxic tumour cells promote malignant progression by secretion of growth factors that may promote angiogenesis, metastasis, or cell proliferation (Hockel and Vaupel, 2001; Rofstad, 2000). We demonstrated how exposure of the spheroids to transient levels of external hypoxia induced an extremely large differential in the levels of cyclin D (Figure 4.1). Western blot analysis using whole spheroids or those separated into two cell subpopulations based on their regional location in the spheroid produced an obvious indication that the outer cells of the spheroid had much greater levels of cyclin D compared to the inner population or compared to the spheroid as a whole. When creating these subpopulations, the inner and outer fractions were partially trypsinized from the same sample and accordingly, both these samples had fewer cells than the sample trypsinized with the entire spheroid. However, the protein concentration of each sample was tested and each lane of the western blot had the same concentration of protein, as verified with the actin protein levels. The whole spheroids were intermediate in cyclin levels between the inner and outer fractions of the spheroid and we rationalize this to further indicate that the outer cells had higher levels of cyclin D because the unsorted sample was a mixture of inner and outer cells. The western blots showed three sets of data where the spheroids were gassed for 4 hours prior to disaggregation with 0%, 5% or 21% oxygen. Atmospheric air 110 composed of 21% oxygen served as the control and comparison for this series of experiments, with the objective to determine whether short exposures of hypoxia would affect the cyclin D levels in each region of the spheroid. Higher levels of external oxygen were associated with greater amounts of cyclin D protein. Even though the spheroids were only transiently exposed to the oxygen mixtures, the short time frame of 4 hours was sufficient to produce the differences we observed in Figure 4.1. The drastic differences observed in the western blots encouraged further, more in-depth experiments to be performed to fully determine the effect of oxygen. The combination of FACS and flow cytometry provided a method to selectively distinguish populations within spheroids and generate quantitative results on the cyclin D levels in each subpopulation. Detailed flow cytometry data supported the western blots, where we expanded the range of oxygenation levels starting at hypoxic levels of 0%, 1%, 2%, and 5%, to 10%, 21% and 95% oxygen (Figure 4.2). The seven curves on that graph revealed a definitive trend where more cyclin D was observed in spheroids gassed with higher oxygen concentrations in addition to the trend where the peripheral cells of the spheroid also had higher cyclin D levels. Of particular interest was the flask of spheroids gassed with 0% oxygen, because even the cells on the periphery of the spheroids had levels of cyclin D similar to the inner cells of the same spheroids, which was also very close to the levels for the inner cells of spheroids exposed to 21% oxygen. It appears that even transient exposures of spheroids to hypoxia had a very intense effect on the levels of cyclin D. We have previously reported that hypoxia similarly created extreme decreases in cyclin B1 levels that ultimately had effects on the cell cycle. Decreased levels of cyclin D would also have a large impact on the cell cycle at the Gi checkpoint, where the hypoxia driven decline in cyclin protein may result in cells entering the quiescence (G0) stage of the cell cycle. The regulation of cyclin D is largely co-dependent on its cyclin dependent kinase, Cdk4. The levels for Cdk4 also showed a dependence on the spatial location of 111 the cells as well as the external oxygen levels (Figure 4.1). Their patterns followed those of the cyclin D indicating that these differences were seen for the active protein complex of cyclin D and Cdk4. Another key component of the cyclin and Cdk interaction involves the cyclin dependent kinase inhibitors (Cki). Cki species are increasingly being considered as novel anticancer agents due to their inherent ability to induce growth arrest or apoptosis in tumour cells. Overexpression of one such Cki, p27K,p, induced autophagic cell death, but not apoptosis of tumour cells (Komata et al., 2003). p27Kip plays a central role in the negative control of cell growth. Autophagic cell death involves the total destruction of the cell and is one of the main types of programmed cell death and has been shown in yeast to be controlled by the Tor kinase (Klionsky, and Emr, 2000) (more details on the Tor protein in Chapter 5 and 6). With the observation that hypoxia had a significant effect on the spheroids, we hypothesized that the same cell lines as single cells would also be affected by hypoxia, and most likely within faster time frames. The nutrient medium and gassing conditions were identical to that with the spheroid experiments, with the only change being that the cells were single cells from monolayer culture as opposed to the spheroids that we have been using for prior experiments. The single cells were tested at 5 time points; starting at V- hour, to hourly time points up to 4 hours of continuous external exposure to oxygen mixtures at 1%, 21%, and 95% oxygen (Figure 4.4). Interestingly, not only did we not observe a faster response to hypoxia with the single cells, we did not observe the dramatic difference in cyclin D levels seen in the spheroids even at the 4 hour time point. In fact, there wasn't any significant difference seen in varying the oxygen tension, between time points, or across cell lines. The greatest variations in the values due to changing the oxygen concentration were observed at the 4 hour time frame, however, even these values were within 5 to 10% of one another. The effects of hypoxia on single cells might require a longer incubation time with the lowered oxygen tension to be able 112 to quantify larger differences in cyclin D levels, since we did observe a very slight decrease at our longest time point. The other interesting observation from these data was the fact that in cells grown as 3-dimensional multicellular spheroids, there were dramatic differences in the cyclin D levels and an incubation time of 4 hours was enough to generate this effect. However, in the same time frame of 4 hours, the levels of cyclin D in single cells remained relatively unchanged, suggesting that the structural properties of the spheroid cells, either the contact and/or aggregation of the cells, changed the cellular dependence on oxygen. There were also some technical differences in the experiments comparing the monolayers with spheroids, where monolayers were trypsinized immediately prior to hypoxia exposure whereas the spheroids were exposed to trypsin after the hypoxia exposure. Even though we used a low concentration of 0.1% trypsin, this concentration was sufficient to dislodge adherent cells from the bottom of cell culture plates as well as disaggregate spheroids back into single cells. This enzyme might have physiological effects on the cells as well, even if only to shock the cells in the new environment, and this cellular stress could cause changes in the cell cycle. This was even more plausible since the cells were under hypoxic conditions for only a transient period of 4 hours, yet the differences between the spheroids and the single cells were very strongly evident. Since spheroids are close in geometry to nodular areas of solid tumours, we anticipate that the effects seen in solid tumours would resemble that of the spheroids. These results suggested that cyclin D may be a candidate for a reversible marker of proliferation that has the ability to react and be responsive to the conditions of the environment. For spheroids, the trends in the results would then further suggest that the differences shown were an effect of the treatment, i.e. hypoxia, and could'be manipulated for diagnostic value. 113 In summary, the present findings indicate the importance of the cyclin D protein in relation to both its differential expression found in spheroids versus monolayers, as well as the subsequent alterations found in cyclin D levels under the influence of hypoxia. Such findings suggest that the conformation of spheroids, which highly resembles the cellular organization of a tumour, provided an environment where the regional placement of a cell within the spheroid changed that cell's production and activation of cyclin D. In addition, transient exposure to hypoxia further increased this differential expression of cyclin D, even in relatively short time frames of 4 hours. While we expected the impact of hypoxia to be even greater and observed in shorter timepoints for single cells of the same tumourgenic lines, that expectation proved to be incorrect. Over a course of 4 hours under hypoxic, normal and hyperoxic conditions, 5 time points revealed a pattern that was not significantly different from one another. In this regard, it could be a possibility that single cells needed more time under different oxygen tensions to generate an effect. In any event, the dramatic increases in cyclin D level were not observed at the 4 hour timepoint in single cells but were shown repeatedly in spheroids. We cannot discount the other microenvironmental factors that could be simultaneously contributing to the observed effects on the cyclin proteins. In particular, glucose levels have been shown to be affected by the presence of hypoxia (Ling and Sutherland, 1987) and lowered glucose levels could create a synergistic effect on the cell cycle proteins. Glucose, an initial substrate for glycolysis and the hexose monophosphate pathway, is one major exogenous substrate known to affect the availability of reducing equivalents. In solid tumours, glucose concentration may be low due to the poorly organized vasculature which gives rise to inefficient blood flow and poor tissue oxygenation (Ling and Sutherland, 1987). In addition to oxygen, concentrations of glucose, lactate and pH are known to vary among tumours and probably spatially and temporally in the same tumour (Thomlinson and Gray, 1955; Tannock, 1968; Streffer et al., 1980). In addition, 114 variable glucose concentrations have been demonstrated within histological sections of tumours using a bioluminescence assay (Walenta and Mueller-Klieser, 1987). With our data, the spheroids were exposed to decreased oxygen concentrations for only 4 hours, a time frame unlikely to significantly affect the levels of available glucose in the culture medium. As a G-i cyclin, cyclin D is activated when bound to Cdk4, at the Gi phase checkpoint prior to entry into the S-phase for replication. Without the activation of these Gi cyclin proteins, cells enter quiescence, or G 0 phase, and proliferation ceases. This process in normal tissues is highly regulated to keep cells at an optimal equilibrium where there would essentially be a feedback loop with positive and negative regulators to keep cells under controlled growth. In a solid tumour system, this feedback loop is not tightly regulated resulting in the overproduction and uncontrolled renewal of cells. It has been a continual goal to try to control this mechanism for each tumour system if not to eliminate the tumour, then at least to restrain it from further growth. In this chapter, our data suggest that cyclin D levels are greatly affected by hypoxia, thus if therapy is targeting the proliferating, anoxic cells, those cells may be eliminated, but patients are left with a portion of the tumour that was not initially responsive to therapy. After treatment, the residual tumour cells would have greater access to nutrients and oxygen, allowing those once-quiescent tumour cells to repopulate and metastasize to secondary locations, causing further problems for the patient. For this reason, there is an obvious interest in cell cycle inhibitors as an adjuvant to conventional therapies. We introduce rapamycin in the upcoming chapters as a chemical agent that has a high activity for quiescent cells as well as causing Gi arrest by inhibiting active cyclin D protein. This agent could have profound effectiveness on tumours since it is targeting a population of cells that would normally be resistant to therapy as well as being a cell cycle inhibitor. The work with rapamycin will be further 115 described in detail in the following two chapters using both an in vitro and in vivo model system. In theory, rapamycin has the potential to inhibit the cell cycle and promote cells to enter quiescence, resulting in a population of cells that are more sensitive to the toxicity of rapamycin itself. Similar to cyclin B1 levels as described in Chapter 3, cyclin D levels were responsive to the environmental status of a cell, making the levels of this protein also a potential predictor for hypoxia. The combination of determining the relationship of hypoxia and both cyclin proteins, and it's applicable use as a predictive assay, supported the initiative to investigate cell cycle inhibitors as an additional therapeutic modality. 116 C H A P T E R 5 : E F F E C T S O F R A P A M Y C I N O N C Y C L I N D L E V E L S IN VITRO 117 5.1 INTRODUCTION Inhibition of the cell cycle could be very beneficial for therapeutic purposes, especially if it can be shown to have an additive effect with current established therapeutic regimens. We introduced our drug of choice, rapamycin (Figure 1.12), to spheroids to determine the effect it would have on the cyclin D levels in each region of the model. Rapamycin, a natural product derived from bacteria, was first tested on yeast genetic screens and was found to induce G^ arrest in some cell types, through inhibiting active cyclin D (Heitman et al., 2002). A serine/threonine kinase named target of rapamycin (TOR) was identified (Figure 5.1), and shown to be a member of the phosphatidylinositol 3-kinase (PIK3) related family of kinases (Heitman et al., 2002). Rapamycin inhibits mTOR (mammalian TOR) by initially binding to the immunophilin FK506 binding protein (FKBP12). This bound FKBP12/rapamycin complex would then bind to mTOR, preventing the phosphorylation of downstream targets such as the S6 kinase (S6K) for cell cycle regulation and 4EBP1 for translational control (Abraham, 2002; Schmelzle and Hall, 2000; Shamji et al., 2003). Just recently, using in vitro mouse mammary tumour cell lines and a derivative of rapamycin, RAD001, it was also reported that rapamycin-induced antiproliferative effects correlated with the down-regulation of cellular p21 levels and the levels of p21 in Cdk2 and Cdk4 complexes (Law et al., 2006). This report by Law et al. showed that cyclin D1/Cdk2 complexes are inactivated by rapamycin treatment in a relatively selective manner through p21 down-regulation. They created a cyclin D1-Cdk2 fusion protein that may be effective in reversing rapamycin-mediated cell cycle arrest through the stabilization of p21 levels as observed with oyerexpression of cyclin D1 alone. However, there is the possibility that because of the physical linkage of the cyclin D1 and Cdk2 domains together, the cyclin D1-CKD2 fusion protein does not require p21-dependent assembly proteins (Sherr, 1995; Law et al., 2006). 118 Source: Based on Wullschleger et al., 2006 B Insulin/IGF Growth factor • receptor synthesis, Ribosome Metabolism biogenesis Transcription Autophagy Actin organization Source: Based on Wullschleger et al., 2006 Figure 5.1 mTOR protein. (A) Diagrammatic representation of the mTOR protein, with the domains listed. (6) mTOR signalling pathway Rapamycin, as well as other comparable drugs that target and inhibit the serine/threonine kinase mTOR protein, has garnered considerable attention as a potential new drug treatment for cancer therapy. In the 1970s, the natural products program at the National Cancer Institute had already identified rapamycin as a potential anticancer agent (Douros and Suffness, 1981). However, it wasn't until recently within the last 3 to 4 years, that many groups seriously studied the molecular pathways. In fact, it was just within the past few years that the Tor protein has been identified as two separate complexes, the TOR complex 1 (TORC1) and TOR complex 2 (TORC2) (Sawyers, 2003; Wullschleger et al., 2006). Each complex comprises of a cluster of proteins with two key identifying differences between TORC1 and TORC2. TORC1 comprises either the TOR1 or TOR2 protein with the raptor protein whereas TORC2 contains only the TOR2 protein and is bound to the rictor protein (Wullschleger et al., 2006). Most importantly, TORC1 is found to be rapamycin sensitive whereas TORC2 is thought to not be inhibited by this drug. TORC1 was identified first (Loewith et al., 2002; Reinke et al., 2002) and associates with a large 150kDa protein, raptor (Hara et al., 2002; Kim et al., 2002). It is still unclear which protein domains interact between raptor and mTOR with mapping results indicating that there might be multiple contact or binding sites for the two proteins. The role of raptor and its involvement in rapamycin sensitivity is also unclear, where some groups are suggesting that raptor serves as an adaptor to recruit substrates to mTOR (Choi et al., 2003, Hara et al., 2002; Nojima et al., 2003; Schalm et al., 2003) but others have also reported that upstream signals are involved to regulate the raptor-mTOR interaction and subsequently the activity of mTORCI (Kim et al., 2002). The mechanism of how rapamycin inhibits mTORCI is not fully understood but Kim et al. have suggested that the binding of rapamycin to FKBP12 dissociates the raptor-mTOR affinity and blocks the access to substrates. 120 mT0RC2 differs from mTORCI by binding to rictor, a protein approximately 200kDa in size. Not much is known about this complex, however two groups have shown that mTORC2 is neither bound by FKBP12-rapamycin nor does FKBP12-rapamycin affect mTORC2 in vitro kinase activity (Jacinto et al., 2004; Sarbassov et al., 2004). This observation indicates that the effects of rapamycin are primarily affecting mTORCI and its subsequent process with a large component of this difference due to the binding of the either the rictor or raptor protein. As shown in Figure 5.1 (adapted from Wullschleger et al., 2006), although it appears that rapamycin has effects only on mTORCI, mTORC2 still signals to other cell cycle proteins which ultimately are affected by external stresses such as hypoxia; stresses that have effects on proteins upstream of both mTOR complexes. Thus even though rapamycin may not directly interact with mTORC2, both mTORCI and mTORC2 send signals downstream to activate the Akt/PKB pathway, and if rapamycin had inhibitory effects on mTORCI, then Akt/PKB activity should also decrease, leading to decreased inhibition of the TSC1/TSC2 complex, thus increasing inhibition on the Rheb protein, and ultimately lowering positive signalling to activate both the mTOR complexes once again. Rapamycin also has effects on cyclin D as well as other cell cycle regulatory proteins (Sawyers, 2003). Treatment with rapamycin using spheroids revealed that the drug does have an effect on the cells in all populations of the spheroid. Cells on the periphery of the spheroid showed a decrease in cyclin D levels compared to the control, and cells towards the interior of the spheroid maintained a low level of cyclin D similar to the control. Previous studies show that cell lines derived from different cancer types were noted to undergo G^ arrest when exposed to 1 nM rapamycin, a concentration which closely matches that required for biochemical inhibition of mTOR in cells. Some cell lines that failed to respond to the 1 nM dose did undergo growth arrest at significantly higher concentrations (-1000 nM) (Sawyers, 2003). While these studies revealed the 121 broad potential for rapamycin as an antiproliferative agent, the mechanism is still unclear. Nonetheless, the results define two groups of rapamycin-sensitive mTOR complexes, the TOR complex 1 (TORC1) also referred to as mTOR/raptor, as well as TOR complex 2 (TORC2) or mTOR/rictor (Wullschleger et al., 2006). This separation of classes was established because mTOR/raptor is very sensitive to rapamycin, whereas, mTOR/rictor is not. In this context, we will show that low dose treatment with rapamycin affected the cyclin D levels via the mTOR/raptor pathways and higher doses resulted in even lower levels of cyclin D, an indication that both the mTOR/raptor and the mTOR/rictor pathways were inhibited. Given these considerations, attempts to identify the effect of rapamycin on spheroids subjected to hypoxia are clearly warranted. 122 5.2 MATERIALS AND METHODS 5.2.1 Spheroids and reagents The preparation of spheroids was similar to the methods already described in Section 2.2.1. Rapamycin (Sigma Chemical Company, St. Louis, MO) was purchased as a fine grade powder that was initially dissolved in 100% DMSO (dimethyl sulfoxide) to make a 1 mg/ml stock solution. Further dilutions of the rapamycin were created using a 60% DMSO solution as the solvent. Stock solutions were stored in UV light protected amber vials at 4°C, while diluted samples were made fresh for each experiment and also stored in these amber vials during the course of the experiment. 5.2.2 Toxicity, dose response, and time course of Rapamycin on spheroids Rapamycin dilutions were made from a 1 mg/ml stock solution as described in section 5.2.1. For all studies, rapamycin was directly inoculated into the growth media of the spheroids and/or single cells for the duration of the experiment. The cells were in spinner flasks and constantly suspended and stirred in this medium. Upon completion, the cells were washed twice with MEM containing 10% FBS prior to further manipulations. For time point experiments with multiple sampling, spinner flasks were quickly sampled and resealed, and data gathered at each designated time interval. The toxicity experiments were survival assays using methods well established in many laboratories including our own where predetermined numbers of cells were collected in test tubes containing culture medium; these tubes were then poured and rinsed into conventional Petri dishes. We did clonogenic survival assays over a time course where plates with equal numbers of cells were exposed to rapamycin for up to 24 hours. Sample measurements were taken at 0, 1, 4, 8, and 24 hours of drug contact with the cells, washed and incubated at 37°C for 2-3 weeks for colony formation. 123 5.2.3 Fluorescence activated cell sorting FACS was performed in a similar manner to that described in Section 3.2.5. For the rapamycin in vitro studies, cells were sorted in two different stages. Some experiments were done similar to Section 3.2.5 and 4.2.5, where all manipulations on the spheroids were done prior to sorting. However, in some of our studies, we also initially separated out cells in regions within spheroids using the Hoechst 333423 diffusion gradient to generate two populations; a bright and a dim sample, which were then subsequently transported back into the spinner flasks to be treated with rapamycin. 5.2.4 Flow cytometry with antibodies Single cell samples were prepared for flow cytometry analysis as described in Section 3.2.6. 5.2.5 Cytospin slides of cells treated with Rapamycin The cells used in the cytospins were processed similarly to the single cells that were probed with cyclin D antibodies and analyzed with flow cytometry. However, instead of using the flow cytometer, the cells were observed on slides. To do this, WiDr single cells were placed in water-jacketed spinner flasks and gassed for various timepoints starting at no gas contact up to 24 hours of gassing with 21% oxygen (air). At the desired timepoint, the cells were removed from the flask and immediately placed on ice to be centrifuged. They were centrifuged at 1230 rpm for 10 minutes, aspirated, and fixed with ice cold 66% ethanol to a final concentration of 2 x 105 cells/ml. After storing the samples in the -20°C freezer for a minimum of 1/2 hour, 100 pi of the cells were added to the cytospin apparatus. This apparatus consists of a plastic funnel, filter paper and a glass slide that are clamped tightly together in that order. The funnel is the input location of the samples and the narrow end of this funnel channels the sample onto the slide in a 124 circular shape. The filter paper between the funnel and the glass slide serves to absorb excess solution and creates a perfectly formed circle of cells. The cytospin apparatus is placed inside the cytospin spin tray and centrifuged at 400rpm for 8 minutes. After cells had been spun onto the slides, the slides were dried briefly and the primary antibody solution against cyclin D at a concentration of 1:100 was added and left for an overnight incubation at 4°C. Before the addition of the secondary antibody, Alexa 488 at a 1:100 concentration, the slides were washed three times with PBS by careful immersion of the slides into a glass slide jar. The secondary antibody was added to the samples at room temperature for 2 hours prior to another series of two PBS washes for one minute, and a final PBS wash for 30 minutes, gently rocking on a benchtop shaker. DAPI was then added at a concentration of 0.05 ug/ml for 5 minutes at room temperature, another 5 minute rinse in PBS, and a final 20 second dip into 2% paraformaldehyde. Slides were allowed to drain fully and a drop of Fluorogard® was added on top of the samples and covered with a sealed cover slip. All slides were subsequently viewed using a florescence microscope for the various fluorescently-tagged antibodies and stored at 4°C. 5.2.6 Levels of cyclin D FACS analyzed prior to rapamycin treatment This method was the reverse of our normal FACS procedure with spheroids where for this series of experiments, spheroids were first FACS sorted to generate three subpopulations; an "all sort" population, a bright (outer) cell population, and a dim (inner) cell population based on a 6 fraction depth sort. After passing through the cell sorter, each subpopulation was then put back into the experimental water-jacketed spinner flasks filled with 50 ml nutrient media and inoculated with either 0.2 ug/ml or 0.05 ug/ml rapamycin. The flasks were gassed with either 0% oxygen or 21% oxygen, where the first series of experiments show all three subpopulations gassed with 21% oxygen, but in 125 the second series, the all sort and outer cells were gassed with air, and the inner cells were gassed with 0% oxygen. After four hours of rapamycin contact, where the single cells were continuously stirred in the spinner flasks to prohibit the cells from adhering to any surface in the spinner flask, each sample was collected in 50 ml Falcon tubes (VWR, Mississauga, ON) and centrifuged for 10 minutes at 1230 rpm. After gently aspirating the supernatant, cells were resuspended in 4 ml of MEM containing 10% FBS and transferred to 4 ml Falcon tubes (VWR, Missisauga, ON). All samples were fixed and stored at -20°C prior to flow cytometry analysis. 126 5.3 R E S U L T S 5.3.1 Rapamycin toxicity in cells Our initial study with rapamycin was to investigate the levels of toxicity of the drug when administered to our cells in vitro (Figure 5.2). The range of doses had a span of 20 fold where colony formation was most affected when cells were exposed to the drug for longer durations, and increased doses of rapamycin resulted in the most inhibition of colony formation. However, our initial purpose for rapamycin was to use this agent as a cytostatic agent in conjunction with other forms of inhibition (or therapy), thus we were more interested in intermediate doses at shorter time intervals. The data presented on this figure were the means of two repeated experiments. 5.3.2 Influence of rapamycin on cyclin D levels in spheroids Using the same range of doses as shown in Figure 5.3, we observed the effects of the drug on the levels of cyclin D in the various regions within a spheroid. The spheroids were fluorescently cell sorted based on the diffusion of Hoechst 33342 in the spheroids, and each flask was subjected to the rapamycin for a period of 4 hours. These experiments were repeated in triplicate and the results in this figure indicate the means of three repeated measurements. The cells near the interior of the spheroid had low levels of cyclin D and this effect could have been observed either because the cyclin levels in the centre of spheroids were low, or due to rapamycin in addition to the microenvironmental stresses occurring near the necrotic centre. For the fraction of cells on the periphery of each spheroid sample, increasing the dosage of rapamycin resulted in lower cyclin D protein levels. With increasing depth into the spheroid, the levels of cyclin D were lower compared to those near the periphery and this decrease was seen for all doses. Once again, there could be two explanations for this trend, either the drug was mostly taken up by the outer cells in the spheroid system or the levels of cyclin D 127 A B i i i i i i i r 10 12 14 16 18 20 22 24 #ofhburs l — i — i — i — i — i — i — r 10 12 14 16 18 20 22 24 #of Hours Figure 5.2 Toxicity of Rapamycin on human tumour cells. SiHa (A) and WiDr (S) cells tested for the rapamycin toxicity on growth at a range of treatment doses. 128 SiHa WiDr 180-170-~ 150-§ 140-£ 1 3 0 -Q 120-110-1 60 -| 50 -4 0 -• Cbrtrd • 0.05|jc / iTi-Rap3rrvcin 0.1 ijg'rrL Raparn/an T 0.2 |jo/rrL Rgparrya'n 0.5 ue/rrt Raparrydn 1.0 ugrr t Rgparrya'n 180 170 160 150 140 130 120 110 100 90 80 70 60 50 40 • Central » 0.05|jg /rTtRaparrvdn * 0.1 ^ r r t Raparrvdn • 0.2pg /nrLRaparrvdn - * - O.Spg'rrLRaparrycin • lOMC/rrLRaparrvdn S o r t F r a c t i o n ( B r i g h t t o D i m ) Figure 5.3 Cyclin D levels with varying doses of rapamycin treatment. Using the range of doses similar to Figure 5.2, levels of cyclin D protein were evaluated using FACS and flow cytometry methods to observe the effects of the drug on cyclin production after 4 hours of incubation. 129 were suppressed under a more stressful environment. Regardless, all curves compared to the control curve were observed to be lower indicating that there was a real effect of the rapamycin on the cells; what remains to be determined was the order in which the processes occurred. This will be described in more detail in section 5.3.5. 5.3.3 Duration of Rapamycin contact influences cyclin D activity From the observations shown in Figure 5.2, both the dose of rapamycin and the duration of the drug contact suggested an effect on the cells resulting in a decreased survival as assessed by colony formation. Measuring the levels of cyclin D in each region of the spheroid using a range of doses (as described in the previous section) examined the dose dependence of cyclin D, but we wanted to observe the time dependence as well. From Figure 5.3, we observed that SiHa and WiDr cells exposed to rapamycin doses higher than 0.2 ug/ml were showing suppressed levels of cyclin Dr, thus we chose 0.2 ug/ml as the constant dose for the time course studies. The range of time points varied from 0 hours up to 24 hours (Figure 5.4). "0 hours" represents rapamycin at a concentration of 0.2 ug/ml being added to the growth medium containing the spheroids, only to be immediately washed off with MEM containing 10% FBS, and the spheroids then prepared for FACS analysis. The control curve for Figure 5.4 served as a comparison to the experimental curves where these spheroids had identical treatment as the experimental flasks but without any addition of rapamycin into the nutrient media. For both SiHa and WiDr cells the cell cycle time is approximately 24 hours, thus over the period of this 24 hour time course, the drug would have been in contact with every cell for at least one cell division, unless there were hindrances that created cell cycle delays. Interestingly, a longer duration of rapamycin contact did not significantly alter the levels of cyclin D when compared to the shorter timepoints, especially noted for the WiDr cell line. Even more 130 SiHa WiDr 60-50-40 30-©- Control i 0.2 (jg'rrL RapaTya'n fa Ohrs A 0.2(acynrtr^ parrvdnfor4hrs v 0.2|jcyiTLFfeparT>dnfa8frs 0.2 ng/rrL Ffeparryan fa 16hrs 0.2 po/rrL Raparrydn fa 24hrs 0 180 170-160-150 140 130 120 110-1 100 90H 80 70 60 50 40 30 - 0 - Cairo! 0.2|^rTtRaparrva'nfa0hrs O^pg^rrtRaparrydnfa^ v • 0.2^rrLRspan^nfa8rrs • O^^rTLRaparryrinfaiehrs - • - 0.2|^irLRapaiTyanfa24rrs -HJ-0 S o r t F r a c t i o n ( B r i g h t t o D i m ) Figure 5.4 Levels of Cyclin D after treatment with 0.2ug/ml Rapamycin. 131 surprising were the curves for "0 hours" as described already above, because it appears that even an immediate contact of the drug could produce an effect on the cyclin D levels. This observation gives rise to a few possibilities, where perhaps the efficacy of rapamycin on SiHa and WiDr cells is quite potent and immediate, as well as the possibility that a lower dose of rapamycin is sufficient to produce an effect. 5.3.4 Cytospin slides of single cells in contact with rapamycin Figure 5.4 graphically displays the levels of cyclin D after 0.2 ug/ml rapamycin treatment. The distribution of the curves was from a range of 0 to 24 hours, and the results were not significantly different from one another, thus we sought to confirm this result using fluorescent images to visualize this effect. Figure 5.5 shows a multi-panel display of cytospin slides where the cells were exposed to 0.2 ug/ml rapamycin from (S) 0 hours to (C) 4 hours, (D) 8 hours, (£) 16 hours, and (F) 24 hours. Panel (A) was a control sample where the cells had no contact with rapamycin at all and serves as a comparison for panels B to F. The green labelling seen in the images reflects the fluorescent antibody against cyclin D and the blue stain was the fluorescence from the DAPI, a DNA staining agent. In some panels, there was a slightly higher concentration of cells but it would be more important to focus on the staining intensities of each cell which would indicate the amount of cyclin D protein available in each sample. Compared to the control panel (A), the level of cyclin D was slightly less even for 0 hours (Figure 5.56) and further decreased as the time frame of rapamycin contact got longer (Figure 5.5C - F). 5.3.5 Determining the order of rapamycin effects In all the results data generated for rapamycin thus far, we have subjected the spheroids to drug treatment prior to FACS analysis. The rationale behind this was to take 132 A B C D E F Figure 5.5 Cytospin slides of WiDr cells treated with rapamycin. A time course of 0.2 ug/ml rapamycin contact on cells from (S) 0 hours, (C) 4 hours, (D) 8 hours, (£) 16 hours, and (F) 24 hours. Panel (A) was the control where cells were not exposed to rapamycin for any length of time. The green fluorescence was the signal for cyclin D and the blue colour was the DAPI stain for DNA. 133 advantage of the three dimensional properties of the tumour model and simulate the environment found in tumours in vivo. Upon review of Figures 5.3 and 5.4, we noticed a constant low level of cyclin D found in the inner fractions of the spheroids and this could be a consequence of two possibilities, either this was an effect of the drug itself, or perhaps, there wasn't much (if any) drug reaching the inner cells since all of it might have been sequestered by the outer cells and thus, the cyclin D levels would have been low already (from microenvironmental factors) and not altered by rapamycin. By reversing the process of pre-sorting the spheroids prior to drug treatment, we ensured that all populations of cells would come into contact with rapamycin and any differences observed would be a result of the drug itself. Both concentrations of rapamycin were tested for the entire series of the three subpopulations so that there was a comparison within one drug dose treatment series as well as the comparison between two differing doses. For some of our other data presented in this chapter, we chose a concentration of 0.2 ug/ml, and as the comparison, we picked a concentration four times lower due to the results we were generating where lower concentrations appeared to generate an equally effective result. When comparing the populations with varying doses of rapamycin, the p-values between the SiHa control curve and the SiHa cells exposed to 0.05 ug/ml or 0.2 ug/ml were 0.0194 and 0.0162 respectively and the p-values for WiDr cells were 0.0095 and 0.0098 respectively. If we compared the curve of 0.05 ug/ml versus 0.2 ug/ml rapamycin, SiHa cells were statistically significant with a p-value of 0.0237 but WiDr cells were not with a p-value of 0.1068. Levels of cyclin D appear to be mostly significantly different when comparing any combination of rapamycin treatment (Figure 5.6). Sort fraction 1 was the "all sort" where every viable and healthy cell would be collected, fraction 2 was the "bright" or outer cells and fraction 3 reflects the "dim" or inner cells. When we compared the different regions of the spheroid against one another, this generated p-values greater than 0.05 for all combinations (fraction 1 versus 134 A Cyclin D levels in pre-sorted SiHa cells 100 Cyclin D levels in pre-sorted WiDr cells 100 v> 80 d) > Q 60 c o CJ 40 c to a> 20 I I WiDr Control V77X 0.05 >ig/ml Rap 0.2 ug/ml Rap 1 2 3 4 Sort Fraction Figure 5.6 Rapamycin treatment of cells after FACS analysis using 21 % oxygen. The three subpopulations of the spheroid were graphically displayed as sort fractions 1 to 3. Sort fraction 1 was the "all sort", 2 was the "bright" or outer cells, and 3 was the column for "dim" or inner spheroid cells. After sorting, each flask was gassed at 21% oxygen and treated with rapamycin for 4 hours for (A) SiHa and (6) WiDr cells. 135 fraction 2, 1 versus 3, or 2 versus 3) in both the SiHa and WiDr cell lines except when comparing the outside population with the inside population for WiDr cells which had a p-value of 0.0189 (Figure 5.6). Thus, this figure indicated that the concentration of the rapamycin treatment resulted in statistically significant data; however, this there wasn't a statistically significant difference when comparing the three regions within a spheroid. These experiments had controlled all the other parameters including the concentration of oxygen during the four hours of drug treatment. The oxygen concentration used was 21% (i.e. normal air) for all three subpopulations, however, normally in a spheroid (and tumour) environment, the inner cells would be under hypoxic conditions, thus we also tested the same experiment while gassing the inner cells with 0% oxygen, while the outer and all sort cells with 21% oxygen (Figure 5.7). In this scenario, most treatments with rapamycin produced statistically significant differences with p-values less than 0.05 in both the SiHa and WiDr cells. However, there were two exceptions to this generalization, firstly, the comparison of the SiHa control and the SiHa cells exposed to 0.05 ug/ml rapamycin which had a p-value of 0.1204 as well as the comparison of 0.05 ug/ml versus 0.2 ug/ml of rapamycin for the WiDr cells with a p-value of 0.1489. Once again, sort fractions 1 to 3 were the same as that described for Figure 5.6. When we analyzed the p-values for the three regions within the spheroid, most p-values were less than 0.05 (except for WiDr fraction 1 versus 2, p=0.1231) but most importantly in this scenario, we wanted to know if there would be a statistically significant difference between the outside cells exposed to 21% oxygen versus the inside cells exposed to 0% oxygen. The p-value was 0.0412 and 0.0212 for SiHa and WiDr cells respectively indicating that the results were statistically significant in this figure whereas the respective data in Figure 5.6 were not significant. Thus the amount of oxygen exposure to the cells in addition to the concentration of the rapamycin had an effect on the cyclin D levels. When comparing between the two cell lines, WiDr cells appeared to be more 136 A B 100 80 H 60 40 20 Cyclin D levels in pre-sorted SiHa cells 1 2 Sort Fraction I I SiHa Control 2223 0.05 u.g/ml Rap £ s S 3 0.2 ng/ml Rap Cyclin D levels in pre-sorted WiDr cells 100 80 60 40 20 I I WiDr Control V77X 0.05 jig/ml Rap ESS 0.2 u.g/ml Rap Sort Fraction Figure 5.7 Rapamycin treatments on cells after FACS using 0% and 21% oxygen. The three subpopulations of the spheroid were graphically displayed as sort fractions 1 to 3. Sort fraction 1 was the "all sort", 2 was the "bright" or outer cells, and 3 was the "dim" or inner spheroid cells. After sorting, sort fractions 1 and 2 were gassed at 21% oxygen, while sort fractions 3 were gassed with 0% oxygen. All three flasks were treated with rapamycin for 4 hours for (A) SiHa and (S) WiDr cells. 137 sensitive to rapamycin since low concentrations of the drug yielded significant results and increasing the concentration did not create significant differences. 138 5.4 DISCUSSION We clearly demonstrated that the addition of rapamycin from doses ranging from 0.05 ug/ml to 1.0 ug/ml was not very toxic to the cells; however, it did cause growth delays as shown in Figure 5.2. With cells in contact with the drug for 24 hours, or approximately one cell cycle time, even the low doses of 0.05 ug/ml and 0.1 ug/ml led to inhibition of colony growth. With the increase in rapamycin concentration, there was a correlation with the growth delay as it also lowered colony numbers. At our highest concentration of 1.0 ug/ml, there was a 6- or 10-fold decrease for SiHa and WiDr cells respectively. At shorter time points of 4 and 8 hours where the cells would not have even gone through one cell cycle, the changes in colony formation was not at the 10-fold scale, however, there were markedly lower numbers of colonies especially in the WiDr cell line. In reference to Chapters 3 and 4, we had further investigated the effects and influence of the microenvironment on the cell cycle via two critical cell cycle cyclins; cyclin D at the Gi checkpoint, and cyclin B1 as a mitotic protein. Rapamycin is an inhibitor of the serine/threonine kinase mTOR as well as inducing Gi arrest and/or apoptosis and we have shown the effects of rapamycin on the levels of cyclin D (Figure 5.3). Increasing the concentration of rapamycin did have an effect on SiHa and WiDr cells with the suggestion that there might be a saturation concentration at around 0.2 ug/ml, where further increases in dosage kept the levels of cyclin D relatively low. These results were seen to be most dramatic on the peripheral cells of the spheroid. With the colony growth assays (Figure 5.2), there were slight differences of colony numbers at the shorter timepoints of 1 and 4 hours, but there were more dramatic decreases in colony number when the cells were incubated with rapamycin for 8 hours, and especially at 24 hours. With the information provided from Figure 5.2, we also did a time course treatment of rapamycin from 0 hours to 24 hours at a concentration of 0.2 ug/ml. Instead of recording 139 the colony formation (i.e. the longer term, effects of the treatment), we analyzed the levels of cyclin D at each time point in relation to the cell position in the spheroid model. These experiments not only provide more information regarding the cyclin D status in spheroids under rapamycin treatment, but also serve as an estimate of the cell proliferation at an earlier time frame. Whereas colony formation had taken two weeks of incubation before garnering any results, measuring the levels of cyclin D were more immediate and alluded to signals molecularly upstream that contributed to the lack of colony growth. Figure 5.4 details the effects of the 0.2 ug/ml rapamycin time course on both the SiHa and WiDr spheroids. For SiHa cells at this concentration, by 8 hours, there was no difference in the cyclin D levels between the outer and inner cells of the spheroid. WiDr cells were found to be more sensitive to rapamycin because at the same concentration of 0.2 ug/ml at 0 hours, there already was a drop in cyclin D levels which was comparable to the cyclin D level of the inner cells. For SiHa spheroids, the 0 hour curve was not nearly as drastic, however, there still was a slight decrease in the levels of cyclin D. This was surprising to us since we had hypothesized that this time point would most likely mirror that of the control, but in reality, it revealed that a quick pulse of rapamycin (approximately 5 minutes), did have an impact on cells. It needs to be clarified, however, that although the rapamycin was washed off using fresh MEM and 10% FBS, the spheroids were still under incubation in a 37°C water bath for an additional 20 minutes (for the contact of the Hoechst 33342) as well as a 10-15 minute trypsin disaggregation procedure also at 37°C. Thus the total minimum amount of time that these "0 hour" cells have been exposed to the rapamycin is more accurately in the 45 minutes to 1 hour time frame. Since the WiDr cells appeared to be more sensitive, we also imaged cytospin slides using single cells from the WiDr cell line. Compared to the control, the level of green fluorescence (probed for cyclin D) was slightly dimmer for 140 0 hours (Figure 5.56) and further decreased in intensity as the time frame of rapamycin contact increased (Figure 5.5C - F). With the two structures of mTORC we mentioned earlier in this section, these results suggest that we have generated results that reflect the influence of both mTORCI/raptor as well as the cyclin D active complex itself. Those instantaneous results of decreased levels of cyclin D at the 0 hour time point suggest that rapamycin affected the cyclin D active complex itself, where some studies (Hidalgo and Rowinsky, 2000) have shown that the bound complex of rapamycin with FKBP12 stimulates an increase in cyclin D turnover, thus preventing it from binding to Cdk4, its active kinase partner. Lack of colony formation when cells were subjected to longer periods of rapamycin exposure suggest that this would be more an effect of the mTORCI where that cascade of cell cycle regulators were inhibited, such as the S6 kinase and 4E-BP, leading to changes in the cell cycle and cyclin D translation respectively, as well as a host of other mTOR effectors as shown in Figure 5.1. Although many of the mechanisms are not fully understood yet, it has been established that there is another mTOR complex, mTORC2, that is not rapamycin sensitive (Sarbassov et al., 2005). Instead of the raptor protein, mTORC2 binds the rictor protein, and together, there is evidence showing an affect on actin organization as well as the Akt/PKB pathway (Wullschleger et al., 2006). From the present state of knowledge of mTORCI and 2, we speculate that our cyclin D and rapamycin data had effects on mTORCI only. However, yet another interesting observation relates to the effects of exterior stresses and nutrients on this system (Figure 5.1). We have used spheroids for most of our studies, which is a three-dimensional tumour model that embodies variation in oxygen and nutrient availability purely due to the physical nature of the model. Both hypoxia and nutrients have effects on proteins upstream of both the mTOR complexes, 141 thus indirectly through the use of our model system, we simulated effects on mTORC-1 with rapamycin as well as both mTORC-1 and 2, with the microenvironmental stresses. The importance of mTOR in cancer cell biology and clinical therapeutics has progressively been evolving as more studies show the implications of mTOR on cell growth, proliferation, and bioenergetic metabolism (Mita et al., 2003). The discovery of two mTOR complexes now gives rise to the question of whether rapamycin (and its derivatives) will be effective in all cancers or if successful treatment will depend on the extent of PI3K/mTOR activation. With little known about mTORC2, there is the possibility that rapamycin treatment could create an imbalance of the TOR complexes, where functional mTORCI may be blocked, and a surge of activation results from mTORC2. At present, it appears that the Akt/PKB pathway is upregulated by mTORC2, thus rapamycin treatment may result in enhanced Akt activation and enhanced tumour growth and survival (Abraham, 2002; Sarbassov et al., 2004; Hay, 2005) Agents such as rapamycin have demonstrated growth inhibition or tumour regression in some xenograft systems (Owa et al., 2001; Yu et al., 2001), but there has been no attempt to reconcile such observations with cellular analysis. In other words, we do not know which cells specifically responded to the drug and were responsible for the tumour growth inhibition or regression in these systems. Moreover, little is known about the translation of the in vitro mechanism of action of these agents to a more complex in vivo system. By utilizing the in vitro spheroid system, one that is intermediate in complexity between monolayers and xenografts, experiments were performed to investigate the effects of rapamycin within different microenvironmental regions. Our data suggested that rapamycin does have growth inhibitory effects on both the SiHa and WiDr tumour cell lines using the levels of cyclin D as a measure of proliferation. Not only did we determine that rapamycin could be cytostatic at non-cytotoxic concentrations in our culture system, we also determined 0.2 ug/ml to be adequate for a maximal effect. 142 Time course experiments also suggested that rapamycin has a relatively fast mode of action where longer periods of exposure positively correlated with reductions in cyclin D levels up to 4 hours of exposure. Longer exposure times up to 24 hours resulted in little or no additional effect. Our results also indicate that rapamycin lowered the levels of cyclin D even when the drug was given as a "short" pulse (0 hour curve in Figure 5.4), indicating the rapid mode of action and also suggesting that we were measuring the effects of rapamycin acting directly on the cyclin D protein, and not through indirect protein-protein interactions. The objective of this chapter was to target specific aim #4 where we used FACS to differentiate regions within the multicelluar aggregates to determine which cells specifically responded to the drug and were responsible for the tumour growth inhibition or regression in these systems. Figures 5.3 and 5.4 showed that the peripheral cells (composed of mostly proliferating cells) were definitely responding to the drug whereas the inner cells (composed mostly of quiescent cells) were constantly at low cyclin levels leading to ambiguity on the responsiveness of these cells. Either rapamycin was responsible for keeping the cyclin levels low in these subpopulations, or they were potentially not exposed to rapamycin due to drug diffusion limitations within the spheroid model. The aim of the series of experiments graphed on Figures 5.6 and 5.7 was targeting this question where the spheroids were first pre-sorted and then subjected to drug exposure. The conclusion from these studies revealed that rapamycin indeed does target both the proliferating and the quiescent populations and the additional microenvironmental stress of hypoxic conditions further lowered the observed levels of cyclin D protein. 143 C H A P T E R 6: R A P A M Y C I N U P T A K E IN VIVO A N D U S E S F O R C O M B I N A T I O N T H E R A P Y W I T H R A D I A T I O N 144 6.1 INTRODUCTION The stem cell model of cancer suggests that solid tumours contain clonogenic cells that are capable of differentiation or self-renewal, much like hematopoietic stem cells in bone marrow (Bush and Hill, 1975; Mackillop ef al., 1983). Thus it is the clonogenic cell component of a tumour that is the ultimate therapeutic target since it is these cells that are capable of driving tumour repopulation and regrowth after therapy. There is some debate, however, regarding the cells that can be classified as clonogenic in solid tumours. Many researchers believe that a relatively small fraction (on the order of 0.1%) of cells in a clinical solid tumour is clonogenic. The wide heterogeneity in the clonogenic fraction of mouse tumours, however, has led others to believe that the majority of tumour cells have the potential to become clonogenic if provided with adequate microenvironmental stimuli (Trott, 1994). Despite the debate over the clonogenic fraction of tumours, the clonogenic cells that are present are thought to spend the majority of time in a quiescent state. As a tumour increases in size, clonogenic cells are distanced from the nutrient-providing tumour vasculature by the faster proliferation of better-nourished cells. During therapy, clonogenic tumour cells are recruited into the cell cycle and act to repopulate the tumour. The stimulus for this response is unknown (Trott, 1999), though it may be related to the restart of previously quiescent cells (Fowler, 1991), the inflammatory response induced by therapy (Trott, 1990), or the inherent ability of some normal tissues to maintain a cell density of 50-60% (Trott and Kummermehr, 1993; Trott et al., 1999). Regardless of the initiation mechanism, this accelerated tumour repopulation is observed experimentally and clinically as an increasing growth rate of the tumour during continued treatment, reminiscent of the "emergence of treatment resistance". Accelerated tumour repopulation has been observed in response to fractionated radiotherapy to varying 145 degrees in both experimental (Durand, 1997; Sham and Durand, 1999) and clinical tumours (Withers etal., 1988). As suggested by Peters and Withers (Peters and Withers, 1997), strategies for combating accelerated repopulation during radiation therapy include reducing the duration of maximal proliferation, or intensifying treatment during periods of maximal proliferation. However, more specific methods to limit the effect of accelerated tumour repopulation during therapy through the application of appropriate pharmaceutical agents may provide a superior alternative (Tannock, 1998). In particular, hypoxic regions of solid tumours are known to be highly radioresistant and have the ability to escape treatment and repopulate. One potential strategy to improve tumour response to therapy is to temporarily block or limit tumour cell division and hence repopulation with a cytostatic agent between doses of radiation therapy. A further extension of this strategy lies in the differential sensitivity of tumour cells in various stages of the cell cycle to certain agents. Depending on the cytostatic agent used, tumour cells could be blocked in a given phase of the cell cycle and then allowed to re-enter the cycle more or less in synchrony. The tumour cells would therefore be synchronized for a short period of time, although the inherent heterogeneity of tumour cell cycle progression would reduce this phenomenon rather rapidly upon resumption of cycling. Nonetheless, there remains a powerful therapeutic potential for cytostatic agents, either in the timing for cell progression or for specific targeting. Our cytostatic agent of choice was rapamycin, not only for its effects on the cell cycle and the cyclin D protein as indicated in Chapter 5, but also because it was reported that quiescent cells have specific sensitivity to rapamycin (Jayaraman and Marks, 1993; Wiederrecht et al., 1995). With the caveat that a majority of hypoxic cells are in a quiescent state, the activity of this agent against quiescent cells may target those cells that escape radiotherapy. In this chapter, we attempt to control accelerated repopulation 146 using a combination of rapamycin in conjunction with radiation therapy. In contrast to the entire thesis thus far, instead of the in vitro multicellular spheroid system, we used xenografted animal models with human tumours from the WiDr cell line. The justification of switching to xenografts was to use a model that was even closer in tumour physiology and microenvironment to in situ human tumours since we were trying to elucidate the effects of using a combination of two tumour therapies. Experiments were performed in a two part series, first to determine the effects of rapamycin on the WiDr xenografts using pulses of BrdUrd and IdUrd over an interval of time, followed by studies of the cytostatic agent in combination with a high and low dose of radiation. 147 6.2 MATERIALS AND METHODS 6.2.1 Tumours Tumours derived from WiDr, a human colon adenocarcinoma (Noguchi et al., 1979), were used for all experiments in this Chapter. This cell line was obtained as a cultured cell line (American Type Culture Collection, Rockville, MD), grown in severe combined immunodeficient (SCID) mice and maintained by intramuscular transplant. Experimental tumours were grown as dorsal subcutaneous implants in 7-8 week old SCID mice (bred and maintained in-house at the BCCA Cancer Research Centre), and were used approximately 3-4 weeks after implantation at an average weight of -500 mg. All procedures were performed in accordance with the ethical standards of the University of British Columbia Committee on Animal Care and the Canadian Council on Animal Care. 6.2.2 Reagents Rapamycin (Sigma Chemical Company, St. Louis, MO) was purchased as a fine grade powder as previously described in Section 5.2.1. In order to label cells that were actively synthesizing DNA, BrdUrd (Sigma-Aldrich Ltd., St. Louis, MO) and IdUrd (Sigma-Aldrich Ltd., St. Louis, MO) were injected intraperitoneally (i.p.) into the mice as described in more detail in Section 6.2.5 and Section 6.2.6. To decipher the proximity of tumour cells from the tumour blood vessels, Hoechst 33342 (Sigma-Aldrich Ltd., St. Louis, MO) was injected intravascularly (i.v.) into the animals 20 minutes prior to sacrifice, at a concentration of 20 mg/ml (1 mg in 0.05 ml double-distilled water). The brightest Hoechst-stained cells, designated fraction 1, were proximal to functional vasculature while the dimmest Hoechst-stained cells, designated fraction 6, were the furthest from functional vasculature at the time of Hoechst injection (Chaplin et al., 1986). Hoechst 33342 at this concentration has been previously tested and reported to not be toxic to host animals or tumour cells (Durand and LePard, 1995). 148 6.2.3 Fluorescence activated cell sorting Upon sacrifice, the tumours were excised and washed in ice-cold PBS and kept on ice until disaggregation of the tumour mass. Single cell suspensions were prepared by finely mincing the solid tumours with dual scalpels, followed by a 40 minute agitated incubation at 37°C. During this incubation, the cells were suspended in a cocktail of enzymes, including 0.5 % trypsin and 0.08% collagenase in PBS. Once the suspension was taken out of the 37°C shaker, 0.06% DNAse was added and the cell suspension gently vortexed, and filtered through a 30 um pore nylon mesh filter to remove any residual cellular clumps. Samples were washed by centrifugation and resuspended in MEM (Gibco Invitrogen Group) containing 10% FBS (Hyclone Biotechnologies) and immediately placed on ice for FACS. Each tumour was volume sorted (see Section 1.4.2 for description) into six fractions according to the intensity of the Hoechst 33342, where more Hoechst 33342 is bound to cells closer to the blood vessels than in those cells further away from the tumour cord. 6.2.4 Flow cytometry with antibodies To facilitate the flow cytometry analysis of incorporated BrdUrd, alcohol-fixed samples were denatured in 2N HCI containing 0.5% Triton X-100 for 20 minutes followed by multiple washes with minimum essential medium (MEM) and contact with the appropriate antibodies. BrdUrd was detected with Bu1/75 (Sigma-Aldrich, St. Louis, MO) primary rat monoclonal antibody at a 1:150 dilution and detected with a fluorescent goat anti-rat Alexa 594 (Molecular Probes, Eugene, OR) secondary antibody at a 1:100 dilution. The IdUrd was detected using a B44 clone-conjugated FITC antibody at a 1:50 dilution and all samples were stained with DAPI at 1 pg/ml. Fluorescence of all samples was detected using an Epics Elite flow cytometer and subsequently reprocessed for analysis from the list mode files. 149 6.2.5 Rapamycin with lododeoxyuridine and Bromodeoxyuridine IdUrd was injected first at a dose of 90 mg/kg (6 mg/ml stock solution in 40mM Tris buffer, pH 10), followed by rapamycin treatment at doses of 1 or 3 mg/kg, and the series ended with a BrdUrd injection 3 hours prior to tumour excision at a dose of 90 mg/kg (6 mg/ml stock solution in Dulbecco's PBS [Gibco Invitrogen Corp., Burlington, ON]). A second series reversed the order of the IdUrd and rapamycin, where rapamycin was given daily either one day prior or two daily doses prior to an injection of IdUrd, followed once again with a 24 hour interval between the BrdUrd injection and an additional 3 hours before the animals were sacrificed. 6.2.6 Rapamycin in combination with radiotherapy Mice were irradiated with doses of either 6 Gy or 12 Gy before being subjected to doses of rapamycin either at 1 or 3 mg/kg. Twenty-two hours after rapamycin, IdUrd at 90 mg/kg was administered i.p. A second marker for DNA proliferation, BrdUrd, was then also injected i.p. approximately 24 hours after IdUrd. Mice were then given an additional 3 hours for BrdUrd incorporation prior to sacrifice. 150 6.3 RESULTS 6.3.1 Rapamycin inhibits cell proliferation in vivo The effects of rapamycin on proliferation were examined in xenografted WiDr tumours using flow cytometric analysis of incorporated iododeoxyuridine (IdUrd) and bromodeoxyuridine (BrdUrd) during a regimen of IdUrd and BrdUrd injections interspersed with rapamycin doses of either 1 mg/kg or 3 mg/kg. As shown in Figure 6.1 (A - C), when 1 mg/kg rapamycin was injected prior to the IdUrd label, many less cells stained positive for IdUrd than in those tumours that received the single dose injection of rapamycin immediately after the IdUrd label. However, after 24 hours, or approximately one cell cycle, regardless of timing or frequency of the rapamycin injections, an injection of BrdUrd was added to determine the proliferation status at this point, and all tumours had a smaller percentage of cells in the S-phase of the cell cycle compared to the control. Although not drastic, the curves do indicate that increasing the exposure repetitions of rapamycin resulted in a decreased percentage of positive BrdUrd labelling. Using a higher dose of rapamycin at 3 mg/kg, the patterns for the graphs were different, where the number of rapamycin repetitions did not appear to change the percentage of positively stained cells for BrdUrd, since all the curves had a low level of BrdUrd positive cells compared to the curves at 1 mg/kg rapamycin dose. The comparisons between the positive labelling of BrdUrd and IdUrd when using 3 mg/kg instead of 1 mg/kg, indicated that when using a higher concentration of rapamycin, the labelling of BrdUrd and IdUrd is lower in all the varying timepoint and frequency regimens. Calculations for the Tpot (potential doubling time) of the WiDr cells were determined (Figure 6.2) and the results suggests that rapamycin prolongs the value for Tpot, and increasing its concentration also serves to further prolong the time frame. 151 A B C Figure 6.1 Labelling indices of WiDr xenografts after treatment with either 1 mg/kg (1R) or 3 mg/kg (3R) rapamycin. A dose of rapamycin was injected immediately prior or post IdUrd injection followed by an interval of 24 hours before another injection of BrdUrd. Animals were sacrificed 3 hours after tumours were removed for analysis of IdUrd and BrdUrd uptake. This figure shows the percentage of cells with (A) BrdUrd only labelling, (S) IdUrd only labelling, or (C) those cells that labelled with both BrdUrd and IdUrd, within each sub region of the tumours from cells closest to tumour blood vasculature to cells more central within the tumours. 152 1000 C/5 ZJ o E i -100 1R1 daily,'2 days pre-ltiUrd 1R daily, one day pre-ldUrd 1Rpost-ldUrd 3R daily, 2 days pre-ldUrd 3R daily, one day pre-ldUr| 3R post-ldUrd 24hr interval 0 1 Sort Fraction (bright to dim) Figure 6.2 Potential doubling times (Tpot) of WiDr xenograft tumours. Tumours were subjected to a fluorescent dye, Hoechst 33342, and fluorescence-activated cell sorted dependent on its proximity to the blood vessels where cells closest to the blood vessels have a high intensity of fluorescence whereas those cells further away from the blood supply, have dim fluorescence. Tpot values were calculated using the knowledge of the percentage of cells in S-phase over a course of two cell cycles. 153 6.3.2 Rapamycin sensitizes WiDr xenografts to Radiation In addition to effects on tumour cell death or proliferation, rapamycin may affect host-tumour interactions, such as angiogenesis, cytokine production, or immune responses that could influence the response to radiation. To assess whether the introduction of rapamycin was sufficient to inhibit tumour proliferation, the fraction of cells in S phase was determined in tumour-bearing animals treated with or without a dose of rapamycin in conjunction with a dose of 6 or 12 Gy of radiation. Twenty-two hours after radiation and drug dose, animals were injected with IdUrd and after another 24 hours, an injection of BrdUrd was injected 3 hours prior to euthanasia. The effects of rapamycin and radiation on tumour growth were evaluated by calculating the potential doubling time for each region within a tumour. Upon analysis of Figure 6.3, rapamycin treatment remained the most significant factor influencing labelling index, where differences in LI were more evident when varying the concentration of rapamycin, as opposed to doubling the dose of radiation. Interestingly, although this rapamycin-mediated cell cycle arrest in the xenografts translated into a longer tumour doubling time for the rapamycin-treated animals, the difference was small between those tumour bearing animals treated with 1 mg/kg rather than 3 mg/kg of rapamycin (Figure 6.4). Nonetheless, when observing the data by varying only the presence or absence of rapamycin, the Tpot values were all significantly higher with the addition of rapamycin. Radiation has therapeutic advantages, but the addition of rapamycin further sensitized the WiDr xenografts. The in vivo cell cycle arrest data were consistent with the biochemical inhibition of mTOR activity by rapamycin demonstrated in vitro in Chapter 5, and indicate an advantage for combination therapy. 154 A B C Figure 6.3 Labelling indices of WiDr xenografts after treatment with radiation and rapamycin. A dose of rapamycin (either 1 mg/kg or 3mg/kg) was injected immediately into tumour-bearing animals after a dose of either 6 Gy or 12 Gy of radiation. An IdUrd injection was followed by an interval of 24 hours before another injection of BrdUrd, 3 hours prior to sacrifice. This figure shows the percentage of cells with (A) BrdUrd only labelling, (S) IdUrd only labelling, or (C) those cells that labelled with both BrdUrd and IdUrd, within each sub region of the tumours from cells closest to tumour blood vasculature to cells more central within the tumours. 155 10000 _ 1000 l_ o .1 100 10 T 6Gy & no Rap : 6 G y & 1 R 6Gy & 3R 12Gy & no Rap 12Gy& 1R 1 2 G y & 3 R 0 1 2 3 4 5 6 Sort Fraction (bright to dim) Figure 6.4 Potential doubling times (Tpot) of WiDr xenograft tumours after treatment with doses of radiation (6Gy or 12Gy) and rapamycin (1 mg/kg or 3mg/kg). 156 6.4 DISCUSSION A number of cytostatic agents have demonstrated varying abilities to halt cell progression at some point in the cycle, including drugs able to block cells in d (Owa et al., 2001). For use in conjunction with radiation therapy, consideration of the selective targeting of radioresistant quiescent cells is highly desirable. One agent that has demonstrated effectiveness on quiescent cells in vitro is the bacterial metabolite rapamycin. Rapamycin uniquely interferes with cell cycle progression from Gi to S phase in the response to proliferative stimuli by blocking the mRNA translation of essential cell cycle proteins (Wiederrecht et al., 1995). The most dramatic rapamycin sensitivity has been observed in quiescent cells (in G0) in vitro that were activated with mitogens after rapamycin treatment. Cells that were exponentially growing have demonstrated a rapamycin sensitivity that is cell type dependent (Kawamata era/., 1998). Since the majority of tumour clonogenic cells are thought to be in a quiescent state prior to therapy, an agent that is most effective against cells entering the cell cycle could provide a level of tumour specificity. This may be essentially important when compared to hematopoietic cells or other rapidly cycling and/or continually renewed normal cells. Indeed, studies in nude mice bearing human glioblastoma xenografts have shown a resumption of normal haematopoietic function immediately after cessation of a multi-dosing regimen while tumour growth suppression lasted much longer (Owa et al., 2001). mTOR is becoming an important target for the new lines of anticancer drugs such as rapamycin and its derivatives, CCI-779 and RAD001, which have exhibited significant anticancer activity in various tumour cell lines (Huang and Houghton, 2003; Boulay et al., 2004). The activation of mTOR enhances protein translation via phosphorylating eukaryotic initiation factor 4E-binding protein 1 (4E-BP1) and S6 kinase (S6K1) (Raught et al., 2001). Certain glioma cell lines have been shown to be sensitive to rapamycin and its derivatives (Hosoi et al., 1998). Furthermore, rapamycin has been reported to 157 sensitize U87 glioma xenografts to radiotherapy (Eshleman et al, 2002). Interestingly, Eshleman et al. also observed that a monolayer of U87 glioma cells treated with rapamycin had no effect on the cells radiosensitivity, however, when glioma cells were cultured as spheroids, treatment with rapamycin induced radiosensitization. In our study, we addressed the effect of rapamycin on cell cycle kinetics of WiDr colon adenocarcinomas and more specifically, the beneficial or negligible effects of rapamycin in combination with radiation. To measure the cell cycle kinetics of the WiDr xenografts, we measured the uptake of two thymidine analogues, BrdUrd and IdUrd (refer to Section 2.2.3), and calculated the potential doubling time (Tpot) of the tumours. The labelling of the BrdUrd and IdUrd indicated the percentage of cells in the S-phase of the cell cycle at that specific moment in time and the proportion of cells that had only one label or another versus both labels, which created ratios of the percentage of cells in the S-phase at various time points. This differential in the ratio provided important information regarding the cellular distribution of the cells within the cell cycle as previously described in Section 2.2.4 in reference to the cell cycle time (Tc). For tumour xenografts, of all the possible measures of growth kinetics, the one that has often been touted to be the most predictive of therapeutic response is the potential doubling time (Tpot) (Steel, 1977). The Tpot value is a measure of time that would be required by a tumour to double in cell number under the two caveats that there was no cell loss and that quiescent cells were included in the population. The cells that would have been lost were assumed to have remained in the population without changing the growth fraction. Several procedures have been proposed (Begg et al., 1985; White and Meistrich 1986; White, 1989; Terry et al, 1990) which allow, in principle, the calculation of Tp o t from flow cytometric data obtained from xenografted tumours. The classical standard equation for Tpo, was: ATs Tpo, = (6) LI 158 where X = 0.693, T s = time of S-phase, and LI = labelling index. It is known that radiation activates the phosphoinositol-3 kinase (PI3K)/Akt pathway and that inhibition of PI3k or Akt sensitizes tumour vasculature to radiotherapy (Holland et al., 2000; Sonoda et al., 2001; Chakravarti et al., 2002; Ermoian et al., 2002). Mammalian target of rapamycin (mTOR) is a downstream target of Akt (Figure 5.1), and we hypothesize that irradiation activates mTOR signalling in colon adenocarcinomas and that radiosensitization results from inhibiting mTOR signalling. mTOR inhibitors, such as rapamycin, were found to radiosensitize WiDr xenografts determined by decreased BrdUrd and IdUrd uptake as well as prolonged cell cycle kinetic measurements. In the present study, we have shown the additive therapeutic benefits of using irradiation in combination with an mTOR signalling inhibitor, rapamycin. We found that rapamycin and radiation together lowered the uptake of BrdUrd (Figure 6.3>A), irrespective of the concentration of rapamycin we used. Figure 6.3C was a slightly different measurement of the S-phase cells, where only those labelled with both the IdUrd and BrdUrd were graphed. This graph was in agreement with Figure 6.3A where the addition of rapamycin with irradiation, resulted in a lower percentage of cells undergoing S-phase when compared to cells that were exposed to the same dose of irradiation only. This pattern was observed for both sets of tumours undergoing 6 Gy or 12 Gy of irradiation and increasing the concentration of rapamycin from 1 mg/ml to 3 mg/ml created no additional inhibition to BrdUrd or IdUrd uptake into the tumour cells. However, when comparing the effects of increasing the dose of irradiation, the higher dose of 12 Gy did result in an even lower percentage of BrdUrd uptake as well as a decrease in the percentage of cells labelled with both BrdUrd and IdUrd. A decrease in uptake of the BrdUrd or IdUrd would signify that a smaller percentage of cells were in the S-phase of the cell cycle, thus ultimately, we would expect tumour growth to be slower, and this was exactly what we observed when the Tp o t was calculated. The results from the Tp0, 159 calculations were graphed in Figure 6.4, where we can conclude that the Tpot values were increased (in the measurement of hours) when cells were subjected to higher concentrations of rapamycin and/or a higher dose of irradiation. When rapamycin injections and irradiation were implemented together, this created the greatest influence on prolonging the Tpot values. We were not only interested in determining the effect of combining irradiation with an mTOR inhibitor, but also to determine if there was a differential effect within the subregions of each xenografted tumour, in an effort to include the distance of cells from tumour blood vessels for an added level of complexity. In previous chapters, we had described the use of spheroids as an in vitro model system to mimic the three-dimensional properties of the tumour systems. With most spheroid experiments, we had sorted cells from each spheroid into subpopulations according to increasing depth into the spheroid. This enabled us to study the microenvironmental effects, such as hypoxia, and we determined from these experiments that increasing distance from the surface of the spheroids resulted in prolonged cell cycle kinetics. Similarly, with the tumour xenografts, the distance of the cells within a tumour might create significant differences in response to irradiation and towards drug treatment. In particular, it has been classically reported for years how hypoxic cells are radioresistant, enabling them the ability to survive radiation therapy and create the potential for tumour cell repopulation (Gray et al., 1953; Gatenby et al., 1988; Olive et al., 2002). Thus, by including assessment of the distance between tumour cells and tumour blood vessels, we also get an indication of the hypoxic status and how that contributes to the therapeutic response to doses of rapamycin and radiation. To garner information on the distance of cells from the blood vessels, every animal received an i.v. injection of Hoechst 33342, thus tumour cells came into contact with Hoechst 33342 via the delivery of this fluorescent marker from the host animal's blood vessels. For all the graphs presented in this chapter, we 160 have plotted the effects of irradiation and rapamycin over a range of six sub-fractions of tumour cells, each fluorescently activated cell sorted based on the volume of cells and Hoechst 33342 brightness, where brighter cells were closest to the blood supply whereas dim cells were the farthest. The relationship between hypoxia and cell cycle kinetics in tumour xenografts can be observed in each figure of this chapter. The x-axis on every graph separated the tumour cells in reference to their distance from the tumour blood vessels. Tumour cells farther away from the blood vessels were shown to have a much reduced level of BrdUrd and IdUrd labelling, as well as a longer Tpot compared to the cells from the same tumour but closer to the blood vessels. The additional effects of adding doses of rapamycin did not change the uptake levels of BrdUrd or IdUrd, however, the cell kinetics Tpot time was further prolonged with higher concentrations of the rapamycin drug (Figure 6.1 and 6.2). Figures 6.3 and 6.4 show data from the tumour xenografts that received both the rapamycin treatment as well as doses of irradiation. Similar to Figures 6.1 and 6.2, the trends show that fewer cells took up BrdUrd and IdUrd in the dim cell population, but the numbers were not further decreased when rapamycin and radiation doses were applied. However, the Tpot values were also calculated to be longer in those populations of cells farther from the tumour blood vessels. We conclude that the combination of rapamycin treatment in conjunction with a dose regimen of irradiation led to additive therapeutic benefits by slowing down the cell cycle time of the growing tumour cells, as well increasing sensitivity to those quiescent and normally radioresistant hypoxic cells through the use of rapamycin. The addition of rapamycin alone prolonged the Tpot time signifying that rapamycin slows down the cell cycle rate of tumour cell growth. For poorly perfused cells, two repetitions with 1 mg/ml of rapamycin produced even longer Tpot values, but this effect was unnoticed in the xenografts exposed to 3 mg/ml. However, there was no effect on Tpot 161 for the well-perfused cells. This is of particular interest since this would suggest that rapamycin is targeting the quiescent cells and a higher concentration of 3 mg/ml does have a more potent effect than that of 1 mg/ml rapamycin. However, even though a higher dose did have a greater cell cycle inhibitory effect, adding more repetitions of rapamycin doses at 3 mg/ml, did not further prolong the cell cycle (in terms of Tpo.) in the proliferating or quiescent cells. Interestingly, using the combination of rapamycin treatment with irradiation resulted in similar general results but further suggests the dynamics on the roles of the drug and radiation on tumour cells. The addition of rapamycin and irradiation resulted in increases in the Tpot not unlike that of rapamycin alone, however the effect was more pronounced. When using 6 Gy of irradiation with rapamycin, the values for Tpo, were the same throughout the xenograft tumours except for the cells furthest from the blood vessels, the most quiescent population of cells in the sorted populations of cells, which displayed longer Tpot values with a higher concentration of rapamycin. When cells were irradiated with 12 Gy, as cell populations were farther from the blood supply, the Tpot values also gradually increased, and the graphs were almost identical regardless of the concentration of rapamycin added. Alternatively graphed, if we vary the dosage of irradiation keeping the concentration of rapamycin constant at 1 mg/ml rapamycin, the tumours show a difference of Tpot only in the dim (quiescent) fractions, whereas increasing the concentration to 3 mg/ml produced overlapping curves for Tpot. This pattern strongly suggests that at the lower dose of rapamycin, increasing the radiation dose from 6 Gy to 12 Gy, increases the Tpot in the quiescent cell populations (dim fractions). However, when using the higher dosage of rapamycin, even those cells that were also treated with only 6 Gy of radiation had an increase in the Tp o t value, in fact, the results indicated that the curves overlapped and the Tpot was the same when treated with 6 Gy or 12 Gy of radiation. This suggests that even though doses of radiation affected 162 the Tp0t, for the quiescent cells that might have been initially radioresistant at the lower radiation dose, it was the higher doses of rapamycin that further prolonged the Tpot values. The most effective combination from our studies would be to use 12 Gy of irradiation in combination with 1 mg/ml of rapamycin. Our data provide additional evidence to the growing amount of research that indicates the effectiveness of mTOR inhibition in treating cancer. Specifically, we suggest that mTOR inhibition with rapamycin provides an additional means of targeting tumour cells that were radioresistant due to their hypoxic properties. The added level of treatment would create a more thorough approach to treatment, also in the hopes of eradicating the potential for regrowth of the primary tumour, or even secondary tumours within the host. Of note, mTOR inhibitors are likely to be more effective at targeting tumour cells than normal cells. This is because transformed cells show increased activation of the Akt/mTOR pathway (Aoki et al., 2001) and therefore present a better target with larger potential for the effects of mTOR inhibition. It is possible that mTOR inhibition with rapamycin would show a greater decrease in radiosensitivity in colon adenocarcinoma cells in vivo. Glioma models using rapamycin derivatives have previously shown that there is no increased radiosensitivity in vitro (Eshleman et al., 2002; Shinohara et al., 2005) but have shown to be significantly increased in vivo (Eshleman et al., 2002). This suggests that rapamycin has an additional effect that is dependent on its action within whole tumours, as opposed to a tumour cell monolayer. Additionally, it has been previously shown that rapamycin inhibits angiogenesis (Guba et al., 2002), and we have established (Chapter 5) that rapamycin also inhibits the cell cycle protein, cyclin D, affecting the cell kinetics of tumours. Overall, this suggests that combining mTOR inhibition with radiation may result in a dual mechanism of tumour inhibition, promoting both tumour cell cytotoxicity and inhibiting tumour angiogenesis. 163 C H A P T E R 7: S U M M A R Y A N D F U T U R E D I R E C T I O N S 164 7.1 SUMMARY AND SIGNIFICANCE OF RESULTS Cancer is ultimately a problem of the uncontrolled growth of cells. Many factors could contribute to changes in cellular growth, ranging from large scale differences such as chromosomal abnormalities and gene expression to minute changes like point mutations in the DNA (Hanahan and Weinberg, 2000; Blume-Jensen and Hunter, 2001). Not only would these varying mechanistic factors contribute to the complexity of cancer, but in addition, there are many cases of failure in cancer therapy due to the emergence of cell resistance to therapeutic strategies. Many theories on the mechanisms behind this resistance have evolved over years of research, and it has been shown where certain conditions, such as hypoxia, create an environment for tumour cells that would lead to radiation resistance. As described throughout this thesis, the majority of our studies were performed using the in vitro multicellular spheroid system as a model for tumour growth. Due to the in vitro nature of the system, not only were the parameters affecting the model easier to manipulate but we also had the ability to control the population of tumourgenic cells which were all derived from identical single cells. With a cellular model that should have similar characteristics throughout the entire system, we discovered there were differences dependent on the region of the cell within the spheroid, and in addition, the cell kinetics of each subregion also proved to have differing values. Therefore, the main overall objective for this project was to gain a better understanding of the mechanisms underlying the control of tumour cell kinetics as well as introduce cell cycle inhibitors to the system in attempts to improve therapeutic response to radiotherapy. The working hypothesis for this thesis is that the cell cycle of tumour cells is a multi-faceted process with many key players, but in particular, that tumour hypoxia has a rapid impact on cell cycle regulatory proteins as well as changing the effectiveness of therapeutic agents. Radioresistance of hypoxic cells is an important limiting factor in tumour response to 165 radiotherapy but can be targeted by the administration of cell cycle inhibitors such as rapamycin to manipulate intracellular targets that control cell cycle progression. We first measured changes in the amount of time it would take for cells to double in population number by initiating studies on the kinetics of cellular subregions within a spheroid. To measure the kinetics of the cells in spheroids, we used an indirect procedure that immunofluorescently dual-stains cells for flow cytometric measurements of two thymidine analogues, iododeoxyuridine and bromodeoxyuridine as detailed earlier in Chapter 2 (Begg et al., 1985; White et al., 1994). Even though the cells in the spheroid were all derived from the same parental line, Figure 2.5 details the differences in cell cycle time between six subpopulations within a spheroid. Cells nearest to the periphery of the spheroid were found to have the shortest cell cycle time of approximately 14 hours for the V79 cells and 21 hours for both SiHa and WiDr spheroids. Conversely, those cycling cells closer to the centre of the spheroid had a much slower cell cycle time averaging 18 and 28 hours for the V79 and for the human lines respectively. We have thus concluded in Chapter 2 that the proliferative status of cells in a solid tumour model highlights the importance of the cell microenvironment. In therapeutic protocols used in cancer clinics, most patients receive a standardized treatment of radiation therapy, commonly in addition to doses of chemotherapy. While this approach is utilized to target the majority of solid tumours, the outcomes of the therapies range tremendously between patients. Without a more individualistic approach, patients may be subjected to treatments that have an anticipated high rate of failure and more importantly, stimulation of the tumour cells from chemotherapy or radiation could initiate accelerated repopulation leading to both the potential growth of the primary, and of secondary tumours in the patients. Our studies also indicated the importance of the 166 microenvironment on growth control and the advantageous use of proliferation studies to better understand solid tumours prior to treatment. To assess the influence of the cellular microenvironment on cell proliferation, we initially investigated the levels of two important cell cycle regulatory proteins, cyclin B1 and cyclin D, in cell subpopulations from spheroids of different species. Most notably, cells near the interior of the spheroid were naturally situated in a hypoxic environment due to the three-dimensional spatial growth of the spheroid model. Recognizing the limitations of quantitatively measuring the levels of oxygen in each subpopulation, we instead chose to manipulate the external level of oxygenation. We determined that cycling cells near the necrotic centre of the spheroid have a longer cell cycle time and similarly, the levels of cyclin B1 and cyclin D were also lower in these regions. The levels of cyclin expression oscillate throughout the cell cycle and the mRNA and protein expressions peak at the time of maximum kinase activation, and depending on the type of cyclin, discrete bursts of kinase activity at specific cell cycle transitions are exhibited (Paschoa et al., 2005). We speculated that regions of cells within a spheroid would have different activity levels for both cyclin B1 and cyclin D. Furthermore, we hypothesized that this change was not entirely dependent on the spatial location of a cell within a spheroid, but also a consequence of external factors such as hypoxia. By altering the external levels of oxygenation of each sub-region of the spheroid, our studies clearly indicated a critical influence of oxygen on cyclin B1 and D levels, and ultimately on cell proliferation. We expected to observe a similar outcome from hypoxic studies using monolayers of the same cell lineage, but found that the dramatic effects were unique to the spheroids since oxygen deprivation studies with single cells resulted in only very minor changes in cyclin D levels for the same cell lines. Even short, transient changes in ambient oxygenation had a significant effect on the levels of both cyclins in spheroids. Upon closer examination of the spheroid cryosections in Figures 167 3.7 and 4.4, cyclin labelling was observed in foci, or groups of clustered cells. This clustering of cells probably represents semi-synchronous subsets of daughter cells from a common parent cell, thus having a very similar growth status and cellular physiology. Short exposures to hypoxia changed the levels of cyclin in 4 hours, and although this was fast, it was not unexpected considering that cyclin levels can rapidly and naturally change as cells progress through the cell cycle. For example, cyclin B1 essentially disappears as cells exit from the G2/M phase into Gi - a time frame of one hour or less. Thus, although a 4 hour hypoxia treatment could be perceived as short, it was not unreasonable anticipate causative changes in the cyclin levels. Decreased levels of cyclin D would have a large impact on the cell cycle at the Gi checkpoint, where the hypoxia driven decline in cyclin protein may result in cells entering the quiescence (G0) stage of the cell cycle. Similar to cyclin B1 levels as described in Chapter 3, cyclin D levels were responsive to the environmental status of a cell, making the levels of this protein also a good predictor for hypoxia-induced proliferation changes. The combination of determining the relationship of hypoxia and both cyclin proteins, and its potential use as predictive assay, supported the initiative to investigate cell cycle inhibitors as a method of reducing the proliferation of surviving, radioresistant hypoxic cells during therapy. In this thesis, we have utilized various methods to measure and quantify the levels of cyclin protein in various microenvironmental situations and in various regions within our model system. Specifically, we used immunoblotting, flow cytometry, and immuno-histochemical sectioning. Using any of the three methods, differences in the levels of cyclin B1 or cyclin D were observed. Likewise, similar trends were observed with each method, but sometimes in a counter-intuitive manner. Consider, for example, the extreme situation of immunoblotting versus immunohistochemical sectioning. In the former, the signal is "averaged" over heterogeneous cells, where an increased signal 168 would result either from modest protein upregulation in all cells, or more marked upregulation in a small subset only. The cryosections clearly showed that the correct interpretation is the latter (notably, however, technical difficulties make quantitation difficult in both methods, including band localization and integration difficulties in the blots, and section thickness and image saturation issues in the cryosections). Flow cytometry has the advantages of being inherently quantitative, inclusive of all cells, and well suited to integrating signals from widely heterogeneous single cells. Consequently, we place the greatest emphasis on the data generated with the flow cytometry methodology. As previously outlined in Chapters 3 and 4, we investigated the effects and influence of the microenvironment on the cell cycle via two critical cell cycle cyclins: cyclin D at the GT checkpoint, and cyclin B1 at the mitosis entry point. Regulation of the cell cycle would be very beneficial for therapeutic purposes, especially if it can be shown to have an additive effect with current established therapeutic regimens. We introduced rapamycin (Figure 1.12) to spheroids to determine the effect it would have on the cyclin D levels in each region of the model, and found that it indeed had an effect on most cells in the spheroid. Cells on the periphery of the spheroid showed a decrease in cyclin D levels compared to controls, and cells more towards the interior of the spheroid maintained a low level of cyclin D similar to the control. Rapamycin is an inhibitor of the serine/threonine kinase mTOR in addition to initiating Gi arrest and we have shown the effects of rapamycin on the levels of cyclin D (Figure 5.3). Previous studies show that cell lines derived from different cancer types were noted to undergo G, arrest when exposed to 1 nM rapamycin, a concentration which closely matches that required for biochemical inhibition of mTOR in cells. Some cell lines that failed to respond to the 1 nM dose did undergo growth arrest at significantly higher concentrations (-1000 nM) (Sawyers, 2003). Increasing the concentration of rapamycin 169 had an increasing effect on the cyclin D protein levels of SiHa and WiDr cells with the suggestion that there might be a saturation of effect at a concentration of 0.2 ug/ml. These results were most dramatic on the peripheral cells of the spheroid. Agents such as rapamycin have demonstrated growth inhibition or tumour regression in some xenograft models (Owa et al., 2001; Yu et al., 2001), but little is known as to which cells specifically responded to the drug and were responsible for the tumour growth inhibition or regression in these systems. Moreover, more studies need to be done regarding the translation of the in vitro mechanism of action of these agents to a more complex in vivo system. Our data suggested that rapamycin has growth inhibitory effects on both the SiHa and WiDr tumour cell lines using the levels of cyclin D as a measure of proliferation, and that those effects were produced by non-cytotoxic doses of rapamycin. Time course experiments suggested a relatively fast mode of action for rapamycin as exposures longer than 4 hours produced little additional effect. As our results indicate that rapamycin rapidly lowered the levels of cyclin D (0 hour curve in Figure 5.4), we conclude that its effect is the result of a relatively direct action on the cyclin D protein, and not through indirect protein-protein interactions. The conclusion from these studies is that rapamycin indeed does target both the proliferating and the quiescent populations and the additional microenvironmental stress of hypoxic conditions further lowered the observed levels of cyclin D protein. This suggests a complex interaction pathway in which rapamycin has direct effects on cyclin D, as well as inhibiting the mTORCI cascade, with possible effects on mTORC2 either indirectly or via effects from microenvironmental stresses. Methods used to counteract accelerated repopulation during radiation therapy include manipulations that directly reduce cellular proliferation, or conversely, intensifying the treatment during periods of maximal proliferation (Peters and Withers, 1997). The 170 former, especially with the application of appropriate pharmaceutical agents, is the most specific approach to limiting the effect of accelerated tumour repopulation during therapy (Tannock, 1998). In particular, hypoxic regions of solid tumours are known to be highly radioresistant and have the ability to escape treatment and repopulate. One potential strategy to improve tumour response to therapy is to temporarily block or limit tumour cell division, and hence repopulation, with a cytostatic agent between doses of radiation therapy. Thus, in reference to radiation therapy, selective targeting of radioresistant cells become highly desirable. One agent that has demonstrated effectiveness on quiescent cells in vitro is the bacterial metabolite rapamycin. The most dramatic rapamycin sensitivity has been observed in quiescent cells (in G0) in vitro that were activated with mitogens after rapamycin treatment. It is known that radiation activates the phosphoinositol-3 kinase (PI3K)/Akt pathway and that inhibition of PI3k or Akt sensitizes tumour vasculature to radiotherapy (Holland et al., 2000; Sonoda et al., 2001; Chakravarti et al., 2002; Ermoian et al., 2002). Mammalian target of rapamycin (mTOR) is a downstream target of Akt (Figure 5.1), and since irradiation indirectly activates mTOR signalling in colon adenocarcinomas through cytoreduction, we hypothesized that inhibiting mTOR signalling and subsequent tumour regrowth would improve the anti-tumour efficacy of radiation. Rapamycin was indeed found to inhibit repopulation of WiDr xenografts as determined by decreased BrdUrd and IdUrd uptake and prolonged cell cycle kinetic measurements. In addition, we demonstrated that there was a differential effect on the heterogeneous cell subpopulations within each xenografted tumour, related to the distance of the cells from tumour blood vessels. Our data provided additional evidence to the growing amount of research that indicated the effectiveness of mTOR inhibition in treating cancer. Specifically, we suggest that mTOR inhibition with rapamycin provides an additional means of targeting tumour cells that were initially radioresistant due to their hypoxic microenvironment. This additional 171 treatment option represents a more comprehensive approach to therapy, where greater therapeutic efficiency is anticipated if the tumour is prohibited from proliferating during therapy. Importantly, a therapeutic advantage is expected since tumour cell proliferation is likely greater than that of the surrounding normal tissues; also, mTOR inhibitors should be more effective at targeting tumour cells as they often show increased activation of the Akt/mTOR pathway (Aoki et al., 2001). 172 7.2 FUTURE DIRECTIONS The work described in this thesis gives rise to a number of potential research avenues, both from the use of the methodology presented in Chapters 3 and 4, and extensions from the data presented in the later Chapters 5 and 6. It was of great interest to us that the patterns of cyclin B1 and D levels oscillate in accordance to available oxygen within a brief time frame. It is thus intriguing that the observations in Chapters 3 and 4 may potentially form the basis of a diagnostic tool for hypoxia as the regulation of cell proliferation. Such assays could easily be applied to tumour biopsies in conjunction with endogenous markers of, for example, tumour hypoxia (Kennedy et al., 1997; Evans et al., 2000; Ljungkvist et al., 2006). Likewise, our work suggests potential new probes to enhance non-invasive function imaging methodologies that are currently generating considerable attention in both the laboratory and clinic. Naturally, the ultimate goal is creating more individualized regimens of cancer treatment for maximal effectiveness on every cancer patient. While cyclin protein levels were demonstrated to be differentially expressed in spheroids subjected to hypoxia, the cell cycle mechanisms directly involved have yet to be determined. Other genes that are affected by this intervention could be identified using methods like microarray analysis or serial analysis of gene expression (SAGE). In fact, another member of our laboratory did some SAGE work using SiHa spheroids modelling chronic or transient hypoxia, and analyzing that data, we observed that cyclin D had significant differential expression. Further molecular and methodological work is clearly indicated. The rapamycin data presented in Chapter 5 using spheroids as an in vitro model of solid tumours represents a solid foundation for future work. While these studies revealed a potential for rapamycin as an antiproliferative agent, the underlying mechanism is still unclear. Further work needs to be done to clearly identify the role of each mTORC with 173 rapamycin, as well as how the overall effect of microenvironmental stresses can affect the processes. As antibodies become increasingly available for not only the mTOR protein, but more specifically, antibodies against the raptor or rictor protein, the individual effects of each mTOR complex can be further examined. Furthermore, the kinetics of the rapamycin-induced cellular arrest should be studied in the spheroid system using methodology described previously in Chapter 2. Using the combination of radiation therapy and doses of rapamycin drug to treat solid tumour xenografts in murine animals, our objective was to increase the effectiveness of therapy by inhibiting accelerated repopulation (Chapter 6). Our results indicated that cytostatic agents have the potential to significantly impact current cancer therapy protocols, and are potentially of increased significance since the vast majority of other studies with cytostatic agents have focussed on deducing their mechanism of action. Clearly, further work with other tumour types, as well as further elucidation of the effects and the mode of action and interaction of these agents is imperative. In summary, data from this thesis reinforce the importance of the microenvironment on the cell cycle proteins and ultimately, the cellular growth of tumours. Measurements of cell cycle time through the spheroid had not been previously reported, but using three cell lines, we reported a gradient of cell cycle time dependent on cellular location within a spheroid. Additionally, we described potentially important methodological and therapeutic manipulations that may ultimately form the basis for more individualized cancer therapy protocols in the clinic. In particular, the introduction of rapamycin in combination with radiotherapy indicated the potential for improving current therapeutic strategies by using a cytostatic agent to selectively target tumour cells that were initially radioresistant, and to inhibit the process of accelerated repopulation. 174 REFERENCES Abraham, R. (2002). "Identification of TOR signaling complexes: More TORC for the cell growth engine." Cell 111: 9-12. Adams, C. and B. Storrie (1981). "A Simple DNA-dependent Fluorescence Enhancement Assay for Cell Number." J Histochem. Cvtochem. 29(2): 326-28. Allen, R. and M. Moskowitz (1978). "Arrest of Cell Growth in the G1 Phase of the Cell Cycle by serine deprivation." Exp. Cell Res. 116: 127-37. Aoki, M., E. Blazek, et al. (2001). "A role of the kinase mTOR in cellular transforation induced by the oncoproteins P3k and Akt." Proc. Natl. Acad. Sci. USA 98: 136-141. Arguello, F., M. Alexander, et al. (1998). "Flavopiridol induces apoptosis of normal lymphoid cells, causes immunosuppression, and has potent antitumor activity In vivo against human leukemia and lymphoma xenografts." Blood 91: 1482-2409. Ash, D., M. Peckham, et al. (1979). "The quantitative response of human tumours to radiation and misonidazole." Br. J Cancer 40: 883-889. Awwad, H., M. El Naggar, et al. (1986). "Intercapillary distance measuremnet as an indicator of hypoxia in carcinoma of the cervix uteri." Int. J. Radiat. Oncol. Biol. Phvs. 12: 1329-1333. Bakker, P., J. Stap, et al. (1991). "An Indirect Immunofluorescence Double Staining Procedure for the Simultaneous Flow Cytometric Measurement of lodo- and Chlorodeoxyuridine Incorporated into DNA." Cytometry 12: 366-72. Bassukas, I. (1993). "A generalized 'tumor growth delay" assay for quantifying alterations of tumor regrwoth patterns." Anticancer Res. 13(5A): 1601-6. Bauer, K., P. Keng, et al. (1982). "Isolation of quiescent cells from multicellular tumor spheroids using centrifugal elutriation." Cancer Res 42(1): 72-8. Bedford, J. and J. Mitchell (1977). "Mitotic accumulation of HeLa cells during continuous irradiation. Observations using time-lapse cinemicrography." Radiat Res. 70(1): 173-86. Begg, A., I. Hofland, et al. (1991). "Tumour cell repopulation during fractionated radiotherapy: correlation between flow cytometric and radiobiological data in three murine tumours." Eur. J. Cancer 27(5): 537-43. Begg, A., N. McNally, et al. (1985). "A method to measure the duration of DNA synthesis and the potential doubling time from a single sample." Cytometry 6: 620-26. Bennewith, K. and R. Durand (2004). "Quantifying Transient Hypoxia in Human Tumor Xenografts by Flow Cytometry." Cancer Res. 64(6183-6189). Bernhard, E., W. McKenna, et al. (1994). "Cyclin expression and G2-phase delay after irradiation." Radiat Res. 138(1 Suppl): S64-7. Beuvink, I., A. Boulay, et al. (2005). "The mTOR inhibitor RAD001 sensitizes tumor cells to DNA-damaged induced apoptosis through inhibition of p21 translation." Cell 120: 747-759. Blume-Jensen, P. and T. Hunter (2001). "Oncogenic kinase signalling." Nature 411: 355-65. Bondi, J., A. Husdal, et al. (2005). "Expression and gene amplification of primary (A, B1, D1, D3, and E) and secondary (C and H) cyclins in colon adenocarcinomas and correlation with patient outcome." J Clin Pathol 58: 509-514. 175 Boulay, A., S. Zumstein-Mecker, et al. (2004). "Antitumor efficacy of intermittent treatment schedules with the rapamycin derivative RAD001 correlates with prolonged inactivation of ribosomal protein S6 kinase 1 in peripheral blood mononuclear cells." Cancer Res. 64(1): 252-261. Braun, R., L. JL, et al. (1999). "Fourier analysis of fluctuations of oxygen tension and blood flow in R3230Ac tumours and muscle in rats." Am J Physiol 277: H551-568. Brown, J. (1979). "Evidence for acutely hypoxic cells in mouse tumours, and a possible mechanism of reoxygenation." Br J Radiol 52: 650-656. Brown, J. and A. Giaccia (1998). "The unique physiology of solid tumors: opportunities (and problems) for cancer therapy." Cancer Res. 58: 1408-1416. Brugarolas, J., K. Lei, et al. (2004). "Regulation of mTOR function in response to hypoxia by REDD1 and the TSC1/TSC2 tumor suppressor complex." Genes Dev. 18: 2893-2904. Brugarolas, J., K. Lei, et al. (2004). "Regulation of mTOR function in response to hypoxia by REDD1 and the TSC1/TSC2 tumor suppressor complex." Genes Dev. 18: 2893-2904. Bush, R. and R. Hill (1975). "Biologic discussions augmenting radiation effects and model systems." Laryngoscope 85(7): 1119-33. Bush, R., R. Jenkin, et al. (1978). "Definitive evidence for hypoxic cells influencing cure in cancer therapy." Br. J. Cancer 37(Suppl. Ill): 302-306. Casciari, J., M. Graham, et al. (1995). "A modeling approach for quantifying tumor hypoxia with [F-18]fluoromisonidazole PET time-activity data." Med Phys. 22(7): 1127-39. Chakravarti, A., A. Chakladar, et al. (2002). "The epidermal growth factor receptor pathway mediates resistance to sequential administration of radiation and chemotherapy in primary human glioblastoma cells in a RAS-dependent manner." Cancer Res. 62: 4307-4315. Chance, B., E. Borer, et al. (1988). "Optical and nuclear magnetic resonance studies of hypoxia in human tissue and tumors." Ann N Y Acad Sci. 551: 1-16. Chaplin, D., R. Durand, et al. (1985). "Cell selection from a murine tumor using the fluorescent probe Hoechst 33342." Br J Cancer 51: 569-572. Chaplin, D., R. Durand, et al. (1986). "Acute hypoxia in tumours: implications for modifiers of radiation effect." Int. J. Radiat. Oncol. Biol. Phys. 12: 1279-1282. Chaplin, D., S. Hill, et al. (1998). "Modification of tumor blood flow: current status and future directions." Semin Radiat Oncol 8: 151-163. Chapman, J. (1984). "The detection and measurement of hypoxic cells in solid tumours." Cancer 54(11): 2441-9. Chaung, W., L. Mi, et al. (1997). "The p53 status of Chinese hamster V79 cells frequently used for studies on DNA damage and DNA repair." Nucleic Acids Res. 25(5): 992-4. Chignola, R., R. Foroni, et al. (1995). "Heterogeneous response of individual multicellular tumour spheroids to immunotoxins and ricin toxin." Br. J. Cancer 72(3): 607-14. Chignola, R., A. Schenetti, et al. (1999). "Oscillating growth patterns of multicellular tumor spheroids." Cell Proliferation 32: 39-48. Choi, K., L. McMahon, et al. (2003). "Two motifs in the translational repressor PHAS-I required for efficient phosphorylation by mammalian target of rapamycin and for recognition by raptor." J. Biol. Chem. 278: 19667-19673. 176 Clevenger, C , K. Bauer, et al. (1985). "A method for simultaneous nuclear immunofluorescence and DNA content quantitation using monoclonal antibodies and flow cytometry." Cytometry 6: 208-14. Cline, J., D. Thrall, et al. (1990). "Immunohistochemical detection of a hypoxia marker in spontaneous canine tumours." Br J Cancer 62(6). Coco Martin, J., A. Balkenende, et al. (1999). "Cyclin D1 overexpression enhances radiation-induced apoptosis and radiosensitivity in a breast tumor cell line." Cancer Res. 59(5): 1134-40. Conger, A. and M. Ziskin (1983). "Growth of mammalian multicellular tumor spheroids." Cancer Res 43: 556-560. Crissman, H., M. Wilder, et al. (1988). "Flow cytometric localization within the cell cycle and isolation of viable cells following exposure to cytoxic agents." Cancer Res. 48(20): 5742-5746. Dahlberg, W., E. Azzam, et al. (1999). "Response of human tumor cells of varying radiosensitivity and radiocurability to fractionated irradiation." Cancer Res. 59(20): 5365-9. Davies, C , H. Muller, et al. (1997). "Comparison of extracellular matrix in human osteosarcomas nad melanomas growing as xenografts, multicellular spheroids, and monolayer cultures." Anticancer Research 17: 4317-4326. Deen, D., T. Hoshino, et al. (1980). "Development of a 9L rat brain tumor cell multicellular spheroid system and its response to 1,3-Bis(2-chloroethyl)-1-nitrosourea and radiation." J. Natl. Cancer Inst. 64: 1373-82. Delsanto, P., M. Griffa, et al. (2005). "Bridging the Gap between Mesoscopic and Macroscopic Models: The Case of Multicellular Tumor Spheroids." PhvsRevLett. 94(14): 148105. Demicheli, R., R. Foroni, et al. (1989). "An exponential-Gompertzian description of LoVo cell tumor growth from in vivo and in vitro data." Cancer Res 49: 6543-6546. Demicheli, R., G. Pratesi, et al. (1991). "The exponential-Gompertzian tumor growth model: Data from six tumor cells lines in vitro and in vivo." Tumori 77: 189-195. Dewhirst, M., H. Kimura, et al. (1996). "Microvascular studies on the origins of perfusion-limited hypoxia." Br J Cancer Suppl 27: S247-251. Dirks, P. B. and J. T. Rutka (1997). "Current concepts in neuro-oncology: the cell cycle -a review." Neurosurgery 40: 1000-15. Douros, J. and M. Suffness (1981). "New antitumor substances of natural origin." Cancer Treat. Rev. 8: 63-87. Downward, J. (1998). "Mechanisms and consequences of activation of protein kinase B/Akt." Curr Opin Cell Biol 10(2): 262-7. Drewinko, B., M. Patchen, et al. (1981). "Differential killing efficacy of twenty antitumor drugs on proliferating and nonproliferating human tumor cells." Cancer Res 41: 2328-2333. Dulic, V., G. Stein, et al. (1998). "Nuclear accumulation of p21Cip1 at the onset of mitosis: a role at the G2/M-phase transition." Mol Cell Biol. 18(1): 546-57. Durand, R. (1976). "Cell cycle kinetics in an in vitro tumour model." Cell Tissue Kinet. 9: 403-12. Durand, R. (1976). "Radiation response of multicell spheroids-an in vitro tumour model." Current topics in radiation research quarterly 11(1): 87-139. Durand, R. (1982). "Use of Hoechst 33342 for cell selection from multicell systems." The Journal of Histochemistry and Cytochemistry 30(2): 117-122. Durand, R. (1986). "Chemosensitivity Testing in V79 Spheroids: Drug Delivery and Cellular Microenvironment." Journal of Clinical Investigation 77(1): 247-252. 177 Durand, R. (1986). "Use of a cell sorter for assays of cell clonogenicity." Cancer Res. 46: 2775-2778. Durand, R. (1990). "Multicell Spheroids as a model for cell kinetic studies." Cell Tissue Kinetics 23: 141-159. Durand, R. (1991). "The Influence of Microenvironmental Factors on the Activity of Radiation and Drugs." Int. J. Rad. Oncol. Biol. Phys. 20: 253-58. Durand, R. (1993). "Cell Kinetics and repopulation during multifraction irradiation of spheroids: implications for clinical radiotherapy." Seminars in Radiation Oncology 3(2): 105-114. Durand, R. (1993). "Determining Tpot in heterogeneous systems: a new approach illustrated with multicell spheroids." Cytometry 14: 527-534. Durand, R. (1994). "The Influence of Microenvironmental Factors during Cancer Therapy." in vivo 8: 691-702. Durand, R. (1997). "Tumour repopulation during radiotherapy: quantitation in two xenografted human tumours." Int. J. Radiat. Oncol. Biol. Phys. Durand, R. and C. Aquino-Parsons (2001). "Clinical Relevance of Intermittent Tumour Blood Flow." Acta Oncol 40(8): 929-936. Durand, R. and C. Aquino-Parsons (2001). "Non-constant Tumour Blood Flow." Acta Oncol 40(7): 862-869. Durand, R. and C. Aquino-Parsons (2004). "Predicting response to treatment in human cancers of the uterine cervix: sequential biopsies during external beam radiotherapy." Int. J. Radiation Oncology Biol. Phys. 58(2): 555-560. Durand, R. and N. LePard (1994). "Modulation of tumor hypoxia by conventional chemotherapeutic agents." International journal of radiation oncology biology physics. 29(3): 481-486. Durand, R. and N. LePard (1995). "Contribution of transient blood flow to tumor hypoxia in mice." Acta Oncologica 34(3): 317-323. Durand, R. and N. LePard (1997). "Tumour blood flow influences combined radiation and doxorubicin treatments." Radiother. Oncol. 42: 171-179. Durand, R. and N. LePard (2000). "Effects of mitomycin C on the oxygenation and radiosensitivity of murine and human tumours in mice." Radiother Oncol 56(2): 245-52. Durand, R. and P. Olive (1982). "Cytotoxicity, Mutagenicity and DNA Damage by Hoechst 33342." The Journal of Histochemistry and Cytochemistry 30(2): 111-116. Durand, R. and P. Olive (1992). "Tumour cell kinetics and heterogeneity: insights from multicell spheroids." BJR Suppl. 24: 79-83. Durand, R. and P. Olive (2001). "Resistance of tumor cells to chemo- and radiotherapy modulated by the three-dimensional architecture of solid tumors and spheroids." Methods Cell Biol 64: 211-33. Elkind, M., A. Han, et al. (1963). "Radiation response of mammalian cells grown in culture: Dose dependence of division delay and post-irradiation growth of surviving and non-surviving Chinese hamster cells." J Natl Cancer Inst. 30: 705-711. Ellwart, J. and P. Dormer (1990). "Vitality Measuremnet Using Spectrum Shift in Hoechst 33342 Stained Cells." Cytometry 11(2): 239-243. Ermoian, R., C. Furniss, et al. (2002). "Dysregulation of PTEN and protein kinase B is associated with glioma histology and patient survival." Clin. Cancer Res. 8: 1100-1106. 178 Eshleman, J., B. Carlson, et al. (2002). "Inhibition of the mammalian target of rapamycin sensitizes U87 xenografts to fractionated radiation therapy." Cancer Res. 62(24): 7291-7297. Evans, S., S. Hahn, et al. (2000). "Detection of hypoxia in human squamous cell carcinoma by EF5 binding." Cancer Res. 60: 2018-2024. Falk, S., R. Ward, et al. (1991). "The influence of carbogen breathing on tumour tissue oxygenation in man evaluated by computerised p02." Br. J. Cancer 66: 919-924. Fowler, J. (1991). "Rapid repopulation in radiotherapy: a debate on mechanism. The phantom of tumor treatment-continually rapid proliferation unmasked." Radiother Oncol 22(3): 156-8. Franko, A. and J. Chapman (1982). "Binding of 14C-misonidazole to hypoxic cells in V79 spheroids." Br J Cancer 45(5): 694-9. Freyer, J. (1988). "Role of necrosis in regulating the growth saturation of multicellular spheroids." Cancer Res. 48: 2432-2439. Freyer, J. (1994). "Rates of oxygen consumption for proliferating and quiescent cells isolated from multicellular tumor spheroids." Adv. Exp. Med. Biol. 345: 335-342. Gatenby, R., H. Kesler, et al. (1988). "Oxygen distribution in squamous cell carcinoma metastases and it's relationship to outcome of radiation therapy." Int. J. Radiat. Oncol. Biol. Phvs. 14: 831-838. Gopfert, T., K. Eckardt, et al. (1997). "Oxygen-dependent regulation of erythropoietin gene expression in rat hepatocytes." Kidney Int. 51(2): 502-6. Graeber, T., J. Peterson, et al. (1994). "Hypoxia induces accumulation of p53 protein, but activation of a G1-phase checkpoint by low-oxygen conditons is independent of p53 status." Mol. Cell. Biol. 14: 6264-6277. Gray, L, A. Conger, et al. (1953). "Concentration of oxygen dissolved in tissues at time of irradiation as a factor in radiotherapy." Br. J. Radiol. 26: 638-648. Guba, M., P. von Breitenbuch, et al. (2002). "Rapamycin inhibits primary and metastatic tumor growth by antiangiogenesis: involvement of vascular endothelial growth factor." Nat. Med 8: 128-35. Hain, J., E. Weller, et al. (1996). "Effects of ionizing- and UV B-radiation on proteins controlling cell cycle progression in human cells: comparison of the MCF-7 adenocarcinoma and the SCL-2 squamous cell carcinoma cell line." Int J Radiat Biol. 70(3): 261-71. Hall, E. (1994). Radiobioloqy for the Radiobioloqist, J. B. Lippincott Company. Hall, M., S. Bates, et al. (1995). "Evidence for different modes of action of cyclin-dependent kinase inhibitors: p15 and p16 bind to kinases, p21 and p27 bind to cvclins." Oncogene 11(8): 1581-8. Hall, M. and G. Peters (1996). "Genetic alterations of cyclins, cyclin-dependent kinases, and cdk inhibitors in human cancers." Adv Cancer Res 68: 67-108. Hanahan, D. and R. Weinberg (2000). "The hallmarks of cancer." CeN 100: 57-70. Hara, K., Y. Maruki, et al. (2002). "Raptor, a binding partner of target of rapamycin (TOR), mediates TOR action." Cell 110: 177-189. Hartwell, L. (1992). "Defects in a cell cycle checkpoint may be responsible for the genomic instability of cancer cell." Cell 71: 543-546. Hay, N. (2005). "The Akt-mTOR tango and its relevance to cancer." Cancer Cell 8: 179-183. Heitman, J., N. Movva, et al. (1991). "Targets for cell cycle arrest by the immunosuppressant rapamycin in yeast." Science 253: 905-909. Hidalgo, M. and E. Rowinsky (2000). "The rapamycin-sensitive signal transduction pathway as a target for cancer therapy." Oncogene 19: 6680-6686. 179 Higashikubo, R., M. Ragouzis, et al. (1996). "Flow cytometric BrdUrd-pulse-chase study of X-ray-induced alterations in cell cycle progression." Cell Prolif. 29(1): 43-57. Hirst, D. (1986). "Oxygen delivery to tumors." Int J Radiat Oncol Biol Phvs. 12(8): 1271-7. Hockel, M. and P. Vaupel (2001). "Tumor hypoxia: Definitions and current clinical, biologic, and molecular aspects." J. Natl. Cancer Inst. 93: 266-276. Hodgkiss, R., L. Webster, et al. (1997). "Measurement of hypoxia in vivo using a 2-nitroimidazole (NITP)." Adv Exp Med Biol. 428: 61-7. Holland, E., J. Celestino, et al. (2000). "Combined activation of Ras and Akt in neural progenitors induces glioblastoma formation in mice." Nat. Genet. 25: 55-57. Horas, J., O. OR, et al. (2005). "On the surviving fraction in irradiated multicellular tumour spheroids: calculation of overall radiosenitivity parameters, influence of hypoxia and volume effects." Phvs Med Biol 50: 1689-1701. Hosoi, H., M. Dilling, et al. (1998). "Studies on the mechanism of resistance to rapamycin in human cancer cells." Mol. Pharmacol. 54(5): 815-824. Houba-Herin, N., C. Calberg-Bacq, et al. (1984). "Photodynamic activity of acridine orange: peroxide radical induction in DNA and synthetic polynucleotides." Int J Radiat Biol Relat Stud Phvs Chem Med 45(5). Huang, P., P. Olive, et al. (1998). "Use of the comet assay for assessment of drug resistance and its modulaiton in vivo." British Journal of Cancer 77(3): 412-416. Huang, S. and P. Houghton (2003). "Targeting mTOR signaling for cancer therapy." Curr. Opin. Pharmacol. 3(4): 371-377. Hunter, T. and J. Pines (1994). "Cyclins and Cancer II: cyclin D and CDK inhibitors come of age." CeH 79: 573-582. Inch, W., J. McCredie, et al. (1970). "Growth of nodular carcinomas in rodents compared with multi-cell spheroids in tissue culture." Growth 34(3): 271-82. Jacinto, E., R. Loewith, et al. (2004). "Mammalian TOR complex 2 controls the actin cytoskeleton and is rapamycin insensitive." Nat. Cell Biol. 6: 1122-1128. Jayaraman, T. and A. Marks (1993). "Rapamycin-FKBP12 blocks proliferation, induces differentiation, and inhibits cdc2 kinase activity in a myogenic cell line." J Biol Chem. 268(34): 25385-8. Jeney, A., I. Kovalszky, et al. (1995). "The biological activity of cisplatin and dibromdulcitol in combination therapy." British Journal of Cancer 71: 317-321. Jensen, P., B. Mortensen, et al. (2000). "Increased cellular hypoxia and reduced proliferation of both normal and leukaemic cells during progression of acute myeloid leukaemia in rats." Cell Prolif. 33(6): 381-95. Kawamata, S., H. Sakaida, et al. (1998). "The upregulation of p27Kip1 by rapamycin results in G1 arrest in exponentially growing T-cell lines." Blood 91(2): 561-9. Kennedy, A., J. Raleigh, et al. (1997). "Proliferation and hypoxia in human squamous cell carcinoma of the cervix: first report of combined immunohistochemical assays." Int J Radiat Oncol Biol Phvs. 37: 897-905. Kesari, S., N. Ramakrishna, et al. (2005). "Targeted molecular therapy of malignant gliomas." Curr Neurol Neurosci Rep. 5(3): 186-97. Kim, D., D. Sarbassov, et al. (2002). "mTOR interacts with raptor to form a nutrient-sensitive complex that signals to the cell growth machinery." Cell 110: 163-175. King, R., P. Jackson, et al. (1994). "Mitosis in Transition." Cell 79: 563-71. Klionsky, D. and S. Emr (2000). "Autophagy as a regulated pathway of cellular degradation." Science 290: 1717-21. Koch, C , J. Kruuv, et al. (1973). "The effect of hypoxia on the generation time of mammalian cells." Radiat Res. 53(1): 43-8. 180 Kochling, J., P. Curtin, et al. (1998). "Regulation of human erythropoietin gene induction by upstream flanking sequences in transgenic mice." 103(4): 960-8. Komata, T., T. Kanzawa, et al. (2003). "Antitumour effect of cyclin-dependent kinase inhibitors (p16INK4A, p18INK4C, p19INK4D, p21WAF1/CIP1 and p27KIP1) on malignant glioma cells." Br J Cancer 88: 1277-80. Kubbies, M. (1990). "Flow Cytometric Recognition of Clastogen Induced Chromatin Damage in G0/G1 Lymphocytes by Non-Stoichiometric Hoechst Fluorochrome Binding." Cytometry 11 (2): 386-394. LaBaer, J., M. Garrett, et al. (1997). "New functional activities for the p21 family of CDK inhibitors." Genes Dev. 11(7): 847-62. Laird, A. (1964). "Dynamics of tumor growth." Br J Cancer 13: 490-502. LaRue, K., M. Khalil, et al. (2004). "Microenvironmental Regulation of Proliferation in Multicellular Spheroids Is Mediated through Differential Expression of Cyclin-Dependent Kinase Inhibitors." Cancer Res 64: 1621-1631. Law, M., E. Forrester, et al. (2006). "Rapamycin disrupts cyclin/cyclin-dependant kinase/p21/proliferating cell nuclear antigen complexes and cyclin D1 reverses rapamycin action by stabilizing these complexes." Cancer Res 66(2): 1070-1080. Lebo, R. (1982). "Chromosome Sorting nad DNA Sequence Localization." Cytometry 3(3): 145-154. Lenormand, J., R. Dellinger, et al. (1999). "Speedy: a novel cell cycle regulator of the G2/M transition." EMBO 18: 1869-77. LePard, R. E. D. a. N. (1994). "Modulation of tumor hypoxia by conventional chemotherapeutic agents." International journal of radiation oncology biology physics. 29(3): 481-486. LePard, R. E. D. a. N. (1995). "Contribution of transient blood flow to tumor hypoxia in mice." Acta Oncolooica 34(3): 317-323. Ling, L. and R. Sutherland (1987). "Dependence of misonidazole binding on factors associated with hypoxic metabolism." Br J Cancer 56: 389-93. Liotta, L. and E. Kohn (2001). "The microenvironment of the tumour-host interface." Nature 411: 375-79. Liotta, L. and E. Kohn (2003). "Cancer's deadly signature." Nat Genet. 33(1): 10-11. Ljungkvist, A., J. Bussink, et al. (2006). "Dynamics of hypoxia, proliferation and apoptosis after irradiation in a murine tumor model." Radiat Res. 165(3): 326-36. Lloyd, H. (1975). "Esitmation of tumor cell kill from Gompertz growth curves." Cancer Chemother. Rep. 59: 267-77. Loewith, R., E. Jacinto, et al. (2002). "Two TOR complexes, only one of which is rapamycin sensitive, have distinct roles in cell growth control." Mol. Cell 10: 457-468. Lord, E., L. Harwell, et al. (1993). "Detection of hypoxic cells by monclonal antibody recognizing 2-nitroimidazole adducts." Cancer Res. 53: 5721-5726. Luca, F., E. Shibuya, et al. (1991). "Both cyclin A delta 60 and B delta 97 are stable and arrest cells in M-phase, but only cyclin B delta 97 turns on cyclin destruction." EMBO J. 10.(13): 4311-20. Luk, C. and R. Sutherland (1987). "Nutrient modification of proliferation and radiation response in EMT6/R0 spheroids." Int J Radiat Oncol Biol Phys. 13(6): 885-95. Lund-Johansen, M., R. Bjerkvig, et al. (1992). Tumor spheroids from monolayer culture. Spheroid Culture in Cancer Research. R. Bjerkvig, CRC Press Boca Raton: 4-18. Luo, Y., S. Marx, et al. (1996). "Rapamycin resistance tied to defective regulation of p27Kip1." Mol Cell Biol 16(12): 6744-51. Mackillop, W., A. Ciampi, et al. (1983). "A stem cell model of human tumor growth: implication for tumor cell clonogenic assays." J Natl Cancer Inst 70(1): 9-16. 181 Maebayashi, K., N. Mitsuhashi, et al. (1999). "p53 mutation decreased radiosensitivity in rat yolk sac tumor cell lines." Int J Radiat Oncol Biol Phvs. 44(3): 677-82. Maity, A., W. McKenna, et al. (1994). "The molecular basis for cell cycle delays following ionizing radiation: a review." Radiother Oncol. 31(1): 1-13. McCredie, J., W. Inch, et al. (1965). "The rate of tumor growth in animals." Growth 29: 331-347. Mocharla, R., H. Mocharla, et al. (1987). "A novel, sensitive fluorometric staining technique for the detection of DNA in RNA preparations." Nucl. Acids Res. 15: 10589. Morgan, D. O. (1995). "Principles of CDK regulation." Nature 374: 131-134. Morice, W., G. Wiederrecht, et al. (1993). "Rapamycin inhibition of interleukin-2-dependent p33cdk2 and p34cdc2 kinase activation in T lymphocytes." J Biol Chem 268(30): 22737-45. Mueller-Klieser, W. (1987). "Multicellular spheroids. A review on cellular aggregates in cancer research." J. Cancer Res. Clin. Oncol. 113: 101-122. Mueller-Klieser, W. (1997). "Three-dimensional cell cultures: from molecular mechanisms to clinical applications." Am. J. Physiol. Cell. Physiol. 42: C1109-C1123. Nagasawa, H., P. Keng, et al. (1998). "Absence of a radiation-induced first-cycle G1-S arrest in p53+ human tumor cells synchronized by mitotic selection." Cancer Res. 58(9): 2036-41. Nave, B., M. Ouwens, et al. (1999). "Mammalian target of rapamycin is a direct target for protein kinase B: identification of a convergence point for opposing effects of insulin and amino-acid deficiency on protein translation." Biochem J 344(Pt 2): 427-31. Ng, C , S. Banerjee, et al. (1999). "p53 status, cellular recovery and cell cycle arrest as prognosticators of in vitro radiosensitivity in human pancreatic adenocarcinoma cell lines." Int J Radiat Biol. 75(11): 1365-76. Nioka, S. and B. Chance (2005). "NIR spectroscopic detection of breast cancer." Technol Cancer Res Treat. 4(5): 497-512. Noguchi, P., R. Wallace, et al. (1979). "Characterization of the WIDR: a human colon carcinoma cell line." In Vitro 15(16): 401-8. Nojima, H., C. Tokunaga, et al. (2003). "The mammalian target of rapamycin (mTOR) partner, raptor, bindes the mTOR substrates p70 S6 kinase and 4E-BP1 through their TOR signaling (TOS) motif." J. Biol. Chem. 278: 15461-15464. Norton, L, R. Simon, et al. (1976). "Predicting the course of Gompertzian growth." Nature 264(5586): 542-5. Nourse, J., E. Firpo, et al. (1994). "lnterleukin-2-mediated elimination of the p27Kip1 cyclin-dependent kinase inhibitor prevented by rapamycin." Nature 372(6506): 570-3. Nurse, P. (1994). "Ordering S Phase and M Phase in the Cell Cycle." Cell 79: 547-550. Ohtsubo, T., X. Wang, et al. (1997). "p53-dependent induction of WAF1 by a low-pH culture condition in human glioblastoma cells." Cancer Res. 57: 3910-3913. Olive, P., J. Banath, et al. (1996). "Development of Apoptosis and polyploidy in human lymphoblast cells as a function of position in the cell cycle at he tie of irradiation." Radiation Research 146: 595-602. Olive, P., J. Banath, et al. (1997). "Detection of subpopulations resistant to DNA-damaging agents in spheroids and murine tumours." Mutation Research 375: 157-165. Olive, P., J. Banath, et al. (2002). "The range of oxygenation in SiHa tumor xenografts." Radiat Res. 158(2): 159-66. 182 Olive, P., D. Chaplin, et al. (1985). "Pharmacokinetics, binding and distribution of Hoechst 33342 in spheroids and murine tumours." British Journal of Cancer 52: 739-746. Olive, P. and R. Durand (1994). "Drug and radiation resistance in spheroids: cell contact and kinetics." Cancer Metastasis Rev. 13: 121-138. Olive, P. and R. Durand (1997). "Apoptosis: an indicator of radiosensitivity in vitro?" International Journal Radiation Biology 71(6): 695-707. Olive, P., P. Johnston, et al. (1998). "The comet assay: a new method to examine heterogeneity associated with solid tumors." Nature medicine 4(1): 103-104. Olive, P., C. Luo, et al. (2002). "Local hypoxia is produced at sites of intratumour injection." Br J Cancer 86(3): 429-35. Oloumi, A., W. Lam, et al. (2002). "Identification of genes differentially expressed in V79 cells grown as multicellular spheroids." Int J Radiat Biol 78(6): 483-92. Oloumi, A., S. MacPhail, et al. (2000). "Changes in subcellular distribution of topoisomerase llalpha correlate with etoposide resistance in multicell spheroids and xenograft tumors." Cancer Research 60: 5747-5753. O'Reilly, K., F. Rojo, et al. (2006). "mTOR Inhibition Induces Upstream Receptor Tyrosine Kinase Signaling and Activates Akt." Cancer Res 66(3): 1500-1508. Owa, T., H. Yoshino, et al. (2001). "Cell cycle regulation in the G1 phase: a promising target for the development of new chemotherapeutic anticancer agents." Curr Med Chem 8(12): 1487-503. Pardo, F., M. Su, et al. (1996). "Cyclin D1 induced apoptosis maintains the integrity of the G1/S checkpoint following ionizing radiation irradiation." Somat Cell Mol Genet. 22(2): 135-44. Peschoa, D., D. Stefanou, et al. (2005). "Cell cycle proteins in laryngeal cancer: role in proliferation and prognosis." J Exp Clin Cancer Res 24(3): 431-7. Peters, L. and H. Withers (1997). "Applying radiobiological principles to combined modality treatment of head and neck cancer-the time factor." Int J Radiat Oncol Biol Phvs. 39(4): 831-6. Petrocelli, T., R. Poon, et al. (1996). "UVB radiation induces p21Cip1A/VAF1 and mediates G1 and S phase checkpoints." Oncogene 12(7): 1387-96. Pollack, A., R. White, et al. (1993). "Calculating potential doubling time using monoclonal antibodies specific for two halogenated thymidine analogues." Int. J. Radiation Oncology Biol. Phvs. 27: 1131-1139. Radu, L., D. Gazdaru, et al. (2004). "Spectroscopic study of fast-neutron-irradiated chromatin." Can J Physiol Pharmacol. 82(2): 79-83. Raghunand, N., R. Gatenby, et al. (2003). "Microenvironmental and cellular consequences of altered blood flow in tumours." Br J Radiol 76: S11-S22. Raleigh, J., G. Miller, et al. (1987). "Fluorescence immunohistochemical detection of hypoxic cells in spheroids and tumours." Br. J. Cancer 56: 395-400. Raught, B., A. Gingras, et al. (2001). "The target of rapamycin (TOR) proteins." Proc. Natl. Acad. Sci. USA 98(13): 7037-7044. Reed, S. (1997). "Control of the G1/S transition." Cancer Surv. 29: 7-23. Reed, S., E. Bailly, et al. (1994). "G1 control in mammalian cells." J Cell Sci Suppl. 18: 69-73. Reinke, A., S. Anderson, et al. (2004). "TOR complex 1 includes a novel component, Tco89p (YPL180w), and cooperates with Ssdlp to maintain cellular integrity in Saccharomyces cerevisiae." J. Biol. Chem. 279: 14752-14762. Revesz, L., E. Siracka, et al. (1989). "Variation of vascular density within and between tumours of the uterine cervix andits predictive value for radiotherapy." Int. J. Radiat. Oncol. Biol. Phvs. 16: 1161-1163. 183 Reynisdottir, I., K. Polyak, et al. (1995). "Kip/Cip and Ink4 Cdk inhibitors cooperate to induce cell cycle arrest in response to TGF-beta." Genes Dev. 9(15): 1831-45. Rijken, P., H. Bernsen, et al. (2000). "Spatial relationship between hypoxia and the (perfused) vascular network in a human glioma xenograft: a quantitative multi-parameter analysis." Int J Radiat Oncol Biol Phvs. 48(2): 571-82. Rodriguez-Pinilla, M., J. Rodriguez-Peralto, et al. (2004). "Cyclin A as a Predictive Factor for Chemotherapy Response in Advance Head and Neck Cancer." Clin Cancer Res 10: 8486-8492. Rofstad, E. (2000). "Microenvironment-induced cancer metastasis." Int. J. Radiat. Biol. 76: 589-65. Rosai, J. (1996). Ackerman's Surgical Pathology. St. Louis, Mosby-Year Book. Sandhu, C , J. Garbe, et al. (1997). "Transforming growth factor beta stabilizes p15INK4B protein, increases p15INK4B-cdk4 complexes, and inhibits cyclin D1-cdk4 association in human mammary epithelial cells." Mol Cell Biol. 17(5): 2458-67. Santini, M. and G. Rainaldi (1999). "Three-dimensional spheroid model in tumor biology." Pathobiology 67: 148-157. Sarbassov, D., S. Ali, et al. (2004). "Rictor, a novel binding partner of mTOR, defines a rapamycin-insensitive and raptor-independent pathway that regulates the cytoskeleton." Curr. Biol. 14: 1296-1302. Sarbassov, D., D. Guertin, et al. (2005). "Phosphorylation and Regulation of Akt/PKB by the Rictor-mTOR Complex." Science 307: 1098-1101. Sarkaria, J., R. Tibbetts, et al. (1998). "Inhibition of phosphoinositide 3-kinase related kinases by the radiosensitizing agent wortmannin." Cancer Res 58(19): 4375-82. Sawyers, C. (2003). "Will mTOR inhibitors make it as cancer drugs?" Cancer Cell 4: 343-348. Schalm, S., D. Fingar, et al. (2003). "TOS motif-mediated raptor binding regulates 4E-BP1 multisite phosphorylation and function." Curr. Biol. 13: 797-806. Schmelzle, T. and M. Hall (2000). "TOR, a central controller of cell growth." Cell 103: 253-262. Schwartz, D., N. Almog, et al. (1997). "Role of wild type p53 in the G2 phase: regulation of the gamma-irradiation-induced delay and DNA repair." Oncogene 15(21): 2597-607. Scott, P., G. Brunn, et al. (1998). "Evidence of insulin-stimulated phosphorylation and activation of the mammalian target of rapamycin mediated by a protein kinase B signaling pathway." Proc Natl Acad Sci USA 95(13): 7772-7. Sham, E. and R. Durand (1998). "Cell kinetics and repopulation mechanisms during multifraction irradiation of spheroids." Radiotherapy and oncology 46: 201-207. Sham, E. and R. Durand (1999). "Cell kinetics and repopulation parameters of irradiated xenograft tumors in scid mice: comparison of two dose-fractionation regimens." European Journal of Cancer 35(5): 850-858. Sham, E. and R. Durand (1999). "Repopulation characteristics and cell kinetic parameters resulting from multi-fraction irradiationof xenograft tumors in SCID mice." Int. Journal of Radiation Oncology Biology Physics 43(3): 617-22. Shamji, A., P. Nghiem, et al. (2003). "Integrationof growth factor and nutrient signaling: implication for cancer biology." Mol. Cell 12: 271-280. Sherr, C. (1994). "G1 phase progression: cycling on cue." Cell 79: 551-555. Sherr, C. (1995). "Inhibitors of mamalian G1 cyclin-dependent kinases." Genes Dev 9: 1149-63. Sherr, C. (1996). "Cancer Cell Cycles." Science 274: 1672-77. 184 Sherr, C. (2000). "The Pezcoller Lecture: Cancer Cell Cycles Revisited." Cancer Res 60: 3689-95. Shinohara, E., C. Cao, et al. (2005). "Enhanced radiation damage of tumor vasculature by mTOR inhibitors." Oncogene 24: 5414-22. Siracka, E., L. Revesz, et al. (1988). "Vascular density in carcinoma of the uterine cervix and its predictive value for radiotherapy." Int. J. Cancer 41: 819-822. Skarsgard, L., M. Skwarchuk, et al. (1996). "Substructure in the radiation survival response at low dose in cells of human tumor cell lines." Radiation research: 388-397. Solomon, M. (1993). "Activation of the various cyclin/cdc2 protein kinases." Curr Opin Cell Biol. 5(2): 180-6. Sonenberg, N. and A. Gingras (1998). "The mRNA 5' cap-binding protein elF4E and control of cell growth." Curr Opin Cell Biol 10(2): 268-75. Sonoda, Y., T. Ozawa, et al. (2001). "Akt pathway activation converts anaplastic astrocytoma to glioblastoma multiforme in a human astrocyte model of glioma." Cancer Res. 61: 6674-6678. Steel, G. (1977). Growth kinetics of tumours: Cell population kinetics in relation to the growth and treatment of cancer. Oxford, Clarendon Press. Stone, H., W. McBride, et al. (2002). "Modifying normal tissue damage postirradiation report of a workshop sponsored by the Radiation Research program, National Cancer Institute, Bethesda, Maryland." Radiat Res. 157: 204-223. Streffer, O, S. Henstebeck, et al. (1980). Glucose metabolism in liver and an adenocarcinoma of mice with and without hyperthermia, Henry Ford Hospital, Special Issue: 77. Sutherland, R. (1986). "Importance of critical metabolites and cellular interactions in the biology of microregions of tumors." Cancer 58(8): 1668-80. Sutherland, R. (1988). "Cell and environment interactions in tumor microregions: the multicell spheroid model." Science 240: 177-184. Sutherland, R., W. Ausserer, et al. (1996). "Tumor hypoxia and heterogeneity: challenges and opportunities for the future." Semin. Radiat. Oncol. 6: 59-70. Sutherland, R., J. Carlsson, et al. (1981). "Spheroids in cancer research." Cancer Research 41: 2980-84. Sutherland, R. and R. Durand (1976). "Radiation response of multicell spheroids - an in vitro tumour model." Curr, Topics Radiat. Res. Q. 11: 87-139. Sutherland, R., J. McCredie, et al. (1971). "Growth of mulicell spheroids in tissue culture as a model of nodular carcinomas." J Natl Cancer Inst. 46(1): 113-20. Swartling, J., A. Bassi, et al. (2005). "Dynamic time-resolved diffuse spectroscopy based on supercontinuum light pulses." Appl Opt. 44(22): 4684-92. Sweigert, S. and E. Alpen (1986). "Dynamics of cell kinetic parameters during 9L spheroid growth." Cell Tissue Kinet. 19(5): 567-76. Tannock, I. (1968). "The relation between cell proliferation and the vascular system in a transplanted mouse mammary tumor." Br J Cancer 22: 258. Tannock, I. (1998). "Conventional cancer therapy: promise broken or promise delayed?" Lancet 351(Suppl 2): SN 9-16. Tannock, I. and I. Kopelyan (1986). "Influence of glucose concentration on growth and formation of necrosis in spheroids derived from a human bladder cancer cell line." Cancer Res 46(6): 3105-10. Taylor, I. and P. Hodson (1984). "Cell cycle regulation by environmental pH." J Cell Physiol. 121(3): 517-25. Terry, N., R. White, et al. (1990). "A new method to calculate the potential doubling times of humna squamous cell carcinomas." Cytometry 99: Suppl 4. 185 Thomlinson, R. and L. Gray (1955). "The historical structure of some human lung cancers and the possible implications for radiotherapy." Br J Cancer 9: 539-549. Torricelli, A., V. Quaresima, et al. (2004). "Mapping of calf muscle oxygenation and haemoglobin content during dynamic plantar flexion exercise by multi-channel time-resolved near-infrared spectroscopy." Phys Med Biol. 49(5): 685-99. Trott, K. (1990). "Cell repopulation and overall treatment time." Int J Radiat Oncol Biol Phvs 19(4): 1071-5. Trott, K. (1994). "Tumour stem cells: the biological concept adn its application in cancer treatment." Radiother Oncol 30(1): 1-5. Trott, K. (1999). "The mechanisms of acceleration of repopulation in squamous epithelia during daily irradiation." Acta Oncol 38(2): 153-7. Trott, K. and J. Kummermehr (1993). "The time factor and repopulation in tumors and normal tissues." SeminRadiat Oncol 3(2): 115-125. Urtasun, R., J. Chapman, et al. (1986). "Binding of 3H-misonidazole to solid human tumors as a measure of tumor hypoxia." Int J Radiat Oncol Biol Phys. 12(7): 1263-7. Urtasun, R., C. Koch, et al. (1986). "A novel technique for measuring human tissue p02 at the cellular level." Br J Cancer 54(3): 453-7. Vaupel, P., F. Kallinowski, et al. (1989). "Blood flow, oxygen and nutrient supply, and metabolic microenvironment of human tumors: a review." Cancer Res. 49(23): 6449-65. Vignot, S., S. Faivre, et al. (2005). "mTOR-targeted therapy of cancer with rapamycin derivatives." Ann Oncol. 16(4): 525-37. Vogelstein, B. and K. Kinzler (2004). "Cancer genes and the pathways they control." Nature Med 10(8): 789-799. Walenta, S. and W. Mueller-Klieser (1987). "Oxygen consumption rate of tumour cells as a function of their proliferative status." Adv Exp Med Biol. 215: 389-91. Webster, L, R. Hodgkiss, et al. (1998). "Cell cycle distribution of hypoxia and progression of hypoxic tumour cells in vivo." Br J Cancer 77(2): 227-34. Weinberg, R. (1995). "The retinoblastoma protein and cell cycle control." Cell 81: 323-330. White, R. (1989). "Computing multiple cell kinetic properties from a single time point." J Theor Biol 141: 429-446. White, R. and M. Meistrich (1986). "A comment on "A method to measure the duration of DNA synthesis and the potential doubling time from a single sample." Cytometry 7: 486-490. White, R., A. Pollack, et al. (1994). "Double labeling to obtain S phose subpopulations: applicaiton to determine cell kinetics of diploid cells in an aneuploid tumour." Cell Proliferation 27: 123-137. White, R. and N. Terry (1992). "A quantitative method for evaluating bivariate flow cytometric data obtained using monoclonal antibodies to bromodeoxyuridine." Cytometry 13(5): 490-5. White, R., N. Terry, et al. (1990). "Improved method for computing potential doubling time from flow cytometric data." Cytometry 11: 314-317. Wibe, E., T. Lindmo, et al. (1981). "Cell kinetic characteristics in different parts of multicellular spheroids of human origin." Cell Tissue Kinet. 141 (6): 639-51. Wiederrecht, G., C. Sabers, et al. (1995). "Mechanism of action of rapamycin: new insights into the regulation of G1-phase progression in eukaryotic cells." Prog Cell Cycle Res. 1: 53-71. 186 Winter, S., K. Shah, et al. (2005). "Relation of erythropoietin and erythropoietin receptor expression to hypoxia and anemia in head and neck squamous cell carcinoma." Clin Cancer Res. 11(21): 7614-20. Withers, H., J. Taylor, et al. (1988). "The hazard of accelerated tumor clonogen repopulation during radiotherapy." Acta Oncol 27(2): 131-46. Witz, I. (2002). "The tumour microenvironment-lntroduction." Semin Cancer Biol. 12: 87-88. Wullschleger, S., R. Loewith, et al. (2006). "TOR Signaling in Growth and Metabolism." Cell 124: 471-484. Yen, A. and A. Pardee (1978). "Exponential 3T3 cells escape in mid-G1 from their high serum requirement." Exp Cell Res. 116(1): 103-13. Yu, C , M. Rahmani, et al. (2003). "The lethal effects of pharmacological cyclin-dependent kinase inhibitors in human leukemia cells proceed through a Phosphatidylinositol 3-kinase/Akt-dependent process." Cancer Res 63: 1822-1833. Yu, K., L. Toral-Barza, et al. (2001). "mTOR, a novel target in breast cancer: the effect of CCI-779, an mTOR inhibitor, in preclinical models of breast cancer." Endocr Relat Cancer 8(3): 249-58. Zanghi, J. A., M. Fussenegger, et al. (1999). "Serum protects protein-free competent Chinese hamster ovary cells against apoptosis induced by nutrient deprivation in batch culture." Biotechnology and Bioenqineerinq 64(1): 108-119. Zhao, S., T. Tsuchida, et al. (2002). "Effect of As203 on cell cycle progression and cyclins D1 and B1 expression in two glioblastoma cell lines differing in p53 status." Int J Oncol 21 (1): 49-55. 187 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0092948/manifest

Comment

Related Items