Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

G₂ DNA damage checkpoint : inhibitors and there potential for anticancer treatment Sturgeon, Christopher Michael 2006

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

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata

Download

Media
831-ubc_2006-201039.pdf [ 15.67MB ]
Metadata
JSON: 831-1.0093035.json
JSON-LD: 831-1.0093035-ld.json
RDF/XML (Pretty): 831-1.0093035-rdf.xml
RDF/JSON: 831-1.0093035-rdf.json
Turtle: 831-1.0093035-turtle.txt
N-Triples: 831-1.0093035-rdf-ntriples.txt
Original Record: 831-1.0093035-source.json
Full Text
831-1.0093035-fulltext.txt
Citation
831-1.0093035.ris

Full Text

G2 DNA Damage Checkpoint Inhibitors and their potential for anticancer treatment by CHRISTOPHER M I C H A E L STURGEON B.Sc.(H), Carleton University, 2001 A THESIS SUBMITTED IN PARTIAL F U L F I L L M E N T OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE F A C U L T Y OF G R A D U A T E STUDIES (Biochemistry and Molecular Biology) THE UNIVERSITY OF BRITISH C O L U M B I A October 2006 © Christopher Michael Sturgeon, 2006 ABSTRACT In response to D N A damage, cell survival can be enhanced by activation of D N A repair mechanisms and of checkpoints that delay cell cycle progression to allow more time for D N A repair. G 2 checkpoint inhibitors can force cells arrested in G 2 phase by D N A damage to enter mitosis thereby enhancing the killing of cancer cells by DNA-damaging approaches. Checkpoint inhibitors represent a promising class of therapeutic agents in the treatment of cancer. The goals of my research were to i) characterize novel checkpoint inhibitors in terms of their therapeutic potential and, i f possible, mode of action, ii) increase the therapeutic properties of checkpoint inhibitors by combining them with D N A repair inhibitors, and iii) develop a therapeutic use of the checkpoint inhibitor, isogranulatimide. Two novel checkpoint inhibitors were psilostachyin and cryptofolione. The target of psilostachyin may be Weel, a kinase which inhibits CDK1. Cryptofolione shares structural and biological properties with leptomycin B, a highly specific inhibitor of Crml -mediated nuclear export. This suggests nuclear export inhibition may be a novel means to achieve checkpoint inhibition. Inhibiting both repair and the checkpoint with drugs might cause cancer cells to undergo cell division in the presence of lethal amounts of unrepaired DNA. However, interfering with D N A repair via inhibition of DNA-dependent protein kinase (DNA-PK) reduces the ability of checkpoint inhibitors to abrogate G 2 arrest, as well as their radiosensitizing activity. Components of the checkpoint pathway were highly 11 overactivated in this situation. This suggests that combining checkpoint inhibition with targeted D N A repair inhibition may not be therapeutically viable. Varying the time of addition of checkpoint inhibitors following D N A damage revealed that G2 checkpoint abrogation alone is insufficient for radiosensitization. Instead, checkpoint inhibitors must be present for the duration of S-phase progression to achieve radiosensitization. Taken together, these studies describe the search for and development of checkpoint inhibitors, ideally for future therapeutic use. These results have relevance to the development of G2 checkpoint inhibitors as experimental therapeutic approaches to the treatment of cancer. This research represents a major step forward in the pre-clinical development of checkpoint inhibitors as anticancer agents. 111 TABLE OF CONTENTS ABSTRACT n TABLE OF CONTENTS iv LIST OF TABLES viii LIST OF FIGURES ix LIST OF ABBREVIATIONS xi PREFACE xv ACKNOWLEDGEMENTS xvii Chapter 1 Introduction 1 1.1 Cell cycle 1 1.1.1 Specific phases of the cell cycle 1 1.1.2 Coordinated regulation of the cell cycle 4 1.1.2.1 Cyclin Dependent Kinases 4 1.1.2.2 Cyclins 5 1.1.2.3 The G 2 /M transition 5 1.2 D N A damage 7 1.2.1 Repair of damaged D N A 8 1.2.1.1 NHEJ 10 1.2.1.2 HRR 13 1.3 Cell cycle checkpoints 14 1.3.1 Damage sensors 18 1.3.2 Checkpoint mediators 19 1.3.3 Checkpoint transducers and effectors 20 1.3.4 The Gi/S checkpoint 21 1.3.5 The intra-S checkpoint 22 1.3.6 The G 2 /M checkpoint 23 1.4 Checkpoint inhibitors 24 1.5 This thesis 27 i v C h a p t e r 2 M a t e r i a l s a n d M e t h o d s 29 2.1 Cell lines 29 2.2 Screening for novel checkpoint inhibitors 30 2.3 Isolation of sesquiterpene lactones from Ambrosia artemisiifolia and cryptofoliones from Cryptocaria concinna 30 2.3.1 Sample N35791 30 2.3.2 Sample N29449 30 2.4 Irradiation, and Checkpoint Inhibitor Treatment 30 2.5 Checkpoint inhibitor activity 31 2.5.1 Mitotic spreads 31 2.5.2 Flow cytometry : 32 2.6 Tubulin polymerization assay 33 2.7 Cell Viability Assay 33 2.8 Nuclear import/export assay '. 34 2.8.1 Import 34 2.8.2 Export 34 2.9 Immunofluorescence microscopy 35 2.10 Yeast drug induced haploinsufficiency screen 35 2.11 Western blotting 36 2.11.1 Histone acid extraction 37 2.12 Cell clonogenicity assays 37 2.13 Cell synchronization 37 2.14 Kinase Assays 38 2.15 Kinexus Bioinformatics "Kinetworks" screen 39 2.16 Reverse-Transcriptase Polymerase Chain Reaction (RT-PCR) amplification 40 2.17 Preparation of PKB-TRES-GFP cells 40 2.18 Senescence detection 41 v 2.19 Xenotransplantation studies in scid mice 41 2.20 Statistical analysis 42 Chapter 3 Identification and characterization of novel checkpoint inhibitors 43 3.1 INTRODUCTION 43 3.2 RESULTS A N D DISCUSSION 45 3.2.1 Psilostachyin 45 3.2.1.1 Checkpoint inhibition by psilostachyins 45 3.2.1.2 Mitotic arrest induced by psilostachyin 48 3.2.1.3 Enhancing checkpoint inhibition by combining with other inhibitors 52 3.2.1.4 Psilostachyins fail to radiosensitize cells 54 3.2.1.5 Target of psilostachyin? 57 3.2.2 Cryptofoliones 67 3.2.2.1 Isolation and identification of cryptofoliones 67 3.2.2.2 Inhibition of ionizing radiation-induced G 2 arrest by cryptofolione 69 3.2.2.3 Structure-activity study 71 3.2.2.4 Inhibition of nuclear export by cryptofolione 72 3.2.2.5 Potent G 2 checkpoint inhibition by leptomycin B 76 3.3 DISCUSSION 76 3.4 S U M M A R Y 80 Chapter 4 Combining DNA repair inhibition and checkpoint inhibition to increase therapeutic effect 83 4.1 INTRODUCTION 83 4.2 RESULTS A N D DISCUSSION 85 4.2.1 D N A - P K inhibition and NHEJ deficiency decrease the ability of cancer cells to respond to G 2 checkpoint inhibitors 85 4.2.2 D N A - P K inhibition leads to reduced radiosensitization by checkpoint inhibitors 94 4.2.3 DNA-PK- and NHEJ-impaired cells enter prolonged G 2 arrest following D N A damage 96 4.2.4 D N A - P K is not required in G 2 for normal G 2 checkpoint recovery 105 4.2.5 IR causes overactivation of Chkl and Chk2 in cells lacking D N A - P K activity 105 4.2.6 Kinexus screen to further clarify mechanism 109 4.2.6.1 M K K 6 does not play a role in M059J prolonged arrest 116 4.2.6.2 P K B does not play a role in M059J prolonged arrest 118 4.2.7 M059J cells enter a senescent-like state after D N A damage... 126 vi 4.3 DISCUSSION 129 4.4 S U M M A R Y 137 Chapter 5 In vivo development of checkpoint inhibitors 139 5.1 INTRODUCTION 139 5.2 RESULTS A N D DISCUSSION 140 5.2.1 Time of checkpoint inhibitor application affects radiosensitivity 140 5.2.2 Isogranulatimide bypasses the G2 checkpoint in vivo 148 . 5.3 DISCUSSION 152 5.4 S U M M A R Y 154 Chapter 6 Future perspectives 157 Chapter 7 References 162 vii LIST OF TABLES Table 1.1: Currently known checkpoint inhibitors and their targets 26 Table 3 . 1 : Saccharomyces cereviscae heterozygous diploid strains exhibiting impaired growth in the presence of 50 uM OZ 66 Table 3 .2 : Summary of activity of cryptofolione analogs 75 Table 4. 1 : Effect of A M A3 7 and checkpoint inhibitors on the clonogenicity of irradiated and unirradiated MCF-7 cells 95 Table 4. 2 : Kinetworks targets exhibiting increased expression in M059J cells compared to M059K cells 16 h after irradiation 113 Table 4. 3 : Kinetworks targets exhibiting decreased expression in M059J cells compared to M059K cells 16 h after irradiation 114 Table 4. 4 : Kinetworks targets exhibiting no change in expression between M059J and M059K cells 16 h after irradiation 115 Table 4. 5 : Checkpoint inhibition of mER-PKB M059 cells 127 Table 4. 6 : G 2 arrest in PKB-IRES-GFP M059 cells over time 128 Table 5 .1 : HCT 116 p537" cell clonogenicity following IR with checkpoint inhibitors. 143 Table 5. 2 : MCF-7 mp53 cell clonogenicity following IR with checkpoint inhibitors. 144 Table 5 .3 : Summary of G 2 / M population from cell suspensions of subcutaneously implanted HCT116 cells in scid mice 151 viii L I S T O F F I G U R E S Figure 1.1: The cell cycle 2 Figure 1.2: CDK1 regulation of the G2 /M transition 6 Figure 1.3: Simplified model of conversion of a SSB to a DSB during D N A replication 9 Figure 1.4: Simplified model of repair of a D N A DSB by NHEJ 11 Figure 1.5: Simplified model of repair of a D N A DSB by HRR 15 Figure 1.6: General model of checkpoint activation in humans 17 Figure 3. 1 : Structural formulae of sesquiterpene lactones isolated from Ambrosia artemisiifolia 46 Figure 3. 2 : G2 checkpoint inhibition by psilostachyins A and C 47 Figure 3 .3 : Effect of psilostachyin C on cell cycle progression 49 Figure 3 .4 : Psilostachyin C does not affect tubulin polymerization in vitro 51 Figure 3 .5 : Checkpoint inhibition by combinations of checkpoint inhibitors 53 Figure 3 .6 : Inhibition of cell proliferation by psilostachyins 55 Figure 3 .7 : Proposed generation of biotinylated psilotachyin 58 Figure 3 .8 : Psilostachyin C does not inhibit nuclear import or export 60 Figure 3 .9 : Psilostachyin C impairs CENP-E localization during mitosis 61 Figure 3.10: Growth curve of wild-type yeast exposed to various concentrations of the checkpoint inhibitor, OZ 63 Figure 3 .11: Growth of 155 diploid yeast strains lacking 1 allele of each kinase in the yeast genome, exposed to 50 uM OZ 65 Figure 3.12: Structural formulae of cryptofolione and related compounds 68 Figure 3. 13 : G2 checkpoint inhibition by cryptofolione and related compounds 70 Figure 3.14: Inhibition of NES-mediated nuclear export by cryptofolione and related compounds 74 Figure 3. 15 : G2 checkpoint inhibition by leptomycin B 77 Figure 4. 1 : Representative 2-dimensional flow cytometry analyses used to monitor checkpoint inhibitor efficacy 87 Figure 4. 2 : Effect of G2 checkpoint inhibitors on human cells lacking D N A - P K or NHEJ activity 88 Figure 4. 3 : AMA37 has little to no effect on A T M function or expression 89 Figure 4. 4 : Effect of G2 checkpoint inhibitors on human cells lacking D N A - P K or NHEJ activity 92 Figure 4. 5 : Effect of G 2 checkpoint inhibitors on CHO cells lacking D N A - P K or NHEJ activity 93 Figure 4. 6 : Sample histograms showing D N A profiles for all cell lines used 97 Figure 4. 7 : Effect of IR on G2 arrest in human cells lacking D N A - P K or NHEJ activity 99 Figure 4. 8 : Cyclin BI accumulation in G2-arrested cells 101 Figure 4. 9 : Effect of IR on G 2 arrest in M059 cells lacking D N A - P K or NHEJ activity 102 ix Figure 4. 10 : Effect of IR on G 2 arrest in CHO cells lacking D N A - P K or NHEJ activity 104 Figure 4. 11 : Effect of IR on G 2 synchronized M059 cells 106 Figure 4. 12 : Overactivation of Chkl in DNA-PK-deficient cells ; 108 Figure 4. 13 : Overactivation of Chk2 in DNA-PK-deficient cells 110 Figure 4. 14 : Sample western blots from Kinexus screen of irradiated M 0 5 9 K and M059J cells 112 Figure 4. 15 : M K K 6 protein levels are much higher in M059J cells compared to M059K cells 117 Figure 4. 16 : Inhibition of MKK6/p38 does not affect checkpoint inhibitor response in M059J cells 119 Figure 4. 17 : P K B activation in M059J cells is insulin-unresponsive but can be inhibited by the M K K 6 inhibitor SKF86002 or p38 inhibitor SB203580 121 Figure 4. 18 : Model for prolonged arrest in M059J cells 122 Figure 4. 19 : Representative GFP-sorting profiles for M059 cells 124 Figure 4. 20 : P K B activation in mER-PKB M059 cell lines after 4-HT treatment 125 Figure 4. 21 : M059J cells exhibit signs of senescence after IR over time 130 Figure 4. 22 : D N A damage at the time of checkpoint inhibitor application decreases checkpoint inhibitor activity 135 Figure 5 . 1 : General approach to discern optimal time of checkpoint inhibitor application 142 Figure 5 .2 : Cell cycle distribution of cells extracted from subcutaneous tumours in scid mice 150 Figure 5 .3 : General overview of liposomal loading of isogranulatimide 155 x LIST OF ABBREVIATIONS 4-HT 4-hydroxytamoxifen 53BP1 p53-binding protein 1 A-T Ataxia-telangiectasia AGP1 Alpha 1 -acid glycoprotein A T M Ataxia-telangiectasia mutated ATR Ataxia-telangiectasia and Rad3-related ATRIP Ataxia-telangiectasia and Rad3-related interacting protein BASC BRCA1-associated genome surveillance complex BER Base excision repair BSA Bovine serum albumin CAF Caffeine C A K CDK-activating kinase CDC Cell-division cycle C D K Cyclin-dependent kinase CENP-E Centromeric protein E CHO Chinese hamster ovary D B H Debromohymenialdisine D M E M Dulbecco's modified Eagle medium DMSO Dimethyl sulphoxide D N A Deoxyribonucleic acid DNA-PK DNA-dependent protein kinase DNA-PKcs DNA-dependent protein kinase catalytic subunit DSB Double-strand break DTT Dithiothreitol EDTA Ethylenediamine tetraacetic acid ELICA Enzyme-linked immunocytochemical assay ER Estrogen receptor FACS Fluorescence-activated cell sorting FITC Fluorescein isothiocyante g Gravity GFP Green fluorescent protein GR Glucocorticoid receptor GRE Glucocorticoid response element Gy Gray HRR Homologous recombination repair IGR Isogranulatimide IR Ionizing radiation MDC1 Mediator of damage checkpoint 1 MOPS 3-(N-Morpholino)propanesulfonic acid mTOR Mammalian target of rapamycin MTT 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium NER Nucleotide excision repair NHEJ Nonhomologous end-joining NLP1 Nucleoporin-like protein 1 xii NOC Nocodazole OZ 13-hydroxy-15-oxozoapatlin PBS Phosphate-buffered saline PCNA Proliferating cell nuclear antigen PDK Phosphoinositide-dependent kinase PFGE Pulsed field gel electrophoresis PI Propidium iodide PI3-K Phosphatidylinositol-3 kinase PIKK Phosphatidylinositol-3 kinase-like kinase PKB Protein kinase B PKC Protein kinase C PP1/2A Protein phosphatase 1/2A RFC Replication factor C R N A Ribonucleic acid RT-PCR Reverse-transcriptase polymerase chain reaction SA (3-gal Senescence-associated acidic p-galactosidase SAB Sodium azide buffer SB SB203580 SD Standard deviation SDS Sodium dodecyl sulphate SKF SKF86002 SSB Single-strand break TBS Tris-buffered saline Tris(hydroxymethyl)aminomethane xiv P R E F A C E This thesis is presented in six chapters. A general overview of the subject area is presented in Chapter 1. Chapter 2 contains the Materials and Methods for all the experiments described in subsequent chapters. Chapters 3 through 5 were written in the general format of a scientific article (i.e. "manuscript format"), as much of the data presented in this thesis have been, or are intended to be, published in scientific journals. Chapter 3 is almost entirely composed of 2 publications, with co-authors Dr. Kyle Craig, Geoffrey Karjala, Colleen Brown, Dr. Natalie Rundle, Dr. Raymond Andersen and Dr. Michel Roberge (Sturgeon et al, 2005), and Dr. Ana Diaz-Marrero, Dr. Bruno Cinel, Lianne McHardy, Michelle Ngo, Dr. Raymond Andersen and Dr. Michel Roberge (Sturgeon et al, 2006a). In this chapter, all data presented is my own work, except for the tubulin polymerization assay performed by Geoffrey Karjala (Figure 3.4), and the chemical isolation, performed by Dr. Kyle Craig, Dr. Bruno Cinel and Dr. Ana Diaz-Marrero in the laboratory of Dr. Ray Andersen. The original identification of psilostachyin and cryptofolione as checkpoint inhibitors was performed before I joined the Roberge laboratory (by Colleen Brown, Dr. Natalie Rundle, Michelle Ngo and Lianne McHardy). Chapter 4 is mostly composed of one publication, with co-authors Dr. Zachary Knight, Dr. Kevan Shokat, and Dr. Michel Roberge (Sturgeon et al, 2006c). A l l work presented in this chapter is my own. AMA37 was provided by Dr. Zachary Knight in the laboratory of Dr. Kevan Shokat, and the Kinexus screen was performed by employees at Kinexus Bioinformatics Corp. through collaboration with Dr. Steven Pelech. xv Chapter 5 is entirely my work, and is partly composed of one publication (Sturgeon and Roberge, 2006d). The animal studies have not yet been submitted for publication at the time of this thesis preparation. Additional work not discussed in this thesis lead to the preparation and publication of a review in (Sturgeon et al., 2006b). x v i A C K N O W L E D G E M E N T S There are many people that need to be thanked for their contribution and help in this thesis. First and foremost, I would like to thank Dr. Michel Roberge for his supervision, guidance, and enthusiasm towards all experiments I undertook. The members of my supervisory committee deserve many thanks for their time, guidance and support (and reference letters!). I extend my appreciation towards all current and past members of both the Roberge and Andersen labs who helped create a friendly, helpful environment: Hilary Anderson, Pamela Austin, Aruna Balgi, Cristina Bigg, Kyle Craig, Ana Diaz-Marrero, Geoffrey Karjala, Danielle Kemmer, Natalie Rundle, Lianne McHardy, Delphine Reberioux, Tamsin Tarling, Yong-Jun Wang, and David Williams. Graduate school would not be possible without the support and friendship of fellow students, particularly Tom, Chris, Cameron, Geoff, Mike and Hyun-Seo. And finally, I would never have been capable of this thesis without the inspiration from my parents, and without the love and support from Kendra. Thank you. xvn Chapter 1 Introduction 1.1 Cell cycle The study of the cell cycle is fundamental to many areas of biology. Its regulation and status can affect many, perhaps all, other cellular functions. When a eukaryotic cell is observed under light microscopy, there are two apparent phases of the cell cycle: mitosis and interphase. The interphase is classified as three general phases: G\, S and G2, each of which must be completed before the next can proceed. A non-dividing cell is not considered to be in the cell cycle, and can be referred to as being in Go. Mitosis is classified as five general phases: prophase, prometaphase, metaphase, anaphase, telophase, and then followed by cytokinesis (reviewed in (Alberts et al, 2002); summarized in Figure 1.1) 1.1.1 Specific phases of the cell cycle The G i phase, or "Gap I" phase, represents a growth period for the newly formed cell from the previous mitosis. A cell in G i does not necessarily have to proceed through to the next phase. In fact, Go and G i can be interchangeably used, as a cell is not committed to a round of the cell cycle until it passes what is called the "restriction point", a period in which the cell is irreversibly committed to undergoing a round of mitosis. Should the cell be actively cycling, G i is used by the cell to grow in size and amass the proteins, food and energy to prepare for the next phase. 1 Figure 1.1 : The cell cycle. Interphase (green) is divided into 3 distinct phases, G l 5 S and G 2 . Mitosis ("M"; blue) is divided into 6 stages, prophase, prometaphase, metaphase, anaphase, telophase and cytokinesis. Entry from G, to S is irreversible upon passing the restriction point (red star), otherwise the cell can exist in G 0 indefinitely until provided the appropriate signal to re-enter the cell cycle. Regulation and timing of cell cycle progression is controlled by CDK/cyclin complexes. Cell cycle checkpoints (red Do-Not-Enter sign) have been identified at the G,/S, intra-S and G 2 / M boundary. 2 S-phase, or the D N A Synthesis phase, is the stage at which D N A replication occurs. A Gi cell normally possesses a In complement of chromosomes. Upon completion of S-phase, the cell now has a An complement of chromosomes in preparation for mitosis. G2, or "Gap 2", is the final period for the cell to continue growth, and production of the necessary proteins, food and energy for mitosis. The first stage of mitosis, termed prophase, represents a period in which the cell prepares for mitosis. Throughout interphase, the chromosomes in the nucleus are not visible. In prophase, the chromosomes condense to their centromere-linked " X " shape and can be visualized by light microscopy. The centrioles prominently appear and migrate to opposite ends of the cell. Spindle microtubules begin to extend from the new centrosomes. The beginning of prometaphase is marked by the complete dissolution of the nuclear evelope, endoplasmic reticulum and Golgi. Proteins and microtubles link the kinetochores to the centrosomes and the chromosomes begin to move throughout the cell. Metaphase marks the alignment of the chromosomes along the midline of the cell (the "metaphase plate"). This is a highly organised process, with feedback checkpoints, to ensure that each daughter cell will receive one copy of each chromosome. In anaphase, the cohesive proteins linking the centromere together are cleaved. One copy of each chromatid is then drawn to opposite ends of the cell, towards the centrosomes. Telophase and cytokinesis occur somewhat simulateously. In telophase, the chromatids arrive at the polar ends of the cell, a nuclear membrane forms and the 3 chromatids decondense while the microtubules disperse. Cytokinesis involves the pinching of the cell into two daughter cells by the action of an actin/myosin ring that forms at the equatorial plane (for an extensive review, see (Alberts et al, 2002)). 1.1.2 Coordinated regulation of the cell cycle 1.1.2.1 Cyclin Dependent Kinases Regulation of the timing of the different phases of the cell cycle is accomplished by the coordinated synthesis and destruction of cyclins with their cyclin-dependent kinase (CDK) partners, and was discovered by the Nobel laureates in Medicine in 2001, Leland Hartwell, Timothy Hunt and Paul Nurse (Hartwell et al, 1970; Culotti and Hartwell, 1971; Hartwell, 1971; Hartwell, 1971; Hereford and Hartwell, 1971; Nurse, 1975; Evans et al, 1983). CDKs consist of a conserved 34 kDa catalytic core, which is inactive in its native state. To-date, there are 10 different CDKs identified in eukaryotic cells. Of those, CDKs 1, 2, 3, 4, and 6 are directly involved in the regulation of the cell cycle (Figure 1.1). CDK7, together with cyclin H, acts as a C D K activating kinase (CAK) for CDKs. Phosphorylation by C A K on the T-loop of a C D K is required to displace the T-loop, which blocks access of ATP to the catalytic domain of the kinase (Fesquet et al., 1993). C A K can only phosphorylate the T-loop of the CDK-cyclin complex, permitting its activation by other factors (for review, see (Nigg, 2001)). C D K expression is constant throughout the cell cycle. The activity of CDKs in the cell, however, is tightly regulated by a variety of factors including C D K inhibitors, proteolysis, subcellular localization, phosphorylation state and cyclins. The latter two, phosphorylation state and cyclins, will be the primary focus in this chapter. 4 1.1.2.2 Cyclins As mentioned above, the activity of CDKs requires binding of a cyclin, first discovered by Timothy Hunt (Evans et al, 1983). As shown in Figure 1.1, there are several different cyclins, each of which interacts with a specific C D K to primarily regulate its activity. Cyclins derive their name from the observation that their expression and destruction is periodic in nature. Briefly, the CDK4/6-cyclin D complex regulates the growth and passage of the restriction point in G i cells, while CDK2-cyclin E regulates the G i / S transition. S-phase is controlled by the CDK2-cyclin A complex, while the G2 /M transition is regulated by CDK1-cyclin B (illustrated in Figure 1.1; reviewed in (Alberts et al, 2002)). 1.1.2.3 T h e G 2 / M transition Of particular relevance to this thesis are the regulatory events that occur at the G2 /M transition. This is controlled by the CDK1-cyclin B complex; however, additional positive regulation is required for mitosis to initiate. CDK1 normally exists in the cell as a hypophosphorylated monomer. When cyclin B production begins in late S / early G 2 , it associates with CDK1 and is then phosphorylated by C A K at Thrl61 in the T-loop, as well as at Thrl4 and Tyr 15 by M y t l (Mueller et al, 1995) and Weel (McGowan and Russell, 1993), respectively (reviewed in (Niida and Nakanishi, 2006)). In contrast to the Thrl61 phosphorylation, phosphorylations of Thrl4 and Tyr 15 keep CDK1 inactive until completion of S and G 2 . Figure 1.2 provides a summary of these regulators of CDK1 activity. 5 Tyr15 CDK1 V T h r U Cyclin B T y r l S CDK1 > T h r 1 4 Cyclin B A 1 ,Tyr15 A / V e e 1 > ) - T h r 1 4 Cyclin H Myt1 t Mitosis CDC25C No Mitosis Figure 1.2 : CDK1 regulation of the G 2 / M transition. Entry into mitosis is dependent on the activity of CDK1, complexed with cyclin B. CDK1 exists in an inactive (grey) hypophosphorylated state until cyclin B production and binding, whereupon it is phosphorylated in its T-loop at Thrl61. It is then immediately phosphorylated at Thrl4 and Tyr 15 by Mytl and Weel (diamonds), maintaining it in an inactive state. CDC25B/C (hexagons) remove these inhibitory phosphorylations at the onset of mitosis, causing full activation (red) of CDK 1. 6 CDK1 can be dephosphorylated by the dual-specificity phosphatases, of the cell-division cycle 25 phosphatase (CDC25) family (Strausfeld et al, 1991). Some studies have suggested that CDC25A may also be required, but this remains inconclusive. Regulation of CDC25B and C is complex and unclear. Briefly, CDC25B activity increases in G2, and possibly mediates the initial activation of CDK1. Once activated, CDK1 may itself activate CDC25C, creating a feedback loop to allow for rapid activation of CDK1. CDC25C can also be negatively regulated by phosphorylation on Ser216, presumably providing a binding site for 14-3-3 proteins to exclude it from the nucleus. Once mitosis is completed, cyclin B is degraded and cyclin D synthesis begins for a new round of replication. 1.2 D N A damage D N A damage can occur naturally in a cell, commonly from either endogenous sources such as free-radical generation from metabolic by-products in oxygen-utilizing organisms, or from exogenous sources such as chemical exposure or ionizing radiation. The rate of D N A damage in human cells, for example, can range from 1,000 to 1,000,000 lesions, per cell per day (Lodish et al, 2004). Although this may seem staggering, this actually only corresponds to 0.000165 % of the human genome. Regardless, the cell has to efficiently and effectively repair this D N A damage, or risk mutagenesis and possible death or oncogenic transformation. D N A damage is the basis for the anticancer activity of ionizing radiation therapy 1 and several chemotherapeutic agents, such as cisplatin and doxorubicin. D N A damage can elicit a variety of cellular responses, notably D N A repair and D N A damage 7 checkpoints (see Section 1.3), as well as downstream events such as apoptosis, mitotic catastrophe and premature senescence (Zhou and Elledge, 2000). The D N A damage illicited by ionizing radiation can exist in many forms : chemical modification of a base or nucleotide, single-strand breaks (SSB), or the most lethal form, double-strand breaks (DSB). If a cell is undergoing D N A replication at the time of generation of a SSB, the lesion may be converted into a DSB (outlined in Figure 1.3). The lethality of a DSB comes from two main factors. First, should a cell undergo mitosis with a DSB, the distribution of one copy of each chromatid to the daughter cells is no longer controlled and a fragment of D N A may be randomly distributed to the wrong daughter cell, or lost completely. Unequal D N A distribution may result in too many copies of an being active gene present, or conversely the lack of essential gene(s). Secondly, the act of D N A repair, or more specifically, non-homologous end joining (described below) repair, may in itself cause mutation(s) in any gene in which the DSB exists. 1.2.1 R e p a i r o f d a m a g e d D N A Many different D N A repair mechanisms exist in eukaryotic cells. The major mechanisms to repair single-strands of D N A include : base-excision repair (BER), which repairs damage on single nucleotides; nucleotide-excision repair (NER), similar to BER except in cases involving damage to 2+ nucleotides; and mismatch repair (MMR), for mismatched base pairs following D N A replication. Repair of D N A double-strand breaks 8 5' 3' 3' 5' DNA 5' 3' 3' * 3' 5' SSB S-phase: DNA Replication DSB Figure 1.3 : Simplified model of conversion of a SSB to a DSB during DNA replication. A SSB in DNA at the time of S-phase, when the polymerase encounters the lesion it will become stalled. The opposite strand may continue DNA replication, resulting in a DSB requiring repair. 9 is accomplished through two major pathways, non-homologous end joining (NHEJ) and homologous recombination-repair (HRR) (reviewed in (Sancar et al, 2004)). 1.2.1.1 N H E J NHEJ is a Gi/S-specific process that is checkpoint-independent and requires DNA-PK, a complex of the PI3-like-kinase (PIKK) family member DNA-PKcs and Ku70/86, as well as D N A Ligase IV and XRCC4 (Lieber et al, 2003). NHEJ machinery is also involved in V(D)J rearrangements during immune system maturation. It is also sometimes referred to as "error-prone DSB repair", as it involves directly joining the two ends of the broken D N A strands without a sister-chromatid template, and often will process and remove damaged nucleotides on the ends of the broken D N A in preparation for ligation. This processing of the broken ends often leads to loss of sequence information. Thus, it is mutagenic in nature. This is unavoidable, as NHEJ occurs primarily in G i and early S-phase, so there is no sister-chromatid in close proximity for recombination repair (described in Section 1.2.1.2). A general model of the repair of a DSB via NHEJ is outlined in Figure 1.4. Upon sensing of a D N A DSB, the Ku heterodimer, consisting of Ku70/Ku86, binds to the broken ends of the D N A and translocates one to two helical turns in the DNA, exposing the broken ends of DNA. This creates a binding site for the DNA-dependent protein kinase catalytic subunit (DNA-PKcs), which binds to both D N A ends with the Ku heterodimer. The complex of Ku70/86 and DNA-PKcs comprises the enzyme commonly referred to as the DNA-dependent protein kinase, or DNA-PK. This displaces the K u heterodimer further down the D N A strands, with the entire complex 10 DSB I Ku binding DNA-PKcs binding •<~Y I M) End processing Synapsis Ligation Figure 1.4 : Simplified model of repair of a DNA DSB by N H E J . The Ku heterodimer (blue circles) binds both ends of the broken DNA and translocates 1-2 helical turn(s), allowing binding of DNA-PKcs (purple oval). The Mrel 1/Rad50/Nbsl complex (red rectangle, green triangle and yellow circle), and other nucleases process the broken ends for ligation, performed by XRCC4/DNA Ligase IV (black and white triangle) as well as other factors. 11 approximately 30 bp from the broken D N A ends. D N A - P K assembly on both broken strands of DNA, followed by "synapsis", or the alignment of both ends together, leads to activation of DNA-PK. Once activated, the broken ends of the D N A can be processed by the nuclease complex of Mrel 1/Rad50/Nbsl (MRN complex), as well as numerous other enzymes,: such as the exonucleases FEN-1, and Artemis, PNK and others, or the unwinding enzymes W R N and B L M , for example. It is this processing step that can lead to the mutagenic properties of NHEJ. After processing, the complex of XRCC4/DNA Ligase IV, as well as others such as D N A polymerase u or X, ligates the broken ends of D N A held in place by the two D N A - P K complexes (for a review of NHEJ, see (Hefferin and Tomkinson, 2005)). D N A - P K is a nuclear serine/threonine kinase with a molecular mass of approximately 470 kDa. The DNA-PKcs kinase domain shares amino acid similarity with that of phosphotidyhnositol-3-kinases (PI3-K) family. However, DNA-PKcs does not possess lipid kinase activity. Therefore, it is referred to as a PI3-K-like kinase, or PIKK. Other members of this family are also very large kinases, such as mTOR, or as discussed later in section 1.3, A T M and ATR. The repair of DSBs via NHEJ in vertebrates is entirely dependent on the kinase activity of DNA-PK. Yeast species, on the other hand, do not possess a DNA-PKcs gene, or homolog (reviewed in (Smith and Jackson, 1999)). Although the specific targets of D N A - P K are still unclear, it is proposed that it regulates NHEJ by i) allowing access of X R C C 4 / D N A Ligase IV to the broken ends of DNA, ii) allowing access to processing enzymes of the broken ends, or iii) phosphorylating Ku (reviewed in (Pastwa and Blasniak, 2003)). Autophosphorylation of DNA-PKcs leads to its inactivation and possible dissociation from DNA, allowing 12 completion of ligation (Chan and Lees-Miller, 1996; Chan et al, 2002; Ding et al, 2003). Regardless of its target, this dependence on kinase activity of D N A - P K can be exploited by the chemical inhibition of DNA-PKcs with small molecule inhibitors, causing increased sensitivity to D N A damage (see below). Inhibitors of D N A - P K include the noncompetitive, general PI3-K inhibitor wortmannin , as well as LY294002 (Rosenzweig et al, 1997). Interestingly, in vivo, inhibition of PI3-K by wortmannin requires concentrations in' the nanomolar range (approximately 200 nM), whereas inhibition of D N A - P K and A T M requires 20 uM, and ATR requires 200 uM. Newer generation inhibitors of D N A - P K are much more specific, such as A M A3 7 (Knight et al, 2004). Chemical inhibition of D N A - P K can radiosensitize several forms of cancer, offering a possible therapeutic approach in combination with D N A damaging therapies such as radiation or chemotherapy (Rosenzweig et al, 1997; Sarkaria et al, 1998; Kim et al, 2002; Allen et al, 2003; Ismail et al, 2003; Kashishian et al, 2003; Willmore et al, 2004; Shinohara et al, 2005). 1.2.1.2 HRR As HRR requires the presence of an identical sequence to be used as a template for DSB repair, it is limited to late S/G2. HRR allows a damaged chromosome to be repaired using the newly created sister chromatid as a template. Because of this, it is considered "error-free DSB repair", in contrast to NHEJ. The enzymatic machinery involved in HRR is nearly identical to the machinery responsible for crossover events during meiosis. HRR is a checkpoint-dependent process that requires numerous Rad proteins, such as Rad52, Rad54, and various Rad51 paralogues (Takata et al, 1998; 13 Thompson and Schild, 2001). A general model of the repair of a DSB via HRR is outlined in Figure 1.5. In contrast to NHEJ, few details of HRR have been elucidated. The first event believed to occur is after sensing the damage is a 5' —> 3' resection of the D N A at the site of the DSB, likely by the M R N complex, leaving single-stranded overhangs. Sensing the DSB is likely accomplished by A T M and Brcal. RPA, Rad51 and paralogs, and Brca2 bind, followed by Rad52 and Rad54 to the overhanging single-stranded DNA. Rad51, Rad52 and Rad54 promote strand exchange between the sister chromatid. The enzymes involved in elongation, gap filling and ligation are currently unknown. Many other proteins are thought to play a role in this process, such as WRN, B L M , Brcal , or A T M . Following strand invasion/elongation, HRR can be resolved in either of two different ways : non-crossing over, from disengagement of the Holliday junctions formed in the process, with subsequent gap filling and ligation is the more common result; the other possibility is a cross-over event, resulting from Holliday junction resolution following endonucleolytic cleavage, with subsequent ligation (for a review of HRR, see (Valerie and Povirk, 2003); outlined in Figure 1.5). 1.3 Cell cycle checkpoints The cell cycle has many levels of organization and control, and D N A damage can cause a temporary halt in cell cycle progression to allow time for either D N A repair or to assess whether the cell should commit apoptosis (programmed suicide). This arrest is termed a "cell cycle checkpoint". Checkpoints are "biochemical pathways that delay or arrest cell cycle progression in response to D N A damage" (Nyberg et al, 2002). 14 ATM? ATR? Brcal? 9-1-1? DSB recognition MRN? Rad17-RFC? Others? Resection RPA, Brca2, Rad51 I 1 Rad52 and Rad54 Sister chromatid pairing OR No crossover Crossover Holliday junction resolution Figure 1.5 : Simplified model of repair of a DNA DSB b y HRR. After DSB detection, DNA ends are resected to allow binding of RPA (red), Rad51 (blue) and Brca2 (yellow). This allows binding of Rad52/54, promoting sister strand invasion and then elongation. HRR is completed by either non-crossing over, with Holliday junctions disengaging and gap filling, or with crossing over with Holliday junction resolution and gap filling. Figure adapted from (Valerie and Porvik, 2003). 15 Although checkpoints were originally thought to be the cellular counterparts to international border checkpoints to halt cell cycle progression (Hartwell and Weinert, 1989; Weinert and Hartwell, 1989), they are more accurately considered constant surveillance systems (Nasmyth, 1996) not only involved in cell cycle arrest, but in D N A repair, senescence and apoptosis (Sancar et al, 2004). There are two major D N A damage checkpoints, at the G i / S and G2 /M transitions. A third checkpoint, the intra-S checkpoint, is not characteristically a full cell cycle halt, but rather a general slow-down of D N A replication during D N A repair in comparison to an undamaged cell. In a healthy cell, there is tight regulation of the entry and exit from each phase maintained by numerous feedback mechanisms. It appears that the pathways controlling normal cell cycle progression are also the same in checkpoint responses. Loss of checkpoint pathway components can be oncogenic in nature (Hartwell et al, 1994). In fact, in cancer cells there are usually genetic alterations that cause major disruptions in cell cycle progression and checkpoints and these differences between normal and cancer cells make checkpoints interesting targets for cancer therapy. Members of the three main checkpoints of concern in this thesis, the G i / S , the intra-S, and the G2 /M, can be classified as damage sensors, mediators, transducers and effectors (outlined in Figure 1.6, reviewed in (Sancar et al, 2004)). As this thesis pertains to the potential of the G2 checkpoint pathway as a target for cancer therapy, I will highlight what is known of these components in vertebrates, specifically in humans. Regardless, the components of the checkpoint pathways are highly conserved across eukaryotic species. 16 DNA Repair ? G 1 • S • G 2 -M Figure 1.6 : General model of checkpoint activation in humans. Components of the DNA damage checkpoint can be divided into Sensors, which sense the DSB, Mediators, which assist in the signaling to Transducers, which in turn activate the Effectors of the checkpoint, ultimately leading to cell cycle arrest. Figure adapted from (Sancar et al., 2004). 17 1.3.1 Damage sensors At the time of this thesis preparation, the mechanism(s) behind D N A damage detection, for the purpose of D N A repair and/or checkpoint activation, remain unclear. What is known is that rapidly after cells are exposed to a D N A damaging agent such as ionizing radiation, many signalling molecules form foci on D N A at or near the sites of D N A damage. Currently there are two groups of proteins considered to be D N A damage sensors: the multiprotein complex Radl7-RFC/9-l- l complex, and the P IKK members A T M and ATR. The multiprotein complex Radl7-RFC is homologous to replication factor C, RFC (Griffiths et al, 1995; Parker et al, 1998; Bao et al, 1999; L i et al, 1999). It is composed of Radl7, p36, p37, p38 and p40. The 9-1-1 complex is composed of Rad9, Radl and Husl_, and is structurally homologous to the proliferating cell nuclear antigen, PCNA (Thelen et al, 1999). In a similar manner to RFC with PCNA, Radl7-RFC loads the ring-like 9-1-1 complex onto damaged DNA, providing an interacting site for checkpoint signalling proteins as well as repair factors (Onge et al, 1999; Volkmer and Karnitz, 1999; Lindsey-Boltz et al, 2001; Zou et al, 2002). The protein mutated in the disease ataxia-telangiectasia (A-T), A T M (ataxia telangiectasia mutated) is a major player in D N A damage sensing, the checkpoint and some forms of D N A repair. A-T is characterized by ataxia (loss of balance), infertility, immunodeficiency, radiosensitivity, genome instability and cancer predisposition (reviewed in (Lavin and Khanna, 1999)). A T M is approximately 350 kDa, and belongs to the same family as DNA-PKcs, the PIKK family which although possessing homology to PI3-K, does not exhibit lipid kinase activity. This protein was initially identified by 18 the fact that, when mutated, cells exhibit defective checkpoint arrest at all three checkpoints. A T M exists as a homodimer in undamaged cells. When a DSB occurs, this leads to changes in chromatin condensation which trigger, through unknown mechanism(s), A T M dimer dissociation with fr-ans-autophosphorylation on Ser 1981 (Bakkenist and Kastan, 2003). Activated A T M then phosphorylates downstream tranducers of the checkpoint, as well as the Radl7-RPC and 9-1-1 complex (Bao et al, 2001) (Chen et al, 2001) (Wang et al, 2001c). A related PIKK, A T M and Rad3 related (ATR) is also considered a D N A damage sensor, although it appears to respond to UV-induced damage and replicative stress, rather than DSBs specifically (as is the case for A T M ; reviewed in (Abraham, 2001)). Although both A T M and A T R possess yeast homologs which require binding partners to bind broken DNA, only A T R has a human binding partner identified - A T R interacting partner (ATRIP) (Cortez et al, 2001). Activated ATR phosphorylates downstream transducers of the checkpoint, as well as Radl7 (Bao et al, 2001). 1.3.2 Checkpoint mediators Mediators of the checkpoint interact with not only the D N A damage sensors such as A T M , but the repair proteins, transducers and effectors. Mediators include, but are not limited to, Claspin, Brcal, p53 binding protein 1 (53BP1), phosphorylated histone 2A.X (y-H 2 AX), M R N , 9-1-1, and the mediator of the D N A damage checkpoint I (MDC1). The specific role of mediators in the D N A damage response varies widely from protein to protein, but generally they tend to provide binding sites for repair components, or 19 mediate the interaction between a damage sensor and a checkpoint transducer (reviewed in (Sancar et al, 2004)). 1.3.3 Checkpoint transducers and effectors The checkpoint transducers are the checkpoint kinases 1 and 2 (Chkl and Chk2). These are Ser/Thr kinases, localized primarily in the nucleus. Chk2 is primarily activated by phosphorylation on Thr68 by A T M (Matsuoka et al, 1998; Chaturvedi et al, 1999; Ann et al, 2000; Matsuoka et al, 2000), while Chkl is activated by phosphorylation on Ser345 by ATR (Liu et al, 2000), although some instances of cross-talk have been observed (Helt et al, 2005). The roles of both kinases in the checkpoint still remain unclear. It seems that Chk2, through A T M , is involved in the initiation of the checkpoint, while Chkl is involved in its longer-term maintenance (Xu et al, 2002). However, gene disruption has shown that Chk2_/" mouse embyonic stem cells could initiate but not maintain arrest (Hirao et al, 2000), while somatic cells of Chk2 - /" mice had no G 2 checkpoint defects (Takai et al, 2002). Chkl deletion is embyronic lethal (Liu et al, 2000), while Chkl inhibition prevents G 2 arrest (Zhao et al, 2002). The downstream targets of Chkl and Chk2 are shared with Weel, CDC25 and p53. As outlined in Figure 1.2, Weel is a negative regulator of CDK1 activity, whereas CDC25 family members are positive regulators of CDK1 (described in Section 1.1.2.3). p53 is a transcription factor with roles in both the Gi/S and G 2 / M checkpoint. It is stabilized by phosphorylation on Ser 15 and 20 by Chkl/2 and A T M / A T R / D N A - P K (see below). 20 1.3.4 The Gi/S checkpoint The G i / S checkpoint prevents the replication of damaged D N A and is entirely dependent on the transcription factor p53. p53 is normally kept at very low levels through MDM2-mediated p53 ubiquitination and degradation. When a DSB is detected, p53 is phosphorylated by the Chk2 kinase on Ser20. This then blocks the p53-MDM2 interaction, stabilizing p53 by preventing it from being sent for degradation. p53 is further activated by phosphorylation on Serl5 by A T M , which enhances its transcriptional transactivation activity. A T M can also phosphorylate M D M 2 at Ser395 to inactivate it. p53 upregulates several genes for the checkpoint, such as GADD45a, p 2 1 C i p l / W a f l , and M D M 2 which functions in a feedback loop to deactivate p53. p 2 1 C i P i / w a f i ^ a l s Q AG CDKNI A) is a cyclin-dependent kinase inhibitor (CDKI), which binds the catalytic domain of the cyclin E-CDK2 complex, preventing it from promoting the G i - S transition. p 2 1 C i p l / W a f l also inhibits cyclin D-CDK4, blocking phosphorylation of Rb. Under normal conditions, Rb phosphorylation by CDK4 releases the E2F transcription factor necessary for S-phase gene transcription (for a review of the G i / S checkpoint, see (Nojima, 2004)). p53 is involved in the maintenance of the G\/S checkpoint. The initiation of this checkpoint is through the ATM-Chk2-CDC25A. CDC25A is required to remove inhibitory phosphorylations on CDK2 in a similar manner as CDC25C to C D K I (see Section 1.1.2.3). Activated Chk2 phosphorylates and inactivates CDC25A on Serl23 by causing nuclear export and degradation (Falck et al, 2001). Approximately 2 h post-IR, the p53-induced genes reach the necessary levels to maintain G i arrest until D N A repair is complete. 21 1.3.5 The intra-S checkpoint The intra-S checkpoint, also referred to as the S-phase checkpoint, does not behave in the same manner as the other two checkpoints in that it is not a complete halt of cell cycle progression, but rather a decrease in the rate of D N A synthesis in comparison to a healthy cell. Of the three checkpoints described in Section 1.3, the intra-S checkpoint is the least clear to-date. Evidence for this checkpoint was first provided in cells with mutated A T M , as they exhibited what is called radioresistant D N A synthesis, or RDS. The mechanism behind this checkpoint can be divided into two parts : firstly, CDC25A is inactivated in the same manner as that for the Gi/S checkpoint, inhibiting CDK2; the second, which is the least clear, involves B A S C (Brcal-associated genome surveillance complex) proteins which includes A T M , Brcal, and Nijmegen breakage syndrome 1 (Nbsl), and structural maintenance of chromosomes 1 (SMC1) (reviewed in (Sancar etal, 2004)). Nbsl , also a component of the M R N complex involved in D N A damage sensing and repair, can be phosphorylated by both A T M and ATR (Lim et al, 2000). SMC1 is also phosphorylated by A T M / A T R at Ser957 and Ser966, as this phosphorylation is dependent on the presence of Nbsl (Yazdi et al, 2002). Once activated, this pathway likely plays a role in double-strand break repair and recovery from checkpoint signalling. Further research into this pathway is necessary to define their role(s) in this checkpoint. However, it is known that mutated/inhibited Brcal, Nbsl , or SMC1 lead to defective S-phase slow down in the presence of D N A DSBs (Lim et al, 2000; X u et al, 2001; Yazdi etal, 2002). 22 1.3.6 The G 2 / M checkpoint As discussed in Section 1.3.4, the G i checkpoint is entirely dependent on the function of one protein, p53, while the G 2 checkpoint has both p53-dependent and -independent components. Since it is estimated that the majority (greater than 50%) of human cancers have a mutated p53 (Hollstein et al, 1991), this leaves only the p53-independent G 2 checkpoint to provide increased time for D N A repair. This checkpoint is particularly important, as it is the last chance for a cell to ensure that its D N A is intact before passing it on to its daughter cells. The G 2 checkpoint negatively regulates entry into mitosis by maintaining inhibitory phosphorylation of Thr-14 and Tyr-15 in C D K I , the key kinase involved in the regulation of mitosis (Smits and Medema, 2001). A general model of G 2 / M checkpoint activation is shown in Figure 1.6. D N A damage activates PIKK family members A T M and A T R that phosphorylate a variety of downstream targets, including the transducer kinases Chkl and Chk2, which phosphorylate and inactivate the phosphatase CDC25A/B/C, thereby preventing the dephosphorylation and activation of C D K I (Flatt and Pientenpol, 2000). Chkl can also phosphorylate and activate Weel, further ensuring the presence of inhibitory phosphorylations on C D K I (O'Connell et al, 1997). However, a growing body of evidence indicates that there are other kinases, such as M A P K s , which may also play a significant role in the maintenance of the checkpoint (Wang et al, 2000; Meng et al, 2004; Mingo-Sion et al, 2004). However, their specific role is still unclear, and is therefore not shown in Figure 1.6. 23 1.4 Checkpoint inhibitors Many forms of cancer are treated using chemotherapy or radiotherapy with the purpose of damaging DNA. The damaged D N A triggers the checkpoint pathways described in Section 1.3. As mentioned above, more than 50% of cancerous cells are estimated to have a mutated or non-functional p53 pathway (Hollstein et ai, 1991). Therefore, the p53-independent G 2 / M checkpoint, outlined in Figure 1.6, is the last chance for cancer cells to repair damaged D N A before mitosis. In the laboratory using cultured cancerous cells, this checkpoint can be turned off using "checkpoint inhibitors", which are small molecules capable of decreasing the repair time available to a cancerous cell. This in turn leads to an increase in cancerous cell death when compared to normal treatment alone, as the cancer cells continue dividing with badly damaged DNA. The use of checkpoint inhibitors is an exciting therapeutic option that could increase the effectiveness of current cancer therapy. Unfortunately, not enough is known about the cell cycle checkpoints to allow for appropriate drug design for use in patients; currently identified checkpoint inhibitors are not specific to only the cell cycle checkpoint, which results in many undesirable side-effects. The use of a chemical to increase the sensitivity of a cancerous cell to the effects of radiation is defined as "radiosensitization", although in this thesis I will refer to it broadly as an increase in cell killing via checkpoint abrogation following either chemically- or radiation-induced D N A damage. The principle behind this effect is simple : cells with damaged D N A are forced into mitosis with broken DNA, which cannot be distributed correctly into 2 daughter cells. Hence, the newly-formed post-24 mitotic cells are either missing, or possessing too many, chromosomal fragments, which can be deleterious to cell survival (reviewed in (Anderson et al, 2003)). Because of this radiosensitizing effect, the use of checkpoint inhibitors, in combination with D N A damaging therapy, could be a valuable therapeutic approach in the treatment of cancer. To-date, there are many different checkpoint inhibitors known (see Table 1.1). Although these inhibitors target a wide range of checkpoint components, the most prevalent is C h k l , and for good reason. Chkl is the main effector kinase of the G 2 checkpoint (Liu et al, 2000; Chen et al, 2003; Latif et al, 2004). Without Chk l , G 2 arrest does not occur, and cells are hyper-sensitive to radiation. However, normal cells do not exhibit any noticeable defects in growth rate when Chkl is inhibited/deleted (reviewed in (Anderson et al, 2003)), making it an ideal target for anti-cancer strategies. Unfortunately, few inroads have been made with respect to utilizing checkpoint inhibitors as therapeutic agents. The most potent of checkpoint inhibitors, caffeine and pentoxifylline, are not appropriate for clinical use due to their numerous side-effects resulting from the high-concentrations required for checkpoint inhibition (Arnaud, 1987; Jiang et al, 2000). To-date, the only published clinical trial utilizing a checkpoint inhibitor (UCN-01) in combination with D N A damage (cisplatin), failed to demonstrate potentiation by checkpoint inhibition (Lara et al, 2005). The study was halted due to toxicity of UCN-01. The clinical development of UCN-01 is further hampered by the fact that it binds non-specifically to human alphal-acid glycoprotein (AGP) (Fuse et al, 1999). Several of the other checkpoint inhibitors listed in Table 1.1 are in clinical or pre-clinical development, however the status of these drugs has not been 25 Table 1.1: Currently known checkpoint inhibitors and their targets. Inhibitor(s) G2-Specific Target(s) Reference Methylxanthines : Caffeine and various analogs A T M / ATR (Walters etal, 1974; Samlaska and Winfield., 1994; Sarkariaetal, 1999) Aminopurines : 2- Unknown, possibly A T M / (Andreassen and Margolis, aminopurine, 6-dimethylaminopurine A T R 1992) Indolocarbazoles : Chkl (Tarn and Schlegel, 1992; Staurosporine, UCN-01, SB218078, Go6976, ICP-1 Wang et al, 1996; Jackson et al, 2000; Eastman et al, 2002; Kohn et al, 2003) Granulatimides : Chkl (Roberge etal, 1998) Granulatimide, Isogranulatimide Hymenialdisines: Hymenialdisine, Debromohymenialdisine Chkl / Chk2 (Curman et al, 2001) Synthetic peptide : TAT-S216A (CBP-501) Chkl / Chk2 (Suganuma et al, 1999) Undisclosed structure : Chkl / Chk2 (Sorensen et al, 2003) CEP-3891 Pyridopyrimidine : PD0166285 Weel (Wang et al, 200Id) Okadaic acid PP2A/PP1 (Yamashitaera/., 1990) Fostriecin PP2A/PP1 (Roberge etal, 1994) Amiloride N a + / H + exchanger (Sailer etal, 1998) Sesquiterpene lactone : Unknown (Weel?) (Sturgeon et al, 2005) Psilostachyin Leptomycin B Unknown (Crml?) (Sturgeon et al, 2006a) Cryptofolione Unknown (Crml?) (Sturgeon et al, 2006a) 13-hydroxy-15-ozoapatlin Unknown (Weel?) (Rundle et al, 2001) 26 made public at the time of preparation of this thesis (Garber, 2005). Should these compounds prove beneficial, it will be of obvious interest to develop methods for in vivo checkpoint inhibitor application and delivery if they are ever to be utilized in a clinical setting. 1.5 This thesis Checkpoints are basic cell cycle signal pathways that have yet to be fully characterized, making them an exciting field for research, both from a clinical perspective and an understanding of basic human cellular processes. The sensitivity of tumour cells to radiotherapy and DNA-damaging chemotherapeutic agents depends in large measure on the level of integrity of their D N A repair and checkpoint mechanisms. Checkpoint abrogation by checkpoint inhibitors increases the efficacy of D N A damaging treatments in killing cancer cells with mutated p53 over cells with wild-type p53. Because of this, my thesis research is primarily concerned with providing a better understanding of the mechanisms behind the G 2 checkpoint, with the hope that results may eventually lead to better therapeutic agents to enhance the efficacy of D N A damaging therapies. Specifically, my work focusses on exit from the G 2 checkpoint, which is much less well understood than its initiation. This work has been divided into three key areas of research, all with the goal of furthering knowledge regarding checkpoint inhibitors as potential therapeutic agents: firstly, to identify new checkpoint inhibitors, with the goal of either identifying new compounds to be used for anti-cancer therapy, or using them as tools to further elucidate the mechanism(s) behind the G 2 / M checkpoint by identifying their target(s); secondly, to 27 attempt to enhance the cell-killing effect of checkpoint inhibitors by combining them with a D N A repair inhibitor; and finally, to develop a novel checkpoint inhibitor identified in the Roberge laboratory, isogranulatimide, as an in vivo checkpoint inhibitor. Published from this work is data in Chapter 3 and 4. Co-authors Dr. Kyle Craig, Geoffrey Karjala, Colleen Brown, Dr. Natalie Rundle, Dr. Raymond Andersen and Dr. Michel Roberge contributed to my first paper (Sturgeon et ai, 2005), regarding psilostachyin as a checkpoint inhibitor. Secondly, from Chapter 4, co-authors Dr. Zachary Knight, Dr. Kevan Shokat, and Dr. Michel Roberge contributed to (Sturgeon et al, 2006c), focussing on the effect when D N A repair inhibitors are combined with checkpoint inhibitors. At the time of this thesis preparation, I have 2 additional papers in review. In Chapter 3, co-authors Dr. Ana Diaz-Marrero, Dr. Bruno Cinel, Lianne McHardy, Michelle Ngo, and Dr. Raymond Andersen and Dr. Michel Roberge contributed to (Sturgeon et al, 2006a), which demonstrates that cryptofolione is an extremely efficacious checkpoint inhibitor which also inhibits nuclear export. From Chapter 5,1 have another submission with Dr. Michel Roberge as a co-author, (Sturgeon and Roberge, 2006d), which explores the effect of radiosensitization achieved by checkpoint inhibitors when applied at different times in combination with D N A damage. 28 Chapter 2 Materials and Methods 2.1 C e l l l i n e s MCF-7 mp53 cells were cultured as previously described (Fan et al, 1995; Anderson et al, 1997), using D M E M supplemented with 50 ng/mL bovine insulin (Gibco BRL), 1 ug/mL hydrocortisone (Sigma-Aldrich), 1 ng/mL hEGF (Gibco) and 1 ng/mL P-estradiol (Sigma-Aldrich) (a gift from Dr. Stephen Friend). HeLa cells were cultured in D M E M (also a gift from Dr. Stephen Friend). HCT116 cells were cultured in McCoy's 5A Media (a gift from Dr. Bert Vogelstein). M059K, M059J and M059J/Fusl cells were cultured in DMEM/F12, with M059J/Fusl cells additionally supplemented with 250 ug/ml G418 (Invitrogen) (M059K and M059J were a gift from Susan Lees-Miller, M059J/Fusl cells were a gift from Cordula Kirchgessner). A l l CHO cells were cultured in D M E M a with 10% FBS. CHO KI and XRCC4 cells were obtained from the Coriell Cell Repository. CHO V3 cells transfected with either empty vector or hDNA-PKcs cDNA (Ding et al, 2003) and were a gift from Susan Lees-Miller. CHO 51D1 and 51D1.3 cells were a gift from Larry H. Thompson. A l l cell culture media were supplemented with 10% FBS (Gibco BRL). 29 2.2 Screening for novel checkpoint inhibitors Screening was performed as described in (Roberge et al, 1998), performed by various previous and current members of the Roberge laboratory. 2.3 Isolation of sesquiterpene lactones from Ambrosia artemisiifolia and cryptofoliones from Cryptocaria concinna Extracts N35791 and N29449 from the US National Cancer Institute Natural Products Repository showed G 2 checkpoint inhibitory activity using a cell-based assay (Roberge et al, 1998). 2.3.1 Sample N35791 Purification of the psilostachyins was performed by Kyle Craig in the laboratory of Ray Andersen, as described in (Sturgeon et al, 2005). For more detail, see (Craig, 2003). 2.3.2 Sample N29449 Purification of the cryptofoliones was performed by Ana Diaz-Marrero and Bruno Cinel in the laboratory of Raymond J. Andersen, and is described in (Sturgeon et al, 2006a). For more detail, see (Cinel, 2001). 2.4 Irradiation, and Checkpoint Inhibitor Treatment Cells were irradiated at an average rate of 0.75 Gy/min using a 6 0 C o source (Gammacell 220, Atomic Energy Commission of Canada). 2 m M caffeine (Sigma-Aldrich), 100 nM 30 UCN-01 (National Cancer Institute), 10 uM isogranulatimide (Roberge et al, 1998) and 40 uM debromohymenialdesine (Curman et al, 2001) were used to abrogate G2 arrest. AMA37 (Calbiochem) was added to cells 30 min prior to treatment, and refreshed every 8h. 2.5 Checkpoint inhibitor activity Checkpoint inhibition was assessed as described in (Rundle et al, 2001), except that cells were irradiated and incubated with checkpoint inhibitors for specified times. Cells were seeded at 2 x 105 cells/dish in 35 mm-diameter dishes, or lx 106 cells/dish in 100 mm-diameter dishes, and subsequently cultured for 24 h. Cells were then irradiated with 10 Gy using a 6 0 C o source. Unless specified otherwise, 16 h later, when 90% of cells were arrested in G 2 (Roberge et al, 1998), drugs were added with 100-300 ng/ml nocodazole (Sigma-Aldrich), and cells were cultured for another 8 h. CHO cells were irradiated with 6.5 Gy and incubated for 8 h, followed by checkpoint inhibitor plus nocodazole addition for an additional 4 h. Standard deviation (SD) was determined from the average of 3 or more data points. 2.5.1 Mitotic spreads Cells were collected via trypsinization, swelled in 75 mM KC1 for 5 min, fixed in Carnoy's fixative for 10 min, spotted onto glass slides, allowed to dry, stained with 200 ug/ml Hoechst 33258 for 5 min and mounted on microscope slides. The percentage of 31 mitotic cells, identified by their condensed chromosomes, was determined by counting 3 random fields of view or a minimum of 300 cells total, whichever first! 2.5.2 Flow cytometry Cells were prepared for flow cytometry analysis exactly as described in (Rundle et al, 2001). Briefly, following checkpoint inhibitor application, cells were collected at the appropriate time post-IR by trypsinization. Cells were washed twice in sodium azide buffer (SAB) buffer (PBS with 1% FBS and 0.05% NaN 3), resuspended in 1 mL SAB and then vortex ed slowly while adding 10 mL 70% ethanol. Cells were left for at least 1 h to a maximum of 7 days before proceeding to the next step. Cells were washed twice in Tw-SAB (SAB with 0.5% Tween-20), rehydrated in 5 mL Tw-SAB for 30 min on ice, then incubated with 1:100 dilution of GF-7 antibody, or 1:100 cyclin BI antibody (Santa Cruz, sc-7393) in Tw-SAB for 1 h on ice (total volume of 200 uL). Cells were washed twice in Tw-SAB, then incubated in 1:500 dilution of goat anti-mouse Alexa-488 conjugate secondary antibody (Molecular Probes A-11029) for 30 min on ice (total volume 200 uL). Cells were washed twice in Tw-SAB, then resuspended at 2 x 106 cells/mL in 1.12% sodium citrate buffer with 100 ug/mL RNAse A for 30 min at 37°C. Cells were supplemented with propidium iodide stain solution (2.5 mg/50 mL) to achieve 1 x 106 cells/mL for 20 min at 37°C. Cells were counted using a FACS LSRi i instrument with FACSDiva or WinMDI software. 32 2.6 Tubulin polymerization assay 90 ul of ice-cold G-PEM (1 m M GTP, 80 mM PIPES pH 6.8, 1 m M EGTA, 1 mM MgCb) was added to 96-well plates. To this was added 10 pi of G-PEM solutions containing either DMSO, paclitaxel or psilostachyin C to give a final concentration of 10 uM paclitaxel or 100 uM, 10 uM, or 100 nM psilostachyin C. 5 ul of 10 mg/ml bovine brain tubulin (Cytoskeleton) in G-PEM was then added to the sample wells. The plate was warmed to 32°C and the optical density at 340 n m was monitored over time for 15 min. This work was performed by Geoffrey Karjala in the lab of Dr. Michel Roberge. 2.7 Cell Viability Assay MCF-7 mp53 cells were seeded at 1000 cells/well in 96-well plates, grown overnight, and treated or not treated with compound for 24 h, immediately followed by irradiation or not. Dimethyl sulphoxide (DMSO) carrier did not exceed 1% final concentration. The drug was removed, and cells were allowed to grow in fresh medium until those not treated with the drug approached confluence, which was typically 4-6 days. Cell proliferation was measured as follows: 25 ul of a 5 mg/ml solution of 3(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide in phosphate-buffered saline was added to cells in the presence of 100 ul of cell culture medium. After a 2 h incubation at 37 °C, 100 ul of 20% sodium dodecyl sulfate (SDS) dissolved in dimethylformamide/water(l:l), pH 4.7, was added, and the absorbance at 570 nm was measured after overnight incubation. SD was determined from the average of 3 or more data points. 33 2.8 Nuclear import/export assay The effect of chemicals on the nuclear import/export of Rev/GR/GFP was determined as described in (Love et al, 1998). HeLa cells stably transfected with an HIV-Rev/glucocorticoid receptor (GR)/GFP construct were a kind gift from Dr. Pamela Silver (Harvard). Cells were fixed with 3.7% formaldehyde in PBS for 5 min, then visualized by fluorescence microscopy after staining of D N A with Hoechst 33342. 2.8.1 Import Cells were treated with chemicals for 30 min, then 1 uM dexamethasone for 30 min. Cells were fixed and visualized by fluorescence microscopy as above. Import inhibition was assessed by the presence or absence of nuclear/nucleoli localization of the GFP marker by visual approximation. 2.8.2 Export Cells were treated with 1 uM dexamethasone for 30 min. Chemicals were then added and the cells were cultured for an additional 30 min. The cells were then washed twice in phosphate-buffered saline (PBS), then cultured for 60 min in the presence of the export inhibitor, without dexamethasone. Cells were fixed and visualized by fluorescence microscopy as above. Export inhibition was assessed as the presence or absence of nuclear/nucleoli retention of GFP marker by visual approximation. 34 2.9 Immunofluorescence microscopy Cells were seeded at 2 x 105 cells per dish on 7X-detergent treated coverslips in 35 mm dishes, then cultured for 24 h. DMSO (1%) or psilostachyin C (50 uM) was added and cells were cultured for another 4 h. Coverslips were washed twice with TBS (10 m M Tris-HCl pH 7.4, 150 m M NaCl), then fixed in 3.7% formaldehyde in TBS, for 10 min at 4°C. Coverslips were washed twice in TBS, then permeabilized in TBS containing 0.1% Triton X-100 and 1% B S A for 30 min. The coverslips were then incubated for 1 h in a humidified chamber with E7 primary antibody against P-tubulin (Developmental Studies Hybridoma Bank) at 1:40 dilution, or CENP-E primary antibody at (a gift from Dr. G. Chan, University of Alberta) at 1:1000 dilution in TBS containing 0.1% Triton X-100 and 1% BSA. Following two washes in the same buffer, they were incubated for 30 min in a humid chamber in the dark with goat anti-mouse Alexa 488-conjugated secondary antibody (Molecular Probes A -11029) or goat anti-rabbit Cy3-conjugated secondary antibody (Molecular Probes 81-6115) in the same buffer at 1:500 dilution each. Coverslips were washed 4 times with the same buffer. D N A was stained by incubation with TOTO-3 (Molecular Probes, 2 uM) or with 200 ug/ml Hoechst 33258. Coverslips were mounted in 90% glycerol in PBS with 0.2 M N -propyl gallate. Cells were visualized with either a Nikon E400 fluorescence microscope or a Bio-Rad Radiance Plus confocal microscope. Confocal images were projected using ImageJ. 2.10 Yeast drug induced haploinsufficiency screen 155 diploid yeast strains, each lacking a different allele of a kinase, were arrayed on 2 different 96-well plates. Wells were grown to saturation and used as stock plates for yeast strains. 50 uM OZ (0.5% DMSO) was added to SC-media with 2% glucose, and 35 100 uL was pipetted into each well of two 96-well plates. Stock yeast strains were then pinned into the drug/media-containing plates using a BioRobotics TAS1 robot with a 0.7-mm diameter 96-pin tool which transfers 100 nL per pin. Plates were grown at 30°C and 24 h later were read using a Tecan Sunrise platereader at O D 5 9 5 . 2.11 Western blotting A l l lysates were prepared using Kinexus Bioinformatics Corp. lysate buffer (20 mM MOPS pH 7.0, 2 m M EGTA, 5 mM EDTA, 30 mM NaF, 40 mM p-glycerophosphate pH 7.2, 20 mM sodium pyrophosphate, 1 m M sodium orthovanadate, 1 m M phenylmethylsulfonylfluoride, 3 m M benzamidine, I X Protease Inhibitor Cocktail (Roche), 0.5% Nonidet P-40). Electrophoresis and blotting was performed using standard measures. Unless otherwise indicated, 10 ug of lysate was loaded in each lane for blotting. Antibodies used were against A T M (Oncogene, PCI 16), ATR (Santa Cruz, sc-1887), p-tubulin (Santa Cruz, sc-9104), CDKI (Santa Cruz, sc-54), Chkl (Santa Cruz, sc-7898), Chk2 (Santa Cruz, sc-9064), y - H 2 A X (Upstate, #07-164), Ku86 (Oncogene, NA54), M K K 6 (Biolegend, #604702), phospho-MKK6 (Cell Signalling, #923IS), PKB (Cell Signalling, #9272),and phospho-PKB (Cell Signalling, #4058S). Antibody application varied according to the manufacturers instructions. Detection was accomplished using Supersignal West Pico Chemiluminescence (Pierce) with Kodak X -OMAT A R film according to the manufacturers instructions. 36 2.11.1 Histone acid extraction Cells were lysed using cold N E T N buffer (100 mM NaCl, 100 m M Tris, pH 7.4, 1 mM EDTA, pH 8.0, 1% NP-40), then spun immediately at 13,000 g. Pellets were incubated with 50 uL of 1M HCI at 37°C, and 10 ug of histone extract was electrophoresed and blotted using standard measures. 2.12 Cell clonogenicity assays To assess radiosensitization by checkpoint inhibitors cells were plated in serial dilutions into 6-well plates. They were allowed to attach for 24 h, and then irradiated in the presence or absence of drug at various times post-IR. After the checkpoint inhibitors were washed off, cells were grown in fresh medium for 10 days. Colonies were stained with Malachite Green and counted. Clonogenicity was expressed as % colonies relative to mock-irradiated cells with 24 h equivalent drug treatment. SD was determined from the average of 3 or more data points. 2.13 Cell synchronization For synchronization into G 2 phase, cells were cultured until 50-60% confluent, then incubated with 1 ug/ml aphidicolin (Sigma-Aldrich) for 24 h, washed twice with PBS, and then released into normal medium for 6 h. At this time, approximately 70-80% cells were in G 2 / M as determined by flow cytometry. 37 2.14 Kinase Assays Following mock irradiation or exposure to 6.5 Gy, cells were further cultured as indicated. Cells were then harvested via trypsinization, washed twice in PBS, then lysed in different ice cold kinase lysis buffers (CDKI / Chkl / Chk2: 50 m M Tris-HCl, pH 7.4, 150 mM NaCl, 1% NP-40, I X Protease Inhibitor Cocktail (Roche Diagnostics), 30 ug/ml DNAse I and RNAse A (Sigma-Aldrich), ImM dithiothreitol, and 1 m M sodium orthovanadate; ATR: 20 m M HEPES, pH 7.4, 100 m M NaCl, 10 m M NaF, 1.5 mM M g C l 2 , ImM EGTA, 1% TritonX-100, I X Protease Inhibitor Cocktail, 10 m M (3-glycerol phosphate and ImM dithiothreitol; A T M : 20 mM HEPES, pH 7.4, 100 mM NaCl, 10 m M NaF, 1.5 m M M g C l 2 , ImM EGTA, 0.2% Tween-20, I X Protease Inhibitor Cocktail, 10 mM P-glycerol phosphate and ImM dithiothreitol) by pipetting up and down 15 times, leaving on ice for 10 ten minutes, then pipetting up and down another 15 times. Samples were cleared by centrifugation for 10 min and protein concentration was determined using the B C A Protein Assay kit (Pierce). 125, 250, or 500 ug of protein lysate (CDKI, Chkl/2 or A T M / A T R , respectively) was transferred to chilled tubes, and pre-cleared with 25 ul 50% protein-A agarose (Invitrogen) for 30 min at 4°C. Beads were recovered by centrifugation and lysates were transferred to new tubes containing 1 ug anti-ATR antibody (sc-1887, Santa Cruz), anti-CDKl antibody (KAP-CC001E, Stressgen) anti-Chk2 antibody (sc-9064, Santa Cruz), or anti-Chkl antibody (sc-7898, Santa Cruz) or 2 L i g anti-ATM antibody (Ab-3 PCI 16, Oncogene). Samples were rotated overnight at 4°C, followed by the addition of 20 ul 50% protein-A agarose. Samples were rotated for a further 30 min at 4°C then washed three times in lysis buffer. Samples were then washed 3 times in kinase buffer (CDKI : 50 m M Tris-HCl, pH 7.4, 100 m M NaCl, 10 mM 38 M g C l 2 , 1 m M dithiothreitol, 50 uM ATP, and 1 mM sodium orthovanadate; Chkl/2: 100 mM HEPES, pH 7.4, 50 m M KC1, 10 mM M g C l 2 , 1 mM EGTA, ImM dithiothreitol, and 10 uM ATP; A T M / A T R : 10 m M HEPES, pH 7.4, 50 mM NaCl, 10 m M M g C l 2 , 20 mM MnCl 2 , 20 uM ATP, and 2 m M dithiothreitol), after which the beads were dried down and kept on ice. 25 ul of kinase reaction mixture (kinase buffer with 5-10 uCi/sample y 3 2 P-ATP and 1 ug/sample of Histone HI (Sigma-Aldrich), GST-CDC25C 2 0o-256 (Wang et al, 2002), or PHAS-1 (AG-Scientific) for C D K I , Chkl/2, or A T M / A T R assays, respectively) was added to each sample, then incubated for 30 minutes at 30°C. Samples were separated via SDS-PAGE, and exposed to Kodak X - O M A T A R film and developed according to the manufacturer's instructions. 2.15 Kinexus Bioinformatics "Kinetworks" screen 5 mg of lysate from either irradiated M059J or M 0 5 9 K cells was prepared using Kinexus Bioinformatics lysate buffer (above) at a concentration of 1 mg/mL. Lysate was electrophoresed, blotted to 215 different antibodies and detected as described in (Pelech et al, 2003). Briefly, lysate is loaded into a large comb well encompassing the entire length of the gel. After electrophoresis and transfer to nitrocellulose membrane, a plastic grating is affixed to the membrane with lengthwise lanes for individual antibody incubations. Signal from a membrane with M 0 5 9 K lysate is compared to a similarly blotted membrane with M059J lysate. This work was carried out by Kinexus employees. Up- or down-regulation of signal was assessed by visual approximation. 39 2.16 Reverse-Transcriptase Polymerase Chain Reaction (RT-PCR) amplification Total RNA was extracted from cells using the Trizol Reagent (Invitrogen) according to manufacturers instructions. 10 ug of total RNA was then subjected to 1-st strand reverse-transcriptase Superscript treatment (Invitrogen) according to the manufacturers instructions. 50 ng of this m R N A x D N A hybrid mix was then subjected to PCR amplification using 3 cycles: 94°C for 30 seconds, 55°C for 30 seconds, and 72°C for 2 minutes, 35 cycles, followed by 15 minutes of 72°C using recombinant Taq polymerase (Invitrogen) according to the manufacturers instructions. The following primers were used: M K K 6 5': C A A A A A ATT GTT A A T A T A CCT C T A T A C TTT A A C GTC A A G G G A A T A T A A T G G A C T A T A A G G A C G A T G A T G A C A TGT CT ; M K K 6 3': CTT G C G G G G TTT TTC A G T A T C TAC GAT T C A T A G A T C TCT C G A GTT AGT CTC C A A G A A T C A GTT TTA C A A A A G A T G ; G A P D H 5': A A C GCC TCC TGC A C C A C C ; G A P D H 3': TTC TCC A G G C G G C A G GTC A A G (a kind gift from Dr. I. Sadowski, UBC). 2.17 Preparation of PKB-IRES-GFP cells A retroviral vector encoding a PKB-IRES-GFP puroR cassette was a gift of Dr. M . Gold, UBC (Kohn et al, 1998). The plasmid was transfected into Phoenix cells (a gift from Dr. M . Wabl, UCSF) using Fugene6 (Roche) according to the manufacturer's instructions. 48 hours post-transfection, the virus-containing medium of the Phoenix cells was collected and filtered through a 0.45 um filter, and applied to the target cells with 5 ug/mL polybrene (Sigma-Aldrich) in 6-well plates. Cells were spun for 1 h at 300 g, then cultured for 48 h in virus-containing medium. Media was replaced with fresh media 40 containing 1 ug/mL puromycin (Sigma-Aldrich) and cultured for an additional 7 days. The remaining viable cells were sorted for GFP expression by Andy Johnson (Biomedical Research Centre, UBC) using a FACS Vantage Turbo SE sorter. GFP + cells were collected in a 15 mL tube with 10% media in PBS, then transferred to a tissue culture dish and cultured normally. 2.18 Senescence detection To assess whether cells enter senescence, cells in 6-well plates were washed with PBS, then fixed in 3.7% formaldehyde in PBS for 5 minutes at room temperature. Cells were washed again in PBS, then incubated overnight in acidic X-gal staining solution at 37°C (1 mg/mL X-gal, 150 m M NaCl, 2 m M M g C l 2 , 5 m M K 3 Fe(CN) 6 , 5 m M K4Fe(CN) 6, and 40 mM sodium pyrophosphate, pH 6.0. Cells were gauged for blue X-gal staining by visual approximation. 3 fields of view were counted or 300 cells, whichever first. 2.19 Xenotransplantation studies in scid mice Wild-type HCT116 cells ( + / +), or p53 double-deletion mutants f/") were collected via trypsinization and resuspended at a concentration of 1 x 10 cells/mL in McCoy's 5A Media. 100 uL of this cell suspension was injected sub-cutaneously into the left ( + / + cells) and right (7" cells) flanks of anaesthetized 8 week-old female scid mice (Jackson Labs). Tumours were allowed to grow until reaching approximately 200 mm 3. Tumours were then injected with 50 uL of 1 mM isogranulatimide intra-tumourally and then immediately irradiated with 6.5 Gy (as described above). Mice were sacrificed 24 h post-41 IR, and cells from each tumour were harvested for flow cytometry one-dimensional flow cytometry. To collect cells, tumours were cut into a fine paste using a razor, then resuspended in 5 mL PBS, with 0.3 mL each of 10 mg/mL DNAse (Sigma-Aldrich), 4 mg/mL collagenase (Roche) and 25 mg/mL trypsin (Roche) and rotated for 30 min at 37°C. Cells were filtered through a 70 uM filter, supplemented with 5 mL of McCoy's 5A Media with 10% FBS and an additional 0.2 mL of 10 mg/mL DNAse. Cells were left on ice for 5 min, then washed twice in SAB and fixed as described above. Cytometric labelling was done as described above, except no GF-7 labelling was used, cells were only labelled with propidium iodide. 2 .20 Sta t is t ica l analysis Standard deviation (SD) was determined from the average of 3 or more data points. When appropriate for marginal data sets, a Student's f-test (p) was used to determine statistical significance. 42 Chapter 3 Identification and characterization of novel checkpoint inhibitors 3.1 I N T R O D U C T I O N As reviewed in Chapter 1, D N A damage activates cell cycle checkpoints that temporarily halt G) progression, slow down S-phase traversal and delay the G2 /M transition (reviewed in (Sancar et al, 2004; Niida and Nakanishi, 2006)). These responses are thought to promote genomic integrity by allowing increased time for D N A repair. In cancer cells, the G i checkpoint is typically inactive because of its dependence on p53, which is mutated in the majority of cancers (Kastan et al, 1991). Therefore, the resistance of tumour cells to radiotherapy and DNA-damaging chemotherapeutic agents is likely to depend significantly on the integrity of the G2 checkpoint. Since the G 2 checkpoint pathway may represent the cancer cell's last defence against D N A damage, inhibition of the checkpoint offers a potential therapeutic approach to increase the efficacy of common DNA-damaging anticancer treatments (reviewed in (Anderson et al., 2003; Kawabe, 2004; Prudhomme, 2004)). A number of G2 checkpoint inhibitors have been discovered in recent years (see Table 1.1). Most target protein kinases in the signalling cascade; caffeine inhibits A T M 1 A section of this chapter has been published (Sturgeon, C. M . , K . Craig, C . Brown, N . T. Rundle, R. J. Andersen and M . Roberge (2005). Modulat ion of the G 2 cell cycle checkpoint by sesquiterpene lactones psilostachyin A and C isolated from the common ragweed Ambrosia artemisiifolia. Planta Medica 71(10): 949-954). Another is in review for publication (Sturgeon, C. M . , A . Diaz-Marrero, B . Cine l , L . McHardy , M . Ngo, R. J. Andersen and M . Roberge (2006a). Abrogation o f ionizing radiation-induced G 2 checkpoint and inhibition o f nuclear export by cryptofolione). 43 (Sarkaria et al, 1999; Zhou et al, 2000), U C N - 0 1 , CBP501 , isogranulatimide and some others inhibit C h k l (Sarkaria et al, 1999; Graves et al, 2000; Y u et al, 2002; Jiang et al, 2004), and debromohymenialdesine inhibits both C h k l and Chk2 (Curman et al, 2001). Several of these checkpoint kinase inhibitors are in preclinical development and the C h k l inhibitory peptide CBP501 recently entered clinical trials (Garber, 2005). It is not yet known which other components of the G 2 checkpoint are targetable by drugs or whether their inhibition w i l l result in the selective ki l l ing of cancer cells. Gaining further insight into the signalling mechanisms involved could lead to the design of better inhibitors, and potentially to the development of therapeutic agents (Anderson et al., 2003). To identify new G 2 checkpoint inhibitors, a phenotypic cell-based screen (Roberge et al, 1998) that does not require a priori knowledge of the checkpoint mechanism was used. G 2 arrested cells were screened against a chemical collection for compounds which lead to checkpoint inhibition. This chapter describes the identification of the sesquiterpene lactones psilostachyin A and C from the common ragweed Ambrosia artemisiifolia, as well as the cryptofoliones from Cryptocaria concinna as novel G 2 checkpoint inhibitors and the characterization of their biological activities. Psilostachyin, in sharing some similarity to the checkpoint inhibitor O Z (Rundle et al, 2001), may inhibit the checkpoint via Wee l inhibition. In the case of the cryptofoliones, checkpoint inhibition correlates with nuclear export inhibition. The highly selective nuclear export inhibitor leptomycin B also potently inhibits the G 2 checkpoint. Therefore, inhibition of nuclear export could be an alternative approach to checkpoint kinase inhibition for abrogating the G 2 checkpoint. 44 3.2 RESULTS AND DISCUSSION 3.2.1 Psilostachyin 3.2.1.1 Checkpoint inhibition by psilostachyins Using the cell-based assay for the identification of compounds that inhibit the G2 DNA damage checkpoint (Roberge et al, 1998), activity in an extract from the common ragweed Ambrosia artemisiifolia (NCI Natural Products Repository extract N35791) was detected. The assay was further used to guide the isolation and identification, by spectroscopic methods, of the active sesquiterpene lactones psilostachyin A and psilostachyin C (Figure 3.1 A). The related inactive compounds psilostachyin B, altamisic acid, paulitin and isopaulitin were also isolated from the extract (Figure 3.IB). Psilostachyin A and C produced a dose-dependent response that had not reached a plateau at 50 uM, the highest concentration that could be used without compound precipitation (Figure 3.2). At this concentration, psilostachyin C and A overrode G2 arrest in approximately 50% and 40% of the cell population, respectively. The efficacy of psilostachyin C as a checkpoint inhibitor was comparable to that of caffeine and UCN-01 (50-70%), the most effective of checkpoint inhibitors, and higher than 13-hydroxy-15-oxozoapatlin (OZ) (15-20%), isogranulatimide and debromohymenialdisine (30-40%) (see Figure 3.5). Psilostachyin B, paulitin, isopaulitin and altamisic acid showed no significant G2 checkpoint inhibition (Figure 3.2 and data not shown). The structures of psilostachyins A and C are unlike those of other G2 checkpoint inhibitors described to date (reviewed in (Anderson et al, 2003; Kawabe, 2004)). However, they possess an a,p-unsaturated carbonyl group also found in OZ. This group 45 Psilostachyin A B \ 1 o Psilostachyin C Psilostachyin 8 Altamisic acid H O Paulitin •if or ^cr H •«<H O Isopaulitin * 0 " " t l O g u r e 3.1 : S t r u c t u r a l f o r m u l a e o f s e s q u i t e r p e n e l a c t o n e s i s o l a t e d f r o m Ambrosia artemisiifolia. A, active compounds. B, inactive compounds. C, chemically modified 0-mercaptoethylated psilostachyin A . 46 Figure 3.2 : G 2 checkpoint inhibition by psilostachyins A and C. MCF-7 mp53 cells were irradiated with 10 Gy, cultured for 16 h, and then incubated with 100 ng/mL nocodazole and the indicated concentrations of psilostachyin A (filled circle), psilostachyin C (dark triangle), psilostachyin B (open circle), paulitin (open triangle), isopaulitin (dark square), or the psilostachyin A-p-mercaptoefhanol adduct (open square) for 8 h. Checkpoint inhibition was measured by mitotic spreads as described in Chapter 2.5.1. Data shown are the means +/- SD (n > 4). 47 made OZ reactive towards nucleophiles, such as cysteine residues in proteins, and elimination of this reactivity, by covalent linkage of OZ to p-mercaptoefhanol, abolished its checkpoint inhibitory activity (Rundle et al, 2001). To determine whether its ot,P-unsaturated carbonyl group is required for activity, psilostachyin A was reacted with P-mercaptoethanol to yield the adduct shown in Figure 3.1C. This compound showed no checkpoint inhibitory activity (Figure 3.2). However, the observation that the inactive compounds psilostachyin B, altamisic acid, paulitin and isopaulitin also possess one or two a,P-unsaturated carbonyl groups (Figure 3.1) indicates that although an a,P-unsaturated carbonyl group is necessary for G 2 checkpoint inhibitory activity, its presence is not sufficient to confer activity to this class of compounds. 3.2.1.2 Mitotic arrest induced by psilostachyin I next determined the effect of psilostachyin A and C on cell cycle progression in unirradiated, proliferating cells. MCF-7 cells were incubated with 50 uM psilostachyin C and the proportion of cells in G], S, G 2 and M was monitored over time by flow cytometry (Rundle et ai, 2001). Psilostachyin C caused no change in the proportion of cells in Gi and a small increase in the S-phase cell population, which happened during the first 3 h of incubation (Figure 3.3 A). Strikingly, psilostachyin C caused a significant decrease in the percentage of G 2 cells accompanied by a large increase in the number of mitotic cells (Figure 3.3A). These data show that psilostachyin C blocks mitotic progression. Compounds that block cell cycle progression only at mitosis cause a linear increase in the M population over time with a proportionate decrease of cells in all other cell cycle phases. This is clearly not the case for psilostachyin C; M-phase accumulation 48 B DMSO PC (50 MM) p-tubulin DNA Merge Figure 3.3 : Effect of psilostachyin C on cell cycle progression. A, MCF-7 mp53 cells were incubated with 50 uM psilostachyin C for the indicated times and their cell cycle distribution was determined by flow cytometry as described in Chapter 2.5.2. Results are the means +/- SD (n = 3). B, HCT116 cells were exposed or not to psilostachyin C (PC) for 4 h and their chromosomes and microtubules were visualized by immunofluorescence microscopy. The panel shows a control (DMSO) mitotic cell with microtubules emanating from spindle poles and condensed chromosomes aligned at the metaphase plate (top row), and a psilostachyin C-treated (PC) cell with condensed chromosomes that are not aligned and a highly abnormal arrangement of microtubules. Results shown are representative of duplicate experiments. Bar, 10 um. 49 was not sustained over time and the proportion of cells in interphase did not decrease after 3 h (Figure 3.3A), indicating that psilostachyin C interferes with interphase progression as well, or that the mitotic arrest is transient and/or toxic. I next examined the mitotic arrest phenotype of cells treated with 50 uM psilostachyin C for 4 h. A l l mitotic cells showed a similar unusual appearance, with metaphase-like condensed chromosomes that failed to align, and major disruption of mitotic spindles (Figure 3.3B). Separate spindle poles were often, but not always, distinguishable, but no clearly organized mitotic spindle was observed. Instead, the microtubules formed long, thick and tortuous fibers (Figure 3.3B). No cells in metaphase, anaphase or telophase were observed. Psilostachyin A showed similar effects (not shown). Most agents that block cells in mitosis bind directly to microtubules or tubulin dimers to either inhibit or stimulate tubulin polymerization (Hadfield et ai, 2003). Agents that inhibit microtubule polymerization cause a disappearance of microtubules and an increase in diffuse tubulin staining throughout the cell, which was not observed here. Agents that stimulate microtubule polymerization can produce thick bundles of microtubules. Therefore, whether psilostachyin C can directly stimulate the polymerization of tubulin in vitro was investigated. Microtubule polymerization did not take place in the presence of DMSO, was stimulated by the known microtubule-stabilizing agent paclitaxel, but was not stimulated by psilostachyin C (Figure 3.4). Therefore, psilostachyin C must cause mitotic arrest via a target other than microtubules. In this respect, psilostachyin C differs from the sesquiterpene lactones costunolide and 50 gure 3.4 : Psi lostachyin C does not affect tubu l in po lymer izat ion in vitro. Tubulin will not spontaneously polymerize at a concentration of 1 mg/mL (open circle). However, taxol (filled circle) will stabilize microtubules, allowing them to form in vitro. This can be measured in a temperature-controlled (37°C) spectrophotometer over time, with an increase in O D 3 4 0 indicative of microtubule formation. 1% D M S O (open circle; negative control), 10 u M taxol (filled circle), 100 u M psilostachyin C (filled triangle), 10 u M psilostachyin C (open triangle), or 0.1 u M psilostachyin C (open square). Spontaneous tubulin polymerization is not stabilized by psilostachyin. This experiment was performed by Geoffrey Karjala. 51 parthenolide, which possess microtubule stabilizing activity (Bocca et al, 2004; Miglietta etal, 2004). 3.2.1.3 Enhancing checkpoint inhibition by combining with other inhibitors No checkpoint inhibitors identified to date are able to induce all G2-arrested cells to enter mitosis. One possible explanation for this observation is that redundancy in the checkpoint signalling pathway requires several targets to be inhibited in order to achieve complete checkpoint inhibition. It is also possible that checkpoint inhibitors are insufficiently selective at high concentrations and target additional proteins or pathways necessary for cell cycle progression or survival, thus antagonizing their checkpoint inhibitory activity. Given the potential therapeutic benefit of complete checkpoint inhibition in cancer therapy, I next asked whether psilostachyin C could act additively or synergistically to cause G 2 checkpoint abrogation with other checkpoint inhibitors. Psilostachyin C was tested at 25 and 50 uM in combination with optimal concentrations of the Chkl inhibitors isogranulatimide and UCN-01, the Chkl and Chk2 inhibitor debromohymenialdisine, the A T M inhibitor caffeine or OZ, a checkpoint inhibitor with an unknown mechanism of action. Alone, 10 uM isogranulatimide induced 29.3 ± 7.1 % cells to enter mitosis and 25 uM psilostachyin C 32.1 ± 3.4 %. In combination they induced 49.1 ± 2.7 % cells to enter mitosis, a slightly less than additive effect (Figure 3.5A) (p < 0.0001). Alone, 100 nM UCN-01 induced 44.3 ± 2.8 % cells into mitosis and this increased to 58.0 ± 6.8 % with 25 uM psilostachyin C (Figure 3.5A), 52 Figure 3.5 : Checkpoin t i nh ib i t i on by combinat ions o f checkpoint inh ib i to rs . A, MCF-7 mp53 cells were treated with DMSO, 25 uM psilostachyin C or 50 uM psilostachyin C as indicated in the x axis, together with nocodazole and no second compound (black), 10 u.M 1GR (grey), 10 uM OZ (horizontal stripes), 2 mM caffeine (white), 100 nM UCN-01 (diagonal stripes), or 40 u.M DBH (crossed stripes), from 16-24 h post-10 Gy. Checkpoint inhibition was assessed by mitotic spreads, as described in Chapter 2.5.1. The combination of psilostachyin C with either IGR or UCN-01 led to a less-than additive effect (p < 0.0001). B, Combinations of psilostachyins A and C or A and B at 25 uM each (black) or 50 uM each (white), on MCF-7 mp53 cells treated as in (A). Checkpoint inhibition was assessed via mitotic spreads as described in Section 2.5.1. Results shown are means +/- SD («>3). 53 again a less than additive effect (p = 0.0001). The effects of caffeine, debromohymenialdisine or OZ alone did not significantly change when combined with 25 uM psilostachyin C, nor was any additional activity observed for any of the checkpoint inhibitors when combined with 50 uM rather than 25 u M psilostachyin C. The combination of 25 uM psilostachyin A with 25 uM psilostachyin C was additive, but not higher than 50 uM psilostachyin C alone (Figure 3.5B). Combination of 50 uM psilostachyin A and 50 u M psilostachyin C showed no additivity at all (Figure 3.5B). Therefore, the psilostachyins show additive effects with two Chkl inhibitors but not with inhibitors acting by other mechanisms. However, no combination was capable of achieving complete checkpoint inhibition. 3.2.1.4 Psilostachyins fail to radiosensitize cells At checkpoint inhibitory concentrations, some checkpoint inhibitors such as caffeine, UCN-01 and isogranulatimide, have little effect on cell proliferation in the absence of D N A damage, but in the presence of D N A damage they can potentiate the antiproliferative effects of the damage in cells lacking p53 function (Fan et al, 1995; Powell et al, 1995; Russell et al, 1995; Wang et al, 1996; Roberge et al, 1998). To determine whether the psilostachyins show similar properties, I measured their effects on cell proliferation in the absence or presence of D N A damage. MCF-7 mp53 cells were subjected or not to y-irradiation (10 Gy) and simultaneously incubated with various concentrations of psilostachyins. After 24 h, the compounds were washed away and cell growth was measured after 5-7 days. In the absence of irradiation, psilostachyin A and C were moderately potent inhibitors of cell proliferation with IC50 of 20 uM and 10.6 uM 54 F i g u r e 3.6 : I n h i b i t i o n o f c e l l p r o l i f e r a t i o n b y p s i l o s t a c h y i n s . MCF-7 mp53 cells were irradiated with 10 Gy (solid symbols) or not (open symbols) and immediately treated with psilostachyin A (circles), psilostachyin B (down triangles), psilostachyin C (squares), paulitin (diamonds), isopaulitin, (up triangle), P-mercaptoethylated psilostachyin A (hexagons). Cell proliferation measured using the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay as described in Chapter 2.7. Results shown are means +/- SD (n = 6). 55 respectively (Figure 3.6), close to the concentrations at which they show checkpoint inhibitory activity. The (3-mercaptoethanol-psilostachyin A adduct, which was inactive as a checkpoint inhibitor and as an antimitotic agent, did not inhibit cell proliferation at any concentration tested, indicating that the Michael acceptor site is required for inhibition of cell proliferation. The closely related psilostachyin B, which was also inactive as a checkpoint inhibitor and as an antimitotic agent, did not inhibit cell proliferation at any concentration tested, indicating precise structural requirements for inhibition of cell proliferation by this class of compounds. On the other hand, paulitin and isopaulitin, which were both inactive as checkpoint inhibitors and antimitotic agents, were potent inhibitors of cell proliferation, with IC 5 0 of 1.5 uM and 4 uM, respectively. In combination with ionizing radiation, none of the psilostachyins showed potentiation of inhibition of cell proliferation (Figure 3.6), even though other checkpoint inhibitors such as caffeine, isogranulatimide, debromohymenialdisine and OZ showed strong potentiation in this same experimental system (Roberge et al, 1998; Jiang et al, 2000; Curman et al, 2001; Rundle et al, 2001). If anything, these compounds seemed to afford a small degree of protection. Therefore, the psilostachyins do not appear promising as agents to enhance the sensitivity of tumour cells to D N A damage. However, the aforementioned checkpoint inhibitor OZ was recently shown to have antitumour activity in vivo in a mouse hollow fiber model when used as a single agent (Braca et al, 2004) and it would be interesting to determine the effects of psilostachyins on tumour growth in vivo when used as single agents. 56 3.2.1.5 Target of psilostachyin? Psilostachyin does not inhibit A T M , ATR, Chkl/2, or PP1/PP2A (data not shown, S. Lees-Miller, personal communication). This then raises the obvious question, what is the target of psilostachyin that is involved in its checkpoint inhibitory activity? Because Michael-type addition is required for its activity, I initially intended to biotinylate psilostachyin (see Figure 3.7) and then immunoprecipitate the protein(s) it is covalently linked to via streptavidin-conjugated agarose beads. The biotin/streptavidin complex formation is one of the strongest non-covalent interactions known to-date. Unfortunately, it was not possible to produce biotinylated psilostachyin (K. Craig, personal communication), so other approaches had to be contemplated. The best candidate-based approach was identified via the structure of psilostachyin : psilostachyins belong to a family of compounds called sesquiterpene lactones, which are known to inhibit N F - K B activation and transcription (Bork et al, 1997). However, when tested using an NF-KB-mediated luciferase reporter assay, psilostachyin did not confer any significant degree of N F - K B inhibition compared to irradiation alone (data not shown, C. Brown, personal communication), indicating that N F - K B inhibition via psilostachyin does not contribute to checkpoint inhibition. The checkpoint inhibitor, OZ, also demonstrates a requirement for its Michael acceptor functionality, and exhibits a similar mitotic arrest phenotype to psilostachyin (Rundle et al, 2001). Extensive work performed by Natalie Rundle in the Roberge laboratory revealed that OZ inhibits nuclear import via RanBP2, and disrupts the localization of CENP-E (Rundle et al, in prep). This was found by using an approach similar to that outlined in Figure 3.7. Using this biotinylated probe, RanBP2 was 57 N-iodoacetyl-N-biotinylhexylenediamine Altamisic acid biotinylated psilostachyin Figure 3.7 : Proposed generation of biotinylated psilotachyin. Altamisic acid, a naturally occurring analog of psilostachyin would be reacted with /V-iodoacetyl-A7-biotinylhexylenediamine, to yield an intermediate product which would then have its double-bond removed to more closely match that of psilostachyin A and C. 58 identified as a target of OZ and related compounds E K A and O K A (Rundle et al, in prep.). This lead to testing for inhibition of nuclear import, as RanBP2 is a component of the nuclear pore complex involved in nuclear membrane transport (Yokoyama et al, 1995). Psilostachyin did not impair nuclear import (Figure 3.8), indicating that it does not target RanBP2 in the same manner as O Z / E K A / O K A . It did, however, disrupt CENP-E localization, as did OZ. Briefly, centromeric protein E (CENP-E) is a kinesin-like motor protein, essential for chromosomal movement to the metaphase plate (Yen et al, 1991; Yen et al, 1992). In normally dividing cells, CENP-E is localized to the kinetochore, a large structure where microtubules bind to chromosomes to orchestrate mitosis, as evidenced by the punctuate pattern exhibited in Figure 3.9 (top panels) (Yen et al, 1991; Yen et al, 1992; Brown et al, 1996; Cooke et al, 1997). Psilostachyin-treated cells do not exhibit this same pattern, with staining appearing diffuse, indicating that kinetochore assembly is impaired (Figure 3.9, lower panels) (Schaar et al, 1997; Wood et al, 1997). Unfortunately, as neither of these observations likely offers any insight into its ability to inhibit the G2 checkpoint, these properties of psilostachyin were considered as beyond the scope of this thesis and were therefore not pursued further. As candidate-based approaches to identifying the relevant target(s) of psilostachyin did not succeed, I turned to an approach using drug-induced haploinsufficiency of yeast as a tool to identify the target of psilostachyin. The assay is based on the simple principle that decreasing the concentration of a protein that is the target of a drug should increase sensitivity to the drug (reviewed in (Sturgeon et al., 2006b)). In yeast, the levels of a protein can be lowered to approximately one half by 59 A DMSO 30 min Dex 0 min Merge B DNA Merge DMSO 30 min Dex 30 min PC 30 min DexO min Dex 30 min Dex 30 min DMSO 30 min DMSO 30 min Dex 30 min PC 30 min Dex 30 min PC 30 min No Dex 0 min No Dex 60 min No Dex 0 min No Dex 60 min ure 3.8 : Psilostachyin C does not inhibit nuclear import or export. A, cells were treated with either DMSO or psilostachyin C (PC, 50 |JM) for 30 min, then 30 min with 1 uM dexamethasone to induce nuclear localization of the Rev/GR/GFP construct. The cells were then fixed and Rev/GR/GFP was visualized by fluorescence microscopy. No inhibition of nuclear import was observed with psilostachyin. B, cells were treated with dexamethasone for 30 min, followed by psilostachyin for an additional 30 min, then washed but maintained in the presence of drug for 60 min. The cells were then fixed and Rev/GR/GFP was visualized by fluorescence microscopy. No inhibition of nuclear export was observed with psilostachyin. DNA is stained with Hoechst 33342 (blue), Rev/GR/GFP is green. Results shown are representative of that observed in 3 random fields of view of one experiment. 60 DNA CENP-E pMubulin Merge DMSO Figure 3 . 9 : Psilostachyin C impairs CENP-E localization during mitosis. HCT116 cells were treated with psilostachyin (PC, 50 u M ) for 8 h, then fixed and analyzed by immunofluorescence staining and microscopy as described in Chapter 2.9. DNA-blue, P-tubulin-green, CENP-E-red. Top panels, a normal mitotic cell exhibits punctate CENP-E localization at the site of kinetochores. Lower panels, a cell arrested in mitosis from psilostachyin treatment exhibits grossly disrupted microtubules and diffuse CENP-E localization throughout the cell. Results shown are representative of 3 random fields of view of one experiment. 61 deleting one copy of a gene in a diploid strain. A set of heterozygous diploid strains covering deletions of -5,000 non-essential and -1,000 essential genes has been generated by the Saccharomyces Genome Deletion Project and can be purchased (Winzeler et al., 1999). This set covers essentially all protein products within the yeast genome. It has been used to test the idea that the mechanism of drug action can be deduced by analyzing the sensitivity of each mutant strain to a drug. This approach has been named drug-induced haploinsufficiency by extension of the concept of haploinsufficiency, whereby deletion of one of the two copies of some genes can cause an abnormal phenotype, and refers to the situation in which a heterozygous diploid strain lacking one copy of a gene is more sensitive to a drug than the wild-type strain. This assay was recently used in the Roberge lab to identify the targets of another compound, dihydromotuporamine C, that had an unknown mechanism of action (Baetz et al., 2004), so I reasoned that it might be a suitable approach for psilostachyin as well. As a first step to this assay, a suitable concentration of drug must be found empirically that will lead to moderate growth inhibition of a wild-type strain of yeast, so that mutant strains lacking the target of the drug will be further impaired in their growth (reviewed in (Sturgeon et ai, 2006b)). Unfortunately, however, psilostachyin did not lead to any growth inhibition at concentrations as high as 100 uM (data not shown), indicating that it is impermeable to the yeast cell wall or that it has no relevant target in yeast. Fortunately, the checkpoint inhibitor OZ, which shares some characteristics with psilostachyin (as described above), exhibited moderate growth inhibition at 50 uM (Figure 3.10) and was therefore chosen for further analysis. 62 m CO Q O DMSO 0.001 uM 0.005 uM 0.01 uM 0.05 uM 0.1 uM 0.5 uM 1.0 uM 5.0 uM 10.0 uM 50.0 uM 100.0 uM Figure 3.10 : Growth curve of wild-type yeast exposed to various concentrations of the checkpoint inhibitor, OZ. Maintained at 30°C in a 96-well plate, yeast growth was monitored via O D 5 9 5 every hour for 24 h. DMSO was 0.5%. 50 uM OZ (highlighted in red) lead to an intermediate growth rate compared to DMSO and lesser concentrations of OZ vs. near-complete growth inhibition in the presence of 100 uM OZ. 63 OZ (50 uM) was added, to a final concentration of 0.5% DMSO, to synthetic liquid media in 96-well plates. 155 strains of yeast, each of which lacks one copy of a specific protein kinase were pinned using a BioRobotics TAS1 robot with a 0.7-mm diameter 96-pin tool which transfers 100 nL into 96-well plates, and then allowed to grow overnight at 30°C. Growth was measured via OD595 n m . Results are summarized in Figure 3.11, and those strains exhibiting less than 50% growth compared to wild-type are listed in Table 3.1. 18 different mutant strains exhibited slow growth in the presence of OZ. The genes are required for various aspects of cellular metabolism/catabolism, which may be related to its cytotoxic properties. Some mutants with high sensitivity to OZ, such as CDC7, KIN4, or Y C K 2 may be relevant to the antimitotic effect of OZ on cells, as these kinases are involved in regulating mitosis. As mentioned previously though, exploring the mode-of-action of psilostachyin (and by extension, OZ) with respect to its antimitotic effect is beyond the scope of this thesis. With respect to mutants which play a role in the G 2 / M checkpoint/transition, only 1 strain was identified, SWE1. The protein encoded by SWE1 is a homolog of Wee 1, a tyrosine kinase that phosphorylates C D K I on tyrosine-15 (Y15), which in concert with Mytl-dependent threonine-14 (T14) phosphorylation, inactivates C D K I until it is dephosphorylated by CDC25 (reviewed in (Niida and Nakanishi, 2006)). During normal cell cycle progression, Chkl phosphorylates Weel on Ser549, activating it to prevent premature mitotic progression with incompletely replicated D N A (Stanford and Ruderman, 2005). However, in the D N A damage checkpoint, Chkl can also phosphorylate Weel, causing G 2 delay (O'Connell et al, 1997). Also, chemical inhibition of Weel, using the selective inhibitor PD0166285, can cause 64 1.2 1.0 'OWt 0.8 CD I - 0.6 o 0.4 0.2 0.0 • • • • n 1 i I I r 1 1 1 1 1 1 1 1 1 1 Strain gure 3.11 : Growth of 155 diploid yeast strains lacking 1 allele of each kinase in the yeast genome, exposed to 50 uM OZ. Various yeast strains were robotically pinned into 96-well plates containing SC-media and 50 uM OZ, then cultured for 24 h at 30°C. Growth was measured via OD 5 95. Expressed as % of growth relative to wild-type. Strains identified as sensitive to OZ are classified as those exhibiting less than 50% wild-type growth (dashed line). 65 Table 3.1 : Saccharomyces cereviscae heterozygous diploid strains exhibiting impaired growth in the presence of 50 uM OZ Strain % of W.T. Growth Function* ADK2 32.6 Adenylate kinase A K L 1 40.2 Actin organization CBK1 44.6 Cell wall biosynthesis CDC7 29.7 Regulation of D N A replication ERG 12 38.2 Ergosterol biosynthesis KIN4 49.3 Spindle organization checkpoint PYK2 45.8 Pyruvate kinase RIM 11 36.3 Similar to GSK-3P SEC59 41.3 Protein glycosylation SKY1 49.5 mRNA metabolism SLN1 43.0 Osmosensor, M A P K STE11 42.3 Pseudohyphal/invasive growth STE7 35.5 Pseudohyphal/invasive growth SWE1 47.3 Weel, G2 /M transition regulation THI22 42.0 Thiamine metabolism VHS1 48.7 Similar to SKS1 Y C K 2 48.3 Endocytosis/cytokinesis YPK2 49.4 Similar to SGK * see Saccharomyces Genome Database (www.yeastgenome.org) 66 checkpoint abrogation in vivo (Wang et al, 200Id). Therefore, it is conceivable that OZ, and likely psilostachyin, act as checkpoint inhibitors by inhibiting Weel kinase activity. Formal demonstration of Weel inhibition by measuring Weel kinase activity in vitro has not yet been accomplished due to insufficient amounts of both OZ and psilostachyin. Although more experiments with purified compound are required before any final conclusions can be drawn, Weel represents a likely target to accomplish checkpoint inhibition. 3.2.2 Cryptofoliones 3.2.2.1 Isolation and identification of cryptofoliones A cell-based assay for G2 checkpoint inhibition was used to screen 20,000 plant extracts from the National Institutes of Health Natural Products Repository. A crude extract of Cryptocarya concinna (N29449), a tree of the laurel family, showed strong G2 checkpoint inhibitory activity. The major active compound was purified from the crude extract using assay-guided fractionation and was identified as cryptofolione (Figure 3.12) by analysis of its Nuclear Magnetic Resonance (NMR) and Mass Spectroscopy (MS) data. Cryptofolione was previously isolated from a plant of the same genus (Sehlapelo et al, 1994) and shown to have mild cytotoxicity towards a variety of cancer cell lines (Schmeda-Hirschmann et al, 2001; Dumontet et al, 2004). To the best of my knowledge, no investigation of its mechanism of action has been described. 67 Figure 3.12 : Structural formulae of cryptofolione and related compounds. Compounds were isolated from C. concinna or prepared by chemical modification of cryptofolione. Compounds 1-6 were naturally occurring, while 7-10 were prepared by chemical modification of natural compounds. Chemical modification was performed by Bruno Cinel. 68 3.2.2.2 Inhibition of ionizing radiation-induced G2 arrest by cryptofolione G2 checkpoint inhibition by cryptofolione was examined in human breast carcinoma MCF-7 cells expressing dominant-negative p53. Cells were exposed to 10 Gy ionizing radiation to induce 90% G2 arrest (Roberge et al, 1998) and then were incubated for 8 h with dimethyl-sulphoxide (DMSO) (negative control), cryptofolione or caffeine (positive control). The drug carrier DMSO did not overcome G2 arrest, but exposure to 30 |j.g/ml cryptofolione (1) caused a reduction in the number of G2-arrested cells accompanied by a large increase in the number of mitotic cells, demonstrating G2 checkpoint abrogation (Figure 3.13). Cryptofolione was more potent but less efficacious than the well-characterized checkpoint inhibitor caffeine, which averages approximately 60% mitotic index under identical conditions (data not shown). The cryptofolione-related compounds 2 and 5 however, exhibited extremely efficacious checkpoint abrogation at 10 ug/mL, resulting in the highest percentage of cells in mitosis, above that observed for caffeine. Interestingly, cells treated with cryptofolione were unable to exit mitosis (not shown), as is the case for the structurally unrelated checkpoint inhibitors 13-hydroxy-15-oxozoapatlin and psilostachyin (Rundle et al, 2001; Sturgeon et al, 2005), but unlike other checkpoint inhibitors described to date, including UCN-01, debromohymenialdisine, caffeine and isogranulatimide (M. Roberge, personal communication; data not shown). This aspect of cryptofolione was not explored further. 69 Figure 3.13 : G 2 checkpoint inhibition by cryptofolione and related compounds. MCF-7 mp53 cells were irradiated with 10 Gy, cultured for 16 h, then exposed to the indicated concentrations of each compound for 8 h with 100 ng/mL nocodazole. G 2 checkpoint inhibition was assessed by flow cytometry as described in Chapter 2.5.2. Compounds 2 and 5 exhibit very strong checkpoint inhibition at 10 ug/mL. Results shown are either means of duplicate experiments (due to limiting amounts of compound), or means +/- SD (n > 3). 70 3.2.2.3 Structure-activity study To gain a better understanding of the structural requirements for checkpoint inhibition by cryptofolione, I compared its activity to that of related compounds isolated from the extract or prepared by chemical modification of cryptofolione, as assessed by mitotic index following D N A damage (n > 1, dependent on compound availability). Compounds 2-6 (Figure 3.12) were isolated as minor components of the C . concinna extract and their activities are shown in Figure 3.1. As mentioned above, compound 2, the (Z) isomer of cryptofolione at the C7'-C8' carbons, showed very high activity at a low concentration, with a peak activity at 10 ug/mL. Compound 3, with a carbonyl at position 6', and compound 4, with a methoxy substituent at position 6', both showed little or no activity (Figure 3.13). Compound 5 and 6, which lack the carbons CI ' -C2' , were both active, although 5 exhibited very strong checkpoint inhibition similar to 2. Additional compounds were prepared by chemical modification of cryptofolione. The acetonide compound 7 showed moderate activity and was slightly more potent than cryptofolione. Compound 8, which is acetylated at positions 4' and 6', was also more potent than cryptofolione but much less efficacious. Compound 9, in which the double bonds at C3-C4, C l ' - C 2 ' and C7'-C8' are reduced, was inactive. Finally, compound 10, which has an open lactone ring, no a,(3 unsaturated carbonyl, and is saturated at CI '-C2' and C7'-C8', showed little or no activity (Figure 3.13). Compounds 2, 5, 6 and 7 showed reduced activity at concentrations above 10 p.g/ml (Figure 3.13). Overall, these results indicate that i) the lactone ring is important as 10 is inactive, and ii) the length of the compound influences activity, as 2 and 5 are much more active than 1, while 6 may be too short as its activity is less than 2. Although 2 has the same 71 number of carbons composing the long chain in 1, the (Z) configuration may cause the benzyl-group to be in closer proximity to the lactone ring. Similarly, lacking the CI '-C2' carbons in 5 may lead to a similar 3-dimensional structure, with the benzyl-ring closer to the lactone ring. Compound 6, which also lacks the CI ' -C2' carbons, is in the (Z) configuration, and may result in the benzyl-ring being too close to the lactone ring for optimal activity. Compounds 2, 5, and 6 were more potent and efficacious than cryptofolione (1). 3.2.2.4 Inhibition of nuclear export by cryptofolione The structure of cryptofolione is unlike that of most other checkpoint inhibitors described to date (reviewed in (Anderson et al, 2003; Zhou and Sausville, 2003)). Similar to psilostachyin, cryptofolione does not inhibit the checkpoint kinases A T M , ATR, Chkl or Chk2 or the protein phosphatases PP1 and PP2A in vitro (data not shown). Cryptofolione bears some structural resemblance to the inhibitors of Crml-mediated nuclear export leptomycin B and ratjadone, in that both possess a lactone-ring followed by a long carbon chain with hydroxyl and carbonyl groups (Kudo et al, 1998; Meissner et al, 2004). In addition, when a C O M P A R E analysis was carried out (Paull et al, 1989) of correlations between gene expression levels and drug sensitivity for different cancer cell lines, NLP-1 (nucleoporin-like protein 1) was found as the gene whose expression levels correlated most highly (Pearson Correlation Coefficient^).616) with the sensitivity to cryptofolione of cancer cell lines with mutated p53. NLP-1 is a protein that interacts with Crml and with nuclear export sequences in proteins (Farjot et al, 1999). These observations prompted me to examine whether cryptofolione affects nuclear export. 72 I used a cellular system in which the HIV-1 Rev protein, which contains a nuclear export sequence, is fused to the hormone-responsive element of the glucocorticoid receptor and to GFP (Love et al, 1998). Addition of dexamethasone causes nuclear accumulation of Rev/GR/GFP in nucleoli at glucocorticoid response elements (GRE). Removal of dexamethasone causes Rev/GR/GFP to relocalize to the cytoplasm, allowing visualization of nuclear export signal-mediated nuclear export. Rev/GR/GFP was localized mostly in the cytoplasm in the absence of dexamethasone (Figure 3.14A) and in the nucleus in the presence of dexamethasone (Figure 3.14B), as previously described (Love et al, 1998). Most of the protein was exported out of the nucleus by 60 min after dexamethasone removal, a process that was not affected by the drug carrier DMSO (Figure 3.14H). Leptomycin B prevented export of the protein to the cytosol (Figure 3.14C), as previously observed (Nishi et al, 1994). At concentrations that inhibited the checkpoint, cryptofolione and other active analogs inhibited nuclear export to varying degrees (for examples, see Figure 3.14E-G). The cryptofolione analogues that showed little or no checkpoint inhibition also showed little or no nuclear export inhibition. Compound 8 showed no activity as a checkpoint inhibitor but inhibited nuclear export (Figure 3.14D). However, this may be because limited compound availability prevented checkpoint abrogation measurement at 30 ug/mL, at which nuclear export inhibition was observed. Overall, there was a fairly consistent correlation between checkpoint inhibition and export inhibition, although strong checkpoint inhibitors were not necessarily strong nuclear export inhibitors. This is summarized in Table 3.2. 73 Figure 3.14 : Inhibition of NES-mediated nuclear export by cryptofolione and related compounds. A, untreated cells. B-H, the cells were treated with 1 uM dexamethasone for 60 min to induce nuclear accumulation of Rev/GR/GFP, then the dexamethasone was washed away and cells were further incubated in the absence of dexamethasone for 0 min (B) or 60 min (C-H) in the presence of leptomycin B (Q, 2 (D), 6 (E), 7 (F), 1 (G), or DMSO (H). The cells were then fixed and Rev/GR/GFP was visualized by fluorescence microscopy, as described in Chapter 2.8.2. The oval shaped structures are individual nuclei and the nucleoli are the bright dots within the nuclei. DNA-blue, Rev/GR/GFP-green. Results shown are representative of 3 fields of view from duplicate experiments. Bar = 10 um 74 Table 3. 2 : Summary of activity of cryptofolione analogs. Relative assessment of checkpoint inhibitor activity and export inhibition is given, with (+++) indicating very strong inhibition, (+) minor inhibition, and (-) indicating no apparent inhibition. Numbers in brackets represent the minimal concentration (ug/mL) used to achieve inhibition. Compound G2 checkpoint Nuclear export inhibition inhibition 1 + (30) + (10) 2 +++(10) + (10) 3 - nd 4 - -5 +++(1) + (10) 6 ++(1) + (10) 7 + (10) + (10) 8 Q -(10*) + (30) y 10 _ _ Leptomycin B ++ + Nd, not determined * highest concentration tested 75 3.2.2.5 Potent G 2 checkpoint inhibition by leptomycin B The observations described above prompted me to ask whether leptomycin B, the most potent and selective export inhibitor described to date (Kudo et al, 1998; Kudo et al, 1999), can also abrogate the G 2 checkpoint. Indeed, leptomycin B showed very potent inhibition of G 2 arrest induced by ionizing radiation, with significant activity at concentrations as low as 0.05 nM (Figure 3.15). 3.3 DISCUSSION Psilostachyin A and C represent a novel class of G 2 checkpoint inhibitors. They resemble the ent-kaurene diterpenoid OZ in their ability to overcome the checkpoint and to also block cells in mitosis. Moreover, an a,(3 unsaturated carbonyl group is also required for the activity of OZ. However, psilostachyin A and C are much more efficacious checkpoint inhibitors than OZ (Rundle et al, 2001). The psilostachyins appear to be more selective checkpoint inhibitors than OZ because the latter inhibits cell proliferation at lower concentrations (IC5o= 0.8 uM) than it inhibits the G 2 checkpoint (IC5o = 6uM) (Rundle et al, 2001), whereas psilostachyin A and C inhibit the checkpoint and proliferation at roughly the same concentrations. In addition, psilostachyin A and C have effects on mitotic microtubules that are different from those of OZ, in that the microtubule disruption appears more severe. Despite enormous advances in the last decade, the understanding of the mechanism of the G 2 checkpoint is still incomplete. In vitro, psilostachyin A and C do 76 Figure 3.15 : G 2 checkpoint inhibition by leptomycin B . MCF-7 mp53 cells were irradiated with 10 Gy, cultured for 16 h, then incubated with the indicated concentration of leptomycin B and 100 ng/mL nocodazole for 8 h. G 2 checkpoint inhibition was assessed by flow cytometry as described in Chapter 2.5.2. Results shown are means of duplicate experiments. 77 not inhibit the checkpoint kinases A T M , ATR, Chkl and Chk2 (not shown), nor are they Ser/Thr phosphatase inhibitors, which have been shown to overcome G2 arrest and induce mitosis (Britton et al, 2003). In addition, the psilostachyins inhibit the checkpoint additively with Chkl inhibitors, indicating that they target a different pathway. The psilostachyins' requirement for an a,P unsaturated carbonyl group for checkpoint inhibition indicates that they bind covalently to their target protein(s). If a stable biotinylated compound can be synthesized, this property may facilitate the identification of their cellular target(s), which may be new proteins of this signalling pathway, if they are not Weel. A distinct checkpoint, called the decatenation checkpoint, is activated by agents such as ICRF-193 that inhibit D N A topoisomerase II without inducing D N A damage and induces G2 arrest (Deming et al, 2001). Leptomycin B has been shown to inhibit the decatenation checkpoint (Haggarty et al, 2003). However, in HeLa cells, leptomycin B was not observed to overcome G2 arrest induced by etoposide, a topoisomerase II inhibitor that induces D N A damage (Toyoshima et al, 1998). Despite this, I have demonstrated that cryptofoliones and leptomycin B do indeed abrogate G 2 arrest induced by ionizing radiation. This suggests that nuclear export is required to prevent premature mitosis in checkpoint-arrested cells. A number of proteins involved in control of the G 2 / M transition shuttle in and out of the nucleus via a Crml -dependent mechanism (Hagting et al, 1998; Toyoshima et al, 1998; Dalai et al, 1999,Graves et al, 2001). Human cyclin BI shuttles between the nucleus and the cytoplasm during interphase and CDKl/cycl in B accumulates in the nucleus at the G 2 / M transition (Hagting et al, 1998). This nuclear accumulation is thought to be important for initiating and coordinating 78 mitotic events (Takizawa and Morgan, 2000). Cdc25B and Cdc25C are also exported from the nucleus via Crml and leptomycin B causes the nuclear accumulation of cyclin BI , Cdc25B and Cdc25C in interphase cells (Hagting et al, 1998; Davezac et ai, 2000; Graves et al., 2001). However, the nuclear translocation of these proteins by leptomycin B treatment is not sufficient to induce premature entry into mitosis in cells that have not been exposed to D N A damage (Toyoshima et al, 1998). Therefore it is unclear whether it is the nuclear translocation of cyclin B I , Cdc25B or Cdc25C or that of other proteins, that is responsible for G2 checkpoint abrogation. G2 checkpoint abrogation in combination with radiotherapy or DNA-damaging chemotherapy can selectively enhance the killing of cancer cells with mutated p53. Does abrogation by nuclear export inhibitors have therapeutic potential for cancer therapy? On the one hand, a large number of proteins and complexes such as the ribosomal subunits are subjected to Crml-dependent export (Thomas and Kutay, 2003; Yashiroda and Yoshida, 2003), indicating that inhibitors of Crm-1-mediated nuclear export might be toxic to both normal and cancer cells. On the other hand, leptomycin B has potent activity in vitro against a range of tumours (Roberts et al, 1986) and it has undergone a Phase I clinical trial under the name elactocin (Newlands et al, 1996). However, elactocin produced serious dose-limiting toxicity and further clinical trials were not recommended (Newlands et al, 1996). These results indicate that Crm-1 is not a viable drug target for cancer therapy. However, it is possible that some of the toxic effects of leptomycin B arise from its chemical reactivity. Leptomycin B binds covalently to cysteine 529 in Crml through Michael addition via its a,(3-unsaturated lactone (reviewed in (Yashiroda and Yoshida, 2003)). Compounds with Michael acceptor functionalities are generally 79 considered undesirable drug candidates. Thus, it is possible that reversible Crm-1 inhibitors might show fewer toxic side effects and greater therapeutic potential. Cryptofolione also possesses an a,(3-unsaturated lactone and is reactive towards sulfhydryls. Unfortunately, disruption of this functional group via reaction with (3-mercaptoethanol produced an unstable compound (Ana Diaz-Marrero, personal communication). Therefore it is unknown if this functional group is necessary in the same manner as that for psilostachyin. Unfortunately, all cryptofoliones are cytotoxic at concentrations that abrogate the checkpoint and do not appear promising as a checkpoint inhibitor for therapeutic intervention. Nevertheless, the demonstration that nuclear export inhibitors can abrogate the G2 checkpoint induced by ionizing radiation should stimulate the search for compounds that display a similar phenotype but with markedly less toxicity. 3.4 SUMMARY A l l G2 checkpoint inhibitors identified to date target protein phosphorylation by inhibiting checkpoint kinases or phosphatases. Of those targeting kinases, they target members of the A T M / A T R - C h k l / 2 pathway. In an effort to identify novel inhibitors targeting both this pathway and possibly others, a phenotypic cell-based assay was used to screen plant extracts from the US National Cancer Institute Natural Products Repository for checkpoint abrogating activity. This screening, performed by many members of the Roberge laboratory, revealed activity in an extract from the common ragweed Ambrosia artemisiifolia, as well as from the leaf of the subtropical plant, Cryptocarya concinna Assay-guided fractionation led to the identification of the 80 sesquiterpene lactones psilostachyin A and C and the cryptofoliones as novel checkpoint inhibitors. In the case of the psilostachyins, elimination of their a, (3-unsaturated carbonyl group caused loss of activity, indicating that the compounds can bind covalently to target proteins through Michael addition. Psilostachyin A and C also blocked cells in mitosis and caused the formation of aberrant microtubule spindles. However, the compounds did not interfere with microtubule polymerization in vitro. The related sesquiterpene lactones psilostachyin B, paulitin and isopaulitin were also isolated from the same extract but showed no checkpoint inhibition. Attempts to relate psilostachyin to a similar checkpoint inhibitor, OZ (Rundle et al., 2001), led to the possible identification of Weel as a target of OZ/psilostachyin, by using yeast drug-induced haploinsufficiency as a screening tool. Confirmation via in vitro assay awaits re-isolation of more psilostachyin. The identification of the target(s) of psilostachyin A and C may provide further insight into the signalling pathways involved in cell cycle arrest and mitotic progression. The work presented in this thesis, from Figures 3.1 to 3.6 has been published in (Sturgeon et al, 2005), while Figures 3.7 to 3.11 and Table 3.1 represent subsequent attempts to elucidate the target of psilostachyin before being hindered by compound availability. In the case of the cryptofoliones, it and related compounds are inhibitors of nuclear export. Leptomycin B, a potent inhibitor of Crml-mediated nuclear export, is also a very potent G2 checkpoint inhibitor. Cryptofolione does not appear promising as a radiosensitizing agent because it was toxic to unirradiated cells at checkpoint inhibitory concentrations. Nevertheless, the results show that inhibition of nuclear export is an alternative to checkpoint kinase inhibition for abrogating the G2 checkpoint and they 81 should stimulate the search for less toxic nuclear export inhibitors. The work presented in this thesis, from Figures 3.12 to 3.15, has been submitted for publication (Sturgeon et al, 2006a). 82 Chapter 4 Combining DNA repair inhibition and checkpoint inhibition to increase 2 therapeutic effect 4.1 I N T R O D U C T I O N D N A damage can elicit a variety of cellular responses (Zhou and Elledge, 2000). Some such as apoptosis, mitotic catastrophe and premature senescence remove damaged cells from proliferating populations. Others such as D N A repair and cell cycle checkpoints, which delay cell cycle progression and allow more time for D N A repair, promote cell survival and subsequent proliferation. Considerable efforts are being invested into identifying chemicals that modulate these responses and investigating whether they have potential as cancer therapy agents. In particular, it has been proposed that drugs that inhibit D N A repair or cell cycle checkpoints might enhance the efficacy of current DNA-damaging cancer treatments, such as ionizing radiation and many chemotherapeutic drugs. As discussed in Section 1.2-1.3, according to current models, D N A damage activates the PIKK family members DNA-PK, A T M and ATR. A T M and ATR phosphorylate and activate the checkpoint kinases Chkl and Chk2, which phosphorylate a variety of downstream targets resulting in cell cycle delay at the G] and G2 phases of the cell cycle and slow down of S-phase traversal. Repair of DSB, the most lethal D N A 2 A section o f this chapter has been published (Sturgeon, C . M . , Z . A . Knight , K . M . Shokat and M . Roberge (2006b). Effect o f combined D N A repair inhibition and G 2 checkpoint inhibition on cell cycle progression after D N A damage. Molecular Cancer Therapeutics 5(4): 885-892). 83 lesions caused by IR, is accomplished through two major pathways, NHEJ and HRR (Sancar et al, 2004). NHEJ is a Gi/S-predominant process controlled by D N A - P K (a complex of the catalytic subunit DNA-PKcs, Ku70/86), D N A Ligase IV, Artemis and XRCC4 (Lieber et al, 2003). HRR is a late S/G2-predominant process that requires Rad52, Rad54, and various Rad51 paralogues (Takata et al, 1998; Thompson and Schild, 2001). Over 50% of human cancers have a mutated form of p53, resulting in an inoperative Gi checkpoint (Kastan et al, 1991) but a functional G 2 checkpoint. Some inhibitors of the G 2 checkpoint can selectively sensitize p53-defective cells to DNA-damaging agents, likely as a result of undergoing mitosis in the presence of damaged D N A (reviewed in (Anderson et al, 2003)). DNA-PK-defective cells show increased sensitivity to ionizing radiation and it was shown that D N A - P K inhibitors can radiosensitize tumours in vitro and in vivo (Kashishian et al, 2003; Shinohara et al, 2005). I reasoned that cells treated with ionizing radiation and a D N A - P K inhibitor would arrest in G 2 phase with more unrepaired D N A than cells exposed to irradiation alone. Exposure of these G2-arrested cells to checkpoint inhibitors would then force them to divide in the presence of more heavily damaged DNA, which should translate into increased killing. However, in the absence of NHEJ repair, D N A damage caused a stronger activation of checkpoint kinases and a more pronounced and prolonged cell cycle arrest. G 2 checkpoint inhibitors were less able to overcome this prolonged arrest and unable to increase cell killing. These observations are relevant to efforts to develop checkpoint inhibitors and D N A - P K inhibitors as tumour-selective radiosensitizers for 84 cancer therapy. I then attempted to delve deeper into the mechanism behind the strengthened G2 arrest through the use of a Kinetworks screen with Kinexus Bioinformatics Corp. Although this lead to the identification of M K K 6 being overexpressed in DNA-PKcs deficient cells, which can cause G2 arrest (Wang et al, 2000), M K K 6 did not appear to play a role in the prolonged G2 arrest in M059J cells. 4.2 RESULTS AND DISCUSSION 4.2.1 DNA-PK inhibition and NHEJ deficiency decrease the ability of cancer cells to respond to G2 checkpoint inhibitors Caffeine, an inhibitor of A T M and ATR (Sarkaria et al, 1999; Zhou et al, 2000), and UCN-01, isogranulatimide and debromohymenialdesine, inhibitors of Chkl can abrogate the G 2 checkpoint and potentiate the lethality of IR (Wang et al, 1996; Roberge et al, 1998; Zhou et al, 2000; Curman et al, 2001; Jiang et al, 2004). D N A - P K inhibitors can radiosensitize tumour cells by impairing D N A repair via the NHEJ system (Allen et al, 2003; Kashishian et al, 2003; Shinohara et al, 2005). The arylmorpholine A M A3 7 is a potent in vitro inhibitor of D N A - P K (IC50 = 0.27 uM) that does not inhibit A T M or ATR and inhibits PI3-K poorly (Knight et al, 2004). I wished to determine whether checkpoint inhibitors are able to overcome G2 arrest in the presence of a DNA-PK inhibitor. The interplay between D N A - P K inhibition and G2 checkpoint inhibition was first examined in human breast carcinoma MCF-7 mp53 cells. These cells are NHEJ-proficient, have an inactive Gi checkpoint (Mercer, 1992; Fan et al, 1995) and have been used extensively to study the G 2 checkpoint (Fan et al, 1995; Roberge et al, 1998; 85 Curman et al, 2001; Rundle et al, 2001). To measure G2 checkpoint inhibition, MCF-7 mp53 cells were exposed to AMA37 or the drug carrier DMSO for 30 min and then irradiated with 6.5 Gy. At 16 h post-IR, when G2 arrest was maximal, cells were treated with checkpoint inhibitors and nocodazole as described in Section 2.4. The cells were then analysed by 2-dimensional flow cytometry using the mitosis-specific antibody GF-7 to distinguish M cells from G2 cells (Rundle et al, 2001) (see Figure 4.1 for examples). Quantitation of G 2 checkpoint inhibition after various treatments is presented in Figure 4.2. Caffeine, UCN-01, isogranulatimide and debromohymenialdesine caused clear G2 checkpoint inhibition. AMA37 reduced the ability of UCN-01, isogranulatimide and debromohymenialdesine, but not caffeine, to overcome G2 arrest (p < 0.05). By contrast, 200 nM wortmannin, which inhibits PI3-K but not A T M or D N A - P K (Sarkaria et al, 1998), did not reduce the activity of the checkpoint inhibitors (Figure 4.2). As a test of the selectivity of AMA37 in vivo, I monitored its effect on phosphorylation of H 2 A X at Ser 139 following IR (y-H2AX), which is primarily phosphorylated by A T M and redundantly by D N A - P K (Stiffs al, 2004). Therefore D N A - P K inhibitors should have minimal effect on y - H 2 A X formation, while chemicals that inhibit both D N A - P K and A T M should strongly inhibit Y -H2AX formation. Exposure to IR caused a strong increase in y - H 2 A X (see Figure 4 .3A). A M A 3 7, LY294002 and 200 nM wortmannin had little affect, but was strongly reduced by 20 uM wortmannin, at which concentration A T M and D N A - P K are inhibited. These results are consistent with a lack of A T M inhibition by AMA37. Furthermore, unlike siRNA knockdown of DNA-PKcs (Peng et al, 2005), long-term AMA37 treatment of cells had little effect on expression of A T M in MCF-7 mp53 cells (Figure 4.3B). Therefore, the effect of AMA37 86 M059K M059J M059J/Fus1 DMSO CAF IGR Figure 4.1 : Representative 2-dimensional flow cytometry analyses used to monitor checkpoint inhibitor efficacy. Cells were exposed to checkpoint inhibitors in the presence of nocodazole between 16 and 24 h after irradiation as described in Chapter 2.4. DNA was stained with PI (x axis), and mitotic cells by FITC-GF-7 immunofluorescence labeling (y axis). The number of events associated with a coordinate on the plot is indicated by the color scale (blue indicating low density and red indicating high density). CAF - caffeine treatment; IGR - isogranulatimide treatment. Plots shown are representative of at least 3 replicates for each condition. 8 7 70 -I ^ 6 0 -U 50 -I s 40 ^ 3 0 20 -I 10 0 D M S O A M A 3 7 7777Zi Wor tmann in O 00 o o > T l i c o z I o 1 t D CD I ;ure 4 .2 : Effect of G 2 checkpoint inhibitors on human cells lacking DNA-PK or NHEJ activity. MCF-7 mp53 cells were exposed to DMSO (black bars), 20 uM AMA37 (white bars) or 200 nM wortmannin (crosshatched bars) 30 min before irradiation with 6.5 Gy. 16 h later, when arrested in G 2 , cells were exposed to checkpoint inhibitors plus 100 ng/mL nocodazole for 8 h, then harvested and analyzed by flow cytometry as described in the Chapter 2.5.2. GF-7 positive cells represent those trapped in mitosis. UCN-01, IGR and DBH were less able to overcome G 2 arrest in the presence of AMA37 (p < 0.05). Data shown are the means +/- SD (n > 3). 88 Y-H2A.X - mmm 4H»m*- -mm, . , „ » -• wmw Histones -:•• • ^•p mmn -^m^" wsg^ B 1 2 ATM Ku86 Tubulin Figure 4.3 : A M A 3 7 has little to no effect on A T M function or expression. A, MCF-7 mp53 cells were either mock-irradiated (No IR) or exposed to 6.5 Gy 30 min after addition of either DMSO, 20 uM AMA37, 20 uM wortmannin, 200 nM wortmannin or 50 uM LY294002. 30 min after IR cells were harvested by trypsinization and analyzed by western blotting with a y-H2AX. Exposure to IR caused a strong increase in y-H2AX. AMA37, LY294002 and 200 nM wortmannin had little effect, but 20 uM wortmannin (which inhibits all PIKK at this concentration) strongly reduced y-H2AX. B, MCF7 mp53 were exposed to DMSO (lane 1) or AMA37 (lane 2). After 16 h ATM, Ku86 and p-tubulin were detected by western blotting. In the AMA37 incubation times used here, the effect on A T M expression is minimal. 89 on y-H2AX phosphorylation was consistent with selective inhibition of D N A - P K in vivo, indicating that the effects of AMA37 were likely due to D N A - P K inhibition and not to modulation of other PI3-K or A T M involved in the checkpoint response. The experiment described in Figure 4.2 was prompted by the premise that G 2 checkpoint inhibitors can induce cells to enter mitosis in the presence of D N A damage and that co-treatment with a D N A - P K inhibitor to reduce D N A repair would cause cells to enter mitosis with even more damaged DNA. It was surprising that cells treated with AMA37 showed a reduced response to G 2 checkpoint inhibitors. This raised the possibility that AMA37 could be antagonizing mitotic entry after D N A damage through inhibition of secondary targets. To further investigate whether the effect was due to reduced D N A - P K activity, I examined the response of human cells lacking the DNA-PKcs to checkpoint inhibitors. Three related cell lines obtained from a human malignant glioma were utilized (Allalunis-Turner et ai, 1993; Hoppe et al, 2000). DNA-PK-deficient M059J cells fail to express the DNA-PKcs while DNA-PK-proficient M 0 5 9 K cells do express it (Allalunis-Turner et al, 1993; Lees-Miller et al, 1995). However, M059J cells express lower levels of A T M than M 0 5 9 K (Chan et al, 1998; Gately et al, 1998; Tsuchida et al, 2002). M059J7Fusl cells, rescued in their expression of DNA-PKcs (Hoppe et al, 2000), have slightly higher A T M expression levels than M059J and are the closest available genetic match to M059J.(Hoppe et al, 2000). The three cell lines are p53-deficient (Mercer, 1992; Allalunis-Turner et al, 1997; Anderson and Allalunis-Turner, 2000). As shown in Figure 4.4A, caffeine, UCN-01 and isogranulatimide caused clear G 2 checkpoint inhibition in M 0 5 9 K cells. However, M059J cells were less responsive to all checkpoint inhibitors. Furthermore, M059J/Fusl 90 cells were more responsive to all inhibitors than M059J cells (Figure 4.4A). Interestingly, at a higher dose of 10 Gy, M059K cells responded similarly to 6.5 Gy, whereas both M059J and M059J/Fusl failed to respond to checkpoint inhibitors (Figure 4.4B). No cell death was observed for any of the cell lines or treatments during the course of these experiments. These data indicated that D N A - P K activity is required for cells to resume cell cycle progression after G2 arrest, or that the reduced activity of checkpoint inhibitors is a consequence of reduced D N A repair via NHEJ in cells lacking D N A - P K activity. To address this issue, I monitored checkpoint inhibition in an additional six related CHO cell lines. CHO K I cells are considered wild-type. CHO XR-1 cells do not express XRCC4, making them defective for NHEJ repair (Li et al, 1995). CHO V3 cells are deficient in DNA-PKcs (Blunt et al, 1995). CHO V3 cells stably transfected with either empty vector (CHO V3) or human DNA-PKcs cDNA (CHO V3 hDNA-PKcs) differ by the presence or absence of DNA-PKcs (Ding et al, 2003). CHO 51D1 lack both alleles expressing Rad51D, while 51D1.3 are a Rad51D reconstituted isogenic match. DNA-PKcs-deficient V3 cells were less responsive to checkpoint inhibitors than their DNA-PK-proficient counterparts (Figure 4.5). Interestingly, XRCC4-deficient XR-1 cells were not responsive at all to checkpoint inhibitors while K I cells with functional NHEJ were. CHO 51D1 cells, deficient in Rad51D, were less responsive to UCN-01 and isogranulatimide, but not caffeine, than their Rad5 ID-expressing counterparts, 51D1.3. These data provide strong evidence that it is a lack of D N A repair that reduces the capacity of checkpoint inhibitors to abrogate G2 arrest. Unfortunately though, complicating this interpretation is the 91 70 "33 o * "55 o CL LL CD B J/5 "53 o > "55 o CL h~ t LL. o —m M059K ( = 3 M059J M059J/Fus1 P I 1 i l l 73 O I I P 3D o DO gure 4.4 : Effect of G 2 checkpoint inhibitors on human cells lacking DNA-PK or NHEJ activity. 16 h post-IR, when arrested in G 2 , M059 cells were exposed to checkpoint inhibitors plus 300 ng/mL nocodazole for 8 h, then harvested as described in the Chapter 2.5.2. GF-7 positive cells represent those trapped in mitosis. A, M059 cells were exposed to 6.5 Gy. B, 10 Gy. M059J cells are less responsive to all checkpoint inhibitors (p < 0.001), while M059J/Fusl are less responsive when exposed to 10 Gy (p = 0.0001). Results shown are means +/- SD (n > 3). 92 [ = 1 XR-1 ezzza V3 Vector ™ V3 hDNAPKcs KSWSI 51D1.3 51D1 Figure 4.5 : Effect of G 2 checkpoint inhibitors on CHO cells lacking DNA-PK or NHEJ activity. Various CHO cells were irradiated with 6.5 Gy. 8 h post-IR, checkpoint inhibitors and 100 ng/mL nocodazole were added for 4 h, and assayed by flow cytometry as described in Chapter 2.5.2. V3 cells respond less to checkpoint inhibitors than V3 cells with hDNA-PKcs cDNA. 51D1.3 cells respond less to UCN-01 and IGR compared to 51D1 cells. Results shown are means of duplicate experiments. 93 observation that D N A repair is complete in M059J cells 24 hours after 10 Gy (DiBiase et al, 2000). Perhaps the repair defect in XR-1 cells is more dramatic than that in V3 (and M059J) cells, as there are documented DNA-PK-independent NHEJ pathways (DiBiase et al, 2000; Williams et al, 2001; Wang et al, 2001b; Wang et al, 2003a; Iliakis et al, 2004; Perrault et al, 2004), which may be evidenced by the more severe G2 arrest in XR-1 cells (Figure 4.5). 4.2.2 DNA-PK inhibition leads to reduced radiosensitization by checkpoint inhibitors To determine whether the decreased ability of checkpoint inhibitors to abrogate G2 arrest translates into reduced radiosensitization, I assessed the clonogenicity of MCF-7 mp53 cells following combined treatment with AMA37 and checkpoint inhibitors. Table 4.1 lists the effect of different drug combinations on clonogenicity. Exposure of cells to 6.5 Gy IR alone lead to a decrease in survival to 15.4 ± 2.0 % of unirradiated controls. Irradiation in the presence of caffeine, UCN-01 or isogranulatimide caused a further decrease in clonogenicity, indicating radiosensitization. Inhibition of D N A - P K with AMA37 also lead to radiosensitization. The checkpoint kinase inhibitors UCN-01 and isogranulatimide were unable to increase radiosensitization in the presence of AMA37, consistent with their inability to overcome G2 arrest. Caffeine, the only compound that retained checkpoint inhibitory activity in the presence of AMA37 (Figure 4.2), was also able to radiosensitize in the presence of AMA37. Caffeine inhibits A T M and ATR and acts upstream of the Chk kinase inhibitors UCN-01 and isogranulatimide. Caffeine is also a less selective checkpoint inhibitor than UCN-01 and isogranulatimide. It inhibits 94 T a b l e 4.1 : E f f e c t o f A M A 3 7 a n d c h e c k p o i n t i n h i b i t o r s o n the c l o n o g e n i c i t y o f i r r a d i a t e d a n d u n i r r a d i a t e d M C F - 7 cells. Cells were treated with drug combinations at the time of irradiation and the % colonies present 10 days post-IR was determined relative to untreated controls. Results are means +/- SD (n = 3) T r e a t m e n t D M S O C a f f e i n e U C N - 0 1 I s o g r a n u l a t i m i d e D M S O M o c k 100 107 95.3 94.4 6.5 G y 15.4 ± 2 . 0 % 4.2 ± 0.7 % 4.5 ± 1.1 % 2.9 ± 1.2 % A M A3 7 M o c k 100 112.1 77.6 84.1 6.5 G y 4.4 ± 1.6% 1.0 ± 1.3 % 5.5 ± 0.7 % 3.6 ± 0 . 2 % 95 several other cellular activities, including HRR D N A repair itself (reviewed in (Anderson et al, 2003)). I speculate that caffeine's radiosensitizing effect in the presence of AMA37 may be due to its ability to downregulate more branches of the checkpoint response than UCN-01 and isogranulatimide and/or its inhibition of both forms of D N A repair. 4.2.3 DNA-PK- and NHEJ-impaired cells enter prolonged G 2 arrest following DNA damage G 2 checkpoint inhibitors can force entry into mitosis in the presence of D N A damage (Wang et al, 1996; Roberge et al, 1998; Zhou et al, 2000; Curman et al, 2001). To investigate why cells with reduced D N A repair were less responsive to checkpoint inhibitors, I used flow cytometry to monitor G 2 arrest over time in the cell lines examined above. As expected, MCF-7 mp53 cells exposed to JR underwent transient G 2 arrest (Figure 4.7 - see Figure 4.6 for example flow profiles). However, treatment with AMA37 and IR resulted in a higher proportion of cells arresting in G 2 and their failure to exit spontaneously from G 2 arrest (Figure 4.7). Cells treated with 200 nM wortmannin and IR underwent G 2 arrest and subsequently exited G 2 arrest similarly to untreated cells (Figure 4.7), indicating that AMA37 does not affect the G 2 checkpoint via PI3-kinase inhibition. LY294002, which inhibits PI3-kinases and D N A - P K but not A T M or A T R in vivo (Stiff et al, 2004), also lead to prolonged G 2 arrest after IR. These results indicate that inhibition of D N A - P K causes a prolonged G 2 arrest. Consistent with NHEJ's role in D N A repair in G\/S, adding AMA37 to G2-arrested cells (16 h post-IR) did not induce prolonged G 2 arrest (Figure 4.7), indicating that AMA37 must exert its effects before G 2 96 200 nM DMSO A M A3 7 Wort O h | . G,/M ! DNA 40 8h " i_. LL'Li 16 h 24 h 32 h I LlU Uu h 'Iii ,')u .'lli , 48 h i M059K M059J M059J/ Fusl Oh h I 16 h 24 h I 32 h i 40 h A. L L L L i i 48 h Figure 4.6 : Sample histograms showing DNA profiles for all cell lines used. A, MCF-7 or B, M059 cells were exposed to 6.5 Gy at time 0 h, sampled at the indicated times and analysed by flow cytometry as described in Chapter 2.5.2. C, CHO lines at 2 or 6.5 Gy (next page). Histograms shown are representative of at least 3 experiments. 97 Gy 2 6.5 2 6.5 2 6.5 2 6.5 2 6.5 2 6.5 10 -i 1 1 1 1 . 1 • 0 8 16 24 32 40 48 56 Time (h) Post-6.5 Gy gure 4.7 : E f f e c t o f IR on G 2 a r r e s t i n h u m a n cells l a c k i n g D N A - P K o r N H E J ac t i v i t y . MCF-7 mp53 cells were irradiated with 6.5 Gy at time 0 h, sampled at the indicated times and the G 2 / M populations were quantified using flow cytometry as described in Chapter 2.5.2. MCF-7 mp53 cells were exposed to DMSO (filled circles), 20 uM A M A3 7 (open circles), 200 nM wortmannin (filled triangles) or 50 uM LY294002 (open triangles) 30 min before irradiation with 6.5 Gy. Another cell population was irradiated with 6.5 Gy and AMA37 was added 16 h later (open squares). Cells were harvested at the indicated times and analysed by flow cytometry. The graph shows quantitation of the G 2 / M populations over time. Results shown are means +/- SD (n > 3). 99 arrest becomes established, and as will be addressed in Section 4.4.4, that D N A - P K activity is not required for cells to resume cell cycle progression after G2 arrest. In order to ensure that the arrest was in G 2 and not late S-phase, I used flow cytometry to monitor the presence of the G 2 marker cyclin B I . Cyclin BI accumulated in cells with An D N A content and its levels remained elevated in AMA37-treated MCF-7 cells during prolonged arrest (see example profiles in Figure 4.8). Similar results were obtained with western blotting (data not shown). This result indicates that the cells were indeed arrested in G 2 phase, not in late S-phase or failed cytokinesis. The observation that cyclin BI levels are stable during prolonged G 2 arrest also indicates that the checkpoint inhibitors did not fail to overcome G 2 arrest because of downregulation/degradation of cyclin B I . Quantitation of the G 2 / M peaks as a function of time after irradiation of M059 cells is shown in Figure 4.9. Exposure to 6.5 Gy caused a large increase in the G 2 / M peak that was maximal at 16 h for all three cell lines (Figure 4.9). Most of the M 0 5 9 K cells escaped G 2 arrest between 16 and 24 h and cycled thereafter as a loosely synchronous population. Similar to AMA37 treatment, M059J cells remained arrested in G 2 for at least 48 h post-IR. Cell populations with sub-Gi D N A content became increasingly evident at time points longer than 48 h, indicating the onset of cell death (not shown). Cyclin BI also accumulated in G 2 - arrested M059J cells (Figure 4.8). M059J/Fusl cells began to exit G 2 arrest 16 h after exposure 6.5 Gy, much like M 0 5 9 K cells (Figure 4.9C). Additional IR doses were tested in these cell lines: a lower dose of 2 Gy caused a significant increase in the G 2 / M peak only in M059J cells and remained for upwards of 100 A DMSO AMA37 O h 24 h 48 h B M059K M059J M059J/ Fusl 0 h ' i L ; I. 1 I IL. ; IL 1 24 h i | ; L L ! J l lL 1 >* 1. 1 48 h i l l —*—V. T*~ • i, j. > — M ure 4.8 : Cyclin BI accumulation in G2-arrested cells. A , MCF-7 mp53 or B, M059 cells were irradiated with 6.5 Gy, then cyclin BI levels were detected using flow cytometry. Cyclin B1 is elevated and not degraded in the G 2 arrested cells with impaired DNA-PK activity. Results shown are representative of duplicate experiments. 101 c M059J/Fus1 100 i Time (h) Post-IR Figure 4.9 : Effect of IR on G 2 arrest in M059 cells lacking DNA-PK or NHEJ activity. M059 cells were irradiated with the 2 Gy (filled circles), 6.5 Gy (open circles), 8.3 Gy (filled triangles) or 10 Gy (open triangles) at time 0 h, sampled at the indicated times and the G 2 /M populations were quantitated using flow cytometry. A, M059K B, M059J, or C , M059J7Fusl cells. M059J cells exhibit permanent G 2 arrest after all IR doses, whereas M059J/Fusl exhibit permanent arrest after 10 Gy only. Results shown are means +/- SD (n >3). 102 48 h (Figure 4.9B). A higher dose of 10 Gy caused a large increase in G 2 / M that peaked at 16-24 h in all cell lines, with M 0 5 9 K and M059J/Fusl cells recovering from this arrest (Figure 4.9). However, M059J/Fusl cells remained arrested longer and fewer cells finally exited G 2 arrest at this dose (p < 0.0001). Interestingly, M059J/Fusl cells have about 60% of the D N A - P K activity of M 0 5 9 K (Hoppe et al, 2000) and they exited G 2 arrest significantly later and to a lesser extent than M059K after exposure.to 10 Gy but not at lower doses, indicating a relationship between the level of cellular D N A - P K activity and the ability of cells to escape G 2 arrest. At 10 Gy, M059J/Fusl cells were equally unresponsive to checkpoint inhibitors as M059J (Figure 4.4B), whereas M 0 5 9 K cells remained equally responsive. The different CHO cell lines were subjected to irradiation with 2 Gy and G 2 arrest was monitored over time. IR elicited a transient G 2 arrest in K I cells and V3 hDNA-PKcs cells, both of which have DNA-PKcs and NHEJ activity (Figure 4.10). However, DNAPKcs-defective V3 cells and NHEJ-defective XR-1 cells showed strong sustained G 2 arrest (Figure 4.10). Exit from G 2 arrest started after 8 h in K I and V3 hDNA-PKcs cells, while V3 and XR-1 cells remained arrested in G 2 for at least 48h. For unclear reasons, G 2 arrest was also accompanied by an increase in >8« population in XR-1 and V3 cells, and a large increase in apoptosis was observed at the higher dose of 6.5 Gy (see Figure 4.6). The G 2 arrest profile of 51D1 and 51D1.3 cells irradiated with 2 Gy was similar to that of wild-type K I cells (Figure 4.10), but 6.5 Gy caused a slightly prolonged G 2 arrest in 51D1 cells (see Figure 4.6). 103 g JCO =5 Q. O 0-- O - XR-1 - • - V3 hDNA-PKcs -V- V3 Vect 51D1.3 0 8 16 24 32 40 48 56 Time (h) Post-2 Gy gure 4.10 : Effect of IR on G 2 arrest in CHO cells lacking DNA-PK or N H E J activity. Various CHO cells were irradiated with 2 Gy at time 0 h, sampled at the indicated times and the G 2 / M populations were quantitated using flow cytometry. Only NHEJ-deficient CHO cells (V3 empty vector and XR-1 cells) exhibit prolonged G 2 arrest. Results shown are means +/- SD in > 3). 104 4.2.4 DNA-PK is not required in G 2 for normal G 2 checkpoint recovery The repair of D N A DSBs caused by IR in Gi phase is carried out mostly by the NHEJ system, while in G 2 it takes place mostly via homologous recombination repair (HRR) (Valerie and Povirk, 2003). NHEJ requires D N A - P K activity while HRR does not (Sancar et al., 2004). When asynchronously dividing cell populations are irradiated, most cells are in Gi and only a relatively small proportion are in S/G 2 (Figure 4.6, top panels). I asked whether irradiation during late-S/G2 also leads to prolonged G 2 arrest in DNA-PK-deficient cells. I synchronized M059K, M059J and M059J/Fusl cells to the G 2 phase by release from aphidicolin arrest as described in Section 2.13. Six hours after release, when 70-80% were in G 2 , cells were irradiated with 6.5 Gy, and G 2 accumulation was monitored over time. M059K, M059J and M059J/Fusl cells all underwent G 2 delay, and exited G 2 at a similar rate of about 1.5% per hour (Figure 4.11) while non-irradiated cells progressed through G 2 at a rate of about 7 % per hour (dashed line, Figure 4.11). This result indicates that D N A - P K (and by extension NHEJ) is not required for cells to recover from G 2 following D N A damage during the G 2 phase. Together with Figures 4.6 - 4.11, these results show that lack of repair at the time of D N A damage causes prolonged G 2 arrest that cannot be abrogated by checkpoint inhibitors, likely caused by the persistence of D N A lesions. 4.2.5 IR causes overactivation of Chkl and Chk2 in cells lacking DNA-PK activity I next wished to examine the mechanism underlying prolonged G 2 arrest in cells lacking D N A repair. The G 2 checkpoint is generally believed to comprise two convergent 105 5? 30 -20 -10 H 1 1 1 1 0 8 16 24 32 Time (h) Post-6.5 Gy Figure 4.11 : Effect of IR on G 2 synchronized M059 cells. Cells were irradiated at time 0 h, six hours after release from aphidicolin, resulting in 80-90% of cells in G 2 , and then sampled at the indicated times and the G , / M populations were quantified using flow cytometry. Cells were synchronized as described in Chapter 2.13. Non-irradiated cells exited G 2 at a rate of 7% per hour, while irradiated cells exited on average at 1.5% per hour. There was no significant G 2 arrest in M059J cells compared to M 0 5 9 K or M059J/Fusl 24 h post-IR. Results shown are the means +/- SD (ti > 3). 106 pathways: A T M activation followed by Chk2 activation and A T R activation followed by Chkl activation, both leading to CDC25 and CDK1 inhibition (Matsuoka et al, 1998; Zhao and Piwnica-Worms, 2001). I therefore monitored the activation of the canonical checkpoint pathways using in vitro kinase assays. I first measured the activity of the kinases ATR, Chkl and CDK1 in M059 cells. Cells were exposed or not to 6.5 Gy, incubated for 16 h to achieve maximal G 2 arrest, and the activity of the immunoprecipitated kinases was measured in vitro using substrate proteins or peptides (see Chapter 2). The protein levels of ATR, Chkl and CDK1 (p34 c d c 2) were similar in the three cell lines and did not change after irradiation (Figure 4.12A). IR caused an increase in ATR activity in M 0 5 9 K and M059J7Fusl cells but not in M059J (Figure 4.12A). No increase in Chkl kinase activity was observed in M 0 5 9 K or M059J/Fusl cells following IR. However, a clear increase in Chkl kinase activity was observed in irradiated M059J cells (Figure 4.12A), which required several hours post-IR to become apparent. This result was unexpected because Chkl is typically activated poorly by IR (Zhao and Piwnica-Worms, 2001), and the upstream activator of Chkl is ATR (Liu et al, 2000). As expected, CDK1 kinase activity decreased following D N A damage in all cell lines, but the decrease was slightly stronger in M059J cells (Figure 4.12A), likely due to increased Chkl activity. Interestingly, an IR dose of 50 Gy was required to elicit a similarly high level of Chkl activation in DNA-PK-complemented M059J/Fusl cells (Figure 4.12C). I also tested the effect of AMA37 on Chkl activity in DNA-PK-proficient MCF-7 mp53 cells. The cells were preincubated with AMA37 for 30 min, irradiated or not, and 107 IR ATR Activity -ATR Protein -Chkl Activity -Chkl Protein -CDK1 Activity -CDK1 Protein -M059J/ M059K M059J F u s 1 B _ + ^00 ^MP DMSO AMA37 IR ATR Activity -ATR Protein -Chkl Activity -Chkl Protein -C M059J/Fus1 M059J Dose(Gy) 0 6 10 25 50 0 6 Chkl Activity - ^ | Load Control - — — — Figure 4.12 : Overactivation of C h k l in DNA-PK-deficient cells. A-B, ATR, Chkl and CDKI activity in irradiated M059 (A) or MCF-7 mp53 (B) cells. Cells were harvested 16 h after mock irradiation (-) or exposure to 6.5 Gy (+). Cellular levels of ATR, Chkl and CDKI protein were measured by Western blotting and their kinase activity was measured in immunoprecipitates as described in Chapter 2.14. Chkl is overactivated after IR in both M059J and AMA37-treated MCF-7 cells. C, Chkl overactivation in M059J cells similar to that in 25-50 Gy irradiated M059J/Fusl cells. Results shown are representative of duplicate experiments. 108 the protein levels and kinase activity of ATR and Chkl were measured after 16 h, when G 2 arrest was maximal. Following IR, Chkl activity increased strongly in cells treated with A M A3 7 but not in cells treated with DMSO (Figure 4.12B). A T R activity also increased, but to a lesser extent (Figure 4.12B). I next examined the activity of the kinases A T M and Chk2. MCF-7 mp53 cells were preincubated with AMA37 and irradiated. One hour after exposure to 6.5 Gy, A T M activity in both DMSO- and AMA37-treated cells increased slightly and then decreased by 16 h (Figure 4.13 A). In the absence of AMA37, Chk2 activity was increased at both 1 and 16 h after irradiation. However, very strong Chk2 activity was observed in AMA37-treated cells 1 h after IR (Figure 4.13A). To confirm that this strong transient increase was indeed due to lack of D N A - P K activity, M059K, M059J and M059J/Fusl cells were irradiated and Chk2 activity was measured 1 and 16 h later. Chk2 was clearly overactivated in DNA-PK-deficient M059J cells compared to DNA-PK-complemented M059J/Fusl cells and M059K cells (Figure 4.13B). I was unable to examine checkpoint kinase activity in CHO cells because available antibodies showed insufficient selectivity towards rodent proteins. 4.2.6 Kinexus screen to further clarify mechanism of G2 arrest As mentioned earlier, the concept of prolonged G 2 arrest simply because of lack of repair in DNA-PK-deficient cells is not supported by D N A repair measurements; M059J cells exhibit complete D N A repair within 24 h post-10 Gy (DiBiase et al, 2000). It could therefore be argued that the observed overactivation of Chkl points towards a role of D N A - P K in cell-cycle resumption following D N A damage, possibly through 109 DMSO AMA37 Load Control -Chk2 Activity -Load Control -Time Post-IR (h) 0 1 16 0 1 16 ATM Activity -B M059K M059J Mf^\JI Time Post-IR (h) o 1 16 0 1 16 0 1 16 Chk2 Activity -Load Control Figure 4.13 : Overactivation of Chk2 in DNA-PK-deficient cells. A, A T M and Chk2 activity in irradiated MCF-7 mp53 cells harvested 1 h or 16 h post-IR. B, Chk2 kinase activity of M059 cells measured at 1 h or 16 h post-IR. Cellular levels of A T M , Chk2, and p-tubulin (loading controls) and the kinase activities were assayed as described in Chapter 2.14. Chk2 exhibits a transient overactivation 1 h post-IR in both cell lines. Results shown are representative of duplicate experiments. 110 some form of signal-transduction mechanism. Indeed, D N A - P K has been detected outside of the nucleus in lipid rafts and different protein phosphorylation patterns were observed in lipid rafts of M059K and M059J cells following irradiation (Lucero et al, 2003). It is for this reason that I collaborated with Dr. Steven Pelech and Dr. Hong Zhang of Kinexus Bioinformatics Corporation to identify upregulated kinases following D N A damage in D N A - P K deficient M059J cells compared to M 0 5 9 K cells. A non-standard "Kinetworks" screen was performed as part of Kinexus's optimization/ development for new antibody screening. As a result, many different antibodies did not provide any useable signal. Others were similar antibodies from different sources, providing overlapping results. Additionally, several others provided conflicting results, (i.e. one antibody source shows target upregulation, whilst another shows downregulation or no change of expression) and were eliminated from subsequent analysis. Figure 4.14 provides example western blot results from the Kinexus screen. Tables 4.2 through 4.4 list the antibody targets that exhibit a change in expression level between M059J and M 0 5 9 K cells. Although there are many (8 upregulated, 17 downregulated), only 2 were considered relevant, as others either had no documented role in the G 2 checkpoint or cell cycle arrest, or their respective up- or down-regulation would be antagonistic to the observed G 2 arrest in M059J cells (i.e. for example, decreased Cdc2 Tyr 15 phosphorylation (Table 4.3) would be indicative of checkpoint release and G2—>M transition). Therefore, attention was paid to M K K 6 and A T M Ser 1981 phosphorylation. In the case of the latter, decreased A T M phosphorylation was considered a likely by-product of reduced total A T M expression in M059J cells, a well 111 Figure 4.14 : Sample western blots from Kinexus screen of irradiated M 0 5 9 K and M059J cells. The large arrow indicates MKK6, with A, low expression in M059K cells, and B, high expression in M059J cells. Lysates were prepared from each cell line 16 h after 6.5 Gy irradiation. This work was performed by employees at Kinexus Bioinformatics Corp. 112 Table 4. 2 : Kinetworks targets exhibiting increased expression in M059J cells compared to M059K cells 16 h after irradiation. Antibody target Cortactin Tyr421 EGFR FAK1 Tyr397 Human NF2 (Merlin) ILK-1 M K K 6 P Y K 2 TAK1 Documented involvement in G2 checkpoint? Confirmed? No No No No No Yes (Wang et al. No No 2000) No* No* No* No* No* Yes No* No* * Results were not re-tested due to no documented role in the G 2 checkpoint 113 Table 4. 3 : Kinetworks targets exhibiting decreased expression in M059J cells compared to M059K cells 16 h after irradiation. Antibody target Documented involvement in G 2 checkpoint? Confirmed? A L K No No* AKT1 Yes (King etal., 2004) See discussion AKT1 Ser473 Yes (King et al., 2004) See discussion A T M Serl981 Yes (Xu et al., 2002) Yes (Peng et al, 2005) Cdc2 Tyr 15 No No DRAK1 (NT) No No* eIF-2A Ser52 No No* eNOS/NOS III Thr495 No No* FosB Ser284 No No* GAP-42 Ser41 No No* IKKalpha No No* Lyn No No* Mekl Ser298/Thr292/386 No No* N M D A Nr2B Tyr 1472 No No* R N A Pol II No No* Smadl Ser463/465 No No* STAT5 Tyr694 No No* Synapsin 1 Ser603/194/9 No No* TauSer296/Thr231 No No* * Results were not re-tested due to no documented role in the G 2 checkpoint 114 Table 4. 4 : Kinetworks targets exhibiting no change in expression between M059J and M059K cells 16 h after irradiation. Alpha-CAM Kinase F K H R Serl 19/22/ MEK1/2 Progesterone II Thr256 Ser218/222 Receptor SI90/294 ASK1 FosB Ser302/303 MST1/KRS2 Rad9 Serl 260 ASK-1 Ser83 G A B A B Receptor Ser892 mTOR Ser2448 Raf-1 ATF2 Thr71 GFAP Myosin S19 Rb Ser780/807/811 Aurora A GluRl Ser831 NEK2 ROCK-1-Cleaved Bad Serl 12/136 HER3 NF-H SH3BP2 Ser427 Beta-CAM Kinase Histone H3 N E - K B Ser276/529 SMC1 Ser957 II SerlO/28/Thrll C A M Kinase II Huntingdon Ser421 p27 SrcTyr419 Thr286/305 Caveolin-2 Ser23/36 IKKe Thr501 p27 SerlO STAT1 Tyr701 CDK7 IRAK p53 Ser6/9/315/392 STAT4 Tyr693 cFOS Thr232 IRAK-2 p90 Rsk STAT5 Tyr694 CREB Serl33 IRS1 Ser307/527 PAK1/2/3 Thr423 Synapson Ser9 c-Src IK(3 Pax2 Ser393 T B K - l / I K K d D N A Topo-IIa i K p a l p h a Ser32/36 Paxillin Tyr31 Tlk-1 D R A K 2 JNK2 PDK1 Topo-IIaThrl342 EGFR Serl 179 Lck PDKpllOdel ta Zip kinase eIF-4E Ser209 M A P K A P K - 2 PKCd Tyr311 Elk-1 Ser383 M A R C K S Serl52/156 PKCg Thr514/674 eNOS Serl 179 niBAD Serl 36 PLK1 F A K M D M 2 Serl 85 P lk l Thr210 115 documented observation (Chan et al, 1998; Hoppe et al, 2000; Tsuchida et al, 2002; Peng et al, 2005) . The remaining candidate target, M K K 6 , was thus explored as having a potential role in the prolonged arrest observed in M059J cells. 4.2.6.1 M K K 6 does not play a role in M 0 5 9 J prolonged arrest Map-kinase kinase 6 (MKK6) is the mammalian homolog of the yeast osmotic sensor Pbs2. It provides the activating phosphorylation on p38, which may contribute to checkpoint arrest via M A P K A P K - 2 (reviewed in (Karlsson-Rosenthal and Millar, 2006)). Indeed, overexpression of a constitutively active M K K 6 or p38 mutant causes permanent G 2 arrest (Wang et al, 2000). Therefore, it was understandably exciting to identify M K K 6 upregulation in M059J cells. With cells expressing abnormally high M K K 6 , activation following IR may lead to G 2 checkpoint activation via a pathway that would not be affected by current checkpoint inhibitors. Although this has not been documented, DNA-PK may in some way inhibit/repress M K K 6 , so without DNA-PK, M K K 6 may lead to a hypersensitive checkpoint response. As shown in Figure 4.15 A , the activating phosphorylation site on M K K 6 was phosphorylated after D N A damage (Serl89/Ser207), and remained phosphorylated for 48 h post-IR, whereas no signal was detected in the D N A - P K proficient cells. RT-PCR of M059J cells revealed that M K K 6 transcripts are at higher levels within the cell compared to M 0 5 9 K cells, indicating that M K K 6 upregulation is at the transcript level, not through stabilization of the M K K 6 protein (Figure 4.15B). However, it is not known whether this is due to increased transcription of the M K K 6 gene, or i f this represents mRNA stabilization. Stable transfection of M059J cells with a constitutively active or kinase-dead M K K 6 cDNA proved difficult due to extremely low transfection efficiencies (data not 116 Time Post-6.5 Gy (h) MKK6 P- MKK3/6 P- tubulin M059K M059J M059J/ Fusl 0 24 48 0 24 48 0 24 48 mat P - M K . K 4 P - M K K 3 P - M K K 6 B K J RT-PCR M K K 6 G A P D H Figure 4.15 : M K K 6 protein levels are much higher in M059J cells compared to M059K cells. A , Western blotting confirms the Kinexus screen result that M059J cells express high levels of MKK6. M059J/Fusl express intermediate/low levels of MKK6. The activating phosphorylation on MK.K.6 (Serl89/Ser207) is present after IR in M059J cells, upwards of 48 h post-IR. Blots shown are representative of duplicate experiments. 5, Relative RT-PCR analysis reveals MKK6 transcript levels are slightly higher in M059J (J) than M059K (K). 117 shown). Luckily, selective inhibitors of both M K K 6 and p38 are commercially available, and were used in combination with checkpoint inhibitors in M059 cells. If M K K 6 expression causes M059J cells to arrest indefinitely, inhibition of M K K 6 or its downstream component p38 should relieve that checkpoint arrest signal. Unfortunately, inhibition of MKK6/p38 did not lead to an increase in checkpoint abrogation in M059J cells, indicating that M K K 6 does not play a role in the strong checkpoint response seen in these cells (Figure 4.16). 4.2.6.2 PKB does not play a role in M 0 5 9 J prolonged arrest During the above-described investigation, it was reported that M059J cells exhibit low, albeit constitutive, protein kinase B (PKB)/Akt (hereafter referred to solely as PKB) activity that is not responsive to typical stimuli such as insulin (Feng et al, 2004). These authors present evidence that DNA-PKcs acts as a putative "PDK2", providing the activating phosphorylation at Ser473 on PKB. P K B requires phosphorylation on 2 sites for activation; Thr308 which is provided by PDK1 (reviewed in (Dong and Liu, 2005)), and Ser473, whose upstream kinase is currently unknown and is broadly referred to as "PDK2". Many different kinases have been proposed to act as PDK2, including D N A - P K (Feng et al, 2004), but also A T M (Viniegra et al, 2005), mTOR (Hresko and Mueckler, 2005), and members of the MKK6/p38/ M A P K A P K - 2 pathway (Rane et al, 2001; Gonzalez et al, 2004). Indeed, P K B Ser473 phosphorylation was identified as lower in M059J cells compared to M 0 5 9 K cells in the Kinexus screen (Table 4.3). As M059J cells exhibit high levels of M K K 6 , and low but constitutive phosphorylation on P K B Ser473 (Feng et al, 2004), I hypothesized that M K K 6 is acting 118 60 •o g 40 in o Q_ i u_ O 30 20 10 0 M059K M059J M059J/Fus1 / / T 1 / / | / / o > o Q 73 W W W CO CO CO S 3 c o rO CD CO 7s CO 7s T l O > T l CO T l O JJ Figure 4.16 : Inhibition of MKK6/p38 does not affect checkpoint inhibitor response in M059J cells. SB203580 (SB) and SKF86002 (SKF) are specific inhibitors of p38 and MKK6, respectively. U0126 is an inhibitor of MEK/ERK. Checkpoint inhibition was assessed as in Figure 4.4. Combining 10 uM of either SB or SKF with 2 mM caffeine or 10 uM isogranulatimide did not increase checkpoint abrogation in M059J cells (white bars). Cells were analyzed by flow cytometry as described in Chapter 2.5.2. Results shown are means +/- SD (n = 3) when error bars are shown. 119 as a PDK2 in M059J cells because its normal PDK2, DNA-PKcs, is not present so the cell line acquired the ability to maintain steady P K B activation for cell survival through M K K 6 upregulation. When M059 cells are serum starved for 24 h, then exposed to 100 nM insulin for 30 minutes, P K B Ser473 phosphorylation is present (Figure 4.17). As reported, M059J cells are not insulin responsive (Figure 4.17) (Feng et al, 2004). However, when serum starved cells are exposed to 10 uM of either SKF86002 or SB203580 for 30 minutes, the constitutive phosphorylation of Ser473 in M059J cells is decreased, whereas levels are unaffected in M 0 5 9 K and M059J/Fusl cells. Activated P K B can phosphorylate and inactivate Chkl/Chk2, possibly acting as a mechanism to "turn o f f the checkpoint (Shtivelman et al, 2002; King et al, 2004). Therefore, a reasonable hypothesis could be that in cells lacking D N A - P K kinase activity, Chkl is maintained in an activated state due to a lack of inhibitory signal from PKB, as PKB is no longer stimulus-responsive due to constitutive MKK6-mediated phosphorylation. This general hypothesis is outlined in Figure 4.18, where D N A - P K normally activated P K B , and in some manner represses M K K 6 expression. When DNA-PK is no longer present, elevated levels of M K K 6 are able to phosphorylate P K B in a stimulus-independent manner. However, normally activated P K B represses the checkpoint effector kinases Chkl and Chk2. Therefore, rescue of P K B activation in M059J cells should "turn o f f the checkpoint. To test this hypothesis, I stably transfected M059 cells with a retrovirus encoding mER-PKB-IRES-GFP (Kohn et al, 1998). This construct has the PH domain of PKB deleted, replaced with an N-terminal myristolation sequence targeting it to the cellular 120 M 0 5 9 K M 0 5 9 J M 0 5 9 J / F u s 1 O CM O CN O 00 O CO O CO S m o I D o L O o c n (D c n to c e o <g 3 CN U_ 3 CM U _ 3 CM LL (/) DO ^ W CO ^ B CQ ^ c (/) (/) £ W C/) J= CO CO P - S e r 4 7 3 • - -p M u b u l i n Figure 4.17 : P K B activation in M059J cells is insulin-unresponsive but can be inhibited by the M K K 6 inhibitor SKF86002 or p38 inhibitor SB203580. M059 cells were serum starved for 24 h, then treated with insulin (100 nM), or 10 uM of either SB203580 or SKF86002 for 30 min. The p38 inhibitor SB203580 and the MKK6 inhibitor SKF86002 reduces phosphorylated Ser473 in serum-starved M059J cells but not M059K or M059J/Fusl. Lysates were prepared and blotted as described in Chapter 2.11. Blots shown are representative of duplicate experiments. 121 A M 0 5 9 K B M 0 5 9 J Figure 4.18 : Model for prolonged arrest in M059J cells. A, DNA-PK typically activates PKB providing inhibitory signals to the checkpoint by inactivating Chkl/2. Through unclear mechanism(s), DNA-PK represses the expression of M K K 6 . B, In M059J cells with no DNA-PK, M K K 6 expression is increased, which in this model provides constitutive phosphorylation on PKB, but at low levels insufficient to inactivate Chkl/2. 122 membrane, and a modified estrogen receptor (mER) at the C-terminus. When the synthetic hormone 4-hydroxytamoxifen (4-HT) is applied to the cells, the receptor shifts to allow the construct to autophosphorylate and activate itself. As these constructs are constitutively expressed, the-IRES-GFP allows for cell sorting of those which express the PKB via GFP expression (example sorting data shown in Figure 4.19). These cells were then tested for P K B activation via Ser473 phosphorylation in western blotting after 4-HT treatment (Figure 4.20). In the presence of serum, Ser473 phosphorylation is clearly evident in all cells lines. M059J cells exhibit lower basal phosphorylation on Ser473, consistent with its proposed role as a PDK2 (Feng et al, 2004). Following 1 uM 4-HT for 30 minutes, the mER-PKB construct is phosphorylated, indicating its activation. The A2-PKB construct exhibits no activation, as expected. Although the endogenous P K B shows differences in phosphorylation levels between cell lines, all 3 cell lines exhibit equivalent activation of mER-PKR following 4-HT treatment. 4-HT did not affect endogenous Ser473 phosphorylation in mock-transfected cells (data not shown). The observed molecular mass of (mER-/A2-)/PKB is slightly higher than expected (approximately 60 kDa), suggesting it may be subject to additional post-translational modification, such as glycosylation for example, in M059 cells. No other bands were observed on the western blots. The cells were irradiated and monitored for G 2 checkpoint abrogation with 4-HT stimulation. Mock-transfected cells exhibited no change in checkpoint abrogation compared to that in Figure 4.4A. However, mER-PKB expressing M 0 5 9 K and M059J/Fusl cells did not exhibit any increase in checkpoint abrogation compared to drug treatment alone. In the case of M059J, there was a very small increase in 123 A B gure 4.19 : Representative GFP-sorting profiles for M059 cells. A, Mock-transfected M059K. cells do not express GFP (no cells in P5 region). B, transfected M059K cells have approximately 10% GFP-positive cells (P5 region). C , enriched M059K cells expressing GFP after FACS, as described in Chapter 2.17. 124 m E R - P K B A 2 - P K B 4 - H T P h o s p h o - S e r 4 7 3 P K B T o t a l P K B 115.5-82.2-64.2-48.8-115.5-82.2-64.2-48.8-M 0 5 9 K M 0 5 9 J M 0 5 9 J / F u s 1 + + - - + + - - + + - -- - + + - - + + - - + + - + - + - + - + - + - + m PKB construct Endogenous PKB PKB construct Endogenous PKB Figure 4.20 : P K B activation in mER-PKB M059 cell lines after 4-HT treatment. M059 cells stably transfected with the PKB-IRES-GFP construct were treated with 1 u.M 4-HT for 30 min. PKB activation is assessed by Ser473 phosphorylation by western blotting as described in Chapter 2.11. Blots shown are representative of duplicate experiments. 125 checkpoint abrogation in mER-PKB M059J cells compared to mock-transfected, but only when treated with caffeine, and the increase was too small to rival that of M 0 5 9 K or M059J/Fusl cells. This likely indicates that P K B activation is incapable of bypassing the checkpoint arrest observed in all M059 cells. These results are summarized in Table 4.5. As a final test, I monitored the G 2 / M population over time in these cells, in case PKB-mediated checkpoint abrogation required longer than the 8 h used in Table 4.5. The observations for this are listed in Table 4.6. Of note, M 0 5 9 K cells normally recover from IR by 24 h, so no difference was detected in the M 0 5 9 K lines. In the case of M059J/Fusl cells, which take slightly longer than M 0 5 9 K cells to recover from G 2 arrest (Figure 4.9, mock-transfected results in Table 4.6), mER-PKB activation failed to decrease checkpoint recovery time. Finally, mER-PKB activation clearly failed to decrease the G 2 / M population of M059J cells, as the mock-transfected, the mER-PKB and the A2-PKB M059J lines exhibit a strong, sustained G 2 arrest. These results unfortunately indicate that insufficient PKB activation is not playing a significant role in the prolonged G 2 arrest observed in M059J cells, supporting the argument that it is indeed lack of D N A repair causing prolonged arrest, not defective signalling mediated by DNA-PK. 4.2.7 M 0 5 9 J cells enter a senescent-like state after D N A damage I concluded my study of DNA-PK-deficient cells by asking the question, what becomes of the G2-arrested cells? If they fail to re-enter the cell cycle 48 h post-IR, is this a truly a permanent exit from the cell cycle? Replicative senescence is defined as a permanent and irreversible exit from the cell cycle 126 Table 4. 5 : Checkpoint inhibition of mER-PKB M059 cells. Cells were treated as in Figure 4.4, except that they were exposed to 1 uM 4-HT at the time of checkpoint inhibitor addition (16 h). Checkpoint inhibition is expressed as the % GF-7 positive cells. Results shown are the means +/- SD (n > 3) DMSO + - -Caffeine - + -Isogranulatimide - - + M059K Mock 16.8 +/-3.9 38.6 +/-2.8 43.6 +/-3.6 mER 17.1 +/-10.1 43.3 +/- 1.1 35.7 +/-3.6 A2 14.8 +/-6.6 34.0 +/- 1.7 33.1 +/- 5.9 M059J Mock 4.0 +/- 1.1 24.5 +/- 1.9 10.2 +/- 2.8 mER 4.6 +/- 1.1 29.7 +/-0.6 14.0 +/- 2.8 A2 3.1 +/- 1.0 18.6+/- 4.6 8.8 +/- 0.8 M059J/Fusl Mock 18.4 +/- 4.0 44.3 +/-4.7 22.1 +/- 1.3 mER 21.1 +/- 1.2 41.1 +/- 15.6 30.3 +/- 1.0 A2 22.6 +/- 9.2 45.5 +/-9.6 40.8 +/- 1.1 127 Table 4. 6 : G 2 arrest in P K B - I R E S - G F P M 0 5 9 cells over time. Cells were exposed to 1 uM 4-HT at the time of irradiation. % G 2 / M population was monitored over time. Results shown are the means +/- SD (n = 3). Time post-6.5 Gy 0 24 48 M 0 5 9 K Mock 25.5 +/- 6.4 34.5 +/- 7.8 34.0 +/- 8.5 mER-PKB 30.0+/- 7.1 30.0 +/- 1.4 30.5 +/- 2.1 A2-PKB 28.5 +/- 6.4 28.5 +/- 2.1 31.0+/- 2.8 M 0 5 9 J Mock 44 +/- 12.0 71.5 +/- 6.4 76.0 +/- 4.2 mER-PKB 29.0+/- 5.7 83.5 +/- 3.5 85.0+/- 5.7 A2-PKB 31.0+/-7.1 87.0 +/- 2.8 87.5 +/- 0.7 M059J /Fus l Mock 28.5 +/- 9.2 58.5 +/- 2.1 38.5 +/- 6.4 mER-PKB 32.5 +/- 10.6 57.0 +/- 4.2 30.5 +/- 4.9 A2-PKB 34.5 +/- 9.2 63.0+/- 2.8 43.0+/- 7.1 128 (Hayflick and Moorhead, 1961; Hayflick, 1965; Matsumura et al, 1979). Senescent cells often have a large and flat morphology while elevating the activity acidic (3-galactosidase (termed senescence associated P-galactosidase, or SA P-gal) (Smith and Lincoln, 1984; Goldstein, 1990; Dimri et al., 1995). As SA P-gal expression is unique to senescent cells and not simply growth arrest, it serves as an ideal biomarker for cells which have entered senescence. I therefore monitored SA P-gal expression over time in equivalent populations of M059 cells after 6.5 Gy IR. As shown in Figure 4.21, M059J cells clearly display an increase in SA P-gal expression over time when compared to both M059K and M059J/Fusl cells. Although not entirely unexpected, this observation may be unique in that M059J cells are arrested in G 2 after IR, while senescence is commonly defined as an arrest in G i . Attempts to correlate SA P-gal expression with cell cycle status (i.e. G | or G 2 arrest) via flow cytometry using fluorescent X-gal analogs were unsuccessful (data not shown). 4.3 DISCUSSION Overall, the results show that irradiation of cells lacking D N A - P K activity or cells exposed to the D N A - P K inhibitor AMA37 causes overactivation of Chkl and Chk2. This abnormally high checkpoint signalling is likely the mechanism underlying the prolonged G 2 arrest. Prolonged G 2 arrest dependent on Chkl overactivation has also been observed in human A T cells after D N A damage (Xu et al, 2002), as well as mouse Ku86-deficient cells (Wang et al, 2002). I also note that the checkpoint inhibitors caffeine and UCN-01 were less able to bypass G 2 arrest in AT and Ku86-/- cells than in wild-type cells (Xu et al, 2002; Wang et al, 2003b). 129 A 0 1 2 3 4 5 6 7 8 Time (Days) Post-6.5 Gy Figure 4.21 : M059J cells exhibit signs of senescence after I R over time. A, example senescence-associated acidic P-galactosidase staining in M059J cells, performed as described in Chapter 2.18. Cells become enlarged, flattened, and express SA-(3-gal after IR. B, Equivalent populations of each cell line were irradiated with 6.5 Gy in 6-well plates, then stained for senescence-associated (SA) acidic P-galactosidase, as in A. An increasing portion of the remaining M059J cells (open circles) exhibit positive SA-acidic P-galactosidase staining over time. Results shown are means +/- SD (n = 3). 130 Curiously, Wang et al. (2002) did not observe a prolonged G 2 arrest in DNA-PK-deficient cells, but did for Ku86-deficient cells. Unfortunately, the cell-type or p53-status of the lines used in that study were not provided, so I cannot account for this different observation. However, the similarly overactivated Chkl during prolonged G 2 arrest indicate that the previously described effects of Ku86 deletion are due to the absence of D N A - P K kinase activity rather than an independent function associated with Ku86 proposed by (Wang et al., 2002). The similarity of the responses of AT cells and NHEJ-deficient cells to D N A damage was unexpected since D N A - P K and A T M , while both members of the PIKK family of protein kinases, have distinct roles in the D N A damage response. D N A - P K is involved mostly in NHEJ D N A repair in humans and is not believed to participate in D N A damage response signalling, while A T M plays checkpoint-signalling roles at all phases of the cell cycle. However, A T M has also been implicated in some forms of D N A repair, as AT cells exhibit a mild repair defect (Nakamura et al, 2006). A similar checkpoint defect has also been noted in cells that are MRE11-/- (Carson et al, 2003), NBS1-/- (Xu et al, 2002), MDC1- / - (Stewart et al, 2003), and possibly others. Indeed, all these components have roles in NHEJ, whether through the detection and activation of repair machinery, or in the repair process itself. Of further interest would be whether this could exemplify the necessary crosstalk between repair machinery and proteins regulating cell cycle control, or i f the observed G 2 accumulation is simply a result of S-phase lesions, as discussed below. The observed increase in A T M expression reported for M059J/Fuslcells (Hoppe et al, 2000) indicates that D N A - P K is somehow involved in the regulation of A T M 131 expression. Indeed, Peng et al. (2005) have demonstrated that D N A - P K expression correlates with A T M expression. Additionally, the failure to observe A T R activation in M059J cells could further indicate a global regulatory role of D N A - P K in other PIKKs. The cause for these changes in expression warrants further examination. However, I consider the reduced expression of A T M to be likely insignificant, as i) M059J/Fusl cells have very low A T M activity (Virsik-Kopp et al, 2004), yet failed to arrest permanently similar to M059J cells at low IR doses, and ii) AMA37 treatment for the times indicated fails to affect A T M activity, yet prolonged G 2 arrest is still observed. The same prolonged arrest is observed in CHO XR-1 cells that possess functional D N A - P K kinase activity, indicating the defective checkpoint signalling is not due to D N A - P K itself, but NHEJ in general. Therefore, what causes the overactivation of Chk2? Several lines of evidence suggest that overactivation is related to the amount of unrepaired D N A damage. First, Chk2 overactivation occurs in NHEJ-defective cells at the time of irradiation when damage is maximal, but not at later times, indicating that other repair pathways can compensate during later stages of the cell-cycle (Figure 4.13). In normal cells, Chk2 activation is probably limited because rapid D N A repair restricts the magnitude of the damage signal. Second, some M059J cells, CHO V3 and CHO XR-1 cells arrest in G 2 even after a very low IR dose of 2 Gy and remain arrested for a very long time. What about Chkl overactivation? This may be related to when in the cell-cycle cells are exposed to IR. M y experiments show that prolonged G 2 arrest does not occur when cells are irradiated in G 2 , or following AMA37 treatment of G2-arrested cells. The observation that upon exposure to a low IR dose of 2 Gy, most M059J cells do not arrest 132 in G 2 but that the small proportion of cells that do arrest in G 2 remain arrested for a very long time (Figure 4.9) might indicate sensitivity during G i . Cells in early Gi may have sufficient time to repair DSBs phase via slow component NHEJ repair before they undergo S-phase (DiBiase et al, 2000; Iliakis et al, 2004). However, cells close the Gi/S transition at the time of irradiation might enter S-phase with significant D N A damage, as they are p53-deficient and lacking a functional Gi checkpoint to arrest the cell cycle. A possible explanation is that the NHEJ-defect leads to a saturated HRR system unable to cope with the type or extent of D N A damage and this induces prolonged G 2 arrest and Chkl overactivation. Indeed, loss of NHEJ leads to increases in HRR activity (Pierce et al, 2001; Allen et al, 2002). Replication of damaged D N A might convert breaks into distinct lesions that signal prolonged G 2 arrest through activation of the HRR system (Xu et al, 2002; Wang et al, 2003b). Furthermore, the larger doses of 6.5 and 10 Gy would cause a significant increase in D N A DSBs at any point in G i , resulting more cells with unrepaired D N A entering S-phase. A requirement for S-phase traversal has been demonstrated for prolonged G 2 -arrest in AT cells (Xu et al, 2002). Further support is evidenced by the complete, albeit very slow, repair of DSBs in scid cells (Evans et al, 1996; Nevaldine et al, 1997); complete repair takes place within 24 hours post-IR as monitored by PFGE (not capable of detecting an S-phase lesion). Additionally, my observations that G 2 synchronization does not result in a checkpoint defect in M059J cells, nor does AMA37 addition to G 2 -arrested cells, suggests S-phase traversal is necessary. Together, these data indicate that the Chkl overactivation and prolonged G 2 arrest are not simply due to the presence of 133 large amounts of unrepaired D N A lesions in these cells, but perhaps a specific type of processed lesion after D N A replication in a repair-compromised system. The overactivation of Chkl and Chk2 may explain why cells lacking D N A - P K activity are less responsive to checkpoint inhibitors that target Chkl and/or Chk2: higher concentrations of inhibitors might be required to efficiently counteract the overactivated pathway. There may also be additional differences between cells undergoing transient G 2 arrest and cells undergoing prolonged G2-arrest that render checkpoint inhibitors less effective, as described below. The degree of checkpoint inhibitor efficacy was observed to differ between CHO V3 and CHO XR-1 cells, with XR-1 cells failing completely to respond to the inhibitors (see Figure 4.5). This could indicate that the degree of unrepaired D N A is a limiting factor in the response to a checkpoint inhibitor, as XRCC4-deficient cells would be expected to have near-complete inhibition of D N A repair, while DNA-PK-deficient cells are capable of a slower but complete repair (Nevaldine et al., 1997; Bryans et al, 1999; Lee et al, 2003; Riballo et al, 2004). Indeed, when secondary D N A damage is inflicted on G2-arrested cells, checkpoint inhibitor efficacy decreases in an IR dose-dependent manner (Figure 4.22). Such a finding is surprising, as checkpoint inhibitors have been thought to bypass cell cycle arrest in the presence of D N A damage, leading to mitotic catastrophe. Instead, checkpoint inhibitors are not uniformly active in the presence of D N A damage; they may be limited by the degree or type of D N A damage present. There may also be additional differences between cells undergoing transient G 2 arrest and cells undergoing prolonged G2-arrest that render checkpoint inhibitors less effective. At time points longer than 48 h, most G2-arrested M059J cells showed sub-Gi 134 gure 4.22 : DNA damage at the time of checkpoint inhibitor application decreases checkpoint inhibitor activity. 16 h post-6.5 Gy, MCF-7 mp 53 cells were irradiated again (mock - black bars, 2 Gy - white bars, 5 Gy - crosshatched bars) and given checkpoint inhibitors plus 100 ng/mL nocodazole for 8 h. With additional DNA damage at the time of checkpoint inhibitor application, checkpoint abrogation decreases. % Mitotic cells were determined by mitotic spreads as described in Chapter 2.5.1. Results shown are means +/-SD (n = 3). 135 D N A peaks, indicating apoptosis. However, many surviving cells were observed and they showed signs of senescence, including a large size, flattened shape and increased acidic (3-galactosidase activity (data not shown). Recently, it has been shown that senescent human fibroblasts exhibit similar markers to D N A damage checkpoint-arrested cells, such as co-localization of phosphorylated H 2 A X with 53BP1, MDC1 and Nbsl , along with overactivation of the checkpoint kinases Chkl/Chk2 (Fagagna et al, 2003). Below a threshold level of G 2 checkpoint signalling, cells may undergo transient G 2 arrest to increase time for D N A repair and then resume cell cycle progression, while above this threshold, cells may arrest permanently and enter a senescent state that is unresponsive to G 2 checkpoint inhibitors. Drugs targeting D N A - P K might be useful to drive cancer cells into a senescent state. Variable levels of D N A - P K activity and DNA-PKcs, Ku70 and Ku86 protein levels have been documented in a variety of cancers (Wilson et al, 2000; Rigas et al, 2001; Stronati et al, 2001; Eriksson et al, 2002; Friesland et al, 2003). The observation that the level of D N A - P K activity influences the ability of cells to resume cell cycle progression after exposure to a clinical dose of 2 Gy may therefore have relevance to the fate of tumour cells after radiotherapy or treatment with DNA-damaging chemotherapeutic agents, and ultimately to the clinical outcome of these therapeutic approaches. In addition, combination therapy with DNA-damaging agents and G 2 checkpoint inhibitors or D N A - P K inhibitors has been proposed as a way to increase the selective killing of tumour cells (Allen et al, 2003; Anderson et al, 2003; Kashishian et al, 2003; Shinohara et al, 2005). The observation that D N A repair inhibition strongly 136 affects the response of cells to checkpoint inhibitors should be taken into consideration when further developing these experimental therapeutic approaches. 4.4 S U M M A R Y In response to D N A damage, cell survival can be enhanced by activation of D N A repair mechanisms and of checkpoints that delay cell cycle progression to allow more time for D N A repair. Inhibiting both responses with drugs might cause cancer cells to undergo cell division in the presence of lethal amounts of unrepaired DNA. However, I show that interfering, with D N A repair via inhibition of DNA-dependent protein kinase (DNA-PK) reduces the ability of checkpoint inhibitors to abrogate G 2 arrest, as well as their radiosensitizing activity. Cells exposed to the D N A - P K inhibitor AMA37, DNA-PK-deficient cells and non-homologous end joining-deficient cells and all enter prolonged G 2 arrest after exposure to ionizing radiation doses as low as 2 Gy. The checkpoint kinase Chk2 becomes rapidly and transiently overactivated while Chkl shows sustained overactivation that parallels the prolonged accumulation of cells in G 2 . Therefore, in irradiated cells, D N A repair inhibition elicits abnormally strong checkpoint signalling that causes essentially irreversible G 2 arrest and strongly reduces the ability of checkpoint kinase inhibitors to overcome G 2 arrest and radiosensitize cells. A Kinexus screen for upregulated phosphorylation-sites in either M 0 5 9 K or M059J cells revealed total M K K 6 levels highly upregulated in M059J cells. Although upregulated M K K 6 can lead to prolonged G 2 arrest, inhibition of M K K 6 in M059J cells did not cause checkpoint abrogation. A potential link between M K K 6 and PKB was explored, but also did not lead to any increased checkpoint escape in M059J cells. Variable levels of proteins 137 controlling D N A repair have been documented in cancer cells. Therefore, these results have relevance to the development of D N A - P K inhibitors and G2 checkpoint inhibitors experimental therapeutic approaches to enhance the selective killing of tumour cells by radiotherapy or DNA-damaging chemotherapeutic agents. The work presented in this thesis, from Figures 4.1 through to 4.13, has been published (Sturgeon et al, 2006c). Figures 4.14 and above, and Tables 4.1 to 4.6 represent my subsequent attempts at delving further into the mechanism behind this prolonged arrest. 138 Chapter 5 In vivo development of checkpoint inhibitors 5.1 I N T R O D U C T I O N As discussed in previous chapters, checkpoint inhibitors are able to bypass cell-cycle arrest in p53-deficient cell lines, causing cells to enter mitosis. There are quite a few checkpoint inhibitors that have been identified to-date, and have been summarized in Table 1.1. Despite the large number of chemical inhibitors known, little inroads have been made with respect to utilizing them as therapeutic agents. As discussed in Section 1.4, the most efficacious of checkpoint inhibitors, caffeine and pentoxifylline, are not appropriate for clinical use due to their numerous side-effects resulting from the high-concentrations required for checkpoint inhibition (Arnaud, 1987; Jiang et al, 2000). The only published clinical trial utilizing a checkpoint inhibitor (UCN-01) in combination with D N A damage (cisplatin), was unable to demonstrate potentiation of the response to cisplatin by UCN-01 (Lara et al, 2005), and was halted prematurely due to toxicity of the drug combination. The clinical development of UCN-01 is further hampered by the fact that it binds non-specifically to human alphal-acid glycoprotein (Fuse et al, 1999). The pre-clinical status of the other inhibitors listed in Table 1.1 has not been made public (Garber, 2005). Should checkpoint inhibitors prove beneficial, it will be of obvious 3 A section of this chapter is in review for publication (Sturgeon, C. M . and M. Roberge (2006). Checkpoint abrogation alone is insufficient for radiosensitization by G 2 checkpoint inhibitors). 139 interest to develop methods for in vivo checkpoint inhibitor application and delivery if they are ever to be utilized in a clinical setting. To accomplish this, I asked whether it was possible to test the Roberge-lab's most advanced and promising checkpoint inhibitor, isogranulatimide (IGR) (Roberge et al, 1998; Jiang et al, 2004), in an in vivo mouse model. Isogranulatimide exhibits strong preference towards Chkl over a panel of kinases, when compared to UCN-01 (Jiang et al, 2004). Before mouse studies began, however, I attempted to optimize single-dose delivery of isogranulatimide in vitro for optimal radiosensitizing effect. Surprisingly, I found that the time of checkpoint inhibitor application after irradiation importantly affects cell survival. The ideal dose was then used in initial attempts to test isogranulatimide in vivo using an isogenic HCT116 lines expressing p53 or not. 5.2 RESULTS AND DISCUSSION 5.2.1 Time of checkpoint inhibitor application affects radiosensitivity Previous studies by the Roberge lab have found that maximal G 2 arrest is attained in 16-24 h post-IR, offering a useful window for the screening of novel compounds that can bypass this arrest (Seynaeve et al, 1993; Wang et al, 1996; Roberge et al, 1998; Curman et al, 2001; Rundle et al, 2001; Sturgeon et al, 2005; Sturgeon et al, 2006a). Such an application eliminates the concern of other possible cell-cycle effects of the drugs, such as an S-phase delay or Gi-arrest (as is the case with UCN-01, a well known G 2 checkpoint inhibitor that can induce Gi arrest in the absence of D N A damage) (Seynaeve et al, 1993). However, others in the Roberge lab have noticed that application of checkpoint inhibitors during the 16-24 h post-IR window fails to effectively 140 radiosensitize cells (unpublished observations). Indeed, when tested in clonogenic assays, checkpoint inhibitors failed to cause radiosensitization in either HCT116 p53"A or MCF-7 mp53 cells (Tables 5.1 and 5.2). This raised the question : does this failure to radiosensitize indicate that G 2 checkpoint abrogation is not necessary for radiosensitization? Presumably checkpoint inhibitors cause radiosensitization via passage through mitosis after D N A damage, and yet bypassing the G 2 checkpoint in this case fails to effectively radiosensitize cells. In order to assess the mechanism of radiosensitization by checkpoint inhibitors, various other dosing schedules were employed, as outlined in Figure 5.1. HCT116 and MCF-7 mp53 cells have an approximate 24 h doubling time in standard cell culture, and have been extensively studied by the Roberge laboratory and others (Roberge et al., 1998; Curman et al, 2001; Rundle et al, 2001; Sturgeon et al, 2005; Sturgeon et al, 2006a). Following exposure to high doses of IR capable of eliciting G 2 arrest, the cells progress to S-phase as a loosely-synchronized population. By 8-h post-IR, the majority of cells are in late-S or G 2 , and remain arrested there for the next 8-16 h. The majority of D N A repair, through NHEJ, is complete within 90 minutes post-IR (DiBiase et al, 2000). The following checkpoint inhibitor applications were used: 0-24 h post-IR, representing one-full doubling time following DNA-damage with inhibitors present during D N A repair via NHEJ, S-phase traversal and G 2 checkpoint inhibition; 16-24 h post-IR, representing a synchronous G2-arrested population of cells solely for checkpoint inhibition; 2-24 h post-IR, similar to 0-24 h, except that checkpoint inhibitors are added to the cells after most D N A repair is complete via NHEJ (DiBiase et al, 2000); and 0-8 h post-IR, representing having checkpoint inhibitors only present during D N A repair and S-141 Figure 5.1 : General approach to discern optimal time of checkpoint inhibitor application. A, After IR, cell cycle distribution moves from an asynchronous profde (time 0) to G 2 arrest (8-16 h) to loosely-synchronized exit (24+ h). G, blue line, S red line, G 2 green line. Histograms are example flow cytometry profiles representative of average cell cycle distribution at the indicated times (h) post-6.5 Gy. B, NHEJ DNA repair is a very fast process, and is mostly complete within 2 hrs (blue bar). Checkpoint inhibitors were added from 0-8 h (yellow bar) to encompass DNA repair and S-phase traversal, from 2-24 h (red bar) to encompass S-phase traversal and G 2 exit but not DNA repair, from 16-24 h (green bar) to encompass G 2 exit only, and from 0-24 h (black bar) to encompass DNA repair, S-phase traversal, and G 2 exit. 142 Table 5. 1 : HCT 116 p53"" cell clonogenicity following I R with checkpoint inhibitors. Checkpoint inhibitors fail to radiosensitize when applied from 16-24 h post-IR, but do when applied from 2-24 h post-IR. Cells were plated in 35 mm plates, and 24 h were irradiated (t = 0 h) in the presence of checkpoint inhibitors or not at the times indicated. Following checkpoint inhibitor treatment cells with fresh media were further cultured undisturbed for 10 days. Colonies were counted after staining with malachite green. % Clonogenic survival is relative to mock-irradiated cells. Results shown are means +/- SD (n = 3). Application Drug % Clonogenic Survival % GF-7 (hrs post - positive cells IR) 2 G y 4 G y 6 Gy 6 Gy - 64.5 +/- 5.0 * 17.1 +/- 2.8 2.4 +/- 0.1 23.5 +/- 3.3 16-24 Caffeine 64.7 +/- 5.7 25.9 +/- 1.7 14.0 +/--0.1 82.0 +/- 1.8 Isogranulatimide 64.8 +/- 8.3 27.2 +/- 1.7 11.8+/-- 1.9 47.2 +/- 7.4 0-24 Caffeine 49.6 +/- 4.3 9.5 +/- 0.1 0.1 +/- 0.1 71.6+/- 0.8 Isogranulatimide 60.9 +/- 8.5 18.9+/- 1.5 2.6 +/- 0.1 66.9 +/- 3.0 2-24 Caffeine 33.2+/- 1.7 3.3 +/- 0.1 0.1 +/- 0.1 70.9 +/- 1.1 Isogranulatimide 34.1 +/-3.4 5.9 +/- 4.4 1.2+/- 0.1 66.5 +/- 4.1 0-8 Caffeine 49.9+/- 1.0 12.0+/- 0.1 0.1 +/- 0.1 50.4 +/- 10.2 Isogranulatimide 60.0+/- 0.1 15.4 +/- 3.4 1.2+/- 0.1 18.9+/- 3.5 * S.D. (n = 3) 143 T a b l e 5. 2 : M C F - 7 mp53 cel l c lonogen ic i t y f o l l o w i n g I R w i t h c h e c k p o i n t i n h i b i t o r s . Checkpoint inhibitors fail to radiosensitize when applied from 16-24 h post-IR, but do when applied at earlier times. Cells were assayed as described in Table 5.1. Results shown are means +/- SD (n = 3). Application Drug % Clonogenic Survival % GF-7 (hrs post - positive cells IR) 2 G y 4 G y 6 Gy 6Gy - 65.4 +/- 13.3 " 44.3 +/- 5.8 6.5 +/- 1.5 19.9+/- 4.8 16-24 Caffeine 72.8 +/- 13.7 36.0 +/- 13.7 32.5 +/-6.1 56.2 +/- 5.1 Isogranulatimide 60.3+/- 13.5 41.3+/- 2.2 18.3+/-3.3 35.6+/- 7.3 0-24 Caffeine 36.2 +/- 3.0 5.3 +/- 1.5 2.1 +/-3.0 57.3 +/- 9.4 Isogranulatimide 42.3 +/- 2.7 13.5 +/- 0.1 4.8+/- 1.3 22.4 +/- 2.6 2-24 Caffeine 38.9+/- 13.5 8.4 +/- 0.1 1.1 +/- 1.2 57.3 +/- 5.0 Isogranulatimide 36.1 +/- 0.1 11.5 +/- 0.1 1.6+/-0.1 16.5 +/- 2.3 0-8 Caffeine 39.3 +/- 2.4 6.0 +/- 3.6 1.7+/-0.1 14.7 +/- 1.6 Iso granul atimi de 49.6 +/- 8.5 13.9 +/- 7.3 3.5 +/- 0.1 4.9 +/- 0.8 G 2-synch - 20.9+/- 1.2 9.2 +/- 1.8 4.6 +/- 0.8 5.2 +/- 0.4 0-24 Caffeine 11.5+/- 1.1 5.3 +/- 0.8 >0.1 6.0 +/- 0.6 Isogranulatimide 13.9+/-2.0 10.3+/- 1.3 >0.1 6.1 +/-4.6 * S.D. (n = 3) 144 phase traversal (Figure 5.1). I followed the radiosensitizing effect of caffeine, and isogranulatimide via clonogenic assays. Tables 5.1 and 5.2 list the percentage viable colonies 10 days post-IR relative to mock-IR treated HCT116 p53v" and MCF-7 mp53 cells. At all doses, caffeine effectively radiosensitized the cells when treated from 0-24 h post-IR. Isogranulatimide, however, failed to effectively radiosensitize HCT116 p53-/" cells but did radiosensitize MCF-7 mp53 cells, indicating there may be cell-line specific differences. As mentioned above, no radiosensitization occurs when checkpoint inhibitors are added solely at the time of G 2 checkpoint inhibition. This difference indicates that checkpoint inhibitors may radiosensitize in a manner that is dependent on early D N A repair via NHEJ, or the S-phase traversal in the p53-deficient cells. To test whether D N A repair plays a role in the cell-killing effect of checkpoint inhibitors, they were added from 2-24 h post-IR. Since NHEJ is considered to be a very fast process in which most repair is complete within 90 minutes post-IR (DiBiase et ai, 2000), application of checkpoint inhibitors 2 h post-IR would consist of only S-phase traversal and G 2 checkpoint abrogation in the presence of checkpoint inhibitors, but not the majority of D N A repair (see Figure 5.1). Both caffeine and isogranulatimide were able to effectively radiosensitize both HCT116 p53"/_ and MCF-7 mp53 cells when added during this timeframe (see Tables 5.1 and 5.2). This indicates that checkpoint abrogation at the time of D N A repair is not necessary for radiosensitization to occur. Of note, isogranulatimide was able to increase cell-killing in HCT116 p53_/" cells in this treatment, but was unable to from 0-24 h. G 2 checkpoint abrogation was similar in both treatments by isogranulatimide (Table 5.1, right column), indicating that it is not due to a cell-cycle 145 delay. Perhaps isogranulatimide possesses some antagonistic effect on cell-killing in HCT116 cells when present during D N A repair. When cells were treated with checkpoint inhibitors from 0-8 h post-IR, during D N A repair and S-phase traversal but not G 2 checkpoint abrogation (Figure 5.1, and Tables 5.1 and 5.2, right column), caffeine was able to effectively radiosensitize both cell lines. Isogranulatimide, as with 0-24 h exposure, was unable to radiosensitize HCT116 cells as it is present during D N A repair, but was able to in MCF-7 mp53 cells at higher doses of 4-6 Gy. This finding is entirely unexpected, as it indicates that S-phase traversal alone, not G 2 checkpoint abrogation, is a determining factor in radiosensitization via checkpoint inhibitors. Although the data presented in Tables 5.1 and 5.2 were unexpected, they may be partly explained by the poor specificity of caffeine as a kinase inhibitor, as well as its putative target(s). Caffeine inhibits the PIKK-family kinases A T M and ATR (Sarkaria et al, 1999; Zhou et al, 2000), which have well-documented roles in cell cycle regulation and the S-phase checkpoint. Indeed, ATM-mutant cells exhibit no S-phase delay after irradiation and are highly radiosensitive. With this in mind, it is possible that caffeine can act as a radiosensitizing agent without G 2 checkpoint inhibition, as it can elicit S-phase checkpoint abrogation as well (Zhou et al, 2002). The same cannot be said for isogranulatimide, a relatively-specific inhibitor of Chk l , as radiosensitization took place with 0-8 h treatment only with higher IR doses. Inhibition of the G 2 checkpoint has been proposed as a therapeutic mechanism for years (reviewed in (Anderson et al, 2003; Kawabe, 2004; Prudhomme, 2004)). However, little mention has been made of S-phase during this time. Although checkpoint 146 inhibitor application from 16-24 h post-IR leads to clear checkpoint abrogation, as determined by the % of cells that bypassed checkpoint arrest into mitosis (see Tables 5.1 and 5.2, right panels), it is not necessary for radiosensitization, as 0-8 h application, with little to no G 2 checkpoint bypass, lead to effective radiosensitization. Clearly though, from my findings presented in Tables 5.1 and 5.2, S-phase traversal in combination with G 2 checkpoint abrogation is required for effective radiosensitization to occur. The implications for these findings are two-fold: firstly, should checkpoint inhibitors be utilized in a therapeutic approach, one must be cognizant of the requirement for checkpoint inhibitors to be applied concurrently with, or near the time of, the D N A damaging agent(s) to ensure that checkpoint bypass is elicited in both S-phase and G 2 . Secondly, this points towards a role of S-phase delay in the G 2 checkpoint itself, as S-phase delay inhibition likely prevents the repair of bulky-lesions and creates secondary D N A damage by synthesis in the presence of D N A SSB (see Section 1.2). Primary NHEJ D N A repair following IR is a very quick process (10-30 min) (DiBiase et al, 2000), so it would not be likely that checkpoint inhibitors radiosensitize cells because of a large number of unrepaired D N A DSBs, as these would be repaired before the cell reaches mitosis. However, lesions generated by S-phase traversal itself likely require a lengthier repair period via HRR (DiBiase et al, 2000; Stewart, 2001; Wang et al, 2001a), hence leading to G 2 accumulation following IR. Increased D N A synthesis by a lack of S-phase delay due to caffeine would likely only increase the amount of lesions present, hence leading to a greater therapeutic effect. In an attempt to explore the requirement for S-phase traversal, MCF-7 cells were synchronized to G 2 phase using aphidicolin prior to irradiation. This was accomplished 147 by 24 h exposure of asynchronous cells to 1 uM aphidicolin, then washing twice and allowing cells to resume growth for 6 hours. At this time approximately 80% of cells were in late-S/G2, and were irradiated and treated with checkpoint inhibitors in the same manner as that used above. As shown in Table 5.2, caffeine was only very weakly able to radiosensitize cells, decreasing the viable cell population by about half with doses of 2 and 4 Gy. This is minimal compared to the nearly ten-fold increase in cell killing observed with caffeine treatment after 4 Gy for the same drug exposure (0-24 h post-IR). Unfortunately, aphidicolin and other synchronization agents are also mild D N A damaging agents (Kurose et al, 2006). Indeed, the survival of G2-synchronized cells after IR is much lower than that of asynchronously dividing cells. Whether this is a result of inherent sensitivity of G 2 cells to D N A damage, or a cumulative effect of synchronization-related D N A damage coupled to ionizing radiation remains to be tested. The use of centrifugal elutriation to prepare G2-synchronized cells without having to use aphidicolin may help answer this question. Therefore, the nature of the proposed S-phase traversal cannot be tested conclusively, although the data presented in Tables 5.1 and 5.2 indicates it may play a significant role. 5.2.2 Isogranulatimide bypasses the G 2 checkpoint in vivo I next focussed on isogranulatimide and its potential as a therapeutic agent. When tested against a panel of kinases in vitro, isogranulatimide exhibited very selective inhibition of Chk l , especially when compared to UCN-01, which, although it is a more potent inhibitor of Chk l , is an even more potent P K C inhibitor; while isogranulatimide does not inhibit P K C (Jiang et al, 2004). 148 I first asked whether isogranulatimide could act as a checkpoint inhibitor in vivo. Subcutaneous HCT116 tumours were implanted in scid mice as described in Materials and Methods (Chapter 2). Mice were transported to the Gammacell 220 in Tupperware containers and irradiated with 6.5 Gy, a sublethal dose (H.-S. Teh, personal communication). Just prior to irradiation, the xenoplanted tumours were injected intra-tumourally with 50 ul , isogranulatimide solution (1% DMSO, 10 m M isogranulatimide in sterile PBS) or carrier (1 % DMSO in PBS). This dose of 6 mg/kg yields a very high tumoural isogranulatimide concentration of approximately 2 mM. As determined in vitro above, 24 hour checkpoint inhibitor application at the time of irradiation offers the strongest checkpoint abrogation, so determining whether checkpoint inhibition takes place in vivo can provide the foundation for further development as a therapeutic approach. Figure 5.2 summarizes the cell cycle distribution of tumour cells recovered from animals after different treatments. HCT116 p53 + / + tumours exposed to 6.5 Gy IR did not exhibit a strong increase in G 2 / M 24 h after IR, as expected, with 12.6 +/- 1.6 % cells in G 2 / M prior to IR, and 14.3 +/- 1.4 % cells after IR (p = 0.231). Isogranulatimide had no effect on the G 2 / M population, with 13.0 +/- 1.6 % cells in mitosis (p = 0.289). Cell cycle arrest was evident though, as after IR S-phase cells were no longer detectable, indicating Gj/S arrest. p53 ~'~ cells, however, did exhibit G 2 / M arrest after IR, with 8.8 +/- 0.4 % cells in G 2 / M prior to IR, and 21.9 + 1.1 % cells after IR (p < 0.001). Isogranulatimide was able to significantly decrease the G 2 / M population to 15.5 +/- 3.3 % (p < 0.05), indicating that isogranulatimide is indeed capable of G 2 checkpoint abrogation in vivo. These results are also summarized in Table 5.3. 149 No IR IR + DMSO IR + IGR p53 p53-A 0 BN L L Figure 5.2 : Cell cycle distribution of cells extracted from subcutaneous tumours in scid mice. HCT116 p53 "/J" or p53_/" cells were implanted on the backs of scid mice as described in Chapter 2.19. When tumours reached 200 mm3, mice were irradiated and injected intra-tumourally with DMSO or isogranulatimide (IGR) solution, then extracted and analyzed by flow cytometry 24 h later. p53 + / + cells did not exhibit a strong G 2 arrest, nor a response to isogranulatimide, as expected, whereas the p53 "'" cells exhibited a G 2 arrest which was decreased by isogranulatimide injection. % G 2 / M population is quantified in Table 5.3. 150 Table 5. 3 : Summary of G2/M population from cell suspensions of subcutaneously implanted HCT116 cells in scid mice. Isogranulatimide lead to a decreased %G2/M population after IR compared to DMSO alone. Results shown are means +/- SD (n > 3). Treatment No IR DMSO Isogranulatimide HCT116p53+/+ 12.6 ± 1.6% 14.3 ± 1 . 4 % 13.0 ± 1.6% HCT116P537 8.8 ± 0.4 % 21.9 ± 1.1 % 15.5 ±3.3 % 151 5.3 DISCUSSION The observation that checkpoint inhibitor application from 16-24 h post-IR fails to radiosensitize cells, even though it causes a large proportion of cells to undergo mitosis is entirely unexpected. The process of checkpoint inhibition in these cells may simply be providing the necessary release to resume cell cycle progression, even though D N A damage is sub-lethal, or may reflect a temporal phenomenon, in which those cells responding to a checkpoint inhibitor would have resumed cell cycle progression normally, albeit at much later times (post-24 h). As mentioned above, the implications from this observation can directly impact the proposed usage of checkpoint inhibitors as radiosensitizing agents in the treatment of cancer. In order to obtain a therapeutic effect, the checkpoint inhibitors should likely be present at the time of D N A damage in order to ensure cell cycle progression with damaged DNA. If the time between D N A damage and checkpoint inhibitor application is too great, there may be no increase in cell killing, but the patient may still suffer from whatever side effects the checkpoint inhibitor may cause. Further development of isogranulatimide as a therapeutic agent was delayed due to limiting amounts of compound. However, some issues were noted in my limited number of in vivo studies. Primarily, a major concern is its formulation. Isogranulatimide is poorly soluble in aqueous solutions, so stock compound must be maintained in 100% DMSO, in which it is still primarily a suspension. Pure DMSO damages the tail vein of the mouse, making intravenous drug delivery impossible. Therefore, intra-tumoural injection was used instead. Consequently, others in the Roberge laboratory have empirically tested several other formulations for isogranulatimide intravenous delivery. 152 Several different formulations have been tested thus far. The first was 15 mg/mL solution of isogranulatimide in 60% DMSO, 30% ethanol, 8% PBS pH 7.4 and 2% dextrose. To achieve a dose of 30 mg/kg in a mouse, 50 pL were injected intravenously. Unfortunately, the isogranulatimide immediately precipitated upon entering the vein. Similarly, 50 uL of a 10 mg/mL solution in 25% DMSO, 56% ethanol, 15% PBS and 4% dextrose mixture was injected to achieve a dose of 13 mg/kg, but it also precipitated immediately upon entering the tail vein of the mouse. However, when 100 uL of the same formulation, but at 5 mg/mL, was injected the drug entered the circulation but the mouse exhibited signs of intoxication, presumably from the high ethanol content in the carrier solution. This work was performed by Jing L i and Pamela Austin of the Roberge laboratory. Although the latter formulation may seem to offer potential for further in vivo studies, little is known about its toxicity to mice at this dose, or whether this dose is sufficient to bypass the G 2 checkpoint. New batches of isogranulatimide should permit for others in the Roberge laboratory to test new formulations. One attractive option is the possibility of loading isogranulatimide into liposomes for drug delivery. Liposomal formulations of D N A damaging agents allow for increased activity and reduced toxicity to healthy tissues compared to the use of free drug injection (reviewed in (Hofheinz et al, 2005)). Basically, drug is encapsulated in a lipid bilayer sphere which passively circulates the body in the blood until exiting the bloodstream at the sites of tumours due to their poorly formed and permeable vasculature. They then slowly leak out their contents directly at the site of the tumour, allowing for exceptionally high, localized doses of cytotoxic drug. Drug solubility is not necessarily a cause for concern 153 with liposomes, as drug can crystallize inside the liposome until release at the tumour site. Therefore, I would consider liposomes promising for localized isogranulatimide delivery, as it may bypass concerns of its toxicity, and limiting amounts of drug, as the injected amount of drug would not necessarily need to be as high as free drug. Dr. Michael Johnston in the Cullis laboratory at U B C has recently been successful in loading stably isogranulatimide into liposomes using an ionophore-mediated pH gradient loading method (summarized in Fig. 5.3) (Fenske et al, 1998). Briefly, liposomes are extruded through a small pore filter in 300 m M MgCb, pH 7.0. After washing the liposomes by dialysis, they are mixed with the ionophore A23187, EDTA (3-15 mM) and isogranulatimide. Isogranulatimide, because of its non-polar nature, freely diffuses across the lipid bilayer of the liposome. However, the ionophore A23187 allows for exchange of M g 2 + ions for 2 H + , slowly causing the interior of the liposome to become acidic. The isogranulatimide inside the liposome then becomes protonated, preventing it from escaping the liposome. It will be of great interest to test whether this formulation can also lead to G 2 checkpoint abrogation in the same manner as that observed by free-drug intra-tumoural injection. This work wil l be performed by Pamela Austin in the Roberge laboratory after completion of my degree. 5.4 S U M M A R Y Checkpoint inhibitors represent a promising class of therapeutic agents in the treatment of cancer. Current models suggest that radiosensitization caused by checkpoint 154 B 3 <3 f a e a o s> o " |Ad(! Drug (D). EDTA ( J). laodA.23187 ($j H 3 H OH" H* "3 DH t i " -D * B" Figure 5.3 : General overv iew of l iposomal loading of isogranulat imide. A, isogranulatimide can be protonated on N15. B, ionophore-mediated pH gradient loading of drug. Liposomes formed in a divalent cation media, such as MgCl 2 , are then mixed with the ionophore A23187, EDTA and the drug isogranulatimide. Isogranulatimide freely diffuses across the membrane, until protonated by the newly-formed acidic interior due to A23187-mediated exchange of M g 2 + for 2H + . Figure adapted from (Fenske et al., 1998) 155 inhibitors is achieved through abrogation of the G 2 / M D N A damage checkpoint. In this chapter I demonstrate that G 2 checkpoint abrogation alone is insufficient for radiosensitization. Instead, in vitro, checkpoint inhibitors must be present for the duration of S-phase progression to achieve radiosensitization, with G 2 checkpoint abrogation actually being unnecessary in some cell lines. This work, comprising Figure 5.1 and Tables 5.1 to 5.3, has been submitted for publication (Sturgeon and Roberge, 2006d). Once an appropriate dosing schedule was determined in vitro, I asked whether isogranulatimide, the Roberge laboratory's most therapeutically promising checkpoint inhibitor to-date, abrogates the checkpoint in vivo. The conclusion to this work is awaiting successful loading of isogranulatimide into liposomes in collaboration with Pieter Cullis (Dept. of Biochemistry and Molecular Biology, UBC). This will be important in the clinical development of checkpoint inhibitors as a therapeutic approach. 156 C h a p t e r 6 Future perspectives M y research in this thesis has specifically dealt with the use of checkpoint inhibitors as potential therapeutic anti-cancer agents. In Chapter 3 I present two new checkpoint inhibitors, psilostachyin and cryptofolione. Identifying new checkpoint inhibitors, in both industrial and academic settings, routinely focusses on identifying inhibitors of A T M , Chkl or Ser/Thr phosphatases. However, I show that psilostachyin may inhibit Weel, a rare target for checkpoint inhibition, while cryptofolione does not inhibit any kinases or phosphatases associated with the G 2 checkpoint. Although both exhibit toxicity towards cells, making them unlikely therapeutic agents, the finding that inhibition of nuclear export may cause checkpoint abrogation provides new clues to the mechanism(s) behind cell cycle arrest, and will hopefully stimulate the search for new nuclear export inhibitors that are less toxic to cells. One additional aspect of cryptofolione remains exciting for future study: until its isolation and identification as a checkpoint inhibitor, the most efficacious checkpoint inhibitor identified to-date has been caffeine and its analogs, causing an average of 60% of cells to enter mitosis after a range of D N A damage. Although in some cell lines U C N -01 nearly rivals caffeine in efficacy, no other compound does. Cryptofolione, on the other hand, can cause greater than 60% of cells to enter mitosis in the same cell line, and in others greater than 70% (data not shown). If possible, the identification of an analog 157 of cryptofolione with reduced toxicity may stimulate the field immensely, leading to a novel therapeutic agent. The observation that leptomycin B acts as a checkpoint inhibitor is particularly surprising, as leptomycin B is commonly believed to be an extremely specific inhibitor of nuclear export with no other cellular activities. If we assume that indeed leptomycin B does not inhibit any other cellular activities than Crml -mediated nuclear export, this unequivocally demonstrates that nuclear export is required for checkpoint function. Current checkpoint models are unclear as to the requirement for nuclear export of cyclin B I , and the CDC25 family phosphatases with 14-3-3. Perhaps the use of the cryptofoliones or leptomycin B can help elucidate the exact role of nuclear export and checkpoint abrogation, as genetic studies thus far have been unable to resolve the issue. Nuclear export inhibition as a means to achieve checkpoint inhibition demonstrates that there are likely many different paradigms than solely kinase/phosphatase inhibition to achieve checkpoint abrogation. Chapter 4 specifically deals with DNA-PK, a key protein involved in D N A repair, and the effect its inhibition has on cell cycle checkpoints. When I initially began these studies, it was unclear as to whether D N A - P K functioned as a cell-cycle regulating PIKK in a similar manner as A T M . Indeed, D N A - P K is capable of phosphorylating p53, RPA1, y - H 2 A X and others (as discussed in Chapter 4). Adding to the confusion was the observation that Ku86-deficient cells arrest permanently in G 2 after D N A damage in the same manner as A T cells, as well as my own observations presented in Chapter 4. Surprisingly, inhibiting not only D N A - P K but D N A repair in general results in cancer cells entering a very strong cell cycle arrest from which they cannot escape after D N A 158 damage, even in the presence of a checkpoint inhibitor. This finding is very significant, as it suggests that the degree of D N A damage correlates to the strength of the checkpoint, with a maximum tolerance threshold before irreversible cell cycle arrest. A role for other PIKKs such as A T M in D N A repair is slowly emerging. Perhaps it is not the cell cycle regulatory activity of A T M that causes AT cells to exhibit similar checkpoint defects, but rather their lack of a required D N A repair activity. My findings also demonstrates 2 unexpected weaknesses in checkpoint inhibitors, in that they are limited in their ability to cause checkpoint abrogation. Firstly, high levels of D N A damage block the inhibitors from functioning. Whether this is due to excessive damage physically preventing the mitotic machinery from functioning, or from a grossly overactivated Chkl remains to be seen. Genetic studies involving overexpression of kinase-dead Chkl or R N A i silencing may offer some resolution to this issue. Secondly, cancer cells typically display a wide-range of mutations with variable expression levels of many proteins, including D N A repair proteins. Barring further advances in the field of pharmacogenetics, checkpoint inhibitor use may be limited to a subset of tumours known to have no D N A repair defects. Traditional anticancer treatments rely on the eradication of all oncogenic cells. However, further study of this permanent cell cycle arrest, or senescence of cancerous cells could be prove to be equally effective. A growth-arrested cancerous cell is no longer proliferating, effectively halting tumour growth within the body. Since the lethality of most tumours is not from their presence, but from their expansion and metastasis, this could be an alternative form of therapy. Additionally, understanding the mechanism behind this stronger cell cycle arrest could allow for appropriately designed 159 "second-generation" checkpoint inhibitors to increase radiation killing in these growth-arrested cells. Although a role for M K K 6 or P K B in the checkpoint-defective M059J cells was not demonstrated, the studies I have presented may stimulate the study of the checkpoint. The general model of G2 checkpoint initiation and maintenance outlined in Chapter 1 may be insufficient. Cell cycle control and the checkpoints should no longer be considered isolated fields; my studies in Chapter 4 may be the starting point to further uncovering the complexities behind checkpoint function and the many proteins which contribute to arrest or checkpoint recovery. Finally, Chapter 5 describes my attempts at advancing the preclinical status of isogranulatimide. In attempting to find an appropriate dosing schedule for optimal in vivo checkpoint abrogation and radiosensitization, it was revealed that radiosensitization achieved by checkpoint inhibitors is not solely dependent on G2 checkpoint abrogation, but also their action during S-phase progression. This suggests that the intra-S checkpoint, or more specifically Chkl activity in S-phase, is equally, i f not more, important for abrogation than the G 2 / M checkpoint to achieve radiosensitization. Such a conclusion is surprising, as checkpoint inhibitors have been considered to radiosensitize cells via G 2 checkpoint abrogation for years. These observations may also correlate with those presented in Chapter 4, in that checkpoint abrogation by checkpoint inhibitors is limited by the degree of damage present. Also, synchronization studies in Chapter 4 revealed that DNA-repair-deficient cells do not undergo prolonged arrest with D N A damage suffered in G2. This further demonstrates a critical role in S-phase progression with relevance to the G2 checkpoint; 160 S-phase traversal in the presence of D N A damage may lead to secondary lesions which are fatal to the cell. Checkpoint abrogation during what would normally be a slow-down of D N A synthesis may increase the frequency or duration of these lesions, leading to increases in cell death. Testing of isogranulatimide in mice revealed that the compound has G 2 checkpoint inhibition activity in vivo. Although this may not seem to be a significant observation, to-date no other checkpoint inhibitor has demonstrated checkpoint abrogation in vivo. UCN-01 has been utilized in several clinical studies, but these were focussed on its PKC-inhibitory properties. When it was used as a checkpoint inhibitor, G 2 checkpoint abrogation itself was not monitored, only an increase in cell killing compared to drug treatment alone. The major hurdle facing the medicinal development of isogranulatimide is its formulation due to poor solubility in aqueous solutions. 161 Chapter 7 References Abraham, R. T. (2001). Cell cycle checkpoint signaling through the A T M and ATR kinases. Genes & Development 15(17): 2177-2196. Ahn, J. Y. , J. K. Schwarz, H . Piwnica-Worms and C. E. Canman (2000). Threonine 68 phosphorylation by ataxia telangiectasia mutated is required for efficient activation of Chk2 in response to ionizing radiation. Cancer Research 60(21): 5934-5936. Alberts, B., A . Johnson, J. Lewis, M . Raff, K. Roberts and P. Walter (2002). Molecular Biology of the Cell. New York, Garland Science: 1463. Allalunis-Turner, J., G. M . Barron and I. Rufus S. Day (1997). Intact G2-phase checkpoint in cells of a human cell line lacking DNA-Dependent Protein Kinase activity. Radiation Research 147(3): 284-287. Allalunis-Turner, M . J., G. M . Barron, I. R.S. Day, K. D. Dobler and R. Mirzayans (1993). Isolation of two cell lines from a human malignant glioma specimen differing in sensitivity to radiation and chemotherapeutic drugs. Radiation Research 134(3): 349-354. Allen, C , J. Halbrook and J. A. Nickoloff (2003). Interactive competition between homologous recombination and non-homologous end joining. Molecular Cancer Research 1(12): 913-920. Allen, C , A. Kurimasa, M . A. Brenneman, D. J. Chen and J. A. Nickoloff (2002). DNA-dependent protein kinase suppresses double-strand break-induced and spontaneous homologous recombination. Proceedings of the National Academy of Sciences of the United States of America 99(6): 3758-3763. Anderson, C. W. and M . J. Allalunis-Turner (2000). Human TP53 from the malignant glioma-derived cell lines M059J and M059K has a cancer-associated mutation in exon 8. Radiation Research 154(4): 473-476. Anderson, H. J., R. J. Andersen and M . Roberge (2003). Inhibitors of the G 2 D N A damage checkpoint and their potential for cancer therapy. Progress in Cell Cycle Research. L. Meijer, A . Jezequel and M . Roberge. Roscoff, France, Editions "Life in Progress". 5: 423-430. Anderson, H. J., J. E. Coleman, R. J. Andersen and M . Roberge (1997). Cytotoxic peptides hemiasterlin, hemiasterlin A and hemiasterlin B induce mitotic arrest and abnormal spindle formation. Cancer Chemotherapy and Pharmacology 39(3): 223-226. Andreassen, P. R. and R. L. Margolis (1992). 2-Aminopurine overrides multiple cell cycle checkpoints in B H K cells. Proceedings of the National Academy of Sciences of the United States of America 89(6): 2272-2276. Arnaud, M . J. (1987). The pharmacology of caffeine. Progress in Drug Research 31(1): 273-313. 162 Baetz, K. , L. McHardy, K. Gable, T. Tarling, D. Reberioux, J. Bryan, et al. (2004). Yeast genome-wide drug-induced haploinsufficiency screen to determine drug mode of action. Proceedings of the National Academy of Sciences of the United States of America 101(13): 4525-4530. Bakkenist, C. J. and M . B. Kastan (2003). D N A damage activates A T M through intermolecular autophosphorylation and dimer dissociation. Nature 421(6922): 499-506. Bao, S., M . S. Chang, D. Auclair, Y . Sun, Y . Wang, W. K. Wong, et al. (1999). HRadl7, a human homologue of the Schizosaccharomyces pombe checkpoint gene radl7, is overexpressed in colon carcinoma. Cancer Research 59(9): 2023-2028. Bao, S., R. S. Tibbetts, K. M . Brumbaugh, Y . Fang, D. A. Richardson, A. A l i , et al. (2001). ATR/ ATM-mediated phosphorylation of human Radl7 is required for genotoxic stress responses. Nature 411(6840): 969-974. Blunt, T., N . J. Finnie, G. E. Taccioli, G. C. Smith, J. Demengeot, T. M . Gottlieb, et al. (1995). Defective DNA-dependent protein kinase activity is linked to V(D)J recombination and D N A repair defects associated with the murine scid mutation. Cell 80(5): 813-823. Bocca, C , L. Gabriel, F. Bozzo and A. Miglietta (2004). A sesquiterpene lactone, costunolide, interacts with microtubule protein and inhibits the growth of MCF-7 cells. Chemico-Biological Interactions 147(1): 79-86. Bork, P. M . , M . L. Schmitz, M . Kuhnt, C. Escher and M . Heinrich (1997). Sesquiterpene lactone containing Mexican Indian medicinal plants and pure sesquiterpene lactones as potent inhibitors of transcription factor NF-kappaB. FEBS Letters 402(1): 85-90. Braca, A. , A . Armenise, I. Morelli, J. Mendez, Q. M i , H. B. Chai, et al. (2004). Structure of kaurane-type diterpenes from Parinari sprucei and their potential anticancer activity. Planta Medica 70(6): 540-550. Britton, R., M . Roberge, C.Brown, R. v. Soest and R. J. Andersen (2003). New okadaic acid analogues from the marine sponge Merriamum oxeato and their effect on mitosis. Journal of Natural Products 66(6): 838-843. Brown, K. D., K. W. Wood and D. W. Cleveland (1996). The kinesin-like protein CENP-E is kinetochore-associated throughout poleward chromosome segregation during anaphase-A. Journal of Cell Science 109(Pt 5): 961-969. Bryans, M . , M . C. Valenzano and T. D. Stamato (1999). Absence of D N A ligase IV protein in XR-1 cells: evidence for stabilization by XRCC4. Mutation Research 433(1): 53-58. Carson, C. T., R. A . Schwartz, T. H. Stracker, C. E. Lilley, D. V . Lee and M . D. Weitzman (2003). The Mrel 1 complex is required for A T M activation and the G 2 / M checkpoint. EMBO Journal 22(24): 6610-6620. Chan, D. W., B. P.-C. Chen, S. Prithivirajsingh, A. Kurimasa, M . D. Story, J. Qin and D. J. Chen (2002). Autophosphorylation of the DNA-dependent protein kinase catalytic subunit is required for rejoining of D N A double-strand breaks. Genes & Development 16(18): 2333-2338. Chan, D. W., D. P. Gately, S. Urban, A. M . Galloway, S. P. Lees-Miller, T. Yen and J. A. Turner (1998). Lack of correlation between A T M protein expression and tumour cell radiosensitivity. International Journal of Radiation Biology 74(2): 217-224. 163 Chan, D. W. and S. P. Lees-Miller (1996). The DNA-dependent Protein Kinase is inactivated by autophosphorylation of the catalytic subunit. Journal of Biological Chemistry 271(15): 8936-8941. Chaturvedi, P., W. K. Eng, Y . Zhu, M . R. Mattern, R. Mishra, M . R. Hurle, et al. (1999). Mammalian Chk2 is a downstream effector of the ATM-dependent D N A damage checkpoint pathway. Oncogene 18(28): 4047-4054. Chen, M . J., Y . T. Lin, H . B. Lieberman, G. Chen and E. Y . Lee (2001). ATM-dependent phosphorylation of human Rad9 is required for ionizing radiation-induced checkpoint activation. Journal of Biological Chemistry 276(19): 16580-16586. Chen, Z., Z. Xiao, J. Chen, S.-C. Ng, T. Sowin, H. Sham, et al. (2003). Human Chkl expression is dispensable for somatic cell death and critical for sustaining G 2 D N A damage checkpoint. Molecular Cancer Therapeutics 2(6): 543-548. Cinel, B. (2001). Biologically active secondary metabolites isolated from marine and terrestrial sources. Chemistry. Vancouver, University of British Columbia: 262. Cooke, C. A. , B. Schaar, T. J. Yen and W. C. Earnshaw (1997). Localization of CENP-E in the fibrous corona and outer plate of mammalian kinetochores from prometaphase through anaphase. Chromasoma 106(7): 446-455. Cortez, D., S. Guntuku, J. Qin and S. J. Elledge (2001). ATR and ATRIP: partners in checkpoint signaling. Science 294(5547): 1713-1716. Craig, K. S. (2003). Terpenoids from marine and terrestrial sources. Chemistry. Vancouver, University of British Columbia: 240. Culotti, J. and L. H . Hartwell (1971). Genetic control of the cell division cycle in yeast. 3. Seven genes controlling nuclear division. Experimental Cell Research 67(2): 389-401. Curman, D., B. Cinel, D. E. Williams, N . Rundle, W. D. Block, A. A . Goodarzi, et al. (2001). Inhibition of the G 2 D N A damage checkpoint and of protein kinases Chkl and Chk2 by the marine sponge alkaloid debromohymenialdisine. Journal of Biological Chemistry 276(21): 17914-17919. Dalai, S. N . , C. M . Schweitzer, J. Gan and J. A . DeCaprio (1999). Cytoplasmic localization of human cdc25C during interphase requires an intact 14-3-3 binding site. Molecular and Cellular Biology 19(6): 4465-4479. Davezac, N . , V . Baldin, B. Gabrielli, A. Forrest, N . Theis-Febvre, M . Yashida and B. Ducommun (2000). Regulation of CDC25B phosphatases subcellular localization. Oncogene 19(18): 2179-2185. Deming, P. B., C. A. Cistulli, H . Zhao, P. R. Graves, H. Piwnica-Worms, R. S. Paules, et al. (2001). The human decatenation checkpoint. Proceedings of the National Academy of Sciences of the United States of America 98(21): 12044-12049. DiBiase, S. J., Z.-C. Zeng, R. Chen, T. Hyslop, J. Walter J. Curran and G. Iliakis (2000). DNA-dependent Protein Kinase stimulates an independently active, nonhomologous, end-joining apparatus. Cancer Research 60(5): 1245-1253. Dimri, G. P., X . Lee, G. Basile, M . Acosta, G. Scott, C. Roskelley, et al. (1995). A biomarker that identifies senescent human cells in culture and in aging skin in vivo. Proceedings of the National Academy of Sciences of the United States of America 92(20): 9363-9367. Ding, Q., Y . V . R. Reddy, W. Wang, T. Woods, P. Douglas, D, A . Ramsden, et al. (2003). Autophosphorylation of the catalytic subunit of the DNA-dependent 164 Protein Kinase is required for efficient end processing during D N A double-strand break repair. Molecular and Cellular Biology 23(16): 5836-5848. Dong, L. Q. and F. Liu (2005). PDK2: the missing piece in the receptor tyrosine kinase signaling pathway puzzle. American Journal of Physiology, Endocrinology and Metabolism 289(2): El87-196. Dumontet, V. , N . V . Hung, M . T. Adeline, C. Riche, A. Chiaroni, T. Sevenet and F. Gueritte (2004). Cytotoxic Flavonoids and a-Pyrones from Cryptocarya obovata. Journal of Natural Products 67(5): 858-862. Eastman, A. , E. A. Kohn, M . K . Brown, J. Rathman, M . Livingstone, D. H. Blank and G. W. Gribble (2002). A novel indolocarbazole, ICP-1, abrogates D N A damage-induced cell cycle arrest and enhances cytotoxicity: similarities and differences to the cell cycle checkpoint abrogator UCN-01. Molecular Cancer Therapeutics 1(12): 1067-1078. Eriksson, A. , R. Lewensoh, R. Larsson and A. Nilsson (2002). DNA-dependent protein kinase in leukaemia cells and correlation with drug sensitivity. Anticancer Research 22(3): 1787-1793. Evans, J. W., X . F. Liu, C. U . Kirchgessner and J. M . Brown (1996). Induction and repair of chromosome aberrations in scid cells measured by premature chromosome condensation. Radiation Research 145(1): 39-46. Evans, T., E. T. Rosenthal, J. Youngblom, D. Distel and T. Hunt (1983). Cyclin: a protein specified by maternal mRNA in sea urchin eggs that is destroyed at each cleavage division. Cell 33(2): 389-396. Fagagna, F. d. A . d., P. M . Reaper, L. Clay-Farrace, H . Fiegler, P. Carr, T. V . Zglinicki, et al. (2003). A D N A damage checkpoint response in telomere-initiated senescence. Nature 426(6963): 194-198. Falck, J., N . Mailand, R. G. Syljuasen, J. Bartek and J. Lukas (2001). The ATM-Chk2-Cdc25A checkpoint pathway guards against radioresistant D N A synthesis. Nature 410(6830): 842-847. Fan, S., M . L. Smith, D. J. R. II, D. Duba, Q. Zhan, K. W. Kohn, J. Albert J. Fornace and P. M . O'Connor (1995). Disruption of p53 function sensitizes breast cancer MCF-7 cells to cisplatin and pentoxifylline. Cancer Research 55(8): 1649-1654. Farjot, G., A. Sergeant and I. Mikaelian (1999). A new nucleoporin-like protein interacts with both HrV-1 Rev nuclear export signal and CRM-1. Journal of Biological Chemistry 274(24): 17309-17317. Feng, J., J. Park, P. Cron, D. Hess and B. A. Hemmings (2004). Identification of a PKB/Akt hydrophobic motif Ser-473 kinase as DNA-dependent protein kinase. Journal of Biological Chemistry 279(39): 41189-41196. Fenske, D. B. , K . F. Wong, E. Maurer, N . Maurer, J. M . Leenhouts, N . Boman, L. Amankwa and P. R. Cullis (1998). Ionophore-mediated uptake of ciprofloxacin and vincristine into large unilamellar vesicles exhibiting transmembrane ion gradients. Biochimica et Biophysica Acta 1414(1-2): 188-204. Fesquet, D., J. C. Labbe, J. Derancourt, J. P. Capony, S. Galas, F. Girard, et al. (1993). The M O 15 gene encodes the catalytic subunit of a protein kinase that activates cdc2 and other cyclin-dependent kinases (CDKs) through phosphorylation of Thrl61 and its homologues. EMBO Journal 12(8): 3111-3121. 165 Flatt, P. M . and J. A . Pientenpol (2000). Mechanisms of cell-cycle checkpoints: at the crossroads of carcinogenesis and drug discovery. Drug Metabolism Reviews 32(3-4): 283-305. Friesland, S., L. Kanter-Lewensohn, R. Tell, E. Munck-Wikland, R. Lewensohn and A. Nilsson (2003). Expression of Ku86 confers favorable outcome of tonsillar carcinoma treated with radiotherapy. Head and Neck 25(4): 313-321. Fuse, E., H. Tanii, K. Takai, K. Asanome, N . Kurata, H. Kobayashi, et al. (1999). Altered pharmacokinetics of a novel anticancer drug, UCN-01, caused by specific high affinity binding to alphal-acid glycoprotein in humans. Cancer Research 59(5): 1054-1060. Garber, K. (2005). New checkpoint blockers begin human trials. Journal of the National Cancer Institute 97(14): 1026-1028. Gately, D. P., J. C. Hittle, G. K. T. Chan and T. J. Chen (1998). Characterization of A T M expression, localization, and associated DNA-dependent Protein Kinase activity. Molecular Biology of the Cell 9(9): 2361-2374. Goldstein, S. (1990). Replicative senescence: the human fibroblast comes of age. Science 249(4973): 1129-1133. Gonzalez, I., G. Tripathi, E. J. Carter, L. J. Cobb, D. A. Salih, F. A . Lovett, et al. (2004). Akt2, a novel functional link between p38 mitogen-activated protein kinase and phosphatidylinositol 3-kinase pathways in myogenesis. Molecular and Cellular Biology 24(9): 3607-3622. Graves, P. R., C. M . Lovly, G. L. Uy and H. Piwnica-Worms (2001). Localization of human Cdc25C is regulated both by nuclear export and 14-3-3 protein binding. Oncogene 20(15): 1839-1851. Graves, P. R., L. Yu, J. K. Schwarz, J. Gales, E. A. Sausville, P. M . O'Connor and H. Piwnica-Worms (2000). The Chkl protein kinase and the Cdc25C regulatory pathways are targets of the anticancer agent UCN-01. Journal of Biological Chemistry 27'5(8): 5600-5605. Griffiths, D. J., N . C. Barbet, S. McCready, A. R. Lehmann and A. M . Carr (1995). Fission yeast radl7: a homologue of budding yeast RAD24 that shares regions of sequence similarity with D N A polymerase accessory proteins. EMBO Journal 14(23): 5812-5823. Hadfield, J. A. , S. Ducki, N . Hirst and A. T. McGown (2003). Tubulin and microtubules as targets for anticancer drugs. Progress in Cell Cycle Research. L. Meijer, A. Jezequel and M . Roberge. Roscoff, France, Editions "Life in Progress". 5: 309-325. Haggarty, S. J., K . M . Koeller, T. R. Kau, P. A. Silver, M . Roberge and S. L. Schreiber (2003). Small molecule modulation of the human chromatid decatenation checkpoint. Chemistry and Biology 10(12): 1267-1279. Hagting, A. , C. Karlsson, P. Clute, M . Jackman and J. Pines (1998). M P F localization is controlled by nuclear export. EMBO Journal 17(14): 4127-4138. Hartwell, L., T. Weinert, L. Kadyk and B. Garvik (1994). Cell cycle checkpoints, genomic integrity, and cancer. Cold Spring Harbor Symposia on Quantitative Biology 59(1): 259-263. 166 Hartwell, L. H . (1971). Genetic control of the cell division cycle in yeast. II. Genes controlling D N A replication and its initiation. Journal of Molecular Biology 59(1): 183-194. Hartwell, L. H. (1971). Genetic control of the cell division cycle in yeast. IV. Genes controlling bud emergence and cytokinesis. Experimental Cell Research 69(2): 265-276. Hartwell, L. H. , J. Culotti and B. Reid (1970). Genetic control of the cell-division cycle in yeast. I. Detection of mutants. Proceedings of the National Academy of Sciences of the United States of America 66(2): 352-359. Hartwell, L. H. and T. A . Weinert (1989). Checkpoints: controls that ensure the order of cell cycle events. Science 246(4930): 629-634. Hayflick, L. (1965). The limited in vitro lifetime of human diploid cell strains. Experimental Cell Research 37(3): 614-636 Hayflick, L. and P. S. Moorhead (1961). The serial cultivation of human diploid cell strains. Experimental Cell Research 25(1): 585-621. Hefferin, M . L. and A. E. Tomkinson (2005). Mechanism of D N A double-strand break repair by non-homologous end joining. DNA Repair 4(6): 639-648. Helt, C. E., W. A. Cliby, P. C. Keng, R. A. Bambara and M . A. O'Reilly (2005). Ataxia telangiectasia mutated (ATM) and A T M and Rad3-related protein exhibit selective target specificities in response to different forms of D N A damage. Journal of Biological Chemistry 280(2): 1186-1192. Hereford, L. M . and L. H . Hartwell (1971). Defective D N A synthesis in permeabilized yeast mutants. Nature New Biology 234(49): 171-172. Hirao, A., Y . Y . Kong, S. Matsuoka, A. Wakeham, J. Ruland, H . Yoshida, et al. (2000). D N A damage-induced activation of p53 by the checkpoint kinase Chk2. Science 287(5459): 1824-1827. Hofheinz, R. D., S. U . Gnad-Vogt, U . Beyer and A. Hochhaus (2005). Liposomal encapsulated anti-cancer drugs. Anticancer Drugs 16(7): 691-707. Hollstein, M . , D. Sidransky, B. Vogelstein and C. C. Harris (1991). p53 mutations in human cancers. Science 253(5015): 49-53. Hoppe, B. S., R. B. Jensen and C. U . Kirchgessner (2000). Complementation of the Radiosensitive M059J Cell Line. Radiation Research 153(2): 125-130. Hresko, R. C. and M . Mueckler (2005). mTOR.RICTOR is the Ser473 kinase for Akt/protein kinase B in 3T3-L1 adipocytes. Journal of Biological Chemistry 280(49): 40406-40416. Iliakis, G., H. Wang, A . R. Perrault, W. Boecker, B. Rosidi, F. Windhofer, et al. (2004). Mechanisms of D N A double strand break repair and chromosome aberration formation. Cytogenetic and Genome Research 104(1-4): 14-20. Ismail, I. H. , S. Martensson, D. Moshinsky, A . Rice, C. Tang, A . Howlett, G. McMahon and O. Hammarsten (2003). SU11752 inhibits the DNA-dependent protein kinase and D N A double-strand break repair resulting in ionizing radiation sensitization. Oncogene 24(4): 873-882. Jackson, J. R., A. Gilmartin, C. Imburgia, J. D. Winkler, L. A . Marshall and A. R. A (2000). An indolocarbazole inhibitor of human checkpoint kinase (Chkl) abrogates cell cycle arrest caused by D N A damage. Cancer Research 60(3): 566-572. 167 Jiang, X . , L. Y . Lim, J. W. Daly, A. H. L i , K. A . Jacobson and M . Roberge (2000). Structure-activity relationships for G 2 checkpoint inhibition by caffeine analogs. International Journal of Oncology 16(5): 971-978. Jiang, X . , B. Zhao, R. Britton, L. Y . Lim, D. Leong, J. Sanghera, et al. (2004). Inhibition of Chkl by the G 2 D N A damage checkpoint inhibitor isogranulatimide. Molecular Cancer Therapeutics 3(10): 1221-1227. Karlsson-Rosenthal, C. and J. B. Millar (2006). Cdc25: mechanisms of checkpoint inhibition and recovery. Trends in Cell Biology 16(6): 285-292. Kashishian, A. , H. Douangpanya, D. Clark, S. T. Schlachter, C. T. Eary, J. G. Schiro, et al. (2003). DNA-dependent protein kinase inhibitors as drug candidates for the treatment of cancer. Molecular Cancer Therapeutics 2(12): 1257-1264. Kastan, M . B. , O. Onyekwere, D. Sidransky, B. Vogelstein and R. W. Craig (1991). Participation of p53 protein in the cellular response to D N A damage. Cancer Research 51(23 Pt 1): 6304-6311. Kawabe, T. (2004). G 2 checkpoint abrogators as anticancer drugs. Molecular Cancer Therapeutics 3(4): 513-519. Kim, C.-H., S.-J. Park and S.-H. Lee (2002). A targeted inhibition of DNA-dependent Protein Kinase sensitizes breast cancer cells following ionizing radiation. The Journal of Pharmacology and Experimental Therapeutics 303(2): 753-759. King, F. W., J. Skeen, N . Hay and E. Shtivelman (2004). Inhibition of Chkl by activated PKB/Akt. Cell Cycle 3(5): 634-637. Knight, Z. A. , G. G. Chiang, P. J. Alaimo, D. M . Kenski, C. B. Ho, R. T. Abraham and K. M . Shokat (2004). Isoform-specific phosphoinositide 3-kinase inhibitors from an arylmorphine scaffold. Bioorganic and Medicinal Chemistry 12(17): 4749-4759. Kohn, A. D., A . Barthel, K. S. Kovacina, A. Boge, B. Wallach, S. A . Summers, et al. (1998). Construction and characterization of a conditionally active version of the serine/threonine kinase Akt. Journal of Biological Chemistry 273(19): 11937-11943. Kohn, E. A. , C. J. Yoo and A. Eastman (2003). The protein kinase C inhibitor Go6976 is a potent inhibitor of D N A damage-induced S and G 2 cell cycle checkpoints. Cancer Research 63(1): 31-35. Kudo, N . , N . Matsumori, H. Taoka, D. Fujiwara, E. P. Schreiner, B. Wolff, M . Yoshida and S. Horinouchi (1999). Leptomycin B inactivates CRMl/exportin 1 by covalent modification at a cysteine residue in the central conserved region. Proceedings of the National Academy of Sciences of the United States of America 96(16): 112-117. Kudo, N . , B. Wolff, T. Sekimoto, E. P. Schreiner, Y . Yoneda, M . Yanagida, S. Horinouchi and M . Yoshida (1998). Leptomycin B inhibition of signal-mediated nuclear export by direct binding to C R M L Experimental Cell Research 242(2): 540-547. Kurose, A. , T. Tanaka, X . Huang, F. Traganos and Z. Darzynkiewicz (2006). Synchronization in the cell cycle by inhibitors of D N A replication induces histone H 2 A X phosphorylation: an indication of D N A damage. Cell Proliferation 39(3): 231-240. 168 Lara, P. N . , P. C. Mack, T. Synold, P. Frankel, J. Longmate, P. H . Gumerlock, J. H . Doroshow and D. R. Gandara (2005). The cyclin-dependent kinase inhibitor UCN-01 plus cisplatin in advanced solid tumors: a California cancer consortium phase I pharmacokinetic and molecular correlative trial. Clinical Cancer Research 11(12): 4444-4450. Latif, C , N . R. d. Elzen and M . J. O'Connell (2004). D N A damage checkpoint maintenance through sustained Chkl activity. Journal of Cell Science 117(Pt 16): 3489-3498. Lavin, M . F. and K . K . Khanna (1999). A T M : the protein encoded by the gene mutated in the radiosensitive syndrome ataxia-telangiectasia. International Journal of Radiation Biology 75(10): 1201-1214. Lee, J. W., S. M . Yannone, D. J. Chen and L. F. Povirk (2003). Requirement for XRCC4 and D N A ligase IV in alignment-based gap filling for nonhomologous D N A end joining in vitro. Cancer Research 63(1): 22-24. Lees-Miller, S. P., R. Godbout, D. W. Chan, M . Weinfeld, I. Rums S. Day, G. M . Barron and J. Allalunis-Turner (1995). Absence of p350 subunit of DNA-Activated Protein Kinase from a radiosensitive human cell line. Science 267(5201): 1183-1185. L i , L., C. A. Peterson, G. Kanter-Smoler, Y . F. Wei, L. S. Ramagli, P. Sunnerhagen, M . J. Siciliano and R. J. Legerski (1999). hRAD17, a structural homolog of the Schizosaccharomyces pombe RAD17 cell cycle checkpoint gene, stimulates p53 accumulation. Oncogene 18(9): 1689-1699. L i , Z., T. Otevrel, Y . Gao, H. L. Cheng, B. Seed, T. D. Stamato, G. E. Taccioli and F. W. Alt (1995). The XRCC4 gene encodes a novel protein involved in D N A double-strand break repair and V(D)J recombination. Cell 83(7): 1079-1089. Lieber, M . R., Y . Ma, U . Pannicke and K. Schwarz (2003). Mechanism and regulation of human non-homologous D N A end-joining. Nature Reviews in Molecular and Cellular Biology 4(9): 712-720. Lim, D. S., S. T. Kim, B. Xu, R. S. Maser, J. Lin, J. H. Petrini and M . B. Kastan (2000). A T M phosphorylates p95/nbsl in an S-phase checkpoint pathway. Nature 404(6778): 613-617. Lindsey-Boltz, L. A . , V . P. Bermudez, J. Hurwitz and A. Sancar (2001). Purification and characterization of human D N A damage checkpoint Rad complexes. Proceedings of the National Academy of Sciences of the United States of America 98(20): 11236-11241. Liu, Q., S. Guntuku, X.-S. Cui, S. Matsuoka, D. Cortez, K . Tamai, et al (2000). Chkl is an essential kinase that is regulated by Atr and required for the G 2 / M D N A damage checkpoint. Genes & Development 14(12): 1448-1459. Lodish, H. , A. Berk, P. Matsudaira, C. A. Kaiser, M . Krieger, M . P. Scott, S. L. Zipursky and J. Darnell, Eds. (2004). Molecular Cell Biology. New York, W.H. Freeman. Love, D. C , T. D. Sweitzer and J. A . Hanover (1998). Reconstitution of HIV-1 rev nuclear export: independent requirements for nuclear import and export. Proceedings of the National Academy of Sciences of the United States of America 95(18): 10608-10613. Lucero, H., D. Gae and G. E. Taccioli (2003). Novel localization of the D N A - P K complex in lipid rafts. Journal of Biological Chemistry 278(24): 22136-22143. 169 Matsumura, T., Z. Zerrudo and L. Hayflick (1979). Senescent human diploid cells in culture: survival, D N A synthesis and morphology. Journal of Gerontology 34(1): 328-334. Matsuoka, S., M . Huang and S. J. Elledge (1998). Linkage of A T M to cell cycle regulation by the Chk2 protein kinase. Science 282(5395): 1893-1897. Matsuoka, S., G. Rotman, A. Ogawa, Y . Shiloh, K. Tamai and S. J. Elledge (2000). Ataxia telangiectasia-mutated phosphorylates Chk2 in vivo and in vitro. Proceedings of the National Academy of Sciences of the United States of America 97(19): 10389-10394. McGowan, C. H . and P. Russell (1993). Human Weel kinase inhibits cell division by phosphorylating p34cdc2 exclusively on Tyrl5. EMBO Journal 12(1): 75-85. Meissner, T., E. Krause and U . Vinkemeier (2004). Ratjadone and leptomycin B block C R M 1-dependent nuclear export by identical mechanisms. FEBS Letters 576(1-2): 27-30. Meng, X . , Y . Yuan, A. Maestas and Z. Shen (2004). Recovery from D N A damage-induced G 2 arrest requires Actin-binding Protein Filamin-A/Actin-binding Protein 280. Journal of Biological Chemistry 279(7): 6098-6105. Mercer, W. E. (1992). Cell cycle regulation and the p53 tumor suppressor protein. Critical Reviews in Eukaryotic Gene Expression 2(3): 251-263. Miglietta, A. , F. Bozzo, L. Gabriel and C. Bocca (2004). Microtubule-interfering activity of parthenolide. Chemico-Biological Interactions 149(2-3): 165-173. Mingo-Sion, A. M . , P. M . Marietta, E. Koller, D. M . Wolf and C. L. V . D. Berg (2004). Inhibition of JNK reduces G 2 / M transit independent of p53, leading to endoreduplication, decreased proliferation, and apoptosis in breast cancer cells. Oncogene 23(2): 596-604. Mueller, P. R., T. R. Coleman, A. Kumagai and W. G. Dunphy (1995). M y t l : a membrane-associated inhibitory kinase that phosphorylates Cdc2 on both threonine-14 and tyrosine-15. Science 270(5233): 86-90. Nakamura, H. , Y . Yasui, N . Saito, A. Tachibana, K. Komatsu and K. Ishizaki (2006). D N A repair defect in AT cells and their hypersensitivity to low-dose-rate radiation. Radiation Research 165(3): 277-282. Nasmyth, K. (1996). Viewpoint: putting the cell cycle in order. Science 274(5293): 1643-1645. Nevaldine, B., J. A . Luongo and P. J. Hahn (1997). The scid defect results in much slower repair of D N A double-strand breaks but not high levels of residual breaks. Radiation Research 147(5): 535-540. Newlands, E. S., G. J. Rustin and M . H. Brampton (1996). Phase I trial of elactocin. British Journal of Cancer 74(4): 648-649. Nigg, E. A. (2001). Mitotic kinases as regulators of cell division and its checkpoints. Nature Reviews Molecular Cell Biology 2(1): 21-32. Niida, H. and M . Nakanishi (2006). D N A damage checkpoints in mammals. Mutagenesis 21(1): 3-9. Nishi, K. , M . Yoshida, D. Fujiwara, M . Nishikawa, S. Horinouchi and T. Beppu (1994). Leptomycin B targets a regulatory cascade of crml, a fission yeast nuclear protein, involved in control of higher order chromosome structure and gene expression. Journal of Biological Chemistry 269(9): 6320-6324. 170 Nojima, H. (2004). Gi and S-phase checkpoints, chromosome instability, and cancer. Methods in Molecular Biology 280(1): 3-49. Nurse, P. (1975). Genetic control of cell size at cell division in yeast. Nature 256(5518): 547-551. Nyberg, K. A., R. J. Michelson, C. W. Putnam and T. A . Weinert (2002). Toward maintaining the genome: D N A damage and replication checkpoints. Annual Review of Genetics 36(1): 617-656. O'Connell, M . J., J. M . Raleigh, H. M . Verkade and P. Nurse (1997). Chkl is a weel kinase in the G 2 D N A damage checkpoint inhibiting cdc2 by Y15 phosphorylation. EMBO Journal 16(3): 545-554. Onge, R. P. S., C. M . Udell, R. Casselman and S. Davey (1999). The human G 2 checkpoint control protein hRAD9 is a nuclear phosphoprotein that forms complexes with h R A D l and hHUSl . Molecular Biology of the Cell 10(6): 1985-1995. Parker, A. E., I. V . d. Weyer, M . C. Laus, P. Verhasselt and W. H. Luyten (1998). Identification of a human homologue of the Schizosaccharomyces pombe radl7+ checkpoint gene. Journal of Biological Chemistry 273(29): 18340-18346; Erratum 274: 24438(1999). Pastwa, E. and J. Blasniak (2003). Non-homologous end joining. Acta Biochimica Polonica 50(4): 891-908. Paull, K. D., R. H. Shoemaker, L. Hodes, A. Monks, D. A . Scudiero, L. Rubinstein, J. Plowman and M . R. Boyd (1989). Display and analysis of patterns of differential activity of drugs against human tumor cell lines: development of mean graph and C O M P A R E algorithm. Journal of the National Cancer Institute 81(14): 1088-1092. Pelech, S., C. Sutter and H. Zhang (2003). Kinetworks protein kinase multiblot analysis. Methods in Molecular Biology 218(1): 99-111. Peng, Y. , R. G. Woods, H. Beamish, R. Ye, S. P. Lees-Miller, M . F. Lavin and J. S. Bedford (2005). Deficiency in the catalytic subunit of DNA-dependent protein kinase causes down-regulation of A T M . Cancer Research 65(5): 1670-1677. Perrault, R., H. Wang, M . Wang, B. Rosidi and G. Iliakis (2004). Backup pathways of NHEJ are suppressed by DNA-PK. Journal of Cellular Biochemistry 92(4): 781-794. Pierce, A. J., P. Hu, M . Han, N . Ellis and M . Jasin (2001). K u D N A end-binding protein modulates homologous repair of double-strand breaks in mammalian cells. Genes & Development 15(24): 3237-3242. Powell, D. N . , J. S. DeFrank, P. Connell, M . Eogan, F. Preffer, D. Dombkowski, W. Tang and S. Friend (1995). Differential sensitivity of p53(-) and p53(+) cells to caffeine-induced radiosensitization and override of G 2 delay. Cancer Research 55(8): 1643-1648. Prudhomme, M . (2004). Combining D N A damaging agents and checkpoint 1 inhibitors. Current medicinal chemistry. And-cancer agents. 4(5): 435-438. Rane, M . J., P. Y . Coxon, D. W. Powell, R. Webster, J. B. Klein, W. Pierce, P. Ping and K. R. McLeish (2001). p38 Kinase-dependent M A P K A P K - 2 activation functions as 3-phosphoinositide-dependent kinase-2 for Akt in human neutrophils. Journal of Biological Chemistry 276(5): 3517-3523. 171 Riballo, E., M . Kuhne, N . Rief, A . Doherty, G. C. Smith, M . J. Recio, et al. (2004). A pathway of double-strand break rejoining dependent upon A T M , Artemis, and proteins locating to gamma-H2AX foci. Molecular Cell 15(5): 715-724. Rigas, B., S. Borgo, A . Elhosseiny, V. Balatsos, Z. Manika, H . Shinya, N . Kurihara, M . Go and M . Lipkin (2001). Decreased expression of DNA-dependent protein kinase, a D N A repair protein, during human colon carcinogenesis. Cancer Research 61(23): 8381-8384. Roberge, M . , R. G. S. Berlinck, L. Xu, H. J. Anderson, L. Y . Lim, D. Curman, et al. (1998). High-throughput assay for G 2 checkpoint inhibitors and identification of the structurally novel compound isogranulatimide. Cancer Research 58(24): 5701-5706. Roberge, M . , C. Tudan, S. M . Hung, K. W. Harder, F. R. Jirik and H . Anderson (1994). Antitumor drug fostriecin inhibits the mitotic entry checkpoint and protein phosphatases 1 and 2A. Cancer Research 54(23): 6115-6121. Roberts, B. J., K. L. Hamelehle, J. S. Sebolt and W. R. Leopold (1986). In vivo and in vitro anticancer activity of the structurally novel and highly potent antibiotic CI-940 and its hydroxy analog (PD 114,721). Cancer Chemotherapy and Pharmacology 16(2): 95-101. Rosenzweig, K . E., M . B. Youmell, S. T. Palayoor and B. D. Price (1997). Radiosensitization of human tumor cells by the phopshatidylinositol 3-Kinase inhibitors wortmannin and LY294002 correlates with inhibition of DNA-dependent Protein Kinase and prolonged G 2 - M delay. Clinical Cancer Research 3(7): 1149-1156. Rundle, N . T., L. Xu, R. J. Andersen and M . Roberge (2001). G 2 D N A damage checkpoint inhibition and antimitotic activity of 13-Hydroxy-15-oxozoapatlin. Journal of Biological Chemistry 276(51): 48231-48236. Russell, K. J., L. W. Wiens, G. W. Demers, D. A . Galloway, S. E. Plon and M . Groudine (1995). Abrogation of the G 2 checkpoint results in differential radiosensitization of G l checkpoint-deficient and Gi checkpoint-competent cells. Cancer Research 55(8): 1639-1642. Sailer, B. L., A . M . Barrasso, J. G. Valdez, J. M . Cobo, J. A . D'Anna and H. A . Crissman (1998). Reduction in the radiation-induced late S phase and G 2 blocks in HL-60 cell populations by amiloride, an efficient inhibitor of the Na+/H+ transporter. Cancer Research 58(3): 413-420. Samlaska, C. P. and E. A . Winfield. (1994). Pentoxifylline, journal of the American Academy of Dermatology 30(4): 603-621. Sancar, A. , L. A . Lindsey-Boltz, K. Unsal-Kacmaz and S. Linn (2004). Molecular mechanisms of mammalian D N A repair and the D N A damage checkpoints. Annual Review of Biochemistry 73(1): 39-85. Sarkaria, J. N . , E. C. Busby, R. S. Tibbetts, P. Roos, Y . Taya, L. M . Karnitz and R. T. Abraham (1999). Inhibition of A T M and ATR kinase activities by the radiosensitizing agent, caffeine. Cancer Research 59(17): 4375-4382. Sarkaria, J. N . , R. S. Tibbetts, E. C. Busby, A. P. Kennedy, D. E. Hi l l and R. T. Abraham (1998). Inhibition of phosphoinositide 3-kinase related kinases by the radiosensitizing agent wortmannin. Cancer Research 58(19): 4375-4382. 172 Schaar, B. T., G. K. Chan, P. Maddox, E. D. Salmon and T. J. Yen (1997). CENP-E function at kinetochores is essential for chromosome alignment. Journal of Cell Biology 139(6): 1373-1382. Schmeda-Hirschmann, G., L. Astudillo, J. Bastida, C. Codina, A . R. D. Arias, M . E. Ferreira, A . Inchaustti and G. Yaluff (2001). Cryptofolione derivatives from Cryptocarya alba fruits. Journal of Pharmacy and Pharmacology 53(4): 563-567. Sehlapelo, B. M . , S. E. Drewes and R. Scott-Shaw (1994). A 6-substituted 5,6-dihydro-a-pyrone from two species of Cryptocarya. Phytochemistry 37(3): 847-849. Seynaeve, C. M . , M . Stetler-Stevenson, S. Sebers, G. Kaur, E. A . Sausville and P. J. Worland (1993). Cell cycle arrest and growth inhibition by the protein kinase antagonist UCN-01 in human breast carcinoma cells. Cancer Research 53(9): 2081-2086. Shinohara, E. T., L. Geng, J. Tan, H. Chen, Y . Shir, E. Edwards, et al. (2005). DNA-dependent protein kinase is a molecular target for the development of noncytotoxic radiation-sensitizing drugs. Cancer Research 65(12): 4987-4992. Shtivelman, E., J. Sussman and D. Stokoe (2002). A role for PI 3-kinase and P K B activity in the G 2 / M phase of the cell cycle. Current Biology 12(11): 919-924. Smith, G. C. M . and S. P. Jackson (1999). The DNA-dependent protein kinase. Genes & Development 13(8): 916-934. Smith, J. R. and D. W. Lincoln (1984). Aging of cells in culture. International Review of Cytology 89(1): 151-176. Smits, V. A . J. and R. H. Medema (2001). Checking out the G2 /M transition. Biochimica et Biophysica Acta 1519(1-2): 1-12. Sorensen, C. S., R. G. Syljuasen, J. Falck, T. Schroeder, L . Ronnstrand, K . K . Klianna, B. B. Zhou, J. Bartek and J. Lukas (2003). Chkl regulates the S phase checkpoint by coupling the physiological turnover and ionizing radiation-induced accelerated proteolysis of Cdc25A. Cancer Cell 3(3): 247-258. Stanford, J. S. and J. V . Ruderman (2005). Changes in regulatory phosphorylation of Cdc25C Ser287 and Weel Ser549 during normal cell cycle progression and checkpoint arrests. Molecular Biology of the Cell 16(12): 5749-5760. Stewart, G. S., B. Wang, C. R. Bignell, A. M . Taylor and S. J. Elledge (2003). MDC1 is a mediator of the mammalian D N A damage checkpoint. Nature 421(6926): 961-966. Stewart, R. D. (2001). Two-lesion kinetic model of double-strand break rejoining and cell killing. Radiation Research 156(4): 365-378. Stiff, T , M . O'Driscoll, N . Rief, K. Iwabuchi, M . Lobrich and P. A . Jeggo (2004). A T M and D N A - P K function redundantly to phosphorylate H 2 A X after exposure to ionizing radiation. Cancer Research 64(7): 2390-2396. Strausfeld, U . , J. C. Labbe, D. Fesquet, J. C. Cavadore, A. Picard, K . Sadhu, P. Russell and M . Doree (1991). Dephosphorylation and activation of a p34cdc2/cyclin B complex in vitro by human CDC25 protein. Nature 351(6323): 242-245. Stronati, L., G. Gensabella, C. Lamberti, P. Barattini, D. Frasca, C. Tanzarella, et al. (2001). Expression and D N A binding activity of the Ku heterodimer in bladder carcinoma. Cancer 92(9): 2484-2492. Sturgeon, C. M . , K. Craig, C. Brown, N . T. Rundle, R. J. Andersen and M . Roberge (2005). Modulation of the G2 cell cycle checkpoint by sesquiterpene lactones 173 psilostachyin A and C isolated from the common ragweed Ambrosia artemisiifolia. Planta Medica 71(10): 949-954. Sturgeon, C. M . , A . Diaz-Marrero, B . Cinel, L. McHardy, M . Ngo, R. J. Andersen and M . Roberge (2006a). Abrogation of ionizing radiation-induced G 2 checkpoint and inhibition of nuclear export by cryptofolione. Cancer Research I n review. Sturgeon, C. M . , D. Kemmer, H. J. Anderson and M . Roberge (2006b). Yeast as a tool to uncover the cellular target of drugs. Biotechnology Journal 1(1): 289-298. Sturgeon, C. M . , Z. A . Knight, K. M . Shokat and M . Roberge (2006c). Effect of combined D N A repair inhibition and G 2 checkpoint inhibition on cell cycle progression after D N A damage. Molecular Cancer Therapeutics 5(4): 885-892. Sturgeon, C. M . and M . Roberge (2006d). Checkpoint abrogation alone is insufficient for radiosensitization by G 2 checkpoint inhibitors. Cancer Research I n prep. Suganuma, M . , T. Kawabe, H . Hori, T. Funabiki and T. Okamoto (1999). Sensitization of cancer cells to D N A damage-induced cell death by specific cell cycle G 2 checkpoint abrogation. Cancer Research 59(23): 5887-6891. Takai, H. , K . Naka, Y . Okada, M . Watanabe, N . Harada, S. Saito, et al. (2002). Chk2-deficient mice exhibit radioresistance and defective p53-mediated transcription. EMBO Journal 21(19): 5195-5205. Takata, M . , M . S. Sasaki, E. Sonoda, C. Morrison, M . Hashimoto, H . Utsumi, et al. (1998). Homologous recombination and non-homologous end-joining pathways of D N A double-strand breaks. EMBO Journal 17(18): 5497-5508. Takizawa, C. G. and D. O. Morgan (2000). Control of mitosis by changes in the subcellular location of cyclin-Bl-Cdkl and Cdc25C. Current Opinion in Cell Biology 12(6): 658-665. Tarn, S. W. and R. Schlegel (1992). Staurosporine overrides checkpoints for mitotic onset in B H K cells. Cell Growth and Differentiation 3(11): 811-817. Thelen, M . P., C. Venclovas and K. Fidelis (1999). A sliding clamp model for the Radl family of cell cycle checkpoint proteins. Cell 96(6): 769-770. Thomas, F. and U . Kutay (2003). Biogenesis and nuclear export of ribosomal subunits in higher eukaryotes depend on the CRM1 export pathway. Journal of Cell Science 116(12): 2409-1419. Thompson, L. H . and D. Schild (2001). Homologous recombinational repair of D N A ensures mammalian chromosome stability. Mutation Research 477(1-2): 131-153. Toyoshima, F., T. Moriguchi, A . Wada, M . Fukuda and E. Nishida (1998). Nuclear export of cyclin BI and its possible role in the D N A damage-induced G 2 checkpoint. EMBO Journal 17(10): 2728-2735. Tsuchida, R., T. Yamada, M . Takagi, A. Shimada, C. Ishioka, Y . Katsuki, et al. (2002). Detection of A T M gene mutation in human glioma cell line M059J by a rapid frameshift/stop codon in yeast. Radiation Research 158(2): 195-201. Valerie, K. and L. F. Povirk (2003). Regulation and mechanisms of mammalian double-strand break repair. Oncogene 22(37): 5792-5812. Viniegra, J. G., N . Martinez, P. Modirassari, J. H . Losa, C. P. Cobo, V . J. Lobo, et al. (2005). Full activation of PKB/Akt in response to insulin or ionizing radiation is mediated through A T M . Journal of Biological Chemistry 280(6): 4029-4036. 174 Virsik-Kopp, P., M . Rave-Frank, H. Hofman-Huther and H. Schmidberger (2004). Role of DNA-dependent protein kinase in the process of radiation-induced aberration formation. International Journal of Radiation Biology 80(2): 125-133. Volkmer, E. and L. M . Karnitz (1999). Human homologs of Schizosaccharomyces pombe radl, husl, and rad9 form a D N A damage-responsive protein complex. Journal of Biological Chemistry 274(2): 567-570. Walters, R. A. , L. R. Gurley and R. A . Tobey (1974). Effects of caffeine on radiation-induced phenomena associated with cell-cycle traverse of mammalian cells. biophysical Journal 14(2): 99-118. Wang, H. , A . R. Perrault, Y . Takeda, W. Qin, H. Wang and G. Iliakis (2003a). Biochemical evidence for Ku-independent backup pathways of NHEJ. Nucleic Acids Research 31(18): 5377-5388. Wang, H., Z. C. Zeng, T. A . Bui, E. Sonoda, M . Takata, S. Takeda and G. Iliakis (2001a). Efficient rejoining of radiation-induced D N A double-strand breaks in vertebrate cells deficient in genes of the RAD52 epistasis group. Oncogene 20(18): 2212-2224. Wang, H. , Z. C. Zeng, A . R. Perrault, X . Cheng, W. Qin and G. Iliakis (2001b). Genetic evidence for the involvement of D N A ligase IV in the DNA-PK-dependent pathway of non-homologous end joining in mammalian cells. Nucleic Acids Research 29(8): 1653-1660. Wang, Q., S. Fan, A. Eastman, P. J. Worland, E. A. Sausville and P. M . O'Connor (1996). UCN-01: a potent abrogator of G 2 checkpoint function in cancer cells with disrupted p53. Journal of the National Cancer Institute 88(14): 956-965. Wang, X . , J. Khadpe, B. Hu, G. Iliakis and Y. Wang (2003b). An overactivated ATR/CHK1 pathway is responsible for the prolonged G 2 accumulation in irradiated A T cells. Journal of Biological Chemistry 278(33): 30869-30874. Wang, X . , G. C. L i , G. Iliakis and Y . Wang (2002). Ku affects the Chkl-dependent G 2 checkpoint after ionizing radiation. Cancer Research 62(21): 6031-6034. Wang, X . , C. H. McGowan, M . Zhao, L. He, J. S. Downey, C. Fearns, Y . Wang, S. Huang and J. Han (2000). Involvement of the MKK6-p38gamma cascade in gamma-radiation-induced cell cycle arrest. Molecular and Cellular Biology 20(13): 4543-4552. Wang, X . , L. Wang, M . D. Callister, J. B. Putnam, L. Mao and L. L i (2001c). Human Radl7 is phosphorylated upon D N A damage and also overexpressed in primary non-small cell lung cancer tissues. Cancer Research 61(20): 7417-7421. Wang, Y. , J. L i , R. N . Booher, A. Kraker, T. Lawrence, W. R. Leopold and Y. Sun (2001d). Radiosensitization of p53 mutant cells by PD0166285, a novel G(2) checkpoint abrogator. Cancer Research 61(22): 8211-8217. Weinert, T. A . and L. H . Hartwell (1989). The RAD9 gene controls the cell cycle response to D N A damage in Saccharomyces cerevisiae. Science 241(4863): 317-322. Williams, C. J., I. Grandal, D. J. Vesprini, U . Wojtyra, J. S. Danska and C. J. Guidos (2001). Irradiation promotes V(D)J joining and RAG-dependent neoplastic transformation in SCED T-cell precursors. Molecular and Cellular Biology 21(2): 400-413. 175 Willmore, E., S. d. Caux, N . J. Sunter, M . J. Tilby, G. H. Jackson, C. A. Austin and B. W. Durkacz (2004). A novel DNA-dependent protein kinase inhibitor NU7026 potentiates the cytotoxicity of topoisomerase II poisons used in the treatment of leukaemia. Blood 103(12): 4659-4665. Wilson, C. R., S. E. Davidson, G. P. Margison, S. P. J. H . Hendry and C. M . West (2000). Expression of Ku70 correlates with survival in carcinoma of the cervix. British Journal of Cancer 83(12): 1702-1706. Winzeler, E. A. , D. D. Shoemaker, A. Astromoff, H . Liang, K . Anderson, B. Andre, et al. (1999). Functional Characterization of the S. cerevisiae Genome by Gene Deletion and Parallel Analysis Science 285(5429): 901-906. Wood, K. W., R. Sakowicz, L. S. Goldstein and D. W. Cleveland (1997). CENP-E is a plus end-directed kinetochore motor required for metaphase chromosome alignment. Ce//91(3): 357-366. Xu, B., S.-T. Kim, D.-S. Lim and M . B. Kastan (2002). Two molecularly distinct G 2 / M checkpoints are induced by ionizing radiation. Molecular and Cellular Biology 22(4): 1049-1059. Xu, B., S. K im and M . B. Kastan (2001). Involvement of Brcal in S-phase and G(2)-phase checkpoints after ionizing irradiation. Molecular and Cellular Biology 21(10): 3445-3450. Yamashita, K. , H. Yasuda, J. Pines, K. Yasumoto, H. Nishitani, M . Ohtsubo, et al. (1990). Okadaic acid, a potent inhibitor of type 1 and type 2A protein phosphatases, activates cdc2/Hl kinase and transiently induces a premature mitosis-like state in BHK21 cells. EMBO Journal 9(13): 4331-4338. Yashiroda, Y . and M . Yoshida (2003). Nucleo-cytoplasmic transport of proteins as a target for therapeutic drugs. Current Medicinal Chemistry 10(9): 741-748. Yazdi, P. T., Y . Wang, S. Zhao, N . P. E. Y . Lee and J. Qin (2002). SMC1 is a downstream effector in the ATM/NBS1 branch of the human S-phase checkpoint. Genes and Development 16(5): 571-582. Yen, T. J., D. A . Compton, D. Wise, R. P. Zinkowski, B. R. Brinkley, W. C. Earnshaw and D. W. Cleveland (1991). CENP-E, a novel human centromere-associated protein required for progression from metaphase to anaphase. EMBO Journal 10(5): 1245-1254. Yen, T. J., G. L i , B. T. Schaar, I. Szilak and D. W. Cleveland (1992). CENP-E is a putative kinetochore motor that accumulates just before mitosis. Nature 359(6395): 536-539. Yokoyama, N . , N . Hayashi, T. Seki, N . Pante, T. Ohba, K. Nishii, et al. (1995). A giant nucleopore protein that binds Ran/TC4. Nature 376(6536): 184-188. Yu, Q., J. L. Rose, H. Zhang, H. Takemura, K. W. Kohn and Y . Pommier (2002). U C N -01 inhibits p53 up-regulation and abrogates gamma-radiation-induced G(2)-M checkpoint independently of p53 by targeting both of the checkpoint kinases, Chk2 and Chkl . Cancer Research 62(20): 5743-5748. Zhao, H. and H . Piwnica-Worms (2001). ATR-mediated checkpoint pathways regulate phosphorylation and activation of human Chkl . Molecular and Cellular Biology 21(13): 4129-4139. Zhao, H., J. L. Watkins and H. Piwnica-Worms (2002). Disruption of the checkpoint kinase 1/cell division cycle 25 A pathway abrogates ionizing radiation-induced S 176 and G2 checkpoints. Proceedings of the National Academy of Sciences of the United States of America 99(23): 14795-14800. Zhou, B.-B. S., P. Chaturvedi, K. Spring, S. P. Scott, R. A. Johanson, R. Mishra, et al. (2000). Caffeine abolishes the mammalian G2 /M D N A damage checkpoint by inhibiting Ataxia-Telangiectasia-mutated kinase activity. Journal of Biological Chemistry 275(14): 10342-10348. Zhou, B. B. and S. J. Elledge (2000). The D N A damage response: putting checkpoints in perspective. Nature 408(6811): 433-439. Zhou, B. B. and E. A . Sausville (2003). Drug discovery targeting Chkl and Chk2 kinases. Progress in Cell Cycle Research. L. Meijer, A . Jezequel and M . Roberge. Roscoff, France, Editions "Life in Progress". 5: 413-421. Zhou, X . - Y . , X . Wang, B. Hu, J. Guan, G. Iliakis and Y . Wang (2002). An A T M -independent S-phase checkpoint response involves Chkl pathway. Cancer Research 62(6): 1598-1603. Zou, L., D. Cortez and S. J. Elledge (2002). Regulation of ATR substrate selection by Radl7-dependent loading of Rad9 complexes onto chromatin. Genes and Development 16(2): 198-208. 177 

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}]}"
                            data-media="{[{embed.selectedMedia}]}"
                            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:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0093035/manifest

Comment

Related Items