CHARACTERIZATION OF SELECT DOWNSTREAM EFFECTORS OF ONCOGENIC NOTCH1 IN T-CELL ACUTE LYMPHOBLASTIC LEUKEMIA by SAMUEL D. GUSSCOTT B.BmedSc. (Honours), Victoria University of Wellington, 2006 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Interdisciplinary Oncology) THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver) March 2016 © Samuel D. Gusscott, 2016 ii Abstract Oncogenic NOTCH1 signalling is a major driver of T cell acute lymphoblastic leukemia (T-ALL) transformation and growth. Although some downstream effectors of this function are known, they cannot explain all observed pro-growth and leukemogenic phenotypes and there are undoubtedly other effectors yet to be described and investigated. This study identifies microRNAs (miRNAs) regulated by NOTCH1 in T-ALL and further characterizes the actions of insulin-like growth factor 1 receptor (IGF1R) and protein kinase C theta (PKCθ), two signalling molecules I was previously involved in identifying as being regulated by NOTCH1 in a T-ALL context. I found that NOTCH1 can negatively regulate miR-223 expression, contributing to its ability to enhance IGF1R expression. In turn, IGF1R signalling is important to maintain growth in a subset of T-ALL cell lines and is a major positive effector of the PI3K/AKT signalling pathway. IGF1R downstream signalling pathways may be negatively affected by PKCθ. As expression of PKCθ is negatively regulated by NOTCH1 in T-ALL, here, I have attempted to identify its direct phosphorylation targets in this context. I have done this through the combined use of an analog sensitive (AS) kinase screen and an ascorbate peroxidase (APEX) based chemical labelling proximity screen. Candidate direct PKCθ phosphorylation targets identified include potential IGF1R downstream signalling components such as IRS4, mTOR, RICTOR, RAF1 and ARAF. Some of these targets were also found to be proximal to PKCθ in a T-ALL cellular context. This suggests that, in addition to regulating IGF1R signalling at the transcript or protein (via miR-223 repression) level, NOTCH1 also has the potential to positively regulate this pathway through repressing PKCθ phosphorylation of downstream components. Further studies are required to validate this hypothesis. Other candidate direct PKCθ phosphorylation targets identified here may also be worth further investigation and may suggest the involvement of PKCθ in additional cellular processes in T-ALL. Further development of my novel combined approach for the identification of direct phosphorylation targets may prove to be useful for the investigation of other kinases in a broad range of cell types.iii Preface Research described in chapters 2-4 was conducted in the laboratory of Dr Andrew Weng in the Terry Fox Laboratory department of the British Columbia Cancer Research Centre by Samuel Gusscott. Dr Weng and I equally contributed to experimental design and I was responsible for performing all the experiments and analyzing the data. A version of chapter 2 has been published. [Gusscott S], Kuchenbauer F, Humphries RK, Weng AP. (2012) Notch-mediated repression of miR-223 contributes to IGF1R regulation in T-ALL. Leuk Res. 2012 Jul; 36(7):905-11. Crown Copyright © 2012. Published by Elsevier Ltd. All rights reserved, doi: 10.1016/j.leukres.2012.02.013 (license number: 3744951199916). I wrote the manuscript in conjunction with Dr Weng, conducted the data collection and interpreted the results. A version of chapter 3 has been submitted for publication. [Gusscott S], Jenkins CE, Lam SH, Giambra V, Pollak M, Weng AP. (2015) Molecular Determinants of IGF Dependence in Human T-cell Acute Lymphoblastic Leukemia. I prepared the manuscript in conjunction with Dr Weng, conducted the data collection and interpreted the results. Re-submission is in progress. For Chapter 4 the experimental approach was conceived by Dr Weng, Dr Gregg Morin and I. With the assistance of two co-op students, Larissa Leenstra and Alyssa Krieger, I performed all experiments. Samples for mass spectrometry analysis were prepared with consultation from Dr Christopher Hughes from the Morin lab. Dr Christopher Hughes also assisted in analysis of mass spectrometry data. iv Table of Contents Abstract ........................................................................................................................... ii Preface ............................................................................................................................ iii Table of Contents ............................................................................................................iv List of Tables ...................................................................................................................xi List of Figures ................................................................................................................. xii List of Abbreviations ...................................................................................................... xvi Acknowledgements ...................................................................................................... xxii Dedication ................................................................................................................... xxiii Chapter 1 General Introduction ....................................................................................... 1 1.1 T-cell acute lymphoblastic leukemia ...................................................................... 1 1.1.1 Acute lymphoblastic leukemia .......................................................................... 1 1.1.2 T-ALL clinical presentation/diagnosis .............................................................. 1 1.1.3 T-ALL prognosis/treatment .............................................................................. 2 1.1.4 T-ALL genetic abnormalities - cytogenetics ..................................................... 2 1.1.5 T-ALL genetic abnormalities - mutations ......................................................... 3 1.2 NOTCH .................................................................................................................. 4 1.2.1 NOTCH signalling ............................................................................................ 4 1.2.2 NOTCH signalling in normal development ....................................................... 6 1.2.3 NOTCH signalling in cancer............................................................................. 7 1.2.4 NOTCH1 in T-ALL - mutations ......................................................................... 7 1.2.5 NOTCH1 in T-ALL - regulation of downstream genes/pathways ..................... 8 1.3 Thesis objectives ................................................................................................. 10 Chapter 2 NOTCH-mediated repression of miR-223 contributes to IGF1R regulation in T-ALL ............................................................................................................................ 12 v 2.1 Introduction .......................................................................................................... 12 2.1.1 miRNAs - general background ....................................................................... 12 2.1.2 T cell development and miRNAs .................................................................... 12 2.1.3 MicroRNAs in hematological malignancies .................................................... 14 2.1.4 Statement of hypothesis and objectives ........................................................ 15 2.2 Materials and methods ......................................................................................... 15 2.2.1 Cell culture ..................................................................................................... 15 2.2.2 miRNA expression profiling ............................................................................ 16 2.2.3 Real-time quantitative PCR ............................................................................ 16 2.2.4 Luciferase reporter assay .............................................................................. 18 2.2.5 Viral vectors and transduction ........................................................................ 18 2.2.6 Western Blot .................................................................................................. 18 2.2.7 Flow cytometry ............................................................................................... 19 2.2.8 Gene expression profile analysis ................................................................... 19 2.2.9 Statistics ........................................................................................................ 19 2.3 Results ................................................................................................................. 19 2.3.1 miRNA microarray experimental setup .......................................................... 19 2.3.2 Microarray profiling of NOTCH1-regulated miRNAs in T-ALL ........................ 22 2.3.3 NOTCH1 regulates miR-223 in human T-ALL cells ....................................... 25 2.3.4 NOTCH1 mediated repression of miR-223 does not involve HES1 or c-MYC 27 2.3.5 miR-223 targets IGF1R .................................................................................. 28 2.3.6 miR-223 regulates IGF1R protein levels in T-ALL ......................................... 34 2.3.7 Manipulation of miR-223 alone does not affect T-ALL cell growth ................. 36 2.4 Discussion ............................................................................................................ 38 2.4.1 NOTCH1 regulated miRNAs .......................................................................... 38 vi 2.4.2 miR-223 ......................................................................................................... 38 2.4.3 miR-223 in cancer .......................................................................................... 39 2.4.4 NOTCH regulation of miR-223 ....................................................................... 40 2.4.5 miR-223 regulation of IGF1R in T-ALL .......................................................... 41 2.4.6 miR-223 regulation of IGF1R in other contexts .............................................. 41 2.4.7 miR-223 in T-ALL ........................................................................................... 42 2.5 Conclusion ........................................................................................................... 44 Chapter 3 Molecular determinants of IGF dependence in human T-ALL ...................... 45 3.1 Introduction .......................................................................................................... 45 3.1.1 IGF1R structure ............................................................................................. 45 3.1.2 IGF1R signalling ............................................................................................ 45 3.1.3 IRS1/PI3K/AKT signalling .............................................................................. 47 3.1.4 MAPK signalling ............................................................................................. 49 3.1.5 IGF1R in T-ALL .............................................................................................. 50 3.1.6 PI3K/AKT and MAPK signalling in T-ALL – mouse models and direct mutational activation ............................................................................................... 51 3.1.7 PI3K/AKT and MAPK signalling in T-ALL – upstream activation .................... 52 3.1.8 IGF1R inhibition ............................................................................................. 54 3.2 Materials and methods ......................................................................................... 54 3.3.1 Cell culture ..................................................................................................... 54 3.2.2 Drugs ............................................................................................................. 55 3.2.3 Viable cell number assay ............................................................................... 55 3.2.4 Viral vectors ................................................................................................... 55 3.2.5 Viral transduction ........................................................................................... 56 3.2.6 Western Blot .................................................................................................. 56 vii 3.2.7 Ligand stimulation assay ............................................................................... 56 3.2.8 Flow cytometry ............................................................................................... 57 3.2.9 Gene expression profile analysis ................................................................... 57 3.2.10 Statistics ...................................................................................................... 57 3.3 Results ................................................................................................................. 57 3.3.1 IGF1R inhibition in T-ALL cell lines ................................................................ 57 3.3.2 Effect of IGF1R expression ............................................................................ 61 3.3.3 Downstream effector pathways ...................................................................... 64 3.3.4 Effect of PTEN ............................................................................................... 67 3.3.5 Role of PI3Kγ/δ .............................................................................................. 74 3.3.6 Non-overlapping roles of IGF-1 and IL-7 ....................................................... 75 3.4 Discussion ............................................................................................................ 79 3.4.1 IGF1R in T-ALL context ................................................................................. 79 3.4.2 Correlations with IGF1R inhibitor resistance/sensitivity ................................. 79 3.4.3 Mechanisms of IGF1R inhibitor resistance .................................................... 80 3.4.4 PI3K/AKT signalling in T-ALL - IGF1R versus IL7R ....................................... 82 3.5 Conclusion ........................................................................................................... 83 Chapter 4 Identification of direct PKCθ phosphotargets in T-ALL ................................. 84 4.1 Introduction .......................................................................................................... 84 4.1.1 PKCs ............................................................................................................. 84 4.1.2 PKCθ ............................................................................................................. 84 4.1.3 PKCθ in T-cells .............................................................................................. 85 4.1.4 PKCθ targets in other cells ............................................................................ 86 4.1.5 PKCθ in cancer .............................................................................................. 87 4.1.6 PKCθ in T-ALL ............................................................................................... 87 viii 4.1.7 Identification of phosphotargets ..................................................................... 89 4.1.8 AS kinase method .......................................................................................... 89 4.1.9 Identification of proximal partners .................................................................. 91 4.1.10 Proximal labelling methods .......................................................................... 93 4.2 Materials and methods ......................................................................................... 95 4.2.1 Cloning PKCθ vectors .................................................................................... 95 4.2.2 Site directed mutagenesis ............................................................................. 99 4.2.3 PKCθ protein production – 293T overexpression ........................................ 101 4.2.4 Lentiviral production and transduction – HPBALL cells ............................... 102 4.2.5 In vitro kinase assay .................................................................................... 102 4.2.6 BioID labelling .............................................................................................. 103 4.2.7 APEX biotin phenol labelling ........................................................................ 103 4.2.8 APEX alkyne phenol labelling – click chemistry ........................................... 104 4.2.9 Western Blotting .......................................................................................... 104 4.2.10 Immunoprecipitation/SA pull down ............................................................. 105 4.2.11 Sample preparation and tryptic digestion ................................................... 106 4.2.12 Dimethyl labelling and peptide clean up .................................................... 106 4.2.13 Mass spectrometry analysis ....................................................................... 107 4.2.14 Data analysis ............................................................................................. 108 4.3 Results ............................................................................................................... 109 4.3.1 Identification and mutation of gatekeeper residue in PKCθ ......................... 109 4.3.2 Overexpression and initial testing of PKCθAS ............................................. 110 4.3.3 Second-site mutation rescue of catalytic activity .......................................... 112 4.3.3.1 Second-site mutation rescue of catalytic activity – mutation to rPKCδ residues ............................................................................................................. 114 ix 4.3.3.2 Second-site mutation rescue of catalytic activity – mutation to common β-branched residues ............................................................................................. 118 4.3.4 PKCθAS385 use with whole cell lysate ....................................................... 124 4.3.5 Mass Spectrometry Identification - PKCθAS................................................ 129 4.3.6 BioID cloning and expression ...................................................................... 133 4.3.7 BioID biotin labelling .................................................................................... 135 4.3.8 APEX cloning and expression ...................................................................... 139 4.3.9 APEX biotin phenol labelling ........................................................................ 141 4.3.10 APEX biotin Phenol IP ............................................................................... 143 4.3.11 Mass spectrometry identification - PKCθAPEX with biotin phenol ............. 145 4.3.12 APEX alkyne phenol labelling .................................................................... 146 4.3.13 Mass spectrometry identification - PKCθAPEX with HexT ......................... 151 4.3.14 Comparison of candidate PKCθ phosphotargets and identified proximal partners ................................................................................................................ 153 4.4 Discussion. ......................................................................................................... 154 4.4.1 Identified candidate direct PKCθ phosphotargets ........................................ 154 4.4.1.1 Candidate direct PKCθ phosphotargets - IRS4 ..................................... 155 4.4.1.2 Candidate direct PKCθ phosphotargets - mTOR, RICTOR ................... 155 4.4.1.3 Candidate direct PKCθ phosphotargets - RAF1, ARAF ......................... 156 4.4.1.4 Other candidate PKCθ phosphotargets of interest ................................ 157 4.4.2 Proximity of candidate direct PKCθ phosphotargets .................................... 158 4.4.3 Method improvement ................................................................................... 159 4.4.3.1 PKCθAS method variations ................................................................... 159 4.4.3.2 PKCθAPEX method variations .............................................................. 163 4.5 Conclusion ...................................................................................................... 164 Chapter 5 General conclusions ................................................................................... 166 x 5.1 Summary of study and findings .......................................................................... 166 5.2 Conclusions regarding the study hypotheses ..................................................... 168 5.3 Strengths and limitations of this study ................................................................ 169 5.4 Future research directions ................................................................................. 170 References .................................................................................................................. 172 Appendices ................................................................................................................. 196 Appendix A. Chemical synthesis and validation of HexT ......................................... 196 Appendix B. Previously described PKCθ phosphotargets, PKCθ Interactors and TCR signalling components ............................................................................................. 198 Appendix C. Proteins identified in 3 out of 4 PKCθAS experiments ......................... 201 Appendix D. Proteins identified in PKCθAPEX experiment using biotin phenol as a substrate .................................................................................................................. 204 Appendix E. PKCθ regulation of IGF1R signalling via IRS1 repression ................... 211 Appendix F. PKCθ regulation of CXCR4 signalling .................................................. 214 xi List of Tables Table 2.1 Ingenuity pathway analysis predicts miR-223 targets regulate IGF signalling30 Table 2.2 Correlation between IGF1R and mature miR-223 ......................................... 32 Table 3.1 PTEN protein expression and IL-7Rα mutational status in human T-ALL cell lines ............................................................................................................................... 69 Table 4.1 Primers used for PKCθ:BirA* fusion cloning .................................................. 97 Table 4.2 Primers used for PKCθ:APEX fusion cloning ................................................ 98 Table 4.3 Oligos used for PKCθ site directed mutagenesis ........................................ 100 Table 4.4 SSM mutations created to attempt catalytic activity rescue ......................... 115 Table 4.5 Candidate direct PKCθ phosphotargets ...................................................... 132 Table 4.6 Criteria for inclusion of protein species into PMA activation states .............. 145 xii List of Figures Figure 1.1 NOTCH1 signalling pathway .......................................................................... 5 Figure 2.1 Kinetics of NOTCH1-regulated mRNA in T-ALL ........................................... 21 Figure 2.2 Microarray profiling of NOTCH1-regulated miRNAs in human T-ALL .......... 23 Figure 2.3 Microarray profiling of NOTCH1-regulated miRNAs in mouse T-ALL ........... 24 Figure 2.4 NOTCH1 represses miR-223 ....................................................................... 27 Figure 2.5 NOTCH1 does not indirectly repress miR-223 through the canonical effectors HES1 or c-MYC ............................................................................................................. 28 Figure 2.6 Ingenuity pathway analysis predicts miR-223 targets regulate IGF signalling ...................................................................................................................................... 31 Figure 2.7 miR-223 can target IGF1R ........................................................................... 33 Figure 2.8 miR-223 can target the 3’UTR of IGF1R to repress expression ................... 34 Figure 2.9 miR-223 regulates IGF1R protein levels, but does not alter cell growth ....... 35 Figure 2.10 miR-223 overexpression does not alter cell growth .................................... 37 Figure 3.1 Canonical IGF1R signalling pathway ........................................................... 47 Figure 3.2 Pharmacological inhibition of IGF1R restricts growth of a subset of human T-ALL cell lines ................................................................................................................. 59 Figure 3.3 BMS-754807 inhibits cell growth to a greater extent than CP-751,871 in a subset of human T-ALL cell lines .................................................................................. 60 Figure 3.4 Sensitivity to IGF1R inhibition correlates with surface IGF1R expression level ...................................................................................................................................... 62 Figure 3.5 Sensitivity to IGF1R inhibition correlates with IGF1R and IRS2 transcript levels ............................................................................................................................. 63 Figure 3.6 Constitutive activation of AKT, but not RAS, rescues T-ALL cells from IGF1R inhibition ........................................................................................................................ 66 Figure 3.7 Pharmalogical inhibition of the PI3K/AKT pathway blocks growth of human T-ALL cell lines ................................................................................................................. 67 Figure 3.8 PTEN protein status in human T-ALL cell lines ............................................ 70 xiii Figure 3.9 Correlation between IGF1R inhibitor efficacy and surface IGF1R expression level holds up in PTEN-positive cell lines, but less so in PTEN-negative cell lines ....... 71 Figure 3.10 Enforced PTEN expression and knockdown in T-ALL cell lines ................. 72 Figure 3.11 PTEN contributes to, but does not define IGF dependence ....................... 73 Figure 3.12 Combined inhibition of IGF1R and PI3Kγ does not block growth of PTEN negative CCRF-CEM cells ............................................................................................ 75 Figure 3.13 Signalling through IL7R does not rescue T-ALL cells from IGF1R inhibition nor maintain prolonged activation of AKT ...................................................................... 77 Figure 3.14 Constitutive activation of IL7R does not confer resistance to IGF1R inhibition ........................................................................................................................ 78 Figure 4.1 Proximity labelling reactions. ........................................................................ 95 Figure 4.2 pcDNA3-nFLAG-hPKCθ plasmid map. ........................................................ 96 Figure 4.3 PKCθ:BirA* and PKCθ:APEX fusion constructs ........................................... 99 Figure 4.4 3D structure of PKCθ ................................................................................. 110 Figure 4.5 PKCθAS variants have reduced expression and catalytic activity .............. 112 Figure 4.6 PKCθ shares highest degree of homology to PKCδ ................................... 113 Figure 4.7 PKCθ protein sequence alignments ........................................................... 114 Figure 4.8 M444L rPKCθ derived SSM variant improves PKCθAS2 expression and activity ......................................................................................................................... 116 Figure 4.9 PKCθAS2 M444L utilizes ATP analog more the ATP ................................ 117 Figure 4.10 PKCθAS2 M444L preferentially utilizes Bn-ATPγS analog ...................... 118 Figure 4.11 Residue distribution of SSM sites ............................................................. 121 Figure 4.12 A410V and M385V β-branched SSM variants improve PKCθAS2 expression ................................................................................................................... 122 Figure 4.13 PKCθAS2-M385V has improved catalytic activity .................................... 123 Figure 4.14 PKCθAS2-M385V has approximately half catalytic activity of wildtype PKCθ .................................................................................................................................... 124 xiv Figure 4.15 PKCθWT and PKCθAS2-M385V can only be stably overexpressed to a moderate degree in the T-ALL cell line HPBALL; not sufficient for in vitro kinase assays .................................................................................................................................... 126 Figure 4.16 PKCθAS2-M385V utilizes ATP analog Bn-ATPγS to produce a unique banding pattern with HPBALL lysate ........................................................................... 127 Figure 4.17 PKCθAS2-M385V produces a similar banding pattern with different T-ALL cell line lysates ............................................................................................................ 128 Figure 4.18 Increasing salt concentration decreases PKCθAS2-M385V phosphotransfer .................................................................................................................................... 129 Figure 4.19 34 PKCθ direct phosphotarget candidates identified ................................ 131 Figure 4.20 No pattern of enrichment in protein species with increased salt concentration ............................................................................................................... 133 Figure 4.21 293T overexpression of PKCθ/BirA* fusion constructs shows some unintended product ..................................................................................................... 134 Figure 4.22 HPBALL expression of PKCθ/BirA* fusion constructs shows specific predicted band with expression equivalent to endogenous PKCθ ............................... 135 Figure 4.23 Biotin labelling by PKCθ/BirA* fusion constructs in 293T ......................... 137 Figure 4.24 Biotin from media utilized by PKCθ/BirA* fusion constructs in HPBALL ... 138 Figure 4.25 Exogenous biotin labelling by PKCθ/BirA* fusion constructs in HPBALL has slow detectable biotin labelling kinetics ....................................................................... 139 Figure 4.26 HPBALL expression of PKCθ:APEX fusion constructs expressed less than endogenous PKCθ and show novel banding with biotin phenol labelling .................... 141 Figure 4.27 HPBALL PKCθ/APEX biotin phenol labelling shows additional novel banding when PKCθ is activated ................................................................................. 142 Figure 4.28 Without removal of free biotin, phenol labelled proteins are unable to be isolated with streptavidin resin .................................................................................... 144 Figure 4.29 Removal of free biotin phenol allows labelled protein to be isolated with streptavidin resin ......................................................................................................... 144 Figure 4.30 3EP is utilized less efficiently than biotin phenol by APEX ....................... 148 Figure 4.31 HexT is utilized more efficiently than biotin phenol by APEX ................... 150 xv Figure 4.32 HexT labelling does not produce observably different banding with PMA activation ..................................................................................................................... 150 Figure 4.33 Number of identified proteins from triplicate PKCθ/APEX HexT experiments .................................................................................................................................... 152 Figure 4.34 Proteins identified as both candidate direct PKCθ phosphotargets and PKCθ/APEX (biotin phenol) interactors ....................................................................... 154 Figure 4.35 PP1 analog can selectively inhibit PKCθAS2-M385V .............................. 162 Figure 4.36 Human PKCθ substrate recognition sequence motif ................................ 163 Figure 5.1 Summarized role of NOTCH1 regulation of the IGF1R signalling pathway in T-ALL .......................................................................................................................... 168 xvi List of Abbreviations 3EP 3-Ethynylphenol AML Acute myeloid leukemia APC Allophycocyanin APEX Ascorbate peroxidase (mutated K14D, W41F, E112K, A134P), monomerized and improved activity AS Analog sensitive ATP Adenosine triphosphate ATPγS Adenosine-5’-(3-thiotriphosphate) BioID Proximity-dependent biotin Identification BirA Biotin-[acetyl-CoA-carboxylase] ligase BirA* BirA mutated (R118G), promiscuous biotinylation BMS-754807 IGF1R/IR kinase inhibitor BN ATPγS N6 -Benzyladenosine-5’-O-(3-thiotriphosphate) BP Biotin Phenol ((3aS,4S,6aR)-hexahydro-N-[2-(4-hydroxyphenyl)ethyl]-2-oxo-1H-thieno[3,4-d]imidazole-4-pentanamide) BSA Bovine serum albumin CA Constitutively active CARMA1 CARD-containing MAGUK Protein 1 CCDC88A Coiled-Coil Domain Containing 88A (also known as Girdin) CCLE Cancer Cell Line Encyclopedia CCR CC chemokine receptor CD Cluster of differentiation CD8-IGF1R Constitutively active IGF1R CDC37 Cell Division Cycle 37, Heat Shock Protein 90 co-chaperone CFSE Carboxyfluorescein succinimidyl ester CLL Chronic lymphocytic leukemia CML Chronic myeloid leukemia CMML Chronic myelomonocytic leukemia CNS Central nervous system xvii CP-751,871 IGF1R blocking antibody (Figitumumab) CXCL C-X-C chemokine Ligand CXCR C-X-C chemokine Receptor DAG Diacyl glycerol DiDBiT Direct Detection of Biotin-containing Tags DMEM Dulbecco's Modified Eagle Medium DMSO Dimethyl sulfoxide DN Double negative (CD4-/CD8-) DNA Deoxyribonucleic acid DnMAML Dominant negative MAML DP Double positive (CD4+/CD8+) DTT Dithiothreitol EDTA Ethylenediaminetetraacetic acid EGTA Ethylene glycol tetraacetic acid ERK Extracellular signal regulated kinases ETP Early T-cell precursor FACS Fluorescence-activated cell sorting FBS Fetal bovine serum FBXW7 F-box/WD repeat-containing protein 7 FISH Fluorescent in situ hybridization FRET Förster resonance energy transfer FSBA 5′-(4-fluorosulfonylbenzoyl)adenosine GDP Guanosine diphosphate GFP Green fluorescent protein GIST Gastrointestinal stromal tumor GPCR G-protein coupled receptor GSI Gamma secretase inhibitor GTP Guanosine triphosphate HA Human influenza hemagglutinin tag HCC Hepatocellular carcinoma HD Heterodimerization domain xviii HEPES 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid HES1 Hairy and enhancer of split-1 HexT N-[2-(4-hydroxyphenyl)ethyl]hex-5-ynamide hexyne-tyramide HNSCC Head and neck squamous cell carcinoma HRP Horse radish peroxidase HSC Hematopoietic stem cell IC50 Half maximal inhibitory concentration ICN Intracellular NOTCH1 (Constitutively active) IGF-1 Insulin-like growth factor IGF1R Insulin-like growth factor-1 receptor IgG Immunoglobulin G IL Interleukin IL7R Interleukin-7 receptor IMDM Iscove's Modified Dulbecco's Medium insLSRC IL7Rα activating mutation, insertion of Leucine (L), Serine (S), Arginine (R), Cysteine (C) at position L242. IP Immunoprecipitation IP3 Inositol triphosphate IR Insulin receptor IRES Internal ribosomal entry site IRS Insulin receptor substrate LBL Lymphoblastic lymphoma LIC Leukemia initiating cell LNA Locked nucleic acid MAML Mastermind-like MAPK Mitogen-activated protein kinase MBP Myelin basic protein MCL Mantle cell lymphoma MHC Major histocompatibility complex MIG MSCV-IRES-GFP retroviral vector miRNA microRNA xix MNDU3 Modified myeloproliferative sarcoma virus long terminal repeat enhancer-promoter MSCV Murine stem cell virus promoter MSH6 MutS homolog 6 mTOR Mammalian target of rapamycin mTORC2 Mammalian target of rapamycin complex myrAKT Myristoylated AKT, constitutively active NA-PP1 1-Naphthyl-PP1 NF-kB Nuclear factor kappa-light-chain-enhancer of activated B cells NGFR Nerve growth factor receptor (truncated) NK Natural killer NMR Nuclear magnetic resonance NSCLC Non-small cell lung cancer PAGE Polyacrylamide gel Electrophoresis PBS Phosphate-buffered saline PE ATPγS N6 - Phenylethyladenosine-5’-O-(3-thiotriphosphate) PEG Polyethylene glycol PEI Polyethyleneimine PEST Degron sequence motif rich in Proline (P), Glutamate (E), Serine (S), Threonine (T), PGK Phosphoglycerate kinase 1 promoter PH Pleckstrin homology PI3K Phosphoinositide 3-Kinase PIP2 Phosphatidylinositol biphosphate PIP3 Phosphatidylinositol triphosphate PKC Protein kinase C PLC Phospholipase C PMA Phorbol-12-myristate 13-acetate PMSF Phenylmethylsulfonyl fluoride PNBM p-Nitrobenzyl mesylate PP1 Pyrazolo[3,4-d]pyrimidine xx PPI Protein-protein Interaction proKALIP Protein Kinase Assay Linked with Phosphoproteomics method PTEN Phosphatase and tensin homolog PVDF Polyvinylidene fluoride qRT-PCR Quantitative real-time polymerase chain reaction RBPJκ/CSL CBF1/Suppressor of Hairless/LAG-1 DNA-binding protein RICTOR Rapamycin-insensitive Companion of mTOR RIPA Radioimmunoprecipitation assay buffer RMA Robust multiarray average RNA Ribonucleic acid ROS Reactive oxygen species RPMI 1640 Roswell Park Memorial Institute 1640 Medium RTK Receptor tyrosine kinase SA Streptavidin SDF Stromal derived factor SDS Sodium dodecyl sulfate sCXCR4 Cell surface expression levels of CXCR4 sh Short hairpin Shc Src homology 2 domain-containing SMZL Splenic marginal zone lymphoma SP Single positive SPAK Ste20-related proline alanine rich kinase SS Serum starve SSM Second-site suppressor mutation STAT Signal transducer and activator of transcription T-ALL T-cell acute lymphoblastic leukemia TCEP Tris(2-carboxyethyl)phosphine TCR T-cell receptor Tfh T follicular helper cell Th T helper cell Treg T regulatory cell xxi UTR Untranslated region WT Wild-type Y2H Yeast two-hybrid ΔE Constitutively active NOTCH1lacking extracellular domain, GSI dependent λPP Lamda phosphatase xxii Acknowledgements I would like to thank my supervisor Dr. Andrew Weng for his mentorship throughout my time in his lab. His support and guidance has been critical to any success I’ve had and the lessons he has taught me will undoubtedly serve me well in my future scientific endeavours. I would also like to thank all the members who have served on my supervisory committee, Dr. Aly Karsan, Dr. Gerald Krystal, Dr. Gregg Morin, and Dr.Keith Humphries. They have all provided invaluable scientific critique throughout my PhD for which I am extremely grateful. I would also like to acknowledge the insightful scientific support and friendship from all past and present Weng Lab colleagues without which this process would have been immensely more difficult. Additionally I would like to extend my thanks to all members of the Terry Fox Laboratory for making my time here enjoyable. I would like to thank all the people throughout the building who have shared ideas, time and reagents, especially members of the Morin Lab, without who any proteomic endeavor would not have been a possibility. Finally I would like to thank my graduate program, the Interdisplenary Oncology Program (IOP), and it director, Dr. Angela Brooks-Wilson, for allowing me to pursue my PhD for as long as I did but also for giving me impetus to finish it. xxiii Dedication To my friends and family1 Chapter 1 General Introduction 1.1 T-cell acute lymphoblastic leukemia 1.1.1 Acute lymphoblastic leukemia Acute lymphoblastic leukemias/lymphomas (ALL/LBL) are an aggressive set of hematological malignancies that are characterized primarily by their similarity to precursor cells of the lymphoid lineage (lymphoblasts). They are the most common malignancy diagnosed in children and adolescents, and account for ~80% of leukemias presenting in children less than six years old [1]. As their name suggests, further classification can be made based on the site of presentation. Disease presenting primarily with confinement to a mass lesion, commonly mediastinal (thymic) or nodal, are termed lymphoblastic lymphoma (LBL), whereas those with additional bone marrow and blood involvement (>25% lymphoblasts in the bone marrow) are termed acute lymphoblastic leukemia (ALL) [2]. Similarly to normal lymphoid development, ALL is further subdivided based on their surface marker phenotypic resemblance to cells committed to either the B or T cell lineage. 15% of childhood and 25% of adult ALL cases present with a precursor T-cell phenotype and are termed T lymphoblastic leukemia (T-ALL) [2]. 1.1.2 T-ALL clinical presentation/diagnosis T-ALL, like other acute leukemias, does not present with any specific symptoms and can often be confused with other common illnesses. More general symptoms include weight loss, fever and fatigue. These symptoms, as well as others such as frequent/persistent infections, easy bruising and frequent bleeding (e.g. nose bleeds), are caused by the low blood cell counts arising from leukemic blasts crowding out normal hematopoietic progenitors in the bone marrow. In addition, persistent enlarged glands due to blast infiltration, abdominal pain due to hepatosplenomegaly and trouble breathing due to thymic enlargement can occur [3]. Pathological diagnosis often begins with a morphological assessment of a blood smear, and subsequently from a bone marrow biopsy. A high blast cell count is indicative of ALL. Final diagnosis and subtyping comes from immunophenotypic and molecular analysis of the blood or bone 2 marrow. Flow cytometric immunophenotyping shows most T-ALL samples to be CD7+ and cytoplasmic CD3+, with variable positive expression of other T cell markers CD1a, CD2, CD4, CD5, CD8, CD10, CD34, CD99, HLA-DR, and TdT. These other markers can be used to stratify T-ALL further based of their resemblance to normal T cell differentiation stages into pro-T; pre-T; cortical T; and medullary T [2-4]. Molecular analysis can involve cytogenetics, fluorescent in situ hybridization (FISH) and RT-PCR. 1.1.3 T-ALL prognosis/treatment Initial induction therapy for T-ALL consists of a cocktail of chemotherapeutic agents such as methotrexate, vincristine, corticosteroids (prednisone or dexamethasone), asparaginase and anthracyclines (doxorubicin) [5]. Higher risk cases can also receive cranial radiation to help prevent central nervous system (CNS) involvement. Although successful at lifting complete remission rates to ~88% in childhood T-ALL, a number of patients (~9%) have induction therapy failure and some (~3%) succumb to treatment-related death. Of those patients that initially achieve complete remission ~11% will go on to relapse resulting in a 5 year event free survival rate of ~75% [6]. Other studies on different cohorts of patients find similar results [7]. Due to the harsh treatment regiment, even those children that achieve continuous complete remission can suffer from treatment late effects, including increased risk of CNS tumors, as well as impaired intellectual function and neuropsychological development [8]. For these reasons there is a continued interest in developing more directed therapies to minimise adverse effects and improve treatment efficacy. To achieve this, a greater understanding of the cellular and molecular characteristics of T-ALL is required. 1.1.4 T-ALL genetic abnormalities - cytogenetics As with other hematological malignancies, some of the first molecular characteristics used to describe and discover the mechanistic origins of T-ALL were cytogenetic. 50-70% of T-ALL/LBL cases present with an abnormal karyotype [9, 10]. Most commonly, these involve a T cell receptor (TCR) locus (α/δ; 14q11.2, β; 7q35, γ; 7p14-15) whereby a translocation can place an oncogene under the transcriptional control of the regulatory element. These loci are thought to be particularity susceptible to translocation due to the normal genetic rearrangements that occur during the development of a functional TCR. 3 The most common TCR translocation partner genes include TLX1, TLX3, c-MYC, TAL1, LMO1, LMO2, LYL1 and LCK [9, 11]. In addition, cytogenetic changes involving sites others than the TCR loci can give rise to aberrant expression of oncogenes, including the SIL-TAL1 interstitial deletion, as well as generating novel fusion oncoproteins such as CALM-AF10 [12] and MLL-ENL [13]. Another much rarer translocation identified from early karyotypic studies was t(7;9)(q34;q34.3) [14]. This translocation positions the NOTCH1 gene next to the TCRβ locus resulting in a truncated version of NOTCH1 that is constitutively active. The importance of NOTCH1 in T-ALL, however, gained elevated appreciation with the discovery of frequent activating mutations [15] and it is now recognized as one of the major pathways that drive this disease. 1.1.5 T-ALL genetic abnormalities - mutations Numerous recurrent molecular abnormalities have been reported in T-ALL and can affect multiple cellular processes. Most frequent are deletions of the CDKN2A locus, found in over 70% of cases [16]. CDKN2A encodes for the tumor suppressors p16/INK4A and p14/ARF which are two of the key G1 cell-cycle checkpoint regulators. Other cell cycle regulators found to be genetically altered in T-ALL include RB1 (4%) [17] and CDKN1B (2%) [18], both of which can be seen to be deleted. In addition to cell cycle regulators, deletion or inactivating mutations can also be seen in transcription factors, chromatin remodelers, and signal transduction components in T-ALL. Recurrently deleted/mutationally inactivated transcription factors include WT1 (10%) [19], LEF1 (10-15%) [20], ETV6 (13%) [21], BCL11B (10%) [22],and RUNX1 (20%) [23]. Recurrently deleted/mutationally inactivated chromatin remodelers include EZH2 (10-15%) [24], SUZ12 (10%) [24], EED (10%) [24] and PHF6 (20-40%) [25]. Recurrently deleted/mutationally inactivated signalling components include FBXW7 (8-16%) [26, 27], PTEN (13-36%) [28, 29] and NF1 (3%) [30]. Activating mutations in T-ALL have also been reported and are often seen in signalling pathway components such as NOTCH1 (60%) [15], N-RAS (5-10%) [31], JAK1/3 (4-18%) [24, 32, 33] and IL7R (10%) [34]. 4 1.2 NOTCH 1.2.1 NOTCH signalling NOTCH proteins are a family of single pass transmembrane receptors with 4 family members represented in mammals (NOTCH1-4). All four of these paralogs are translated as single polypeptides that are subsequently cleaved to produce a heterodimeric complex with each half being held together non-covalently at the heterodimerization domain (HD) [35]. Canonical signalling is initiated through binding of cell surface ligands of the Delta-like (Delta-like 1, 3, and 4) or Jagged (Jagged 1, 2) family on adjacent cells. This binding causes a conformational change that exposes proteolytic cleavage sites at the juxtamembrane region and leads to sequential proteolytic processing by ADAM family members and the gamma secretase complex, resulting in the release of the intracellular NOTCH (ICN) domain [36, 37]. ICN translocates to the nucleus where it binds RBPJκ/CSL [38], converting it from a transcriptional repressor to a transcriptional activator through the recruitment of co-activators such as Mastermind-like family members (MAML1-3) [39]. This in turn leads to upregulation of context dependent target genes (Figure 1.1). NOTCH signalling is, in part, repressed through its subsequent proteosomal degradation, itself directed by the ubiquitination of its C terminal PEST (proline, glutamic acid, serine, threonine rich) domain by the E3 ligase FBXW7. In normal mammalian development, NOTCH1, along with its other family members NOTCH2-4, play important roles in cellular fate decisions (see section 1.2.2). Some of these functions show redundancy between NOTCH paralogs [40] whilst others remain unique [41]. 5 CSL/RBPJMAMLNOTCH1ExtracellularIntracellularICNS2S3 γ-secretaseγ-secretase inhibitor (GSI)MYCIL7RAIGF1RHES1PEST domainHD domain Figure 1.1 NOTCH1 signalling pathway Schematic representation of the NOTCH1 signalling pathway. S2/S3 = activating cleavage sites. Yellow stars indicate common mutation sites that enhance NOTCH1 signalling in T-ALL (activating mutations = NOTCH1 HD and PEST domains, inactivating mutation = FBXW7 mutations. NOTCH1 transcriptional co-activators, MAML and CSL/RBPJ, and examples T-ALL downstream transcriptional targets, MYC, IL7RA, IGF1R and HES1, indicated. 6 1.2.2 NOTCH signalling in normal development NOTCH1 signalling is required very early on in embryonic hematopoietic development, during the development of the first definitive hematopoietic stem cells (HSCs) [42, 43]. In addition subsequent fetal HSC development has also been shown to be dependent on NOTCH1 transcription through the use of transactivation domain (TAD) mutant mice [44]. Initial gain-of-function experiments showed that NOTCH1 increased adult HSC expansion [45, 46], however, subsequent loss-of-function experiments showed NOTCH1 signalling to be dispensable for adult HSC maintenance [47] and suggested that the previous experiments showing expansion to possibly be a result of supraphysiological levels of signalling. One of the most striking and best characterized roles of NOTCH1 in hematopoiesis is in intrathymic early T cell development. This is highlighted by the fact that mouse hematopoietic progenitors completely lacking NOTCH1 do not produce any mature T cells [48] in vivo whilst constitutive activation in this setting results in extrathymic T cell development [49]. NOTCH1 is also thought to play a role in many other developmental binary fate decisions outside hematopoiesis. Examples include the skewing of differentiation into secretory rather than absorptive cell types in intestinal epithelia [50] and the promotion of glial cell differentiation over a neuronal fate [51] in the nervous system. Moreover, during intestinal epithelial differentiation, NOTCH1 has been shown to be functionally redundant with NOTCH2 with inhibition of both required for secretory goblet cell skewing. The functions of NOTCH2-4 are less well characterized than those of NOTCH1. In addition to its redundant role in intestinal epithelium NOTCH2 is also described to be involved in directing proper marginal zone B cell development [52], CD8+ cytotoxic T lymphocyte activation [53], intrahepatic bile duct development [54] and erythroid lineage commitment [55]. NOTCH3 and 4 have a more limited tissue distribution than other NOTCH family members and, unlike NOTCH1 [56] and NOTCH2 [57], their deletion does not lead to embryonic lethality in mice [58, 59]. NOTCH3 is mainly expressed in vascular smooth muscle [60], where it is involved in postnatal maturation and arterial specification [58], as well as in the central nervous system [61] and in some thymocyte subsets [62]. Normal expression of NOTCH4 is primarily restricted to vascular endothelial cells, where its exact role is unknown [59]. Expression and functional 7 importance in megakaryopoiesis has also been suggested [63]. Given these important roles for NOTCH family members it comes as no surprise that they also have been implicated in the development of numerous malignancies. 1.2.3 NOTCH signalling in cancer NOTCH signalling has been described to act in both tumor suppressive and oncogenic fashions in various cancer types [64, 65]. Although evidence for some of these claims is tenuous and based on expression correlations, cases where activating or inactivation mutations are found more clearly suggest a definitive role. NOTCH signalling has been described as tumor suppressive in hepatocellular carcinoma (HCC) [66], chronic myelomonocytic leukemia (CMML) [67] and head and neck squamous cell carcinoma (HNSCC) [68, 69]. In CMML and HNSCC, inactivating mutations in NOTCH family members and critical pathway components have been reported; NOTCH2 and MAML mutations in CMML and NOTCH1/2/3 mutations in HNSCC. Conversely an oncogenic role for NOTCH signalling has been implicated in breast carcinoma, pancreatic adenocarcinoma, medulloblastoma, melanoma, non-small cell lung carcinoma (NSCLC), [70] as well as many hematopoietic malignancies including chronic lymphocytic leukemia (CLL) [71, 72], mantle cell lymphoma (MCL) [73], splenic marginal zone lymphoma (SMZL) [74], diffuse large B cell lymphoma (DLBCL) [75] and, most notably, T-ALL [15]. Many of these are ascribed to activating NOTCH mutations [15, 70-75], the most well characterised being found in NOTCH1 and are present in NSCLC, CLL, MCL and T-ALL. Similar activating mutations in NOTCH2 have also been reported in SMZL [76] and DLBCL [75]. Even NOTCH4, which is poorly described in normal physiology, has been found to be mutationally activated, along with other NOTCH family members, in rare cases of breast carcinoma [77, 78]. 1.2.4 NOTCH1 in T-ALL - mutations In a T-ALL context, NOTCH1 was first described as being aberrantly activated in a small fraction of patients that were shown to harbor the t(7;9)(q34;q34.3) translocation [14]. This translocation places a truncated NOTCH1 gene under the control of the TCRB locus. The resulting protein product is not only aberrantly expressed but, due to N-terminal truncation, is constitutively active. This truncation can either result in 8 membrane bound protein that is constitutively processed to produce ICN or a shorter form that mimics ICN and does not go to the membrane. The functional importance of such constitutively active NOTCH1 was shown in mouse studies where transduction of hematopoietic progenitors would produce murine T-ALL. However, because of the relative rarity of this translocation, identified in <1% of cases, NOTCH1 was given equal standing with many other rare translocation partners in T-ALL. The importance of NOTCH1 as an oncogene in T-ALL was recognized with the subsequent discovery of mutations in >60% of patients [15]. These mutations fall into two categories, both leading to an increase in signalling [15, 79]. Firstly, ~40% of patients harbor mutations in exons 26 and 27 that encode for the heterodimerization domain (HD). These can either be small HD destabilizing mutations leading to ligand hypersensitive/ligand independent activation of NOTCH1 or larger insertions that leave the ADAM cleavage site exposed. Secondly, ~30% of patients harbor frameshift mutations in exon 34. The resulting C terminally truncated protein lacks a PEST domain, preventing turn over of ICN by FBWX7 ubiquitination directed proteosomal degradation and allowing for persistent NOTCH1 signalling. HD and PEST mutations are not mutually exclusive and a subset of patients present with these mutations in cis. Related to these PEST mutations are inactivating mutations in FBWX7 which are found in ~8-16% of patients and also prevent NOTCH1 turn over [26, 27]. FBWX7 can also ubiquitinate and direct other T-ALL oncoproteins for proteosomal degradation and as such, may have more oncogenic potential than NOTCH1 PEST mutations alone. Activating mutations in NOTCH1 and FBXW7 mutations have been investigated extensively with regard to their prognostic potential [80-86]. It is believed that although they are associated with a better initial response to treatment the overall consensus is that they do improve long term outcome [83-86]. 1.2.5 NOTCH1 in T-ALL - regulation of downstream genes/pathways NOTCH1 has been described to regulate many genes and pathways in T-ALL transformation, some through direct transcriptional activation and others indirectly. Perhaps the most potent oncogenic described action of hyperactive NOTCH signalling is the upregulation of the direct transcriptional target c-MYC [87-89]. An increase in c-MYC expression is seen in numerous tumor types, including many lymphoid 9 malignancies, and is believed to be a major oncogenic driver. In T-ALL, c-MYC is thought to transcriptionally upregulate many anabolic genes, often in conjunction with NOTCH1, to drive cellular growth [87]. Its importance is also highlighted by the fact that enforced c-MYC expression alone can drive T-ALL leukemogenesis as well as restore growth to a fraction of T-ALL cell lines after NOTCH inhibition [88]. Other growth promoting pathways shown to be regulated by NOTCH1 in T-ALL include PI3K/AKT/mTOR signalling and NF-κB signalling. Inhibition of NOTCH1 results in a decreased phosphorylation of mTOR targets in T-ALL [90]. Mechanisms proposed for this action include repression of PTEN expression by the direct NOTCH1 target Hes1 as well as by direct transcriptional upregulation of upstream pathway inputs, namely IL7Rα [91] and IGF1R [92]. Likewise, NOTCH1 is described to promote signalling through NF-κB at multiple levels. NOTCH can directly interact with IKK to enhance its activity and furthermore directly promote transcription of NFKB2 and RELB [93]. NOTCH1 has also been reported to indirectly regulate NF-κB signalling through HES1, in this case through its repression of CYLD [94]. Proliferation is thought to be maintained by NOTCH1 through the regulation of multiple cell cycle components. NOTCH1 has been described to drive cell cycle progression by increasing expression of CCND3, CDK4, and CDK6 in T-ALL, with CCND3 being reported as a direct target [95]. Similarly decreased expression of a number of cell cycle inhibitory proteins by NOTCH1, including CDKN2D (p19/INK4d) and CDKN1B (p27/Kip1) [96], has also been reported. Cell migration in T-ALL is also promoted by NOTCH1 through the direct transcriptional upregulation of RHOU [97] as well as the chemokine receptors CCR5, CCR7 and CCR9 [98, 99]. With regard to CCR7, this has been described to be important for the clinically relevant phenotype of CNS infiltration. Given the important role NOTCH signalling plays in the T-ALL leukemogenic phenotype it is unsurprising that there has been great interest in therapeutic inhibition of NOTCH for the treatment of this disease. Strategies described to block NOTCH signalling include inhibition of the proteolytic gamma secretase complex using chemical inhibitors (GSIs) to prevent the release of ICN, small peptide inhibition of NOTCH1 transcriptional complex formation [100] and anti-NOTCH1 antibodies directed to stabilize the HD 10 domain [101, 102]. Of these approaches, the use of GSIs is the most heavily investigated owing to their ready availability. GSIs can effectively cause loss of cleaved ICN and downregulation of NOTCH1 target genes followed by G1/S cell cycle arrest in human T-ALL cell lines as well as primary mouse and human leukemias. However, only a small fraction of T-ALL cell lines are sensitive and an early phase I clinical trial showed no objective success [103]. In addition, gastrointestinal toxicity was observed, mostly likely due to the inhibition of NOTCH signalling driving the accumulation of secretory goblet cells [104]. A more recent trial has shown more encouraging results with GSI as a single agent [105], and early research indicates efficacy could be improved by combination with other treatments [106, 107]. In order to understand how cells might become resistant to GSI treatment, it is important to understand the molecular processes that oncogenic NOTCH1 signalling can control in T-ALL. 1.3 Thesis objectives T-ALL, like all cancers, is a complex disease driven by numerous molecular aberrations. One such aberrant event present in the majority of T-ALL cases is hyperactivation of the NOTCH1 signalling pathway. NOTCH1 can drive the expression and activity of many pro-oncogenic processes. Although many of these have been previously described in the literature, I believe that there is more to learn about how NOTCH1 can drive T-ALL. This area is of particular significance given the interest in therapeutically targeting this pathway in T-ALL. I hoped to do this by identifying not only novel downstream miRNA effectors of NOTCH1, but also by further investigation into pathways previously described by our lab as being altered by oncogenic NOTCH1, namely IGF1R and PKCθ signalling. By learning more about effects downstream of NOTCH1, not only can a greater understanding be gained of how hyperactive NOTCH1 drives this disease, but also the discovery of more broadly applicable processes irrespective of NOTCH1 activity. Ultimately, a greater understanding of aberrant T-ALL events would result in the opening of new avenues for therapeutic targeting, as well as elucidating potential mechansims for NOTCH1 inhibitor resistance observed in patients. 11 Aim 1: Investigate what miRNAs are regulated downstream of NOTCH1 in T-ALL. Aim 2: Investigate the effects of IGF1R inhibition and characterize important IGF1R downstream signalling required for growth in a T-ALL context. Aim 3: Find direct PKCθ phosphorylation targets in a T-ALL context. 12 Chapter 2 NOTCH-mediated repression of miR-223 contributes to IGF1R regulation in T-ALL 2.1 Introduction 2.1.1 miRNAs - general background MicroRNAs (miRNAs) are a class of endogenously encoded small non-coding RNAs. They are involved in the post transcriptional regulation of most mammalian genes through inhibition of mRNA translation and/or modulating the stability of messenger RNA (mRNA) [108]. This regulation is based on complementary base pairing between the mature miRNA and its target mRNA, often in the 3’ untranslated region (3’UTR). The degree of the complementarity between the miRNA and its target can dictate how the gene is repressed. Partial complementarity is generally thought to promote translational repression and exonucleolytic mRNA decay, whereas complete complementarity promotes endonucleolytic mRNA cleavage (although other mechanisms of repression have been suggested [109]). To recognize their target mRNA with partial complementarity, miRNAs must still have perfect complementarity in their 5’ “seed” region. Minimally, this seed region consists of 6 nucleotides (6 mer) of perfect complementarity (nucleotides 2-7), however a 8 mer or 7 mer can increase repressive efficacy [110]. The relatively recent discovery of miRNAs added to the complexity of gene regulation networks, highlighted by the fact that a single miRNA species can target multiple mRNAs, predicted to be on average >100, and a single mRNA being able to be targeted by multiple miRNAs [111]. Some miRNAs are described to have relatively modest effects on a single target gene’s expression but when considering this complex layering, the idea of miRNAs fine tuning the expression of large networks of genes can be appreciated. The miRBase catalogue (v20, June 2013) contains a record of 30,424 mature miRNA from 206 species, including 2588 mature miRNAs from human and 1915 from mouse [112] and they have been implicated in virtually all mammalian cellular processes. 2.1.2 T cell development and miRNAs T-cell development and function are very well characterized processes. Common lymphoid progenitors resident in the bone marrow give rise to T-cell committed 13 progenitors that migrate to the thymus, where they undergo sequential development stages categorized by the expression of the surface markers CD4 and CD8. Immature progenitor cells are negative for both these markers, double negative (DN), and during this stage, begin to genetically rearrange their T-cell receptor (TCR) β-chain locus. Productive re-arrangements pair with the pre-Tα chain and halt further β-chain locus rearranging, upregulate both CD4 and CD8 to become double positive (DP), and initiate a burst of proliferation. At this DP stage, T cells also rearrange their TCR α-chain locus and undergo positive selection to remove αβTCR complexes that are unable to recognize peptide/major histocompatibility complex (MHC) complexes. DP cells that recognize self MHC class II molecules will eventually downregulate CD8 to become mature naïve CD4+ single positive (SP4) T-cells. Likewise, DP cells that recognize self MHC class I molecules will eventually downregulate CD4 to become mature naïve CD8+ single positive (SP8) T-cells. Also, during this process if the affinity for the self peptide/MHC complex is too strong, negative selection occurs to remove potentially harmful self-reactive T-cell clones. Mature naïve T cells emigrate from the thymus and enter the periphery where they are presented antigens. Upon recognition of an appropriately presented antigen, T cells are activated and exert their effector immune functions. CD8+ T-cells have a cytotoxic effector function and mediate direct cell killing of cells presenting their recognized antigen. CD4+ T-cells have a helper effector function and direct an immune response through the secretion of various cytokines. Several variants of CD4+ T helper cells have been described in the literature (Th1, Th2 [113], Th17 [114], TFH [115], Treg). Each is directed in their differentiation by the milieu of cytokines they are activated in, and produce distinct cytokines to direct unique immune responses. Early studies alluded to the importance of miRNAs in hematopoietic cell differentiation through their enforced expression skewing the relative abundance of mature subsets [116]. More specifically, in T-cells, the importance of miRNAs was established through the use of conditional deletion of the requisite miRNA processing RNase dicer1, whereby CD8+ T cell development was blocked and aberrant CD4+ T cell differentiation and cytokine production was observed [117]. Subsequent identification of specific miRNAs involved in T-cell development and function has followed with some shown to 14 be directly regulated by pathways already heavily implicated in these processes [118, 119]. For example, NF-κB mediated transcription downstream of TCR activation has been shown to upregulate miR-146a and miR-155 to inhibit activation-induced T-cell death via Fas-associated death domain (FADD) repression [120] and enhance T-cell IL-2 induced proliferation via repression of suppressor of cytokine signalling 1 (SOCS1) [121], respectively. Other miRNAs well characterized in T-cell development/differentiation include miR-181a and those of the miR-17~92 cluster (miR-17-3p, miR-17-5p, miR-18a, miR-19a, miR-20a, miR-19b and miR-92). miR-181 is critical in regulating TCR signal strength and, as such, positive/negative selection through the repression of a number of phosphatases [122]. miRNAs from the miR-17~92 cluster are described to repress a number of genes critical for maintaining normal differentiation including PTEN, transforming growth factor beta receptor II (TGFBR2), cAMP responsive element binding protein 1 (CREB1), BIM, and E2F1. They impact almost all types of T-cells either positively, e.g. overexpression promoting CD8+ differentiation, or negatively, e.g. knockout increasing Treg numbers. With the ability to induce such profound phenotypic effects, it comes as no surprise that miRNAs such as those from the miR17~92 cluster have been implicated in many hematological malignancies [123]. 2.1.3 MicroRNAs in hematological malignancies miRNAs have become a key focus area in cancer research. Their dysregulation in malignant progression has been investigated and they are found to act as both oncogenes (oncomiRs) and tumor suppressors. In addition, their potential use as disease biomarkers and for therapeutic treatment has also been explored. The very first study reporting dysregulation of miRNAs in cancer was published as recently as 2002 and described the loss of miR-15 and miR-16 in the commonly deleted region of 13q in B-cell chronic lymphocytic leukemia (CLL) [124]. This loss was subsequently confirmed to lead to the malignant phenotype, partly by de-repressing the critical anti-apoptotic regulator BCL2 [125]. Since then, numerous miRNA species have been implicated in nearly all cancer types, including many hematological malignancies. In T-ALL specifically, numerous miRNAs have been implicated in disease progression or maintenance. Mature miRNAs from the miR17~92 cluster are reported to be 15 upregulated by the T-ALL oncogenes NKX2-5, TLX1 and TLX3 to reduce apoptosis via the repression of E2F1 [126]. miR-19, a specific member of this miR17~92 cluster, also supports T-ALL growth via repression of CYLD [127] and several members of the PI3K/AKT signalling axis (BIM, PrkAA1, PTEN and PP2A) [128]. Other examples of potential T-ALL oncomiRs include miR128-3p, via PHF6 repression [129], miR-181a via EGR1 repression [130], and miR-16 via PDCD4 repression[131]. In T-ALL, miRNAs have also been described as acting in a tumor suppressive fashion, such as the action of miR-451 to repress of c-MYC [132]. Other miRNAs have been investigated as potential prognostic markers in T-ALL, for example miR-221 [133] and miR-16 [134] have been found to be associated with overall survival. 2.1.4 Statement of hypothesis and objectives Given that NOTCH1 is an important oncogene in T-ALL, acts as a nuclear transcriptional factor, and has pleotropic effects that are highly cell context-dependent, I hypothesize that NOTCH1 might directly regulate a set of miRNAs that contribute to its cell context-dependent, oncogenic activity. The objectives to address this hypothesis include, 1) deriving a comprehensive map of miRNA expression in the context of T-ALL that is specifically dependent on Notch signaling, 2) to identify candidate mRNA species that are regulated by Notch-dependent miRNAs, and 3) to assess the effect of these miRNAs on cell growth related cellular phenotypes. Aim 1: Investigate what miRNAs are regulated downstream of NOTCH1 in T-ALL. 2.2 Materials and methods 2.2.1 Cell culture Human T-ALL cell lines (ALLSIL, DND41, HPBALL, Jurkat, KOPTK1, P12 Ichikawa, PF382, RPMI 8402) were grown in RPMI 1640 medium supplemented with 10% FBS, 1 mM sodium pyruvate, 2 mM L-glutamine, and antibiotics. Mouse T-ALL leukemias, generated as previously described through the enforced expression of a constitutively active version of NOTCH1 (ΔE) in murine bone marrow [92, 135], were cultured ex-vivo 16 in RPMI 1640 medium supplemented with 20% FBS, 10 ng/mL mouse IL-2 (Peprotech), 10 ng/mL mouse IL-7 (Peprotech), 2 mM L-glutamine, and antibiotics. 2.2.2 miRNA expression profiling Human T-ALL cell lines (Jurkat and P12 Ichikawa) and in vitro adapted mouse primary T-ALL leukemias were treated with compound E (Calbiochem), a potent γ-secretase inhibitor (GSI), to block NOTCH signalling. After 4 days GSI was washed out and replaced with media with or without fresh GSI for 24 hours. Cells were then harvested and total RNA extracted using TRIzol reagent (ThermoFisher). Sample RNAs were profiled by Exigon using their miRCURY LNA™ microRNA array system (v11.0). Briefly, total RNA for each sample was labelled with Hy3, mixed with Hy5-labelled RNA from a common reference pool, and hybridized to miRCURY LNA™ microRNA arrays containing capture probes for 1264 different miRNA species. The common reference pool was generated by mixing all sample RNAs for the same species from the study. Arrays were scanned to obtain both Hy3 and Hy5 signals. Capture probes with both Hy3 and Hy5 signals greater than 1.5x of the median signal intensity received “present” calls, while those failing this detection threshold were excluded from further analysis. Expression level for each miRNA species deemed “present” is reported as log2 (Hy3/Hy5) ratio. Ratios > +/-0.58 (Fold change > 1.5) were deemed meaningful changes. Each sample was assessed by microarray once. 2.2.3 Real-time quantitative PCR To assess GSI inhibition prior to miRNA profiling, a fraction of the RNA sample was reverse transcribed into cDNA using SuperScript III (ThermoFisher) and a mix of random 15-mer and anchored oligo(dT)20+1 primers. Platinum SYBR Green qRT-PCR SuperMix-UDG (ThermoFisher) and the following specific primer sets were used to assess DELTEX and β-actin expression: hDTX1 forward 5′-CTA TGA CAT CCC CAC AGG CAT C-3′, hDTX1 reverse 5′-ACG TGC CGA TAG TGA AGA TGA GT-3′, and hActB forward 5′-CGCGAGAAGATGACCCAGAT-3′, hActB reverse 5′-GAT AGC ACA GCC TGG ATA GCA AC-3′, mDTX1 forward 5′-TGT CTT CTC CCC GTA GAT GG-3′, mDTX1 reverse 5′-TGC TCC GAA ACA AGA GTG TG-3′, mActB forward 5′- CTT CTA CAA TGA GCT GCG TGT G-3′, mActB reverse 5′- TTG AAG GTC TCA AAC ATG ATC 17 TGG-3′. Each sample was assayed in triplicate using a Dyad Disciple thermal cycler equipped with Chromo4 optical head (Bio-Rad Laboratories). Expression levels were calculated by the ΔΔCt method with normalization to β-actin. Initial real-time miR-223 and miR-223* qRT-PCR studies were performed by “Exiqon Services” as contract work. Briefly, 10 ng total RNA was reverse transcribed in 10 µl reactions using the miRCURY LNA™ Universal RT microRNA PCR, Polyadenylation and cDNA synthesis kit (Exiqon); each sample was processed in triplicates. cDNA was diluted 80x and 4 μl was used in 10 μl PCR reactions according to the protocol for miRCURY LNA™ Universal RT microRNA PCR; each miRNA was assayed once by qRT-PCR in triplicate cDNA. The amplification was performed in a LightCycler® 480 Real-Time PCR System (Roche) in 384 well plates. The following assays were performed: miR-223, miR-223*, and as reference miR: miR-103, miR-191, and miR-423-5p. LightCycler® 480 software was used to determine the Cp value, and to generate amplification and melting curves. LinRegPCR (version 11.5) software was used to determine the amplification efficiency. The average amplification efficiency was used to correct the Raw Cp values. Reference miRNAs were analyzed for stability using SLqRT-PCR algorithm (similar to geNorm) and miR-191 selected as the most stable reference miRNA and subsequently used to normalize all measurements on a well-to-well basis. Subsequent miR-223 real-time qRT-PCR studies were performed by generating first-strand cDNA from total RNA by reverse transcription with SuperScript III (ThermoFisher) using stem-loop primers specific for miR-223 (5’-GTT GGC TCT GGT GCA GGG TCC GAG GTA TTC GCA CCA GAG CCA ACG GGG TA-3’ and a normalization control miRNA, miR-191 (5’-GTT GGC TCT GGT GCA GGG TCC GAG GTA TTC GCA CCA GAG CCA ACC AGC TG-3’). qRT-PCR amplification using Platinum SYBR Green qRT-PCR SuperMix-UDG (ThermoFisher) was then carried out using primers specific for miR-223 (5’-CCG CCC GTG TCA GTT TGT CA-3’) and miR-191 (5’-CCC GCA ACG GAA TCC CAA AAG-3’) in combination with a universal stem-loop primer (5’-GTG CAG GGT CCG AGG T-3’). miR-191 was chosen as normalization control based on its expression stability between samples using the miRCURY LNA™ Universal RT system and on the melt curve plots (data not shown). Each sample was assayed in triplicate using a Dyad Disciple thermal 18 cycler equipped with Chromo4 optical head (Bio-Rad). miR-223 expression levels were calculated by the ΔΔCt method with normalization to miR-191. 2.2.4 Luciferase reporter assay A portion of the IGF1R 3’UTR containing putative miR-223 target sites was cloned downstream of the luciferase gene in the pGL3 reporter plasmid (Promega). Luciferase reporter and miRNA expressing plasmids were transfected at a 2:1 ratio into HeLa cells using Lipofectamine Plus (ThermoFisher) along with Renilla luciferase transfection control plasmid. Luciferase reporter activity was measured 42 hours later using the Dual Luciferase Reporter (DLR) assay kit (Promega) and read with a tube luminometer (Lumat LB 9507). 2.2.5 Viral vectors and transduction High titer, replication defective retrovirus was produced by transient transfection of 293T producer cells as described [88]. Lentivirus was produced in a similar manner, but utilizing pCMVΔR8.74, pCMV-VSV-G, and pRSV-Rev packaging vectors. HES1 and c-MYC were overexpressed from the MSCV-IRES-GFP (MIG) retroviral vector. miR-223 was overexpressed from the MSCV-PGK-GFP retroviral vector [136]. miR-223 knockdown was achieved using lentivirus overexpressing miR-223 target sequences from a polymerase II promoter [137]. Doxycycline-inducible expression of dominant negative Mastermind-like 1-GFP fusion protein (DnMAML) [138] was achieved using pLVX-Tet-On Advanced and pLVX-Tight-puro lentivectors (Clontech), the latter containing DnMAML cDNA insert. Viral transduction was performed by spinoculation with 4 μg/mL polybrene as previously described [88]. Transduced cells were selected with G418 and/or puromycin as appropriate, or FACS sorted by GFP expression. Doxycycline-inducible expression was achieved by treatment of cells with 500 ng/mL doxycycline (Sigma) for 4 days. 2.2.6 Western Blot Whole cell protein extracts were prepared using RIPA buffer (50 mM Tris-HCl pH 8, 150 mM NaCl, 1% NP-40, 0.25% Na-deoxycholate, 1 mM EDTA) supplemented with 1 mM NaF, 1 mM Na3VO4, 2.5 mM Na-pyrophosphate, 1 mM phenyl-methylsulfonyl-fluoride and 1x Protease Inhibitor Cocktail Set III (Calbiochem). Lysates were cleared by 19 centrifugation at 14,000 × g for 10 minutes at 4 °C, separated by SDS-PAGE, and transferred to PVDF membranes. Blots were probed with antibodies directed against IGF1Rα (sc-712, Santa Cruz) and ERK2 (sc-154, Santa Cruz), followed by HRP-conjugated secondary antibody and detected by chemiluminescence. Band intensities were quantified using ImageJ software. 2.2.7 Flow cytometry Fresh cells were stained with antibodies directed against IGF1R (αIR3; Calbiochem) followed by APC-conjugated donkey anti-mouse IgG secondary antibody and analyzed by flow cytometry using a FACSCalibur cytometer (BD Biosciences) and FlowJo software (Treestar). Virally transduced cells were sorted by GFP expression using an Influx sorter (Cytopeia). 2.2.8 Gene expression profile analysis Gene and miR-223 expression data for T-ALL cell lines used for correlation analysis was downloaded from Cancer Cell Line Encyclopedia (CCLE) project (http://www.broadinstitute.org/ccle/home) and Sanghvi et al. [139], respectively. CCLE data comes from Affymetrix U133+2 arrays with and raw Affymetrix CEL files converted to a single value for each probe set using Robust Multi-array Average (RMA) and normalized using quantile normalization. miR-223 data comes from qRT-PCR miRNA expression profiles, with mean expression value of all expressed miRNAs in a given sample as the normalization factor 2.2.9 Statistics Prism 6 software (GraphPad) was used for statistical analyses. 2.3 Results 2.3.1 miRNA microarray experimental setup My initial hypothesis was that directly regulated miRNAs would be upregulated by NOTCH1, because of its role as a canonical transcriptional activator. To this end, an approach which would enrich for direct transcriptional miRNA targets of NOTCH1 was designed. This approach involved repressing NOTCH1 activity through the use of a 20 gamma secretase inhibitor (GSI) for a prolonged period of time, followed by washing out with or without GSI drug and miRNA assessment. Such an approach was designed to minimise the possible problem of the long half-life of miRNA and, additionally, favour direct transcriptional miRNA targets, as opposed to indirect/downstream targets that would show regulation at longer time points. I assessed the kinetics of direct NOTCH1 transcription target repression by GSI over time by monitoring mRNA transcript levels of the canonical NOTCH1 transcription target DELTEX1 in both human T-ALL cell lines and mouse derived T-ALLs. In both human and mouse T-ALL cells, near maximal repression of DELTEX1 was seen after 2 days of GSI treatment (Figure 2.1A/1B). Because of the purported longer half-life of some miRNA species [140, 141] relative to mRNA, it was decided to double this timeframe and perform GSI washout at a 4 day time point, followed by miRNA assessment at 24 hours post washout, when mRNA (DELTEX1) levels had increased in the mock washout relative to the culture maintained in GSI. This timeframe is also supported by research showing that, in T-ALL cell lines, shutting off miRNA transcription for 4 days results in ~20% expression compared to unmanipulated cultures [141]. 21 Figure 2.1 Kinetics of NOTCH1-regulated mRNA in T-ALL Real-time qRT-PCR analyses expression level of canonical direct NOTCH1 target DELTEX1. (A) Human T-ALL cell lines Jurkat and P12 Ichikawa and (B) three primary mouse T-ALL leukemias (H15-1, H3-2-6, H26-2) were treated with GSI to block NOTCH (A) (B) 0 2 4 4 8 7 2051 01 52 0P 1 2 Ic h ik a w aH o u rsRelative hDTX1G S I-w a s h -G S I G S I-w a s h -D M S O9 6W a s h o u t1 2 0H a r v e s t0 2 4 4 8 7 201 02 03 04 0J u r k a tH o u rsRelative hDTX1G S I-w a s h -G S I G S I-w a s h -D M S O9 6W a s h o u t1 2 0H a r v e s t0 2 4 4 8 7 205 01 0 01 5 0H 1 5 -1H o u rsRelative mDTX1G S I-w a s h -G S I G S I-w a s h -D M S O9 6W a s h o u t1 2 0H a r v e s t0 24 48 7201234H3-2-6HoursRelative mDTX1GSI-wash-GSI GSI-wash-DMSO96Washout120Harvest0 2 4 4 8 7 205 01 0 01 5 02 0 0H 2 6 -2H o u rsRelative mDTX1G S I-w a s h -G S I G S I-w a s h -D M S O9 6W a s h o u t1 2 0H a r v e s t22 signalling for 4 days in vitro followed by a washout and replacement with media containing GSI or DMSO for 24 hours. Cells were harvested and total RNA isolated for qRT-PCR analysis normalized to β-actin control. hDTX and mDTX expression values are given relative to lowest expressing sample within each experiment. 2.3.2 Microarray profiling of NOTCH1-regulated miRNAs in T-ALL To identify miRNAs that are regulated by NOTCH1 signalling, miRNA expression profiling of two human T-ALL cell lines (Jurkat and P12 Ichikawa), and three mouse derived T-ALLs (H3-2-6, H15-1, and H26-2) was performed. For this initial discovery/screening step, the Exigon miRCURY LNA™ microRNA array platform (v11.0 hmr) which features locked nucleic acid capture probes was utilized. For the human samples, 33-37% of miRBaseV13 human/miRPlus (proprietary) miRNA species were detected in the samples. Unexpectedly, only 3 miRNAs were identified to be significantly altered upon GSI washout in Jurkat cells, whereas no miRNAs appeared significantly changed in P12 Ichikawa cells (Figure 2.2A). miRPlus-A1065 was found to be increased by approximately 1.5-fold when NOTCH1 signalling was activated by GSI washout, whereas miR-223, along with its complementary “star” strand (miR-223*), were found to be decreased approximately 1.5-fold (Figure 2.2B). miRPlus-A1065 is predicted by TargetScan 5.1 software to target lysophospholipase II (LYPLA2) and collagen, type XXII, α1 (COL22A1). At the time of this study miRPlus-A1065 was an Exigon proprietary miRNA species and not otherwise reported in the literature. miRPlus-A1065 was subsequently re-named miR-1307 in later releases of miRBase. In contrast, miR-223 has previously been described to be involved in several aspects of normal and malignant hematopoiesis [142-144] and thus, I elected to focus further studies on miR-223. For the mouse samples, 43-49% of miRBaseV13 mouse miRNA species were detected in the samples. Unfortunately, relatively few miRNAs were also identified to be significantly altered upon GSI washout in the mouse T-ALL samples. One leukemia, H26-2, showed no miRNAs changes, and samples from leukemias H15-1 and H3-2-6 showed only modest changes after GSI washout (0.38-2.44 fold), with no individual species showing significant changes in both. Because of these relatively low fold 23 changes and lack of consistency between the mouse and human derived samples, I decided to pursue miR-223 in a human context. Mature miR-223 expression was not detected in the mouse T-ALL samples (Figure 2.3). Figure 2.2 Microarray profiling of NOTCH1-regulated miRNAs in human T-ALL miRNA microarray profiling. Total RNA from Jurkat and P12 Ichikawa cells 24 hours after washout, as described in Fig.1A, was sent for miRNA microarray analysis using the Exigon miRCURY LNA™ microRNA Array system. Log median fluorescence ratios (LMR) were determined for hybridization of sample miRNAs over control pool miRNAs. (A) Graphical depiction of LMRs of GSI-wash-GSI treated vs. GSI-wash-DMSO treated cells. Hashed lines represent the variation threshold (ΔLMRs > +/-0.58) and red dots identify miRNA species whose variation exceeded this threshold. (B) Graphical Jurkat-3 -2 -1 0 1 2 3-3-2-10123miR-223miR-223*miRPlus-A1065GSI-wash-DMSO treated(Log Median Ratio)GSI-wash-GSI treated(Log Median Ratio)P12 Ichikawa-3 -2 -1 0 1 2 3-3-2-10123GSI-wash-DMSO(Log Median Ratio)GSI-wash-GSI(Log Median Ratio)JurkatmiRPlus-A1065 miR-223 miR-223*0.00.51.01.52.0GSI-wash-DMSOGSI-wash-GSIfold change over GSI-wash-DMSO(A) (B) 24 depiction of fold changes calculated from microarray ΔLMRs for the three miRNA species exceeding the minimum variance threshold (Δ1.5 fold) in Jurkat cells. Figure 2.3 Microarray profiling of NOTCH1-regulated miRNAs in mouse T-ALL miRNA microarray profiling. Total RNA from primary mouse T-ALL leukemias (H15-1, H3-2-6, H26-2) 24 hours after washout, as described in Figure 2.2B, was sent for miRNA microarray analysis using the Exigon miRCURY LNA™ microRNA Array system. Log median fluorescence ratios (LMR) were determined for hybridization of sample miRNAs over control pool miRNAs. Graphical depiction of fold changes calculated from microarray ΔLMRs for the miRNA species exceeding the minimum variance threshold (Δ1.5 fold) in at least one sample (miR-223 also shown). *, >Δ1.5 fold change; nd, not determined. mmu -miR-34 0-5pmmu -miR-21 2mmu -miR-70 8mmu -miR-88 7*mmu -miR-19 3mmu -miR-14 7mmu -miR-71 5mmu -miR-45 5mmu -miR-48 3* 'mmu -miR-29 6-5pmmu -miR-22 30 .00 .51 .01 .52 .02 .53 .0H 1 5 -1fold change over GSI-wash-DMSOG S I-w a s h -D M S OG S I-w a s h -G S I* * *N Dmmu -miR-34 0-5pmmu -miR-21 2mmu -miR-70 8mmu -miR-88 7*mmu -miR-19 3mmu -miR-14 7mmu -miR-71 5mmu -miR-45 5mmu -miR-48 3*mmu -miR-29 6-5pmmu -miR-22 30 .00 .51 .01 .52 .02 .53 .0H 3 -2 -6fold change over GSI-wash-DMSOG S I-w a s h -D M S OG S I-w a s h -G S I**** ** *N D N Dmmu-miR-340-5pmmu-miR-212mmu-miR-708mmu-miR-887*mmu-miR-193mmu-miR-147mmu-miR-715mmu-miR-455mmu-miR-483*mmu-miR-296-5pmmu-miR-2230.00.51.01.52.02.53.0H26-2fold change over GSI-wash-DMSO GSI-wash-DMSOGSI-wash-GSINDND25 2.3.3 NOTCH1 regulates miR-223 in human T-ALL cells Because the microarray showed miR-223 (and miR-223*) to have lower expression in a NOTCH1 active state, the assay to assess expression changes was changed from the GSI washout assay described above to a straight +/- GSI/mock (DMSO) experiment. To validate the microarray results, I performed real-time qRT-PCR analysis for miR-223 and miR-223* in Jurkat and P12 Ichikawa cells, as well as in additional human T-ALL cells lines. Initial validation was carried out using the Exiqon qRT-PCR service that, like the miRCURY LNA™ microRNA array platform, utilizes locked nucleic acid (LNA) technology to allow for specific miRNA detection. The qRT-PCR assay for miR-223* yielded expression results at or below the assay detection limit, as well as indicating of non-specific amplification based on melt curves patterns, and further validation could not be pursued. In contrast, miR-223 was readily detected, and expression levels increased significantly with GSI treatment in Jurkat, but not in P12 Ichikawa cells, thus confirming the microarray findings (Figure 2.4A). As well, 2 out of 3 additional human T-ALL cell lines (PF382, RPMI 8402) showed significantly increased miR-223 expression with GSI treatment, while the remaining cell line (ALLSIL) exhibited miR-223 expression near the limit of detection and, as such, miR-223 expression level could not be determined. qRT-PCR validation with these same 5 cell lines along with 3 additional lines (DND41, KOPTK1, HPBALL) was undertaken using my own miR-223 specific stem-loop primer assay. DND41 and KOPTK1 also showed a significant increase in miR-223 levels after 4 days GSI treatment, whilst HPBALL, (similar to ALLSIL, showed miR-223 at the limit of detection (Figure 2.4B). The similarity of the results between the LNA based qRT-PCR and the stem-loop primer approaches gave me confidence in the accuracy of the assay. Of note, both GSI “sensitive” and “resistant” cell lines, the former of which undergo G0/G1 cell cycle arrest upon treatment with GSI [15, 26, 27], showed regulation of miR-223 by NOTCH, suggesting that changes in miR-223 occur regardless of effects on cell cycle progression. Subsequent time course analysis showed miR-223 expression to increase gradually, reaching a maximum after 3-4 days of GSI treatment (Figure 2.4C). To exclude that miR-223 upregulation was due to effects of GSI on targets other than NOTCH signalling, T-ALL cells were transduced with lentiviral vectors 26 encoding a doxycycline-inducible dominant negative MAML1 (DnMAML) the expression of which blocks NOTCH signalling [15, 138]. After induction of DnMAML we also observe upregulation of miR-223 (Figure 2.4D). Taken together, these results support the conclusion that NOTCH1 signalling negatively regulates the expression of miR-223 in human T-ALL cells. Of note, mouse acute T-cell leukemias generated experimentally by retroviral transduction of bone marrow with activated NOTCH1 [135] showed no detectable miR-223 expression even after GSI treatment suggesting NOTCH1 may only regulate miR-223 in human cells (data not shown). R P M I 8 4 0 20 2 0 4 0 6 0 8 0 1 0 0012345h o u rs o f G S I- tre a tm e n tmiR-223(fold over mock)JurkatP12 IchikawaPF382RPMI 8402ALLSIL01234miR-223(fold over mock)mock GSI*ns****ndJurka tP 12 Ich ik awaP F3 82R PMI 84 02D ND 41K OP TK 102468m o c k G S I************n sG S I re s is ta n t G S I s e n s itiv emiR-223(fold over mock)(A) (C) (B) (D) 01234 **RPMI 8402 TetOn-DnMAMLmock doxycyclinemiR-223(fold over mock)27 Figure 2.4 Figure 2.4 NOTCH1 represses miR-223 Real-time quantitative PCR (qRT-PCR) analyses of mir-223 expression level. Human T-ALL cell lines were treated with GSI to block NOTCH signalling for 3-4 days vs. DMSO vehicle control (mock) and assessed by (A) Exiqon miRCURY LNA™ Universal RT microRNA qRT-PCR or (B) miR-223 (and miR-191) specific stem-loop primers. (C) Time course analysis following NOTCH inhibition with GSI. RPMI 8402 cells were treated with GSI vs. mock and cells harvested at time points indicated. (D) RPMI 8402 cells with doxycycline-inducible expression of dominant negative Mastermind-like 1 (DnMAML) to block NOTCH signalling. Cells were treated with 500 ng/mL doxycycline for 4 days vs. ethanol vehicle control (mock). ns, not significant; nd, not determined; *, p<0.05; **, p<0.01; ***, p<0.001 (Student’s t-test). 2.3.4 NOTCH1 mediated repression of miR-223 does not involve HES1 or c-MYC Because canonical NOTCH signalling results in transcriptional upregulation of its target genes, the observation that inhibition of NOTCH signalling leads to increased miR-223 expression suggests the involvement of an intermediary repressor protein or proteins. That is, it would be hypothesized that NOTCH signalling leads to increased expression of a transcriptional repressor that, in turn, represses transcription of miR-223. Perhaps the best characterized transcriptional repressor induced by NOTCH signalling is HES1 [145-147], although other members of the HES/HEY family of transcriptional repressors are also regulated in T-ALL cells by NOTCH1 [87-89]. To determine if HES1 may be responsible for NOTCH-dependent miR-223 repression, T-ALL cell lines were transduced with a HES1 expressing retrovirus, then treated with GSI to see if enforced HES1 expression might abrogate the increase in miR-223 expression level. In 4 out of 5 cell lines tested, enforced HES1 did not prevent induction of miR-223 with GSI (Figure 2.5A) suggesting other intermediary repressors are involved. Another important direct target of NOTCH1 that can mediate many important downstream expression changes is the strong T-ALL oncogene c-MYC. In fact, c-MYC over expression can reverse the sensitivity of some T-ALL cell line to GSI and allow their continued growth in a NOTCH “off” state [88]. c-MYC is a transcriptional activator and as such, would not be predicted to directly repress miR-223 expression. However, to test if it does play at least an indirect role I assessed miR-223 level changes after GSI treatment in KOPTK1 with 28 enforced c-MYC expression. Despite being previously shown to restore proliferation after GSI treatment, in this cell line, c-MYC did not abrogate the ~3 fold increase in miR-223 expression level (Figure 2.5B). Thus the mechanism for how miR-223 is repressed by NOTCH1 in human T-ALL remains unknown. Figure 2.5 NOTCH1 does not indirectly repress miR-223 through the canonical effectors HES1 or c-MYC Real-time quantitative PCR (qRT-PCR) analyses of miR-223 expression level. T-ALL cell lines were transduced with MIG retroviruses to overexpress (A) HES1 or (B) c-MYC, FACS sorted by expression of the linked retroviral GFP marker, then treated with GSI vs. mock for 3-4 days. Error bars indicate standard deviation of assays performed in triplicate. ns, not significant; *, p<0.05; **, p<0.01; ***, p<0.001 (Student’s t-test). 2.3.5 miR-223 targets IGF1R TargetScan 5.1 software was used to identify potential miR-223 human mRNA targets based on seed region matches to 3’UTRs as well as predicted targeting efficacy (Table 2.1A). The resulting target list was analyzed using Ingenuity Pathway Analysis software to identify signalling pathways that have an overrepresentation of predicted miR-223 targets. This analysis yielded insulin-like growth factor-1 (IGF-1) signalling as the top ranked pathway (Table 2.1B), with 9/107 annotated members, representing both the Jurka tP F3 82RP MI 84 02DND4 1K OP TK 1024681 0m o c k G S I**** *****n sM ig H E S 1 tra n s d u c e dmiR-223(fold over mock)mo ckGS I012345K O P T K 1 M ig c -m y c tra n s d u c e dmiR-223(fold over mock)*(A) (B) 29 PI3K/AKT and MAPK downstream signalling arms, as well as IGF1R itself (Figure 2.6). In order to focus in on which IGF-1 signalling predicted miR-223 targets may be most efficiently repressed, I assessed if there was any correlation between their transcript levels and miR-223 expression from publically available microarray data (Table 2.2) [139, 148]. Of the 9 predicted miR-223 targets in the IGF-1 signalling pathway, only one, IGF1R, significantly negatively correlated with miR-223 in a cohort of 10 T-ALL cell lines (Spearman r = 0.6606, p = 0.0438) (Figure 2.7A). IGF1R is of particular interest given that I have recently been involved in reporting IGF1R to be upregulated by NOTCH1 signalling, and that IGF1R plays an important role in T-ALL cell growth and leukemia-initiating activity [92]. The 3’UTR of human IGF1R contains a 8mer miR-223 seed match that is conserved in many placental mammals (position 224-231 of human IGF1R 3' UTR) (Figure 2.7B/C). To address the possibility that miR-223 may contribute to IGF1R upregulation by NOTCH1 (i.e. NOTCH1 represses miR-223 expression, which leads to upregulation of IGF1R), a portion of the IGF1R 3’ UTR containing predicted miR-223 target sequences was cloned downstream of a luciferase reporter gene. Overexpression of miR-223 resulted in decreased luciferase reporter activity (Figure 2.8), suggesting that miR-223 can indeed target the IGF1R 3’UTR and thus potentially antagonize IGF1R protein expression. Interestingly, IGF1R has also been predicted and experimentally shown to be targeted and repressed by miR-223* [136]. 30 Table 2.1 Ingenuity pathway analysis predicts miR-223 targets regulate IGF signalling Bioinformatic identification of potential miR-223 target genes. (A) Predicted miR-223 targets according to TargetScan 5.1 software analysis. Genes are ranked according to total context score. Top ten genes and selected IGF-1 signalling pathway elements are depicted. IGF-1 pathway elements are indicated by arrowheads. (B) Ingenuity Pathway Analysis performed using the full predicted miR-223 target gene list. The top 5 identified pathways are listed. 1 FBXW7 F-box and WD repeat domain containing 7 -1.362 MYO5B myosin VB -0.853 APC adenomatous polyposis coli -0.784 RHOB ras homolog gene family, member B -0.735 PURB purine-rich element binding protein B -0.76 SLC4A4 solute carrier family 4, sodium bicarbonate cotransporter, member 4 -0.697 FOXO1 forkhead box O1 -0.668 ADCY7 adenylate cyclase 7 -0.669 ECT2 epithelial cell transforming sequence 2 oncogene -0.6510 KPNA1 karyopherin alpha 1 (importin alpha 5) -0.6532 RRAS2 related RAS viral (r-ras) oncogene homolog 2 -0.4666 RPS6KB1 ribosomal protein S6 kinase, 70kDa, polypeptide 1 -0.3867 RASA1 RAS p21 protein activator (GTPase activating protein) 1 -0.3776 IGF1R insulin-like growth factor 1 receptor -0.3684 FOXO3 forkhead box O3 -0.35133 PIK3C2A phosphoinositide-3-kinase, class 2, alpha polypeptide -0.24177 PRKACB protein kinase, cAMP-dependent, catalytic, beta -0.15200 PDPK1 3-phosphoinositide dependent protein kinase-1 -0.04Rank Target gene Gene nameTotal context score Ingenuity Pathway name p-value RatioIGF-1 Signaling 7.57E-07 9/107 (0.084)Role of NFAT in Cardiac Hypertrophy 2.51E-06 11/208 (0.053)ERK5 Signaling 3.93E-05 6/64 (0.094)Insulin Receptor Signaling 5.58E-05 8/140 (0.057)PTEN Signaling 1.19E-04 7/124 (0.056) (A) (B) 31 Figure 2.6 Ingenuity pathway analysis predicts miR-223 targets regulate IGF signalling IGF-1 pathway elements predicted to be miR-223 targets according to Ingenuity Pathway Analysis. Predicted targets shown in purple. 32 Table 2.2 Correlation between IGF1R and mature miR-223 Calculated Spearman correlation between expression of mature miR-223 and predicted target IGF-1 signalling pathway components transcripts in 10 T-ALL cell lines.*, p<0.05. Figure 2.7A 0 500 1000 1500 20004681012mature miR-223 expressionIGF1R mRNA expressionr= -0.6606,p< 0.0438(A) Target Gene Spearman r p valueIGF1R -0.661 0.0438* FOXO1 -0.642 0.3679RRAS2 -0.588 0.0806FOXO3 -0.164 0.6567RASA1 -0.078 0.8382PIK3C2A 0.03 0.946PDPK1 0.067 0.8651RPS6KB1 0.176 0.6321PRKACB 0.333 0.348733 Position 224-231 of IGF1R 3' UTR 5' ...CCUGCCCAAACCCUUAACUGACA...||||||||hsa-miR-223 3' ACCCCAUAAACUGUUUGACUGU Hsa (human) CCCU---G--CCCAAACCCUU-AA---C--------UGACAUG-------GG---CCUUU--AAGAACCUPtr (Chimpanzee) CCCU---G--CCCAAACCCUU-AA---C--------UGGCAUG-------GG---CCUUU--AAGAACCUMml (Rhesus) CCCU---A--CCCAAACCCUU-AA---C--------UGACAUG-------GG---CCUUU--AAGAACCUOga (Bushbaby) ----------------------------------------------------------------------Tbe (Treeshrew) UACAUGGA--CCCGAACCCCU-CA---CG-------UGAGAUC-------GGGUCCCUCU--CAGGACCUMmu (Mouse) CUUU---A--CCCGAACCCUU-AA---C--------UGACAUG-------GG---CCUCU--GCAAACCURno (Rat) CCCU---A--CCCAAACCCUU-AA---C--------UGACAUG-------GG---CCUCU--GCAAACCUCpo (Guinea Pig) CCCU---G--CCCAAGCCCUU-AA---C--------UGACACG-------GG---CCACU--GAGAACCUOcu (Rabbit) CCCC---A--CCCAAACCCUU-AA---C--------UGACACG-------GG---CCUUU--AAGAAACUSar (Shrew) ----------------------------------------------------------------------Eeu (Hedgehog) CCUG---A--CCCAAACCCUU-AA---C--------UGACAUG-------GG---CCUUU--AAGAACCUCfa (Dog) CCCU---A--CCCAAACCCUU-AA---C--------UGACAUG-------GG---CCUUC--AAGAACCUFca (Cat) CCCU---A--CCCGAACCCUU-AA---C--------UGACAUG-------GG---CCUUU--AAGAACCUEca (Horse) CCCU---A--CCCAAACCCUU-AA---C--------UGACAUG-------GG---CCUUU--AAGAACCUBta (Cow) CCCU---A--CCCAAACCCUU-AA---C--------UGACACG-------GG---CCUUU--AAGAACCUDno (Armadillo) CCCU---A--ACCAAACCCUUCAA---C--------UGACAUG-------GG---CCUCU--AAGAACCULaf (Elephant) CCCU---A--ACCAAACCCUCCAA---C--------UGACAUG-------GG---CCUUU--AAGAACCUEte (Tenrec) GCCU---A--ACCAAACCCUCCAA---C--------UGACAUG-------GG---CCUUU--AAGCACCUMdo (Opussum) CAGU---A--AAGGAUUCCUC-AA---A--------UGAGGCG-------GG---CCUUU--AAAAACUUOan (Platypus) CAGU---U--AAC-AACUCCA-AA---CGUGGGGCAUGAAGUG-------GG---CCUUUUAAAAAACUUAca (Lizard) CAAG---CAAAAGAACUCUUG-UUUAAG--------UGGGGCGCACAGAGAG---CCUUC--UAAAACUUGga (Chicken) CGGU---A--AAGGACUCCAU-UC---A--------UGGGGCACACAGUUGG---CCUUU--UAAAACUCXtr (Frog) ----------------------------------------------------------------------Placental mammals Figure 2.7 miR-223 can target IGF1R (A) Correlation plot for miR-223 and IGF1R transcript expression levels in 10 human T-ALL cell lines from publically available data. miR-223 seed alignment with IGF1R 3’UTR. (B) 8mer match of human miR-223 to position 224-231 of human IGF1R 3’UTR. (C) Aligned IGF1R 3’UTR regions corresponding to position 224-231 of human IGF1R 3’UTR from various animal species, placental mammals indicated. Underlined, 6mer miR-223 seed region and complementary IGF1R sequence; red, 8mer (the seed, + position 8, followed by an 'A‘) miR-223 region and complementary IGF1R sequence. (B) (C) 34 Figure 2.8 miR-223 can target the 3’UTR of IGF1R to repress expression Luciferase reporter assay. HeLa cells were transiently transfected with luciferase CDS + IGF1R 3’UTR construct along with either miR-223 or empty vector control plasmids. Cell extracts were prepared 42 hours later and assayed for firefly luciferase activity with normalization to Renilla luciferase transfection control. Error bars indicate standard deviation of assays performed in triplicate. **, p<0.01 (Student’s t-test). 2.3.6 miR-223 regulates IGF1R protein levels in T-ALL To address more directly whether miR-223 can regulate IGF1R expression in T-ALL cells, miR-223 was overexpressed by retroviral transduction. Transduced cells were sorted by virtue of the linked retroviral GFP marker and IGF1R protein levels assessed by Western Blot. Overexpression of miR-223 in this manner was associated with a reduction in steady-state protein levels of both the pro-protein and mature forms of IGF1R in two different cell lines, Jurkat and RPMI 8402 (Figure 2.9A). The degree of enforced miR-223 expression achieved was physiological and similar to that achieved by GSI treatment of parental cells (Figure 2.9B). Conversely, lentiviral knockdown of miR-223 using a complementary miR-223 sponge resulted in increased levels of both pro- and mature IGF1R protein (Figure 2.9C). Taken together, these data support that miR-223 regulates steady-state IGF1R protein levels in human T-ALL cells. MigRI Mig miR-2230.00.20.40.60.81.01.2**luciferase activity(normalized to control)35 Figure 2.9 miR-223 regulates IGF1R protein levels, but does not alter cell growth Western Blot analyses of IGF1R protein levels. (A) miR-223 was overexpressed in T-ALL cell lines by retroviral transduction. Transduced (GFP+) cells were FACS sorted and whole cell lysates prepared. Lysates from non-transduced (parental) cells were prepared in parallel. (B) The level of overexpression of mature miR-223 achieved, as Parental MockParental GSImiR-223 Mock01234RPMI 8402miR-223(fold over parental mock) *** ns*(A) (B) (C) 36 assessed by qRT-PCR, was equivalent to that induced by GSI treatment. (C) miR-223 was knocked-down in T-ALL cells by lentiviral transduction with a miR-223 sponge. Transduced cells exhibiting high and low levels of expression of the linked GFP marker were FACS sorted and whole lysates prepared. Lysates from sorted GFP+ scramble control (miR-Scr)-transduced cells were prepared in parallel. ERK2 loading control is shown in lower panels. Numbers below panels indicate densitometric band quantitation with normalization to ERK2 loading control and expressed as fold-change over respective non-transduced/scramble controls. ns, not significant; *, p<0.05, ***; p<0.001 (Student’s t-test). 2.3.7 Manipulation of miR-223 alone does not affect T-ALL cell growth To determine whether NOTCH-regulated miR-223 expression has a functional effect on T-ALL cells, I overexpressed or knocked down miR-223 by viral transduction of T-ALL cell lines and tracked the percentage of transduced (GFP+) cells in culture over time, effectively performing a growth competition assay versus non-transduced (GFP-) cells. My initial hypothesis was that miR-223 overexpression (or knock-down) would lead to decreased (or increased) IGF1R expression and thus decreased (or increased) T-ALL cell growth. However, in both overexpression and knockdown experiments, there was no detectable change in the %GFP+ fraction during extended passage in culture (Figure 2.10A). It was considered that although steady-state total protein levels, as measured by Western Blot, reflected these changes, that it was possible that cells might have the capacity to recycle/redistribute IGF1R receptors between the cell surface and endosomal compartments to restore signalling homeostasis. In fact, flow cytometric analysis of surface IGF1R levels revealed no significant changes with either miR-223 overexpression or knock-down (Figure 2.10B). In addition an analysis of a cohort of 17 T-ALL cell lines showed that there was no significant correlation between relative surface IGF1R and miR-223 expression (Figure 2.10C). I conclude from these data that while miR-223 can modulate total IGF1R protein levels, the magnitude of change alone is not sufficient to overcome compensatory mechanisms which presumably act to restore cellular homeostasis by maintaining cell surface levels of receptor. 37 5 0 0 1 0 0 0 1 5 0 0 2 0 0 0-0 .50 .00 .51 .01 .52 .02 .5m iR -2 2 3sIGF1Rr= 0 .3 9 7 1p < 0 .1 1 5 6 Figure 2.10 miR-223 overexpression does not alter cell growth Growth competition assay. (A) Flow cytometric assessment of GFP+ fraction in T-ALL cell line cultures containing both transduced (GFP+) and non-transduced (GFP-) cells following exposure to miR-223 overexpressing retrovirus. (B) Flow cytometric analysis of surface IGF1R levels. T-ALL cells were transduced with miR-223 overexpressing retrovirus (upper panel) or mir-223 sponge lentivirus (lower panel). Live cells were stained with anti-IGF1Rα antibody and analyzed by flow cytometry. (C) Calculated Spearman’s correlation r values and associated significance p-values between expression of miR-223 and surface IGF1R protein expression of 17 T-ALL cell lines. 0 1 2 3 4 5 6 70102030405060RPMI 8402miR-223JurkatmiR-223days in culture%GFP+(A) (B) (C) 38 2.4 Discussion 2.4.1 NOTCH1 regulated miRNAs The objective of this study was to determine the set of miRNAs directly regulated by NOTCH1 in T-ALL and assess their contribution to oncogenic activity in this context. Despite the prominent role NOTCH1 plays in T-ALL development and maintenance through the direct transcriptional activation of numerous protein coding genes, such as HES1 and c-MYC, relatively few miRNAs were seen to be regulated. Similarly, few reports in the literature describe miRNAs directly regulated by NOTCH, although c-MYC has been proposed to both positively and negatively regulate numerous miRNAs in various contexts [149]. An early report described miR-61 as a direct transcriptional target of LIN-12/NOTCH in C. elegans [150] and the miR-143/145 cluster has also been described as a direct target in vascular smooth muscle [151]. These miRNA species, however, were not found to be differentially regulated in any of my datasets. At the time this study was initiated very little was known about the turnover of miRNAs, although a relatively long half-life was suggested from some early studies [140] and, for this reason NOTCH inhibition was extended for 4 days. Subsequent studies have shown that although some miRNAs can have a very short half-lives (~1 hour) [152] others can persist a long time (half-life >7 days) if cells are not cycling [153, 154]. The human T-ALL cell lines sent for miRNA profiling were GSI resistant and as such cell division rates (~1 division every 24 hours) would not have been affected by GSI treatment. Primary mouse NOTCH driven leukemias, however, are usually GSI sensitive in vitro and this may be a confounding factor. I decided to focus my study on miR-223 because it (along with its star strand) were the only characterized miRNAs observed to be differently regulated in human T-ALL cells and there was no consistent pattern of miRNA regulation observed in the murine T-ALL cells. Because NOTCH1 is canonically described to act as a transcription activator, I believe its repression of miR-223 to be indirect. 2.4.2 miR-223 In this study, I have identified miR-223 as being regulated by NOTCH1 in a T-ALL context. miR-223 is a relatively well characterized miRNA that is mainly expressed by 39 cells of the hematopoietic lineage. It was initially described to be confined to cells of the myeloid lineage, and much work focusing on its role in this context has been published [116] . In humans, miR-223 has been shown to be upregulated during granulocyte differentiation. This regulation is controlled by the myeloid transcription factors PU.1 and CAAT/enhancer-binding protein C/EBPβ [155]. Perturbing the action of miR-223 through its overexpression or inhibition can affect this differentiation, highlighting the role of miR-223 in promoting myeloid differentiation. Unlike in humans, however, a loss-of-function miR-223 allele in mice results in an expanded compartment of abnormal granulocytes and, as such, suggests a negative regulation of granulocyte differentiation [144]. miR-223 has also been shown to affect other hematopoietic differentiation events; it has been shown to negatively regulate erythroid differentiation [156] whilst it is also reported to positively regulate megakaryotype [157] and eosinophil [158] differentiation. With miR-223 having such wide ranging effects during myeloid cell differentiation, it comes as no surprise that it has also been shown to have an effect on immune cell function. miR-223 may steer macrophages to a M2 phenotype [159], alter cytokine production [160] and influence NF-κB signalling [161]. All these functions may, in part, explain the altered miR-223 expression observed in many inflammatory and infectious states. Additionally, changes in miR-223 expression have also been linked to numerous malignancies, both hematopoietic and non-hematopoietic. 2.4.3 miR-223 in cancer miR-223 has been suggested to play a role in numerous cancer types including hepatoma, cervical carcinoma [162], osteosarcoma [163], gastric carcinoma [164] and oesophageal cancer [165]. It can act both as an oncomiR and a tumor suppressor to promote and suppress malignant phenotypes. For example, miR-223 appears to have opposing roles in the regulation of migration and invasion, important processes in tumor metastasis. In gastric cancer, cell migration and invasion is thought to be promoted by miR-223 overexpression, whereas in oesophageal cancer, migration and invasiveness is suppressed by miR-223. As with solid tumors, miR-223 has also been reported to both promote and suppress malignant phenotypes in hematological malignancies. In AML miR-223 is down regulated, which is thought to prevent its suppression of cell cycle progression whereas in T-ALL it has, subsequent to this study, been reported to 40 be leukemogenic. In T-ALL, miR-223 has also been reported to have higher expression in cases that present with myeloid-like gene features [143], where expression levels are more similar to those in AML. Myeloid-like genes upregulated in this context include C/EBPβ, potentially providing a mechanism for the higher observed miR-223 expression. 2.4.4 NOTCH regulation of miR-223 To my knowledge, this is the first account of the regulation of miR-223 downstream of NOTCH1. A previous report had eluded to this possibility by showing, in an in vitro murine embryonic stem cell hematopoietic differentiation system that enforced expression of constitutively active NOTCH1 resulted in lower miR-223 expression 3 days later [166]. However, in this system as the hemangioblasts are differentiating into hematopoietic progenitors, miR-223 expression levels increase, suggesting that the lower miR-223 level induced by NOTCH1 may represent the more immature state of these cells, in line with other phenotypic observations. Although confounded by changes in differentiation state, this observation is consistent with my own that NOTCH1 represses miR-223 expression. More recently, another group has also described the regulation of miR-223 by NOTCH1 (and NOTCH3) in a T-ALL context [167]. Despite also observing GSI mediated NOTCH inhibition resulting in an increased miR-223 expression level (Jurkat and MOLT3), they reported a decreased level of expression in another T-ALL cell line (DND41). Although I observe some T-ALL cell lines to not express miR-223 (ALLSIL and HPBALL) or not regulate miR-223 in response to GSI (P12 Ichikawa), miR-223 upregulation with NOTCH inhibition was never observed, even using the same cell line reported by Kumar et al. (DND41) [167]. They also describe a mechanism for how GSI inhibition of NOTCH signalling might downregulate miR-223 through a conserved upstream promoter region that can bind NOTCH1 and NOTCH3 (as well as NF-κB) to transcriptionally activate miR-223. To explain the differential miR-223 expression between cell lines after GSI treatment they highlight an observed increase in expression of the well characterised miR-223 regulator C/EBPα that is only seen in the lines which upregulate miR-223 following GSI. In T-cell development C/EBPα has been described to be repressed by the canonical NOTCH1 target gene HES1 [168]. This mechanism could possibly account for the increased C/EBPα 41 expression observed after GSI treatment in T-ALL, however, the data shown here where enforced expression of HES1 did not affect GSI induced miR-223 upregulation would suggest that the NOTCH1/HES1/C/EBPα/miR-223 axis is not functioning in the context I tested. 2.4.5 miR-223 regulation of IGF1R in T-ALL Of the numerous predicted miR-223 targets, I further focused on IGF1R given our labs recent finding of its importance in T-ALL cell growth and its leukemia-initiating activity [92]. In that study, I was involved in demonstrating that signalling through NOTCH1 upregulates surface IGF1R protein expression in T-ALL cells by approximately 2-fold on average, and that NOTCH1 can upregulate IGF1R mRNA expression via binding in complex with CSL and MAML1 to an IGF1R intronic enhancer element. I now have supporting evidence that NOTCH1 negatively regulates miR-223 in T-ALL cells, and that miR-223 negatively regulates total IGF1R protein levels. Taken together, our labs data suggest a model in which NOTCH signalling supports IGF1R expression directly by enhancing its transcription, and indirectly by antagonizing miR-223. Presumably, these effects combine to achieve the physiologically relevant net increase in IGF1R expression that we have previously described to support propagation of T-ALL disease. I observed no effect of manipulating miR-223 alone on T-ALL cell growth and find that surface IGF1R protein levels were unchanged, consistent with recycling/redistribution from intracellular pools which I hypothesize may occur as part of a compensatory mechanism to restore cellular homeostasis. In addition, it may not be entirely surprising to find that modulation of a single miRNA species in isolation does not produce major phenotypic consequences since miRNAs are often described as playing a more subtle “fine-tuning” role in the regulation of gene expression and tend to work in networks where individual miRNAs repress multiple targets and individual targets are repressed by multiple miRNAs [169]. 2.4.6 miR-223 regulation of IGF1R in other contexts To date, miR-223 has now been reported to negatively regulate IGF1R expression in numerous malignant cell types including cervix carcinoma (HeLa), acute promyelocytic leukemia (NB4) and hepatoma (SMMC-7721, BEL-7404, Huh-7) [162], as well as 42 normal tissues including vascular smooth muscle [170], eosinophil progenitors [158], embryonic stem cells [171], endothelial cells (HUVEC) [172]. All of these studies reported IGF1R expression changes at the mRNA and/or total protein levels without assessing surface expression levels. In many of these cases, unlike my own, a reduction in proliferation was also observed with miR-223 overexpression. This difference may be explained by higher overexpression of miR-223, different IGF1R trafficking dynamics that allow for regulation of the functional surface IGF1R pool or the use of cell lines more sensitive to marginal changes in IGF1R signalling. After publication of this study I have determined that the cell lines used to competitively track GFP with miR-223 overexpression, Jurkat and RPMI 8402, are resistant and only marginally sensitive to pharmacological inhibition of IGF1R, respectively, (see Chapter 3) and would not be expected to show an IGF1R specific proliferation disadvantage had surface IGF1R been reduced. In addition other cellular contexts would present a different set of miR-223 targets, whose regulation in conjunction with IGF1R may have different phenotypic outcomes. 2.4.7 miR-223 in T-ALL miR-223 is also predicted to target numerous additional transcripts, some of which have been experimentally validated by other groups. Described miR-223 targets include FBXW7 [173], LMO2 [157], MEF2C [144], FOXO1 [174], RPS6KB1 [175], PAX6 [176], RHOB [177], HSP90B1 [163], RASA1 [162], NLRP3 [178], NFIA [179], PARP1 [180], ECT2 [181] and PTBP2 [182]. Some of these miR-223 targets (FOXO1, RPS6KB1) are known to be involved downstream of IGF1R and may suggest that miR-223 regulates the entire pathway at multiple points. In addition, other described miR-223 targets (FBXW7, LMO2, and MEF2C) have previously characterized roles in T-ALL. LMO2 and MEF2C are both transcription factors whose overexpression has been reported in T-ALL to be leukemogenic [183-185], often as a direct or indirect result of chromosomal rearrangements [186, 187]. Both also show higher expression in phenotycially immature “ETP” T-ALL, along with miR-223 and other myeloid lineage markers [188, 189]. It is thus tempting to speculate that NOTCH1-mediated repression of miR-223 may in some way contribute to extinction of myeloid potential during early T-cell development such that further commitment to the lymphoid lineage can occur [190, 191]. LMO2 seems to 43 only be expressed in a fraction of T-ALL cell lines, and within those expressing lines, no downregulation of LMO2 protein levels was detected after 7 days of GSI treatment (data not shown). This observation is suggestive of no significant regulation by miR-223 in this context. FBXW7 negatively regulates NOTCH1 protein levels (as well as levels of c-MYC, c-JUN and Cyclin E) through PEST domain ubiquitination and proteosomal degradation [192]. It has been strongly predicted to be a miR-223 target (highest TargetScan Total Context+ Score (-1.05) of all predicted human miR-223 targets) and being previously validated as a target in colorectal carcinoma and chronic myelogenous leukemia (CML) cell lines at the time of this study [173]. We did not pursue FBXW7 further in the T-ALL context because the negative repression of miR-223 by NOTCH1 and on the subsequent derepression of FBXW7 would be predicted to diminish NOTCH1 signalling strength and represent a negative feedback loop not promoting leukemic activity. In addition, the frequency of FBXW7 loss of function mutations or NOTCH1 PEST domain mutations also suggests that such a negative feedback loop would not be functioning in many T-ALL cases. After the publication of my study, several publications describing high miR-223 expression and its regulation of FBXW7 in T-ALL have been published. In a panel of 50 T-ALL patient samples and 18 human T-ALL cell lines Mavrakis et al. found miR-223 to be the most abundantly expressed miRNA and cooperated with constitutively active NOTCH1 to reduce T-ALL disease latency in a mouse model [142]. This pro-leukemogenic effect was suggested to be effected via FBXW7 repression, and was further supported by Mansour et al. who also described high miR-223 levels to be positively regulated by the T-ALL oncogene TAL1 and its regulatory partners (HEB, E2A, LMO1/2, GATA3, RUNX1) [193]. It has also been postulated that miR-223 is able to confer GSI resistance through inhibition of the oncosuppressor FBXW7 [167]. Mutational loss of FBXW7 has been proposed to contribute to GSI resistance in the past though the stabilization of c-MYC protein levels. Much of the data supporting this hypothesis is associative and many T-ALL cell lines appear to have an extended c-MYC half life regardless of FBXW7 mutational or GSI sensitivity/resistance status [26]. In addition, enforced c-MYC has been reported to confer GSI resistance to only some GSI sensitive cell lines whilst others remain sensitive, highlighting the complexity of the GSI 44 associated phenotype [88]. This is further supported by the observation of widely varying expression of miR-223 in T-ALL cell lines independent of GSI sensitivity/resistance status [142, 193]. With regard to my own finding that NOTCH1 repression increases miR-223 expression, the idea that this would also increase GSI resistance seems paradoxical. 2.5 Conclusion Here, I describe miR-223 to be downregulated by NOTCH signalling in T-ALL. I suggest that, in concert with other reported NOTCH1 functions, this suppression works to upregulate pro-leukemic IGF1R and, as such, miR-223 may have tumor suppressive activity in T-ALL. Subsequently published conflicting reports suggesting miR-223 can be upregulated by NOTCH1 in T-ALL and that it potentially may promote leukemogenesis through the suppression of FBXW7 suggest that further investigation of the role of miR-223 in T-ALL may be warranted. 45 Chapter 3 Molecular determinants of IGF dependence in human T-ALL 3.1 Introduction 3.1.1 IGF1R structure The type 1 insulin-like growth factor receptor (IGF1R) is a transmembrane receptor tyrosine kinase (RTK). The mature IGF1R is a heterotetrameric complex consisting of two α and two β subunits. Each IGF1R precursor polypeptide is cleaved into α and β subunits, which themselves are cysteine crosslinked before subsequent cysteine crosslinking to a partner to form the mature receptor. β subunits contain the intracellular catalytic tyrosine kinase domain as well as a transmembrane spanning region. α subunits are linked to β subunits on the extracellular surface of the cell and bind ligand. Upon binding its ligand, insulin-like growth factor (IGF) 1 or 2, a signalling cascade is initiated through phosphorylation of its associated partners. On the basis of its kinase domain sequence IGF1R falls into RTK class II, along with its closely related family member, and namesake, the insulin receptor (IR) [194, 195]. Both receptors have evolved from an ancestral receptor [196] with a presumed role in regulation of metabolism, organismal size and longevity, with each retaining some of these characteristics. In addition, each has structural similarities that allow for their association with one another. This allows for the formation of hybrid receptors consisting of α and β subunits from both IGF1R and IR [197] that add an extra level of complexity based on intracellular potential changes in binding/phosphorylation partners as well as altered ligand binding affinities [198]. In addition, there are also reports that IR/IGF1R can form hybrid receptor complexes with epidermal growth factor receptor (EGFR) family members, EGFR, EGFR2 (ErbB2/HER2) and ErbB3 [199, 200]. 3.1.2 IGF1R signalling Canonical signalling events downstream of ligand binding to the α-subunit of IGF1R begins with a conformational change which shifts the receptor from an inhibitory conformation to an active state through the autophosphorylation of the β subunit kinase domain [201]. It is thought that this autophosphorylation is an intermolecular event between each of the β subunits of the tetrameric receptor. Secondary autophosphorylation can then occur outside of the kinase domain in the juxtamembrane 46 region, providing phosphotyrosines that can act as docking sites for adaptor molecules. Tyrosine 950 (Y950) in particular plays a key role in mediating downstream signalling by allowing binding by members of the insulin receptor substrate (IRS) family and Src homology and collagen domain protein (Shc) [202]. The recruitment of these proteins leads to the activation of the phosphatidylinositol-3-kinase (PI3K)/AKT and RAS/RAF/mitogen-activated protein kinase (MAPK) pathways, respectively, described by the classical model (Figure 3.1). In addition to Y950, other less well characterized IGF1R phosphosites C terminal to the kinase domain have been reported to bind various signalling molecules allowing for propagation of additional signalling responses. For example, S1248 with RACK1 [203] and β-arrestin [204] and Y1316 with GRB10 [205]. Other kinases that canonically reside in other signalling pathways have also been described as being activated downstream of the IR/IGF1R, including Src [206], JAK/STAT [207], and alternate MAPK family members (e.g. JNK) [208]. The degree to which these numerous signalling events are happening in individual cell types is unknown, and the complex network of crosstalk and feedback between these pathways makes discerning such information particularly difficult. 47 PI3KPTENRasIRSPIP2PIP3AKTPDK1mTORC2IGF1RRAF1/ARAFMEK1/2ERK1/2ShcSOSGRB2MYC AP1 ETS CREBProliferation and survivalp27 p21 BADFOXOmTORC1Protein synthesisCell cycle/proliferationSurvivalmTOR RICTORPI3K/AKT signalling armMAPK signalling arm Figure 3.1 Canonical IGF1R signalling pathway Schematic representation of the canonical IGF1R signalling pathway the PI3K/AKT and MAPK arms indicated. 3.1.3 IRS1/PI3K/AKT signalling IRS family members can be phosphorylated on multiple tyrosine residues upon activation. These sites provide for binding by SH2 domain containing proteins, including p85 (PI3K class I regulatory subunit) [209], growth factor receptor bound protein 2 (GRB2), SHP-2, FYN and NCK [210]. The binding of p85 and subsequent activation of the PI3K/AKT signalling pathway is thought to provide for much of the known effects 48 downstream of IRS. Class 1 PI3Ks are a family of lipid kinases that catalyses the phosphorylation of phosphoinositols at the 3-position [211]. As such, they are able to create phosphatidylinositol-3-monophosphate (PI(3)P), phosphatidylinositol-3,4-bisphosphate (PI(3,4)P2), and phosphatidylinositol-3,4,5-trisphosphate (PI(3,4,5)P3). They are heterodimeric protein complexes that are comprised of a catalytic domain (p110) and a regulatory domain (p85 or p101). Those PI3K complexes that are comprised of a p85 subunit associate with p110α, p110β or p110δ catalytic subunits and are termed class 1A PI3Ks, whereas those that consist of p101 and p110γ are class 1B. This difference in regulatory subunits also manifests itself in the types of upstream signalling inputs that result in their activation. Class 1B PI3K complexes are typically activated downstream of G-protein-coupled receptors (GPCRs), such as chemokine receptors. Class 1A PI3K complexes are typically activated downstream of tyrosine phosphorylation of various proteins. This includes receptor tyrosine kinases (RTKs) and their adaptors, such as IGF1R/IRS [212], as well as many other types of cell surface protein complexes (T cell receptor (TCR) [213], cytokine receptors [214, 215] and, toll-like receptors (TLRs) [216]). The PI(3,4)P2 and/or PI(3,4,5)P3 created by PI3Ks serves to recruit numerous pleckstrin homology (PH) domain containing proteins to a membrane to initiate downstream signalling events. Lipid phosphatases, such as phosphatase and tensin homolog (PTEN), negatively regulate this process through the removal of phosphates from PI(3,4,5)P3 to abrogate PH mediated recruitment [217, 218]. One of the main functions described downstream of active PI3K is the activation of the serine/threonine kinase AKT, also known as protein kinase B (PKB). AKT is recruited by its PH domain to the plasma membrane by PI3K [212]. There, it is phosphorylated by 3-phosphoinositide dependent protein kinase 1 (PDK1) and mTOR complex 2 (mTORC2) at T308 [219] and S473 [220], respectively. Maximal activation of AKT requires phosphorylation at both these residues. AKT is described to play a key role in numerous cellular processes including protein synthesis (via TSC/mTOR) [221-223], survival (via BAD/BCL2 [224], MDM2/p53 [225], IKKα/NF-κB [226], and FOXOs [227]), proliferation (via CDK inhibitors p21CIP [228], p27KIP [229, 230]) and glucose metabolism (via PFKFB2 [231] and AS160/GLUT4 [232]). With deregulation of such a pathway 49 potentially impacting on many of the hallmarks of cancer, it comes as no surprise that aberrant activation of AKT has been reported in many malignancies [233]. In addition to tyrosine phosphorylation, IRS family members can be serine/threonine phosphorylated on multiple residues. Phosphorylation of many of these sites is thought to play a negative feedback role through the disassociation of IRS from the receptor [234] and its degradation [235]. Numerous kinases have been implicated to be involved in this phosphorylation including PKCδ, PKCθ [236], IKKβ, JNK, S6K, mTOR, AKT, ERK1/2, and AMPK [237]. 3.1.4 MAPK signalling In addition to binding IRS, IGF1R family members phosphorylated at Y950 in the juxtamembrane region of IGF1R can also act as a docking site for Shc [238]. In canonical IGF1R signalling Shc phosphorylation provides a docking site for GRB2, which in turn is bound by SOS1 (Son of sevenless homolog 1). SOS1 is a guanine nucleotide exchange factor (RASGEF) that promotes the release of GDP from RAS proteins, allowing for the binding of GTP and their activation. Another class of proteins termed RASGAPs work to inactivate Ras through promoting GTP hydrolysis into GDP. Active Ras can in turn activate a MAPK phosphorylation signalling cascade through RAF (MAP3K), MEK1/2 (MAP2K) and ERK1/2 (MAPK). Active ERK1/2 has been described to directly or indirectly lead to the phosphorylation of multiple transcription factors through the activation of Ribosomal S6 kinase (RSK). These transcription factors include c-MYC, AP-1, ETS and CREB which upon phosphorylation initiate transcription of numerous target genes [239]. Amongst these targets are cell cycle regulators (e.g. Cyclin D1, RB1) that can promote cell proliferation and apoptosis regulators (e.g. BCL2, p53) that can promote survival. Activation of other MAPK signalling pathways, typically thought to be stress activated (JNK signalling), have been shown to be activated downstream of IGF1R. The exact mechanism for how this is done remains unknown, but is thought to be IRS1/2 independent and only partially dependent on Y950. JNK activation has been described to be both Ras dependent (often downstream of growth factor receptors) and Ras independent [240]. 50 3.1.5 IGF1R in T-ALL Early in vitro experiments showed IGF-1 signalling to be important for neoplastic cell proliferation [241] as well as initial transformation [242], a role reinforced by subsequent in vivo experiments [243]. In addition, correlative population based studies have suggested a link between circulating serum IGF-1 levels and risk of cancer development for numerous cancer types [244]. It has been well established that IGF1R, as well as other IGF-1 signalling components, are expressed by human T-ALL cell lines, patient samples [92, 245, 246] and mouse models of T-ALL [92]. In addition, two separate T-ALL cases of translocations involving TCR loci with Xq22, a region containing IRS4 have been reported [247, 248]. One of these reports further described this translocation as resulting in strong overexpression of IRS4 at the protein level [247]. Given the importance of IRS proteins in propagating signals downstream of IGF1R and IR, and the fact that overexpression of IRS4 in hematopoietic cells can increase proliferation [249], it is tempting to speculate that IGF1R/IRS4 signalling may be driving these leukemias. With regard to the role of IGF1R in T-ALL specifically, previous research has provided conflicting data on its effects on growth and survival. An early study by Smith et al. [250] showed that supplemental IGF-1 was important for the growth and subsequent establishment of T-ALL cell lines in vitro under hypoxic conditions, but not for their continued growth once established. Early antibodies targeting IGF1R were reported to suppress proliferation in established cell lines [251]. Similar studies utilizing purported IGF1R inhibitors also showed suppressed proliferation in T-ALL cell lines [252, 253]), albeit in a limited number of cell lines tested. As well, our lab and others have previously reported that IGF1R is unregulated both transcriptionally [92, 254] and post transcriptionally [255] in T-ALL by NOTCH1 [256], and that IGF1R signalling contributes to growth/survival of bulk cells as well as leukemia-initiating activity [92]. These observations suggest that pharmacologic inhibition of IGF1R signalling may have a therapeutic role in T-ALL, both in terms of treating bulk disease as well as in targeting leukemia stem cells to prevent relapse. Mutations in IGF1R are rare, and to date, none have been definitively characterized to activate signalling [257, 258]. On the other hand, mutations activating both canonical downstream signalling pathways, PI3K/AKT and MAPK/ERK, occur frequently in human cancers and have been implicated in the 51 pathogenesis of T-ALL [28, 259, 260]. In addition, array CGH data reported 29% of T-ALL patients to have gains in regions containing PI3K-AKT pathway components, including IR and IGF2 [17]. 3.1.6 PI3K/AKT and MAPK signalling in T-ALL – mouse models and direct mutational activation Many lines of evidence point to the importance of both PI3K/AKT and MAPK signalling in T-ALL including transgenic and retroviral induced mouse models. Expression of constitutively active K-RAS (G12D) using a conditional knock-in mouse model results in the development of aggressive T cell malignancies that phenotypically resemble T-ALL patients (CD4+, CD8+) and often acquire NOTCH1 activating mutations [261]. In addition, the established leukemias showed reduced growth when treated with a MEK1 inhibitor in vitro. Similarly, when mutations activating PI3K/AKT signalling are introduced into hematopoietic cells, T-ALL-like malignancies can arise. This can either be done through the viral introduction of activating mutations into the bone marrow (constitutively active AKT) [262] or through the deletion of pathway repressors (e.g. PTEN) in early hematopoietic progenitors [263]. In both examples referenced, myeloproliferative diseases were also reported to arise, sometimes in conjunction with T-ALL, highlighting the importance of PI3K/AKT signalling in numerous types of hematopoietic malignancies. Another strong line of evidence for the importance of MAPK and PI3K/AKT signalling in T-ALL comes from frequently occurring mutations. Mutations that have been reported to enhance signalling in either of these two pathways have been correlated with a worse survival outcome for patients with T-ALL [264]. Mutated major components of the MAPK signalling pathway in T-ALL include K-RAS/N-RAS (9.1%) and NF1 (neurofibromin 1) (2.1%) [28]. Reported K-RAS/N-RAS mutations in T-ALL are mostly point mutations that alter amino acids G12 or G13 and of a type found commonly amongst other cancers [265]. Such mutations result in a form of RAS that favours GTP binding and therefore produces constitutive activation. NF1 is a RASGAP and as such, works to inactivate RAS. In T-ALL, disruptive frameshift mutations and chromosome deletions of NF1 have been reported to cause biallalic loss, which is thought to promote leukemogenesis 52 through activation of MAPK signalling [30]. Mutated major components of the PI3K/AKT signalling pathway in T-ALL include AKT (~2.3%), PI3KCA (~4.5%), PI3KR1 (~4.5%) and PTEN (13-36%) [28, 29]. Similarly to RAS, the point mutations of AKT (E17K) and PIK3CA (E542K) as well as the insertion/deletion of nucleotides (indels) found in the iSH2 domain of PIK3R1 found in T-ALL are in well described hotspots reported in numerous other cancers [266-269] and work to increase PI3K/AKT pathway activation. By far the most common and widely studied mutations to affect the PI3K/AKT pathway in T-ALL are those found in PTEN. PTEN mutations are frequent in many cancer types and are most commonly indels resulting in a frameshift and no detectable protein expression in T-ALL. Less frequent are reports of loss of function point mutations that can abrogate phosphatase activity, as well as repressive promoter hypermethylation [28, 29]. Regardless of how PTEN is perturbed, the result is thought to be prolonged hyperactivation of PI3K. In addition to its lipid phosphatase activity, PTEN is also reported to possess protein phosphatase activity in T-ALL. Its loss of expression has recently been reported to additionally contribute to T-ALL development through focal-adhesion kinase (FAK) activation [270]. PTEN loss was initially thought to confer GSI resistance based on a strong correlation between these two attributes in T-ALL cell lines. Such an observation is supported by the reported ability of constitutively active AKT to functionally replace NOTCH1 during β-selection [271]. However, recent findings by our own group and others [29, 135] suggest that loss of PTEN cannot fully explain GSI resistance in T-ALL, and the correlation between these phenotypes does not hold up in primary mouse and human cases. Although not appearing to be directly mutationally activated in the majority of cases, increased MAPK and PI3K/AKT signalling have been reported in many T-ALL patients, 38-50% [272, 273] and 70–85% [274], respectively, suggesting additional possible mechanisms of activation. 3.1.7 PI3K/AKT and MAPK signalling in T-ALL – upstream activation PI3K/AKT and/or MAPK pathway activation is reported in the majority of T-ALL cases. The frequency of mutations in direct pathway components, however, suggests other mechanisms of activation may be involved. In T-ALL, the microenvironment is thought to provide signalling inputs vital to maintain growth and survival. A key component of 53 this is thought to be through the action of common gamma chain-signalling cytokines [275, 276], most notably IL-7 [277-279]. IL-7 signalling is undoubtedly important for sustaining T-ALL cell growth/survival as highlighted by its requirement in the ex vivo expansion of primary human T-ALL cells [275, 280, 281]. In addition to STAT5, IL-7 has also been shown to activate both PI3K/AKT and MEK/ERK signalling in T-ALL [280]. T-ALLs have also been described to enhance IL-7 downstream activation through multiple mechanisms. IL-7Rα itself has been described to be a direct transcriptional target of NOTCH1 [91] and additionally, is mutated in about 9% of T-ALL cases [34]. These mutations consist of inframe indels in the extracellular juxtamembrane-transmembrane region, the majority of which introduce a cysteine residue. It is these cysteine residues that are thought to allow for intermolecular disulfide bonds between mutant IL-7Rα proteins and result in constitutive downstream JAK1 signalling. Enhanced IL-7R downstream signalling in T-ALL can also be achieved through activating mutations in JAK1 [33], as well as mutational inactivation and deletion of the phosphatases CD45 [282] and PTPN2 [283], respectively. Although mostly recognized as an important activator of indispensible PI3K/AKT activation, the role IL-7 plays in activating MEK/ERK signalling in T-ALL is also of interest in the literature. This is highlighted by the recent description of increased RASGRP1 abundance in some T-ALL patients and how this can potentially cooperate with IL-7 signalling to activate Ras and stimulate leukemogenic growth [284]. RASGRP1 is a RASGEF and, like SOS1, promotes the release of GDP from RAS proteins, allowing for the binding of GTP and activation. Additionally, integrations near the RASGRP1 locus in murine retroviral models of T-ALL have been described to be relatively frequent [284], and enforced over expression of RASGRP1 has been shown to promote murine T-ALL leukemogenesis [285]. Other microenvironment factors have been described to active PI3K/AKT and MAPK signalling in T-ALL including integrins [286] and IGF1R [92]. How multiple pathway inputs in combination with diverse genetic abnormalities result in net pathway activation is important for understanding the true requirements for T-ALL growth and better therapeutic treatment design. 54 3.1.8 IGF1R inhibition In numerous pre-clinical studies, IGF1R inhibitors have shown efficacy in solid tumors, including non-small cell lung cancer, breast cancer, adrenocortical carcinoma, and Ewing sarcoma [287], as well as in hematologic malignancies such as myeloma, CLL, B-ALL, T-ALL, and AML [92, 288-291]. Several agents have advanced to clinical trials [292, 293]. To date, however, due to limited efficacy and, in some instances, metabolic toxicity, none have been approved for use outside of investigational studies [294]. It has been suggested that efficacy could be improved in selected patient groups with predictive biomarkers and in combination with complementary therapies that target PI3K/AKT and RAS/RAF/MEK/ERK pathways simultaneously [295-299]. In order to investigate the potential efficacy of IGF signalling inhibitors in human T-ALL, I tested two clinically relevant IGF1R inhibitors, a small molecule tyrosine kinase inhibitor, BMS-754807 [300], and a humanized monoclonal blocking antibody, CP-751,871 [301], against a large panel of 27 human T-ALL cell lines. Here, I describe that a subset of cell lines demonstrates sensitivity to these agents and characterize genetic/phenotypic features that define cellular dependence on IGF signalling. I hypothesize that the growth of some T-ALL cell lines will be negatively impacted by the inhibition of IGF1R signalling and its canonical downstream signalling arms, PI3K/AKT or MAPK, may be responsible for its action. Aim 2: Investigate the effects of IGF1R inhibition and characterize important IGF1R downstream signalling required for growth in a T-ALL context. 3.2 Materials and methods 3.3.1 Cell culture All established cell lines used (Table 3.1) have been reported previously [15, 302]. Human T-ALL cell lines were grown in RPMI 1640 medium supplemented with 10% FBS, 1 mM sodium pyruvate (Stem Cell Technologies), 2 mM L-glutamine (Life Technologies), and Penicillin/Streptomycin (Stem Cell Technologies). Recombinant 55 human IGF-1 and human IL-7 (Peprotech) were resuspended in PBS/1% BSA and added to culture media at 5 ng/mL and 100 ng/mL final concentration, respectively. 3.2.2 Drugs IGF1R inhibitors BMS-754807 and CP-751,871 were obtained under Material Transfer Agreement from their respective manufacturers, Bristol-Myers Squibb and Pfizer, respectively. Dosing was chosen based on previous literature, including phase I clinical trials [300, 301, 303], as well as my own dose response analysis (data not shown). The PI3Kγ-selective inhibitor, AS-604850, was obtained from Cayman Chemical (Cat# 10010175). The pan-PI3K inhibitor (Ly294002), dual PI3K/mTOR inhibitor (PI-103) and the mTOR inhibitor (Rapamycin) were all obtained from Calbiochem (Cat# 440202, 528100, 553210). BMS-754807, AS-604850, Ly294002, PI-103 and Rapamycin were resuspended in DMSO and diluted in PBS or culture media prior to addition to cell cultures. CP-751,871 was diluted in PBS/1% BSA or culture media prior to addition to cell cultures. 3.2.3 Viable cell number assay Viable cell numbers were determined by a resazurin reduction assay, either the CellTiter-Blue cell viability assay (Promega) or the equivalent Alamar Blue assay (ThermoFisher). Briefly, 2 x 104 – 4 x 104 cells/100 μL culture media were seeded per well of a 96-well plate, various cytokines/growth factors/inhibitors were added at the doses indicated, and cells cultured for 3 days at 37 °C in 5% CO2. CellTiter/Alamar Blue reagent was added per manufacturer’s protocol and incubated at 37 °C for 20-40 minutes prior to fluorescence measurement at 590 nm using a GENios FL microplate reader (Tecan). All assays were performed in triplicate. 3.2.4 Viral vectors All lentiviral expression constructs were generated using a pRRL lentiviral backbone with MNDU3 promoter and PGK-GFP or PGK-NGFR marker as previously described [304]. The constitutively active K-RAS G12D mutant was cloned from the murine T-ALL cell line 144CLP [305]. A myristoylated, constitutively active form of AKT1 was cloned from pUSEmyrAKT1 (Upstate). A constitutively active form of IGF1R, CD8-IGF1R, was 56 created by fusing the extracellular/transmembrane domain of human CD8a (amino acids 1 to 218) with the intracellular beta chain of human IGF1R (amino acids 964 to 1367), as previously described [306]. The constitutively active IL7Rα (p.L242_L243insLSRC) mutant was cloned from the human T-ALL cell line DND41, as described [307]. PTEN was cloned from a wild-type patient T-ALL sample. Lentiviral shRNA knockdown constructs targeting PTEN were identified from the RNAi Consortium (TRCN0000002746 and TRCN0000002749) and cloned into pLKO.1-GFP [304]. The non-silencing scramble shRNA was a gift from David Sabatini (Addgene plasmid #1864). All constructs were verified by Sanger sequencing. 3.2.5 Viral transduction High titer, replication defective lentivirus was produced utilizing pCMVΔR8.74, pCMV-VSV-G, and pRSV-Rev packaging vectors by transient transfection of 293T producer cells, where necessary virus was concentrated using PEG-8000. Viral transduction was performed by spinoculation as previously described [255]. Briefly, viral supernatant was added to cells along with 4 μg/mL polybrene and spun at 750 x g at 32 °C for 2 hours. 3.2.6 Western Blot Whole cell protein extracts were prepared using RIPA buffer (50 mM Tris-HCl pH 8, 150 mM NaCl, 1% NP-40, 0.25% Na-deoxycholate, 1 mM EDTA) supplemented with 1 mM NaF, 1 mM Na3VO4, 2.5 mM Na-pyrophosphate, 1 mM phenyl-methylsulfonyl-fluoride and 1x Protease Inhibitor Cocktail Set III (Calbiochem). Lysates were cleared by centrifugation at 14,000 × g for 10 minutes at 4 °C, separated by SDS-PAGE, and transferred to PVDF membranes. Blots were probed with antibodies directed against PTEN (Y184, Abcam), ZAP70 (99F2, Cell Signalling Technologies), and β-actin (AC-15, Sigma) followed by HRP-conjugated secondary antibody and chemiluminescence detection (Pierce). Band intensities were quantified using Image Studio Lite (LI-COR) or ImageJ (NIH) software. 3.2.7 Ligand stimulation assay Cells were serum starved by culturing in serum-free media for 24 hours, then stimulated by addition of recombinant ligand or 10% FBS. Ten minutes later, cells were fixed by 57 addition of formaldehyde to 1.5% final concentration, and then permeabilized with ice-cold methanol for at least 30 minutes before analysis by flow cytometry. 3.2.8 Flow cytometry Cell surface expression levels of IL7Rα and IGF1R were determined by staining fresh cells with antibodies against IL7Rα (Biolegend A019D5, APC conjugate,) or IGF1R (EMD Millipore αIR3,), respectively. The latter was detected by staining with an APC-conjugated goat anti-mouse secondary antibody (BioLegend). Intracellular phospho-AKT levels were assessed by staining formaldehyde fixed, methanol permeabilized cells with an AlexaFluor647-conjugated anti-Phospho-AKT (Ser473) antibody (D9E clone, Cell Signalling) or isotype control (DA1E, Cell Signalling). Data was acquired on FACSCalibur or LSRFortessa cytometers (Becton Dickinson) and analyzed using FlowJo software (Tree Star). 3.2.9 Gene expression profile analysis Gene expression for T-ALL cell lines was downloaded from the publically available Cancer Cell Line Encyclopedia (CCLE) project (http://www.broadinstitute.org/ccle/home). CCLE data comes from Affymetrix U133+2 arrays with and raw Affymetrix CEL files converted to a single value for each probe set using Robust Multi-array Average (RMA) and normalized using quantile normalization. Data was analyzed using GENE-E. 3.2.10 Statistics Prism 6 software (GraphPad) was used for statistical analyses. 3.3 Results 3.3.1 IGF1R inhibition in T-ALL cell lines Prior work from our lab, that I was involved in generating, has revealed that IGF1R signalling is important in T-ALL; however, we addressed this issue by relying most heavily upon genetically defined mouse models where the contribution of other genetic variables can be excluded. In fact, the spectrum of human T-ALL presents several common genetic alterations that may reasonably be expected to modulate the IGF 58 signalling pathway, and in this sense, may potentially limit the clinical efficacy of IGF inhibitors. To begin to address this issue, I screened two clinically relevant IGF inhibitors, CP-751,871 (also known as figitumumab), an IGF1R blocking antibody, and BMS-754807, a small molecule dual IGF1R/IR tyrosine kinase inhibitor, against a broad panel of 27 human T-ALL cell lines and scored for effects on overall cell growth/survival. One-third of cell lines showed a statistically significant effect of CP-751,871 (9/27 with p<0.05 difference between mock and treated by t-test; median 19% inhibition, range 5-49%,) including 4/27 exhibiting more than 20% decrease in cell growth (Figure 3.2A), while over half were affected by BMS-754807 (15/27 with p<0.05 difference between mock and treated by t-test; median 18% inhibition, range 6-63%) including 7/27 exhibiting more than 20% decrease in growth (Figure 3.2B). The 20% growth inhibition after IGF1R abrogation cut-off was chosen arbitrarily. A strong correlation in response to CP-751,871 and BMS-754807 was observed among these cell lines (Pearson r = 0.932, p<0.0001); however, there was also a subset of cell lines which were more responsive to BMS-754807 than CP-751,871 (7/27 with statistically significant difference of greater than 10%; (Figure 3.3), potentially reflecting the contribution of IR, which can activate similar downstream signalling to IGF1R, or other off-target related tyrosine kinases. Of note, there was no significant correlation between the degree to which some cell lines were more sensitive to BMS-754807 over CP-751,871 and their IR transcript levels (data not shown). 59 Figure 3.2 Pharmacological inhibition of IGF1R restricts growth of a subset of human T-ALL cell lines (A) (B) human T-ALL cell linesfluorescence(normalized to mock)ALLSILHPBALLTALL1HSBKOPTK1SUPT1MOLT15RPMI 8402MOLT4SUPT13CUTLL1THP6PF382MKB1BE13MOLT13Karpas45JurkatDND41LoucySKW3REXPEERP12 IchikawaCCRF-CEMSUPT11MOLT160.00.20.40.60.81.0mockCP-751,871human T-ALL cell linesfluorescence(normalized to mock)ALLSILHPBALLTALL1HSBKOPTK1SUPT1MOLT15RPMI 8402MOLT4SUPT13CUTLL1THP6PF382MKB1BE13MOLT13Karpas45JurkatDND41LoucySKW3REXPEERP12 IchikawaCCRF-CEMSUPT11MOLT160.00.20.40.60.81.0mockBMS-75480760 Cell growth as measured by resazurin reduction assay. Twenty-seven human T-ALL cell lines were cultured in vitro for 3 days with either (A) IGF1R blocking antibody (CP-751,871, 1 μg/mL) versus PBS vehicle control (mock) or (B) dual IGF1R/IR tyrosine kinase inhibitor (BMS-754807, 0.5 μM) versus DMSO vehicle control (mock). Mean resorufin (reduced resazurin) fluorescence values +/- SD after normalization to mock-treated controls are plotted for assays performed in triplicate. Cell lines are rank ordered left-to-right by decreasing effect of the CP-751,871 blocking antibody. The horizontal dotted line indicates the 20% growth inhibition level. Statistically significant differences indicated in red (p<0.05, t-test). Figure 3.3 BMS-754807 inhibits cell growth to a greater extent than CP-751,871 in a subset of human T-ALL cell lines Cell growth as measured by resazurin reduction assay. (BMS-754807 data points significantly different than their corresponding CP data points are indicated in red (p<0.05, t-test). The horizontal dotted line demarcates cell lines with greater than 10% difference between (BMS-754807 and CP-751,871 values. Plotted data are identical to those presented in Figure 3.1, but normalized to the CP-751,871 fluorescence values. ALLSILCCRF-CEMKOPTK1JurkatHPBALLSUPT11PF382BE13PEERSUPT1RPMI 8402CUTLL1THP6P12 IchikawaMOLT16HSBSKW3MOLT13MOLT4TALL1REXDND41MKB1MOLT15LoucyKarpas45SUPT130 .60 .70 .80 .91 .01 .11 .21 .3fluorescence ratio (normalized to CP)C P -7 5 1 8 7 1B M S -7 5 4 8 0 761 3.3.2 Effect of IGF1R expression One obvious variable that might be expected to affect a cell’s response to IGF1R inhibition would be the level of IGF1R expressed on the cell surface. Indeed, I found the surface IGF1R level to be inversely correlated with cell growth under inhibition with both CP-751,871 (Pearson r = −0.700, p<0.0001) (Figure 3.4A) and BMS-754807 (Pearson r = −0.705, p<0.0001) (Figure 3.4B) such that cells with higher levels of surface IGF1R expression were more sensitive to IGF1R inhibition. To further explore this observation, I analyzed publically available microarray gene expression profiling data on a smaller cohort of T-ALL cell lines and found that transcript levels of IGF1R were also inversely correlated with cell growth under inhibition with both CP-751,871 (Spearman r = -0.6791, p< 0.0094) (Figure 3.5A) and BMS-754807 (Spearman r = -0.7055, p< 0.0063) (Figure 3.5B). Using this same gene expression data and grouping the lines into two subsets based on their response to CP-751,871, either “CP sensitive” (ALL-SIL, HPBALL, TALL1) or “CP resistant” (DND41, JURKAT, LOUCY, MOLT13, MOLT16, MOLT4, P12 Ichikawa, PEER, PF382 RPMI 8402, SUPT1) as defined by at least 20% reduction in viable cell numbers, the expression of which genes correlated most with this phenotype was determined. This analysis showed IGF1R transcript expression as being ranked 24/18988 for genes correlating with the “CP sensitive” grouping. Interestingly, the second most positively correlated gene was IRS2, an adaptor immediately downstream of IGF1R (Figure 3.5C). This suggests that cells which gain growth/survival advantage from IGF signalling have been selected to upregulate expression of IGF1R on the cell surface and other IGF signalling components. 62 Figure 3.4 Sensitivity to IGF1R inhibition correlates with surface IGF1R expression level Plots of cell growth with IGF1R inhibition, (A) CP-751,871 and (B) BMS-754807, (normalized resazurin fluorescence data from Figure 1) against surface IGF1R expression level (mean fluorescence intensity by flow cytometry). Linear regression lines are depicted with the 95% confidence interval indicated by flanking dotted lines. Pearson correlation r values and associated significance p-values are as indicated. CP-751,871surface IGF1R expression(normalized to HPBALL)fluorescence(normalized to mock)0.0 0.5 1.0 1.5 2.00.40.60.81.01.2r= -0.700,p< 0.0001BMS-754807surface IGF1R expression(normalized to HPBALL)fluorescence(normalized to mock)0.0 0.5 1.0 1.5 2.00.20.40.60.81.01.2r= -0.705,p< 0.0001(A) (B) 63 Row min Row maxALLSILHPBALLTALL1DND41JURKATLOUCYMOLT13MOLT16MOLT4P12ICHIKAWAPEERPF382RPMI8402SUPT1CP sensitive CP resistantGENE RANKIGF1R 24SLMO1 1IRS2 2GPR87 3GPR155 4EVL 5MEX3D 6FANCC 7LOC100131655 8UTY 9ATP6V1B1 10SYNPR 18979PSAT1 18980ANXA5 18981SLC4A1AP 18982NRBP2 18983SLC30A10 18984UBE2D4 18985MRPL41 18986SCN3A 18987BATF3 18988 Figure 3.5 Sensitivity to IGF1R inhibition correlates with IGF1R and IRS2 transcript levels Plots of cell growth with IGF1R inhibition, (A) CP-751,871 and (B) BMS-754807, (normalized resazurin fluorescence data from Figure 1) against IGF1R transcript expression level in 14 human T-ALL cell lines. Linear regression lines are depicted with 6.5 7.0 7.5 8.0 8.5 9.0 9.5 10.00.20.40.60.81.01.2BMS-754807IGF1R transcript expressionfluorescence(normalized to mock)r= -0.706,p< 0.00636.5 7.0 7.5 8.0 8.5 9.0 9.5 10.00.40.60.81.01.2CP-751,871IGF1R transcript expressionfluorescence(normalized to mock)r= -0.679,p< 0.0094(A) (B) (C) 64 the 95% confidence interval indicated by flanking dotted lines. Spearman correlation r values and associated significance p-values are as indicated. (C) Top ranked genes correlated with assigned response to IGF1R inhibition with CP-751,871, CP sensitive versus CP resistant. Transcript expression levels taken from publically available microarray data. 3.3.3 Downstream effector pathways Our lab has previously described that a hypomorphic allele of IGF1R (IGF1Rneo) abrogated serial transplantability of mouse T-ALL [92]. In this study, it was also observed a limited number of leukemic clones bypassed the requirement for high-level IGF1R expression and have restored transplantability. This IGF1Rneo mouse model may be predictive of what might be expected to occur in T-ALL patients following IGF1R inhibitor therapy. To explore this issue in the context of human T-ALL, I elected to test prospectively the ability of candidate downstream signalling elements to render previously sensitive cell lines resistant to IGF1R inhibition. Based on the accumulated literature supporting that IGF signals bifurcate into two major arms, PI3K/AKT and RAS/RAF/MEK/ERK, I first tested whether constitutive activation forms of either would restore cell growth following treatment with CP/BMS. Interestingly, a constitutively active myristoylated AKT construct rescued T-ALL cells from CP-induced growth inhibition on par with a constitutively active CD8-IGF1R fusion protein [306], which was employed as a positive control in this assay (Figure 3.6A). The CD8-IGF1R fusion combines the extracellular/transmembrane domain of human CD8 with the intracellular IGF1Rβ chain [306] and thus, confers constitutive activation of signalling through IGF1R, but yet is immune to the CP-751,871 blocking antibody. In contrast, the constitutively active RAS (G12D) mutant showed no evidence of rescue with CP-751,871 induced inhibition and performed similarly to empty virus, which was employed as a negative control in this assay. To confirm the specificity of the constitutively active CD8-IGF1R positive control, it was demonstrated in the same assay that Y950F or K1003A point mutants of the CD8-IGF1R construct which are unable to interact with downstream effectors IRS1/2 and Shc, or lack kinase activity, respectively, were unable to rescue CP-751,871 induced growth inhibition (Figure 3.6B). The cell growth result was also corroborated with intracellular phospho-AKT levels in a highly quantitative flow cytometry assay which 65 demonstrated that, while RAS (G12D) failed to rescue CP-751,871 induced diminution of pAKT levels, both myr-AKT and CD8-IGF1R did (Figure 3.6C). These results support the notion that signalling downstream of IGF1R via AKT, but not RAS, is responsible for the growth promoting effects of IGF in human T-ALL cells. Extensive research has shown PI3K/AKT/mTOR signalling to be important in T-ALL and many studies have shown inhibition in many T-ALL cell lines to result in reduced cell numbers [308-312]. My own data, using common PI3K/AKT pathway inhibitors, where reduced cell numbers, and, in most cases, no major cell death is observed, supports this (Figure 3.7 A/B). To this end, I believe that IGF1R inhibitor resistant cells can maintain growth through IGF1R independent PI3K/AKT activation. Figure 3.6 pAKT level(normalized to mock empty)e mp ty v irusmy r-AK TR AS (G1 2D )C D8 -IGF 1R0 .00 .51 .01 .52 .0m o c k C P -7 5 1 8 7 1(C) emp ty v irusmy r-AK TR AS (G1 2D )C D8 -IGF 1R0 .00 .20 .40 .60 .81 .01 .2m o c k C P -7 5 1 8 7 1fluorescence(normalized to mock)* * * * * *n s n semp ty v iruswild-typ eY 95 0FK 10 03A0 .00 .20 .40 .60 .81 .01 .2m o c k C P -7 5 1 8 7 1fluorescence(normalized to mock)n s * * * * * * * ** * * *C D 8 -IG F 1 R(A) (B) 66 Figure 3.6 Constitutive activation of AKT, but not RAS, rescues T-ALL cells from IGF1R inhibition (A-B) Cell growth as measured by resazurin reduction assay. HPBALL cells were transduced with lentiviruses carrying (A) constitutively activated AKT, RAS, or IGF1R signalling constructs or (B) signalling-inactive variants (Y950F, K1003A) of the constitutively activated CD8-IGF1R fusion construct. Transduced cells were FACS sorted and then cultured in vitro with IGF1R blocking antibody (CP-751,871, 1 μg/mL) for 3 days. Mean resorufin fluorescence values +/- SD after normalization to respective mock-treated controls are plotted for assays performed in triplicate. ***, p<0.001; ****, p<0.0001; ns, not significant (2-way ANOVA with Sidak’s multiple comparisons test). (C) Flow cytometric analysis for intracellular pAKT level. HPBALL cells were transduced, FACS sorted, and cultured as in (A) before staining for pAKT (Ser473) and assay by flow cytometry. Mean fluorescence intensity after normalization to mock-treated, empty virus control is plotted for a representative example of assays performed in duplicate. Figure 3.7A (A) A LL SILH PB AL LK OP TK 1J urka tP 12 Ich ik awaR PMI 84 020 .00 .51 .0M o c kL y 2 9 4 0 0 2P I-1 0 3R a p a m y c inH u m a n T -A L L c e ll lin e sNormalized relative fluorescence**** **** **** **** **** ****67 Figure 3.7 Pharmalogical inhibition of the PI3K/AKT pathway blocks growth of human T-ALL cell lines (A) Relative normalized cell numbers of a selection of T-ALL cell lines, both IGF1R inhibitor sensitive (ALL-SIL and HPBALL) and strongly resistant (Jurkat and P12 Ichikawa), as measured by resazurin reduction assay after 3 days of in vitro culture with PI3K/AKT/mTOR pathway inhibitors. (B) Cell viability of the same cultures was assessed by flow cytometry and measured as % of cells excluding propidium iodide (PI) dye. Concentrations of inhibitors used were 10 μm Ly294002, 1 μm PI-103 and 0.1 μm Rapamycin. **** = p < 0.0001. (Multiple comparison 2-way ANOVA). 3.3.4 Effect of PTEN Canonical activation of AKT downstream of receptor tyrosine kinases such as IGF1R occurs via PI3K-dependent conversion of PI(4,5)P2 to PI(3,4,5)P3 at the plasma membrane. The lipid phosphatase PTEN dephosphorylates PIP3, converting it back to PI(4,5)P2, and thereby down modulating signalling through AKT. Thus, another major variable that might be expected to modulate a cell’s sensitivity to IGF inhibition would be its PTEN status. Of note, PTEN is deleted or mutated in approximately 13-36% of patient T-ALLs with greater incidence at relapse [28, 29]. Among the 27 cell lines tested, 11 were positive for PTEN, including the 5 most sensitive cell lines (Table 3.1, Figure ALLSILHPBALLKOPTK1JurkatP12 IchikawaRPMI 8402020406080100MockLy294002PI-103RapamycinHuman T-ALL cell linesCell Viability (% PI-)(B) 68 3.8). In fact, there was a significant correlation between PTEN positivity and sensitivity to both CP-751,871 and BMS-754807 (p = 0.007 and 0.004, respectively; one-tailed point-biserial correlation), but this correlation was not absolute as there were indeed examples of PTEN-positive IGF1R inhibitor resistant lines (e.g. DND41) and PTEN-negative IGF1R inhibitor sensitive lines (e.g. SUPT1) (Figure 3.2A/B). Of note, the correlation between IGF1R inhibitor efficacy and surface IGF1R expression noted above continues to hold true in the PTEN-positive subset of cell lines (n = 11; Pearson r for CP-751,871 = −0.872, p = 0.0005; Pearson r for BMS-754807 = −0.837, p = 0.0013); however, within the PTEN-negative subset (n = 15), the correlation holds for BMS-754807 (Pearson r = −0.572, p = 0.026), but loses significance for CP-751,871 (Pearson r = −0.393, p = 0.148) (Figure 3.9 ). To test the association between PTEN and IGF dependence in a more direct manner, isogenic cell lines with and without PTEN were generated, by 1) re-expressing PTEN in PTEN-negative cell lines and 2) knocking down PTEN in PTEN-positive cell lines, and then reassessing their response to IGF1R inhibition. Restoration of PTEN expression in otherwise PTEN-negative P12 Ichikawa and PF382 cells (Figure 3.10 A) had surprisingly limited effects on response to IGF1R inhibition with either CP-751,871 or BMS-754807 (Figure 3.11 A). In one instance (P12 Ichikawa cells treated with BMS), PTEN re-expression had a small (9% increase in sensitivity) but statistically significant effect compared to empty virus control; however, in all other instances cells with re-expressed PTEN trended toward reduced growth compared to empty virus controls, but the effects fell short of statistical significance. Conversely, knock-down of PTEN by 70-90% using lentiviral shRNAs (Figure 3.10 B) had greater magnitude and more consistent effects in increasing resistance to IGF1R inhibition (~20% increase in HPBALL/CP-751,871, ALLSIL/CP-751,871, and ALLSIL/BMS-754807; whereas HPBALL/BMS-754807 was not significant) (Figure 3.11 B): however still fell short of restoring growth completely. From these results, I conclude that while loss of PTEN indeed reduces cellular dependence on IGF, the minimal effects of re-expressing PTEN suggests that other co-occurring genetic variables, perhaps mutations in other PI3K/AKT pathway elements as previously reported in patient T-ALL samples [28], likely function in combination with PTEN loss to confer IGF independence. 69 Table 3.1 PTEN protein expression and IL-7Rα mutational status in human T-ALL cell lines Cell Line PTEN expression PTEN reference IL7Rα mutation status [313] ALLSIL + (WT) Zuurbier [29], Palomero [302] WT HPBALL + (WT) Zuurbier [29], Palomero [302] WT TALL1 + (WT) Zuurbier [29], Palomero [302] WT HSB + (WT) Zuurbier [29] WT KOPTK1 + (WT) You [270], Palomero [302] WT SUPT1 - Zuurbier [29] WT MOLT15 + (ND) This report ND RPMI 8402 - Zuurbier [29] WT MOLT4 - Zuurbier [29] WT SUPT13 - This report WT CUTLL1 + (WT) You [270], Palomero [302] ND THP6 - This report ND PF382 - Zuurbier [29] WT MKB1 - This report ND BE13 + (WT) Zuurbier [29] WT MOLT13 - This report ND Karpas45 - Zuurbier [29] ND Jurkat - You [270] WT DND41 + (WT) Zuurbier [29] mutated Loucy - Zuurbier [29] WT SKW3 - Zuurbier [29] WT PEER + (WT) Zuurbier [29] WT P12 Ichikawa - Zuurbier [29] WT CCRF-CEM - Zuurbier [29] WT SUPT11 + (ND) This report WT MOLT16 - Zuurbier [29] WT 70 Figure 3.8 PTEN protein status in human T-ALL cell lines Western Blot analysis for PTEN in cell lines whose PTEN status was not previously reported. HPBALL is included as a positive staining control. β-actin is shown as a loading control. 71 Figure 3.9 Correlation between IGF1R inhibitor efficacy and surface IGF1R expression level holds up in PTEN-positive cell lines, but less so in PTEN-negative cell lines Linear regression lines are shown separately for PTEN-positive and -negative subsets, with 95% confidence intervals indicated by flanking dotted lines. Pearson correlation r values and associated significance p-values are indicated for each subset. Plotted data are identical to those presented in Figure 3.4, but now segregated into PTEN-positive and -negative subsets. surface IGF1R expression(normalized to HPBALL)fluorescence(no rmalize d t o mock)0.0 0.5 1.0 1.5 2.00.40.60.81.01.2PTEN-negativePTEN-positiveCP-751871r= -0.393, p= 0.148r= -0.872, p= 0.0005s u r f a c e IG F 1 R e x p r e s s io n(n o r m a liz e d to H P B A L L )fluorescence(normalized to mock)0 .0 0 .5 1 .0 1 .5 2 .00 .20 .40 .60 .81 .0P T E N -n e g a tiveP T E N -p o s itiveB M S -7 5 4 8 0 7r = -0 .5 7 2 , p = 0 .0 2 6r = -0 .8 3 7 , p = 0 .0 0 1 3(A) (B) 72 Figure 3.10 Enforced PTEN expression and knockdown in T-ALL cell lines Western Blot analysis for PTEN protein expression. (A) PTEN-negative cell lines P12 Ichikawa and PF-382 were transduced with PTEN-expressing lentivirus. (B) PTEN-positive cell lines ALLSIL and HPBALL were transduced with lentiviral shRNAs against PTEN. Numbers below shRNA panels indicate relative PTEN protein amounts after normalization to ZAP70 loading control. (A) (B) 73 Figure 3.11 PTEN contributes to, but does not define IGF dependence Cell growth as measured by resazurin reduction assay. (A) PTEN-negative cell lines P12 Ichikawa and PF-382 were transduced with lentiviral PTEN expression constructs, FACS sorted, and then cultured in vitro with IGF1R blocking antibody (CP-751,871, 1 μg/mL) or dual IGF1R/IR tyrosine kinase inhibitor (BMS-754807, 0.2 μM) for 3 days. (B) PTEN-positive cell lines HPBALL and ALLSIL were transduced with lentiviral shRNAs against PTEN, FACS sorted, and then cultured as in (A). Transduced HPBALL cells were grown in vitro for 3 days with IGF1R blocking IGF1R kinase inhibitor (BMS). Mean resorufin fluorescence values +/- SD after normalization to respective mock-treated controls are plotted for assays performed in triplicate. *, p<0.05; **, p<0.01; ns, not significant (2-way ANOVA with Sidak’s multiple comparisons test). P 1 2 Ic h ik a w amo ckC P-75 18 71B MS -7 54 80 70 .00 .20 .40 .60 .81 .01 .2e m p ty v iru s P T E Nfluorescence(normalized to mock) n s *P F 3 8 2mo ckC P-75 18 71B MS -7 54 80 70 .00 .20 .40 .60 .81 .01 .2e m p ty v iru s P T E Nfluorescence(normalized to mock) n s n sH P B A L Lmo ckC P-75 18 71B MS -7 54 80 70 .00 .20 .40 .60 .81 .01 .2s h S c ra m b le s h P T E N 46 s h P T E N 49fluorescence(normalized to mock)* *n sn sn sA L L S ILmo ckC P-75 18 71B MS -7 54 80 70 .00 .20 .40 .60 .81 .01 .2s h S c ra m b le s h P T E N 4 6fluorescence(normalized to mock)* **(A) (B) 74 3.3.5 Role of PI3Kγ/δ PI3Ks are composed of a catalytic and a regulatory subunit, the former existing in one of four distinct isoforms, p110α-δ [314]. Receptor tyrosine kinases (RTKs), such as IGF1R, act upon class IA PI3Ks p110α/β/δ, whereas G-protein coupled receptors (GPCRs) act via the sole class IB PI3K, p110γ [315]. Interestingly, it has been shown that in the absence of PTEN, T-ALL cells are reliant upon either p110γ or p110δ such that treatment with a dual specificity inhibitor induces growth arrest and apoptosis [316], implying that p110α and p110β are unable to compensate in this context. Given that the screen identified a correlation between loss of PTEN and resistance to inhibition of IGF1R, and since IGF1R may act via p110δ, I wondered whether resistance might be mediated by activation of p110γ, possibly via an as-yet-uncharacterized GPCR. To test this hypothesis, the PTEN-negative cell line CCRF-CEM, which was resistant to IGF1R inhibition with CP-751,871 in my screen, (Figure 3.2A), but previously reported to be highly sensitive to a dual p110γ/δ inhibitor [316] was selected. These cells were treated with a p110γ-specific inhibitor, AS-604850, at dosing appropriate for in vitro cell-based assays [317], both alone and in combination with CP-751,871 antibody, anticipating that AS-604850 would block p110γ and CP-751,871 would block signalling upstream of p110δ, and thus phenocopy the effect dual p110γ/δ inhibition. In fact, AS-604850 was not observed to render CCRF-CEM cells sensitive to CP-751,871 (Figure 3.12 ); suggesting perhaps that an input other than IGF1R may be responsible for activation of PI3Kδ in these cells. 75 Figure 3.12 Combined inhibition of IGF1R and PI3Kγ does not block growth of PTEN negative CCRF-CEM cells Cell growth as measured by resazurin reduction assay. CCRF-CEM cells were cultured in vitro with IGF1R blocking antibody (CP-751,871, 1 μg/mL) with or without a selective PI3Kγ inhibitor (AS-604850, 50 μM) for 3 days. Mean resorufin fluorescence values +/- SD after normalization to untreated control are plotted for assays performed in triplicate. ns, not significant (2-way ANOVA with Sidak’s multiple comparisons test). 3.3.6 Non-overlapping roles of IGF-1 and IL-7 Prior studies on cytokine/growth factor-dependence in human T-ALL have suggested that IL-7 is a prominent contributor supporting T-ALL cell growth both in vitro [275] and in vivo [34], and that this effect is mediated in part through PI3K/AKT [280]. As well, activating mutations in IL7R occur in 10% of pediatric T-ALL cases [34, 318], and transduction of mouse thymocyte progenitors with patient-derived IL7R mutants develop immature T-cell leukemias resembling human ETP T-ALL [319]. Of note, the only cell line in the panel known to harbor an activating IL7Rα mutation, DND41, is resistant to both CP-751,871 and BMS-754807 inhibitors. In addition DND41 also expresses moderate levels of IGF1R on the surface and is PTEN wild-type; raising the intriguing D M S O AS -6 0 4 8 5 00 .0 00 .2 50 .5 00 .7 51 .0 01 .2 5fluorescence(normalized to untreated)m o c k C P -7 5 1 8 7 1n sn s76 possibility that constitutive activation of IL7R may stimulate PI3K/AKT sufficiently such that input from IGF1R is not required to maintain cell growth/survival. To determine whether there is indeed overlap between IL-7 and IGF signalling with respect to PI3K/AKT activation in human T-ALL cells, I attempted to rescue CP/BMS-induced growth suppression by supplementing culture media with recombinant IL-7. HPBALL was selected as a model for this purpose because it is one of the few T-ALL cell lines that stably express IL7R on the cell surface (Figure 3.13A, data not shown). While supplemental IL-7 indeed had a positive effect on cell growth overall, it did not completely restore growth after inhibition by either CP-751,871 or BMS-754807 (Figure 3.13 B). Interestingly, I found that AKT nonetheless underwent phosphorylation following acute stimulation with recombinant IL-7 to an extent comparable to recombinant IGF-1 (Figure 3.13 C). To address the apparent discrepancy between robust activation of AKT by IL-7 in CP-treated cells and the lack of restoration of cell growth, phospho-AKT levels after 3 days of culture in supplemental IL-7 were examined and, surprisingly, no enhancement compared to non-supplemented control was found (Figure 3.13 D). In contrast, culture for the same period in CP antibody did reveal a reduced level of phospho-AKT compared to mock treated control, correlating with its growth inhibitory effect. Of note, the specificity of the CP-751,871 antibody in this context was confirmed by its inhibition of AKT phosphorylation by IGF-1, but not by IL-7, whereas the BMS-754807 TKI also dampened the effect of IL-7, indicating a degree of cross-reactivity on IL7R/JAK (Figure 3.13 C). It was also notable that CP-751,871 completely inhibited AKT phosphorylation by FBS, suggesting that the majority of AKT stimulating activity in standard FBS-supplemented culture media is signals through IGF1R. I also attempted to rescue IGF1R-induced growth inhibition by enforced expression of the constitutively active IL7Rα mutant, insLSRC, derived from the DND41 T-ALL cell line. The insLSRC mutant was unable to rescue growth or to activate AKT in the presence of IGF1R inhibition (Figure 3.14 A/B). Of note, expression of the mutant construct was not verified in the transduced cells due to the lack of a commercially available antibody reagent capable of distinguishing between the exogenous mutated allele and the endogenous wild-type allele. Taken together, these findings support the 77 conclusion that IGF signalling fulfills an important role in T-ALL cell growth that is distinct from that provided by IL-7, possibly related to a greater perdurance of AKT activation that even a constitutively active IL-7Rα mutant cannot afford. Figure 3.13 Signalling through IL7R does not rescue T-ALL cells from IGF1R inhibition nor maintain prolonged activation of AKT (A) Flow cytometric analysis for surface IL7R expression level in HPBALL cells cultured under routine conditions. (B) Cell growth as measured by resazurin reduction assay. HPBALL cells were cultured in vitro with IGF1R blocking antibody (CP-751,871, 1 μg/mL) or IGF1R/IR tyrosine kinase inhibitor (BMS-754807, 0.2 μM) with or without supplemental recombinant IL-7 (100 ng/mL) for 3 days. Mean resorufin fluorescence values +/- SD after normalization to untreated control are plotted for assays performed in triplicate. *, p<0.05; **, p<0.01; ***, p<0.001 (2-way ANOVA with Sidak’s multiple comparisons test). (C-D) Flow cytometric analysis for intracellular pAKT. (C) HPBALL cells were serum starved with or without IGF1R inhibition for 24 hours, and then pulsed m o c k C P -7 5 1 8 7 1 B M S -7 5 4 8 0 70 .00 .20 .40 .60 .81 .01 .2- IL 7 + IL 7fluorescence(normalized to untreated)* ** * **S S S S + C P S S + B M S0 .00 .51 .01 .52 .02 .53 .0m o c kIG F 1 (5 n g /m l)IL -7 (1 0 0 n g /m l)F B S (1 0 % )pAKT level(normalized to unstarved)pAKT level(normalized to untreated)- IL 7 + IL 70 .00 .51 .01 .5m o c k C P -7 5 1 8 7 1(A) (C) (B) (D) 78 with IL-7, IGF-1 or FBS for 10 minutes. Mean fluorescence intensity after normalization to unstarved, untreated control is plotted for a representative example of assays performed in duplicate. SS, serum starved; CP, CP-751,871 (1 μg/mL); BMS, BMS-754807 (0.2 μM). (D) HPBALL cells were cultured in vitro with IGF1R blocking antibody (CP-751,871, 1 μg/mL) with or without supplemental recombinant IL-7 (100 ng/mL) for 3 days. Mean fluorescence intensity after normalization to untreated control is plotted for a representative example of assays performed in duplicate. Figure 3.14 Constitutive activation of IL7R does not confer resistance to IGF1R inhibition (A) Cell growth as measured by resazurin reduction assay. HPBALL cells were transduced with constitutively active IL-7R (insLSRC) lentivirus, FACS sorted, and cultured in vitro with IGF1R blocking antibody (CP-751,871, 1 μg/mL) for 3 days. Mean resorufin fluorescence values +/- SD after normalization to respective mock-treated emp ty v irusIL7R ( in sL SR C)0 .00 .20 .40 .60 .81 .01 .2m o c kC P -7 5 1 ,8 7 1fluorescence(normalized to mock)n s* * ** *pAKT level(normalized to mock empty)e mp ty v irusIL7R ( in sL SR C)0 .00 .51 .01 .5m o c kC P -7 5 1 ,8 7 1(A) (B) 79 controls are plotted for assays performed in triplicate. **, p<0.01; ***, p<0.001; ns, not significant (2-way ANOVA with Sidak’s multiple comparisons test). (B) Flow cytometric analysis for intracellular pAKT level. HPBALL cells were transduced with constitutively active IL-7R (insLSRC), FACS sorted, and cultured as in (A). Mean fluorescence intensity after normalization to mock-treated, empty virus control is plotted for a representative example of assays performed in duplicate. 3.4 Discussion 3.4.1 IGF1R in T-ALL context In our prior study examining the contribution of IGF1R to leukemia propagation in vivo, it was found that a hypomorphic allele of IGF1R (IGF1Rneo) reproducibly abrogated serial transplantability in a genetically defined mouse model of T-ALL [92]. As might be expected in any biological system, limited number of clones that bypassed the requirement for high-level IGF1R expression and, in doing so, restored transplantability was observed. In a sense, the IGF1Rneo mouse model may be predictive of what might occur in T-ALL patients following IGF1R inhibitor therapy. Accordingly, it is useful to consider prospectively what mechanisms might arise by which cells to circumvent or compensate for reduced IGF signalling. Prior studies have examined this issue in the context of solid tumors and interestingly different pathways were identified in different tumor types [296, 320], highlighting the importance of cellular context. Herein I have taken up this issue in the context of T-ALL and employed a broad screen of 27 established human T-ALL cell lines to capture a spectrum of genetic backgrounds in the disease. 3.4.2 Correlations with IGF1R inhibitor resistance/sensitivity Varied growth responses to IGF1R inhibition across 27 T-ALL cell lines were observed, with some lines displaying clear growth impairment whilst others appear refractory. Expression of MYB has previously been associated with sensitivity to IGF1R inhibition by CP-751,871 in lung, breast, and colorectal cell lines [321]. In T-ALL, MYB can be overexpressed by duplication [26, 322, 323]. Indeed, five cell lines in the panel are known to carry extra chromosomal copies of MYB (ALLSIL, RPMI 8402, MOLT4, P12 Ichikawa, CCRF-CEM) [322]; however, only one of these (ALLSIL) exhibited sensitivity to IGF1R inhibition, suggesting that multiple variables might contribute to the net 80 pharmacological effect [322]. In support of this, a significant correlation between Myb transcript levels and response to IGF1R inhibition was not observed (data not shown). In a further attempt to decipher why some lines were sensitive to IGF1R inhibition whilst others remained refractory, a significant correlation between low levels of IGF1R at both the transcript and surface protein levels and resistance to IGF1R inhibition was found. In addition, expression of another member of the IGF1R signalling pathway, IRS2, was observed to be strongly inversely correlated with resistance to IGF1R inhibition. These same correlations with IGF1R and IRS2 have also been reported recently in colorectal cancer cell lines, suggesting that our independent observations may be applicable across various cancer types [324]. In addition, IRS2 has been previously implicated in T-ALL, whereby its locus, along with that of LMO2, have been reported as being a recurrent site for retroviral insertions in mice that go on to develop T-ALL [325]. This observation, along with the reports of rare translocations between TCR loci and Xq22, a region containing IRS4 [247, 248], suggests that upregulation of IRS family members might promote advantageous IGF signalling in T-ALL. From this, I conclude that cells which gain growth/survival advantage from IGF signalling have been selected to upregulate expression of IGF1R signalling components and thus maximize their ability to respond to ambient levels of IGF factors in the surrounding environment. Accordingly, high levels of IGF1R expression or IRS family members may be taken as a feature which would suggest a given tumor is likely to respond to inhibition of IGF signalling. Further examination of the inverse correlation between surface IGF1R and resistance to IGF1R inhibition found that it was only significant in PTEN expressing cells, suggesting that PTEN expression may aid in conferring IGF1R inhibitor resistance. Despite this, knock-down or enforced re-expression of PTEN failed to confer or reverse resistance, respectively, revealing that other genetic variables must be contributing in combination with PTEN to modulate IGF dependence. 3.4.3 Mechanisms of IGF1R inhibitor resistance One mechanism by which other groups have reported resistance to IGF1R inhibition is through signalling from other cell surface receptors [326, 327]. Possible IGF1R inhibitor resistance due to utilization of IR has been reported by Garofalo et al. [326]. To address 81 such a problem upfront, the dual IGF1R/IR tyrosine kinase inhibitor BMS-754807 in addition to the IGF1R blocking antibody (CP-751,871) was used. Although some cell lines showed statistically greater response to BMS treatment at the dose that was used relative to CP treatment, I did not observe a resistant cell line switching dramatically to become as sensitive as the BMS sensitive cell lines. To this end, I believe that IR is not being used by T-ALL cells to compensate for IGF1R inhibition and that any additional efficacy may be coming from off target effects, most likely inhibition of other kinases with similarly low IC50 values for the compound (e.g. c-MET, NTRK1, NTRK2, AurA) [300]. This same report also suggested that it was MAPK, specifically ERK1/2, activation downstream of IR that may be mediating resistance. Further support for this idea comes from the observation that activating K-RAS and B-RAF mutations have been correlated with IGF1R inhibitor resistance in other cancer contexts [324]. The characterization of the dominance of PI3K/AKT over RAS/RAF/MEK/ERK in communicating IGF signals to downstream growth/survival effectors argues against a major growth function of RAS/RAF/MEK/ERK signalling downstream of IGF1R in T-ALL. This is further enforced by known Ras activating mutations being reported in cell lines found to be both IGF1R inhibitor sensitive and resistant [259]. In addition to not supporting a growth promoting role for RAS/RAF/MEK/ERK signalling downstream of IGF1R, the observation that activation of PI3K/AKT alone can restore growth also argues against a growth promoting function from other AKT independent, non-canonical IGF1R signalling such as Src, JAK, stress-activated MAPK (JNK and p38) pathways. Recently, downstream of IGF-1 ligand binding, IRS1 has been described to translocate to the nucleus [328] and act as a transcription factor for CyclinD1 and c-MYC [329], cell cycle progression genes both described to be regulated by NOTCH1 in T-ALL [88, 330]. This activity, however, is also described to be independent of the action of PI3K and as such, would also not be providing the major growth function downstream of IGF1R [329]. Although not contributing to the major growth of bulk leukemia cells, it would be interesting to investigate whether any of these pathways play more subtle roles in leukemia initiating cell activity. The observation that AKT activation alone can restore growth after IGF1R inhibition along with the fact PTEN knockdown alone was not able to impart IGF1R inhibitor 82 resistance suggest that an activatory input to the PI3K/AKT pathway may be required above and beyond just removal of its repression. PTEN null T-ALL cells still require PI3K/AKT signalling to maintain growth, and it has been suggested that both class IA PI3Ks, namely p110δ, and class IB PI3K, p110γ, can lead to non-redundant AKT activation and cell growth maintenance [316]. The inability of p110γ inhibition to confer resistance to IGF1R inhibition suggests that input from kinases other than IGF1R likely support activation of the class 1A PI3K isoforms. The data would also suggest that reported activating AKT and PIK3CA mutations found in T-ALL [28] would in theory render cells resistant to IGF1R inhibition. These mutations, however, are relatively rare in patients and, to my knowledge, not reported in any of the T-ALL cell lines tested. I therefore hypothesize that IGF1R inhibitor resistant T-ALL cell lines maintain growth through the activation of class 1A PI3Ks by alternate upstream inputs. 3.4.4 PI3K/AKT signalling in T-ALL - IGF1R versus IL7R In an attempt to address if PI3K/AKT signalling downstream from other cell surface receptors can compensate I decided to focus on IL7R signalling. Trophic signals from IGF and IL-7 were not equivalent in their ability to support T-ALL growth, despite both having the ability to activate PI3K/AKT. Although most IGF1R inhibitor resistant T-ALL cell lines express no (or little) IL7R on their surface (data not shown), and therefore would not be expected to be using this as a compensatory mechanism, it is expressed by many primary patient T-ALLs [331] and is described as a key activator of PI3K/AKT [280]. IL-7 signalling is undoubtedly important for sustaining T-ALL cell growth/survival, as highlighted by its requirement for ex vivo expansion of primary human T-ALL cells [275, 280, 281]; however, IL-7 was found to be unable to maintain pAKT at a sufficient level and duration needed to sustain growth under conditions where IGF1R was inhibited. This may suggest that negative feedback mechanisms occur downstream of IL7R, possibly mediated by suppressor of cytokine signalling 1 (SOCS1) [332], but which do not exist for IGF signalling in this context. As well, the occurrence of activating mutations in STAT5B in T-ALL [333] suggests that activation of the STAT5 arm downstream of IL7R also supports leukemogenic activity. I have previously attempted to expand the investigation of the effect of IGF1R inhibition into xenograft expanded primary T-ALL patient samples and have observed IGF1R inhibition with CP-751,871 to 83 significantly reduce growth in ex-vivo culture [281]. Interestingly, the culture conditions used for ex-vivo expansion include IL-7, lending further support to the idea that AKT activation from IGF and IL-7 are not equivalent in their ability to support T-ALL growth; incidentally this case also carries biallelic frameshift mutations in PTEN that would be predicted to not produce a functional protein product [281] and adding support for the idea that PTEN loss alone cannot account for IGF1R resistance. 3.5 Conclusion Here, I suggest that the use of IGF1R inhibitors may be of use in a subset of T-ALL cases that derive their oncogenic AKT activity from IGF1R signalling. The precise mechanisms by which IGF1R inhibitor resistant cells maintain high AKT signalling to support cellular growth remains to be revealed, but will offer insight into combinatorial therapies to improve IGF1R inhibitor efficacy. 84 Chapter 4 Identification of direct PKCθ phosphotargets in T-ALL 4.1 Introduction 4.1.1 PKCs Protein Kinase C θ (PKCθ) is a member of the Protein Kinase C (PKC) family, itself part of the larger AGC family of protein serine/threonine kinases. PKC isoforms can be split into one of three subfamilies based on their activation requirements. Conventional PKCs (PKCα, PKCβ1, PKCβ2, and PKCγ) require both diacylglycerol (DAG) and calcium ion (Ca2+) secondary messenger molecules to achieve full activation. On the other hand, novel PKC family members (PKCδ, PKCδ1, PKCδ2, PKCδ3, PKCε, PKCη and PKCθ) are defined by not requiring DAG, whereas atypical isoforms (PKCι, PKCζ, PKN1, PKN2, PKN3) do not require either DAG or Ca2+. All PKC isoforms share a common structure consisting of an N-terminal regulatory region joined via a flexible hinge to a C-terminal catalytic region [334]. The catalytic regions between PKC isoforms remain relatively highly conserved, whereas the presence or absence of domains that bind DAG and Ca2+ in the regulatory region, C1a/C1b and C2 domain, respectively, determine second messenger binding and as such, subfamily classification [334]. DAG and Ca2+ production/release are common events downstream of phospholipase C (PLC) activation, itself commonly turned on by G-protein coupled receptors (GPCR). PLC class enzymes cleave the phospholipid phosphatidylinositol 4,5-bisphosphate (PIP2) in the inner surface of the cell membrane into DAG and inositol 1,4,5-trisphosphate (IP3). In the process, IP3 is released into the cytosol, where it can bind IP3 receptors on the endoplasmic reticulum, opening calcium channels and increasing cytosolic Ca2+ concentrations. 4.1.2 PKCθ PKCθ, like other PKC family members, can exist in two different conformational states, “open/active” and “closed/inactive”. In order to transition from the closed state to the open state, an allosteric change that includes displacement of the pseudosubstrate domain from the substrate binding site occurs; dependent on both DAG binding and phosphorylation of Thr-538 (T538) in the activation loop [335]. It has been suggested that T538 is constitutively phosphorylated [336], and as such, PKCθ does not reside in a 85 completely inactive state but rather a primed state waiting for full activation. However, this fact remains controversial, and others groups argue for induced T538 phosphorylation being reliant upon an upstream kinase, possibly GLK during TCR signalling [337]. Upon DAG binding, PKCθ is further phosphorylated, allowing for its full activation and plasma membrane localization. Phosphorylation at S676 and S695, possibly via auto-phosphorylation, is thought to enhance catalytic activation whereas phosphorylation of Y90, via LCK in TCR signalling [338], and T219, has been shown to regulate membrane translocation. Much of the function for PKCθ has been studied in a T cell context, namely downstream of the TCR, however, PKCθ expression and function has also been described in other cell contexts [339-343]. 4.1.3 PKCθ in T-cells In T cells, PKCθ translocates to the immunological synapse upon T cell antigen stimulation and mediates T-cell activation downstream of the T-cell receptor (TCR) complex. Engagement of the TCR/CD28 complex by a recognized peptide/MHC complex leads to initial phosphorylation by the Src family kinases LCK and FYN, as well as the SYK family kinase ZAP70 [344]. Multiple downstream pathways are initiated by these events, leading to cytoskeletal polarization and activation/nuclear translocation of transcription factors such as NF-κB [345], NFAT [346], AP1 (c-FOS/c-JUN). The ensuing gene expression changes lead to T cell activation, proliferation, differentiation, and the subsequent generation of effector or memory immune responses. Despite T cells expressing up to eight different PKC isoforms (PKCα, β, δ, ε, η, θ, ζ and ι) and different PKCs have been reported to show some functional redundancy in certain systems, a vital non-redundant role for PKCθ in TCR signalling has been established [347, 348]. Studies using Prkcq knockout mice have shown that, in the absence of PKCθ protein, activation of mature T cells is impaired [349]. PKCθ has been shown to activate NF-κB, NFAT, and AP1, its role in the activation of the NF-κB being the best characterized. PKCθ has been reported to directly phosphorylate CARMA1 on several serine residues [350], leading to the recruitment of the BCL10–MALT1 complex and to the activation of the IKK complex. IKK then phosphorylates IκBα, leading to its ubiquitination and subsequent proteosomal degradation, allowing for the release and 86 nuclear translocation of NF-κB (NF-κB1/2 and Rel family heterodimer). The activation of AP1 and NFAT transcription factors by PKCθ is less understood, but involves activation of MAPK signalling, possibly via SPAK (Ste20-related MAP kinase) [351] phosphorylation and in conjunction with calcineurin [346], respectively. A role for PKCθ in various other molecular signalling pathways in T-cells has been described. Although these may not be entirely independent of its association with TCR components they include, but are not limited to, roles in IL-7 signalling [352], CXCR4 signalling [353], interferon signalling [354], integrin adhesion [355], and direct transcriptional activation of immune genes through its action in a chromatin anchored complex [356]. 4.1.4 PKCθ targets in other cells Although primarily described in a T cell context, PKCθ is also found to be expressed in other cell types. These include both hematopoietic cells such as mast cells [339], natural killer (NK) cells [340], and platelets [341, 342], as well as non-hematopoietic cells, including skeletal muscle myocytes [342, 343] and possibly the embryonic nervous system [347]. In mast cells, PKCθ has been observed to localize to the plasma membrane and interact with Src kinases upon Fcε receptor I engagement [339]. In this setting, active PKCθ was also shown to enhance MAPK signalling, IL-3 transcription and degranulation [339]. In NK cells, PKCθ is recruited to the immune synapse during activation by target cell recognition where it is described to be involved in actin and myosin IIA recruitment through the phosphorylation of WASP-interacting protein (WIPF1) [357], as well as the induction of downstream IFN-γ and TNF-α secretion [358]. In platelets, PKCθ is activated downstream of glycoprotein VI (GPVI) and protease-activated receptor (PAR) engagement [359]. Without PKCθ activation, these receptors cannot cause platelet aggregation and granule secretion, possibly via impaired MAPK activation and reduced syntaxin-4 phosphorylation [359]. Interestingly, PKCθ transcription in this context has been described to be regulated by RUNX1 [360]. In skeletal muscle, PKCθ is found to be highly expressed [342, 343], and may play a role in insulin signalling. PKCθ is enriched in the membrane fraction after insulin exposure [361] and is thought to play a negative feedback role through the phosphorylation of IRS1 [236, 362, 363]. This function is highlighted by the observation that free fatty acid 87 induced insulin resistance in diabetic patients is associated with PKCθ activation [364, 365] and that PKCθ knockout mice are protected from this phenotype [366]. 4.1.5 PKCθ in cancer Although not as heavily implicated in malignant transformation as other kinases and having no described oncogenic mutations to date, roles for PKCθ in various cancer types have been proposed. These include breast and renal carcinoma, gastrointestinal stromal tumors (GIST) and T-ALL. In breast carcinoma expression of PKCθ has been associated with multidrug resistance [367, 368], and more recently described to promote epithelial to mesenchymal (EMT) transition through its part in a transcriptional complex that upregulates EMT genes [369]. In glucose dependent renal cell carcinoma cell lines, a purported selective PKCθ activator has been shown to be cytotoxic and is thought to act through the induction of a metabolic catastrophe brought on by the activity of PKCθ. Its described mechanism of action involves a simultaneous reduction in glucose uptake as a result of the insulin resistance imparted by PKCθ phosphorylation and repression of IRS1, as well as enforced glucose dependence by the PKCθ phosphorylation and enhanced transcription activity of heat shock factor 1 (HSF1) [370]. In GIST, PKCθ has been proposed as a diagnostic marker, owing to its strong expression in most (~85%) cases, and may be of particularly use in those cases negative for other diagnostic markers such as c-KIT [371, 372]. In addition, a functional role for PKCθ has also been suggested in GIST, whereby knockdown of PKCθ in patient derived cell lines has been shown to reduce proliferation possibly through the down regulation of c-KIT [373]. Unsurprisingly, given its well described function in T cell signalling, PKCθ has also been investigated with regard to its role in T cell leukemias. 4.1.6 PKCθ in T-ALL Shortly after the first report to identify and characterize PKCθ in the literature [343] its expression in leukemic cell lines relative to those from other malignancies was described. Specifically, T-cell leukemia lines (Jurkat, CEM, HPBMLT, HUT-78, L.12.37.8.), an erythroleukemia (K562) and histiocytic lymphoma ceIIs (U-937) showed expression, while promyelocytic leukemia (HL-60), Burkitt’s lymphoma (Raji), non-differentiated colon carcinoma (SW620), goblet cell like colon carcinoma (HT-29-N2), 88 astrocytoma (U251) and schwanoma (TC620) lines did not [374]. Although much work focusing on the mechanistic role PKCθ plays in TCR signalling has been performed in T-ALL cell lines, there are relatively few reports describing its role in leukemic cell physiology. The first report suggesting such a role described a correlation between PKCθ, along with many other PKC family members, and acquired multi-drug resistance [375]. However no other reports have described similar findings, and expression of PKCθ was not reported to be a predictor of treatment failure in a large patient study [376]. PKCθ null mice have been shown to have a longer latency and reduced penetrance in a NOTCH3 induced transgenic mouse model of T-ALL [377]. Results from our own lab in an arguably more relevant NOTCH1 mouse model of T-ALL did not find such a trend, with leukemias derived from PKCθ null bone marrow developing similarly to those from wildtype bone marrow [304]. In competitive transplantation assays, however, Dr Giambra and I have found that restoration of PKCθ expression resulted in compromised leukemia initiating cell activity. We also found that PKCθ expression is lower in the CD44+/reactive oxygen species (ROS) low fraction, a fraction we showed to contain leukemia initiating cell activity in our model. Functionally, PKCθ can increase ROS levels and this suggests that in the LIC fraction ROS accumulation is in part restrained by downregulation of PKCθ. The ability of PKCθ to regulate ROS was also linked to the observation that enforced PKCθ expression actually increased radio/chemo-sensitivity in T-ALL cells [304]. One mechanism T-ALL cells use to reduce the PKCθ expression level is through indirect repression by NOTCH1, via RUNX1 and RUNX3. In order to investigate how PKCθ contributes to the phenotypes above, namely decreased LIC activity, increased ROS and increased radio/chemo-sensitivity, I focused on identifying direct PKCθ phosphorylation targets in a T-ALL context. In doing so I also hoped to be able to identify novel PKCθ regulated phenotypes in T-ALL. 89 4.1.7 Identification of phosphotargets Identification of the direct phosphorylation targets of a specific kinase is important for understanding its role in cellular signalling pathways. In the past, numerous techniques have been employed to elucidate such partnerships; however, it still remains a difficult task. Phosphosite-antibody detection is the mostly commonly used method to determine protein phosphorylation. Methods using this technology, such as intracellular flow cytometry and antibody microarrays, however, are limited in their ability to only detect phosphosites for which antibodies are available and being unable to determine the kinase responsible for the direct phosphorylation. The gold standard assay to determine if a kinase directly phosphorylates a target is an in vitro kinase assay. Purified kinase and target substrate are combined with radio-labelled ATP, and phosphorylation is allowed to proceed before assessment for a transferred radio-labelled phosphate group. Although efficient at determining if a substrate can be directly phosphorylated, in its simplest form only one potential substrate can be screened at a time. Strategies have been developed to scale up these reactions to proteome chips although they are still limited by the number of proteins that can be assessed at one time, as well as availability for different organisms and cell types [378]. Detection of direct phosphorylation of targets in a whole cell lysate is complicated by the presence of other endogenously expressed kinases. Procedures to reduce this problem have been developed, but come with their own disadvantages. For example, instead of using proteins as the substrate, peptide mixtures, where endogenous kinase activity is lost, can be used [379]. However, in such an assay, the loss of substrate structure can also decrease kinase specificity [379]. Recently a method using analog sensitive (AS) kinases has been developed to attempt to identify direct phosphorylation targets of particular kinases in whole cell lysates whilst minimizing endogenous kinase interference [380, 381]. 4.1.8 AS kinase method The identification of direct phosphotargets using the AS kinase method relies on the specific modification of a kinase of interest. This modification allows for selective utilization of bulky ATP analogs that are not used by endogenous wildtype kinases [382, 383]. Bulky ATP analogs also have a modified γ-phosphate group such that transfer of a 90 phosphate onto substrates by a kinase utilizing these analogs will specifically label them. This label can subsequently be used for substrate identification [380]. AS kinases are generated through the modification of an amino acid residue termed the “gatekeeper”. This gatekeeper lies within a hydrophobic pocket that interacts with the adenosine portion of ATP. Because of the highly conserved primary sequences and three-dimensional structures of ATP binding sites [384], the position of this gatekeeper residue is well conserved between all kinases and among biological taxa. For this reason, gatekeeper identification based on catalytic domain alignment and positioning compared to previously described gatekeeper residues (e.g. I338 of v-Src [385]) is relatively straight forward. By modifying the “large” hydrophobic amino acid at the gatekeeper position to a “small” residue, the ATP binding pocket is extended into the position once occupied by the hydrophobic side chain (“small” being glycine or alanine and “large” being any other amino acid). This extra space allows for the acceptance of ATP analogs that have a bulky substitution at the N6 position of the adenine ring that do not usually fit into the vast majority of unmodified kinases. The gatekeeper residue can be modified to either a glycine or an alanine and termed either analog-sensitive-1 (AS1) or analog-sensitive-2 (AS2), respectively. Despite being in principle applicable to all kinases, some limitations of the AS methodology have been observed. Firstly some kinases have been reported to not be efficient at utilizing ATPγS for phosphorylation, even when the gatekeeper is unmodified [386]. Secondly, and more commonly, the modification of the gatekeeper residue can result in the loss of kinase catalytic activity for normal ATP and other ATP variants [387]. In the case of AS1 modifications, the loss of the catalytic activity with the change of the gatekeeper to a glycine could possibly be explained by main-chain conformational changes [388]. Such conformational changes should be reduced by the alanine introduced with AS2 modifications but, despite this, ~30% of kinases remain intolerant to both AS1 and AS2 gatekeeper modifications [387]. A strategy to overcome such intolerance and broaden the scope of kinases that are amendable to AS kinase conversion has been developed. The second-site suppressor mutation strategy, as the name suggests, aims to restore catalytic activity after gatekeeper mutation through the introduction of secondary amino acid changes. A screen performed by Zhang et al. 91 [387] identified “hotspots” for second-site suppressor mutations able to restore catalytic activity in the antiparallel β-sheets of the N-terminal subdomain. It is thought that, because these mutations are not directly adjacent to the gatekeeper, they do not directly affect the gatekeeper pocket space but rather restore catalytic activity through the stabilization of the β-sheets. There is no universal site that can be mutated to restore activity but rather, second site suppressor mutations sites must be determined empirically from second-site suppressor mutation “hotspots” and conservation with gatekeeper mutation tolerant kinases [389]. Initial experiments using AS kinases used radio labelled γ-phosphate groups on the ATP analogs to visualize substrates [390] and aid in their identification [391]. A significant adaption of this method occured with the development of an antibody directed against thiophosphate esters [380]This allows for the direct enrichment of phosphotargets for a kinase of interest by immunoprecipitation when γ-thiophosphate ATP analogs are used. This method has been employed for many different kinases, including PKC family members [392-395]. An inherent caveat in all in vitro kinase assays, compared to in vivo assays, is the loss of spatial localization when lysates are generated. This can potentially result in the identification of substrates that may be phosphorylated in vitro but not necessarily in a cellular context where the kinase and substrate do not have the chance to come into contact (e.g. kinase and substrate localized to different subcellular organelles). To gain higher confidence that the phosphorylation of substrates indentified in vitro may actually be occurring in a T-ALL cell context their proximity to PKCθ in live cells can also be assessed. 4.1.9 Identification of proximal partners In vivo, individual proteins act in complex interconnected networks. To understand how a protein might function within a cell, it is important to gain an insight into the complex network which it inhabits. One approach used to gain such insight is to identify a protein of interest’s direct physical neighbors. An understanding of the proximal partner proteins can expand depth of knowledge about a particular molecular network, as well as allude to previously undescribed functions. Many methods have been developed to probe protein-protein interaction (PPI) networks, each with their own advantages and 92 disadvantages. Methods to determine protein interactions can broadly be based into two categories, theoretical and experimental. Theoretical determination of PPIs can be performed many ways including computational analysis of protein structural features and inference from co-evolution in order to predict candidate protein partners. Although fast and cheap to perform, theoretical methods to predict protein interactions often produce many false positives. Experimental PPI methods can be further divided into those that are hypothesis driven or those that are non-hypothesis-based screens. Hypothesis driven experiments rely on a priori information to suggest an interaction between two species. Examples of such methods include Forster (or fluorescence) resonance energy transfer (FRET), which works at a 10-100 Å (1-10nm) range [396], and protein-fragment complementation (e.g., split-GFP or split-luciferase). Hypothesis driven PPI methods are a good way to confirm a reaction and some, such as the methods mentioned above, can have added power by being performed in live cells. In order to discover novel protein interactors and understand large complexes/networks unbiased PPI screens are required. Some screens rely on strong/stable interactions such as co-immunoprecipitation (co-IP) and Yeast- 2-hybrid (Y2H) [397] assays. In order to detect more weak interactors, such as kinase-phosphosubstrates pairs, a proximity-dependent protein labelling can be employed [398]. Proximity dependent labelling investigates the spatially neighboring protein species of a protein of interest through their covalent modification and, unlike some of the other methodologies described above, does not rely on direct physical interaction between two protein species. The most common proximity PPI method is chemical crosslinking that uses reagents, such as formaldehyde, to create physical covalent linkages between neighbouring proteins within <2–20 Å (0.2-2 nm) [399, 400]. A disadvantage of chemical crosslinking, however, is it can be difficult to identify transient low abundance interactions, a characteristic of kinase/phosphosubstrates pair interactions. Additionally chemical crosslinking has the disadvantage of high background labelling resulting in the identification of many false positive protein species. 93 4.1.10 Proximal labelling methods Recently, methods that have the potential to capture weak and transient interactors in non-hypothesis-based screens have been developed. These methods utilize enzymatic proximity dependent labelling and can be performed in live cells. One such method, BioID, is derived from the hypothesis driven BirA/BioTag system [398]. BirA is a biotin ligase from E. coli that is involved in regulating metabolism through the biotinylation of specific lysine rich protein sequences in acetyl-CoA carboxylase [401, 402]. Through the fusion of a protein of interest to a mutated BirA (BirA*), capable of promiscuous biotinylation, proximal proteins can be labelled with biotin upon its supplementation to the growth media [398] (Figure 4.1A). The extended labelling time offered with this method allows for the accumulation of biotin labelled transient interactions. Biotin adducts can subsequently be used as a molecular handle to pull down labelled proteins for identification. The precise labelling radius of BirA* remains unknown, however, the biotinoyl-AMP intermediate is thought to be rapidly hydrolyzed owing to the numerous reactive primary amines of proteins. Initial reports have proposed a labelling radius <20nM (<200 Å) [403]. In addition to being used for the identification of proximal partners of structural proteins, BioID has also recently been employed to probe signalling pathways [404]. A similar method to BioID involves the use of a modified soybean ascorbate peroxidase (APEX) in place of BirA*. Mutations engineered into this ascorbate peroxidase ensure it monomerizes (K14D/E112K) and has improved biotinylation activity (W41F/A134P) (= APEX2 [405]). Ascorbate peroxidases are normally involved in the detoxification of hydrogen peroxide through the oxidation of ascorbate. In the presence of H2O2, they can oxidize phenol derivatives to phenoxyl radicals. These phenoxyl radicals can covalently attach themselves to electron-rich amino acids such as tyrosine, tryptophan, histidine and cysteine [406]. By using a molecular group such as biotin attached to a phenolic group, APEX2 can covalently attach biotin to proximal proteins (Figure 4.1B). Such radicals are very short lived (~1 millisecond) [407] and are thought to result in a labelling radius of <20 nM (<200 Å) [408]. Originally developed as a reporter for electron microscopy [409], APEX has been developed into a tool for probing proximal proteins [406, 410]. A reported benefit of APEX2 proximal labelling has over BirA* is a greatly 94 reduced labelling time, 1 minute compared to 6-24 hours, respectively. With respect to proteins that can be induced to change their functional states, this can potentially be very advantageous as the proximal partners for each state can be assessed in a reasonable timeframe. LysLysPPiBirA*+ ATPAMP+BiotinBiotinoyl-AMPBiotinylated Protein TyrTyrH2O22H2OAPEX+Biotin PhenolReactive Biotin PhenolBiotinylated Protein Figure 4.1 (A) (B) 95 Figure 4.1 Proximity labelling reactions. Schematic representation of reactions for the biotinylation of proximal proteins (orange) with (A) BirA* and (B) APEX. By using proximity labelling (PKCθ/BirA* or PKCθ/APEX2 fusions) to identify those protein species that reside in the vicinity of PKCθ and comparing these to those identified in vitro with an analog specific PKCθ, I attempted to characterize those protein species with the spatial potential to be direct phosphorylation targets in a T-ALL context. I hypothesize that, in addition to identifying previously described direct PKCθ targets, novel targets will be found. The list of direct targets identified will hopefully allow for greater insight into the PKCθ phenotypes previously characterized in a T-ALL context [304], as well as suggesting additional phenotypes which could be investigated. Aim 3: Find direct PKCθ phosphorylation targets in a T-ALL context. 4.2 Materials and methods 4.2.1 Cloning PKCθ vectors The cDNA sequence encoding human PKCθ was originally sourced from OpenBiosystems in a pCR4-TOPO vector. This sequence corresponds to GenBank: BC113359.1 and encodes the canonical 706 amino acid PKCθ isoform 1 (GenBank: AAI13360.1). For the cloning of PKCθ plasmids for AS method, I PCR amplified using primers that replaced the PKCθ start codon with a canonical Kozak sequence (CCACC) followed by a new start codon in front of bases coding for an in-frame FLAG sequence (DYKDDDDK) and a GSG linker, nFLAG-hPKCθ. Restriction sites also embedded within these primers allowed for digestion and ligation into complimentary sites within multiple cloning sites (MCSs) of pSP72 (cloning vector) and pcDNA3.1 (+) (mammalian expression vector) (Figure 4.2). pSP72-nFLAG- PKCθ variants are relatively small vectors (4615bp), and thus improved the efficiency of site-directed mutagenesis of PKCθ. pcDNA3-nFLAG-hPKCθ variants were used to express recombinant FLAG tagged PKCθ proteins in 293T cells using a cytomegalovirus (CMV) promoter. 96 pcDNA3/nFLAG-PKC theta7529 bpPKC theta (human) coding sequenceAmp(R)Neo(R)FLAG tagGatekeeper: M458SSM site: M444SSM site: F445SSM site: A410SSM site: M385SSM site: H383SV40 polyABGH polyACMV promoterbla promoterSV40 early promoterpUC origin Figure 4.2 pcDNA3-nFLAG-hPKCθ plasmid map. Functional elements highlighted include: coding sequences (orange), promoter for eukaryotic expression (green), promoters for prokaryotic expression (maroon) expression, sequence variations (gatekeeper and second site suppressor mutations (SSM)), and FLAG tag. For the cloning of PKCθ constructs for both the BirA* and APEX methodologies human PKCθ (GenBank: BC113359.1) was also used. Briefly, PKCθ was cloned with the purpose of creating in-frame fusion products with BirA* either at the N or C terminus. Primers employing the following setup were used to PCR amplify PKCθ (Table 4.1A/B). 97 Table 4.1 Primers used for PKCθ:BirA* fusion cloning (A) N-terminal BirA* and (B) C-terminal BirA*. Restriction sites shown in bold and linkers underlined. (A) BirA*/ PKCθ (N term) Primer (5'>3') "Long linker" Fw (XhoI) CTACTCGAGTCGGGTGGTGGAGGATCAGGAGGCGGAGGTTCGCCATTTCTTCGGATTG "Short linker" Fw (XhoI) CTACTCGAGGGAAGCGGATCGCCATTTCTTCGGATTG "No linker" Fw (XhoI) CTACTCGAGTCGCCATTTCTTCGGATTG Rv (PmeI) GCCGTTTAAACTCAGGATATCAGCCGCTC (B) PKCθ /BirA* (C term) Primer (5'>3') Fw (NheI) CGGGCTAGCCAACCATGTCGCCATTTCT "Long linker" Rv (BamHI) CTAGGATCCTCCGCCACCAGAACCTCCGCCACCGGATATCAGCCGCTCCAT "Short linker" Rv (BamHI) CTAGGATCCTCCGCTTCCGGATATCAGCCGCTCCAT "No linker" Rv (BamHI) CTAGGATCCGGATATCAGCCGCTCCAT Three versions for both N and C terminal BirA* fusions were created with variable linker length resulting in either “no linker”, “short linker” (GSG), or “long linker” (SGGGGSGGGG) variants for each. PKCθ amplicons were digested with either XhoI/PmeI or NheI/BamHI ligated into pcDNA3.1 mycBioID (Addgene ID: 35700) and pcDNA3.1 MCS-BirA(R118G)-HA (Addgene ID: 36047) to produce N terminal BirA* and C terminal BirA* fusion constructs, respectively (Figure 4.3A). The resulting fusion constructs were used for transient mammalian expression in 293T and as a source for subsequent subcloning into a lentiviral vector backbone 98 (pRRLsin.cPPTCTS.MNDU3.BNhXhPXbE.PGK.GFP.WPRE) for expression in T-ALL cell lines. In order to create PKCθ:APEX fusion constructs, APEX2 cDNA was PCR amplified with the following primers (Table 4.2A/B) from pcDNA3 APEX2-NES (Addgene ID: 49386) and used to replace the BirA* in the pcDNA3.1 expression vectors before they too were subcloned into the same lentiviral backbone (Figure 4.3B). Table 4.2 Primers used for PKCθ:APEX fusion cloning (A) N-terminal APEX and (B) C-terminal APEX. Restriction sites showed in bold. (A) APEX/ PKCθ (N term) Primer (5'>3') Fw (NheI) GGCGCTAGCCACCATGGACTACAAGGATG Rv (XhoI) AATCTCGAGGTCCAGGGTCAGGCGCTC (B) PKCθ /APEX (C term) Primer (5'>3') Fw (BamHI) CTTGGATCCGGAAAGTCTTACCCAACTGTGAG Rv (PmeI) TCCGTTTAAACTTAGTCCAGGGTCAGGCG 99 (A) Myc --BirA*-- ----------------- hPKCθ ----------------------------------hPKCθ ----------------- --BirA*-- HA NNCCLINKERLINKER (B) --APEX2-- ----------------- hPKCθ ----------------------------------hPKCθ ----------------- --APEX2--NNCCSHORTLINKERSHORTLINKER Figure 4.3 PKCθ:BirA* and PKCθ:APEX fusion constructs (A) Graphical depiction of N-terminal myc tagged BirA* and C terminal HA tagged BirA* PKCθ fusion constructs, myc-BirA*/PKCθ and PKCθ/BirA*-HA, respectively. (B) Graphical depiction of N-terminal APEX2 and C-terminal APEX2 PKCθ fusion constructs APEX/PKCθ and PKCθ/APEX respectively. 4.2.2 Site directed mutagenesis Site-directed mutagenesis for both gatekeeper mutations as well as second-site mutations was performed as described in QuikChange Site Directed Mutagenesis Kit (Agilent Technologies)). Briefly, oligos coding for the amino acid substitution required were designed using the Agilent Primer Design Tool 100 (https://www.genomics.agilent.com/primerDesignProgram.jsp), see Table 4.3. 50 μL reactions containing 50 ng of DNA vector template (psp72-nFLAG- PKCθ /AS2 or pcDNA3-nFLAG- PKCθ /AS2), 125 ng of each appropriate oligo, reaction buffer, 200 μm ATP, 2.5U PfuUltra HF DNA polymerase (Agilent Technologies cat#600380) were run using suggested cycling conditions. 10U of DpnI was subsequently added to digest unmutated parental strand before transformation. Correctly mutated clones were confirmed through Sanger sequencing of the relevant bases. Table 4.3 Oligos used for PKCθ site directed mutagenesis Oligos used for site directed mutagenesis of gatekeeper residues (M458) and SSM sites (H383, A410, M444, M385, F445) as designed by Agilent Primer Design Tool. Base pairs altered from wildtype sequence indicated, bold/underlined. Name (amino acid change)Base pair changeSequence (5'>3')hPKCq M458G sense a1372g_t1373g GGA AAA CCT CTT TTT TGT GGG GGA GTA CCT CAA CGG AGG GhPKCq M458G antisense CCC TCC GTT GAG GTA CTC CCC CAC AAA AAA GAG GTT TTC ChPKCq M458A sense a1372g_t1373c GGA AAA CCT CTT TTT TGT GGC GGA GTA CCT CAA CGG AGG GhPKCq M458A antisense CCC TCC GTT GAG GTA CTC CGC CAC AAA AAA GAG GTT TTC ChPKCq H383Q sense c1149g AAA ATT GAG GAT TTT ATC TTG CAG AAA ATG TTG GGG AAA GGA AGT TThPKCq H383Q antisense AAA CTT CCT TTC CCC AAC ATT TTC TGC AAG ATA AAA TCC TCA ATT TThPKCq A410Y sense g1228t_c1229a_c1230tGAA AAC CAA TCA ATT TTT CGC AAT AAA GTA TTT AAA GAA AGA TGT GGT CTT GAT GGA CGhPKCq A410Y antisense CGT CCA TCA AGA CCA CAT CTT TCT TTA AAT ACT TTA TTG CGA AAA ATT GAT TGG TTT TChPKCq M444L sense a1330t CAT CCG TTT CTG ACG CAC TTG TTT TGT ACA TTC CAG AhPKCq M444L antisense TCT GGA ATG TAC AAA ACA AGT GCG TCA GAA ACG GAT GhPKCq H383I sense c1147a_a1148t AAA CTA AAA ATT GAG GAT TTT ATC TTG ATC AAA ATG TTG GGG AAA GGA AGT TTT GGhPKCq H383I antisense CCA AAA CTT CCT TTC CCC AAC ATT TTG ATC AAG ATA AAA TCC TCA ATT TTT AGT TThPKCq M385V sense a1153g ACT AAA AAT TGA GGA TTT TAT CTT GCA CAA AGT GTT GGG GAA AGG AAGhPKCq M385V antisense CTT CCT TTC CCC AAC ACT TTG TGC AAG ATA AAA TCC TCA ATT TTT AGThPKCq A410V sense c1229t CAA TCA ATT TTT CGC AAT AAA GGT CTT AAA GAA AGA TGT GGT CTT GAhPKCq A410V antisense TCA AGA CCA CAT CTT TCT TTA AGA CCT TTA TTG CGA AAA ATT GAT TGhPKCq M444I sense g1332a TCC GTT TCT GAC GCA CAT ATT TTG TAC ATT CCA GAC ChPKCq M444I antisense GGT CTG GAA TGT ACA AAA TAT GTG CGT CAG AAA CGG AhPKCq F445I sense t1333a CGT TTC TGA CGC ACA TGA TTT GTA CAT TCC AGA CChPKCq F445I antisense GGT CTG GAA TGT ACA AAT CAT GTG CGT CAG AAA CG 101 4.2.3 PKCθ protein production – 293T overexpression In order to produce recombinant PKCθ, a 293T overexpression system was employed. 8 μg of pcDNA3-nFLAG- PKCθ/AS2 plasmid DNA was transiently transfected into ~70% confluent 10 cm plates of 293T cells using polyethylenimine (PEI) at a 3:1 ratio (PEI:DNA). After 48 hours, plates were gently washed with D-PBS, and on plate lysis was performed with a non-denaturing lysis buffer (50 mM Tris-HCl pH 7.4, 150 mM NaCl, 1 mM EDTA, 1% Triton X-100, 1 mM NaF, 2.5 mM sodium pyrophosphate, 1 mM PMSF, 1 mM Na3VO4, 1x Protease Inhibitor Cocktail Set III (Calbiochem cat# 539134)). Lysates were clarified by 16,000 x g centrifugation at 4 °C for 10 minutes before pre-clearing overnight with pre-washed Sepharose 4B beads (Sigma cat# 4B200), 20 μL of bead volume per 10 cm plate. Pre-cleared lysates were prepared by centrifugation at 1,000 x g at 4 °C for 2 minutes and added to pre-washed FLAG-M2 agarose beads (Sigma cat #A2220), 10 μL of bead volume per 10 cm plate. FLAG immunoprecipitation (FLAG-IP) was performed overnight with rotation at 4 °C. FLAG-IP solutions were centrifuged at 1,000 x g at 4 °C for 2 minutes and the unbound fraction was removed. Beads were resuspended in 500 μL of chilled non-denaturing lysis buffer using a wide bore pipette tip and loaded into SigmaPrep spin columns (Sigma cat# SC1000). Beads were washed 3x with 500 μL non-denaturing lysis buffer followed by 3x 500 μL wash/kinase buffer (20 mM HEPES pH 7.4, 150 mM NaCl, 1 mM EGTA, 1 mM NaF, 2.5 mM sodium pyrophosphate, 1 mM PMSF, 1 mM Na3VO4, 1x Protease Inhibitor Cocktail Set III (Calbiochem)). FLAG tagged PKCθ was eluted off the beads by adding FLAG elution buffer (250 ng/mL FLAG peptide (Sigma cat# F4799), 20 mM HEPES pH 7.4, 150 mM NaCl, 1 mM EGTA, 1 mM NaF, 2.5 mM sodium pyrophosphate, 1 mM PMSF, 1 mM Na3VO4, 1x Protease Inhibitor Cocktail Set III (Calbiochem)) with agitation at room temperature for 30 minutes, 20 μL per 10 μL of bead volume (per 10 cm plate). In-gel semi-quantitative analysis, relative to a BSA standard, shows a consistent concentration of 50-100 ng/mL for PKCθ WT and 250-500 ng/mL for PKCθ AS2-M385V. 102 4.2.4 Lentiviral production and transduction – HPBALL cells High titer, replication defective lentivirus was produced utilizing pCMVΔR8.74 (Addgene# 22036), pCMV-VSV-G (Addgene# 8454), and pRSV-Rev (Addgene# 12253) packaging vectors by transient transfection of 293T producer cells, where necessary virus was concentrated using PEG-8000. Viral transduction of the T-ALL cell line HPBALL was performed by spinoculation with 4 μg/mL polybrene as described previously. Transduced HPBALL cells were FACS sorted by GFP expression and cultured in RPMI 1640 medium supplemented with 10% FBS, 1 mM sodium pyruvate, 2 mM L-glutamine, and antibiotics unless otherwise indicated. 4.2.5 In vitro kinase assay In vitro kinase assays were performed using dephosphorylated myelin basic protein (dMBP) (ActiveMotif cat# 31314) or whole cell lysate as substrates. When dMBP was use as a substrate, the following reaction conditions were used: 20 mM HEPES pH 7.4, 150 mM NaCl, 1 mM EGTA, 10 mM MgCl2, 1 μm PMA, 500 μg/mL phosphatidylserine, 2 μg dMBP, 100 μM ATPγS (Abcam cat# ab92570) or ATPγS analog (N6- Benzyladenosine (Bn) or N6- Cyclopentyladenosine (Pe) (BioLog cat# B 072, C 116)), 1 mM NaF, 2.5 mM sodium pyrophosphate, 1 mM PMSF, 1 mM Na3VO4, 1x Protease Inhibitor Cocktail Set III) along with the appropriate amount of immunoprecipitated PKCθ (1-10 μL) in a final reaction volume of 30 μL. For in vitro kinase assays on whole cell lysates, T-ALL cell lines grown as previously indicated were lysed using a kinase assay compatible, non-denaturing lysis buffer (20 mM HEPES pH 7.4, 150 mM NaCl, 1 mM EGTA, 0.9% Triton X-100, 1 mM NaF, 2.5 mM sodium pyrophosphate, 1 mM PMSF, 1 mM Na3VO4, 1x Protease Inhibitor Cocktail Set III). For small scale testing, most in vitro kinase assays were performed with 20 μg of lysate as a substrate using the following reaction conditions in a 30 μL final reaction volume: 20 mM HEPES pH 7.4, 150 mM NaCl, 1 mM EGTA, 10 mM MgCl2, 0.3% Triton X-100, 1 μM PMA, 100 μM ATPγS (or Bn-ATPγS)), 50-100 μM ATP, 1 mM GTP, 1 mM NaF, 2.5 mM sodium pyrophosphate, 1 mM PMSF, 1 mM Na3VO4, 1x Protease Inhibitor Cocktail Set III (Calbiochem) along with the appropriate amount of immunoprecipitated PKCθ (1-10 μL). Unless otherwise indicated, large scale reactions used for mass spectrometry used the same reaction conditions, but with 100-150 μg of lysate with 50-75 μL of 103 immunoprecipitated PKCθAS2-M385V in a final volume of 150-225 μL. Reactions were performed at room temperature for 2 hours with agitation, followed by alkylation with p-Nitrobenzyl mesylate (PNBM) (Abcam cat# ab138910) for 30 minutes at room temperature. Initial in vitro kinase assay involved using PKCθ still attached to the FLAG immunoprecipitation beads in an attempt to allow isolation of the substrate fraction. However, later experiments used FLAG peptide eluted PKCθ because of observed assay efficacy. 4.2.6 BioID labelling PKCθ/BirA* transduced variants or empty transduced HPBALL cells were grown in RPMI 1640 medium supplemented with 10% FBS, 1 mM sodium pyruvate, 2 mM L-glutamine, and antibiotics. Where indicated, FBS was reduced to 5% and base media was changed to DMEM, IMDM or DMEM/F12. Labelling was initiated with the addition of biotin (Sigma) to a final concentration of 50 μm. Labelling was allowed to proceed for 24 hours (or less where indicated) before whole cell lysates were prepared with RIPA/SDS lysis buffer (50 mM Tris-HCl pH 7.4, 150 mM NaCl, 1 mM EDTA, 1% NP-40, 0.25% Sodium Deoxycholate 0.1% SDS, 1 mM NaF, 2.5mM Na pyrophosphate, 1 mM PMSF, 1 mM Na3VO4, 1x Protease Inhibitor Cocktail Set III (Calbiochem)). 4.2.7 APEX biotin phenol labelling PKCθ/APEX transduced variants or empty transduced HPBALL cells were grown in RPMI 1640 medium supplemented with 10% FBS, 1 mM sodium pyruvate, 2 mM L-glutamine, and antibiotics. Prior to labelling, cells were resuspended in growth media containing 500 μm biotin phenol ((3aS,4S,6aR)-hexahydro-N-[2-(4-hydroxyphenyl)ethyl]-2-oxo-1H-thieno[3,4-d]imidazole-4-pentanamide) (Iris Biotech) at a concentration of 3e6 cells/mL and incubated at 37 °C /5 %CO2 for 30 minutes. If required, 5 minutes prior to labelling PMA was added to final concentration of 100 nM in order to activate PKCθ. Labelling was initiated by the addition of 1 mM H2O2 and the cells were immediately centrifuged at 500 x g for 1 minute before the supernatant was quickly removed and the cell pellet resuspended in 5 mL cold quencher solution (10 mM sodium azide, 10 mM sodium ascorbate, and 5 mM Trolox in D-PBS) to halt labelling. Cell were washed four additional times in 5 mL of quencher solution before whole cell 104 lysates were prepared with RIPA/SDS lysis buffer with addition of quenchers at concentration indicated above. 4.2.8 APEX alkyne phenol labelling – click chemistry Labelling with alkyne phenols (3-ethynylphenol (3EP) (Sigma) or N-[2-(4-hydroxyphenyl)ethyl]hex-5-ynamide hexyne-tyramide (HexT) (courtesy of Dr Kuo-Shyan Lin) was performed as described above for biotin phenol, with the following modifications. Sodium azide was removed from the quencher solution, and lysis was performed using a click chemistry compatible lysis buffer 50 mM HEPES, 1% Triton X-100, 1 mM NaF, 2.5 mM Na pyrophosphate, 1 mM PMSF, 1 mM Na3VO4, 1x Protease Inhibitor Cocktail Set III (Calbiochem), 10 mM sodium ascorbate, and 5 mM Trolox). Click reactions were performed with final concentrations of 1 mg/mL lysate, 800 μm CuSO4, 500 μm THPTA, 40 μM biotin azide or azide beads (50 μL/mL of reaction), 1 mM NaF, 2.5 mM Na pyrophosphate, 1 mM PMSF, 1 mM Na3VO4, 1x Protease Inhibitor Cocktail Set III (Calbiochem), 10 mM sodium ascorbate, and 5 mM Trolox. Reactions were incubated at room temperature for 1 hour with rotation. HexT was synthesized from 5-hexynoic acid (Sigma; cat# 544000) and tyramine (Sigma; cat# T90344) using the methodology described by Rhee et al [406] (Figure. A.1A). Identity and purity of the synthesized HexT compound was confirmed using NMR and mass spectrometry (courtesy of Dr Chris Hughes and Dr Kuo-Shyan Lin) (Figure. A.1B/1C). 4.2.9 Western Blotting Whole cell protein lysates, immunoprecipitated proteins and in vitro kinase assays were separated by SDS-PAGE, and transferred to PVDF membranes. Blots were probed with antibodies directed against FLAG (Sigma clone# M2), HA tag (Sigma clone# HA-7), MYC tag (Cell Signalling Technologies clone# 9B11), PKCθ (Santa Cruz #C-19), thiophosphate ester (Abcam clone# 51-8) and β-actin (Sigma clone# AC-15), followed by HRP-conjugated secondary antibody, or directly detected with streptavidin conjugated HRP (Calbiochem) and detected by chemiluminescence. Where necessary to prevent detection of IP antibodies, TrueBlot Band HRP-conjugated secondary antibodies were used (Rockland). Intensities were quantitated using ImageJ software. 105 4.2.10 Immunoprecipitation/SA pull down For the PKCθAS IP experiments, the large scale in vitro kinase assays were first run through PD10 desalting columns (GE Healthcare) to remove PNBM prior to immunoprecipitation. Fractions of lysates eluted off the columns were quantified for protein content using the BCA protein assay kit (Pierce cat# 23225) to determine which fractions to pool for IP. Equal amounts of protein from each condition (100-200 μg) were pre-cleared with Sepharose 4B (Sigma cat# 4B200) beads for 2-15 hours before a 15 hours incubation with 50 μL Protein G magnetic (ThermoFisher cat# 10003D), or agarose, (ProteinMods cat# PGGH) beads pre-loaded with 3-4 μg of anti-thiophosphate ester antibody (Abcam cat# ab133473) to enrich for thiophosphorylated proteins. Beads were washed five times in RIPA/SDS lysis buffer before being resuspended in 1x LDS loading dye (ThermoFisher) and heated to 95 °C for 10 minutes to elute proteins. For the PKCθAPEX IP experiments, biotin phenol labelled lysates were first run through PD10 desalting columns (GE cat# 28-9180-08) to remove free biotin phenol. Fractions eluted off the columns were quantified for protein content using the 660 nm Protein Assay Reagent (Pierce cat# 22662) to determine which fractions to pool for streptavidin pull down. This cleanup step was not necessary for HexT, most likely due to improved membrane permeability. Equal amounts of protein from each condition (500-1000 μg) were pre-cleared with Sepharose 4B beads for 2-15 hours before a 15 hours incubation with 50 μL streptavidin (Pierce cat# 88816) or azide (Jena Biosciences cat# CLK-1036) magnetic beads to enrich for biotinylated or alkylated proteins, respectively. For PKCθAPEX biotin phenol labelled samples, beads were then washed twice with RIPA/SDS lysis buffer followed by a 2 M urea wash and two additional RIPA/SDS lysis buffer washes before being resuspended in 50 μL 1x LDS loading dye with 3 mM biotin and heated to 95 °C for 10 minutes to elute proteins. For PKCθAPEX HexT labelled samples, beads were washed three times with RIPA/SDS lysis buffer followed by three 8 M urea washes before being resuspended in 50 μL in 50 mM HEPES buffer pH 8. 106 4.2.11 Sample preparation and tryptic digestion PKCθAS thiophosphate enriched proteins, PKCθAPEX biotin phenol labelled biotin enriched proteins and PKCθAPEX HexT labelled alkyne enriched azide bead bound proteins were reduced with 5 mM TCEP at 60 °C for 45 minutes and alkylated with 10 mM iodoacetamide at room temperature for 1 hour. To quench the reaction, DTT was added to 10 mM for 10 minutes. PKCθAS thiophosphate enriched proteins and PKCθAPEX biotin phenol labelled biotin enriched proteins were processed for mass spectrometry analysis using the SP3 protocol as described by Hughes et al. [411]. Briefly, 20 μg of Sera-Mag bead mix (1:1 Sera-Mag Speed Beads, GE Healthcare, cat# 45152105050250 & cat# 65152105050250) was added to proteins along with acetonitrile to 50% (v/v) were added to proteins, and then incubated for 8 minutes at room temperature. The protein-beads complexes were applied for 2 minutes to a magnetic rack, and then washed twice with 200 μL 70% ethanol, followed by one wash with 180 μL 100% acetonitrile before being resuspended in 10 μL 50 mM HEPES buffer pH 8. These samples, along with PKCθAPEX HexT labelled alkyne enriched azide bead bound proteins, were digested into tryptic fragments. Trypsin/rLysC enzyme mix (Promega cat# V5073) at a ~1:25 ratio of enzyme to protein was added to each sample and digests were incubated for 14 hours at 37 °C. Following digestion, tryptic peptides were either directly cleaned up for mass spectrometry analysis or dimethyl labelled before clean up. 4.2.12 Dimethyl labelling and peptide clean up For dimethyl labelling, solutions of 155 mM sodium cyanoborohydride (NaBH3CN or NaBD3CN) and 1% formaldehyde (CH2O, CD2O, 13CD2O) were prepared fresh. “Light” labels are a combination of CH2O with NaBH3CN, “Medium” labels are a combination of CD2O with NaBH3CN, and “Heavy” labels are a combination of 13CD2O with NaBD3CN and add 28.0313 Da, 32.0564 Da and 36.0757 Da of mass per peptide, respectively. 1 μL of appropriate sodium cyanoborohydride solution and 1 μL of formaldehyde solution were added to the tryptic peptides and incubated for 30 minutes. An additional 1 μL of appropriate sodium cyanoborohydride and formaldehyde solutions were added and incubated for another 30 minutes to ensure complete labelling, followed by quenching the reaction by the addition of 1 μL 100 mM glycine solution for 5 minutes. To clean up 107 labelled or unlabelled peptides, acetonitrile was added to >95% and incubated for 8 minutes, followed by a further 2 minutes on a magnetic rack and one wash with 180 μL 100% acetonitrile. Peptides were released from the beads by reconstitution in 6 μL H2O/2% DMSO and transferred into a new tube. For dimethyl labelled samples the peptide solution from one sample was used to reconstitute the other. For PKCθAPEX HexT dimethyl labelled peptides samples were cleaned up using SepPak cartridges (C18-t, 50 mg, Waters) before being combined and concentrated in a SpeedVac centrifuge. All samples were acidified with formic acid prior to mass spectrometry analysis. 4.2.13 Mass spectrometry analysis Tryptic peptides were analysed out on an Orbitrap Fusion Tribrid mass spectrometer (ms) platform (Thermo Scientific). Samples were introduced into the ms by coupled liquid chromatography nano-electrospray using an Easy-nLC 1000 system (Thermo Scientific). Columns used for trapping and separations were packed in-house. Trapping columns were packed in 100 μm internal diameter capillaries to a length of 20 mm with C18 beads (Reprosil-Pur, Dr. Maisch, 3 μm particle size). After trapping, gradient elution of peptides was performed on a C18 (Reprosil-Pur, Dr. Maisch, 3 μm particle size) column packed in-house in Pico-Frit capillaries (New Objective, 75 μm internal diameter). Elution was performed with a gradient of mobile phase A (98.9% water, 1% DMSO, and 0.1% formic acid) to 22% B (98.9% acetonitrile, 1% DMSO and 0.1% formic acid) over 90 minutes, and to 40% B over 20 minutes, with final elution and equilibration using a further 25 minutes at a flow rate of 350 nL/minute. Data acquisition on the Orbitrap Fusion (control software version 2.0) was carried out using a data-dependent method. Survey scans covering the mass range of 350 – 1500 were acquired at a resolution of 120,000 (at m/z 200), with quadrupole isolation enabled, an S-Lens RF Level of 60%, a maximum fill time of 50 milliseconds, and an automatic gain control (AGC) target value of 2e5. For MS2 scan triggering, monoisotopic precursor selection was enabled, charge state filtering was limited to 2 – 5, an intensity threshold of 5eE3 was employed, and dynamic exclusion of previously selected masses was enabled for 45 seconds with a tolerance of 10 ppm. MS2 scans 108 were acquired in the ion trap after HCD fragmentation with a maximum fill time of 60 milliseconds, quadrupole isolation, an isolation window of 1 m/z, collision energy of 32%, activation Q of 0.25, injection for all available parallelizable time, and an AGC target value of 5e3. 4.2.14 Data analysis Mass spectrometry data was processed using Proteome Discoverer Software (ver. 2.1.0.62). MS2 spectra were searched using Sequest HT against the UniProt Human proteome database as well as a list of common contaminants (20,239 total sequences). The following criterion were specified: trypsin enzyme, 2 missed cleavages allowed, minimum peptide length of 6, precursor mass tolerance of 20 ppm, and a fragment mass tolerance of 0.8 Da. Oxidation of methionine was set as a variable modification, and carbamidomethylation of cysteine was set as a fixed modification. Peptide spectral match error rates were determined using the target-decoy strategy coupled to Percolator modeling of positive and false matches using the software Percolator [412, 413]. Only peptide spectral matches reaching the confidence threshold of ‘High’ (q-value<0.01) were used for further analysis. Peptide spectral match data were processed further in the R scientific computing language. Redundant peptides and those matching to contaminant proteins were removed. Only unique peptides were retained for all quantification analyses. Raw peptides were normalized using the Limma package [414] in R with the setting ‘VSN’ [415, 416] and aggregated into proteins using a median representation of all those assigned to a given protein. Fold change values were calculated at the protein level and used to assign protein species as described in results section. Protein lists created were compared to those of previously described PKCθ phosphotargets, PKCθ interactors and TCR signalling components. Previously described PKCθ phosphorylation substrates were compiled from http://www.phosphosite.org [417] and http://140.138.144.141/~RegPhos/index.php [418] databases (Table B.1). Previously described PKCθ interactors were compiled from http://thebiogrid.org/ [419] and http://mentha.uniroma2.it/index.php [420] (Table B.2). Known TCR signalling components were taken from 109 http://amigo.geneontology.org/amigo/landing [421] (GO: 0050852 (mammalian)) (Table B.3). Additional comparisons were made with membrane proteins (GO:0016020) and plasma membrane proteins (GO:0005886). 4.3 Results 4.3.1 Identification and mutation of gatekeeper residue in PKCθ To identify the PKCθ gatekeeper residue, a search was performed on the Kinase Sequence Database (http://sequoia.ucsf.edu/ksd/) [422]. This database aligns kinases by the amino acid sequence similarity of their catalytic domains and groups them into families. Currently, there are 7128 protein kinases from 948 organisms grouped into 287 families. Residues contacting ATP in the active site as well as those in the gatekeeper position are identified. The methionine at position 458 (M458) sits deep inside the ATP binding pocket (Figure 4.4A/B) and acts as the gatekeeper residue in human PKCθ [334]. To create AS versions of PKCθ, site-direct mutagenesis was employed to change base pairs corresponding to M458 in a PKCθ expression vector (pcDNA3-nFLAG-PKCθ). PKCθAS1, where the gatekeeper methionine is changed to a glycine, was achieved by mutating AT to GG (ATG to GGG), whilst PKCθAS2, where the gatekeeper methionine is changed to an alanine, was achieved by mutating AT to GC (ATG to GCG). 110 Figure 4.4 3D structure of PKCθ Structure of human PKCθ (residues 362-706), including kinase domain (residues 380-634) resolved from crystal structure of human PKCθ complexed with staurosporine at 2A resolution (Protein Data Bank ID number 1XJD). (A) Calculated surface of PKCθ showing ATP binding pocket containing M458 gatekeeper residue (red). (B) Ribbon structure of PKCθ showing M458 gatekeeper residue (red), β-sheets (purple), and residues mutated in an attempt to rescue catalytic activity (SSM; H383, A410, M444, M385, F445) (green). 4.3.2 Overexpression and initial testing of PKCθAS When PKCθ and PKCθAS variants were transiently overexpressed in 293T both PKCθAS1 and PKCθAS2 were observed to be significantly impaired in their protein expression level, as determined by FLAG immunoblot (Figure 4.5). The expression appeared to be slightly more reduced for PKCθAS1 compared to PKCθAS2. After FLAG immunoprecipitation, equal volumes were used for in vitro kinase assays using myelin basic protein (MBP) as a substrate and ATPγS. The immunoprecipitation procedure seemed to somewhat minimize the differences in the amount of input PKCθ used in (A) (B) 111 each assay relative to what was originally expressed. Despite this, the efficiency of the assay with PKCθAS appeared to be significantly impaired relative to PKCθWT, as indicated by a greatly reduced intensity of thiophosphate ester signal on MBP (Figure 4.5). The fact that PKCθWT was able to use the ATPγS so efficiently suggests that the gatekeeper mutation is causing the lack of activity, as opposed to the utilization of the γ-thiophosphate containing ATPγS, which has been reported to not be utilized by some kinases [380]. Because of how significant the loss of catalytic activity was for both PKCθAS1 and PKCθAS2 variants, it is very difficult to determine if one version retains more catalytic activity than the other. A faint band with PKCθAS2 was discernible; possibly suggesting PKCθAS2 may retain more catalytic activity. However, as there appears to be more PKCθAS2 loaded into the assay compared to PKCθAS1, this conclusion cannot be made with confidence. Empty WTAS1 AS2ASPKCθFLAG/PKCθThiophosphate ester/MBPActinWhole cell lysateFLAG/PKCθInvitro kinase assay Figure 4.5 112 Figure 4.5 PKCθAS variants have reduced expression and catalytic activity In vitro kinase assay. FLAG tagged PKCθWT, PKCθAS1 (M458G) or PKCθAS2 (M458A) was transiently overexpressed in 293T cells followed by FLAG immunoprecipitation and an in vitro kinase assay using ATPγS and MBP as a substrate. Whole cell lysates were probed with antibodies directed against FLAG (PKCθ) and β-actin, in vitro kinase assays were probed with antibodies directed against FLAG (PKCθ) and thiophosphate ester (MBP). 4.3.3 Second-site mutation rescue of catalytic activity In an attempt to rescue the catalytic activity of PKCθAS, a second-site mutation suppressor (SSM) strategy was attempted. Loss of catalytic activity after gatekeeper mutation has been reported by other groups with different kinases [423-427] and may affect ~30% of kinases [387]. In some cases, catalytic activity can be restored through SSM [387]. Because PKCθAS2 showed slightly better expression and possibly better catalytic activity than PKCθAS1 we decided to move forward with the second-site mutation strategy on PKCθAS2. There is no universal site that, when mutated, restores catalytic activity to kinases that have lost activity after gatekeeper mutations, so a trial and error approach must be used. Firstly, I ensured that residues at positions -11 and -1 relative to the conserved DFG kinase motif are leucine and alanine in hPKCθ, respectively, as these residues at those positions have been found to be beneficial to maintain kinase activity [428]. Sequence conservation between hPKCθ and rat PKCδ (rPKCδ) was then compared. PKCδ is the closest PKC family member to PKCθ based on their amino acid sequence similarity for both their full-length (66.02%) (Figure 4.6A) and catalytic domains (76.86%) (Figure 4.6B). Fortunately rPKCδ has previously been reported to be successfully mutated to rPKCδAS2 without loss of activity [380, 393]. Amino acids discordant between hPKCθ and rPKCδ in the antiparallel β-sheets N-terminal of the gatekeeper residues were focused on, as these regions have been reported to be hotspots for second-site mutations that restore activity after gatekeeper mutation (Figure 4.7A) [387]. H383, A410 and M444 were selected for the initial round of second-site mutations because sites analogous to these in other kinases have been previously reported as restorative second-site mutations [423-425, 427]. For example, M444 from human PKCθ is at an equivalent position to M429 from mouse Syk and 113 M152 from human STK38, both of which have been reported to restore activity after gatekeeper mutation (Figure 4.7B). Figure 4.6 PKCθ shares highest degree of homology to PKCδ Phylogenetic analysis of human PKC family protein sequences. Analysis of (A) full-length and (B) catalytic domain sequence homology between PKC isoforms. PKC subfamilies indicated (conventional (PKCα, PKCβ, PKCγ); novel (PKCε, PKCη, PKCθ, and PKCδ); atypical (PKCι PKCζ)). hPKCθ indicated in red and closest related isoform (hPKCδ) indicated in green. Uniprot accession numbers for each protein sequence, as well as catalytic domains residue numbers, shown. (A) (B) 114 Q04759 hPKCθ 361 EPELNKERPSLQIKLKIEDFILHKMLGKGSFGKVFLAEFKKTNQFFAIKALKKDVVLMDD 420P09215 rPKCδ 340 ------------NRCRLENFTFQKVLGKGSFGKVLLAELKGKERYFAIKYLKKDVVLIDD 387: ::*:* ::*:*********:***:* .:::**** *******:**Q04759 hPKCθ 421 DVECTMVEKRVLSLAWEHPFLTHMFCTFQTKENLFFVMEYLNGGDLMYHIQSCHKFDLSR 480P09215 rPKCδ 388 DVECTMVEKRVLALAWENPFLTHLICTFQTKDHLFFVMEFLNGGDLMFHIQDKGRFELYR 447************:****.*****::******:.******:*******:***. :*:* * Q04759 hPKCθ 481 ATFYAAEIILGLQFLHSKGIVYRDLKLDNILLDKDGHIKIADFGMCKENMLGDAKTNTFC 540P09215 rPKCδ 448 ATFYAAEIICGLQFLHGKGIIYRDLKLDNVMLDKDGHIKIADFGMCKENIFGENRASTFC 507********* ******.***:********::******************::*: ::.***383385410444445458DFG-1DFG-11 444458444458Q04759 hPKCθ 417 LMDDDVECTMVEKRVLSLAWEHPFLTHMFCTFQTKENLFFVMEYLNGGDLMYHIQSCHKF 476Q15208 hSTK38 126 LEKEQVGHIRAERD-ILVEADSLWVVKMFYSFQDKLNLYLIMEFLPGGDMMTLLMKKDTL 184* .::* .*: : : : ::.:** :** * **:::**:* ***:* : . ..:Q04759 hPKCθ 433 SLAWEHPFLTHMFCTFQTKENLFFVMEYLNGGDLMYHIQSCHKFDLSRATFYAAEIILGL 492P48025 mSYK 418 MQQLDNPYIVRMIGICE-AESWMLVMEMAELGPLNKYLQQNRHIKDKNIIELVHQVSMGM 476:.*::.:*: : *. ::*** : * * ::*. :::. .. . :: :*:429152442166 Figure 4.7 PKCθ protein sequence alignments Alignments with human PKCθ. hPKCθ protein sequence aligned with (A) the successfully gatekeeper mutated rat PKCδ and (B) the successfully SSM catalytically rescued human STK38 and mouse SYK kinases. Features indicated include β-sheets (blue), gatekeeper residues (red), SSM (green) and GXGXXG, invariant lysine, and DFG (bold/underlined). 4.3.3.1 Second-site mutation rescue of catalytic activity – mutation to rPKCδ residues Because rPKCδ is catalytically competent after AS2 mutation, I decided to mutate the hPKCθ second-site mutation candidates to the corresponding amino acid in rPKCδ, PKCθAS2-H383Q, PKCθAS2-A410Y, PKCθAS2-M444L (Table 4.4). Of these three SSM sites M444L not only had the best expression (Figure 4.8A), but also the best (A) (B) 115 utilization of ATPγS, suggesting some degree of restored catalytic activity (Figure 4.8B). However, when compared to PKCθWT, the improvement still fell well short of complete rescue. In addition to utilizing ATPγS, albeit at a reduced level compared to PKCθWT, PKCθAS2-M444L was also able to utilize one of the N6 substituted ATP analog tested, N6- benzyladenosine- 5'- O- (3- thiotriphosphate) (Bn-ATPγS ), whereas PKCθWT was not (Figure 4.9A/B/C). In fact, the utilization of Bn-ATPγS by PKCθAS2-M444L seemed to be higher that of ATPγS. A second N6 substituted ATP analog was tested, N6- cyclopentyladenosine- 5'- O- (3- thiotriphosphate) (Pe-ATPγS), but was not utilized by any variant of PKCθ (Figure 4.10). Because the catalytic activity of PKCθAS2-M444L still remained significantly lower than PKCθWT for ATP, I decided to attempt a second round of second-site mutations on PKCθAS2. Table 4.4 SSM mutations created to attempt catalytic activity rescue Mutated to rPKCδ equivalent residueSSM: H383QSSM: A410YSSM: M444LMutated to β-branched residueSSM: H383ISSM: M385VSSM: A410VSSM: M444ISSM: F445I 116 Figure 4.8 M444L rPKCθ derived SSM variant improves PKCθAS2 expression and activity In vitro kinase assay. FLAG tagged PKCθWT or PKCθAS2 SSM (H383Q, A410Y, or M444L) was transiently overexpressed in 293T cells followed by FLAG immunoprecipitation and an in vitro kinase assay using ATPγS and MBP as a substrate. Whole cell lysates were probed with antibodies directed against FLAG (PKCθ) and β-actin, in vitro kinase assays were probed with antibodies directed against FLAG (PKCθ) and thiophosphate ester (MBP). WT AS2PKCθFLAG/PKCθThiophosphate ester/MBPActinWhole cell lysateFLAG/PKCθH383Q A410Y M444LInvitro kinase assay(ATP-γ-S)(A) (B) 117 BN-ATPgSFLAG/PKCθThiophosphate ester/MBPActinWhole cell lysateFLAG/PKCθInvitro kinase assay(ATP-γ-S)WT ASAS1 AS2 M444LPKCθInvitro kinase assay(Bn-ATP-γ-S)Thiophosphate ester/MBPFLAG/PKCθ Figure 4.9 PKCθAS2 M444L utilizes ATP analog more the ATP In vitro kinase assay. FLAG tagged PKCθWT or PKCθAS2 M444L was transiently overexpressed in 293T cells followed by FLAG immunoprecipitation and an in vitro kinase assay using ATPγS or Bn-ATPγS and MBP as a substrate. Whole cell lysates were probed with antibodies directed against FLAG (PKCθ) and β-actin, in vitro kinase assays were probed with antibodies directed against FLAG (PKCθ) and thiophosphate ester (MBP). 118 WT AS2-M444LPKCθThiophosphate ester/MBPFLAG/PKCθInvitro kinase assay- Bn Pe - Bn Pe - Bn Pe ATPγs1.00 1.10 2.27 4.13 Figure 4.10 PKCθAS2 M444L preferentially utilizes Bn-ATPγS analog In vitro kinase assay. FLAG tagged PKCθWT or PKCθAS2 M444L was transiently overexpressed in 293T cells followed by FLAG immunoprecipitation and an in vitro kinase assays using ATPγS, Bn-ATPγS or Pe-ATPγS and MBP as a substrate. Whole cell lysates were probed with antibodies directed against FLAG (PKCθ) and β-actin, in vitro kinase assays were probed with antibodies directed against FLAG (PKCθ) and thiophosphate ester (MBP). Values indicate intensity of thiophosphorylated MBP bands normalized to FLAG/PKCθ input and relative to PKCθAS2-M444L/ATPγS. 4.3.3.2 Second-site mutation rescue of catalytic activity – mutation to common β-branched residues For the second round of SSM, I decided to mutate the chosen residues to β-branched amino acids (valine, threonine and isoleucine). β-branched amino acids tend to be stabilizing in β sheet protein structures [429], and in some cases are able to counteract detrimental gatekeeper mutations [387]. In addition to the originally chosen sites (H383, A410, M444), an additional two sites in the antiparallel β-sheets for SSM (M385, F445) were chosen. These sites were chosen based on their discordance between hPKCθ and rPKCδ and being a β-branched amino acid only in rPKCθ (Figure 4.7A). The decision of which β-branched amino acid to mutate these sites to was in part dictated by the prevalence of β-branched amino acids conserved positions in a closely related set of kinases (Family 17), PKCθAS2-H383I, PKCθAS2-M385V, PKCθAS2-M410V, PKCθAS2-M444I, PKCθAS2-F445I (Figure 4.11, Table 4.4). Both PKCθAS2-M410V 119 and PKCθAS2-M385V were able to improve PKCθAS2 expression levels (Figure 4.12) and were selected for further testing by an in vitro kinase assay. Only PKCθAS2-M385V looked to slightly improve on the catalytic activity of PKCθAS2-M444L and was able to do this with both ATPγS (Figure 4.13A) and Bn-ATPγS (Figure 4.13B). PKCθAS2-M410V may also improve catalytic activity, but did not immunoprecipitate as efficiently in this experiment and so was difficult to assess accurately. Although not able to restore ATPγS utilization back to PKCθWT levels, PKCθAS2-M385V was able to utilize Bn-ATPγS to similar levels as PKCθWT utilization of ATPγS when PKCθ input was equalized (~50%) (Figure 4.14). Because the utilization of an N6 substituted ATPγS analog is required for downstream applications and thus not utilizing ATPγS should theoretically reduce competition with normal ATP in a whole cell lysate context. Combining the M444L and M385V did not show an improvement in PKCθAS2 expression over M385V and it was decided to move forward with PKCθAS2-M385V as the final PKCθAS variant. 120 Distribution of residues at position 383 in hPKCθ Distribution of residues at position 385 in hPKCθ Figure 4.11A/B (A) (B) 121 Distribution of residues at position 410 in hPKCθ Distribution of residues at position 445 in hPKCθ Figure 4.11 Residue distribution of SSM sites Analysis of most common amino acids at potential SSM sites. For the 4/5 SSM sites chosen for mutation to β-branched residues the residue was chosen was based on being the most frequently utilized β-branched residue at that site among closely related (family 17) kinases. (A) H383I, (B) M385V, (C) A410V, (D) F445I. Chosen β-branched residues indicated (hashed box). Kingdom kinases derived from Animalia (red), Plantae (green), Fungi (green), Bacteria (blue). Generated from http://sequoia.ucsf.edu/ksd/. (C) (D) 122 M444L A410V M385V F445I M444I L383IFLAG/PKCθActinWhole cell lysateWT AS2PKCθEmpty1.00 0.13 0.28 0.42 0.39 0.06 0.07 0.05 Figure 4.12 A410V and M385V β-branched SSM variants improve PKCθAS2 expression FLAG Western Blot for PKCθAS2 β-branched SSM expression. FLAG tagged PKCθWT, PKCθAS2, PKCθAS2-M444L or PKCθAS2 β-branched SSM variants (M410V, M385V, F445I, M444I or H383I) were transiently overexpressed in 293T cells. Whole cell lysates were probed with antibodies directed against FLAG (PKCθ) and β-actin. Values indicate intensity of FLAG/PKCθ bands normalized to actin and relative to PKCθWT. 123 Thiophosphate ester/MBPFLAG/PKCθInvitro kinase assay(ATP-γ-S)Invitro kinase assay(Bn-ATP-γ-S)Thiophosphate ester/MBPFLAG/PKCθWT AS2PKCθM444L M385V A410VEmpty1.00 1.54 0.76 Figure 4.13 PKCθAS2-M385V has improved catalytic activity In vitro kinase assay. FLAG tagged PKCθWT or PKCθAS2 (M444L, M385V and A410V) was transiently overexpressed in 293T cells followed by FLAG immunoprecipitation and an in vitro kinase assay using ATPγS or Bn-ATPγS and MBP as a substrate. Whole cell lysates were probed with antibodies directed against FLAG (PKCθ) and β-actin, in vitro kinase assays were probed with antibodies directed against FLAG (PKCθ) and thiophosphate ester (MBP). Values indicate intensity of thiophosphorylated MBP bands relative to PKCθAS2-M444L. 124 Thiophosphate ester/MBPFLAG/PKCθWTAS2M385VEmptyInvitro kinase assay(ATP-γ-S)Invitro kinase assay(Bn-ATP-γ-S)WTAS2M385VEmpty1.00 0.04 0.03 0.50PKCθ PKCθ Figure 4.14 PKCθAS2-M385V has approximately half catalytic activity of wildtype PKCθ In vitro kinase assay. FLAG tagged PKCθWT or PKCθAS2-M385V was transiently overexpressed in 293T cells followed by FLAG immunoprecipitation and an in vitro kinase assay using ATPγS or Bn-ATPγS and MBP as a substrate, input mass of PKCθ into each reaction was attempted to similar. Probed with antibodies directed against FLAG (PKCθ) and thiophosphate ester (MBP). Values indicate intensity of thiophosphorylated MBP bands normalized to PKCθ FLAG/PKCθ and relative to PKCθWT/ATPγS. 4.3.4 PKCθAS385 use with whole cell lysate To identify PKCθ direct phosphotargets in T-ALL, I performed kinase assays using PKCθAS2-M385V and Bn-ATPγS on HPBALL lysates. In my hands PKCθWT or PKCθAS2-M385V cannot be overexpressed to a high degree in T-ALL cells (Figure 4.15A). Initial attempts using lysates from HPBALL cells overexpressing PKCθAS2-M385V and using Bn-ATPγS did not show any additional banding compared to overexpression of PKCθWT or empty transduced control (Figure 4.15B). It has been reported that high expression of an analog sensitive kinase is required to achieve an acceptable signal to noise ratio [428, 430]. For this reason, I performed an in vitro kinase assay on HPBALL whole cell lysates with recombinant PKCθAS2-M385V obtained from transiently transfected 293Ts cells, similarly to the initial MBP optimization experiments. When roughly equal amounts of PKCθ (WT or AS-M385V) 125 are used for in vitro kinase assays with HPBALL whole cell lysate and Bn-ATPγS as substrates, many unique bands are observed with PKCθAS2-M385V (Figure 4.16). This result shows that under these conditions PKCθWT cannot utilize Bn-ATPγS efficiently, even for its own auto-phosphorylation, and that Bn-ATPγS is not being utilized by other kinases in the HPBALL lysate. A similar banding pattern was observed when lysates from other T-ALL cell lines were used as substrates, suggesting that major phosphotarget species may be commonly expressed in T-ALL (Figure 4.17); further analysis would be required to definitively make such a statement. In addition, to improve the transfer of thiophosphates onto targets, I attempted to clear the substrate proteome of pre-existing phosphorylation by lambda phosphatase treatment. However, this treatment actually resulted in reduced band intensity and was not pursued further. In this same experiment it was also shown that in vitro kinase assays using PKCθAS2-M385V along with BSA or no substrate resulted in no additional banding beyond that of auto-phosphorylation, and gives me further confidence that banding seen is not coming from proteins pulled down with PKCθ. In an attempt to select for stronger PKCθ interactors and thus potentially enrich for “true” PKCθ phosphotargets, the utility of increasing the salt concentration (from 150 mM NaCl to 300 mM or 600 mM NaCl) used during the in vitro kinase assay was investigated. At high salt concentration no major changes in the banding pattern were observed, but we did see a reduction in banding intensity suggesting an effect in thiophosphate deposition (Figure 4.18). I attempted to incorporate high salt conditions into some of the subsequent enrichment and mass spectrometry identification experiments. 126 Whole cell lysateActinPKCθWTAS2M385VEmpty1.00 2.03 1.85PKCθ WTAS2M385VEmptyThiophosphate esterPKCθInvitrokinaseassay(Bn-ATP-γ-S) Figure 4.15 PKCθWT and PKCθAS2-M385V can only be stably overexpressed to a moderate degree in the T-ALL cell line HPBALL; not sufficient for in vitro kinase assays PKCθ blot and in vitro kinase assay. HPBALL cells were retrovirally transduced with PKCθWT or PKCθAS2-M385V. (A) Lysates from cells were probed with antibodies directed against PKCθ and β-actin. Values indicate intensity of total PKCθ normalized to β-actin and relative to endogenous PKCθ (Empty). (B) 20 μg of lysates were used for in vitro kinase assay using Bn-ATPγS as a substrate. Probed with antibody directed against thiophosphate ester. (A) (B) 127 FLAG/PKCθThiophosphate esterActinInvitro kinase assay(Bn-ATP-γ-S)WTAS2 M385VEmpty PKCθ10 5 2 1 10 5 2 1 uLFLAG-PKCθ input2.59 1.55 2.02 1.00 Figure 4.16 PKCθAS2-M385V utilizes ATP analog Bn-ATPγS to produce a unique banding pattern with HPBALL lysate In vitro kinase assay. FLAG tagged PKCθWT or PKCθAS2-M385V was transiently overexpressed in 293T cells followed by FLAG immunoprecipitation and an in vitro kinase assay using Bn-ATPγS and HPBALL whole cell lysate as a substrate. Serial dilutions of eluted PKCθ were used as input into each reaction. Probed with antibodies directed against FLAG (PKCθ), thiophosphate ester and β-actin (loading control). Values indicate intensity of FLAG/PKCθ bands normalized to β-actin and relative to 1 μL PKCθWT input. 128 - + - + - + - +ALLSIL CUTTL1 JURKAT RPMI HPBALLHPBALLλPP BSAPKCθAS2-M385V FLAG/PKCθActinThiophosphate ester- + - + - + +Invitrokinaseassay(Bn-ATP-γ-S) Figure 4.17 PKCθAS2-M385V produces a similar banding pattern with different T-ALL cell line lysates In vitro kinase assay. FLAG tagged PKCθAS2-M385V was transiently overexpressed in 293T cells followed by FLAG immunoprecipitation and an in vitro kinase assay using Bn-ATPγS and T-ALL cell line whole cell lysates (with or without prior phosphatase treatment (λPP)) or BSA as a substrate. Equal mass of protein substrate was used for each reaction. Probed with antibodies directed against FLAG (PKCθ), thiophosphate ester and β-actin (loading control). 129 600 450 300 150mM NaClActinThiophosphate esterFLAG/PKCθInvitrokinaseassay(Bn-ATP-γ-S) Figure 4.18 Increasing salt concentration decreases PKCθAS2-M385V phosphotransfer In vitro kinase assay. FLAG tagged PKCθAS2-M385V was transiently overexpressed in 293T cells followed by FLAG immunoprecipitation and an in vitro kinase assay using Bn-ATPγS and HPBALL whole cell lysate as a substrate. Increasing concentrations of NaCl were used for each for each reaction. Probed with antibodies directed against FLAG (PKCθ), thiophosphate ester and β-actin (loading control). Values indicate intensity of FLAG/PKCθ bands normalized to β-actin and relative to 1uL PKCθWT input. 4.3.5 Mass Spectrometry Identification - PKCθAS Four separate in vitro kinase/mass spectrometry experiments were performed in order to identify direct PKCθ phosphotargets, all comparing HPBALL lysates with and without exogenous PKCθAS2-M385V. Variations between the experiments included running without dimethyl quantitative labelling (Experiment 1), and including an additional high 130 salt condition in the hope of revealing high affinity targets (Experiments 3 and 4). From each experiment, hundreds of individual protein species were identified (Exp1 = 2426, Exp2 = 1449, Exp3 = 1410, Exp4 = 866). To be considered a potential PKCθ direct phosphotarget, proteins had to be unique to PKCθAS2-M385V samples (regardless of salt condition) or be > 10-fold enriched over control (no kinase added). In each experiment, hundreds of individual protein species matched this criterion (Exp1 = 562, Exp2 = 926, Exp3 = 555, Exp4 = 599), and a comparison between these four protein lists revealed an overlap of 34 potential PKCθ direct phosphosubstrates that I considered to be the list of top candidates (Figure 4.19, Table 4.5). In addition, a number of protein species were identified in 3out of 4 separate experiments (Table C.1/2/3/4), suggesting them to also be potential PKCθ phosphotargets. No pattern with regard to the salt concentrations used in the in vitro kinase assays for Exp3 and Exp4 was seen. This is highlighted by the observation that in these experiments proteins from the list of candidates that were identified in both salt conditions were not consistently enriched in either condition (Figure 4.20). I do not consider that increasing the salt concentration in this assay was enriching for certain protein species and as such, both salt conditions were considered as one for these experiments. Given that PKCθ, along with other novel PKC family members, is described to localize to membranes and bind DAG for full activation it is interesting to note that 21 of the 34 candidate PKCθ direct phosphosubstrates are associated with membranes (GO:0016020) and more specifically, 9 of these are associated with the plasma membrane (GO:0005886). Further analysis of the list shows identification of multiple proteins from the same complex, namely mTORC2 (mTOR and RICTOR) and the nuclear pore complex (NUP133 and NUP160). None of the 34 candidate PKCθ phosphotargets identified here correspond to previously described PKCθ substrates. To gain higher confidence that these candidate PKCθ phosphotargets identified in vitro may actually be occurring in a T-ALL cell context I also aimed to identify them based on their proximity to PKCθ in live cells. 131 Figure 4.19 34 PKCθ direct phosphotarget candidates identified Venn diagram of potential PKCθ direct phosphotargets identified from four independent PKCθAS2-M385V in vitro kinase assay/thiophosphate ester immunoprecipitation experiments (Exp#1-4). 34 candidate PKCθ direct phosphotargets found in all four experiments, outlined in bold. 132 Table 4.5 Candidate direct PKCθ phosphotargets 34 candidate direct PKCθ phosphotargets found common between four independent PKCθAS2-M385V in vitro kinase assay/thiophosphate ester immunoprecipitation experiments (Exp#1-4). UniProtID Gene name Gene DescriptionQ9P2R3 ANKFY1 Rabankyrin-5P10398 ARAF Serine/threonine-protein kinase A-RafQ9Y6D6 ARFGEF1 Brefeldin A-inhibited guanine nucleotide-exchange protein 1Q8N3C0 ASCC3 Activating signal cointegrator 1 complex subunit 3Q16543 CDC37 Hsp90 co-chaperone Cdc37Q5VV42 CDKAL1 Threonylcarbamoyladenosine tRNA methylthiotransferaseQ9H9E3 COG4 Conserved oligomeric Golgi complex subunit 4Q5JPH6 EARS2 Probable glutamate--tRNA ligase, mitochondrialP41250 GARS Glycine--tRNA ligaseO94927 HAUS5 HAUS augmin-like complex subunit 5Q9NSE4 IARS2 Isoleucine--tRNA ligase, mitochondrialQ9UI26 IPO11 Importin-11O14654 IRS4 Insulin receptor substrate 4Q15334 LLGL1 Lethal(2) giant larvae protein homolog 1Q9NU22 MDN1 MidasinP52701 MSH6 DNA mismatch repair protein Msh6P42345 MTOR Serine/threonine-protein kinase mTORQ8WUM0 NUP133 Nuclear pore complex protein Nup133Q12769 NUP160 Nuclear pore complex protein Nup160Q9NVE7 PANK4 Pantothenate kinase 4P42356 PI4KA Phosphatidylinositol 4-kinase alphaQ07864 POLE DNA polymerase epsilon catalytic subunit AP25788 PSMA3 Proteasome subunit alpha type-3P48651 PTDSS1 Phosphatidylserine synthase 1P04049 RAF1 RAF proto-oncogene serine/threonine-protein kinaseQ6R327 RICTOR Rapamycin-insensitive companion of mTORQ14684 RRP1B Ribosomal RNA processing protein 1 homolog BQ9Y3Z3 SAMHD1 Deoxynucleoside triphosphate triphosphohydrolase SAMHD1Q6P4A7 SFXN4 Sideroflexin-4Q9UJS0 SLC25A13 Calcium-binding mitochondrial carrier protein Aralar2Q15208 STK38 Serine/threonine-protein kinase 38Q9BTW9 TBCD Tubulin-specific chaperone DQ13049 TRIM32 E3 ubiquitin-protein ligase TRIM32Q9BQA1 WDR77 Methylosome protein 50 133 Figure 4.20 No pattern of enrichment in protein species with increased salt concentration Fold enrichment of candidate PKCθ direct phosphotargets identified in both low and high salt concentrations from Experiments #3 and #4. Joining line indicated the protein species was found to be differentially enriched between low and high salts in both experiments. 4.3.6 BioID cloning and expression A modified (R118G) bacterial biotin ligase (BirA*) was cloned, along with MYC and HA tags, onto both the N and C termini on human PKCθ, respectively, (Figure 4.3A) to identify PKCθ proximal proteins. Initially three versions of each construct were cloned with either no linker, a short linker, or a long linker separating the PKCθ and the BirA* in the hopes of indentifying a version that would be the most stably expressed. Transient transfection of these constructs into 293T cells showed multiple bands laddering down from a major band at the predicted molecular weight of the fusion product (~115kDa) (Figure 4.21). The lower molecular weight bands had a slightly increased intensity with the N terminal BirA* fusion when a long linker was employed. It was thus decided to proceed with the no linker fusion constructs, where no lower molecular weight laddering was observed when they were lentivirally transduced into the T-ALL cell line HPBALL E x p # 3 E x p # 4-3-2-10123C a n d id a te P K C θ p h o s p h o ta rg e ts e n r ic h m e n t f ro m" lo w " v s " h ig h " s a lt c o n d it io n sLow salt/high salt(Log2 fold enrichment)134 (Figure 4.22). I hypothesize that the much lower expression levels achieved through transduction of HPBALL cells as opposed to transient transfection of 293T cells may explain the loss of any low molecular weight bands. For both the N terminal and C terminal fusion products, expression levels achieved with lentiviral transduction of HPBALL cells were similar (1.21 and 0.91 fold) relative to endogenous PKCθ levels. As excessive expression of the fusion products could theoretically lead to mislocalization, low levels of expression of the transgenes should limit artifactual interactions. EmptyLong linkerShort linkerNo linkerpcDNA3-Myc-BirA*/PKCθpcDNA3-PKCθ/BirA*-HAPKCθ/BirA* fusion ~115kDaPKCθMYC HAPKCθ/BirA* fusion ~115kDaLong linkerShort linkerNo linker Figure 4.21 293T overexpression of PKCθ/BirA* fusion constructs shows some unintended product Western Blot. Transient overexpression of pcDNA3 based myc-BirA*/PKCθ and PKCθ/BirA*-HA fusion construct variants with different linker lengths. Detected with antibodies directed against PKCθ, myc and HA. 135 EmptyMyc-BirA*-PKCθ EmptyPKCθ-BirA*-HA EmptyMyc-BirA*-PKCθPKCθ -BirA*-HAPKCθMYC HAPKCθ/BirA* fusion ~115kDaPKCθ1.001.211.000.91 Figure 4.22 HPBALL expression of PKCθ/BirA* fusion constructs shows specific predicted band with expression equivalent to endogenous PKCθ Western Blot. Stable overexpression of lentiviral based myc-BirA*/PKCθ and PKCθ/BirA*-HA fusion construct variants short linker lengths. Probed with antibodies directed against PKCθ, myc and HA. Values indicate intensity of PKCθ/BirA* fusion bands relative to endogenous PKCθ. 4.3.7 BioID biotin labelling Labelling of proximal proteins using the BirA* system is initiated through the addition of biotin to the cell culture media. When testing the system in the transiently transfected 293T system the appearance of novel bands after the addition of 50 µM biotin was seen (Figure 4.23). Initially, this assay could not be directly translated to the HPBALL lentiviral transduction system because, in contrast to 293T which are cultured in media lacking biotin (DMEM), the base media used to grow HPBALL cells (RPMI 1640) contains biotin. The amount of biotin in RPMI 1640 (820 nM) was sufficient to cause a background level of biotinylation of protein species by the PKCθ/BirA* fusion construct, which was further increased with the addition of more biotin to the media. To overcome this problem and reduce the noise caused by the RPMI 1640 media derived biotin PKCθ/BirA* fusion expressing HPBALL cells were grown for 6 days in other base media that contain lower levels of biotin; DMEM (0 nM biotin), IMDM (53 nM biotin) and DMEM/F12 (14 nM biotin) prior to initiating biotinylation. In addition, as FBS has been reported to contain up to ~200 nM biotin, serum levels were reduced from the usual 136 10% to 5%, resulting in effective biotin concentrations for RPMI, DMEM, IMDM, DMEM/F12 of 789 nM, 10 nM, 60 nM, and 23 nM, respectively. The degree of background biotinylation seen with DMEM, IMDM, DMEM/F12 based media looked to be significantly reduced compared to RPMI 1640, and possibly only marginally above that of HPBALL transfected with an empty vector (Figure 4.24). However, cellular growth in each of these media was altered, with DMEM and DMEM/F12 showing the most significant growth reduction. IMDM, although causing cells to grow in clumped aggregates, was able to maintain a proliferating culture in the short time periods tested (data not shown). As PKCθ signalling is a dynamic event involving re-localization and degradation of signalling components, I reduced the 24 hour biotin labelling time that was initially used for optimization. Signal intensity appears to drop off significantly when supplemental biotin is added for less than 24 hours with both N and C terminal BirA* PKCθ fusion constructs, showing little significant signal intensity increase at 6 hours compared to base line (0 hours) (Figure 4.25). In addition, the mechanism by which BirA* labelling occurs is thought to be through an adenylate ester, which has a half-life of minutes. This longer half-life, relative to APEX labelling (<1 millisecond), would theoretically give more time for the activated label to travel further from the fusion protein and potentially allow for a larger labelling radius, and as such, identification of proteins that are not proximal to PKCθ. Because of this potentially large labelling radius, slow labelling kinetics, and low signal intensity over background caused by the utilization of IMDM and serum derived biotin I looked to improve the proximity labelling method by switching to APEX fusion PKCθ proteins. 137 EmptyLong linkerShort linkerNo linkerBiotin- + - + - + - +SA-HRPpcDNA3-Myc-BirA*/PKCθ Figure 4.23 Biotin labelling by PKCθ/BirA* fusion constructs in 293T Western Blot. Transient overexpression of pcDNA3 based myc-BirA*/PKCθ and fusion construct variants with different linker lengths. Labelled with biotin for 24 hours. Biotin detected with SA-HRP. 138 Myc-BirA*-PKCθ- + - + - + - +RPMI DMEM IMDMDMEM/F12EmptySA-HRPBiotin Figure 4.24 Biotin from media utilized by PKCθ/BirA* fusion constructs in HPBALL Western Blot. Stable overexpression of lentiviral based myc-BirA*/PKCθ fusion (no linker) construct in HPBALL cells grown in different base media. Labelled with exogenous biotin for 24 hours. Biotin detected with SA-HRP. 139 myc-BirA*-PKCθPKCθ-BirA*-HAEmtpy PKCθ/BirA* fusionTime Biotin (hr)0 24 0 0.5 1 2 4 6 24SA-HRP Figure 4.25 Exogenous biotin labelling by PKCθ/BirA* fusion constructs in HPBALL has slow detectable biotin labelling kinetics Western Blot. Stable overexpression of lentiviral based myc-BirA*/PKCθ and PKCθ/BirA*-HA fusion (no linker) constructs in HPBALL cells grown in IMDM base media. Labelled with biotin for 0-24 hours. Biotin detected with SA-HRP. 4.3.8 APEX cloning and expression A modified soybean ascorbate peroxidase (APEX2) was cloned with a short linker onto both the N and C termini on human PKCθ (Figure 4.3B). When lentivirally transduced into the T-ALL cell line HPBALL, both N and C termini APEX fusion proteins were expressed at lower levels than endogenous PKCθ (0.15 and 0.54 fold) (Figure 4.26A). 140 As the C terminal APEX fusion showed a higher expression level, this construct was selected for all future experiments. An additional rationale in favor of fusion of APEX to the C terminus of PKCθ is the previously reported continued immunological synapse localization after TCR engagement of PKCθ C terminal GFP fusions [431]. This suggests that the C terminal positioning of a ~27 kDa protein (GFP = 27 kDa, APEX = 28 kDa) does not majorly affect normal protein function and interactions due to steric hindrance. The difference in expression between the fusion proteins was also reflected in the degree of biotin phenol (BP) labelling they could invoke. The higher expressed C terminal APEX fusion protein yielded stronger intensity, as well as novel banding relative to the lower expressed N terminal APEX fusion (Figure 4.26B). PKCθPKCθ/APEXEmpty APEX/PKCθ1.000.151.000.54SA-HRPPKCθ/APEXEmpty APEX/PKCθ Figure 4.26 (A) (B) 141 Figure 4.26 HPBALL expression of PKCθ:APEX fusion constructs expressed less than endogenous PKCθ and show novel banding with biotin phenol labelling Western Blot. (A) Stable overexpression of lentiviral based APEX/PKCθ and PKCθ/APEX fusion construct variants with short linker lengths in HPBALL. Probed with antibody directed against PKCθ. Values indicate intensity of PKCθ/APEX or APEX/PKCθ fusion bands relative to endogenous PKCθ. (B) Labelling with BP/H2O2. Biotin detected with SA-HRP. Triangles indicated examples of novel/stronger intensity banding with PKCθ/APEX. 4.3.9 APEX biotin phenol labelling To initiate biotin phenol labelling of proximal proteins by APEX, hydrogen peroxide is added to initiate the catalysis of a phenol group to phenoxyl radicals. These phenoxyl radicals are reported to have an extremely short half-life (<1 millisecond [407]) and have a small labelling radius (<20 nm [408]). As expected, initiation of labelling with biotin phenol appeared to be entirely dependent upon both expression of the PKCθ APEX fusion protein and addition of hydrogen peroxide (Figure 4.27). As such, any background signal observed was due to endogenously biotinylated proteins/non-specific signal rather than non-specific or premature utilization of the substrate as with the BirA* system. Due to the possibility of a short labelling time (1 minute), I explored the dynamic changes in proximal protein partners in a steady state versus a ligand activated state. PKC activators such as phorbol 12-myristate 13-acetate (PMA) mimic DAG to activate PKCs, as well as other DAG binding proteins (chimaerins [432], MUNC13 [433] and RasGRPs [434]). Although non-specific for PKCθ, the fusion with APEX ensures only PKCθ proximal changes should be observed. PMA can cause membrane localization of PKCθ in under 2 minutes and can persist at the membrane for over 15 minutes [374, 435, 436]. To determine the most appropriate time to activate PKCθ before initiating labelling, the banding pattern/intensity in steady state versus 5 minutes or 30 minutes of PMA activation (1 minute labelling initiated at 4 and 29 minutes post PMA addition, respectively) was tested. Stronger banding intensity was observed when PMA was activated for 5 minutes as opposed to the extended 30 minutes (Figure 4.27). This may represent the fact that despite PKCθ being retained at the membrane, interacting 142 partners, including phosphotargets, may rapidly transit away from the membrane following activation to fulfill their biological function. SA-HRPPKCθActinPKCθ/APEXEmpty- - + + + + Biotin PhenolH2O2PMA (min)- - - + + +- - - - (5) (30) Figure 4.27 HPBALL PKCθ/APEX biotin phenol labelling shows additional novel banding when PKCθ is activated Western Blot. Stable lentiviral overexpression of PKCθ/APEX fusion construct in HPBALL cells. Cells cultured with or without biotin phenol and labelling initiated with addition of with H2O2 with or without prior PMA activation (5 minutes or 30 minutes). Probed with antibodies directed against PKCθ and β-actin. Biotin detected with SA-HRP. Triangles indicated examples of stronger intensity banding with 5 minutes of PMA treatment. 143 4.3.10 APEX biotin Phenol IP In order to identify proteins proximal to PKCθ, proteins labelled with biotin phenol are first pulled down with a streptavidin resin. Despite numerous washes of the cells with PBS after the labelling event, there may still be free-biotin phenol in the resulting lysate owing to its poor membrane permeability that saturated the SA resin. For this reason, I believe the initial pull-down attempt failed, as the same intensity of bands in our unbound fraction was observed (Figure 4.28). To remove this free biotin phenol prior to streptavidin pull down and prevent competition with the labelled protein for streptavidin binding sites, free biotin phenol was removed with PD-10 size exclusion columns prior to streptavidin pull down. After this removal of free biotin phenol, the streptavidin pull-down fraction showed an increased SA-HRP signal, as well as reduced intensity in the unbound fraction (Figure 4.29) indicating an improvement in pull-down efficiency. Pre-Pulldown (1.25/500)unbound (1.25/500)SA-Pulldown (2/40)+ + + + + + - + +- - + - - + - + +- + + - + + - + +SA-HRPActinBPH2O2PMA Figure 4.28 144 Figure 4.28 Without removal of free biotin, phenol labelled proteins are unable to be isolated with streptavidin resin SA pull-down and blot. Stable lentiviral overexpression of PKCθ/APEX fusion construct in HPBALL cells. Cells were cultured in biotin phenol containing media for 30 minutes before initiation of labelling with H2O2 with or without prior PMA activation. Lysates were used for SA-pull-down. Pre-pull-down, unbound and SA-pull-down samples probed with antibody directed against β-actin and biotin detected with SA-HRP. preIP (5/1000)unbound (5/1000)SA-IP (1/50)+ + + + + + + + +- - + - - + - - +- + + - + + - + +BPH2O2PMASA-HRPActin Figure 4.29 Removal of free biotin phenol allows labelled protein to be isolated with streptavidin resin SA pull-down and blot. Stable lentiviral overexpression of PKCθ/APEX fusion construct in HPBALL cells. Cells were cultured in biotin phenol containing media for 30 minutes before initiation of labelling with H2O2 with or without prior PMA activation. Free biotin was removed from lysates using PD-10 desalting columns followed by SA-pull-down. Pre-pull-down, unbound and SA-pull-down samples probed with antibody directed against β-actin and biotin detected with SA-HRP. 145 4.3.11 Mass spectrometry identification - PKCθAPEX with biotin phenol In order to identify PKCθ proximal proteins, mass spectrometry was employed. Streptavidin pull-down was performed on lysates from PKCθAPEX expressing HPBALL cells cultured in the presence of biotin phenol and subjected to three conditions: (A) PMA un-activated/unlabelled, (B) PMA un-activated/labelled, and (C) PMA activated/labelled. For each condition, hundreds of protein species were identified (A) 613, (B) 495 and (C) 771. This initial run did not utilize any dimethyl labelling. Identified proteins were classified into those found to be enriched in a PMA “activated” state, enriched in PMA “un-activated” state or found in “both” the PMA un-active and active states equally, based on the following criteria (Table 4.6). Table 4.6 Criteria for inclusion of protein species into PMA activation states A = expression of a protein species in PMA un-activated/unlabelled condition, B = expression of a protein species in the PMA un-activated/labelled condition, C = expression of a protein species in the PMA activated/labelled condition. “Activated” state criteria “Un-activated” state criteria “Both” activated and un-activated states criteria C ≥ 4A B ≥ 4A B ≥ 4A or C ≥ 4A C ≥ 2B B ≥ 2C 0.5C ≤ B ≤ 2C or 0.5B ≤ C ≤ 2B The number of unique protein species in each state were, “Un-activated” = 73, “Activated” = 342, “Both” = 120. (Table D.1/2/3). While Reactome analysis showed no significant overrepresentation of components of any particular pathway/process in the “un-activated” state, a number of pathways/processes, mainly relating to cell cycle, are overrepresented in the “activated” and “both” states. With regard to TCR signalling (GO: 0050852(mammalian)), where the canonical function of PKCθ is described, one protein (PELO) in the “un-activated” state was seen to be involved in TCR signalling. In the “activated” state, where PKCθ should be brought to the plasma membrane, eleven TCR signalling components in the list of identified protein species from this state (CD28, 146 CD3D, CSDE1, EZR, FYB, PIK3R1, PLCG1, PRKD2, PTPN2, TRAT1, ZAP70) and three in the “both” state (FYN, IPP5D, LCK) were found. When comparing previously described PKCθ interactors to the protein lists no matches with the “un-activated” state proteins were seen but “activated” state (CBL, FYB, MARK3, VAV1) and “both” state (FYN, LCK) proteins were observed. Additionally in the “activated” state, two protein targets, CCDC88A and EZR, also corresponded to known PKCθ phosphotargets. In addition to detecting some previously described interactors as well as canonical pathway components; many additional proteins were identified by this experiment. While many of these may be novel proximal partners of PKCθ, the high background level of endogenously biotinylated protein species may also be a confounding factor. Although the degree of endogenous biotin signal in each sample should be theoretically equal and therefore not adversely skew the final result, it still represents the majority of the overall signal, making the less abundant species harder to detect. In addition, it would make difficult or impossible to identify true PKCθ proximal proteins that are endogenously biotinylated. 4.3.12 APEX alkyne phenol labelling In an attempt to reduce background noise in the APEX labelling system, I attempted to remove its dependency on biotin-streptavidin detection by replacing the biotin of my phenol based substrate with a different molecular handle. . I adapted my system to click chemistry, more specifically azide/alkyne cycloaddition. This reaction uses copper as a catalyst to covalently link an azide group containing molecule to an alkyne group containing molecule through the formation of a triazole product. This type of reaction has been used extensively for labelling biomolecules owing to its extremely specific and biorthogonal nature [437]. I made the detection/capture reagent an azide and the phenol based substrate an alkyne. The opposite combination is not possible because an azide would quench the singlet oxygen created by the APEX catalysis of the phenol group and prevent labelling. In addition, the covalent linkage resulting from the alkyne labelled protein and azide resin allows for extremely stringent wash conditions, further reducing non-specific protein interactors. Initial experiments were attempted with the only readily available commercial alkyne phenol, 3-Ethynylphenol (3EP) detected with biotin azide/streptavidin-HRP (Figure 4.30A). Although the resulting blots from such 147 experiments show the endogenously biotinylated proteins also seen when biotin phenol is used as a substrate, these should not be pulled down with the azide capture resin. Testing of 3EP in 293T cells transiently transfected with an APEX overexpression construct showed its potential utilization as a substrate by APEX, albeit with reduced efficiency compared to biotin phenol (Figure 4.30B). When this was translated to the HPBALL lentiviral PKCθAPEX fusion expression system the utilization of 3EP was greatly reduced compared to biotin phenol, so much so that it was not obvious above the background detection of endogenously biotinylated proteins (Figure 4.30C). One potential reason for the reduced utilization of 3EP as a substrate for APEX relative to biotin phenol may be the position of the hydroxyl group. As the name suggests, 3EP has the hydroxyl group at third carbon relative to the ethynyl chain, a meta-position. Conversely, biotin phenol has the hydroxyl group at fourth carbon, a para-position. Of potential substrates previously tested, subtle variations upon the biotin phenol structure prevented their utilization by APEX in a cell based system [406]. In order to provide a substrate more similar in structure to biotin phenol (Figure 4.31A), synthesizing an alkyne phenol compound with the hydroxyl group in the para-position was investigated. A structure similar to that of biotin phenol was decided on and the same synthesis method described by Rhee et al [406] was used to create HexT (Figure 4.31B). This compound was seen to be utilized extremely well in both the transient 293T system as well as the transduced HPBALL system by APEX and PKCθAPEX, respectively (Figure 4.31C/D). Band intensity was even greater than that observed with biotin phenol, which suggested good labelling of proximal proteins when HexT is used as a substrate should be possible. Unlike biotin phenol, there was no banding pattern difference seen with PMA activation (Figure 4.32). Despite this difference HexT was chosen to move forward with for sample preparation for mass spectrometry analysis. 148 3-Ethynylphenol (3EP) 293TAPEXtransfected- BP 3EPBioAzide/SA-HRPActinHPBALLPKCθ/APEXtransducedBioAzide/SA-HRPActin- BP 3EP Figure 4.30 3EP is utilized less efficiently than biotin phenol by APEX Western Blot. (A) Chemical structure of 3EP. (B) Transient overexpression in 293T or (C) stable lentiviral overexpression in HPBALL cells of PKCθ/APEX fusion construct. Cells were cultured in media containing biotin phenol or 3EP before initiation of labelling with H2O2. Lysates from 3EP labelled cells were subjected to click chemistry reaction with biotin-azide before being run on a gel. Probed with antibody directed against β-actin and biotin detected with SA-HRP. (B) (C) (A) 149 Biotin Phenol (BP) Hexyne-Tyramide (HexT) 293TAPEXtransfectedHPBALLPKCθ/APEXtransduced- BP 3EP HexT - BP 3EP HexTBioAzide/SA-HRPActinBioAzide/SA-HRPActin Figure 4.31 (C) (B) (D) (A) 150 Figure 4.31 HexT is utilized more efficiently than biotin phenol by APEX (A) Chemical structure of biotin phenol (BP) and (B) HexT. Western Blot. (C) Transient overexpression in 293T, or (D) stable lentiviral overexpression in HPBALL, of the PKCθ/APEX fusion construct. Cells were cultured in media containing biotin phenol, 3EP, or HexT before initiation of labelling with H2O2. Lysates from 3EP and HexT labelled cells were subjected to click chemistry reaction with biotin-azide before being run on a gel. Probed with antibody directed against β-actin and biotin detected with SA-HRP. SA-HRPActin- + + +- - - + - - + +HexTH2O2PMAPKCθ/APEX transduced Figure 4.32 HexT labelling does not produce observably different banding with PMA activation Western Blot. HPBALL cells stably transduced with the PKCθ/APEX fusion construct were cultured in media containing HexT before PMA activation and initiation of labelling with H2O2. Lysates from HexT labelled cells were subjected to click chemistry reaction with biotin-azide before being run on a gel. Probed with antibody directed against β-actin and biotin detected with SA-HRP. 151 4.3.13 Mass spectrometry identification - PKCθAPEX with HexT In order to generate samples for mass spectrometric identification of PKCθ proximal proteins using HexT, three separate cultures of PKCθAPEX transduced HPBALL cells were prepared from three separate transductions. These biological triplicate cultures showed expression of the PKCθAPEX fusion product, albeit at different endogenous/exogenous ratios (Figure 4.33A). For each culture, three conditions of APEX labelling were set up in the same fashion as the initial biotin phenol experiment but substituted with HexT; (A) “PMA un-activated/unlabelled”, (B) “PMA un-activated/labelled”, and (C) “PMA activated/labelled”. From these samples, lysates were generated, alkyne labelled proteins enriched and the resulting peptides submitted for mass spectrometry identification. From each of the biological triplicate experiments, hundreds of total individual protein species were identified (Exp1 = 1848, Exp2 = 1916, Exp3 = 1837) with 1356-1676 for each condition (A, B, or C). The same criteria as with the biotin phenol labelling mass spectrometry experiment (Table 4.6) was employed to separate the identified proteins into “activated”, “un-activated” and “both” states. Very little overlap between the three experiments was found and only 2/317, 3/349 and 1/128 of proteins were found to be common between the triplicates for the “activated”, “un-activated” and “both” states, respectively. Even when disregarding PMA activation states, only 26/716 were enriched ≥4-fold in all three independent experiments (Figure 4.33B), none of which were found to correspond to TCR signalling components or known PKCθ interactors. When I relaxed the criteria further to targets enriched in at least two experiments (143/716) only two TCR signalling components (LCK, PHPT1) and two known PKCθ interactors (LCK, YWHAG) were revealed. 152 ActinPKCθPKCθ endogenousPKCθ/APEX fusionParental #1 #2 #3PKCθ/APEX1.00 1.00 1.000.37 1.10 1.47 Figure 4.33 Number of identified proteins from triplicate PKCθ/APEX HexT experiments (A) Western Blot. Stable overexpression of lentiviral based PKCθ/APEX fusion construct in overexpression in HPBALL cells from three separate transduction events used for click pull down and mass spectrometry identification (#1, #2, #3). Probed with antibodies directed against PKCθ and β-actin. Values indicate intensity of PKCθ/APEX or APEX/PKCθ fusion bands relative to endogenous PKCθ. (B) Comparison of protein species identified from three PKCθ/APEX replicate experiments (Exp#1, Exp#2, Exp#3) uniquely found/enriched in HexT labelled relative to their respective unlabelled sample regardless of activation state (“activate”, “un-activated”, or “both” states). (B) (A) 153 4.3.14 Comparison of candidate PKCθ phosphotargets and identified proximal partners To gain higher confidence that the candidate PKCθ phosphotargets identified from the in vitro PKCθAS2-M385V screen may actually be occurring in a T-ALL cell context, I compared them to the protein species identified in the PKCθAPEX experiments. Unfortunately, none of the 26 protein species identified in triplicate experiments or (143 protein species identified in 2/3 experiments) from the HexT labelling experiments corresponded with the candidate PKCθ phosphotargets or previously described PKCθ phosphotargets. However, when comparing with those proteins identified from the biotin phenol labelling experiment, I see 7/34 of the candidate PKCθ phosphotargets identified from the PKCθAS2-M385V screen: MTOR in the “un-activated” state, HAUS5, PANK4, LLGL1, CDC37, ARAF in the “activated” state and PI4KA in “both” the activated and un-activated states (Figure 4.34). From this data, I conclude that APEX HexT labelling may not reflect true proximal PKCθ partners, and that candidate PKCθ phosphotargets also identified by APEX biotin phenol labelled experiments may be of interest for future investigation. 154 MTORHAUS5PANK4LLGL1CDC37ARAFPI4KA2772 337 119Candidate direct PKCθAS phosphotargets“Un-activated” “Activated” “Both”PKCθ interactors (biotin phenol) Figure 4.34 Proteins identified as both candidate direct PKCθ phosphotargets and PKCθ/APEX (biotin phenol) interactors Comparison of protein species identified as candidate direct PKCθ phosphotargets (red) with those identified from PKCθ/APEX interactor screen with biotin phenol used as a substrate. Interactors subdivided into those identified in “un-activated” (blue), “activated” (yellow) and “both” activated and un-activated states (green). 4.4 Discussion. 4.4.1 Identified candidate direct PKCθ phosphotargets 34 candidate direct PKCθ phosphotarget protein species were identified from the PKCθAS screens, none of them previously described as PKCθ direct phosphotargets. This may be due to the cellular context used, where some of these previously described PKCθ direct phosphotargets proteins may be unexpressed or may already the 155 phosphorylated at the specific PKCθ phosphosites, preventing further phosphorylation and identification. A number of these my candidates, however, may be interesting in a T-ALL context based on their described function in the literature (IRS4, mTOR, RICTOR, RAF1, ARAF, CDC37, LLGL1, MSH6). 4.4.1.1 Candidate direct PKCθ phosphotargets - IRS4 IRS4 is a member of the insulin receptor substrate family, which mediate much of the signalling events downstream of IR and IGF1R. Numerous serine/threonine phosphorylation sites have been described for IRS family members that have been described to both enhance and repress downstream signalling [237]. More specifically, various PKC family members have been described to phosphorylate IRS proteins [438], including PKCθ [236], with most studies showing a repressive effect on downstream pathway activation. Multiple mechanisms of action have been proposed to account for the repressive phosphorylation of IRS proteins, including mediating tyrosine dephosphorylation, receptor dissociation and degradation. My own data shows that activation of PKCθ can result in impaired IGF-1 induced downstream PI3K/AKT signalling (Fig. E.1A/1B/1C), possibly as a consequence of reduced IRS1 levels in T-ALL (Fig. E.2A/2B). Due to its less ubiquitous expression, IRS4 is less well characterized than other family members, however, it has been previously implicated in IGF1R mitogenic signalling [439]. In those cells that express high levels of IRS4, its knockdown has been shown to drastically negatively affect PI3K/AKT signalling, even in the presence of other IRS family members [440]. Additionally, IRS4 has been shown to be phosphorylated by PKCζ, leading to reduced PI3K activation [441]. In the context of T-ALL, I have shown here (see Chapter 3) and previously [92] that IGF1R signalling plays important leukemia promoting roles. Moreover, IRS4 has been shown to be overexpressed in T-ALL as a result of rare chromosomal translocations [247, 248]. 4.4.1.2 Candidate direct PKCθ phosphotargets - mTOR, RICTOR mTOR is the kinase catalytic component of the mTORC1 and mTORC2 signalling complexes. mTORC1 is activated indirectly downstream of AKT and in response to growth factors such as IGF-1. mTORC1 regulates protein translation by phosphorylating S6K1 and 4EBP1. The mTORC2 complex also contains another PKCθ candidate 156 phosphotarget, RICTOR. mTORC2 functions upstream of AKT and directly causes its activation by phosphorylation of S473. In addition, other AGC family kinases, including PKC family members, have also been described to be regulated through direct phosphorylation by mTORC2 [442]. PKCθ has been suggested as a direct target of mTORC2, whereby conditional deletion of rictor reduces PKC hydrophobic motif phosphorylation following TCR activation [443]. It is tempting to speculate that this interaction might also result in reciprocal signalling back to the mTORC2 complex by phosphorylation of mTOR and/or RICTOR. RICTOR has been previously described to be phosphorylated at multiple serine/threonine sites. In particular, T1135 has been shown to be phosphorylated by multiple AGC family kinases, resulting in the disruption of the RICTOR/Cullin-1 E3 ligase complex [444]. In addition, when this site (T1135A) is mutated, an increase in pAKT can be observed [445, 446]. GSK3β has been shown to phosphorylate RICTOR at two sites, S1235 and T1695, which were described to disrupt AKT binding [447] and promote degradation [448], respectively. mTOR has also been reported to be phosphorylated at multiple serine/threonine sites described to increase interaction with other mTORC complex components and to promote downstream signalling [449-453]. The upstream kinases that phosphorylate some of these sites are known, including mTOR itself [449] and S6K1 [450, 452]. As with other vital components of the PI3K/AKT signalling pathway, inhibition of the mTORC complexes is of therapeutic interest in T-ALL [454]. 4.4.1.3 Candidate direct PKCθ phosphotargets - RAF1, ARAF RAF1 (also known as CRAF) and ARAF are two members of the RAF serine/threonine kinase family. They are thought to form homo or heterodimers with each other, or the third family member BRAF, and are canonically described to play an integral role in the MAPK (RAS/RAF/MEK/ERK) signalling pathway by acting as MAP3Ks to phosphorylate/activate MAP2Ks, such as MEK1/2. The recurrent mutational activation of this pathway in T-ALL [28] and leukemogenic role it may play [261] makes RAF1 and ARAF interesting candidate PKCθ phosphotargets. Owing to its earlier discovery, the role and regulation of RAF1 is better characterized than that of ARAF. A number of phosphorylation sites appear to be important for both enhancing activation and repressing RAF1 activity. Phosphorylation at S259 and S621 affects its binding of 14-3-157 3 proteins, and, along with phosphorylation of S43 by PKA [455], are thought to stabilize RAF1 in an inactive state [456]. Following upstream receptor engagement and active RAS-GTP binding, RAF1 is further phosphorylated at S338/9 to enhance activation [457]. Additionally, a number of other serine/threonine sites have been described to be phosphorylated by ERK1/2 to act in both a positive [458] and negative feedback fashion [459]. PKCs have also been implicated in the direct phosphorylation of RAF1 (at S497/9) [460], however, the functional importance of these sites is uncertain. More interesting still is the observation that PKCθ can interact with RAF1 to enhance PKCθ activity. Reciprocal phosphorylation, however, was not observed [461]. Much of how ARAF functions and is regulated is extrapolated from RAF1, based on sequence similarity, but unique phosphorylation sites have also been described [462]. Unlike their family member BRAF, activating mutations in RAF1 and ARAF appear to be extremely rare in cancer [463]. Interestingly, until recently [464], the only reported mutation of ARAF in the literature had been in the T-ALL cell line MOLT4 [465]. 4.4.1.4 Other candidate PKCθ phosphotargets of interest Other candidate PKCθ phosphotargets of interest based on their known function include CDC37, LLG1 and MSH6. CDC37 forms a chaperone complex with HSP90 to stabilize multiple kinases (including AKT [466], RAF1 [467] and PKC [468]), and is potently oncogenic [469]. Loss of LLG1 has been shown to increase HSC cycling [470], and MSH6 is a key component of the DNA mismatch repair machinery and is mutationally lost in many cancers, including acute leukemias [471]. In vitro kinase assay experiments on purified protein substrates will be required to confirm the validity of candidates of direct PKCθ phosphotargets as well as the specific site of phosphorylation. Additionally, the functional consequence of these phosphorylation events would also be of interest. Whether PKCθ promotes or represses the described functional role of these candidates may illuminate how PKCθ may drive potential oncogenic or tumor suppressive activities in T-ALL, as well as shed light on its action in other cellular contexts. Because of our previously published observation that PKCθ can increase ROS levels [304] I was interested to see if any of the 34 candidate PKCθ direct phosphotargets may suggest a mechanism for this phenotype. Unfortunately, none of the candidates have 158 been described to be involved in the regulation of ROS levels (GO: 0016209 antioxidant activity, GO: 0042554 superoxide anion generation). PKCθ (and other PKC isoforms) have been implicated in ROS generation previously, often through NADPH oxidases (NOX) [472-474]. However, PKCθ and other novel isoforms have also been implicated in mitochondrial generated ROS [304, 472, 475, 476], components of which are not encompassed by GO: 0016209. Some mitochondrial proteins in the list of candidates, but none with obvious functions in ROS generation, were observed. An additional phenotype I have recently attributed to PKCθ is the reduction of surface CXCR4 expression (Figure. F.1A/1B/1C), and signalling (Figure. F.2A/2B), although this does not translate into reduced cellular migration (Figure. F.3A/3B, Fig. F.4). PKC isoforms have been previously implicated in the internalization of CXCR4, although somewhat unspecifically [477, 478], and novel isoforms may directly phosphorylate its serine/threonine-rich C-terminal tail [479]. Despite this, CXCR4, or other components implicated in CXCR4 internalization [479, 480], were not observed in the list of candidate direct PKCθ phosphotargets. 4.4.2 Proximity of candidate direct PKCθ phosphotargets To further refine the candidate list of direct PKCθ phosphotargets, I attempted to define the protein species proximal to PKCθ in live T-ALL cells. The rationale for such an approach was that, in order to have the potential to be phosphorylated by PKCθ, a target must first reside in a close spatial area. In order to theoretically reduce non-specific identification of protein species and to be able to assess protein species proximal to PKCθ in both activated and un-activated states, I shifted focus from the initial BioID strategy to a relatively novel APEX strategy. Further refinement of this APEX strategy lead me to attempt to identify PKCθ proximal proteins using alkyne substrates and click chemistry. Although very strong labelling with this approach was observed, the variability of the identified protein species leads me to believe that the labelling may be too promiscuous and that the protocol may not be enriching for proteins proximal to PKCθ. Very little overlap between triplicate experiments within each activation state was seen, and the described functions of those candidates that were found in all three experiments regardless of activation state did not seem to be consistent with known functions or localizations of PKCθ. For example, very few 159 associated TCR components were observed with HexT relative to when biotin phenol was used as a substrate. For these reasons, I have more confidence in data obtained from the single mass spectrometry experiment using the PKCθAPEX strategy with biotin phenol as a substrate. In this experiment, 7/34 of the candidate PKCθ phosphotargets identified from the PKCθAS2-M385V screen, including the four functionally interesting candidates mentioned above (mTOR, ARAF, CDC37 and LLGL1) were observed. To narrow down further to true proximal interactors and to reduce inclusion of false positives, I intend to repeat this experiment in triplicate. In addition, modifications to the methodology might also allow for further improvements. 4.4.3 Method improvement Both analog sensitive kinase screening for direct phosphotargets and BioID/APEX proximity labelling are recently developed methods with numerous possible procedural variations available. It is important to consider such variations in the attempt to improve future results when using these methods. 4.4.3.1 PKCθAS method variations Alterations and additions to various steps throughout the AS kinase procedure can potentially improve identification of direct phosphotargets. For example an AS kinase assay performed in live cells has been described to identify potentially more physiologically relevant direct phosphotargets of AMPK in vivo [481]. However, because of my inability to strongly overexpress PKCθAS in a T-ALL cellular context, I am limited to in vitro PKCθAS kinase assays using exogenously produced PKCθAS. With this approach, alternatives that can be explored to improve target identification include alternative conditions for lysate or kinase preparation. For example, T-ALL lysates could be treated with a generic kinase inhibitor such as 5′-(4-fluorosulfonylbenzoyl)adenosine (FSBA) as described in the proKALIP method [482]. FSBA is an ATP analog that covalently binds the invariant lysine present in all kinases and irreversibly inhibits their activity. Treatment of lysates prior to addition of PKCθAS would help to ensure that endogenously expressed kinases are not utilizing ATP analogs. Although, with the exception of some autophosphorylation by the PKCθ itself, I did not observe any major banding from in vitro kinases assays using wild-type PKCθ with lysates as a substrate, I 160 cannot rule out that banding below the detection of the blot could be picked up by the mass spectrometry analysis. Prior phosphorylation of phosphosites would be another factor affecting deposition of thiophosphates onto target proteins. The proKALIP method also includes a phosphatase treatment step to remove pre-existing phosphorylation [482] . If lysates are de-phosphorylated prior to their use in in vitro kinases assays, it might be expected that the overall level of thiophosphate deposition by PKCθAS would be enhanced. However, in my hands, I saw less intense banding with prior λPP treatment of lysates, suggesting an unintended reduction in thiophosphate transfer after phosphatase treatment. The main variable between the proKALIP phosphatase treatment and my own was the type of phosphatase used. Because PKCθ is a serine/threonine kinase, I chose to use λPP because of its greater reactivity against phospho-serine/threonine residues. In contrast the proKALIP method was designed to assess tyrosine kinases (e.g. SYK) and uses an alkaline phosphatase for the dephosphorylation of phospho-tyrosine residues, although it can also dephosphorylate serine and threonine residues. Protein species immunoprecipitated along with PKCθAS2-M385V from 293T cells may also be a source for an unintended thiophosphorylated protein species. Although I do not see any major banding with PKCθAS2-M385V alone in the in vitro kinase assay or when BSA is added to normalize the total protein mass in the reaction, there is the possibility that thiophosphorylated PKCθAS2-M385V co-immunoprecipitated proteins may be below the detection of the blot but could be picked up by the mass spectrometry analysis. This scenario might be mitigated by optimizing the reaction conditions or through the use of recombinant PKCθAS2-M385V produced from another cell source (i.e. Sf9 insect cells using baculovirus) where purity can be higher and contaminating non-mammalian proteins could be filtered out. Another tool that can be employed in AS kinase screens are analog selective inhibitors. Many such inhibitors are based on the structure of the semi-promiscuous kinase inhibitor Pyrazolo[3,4-d]pyrimidine (PP1) [483]. PP1 is an ATP competitive inhibitor that mimics that adenine ring of ATP with its C3 p-tolyl group adjacent to the gatekeeper residue [484]. Adding a bulky side chain at the C3 position prevents binding to wild-type ATP binding pockets due to steric hindrance with the gatekeeper residue. However, AS1 and AS2 mutated ATP binding 161 pockets provide space for these bulky side chains and allows certain PP1 analogs to selectively inhibit AS kinases whilst preserving the activity of their wildtype counterparts [382, 383]. This strategy has been mainly used to probe the function of other specific kinases in intact cells, for example inhibition of ZAP70 kinase activity in ZAP70-AS2 expressing mice [485], but has also been used with lysates to assess direct phosphorylation of substrates [392]. Use of such analogs in the PKCθAS2-M385V in vitro kinase assays might be useful as another comparator in a screen of a lysate or a later downstream validation. I have already tested one of the more common analogs 1NA-PP1 in a preliminary experiment and seen increased inhibition of PKCθAS2-M385V compared to PKCθWT (Figure 4.35). Further assessment of other available analogs that might offer improved efficacy and less off-target effect may be warranted [486]. In addition to variations to in vitro kinase assay conditions, alternative phosphotarget enrichment strategies can be employed. There are two alternative strategies for identification of thiophosphorylated proteins [380, 487]. The first, which I undertook, is based on a more traditional immunoprecipitation/mass spectrometry (IP/MS) approach. Proteins containing thiophosphates are alkylated with PNBM, immunoprecipitated with the thiophosphate ester antibody [380], then trypsinized to be identified via mass spectrometry. Each protein can then potentially produce multiple tryptic peptides suitable for mass spectrometric identification. Alternatively, trypsinzation can be performed after the in vitro kinase assay, and thiopeptides can be pulled down with an iodoacetyl resin [487]. However, iodoacetyl beads will also bind thio-cysteine resides, and care must taken to specifically elute off only thiophosphate ester linked peptides before mass spectrometry identification [430]. The advantage of this second approach is that the serine/threonine that is thiophosphorylated can be directly identified as it will be contained within the specific peptide identified. However, this also leads to the significant limitation that only those thiophosphate tryptic peptides suitable for mass spectrometric identification will provide information. This is particularly important in the case of PKCs, which have a predicted target recognition motif of K/R-X-S/T-X-K/R [417, 488] (Figure 4.36). With many target sequences being flanked by lysine and/or arginine residues, the resulting short peptide fragments might prove difficult for identification purposes, and as such, the 162 protein as a whole would not be found to be a PKCθ phosphotarget. Examples of PKCθ substrates that have their described target serine/ threonine within tryptic fragments of <6 amino acids include RASGRP1, RASGRP3, STK39, EZR, WIPF. To limit this problem, alternative digestion enzymes could be employed to generate unique, identifiable peptide fragments, for example Glu-C, Lys-N, Lys-C, Asp-N, or chymotrypsin. Figure 4.35 PP1 analog can selectively inhibit PKCθAS2-M385V In vitro kinase assay. FLAG tagged PKCθWT and PKCθAS2-M385V transiently overexpressed in 293T cells followed by FLAG immunoprecipitation and in vitro kinase assay using ATPγS and BN-ATPγS, respectively, and MBP as a substrate. Increasing concentrations of 1-NA-PP1 inhibitor were used in each reaction. Probed with antibodies directed against thiophosphate ester. Values indicate intensity of MBP/thiophosphate ester bands normalized to their respective 0 µM 1-NA-PP1 control, n= 1. 0 1 2 3 4 5 6 7 8 9 1 00 .00 .51 .01 .52 .0u M 1 -N A -P P 1Density of thiophosphrylated MBP(normalized to 0uM 1-NA-PP1)P K C q W T /A T P -χ -SP K C θ A S 2 -M 3 8 5 V /B N -A T P -χ -S163 Figure 4.36 Human PKCθ substrate recognition sequence motif Human PKCθ substrate recognition sequence motif generated from 36 input sequences of described human PKCθ phospho-serine/threonine targets as identified by http://www.phosphosite.org. 4.4.3.2 PKCθAPEX method variations In an attempt to reduce enrichment of endogenously biotinylated proteins when biotin phenol was used as a substrate for PKCθAPEX, I switched to an alkyne-phenol (HexT)/click strategy. Although I was able to detect strong labelling of protein species and observed lower background in my blots I believe that the labelling may be too efficient and result in the enrichment of protein species beyond the proximity of the fusion protein. Because biotin phenol was used less efficiently as an APEX2 substrate this may result in less promiscuous labelling, suggesting that this approach may be a path worth investigating further. Non-specific protein pull-down can confound identification and discrimination of true candidates from background contaminating proteins. These contaminating species can come from both non-specific interactions with the bead matrix or from non-specific interactions with target proteins during enrichment. Using non-specific, harsh elution techniques or on bead digestion results in these contaminating proteins being enriched alongside true target proteins. Relatively recently, a technique called “Direct Detection of Biotin-containing Tags” (DiDBiT) has 164 been developed to mitigate these contamination issues [489]. DiDBiT uses tryptic digestion of lysates prior to enrichment of the resulting peptides and identification by mass spectrometry. This process was shown to result in the identification of more total protein species than conventional methods, with the majority being biotin modified [489]. Unlike the AS kinase method that enriches thiophosphorylated peptides, the use of digestion prior to enrichment with the APEX method would in theory still allow for identification of multiple peptides from an individual protein species. Knowing the specific region on which the identified protein is labelled and having it match to logical expectations might also add another layer of confidence to the results. For example, identification of intracellular peptides, as opposed to extracellular peptides, from transmembrane proteins might be expected for the cytoplasmically located PKCθ. An alternative to using biotin-phenol as a substrate may be the use of desthiobiotin phenol. Desthiobiotin is a biotin precursor that lacks sulfur. This change reduces its binding affinity for streptavidin from KD = ~10-15 M to KD = ~10-11 M [490], closer to that of monoclonal antibodies. Assuming comparable efficacy as an APEX substrate, the use of desthiobiotin phenol would allow for the biotin elution of labelled proteins from streptavidin beads. This in turn would allow for the specific enrichment of desthiobiotin labelled proteins over endogenously biotinylated proteins and additionally would reduce the elution of non-specifically bead bound protein species by the harsh conditions biotin elution entails. Although the PKCθAS and PKCθAPEX screens were performed as separate assays with certain procedural variations, it may be possible to combine them into one workflow. For example, use of desthiobiotin-phenol as a PKCθAPEX substrate may allow for the elution of biotin enriched proteins in their native form. This in turn would allow these proteins to be used as a substrate source for the PKCθAS2-M385V in vitro kinase assay. Reducing the substrate complexity to only proximal partners would hopefully reduce many of the potential non-specific protein species identified from each assay when done individually. 4.5 Conclusion Here, I present preliminary data from two screens that aimed to identify direct PKCθ phosphorylation targets in a T-ALL context. I describe a set of 34 candidate PKCθ 165 phosphorylation targets, some of which I also found to be proximal to PKCθ. The known function of some of these candidates in disease relevant pathways suggests possible mechanisms for how PKCθ activity may alter T-ALL phenotypic characteristics. Confirmation of such actions will require future validation and/or assay development. 166 Chapter 5 General conclusions 5.1 Summary of study and findings The NOTCH1 signalling pathway is an important oncogenic driver of T-ALL. Despite many studies on how NOTCH1 may direct this phenotypic program and the identification of important downstream effectors such as c-MYC and HES1, there is still much to elucidate from this complex pathway in a T-ALL context. In these studies, I have attempted to probe two signalling components recently shown to be NOTCH1 regulated by our group, IGF1R [92] and PKCθ [304]. In addition, I attempted to identify novel downstream effectors of NOTCH1 in T-ALL by characterising the miRNAs differentially expressed following NOTCH1 signalling perturbation. From these three studies, I reveal that the IGF1R signalling pathway has the potential to be positively regulated by NOTCH1 through multiple mechanisms. I also show that this pathway, and more specifically its signalling through the PI3K/AKT signalling arm, may be important for driving leukemic growth in some T-ALLs (Figure 5.1). An investigation, in which I took part, has previously reported that NOTCH1 may act to enhance IGF1R transcription through its binding of an intronic enhancer [92]. This idea was subsequently expanded upon by the suggestion that this may be through the regulation of a long non-coding RNA (LUNAR1) [254]. Here, I suggest that NOTCH1 can also affect IGF1R at the protein level through the indirect repression of the IGF1R targeting miRNA, miR-223. We have also previously shown that NOTCH1 can indirectly negatively regulate the expression of PKCθ via intermediate RUNX1/3 transcriptional regulation [304]. In a separate observation, I see that activation of PKCθ in a T-ALL context represses IGF-1 mediated PI3K/AKT signalling, possibly via the previously described mechanism of inhibition/downregulation of IRS1. Although I do not observe IRS1 as a direct PKCθ phosphotarget in my PKCθAS screen, IRS4, a family member previously implicated in T-ALL [247, 248], as well as other potential downstream effectors of IGF1R/PI3K/AKT signalling, mTOR and RICTOR were identified. Additionally, I identified members of the other canonical signalling arm downstream of IGF1R, the MAPK pathway, as candidate PKCθ phosphotargets, RAF1 and ARAF. 167 Along with IGF1R itself, I was able to identify mTOR and ARAF as being potentially spatially related to PKCθ in a preliminary PKCθAPEX biotin proximity labelling screen, suggesting that in addition to playing a role in TCR signalling, PKCθ may also be directly impacting the IGF1R signalling pathway in T-ALL. Additionally, my APEX screen showed the previously described direct PKCθ target CCDC88A [491] as being proximal to PKCθ when PMA activated and, although not one of the 34 candidate phosphotargets, it was identified in 3/4 of the PKCθAS mass spectrometry runs. PKCθ phosphorylation inhibits the guanine nucleotide exchange factor activity of CCDC88A, which in turn impairs insulin induced PI3K/AKT activation [491], suggesting another possible mechanism for my observed PKCθ mediated repression of IGF-1 mediated PI3K/AKT signalling. Validation, specific phospho-residue identification and functional consequence for the above mentioned potential candidates still remain to be performed. Another property we have previously ascribed to IGF1R is its importance in the maintenance of leukemia initiating cell activity, whereby reduced IGF1R expression can reduce this activity [92]. Interestingly, we have also ascribed this same property to PKCθ, whereby the LIC enriched population has lower PKCθ expression and, by increasing PKCθ activity leukemia initiating cell activity, is reduced [304]. It is tempting to speculate that these two observations may be related and that one mechanism by which T-ALL propagation is maintained is through the repression of PKCθ and resulting sustained IGF1R signalling. Other hypotheses for how PKCθ might abrogate leukemia initiating cell activity include through the upregulation of reactive oxygen species or the downregulation of CXCR4 surface expression, as published previously by our group and described in the appendix, respectively. I was unable to find any direct PKCθ phosphotargets that could account for these phenotypes, but this may reflect an indirect action or require further refinement of the PKCθAS and PKCθAPEX methodologies. 168 ?PI3KJAK1PTEN RasIRS1/2/4STAT5PKCθmiR-223NOTCH1Runx1Runx3PIP2PIP3AKTPDK1mTORC2mTORRICTORRAF1/ARAFIGF1RIL7R CXCR4T-ALL growth Figure 5.1 Summarized role of NOTCH1 regulation of the IGF1R signalling pathway in T-ALL Red lines indicate potential direct PKCθ phosphorylation as identified by my AS kinase screen. 5.2 Conclusions regarding the study hypotheses The overall aim of this thesis was to expand and characterize downstream effectors of oncogenic NOTCH1 signalling in T-ALL. In chapter 2, my aim was to investigate the miRNAs regulated downstream of NOTCH1 in T-ALL. I hypothesized that NOTCH1 might regulate a set of miRNAs that contribute to its oncogenic activity in this context. Unfortunately, in the initial screen no consistent pattern of differential miRNA expression was observed in mouse derived T-ALLs, and only one characterized miRNA, miR-223, was observed in human T-ALL cell lines. I was, however, able to characterize IGF1R as a target of miR-223 in T-ALL. Although the functionally relevant surface level of IGF1R alone remained unaltered, I believe that, in conjunction with other mechanisms, the 169 repression of miR-223 by NOTCH1 can enhance IGF1R expression to support T-ALL leukemic growth. In chapter 3, my aim was to investigate the effects of IGF1R inhibition and characterize important IGF1R downstream signalling required for growth in a T-ALL context. Based on the above mentioned and previously described regulation of IGF1R by NOTCH1, I hypothesized that the growth of some T-ALL cell lines would be negatively impacted by the inhibition of IGF1R signalling and that canonical downstream signalling arms may be responsible for its action. By testing a broad panel of T-ALL cell lines a subset was observed to be sensitive to clinically relevant IGF1R inhibition. I further showed that IGF1R can maintain this cell growth through the activation of PI3K/AKT signalling and that previously described activators of this pathway in T-ALL, IL-7 signalling and loss of PTEN, could not functionally substitute for IGF1R to recover growth. In chapter 4, my aim was to find direct PKCθ phosphorylation targets in a T-ALL context. I hypothesized that, in addition to identifying previously described direct PKCθ targets, I would find novel targets and hoped to mechanistically link some of these to phenotypes observed with PKCθ perturbation. 34 candidate direct PKCθ phosphorylation targets were able to be identified using an AS kinase screen, some of which were also able to be observed to have a spatial relationship with PKCθ in a T-cell cellular context using APEX proximity labelling. None of these candidates have been previously described to be directly phosphorylated by PKCθ, but interestingly, a number are described to potentially act in the IGF1R signalling pathway, which I observe to be repressed by the action of PKCθ in T-ALL cell lines. 5.3 Strengths and limitations of this study There were a number of strengths and limitations in various aspects of this study. Strengths include the characterization of miRNAs regulated downstream of NOTCH1 in human T-ALL, investigation of two clinically relevant IGF1R inhibitors in a broad panel of human T-ALL cell lines, and the use of a novel strategy to identify direct PKCθ phosphotargets. Overall, these three projects originally conceived to be separate from one another were able to converge along with previously published data from our lab, 170 allowing for a thorough examination of mechanisms by which oncogenic NOTCH may promote T-ALL leukemic growth via IGF1R signalling. In the AS kinase screen for direct PKCθ phosphotargets, difficulty in overexpressing PKCθAS and its reduced activity made adaption of previously described protocols particularly challenging. Ideally, this screen should be performed in intact live cells or failing that, AS kinase overexpressing lysates. However, due to the inability to successfully overexpress PKCθAS, I was forced to rely on recombinant PKCθAS. Additionally, in attempting to improve the APEX proximity labelling strategy by converting it to a click chemistry based detection system, I believe that I may have reduced its PKCθ specificity. For this reason a single previous PKCθAPEX experiment utilizing biotin based detection to draw conclusions was relied on. I hope to repeat this experiment as well as validate the candidates in the future using standard in vitro kinase assays. Moreover, this study could be strengthened by the use of primary/xenograft expanded patient samples. Due to the difficulty in achieving robust in vitro growth with such samples, as well as poor transduction efficiency, these experiments become particularly challenging to perform and interpret. 5.4 Future research directions In my investigation into the identification of direct PKCθ phosphotargets in T-ALL I have produced an analog specific variant of human PKCθ that retains acceptable activity, PKCθAS2-M385V. This kinase will be useful in our own labs further investigation into PKCθ, as well as being useful to the general field. Future research carrying on from work presented in this thesis will mainly focus on the validation and characterization of direct PKCθ phosphotargets. In order to confirm PKCθ dependent phosphosites I intend to use site-directed mutagenesis of phosphosites on the identified target proteins to prevent phosphotransfer. In addition, site-directed mutagenesis could also be employed to change these confirmed serine/threonine residues to glutamic acid or aspartic acid residues in the hope of being able to mimic phosphorylation and help interrogate the function of each site. 171 I have already been able to achieve impressive PKCθ knockdown in human T-ALL cell lines and have investigated its impact on IGF1R/IRS1/PI3K/AKT signalling as well as CXCR4 surface expression. I hope to employ this same system to further investigate the molecular actions downstream of PKCθ that may be mediated through direct phosphorylation of the candidate target protein I have identified. My results showing a reduction in IGF-1 induced signalling in response to PKCθ activation already hints at the possible involvement of a number of my candidate targets. I am interested in pursuing this angle further as well as investigating this action in a broader set of models, including xenograph expanded primary T-ALL patient samples and NOTCH1 driven T-ALL mouse leukemias. In addition to knockdown human cell line T-ALL models our lab has also generated murine T-ALL leukemia on a PKCθ knockout background that will provide a valuable resource on which to test my hypotheses. Beyond elucidating proteins acting directly downstream from PKCθ I have some interest in investigating upstream regulators. Much of what is known about the activation of PKCθ in T cells is focused on its role downstream of TCR. However, PKCθ has been proposed to be activated downstream of other physiological stimuli, including IGF-1 [492] and interferon [354]. These TCR independent interactions are very poorly characterized and our PKCθAPEX proximity screen could be used explore them in a T-ALL context, as well as in other cell types where PKCθ signaling is functionally implicated. 172 References 1. Ward, E., et al., Childhood and adolescent cancer statistics, 2014. CA Cancer J Clin, 2014. 64(2): p. 83-103. 2. Swerdlow, S.H., International Agency for Research on Cancer., and World Health Organization., WHO classification of tumours of haematopoietic and lymphoid tissues. 4th ed. World Health Organization classification of tumours. 2008, Lyon, France: International Agency for Research on Cancer. 439 p. 3. Pui, C.-H., Childhood leukemias. 2nd ed. 2006, Cambridge ; New York: Cambridge University Press. xvi, 960 p., 12 p. of plates. 4. Onciu, M., Acute lymphoblastic leukemia. Hematol Oncol Clin North Am, 2009. 23(4): p. 655-74. 5. Cooper, S.L. and P.A. Brown, Treatment of pediatric acute lymphoblastic leukemia. Pediatr Clin North Am, 2015. 62(1): p. 61-73. 6. Goldberg, J.M., et al., Childhood T-cell acute lymphoblastic leukemia: the Dana-Farber Cancer Institute acute lymphoblastic leukemia consortium experience. J Clin Oncol, 2003. 21(19): p. 3616-22. 7. Gaynon, P.S., et al., Long-term results of the children's cancer group studies for childhood acute lymphoblastic leukemia 1983-2002: a Children's Oncology Group Report. Leukemia, 2010. 24(2): p. 285-97. 8. Laver, J.H., et al., Effects of cranial radiation in children with high risk T cell acute lymphoblastic leukemia: a Pediatric Oncology Group report. Leukemia, 2000. 14(3): p. 369-73. 9. Graux, C., et al., Cytogenetics and molecular genetics of T-cell acute lymphoblastic leukemia: from thymocyte to lymphoblast. Leukemia, 2006. 20(9): p. 1496-510. 10. Han, X. and C.E. Bueso-Ramos, Precursor T-cell acute lymphoblastic leukemia/lymphoblastic lymphoma and acute biphenotypic leukemias. Am J Clin Pathol, 2007. 127(4): p. 528-44. 11. De Keersmaecker, K., P. Marynen, and J. Cools, Genetic insights in the pathogenesis of T-cell acute lymphoblastic leukemia. Haematologica, 2005. 90(8): p. 1116-27. 12. Asnafi, V., et al., CALM-AF10 is a common fusion transcript in T-ALL and is specific to the TCRgammadelta lineage. Blood, 2003. 102(3): p. 1000-6. 13. Hayette, S., et al., Recurrent involvement of the MLL gene in adult T-lineage acute lymphoblastic leukemia. Blood, 2002. 99(12): p. 4647-9. 14. Ellisen, L.W., et al., TAN-1, the human homolog of the Drosophila notch gene, is broken by chromosomal translocations in T lymphoblastic neoplasms. Cell, 1991. 66(4): p. 649-61. 15. Weng, A.P., et al., Activating mutations of NOTCH1 in human T cell acute lymphoblastic leukemia. Science, 2004. 306(5694): p. 269-71. 16. Hebert, J., et al., Candidate tumor-suppressor genes MTS1 (p16INK4A) and MTS2 (p15INK4B) display frequent homozygous deletions in primary cells from T- but not from B-cell lineage acute lymphoblastic leukemias. Blood, 1994. 84(12): p. 4038-44. 17. Remke, M., et al., High-resolution genomic profiling of childhood T-ALL reveals frequent copy-number alterations affecting the TGF-beta and PI3K-AKT pathways and deletions at 6q15-16.1 as a genomic marker for unfavorable early treatment response. Blood, 2009. 114(5): p. 1053-62. 18. Mullighan, C.G., et al., Genome-wide analysis of genetic alterations in acute lymphoblastic leukaemia. Nature, 2007. 446(7137): p. 758-64. 19. Tosello, V., et al., WT1 mutations in T-ALL. Blood, 2009. 114(5): p. 1038-45. 20. Gutierrez, A., et al., Inactivation of LEF1 in T-cell acute lymphoblastic leukemia. Blood, 2010. 115(14): p. 2845-51. 21. Van Vlierberghe, P., et al., ETV6 mutations in early immature human T cell leukemias. J Exp Med, 2011. 208(13): p. 2571-9. 173 22. Gutierrez, A., et al., The BCL11B tumor suppressor is mutated across the major molecular subtypes of T-cell acute lymphoblastic leukemia. Blood, 2011. 118(15): p. 4169-73. 23. Grossmann, V., et al., Prognostic relevance of RUNX1 mutations in T-cell acute lymphoblastic leukemia. Haematologica, 2011. 96(12): p. 1874-7. 24. Zhang, J., et al., The genetic basis of early T-cell precursor acute lymphoblastic leukaemia. Nature, 2012. 481(7380): p. 157-63. 25. Van Vlierberghe, P., et al., PHF6 mutations in T-cell acute lymphoblastic leukemia. Nat Genet, 2010. 42(4): p. 338-42. 26. O'Neil, J., et al., FBW7 mutations in leukemic cells mediate NOTCH pathway activation and resistance to gamma-secretase inhibitors. J Exp Med, 2007. 204(8): p. 1813-24. 27. Thompson, B.J., et al., The SCFFBW7 ubiquitin ligase complex as a tumor suppressor in T cell leukemia. J Exp Med, 2007. 204(8): p. 1825-35. 28. Gutierrez, A., et al., High frequency of PTEN, PI3K, and AKT abnormalities in T-cell acute lymphoblastic leukemia. Blood, 2009. 114(3): p. 647-50. 29. Zuurbier, L., et al., The significance of PTEN and AKT aberrations in pediatric T-cell acute lymphoblastic leukemia. Haematologica, 2012. 97(9): p. 1405-13. 30. Balgobind, B.V., et al., Leukemia-associated NF1 inactivation in patients with pediatric T-ALL and AML lacking evidence for neurofibromatosis. Blood, 2008. 111(8): p. 4322-8. 31. Bar-Eli, M., et al., N-RAS mutations in T-cell acute lymphocytic leukaemia: analysis by direct sequencing detects a novel mutation. Br J Haematol, 1989. 72(1): p. 36-9. 32. Asnafi, V., et al., JAK1 mutations are not frequent events in adult T-ALL: a GRAALL study. Br J Haematol, 2010. 148(1): p. 178-9. 33. Flex, E., et al., Somatically acquired JAK1 mutations in adult acute lymphoblastic leukemia. J Exp Med, 2008. 205(4): p. 751-8. 34. Zenatti, P.P., et al., Oncogenic IL7R gain-of-function mutations in childhood T-cell acute lymphoblastic leukemia. Nat Genet, 2011. 43(10): p. 932-9. 35. Blaumueller, C.M., et al., Intracellular cleavage of Notch leads to a heterodimeric receptor on the plasma membrane. Cell, 1997. 90(2): p. 281-91. 36. Brou, C., et al., A novel proteolytic cleavage involved in Notch signaling: the role of the disintegrin-metalloprotease TACE. Mol Cell, 2000. 5(2): p. 207-16. 37. Mumm, J.S., et al., A ligand-induced extracellular cleavage regulates gamma-secretase-like proteolytic activation of Notch1. Mol Cell, 2000. 5(2): p. 197-206. 38. Fortini, M.E. and S. Artavanis-Tsakonas, The suppressor of hairless protein participates in notch receptor signaling. Cell, 1994. 79(2): p. 273-82. 39. Wu, L., et al., MAML1, a human homologue of Drosophila mastermind, is a transcriptional co-activator for NOTCH receptors. Nat Genet, 2000. 26(4): p. 484-9. 40. Krebs, L.T., et al., Notch signaling regulates left-right asymmetry determination by inducing Nodal expression. Genes Dev, 2003. 17(10): p. 1207-12. 41. Cheng, H.T., et al., Notch2, but not Notch1, is required for proximal fate acquisition in the mammalian nephron. Development, 2007. 134(4): p. 801-11. 42. Kumano, K., et al., Notch1 but not Notch2 is essential for generating hematopoietic stem cells from endothelial cells. Immunity, 2003. 18(5): p. 699-711. 43. Hadland, B.K., et al., A requirement for Notch1 distinguishes 2 phases of definitive hematopoiesis during development. Blood, 2004. 104(10): p. 3097-105. 44. Gerhardt, D.M., et al., The Notch1 transcriptional activation domain is required for development and reveals a novel role for Notch1 signaling in fetal hematopoietic stem cells. Genes Dev, 2014. 28(6): p. 576-93. 174 45. Stier, S., et al., Notch1 activation increases hematopoietic stem cell self-renewal in vivo and favors lymphoid over myeloid lineage outcome. Blood, 2002. 99(7): p. 2369-78. 46. Varnum-Finney, B., et al., Pluripotent, cytokine-dependent, hematopoietic stem cells are immortalized by constitutive Notch1 signaling. Nat Med, 2000. 6(11): p. 1278-81. 47. Maillard, I., et al., Canonical notch signaling is dispensable for the maintenance of adult hematopoietic stem cells. Cell Stem Cell, 2008. 2(4): p. 356-66. 48. Radtke, F., et al., Deficient T cell fate specification in mice with an induced inactivation of Notch1. Immunity, 1999. 10(5): p. 547-58. 49. Pui, J.C., et al., Notch1 expression in early lymphopoiesis influences B versus T lineage determination. Immunity, 1999. 11(3): p. 299-308. 50. Stanger, B.Z., et al., Direct regulation of intestinal fate by Notch. Proc Natl Acad Sci U S A, 2005. 102(35): p. 12443-8. 51. Morrison, S.J., et al., Transient Notch activation initiates an irreversible switch from neurogenesis to gliogenesis by neural crest stem cells. Cell, 2000. 101(5): p. 499-510. 52. Saito, T., et al., Notch2 is preferentially expressed in mature B cells and indispensable for marginal zone B lineage development. Immunity, 2003. 18(5): p. 675-85. 53. Maekawa, Y., et al., Notch2 integrates signaling by the transcription factors RBP-J and CREB1 to promote T cell cytotoxicity. Nat Immunol, 2008. 9(10): p. 1140-7. 54. Geisler, F., et al., Liver-specific inactivation of Notch2, but not Notch1, compromises intrahepatic bile duct development in mice. Hepatology, 2008. 48(2): p. 607-16. 55. Oh, P., et al., In vivo mapping of notch pathway activity in normal and stress hematopoiesis. Cell Stem Cell, 2013. 13(2): p. 190-204. 56. Swiatek, P.J., et al., Notch1 is essential for postimplantation development in mice. Genes Dev, 1994. 8(6): p. 707-19. 57. Hamada, Y., et al., Mutation in ankyrin repeats of the mouse Notch2 gene induces early embryonic lethality. Development, 1999. 126(15): p. 3415-24. 58. Domenga, V., et al., Notch3 is required for arterial identity and maturation of vascular smooth muscle cells. Genes Dev, 2004. 18(22): p. 2730-5. 59. Krebs, L.T., et al., Notch signaling is essential for vascular morphogenesis in mice. Genes Dev, 2000. 14(11): p. 1343-52. 60. Joutel, A., et al., The ectodomain of the Notch3 receptor accumulates within the cerebrovasculature of CADASIL patients. J Clin Invest, 2000. 105(5): p. 597-605. 61. Lardelli, M., J. Dahlstrand, and U. Lendahl, The novel Notch homologue mouse Notch 3 lacks specific epidermal growth factor-repeats and is expressed in proliferating neuroepithelium. Mech Dev, 1994. 46(2): p. 123-36. 62. Felli, M.P., et al., Expression pattern of notch1, 2 and 3 and Jagged1 and 2 in lymphoid and stromal thymus components: distinct ligand-receptor interactions in intrathymic T cell development. Int Immunol, 1999. 11(7): p. 1017-25. 63. Mercher, T., et al., Notch signaling specifies megakaryocyte development from hematopoietic stem cells. Cell Stem Cell, 2008. 3(3): p. 314-26. 64. Ranganathan, P., K.L. Weaver, and A.J. Capobianco, Notch signalling in solid tumours: a little bit of everything but not all the time. Nat Rev Cancer, 2011. 11(5): p. 338-51. 65. Lobry, C., P. Oh, and I. Aifantis, Oncogenic and tumor suppressor functions of Notch in cancer: it's NOTCH what you think. J Exp Med, 2011. 208(10): p. 1931-5. 66. Viatour, P., et al., Notch signaling inhibits hepatocellular carcinoma following inactivation of the RB pathway. J Exp Med, 2011. 208(10): p. 1963-76. 67. Klinakis, A., et al., A novel tumour-suppressor function for the Notch pathway in myeloid leukaemia. Nature, 2011. 473(7346): p. 230-3. 175 68. Agrawal, N., et al., Exome sequencing of head and neck squamous cell carcinoma reveals inactivating mutations in NOTCH1. Science, 2011. 333(6046): p. 1154-7. 69. Stransky, N., et al., The mutational landscape of head and neck squamous cell carcinoma. Science, 2011. 333(6046): p. 1157-60. 70. Westhoff, B., et al., Alterations of the Notch pathway in lung cancer. Proc Natl Acad Sci U S A, 2009. 106(52): p. 22293-8. 71. Puente, X.S., et al., Whole-genome sequencing identifies recurrent mutations in chronic lymphocytic leukaemia. Nature, 2011. 475(7354): p. 101-5. 72. Fabbri, G., et al., Analysis of the chronic lymphocytic leukemia coding genome: role of NOTCH1 mutational activation. J Exp Med, 2011. 208(7): p. 1389-401. 73. Kridel, R., et al., Whole transcriptome sequencing reveals recurrent NOTCH1 mutations in mantle cell lymphoma. Blood, 2012. 119(9): p. 1963-71. 74. Rossi, D., et al., The coding genome of splenic marginal zone lymphoma: activation of NOTCH2 and other pathways regulating marginal zone development. J Exp Med, 2012. 209(9): p. 1537-51. 75. Lee, S.Y., et al., Gain-of-function mutations and copy number increases of Notch2 in diffuse large B-cell lymphoma. Cancer Sci, 2009. 100(5): p. 920-6. 76. Kiel, M.J., et al., Whole-genome sequencing identifies recurrent somatic NOTCH2 mutations in splenic marginal zone lymphoma. J Exp Med, 2012. 209(9): p. 1553-65. 77. Wang, K., et al., PEST domain mutations in Notch receptors comprise an oncogenic driver segment in triple-negative breast cancer sensitive to a gamma-secretase inhibitor. Clin Cancer Res, 2015. 21(6): p. 1487-96. 78. Imatani, A. and R. Callahan, Identification of a novel NOTCH-4/INT-3 RNA species encoding an activated gene product in certain human tumor cell lines. Oncogene, 2000. 19(2): p. 223-31. 79. Malecki, M.J., et al., Leukemia-associated mutations within the NOTCH1 heterodimerization domain fall into at least two distinct mechanistic classes. Mol Cell Biol, 2006. 26(12): p. 4642-51. 80. Breit, S., et al., Activating NOTCH1 mutations predict favorable early treatment response and long-term outcome in childhood precursor T-cell lymphoblastic leukemia. Blood, 2006. 108(4): p. 1151-7. 81. Park, M.J., et al., FBXW7 and NOTCH1 mutations in childhood T cell acute lymphoblastic leukaemia and T cell non-Hodgkin lymphoma. Br J Haematol, 2009. 145(2): p. 198-206. 82. Asnafi, V., et al., NOTCH1/FBXW7 mutation identifies a large subgroup with favorable outcome in adult T-cell acute lymphoblastic leukemia (T-ALL): a Group for Research on Adult Acute Lymphoblastic Leukemia (GRAALL) study. Blood, 2009. 113(17): p. 3918-24. 83. Mansour, M.R., et al., Prognostic implications of NOTCH1 and FBXW7 mutations in adults with T-cell acute lymphoblastic leukemia treated on the MRC UKALLXII/ECOG E2993 protocol. J Clin Oncol, 2009. 27(26): p. 4352-6. 84. Zuurbier, L., et al., NOTCH1 and/or FBXW7 mutations predict for initial good prednisone response but not for improved outcome in pediatric T-cell acute lymphoblastic leukemia patients treated on DCOG or COALL protocols. Leukemia, 2010. 24(12): p. 2014-22. 85. Kox, C., et al., The favorable effect of activating NOTCH1 receptor mutations on long-term outcome in T-ALL patients treated on the ALL-BFM 2000 protocol can be separated from FBXW7 loss of function. Leukemia, 2010. 24(12): p. 2005-13. 86. Clappier, E., et al., NOTCH1 and FBXW7 mutations have a favorable impact on early response to treatment, but not on outcome, in children with T-cell acute lymphoblastic leukemia (T-ALL) treated on EORTC trials 58881 and 58951. Leukemia, 2010. 24(12): p. 2023-31. 87. Palomero, T., et al., NOTCH1 directly regulates c-MYC and activates a feed-forward-loop transcriptional network promoting leukemic cell growth. Proc Natl Acad Sci U S A, 2006. 103(48): p. 18261-6. 176 88. Weng, A.P., et al., c-Myc is an important direct target of Notch1 in T-cell acute lymphoblastic leukemia/lymphoma. Genes Dev, 2006. 20(15): p. 2096-109. 89. Sharma, V.M., et al., Notch1 contributes to mouse T-cell leukemia by directly inducing the expression of c-myc. Mol Cell Biol, 2006. 26(21): p. 8022-31. 90. Chan, S.M., et al., Notch signals positively regulate activity of the mTOR pathway in T-cell acute lymphoblastic leukemia. Blood, 2007. 110(1): p. 278-86. 91. Gonzalez-Garcia, S., et al., CSL-MAML-dependent Notch1 signaling controls T lineage-specific IL-7R{alpha} gene expression in early human thymopoiesis and leukemia. J Exp Med, 2009. 206(4): p. 779-91. 92. Medyouf, H., et al., High-level IGF1R expression is required for leukemia-initiating cell activity in T-ALL and is supported by Notch signaling. J Exp Med, 2011. 208(9): p. 1809-22. 93. Vilimas, T., et al., Targeting the NF-kappaB signaling pathway in Notch1-induced T-cell leukemia. Nat Med, 2007. 13(1): p. 70-7. 94. Espinosa, L., et al., The Notch/Hes1 pathway sustains NF-kappaB activation through CYLD repression in T cell leukemia. Cancer Cell, 2010. 18(3): p. 268-81. 95. Joshi, I., et al., Notch signaling mediates G1/S cell-cycle progression in T cells via cyclin D3 and its dependent kinases. Blood, 2009. 113(8): p. 1689-98. 96. Rao, S.S., et al., Inhibition of NOTCH signaling by gamma secretase inhibitor engages the RB pathway and elicits cell cycle exit in T-cell acute lymphoblastic leukemia cells. Cancer Res, 2009. 69(7): p. 3060-8. 97. Bhavsar, P.J., et al., Analysis of Rho GTPase expression in T-ALL identifies RhoU as a target for Notch involved in T-ALL cell migration. Oncogene, 2013. 32(2): p. 198-208. 98. Mirandola, L., et al., Notch1 regulates chemotaxis and proliferation by controlling the CC-chemokine receptors 5 and 9 in T cell acute lymphoblastic leukaemia. J Pathol, 2012. 226(5): p. 713-22. 99. Buonamici, S., et al., CCR7 signalling as an essential regulator of CNS infiltration in T-cell leukaemia. Nature, 2009. 459(7249): p. 1000-4. 100. Moellering, R.E., et al., Direct inhibition of the NOTCH transcription factor complex. Nature, 2009. 462(7270): p. 182-8. 101. Aste-Amezaga, M., et al., Characterization of Notch1 antibodies that inhibit signaling of both normal and mutated Notch1 receptors. PLoS One, 2010. 5(2): p. e9094. 102. Wu, Y., et al., Therapeutic antibody targeting of individual Notch receptors. Nature, 2010. 464(7291): p. 1052-7. 103. Deangelo, D.J., et al., A phase I clinical trial of the notch inhibitor MK-0752 in patients with T-cell acute lymphoblastic leukemia/lymphoma (T-ALL) and other leukemias. ASCO Meeting Abstracts, 2006. 24(18_suppl): p. 6585. 104. Wei, P., et al., Evaluation of selective gamma-secretase inhibitor PF-03084014 for its antitumor efficacy and gastrointestinal safety to guide optimal clinical trial design. Mol Cancer Ther, 2010. 9(6): p. 1618-28. 105. Zweidler-McKay, P.A., et al., The Safety and Activity of BMS-906024, a Gamma Secretase Inhibitor (GSI) with Anti-Notch Activity, in Patients with Relapsed T-Cell Acute Lymphoblastic Leukemia (T-ALL): Initial Results of a Phase 1 Trial. Vol. 124. 2014. 968-968. 106. Real, P.J., et al., Gamma-secretase inhibitors reverse glucocorticoid resistance in T cell acute lymphoblastic leukemia. Nat Med, 2009. 15(1): p. 50-8. 107. Samon, J.B., et al., Preclinical analysis of the gamma-secretase inhibitor PF-03084014 in combination with glucocorticoids in T-cell acute lymphoblastic leukemia. Mol Cancer Ther, 2012. 11(7): p. 1565-75. 108. Friedman, R.C., et al., Most mammalian mRNAs are conserved targets of microRNAs. Genome Res, 2009. 19(1): p. 92-105. 177 109. Morozova, N., et al., Kinetic signatures of microRNA modes of action. RNA, 2012. 18(9): p. 1635-55. 110. Lewis, B.P., C.B. Burge, and D.P. Bartel, Conserved seed pairing, often flanked by adenosines, indicates that thousands of human genes are microRNA targets. Cell, 2005. 120(1): p. 15-20. 111. Brennecke, J., et al., Principles of microRNA-target recognition. PLoS Biol, 2005. 3(3): p. e85. 112. Kozomara, A. and S. Griffiths-Jones, miRBase: annotating high confidence microRNAs using deep sequencing data. Nucleic Acids Res, 2014. 42(Database issue): p. D68-73. 113. Mosmann, T.R., et al., Two types of murine helper T cell clone. I. Definition according to profiles of lymphokine activities and secreted proteins. J Immunol, 1986. 136(7): p. 2348-57. 114. Harrington, L.E., et al., Interleukin 17-producing CD4+ effector T cells develop via a lineage distinct from the T helper type 1 and 2 lineages. Nat Immunol, 2005. 6(11): p. 1123-32. 115. Schaerli, P., et al., CXC chemokine receptor 5 expression defines follicular homing T cells with B cell helper function. J Exp Med, 2000. 192(11): p. 1553-62. 116. Chen, C.Z., et al., MicroRNAs modulate hematopoietic lineage differentiation. Science, 2004. 303(5654): p. 83-6. 117. Muljo, S.A., et al., Aberrant T cell differentiation in the absence of Dicer. J Exp Med, 2005. 202(2): p. 261-9. 118. Kroesen, B.J., et al., Immuno-miRs: critical regulators of T-cell development, function and ageing. Immunology, 2015. 144(1): p. 1-10. 119. Liu, J., et al., Mechanism of T cell regulation by microRNAs. Cancer Biol Med, 2013. 10(3): p. 131-7. 120. Curtale, G., et al., An emerging player in the adaptive immune response: microRNA-146a is a modulator of IL-2 expression and activation-induced cell death in T lymphocytes. Blood, 2010. 115(2): p. 265-73. 121. Dudda, J.C., et al., MicroRNA-155 is required for effector CD8+ T cell responses to virus infection and cancer. Immunity, 2013. 38(4): p. 742-53. 122. Li, Q.J., et al., miR-181a is an intrinsic modulator of T cell sensitivity and selection. Cell, 2007. 129(1): p. 147-61. 123. Lawrie, C.H., MicroRNAs in hematological malignancies. Blood Rev, 2013. 27(3): p. 143-54. 124. Calin, G.A., et al., Frequent deletions and down-regulation of micro- RNA genes miR15 and miR16 at 13q14 in chronic lymphocytic leukemia. Proc Natl Acad Sci U S A, 2002. 99(24): p. 15524-9. 125. Cimmino, A., et al., miR-15 and miR-16 induce apoptosis by targeting BCL2. Proc Natl Acad Sci U S A, 2005. 102(39): p. 13944-9. 126. Nagel, S., et al., Activation of miR-17-92 by NK-like homeodomain proteins suppresses apoptosis via reduction of E2F1 in T-cell acute lymphoblastic leukemia. Leuk Lymphoma, 2009. 50(1): p. 101-8. 127. Ye, H., et al., MicroRNA and transcription factor co-regulatory network analysis reveals miR-19 inhibits CYLD in T-cell acute lymphoblastic leukemia. Nucleic Acids Res, 2012. 40(12): p. 5201-14. 128. Mavrakis, K.J., et al., Genome-wide RNA-mediated interference screen identifies miR-19 targets in Notch-induced T-cell acute lymphoblastic leukaemia. Nat Cell Biol, 2010. 12(4): p. 372-9. 129. Mets, E., et al., MicroRNA-128-3p is a novel oncomiR targeting PHF6 in T-cell acute lymphoblastic leukemia. Haematologica, 2014. 99(8): p. 1326-33. 130. Verduci, L., et al., microRNA-181a enhances cell proliferation in acute lymphoblastic leukemia by targeting EGR1. Leuk Res, 2015. 39(4): p. 479-85. 131. Junker, F., et al., Dicer1 imparts essential survival cues in Notch driven T-ALL via miR-21 mediated tumor suppressor Pdcd4 repression. Blood, 2015. 132. Li, X., et al., Repression of tumor suppressor miR-451 is essential for NOTCH1-induced oncogenesis in T-ALL. J Exp Med, 2011. 208(4): p. 663-75. 178 133. Gimenes-Teixeira, H.L., et al., Increased expression of miR-221 is associated with shorter overall survival in T-cell acute lymphoid leukemia. Exp Hematol Oncol, 2013. 2(1): p. 10. 134. Xi, Y., et al., Micro-RNA-16 expression in paraffin-embedded specimen correlates with overall survival of T-lymphoblastic lymphoma/leukemia. Hum Pathol, 2013. 44(6): p. 1011-6. 135. Medyouf, H., et al., Acute T-cell leukemias remain dependent on Notch signaling despite PTEN and INK4A/ARF loss. Blood, 2010. 115(6): p. 1175-84. 136. Kuchenbauer, F., et al., Comprehensive analysis of mammalian miRNA* species and their role in myeloid cells. Blood, 2011. 118(12): p. 3350-8. 137. Gentner, B., et al., Stable knockdown of microRNA in vivo by lentiviral vectors. Nat Methods, 2009. 6(1): p. 63-6. 138. Weng, A.P., et al., Growth suppression of pre-T acute lymphoblastic leukemia cells by inhibition of notch signaling. Mol Cell Biol, 2003. 23(2): p. 655-64. 139. Sanghvi, V.R., et al., Characterization of a set of tumor suppressor microRNAs in T cell acute lymphoblastic leukemia. Sci Signal, 2014. 7(352): p. ra111. 140. Gregory, R.I., et al., The Microprocessor complex mediates the genesis of microRNAs. Nature, 2004. 432(7014): p. 235-40. 141. Rainer, J., et al., Glucocorticoid-regulated microRNAs and mirtrons in acute lymphoblastic leukemia. Leukemia, 2009. 23(4): p. 746-52. 142. Mavrakis, K.J., et al., A cooperative microRNA-tumor suppressor gene network in acute T-cell lymphoblastic leukemia (T-ALL). Nat Genet, 2011. 43(7): p. 673-8. 143. Chiaretti, S., et al., Gene expression profiling identifies a subset of adult T-cell acute lymphoblastic leukemia with myeloid-like gene features and over-expression of miR-223. Haematologica, 2010. 95(7): p. 1114-21. 144. Johnnidis, J.B., et al., Regulation of progenitor cell proliferation and granulocyte function by microRNA-223. Nature, 2008. 451(7182): p. 1125-9. 145. Wendorff, A.A., et al., Hes1 is a critical but context-dependent mediator of canonical Notch signaling in lymphocyte development and transformation. Immunity, 2010. 33(5): p. 671-84. 146. Jarriault, S., et al., Signalling downstream of activated mammalian Notch. Nature, 1995. 377(6547): p. 355-8. 147. Jarriault, S., et al., Delta-1 activation of notch-1 signaling results in HES-1 transactivation. Mol Cell Biol, 1998. 18(12): p. 7423-31. 148. Barretina, J., et al., The Cancer Cell Line Encyclopedia enables predictive modelling of anticancer drug sensitivity. Nature, 2012. 483(7391): p. 603-7. 149. Jackstadt, R. and H. Hermeking, MicroRNAs as regulators and mediators of c-MYC function. Biochim Biophys Acta, 2015. 1849(5): p. 544-53. 150. Yoo, A.S. and I. Greenwald, LIN-12/Notch activation leads to microRNA-mediated down-regulation of Vav in C. elegans. Science, 2005. 310(5752): p. 1330-3. 151. Boucher, J.M., et al., The miR-143/145 cluster is a novel transcriptional target of Jagged-1/Notch signaling in vascular smooth muscle cells. J Biol Chem, 2011. 286(32): p. 28312-21. 152. Ruegger, S. and H. Grosshans, MicroRNA turnover: when, how, and why. Trends Biochem Sci, 2012. 37(10): p. 436-46. 153. Gantier, M.P., et al., Analysis of microRNA turnover in mammalian cells following Dicer1 ablation. Nucleic Acids Res, 2011. 39(13): p. 5692-703. 154. Baccarini, A., et al., Kinetic analysis reveals the fate of a microRNA following target regulation in mammalian cells. Curr Biol, 2011. 21(5): p. 369-76. 155. Fukao, T., et al., An evolutionarily conserved mechanism for microRNA-223 expression revealed by microRNA gene profiling. Cell, 2007. 129(3): p. 617-31. 179 156. Felli, N., et al., MicroRNA 223-dependent expression of LMO2 regulates normal erythropoiesis. Haematologica, 2009. 94(4): p. 479-86. 157. Yuan, J.Y., et al., MicroRNA-223 reversibly regulates erythroid and megakaryocytic differentiation of K562 cells. J Cell Mol Med, 2009. 13(11-12): p. 4551-9. 158. Lu, T.X., et al., MiR-223 deficiency increases eosinophil progenitor proliferation. J Immunol, 2013. 190(4): p. 1576-82. 159. Zhuang, G., et al., A novel regulator of macrophage activation: miR-223 in obesity-associated adipose tissue inflammation. Circulation, 2012. 125(23): p. 2892-903. 160. Chen, Q., et al., Inducible microRNA-223 down-regulation promotes TLR-triggered IL-6 and IL-1beta production in macrophages by targeting STAT3. PLoS One, 2012. 7(8): p. e42971. 161. Li, T., et al., MicroRNAs modulate the noncanonical transcription factor NF-kappaB pathway by regulating expression of the kinase IKKalpha during macrophage differentiation. Nat Immunol, 2010. 11(9): p. 799-805. 162. Jia, C.Y., et al., MiR-223 suppresses cell proliferation by targeting IGF-1R. PLoS One, 2011. 6(11): p. e27008. 163. Li, G., et al., Heat shock protein 90B1 plays an oncogenic role and is a target of microRNA-223 in human osteosarcoma. Cell Physiol Biochem, 2012. 30(6): p. 1481-90. 164. Li, J., et al., MicroRNA-223 functions as an oncogene in human gastric cancer by targeting FBXW7/hCdc4. J Cancer Res Clin Oncol, 2012. 138(5): p. 763-74. 165. Kurashige, J., et al., Overexpression of microRNA-223 regulates the ubiquitin ligase FBXW7 in oesophageal squamous cell carcinoma. Br J Cancer, 2012. 106(1): p. 182-8. 166. Hamidi, H., et al., Identification of novel targets of CSL-dependent Notch signaling in hematopoiesis. PLoS One, 2011. 6(5): p. e20022. 167. Kumar, V., et al., Notch and NF-kB signaling pathways regulate miR-223/FBXW7 axis in T-cell acute lymphoblastic leukemia. Leukemia, 2014. 28(12): p. 2324-35. 168. De Obaldia, M.E., et al., T cell development requires constraint of the myeloid regulator C/EBP-alpha by the Notch target and transcriptional repressor Hes1. Nat Immunol, 2013. 14(12): p. 1277-84. 169. Sevignani, C., et al., Mammalian microRNAs: a small world for fine-tuning gene expression. Mamm Genome, 2006. 17(3): p. 189-202. 170. Song, L., et al., Downregulation of miR-223 and miR-153 mediates mechanical stretch-stimulated proliferation of venous smooth muscle cells via activation of the insulin-like growth factor-1 receptor. Arch Biochem Biophys, 2012. 528(2): p. 204-11. 171. Yu, Y.H., et al., MiR-223 regulates human embryonic stem cell differentiation by targeting the IGF-1R/Akt signaling pathway. PLoS One, 2013. 8(11): p. e78769. 172. Pan, Y., et al., Platelet-secreted microRNA-223 promotes endothelial cell apoptosis induced by advanced glycation end products via targeting the insulin-like growth factor 1 receptor. J Immunol, 2014. 192(1): p. 437-46. 173. Xu, Y., et al., MicroRNA-223 regulates cyclin E activity by modulating expression of F-box and WD-40 domain protein 7. J Biol Chem, 2010. 285(45): p. 34439-46. 174. Wu, L., et al., MicroRNA-223 regulates FOXO1 expression and cell proliferation. FEBS Lett, 2012. 586(7): p. 1038-43. 175. Dai, G.H., et al., MicroRNA-223-3p inhibits the angiogenesis of ischemic cardiac microvascular endothelial cells via affecting RPS6KB1/hif-1a signal pathway. PLoS One, 2014. 9(10): p. e108468. 176. Huang, B.S., et al., microRNA-223 promotes the growth and invasion of glioblastoma cells by targeting tumor suppressor PAX6. Oncol Rep, 2013. 30(5): p. 2263-9. 177. Sun, G., H. Li, and J.J. Rossi, Sequence context outside the target region influences the effectiveness of miR-223 target sites in the RhoB 3'UTR. Nucleic Acids Res, 2010. 38(1): p. 239-52. 180 178. Bauernfeind, F., et al., NLRP3 inflammasome activity is negatively controlled by miR-223. J Immunol, 2012. 189(8): p. 4175-81. 179. Fazi, F., et al., A minicircuitry comprised of microRNA-223 and transcription factors NFI-A and C/EBPalpha regulates human granulopoiesis. Cell, 2005. 123(5): p. 819-31. 180. Streppel, M.M., et al., MicroRNA 223 is upregulated in the multistep progression of Barrett's esophagus and modulates sensitivity to chemotherapy by targeting PARP1. Clin Cancer Res, 2013. 19(15): p. 4067-78. 181. Xu, J., et al., MiR-223/Ect2/p21 signaling regulates osteosarcoma cell cycle progression and proliferation. Biomed Pharmacother, 2013. 67(5): p. 381-6. 182. Agatheeswaran, S., et al., BCR-ABL mediated repression of miR-223 results in the activation of MEF2C and PTBP2 in chronic myeloid leukemia. Leukemia, 2013. 27(7): p. 1578-80. 183. Neale, G.A., J.E. Rehg, and R.M. Goorha, Ectopic expression of rhombotin-2 causes selective expansion of CD4-CD8- lymphocytes in the thymus and T-cell tumors in transgenic mice. Blood, 1995. 86(8): p. 3060-71. 184. Hacein-Bey-Abina, S., et al., LMO2-associated clonal T cell proliferation in two patients after gene therapy for SCID-X1. Science, 2003. 302(5644): p. 415-9. 185. Homminga, I., et al., Integrated transcript and genome analyses reveal NKX2-1 and MEF2C as potential oncogenes in T cell acute lymphoblastic leukemia. Cancer Cell, 2011. 19(4): p. 484-97. 186. Garcia, I.S., et al., A study of chromosome 11p13 translocations involving TCR beta and TCR delta in human T cell leukaemia. Oncogene, 1991. 6(4): p. 577-82. 187. Nagel, S., et al., MEF2C is activated by multiple mechanisms in a subset of T-acute lymphoblastic leukemia cell lines. Leukemia, 2008. 22(3): p. 600-7. 188. Chiaretti, S., et al., Myeloid/T-cell acute lymphoblastic leukemia in children and adults. Pediatr Rep, 2011. 3 Suppl 2: p. e3. 189. Zuurbier, L., et al., Immature MEF2C-dysregulated T-cell leukemia patients have an early T-cell precursor acute lymphoblastic leukemia gene signature and typically have non-rearranged T-cell receptors. Haematologica, 2014. 99(1): p. 94-102. 190. Bell, J.J. and A. Bhandoola, The earliest thymic progenitors for T cells possess myeloid lineage potential. Nature, 2008. 452(7188): p. 764-7. 191. Maillard, I., T. Fang, and W.S. Pear, Regulation of lymphoid development, differentiation, and function by the Notch pathway. Annu Rev Immunol, 2005. 23: p. 945-74. 192. Oberg, C., et al., The Notch intracellular domain is ubiquitinated and negatively regulated by the mammalian Sel-10 homolog. J Biol Chem, 2001. 276(38): p. 35847-53. 193. Mansour, M.R., et al., The TAL1 complex targets the FBXW7 tumor suppressor by activating miR-223 in human T cell acute lymphoblastic leukemia. J Exp Med, 2013. 210(8): p. 1545-57. 194. Robinson, D.R., Y.M. Wu, and S.F. Lin, The protein tyrosine kinase family of the human genome. Oncogene, 2000. 19(49): p. 5548-57. 195. Lemmon, M.A. and J. Schlessinger, Cell signaling by receptor tyrosine kinases. Cell, 2010. 141(7): p. 1117-34. 196. Hernandez-Sanchez, C., et al., Evolution of the insulin receptor family and receptor isoform expression in vertebrates. Mol Biol Evol, 2008. 25(6): p. 1043-53. 197. Soos, M.A., et al., Receptors for insulin and insulin-like growth factor-I can form hybrid dimers. Characterisation of hybrid receptors in transfected cells. Biochem J, 1990. 270(2): p. 383-90. 198. Soos, M.A., C.E. Field, and K. Siddle, Purified hybrid insulin/insulin-like growth factor-I receptors bind insulin-like growth factor-I, but not insulin, with high affinity. Biochem J, 1993. 290 ( Pt 2): p. 419-26. 181 199. Morgillo, F., et al., Heterodimerization of insulin-like growth factor receptor/epidermal growth factor receptor and induction of survivin expression counteract the antitumor action of erlotinib. Cancer Res, 2006. 66(20): p. 10100-11. 200. Nahta, R., et al., Insulin-like growth factor-I receptor/human epidermal growth factor receptor 2 heterodimerization contributes to trastuzumab resistance of breast cancer cells. Cancer Res, 2005. 65(23): p. 11118-28. 201. Kavran, J.M., et al., How IGF-1 activates its receptor. Elife, 2014. 3. 202. Tartare-Deckert, S., et al., Evidence for a differential interaction of SHC and the insulin receptor substrate-1 (IRS-1) with the insulin-like growth factor-I (IGF-I) receptor in the yeast two-hybrid system. J Biol Chem, 1995. 270(40): p. 23456-60. 203. Kiely, P.A., A. Sant, and R. O'Connor, RACK1 is an insulin-like growth factor 1 (IGF-1) receptor-interacting protein that can regulate IGF-1-mediated Akt activation and protection from cell death. J Biol Chem, 2002. 277(25): p. 22581-9. 204. Girnita, L., et al., Beta-arrestin and Mdm2 mediate IGF-1 receptor-stimulated ERK activation and cell cycle progression. J Biol Chem, 2007. 282(15): p. 11329-38. 205. Wang, J., et al., Grb10, a positive, stimulatory signaling adapter in platelet-derived growth factor BB-, insulin-like growth factor I-, and insulin-mediated mitogenesis. Mol Cell Biol, 1999. 19(9): p. 6217-28. 206. Sekharam, M., et al., Insulin-like growth factor 1 receptor activates c-SRC and modifies transformation and motility of colon cancer in vitro. Anticancer Res, 2003. 23(2B): p. 1517-24. 207. Giorgetti-Peraldi, S., et al., Involvement of Janus kinases in the insulin signaling pathway. Eur J Biochem, 1995. 234(2): p. 656-60. 208. Monno, S., et al., Insulin-like growth factor I activates c-Jun N-terminal kinase in MCF-7 breast cancer cells. Endocrinology, 2000. 141(2): p. 544-50. 209. Myers, M.G., Jr., et al., IRS-1 activates phosphatidylinositol 3'-kinase by associating with src homology 2 domains of p85. Proc Natl Acad Sci U S A, 1992. 89(21): p. 10350-4. 210. Hanke, S. and M. Mann, The phosphotyrosine interactome of the insulin receptor family and its substrates IRS-1 and IRS-2. Mol Cell Proteomics, 2009. 8(3): p. 519-34. 211. Fruman, D.A., R.E. Meyers, and L.C. Cantley, Phosphoinositide kinases. Annu Rev Biochem, 1998. 67: p. 481-507. 212. Alessi, D.R., et al., Mechanism of activation of protein kinase B by insulin and IGF-1. EMBO J, 1996. 15(23): p. 6541-51. 213. Genot, E.M., et al., The T-cell receptor regulates Akt (protein kinase B) via a pathway involving Rac1 and phosphatidylinositide 3-kinase. Mol Cell Biol, 2000. 20(15): p. 5469-78. 214. Venkitaraman, A.R. and R.J. Cowling, Interleukin-7 induces the association of phosphatidylinositol 3-kinase with the alpha chain of the interleukin-7 receptor. Eur J Immunol, 1994. 24(9): p. 2168-74. 215. Ahmed, N.N., et al., Transduction of interleukin-2 antiapoptotic and proliferative signals via Akt protein kinase. Proc Natl Acad Sci U S A, 1997. 94(8): p. 3627-32. 216. Yoshimura, A., et al., Structural requirements of muramylpeptides for induction of Toll-like receptor 2-mediated NF-kappaB activation in CHO cells. J Endotoxin Res, 2000. 6(5): p. 407-10. 217. Stambolic, V., et al., Negative regulation of PKB/Akt-dependent cell survival by the tumor suppressor PTEN. Cell, 1998. 95(1): p. 29-39. 218. Maehama, T. and J.E. Dixon, The tumor suppressor, PTEN/MMAC1, dephosphorylates the lipid second messenger, phosphatidylinositol 3,4,5-trisphosphate. J Biol Chem, 1998. 273(22): p. 13375-8. 219. Alessi, D.R., et al., Characterization of a 3-phosphoinositide-dependent protein kinase which phosphorylates and activates protein kinase Balpha. Curr Biol, 1997. 7(4): p. 261-9. 182 220. Sarbassov, D.D., et al., Phosphorylation and regulation of Akt/PKB by the rictor-mTOR complex. Science, 2005. 307(5712): p. 1098-101. 221. Inoki, K., et al., TSC2 is phosphorylated and inhibited by Akt and suppresses mTOR signalling. Nat Cell Biol, 2002. 4(9): p. 648-57. 222. Brunn, G.J., et al., Phosphorylation of the translational repressor PHAS-I by the mammalian target of rapamycin. Science, 1997. 277(5322): p. 99-101. 223. Jefferies, H.B., et al., Rapamycin suppresses 5'TOP mRNA translation through inhibition of p70s6k. EMBO J, 1997. 16(12): p. 3693-704. 224. Datta, S.R., et al., Akt phosphorylation of BAD couples survival signals to the cell-intrinsic death machinery. Cell, 1997. 91(2): p. 231-41. 225. Zhou, B.P., et al., HER-2/neu induces p53 ubiquitination via Akt-mediated MDM2 phosphorylation. Nat Cell Biol, 2001. 3(11): p. 973-82. 226. Romashkova, J.A. and S.S. Makarov, NF-kappaB is a target of AKT in anti-apoptotic PDGF signalling. Nature, 1999. 401(6748): p. 86-90. 227. Brunet, A., et al., Akt promotes cell survival by phosphorylating and inhibiting a Forkhead transcription factor. Cell, 1999. 96(6): p. 857-68. 228. Zhou, B.P., et al., Cytoplasmic localization of p21Cip1/WAF1 by Akt-induced phosphorylation in HER-2/neu-overexpressing cells. Nat Cell Biol, 2001. 3(3): p. 245-52. 229. Shin, I., et al., PKB/Akt mediates cell-cycle progression by phosphorylation of p27(Kip1) at threonine 157 and modulation of its cellular localization. Nat Med, 2002. 8(10): p. 1145-52. 230. Liang, J., et al., PKB/Akt phosphorylates p27, impairs nuclear import of p27 and opposes p27-mediated G1 arrest. Nat Med, 2002. 8(10): p. 1153-60. 231. Deprez, J., et al., Phosphorylation and activation of heart 6-phosphofructo-2-kinase by protein kinase B and other protein kinases of the insulin signaling cascades. J Biol Chem, 1997. 272(28): p. 17269-75. 232. Sano, H., et al., Insulin-stimulated phosphorylation of a Rab GTPase-activating protein regulates GLUT4 translocation. J Biol Chem, 2003. 278(17): p. 14599-602. 233. Vivanco, I. and C.L. Sawyers, The phosphatidylinositol 3-Kinase AKT pathway in human cancer. Nat Rev Cancer, 2002. 2(7): p. 489-501. 234. Paz, K., et al., A molecular basis for insulin resistance. Elevated serine/threonine phosphorylation of IRS-1 and IRS-2 inhibits their binding to the juxtamembrane region of the insulin receptor and impairs their ability to undergo insulin-induced tyrosine phosphorylation. J Biol Chem, 1997. 272(47): p. 29911-8. 235. Haruta, T., et al., A rapamycin-sensitive pathway down-regulates insulin signaling via phosphorylation and proteasomal degradation of insulin receptor substrate-1. Mol Endocrinol, 2000. 14(6): p. 783-94. 236. Li, Y., et al., Protein kinase C Theta inhibits insulin signaling by phosphorylating IRS1 at Ser(1101). J Biol Chem, 2004. 279(44): p. 45304-7. 237. Copps, K.D. and M.F. White, Regulation of insulin sensitivity by serine/threonine phosphorylation of insulin receptor substrate proteins IRS1 and IRS2. Diabetologia, 2012. 55(10): p. 2565-82. 238. Dews, M., et al., Domains of the insulin-like growth factor I receptor required for the activation of extracellular signal-regulated kinases. Endocrinology, 2000. 141(4): p. 1289-300. 239. Chang, F., et al., Signal transduction mediated by the Ras/Raf/MEK/ERK pathway from cytokine receptors to transcription factors: potential targeting for therapeutic intervention. Leukemia, 2003. 17(7): p. 1263-93. 240. Minden, A., et al., Differential activation of ERK and JNK mitogen-activated protein kinases by Raf-1 and MEKK. Science, 1994. 266(5191): p. 1719-23. 241. Myal, Y., et al., Receptor binding and growth-promoting activity of insulin-like growth factors in human breast cancer cells (T-47D) in culture. Cancer Res, 1984. 44(12 Pt 1): p. 5486-90. 183 242. Sell, C., et al., IGF-1 receptor levels and the proliferation of young and senescent human fibroblasts. Biochem Biophys Res Commun, 1993. 194(1): p. 259-65. 243. Yang, X.F., et al., Reduced growth of human breast cancer xenografts in hosts homozygous for the lit mutation. Cancer Res, 1996. 56(7): p. 1509-11. 244. Pollak, M., The insulin and insulin-like growth factor receptor family in neoplasia: an update. Nat Rev Cancer, 2012. 12(3): p. 159-69. 245. Vorwerk, P., et al., Expression of components of the IGF signalling system in childhood acute lymphoblastic leukaemia. Mol Pathol, 2002. 55(1): p. 40-5. 246. Baier, T.G., et al., Characterisation of insulin-like growth factor I receptors of human acute lymphoblastic leukaemia (ALL) cell lines and primary ALL cells. Eur J Cancer, 1992. 28A(6-7): p. 1105-10. 247. Karrman, K., et al., The t(X;7)(q22;q34) in paediatric T-cell acute lymphoblastic leukaemia results in overexpression of the insulin receptor substrate 4 gene through illegitimate recombination with the T-cell receptor beta locus. Br J Haematol, 2009. 144(4): p. 546-51. 248. Kang, D.H., et al., Simultaneous translocation of both TCR Loci (14q11) with rare partner loci (Xq22 and 12p13) in a case of T-lymphoblastic leukemia. Ann Lab Med, 2012. 32(3): p. 220-4. 249. Uchida, T., M.G. Myers, Jr., and M.F. White, IRS-4 mediates protein kinase B signaling during insulin stimulation without promoting antiapoptosis. Mol Cell Biol, 2000. 20(1): p. 126-38. 250. Smith, S.D., et al., Long-term growth of malignant thymocytes in vitro. Blood, 1989. 73(8): p. 2182-7. 251. Baier, T.G., et al., Influence of antibodies against IGF-I, insulin or their receptors on proliferation of human acute lymphoblastic leukemia cell lines. Leuk Res, 1992. 16(8): p. 807-14. 252. Huang, Z., et al., Inhibition of type I insulin-like growth factor receptor tyrosine kinase by picropodophyllin induces apoptosis and cell cycle arrest in T lymphoblastic leukemia/lymphoma. Leuk Lymphoma, 2014. 55(8): p. 1876-83. 253. Leclerc, G.M., et al., AMPK-induced activation of Akt by AICAR is mediated by IGF-1R dependent and independent mechanisms in acute lymphoblastic leukemia. J Mol Signal, 2010. 5: p. 15. 254. Trimarchi, T., et al., Genome-wide mapping and characterization of Notch-regulated long noncoding RNAs in acute leukemia. Cell, 2014. 158(3): p. 593-606. 255. Gusscott, S., et al., Notch-mediated repression of miR-223 contributes to IGF1R regulation in T-ALL. Leuk Res, 2012. 36(7): p. 905-11. 256. Aster, J.C., S.C. Blacklow, and W.S. Pear, Notch signalling in T-cell lymphoblastic leukaemia/lymphoma and other haematological malignancies. J Pathol, 2011. 223(2): p. 262-73. 257. Forbes, S.A., et al., COSMIC: exploring the world's knowledge of somatic mutations in human cancer. Nucleic Acids Res, 2015. 43(Database issue): p. D805-11. 258. Craddock, B.P. and W.T. Miller, Effects of somatic mutations in the C-terminus of insulin-like growth factor 1 receptor on activity and signaling. J Signal Transduct, 2012. 2012: p. 804801. 259. Kawamura, M., et al., Alterations of the p53, p21, p16, p15 and RAS genes in childhood T-cell acute lymphoblastic leukemia. Leuk Res, 1999. 23(2): p. 115-26. 260. Perentesis, J.P., et al., RAS oncogene mutations and outcome of therapy for childhood acute lymphoblastic leukemia. Leukemia, 2004. 18(4): p. 685-92. 261. Kindler, T., et al., K-RasG12D-induced T-cell lymphoblastic lymphoma/leukemias harbor Notch1 mutations and are sensitive to gamma-secretase inhibitors. Blood, 2008. 112(8): p. 3373-82. 262. Kharas, M.G., et al., Constitutively active AKT depletes hematopoietic stem cells and induces leukemia in mice. Blood, 2010. 115(7): p. 1406-15. 263. Guo, W., et al., Multi-genetic events collaboratively contribute to Pten-null leukaemia stem-cell formation. Nature, 2008. 453(7194): p. 529-33. 184 264. Trinquand, A., et al., Toward a NOTCH1/FBXW7/RAS/PTEN-based oncogenetic risk classification of adult T-cell acute lymphoblastic leukemia: a Group for Research in Adult Acute Lymphoblastic Leukemia study. J Clin Oncol, 2013. 31(34): p. 4333-42. 265. Prior, I.A., P.D. Lewis, and C. Mattos, A comprehensive survey of Ras mutations in cancer. Cancer Res, 2012. 72(10): p. 2457-67. 266. Karakas, B., K.E. Bachman, and B.H. Park, Mutation of the PIK3CA oncogene in human cancers. Br J Cancer, 2006. 94(4): p. 455-9. 267. Kim, M.S., et al., Mutational analysis of oncogenic AKT E17K mutation in common solid cancers and acute leukaemias. Br J Cancer, 2008. 98(9): p. 1533-5. 268. Shekar, S.C., et al., Mechanism of constitutive phosphoinositide 3-kinase activation by oncogenic mutants of the p85 regulatory subunit. J Biol Chem, 2005. 280(30): p. 27850-5. 269. Chalhoub, N. and S.J. Baker, PTEN and the PI3-kinase pathway in cancer. Annu Rev Pathol, 2009. 4: p. 127-50. 270. You, D., et al., FAK mediates a compensatory survival signal parallel to PI3K-AKT in PTEN-null T-ALL cells. Cell Rep, 2015. 10(12): p. 2055-68. 271. Ciofani, M. and J.C. Zuniga-Pflucker, Notch promotes survival of pre-T cells at the beta-selection checkpoint by regulating cellular metabolism. Nat Immunol, 2005. 6(9): p. 881-8. 272. von Lintig, F.C., et al., Ras activation in normal white blood cells and childhood acute lymphoblastic leukemia. Clin Cancer Res, 2000. 6(5): p. 1804-10. 273. Gregorj, C., et al., ERK1/2 phosphorylation is an independent predictor of complete remission in newly diagnosed adult acute lymphoblastic leukemia. Blood, 2007. 109(12): p. 5473-6. 274. Silva, A., et al., PTEN posttranslational inactivation and hyperactivation of the PI3K/Akt pathway sustain primary T cell leukemia viability. J Clin Invest, 2008. 118(11): p. 3762-74. 275. Barata, J.T., et al., Common gamma chain-signaling cytokines promote proliferation of T-cell acute lymphoblastic leukemia. Haematologica, 2004. 89(12): p. 1459-67. 276. Cardoso, B.A., et al., Interleukin-4 stimulates proliferation and growth of T-cell acute lymphoblastic leukemia cells by activating mTOR signaling. Leukemia, 2009. 23(1): p. 206-8. 277. Scupoli, M.T., et al., Interleukin 7 requirement for survival of T-cell acute lymphoblastic leukemia and human thymocytes on bone marrow stroma. Haematologica, 2007. 92(2): p. 264-6. 278. Scupoli, M.T., et al., Thymic epithelial cells promote survival of human T-cell acute lymphoblastic leukemia blasts: the role of interleukin-7. Haematologica, 2003. 88(11): p. 1229-37. 279. Barata, J.T., et al., Interleukin-7 promotes survival and cell cycle progression of T-cell acute lymphoblastic leukemia cells by down-regulating the cyclin-dependent kinase inhibitor p27(kip1). Blood, 2001. 98(5): p. 1524-31. 280. Barata, J.T., et al., Activation of PI3K is indispensable for interleukin 7-mediated viability, proliferation, glucose use, and growth of T cell acute lymphoblastic leukemia cells. J Exp Med, 2004. 200(5): p. 659-69. 281. Yost, A.J., et al., Defined, serum-free conditions for in vitro culture of primary human T-ALL blasts. Leukemia, 2013. 27(6): p. 1437-40. 282. Porcu, M., et al., Mutation of the receptor tyrosine phosphatase PTPRC (CD45) in T-cell acute lymphoblastic leukemia. Blood, 2012. 119(19): p. 4476-9. 283. Kleppe, M., et al., Deletion of the protein tyrosine phosphatase gene PTPN2 in T-cell acute lymphoblastic leukemia. Nat Genet, 2010. 42(6): p. 530-5. 284. Shieh, A., et al., Defective K-Ras oncoproteins overcome impaired effector activation to initiate leukemia in vivo. Blood, 2013. 121(24): p. 4884-93. 285. Oki, T., et al., Aberrant expression of RasGRP1 cooperates with gain-of-function NOTCH1 mutations in T-cell leukemogenesis. Leukemia, 2012. 26(5): p. 1038-45. 185 286. Naci, D. and F. Aoudjit, Alpha2beta1 integrin promotes T cell survival and migration through the concomitant activation of ERK/Mcl-1 and p38 MAPK pathways. Cell Signal, 2014. 26(9): p. 2008-15. 287. Gombos, A., et al., Clinical development of insulin-like growth factor receptor--1 (IGF-1R) inhibitors: at the crossroad? Invest New Drugs, 2012. 30(6): p. 2433-42. 288. Mitsiades, C.S., et al., Inhibition of the insulin-like growth factor receptor-1 tyrosine kinase activity as a therapeutic strategy for multiple myeloma, other hematologic malignancies, and solid tumors. Cancer Cell, 2004. 5(3): p. 221-30. 289. Chapuis, N., et al., Autocrine IGF-1/IGF-1R signaling is responsible for constitutive PI3K/Akt activation in acute myeloid leukemia: therapeutic value of neutralizing anti-IGF-1R antibody. Haematologica, 2010. 95(3): p. 415-23. 290. Yaktapour, N., et al., Insulin-like growth factor-1 receptor (IGF1R) as a novel target in chronic lymphocytic leukemia. Blood, 2013. 122(9): p. 1621-33. 291. Yamada, H., et al., Effects of insulin-like growth factor-1 on B-cell precursor acute lymphoblastic leukemia. Int J Hematol, 2013. 97(1): p. 73-82. 292. Arcaro, A., Targeting the insulin-like growth factor-1 receptor in human cancer. Front Pharmacol, 2013. 4: p. 30. 293. Jin, M., E. Buck, and M.J. Mulvihill, Modulation of insulin-like growth factor-1 receptor and its signaling network for the treatment of cancer: current status and future perspectives. Oncol Rev, 2013. 7(1): p. e3. 294. Yee, D., Insulin-like growth factor receptor inhibitors: baby or the bathwater? J Natl Cancer Inst, 2012. 104(13): p. 975-81. 295. Villanueva, J., et al., Acquired resistance to BRAF inhibitors mediated by a RAF kinase switch in melanoma can be overcome by cotargeting MEK and IGF-1R/PI3K. Cancer Cell, 2010. 18(6): p. 683-95. 296. Flanigan, S.A., et al., Overcoming IGF1R/IR resistance through inhibition of MEK signaling in colorectal cancer models. Clin Cancer Res, 2013. 19(22): p. 6219-29. 297. Molina-Arcas, M., et al., Coordinate direct input of both KRAS and IGF1 receptor to activation of PI3 kinase in KRAS-mutant lung cancer. Cancer Discov, 2013. 3(5): p. 548-63. 298. Britten, C.D., PI3K and MEK inhibitor combinations: examining the evidence in selected tumor types. Cancer Chemother Pharmacol, 2013. 71(6): p. 1395-409. 299. Renshaw, J., et al., Dual blockade of the PI3K/AKT/mTOR (AZD8055) and RAS/MEK/ERK (AZD6244) pathways synergistically inhibits rhabdomyosarcoma cell growth in vitro and in vivo. Clin Cancer Res, 2013. 19(21): p. 5940-51. 300. Carboni, J.M., et al., BMS-754807, a small molecule inhibitor of insulin-like growth factor-1R/IR. Mol Cancer Ther, 2009. 8(12): p. 3341-9. 301. Cohen, B.D., et al., Combination therapy enhances the inhibition of tumor growth with the fully human anti-type 1 insulin-like growth factor receptor monoclonal antibody CP-751,871. Clin Cancer Res, 2005. 11(5): p. 2063-73. 302. Palomero, T., et al., Mutational loss of PTEN induces resistance to NOTCH1 inhibition in T-cell leukemia. Nat Med, 2007. 13(10): p. 1203-10. 303. Haluska, P., et al., Phase I dose escalation study of the anti insulin-like growth factor-I receptor monoclonal antibody CP-751,871 in patients with refractory solid tumors. Clin Cancer Res, 2007. 13(19): p. 5834-40. 304. Giambra, V., et al., NOTCH1 promotes T cell leukemia-initiating activity by RUNX-mediated regulation of PKC-theta and reactive oxygen species. Nat Med, 2012. 18(11): p. 1693-8. 305. Chiang, M.Y., et al., Leukemia-associated NOTCH1 alleles are weak tumor initiators but accelerate K-ras-initiated leukemia. J Clin Invest, 2008. 118(9): p. 3181-94. 306. Carboni, J.M., et al., Tumor development by transgenic expression of a constitutively active insulin-like growth factor I receptor. Cancer Res, 2005. 65(9): p. 3781-7. 186 307. Yokoyama, K., et al., In vivo leukemogenic potential of an interleukin 7 receptor alpha chain mutant in hematopoietic stem and progenitor cells. Blood, 2013. 122(26): p. 4259-63. 308. Levy, D.S., J.A. Kahana, and R. Kumar, AKT inhibitor, GSK690693, induces growth inhibition and apoptosis in acute lymphoblastic leukemia cell lines. Blood, 2009. 113(8): p. 1723-9. 309. Bressanin, D., et al., Harnessing the PI3K/Akt/mTOR pathway in T-cell acute lymphoblastic leukemia: eliminating activity by targeting at different levels. Oncotarget, 2012. 3(8): p. 811-23. 310. Lonetti, A., et al., PI3K pan-inhibition impairs more efficiently proliferation and survival of T-cell acute lymphoblastic leukemia cell lines when compared to isoform-selective PI3K inhibitors. Oncotarget, 2015. 6(12): p. 10399-414. 311. Lonetti, A., et al., Activity of the pan-class I phosphoinositide 3-kinase inhibitor NVP-BKM120 in T-cell acute lymphoblastic leukemia. Leukemia, 2014. 28(6): p. 1196-206. 312. Chiarini, F., et al., Activity of the novel dual phosphatidylinositol 3-kinase/mammalian target of rapamycin inhibitor NVP-BEZ235 against T-cell acute lymphoblastic leukemia. Cancer Res, 2010. 70(20): p. 8097-107. 313. Mansour, M.R., et al., Targeting oncogenic interleukin-7 receptor signalling with N-acetylcysteine in T cell acute lymphoblastic leukaemia. Br J Haematol, 2015. 168(2): p. 230-8. 314. Zhao, L. and P.K. Vogt, Class I PI3K in oncogenic cellular transformation. Oncogene, 2008. 27(41): p. 5486-96. 315. Vanhaesebroeck, B., L. Stephens, and P. Hawkins, PI3K signalling: the path to discovery and understanding. Nat Rev Mol Cell Biol, 2012. 13(3): p. 195-203. 316. Subramaniam, P.S., et al., Targeting nonclassical oncogenes for therapy in T-ALL. Cancer Cell, 2012. 21(4): p. 459-72. 317. Camps, M., et al., Blockade of PI3Kgamma suppresses joint inflammation and damage in mouse models of rheumatoid arthritis. Nat Med, 2005. 11(9): p. 936-43. 318. Shochat, C., et al., Gain-of-function mutations in interleukin-7 receptor-alpha (IL7R) in childhood acute lymphoblastic leukemias. J Exp Med, 2011. 208(5): p. 901-8. 319. Treanor, L.M., et al., Interleukin-7 receptor mutants initiate early T cell precursor leukemia in murine thymocyte progenitors with multipotent potential. J Exp Med, 2014. 211(4): p. 701-13. 320. Wan, X., et al., IGF-1R Inhibition Activates a YES/SFK Bypass Resistance Pathway: Rational Basis for Co-Targeting IGF-1R and Yes/SFK Kinase in Rhabdomyosarcoma. Neoplasia, 2015. 17(4): p. 358-66. 321. Pavlicek, A., et al., Molecular predictors of sensitivity to the insulin-like growth factor 1 receptor inhibitor Figitumumab (CP-751,871). Mol Cancer Ther, 2013. 12(12): p. 2929-39. 322. Lahortiga, I., et al., Duplication of the MYB oncogene in T cell acute lymphoblastic leukemia. Nat Genet, 2007. 39(5): p. 593-5. 323. Clappier, E., et al., The C-MYB locus is involved in chromosomal translocation and genomic duplications in human T-cell acute leukemia (T-ALL), the translocation defining a new T-ALL subtype in very young children. Blood, 2007. 110(4): p. 1251-61. 324. Huang, F., et al., IRS2 copy number gain, KRAS and BRAF mutation status as predictive biomarkers for response to the IGF-1R/IR inhibitor BMS-754807 in colorectal cancer cell lines. Mol Cancer Ther, 2015. 14(2): p. 620-30. 325. Dave, U.P., et al., Murine leukemias with retroviral insertions at Lmo2 are predictive of the leukemias induced in SCID-X1 patients following retroviral gene therapy. PLoS Genet, 2009. 5(5): p. e1000491. 326. Garofalo, C., et al., Identification of common and distinctive mechanisms of resistance to different anti-IGF-IR agents in Ewing's sarcoma. Mol Endocrinol, 2012. 26(9): p. 1603-16. 327. Bielen, A., et al., Enhanced efficacy of IGF1R inhibition in pediatric glioblastoma by combinatorial targeting of PDGFRalpha/beta. Mol Cancer Ther, 2011. 10(8): p. 1407-18. 187 328. Tu, X., et al., Nuclear translocation of insulin receptor substrate-1 by oncogenes and Igf-I. Effect on ribosomal RNA synthesis. J Biol Chem, 2002. 277(46): p. 44357-65. 329. Wu, A., J. Chen, and R. Baserga, Nuclear insulin receptor substrate-1 activates promoters of cell cycle progression genes. Oncogene, 2008. 27(3): p. 397-403. 330. Guo, D., et al., Notch-1 regulates Akt signaling pathway and the expression of cell cycle regulatory proteins cyclin D1, CDK2 and p21 in T-ALL cell lines. Leuk Res, 2009. 33(5): p. 678-85. 331. Maude, S.L., et al., Efficacy of JAK/STAT pathway inhibition in murine xenograft models of early T-cell precursor (ETP) acute lymphoblastic leukemia. Blood, 2015. 125(11): p. 1759-67. 332. Cornish, A.L., et al., Suppressor of cytokine signaling-1 regulates signaling in response to interleukin-2 and other gamma c-dependent cytokines in peripheral T cells. J Biol Chem, 2003. 278(25): p. 22755-61. 333. Bandapalli, O.R., et al., The activating STAT5B N642H mutation is a common abnormality in pediatric T-cell acute lymphoblastic leukemia and confers a higher risk of relapse. Haematologica, 2014. 99(10): p. e188-92. 334. Steinberg, S.F., Structural basis of protein kinase C isoform function. Physiol Rev, 2008. 88(4): p. 1341-78. 335. Seco, J., et al., Allosteric regulation of PKCtheta: understanding multistep phosphorylation and priming by ligands in AGC kinases. Proteins, 2012. 80(1): p. 269-80. 336. Liu, Y., et al., Phosphorylation of the protein kinase C-theta activation loop and hydrophobic motif regulates its kinase activity, but only activation loop phosphorylation is critical to in vivo nuclear-factor-kappaB induction. Biochem J, 2002. 361(Pt 2): p. 255-65. 337. Chuang, H.C., et al., The kinase GLK controls autoimmunity and NF-kappaB signaling by activating the kinase PKC-theta in T cells. Nat Immunol, 2011. 12(11): p. 1113-8. 338. Liu, Y., et al., Regulation of protein kinase Ctheta function during T cell activation by Lck-mediated tyrosine phosphorylation. J Biol Chem, 2000. 275(5): p. 3603-9. 339. Liu, Y., et al., Protein kinase C theta is expressed in mast cells and is functionally involved in Fcepsilon receptor I signaling. J Leukoc Biol, 2001. 69(5): p. 831-40. 340. Vyas, Y.M., et al., Spatial organization of signal transduction molecules in the NK cell immune synapses during MHC class I-regulated noncytolytic and cytolytic interactions. J Immunol, 2001. 167(8): p. 4358-67. 341. Meller, N., Y. Elitzur, and N. Isakov, Protein kinase C-theta (PKCtheta) distribution analysis in hematopoietic cells: proliferating T cells exhibit high proportions of PKCtheta in the particulate fraction. Cell Immunol, 1999. 193(2): p. 185-93. 342. Chang, J.D., et al., Molecular cloning and expression of a cDNA encoding a novel isoenzyme of protein kinase C (nPKC). A new member of the nPKC family expressed in skeletal muscle, megakaryoblastic cells, and platelets. J Biol Chem, 1993. 268(19): p. 14208-14. 343. Osada, S., et al., A new member of the protein kinase C family, nPKC theta, predominantly expressed in skeletal muscle. Mol Cell Biol, 1992. 12(9): p. 3930-8. 344. Salmond, R.J., et al., T-cell receptor proximal signaling via the Src-family kinases, Lck and Fyn, influences T-cell activation, differentiation, and tolerance. Immunol Rev, 2009. 228(1): p. 9-22. 345. Coudronniere, N., et al., NF-kappa B activation induced by T cell receptor/CD28 costimulation is mediated by protein kinase C-theta. Proc Natl Acad Sci U S A, 2000. 97(7): p. 3394-9. 346. Pfeifhofer, C., et al., Protein kinase C theta affects Ca2+ mobilization and NFAT cell activation in primary mouse T cells. J Exp Med, 2003. 197(11): p. 1525-35. 347. Bauer, B., et al., T cell expressed PKCtheta demonstrates cell-type selective function. Eur J Immunol, 2000. 30(12): p. 3645-54. 348. Pfeifhofer-Obermair, C., N. Thuille, and G. Baier, Involvement of distinct PKC gene products in T cell functions. Front Immunol, 2012. 3: p. 220. 188 349. Sun, Z., et al., PKC-theta is required for TCR-induced NF-kappaB activation in mature but not immature T lymphocytes. Nature, 2000. 404(6776): p. 402-7. 350. Matsumoto, R., et al., Phosphorylation of CARMA1 plays a critical role in T Cell receptor-mediated NF-kappaB activation. Immunity, 2005. 23(6): p. 575-85. 351. Li, Y., et al., SPAK kinase is a substrate and target of PKCtheta in T-cell receptor-induced AP-1 activation pathway. EMBO J, 2004. 23(5): p. 1112-22. 352. Li, W.Q., et al., IL-7 promotes T cell proliferation through destabilization of p27Kip1. J Exp Med, 2006. 203(3): p. 573-82. 353. Shahabi, N.A., K. McAllen, and B.M. Sharp, Stromal cell-derived factor 1-alpha (SDF)-induced human T cell chemotaxis becomes phosphoinositide 3-kinase (PI3K)-independent: role of PKC-theta. J Leukoc Biol, 2008. 83(3): p. 663-71. 354. Srivastava, K.K., et al., Engagement of protein kinase C-theta in interferon signaling in T-cells. J Biol Chem, 2004. 279(29): p. 29911-20. 355. Letschka, T., et al., PKC-theta selectively controls the adhesion-stimulating molecule Rap1. Blood, 2008. 112(12): p. 4617-27. 356. Sutcliffe, E.L., et al., Chromatin-associated protein kinase C-theta regulates an inducible gene expression program and microRNAs in human T lymphocytes. Mol Cell, 2011. 41(6): p. 704-19. 357. Krzewski, K., et al., Formation of a WIP-, WASp-, actin-, and myosin IIA-containing multiprotein complex in activated NK cells and its alteration by KIR inhibitory signaling. J Cell Biol, 2006. 173(1): p. 121-32. 358. Tassi, I., et al., NK cell-activating receptors require PKC-theta for sustained signaling, transcriptional activation, and IFN-gamma secretion. Blood, 2008. 112(10): p. 4109-16. 359. Nagy, B., et al., Impaired activation of platelets lacking protein kinase C-theta isoform. Blood, 2009. 113(11): p. 2557-2567. 360. Jalagadugula, G., et al., Platelet protein kinase C-theta deficiency with human RUNX1 mutation: PRKCQ is a transcriptional target of RUNX1. Arterioscler Thromb Vasc Biol, 2011. 31(4): p. 921-7. 361. Yamada, K., et al., Effects of insulin on the translocation of protein kinase C-theta and other protein kinase C isoforms in rat skeletal muscles. Biochem J, 1995. 308 ( Pt 1): p. 177-80. 362. Kellerer, M., et al., Protein kinase C isoforms alpha, delta and theta require insulin receptor substrate-1 to inhibit the tyrosine kinase activity of the insulin receptor in human kidney embryonic cells (HEK 293 cells). Diabetologia, 1998. 41(7): p. 833-8. 363. Yu, C., et al., Mechanism by which fatty acids inhibit insulin activation of insulin receptor substrate-1 (IRS-1)-associated phosphatidylinositol 3-kinase activity in muscle. J Biol Chem, 2002. 277(52): p. 50230-6. 364. Itani, S.I., et al., Increased protein kinase C theta in skeletal muscle of diabetic patients. Metabolism, 2001. 50(5): p. 553-7. 365. Griffin, M.E., et al., Free fatty acid-induced insulin resistance is associated with activation of protein kinase C theta and alterations in the insulin signaling cascade. Diabetes, 1999. 48(6): p. 1270-4. 366. Kim, J.K., et al., PKC-theta knockout mice are protected from fat-induced insulin resistance. J Clin Invest, 2004. 114(6): p. 823-7. 367. Gill, P.K., A. Gescher, and T.W. Gant, Regulation of MDR1 promoter activity in human breast carcinoma cells by protein kinase C isozymes alpha and theta. Eur J Biochem, 2001. 268(15): p. 4151-7. 368. Budworth, J., T.W. Gant, and A. Gescher, Co-ordinate loss of protein kinase C and multidrug resistance gene expression in revertant MCF-7/Adr breast carcinoma cells. Br J Cancer, 1997. 75(9): p. 1330-5. 369. Zafar, A., et al., Chromatinized protein kinase C-theta directly regulates inducible genes in epithelial to mesenchymal transition and breast cancer stem cells. Mol Cell Biol, 2014. 34(16): p. 2961-80. 189 370. Sourbier, C., et al., Englerin A stimulates PKCtheta to inhibit insulin signaling and to simultaneously activate HSF1: pharmacologically induced synthetic lethality. Cancer Cell, 2013. 23(2): p. 228-37. 371. Motegi, A., et al., PKC theta, a novel immunohistochemical marker for gastrointestinal stromal tumors (GIST), especially useful for identifying KIT-negative tumors. Pathol Int, 2005. 55(3): p. 106-12. 372. Wang, C., et al., Diagnostic significance of DOG-1 and PKC-theta expression and c-Kit/PDGFRA mutations in gastrointestinal stromal tumours. Scand J Gastroenterol, 2013. 48(9): p. 1055-65. 373. Ou, W.B., et al., Protein kinase C-theta regulates KIT expression and proliferation in gastrointestinal stromal tumors. Oncogene, 2008. 27(42): p. 5624-34. 374. Baier, G., et al., Expression and biochemical characterization of human protein kinase C-theta. Eur J Biochem, 1994. 225(1): p. 195-203. 375. Beck, J.F., et al., Anticancer drug-mediated induction of multidrug resistance-associated genes and protein kinase C isozymes in the T-lymphoblastoid cell line CCRF-CEM and in blasts from patients with acute lymphoblastic leukemias. Jpn J Cancer Res, 2001. 92(8): p. 896-903. 376. Winter, S.S., et al., Identification of genomic classifiers that distinguish induction failure in T-lineage acute lymphoblastic leukemia: a report from the Children's Oncology Group. Blood, 2007. 110(5): p. 1429-38. 377. Felli, M.P., et al., PKC theta mediates pre-TCR signaling and contributes to Notch3-induced T-cell leukemia. Oncogene, 2005. 24(6): p. 992-1000. 378. Ptacek, J., et al., Global analysis of protein phosphorylation in yeast. Nature, 2005. 438(7068): p. 679-84. 379. Wang, C., et al., Determination of CK2 specificity and substrates by proteome-derived peptide libraries. J Proteome Res, 2013. 12(8): p. 3813-21. 380. Allen, J.J., et al., A semisynthetic epitope for kinase substrates. Nat Methods, 2007. 4(6): p. 511-6. 381. Allen, J.J., S.E. Lazerwith, and K.M. Shokat, Bio-orthogonal affinity purification of direct kinase substrates. J Am Chem Soc, 2005. 127(15): p. 5288-9. 382. Blencke, S., et al., Characterization of a conserved structural determinant controlling protein kinase sensitivity to selective inhibitors. Chem Biol, 2004. 11(5): p. 691-701. 383. Liu, Y., et al., Structural basis for selective inhibition of Src family kinases by PP1. Chem Biol, 1999. 6(9): p. 671-8. 384. Hanks, S.K. and T. Hunter, Protein kinases 6. The eukaryotic protein kinase superfamily: kinase (catalytic) domain structure and classification. FASEB J, 1995. 9(8): p. 576-96. 385. Shah, K., et al., Engineering unnatural nucleotide specificity for Rous sarcoma virus tyrosine kinase to uniquely label its direct substrates. Proc Natl Acad Sci U S A, 1997. 94(8): p. 3565-70. 386. Sondhi, D., et al., Peptide and protein phosphorylation by protein tyrosine kinase Csk: insights into specificity and mechanism. Biochemistry, 1998. 37(1): p. 165-72. 387. Zhang, C., et al., A second-site suppressor strategy for chemical genetic analysis of diverse protein kinases. Nat Methods, 2005. 2(6): p. 435-41. 388. Cunningham, B.C. and J.A. Wells, High-resolution epitope mapping of hGH-receptor interactions by alanine-scanning mutagenesis. Science, 1989. 244(4908): p. 1081-5. 389. Lopez, M.S., J.I. Kliegman, and K.M. Shokat, The logic and design of analog-sensitive kinases and their small molecule inhibitors. Methods Enzymol, 2014. 548: p. 189-213. 390. Witucki, L.A., et al., Mutant tyrosine kinases with unnatural nucleotide specificity retain the structure and phospho-acceptor specificity of the wild-type enzyme. Chem Biol, 2002. 9(1): p. 25-33. 391. Ubersax, J.A., et al., Targets of the cyclin-dependent kinase Cdk1. Nature, 2003. 425(6960): p. 859-64. 190 392. Wu, D.F., et al., PKCepsilon phosphorylation of the sodium channel NaV1.8 increases channel function and produces mechanical hyperalgesia in mice. J Clin Invest, 2012. 122(4): p. 1306-15. 393. Choi, D.S., et al., Protein kinase Cdelta regulates ethanol intoxication and enhancement of GABA-stimulated tonic current. J Neurosci, 2008. 28(46): p. 11890-9. 394. Weng, Y.C., et al., Identification of lipocalin-2 as a PKCdelta phosphorylation substrate in neutrophils. J Biomed Sci, 2015. 22: p. 21. 395. Cibrian Uhalte, E., et al., In vivo conditions to identify Prkci phosphorylation targets using the analog-sensitive kinase method in zebrafish. PLoS One, 2012. 7(6): p. e40000. 396. Sekar, R.B. and A. Periasamy, Fluorescence resonance energy transfer (FRET) microscopy imaging of live cell protein localizations. J Cell Biol, 2003. 160(5): p. 629-33. 397. Fields, S. and O. Song, A novel genetic system to detect protein-protein interactions. Nature, 1989. 340(6230): p. 245-6. 398. Roux, K.J., et al., A promiscuous biotin ligase fusion protein identifies proximal and interacting proteins in mammalian cells. J Cell Biol, 2012. 196(6): p. 801-10. 399. Corgiat, B.A., J.C. Nordman, and N. Kabbani, Chemical crosslinkers enhance detection of receptor interactomes. Front Pharmacol, 2014. 4: p. 171. 400. Sutherland, B.W., J. Toews, and J. Kast, Utility of formaldehyde cross-linking and mass spectrometry in the study of protein-protein interactions. J Mass Spectrom, 2008. 43(6): p. 699-715. 401. Barker, D.F. and A.M. Campbell, The birA gene of Escherichia coli encodes a biotin holoenzyme synthetase. J Mol Biol, 1981. 146(4): p. 451-67. 402. Chapman-Smith, A. and J.E. Cronan, Jr., Molecular biology of biotin attachment to proteins. J Nutr, 1999. 129(2S Suppl): p. 477S-484S. 403. Kim, D.I., et al., Probing nuclear pore complex architecture with proximity-dependent biotinylation. Proc Natl Acad Sci U S A, 2014. 111(24): p. E2453-61. 404. Couzens, A.L., et al., Protein interaction network of the mammalian Hippo pathway reveals mechanisms of kinase-phosphatase interactions. Sci Signal, 2013. 6(302): p. rs15. 405. Lam, S.S., et al., Directed evolution of APEX2 for electron microscopy and proximity labeling. Nat Methods, 2015. 12(1): p. 51-4. 406. Rhee, H.W., et al., Proteomic mapping of mitochondria in living cells via spatially restricted enzymatic tagging. Science, 2013. 339(6125): p. 1328-31. 407. Mortensen, A. and L.H. Skibsted, Importance of carotenoid structure in radical-scavenging reactions. Journal of Agricultural and Food Chemistry, 1997. 45(8): p. 2970-2977. 408. Bendayan, M., Tech.Sight. Worth its weight in gold. Science, 2001. 291(5507): p. 1363-5. 409. Martell, J.D., et al., Engineered ascorbate peroxidase as a genetically encoded reporter for electron microscopy. Nat Biotechnol, 2012. 30(11): p. 1143-8. 410. Hung, V., et al., Proteomic mapping of the human mitochondrial intermembrane space in live cells via ratiometric APEX tagging. Mol Cell, 2014. 55(2): p. 332-41. 411. Hughes, C.S., et al., Ultrasensitive proteome analysis using paramagnetic bead technology. Molecular Systems Biology, 2014. 10(10). 412. Kall, L., et al., Semi-supervised learning for peptide identification from shotgun proteomics datasets. Nature Methods, 2007. 4(11): p. 923-925. 413. Spivak, M., et al., Improvements to the Percolator Algorithm for Peptide Identification from Shotgun Proteomics Data Sets. Journal of Proteome Research, 2009. 8(7): p. 3737-3745. 414. Ritchie, M.E., et al., limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Research, 2015. 43(7). 415. von Heydebreck, A., et al., Variance stabilization and robust normalization for microarray gene expression data. Compstat 2002: Proceedings in Computational Statistics, 2002: p. 623-628. 191 416. Karp, N.A., et al., Addressing Accuracy and Precision Issues in iTRAQ Quantitation. Molecular & Cellular Proteomics, 2010. 9(9): p. 1885-1897. 417. Hornbeck, P.V., et al., PhosphoSitePlus, 2014: mutations, PTMs and recalibrations. Nucleic Acids Res, 2015. 43(Database issue): p. D512-20. 418. Lee, T.Y., et al., RegPhos: a system to explore the protein kinase-substrate phosphorylation network in humans. Nucleic Acids Res, 2011. 39(Database issue): p. D777-87. 419. Stark, C., et al., BioGRID: a general repository for interaction datasets. Nucleic Acids Res, 2006. 34(Database issue): p. D535-9. 420. Calderone, A., L. Castagnoli, and G. Cesareni, mentha: a resource for browsing integrated protein-interaction networks. Nat Methods, 2013. 10(8): p. 690-1. 421. Ashburner, M., et al., Gene ontology: tool for the unification of biology. The Gene Ontology Consortium. Nat Genet, 2000. 25(1): p. 25-9. 422. Buzko, O. and K.M. Shokat, A kinase sequence database: sequence alignments and family assignment. Bioinformatics, 2002. 18(9): p. 1274-5. 423. Oh, H., et al., Generation of an analog-sensitive Syk tyrosine kinase for the study of signaling dynamics from the B cell antigen receptor. J Biol Chem, 2007. 282(46): p. 33760-8. 424. Niswender, C.M., et al., Protein engineering of protein kinase A catalytic subunits results in the acquisition of novel inhibitor sensitivity. J Biol Chem, 2002. 277(32): p. 28916-22. 425. Kenski, D.M., et al., Chemical genetic engineering of G protein-coupled receptor kinase 2. J Biol Chem, 2005. 280(41): p. 35051-61. 426. Joseph, R.E. and A.H. Andreotti, Controlling the activity of the Tec kinase Itk by mutation of the phenylalanine gatekeeper residue. Biochemistry, 2011. 50(2): p. 221-9. 427. Ultanir, S.K., et al., Chemical genetic identification of NDR1/2 kinase substrates AAK1 and Rabin8 Uncovers their roles in dendrite arborization and spine development. Neuron, 2012. 73(6): p. 1127-42. 428. Hertz, N.T., et al., Chemical genetic approach for kinase-substrate mapping by covalent capture of thiophosphopeptides and analysis by mass spectrometry. Curr Protoc Chem Biol, 2010. 2(1): p. 15-36. 429. Otzen, D.E. and A.R. Fersht, Side-chain determinants of beta-sheet stability. Biochemistry, 1995. 34(17): p. 5718-24. 430. Carlson, S.M. and F.M. White, Labeling and identification of direct kinase substrates. Sci Signal, 2012. 5(227): p. pl3. 431. Quann, E.J., et al., A cascade of protein kinase C isozymes promotes cytoskeletal polarization in T cells. Nat Immunol, 2011. 12(7): p. 647-54. 432. Caloca, M.J., et al., Phorbol esters and related analogs regulate the subcellular localization of beta 2-chimaerin, a non-protein kinase C phorbol ester receptor. J Biol Chem, 2001. 276(21): p. 18303-12. 433. Betz, A., et al., Munc13-1 is a presynaptic phorbol ester receptor that enhances neurotransmitter release. Neuron, 1998. 21(1): p. 123-36. 434. Tognon, C.E., et al., Regulation of RasGRP via a phorbol ester-responsive C1 domain. Mol Cell Biol, 1998. 18(12): p. 6995-7008. 435. Bi, K., et al., Antigen-induced translocation of PKC-theta to membrane rafts is required for T cell activation. Nat Immunol, 2001. 2(6): p. 556-63. 436. Carrasco, S. and I. Merida, Diacylglycerol-dependent binding recruits PKCtheta and RasGRP1 C1 domains to specific subcellular localizations in living T lymphocytes. Mol Biol Cell, 2004. 15(6): p. 2932-42. 437. McKay, C.S. and M.G. Finn, Click chemistry in complex mixtures: bioorthogonal bioconjugation. Chem Biol, 2014. 21(9): p. 1075-101. 438. Turban, S. and E. Hajduch, Protein kinase C isoforms: mediators of reactive lipid metabolites in the development of insulin resistance. FEBS Lett, 2011. 585(2): p. 269-74. 192 439. Qu, B.H., et al., Insulin receptor substrate-4 enhances insulin-like growth factor-I-induced cell proliferation. J Biol Chem, 1999. 274(44): p. 31179-84. 440. Hoxhaj, G., K. Dissanayake, and C. MacKintosh, Effect of IRS4 levels on PI 3-kinase signalling. PLoS One, 2013. 8(9): p. e73327. 441. Lee, S., et al., Protein kinase C-zeta phosphorylates insulin receptor substrate-1, -3, and -4 but not -2: isoform specific determinants of specificity in insulin signaling. Endocrinology, 2008. 149(5): p. 2451-8. 442. Ikenoue, T., et al., Essential function of TORC2 in PKC and Akt turn motif phosphorylation, maturation and signalling. EMBO J, 2008. 27(14): p. 1919-31. 443. Lee, K., et al., Mammalian target of rapamycin protein complex 2 regulates differentiation of Th1 and Th2 cell subsets via distinct signaling pathways. Immunity, 2010. 32(6): p. 743-53. 444. Gao, D., et al., Rictor forms a complex with Cullin-1 to promote SGK1 ubiquitination and destruction. Mol Cell, 2010. 39(5): p. 797-808. 445. Dibble, C.C., J.M. Asara, and B.D. Manning, Characterization of Rictor phosphorylation sites reveals direct regulation of mTOR complex 2 by S6K1. Mol Cell Biol, 2009. 29(21): p. 5657-70. 446. Julien, L.A., et al., mTORC1-activated S6K1 phosphorylates Rictor on threonine 1135 and regulates mTORC2 signaling. Mol Cell Biol, 2010. 30(4): p. 908-21. 447. Chen, C.H., et al., ER stress inhibits mTORC2 and Akt signaling through GSK-3beta-mediated phosphorylation of rictor. Sci Signal, 2011. 4(161): p. ra10. 448. Koo, J., et al., Rictor Undergoes Glycogen Synthase Kinase 3 (GSK3)-dependent, FBXW7-mediated Ubiquitination and Proteasomal Degradation. J Biol Chem, 2015. 290(22): p. 14120-9. 449. Soliman, G.A., et al., mTOR Ser-2481 autophosphorylation monitors mTORC-specific catalytic activity and clarifies rapamycin mechanism of action. J Biol Chem, 2010. 285(11): p. 7866-79. 450. Chiang, G.G. and R.T. Abraham, Phosphorylation of mammalian target of rapamycin (mTOR) at Ser-2448 is mediated by p70S6 kinase. J Biol Chem, 2005. 280(27): p. 25485-90. 451. Ekim, B., et al., mTOR kinase domain phosphorylation promotes mTORC1 signaling, cell growth, and cell cycle progression. Mol Cell Biol, 2011. 31(14): p. 2787-801. 452. Holz, M.K. and J. Blenis, Identification of S6 kinase 1 as a novel mammalian target of rapamycin (mTOR)-phosphorylating kinase. J Biol Chem, 2005. 280(28): p. 26089-93. 453. Acosta-Jaquez, H.A., et al., Site-specific mTOR phosphorylation promotes mTORC1-mediated signaling and cell growth. Mol Cell Biol, 2009. 29(15): p. 4308-24. 454. Evangelisti, C., et al., Therapeutic potential of targeting mTOR in T-cell acute lymphoblastic leukemia (review). Int J Oncol, 2014. 45(3): p. 909-18. 455. Ramstad, C., et al., cAMP-dependent protein kinase (PKA) inhibits T cell activation by phosphorylating ser-43 of raf-1 in the MAPK/ERK pathway. Cell Signal, 2000. 12(8): p. 557-63. 456. Tzivion, G., Z. Luo, and J. Avruch, A dimeric 14-3-3 protein is an essential cofactor for Raf kinase activity. Nature, 1998. 394(6688): p. 88-92. 457. King, A.J., et al., The protein kinase Pak3 positively regulates Raf-1 activity through phosphorylation of serine 338. Nature, 1998. 396(6707): p. 180-3. 458. Balan, V., et al., Identification of novel in vivo Raf-1 phosphorylation sites mediating positive feedback Raf-1 regulation by extracellular signal-regulated kinase. Mol Biol Cell, 2006. 17(3): p. 1141-53. 459. Dougherty, M.K., et al., Regulation of Raf-1 by direct feedback phosphorylation. Mol Cell, 2005. 17(2): p. 215-24. 460. Kolch, W., et al., Protein kinase C alpha activates RAF-1 by direct phosphorylation. Nature, 1993. 364(6434): p. 249-52. 461. Hindley, A. and W. Kolch, Raf-1 and B-Raf promote protein kinase C theta interaction with BAD. Cell Signal, 2007. 19(3): p. 547-55. 193 462. Baljuls, A., et al., Positive regulation of A-RAF by phosphorylation of isoform-specific hinge segment and identification of novel phosphorylation sites. J Biol Chem, 2008. 283(40): p. 27239-54. 463. Emuss, V., et al., Mutations of C-RAF are rare in human cancer because C-RAF has a low basal kinase activity compared with B-RAF. Cancer Res, 2005. 65(21): p. 9719-26. 464. Sia, D., et al., Massive parallel sequencing uncovers actionable FGFR2-PPHLN1 fusion and ARAF mutations in intrahepatic cholangiocarcinoma. Nat Commun, 2015. 6: p. 6087. 465. Lee, J.W., et al., Mutational analysis of the ARAF gene in human cancers. APMIS, 2005. 113(1): p. 54-7. 466. Gray, P.J., Jr., et al., Targeting the oncogene and kinome chaperone CDC37. Nat Rev Cancer, 2008. 8(7): p. 491-5. 467. Grammatikakis, N., et al., p50(cdc37) acting in concert with Hsp90 is required for Raf-1 function. Mol Cell Biol, 1999. 19(3): p. 1661-72. 468. Gould, C.M., et al., The chaperones Hsp90 and Cdc37 mediate the maturation and stabilization of protein kinase C through a conserved PXXP motif in the C-terminal tail. J Biol Chem, 2009. 284(8): p. 4921-35. 469. Stepanova, L., et al., The oncoprotein kinase chaperone CDC37 functions as an oncogene in mice and collaborates with both c-myc and cyclin D1 in transformation of multiple tissues. Mol Cell Biol, 2000. 20(12): p. 4462-73. 470. Heidel, F.H., et al., The cell fate determinant Llgl1 influences HSC fitness and prognosis in AML. J Exp Med, 2013. 210(1): p. 15-22. 471. Yang, J.J., et al., Genome-wide copy number profiling reveals molecular evolution from diagnosis to relapse in childhood acute lymphoblastic leukemia. Blood, 2008. 112(10): p. 4178-83. 472. Kaminski, M., et al., Novel role for mitochondria: protein kinase Ctheta-dependent oxidative signaling organelles in activation-induced T-cell death. Mol Cell Biol, 2007. 27(10): p. 3625-39. 473. Nitti, M., et al., PKC delta and NADPH oxidase in retinoic acid-induced neuroblastoma cell differentiation. Cell Signal, 2010. 22(5): p. 828-35. 474. Chen, F., et al., Regulation of NADPH oxidase 5 by protein kinase C isoforms. PLoS One, 2014. 9(2): p. e88405. 475. Majumder, P.K., et al., Mitochondrial translocation of protein kinase C delta in phorbol ester-induced cytochrome c release and apoptosis. J Biol Chem, 2000. 275(29): p. 21793-6. 476. Li, L., et al., Protein kinase Cdelta targets mitochondria, alters mitochondrial membrane potential, and induces apoptosis in normal and neoplastic keratinocytes when overexpressed by an adenoviral vector. Mol Cell Biol, 1999. 19(12): p. 8547-58. 477. Guinamard, R., et al., B cell antigen receptor engagement inhibits stromal cell-derived factor (SDF)-1alpha chemotaxis and promotes protein kinase C (PKC)-induced internalization of CXCR4. J Exp Med, 1999. 189(9): p. 1461-6. 478. Signoret, N., et al., Phorbol esters and SDF-1 induce rapid endocytosis and down modulation of the chemokine receptor CXCR4. J Cell Biol, 1997. 139(3): p. 651-64. 479. Busillo, J.M., et al., Site-specific phosphorylation of CXCR4 is dynamically regulated by multiple kinases and results in differential modulation of CXCR4 signaling. J Biol Chem, 2010. 285(10): p. 7805-17. 480. Mueller, W., et al., Hierarchical organization of multi-site phosphorylation at the CXCR4 C terminus. PLoS One, 2013. 8(5): p. e64975. 481. Banko, M.R., et al., Chemical genetic screen for AMPKalpha2 substrates uncovers a network of proteins involved in mitosis. Mol Cell, 2011. 44(6): p. 878-92. 482. Xue, L., R.L. Geahlen, and W.A. Tao, Identification of direct tyrosine kinase substrates based on protein kinase assay-linked phosphoproteomics. Mol Cell Proteomics, 2013. 12(10): p. 2969-80. 483. Hanke, J.H., et al., Discovery of a novel, potent, and Src family-selective tyrosine kinase inhibitor. Study of Lck- and FynT-dependent T cell activation. J Biol Chem, 1996. 271(2): p. 695-701. 194 484. Schindler, T., et al., Crystal structure of Hck in complex with a Src family-selective tyrosine kinase inhibitor. Mol Cell, 1999. 3(5): p. 639-48. 485. Au-Yeung, B.B., et al., A genetically selective inhibitor demonstrates a function for the kinase Zap70 in regulatory T cells independent of its catalytic activity. Nat Immunol, 2010. 11(12): p. 1085-92. 486. Zhang, C., et al., Structure-guided inhibitor design expands the scope of analog-sensitive kinase technology. ACS Chem Biol, 2013. 8(9): p. 1931-8. 487. Blethrow, J.D., et al., Covalent capture of kinase-specific phosphopeptides reveals Cdk1-cyclin B substrates. Proc Natl Acad Sci U S A, 2008. 105(5): p. 1442-7. 488. Nishikawa, K., et al., Determination of the specific substrate sequence motifs of protein kinase C isozymes. J Biol Chem, 1997. 272(2): p. 952-60. 489. Schiapparelli, L.M., et al., Direct detection of biotinylated proteins by mass spectrometry. J Proteome Res, 2014. 13(9): p. 3966-78. 490. Parandoosh, Z., et al., Encoded chemical synthesis coupled to screening: "Pot Assay". Comb Chem High Throughput Screen, 1998. 1(3): p. 135-42. 491. Lopez-Sanchez, I., et al., Protein kinase C-theta (PKCtheta) phosphorylates and inhibits the guanine exchange factor, GIV/Girdin. Proc Natl Acad Sci U S A, 2013. 110(14): p. 5510-5. 492. Thimmaiah, K.N., J.B. Easton, and P.J. Houghton, Protection from rapamycin-induced apoptosis by insulin-like growth factor-I is partially dependent on protein kinase C signaling. Cancer Res, 2010. 70(5): p. 2000-9. 493. Moeschel, K., et al., Protein kinase C-zeta-induced phosphorylation of Ser318 in insulin receptor substrate-1 (IRS-1) attenuates the interaction with the insulin receptor and the tyrosine phosphorylation of IRS-1. J Biol Chem, 2004. 279(24): p. 25157-63. 494. Greene, M.W., et al., PKCdelta-mediated IRS-1 Ser24 phosphorylation negatively regulates IRS-1 function. Biochem Biophys Res Commun, 2006. 349(3): p. 976-86. 495. Chatterjee, S., B. Behnam Azad, and S. Nimmagadda, The intricate role of CXCR4 in cancer. Adv Cancer Res, 2014. 124: p. 31-82. 496. Teicher, B.A. and S.P. Fricker, CXCL12 (SDF-1)/CXCR4 pathway in cancer. Clin Cancer Res, 2010. 16(11): p. 2927-31. 497. Nagasawa, T., CXC chemokine ligand 12 (CXCL12) and its receptor CXCR4. J Mol Med (Berl), 2014. 92(5): p. 433-9. 498. Balabanian, K., et al., The chemokine SDF-1/CXCL12 binds to and signals through the orphan receptor RDC1 in T lymphocytes. J Biol Chem, 2005. 280(42): p. 35760-6. 499. Rajagopal, S., et al., Beta-arrestin- but not G protein-mediated signaling by the "decoy" receptor CXCR7. Proc Natl Acad Sci U S A, 2010. 107(2): p. 628-32. 500. Bendall, L.J., et al., Defective p38 mitogen-activated protein kinase signaling impairs chemotaxic but not proliferative responses to stromal-derived factor-1alpha in acute lymphoblastic leukemia. Cancer Res, 2005. 65(8): p. 3290-8. 501. Barbero, S., et al., Stromal cell-derived factor 1alpha stimulates human glioblastoma cell growth through the activation of both extracellular signal-regulated kinases 1/2 and Akt. Cancer Res, 2003. 63(8): p. 1969-74. 502. Kremer, K.N., et al., Stromal cell-derived factor-1 signaling via the CXCR4-TCR heterodimer requires phospholipase C-beta3 and phospholipase C-gamma1 for distinct cellular responses. J Immunol, 2011. 187(3): p. 1440-7. 503. Vila-Coro, A.J., et al., The chemokine SDF-1alpha triggers CXCR4 receptor dimerization and activates the JAK/STAT pathway. FASEB J, 1999. 13(13): p. 1699-710. 504. Feng, Y., et al., HIV-1 entry cofactor: functional cDNA cloning of a seven-transmembrane, G protein-coupled receptor. Science, 1996. 272(5263): p. 872-7. 195 505. Zou, Y.R., et al., Function of the chemokine receptor CXCR4 in haematopoiesis and in cerebellar development. Nature, 1998. 393(6685): p. 595-9. 506. Tachibana, K., et al., The chemokine receptor CXCR4 is essential for vascularization of the gastrointestinal tract. Nature, 1998. 393(6685): p. 591-4. 507. Ara, T., et al., Impaired colonization of the gonads by primordial germ cells in mice lacking a chemokine, stromal cell-derived factor-1 (SDF-1). Proc Natl Acad Sci U S A, 2003. 100(9): p. 5319-23. 508. Nagasawa, T., et al., Defects of B-cell lymphopoiesis and bone-marrow myelopoiesis in mice lacking the CXC chemokine PBSF/SDF-1. Nature, 1996. 382(6592): p. 635-8. 509. Ara, T., et al., Long-term hematopoietic stem cells require stromal cell-derived factor-1 for colonizing bone marrow during ontogeny. Immunity, 2003. 19(2): p. 257-67. 510. Sugiyama, T., et al., Maintenance of the hematopoietic stem cell pool by CXCL12-CXCR4 chemokine signaling in bone marrow stromal cell niches. Immunity, 2006. 25(6): p. 977-88. 511. Greenbaum, A., et al., CXCL12 in early mesenchymal progenitors is required for haematopoietic stem-cell maintenance. Nature, 2013. 495(7440): p. 227-30. 512. Janas, M.L., et al., Thymic development beyond beta-selection requires phosphatidylinositol 3-kinase activation by CXCR4. J Exp Med, 2010. 207(1): p. 247-61. 513. Trampont, P.C., et al., CXCR4 acts as a costimulator during thymic beta-selection. Nat Immunol, 2010. 11(2): p. 162-70. 514. Kumar, A., et al., CXCR4 physically associates with the T cell receptor to signal in T cells. Immunity, 2006. 25(2): p. 213-24. 515. Passaro, D., et al., CXCR4 Is Required for Leukemia-Initiating Cell Activity in T Cell Acute Lymphoblastic Leukemia. Cancer Cell, 2015. 27(6): p. 769-79. 516. Pitt, L.A., et al., CXCL12-Producing Vascular Endothelial Niches Control Acute T Cell Leukemia Maintenance. Cancer Cell, 2015. 27(6): p. 755-68. 517. Cannon, J.L., et al., PKCtheta regulates T cell motility via ezrin-radixin-moesin localization to the uropod. PLoS One, 2013. 8(11): p. e78940. 196 Appendices Appendix A. Chemical synthesis and validation of HexT HexT was synthesized as described in Rhee et al. [406] (Figure A.1A). Identity and purity of the synthesized HexT compound was confirmed using NMR and mass spectrometry (courtesy of Dr Chris Hughes and Dr Kuo-Shyan Lin) (Figure A.1B/C). HATU, DIPEADMSOHexynoic acid Tyramine Hexyne-Tyramide (HexT) Figure A.1B/C (A) (B) 197 Figure A.1 Chemical synthesis and confirmation of HexT (A) Graphical representation of HexT synthesis from tyramine and hexynoic acid. Confirmation of HexT purity using (B) mass spectrometry and (C) nuclear magnetic resonance (NMR). (C) 198 Appendix B. Previously described PKCθ phosphotargets, PKCθ Interactors and TCR signalling components Table B.1 Previously described PKCθ phosphotargets Compiled from http://www.phosphosite.org [417] and http://140.138.144.141/~RegPhos/index.php [418] database. ARHGEF6 CHUKAICDA IKBKBBADCACNA1CCARD11CBLBCCDC88ACD4EZRFOSL1GRM5HABP4HIST1H3AICAM3IRS1MSNNR2F6PPP1R12APTPN7RAPGEF2RASGRP1RASGRP3RDXSTK39WIPF1NDRG2PDPK1PIP5K1BPTPN6SPNKnown PKCθ phosphotargetsGene Name 199 Table B.2. Previously described PKCθ interactors Compiled from http://thebiogrid.org/ [419] and http://mentha.uniroma2.it/index.php [420]. AKT1 HLA-DQB1 KCNA5 TEC V5-4BCL10 HLA-DQB2 LCK TRAC V5-6BTK HLA-DRA LCP2 TRAF6 VAV1CARD11 HLA-DRB1 LYN TRBC1 VAV2CASP3 HLA-DRB3 MALT1 TRBV12-3 VAV3CBL HLA-DRB4 MAP3K7 UBA52 YWHAGCD3D HLA-DRB5 MARK3 UBB YWHAQCD3E HSF1 MEF2A UBCCD3G HSP90AA1 MS4A2 UBE2NCD4 HSP90AB1 NFE2L2 UBE2V1CHUK IGHE PDPK1 V1-11DCC IGHV PLCB1 V1-13DEF6 IGHV3-23 PLCB2 V1-16EZR IGHV7-81 PLCB3 V1-20FCER1A IGKC PLCB4 V1-3FCER1G IGKV PLCG1 V1-5FYB IGKV1-5 PLCG2 V1-7FYN IGKV1D-16 PRKCD V1-9GLRX3 IGKV4-1 PRKCZ V2-11GNA11 IGLC1 RASGRP3 V2-15GNA14 IGLC2 RGS9BP V2-17GNA15 IGLC3 RPS27A V2-19GNAQ IGLC6 SELL V2-8GNB1 IGLC7 SOS1 V3-2GNG2 IGLV TAB1 V3-3GRAP2 IKBKB TAB2 V3-4HLA-DPA1 IKBKG TAB3 V4-1HLA-DPB1 Irs1 TAL1 V4-2HLA-DQA1 ITK TCRA V4-6HLA-DQA2 JUND TCRB V5-1Gene NameKnown PKCθ Interactors 200 Table B.3. Previously described TCR signalling components Taken from http://amigo.geneontology.org/amigo/landing [421] (GO: 0050852(mammalian)) ADA DRD4 LCK PSEN1 TRACBCAR1 DUSP22 LCP2 PSEN2 TRAF6BCL10 DUSP3 LGALS3 PTEN TRAT1BCL2A1D ELF1 LIME1 PTPN2 TRBC1BMX ENAH MALT1 PTPN22 TRBV12-3BTK EZR MAP3K7 PTPN6 TXKBTN2A2 FOXP3 MAPK1 PTPRC UBA52BTN3A1 FYB NCK1 PTPRJ UBASH3ABTNL2 FYN NFKB1 RBCK1 UBBCACNA1F GATA 3 NFKBIA RC3H1 UBCCACNB3 GATA3 PAG1 RELA UBE2NCACNB4 GBP1 PAK1 RFTN1 VASPCARD11 GRAP2 PAK2 RIPK2 WASCBLB GTPBP2 PAK3 RLTPR ZAP70CCR7 HLA-DPA1 PAWR RNF31CD247 HLA-DPB1 PDE4B SHBCD28 HLA-DQA1 PDE4D SKAP1CD300A HLA-DQA2 PDPK1 SPPL3CD3D HLA-DQB1 PELO STK11CD3E HLA-DQB2 PHPT1 STOML2CD3G HLA-DRA PIK3CA TAB2CD4 HLA-DRB1 PIK3CB TCRACD90 HLA-DRB3 PIK3CD TCRBCHUK HLA-DRB4 PIK3R1 TECCLEC2I HLA-DRB5 PIK3R2 TESPA1CNOT6 IKBKG PLCG1 THEMISCSDE1 INPP5D PLCG2 THEMIS2CSK ITK PRKCQ THY1DRD2 KCNN4 PRKD2 THY-1DRD3 LAT PRNP TNFRSF21Gene NameTCR signal components 201 Appendix C. Proteins identified in 3 out of 4 PKCθAS experiments As a requirement for consideration as a candidate PKCθ direct phosphotarget, a protein species needed to be identified identification in all four PKCθAS experiments. Many protein species were identified in 3 out of 4 PKCθAS experiments and may still represent interesting potential candidates with examples being previously implicated in T-ALL (PHF6), mTORC2 complex (TELO2) and previously being described as a PKCθ phosphotarget (CCDC88A). Table C.1 Proteins identified in PKCθAS experiments #1, #2 and #3 only UniProtID Gene name UniProtID Gene nameQ15057 ACAP2 Q5W0B1 RNF219Q9NUQ2 AGPAT5 P50454 SERPINH1P46736 BRCC3 P53814 SMTNQ3V6T2 CCDC88A P62306 SNRPFO00311 CDC7 Q9Y2H1 STK38LQ96EY1 DNAJA3 P53597 SUCLG1P63167 DYNLL1 O14787 TNPO2Q9NV70 EXOC1 Q8NFQ8 TOR1AIP2O60645 EXOC3 O14545 TRAFD1Q8IYI6 EXOC8 Q6P3X3 TTC27Q8WVX9 FAR1 Q9BQE3 TUBA1CQ9NYY8 FASTKD2 Q9BVA1 TUBB2BP23142 FBLN1 Q15386 UBE3CQ9H583 HEATR1 O43709 WBSCR22P51617 IRAK1Q8WUY8 NAT14O75251 NDUFS7P21359 NF1Q9Y697 NFS1P49757 NUMBQ5SRE5 NUP188P51003 PAPOLAP49585 PCYT1AQ96HC4 PDLIM5P52209 PGDQ8IWS0 PHF6O75688 PPM1BP53041 PPP5CP24723 PRKCHO43172 PRPF4PKCθAS Exp# 1+2+3 202 Table C.2 Proteins identified in PKCθAS experiments #1, #2 and #4 only UniProtID Gene nameQ9UNQ2 DIMT1Q92796 DLG3Q9ULK4 MED23A2RRP1 NBASO96011 PEX11BQ99570 PIK3R4Q99808 SLC29A1Q5BJF2 TMEM97Q9UGJ1 TUBGCP4PKCθAS Exp# 1+2+4 Table C.3 Proteins identified in PKCθAS experiments #1, #3 and #4 only UniProtID Gene nameO14981 BTAF1O96005 CLPTM1Q9UBM7 DHCR7Q9UPY3 DICER1Q14185 DOCK1Q96P70 IPO9P29966 MARCKSQ09161 NCBP1P55786 NPEPPSQ9UBU9 NXF1O15027 SEC16AP12270 TPRQ9Y277 VDAC3PKCθAS Exp# 1+3+4 203 Table C.4 Proteins identified in PKCθAS experiments #2, #3 and #4 only UniProtID Gene name UniProtID Gene name UniProtID Gene nameP49588 AARS Q3KQU3 MAP7D1 P51571 SSR4P61160 ACTR2 Q14149 MORC3 Q8N3U4 STAG2Q8NI60 ADCK3 P43246 MSH2 Q9BW92 TARS2P18085 ARF4 Q13459 MYO9B Q9Y4R8 TELO2Q9Y6D5 ARFGEF2 Q9BTX1 NDC1 Q9UNS1 TIMELESSQ9Y679 AUP1 Q16795 NDUFA9 P57088 TMEM33Q6PJG6 BRAT1 P57740 NUP107 Q8IZ69 TRMT2AQ8WWC4 C2orf47 Q92621 NUP205 Q14166 TTLL12O95674 CDS2 Q15084 PDIA6 Q9BSJ2 TUBGCP2O76071 CIAO1 P53350 PLK1 Q96RT7 TUBGCP6Q9UBF2 COPG2 Q9NRX1 PNO1 Q8NBM4 UBAC2P31689 DNAJA1 Q16537 PPP2R5E Q05086 UBE3AO60762 DPM1 P49642 PRIM1 P31930 UQCRC1Q5VYK3 ECM29 Q8WWY3 PRPF31 O94782 USP1P18074 ERCC2 Q86UA1 PRPF39 Q53GS9 USP39Q969X5 ERGIC1 P28066 PSMA5 P45974 USP5Q96CS3 FAF2 P60900 PSMA6 Q9UID3 VPS51Q14C86 GAPVD1 P11216 PYGB Q9H4A3 WNK1Q12789 GTF3C1 Q9H2M9 RAB3GAP2 Q8IZH2 XRN1Q7Z4H7 HAUS6 Q14257 RCN2 Q969M3 YIPF5O75146 HIP1R P40429 RPL13A Q96TA2 YME1L1Q8TCT9 HM13 P62269 RPS18 Q5VZL5 ZMYM4Q14103 HNRNPD Q71UM5 RPS27LQ8TEX9 IPO4 O95248 SBF1Q13303 KCNAB2 P43007 SLC1A4P52294 KPNA1 Q02978 SLC25A11O43813 LANCL1 O75746 SLC25A12P50851 LRBA Q6P1M0 SLC27A4Q13257 MAD2L1 P11166 SLC2A1Q9UNF1 MAGED2 Q9NXE4 SMPD4PKCθAS Exp# 2+3+4 204 Appendix D. Proteins identified in PKCθAPEX experiment using biotin phenol as a substrate Protein species form the PKCθAPEX experiment that used biotin phenol as a substrate were separated into found to be enriched in a PMA “activated” state, enriched in PMA “un-activated” state or found in “both” the PMA un-active and active states. Table D.1 Proteins identified to be enriched in PMA “un-activated” state UniProtID Gene name UniProtID Gene name UniProtID Gene nameQ5VUY0 AADACL3 Q6ZS17 FAM65A P62140 PPP1CBQ562R1 ACTBL2 Q75N90 FBN3 Q14498 RBM39Q99996 AKAP9 Q9NWN3 FBXO34 O43166 SIPA1L1Q3KRA9 ALKBH6 Q9BYG8 GSDMC P53814 SMTNQ96Q42 ALS2 Q8TF76 GSG2 P10124 SRGNQ9NRW3 APOBEC3C P09651 HNRNPA1 Q13148 TARDBPP40616 ARL1 Q53GQ0 HSD17B12 P34981 TRHRQ96GX2 ATXN7L3B P16871 IL7R Q8WZ42 TTNQ10589 BST2 P08514 ITGA2B P26368 U2AF2E9PRG8 C11orf98 O60229 KALRN Q9Y2K6 USP20Q8N5I9 C12orf45 O00505 KPNA3 Q9P2S5 WRAP73Q5SQH8 C6orf136 Q8WWI1 LMO7 Q96KR1 ZFRQ6ZV77 C9orf139 P06858 LPL Q9UBW7 ZMYM2O14936 CASK O75197 LRP5Q9P219 CCDC88C P50579 METAP2P51684 CCR6 Q86YT6 MIB1P19256 CD58 P39900 MMP12O14578 CIT Q9Y483 MTF2Q16643 DBN1 P42345 MTORQ9UJW0 DCTN4 Q9Y2K3 MYH15Q92841 DDX17 A7E2Y1 MYH7BQ7L2E3 DHX30 P43490 NAMPTQ8WXX0 DNAH7 P20929 NEBQ9ULA0 DNPEP Q6X4W1 NSMFP23588 EIF4B Q9NPF4 OSGEPQ8IUD2 ERC1 Q9BRX2 PELOA0FGR8 ESYT2 O15067 PFASP14921 ETS1 P08237 PFKMO60447 EVI5 Q9H307 PNNQ01844 EWSR1 Q8NA72 POC5"Unactivated" state 205 Table D.2 Proteins identified to be enriched in PMA “activated” state UniProtID Gene name UniProtID Gene name UniProtID Gene nameQ15057 ACAP2 Q16566 CAMK4 Q8NE01 CNNM3P53396 ACLY Q86VP6 CAND1 A5YKK6 CNOT1P55265 ADAR P49589 CARS Q5TZA2 CROCCP35611 ADD1 P22681 CBL O75534 CSDE1Q9UEY8 ADD3 Q9P2K1 CC2D2A O15320 CTAGE5Q8N7X0 ADGB Q6ZN84 CCDC81 Q7Z7A3 CTU1Q9H2P0 ADNP Q3V6T2 CCDC88A Q96F07 CYFIP2P25098 ADRBK1 P14635 CCNB1 O43293 DAPK3O95831 AIFM1 Q8WWL7 CCNB3 P27707 DCKO00170 AIP O43303 CCP110 Q9GZR7 DDX24Q12802 AKAP13 P10747 CD28 P17844 DDX5P54886 ALDH18A1 P04234 CD3D P35659 DEKQ8IWZ3 ANKHD1 P09326 CD48 Q8IWF6 DENND6AO75179 ANKRD17 P06127 CD5 O60610 DIAPH1Q9NQW6 ANLN P30203 CD6 Q9UPY3 DICER1O95782 AP2A1 P60033 CD81 Q12959 DLG1Q13367 AP3B2 P27701 CD82 Q96JB1 DNAH8O14617 AP3D1 Q12834 CDC20 O75190 DNAJB6Q7Z5R6 APBB1IP Q16543 CDC37 Q5F1R6 DNAJC21P10398 ARAF P24941 CDK2 P26358 DNMT1Q7Z6I6 ARHGAP30 Q00534 CDK6 Q96N67 DOCK7P52566 ARHGDIB Q9NXV6 CDKN2AIP Q9Y295 DRG1P05023 ATP1A1 P49454 CENPF Q9NZJ0 DTLP36542 ATP5C1 Q5SW79 CEP170 Q8TDB6 DTX3LQ9UI12 ATP6V1H Q9UHD1 CHORDC1 Q6P2E9 EDC4Q9NYF8 BCLAF1 Q8WWK9 CKAP2 P57772 EEFSECP11274 BCR Q14008 CKAP5 Q14156 EFR3AQ8NFC6 BOD1L1 Q7Z460 CLASP1 Q7Z2Z2 EFTUD1O43684 BUB3 O75122 CLASP2 Q9UI10 EIF2B4P02746 C1QB Q9NY35 CLDND1 O15371 EIF3D"Activated" state (Table continued on next page) 206 Table D.2 UniProtID Gene name UniProtID Gene name UniProtID Gene nameQ9Y262 EIF3L Q08378 GOLGA3 Q8NHM5 KDM2BO60841 EIF5B Q14789 GOLGB1 A2VDJ0 KIAA0922P0C7U0 ELFN1 Q9NQX3 GPHN Q8TCG1 KIAA1524P50402 EMD P15170 GSPT1 Q15058 KIF14Q9HC35 EML4 P00390 GSR Q9NS87 KIF15P11171 EPB41 Q8N3Y3 GYLTL1B Q86Y91 KIF18BO43491 EPB41L2 Q92522 H1FX O95239 KIF4AQ9HCM4 EPB41L5 O14929 HAT1 P52294 KPNA1O15197 EPHB6 O94927 HAUS5 Q6GTX8 LAIR1P42566 EPS15 P51610 HCFC1 Q6PKG0 LARP1O75477 ERLIN1 Q00341 HDLBP P07195 LDHBP62495 ETF1 Q15751 HERC1 Q15334 LLGL1Q96A65 EXOC4 O75146 HIP1R Q8NF37 LPCAT1P15311 EZR Q14527 HLTF Q6P1A2 LPCAT3P00451 F8 P52597 HNRNPF Q12912 LRMPQ86UX7 FERMT3 O43390 HNRNPR Q15345 LRRC41Q13642 FHL1 Q5SSJ5 HP1BP3 Q96NW7 LRRC7Q5T1M5 FKBP15 Q92598 HSPH1 Q32MZ4 LRRFIP1Q02790 FKBP4 Q7Z6Z7 HUWE1 Q9UNF1 MAGED2Q14315 FLNC P08069 IGF1R P46734 MAP2K3O95466 FMNL1 O95163 IKBKAP P27816 MAP4Q8IVF7 FMNL3 P06213 INSR P27448 MARK3Q96RU3 FNBP1 Q9H095 IQCG P56192 MARSO15117 FYB P13612 ITGA4 P31153 MAT2AP11413 G6PD P05556 ITGB1 Q8N4C8 MINK1Q14C86 GAPVD1 P23458 JAK1 Q9Y6F6 MRVI1P50395 GDI2 Q5T7N3 KANK4 Q765P7 MTSS1LP57678 GEMIN4 Q15046 KARS P20591 MX1Q8TEQ6 GEMIN5 Q8IYT4 KATNAL2 Q13459 MYO9BQ14344 GNA13 O94819 KBTBD11 Q9BXJ9 NAA15"Activated" state(Table continued on next page) 207 Table D.2 UniProtID Gene name UniProtID Gene name UniProtID Gene nameQ9BPX3 NCAPG Q9NRX1 PNO1 Q14644 RASA3Q15003 NCAPH P24928 POLR2A P06400 RB1P55160 NCKAP1L P30876 POLR2B P35250 RFC2O75376 NCOR1 Q06203 PPAT P40938 RFC3Q8NHV4 NEDD1 Q13136 PPFIA1 P84095 RHOGQ8TD19 NEK9 Q9BZL4 PPP1R12C Q5UIP0 RIF1O75161 NPHP4 P67775 PPP2CA Q13464 ROCK1Q9Y266 NUDC Q08209 PPP3CA Q8IXW5 RPAP2Q14980 NUMA1 Q9UPN7 PPP6R1 Q92600 RQCD1Q8TEM1 NUP210 Q5H9R7 PPP6R3 P23921 RRM1Q96CV9 OPTN Q13131 PRKAA1 Q96T51 RUFY1Q9BXB4 OSBPL11 P54619 PRKAG1 Q9Y265 RUVBL1Q9H0X9 OSBPL5 P24723 PRKCH Q9Y230 RUVBL2P11940 PABPC1 Q04759 PRKCQ Q9BXA9 SALL3Q9NVE7 PANK4 Q9BZL6 PRKD2 O43290 SART1Q96RG2 PASK O94806 PRKD3 O95248 SBF1Q9HC56 PCDH9 O94906 PRPF6 O15027 SEC16AQ16822 PCK2 O75475 PSIP1 Q15436 SEC23AQ15116 PDCD1 Q06323 PSME1 O95487 SEC24BQ29RF7 PDS5A Q9UL46 PSME2 P61619 SEC61A1P52209 PGD Q13308 PTK7 Q96T21 SECISBP2Q9C0D0 PHACTR1 Q06124 PTPN11 Q15019 SEPT2P27986 PIK3R1 P17706 PTPN2 Q16181 SEPT7Q99755 PIP5K1A Q9H0U4 RAB1B Q9UHD8 SEPT9P19174 PLCG1 Q92930 RAB8B O75533 SF3B1Q6IQ23 PLEKHA7 O60216 RAD21 Q9Y3P8 SIT1Q8IWE5 PLEKHM2 Q92878 RAD50 P41440 SLC19A1O60664 PLIN3 P11233 RALA P43007 SLC1A4P53350 PLK1 Q8TEU7 RAPGEF6 O75746 SLC25A12Q14651 PLS1 Q15283 RASA2 Q7Z769 SLC35E3"Activated" state(Table continued on next page) 208 Table D.2 UniProtID Gene name UniProtID Gene name UniProtID Gene nameQ96QD8 SLC38A2 Q9Y2W1 THRAP3 Q4G0F5 VPS26BP08195 SLC3A2 Q9UDY2 TJP2 Q7Z5K2 WAPALQ9Y6M7 SLC4A7 P04183 TK1 O43379 WDR62P31641 SLC6A6 Q9BRN9 TM2D3 Q9UIA9 XPO7Q01650 SLC7A5 P42166 TMPO P13010 XRCC5Q9H2G2 SLK Q92973 TNPO1 P54577 YARSP51532 SMARCA4 Q9Y5L0 TNPO3 P16989 YBX3O60264 SMARCA5 P11388 TOP2A Q04917 YWHAHQ92925 SMARCD2 Q8N9V7 TOPAZ1 P43403 ZAP70Q9NTJ3 SMC4 Q12888 TP53BP1 Q7Z2W4 ZC3HAV1O75643 SNRNP200 P29144 TPP2 Q96H79 ZC3HAV1LQ96L92 SNX27 Q9H4I3 TRABD P0C7V5 ZNF812Q96R06 SPAG5 Q6PIZ9 TRAT1O60271 SPAG9 O75962 TRIOQ8N0X7 SPG20 Q15645 TRIP13O76094 SRP72 Q9NXH9 TRMT1Q96SB4 SRPK1 Q9BQE3 TUBA1CP78362 SRPK2 O15042 U2SURPQ9UQ35 SRRM2 A0AVT1 UBA6Q76I76 SSH2 Q05086 UBE3AP43307 SSR1 Q14139 UBE4AO60284 ST18 Q9NWZ5 UCKL1Q9P2P6 STARD9 Q96T88 UHRF1O94901 SUN1 Q9HAU5 UPF2Q5T011 SZT2 Q9BZI7 UPF3BQ9H2K8 TAOK3 P51784 USP11P26639 TARS P54578 USP14Q9BXI6 TBC1D10A Q93008 USP9XQ8TC07 TBC1D15 Q8TAA9 VANGL1Q9UHD2 TBK1 P15498 VAV1"Activated" state 209 Table D.3 Proteins identified to be enriched in “both” PMA activated and un-activated states UniProtID Gene name UniProtID Gene name UniProtID Gene nameP61221 ABCE1 P55884 EIF3B Q9BW19 KIFC1Q8NE71 ABCF1 Q13347 EIF3I P52292 KPNA2Q9UG63 ABCF2 P38919 EIF4A3 P06239 LCKQ96SZ5 ADO Q15717 ELAVL1 P13796 LCP1P30566 ADSL Q92556 ELMO1 Q9Y2L9 LRCH1O95433 AHSA1 Q9UI08 EVL Q9Y383 LUC7L2Q9NWB6 ARGLU1 Q15024 EXOSC7 O95232 LUC7L3P98171 ARHGAP4 Q9NZB2 FAM120A Q86V48 LUZP1Q92888 ARHGEF1 P49327 FASN P78559 MAP1AQ92974 ARHGEF2 P51114 FXR1 P51608 MECP2Q93084 ATP2A3 Q9BQS8 FYCO1 P46013 MKI67Q9UBB4 ATXN10 P06241 FYN Q9BU76 MMTAG2P35520 CBS Q92616 GCN1L1 Q86UE4 MTDHQ96JN2 CCDC136 Q9Y2X7 GIT1 Q9NVV4 MTPAPP08174 CD55 P49915 GMPS Q8IUG5 MYO18BP06493 CDK1 P08754 GNAI3 Q15021 NCAPD2Q14839 CHD4 P78347 GTF2I Q08J23 NSUN2Q8NFW8 CMAS Q7Z4H7 HAUS6 Q9BXS6 NUSAP1P53618 COPB1 P26583 HMGB2 Q96R67 OR4C12Q9Y678 COPG1 P52272 HNRNPM O95747 OXSR1Q13616 CUL1 P41252 IARS O95613 PCNTP81605 DCD Q27J81 INF2 Q9BUL8 PDCD10Q9UHI6 DDX20 Q92835 INPP5D Q53EL6 PDCD4O00148 DDX39A O95373 IPO7 P42356 PI4KAO00571 DDX3X Q14571 ITPR2 Q99569 PKP4O43143 DHX15 P52732 KIF11 P0CG38 POTEIQ9P265 DIP2B Q02241 KIF23 O60237 PPP1R12BQ6XZF7 DNMBP O00139 KIF2A P17252 PRKCAQ92608 DOCK2 Q99661 KIF2C Q02156 PRKCEP05198 EIF2S1 P33176 KIF5B O14744 PRMT5"Both" state(Table continued on next page) 210 Table D.3 UniProtID Gene nameQ9Y520 PRRC2CO43242 PSMD3P49792 RANBP2P61225 RAP2BQ86YV0 RASAL3Q96T37 RBM15Q15382 RHEBQ63HN8 RNF213P31350 RRM2Q7L099 RUFY3P53992 SEC24CP53985 SLC16A1Q9UBX3 SLC25A10Q14683 SMC1AO95347 SMC2Q9UQE7 SMC3Q8NA61 SPERTQ13813 SPTAN1Q8IYB3 SRRM1Q9BXP5 SRRTO60506 SYNCRIPP42167 TMPOQ14258 TRIM25Q15642 TRIP10Q01081 U2AF1Q14694 USP10Q9H4A3 WNK1O14980 XPO1O43592 XPOTQ9H0D6 XRN2"Both" state 211 Appendix E. PKCθ regulation of IGF1R signalling via IRS1 repression Introduction Insulin Receptor Substrate 1 (IRS1) is a member of a family of adaptor proteins that mediate signalling (namely PI3K/AKT) downstream of IGF1R/IR [209]. As described in Chapter 3, signalling downstream of IGF1R requires IRS binding in order to activate PI3K/AKT and maintain growth in a subset of T-ALL cell lines. This suggests that inhibition of IRS, like IGF1R, might reduce growth in IGF1R inhibitor sensitive cell lines. IRS proteins are large adaptor proteins that, in addition to having key tyrosine residues that are phosphorylated to propagate positive downstream signalling, also contain numerous serine/threonine sites which, upon phosphorylation, have been described to negatively impact IGF/insulin signalling [237]. Dissociation of IRSs from IR/IGF1R and IRS degradation are thought to be two of the main mechanisms that mediate this attenuation of signalling. Numerous IGF/insulin-stimulated kinases (atypical PKC, AKT, SIK2, mTOR, S6K1, ERK1/2 and ROCK1) and IGF/insulin independent serine/threonine kinases (GRK2, novel and conventional PKCs, JNK, IKKβ, mPLK) have been implicated in IRS phosphorylation. Of particular interest to our group is the suggestion that PKCθ can phosphorylate IRS1 and repress insulin signalling [236]. This interaction has been mainly reported thus far in skeletal muscle, where it is described as a mechanism for free fatty acid induced insulin resistance. This is highlighted by the fact that prkcq knockout mice are protected from this phenotype [366]. Given this reported function and our interest in the role PKCθ plays in T-ALL I looked to investigate the impact PKCθ has on IGF1R signalling in this context. Results and Discussion PKCθ repressors IGF1R mediated PI3K/AKT activation Increasing PKCθ activity through the induction of constitutively active PKCθ (PKCθCA) reduced pAKT levels under both steady state growth conditions and after serum starvation/IGF pulse in an IGF1R inhibitor sensitive T-ALL cell line (HPBALL) (Figure E.1A/1B). This reduction in PI3K/AKT signalling was accompanied by a marginal decrease in relative cells numbers after 3 days of PKCθCA induction, however this result was not consistently seen (Figure E.1C). Of note, complete IGF1R or pan PI3K 212 inhibition repressed growth more than PKCθCA induction, but combination of the two did not reduce growth more than IGF1R/PI3K inhibition alone, further suggesting that PKCθ may be acting through the IGF1R/IRS1/PI3K/AKT pathway. The cell line used for these assays, HPBALL, does not express PTEN and is IGF1R inhibitor sensitive. Figure E.1 PKCθ reduces PI3K/AKT activation via IGF in an IGF1R inhibitor sensitive T-ALL cell line Flow cytometric analysis of AKT activation in HPBALL cells. 1 μg/mL doxycycline for 2 days was used to induce expression of constitutively active PKCθ and intracellular flow cytometry for pAKT (S473) was used to assess PI3K/AKT activation at either (A) steady state or (B) following 24 hour serum starvation and 10 minute 10 ng/mL IGF-1 pulses. (C) Comparison of relative normalized cell numbers of PKCθCA inducible HPBALL cells as measured by resazurin reduction assay after 3 days of in vitro culture with 1 μg/mL S te a d y S ta te0 .00 .20 .40 .60 .81 .0m o c k d o x y c y c lin eNormalized pAKT levelS e r u m S ta r v e d + IG F -1 p u ls e0 .00 .20 .40 .60 .81 .0m o c k d o x y c y c lin eNormalized pAKT levelmo ckC P7 51 ,8 71B MS -7 54 80 7L y2 94 00 20 .00 .20 .40 .60 .81 .01 .2m o c kd o x y c y c lin eNormalized relative fluorescencen s n s n s**(A) (B) (C) 213 doxycycline. Cells also grown under IGF1R inhibiting (1 μg/mL CP-751,871 or 0.2 μM BMS-754807) or pan-PI3K inhibiting (5 μM LY294002) conditions. Normalized to mock treated control. Error bars indicate standard deviation of assays performed in triplicate. ** = p < 0.01; ns = not significant. (Multiple comparison 2-way ANOVA). PKCθ activation decreases IRS1 expression Surface IGF1R levels were unaffected by PKCθCA induction, suggesting involvement further down the pathway (Figure E.2A). IRS1 total protein levels in PKCθCA induced cells, however, were about half of non-induced cells, suggesting PKCθ may be repressing IGF1R signalling at the IRS1 level, possibly via degradation (Figure E.2B). PKCθ may be directly phosphorylating IRS1 as has been reported in the literature in myocytes (Ser1101) [236], or possibly indirectly via the numerous other kinases above, many of which PKCθ has been previously reported to impact. Of note, other PKC isoforms (PKCδ and PKC ζ) have been reported to directly phosphorylate IRS1 [493, 494]. It is tempting to speculate the importance of PKCθ in repressing IGF1R signalling in T-ALL may extend outside the bulk cell context and also impact on LIC activity. Such a hypothesis is based on two published findings from our lab that both reduced IGF1R [92] as well as increased PKCθ [304] levels can negatively impact the ability of T-ALL leukemia from transplanting into secondary recipients. 214 IRS1PKCθ/PKCθCAmock doxycyclineHPBALL TetON hPKCθCAActin0.5811 2.09sIGF1RHPBALL TetON hPKCθCAmockdoxycycline Figure E.2 PKCθ does not change sIGF1R expression but can downregulate IRS1 in T-ALL Flow cytometric and Western Blot assessment of surface IGF1R and IRS1 expression levels in HPBALL cells. Expression of constitutively active PKCθ used induced by the addition of 1 μg/mL doxycycline for 2 days. (A) Surface expression of IGF1R as measured by flow cytometry. (B) IRS1 total protein expression as measure by Western Blot. Total PKCθ relative to non-induced cells shown. Numbers indicate densitometric band quantitation with normalization to β-actin loading control and expressed as fold-change over respective non-induced cells. Appendix F. PKCθ regulation of CXCR4 signalling Introduction CXCR4 signalling CXC chemokine receptor 4 (CXCR4) is a member of the CXC subfamily of GPCR chemokine receptors that bind CXC chemokines [495-497]. Chemokines are characterized by the spacing between the N terminal cysteine amino acids that form disulfide bonds with C terminal cysteines to maintain tertiary structure. The N terminal cysteines of CXC chemokine are spaced by one amino acid. CXC chemokine ligand (A) (B) 215 (CXCL)12, also known as stromal cell-derived factor (SDF)-1 is the ligand for CXCR4 and was until recently considered the sole receptor for CXCL12. Subsequent studies, however, have shown that CXCR7 can also bind CXCL12 [498] and has been suggested to act as a decoy receptor that scavenges CXCL12 away from CXCR4 [499]. However, the true function of CXCL12 binding CXCR7 is still a contentious topic in the field, with reports of relevant downstream signalling occurring in various cell contexts in response to ligand. Canonical CXCR4 downstream signalling, like all GPCR signalling, involves the activation of heterotrimeric G protein complexes upon ligand binding leading to the dissociation into Gα and Gβγ subunits. Gαi is thought to be that main Gα subunit activated downstream of CXCR4 and is described to activate MAPK (p38 and ERK1/2) [500] and PI3K (p110γ)/AKT [501] signalling, whereas the Gβγ subunit activates PLC/PKC signalling [502]. All three of these downstream signalling arms have been reported to play a role in mediating chemotactic activity downstream of CXCR4. G protein complex independent JAK/STAT signalling has also been reported downstream of CXCR4 [503]. Upon ligand binding, CXCR4 also activates G-protein receptor kinases (GRK) (and possibly PKCs) that phosphorylate CXCR4 and act to desensitize signalling by causing its internalization via β-arrestin/clathrin-mediated endocytosis [479, 480]. CXCR4 function Initially, CXCR4 was described as a chemokine receptor highly expressed in various leukocytes subsets and as a co-factor for human immunomodulatory virus (HIV) entry into T-cells [504]. However, its important role in numerous normal physiological processes has now been described. These include, but are not limited to, neurogenesis [505], vascular formation [506], germ cell development [507], immune cell development [508] and hematopoietic stem cell (HSC)/bone marrow niche interactions [509]. Most notably, mouse knockout experiments have shown that CXCR4 and CXCL12 are essential for homing of HSC and early progenitor cells to the bone marrow [510, 511]. With specific regard to T-cells, development beyond β-selection has been shown to require PI3K activation downstream of CXCR4 [512]. In addition, CXCR4 associates with TCR complexes and acts to co-stimulate downstream signalling [513, 514]. 216 CXCR4 in T-ALL Chemokine receptors signalling have previously been reported to play a role in T-ALL by allowing infiltration into the central nervous system (CCR7), a clinically relevant site [99]. In addition, it has been shown that the expression of CCR7, as well as other chemokine receptors (CCR5 and CCR9), is positively regulated by oncogenic NOTCH1 in T-ALL [98, 99]. CXCR4, although not reported to not be transcriptionally regulated by NOTCH1, does appear to be very important in T-ALL disease maintenance. Recently, two papers have highlighted this by showing that SDF1 released from the bone marrow stroma maintains T-ALL and that genetic or pharmalogical perturbation of CXCR4 results in the loss of leukemia initiating cell activity [515, 516]. This suggests that finding mechanisms by which T-ALL can negatively regulate CXCR4 signalling may prove effective in mitigating disease burden. Results and Discussion PKCθ decreases surface expression of CXCR4 and downstream signalling We have previously published the observation that enforced PKCθ expression in both mouse and human primary T-ALLs compromises leukemia initiating cell activity [304]. In an attempt to explain this phenomenon, I investigated the role PKCθ might play in CXCR4 mediated migration. When expression of a constitutively active version of PKCθ (PKCθCA) was induced, we observe a reduction in surface CXCR4 expression (Figure F.1A). Conversely, with PKCθ knockdown (Figure F.1B), a mild increase in sCXCR4 expression is seen (Figure F.1C). This phenotype becomes more prevalent after SDF1α ligand exposure, where a reduction in surface CXCR4 after 10 minutes in control cells but not in PKCθ knockdown cells is seen. After longer ligand exposure, this effect is lost, suggesting that while PKCθ may play a role in the initial internalization of CXCR4, other factors may be involved. The decrease in surface CXCR4 seen with increased PKC activity also corresponded to the level of downstream MAPK activation that was achievable after ligand stimulation whereby inducing PKCθCA expression reduced the phosphorylation of ERK in response to SDF1α (Figure F.2A/2B). 217 ActinPKCθParental scramble shPKCθ54sCXCR4- mock+ doxycycline Figure F.1 PKCθ regulates the surface expression of CXCR4 Flow cytometric analysis of surface CXCR4 in HPBALL cells. (A) Steady state CXCR4 surface expression levels were measured after doxycycline induced expression of constitutively active PKCθ (red), 1 μg/mL doxycycline for 2 days. (B) Western Blot analysis of PKCθ with antibodies directed against PKCθ and β-actin as a loading control show efficient lentiviral knockdown of endogenous PKCθ. (C) CXCR4 surface expression levels in knockdown or non-silencing control cells were measured overtime after SDF1α ligand exposure. 0 1 0 2 0 3 0 4 0 5 0 6 00 .00 .20 .40 .60 .81 .01 .2M in u te s p o s t S D FNormalized mean sCXCR4 levelS c ra m b les h P K C θ 54(A) (B) (C) 218 sCXCR4- mock+ doxycycline Figure F.2 PKCθ induced surface CXCR4 decrease reduces downstream ligand stimulated signalling Flow cytometric assessment of downstream ERK activation in HPBALL cells. (A) Surface CXCR4 expression after doxycycline induced expression of constitutively active PKCθ expression (red), 1 μg/mL doxycycline for 2 days. Cells with reduced surface CXCR4 indicated (“CXCR4 low”). (B) 100 ng/mL SDF1α for 10 minutes induced activation of ERK as assessed by Thr202/Tyr204 phosphorylation. CXCR4 promotes ligand and stromal cell induced migration of T-ALL cells From these observations, I hypothesized that PKCθ might reduce CXCR4 mediated migration/homing by decreasing sCXCR4 expression, which in turn might contribute to the compromised LIC activity with PKCθ overexpression. This idea was further supported by recent publications [515, 516] which describe both the expression of CXCR4 by T-ALL cells as well as CXCL12 by the vascular endothelial cells as being essential to LIC activity. This is consistent with our own preliminary work showing knockdown of CXCR4 caused a human T-ALL cell line (PF382) to be out competed by untransduced cells when transplanted into immunocompromised mice (data not shown; courtesy of Dr. Vincenzo Giambra). To test this, I decided to perform Transwell migration assays to see if the reduction in sCXCR4 caused by PKCθ would impede SDF1 mediated migration. As a control lentiviral shRNA knockdown of sCXCR4 impeded migration toward SDF1 (as well as the SDF1 producing murine feeder cell line (A) (B) Normalized pERK level012345m o c kS D F 1αm o c k d o x y c y c l in eCXCR4 "high"CXCR4 "low"219 MS5) was able to be shown (Figure F.3A/3B). When a competitive migration assay between PKCθCA induced and un-induced cells was performed, however, I saw that despite reduced sCXCR4 levels, migration towards SDF1 was actually increased after PKCθCA induction (Figure F.4). PKCθ having a positive effect on cell motility is also supported in the literature, and various PKC isoforms, including PKCθ, have been described to activated downstream of CXCR4 and positively contribute to cell migration [353, 517]. In T cells, this PKCθ mediated SDF1/CXCR4 migration is thought to be PI3K/AKT independent [353]. My own observations show that SDF1 can elicit both AKT and ERK phosphorylation in T-ALL, though AKT is rapidly (~20 minutes) turned off (data not shown). The observation that PKCθ can downregulate sCXCR4 may therefore represent a negative feedback loop whereby in addition to promoting cell migration towards an SDF1 gradient, PKCθ also dampens the response to ligand. PKCs have long been implicated in CXCR4 internalization through either direct phosphorylation or indirect activation of GRKs; however this is the first report of this function by PKCθ. UnstainedParentalNon-silencingshCXCR4sCXCR4 Figure F.3 T ra n s w e ll M ig r a t io n A s s a yRelative fraction ofshCXCR4(NGFR+) cellsN on -s ils hC XC R4N on -s ils hC XC R40 .00 .20 .40 .60 .81 .0S D F 1 α M S 5 fe e d e rs(A) (B) 220 Figure F.3 A decrease in surface CXCR4 expression reduces migration towards SDF ligand and stromal feeders Transwell migration assay on HPBALL cells. (A) CXCR4 expression was knocked down using a lentiviral vector with a NGFR marker (red). (B) A mixed population of CXCR4 knockdown or non-silencing control (NGFR+) and parental cells were loaded into the upper chamber of a transwell plate and allowed to migrate towards 50 nM SDF1 or MS5 feeder cells in the lower chamber. Change in fraction of NGFR+ cells migrating to lower chamber relative to input indicated. Figure F.4 PKCθ activity increases migration towards SDF ligand Transwell migration assay on HPBALL cells. Doxycycline induced expression of constitutively active PKCθ cells were labelled with CFSE (CFSE positive/+PKCθCA) mixed with non induced cells (CFSE negative/-PKCθCA) and loaded into the upper chamber of a transwell plate. Varying concentration of SDF1α ligand was added to lower chamber and cells were allowed to migrate. Percentage of CFSE positive cells in lower chamber after migration indicated. 0 1 0 5 0 2 5002 55 07 51 0 0C F S E p o s (+ P K C θ C A )C F S E n e g ( -P K C θ C A ) S D F 1 α (n g /m L )% CFSE positive/negativein lower chamber