SUPPRESSION OF NEUROBLASTOMA TUMORIGENESIS USING ENU MUTAGENESIS IN THE TH-MYCN MOUSE MODEL OF NEUROBLASTOMA
A thesis submitted to the University of New South Wales in fulfilment of the requirements for the degree of Doctor of Philosophy
Jayne Murray
Children’s Cancer Institute and School of Women’s and Children’s Health Faculty of Medicine
University of New South Wales
December 2017
i THESIS/DISSERTION SHEET
PLEASE TYPE THE UNIVERSITY OF NEW SOUTH WALES Thesis/Dissertation Sheet
Surname or Family name: Murray
First name: Jayne Other name/s: Elizabeth
Abbreviation for degree as given in the University calendar: PhD
School: School of Women’s and Children’s Health Faculty: Faculty of Medicine
Title: Suppression of neuroblastoma tumorigenesis using ENU mutagenesis in the Th-MYCN mouse model of neuroblastoma
Abstract 350 words maximum: (PLEASE TYPE) Neuroblastoma, a disease of the neural crest and the most common solid tumour of infancy, accounts for 7-10% of all childhood cancer. The disease arises from primitive cells within the neural crest, and can develop at any point along the sympathetic nervous system, including sites within the abdomen, pelvis, neck and thoracic cavity. Neuroblastoma tumours with MYCN oncogene amplification are highly aggressive, leading to a dismal outcome in most patients. A mouse model of neuroblastoma previously generated by targeting the human MYCN oncogene to neural crest cells (Th-MYCN), leads to 100% of MYCN homozygous mice developing tumours at an early age. This thesis describes a forward genetic screen involving the use of N-ethyl-N-nitrosourea (ENU) to generate heritable mutations causing delayed tumorigenesis in this model, as well as genomic sequencing to identify the mutated gene(s) responsible for the delayed tumorigenesis observed.
Of 1716 ENU progeny screened, two mice demonstrated an unprecedented delay in tumour development. Subsequent breeding generated lines from these founders in which the delayed phenotypes were inherited in a dominant Mendelian fashion. Mapping, and whole exome and genome sequencing, identified a single gene in each of the lines as strong candidates responsible for delayed tumour formation, namely Runt-related transcription factor-1; translocated-to-1 (Runx1t1) and RING finger protein 121 (Rnf121). Runx1t1 was investigated further. Confirming the ENU mutagenesis findings, crossing of Runx1t1 knock-out mice with Th-MYCN mice, showed that loss of one allele of Runx1t1 abrogated tumour formation. The mechanism by which mutant Runx1t1 prevents neuroblastoma tumorigenesis was found to involve inhibition of neuroblast hyperplasia in mouse ganglia, a known prerequisite for neuroblastoma initiation. There was little evidence of MYCN directly regulating RUNX1T1 transcription, but rather increased RUNX1T1 protein levels were observed in human MYCN-amplified tumours. Knock-down of RUNX1T1 using shRNA resulted in decreased neuroblastoma colony formation in vitro, suggesting that this gene may also have a role in neuroblastoma maintenance.
Collectively this study has employed ENU mutagenesis in an in vivo forward genetic screen to identify RUNX1T1 and RNF121 as novel neuroblastoma-associated genes. These genes represent potential new therapeutic targets for the treatment and ultimately prevention of this disease.
ii Declaration relating to disposition of project thesis/dissertation
I hereby grant to the University of New South Wales or its agents the right to archive and to make available my thesis or dissertation in whole or in part in the University libraries in all forms of media, now or here after known, subject to the provisions of the Copyright Act 1968. I retain all property rights, such as patent rights. I also retain the right to use in future works (such as articles or books) all or part of this thesis or dissertation.
I also authorise University Microfilms to use the 350 word abstract of my thesis in Dissertation Abstracts International (this is applicable to doctoral theses only).
……………………………………………………… ……………………………………..………… ……….……………………...… Signature Witness Date
The University recognises that there may be exceptional circumstances requiring restrictions on copying or conditions on use. Requests for restriction for a period of up to 2 years must be made in writing. Requests for a longer period of restriction may be considered in exceptional circumstances and require the approval of the Dean of Graduate Research.
FOR OFFICE USE ONLY Date of completion of requirements for Award:
iii ORIGINALITY STATEMENT
‘I hereby declare that this submission is my own work and to the best of my knowledge it contains no materials previously published or written by another person, or substantial proportions of material which have been accepted for the award of any other degree or diploma at UNSW or any other educational institution, except where due acknowledgement is made in the thesis. Any contribution made to the research by others, with whom I have worked at UNSW or elsewhere, is explicitly acknowledged in the thesis. I also declare that the intellectual content of this thesis is the product of my own work, except to the extent that assistance from others in the project's design and conception or in style, presentation and linguistic expression is acknowledged.’
Signed ...... Date ...... 04052018
iv COPYRIGHT STATEMENT
‘I hereby grant the University of New South Wales or its agents the right to archive and to make available my thesis or dissertation in whole or part in the University libraries in all forms of media, now or here after known, subject to the provisions of the Copyright Act 1968. I retain all proprietary rights, such as patent rights. I also retain the right to use in future works (such as articles or books) all or part of this thesis or dissertation. I also authorise University Microfilms to use the 350 word abstract of my thesis in Dissertation Abstract International (this is applicable to doctoral theses only). I have either used no substantial portions of copyright material in my thesis or I have obtained permission to use copyright material; where permission has not been granted I have applied/will apply for a partial restriction of the digital copy of my thesis or dissertation.'
Signed ...... Date ...... 04052018
v AUTHENTICITY STATEMENT
‘I certify that the Library deposit digital copy is a direct equivalent of the final officially approved version of my thesis. No emendation of content has occurred and if there are any minor variations in formatting, they are the result of the conversion to digital format.’
Signed ...... Date ...... 04052018
vi ACKNOWLEDGEMENTS
Firstly I would like to thank my supervisors Professor Murray Norris and Professor Michelle Haber for entertaining the crazy idea of doing a part-time PhD while also working at Children’s Cancer Institute full-time as Program Officer for Molecular Diagnostics. I could not have done this without your supervision and guidance and allowing me time to write and do part of my PhD overseas in the laboratory of Professor Giovanni Perini. I cannot say thank you enough. Also, to my co-supervisor, Professor Douglas Hilton for suggesting the idea of doing an ENU mutagenesis study in the first place, and for performing the ENU injections.
A project of this scope and longevity could not have been done without the support of a number of people. To the ‘mouse team’; Ashleigh Clark, Hannah Webber, Michelle Ruhle, Tony Huynh, Georgina Eden, Sophie Allan, Aaminah Khan, Sara Sarraf and new recruits Steph Alfred and Jenn Brand, that have battled through the huge number of experiments that were part of my full-time job and supported me in this endeavour, allowing me time to do my PhD. Whilst some of you have come and gone through this journey, we remain in contact and I could not have gotten to this point without each, and every one of you. Dr Emanuele Valli, not only for your expertise in creating any construct you could ask for and help with experiments, but also for welcoming me in Bologna when I came to do part of my PhD there. I value your friendship enormously. Also to Professor Giovanni Perini and Dr Daniela Erriquez, for allowing me to come to your laboratory and for working with me on the fractionation portion of this thesis, thank you.
Thanks must also go to the collaborators that this project has collected over the years. To Dr Benjamin Kile and Dr Janelle Collinge, for your valuable insight into ENU mutagenesis, and for performing the exome and whole genome sequencing. To Dr Jamie Fletcher, for the protein modelling and for insightful scientific advice. To Amanda Russell and Professor Joel Mackay, for the NMR work and to Dr Andrew Gifford and Dr Laura Gamble for the histology analysis. To all the members of Experimental Therapeutics/Molecular Diagnostics, both past and present, thank you for providing a great working environment. To my panel members Professors Richard Lock and Maria Kavallaris and Dr Belamy Cheung, thank you for all your support.
vii
To Jessica Koach, Claudia Flemming, Joanna Keating, Leanna Cheung, Kathleen Kimpton and Tanya Dwarte: I could not have done this without your friendship, help, training, ideas and the occasional tap takeover. You all helped keep me sane during this period, whether it was just a chat over coffee or travelling to a great destination together, believe me, you helped.
Finally, to my family, that have put up with the constant weekend work or late nights. I promise, life will get back to some sense of normalcy from now on.
viii CONFERENCES, PUBLICATIONS AND AWARDS
Conferences: Oral Presentations Murray, J., Kile, B.T., Collinge, J.E., Marshall, G.M., Hilton, D.J., Haber, M. and Norris, M.D. “Suppression of neuroblastoma tumorigenesis using ENU mutagenesis in the Th-MYCN mouse model of neuroblastoma.” Oral Presentation at the Advances in Neuroblastoma Research meeting, Toronto, Canada, June 2012.
Murray, J., Haber, M., Carnegie-Clark, A., Webber, H., Marshall, G.M., Gurova, K., Burkhart, C., Purmal, A., Gudkov, A. and Norris, M.D. “CBL0137, an anti-cancer compound that simultaneously suppresses NFκB and activates p53, is highly effective in two independent mouse models of neuroblastoma.” Oral Presentation at the Australian Society for Medical Research NSW Annual Scientific Meeting, June 2013
Conferences: Poster Presentations: Murray, J., Valli, E., Troung, A., Eden, G., Allan, S., Webber, H., Tivnan, A., Tan, A., Flemming, C., Cheung, L., Hanssen, K., Carnegie-Clark, A., Ruhle, M., Henderson, M., Marshall, G.M., Yu, D., Norris, M.D., Fletcher, J.I. and Haber, M. “Supression of Multidrug resistance protein 4 inhibits neuroblastoma growth in vitro and in vivo.” Poster Presentation at the Advances in Neuroblastoma Research meeting, Cairns, Australia, June 2016.
Murray, J., Haber, M., Gamble, L., Carnegie-Clark, A., Webber, H., Ruhle, M., Gifford, A., Carter, D., Oberthuer, A., Fischer, M., Ziegler, D., Marshall, G.M., Gurova, K., Burkhart, C., Purmal, A., Gudkov, A. and Norris, M.D. “CBL0137, a novel NFkB suppressor and p53 activator, is highly effective in two independent pre-clinical mouse models of neuroblastoma.” Poster Presentation at The Lowy Cancer Symposium, Sydney, May 2015.
Awards
Postgraduate Research Support Scheme: Travel grant to attend the Advances in Neuroblastoma meeting, Toronto, 2012.
ix University of Western Sydney Prize for Best Student Oral Presentation, Australian Society for Medical Research, NSW Annual Scientific Meeting, 2013
Publications arising from this thesis Book Chapter: Cheung, L., Murray, J.E., Haber, M., and Norris, M.D. (2013). The MYCN Oncogene, Oncogene and Cancer - From Bench to Clinic, Dr. Yahwardiah Siregar (Ed.), InTech, DOI: 10.5772/54813. Available from: https://www.intechopen.com/books/oncogene- and-cancer-from-bench-to-clinic/the-mycn-oncogene
Publications arising during this thesis
Dagg, R.A., Pickett, H.A., Neumann, A.A., Napier, C.E., Henson, J.D., Teber, E.T., Arthur, J.W., Reynolds, C.P., Murray, J., Haber, M., Sobinoff, A.P., Lau, L.M.S., and Reddel, R.R. (2017). Extensive Proliferation of Human Cancer Cells with Ever-Shorter Telomeres. Cell Reports; 19, 2544-2556.
Carter, D.R., Sutton, S.K., Pajic, M., Murray, J., Sekyere, E.O., Fletcher, J., Beckers, A., De Preter, K., Speleman, F., George, R.E., Haber, M., Norris, M.D., Cheung, B.B., and Marshall, G.M. (2016). Glutathione biosynthesis is upregulated at the initiation of MYCN-driven neuroblastoma tumorigenesis. Molecular Oncology; 10, 866-878.
Decock, A., Ongenaert, M., Cannoodt, R., Verniers, K., De Wilde, B., Laureys, G., Van Roy, N., Berbegall, A.P., Bienertova-Vasku, J., Bown, N., Clement, N., Combaret, V., Haber, M., Hoyoux, C., Murray, J., Noguera, R., Pierron, G., Schleiermacher, G., Schulte, J.H., Stallings, R.L., Tweddle, D.A., De Preter, K., Speleman, F., and Vandesompele, J. (2016). Methyl-CpG-binding domain sequencing reveals a prognostic methylation signature in neuroblastoma. Oncotarget; 7, 1960-1972.
Evageliou, N.F., Haber, M., Vu, A., Laetsch, T.W., Murray, J., Gamble, L.D., Cheng, N.C., Liu, K., Reese, M., Corrigan, K.A., Ziegler, D.S., Webber, H., Hayes, C.S., Pawel, B., Marshall, G.M., Zhao, H., Gilmour, S.K., Norris, M.D., and Hogarty, M.D. (2016). Polyamine Antagonist Therapies Inhibit Neuroblastoma Initiation and Progression. Clinical Cancer Research; 22, 4391-4404.
x O'Brien, R., Tran, S.L., Maritz, M.F., Liu, B., Kong, C.F., Purgato, S., Yang, C., Murray, J., Russell, A.J., Flemming, C.L., von Jonquieres, G., Pickett, H.A., London, W.B., Haber, M., Gunaratne, P.H., Norris, M.D., Perini, G., Fletcher, J.I., and MacKenzie, K.L. (2016). MYC-Driven Neuroblastomas Are Addicted to a Telomerase- Independent Function of Dyskerin. Cancer Research; 76, 3604-3617.
Carter, D.R.*, Murray, J.*, Cheung, B.B., Gamble, L., Koach, J., Tsang, J., Sutton, S., Kalla, H., Syed, S., Gifford, A.J., Issaeva, N., Biktasova, A., Atmadibrata, B., Sun, Y., Sokolowski, N., Ling, D., Kim, P.Y., Webber, H., Clark, A., Ruhle, M., Liu, B., Oberthuer, A., Fischer, M., Byrne, J., Saletta, F., Thwe le, M., Purmal, A., Haderski, G., Burkhart, C., Speleman, F., De Preter, K., Beckers, A., Ziegler, D.S., Liu, T., Gurova, K.V., Gudkov, A.V., Norris, M.D., Haber, M., and Marshall, G.M. (2015). Therapeutic targeting of the MYC signal by inhibition of histone chaperone FACT in neuroblastoma. Science Translational Medicine; 7, 312ra176. * Equal first
Pasquier, E., Street, J., Pouchy, C., Carre, M., Gifford, A.J., Murray, J., Norris, M.D., Trahair, T., Andre, N., and Kavallaris, M. (2013). beta-blockers increase response to chemotherapy via direct antitumour and anti-angiogenic mechanisms in neuroblastoma. British Journal of Cancer; 108, 2485-2494.
Fletcher, J.I., Gherardi, S., Murray, J., Burkhart, C.A., Russell, A., Valli, E., Smith, J., Oberthuer, A., Ashton, L.J., London, W.B., Marshall, G.M., Norris, M.D., Perini, G., and Haber, M. (2012). N-Myc regulates expression of the detoxifying enzyme glutathione transferase GSTP1, a marker of poor outcome in neuroblastoma. Cancer Research; 72, 845-853.
Henderson, M.J., Haber, M., Porro, A., Munoz, M.A., Iraci, N., Xue, C., Murray, J., Flemming, C.L., Smith, J., Fletcher, J.I., Gherardi, S., Kwek, C.K., Russell, A.J., Valli, E., London, W.B., Buxton, A.B., Ashton, L.J., Sartorelli, A.C., Cohn, S.L., Schwab, M., Marshall, G.M., Perini, G., and Norris, M.D. (2011). ABCC multidrug transporters in childhood neuroblastoma: clinical and biological effects independent of cytotoxic drug efflux. Journal of the National Cancer Institute; 103, 1236-1251.
Pajic, M., Murray, J., Marshall, G.M., Cole, S.P., Norris, M.D., and Haber, M. (2011). ABCC1 G2012T single nucleotide polymorphism is associated with patient outcome in
xi primary neuroblastoma and altered stability of the ABCC1 gene transcript. Pharmacogenetics and Genomics; 21, 270-279.
Burkhart, C.A., Watt, F., Murray, J., Pajic, M., Prokvolit, A., Xue, C., Flemming, C., Smith, J., Purmal, A., Isachenko, N., Komarov, P.G., Gurova, K.V., Sartorelli, A.C., Marshall, G.M., Norris, M.D., Gudkov, A.V., and Haber, M. (2009). Small-molecule multidrug resistance-associated protein 1 inhibitor reversan increases the therapeutic index of chemotherapy in mouse models of neuroblastoma. Cancer Research; 69, 6573- 6580.
xii TABLE OF CONTENTS THESIS/DISSERTION SHEET ...... ii ORIGINALITY STATEMENT ...... iv COPYRIGHT STATEMENT ...... v AUTHENTICITY STATEMENT ...... vi ACKNOWLEDGEMENTS ...... vii CONFERENCES, PUBLICATIONS AND AWARDS ...... ix LIST OF FIGURES ...... xviii LIST OF TABLES ...... xxi ABBREVIATIONS AND ACRONYMS ...... xxiii ABSTRACT ...... xxviii CHAPTER 1 INTRODUCTION ...... 1 1.1 Neuroblastoma ...... 1 1.1.1 Incidence ...... 1 1.1.2 Staging ...... 2 1.1.3 Clinical Factors ...... 5 1.1.3.1 Age ...... 5 1.1.3.2 Chromosomal Gains and Losses ...... 6 1.1.3.3 Histology ...... 9 1.1.3.4 Ploidy ...... 11 1.1.3.5 MYCN Amplification ...... 11 1.1.3.6 Other Biological Factors ...... 12 1.1.4 Risk Stratification and Treatment Outcome ...... 13 1.2 The MYCN Oncogene ...... 16 1.2.1 Discovery and Structure of the MYC Family ...... 16 1.2.2 Activator and Repressor Functions of MYCN ...... 18 1.2.3 Tissue Expression of Myc Family Genes ...... 18 1.2.4 MYCN Dysregulation in Cancer ...... 20 1.3 Mouse Models of Neuroblastoma ...... 21 1.3.1 Early Mouse Models ...... 21 1.3.2 The Th-MYCN Mouse Model of Neuroblastoma ...... 22 1.3.2.1 Creation of the Th-MYCN mouse ...... 22 1.3.2.2 Characterisation of the Th-MYCN mouse model ...... 25 1.3.2.3 Therapeutic targeting of neuroblastoma using the Th-MYCN mouse . 26 1.3.2.4 Therapeutic Targeting of the MYCN pathway in Th-MYCN mice ..... 28 xiii 1.3.2.5 Testing of novel drug targets in Th-MYCN mice ...... 30 1.3.2.6 Th-MYCN mouse models crosses ...... 32 1.3.3 Other mouse models of neuroblastoma ...... 34 1.4 Gene Discovery Approaches ...... 35 1.4.1 Reverse genetics ...... 36 1.4.1.1 Site-directed mutagenesis and genome editing ...... 36 1.4.1.2 Transposable element excision ...... 36 1.4.1.3 Silencing using RNAi ...... 37 1.4.1.4 Knock-out mouse models ...... 37 1.4.2 Forward genetics ...... 38 1.4.2.1 Mutagenesis using X-Rays ...... 38 1.4.2.2 Mutagenesis using chemicals ...... 39 1.4.2.3 Mutagenesis using insertional transposable elements ...... 43 1.4.2.4 Quantitative trait loci mapping ...... 45 1.4.2.5 Emerging technologies using CRISPR/Cas9 ...... 46 1.5 Summary and Thesis perspectives ...... 46 CHAPTER 2 MATERIALS AND METHODS ...... 49 2.1 Materials ...... 49 2.1.1 Chemicals and Reagents ...... 49 2.1.1.1 Tissue Culture ...... 49 2.1.1.2 Cytotoxic drugs and chemicals ...... 49 2.1.1.3 siRNA transfection ...... 49 2.1.1.4 DNA and RNA isolation, ...... 50 2.1.1.5 Complementary DNA (cDNA) synthesis, polymerase chain reaction (PCR) and sequencing ...... 50 2.1.1.6 Protein isolation and western blotting ...... 50 2.1.1.7 Electrophoresis ...... 51 2.1.1.1 Cloning and transfection ...... 53 2.1.2 Equipment ...... 53 2.1.2.1 Tissue culture ...... 53 2.1.2.2 Plasmid DNA and nucleic acid isolation ...... 53 2.1.2.3 PCR and sequencing ...... 54 2.1.2.4 Spectrophotometer ...... 54 2.1.2.5 Electrophoresis ...... 54 2.1.2.6 Visualisation of gels and blots ...... 55 xiv 2.1.3 Cell lines ...... 55 2.1.4 Patient cohorts ...... 56 2.1.1 Mice...... 56 2.2 Methods ...... 59 2.2.1 Cell biology ...... 59 2.2.1.1 Maintenance of cell lines ...... 59 2.2.1.2 Overexpression and siRNA knock-down ...... 59 2.2.1.3 Colony formation assays ...... 59 2.2.1.4 Fractionation and western blotting ...... 60 2.2.1.5 Histology ...... 65 2.2.1.6 Immunohistochemistry ...... 65 2.2.1.7 Computational modelling ...... 66 2.2.2 Molecular biology ...... 68 2.2.2.1 Isolation of genomic DNA from mouse livers ...... 68 2.2.2.2 Genotyping of mice ...... 68 2.2.2.3 Exome and whole genome sequencing ...... 69 2.2.2.4 PCR and sequencing of mutant mouse lines ...... 71 2.2.2.5 RNA Isolation using TRIzol™ ...... 73 2.2.2.6 cDNA synthesis ...... 74 2.2.2.7 Real time quantitative PCR for expression studies ...... 74 2.2.2.8 Fluorescence in situ hybridisation (FISH) ...... 76 2.2.2.9 Cloning of overexpression and shRNA constructs ...... 76 2.2.2.10 Cloning of MYND domain ...... 80 2.2.2.11 Plasmid DNA isolation ...... 80 2.2.3 In vivo mouse studies ...... 81 2.2.3.1 Fertility testing after ablation of tumours with cyclophosphamide .... 81 2.2.3.2 ENU mutagenesis ...... 81 2.2.3.3 Mating for exome and whole genome sequencing ...... 82 2.2.3.4 Th-MYCN cross with Runx1t1 knock-out mice ...... 83 CHAPTER 3 ENU MUTAGENESIS OF THE TH-MYCN MOUSE
MODEL OF NEUROBLASTOMA ...... 84 3.1 Introduction ...... 84 3.2 Results ...... 86 3.2.1 Testing fertility of cyclophosphamide treated Th-MYCN mice ...... 86
xv 3.2.2 ENU Mutagenesis ...... 90 3.2.3 Fate of the ENU Treated Males ...... 94 3.2.4 Tumour development of ENU Progeny ...... 98 3.2.5 Heritability testing of delayed tumour phenotype ...... 104 3.3 Discussion ...... 110 CHAPTER 4 ELUCIDATION OF THE GENES RESPONSIBLE FOR DELAYED TUMOUR FORMATION IN ENU-TREATED TH-MYCN
MICE…………...... 114 4.1 Introduction ...... 114 4.2 Results ...... 115 4.2.1 Mouse 1590: Mapping and sequencing of progeny ...... 115 4.2.1.1 Appearance and histology of 1590 tumour ...... 115 4.2.1.2 Confirmation of MYCN genotype ...... 115 4.2.1.3 Exome sequencing of 1590 progeny ...... 119 4.2.1.4 Backcrossing for whole genome sequencing ...... 127 4.2.1.5 Runx1t1 YH mutation ...... 143 4.2.2 Mouse 1929: Mapping and sequencing of progeny ...... 148 4.2.2.1 1929 Exome sequencing ...... 148 4.2.2.1 1929 Backcrossing for confirmatory sequencing ...... 151 4.3 Discussion ...... 162 CHAPTER 5 CHARACTERISATION OF THE ROLE OF RUNX1T1 IN
NEUROBLASTOMA ...... 169 5.1 Introduction ...... 169 5.2 Results ...... 174 5.2.1 Th-MYCN mice crossed with Runx1t1 knock-out mice ...... 174 5.2.2 Neuroblast hyperplasia in 1590 and Runx1t1 knock-out mice...... 183 5.2.3 Inheritance of two copies of the 1590 mutation ...... 186 5.2.4 RUNX1T1 expression levels in human cell lines and mouse tissues ...... 189 5.2.5 Overexpression of Runx1t1 ...... 196 5.2.6 Knock-down of RUNX1T1 in neuroblastoma cells ...... 196 5.2.7 RUNX1T1 expression after MYCN knock-down in neuroblastoma cells .. 202 5.2.8 Fractionation of Runx1t1 protein complexes ...... 205 5.2.9 Correlation of RUNX1T1 expression with neuroblastoma outcome ...... 213 5.2.10 Correlation of RUNX1T1 protein in human neuroblastoma samples ...... 217
xvi 5.3 Discussion ...... 223 CHAPTER 6 CONCLUSIONS AND FUTURE DIRECTIONS ...... 228 REFERENCES ...... 231 APPENDICES ...... 271
xvii LIST OF FIGURES
Figure 1-1 Domains of the MYC protein family...... 17 Figure 1-2 Schematic of the Th-MYCN cassette...... 24 Figure 3-1 Survival curves of progeny from individual harems...... 88 Figure 3-2 Combined survival graphs for progeny from both homozygote and hemizygote females...... 91 Figure 3-3 ENU mutagenesis and homozygote Th-MYCN mouse mating scheme ...... 92 Figure 3-4 Unusual phenotypes produced as a product of ENU mutagenesis...... 95 Figure 3-5 Conditions requiring euthanasia in the ENU treated males...... 97 Figure 3-6 Tumour development of 1716 mice from ENU treated males...... 100 Figure 3-7 Mating diagram for testing inheritance to subsequent generations ...... 105 Figure 3-8 Time to small palpable tumour (weeks) of progeny of 31 mice in mouse number order...... 106 Figure 3-9 Mendelian inheritance in offspring of Mouse 1590...... 107 Figure 3-10 Inheritance of tumour phenotype in offspring of Mouse 1929...... 109 Figure 4-1 Comparison of 1590 mouse tumour with a tumour from a normal homozygote Th-MYCN mouse...... 116 Figure 4-2 Close-up of the 1590 tumour...... 117 Figure 4-3 Histology of the 1590 tumour...... 118 Figure 4-4 MYCN FISH of 1590 mouse blood...... 120 Figure 4-5 Mating diagram for Exome sequencing...... 121 Figure 4-6 Offspring from crossing together two 1590 suppressed mice...... 122 Figure 4-7 Single nucleotide changes in the 3 exomes that were sequenced...... 123 Figure 4-8 Mating diagram for 1590 crosses to BALB/c or C57BL/6...... 128 Figure 4-9 Tumour development in 1590 mice crossed to BALB/c mice...... 130 Figure 4-10 Tumour development of 1590 mice crossed to C57BL/6 mice...... 132 Figure 4-11 Mating diagram for crosses to BALB/c or C57BL/6 and backcross to Th- MYCN...... 134 Figure 4-12 Tumour development of 1590 mice crossed to BALB/c and backcrossed to Th-MYCN mice...... 135 Figure 4-13 Tumour development of 1590 mice crossed to C57BL/6 and backcrossed to Th-MYCN mice...... 136 Figure 4-14 Genotyping of suppressed and unsuppressed 1590 mice...... 142
xviii Figure 4-15 Nervy homology regions (NHR) of the myeloid translocation gene (MTG) family...... 144 Figure 4-16 Computational modelling of the RUNX1T1 protein structure...... 145 Figure 4-17 One-dimensional 1H NMR spectra of the RUNX1T1 MYND domain wild- type and YH point mutant...... 147 Figure 4-18 Mating diagram for exome sequencing of 1929 progeny...... 149 Figure 4-19 Offspring from crossing a 1929 suppressed mouse with a homozygote Th- MYCN mouse...... 150 Figure 4-20 Single nucleotide changes in the 4 exome sequencing samples from 1929 mice...... 152 Figure 4-21 Tumour development of 1929 mice crossed to C57BL/6 and backcrossed to Th-MYCN mice...... 155 Figure 4-22 Tumour development of 1929 mice crossed to C57BL/6 and backcrossed to Th-MYCN...... 157 Figure 4-23 Sequence of Rnf121 mutation...... 160 Figure 4-24 Transmembrane domains of RNF121...... 161 Figure 5-1 Generation of the Runx1t1 knock-out mouse model...... 170 Figure 5-2 Schematic of RUNX1T1 and its binding partners...... 172 Figure 5-3 Kaplan-Meier survival analysis in Th-MYCN homozygote mice crossed with Runx1t1 knock-out mice...... 175 Figure 5-4 Kaplan Meier survival analysis in Th-MYCN hemizygote mice crossed with Runx1t1 knock-out mice...... 176 Figure 5-5 Weights of Th-MYCN x Runx1t1 mice...... 178 Figure 5-6 Comparison of Runx1t1 wild-type and double knock-out mice...... 181 Figure 5-7 One week old Th-MYCN x Runx1t1 littermates...... 182 Figure 5-8 Comparison of suppressed and unsuppressed 1590 mice...... 184 Figure 5-9 Scoring of ganglia and tumours from 1590 suppressed and unsuppressed mice...... 185 Figure 5-10 Scoring of ganglia in day 0 Th-MYCN x Runx1t1 mice...... 187 Figure 5-11 Expression of RUNX1T1 in neuroblastoma cell lines...... 190 Figure 5-12 Gene expression of Runx1t1 across different mouse tissues...... 191 Figure 5-13 Gene expression of Runx1t1 in intestinal tissue and heart of 6 week old mice...... 192 Figure 5-14 Gene expression levels of MYCN and Runx1t1 in tumours from the 1590 line mice...... 194
xix Figure 5-15 Runx1t1 expression in ganglia from Th-MYCN mice...... 195 Figure 5-16 Colony formation in BE(2)-C cells transfected with Runx1t1 overexpression constructs...... 197 Figure 5-17 Knock-down of RUNX1T1 using SMARTPOOL siRNA...... 198 Figure 5-18 RUNX1T1 knock-down in KELLY cells using Ambion siRNA...... 200 Figure 5-19 Colony formation after induction of RUNX1T1 shRNA using doxycycline...... 201 Figure 5-20 Knock-down of MYCN in BE(2)-C cells...... 203 Figure 5-21 MYCN and RUNX1T1 expression in SH-EP Tet-21/N cells after induction of MYCN knock-down using tetracycline...... 204 Figure 5-22 Coomassie Blue staining of protein fractions 1-40 of untransfected BE(2)-C cells ...... 206 Figure 5-23 Coomassie Blue staining of protein fractions 1-40 of BE(2)-C cells transfected with a wild-type Runx1t1 construct...... 207 Figure 5-24 Coomassie Blue staining of protein fractions 1-40 of BE(2)-C cells transfected with the YH mutant Runx1t1 construct...... 208 Figure 5-25 Western blot of fractionated BE(2)-C cells using a MYCN antibody...... 209 Figure 5-26 Western blot of transfected and fractionated BE(2)-C cells using a FLAG antibody...... 211 Figure 5-27 Western blots of RUNX1T1 protein complex partners...... 212 Figure 5-28 Kaplan-Meier survival analysis based on RUNX1T1 expression in 649 primary neuroblastomas...... 214 Figure 5-29 Kaplan-Meier survival analysis based on RUNX1T1 expression in 101 Stage 4 single copy MYCN human neuroblastoma samples ...... 215 Figure 5-30 RUNX1T1 expression levels in 498 human neuroblastoma samples...... 216 Figure 5-31 Clinical characteristics of the tumour microarray cohort...... 218 Figure 5-32 Immunohistochemical staining of neuroblastoma tumour microarrays for RUNX1T1...... 219 Figure 5-33 RUNX1T1 expression in a TMA of primary human neuroblastoma...... 220 Figure 5-34 RUNX1T1 expression and Kaplan-Meier survival analysis of neuroblastoma samples from the tumour microarray...... 221 Figure 5-35 Proposed model of RUNX1T1 in neuroblastoma development...... 225
xx LIST OF TABLES
Table 1-1 Chromosomal gains and losses reported in neuroblastoma patients ...... 7 Table 2-1 Primary antibodies used in Western blotting and immunohistochemistry ..... 52 Table 2-2 Human tissue microarrays used in immunohistochemistry ...... 57 Table 2-3 Hypotonic buffer recipe for nuclear protein extraction ...... 61 Table 2-4 Co-IP buffer recipe for nuclear protein extraction ...... 62 Table 2-5 Antibodies used for western blotting ...... 64 Table 2-6 Scoring system for TMA ...... 67 Table 2-7 Primers and probes used in real-time PCR analysis for mouse genotyping ... 70 Table 2-8 Primers used in PCR for confirmatory sequencing ...... 72 Table 2-9 Gene expression assays and control genes used in real-time quantitative PCR ...... 75 Table 2-10 Primers used to create mutant Runx1t1 constructs ...... 77 Table 2-11 Primers used to create RUNX1T1 shRNA constructs ...... 79 Table 3-1 Number of litters and average litter size of offspring from harem matings ... 87 Table 3-2 Statistical comparisons of time to first palpable tumour and tumour cull between offspring from homozygous and hemizygous female...... 89 Table 3-3 All mice born to ENU treated males...... 93 Table 3-4 Primary reason for euthanasia of 81 ENU treated males ...... 96 Table 3-5 Viable progeny of ENU treated males split according to normal or delayed tumour development...... 99 Table 3-6 Th-MYCN progeny with delayed or absent tumour development ...... 101 Table 4-1 List of single nucleotide variants that were restricted to the two mice with a suppressed tumour phenotype ...... 125 Table 4-2 Comparisons between each 1590 and MYCN backcrossed lines...... 131 Table 4-3 Sequencing of XWM samples and proximal end of chromosome 4...... 138 Table 4-4 Sequencing of XBlM samples and proximal end of chromosome 4...... 139 Table 4-5 Whole genome sequencing of the proximal end of chromosome 4 of backcrossed mice...... 141 Table 4-6 List of variants in genes detected in suppressed samples from 1929 mice that were not detected in the unsuppressed mice...... 153 Table 4-7 Comparisons between each 1929 and MYCN backcrossed lines...... 156 Table 4-8 Number of variants found in backcrossed mice for the 13 genes selected from exome sequencing...... 158
xxi Table 5-1 Runx1t1 double knock-out colony statistics ...... 180 Table 5-2 Outcome table for mice with two copies of the suppressor mutation...... 188
xxii ABBREVIATIONS AND ACRONYMS
µg microgram µL microlitre µM micromolar aa amino acid A adenine ACEC Animal Care and Ethics Committee ALK Anaplastic Lymphoma Kinase ALL acute lymphoblastic leukaemia AML acute myeloid leukaemia ANR Advances in Neuroblastoma Research BCA bicinchoninic acid assay BDNF brain-derived neurotrophic factor BRD bromodomain BR-HLH-LZ basic region helix loop helix leucine zipper BET bromodomain and extraterminal domain C cytosine C57BL/6 C57 Black 6 CAMTA1 calmodulin binding transcription activator 1 Cas9 CRISPR associated protein 9 CASZ1 castor zinc finger1 cDNA complementary DNA CCG Children’s Cancer Study Group Chr chromosome CG coeliac ganglia CGH comparative genomic hybridisation COG Children’s Oncology Group Co-IP co-immunoprecipitation CRISPR clustered regularly interspaced short palindromic repeats CT computerised tomography DAPI 2-(4-amidinophenyl)-1H -indole-6-carboxamidine Dbh dopamine-β-hydroxylase DFMO difluoromethylornithine DMEM Dulbecco’s modified eagle medium
xxiii DMSO dimethyl sulfoxide DNA deoxyribonucleic acid dNTP deoxynucleoside triphosphates DOPA dihydrophenylalanine DOX doxycycline DTT dithiothreitol E embryonic day E-boxes enhancer boxes EDTA ethylenedinitrilotetraacetic acid EFS event free survival eIF4E eukaryotic initiation factor 4E ENU N-ethyl-N-nitrosourea ETO eight twenty one EV empty vector FACT Facilitates Chromatin Transcription FCS foetal calf serum FISH fluorescence in situ hybridisation FVB sensitive to friend leukaemia virus B G guanine GEMM genetically engineered mouse model GFI1 growth factor independence 1 GFP green fluorescent protein GN ganglioneuroma GNB ganglioneuroblastoma h hour H histidine HDAC histone deacetylase HEPES 2-[4-(2-hydroxyethyl)piperazin-1-yl]ethanesulfonic acid HLP hind limb paralysis HVA homovanillic acid IDRF Image Defined Risk Factors IHC immunohistochemistry INPC International Neuroblastoma Pathology Committee INRG International Neuroblastoma Risk Group INRGSS International Neuroblastoma Risk Group Staging System
xxiv INSS International Neuroblastoma Staging System kB kilo base LB Luria Bertani LOH loss of heterozygosity LSL loxP-STOP-loxP M molar MAD MAX dimerization protein 1 MAX MYC associated factor X Mb mega base MBl MYCN crossed to C57BL/6 MBlM MYCN crossed to C57BL/6 backcrossed to MYCN MDM2 mouse double minute 2 homolog MGB-NFQ minor groove binder non-fluorescent quencher MIBG meta-iodobenzylguanine M methionine min minute miR micro RNA mL millilitre MLL mixed linage leukaemia mM millimolar MMLV Moloney murine leukaemia virus MKI mitosis karyorrhexis index MRI magnetic resonance imaging MRP multidrug resistance-associated protein 1 MTG myeloid translocation gene MYC myelocytomatosis virus 29 MW MYCN crossed to BALB/c MBlM MYCN crossed to BALB/c backcrossed to MYCN MXI MAX interactor NCBI National Center for Biotechnology Information neo neomycin N-CoR nuclear receptor co-repressor NGF nerve growth factor NHR nervy homology region nm nanometre
xxv NMR nuclear magnetic resonance NON-SYN non-synonymous NP-40 nonyl phenoxypolyethoxylethanol NSE neurone-specific enolase ODC ornithine decaboxylase OS overall survival PBS phosphate-buffered saline PCR polymerase chain reaction PEI polyethylenimine POG Pediatric Oncology Group pml promyelocytic leukaemia gene PMSF phenylmethylsulfonyl fluoride PNK polynucleotide kinase PTEN phosphatase and tensin homolog PWD pre-wean death QTL quantitative trait loci R arginine Ref Reference RefSeq reference sequence RNA ribonucleic acid RPMI-1640 Roswell Park Memorial Institute-1640 Rnf121 RING finger protein 121 Runx1t1 Runt-related transcription factor 1; translocated to, 1 RNAi RNA interference SCG superior cervical ganglia SDS sodium dodecyl sulphate shRNA short hairpin RNA SIN3A SIN3 transcription regulator family member A siRNA short interferring RNA SMRT silencing mediator for retinoid or thyroid-hormone receptors SNP single nucleotide polymorphism SNV single nucleotide variation SON sQTL splicing quantitative trait loci T thymine
xxvi TALEN transcription activator-like effector nuclease TEMED N,N,N',N'-tetramethylethylenediamine Th tyrosine hydroxylase tk thymidine kinase TBS tris buffered saline TBST TBS Tween-20 TG2 tissue transglutaminase TMA tumour microarray Trk tropomyosin-receptor kinase TTRAP transactivation/transformation-associated protein Var Variant VBlM 1929 crossed to C57BL/6 and backcrossed to MYCN UTR untranslated region VCR vincristine VP-16 etoposide VMA vanillylmandelic acid VWM 1929 crossed to BALB/c and backcrossed to MYCN WT wild-type XBl 1590 crossed to C57BL/6 XBlM 1590 crossed to C57BL/6 and backcrossed to MYCN XW 1590 crossed to BALB/c XWM 1590 crossed to BALB/c and backcrossed to MYCN Y tyrosine
xxvii ABSTRACT
Neuroblastoma, a disease of the neural crest and the most common solid tumour of infancy, accounts for 7-10% of all childhood cancer. The disease arises from primitive cells within the neural crest, and can develop at any point along the sympathetic nervous system, including sites within the abdomen, pelvis, neck and thoracic cavity. Neuroblastoma tumours with MYCN oncogene amplification are highly aggressive, leading to a dismal outcome in most patients. A mouse model of neuroblastoma previously generated by targeting the human MYCN oncogene to neural crest cells (Th- MYCN), leads to 100% of MYCN homozygous mice developing tumours at an early age. This thesis describes a forward genetic screen involving the use of N-ethyl-N- nitrosourea (ENU) to generate heritable mutations causing delayed tumorigenesis in this model, as well as genomic sequencing to identify the mutated gene(s) responsible for the delayed tumorigenesis observed. Of 1716 ENU progeny screened, two mice demonstrated an unprecedented delay in tumour development. Subsequent breeding generated lines from these founders in which the delayed phenotypes were inherited in a dominant Mendelian fashion. Mapping, and whole exome and genome sequencing, identified a single gene in each of the lines as strong candidates responsible for delayed tumour formation, namely Runt-related transcription factor-1; translocated-to-1 (Runx1t1) and RING finger protein 121 (Rnf121). Runx1t1 was investigated further. Confirming the ENU mutagenesis findings, crossing of Runx1t1 knock-out mice with Th-MYCN mice, showed that loss of one allele of Runx1t1 abrogated tumour formation. The mechanism by which mutant Runx1t1 prevents neuroblastoma tumorigenesis was found to involve inhibition of neuroblast hyperplasia in mouse ganglia, a known prerequisite for neuroblastoma initiation. There was little evidence of MYCN directly regulating RUNX1T1 transcription, but rather increased RUNX1T1 protein levels were observed in human MYCN-amplified tumours. Knock-down of RUNX1T1 using shRNA resulted in decreased neuroblastoma colony formation in vitro, suggesting that this gene may also have a role in neuroblastoma maintenance. Collectively this study has employed ENU mutagenesis in an in vivo forward genetic screen to identify RUNX1T1 and RNF121 as novel neuroblastoma-associated genes. These genes represent potential new therapeutic targets for the treatment and ultimately prevention of this disease.
xxviii CHAPTER 1 INTRODUCTION
1.1 Neuroblastoma
1.1.1 Incidence
In Australia, 14 new cases of cancer are diagnosed per 100,000 children aged 0-14 years, with an average of 575 new cases per annum (AIHW, 2009). Cancer deaths are the highest disease-related cause of death in Australian children (17%), second only to injuries (37%) (AIHW, 2009) and this is also seen in other developed countries (Kaatsch, 2010). Neuroblastoma, a disease of the neural crest and the most common tumour of infancy, contributes approximately 15% of paediatric cancer-related deaths (Esiashvili et al., 2009), despite accounting for only for 7-10% of all childhood cancer cases in developed countries (Baade et al., 2010; Brodeur and Castleberry, 1997; Brodeur, 2003; Maris et al., 2007). Neuroblastoma is thought to arise from primitive sympathetic neural precursor cells, and as such, tumours can develop at any point along the sympathetic nervous system (Brodeur, 2003; Maris et al., 2007), including sites within the abdomen, pelvis, neck and thoracic cavity (Esiashvili et al., 2009). However, the most common site for tumour development is in the abdomen (65%), with approximately half of these arising specifically from the adrenal medulla (Maris et al., 2007). Neuroblastoma tumours can be highly aggressive and children often present with advanced disease, with greater than 50% of patients already having metastatic spread via the lymphatic and circulatory systems to bone, bone marrow, liver and skin (Brodeur and Castleberry, 1997; Park et al., 2010). The median age at diagnosis is 18- 23 months, with 40% diagnosed by 1 year of age, 75% by the age of 4 years and 98% by the age of 10 years (Brodeur, 2003; Park et al., 2010).
The cause of neuroblastoma remains an enigma. Only 1-2% of neuroblastoma cases can be attributed to a genetic predisposition, with activation of the Anaplastic Lymphoma Kinase (ALK) oncogene by gene fusion/amplification/mutation as the main contributing factor in neuroblastoma pedigrees (Ogawa et al., 2011). The remaining 98% of neuroblastoma tumours are presumed to be spontaneously occurring, and although many environmental studies have been undertaken examining a wide range of potential toxins including drug and alcohol exposure, use of maternal hair dyes and radiation
Chapter 1- Introduction 1 among others, no clear link has been observed and it is believed that an environmental exposure is not the cause (Brodeur and Castleberry, 1997).
1.1.2 Staging
Clinical stage is an important factor to consider in the biology of neuroblastoma as it describes the extent of disease spread and allows for stratification into treatment groups. Several staging systems were proposed in the late 1960s up until the early 1980s (Carlsen et al., 1986). Three main systems were developed and used by different cancer bodies including Children’s Cancer Study Group (CCG), the system used by Pediatric Oncology Group (POG) and the St Jude’s Hospital modification of this, and the International Union Against Cancer (Evans et al., 1990). One of the first staging systems was described by Evans and colleagues at the CCG in 1971 (Evans et al., 1971). This split neuroblastoma patients into one of 5 stages. In Stage I, the tumour remains localised to the organ/structure of origin. Stage II of the disease, the tumour extends beyond the structure of origin, but does not cross the midline and there may also be homolateral (i.e. on the same side) lymph node involvement. For Stage III patients, the primary tumour has crossed the midline and bilateral involvement of the lymph nodes may occur. Large primary tumours and metastasis to multiple sites including bone, bone marrow, liver and distal lymph nodes, occurs in Stage IV patients. The final stage has been named IVs, with the ‘s’ standing for ‘special’. Patients in this stage have a localised primary mass and would usually be classed as Stage I or II, except for the presence of metastases to the liver, skin or bone marrow. Surprisingly, Stage IVs patients have an excellent prognosis and these tumours often spontaneously regress with little or no treatment. The mechanism of spontaneous regression remains unknown.
The 2 year survival rates of 100 patients using this staging criteria was assessed in 1971. Stage I patients had the best prognosis at 80%, Stage II 60% and Stage IVs 75%. Once the tumour had crossed the midline (Stage III) and multiple metastases had occurred (Stage IV), survival dropped to 13% and 7% respectively (Evans et al., 1971).
With so many different systems being used worldwide, comparisons between patients in different countries, and even hospitals within the same country, were becoming more
Chapter 1- Introduction 2 difficult. In two different comparisons, all staging systems tested were able to predict outcome in low risk patients with unilateral tumours or minor residual disease, however, the Evans staging system was found to more accurately predict outcome in later stage patients. One review hypothesized this was due to tumour size or factors other than resectability, while Evans et al believed it was due to the classification of whether the neuroblastoma tumours crossed the midline or not, something the other staging systems lacked (Carlsen et al., 1986; Evans et al., 1990).
The Evans staging system (Evans et al., 1971) formed the basis for the International Neuroblastoma Staging System (INSS) that was developed following the Advances In Neuroblastoma Research (ANR) meeting in 1987 and then revised in 1993 (Brodeur et al., 1993; Brodeur et al., 1988). The revisions clarified the language used, redefined midline as crossing the vertebral column and recommended the inclusion of meta- iodobenzylguanine (MIBG) scanning to distinguish between active tumour and scar tissue, and also to assess the extent of metastases to cortical bone. Neuroblastoma patients should be assessed by computerised tomography (CT) and/or magnetic resonance imaging (MRI) coupled with MIBG scanning, bone marrow aspirate and lymph node examination for correct staging to occur. The new INSS uses Arabic numbers rather than the Roman numerals of the CCG staging system. Thus, Stage 1 was defined as a localised tumour with complete gross excision, either with or without microscopic residual disease and a negative biopsy of the representative homolateral lymph nodes. Stage 2A was still a localised tumour, but gross excision was incomplete with a negative biopsy of the homolateral lymph nodes. For Stage 2B, the excision of the localised tumour could be either completely or incompletely removed, however the homolateral lymph node biopsy was positive. Enlarged contralateral lymph nodes must be negative for this stage. Stage 3 patients have either an unresectable unilateral tumour that crosses the midline (where the midline is defined as the vertebral column and tumours that originate on one side must infiltrate to or beyond the side of the vertebral column), with or without lymph node involvement; or a localised unilateral tumour with positive contralateral lymph node; or a midline tumour with bilateral extension by infiltration that is unresectable or has lymph node involvement. Stage 4 patients are defined as any primary tumour that has spread to distant lymph nodes, bone, bone marrow, skin, liver and/or any other organ, except as defined for Stage 4S patients. Stage 4S patients have a localised primary tumour and would otherwise be defined as Stage I, 2A or 2B, except for the presence of metastases which is limited to skin, liver
Chapter 1- Introduction 3 and/or bone marrow, and the patient is <1 year of age. Most 4S patients will have ≤1% marrow involvement, but an upper limit of 10% is used, with anything higher resulting in a Stage 4 classification (Brodeur et al., 1993). Analysis of patients in a multicentre study in Spain concluded that the INSS was easy to use and results across the different centres were able to be combined, with Stage 1, 2B and 4S patients having 100% 5 year survival, 75% and 73% for Stage 2A and 3 respectively, and Stage 4 patients having the worst prognosis at 33% (Castel et al., 1999).
More recently, a new staging system was proposed by the International Neuroblastoma Risk Group (INRG) Task Force that allows risk stratification to occur prior to surgical resection. The INRG recognised that the INSS had some limitations in that the same tumour could be classified as either Stage 1 or 3 depending on the success of the resection, while the assessment of lymph node involvement was up to the individual surgeon (Monclair et al., 2009). The INRG Staging System (INRGSS) on the other hand, is based solely on image defined risk factors (IDRFs) using CT and/or MRI and MIBG scintigraphy of primary and metastatic tumour sites. There are 4 INRGSS stages, namely, L1, L2, M and MS. Stage L1 is a localised tumour confined to one body compartment (either the neck, chest, abdomen or pelvis) and does not involve vital structures (ie major veins and arteries, trachea and skull among others) as defined by the IDRFs. Stage L2 disease represents a locoregional tumour (confined to one side of the body) with 1 or more IDRFs. Stage M represents distant metastatic disease, except as is defined for MS, while Stage MS presents as metastatic disease in patients aged <18 months with the metastatic sites limited to the skin, liver and/or bone marrow, with a negative MIBG scan in bone and bone marrow. Like Stage 4S in the INSS classification, <10% bone marrow involvement is required to be classified as MS (Monclair et al., 2009). The INRGSS is not intended to work in isolation, rather it complements the INSS and the two should be used together to stratify patients (Brisse et al., 2011).
Chapter 1- Introduction 4 1.1.3 Clinical Factors
1.1.3.1 Age
Age at diagnosis as a clinical factor in neuroblastoma was first described by Breslow and McCann in 1971 when they observed that older patients had a significantly worse outcome (Breslow and McCann, 1971). They split the 246 patients into 3 age ranges: 0– 11months, 12–23 months and 24+ months. Probability modelling suggested that those children in the less than 11 month group had a better 2 year survival prognosis than those in the other two groups, with decreasing survival as age increased. Age ≤1 against ≥1 year of age was then used in the classification of patients for more than 30 years. This was a convenient cut-off for clinicians, but those aged between 12 and 24 months represented a grey area of gradual decline to poorer outcome (London et al., 2005). The Children’s Oncology Group (COG) revisited the age cut-off in 2005 by analysing up to 3666 neuroblastoma patients enrolled under POG and CCG studies from 1986 to 2001. Although there was no clear cut number where age is the critical determinate, the 1 year cut-off was deemed too low and 15 months was suggested as an alternate (London et al., 2005). Due to changing treatment regimens over time, the INRG recently reviewed 11037 children diagnosed across several continents from 1974 to 2002. Regardless of the era the children were diagnosed in, aged >18 months always had the poorest outcome when compared to 0-12 or >12-18 months at diagnosis (Moroz et al., 2011). For those diagnosed during the study period the 3 year event free survival (EFS) was 84% (hazard ratio (HR) = 1) in the children aged 0-12 months, 68% (HR=2.11) in those 13-18 months and 43% (HR=4.47) in the children older than 18 months (Moroz et al., 2011). An age cut-off of 18 months (547 days) is now widely recognised and accepted as the stratification criteria for patients, with one exception. Children diagnosed with diploid, Stage M tumours with single copy MYCN still retain the conservative 12 month cut-off, as clinicians believe that a reduction in therapy intensity for this group would reduce survival outcomes (Cohn et al., 2009). The implications of this change in age stratification to 18 months is that children who are aged between 12-18 months now get less intensive therapy than they would have under the old classification, with hopefully less long term side effects (London et al., 2005).
Chapter 1- Introduction 5 1.1.3.2 Chromosomal Gains and Losses
The most characterised deletion or chromosomal aberration in neuroblastoma belongs to the short arm of chromosome 1 (1p), where up to 35% of patients have been found to possess some form of abnormality (Table 1-1), with the most common region deleted located at 1p36.3 (Attiyeh et al., 2005; Brodeur, 2003). In a large study of neuroblastoma patients, this loss of heterozygosity (LOH) resulted in a decrease in 3 year EFS from 77% (no loss) to 47% (LOH at 1p36) (Attiyeh et al., 2005). Although loss of 1p is an independent prognostic factor when combined with age and stage, it is not an independent prognostic factor for neuroblastoma when MYCN amplification is used in the stratification (Caron et al., 1996; Maris et al., 1995). Most patients that are MYCN amplified also have a deletion in 1p, however not all 1p deleted patients will be amplified for MYCN (Caron et al., 1996). The prognostic value of 1p loss rests in its ability to predict poor outcome in patients with single copy MYCN (Caron et al., 1996).
It has previously been hypothesized that one or more tumour suppressor genes reside in this 1p36 locus (Maris et al., 1995) and recently, a number of studies have identified potential candidate suppressor genes within this region. Located in the region of 1p36.31, a gene responsible for chromatin remodelling, CHD5, was found to be preferentially expressed in neural and adrenal tissue and lacked expression in 1p deleted cells. Moreover, high expression predicted for favourable survival and when overexpressed in neuroblastoma cells, tumour growth and clonogenic potential were reduced (Fujita et al., 2008; Okawa et al., 2008).
A number of other potential tumour suppressor candidates have also been identified in this region. Castor zinc finger 1 (CASZ1), a zinc finger transcription factor, is required for neuronal differentiation and when overexpressed in neuroblastoma cells it inhibits migration and reduces tumorigenic potential both in vitro and in vivo (Liu et al., 2011b). In further in vitro experiments, CASZ1 expression was also capable of blocking cell cycle progression at G1 and is able to restore the retinoblastoma tumour suppressor protein making it a strong 1p loss candidate (Liu et al., 2013). Located at 1p36.22 is microRNA-34a (miR-34a), which targets a number of cellular proliferation and apoptosis genes including MYCN and calmodulin binding transcription activator 1 (CAMTA1), and over-expression of this miR also results in reduced tumour burden in-
Chapter 1- Introduction 6 Table 1-1 Chromosomal gains and losses reported in neuroblastoma patients
Frequency Chromosome Change Reported
1p36 Loss ~35%
2p24 Gain 14%
3p26 Loss 18%
4p Loss ~20%
10q Loss 18-20%
11q23 Loss 35-40%
12p Gain Unknown
14q Loss 18%
17q Gain 63-83%
Chapter 1- Introduction 7 vivo (Tivnan et al., 2011). In addition, CAMTA1, is also located on 1p36, and has its own role in neuronal differentiation and neuroblastoma outcome (Henrich et al., 2011).
The reason there has not been a definitive candidate implicated for 1p loss may be that the answer is not simply the loss of one factor, but rather the loss of multiple factors and their effect on downstream targets, some of which may be unknown at this stage. Clearly this is a very interesting field and our knowledge of chromosome 1p deletion and its role in neuroblastoma outcome will undoubtedly increase over the coming years.
Chromosomal deletions also occur in other regions of the neuroblastoma genome. LOH is found in the long arm of chromosome 11 at 11q23 and is present in 35-40% of neuroblastoma patients (Attiyeh et al., 2005; Guo et al., 2000). Interestingly, unlike 1p deletions, 11q LOH is inversely correlated with MYCN status where patients are more likely to be single copy for MYCN. Regardless of this, patients with 11q loss still have a poor prognosis as this feature correlates with both advanced stage and unfavourable histopathology (Guo et al., 1999). It has been hypothesized that 11q loss may be a useful marker for single copy MYCN tumours that are nevertheless aggressive and have poorer outcome (Attiyeh et al., 2005).
Like loss of 1p, it has been speculated that one or more tumour suppressors reside in this 11q locus. Only two genetic factors have been postulated to be tumour suppressors in this region, namely miR-34c and cell adhesion molecule 1 (Cole et al., 2008; Nowacki et al., 2008). However, further work is needed to confirm their role in neuroblastoma. A second hypothesis is that 11q loss does not result in the deletion of a potential tumour suppressor but rather causes fusion of genes in this region. The mixed- lineage leukaemia (MLL) gene is located on 11q23 and is a frequent fusion partner in leukaemias (De Braekeleer et al., 2005). In neuroblastoma, it has been postulated that the MLL gene fuses with the forkhead-box oncogene causing upregulation of this oncogene, which in turn regulates the normally tumour suppressive forkhead-box transcription factors that control both cell fate and tumorigenesis (Santo et al., 2011).
Chromosome 17 gain, either complete gain of a chromosome or part thereof, occurs in 65-85% of neuroblastoma patients, the most common cytogenetic abnormality in neuroblastoma. Specifically, unbalanced 17q gain occurs in approximately 50% of patients (Bown et al., 1999) and is associated with advanced Stage 4 disease and with
Chapter 1- Introduction 8 older age at diagnosis. Interestingly, 17q gain is also associated with both 1p deletion and amplification of MYCN, such that, all MYCN amplified tumours have either 1p loss or 17q gain or both (Bown et al., 1999). Cytogenetic analysis has shown that 17q can translocate with chromosome 1p while still retaining two normal copies of chromosome 17 (Bown et al., 1999). Within the commonly gained portion of 17q a number of important genes reside. Nerve growth factor receptor and survivin, a member of the inhibitor of apoptosis proteins, are located within this region and increased survivin expression has been correlated with poor clinical outcome and increased cell survival (Bown et al., 1999; Islam et al., 2000). Furthermore, studies have shown that survivin can be gained and translocated from 17q to 1p, making it a potential candidate for future therapy (Islam et al., 2000).
There have also been reports of other chromosomal changes within neuroblastoma cell lines and patients, with less frequency than those mentioned above (Table 1-1). Gain of 2p results in a gain of the MYCN gene, as well as other genes known to influence neuroblastoma, like ALK (Jeison et al., 2010). Children that had gain of 2p had a similar EFS to those that had MYCN amplification, and this 2p gain was also frequently observed alongside 11q loss, 1p deletion and advanced stage disease (Jeison et al., 2010). LOH, imbalance and gain have been reported for chromosomes 3p, 4p, 10q, 12p and 14q (Caron et al., 1996; Jeison et al., 2010; Lastowska et al., 2007; Lázcoz et al., 2007; Spitz et al., 2003). These are less frequently observed than 1p, 11q and 17q and their significance is unclear at this time.
1.1.3.3 Histology
In 1984, Shimada and colleagues developed a classification system based on the histopathology of primary neuroblastoma tumours, called the Shimada system (Shimada and Nakagawa, 1999). This system was further refined in 1994 when the International Neuroblastoma Pathology Committee (INPC) was created (Shimada et al., 1999b). The system grades tumours into 3 groups, namely undifferentiated, poorly differentiated and differentiating, depending on the level of differentiation of the cells within the biopsy sample (Shimada et al., 1999a). For the undifferentiated subtype, neuroblastoma cells are small to medium in size, are undifferentiated, have an ill-defined cytoplasm and neuropil (thin neuritic processes containing dendrites, glial cells and unmyelinated
Chapter 1- Introduction 9 axons that form a dense region) is almost absent (Joshi et al., 1992; Shimada et al., 1999a). In the poorly differentiated group, less than 5% of cells have undergone differentiation, with some neuropil present (Joshi et al., 1992; Shimada et al., 1999a). The differentiating group is classified as having greater than 5% of the biopsy sample with features of maturation to a ganglion cell (Shimada et al., 1999a).
Tumours can be further classified as being Schwannian stroma-rich (>50% positive) or Schwannian stroma-poor (<50% positive), graded according to the level of background stroma (Shimada et al., 1999a). In neuroblastoma, Schwann cells (glial cells of the peripheral nervous system responsible for the preservation of nerve fibres) are thought to inhibit neuroblastoma proliferation and stimulate the differentiation process into mature ganglion (Ambros et al., 1996). The mitosis karyorrhexis index (MKI) assesses the number of tumour cells undergoing mitosis and karyorrhexis (chromatin in the cytoplasm from nuclear fragmentation) and can be split into low (<2%), intermediate (2-4%) and high (>4%) (Shimada et al., 1999a).
Tumours are then defined as either having favourable or unfavourable histology based on the previously mentioned criteria, i.e. percentage of differentiating cells, Schwannian cell content, the MKI as well as the age at diagnosis of the patient (Esiashvili et al., 2009; Jiang et al., 2011). The INPC defines favourable histology as either less than 18 months with poorly differentiated or differentiating subtype and a low or intermediate MKI; or aged 18 months to 5 years of age with a differentiating subtype and a low MKI. An unfavourable histology is defined in the less than 18 month age group as having either an undifferentiated subtype or a high MKI while those aged 18 months to 5 years have either an undifferentiated/poorly differentiated subtype or an intermediate/high MKI. All children aged >5 years of age with neuroblastoma (Schwannian stroma-poor) are classified in the unfavourable histology group (Shimada and Nakagawa, 1999). Analysis of 746 children enrolled under CCG studies between 1991-1995 demonstrated the importance of histological grading with a 9-year overall survival (OS) rate in the favourable histology group being significantly better than the unfavourable histology group (96% verses 32%) (Sano et al., 2006).
Chapter 1- Introduction 10 1.1.3.4 Ploidy
DNA content of neuroblastoma tumour cells is a controversial predictor of outcome. Hypodiploidy (decreased tumour cell DNA content) is rarely seen in neuroblastoma patients and most tumours analysed are hyperdiploid (increased tumour-cell DNA content) in the triploid to pentaploid range, with only approximately a third being normal diploid (Look et al., 1991). Ploidy as a predictor of clinical outcome is not valuable for infants with Ds tumours, nor is it predictive in Stage A patients who only have surgical resection or children older than 24 months (Look et al., 1991). The predictive value lies in the more advanced stage patients aged under 24 months, where tumours with hyperdiploidy correlated to a better clinical outcome than those patients with diploid tumours (Look et al., 1991). Look and colleagues also observed a favourable correlation between response to DNA damaging agents, cyclophosphamide and doxorubicin in patients with hyperdiploid tumours (Look et al., 1984). Tumour DNA content will continue to be analysed in neuroblastoma patients as it forms part of the International Consensus for Neuroblastoma molecular diagnostics from the INRG (Ambros et al., 2009).
1.1.3.5 MYCN Amplification
MYCN was first discovered and described in the early 1980s by Schwab and colleagues when they examined a number of neuroblastoma cell lines looking for karyotypic abnormalities (Schwab et al., 1983). MYCN is a member of the MYC family of proto- oncogenes and is required for normal neuronal development, highlighted by the fact that mice lacking MYCN die during embryogenesis with severe defects in the brain, heart, lungs and central and peripheral nervous systems (Jiang et al., 2011; Westermark et al., 2011). MYCN is located at chromosome 2p24 and increased number of gene copies or amplification of this gene occurs in approximately 25% of neuroblastomas and can be amplified by up to 500-fold (Jiang et al., 2011; Shimada and Nakagawa, 1999). Amplification of MYCN is associated with advanced disease and rapid tumour progression (Seeger et al., 1985), with almost half of Stage D tumours having MYCN amplification, and almost all of the patients studied failing therapy (Look et al., 1991),
Chapter 1- Introduction 11 and as such remains one of the strongest predictors of treatment failure to date (Park et al., 2010).
1.1.3.6 Other Biological Factors
Other biological features known to be prognostic indicators include levels of ferritin, neurone-specific enolase (NSE), serum catecholamines and tropomyosin-receptor- kinases (Trks). High levels of ferritin are secreted by Stage III and IV neuroblastoma tumours and the level of ferritin at diagnosis is indicative of outcome in these patients, with higher levels having poorer outcome (Hann et al., 1985). NSE is a marker of neuronal differentiation and the levels are higher in advanced stage metastatic neuroblastoma. NSE levels also correlate to clinical outcome, with high levels showing a poor prognosis and these levels can be used as a marker to follow treatment success or failure (Zeltzer et al., 1986). Both ferritin and NSE levels are useful tools for distinguishing between Stage IV and IVs patients (Hann et al., 1985; Zeltzer et al., 1986).
Urinary catecholamines are a useful tool for neuroblastoma diagnosis (Brodeur and Castleberry, 1997). Catecholamines are secreted by the adrenal medulla and act as neurotransmitters in the sympathetic and central nervous systems (Erdelyi et al., 2011). Urinary catecholamines can be detected across all stages of the disease and with a frequency of between 78-100% (Erdelyi et al., 2011). In a very complex process, tyrosine is converted to dihydroxyphenylalanine (DOPA) by tyrosine hydroxylase, an enzyme that is highly expressed in neuroblastoma tumours (Brodeur and Castleberry, 1997; Erdelyi et al., 2011). DOPA is converted to dopamine and eventually to adrenaline and noradrenaline. Dopamine and DOPA are converted to homovanillic acid (HVA), while adrenaline and noradrenaline themselves are converted to vanillylmandelic acid (VMA). Urinary testing of catecholamines examines both HVA and VMA (Brodeur and Castleberry, 1997). The absolute values of these two metabolites are not themselves predictive of outcome, it is the ratio of the two together that predicts survival, with a lower ratio having a poorer prognosis (Laug et al., 1978).
The Trk family are transmembrane glycoproteins that are part of the nerve growth factor (NGF) signalling-transduction pathway and are critical for neural cell development
Chapter 1- Introduction 12 (Brodeur, 2003; Svensson et al., 1997). There are 3 family members that have been investigated in neuroblastoma biology, namely TrkA, TrkB and TrkC and their 3 respective receptors are NGF, brain-derived neurotrophic factor (BDNF) and neurotrophin-3 (Brodeur, 2003). Activation of TrkA promotes differentiation and it has been postulated that it may also be involved in spontaneous regression (Brodeur, 2003). TrkA is expressed in approximately 45% of neuroblastoma patients and high levels correlate with good clinical outcome and favourable histology and are inversely correlated with MYCN amplification (Nakagawara et al., 1993; Nakagawara et al., 1992; Souza et al., 2011). Recently, MYCN was discovered to be able to repress the expression of TrkA in neuroblastoma cells, giving a rationale as to why TrkA expression and MYCN amplification are inversely correlated (Iraci et al., 2011). However, in a large study of patients, TrkA expression was not an independent predictor of outcome when adjusted for stage, histology and MYCN status, and as such is unlikely to add any further value for patient stratification at this time (Shimada et al., 2004). Another member of the Trk family, TrkC is expressed in approximately 25% of patients, is often co-expressed with TrkA, and also correlates with favourable biology and stage (I, II and IVs) and lack of MYCN amplification (Brodeur et al., 1997; Fung et al., 2011; Svensson et al., 1997). However, TrkC expression was also found not to be an independent predictor of outcome (Fung et al., 2011). In contrast to TrkA and TrkC, high expression of full length TrkB promotes angiogenesis, migration, drug resistance and tumorigenicity (Brodeur, 2003; Park et al., 2010). TrkB, and its receptor BDNF, are often expressed in highly aggressive, MYCN amplified tumours (Nakagawara et al., 1994).
1.1.4 Risk Stratification and Treatment Outcome
Treatment for neuroblastoma involves a combination of surgery, chemotherapy and radiotherapy, and is dependent upon their stratification into low, intermediate and high risk groupings using the previously described factors such as age, stage, histology, chromosomal gains and losses, ploidy and MYCN amplification status (Park et al., 2010). Using the INRG staging, patients are assigned to a pre-treatment risk group, which ranges from very low to high. As outlined by Cohn and colleagues, a very low- risk patient has a maturing or intermixed histology and is Stage L1/L2; or is Stage L1 with any histology other than ganglioneuroma (GN) maturing or ganglioneuroblastoma
Chapter 1- Introduction 13 (GNB) intermixed; or is an MS stage patient with non-amplified MYCN and is normal at 11q (Cohn et al., 2009). When clustered using these criteria, 28% of patients were grouped into this category and patients in this group have a 5-year EFS of >85% (Cohn et al., 2009).
The low pre-treatment risk group consists of patients with L2 stage, non-amplified MYCN and normal 11q either aged <18 months of age with any histology other than GN maturing or GNB intermixed, or aged >18 months with differentiating histology; or Stage M, non-amplified MYCN, hyperdiploid patients; or Stage MS patients with non- amplified MYCN and normal 11q (Cohn et al., 2009). Low-risk patients often have very little intervention other than surgery and chemotherapy is reserved only for those patients who relapse (Brodeur and Castleberry, 1997). Low-risk patients have a >75% to ≤85% 5 year EFS (Cohn et al., 2009).
Intermediate-risk disease comprises patients in Stage L2 with non-amplified MYCN and 11q aberration, either <18 or >18 months; or Stage M tumours that are non-amplified and diploid (Cohn et al., 2009). These patients receive surgery and chemotherapy, either 4 cycles for tumours with favourable biology or 8 cycles with unfavourable biology (Esiashvili et al., 2009; Park et al., 2010). Another study demonstrated that for intermediate risk patients, chemotherapy could be reduced to 4 or 8 cycles, which reduced the long-term side effects while still achieving high success rates (Baker et al., 2010). Chemotherapy cycles utilise cyclophosphamide, carboplatin, etoposide (VP-16) and doxorubicin (Esiashvili et al., 2009). The 5 year EFS for intermediate patients ranges from 50-75% (Cohn et al., 2009).
Any patient with a MYCN amplified tumour is immediately placed into the high-risk neuroblastoma group regardless of stage (including MS) or any other factor (Cohn et al., 2009). This group also includes patients that are non-amplified for MYCN and either Stage M aged ≥18 months or patients that are Stage MS with a 11q aberration (Cohn et al., 2009). The majority of neuroblastoma patients diagnosed fall into this category, and these patients have the poorest prognosis with a 5 year EFS at less than 50% (Cohn et al., 2009). High-risk neuroblastoma patients receive at least 4 components of therapy: induction therapy, local control, myeloablative consolidation therapy, and maintenance treatment with biological agents (Esiashvili et al., 2009; Park et al., 2010). Induction therapy consists of 4-7 cycles of anthracyclins, platinum compounds, alkylators and
Chapter 1- Introduction 14 topoisomerase II inhibitors (Park et al., 2010). Typically these drugs are doxorubicin, cyclophosphamide, VP-16, vincristine (VCR) and cisplatin, and following this treatment, surgical resection of the primary tumour is performed (Esiashvili et al., 2009; Jiang et al., 2011). Following induction and surgery, patients then move into the local control phase which consists of daily radiotherapy targeting the site of the primary tumour (Park et al., 2010). In the myeloablative consolidation therapy phase, high dose chemotherapy of carboplatin, VP-16 and melphalan is given to kill all cells in the bone marrow, followed by a bone marrow transplant (Esiashvili et al., 2009; Park et al., 2010). Total body irradiation can also be used, although this practice is losing favour due to toxicity without significantly improving outcome (Esiashvili et al., 2009).
Maintenance therapy is used with the goal of eliminating any minimal residual disease that resides in the bone marrow and is resistant to prior therapy (Esiashvili et al., 2009). 13-cis-retinoic acid which is used for up to 6 months in maintenance therapy has been shown in vitro to both down regulate MYCN expression and cause neuronal differentiation in neuroblastoma cells (Esiashvili et al., 2009; Westermark et al., 2011). Biologic therapy may also include treatment with monoclonal antibodies and cytokines to reduce disease recurrence (Park et al., 2010). Surface antigens on neuroblastoma cells have been used as targets for immunotherapy. The most successful example of this is the use of anti-disialoganglioside therapy, involving dinutuximab (a monoclonal antibody that binds ganglioside GD2), which resulted in improved EFS (McGinty and Kolesar, 2017). Dinutuximab is combined with 13-cis-retinoic acid and is given with an interleukin-2 inhibitor and granulocyte-macrophage colony stimulating factor, to reduce toxicity observed when dinutuximab is used. This combination is the current standard used in maintenance therapy (McGinty and Kolesar, 2017).
Despite this intensive combination of chemotherapy, radiotherapy, surgery, bone marrow transplant and maintenance therapy, these patients still have a poor survival rate and those that do survive have long term health effects including the development of second cancers and infertility as well as chronic health conditions affecting neurological, musculoskeletal and endocrine systems (Esiashvili et al., 2009; Laverdiere et al., 2009). Alternate treatments with less long term side effects need to be developed and if successful incorporated into induction therapy.
Chapter 1- Introduction 15 1.2 The MYCN Oncogene
1.2.1 Discovery and Structure of the MYC Family
MYCN was first described by Manfred Schwab and colleagues in 1983 when they detected 140-fold upregulation of a c-MYC homolog in neuroblastoma cell lines (Schwab et al., 1983). c-MYC had been described only a few years earlier when it was isolated from chicken DNA as one of a number of genes that were capable of transforming cells into tumours (Hayward et al., 1981). MYC, named after the myelocytomatosis virus 29, was detected at very high levels in chickens with avian- leukosis virus-induced lymphoma, suggesting that c-MYC was the gene capable of causing this transformation (Hayward et al., 1981). Soon after, c-MYC was found to be upregulated in a number of human haematopoietic malignancies including acute lymphoblastic leukaemia (ALL), myeloid leukaemia and Burkitt’s lymphoma (Rothberg et al., 1984). There are currently 3 described members of the MYC family, namely c- MYC, L-MYC and MYCN. The three members are highly conserved, although differences do exist, and the proteins are of slightly different lengths. MYCN is the longest at 464 amino acids (aa), L-MYC is the shortest at 364 aa and c-MYC is 439 aa (reviewed in (Tansey, 2014)). All 3 possess a DNA binding domain containing a basic region helix-loop-helix leucine zipper (BR-HLH-LZ) motif at the carboxy terminus. BR-HLH-LZ containing proteins recognise consensus sequences called enhancer boxes (E-boxes). Both c-MYC and MYCN also contain 5 MYC homology boxes termed I, II, IIIa, IIIb and IV while L-MYC lacks IIIa (Figure 1-1). MYC boxes I and II are required for cellular transformation and tumorigenicity, however, little is known about the specific role of MYC boxes III and IV (Tansey, 2014). c-MYC and MYCN can form dimers with another BR-HLH-LZ protein called MYC associated factor X (MAX), which results in gene transcription (Tansey, 2014). MAX can also homodimerise to itself without a MYC family member as well as form complexes with other BR-HLH- LZ proteins such as MAX dimerization protein 1 (MAD) and MAX Interactor (MXI) (Ayer et al., 1993; Zervos et al., 1993). MXI, which is a transcriptional repressor, has been shown to antagonise MYCN by competing for MAX in neuroblastoma cells (Erichsen et al., 2015). MYC-MAX and MAD-MAX complexes are capable of
Chapter 1- Introduction 16
Figure 1-1 Domains of the MYC protein family. The MYC homology boxes (I-IV) are at the N-terminus while the carboxy-terminus contains the basic region helix-loop-helix leucine zipper (Adapted from Tansey, 2014).
Chapter 1- Introduction 17 competing with one another for target sequences (Nair and Burley, 2003). Thus far, only the BR-HLH-LZ portion of the MYC-MAX and MAD-MAX protein heterodimers have been crystallised (Nair and Burley, 2003).
1.2.2 Activator and Repressor Functions of MYCN It is estimated that the MYC family can bind to and regulate between 10-15% of the human genome while approximately 25% of all known gene promoters contain E-boxes (Gustafson and Weiss, 2010; Meyer and Penn, 2008). The MYC family genes have dual roles in that they can both activate and repress gene transcription. MYCN specifically has roles in promoting proliferation, survival, migration, metastasis and angiogenesis while repressing genes that are involved in immune surveillance, differentiation and cell cycle arrest (reviewed in (Huang and Weiss, 2013). Gene activation occurs through the binding of MYC to its partner MAX, which has been described above. The discovery of transactivation/transformation-associated protein (TTRAP) in 1998 helped to elucidate what occurs after the MYC-MAX binding. By using deletion mutations for MYC box II, McMahon et al proved that TTRAP binds to c-MYC at MYC box II and this interaction is critical for transformation to occur (McMahon et al., 1998). TTRAP has also been shown to interact with MYCN in the same fashion (Nikiforov et al., 2002). For MYC’s repressive abilities, it was shown that a fully functional MYC-MAX complex was required for this to occur (Mao et al., 2003). The MYC-MAX complex then interacts with enhancer elements, such as MYC-interacting zinc finger-1, the transcription factor specificity protein 1 or polycomb repression complexes which then causes both displacement of co-activators and the recruitment of co-repressors to the complex (Cole, 2014; Iraci et al., 2011; Meyer and Penn, 2008).
1.2.3 Tissue Expression of Myc Family Genes
Using northern blotting, the expression pattern of the Myc family members was examined in both newborn and adult mouse tissues (Zimmerman et al., 1986). Mycn mRNA expression was highest in newborn brain, kidney and intestine and levels were low to undetectable in adult tissues. The expression pattern of L-Myc was similar to that of Mycn, however, c-Myc expression was widespread across the tissues studied and the levels in the adrenal gland, spleen and thymus were similar in both newborn and adult
Chapter 1- Introduction 18 tissues. Interestingly, in the brain and kidney, c-Myc was detected only in the newborn tissue and like Mycn and L-Myc, disappeared in the adult samples. To examine this further, the pre- and post-natal expression levels of all three Myc family members in the brain from embryonic day (E) 15 through to 21 days post birth was examined. Mycn and L-Myc expression could be detected in the fore- and hindbrain for up to 13 and 15 days post birth, respectively. However, c-Myc expression declined at a more rapid rate in these tissues and could only be detected for up to 11 days post birth in the forebrain and 1 day in the hindbrain. Interestingly, in the kidney, the opposite was true. Mycn and L- Myc expression rapidly declined and was undetectable by 7 days post birth, however, c- Myc was observed at all time points tested (Zimmerman et al., 1986). In more extensive studies, Mycn was analysed early in mouse gestation where it was found to be expressed in the ectoderm during gastrulation and in the central and peripheral nervous systems and ganglia as well as in the gut, kidney and lung during neurulation (the formation of the neural tube from the neural plate), organogenesis (the internal organ formation from the ectoderm, endoderm and mesoderm) and late foetal development (Stanton et al., 1992). Mice that were homozygous mutant did not survive and had life threatening abnormalities by E11.5, with hypoplasia of telencephalic structures, genitourinary defects and spontaneous bleeding resulting in death by E12.5 (Stanton et al., 1992). This was further validated by a conditional Mycn mouse that had disruption of Mycn in the nervous system. Four weeks after birth, null MYCN mice displayed abnormal behaviour (excessive grooming and digging) and ataxia and between 8-16 weeks of age showed signs of microencephaly (reduced head size) (Knoepfler et al., 2002). Complete loss of MYCN function through creation of mutant or null Mycn alleles demonstrated that MYCN was required for normal foetal development. To determine whether MYCN and c-MYC have redundant functions, part of the c-Myc coding sequence was replaced by Mycn coding sequence in mice. These mice did not demonstrate embryonic lethality, that was known to happen when either c-Myc or Mycn was knocked out, and they survived through to adulthood. They were, however, smaller in size and had some dystrophy in skeletal muscle indicating that MYCN and c-MYC cannot completely compensate for one another (Malynn et al., 2000).
Chapter 1- Introduction 19 1.2.4 MYCN Dysregulation in Cancer
Approximately 50% of human cancers possess some form of MYC dysregulation (Tansey, 2014). c-MYC is the most common MYC family member that has been associated with cancer, with overexpression or amplification described in a range of cancers including leukaemia, lymphoma, melanoma, colon and prostate cancer among others (Vita and Henriksson, 2006). However, although MYCN has been best described in neuroblastoma, as discussed earlier, dysregulation of this gene has also been observed in other cancers. For example, MYCN amplification and/or over-expression has been observed in high grade C5 serous ovarian tumours, small cell lung cancer, rhabdomyosarcoma and neuroendocrine prostate cancer (Beltran et al., 2011; Helland et al., 2011; Kim et al., 2006; Tonelli et al., 2012), while gain of 2p (and MYCN) plays a role in chronic lymphocytic leukaemia (Ma et al., 2011). In childhood medulloblastoma, MYCN, c-MYC, and to a lesser extent L-MYC, appear to be involved in the biology of this disease (Ryan et al., 2012). MYCN amplification occurs in up to 10% of medulloblastoma patients and is associated with poor clinical outcome, and like neuroblastoma, the risk of death increases with increasing copy number (Ryan et al., 2012). Furthermore, MYCN expression was found to be high in foetal cerebella, with the levels decreasing to almost absent in adult cerebella, suggesting that MYCN is essential to normal foetal development (Swartling et al., 2010). Interestingly, in this study, MYCN expression was absent from the medulloblastoma cell lines tested, which differed from the expression pattern observed in the primary tumours (Swartling et al., 2010). Finally, as with neuroblastoma, the association of MYCN mRNA levels with clinical outcome remains unclear (Eberhart et al., 2004) and it has been postulated that mRNA levels of both c-MYC and MYCN may only be clinically relevant in subgroups of medulloblastoma (Ryan et al., 2012).
The most compelling evidence for a role of MYCN in the biology of medulloblastoma comes from two mouse models of this disease. Firstly, targeted expression of MYCN to the cerebellum in transgenic mice has demonstrated the importance of MYCN in contributing to the initiation and progression of medulloblastoma and also in the metastatic spread of disease to the spinal and paraspinal tissues via cerebral spinal fluid. Furthermore, the MYCN downstream targets ornithine decarboxylase (Odc1), mouse double minute 2 homolog (Mdm2) and fibrillarin were upregulated and correlated with
Chapter 1- Introduction 20 MYCN mRNA levels (Swartling et al., 2010). The second model used targeted Smoothened to the cerebellar of transgenic mice and then crosses of these mice with conditional knock-outs of MYCN, to demonstrate that MYCN was essential for medulloblastoma tumorigenesis (Hatton et al., 2006). These two models thus serve to demonstrate the importance of MYCN in the initiation and progression of this disease.
1.3 Mouse Models of Neuroblastoma
1.3.1 Early Mouse Models
Prior to the development of transgenic mouse models, which will be discussed below in detail, in vivo studies utilised immunocompromised nude mice to inject human cells subcutaneously into the dorsal flank, to study tumour biology and treatment regimens (Vassal et al., 1996). Newer models now use orthotopic injections of human cells into the adrenal fat pad or space around the kidney to recapitulate the site of neuroblastoma in humans, however, this also requires an immunocompromised mouse (Daudigeos- Dubus et al., 2014). Both of these models rely on the lack of certain immune molecules to ensure that the grafts are not rejected by the mouse’s immune system. With increasing evidence that the tumour microenvironment is important in the biology of cancer, and that immune surveillance and inflammation can play a role in tumour development (Hanahan and Weinberg, 2011), a mouse model of neuroblastoma was needed that not only recapitulated the human disease but was done so in an immunocompetent animal. In the early 1990s several transgenic mouse models of neuroblastoma were developed utilising either Polyoma middle T or SV40 large T antigens (Westermark et al., 2011). Tumour onset ranged from 2 to 10 months and originated in the sympathetic ganglia, adrenal medulla and brain. In one model, metastases in the liver, lung and lymph nodes was also observed. Interestingly, MYCN was upregulated in two of these models (Westermark et al., 2011). These early models provided information about neuroblastoma, but it wasn’t until the creation of the Th- MYCN mouse model, and models that followed this, that larger scale testing of drugs and insight into the biology of the disease increased our understanding of this cancer. These models will be discussed further below.
Chapter 1- Introduction 21 1.3.2 The Th-MYCN Mouse Model of Neuroblastoma
1.3.2.1 Creation of the Th-MYCN mouse
MYCN expression is important during normal foetal development in both humans and mice. The MYCN gene is expressed during embryogenesis along the primitive streak and in cells of neural crest origin (Downs et al., 1989; Zimmerman et al., 1986). Mice with targeted disruption of two copies of the MYCN gene showed embryonic lethality between E10.5 and E12.5 (Charron et al., 1992). Furthermore, prior to the embryos being reabsorbed in utereo, they were shown to be smaller in size, with fewer cells and exhibiting signs of delayed organ development, particularly in the heart and central nervous system, compared to their heterozygous and wild-type littermates (Charron et al., 1992). Taking this evidence and the fact that MYCN amplification occurs in human neuroblastoma tumours, the Th-MYCN mouse model was created by Weiss and colleagues in the late 1990s, providing the first non-viral transgenic mouse model of neuroblastoma (Weiss et al., 1997). The line was developed by placing full length human MYCN cDNA, 4-6 copies inserted head-to-tail, under the control of a rat tyrosine hydroxylase (Th) promoter (Haraguchi and Nakagawara, 2009; Weiss et al., 1997). The Th promoter allowed targeting of MYCN expression to the neural crest, the tissue of origin for neuroblastoma. As Th requires a TATA box for 100% expression, a rabbit beta-globin (β-Globin) intron was inserted between the Th promoter and the MYCN cDNA, to increase the promoting effects of Th (Patankar et al., 1997; Weiss et al., 1997). A schematic of the transgene is shown in Figure 1-2. The transgenic cassette was introduced to murine oocytes and resulting mice demonstrated high expression of human MYCN specifically in the adrenal gland, with minimal expression in other tissues (Weiss et al., 1997).
Several different lines have been developed on different murine backgrounds. There were 3 different strains used in the original experiments, C57BL/6J mice, either alone or crossed with BALB/c, and FVB (sensitive to Friend leukaemia virus B). Out of 24 founder mice, only 4 lines produced mice with tumours. One of the C57BL/6J x BALB/c lines produced mice that had not only expression of the MYCN transgene but also tumour development that was inherited in subsequent generations. These mice were backcrossed to C57BL/6J mice twice and 13 homozygous offspring were followed which demonstrated 100% tumour incidence by 4 months of age (Weiss et al., 1997).
Chapter 1- Introduction 22 Mice developed both abdominal and thoracic masses and the histology was characteristic of neuroblastoma in humans. Comparative Genomic Hybridisation (CGH) analysis of these tumours showed gain of mouse chromosomes 3, 11 and 17 and loss in chromosomes 5, 9, 16 and X (Hackett et al., 2003). Interestingly mouse chromosome 11 is syntenic to human chromosome 17q which is frequently gained in human neuroblastoma tumours and is associated with advanced disease and MYCN amplification (Bown et al., 1999; Hackett et al., 2003). Also noted from the CGH array was amplification of Chromosome 18, which was hypothesized to harbour the MYCN transgenic cassette (Hackett et al., 2003). This region was later sequenced and the transgene was found to be integrated at 18qE4 and matched the original pTh-MYCN plasmid created by Weiss and colleagues (Haraguchi and Nakagawara, 2009; Rasmuson et al., 2012).
To increase the penetrance of the phenotype, the tumour bearing Th-MYCN lines were backcrossed to 129/SvJ mice, and subsequent publications have used this strain of mice. This strain has a greater tumour penetrance than other strains tested, and it has been hypothesized that this is due to a modifier that is unique to 129/SvJ (Teitz et al., 2011). To determine the unique modifier, Th-MYCN mice on a 129/SvJ background were crossed to FVB/NJ mice, which had already demonstrated poor tumour incidence in the original studies (Hackett et al., 2014; Weiss et al., 1997). Linkage mapping using microsatellites and single nucleotide polymorphism (SNP) markers of these mouse populations revealed Liver arginase (Arg1) on chromosome 10 as a potential candidate. Mice with higher Arg1 expression demonstrated higher tumour penetrance. Arg1 is a member of the gamma-amino butyric acid synthesis pathway that produces polyamines, a pathway necessary for cell growth (Hackett et al., 2014). A secondary susceptibility loci was also identified on chromosome 4, however any potential candidate gene(s) in this region remains to be determined (Hackett et al., 2014). Identification of genes and pathways involved in the biology of Th-MYCN tumorigenesis has the potential to provide novel targets for therapy of this disease.
Chapter 1- Introduction 23
0.1kb Gene Terminator 0.7kb β-Globin
4.8kb Tyrosine Hydroxylase Promoter 1.7kb Human MYCN cDNA
Figure 1-2 Schematic of the Th-MYCN cassette. A 4.8 kb rat tyrosine hydroxylase promoter was inserted upstream of human MYCN cDNA (1.7kb). The rabbit beta-globin intron (0.7 kb) was used to increase expression of the promotor while transcription was terminated by the thymidine kinase gene from herpes simplex virus (0.1 kb). Adapted from Weiss et al., 1997 and Haraguchi and Nakagawara, 2009.
Chapter 1- Introduction 24
1.3.2.2 Characterisation of the Th-MYCN mouse model
To further investigate the applicability of this model for studying neuroblastoma, several groups have examined the biology of this mouse line as well as the response to therapeutic regimens, both established treatments for neuroblastoma, as well as experimental drugs. Th-MYCN mice on a background of 129/SvJ develop both abdominal and thoracic tumours, and the penetrance of the disease is dependent on the gene dosage (Hansford et al., 2004). Mice with two copies of the MYCN transgene (homozygotes) all develop neuroblastoma by 7 weeks of age, while those harbouring only one copy of the transgene (hemizygotes) have reduced tumour incidence and longer latencies with only 25-35% developing tumours at a median age of 13 weeks (Burkhart et al., 2003; Hansford et al., 2004). A histological audit of these mice, beginning at E14 and continuing until 20 weeks of age, revealed that all mice, regardless of transgene dosage, had the same percentage of paravertebral ganglia with neuroblast hyperplasia, defined as >30 small round blue cells in the ganglia, at day 0 (Hansford et al., 2004). In the wild-type mice (those that lacked the transgene) this hyperplasia dissipated by two weeks of age and was not observed in subsequent time points (Hansford et al., 2004). However, mice carrying the MYCN transgene underwent incomplete regression such that those hemizygous for the transgene showed a delay in regression with neuroblast hyperplasia extending out to 6 weeks of age, while mice homozygous for the transgene failed to show complete regression before developing a tumour by 6.5 weeks (Hansford et al., 2004). Furthermore, in the hemizygous mice that did develop tumours, the level of MYCN expression in their tumours was found to be similar to that observed in homozygote tumours, due to the human MYCN transgene undergoing amplification in these cells. This result suggests a threshold level of MYCN expression that needs to be achieved for tumorigenesis to occur (Hansford et al., 2004; Norris et al., 2000). Primary cultures of superior cervical (SCG) and coeliac (CG) ganglia from Th-MYCN mice at 2 weeks of age demonstrated that wild-type ganglia responded to NGF withdrawal and died, however homozygotes were resistant to this type of death stimuli (Hansford et al., 2004).
Pathology studies of the neuroblastoma tumours arising from Th-MYCN mice have revealed these tumours mimic the biology of aggressive, undifferentiated or poorly
Chapter 1- Introduction 25 differentiated human neuroblastoma (Teitz et al., 2011). As the tumours increase in size, they envelop the kidneys, adrenal glands and other vital structures (Rasmuson et al., 2012). Furthermore, microarray data comparing human and mouse neuroblastomas revealed this model has a gene expression pattern that mostly resembles Stage 3 or 4, MYCN amplified neuroblastoma (Teitz et al., 2011). Interestingly, using ultrasound and MRI, it has also been noted that 10% of hemizygous mice also demonstrate spontaneous regression of their tumours, as observed in Stage 4S patients, between 8-10 weeks of age (Teitz et al., 2011).
In the original mouse lines, a small number of animals had gross metastatic lesions to the liver, lung and ovary (Weiss et al., 1997). Subsequent microscopic evaluation of organs from tumour bearing 129/SvJ mice aged between 3-5 months revealed metastatic lesions of <1mm3 in several tissues including the lungs, ovaries and lymph nodes (Teitz et al., 2011). However, the one limitation of this model is the lack of substantial bone marrow metastasis with fewer than 5% of mice having microscopic lesions at this site, which does not recapitulate the human disease where metastasis to cortical bone and bone marrow is frequently observed (Brodeur and Castleberry, 1997; Park et al., 2010; Teitz et al., 2011).
1.3.2.3 Therapeutic targeting of neuroblastoma using the Th-MYCN mouse
As the Th-MYCN mouse represents an excellent model of the human disease, it is ideal for understanding neuroblastoma tumour initiation and for testing both established and novel treatment options. Although the high level and irregular tumour vasculature observed in these mice makes surgical resection an extremely difficult proposition (Chesler et al., 2007), standard chemotherapy can be used and the 5 main drugs utilised in high-risk induction therapy have been tested in this model. Treatment of homozygous Th-MYCN mice with VP-16 or VCR, results in only a short extension of life span (Burkhart et al., 2009). This poor response is due, in part, to the high levels of the multidrug resistance-associated protein 1 (Mrp1) that is over-expressed in this model (Cheng et al., 2007; Norris et al., 2000). In neuroblastoma patients, high levels of MRP1 in the tumour have been shown to correlate with poor clinical outcome (Haber et al., 2006; Norris et al., 1996). Cyclophosphamide and cisplatin, both non-MRP1
Chapter 1- Introduction 26 substrates, are more effective than either VP-16 or VCR in this model. Mice not only have an increased median life span when treated with either of these drugs, but they can also achieve long-term cure (Chesler et al., 2008; Hogarty et al., 2008). The final drug used in high risk therapy is doxorubicin. While doxorubicin is an MRP1 substrate in humans, it is not a substrate for this transporter in mice (Stride et al., 1997) and as such, doxorubicin behaves more like cisplatin and cyclophosphamide in the murine setting, although mice do not become long-term tumour free survivors.
The current relapse protocol for the treatment of neuroblastoma is a combination of the frontline alkylating agent cyclophosphamide and the topoisomerase inhibitor topotecan (London et al., 2010). In a recent COG study, the relapse protocol of cyclophosphamide/topotecan showed that 32% of patients treated with this combination had a complete or partial response (London et al., 2010). This type of chemotherapy drug combination can also be applied to the Th-MYCN mouse model. By lowering the dosages of the individual drugs used in treating the mice, an average therapeutic response of 20-30 days can be achieved similar to higher dose cyclophosphamide (Carter et al., 2015). This combination has proven to be so effective when treating relapsed neuroblastoma in children that it is likely to become part of frontline therapy for high risk disease (Park et al., 2011) and as such a new relapse protocol is being sought. Two related drugs, irinotecan, a topoisomerase inhibitor and temozolomide, an alkylating agent are likely to become the new relapse protocol when cyclophosphamide/topotecan becomes part of the initial upfront treatment (Bagatell et al., 2011). As this is likely to become the relapse protocol, testing in the Th-MYCN model was undertaken, and this combination, while not being as beneficial as cyclophosphamide/topotecan was still able to prolong survival for an average 20-25 days (Carter et al., 2015).
After establishing dose and regimens of clinically relevant drugs to use in this model, small molecule inhibitors, modifiers or novel compounds can be tested in combination with standard-of-care therapy to investigate and hopefully enhance the effects of the chemotherapeutic compounds, such as those discussed below.
Chapter 1- Introduction 27 1.3.2.4 Therapeutic Targeting of the MYCN pathway in Th-MYCN mice
Direct targeting of MYC or MYCN itself has proven difficult. The development of molecules or drugs that target these transcription factors has been hampered by a number of factors, including the lack of a crystal structure to design small molecule inhibitors to, the localisation of MYC in the nucleus, and the need for molecules that interfere with large protein-DNA surface areas rather than binding in an active site (Posternak and Cole, 2016). Coupled with the fact that the MYC family can regulate up to 15% of the human genome has led to the belief, up until recently, that the MYC family are undruggable targets (Horiuchi et al., 2014). There is now increasing evidence that this can be possible with several studies using the Th-MYCN mouse model focussing on targeting MYCN directly with some success.
Inhibition of human MYCN using antisense oligonucleotides resulted in a significant decrease in hemizygous Th-MYCN mice developing tumours with less than 10% tumour incidence observed compared to 42% and 31% in the scrambled oligonucleotide and untreated control mice, respectively (Burkhart et al., 2003). All of the homozygous mice, regardless of treatment developed neuroblastoma, however, the tumour burden in the antisense treated group was significantly decreased with smaller tumours observed by comparison with the control groups (Burkhart et al., 2003). Taken together, these results highlighted the importance of MYCN in driving tumour development in this model, although antisense oligonucleotides are unlikely to be used clinically and hence other methods of MYCN inhibition must be sought.
The Bromodomain (BRD) and Extraterminal Domain (BET) family are epigenetic readers that activate transcription through binding of modified histones in chromatin (Braun and Gardin, 2017). BET inhibitors have proved the most successful in inhibiting MYC and MYCN, with JQ1 and derivatives currently in pre-clinical testing in multiple models (Horiuchi et al., 2014). JQ1 displaces one of the BET family members, BRD4, from chromatin in the MYCN promoter by competitive binding (Puissant et al., 2013). Th-MYCN mice treated with JQ1 for 28 days resulted in a decrease in the level of MYCN protein and a 30-60 day extension of life span, compared to the controls, which survived less than 20 days (Puissant et al., 2013). 10058-F4, an inhibitor of the MYC-
Chapter 1- Introduction 28 MAX and MYCN-MAX interactions was found to also cause cell cycle arrest, apoptosis and neuronal differentiation (Zirath et al., 2013). When tested in Th-MYCN mice with palpable tumours, a decrease in the level of MYCN protein and a small, but significant delay of 21 days survival, was observed compared to the controls which survived for only 11 days post treatment (Zirath et al., 2013). These studies not only highlight the relevance of the Th-MYCN model in pre-clinical testing but also the importance of MYCN in the biology of this disease. If we can effectively inhibit MYCN in amplified patients, then a new avenue of therapeutic options would be available for patients with high risk disease.
MYCN can act on a range of genes involved in many different cellular functions including cell cycle, migration and invasion, proliferation, angiogenesis and anti- angiogenesis, apoptosis and differentiation as either a transcriptional activator or a repressor (Gherardi et al., 2013; Huang and Weiss, 2013). MYCN can also regulate ABC transporters, histone deacetylases (HDACs) and polyamine synthesis (Henderson et al., 2011; Hogarty et al., 2008; Liu et al., 2007a; Marshall et al., 2011; Porro et al., 2010).
To overcome drug resistance to VCR and VP-16 in neuroblastoma, a small molecule inhibitor to MRP1 (Reversan) was developed. Treatment with Reversan was able to extend life in Th-MYCN mice when co-treated with either VCR or VP-16, a finding that was also replicated in BALB/c nude mice bearing human BE(2)-C xenografts (Burkhart et al., 2009).
HDACs remove acetyl groups from histones, thereby inactivating chromatin, which in turn regulates gene transcription. HDACs and their counterparts, histone acetylases, are often regarded as the chief epigenetic regulator (Parbin et al., 2014). Targeting HDACs that are MYCN regulated may also be useful in neuroblastoma treatment. Tenovin-6 and Cambinol, two inhibitors to the class III HDAC, SIRT1 (Silent mating type information regulation 2 homolog 1) have been tested in the Th-MYCN mouse model. Both Cambinol and Tenovin-6 were able to decrease tumour volume in this model when given prophylactically at 5 days and as a therapeutic option at 28 days, respectively (Marshall et al., 2011). In addition it has been shown that MYCN is able to down- regulate tissue transglutaminase (TG2) and treatment of Th-MYCN homozygotes with the HDAC inhibitor Trichostatin A is able to increase TG2 expression and decrease the
Chapter 1- Introduction 29 tumour volume by 8-fold (Liu et al., 2007a). Finally, the pan-HDAC inhibitor panobinostat, was not only able to significantly extend life span in Th-MYCN mice, but was also able to cause an increase in cellular differentiation in the tumour as well as a reduction in MYCN protein after only 7 days of treatment (Waldeck et al., 2016).
Polyamines, such as spermidine, spermine and putrescine, are polycations required for cell growth, and MYC driven cancers appear to rely heavily on polyamine synthesis to drive disease (Bassiri et al., 2015). The rate-limiting step in the polyamine pathway is controlled by ornithine decarboxylase (ODC1), which is also regulated by MYCN and is itself a predictor of poor outcome in neuroblastoma patients (Hogarty et al., 2008). Therapeutic targeting of Odc1 using difluoromethylornithine (DFMO), either at birth or after weaning was able to increase the survival time of homozygote Th-MYCN mice, and when combined with cisplatin or cyclophosphamide was able to significantly extend survival further (Hogarty et al., 2008). A clinical trial in relapsed neuroblastoma patients involving the use of DFMO in combination with the relapse backbone of cyclophosphamide/topotecan is currently underway (NCT02030964).
1.3.2.5 Testing of novel drug targets in Th-MYCN mice
The importance of having an intact immune system and the correct microenvironment for therapeutic applications was highlighted in a study using this model by treating with the anti-angiogenic fumagillin analogue TNP-470. In a pancreatic animal model, TNP- 470 was shown to be effective when used in combination with an immune stimulator, which stimulated cytotoxic T-cells (Matsumoto et al., 2003). Previous studies in nude mice, which largely lack T-cell populations, had found that TNP-470 was only effective against the sub-cutaneously xenografted neuroblastoma cell line NBL-W-N, when the tumours were treated either soon after inoculation or before a palpable tumour was evident (Katzenstein et al., 1999; Katzenstein et al., 2001). Once a small tumour (<400mm3) was detected, the effect was still significant but diminished and there was no effect if treatment started once a tumour was greater than 400mm3 (Katzenstein et al., 1999). In contrast, TNP-470 was effective against an established Th-MYCN tumour. Upon euthanasia, the tumour burden in treated mice was greatly reduced compared to controls and animals in the treatment group were only euthanased due to significant weight loss rather than for tumour end-point (Chesler et al., 2007). Although this effect
Chapter 1- Introduction 30 may be due to the site of the tumour and the mice having an intact immune system, it may also be due to the differences between murine and human neuroblastoma. A decrease in tumour burden was also observed in Th-MYCN mice treated with low dose aspirin. Interestingly, the authors also noted the presence of tumour promoting immune molecules such as tumour-associated macrophages, dendritic cells and myeloid derived suppressor cells in control mice that were reduced in the aspirin treated group (Carlson et al., 2013).
Beta-blockers (β-blockers) are responsible for blocking the binding sites for adrenaline and noradrenaline in the nervous system. Two β-blockers, propranolol and carvedilol, also have both anti-proliferative and anti-angiogenic effects in cancer cells. Treatment of the Th-MYCN mouse with either of these two agents slowed tumour progression, and when used in combination with VCR, increased median survival to 30 and 29 days, respectively, compared to 7 days for the VCR alone control (Pasquier et al., 2013).
CBL0137, a small molecule inhibitor that activates p53 and inhibits nuclear factor kappa-light-chain-enhancer of activated B cells (abbreviated to NF-κB), has been shown to be effective against a variety of cancers as a single agent. CBL0137 exerts its action by trapping the Facilitates Chromatin Transcription (FACT) complex. Despite its intimate association with DNA and FACT, CBL0137, unlike other DNA damaging agents, has been shown to be non-genotoxic (Gasparian et al., 2011). CBL0137 has also demonstrated efficacy in combination with gemcitabine in a mouse model of pancreatic cancer (Burkhart et al., 2014). Treatment of Th-MYCN mice with CBL0137 monotherapy resulted in an extension of life span when given orally, comparable to cisplatin or cyclophosphamide or when given intravenously, in complete cure of the majority of mice (Carter et al., 2015). When a lower dose of CBL0137 was combined with the five drugs given in frontline therapy (cisplatin, cyclophosphamide, VP16, VCR or doxorubicin), a significant extension of life span was achieved and a large proportion of mice were completely cured compared to tumour progression observed with the single agent drugs. When CBL0137 was combined with either the current relapse protocol of cyclophosphamide/topotecan or the new relapse protocol irinotecan/temozolomide, a significant extension of life and complete cure in a large proportion of mice was also observed. These findings were repeated in BE(2)-C xenografted nude mice and a similar trend was found for the two relapse protocols further highlighting the applicability of this model (Carter et al., 2015).
Chapter 1- Introduction 31
1.3.2.6 Th-MYCN mouse models crosses
The Th-MYCN mouse can also be used to discover the importance of a gene on neuroblastoma initiation, progression or therapeutic response, by crossing this model with other mutant or knock-out models of interest. Several studies have undertaken this approach and are highlighted below.
Mutation, amplification or fusion of the ALK oncogene has previously been described in the 1-2% of neuroblastoma tumours that can be attributed to a genetic pedigree rather than spontaneously arising (Ogawa et al., 2011). Mutations in this gene have been observed in approximately 8% of neuroblastoma patients across a number of studies (Berry et al., 2012). To investigate the role of ALK mutations further, the most common mutation, consisting of a phenylalanine to leucine change at position 1174 in the kinase domain of the ALK protein (F1174L), was cloned and injected into murine blastocysts under the control of the rat Th promoter, the same as in the Th-MYCN mice (Berry et al., 2012). The mice were then crossed with the Th-MYCN mouse model. Only one copy of MYCN and one copy of the ALK mutation were sufficient to cause 100% tumour penetrance in these mice (Berry et al., 2012). In the two founder lines where 25-35% tumour incidence was observed in the hemizygous MYCN mice without the ALK mutation, the median latency was significantly shorter in those that possessed the ALK mutation, indicating that the ALK mutation not only increased penetrance but also progression of the disease (Berry et al., 2012). It has previously been reported that mice that are hemizygous for the MYCN transgene show an increase in MYCN copy number and expression in their tumours (Hansford et al., 2004; Norris et al., 2000). In the MYCN/ALK mice, not only an increase in the human MYCN transgene was observed, but also an upregulation in the murine Mycn gene, which was not observed in the hemizygous MYCN mice without the ALK mutation (Berry et al., 2012).
Caspase-8, a gene that cleaves peptide bonds, and is critically involved in the programmed cell death cascade, has also been shown to have roles in migration, growth and differentiation, among others (Teitz et al., 2013). Caspase-8 mice with a LoxP site were mated to both Th-Cre and Th-MYCN mice to produce a knockout model of caspase-8 with targeted loss of expression in the peripheral neural crest lineage cells
Chapter 1- Introduction 32 that develop neuroblastoma tumours. Knockout of caspase-8, either heterozygote or homozygote, did not significantly alter tumour incidence in the Th-MYCN mouse with the majority of tumours developing near the aorta, and then in the adrenal gland or kidney (Teitz et al., 2013). However, interestingly, despite not seeing a difference in primary tumour formation, there was an increase in metastasis to bone marrow in approximtely a third of the mice that had caspase-8 deficiencies. Metastasis to bone and bone marrow is frequently reported in advanced stage neuroblastoma, making this mouse model potentially extremely important for studying metastatic disease, something which the Th-MYCN mouse model lacks (Brodeur and Castleberry, 1997; Park et al., 2010; Teitz et al., 2013; Teitz et al., 2011).
Mutations in the tumour suppressor p53 in neuroblastoma are extremely rare at diagnosis and are often only seen following relapse (Imamura et al., 1993). To examine the direct effect of loss of p53 in neuroblastoma biology, heterozygote p53 null mice were crossed with Th-MYCN mice. Mice with loss of one copy of p53 not only developed tumours at a higher incidence, but they also had shorter latency compared to mice with intact p53 (Chesler et al., 2008). Although the macroscopic appearance of the tumours was similar to those with 2 copies of p53, on microscopic investigation, there was an increase in proliferating neuroblasts as well as reduced apoptosis and necrosis, indicating that by inactivating p53, MYCN was able to exert its function to further drive proliferation (Chesler et al., 2008). A second study also examined loss of MDM2, an inhibitor of p53 function, in Th-MYCN mice. Loss of one copy of Mdm2 was able to decrease the incidence and increase the latency of neuroblastoma tumour formation compared to mice with two copies of Mdm2 (Chen et al., 2009). Interestingly, there was a difference observed in the male and female mice with the females having a significantly greater decrease in tumour incidence and the authors hypothesized that this may be due to the interactions between MDM2 and oestrogen-dependant signalling (Chen et al., 2009).
Crossing of the Th-MYCN mouse model with a knock-out model of clusterin, a gene that is both negatively regulated by MYCN and a tumour suppressor, demonstrated that loss of clusterin decreased survival in both homozygote and hemizygote MYCN mice, although significance was only reached when the homozygote and hemizygote mice were pooled, or in the homozygotes when two copies of clusterin were lost (Chayka et al., 2009). When the tumours were investigated for clusterin protein, almost no
Chapter 1- Introduction 33 expression was detected in the homozygote tumours independent of their clusterin genotype, further validating the suppression of clusterin by MYCN (Chayka et al., 2009). Finally knock-out of the heparin-binding growth factor midkine resulted in a 1-2 week delay in tumour progression in hemizygous MYCN mice and a small increase in tumour incidence (Kishida et al., 2013).
All these studies highlight the usefulness of the Th-MYCN model in elucidating the role of genes in neuroblastoma biology.
1.3.3 Other mouse models of neuroblastoma
Despite being an excellent mouse model to use, there are several limitations of the Th- MYCN mouse model that have been discussed above including the lack of metastasis to bone and bone marrow, the majority of tumours arising from the sympathetic ganglia and not the adrenal medulla which differs from the clinical setting, the lower incidence of tumour formation in strains other than the 129/SvJ and the lack of a reporter gene to image the mice. The lack of metastasis may, in part, be rectified by using the Th- MYCN/caspase-8 model discussed above, but the site of tumour formation and lack of a reporter gene would require significant modification or creation of a new model. Currently, to overcome the strain variation in tumour incidence, the mice carrying the knock-out or transgenic gene of interest are backcrossed onto 129/SvJ mice before crossing them to the Th-MYCN mice. However, different laboratories have tried to correct these issues by creating new transgenic mouse models of neuroblastoma for study.
Palpation is the current method of tumour detection and while this can be done successfully using Th-MYCN mice, it requires experienced personnel. Other forms of tumour detection include the use of ultrasound, small animal positron emission tomography and MRI (Quarta et al., 2013; Teitz et al., 2011). A Cre-conditional, MYCN driven mouse model has also been created that carries the construct for luciferase expression and can therefore be imaged. The mice, termed LSL-MYCN have a loxP-flanked strong transcriptional termination site (loxP-STOP-loxP or LSL) that was inserted along with the human MYCN gene under a chicken actin gene promoter as well as a luciferase gene construct (Althoff et al., 2015). These mice were then crossed with
Chapter 1- Introduction 34 the Dbh-iCre mice, which contain an iCre sequence and a dopamine-β-hydroxylase (Dbh) promoter (Stanke et al., 2006). Mice develop tumours that are bioluminescent in the SCG, CG and adrenals. Tumour incidence is approximately 75% in double transgenic LSL-MYCN;Dbh-iCre mice, with a latency between 26-337 days (Althoff et al., 2015). Gain of mouse chromosome 11q, syntenic to human chromosome 17q, has been observed in this model, similar to the Th-MYCN mouse model (Althoff et al., 2015; Hackett et al., 2003). This model provides an alternative to the Th-MYCN mice, with the one advantage being the ability to image the animals using bioluminescence.
A common ALK mutation in humans (F1174L) was introduced to mice using a Cre inducible model under control of tyrosine hydroxylase, to direct expression to the neural crest. Five out of 12 founder mice containing both the ALK mutation and Cre recombinase developed poorly differentiated neuroblastoma tumours, primarily in the abdominal region, and arising specifically from the adrenal gland (Heukamp et al., 2012). No bone marrow metastases were identified, however, two of the mice developed liver metastases. In one line, high ALK expression was accompanied by MYCN amplification. When ALK mutated mice were then crossed with Th-MYCN mice, 100% tumour penetrance by 48 days of age was observed. The Th-MYCN colony used in this cross only develops neuroblastoma at a rate of one in two by 100 days of age, hence the addition of the ALK mutation greatly accelerated both tumour formation and progression (Heukamp et al., 2012).
1.4 Gene Discovery Approaches
Forward and reverse genetics methodologies are two well established approaches that provide the tools to understand and elucidate the function of genes within organisms and in the disease process. In reverse genetics, a gene is targeted using a range of available techniques to alter the function of the gene, a phenotype is followed and a biological function for that gene is elucidated. In forward genetics, a phenotype-driven methodology is used to discover the gene responsible for that phenotype. These types of screens use mutagenesis to create mutations in the model of interest to discover the causative gene. Both approaches have advantages and disadvantages and both can be used to discover the role of genes within your organism or disease of interest.
Chapter 1- Introduction 35 1.4.1 Reverse genetics
Reverse genetics methodologies have been used for many years, either as a candidate gene approach or as confirmation of the genes discovered in forward genetics screens and will be discussed below.
1.4.1.1 Site-directed mutagenesis and genome editing
Flavell and colleagues developed site-directed mutagenesis in the mid-1970s, where they successfully targeted mutations into regions of the genome that were highly conserved in bacteriophage (Flavell et al., 1974, 1975). The ability to target regions of interest, rather than random mutagenesis, meant the function of a gene could be studied in detail. Site-directed mutagenesis was used for many years, with success, and in recent years more sophisticated methods of genome editing have been developed. Such genome editing can be used to generate point mutations, knock-out gene function or even to introduce foreign DNA into the genome. The technology uses 4 main nucleases to achieve this, namely zinc-finger nucleases, clustered regularly interspaced short palindromic repeats (CRISPR)/CRISPR associated protein 9 (Cas9) (CRISPR/Cas9), transcription activator-like effector nucleases (TALENs) or RNA-guided endonucleases (reviewed in (Hendel et al., 2015).
Genome editing can be used to complement forward genetics screens. To confirm the findings of a mutagenesis screen in osteosarcoma, TALEN was used to knock-out the genes found to be responsible for osteosarcoma development in an immortalized osteoblast cell line. Mutation of phosphatase and tensin homolog (PTEN) occurred in 60% of cells and showed an increase in soft-agar colonies, further validating the gene identified in the forward genetics screen (Moriarity et al., 2015).
1.4.1.2 Transposable element excision
Transposable elements are DNA sequences that can move locations in the genome. Instead of taking a genome wide approach to mutation using transposable elements, where insertion is random, a more targeted approach can be undertaken. Generating
Chapter 1- Introduction 36 mouse models, while extremely useful in determining the role of a gene, are both time consuming and costly to generate. Transposable elements like Sleeping Beauty, which will be discussed further in the forward genetics section below, can be engineered to deliver vectors containing the gene of interest or to knock down a gene. Thus, this method has been successfully used to deliver angiogenic inhibitors to both subcutaneous and intracranial models of glioblastoma and to colorectal cancer liver metastasis demonstrating an increase in survival in all three models (Belur et al., 2011; Ohlfest et al., 2005). Alternatively, transposons can also be used to deliver oncogenes to cause spontaneous brain cancer in the cerebral hemisphere (Wiesner et al., 2009).
1.4.1.3 Silencing using RNAi
The generation of tools for reverse genetics, such as RNA interference (RNAi) using short interfering RNA (siRNA) or short hairpin RNA (shRNA) and CRISPR/Cas9 made silencing of genes possible. RNAi technology can be used to silence a gene in mice and provide valuable tools for determining the role of a gene, and can be used either to confirm a gene isolated from a forward genetics screen or to investigate a gene of interest. For example, to confirm the role of Foxp2, a transcription factor isolated from a forward genetics transposable element screen in pancreatic cancer, both siRNA and shRNA knock down cell lines were established. Regardless of the technique employed to knock down the gene, a decrease in cell viability and colony formation was observed further validating these screening approaches (Rad et al., 2015). shRNA targeting of LIN28B, an RNA-binding protein, in neuroblastoma demonstrated cell cycle arrest, and neuronal differentiation (Molenaar et al., 2012).
1.4.1.4 Knock-out mouse models
Knock-out or knock-in mouse models have been widely used to validate the role of a gene in a specific disease model. Various methods of creating these mouse models have been employed. Targeted disruption of the gene by using bacterial resistance genes, for example neomycin or hygromycin, is a common method of disrupting the gene function as well as providing a means to genotype the mice, and have been used in many models, as for example in studying the role of multidrug transporters in conferring drug
Chapter 1- Introduction 37 resistance or sensitivity (Lorico et al., 1997; Zelcer et al., 2006). Knock–out models crossed with the Th-MYCN model were discussed in detail above. Newer technologies using CRISPR/Cas9 to create knock-out mouse models have recently been published (Hara et al., 2015; Miura et al., 2015). Briefly, for a transposable element forward genetics screen validation, PTEN was identified as a candidate gene in osteosarcoma, and Pten was knocked-out using a conditional Cre model. Loss of PTEN, resulted in not only increased tumour penetrance but also a decreased latency, further validating this gene in osteosarcoma and the robustness of the screening approach (Moriarity et al., 2015).
1.4.2 Forward genetics
Forward genetics screens have used 4 main mutagenesis technologies including X-rays, chemicals, transposons or qualitative trait locus mapping, and each will be discussed below.
1.4.2.1 Mutagenesis using X-Rays
The first work on the role of X-rays causing heritable mutations came from studies in Drosophila by Muller in 1927 and in tobacco plants by Goodspeed and Olson in 1928. Muller’s studies described the treatment of Drosophila sperm cells with high doses of radiation that resulted in a seventh of the offspring acquiring mutations (Muller, 1927). He also noted that the numbers of lethal mutations outweighed the non-lethal mutations, and therefore there was a need to test a range of radiation doses to reduce the lethality while still obtaining a sufficient mutation rate for study. In Nicotiana plants, Goodspeed and Olson studied over 1000 offspring that were grown from two mature X-rayed tobacco plants, observing a 20% mutation rate (Goodspeed and Olson, 1928). Both studies noted the existence of both advantageous (increased growth and fertility) and non-advantageous (sterility and lethality) mutations resulting from the X-Rays. Olson and Lewis also theorised in 1928 that natural sources of radiation were responsible for producing variants, at a rate of 2 variants out of 1000 tobacco plants, and were the likely cause of evolution as described by Darwin almost 70 years earlier (Olson and Lewis, 1928). Studies then moved to higher order species like mice, rats and guinea pigs. To ascertain the period of sterility in the male mouse following X-Ray exposure, a series of
Chapter 1- Introduction 38 experiments involving either increased single dose or fractionated doses were carried out and fertility tested. With increasing doses, the period of sterility increased and at 1000 Rad, the median sterility was over 100 days (Cattanach, 1974). In fractionated experiments, a combined dose of 1000 Rad also resulted in permanent sterility in a number of mice as their testes were smaller and there was patchy spermatogenesis (Cattanach, 1974). In both methods, smaller litters were produced with increasing doses, and it has been postulated that the number of dominant lethal mutations may be responsible for this, not only in the mouse, but also in larger mammals such as the guinea pig and hamster where similar results were observed (Cattanach, 1974; Cox and Lyons, 1975). The specific locus mutation rate in mice at a single dose of 1000 Rad was less than 1/10000, but could be increased to between ~1/2000 to 1/5000 if fractionated doses to 1000 Rad were given with an interval ranging between 24 hours to 7 days (Cattanach and Moseley, 1974). Balancing sterility with effectiveness, X-Ray mutagenesis studies have been carried out in a number of mouse models of disease, but mainly ones involving either leukaemia or lymphoma. For example, to induce thymic lymphoma, mice were exposed to increasing doses of X-rays resulting in increased tumorigenesis, with the maximum incidence peaking at 36% in the mice studied (Hirano et al., 2013). X-ray and chemical mutagenesis were combined in this model as discussed below.
1.4.2.2 Mutagenesis using chemicals
The field of mutagenesis exploded with the discovery of chemical mutagens. The alkylating agent N-ethyl-N-nitrosourea (ENU) was discovered to be a powerful chemical mutagen in mice by William Russell and colleagues in the late 1970s (Russell et al., 1979). ENU works by transferring the ethyl group (CH2CH3) of ENU to either nitrogen or oxygen radicals in DNA, and DNA adducts are created by this transfer, which causes mispairing during DNA replication (Noveroske et al., 2000). A dose of 250mg/kg of ENU was found to be 5 times more mutagenic than 600 Rad of X-rays, which had previously been determined to be the effective dose, and 15 times more effective than the most powerful chemical mutagen to date, procarbazine (Russell et al., 1979). In specific locus mutation tests, sperm were found to be mutated at a rate of 1.2 in a 1000 (Hitotsumachi et al., 1985). To determine the nucleotide substitution rate using ENU, sequencing of 9.6 mega bases revealed mutations occurred at 1.06 in 1
Chapter 1- Introduction 39 million bases (Concepcion et al., 2004). This mutation rate would result in each gamete harbouring approximately 1000 mutations, of which only 1/500 would be functional (Nguyen et al., 2011). In the murine pre-meiotic germline, the mutations are more likely to be either AT to TA transversions or AT to GC transitions, however, in transgenic mice, the transition is more likely to be an GC to AT (Noveroske et al., 2000). The majority of mutations that result from injection with ENU are point mutations, with over half of these being missense mutations and the remainder comprising splicing or nonsense mutations (Nguyen et al., 2011).
One important factor to consider in an ENU mutagenesis screen is the balance between achieving an effective dose that causes enough mutations and the sterility that is caused by the mutagen. A cumulative dose that is too high, or repeated doses with intervals that are too short, can cause permanent or long periods of sterility of up to 40 weeks in the mice (Hitotsumachi et al., 1985). A 7 day interval between ENU injections is recommended, which reduces the risk of death within one month of injection and allows the majority of mice to survive for at least 6 months post injection to generate offspring (Hitotsumachi et al., 1985). However, a dose and interval rate of ENU cannot be applied to all mouse strains and careful consideration should be made prior to injection. Mice with an FVB/N background are the most sensitive to the effects of ENU injection with fertility being completely lost and an extremely high rate of death after injection with a dose and schedule that is well tolerated by C57BL/6J and BALB/cR1 mice (Justice et al., 2000). Regardless of the strain, mice injected with ENU will have a decreased life span compared to their untreated counterparts, are immunosuppressed and more likely to develop infections as a result, and be prone to developing cancer, unrelated to the mouse model (Justice et al., 2000).
To test for heritability, mice are mated to an unrelated mouse of either the same background or different backgrounds to elucidate dominant mutations. By crossing to a different genetic background at this point, later mapping of the mutation becomes easier, however, the phenotype may be more robust when the same background is used (Rubio-Aliaga et al., 2007). If inheritance is Mendelian, 50% of the offspring should possess the mutated phenotype, while 50% should not (Acevedo-Arozena et al., 2008). For recessive mutations, mice that do not possess a visible phenotype are mated to either their parents or to a sibling. When the progeny are then crossed to normal mice, if
Chapter 1- Introduction 40 an abnormal phenotype is observed and the inheritance is Mendelian, 25% of the offspring should have the unusual phenotype (Nguyen et al., 2011).
To determine the gene responsible for the change in phenotype a combination of mapping and sequencing is required. The background strain of the mouse is important in this process. Mice must be outcrossed to a background that is well described, for example C57BL/6 or BALB/c, as these strains have already been fully sequenced (Nguyen et al., 2011). The phenotype must be outcrossed to a well described inbred background, and then the progeny are either backcrossed or intercrossed and the phenotype followed (Nguyen et al., 2011). Exome sequencing can be used as a tool as almost all ENU mutation screens have yielded mutations only in the coding sequence (Kile and Hilton, 2005; Nguyen et al., 2011). Once the region has been narrowed, whole genome sequencing can be used to determine the exact mutation responsible for the change in phenotype.
Large scale ENU mutagenesis libraries have been created at a number of institutions including UK ENU Mutagenesis Programme at Harwell, the German ENU-mouse mutagenesis screen project, the Australian Phenomics Facility, and at RIKEN in Japan, which are accessible to the public as a resource (Andrews et al., 2012; Gondo, 2010; Hrabe de Angelis et al., 2000; Nolan et al., 2000). In the UK mutagenesis project, Nolan and colleagues analysed 26,047 progeny from normal ENU treated BALB/c mice and assessed them for visible abnormalities of the limbs, tail and craniofacial regions (Nolan et al., 2000). Visible changes or defects were observed in motor neuron skills and sensory, neuropsychiatric, spinocerebellar and autonomic functions of 1089 mice. Of those mice, 126 phenotypes were heritable and 50% of these showed dominant heritability traits. Thirty mutants were mapped which revealed 13 novel mutations and 3 of these were found to be X-linked. Interestingly, the proximal end of mouse Chromosome 4 was found to harbour several mutations, 6 that resulted in circling characteristics (the mouse lines were named tornado, dizzy, cyclone, eddy, orbitor and ferris) and 1 that resulted in dark footpads. The German project took a similar approach with normal C3HeB/FeJ mice, and looked for both dominant and recessive traits, with 14000 mice screened for a phenotype, yielding 182 unique lines (Hrabe de Angelis et al., 2000). The majority examined had skeletal defects in the limbs and tail, or abnormalities in coat colour, behaviour or skin.
Chapter 1- Introduction 41 The same mouse model discussed above in 1.4.2.1 that was used to examine thymic lymphoma when exposed to X-rays, was also administered ENU in the drinking water and comparisons were made between the two methods. Interestingly, ENU mutagenesis resulted in less thymic lymphoma incidence than the X-rays at the highest dose of each used (<36% compared to <28%), however, when both X-Rays and ENU were combined in the same mouse, a synergistic effect was observed which resulted in not only increased incidence of lymphoma to around 80% in all dosages of both agents tested, but also a significant reduction in latency (Hirano et al., 2013). The combination of both of these methods resulted in almost half the lymphoma mice studied having point mutations in the tumour suppressor gene Ikaros and insertions or deletions in the Notch1 gene, involved in signalling. Both of these genes are dysregulated in cancers. Also implicated were mutations in the tumour suppressor p53 and the proto-oncogene Kras, albeit at a much lower frequency (Hirano et al., 2013).
Creation of large scale mutagenesis projects such as these have yielded important new information about the roles of genes and pathways, however, more targeted approaches are also needed to complement these large scale screens focusing on the disease of interest. Modifier screens looking for suppressors or enhancers use a model that already has a phenotype of interest and looks to modify that phenotype (Kile and Hilton, 2005). Very few papers have published this approach and most of the ENU mouse mutants have come from the large scale mutagenesis projects. To date, none have focused on cancer, however, other disease models have been studied including diabetes (Goldsworthy et al., 2008; Tchekneva et al., 2007), Alzheimer’s (Liu et al., 2011a), bone abnormalities (Edderkaoui et al., 2013), thrombocytopaenia (Carpinelli et al., 2004; Kauppi et al., 2008) and diseases that arise from defects in the neural crest (Buac et al., 2008; Stottmann et al., 2011). An ENU treated myeloproliferative leukaemia virus mouse model of thrombocytopaenia (Mpl-/-), which have low platelet and megakaryocyte counts, yielded several genes involved in this disease. The inheritance of a semi-dominant suppressor disrupted the interaction between c-Myb, a gene involved in haematopoiesis, and p300 which in turn restored platelet production (Carpinelli et al., 2004; Kauppi et al., 2008).
The neural crest cell lineage gives rise to many cell types ranging from the nervous system, smooth muscle and melanocytes. Defects in the neural crest cell lineage results in a range of syndromes that are called neurocristopathies. An ENU mutagenesis screen
Chapter 1- Introduction 42 was undertaken by Buac and colleagues to elucidate genes involved in neural crest development by crossing ENU treated BALB/c mice with heterozygous knock-out Sox10 mice, a gene known to be expressed early in neuronal development (Buac et al., 2008). The 4 lines that arose from this screen resulted in a range of phenotypes from disruption of the neural tube and abnormal expression of Sox10. All of the mutations were recessive and two of the mutations were lethal before weaning while the second 2 had reduced viability, with less than 10% surviving to weaning age. For one of the mutant lines, the receptor tyrosine kinase Erbb3 was mutated and although protein expression was not compromised, downstream signalling was affected.
Taken together, these studies show that ENU mutagenesis has become a powerful tool for elucidating new genes and pathways involved in normal and disease development.
1.4.2.3 Mutagenesis using insertional transposable elements
Whilst mutagenesis using X-Ray or chemicals has proven successful in a number of studies, the putative cancer suppressor or oncogene responsible for the phenotype can be difficult to ascertain and requires backcrossing of mice and sequencing of a number of progeny to elucidate the gene of interest. Insertional transposable elements provide a new methodology for forwards genetics screens. The first of this type used retrotransposons from viruses in the 1990s to successfully integrate in a ‘copy-and- paste’ manner into transgenic mice (Woodard and Wilson, 2015). The advantage of using retrotransposons from retroviruses is that integration is permanent. However, only a small number of insertions appeared to occur in genes uploaded into the National Center for Biotechnology Information Reference Sequence database (NCBI RefSeq) making analysis difficult at the time (Collier and Largaespada, 2007). Retrotransposons were quickly replaced by more sophisticated non-viral transposable elements like Sleeping Beauty, piggyBAC and Frog Prince. Each method has its differences that both confer advantages and disadvantages over each other.
Non-viral transposons were originally isolated from species of insects and fish in the 1980s, and are elements that can perform horizontal gene transfer by performing a ‘cut- and-paste’ methodology that removes a section of DNA and relocates it to another position in the genome (Ivics et al., 1997; Yusa, 2015). However, they did not become
Chapter 1- Introduction 43 more widespread in their use until modifications were made that allowed them to cross species barriers (Ivics et al., 1997). The two biggest advantages of non-viral transposons over viral transposons are the increased cargo size and the ability to tag the transposon with a marker to elucidate the gene responsible for the change in phenotype (Woodard and Wilson, 2015). The Sleeping Beauty transposon was isolated from the Salmonid subfamily, and was named as such due to the long evolutionary sleep that the transposon had been awakened from given its ancient nature (Ivics et al., 1997). After mutating the transposon by restoring open reading frames, and adding the lacZ gene, Ivics and colleagues demonstrated transfer into mouse cells in culture. Following further modification and co-transfection with donor plasmids, the transposon demonstrated integration into carp, mouse and human cells (Ivics et al., 1997).
Although Sleeping Beauty was the first non-viral transposon developed, others soon followed. piggyBac, from insect cells and Frog Prince from the Leopard frog, provide alternatives to Sleeping Beauty, and while they all work in similar fashions, each has unique differences (Miskey et al., 2003; Woodard and Wilson, 2015). All three use a ‘cut-and-paste’ integration mechanism. While Sleeping Beauty and Frog Prince integrate at TA sites, piggyBac recognises TTAA sites (Ivics and Izsvak, 2010; Woodard and Wilson, 2015). piggyBac transposons may result in more reliable transgene expression due to its preference for transcriptional start sites, while Sleeping Beauty’s integration is more random (Woodard and Wilson, 2015). Sleeping Beauty has higher integration in human cells over Frog Prince, while Frog Prince integrates more efficiently into zebrafish genomes (Miskey et al., 2003). piggyBAC transposons have a larger cargo capacity than Sleeping Beauty, being able to carry approximately 9.1kilo bases (kB) of sequence before efficiency is compromised (Ding et al., 2005). Transposon screens can result in both loss- and gain-of-function mutations, exemplified in a medulloblastoma screen where insertions in tumour suppressor genes were hypothesised to be loss-of-function mutations and insertions in candidate oncogenes were seen as gain-of-function mutations (Wu et al., 2012). Conditional mouse models can also be used for transposon screens, which target the mutations only to the tissue of interest, overcoming the issue with embryonic lethality, and have been used in both melanoma and pancreatic cancer to discover new drivers of these diseases (Ni et al., 2013; Rad et al., 2015). Sleeping Beauty can be used together with Frog Prince in the same system, as can Sleeping Beauty and piggyBac (Miskey et al., 2003; Woodard and Wilson, 2015). Selection of the transposon to use depends upon the model of interest
Chapter 1- Introduction 44 and species used, but all can be used to discover new genes and pathways involved in disease.
Perhaps the most wide spread use of transposon screens in disease has been in the field of cancer. Studies using transposons in cancer first began to appear in 2005 and have increased since across a number of cancer including leukaemia, melanoma, glioma, medulloblastoma, colorectal carcinoma and osteosarcoma, among others (Moriarity and Largaespada, 2015). Several screens have resulted in the identification of genes that already have been described in the disease of interest, however, they also revealed novel genes (Moriarity et al., 2015; van der Weyden et al., 2015). For example, in a screen of B-cell leukaemia, Sleeping Beauty mutations identified 6 common insertion sites, 4 of which were already known to play a role in B-cell maturation, however 2 were novel genes (van der Weyden et al., 2015). Using mice that were pre-disposed to osteosarcoma and harbouring either wild-type or mutant p53, a Sleeping Beauty transposon screen yielded 20 known and numerous novel genes involved in both tumorigenesis and metastasis of osteosarcoma (Moriarity et al., 2015). By ‘discovering’ genes that are already known to be involved in the disease, confidence in the robustness of the screen is increased.
1.4.2.4 Quantitative trait loci mapping
Quantitative trait loci (QTL) mapping studies are used to determine which alleles are responsible for what phenotype, however this method of forward genetics has proven to be challenging as multiple alleles are responsible for a phenotype (Gonzales and Palmer, 2014). QTL mapping studies identify regions that span 30-40 mega bases (Mb), making identification of the actual gene responsible for the phenotype of interest difficult (Solberg Woods, 2014). Expression QTL (eQTL) studies have been undertaken using the Th-MYCN mouse model, as mentioned above. Briefly, Hackett and colleagues demonstrated that higher Arg1 expression could result in increased tumour penetrance and that a second susceptibility loci was found on chromosome 4 (Hackett et al., 2014). The same group also examined splicing QTL (sQTL), by examining differentially spliced genes in the cerebellum and SCG tissue and identified more than 2500 candidates (Chen et al., 2015). Known direct targets of the MYC oncogene were identified, however, the expression of the MYCN transgene remained unchanged
Chapter 1- Introduction 45 between the different strains tested indicating that sQTL may not be the primary driver in this model (Chen et al., 2015).
1.4.2.5 Emerging technologies using CRISPR/Cas9
The generation of tools for reverse genetics, such as RNAi, shRNA and CRISPR)/ Cas9, can also be used in a forwards genetics screen. While RNAi technology can be used in this way, the issues that are seen in reverse genetics approaches, such as low knock-down efficiency and off-target effects, also cause problems in these types of screens (reviewed in (Shalem et al., 2015). The CRISPR/Cas9 system however, provides a more robust tool. To generate hepatocellular carcinoma and intrahepatic choloangiocarcinoma in mice and elucidate the genes and pathways responsible, Weber et al combined a CRISPR system with Sleeping Beauty that contained both single guide RNA and Cas9 sequences. They targeted 10 known tumour suppressor genes and demonstrated that a CRISPR/Cas9 system could be used to cause cancer in a mouse (Weber et al., 2015). As technology progresses, the use of this technology will increase and more of these types of screens will emerge.
1.5 Summary and Thesis perspectives
Neuroblastoma is a disease where outcomes remain dismal for high risk disease despite intensive chemotherapy for patients and extensive research over the past several decades. The cause of neuroblastoma still remains a mystery for the 98% of patients that don’t have a genetic predisposition to the disease. The failure clinically is likely due to multiple aspects including tumour heterogeneity, the lack of targetable mutations and multidrug resistance of the tumours to chemotherapy. For those that do survive the disease, there are long-term health complications. By increasing our understanding of the mechanisms of initiation of this disease, and factors affecting progression of the tumour, novel strategies can be designed to overcome these hurdles and potentially lead to increased patient outcomes.
The Th-MYCN mouse model of neuroblastoma has allowed insight into the biology of the disease. While it does have several negatives, including the lack metastasis to bone
Chapter 1- Introduction 46 marrow, and the tumours arising from the sympathetic ganglia rather than the adrenal medulla, the negatives are far outweighed by the positive aspects of this model. The tumours represent aggressive, undifferentiated or poorly differentiated Stage 3 or 4 neuroblastoma, with amplification of the MYCN oncogene, all aspects of high-risk disease in children. The most powerful aspect of this model is the 100% tumour incidence by 7 weeks of age in mice homozygous for the MYCN oncogene, meaning that experiments can be undertaken in a short time-frame with confidence that the outcomes are robust.
Each of the above forward genetics methods discussed has advantages and disadvantages. ENU mutagenesis screens will likely continue to be the method of choice for forward genetics screens, at least in the short term. While screening for insertional mutations using tagged transposons is easier and more cost effective as a smaller number of mice are needed, ENU mutagenesis studies result in a greater mutation rate and more chance that a gene involved in the model is responsible for the phenotype. Combinations of two of the methods listed above may provide more insight into disease models.
Forward and reverse genetics screens are designed to be used alone, however, it is their use in combination that provides a robust screening methodology. By not biasing oneself to a gene or pathway and following a phenotype, finding a gene that has a real role in your disease of interest may have greater implications. The gene elucidated, must be investigated using some or all of the reverse genetics methods listed to confirm the phenotype is real, and thus independent of the screening methodology.
The initial aim of this work was to use ENU mutagenesis in the Th-MYCN mouse model of neuroblastoma to elucidate genes and pathways that are important in the initiation and progression of this disease. By taking homozygote MYCN mice and treating them with ENU, any change in the tumour phenotype of their offspring, either a delay or complete abrogation of the tumour, could easily be recognised. Once tumour phenotypes were identified, heritability of the phenotypes was tested to ensure robust inheritance.
Once mouse lines were established, the thesis was then directed towards elucidating the genes responsible for the phenotypes. This was achieved through a series of
Chapter 1- Introduction 47 backcrossing to well-described strains and a combination of exome and whole genome sequencing. Two candidate genes for the lines were hypothesized to be responsible for the phenotypes observed.
The final aim of the thesis was to confirm the gene responsible for the tumour delay for one of the mouse lines. This involved a series of experiments including knock-down of the gene and its interaction with MYCN in vitro and genetic knock-out and characterisation of tumour initiation in the mutated line in vivo.
Collectively, this thesis aimed at identifying new genes involved in neuroblastoma initiation and new targets for neuroblastoma therapy. It is hoped that these studies will eventually lead to better understanding of the disease and ultimately a new avenue for therapy for high-risk neuroblastoma.
Chapter 1- Introduction 48 CHAPTER 2 MATERIALS AND METHODS
2.1 Materials
2.1.1 Chemicals and Reagents
2.1.1.1 Tissue Culture
Gibco™ Dulbecco’s modified Eagle medium (DMEM), Roswell Park Memorial Institute-1640 (RPMI-1640) medium, Dulbecco’s phosphate buffered saline (PBS), trypsin solution (0.25% (w/v) trypsin in Hank’s solution), Trypan Blue (0.4%) and foetal calf serum (FCS) were purchased from Thermo Fisher Scientific (Australia). Tissue culture flasks were from Corning® and 6-well plates from Greiner CELLSTAR® (Sigma-Aldrich, Australia).
2.1.1.2 Cytotoxic drugs and chemicals
Cyclophosphamide (Endoxan) was purchased from Baxter (Sydney, Australia). N-ethyl- N-nitrosourea, sodium thiosulfate, Crystal Violet, sodium citrate, sodium chloride, ethylenedinitrilotetraacetic acid (EDTA), Tween-20 and doxycycline were purchased from Sigma-Aldrich (Australia). UNIVAR® sodium hydroxide, UNILAB® absolute ethanol and reagents for Carnoy fixative (60% ethanol, 30% chloroform, 10% glacial acetic acid) were purchased from Ajax Fine Chemicals (Thermo Fisher Scientific, Australia). Polyethylenimine (PEI) was purchased from Polysciences (Hirschberg an der Bergstraße, Germany). Neutral buffered formalin (10%), methanol and formamide were purchased from Fronine (Thermo Fisher Scientific). 2-(4-amidinophenyl)-1H - indole-6-carboxamidine (DAPI II) was purchased from Vysis (Abbott Laboratories; Abbott Park, Illinois).
2.1.1.3 siRNA transfection
Human SMARTpool: ON-TARGETplus RUNX1T1 siRNA (Cat# L-011824-00-0005) and scrambled control siRNA were purchased from DharmaconGE (Millennium
Chapter 2- Materials and Methods 49 Science Pty Ltd, Victoria, Australia). Ambion™ siRNA for RUNX1T1 (AM16708), Lipofectamine® 2000 and RNAiMAX transfection reagents were purchased from Invitrogen™ (Thermo Fisher Scientific, Australia). MYCN siRNA (MYCN_6) was purchased from QIAGEN (Victoria, Australia).
2.1.1.4 DNA and RNA isolation,
RNA and DNA isolation reagents were purchased as follows: TRIzol™ from Invitrogen and buffer-saturated phenol from Life Technologies (Thermo Fisher Scientific, Australia); proteinase K from Sigma-Aldrich; Univar® Tris HCl, sodium chloride (NaCl) and EDTA Disodium salt from Ajax Fine Chemicals (Thermo Fisher Scientific, Australia).
2.1.1.5 Complementary DNA (cDNA) synthesis, polymerase chain reaction (PCR) and sequencing cDNA synthesis, PCR and sequencing reagents were purchased as follows: Moloney murine leukaemia virus (MMLV) reverse transcriptase, 5x first strand buffer and 0.1M dithiothreitol (DTT) from Life Technologies (Thermo Fisher Scientific, Australia);
AmpliTaq® Gold DNA polymerase, AmpliTaq® Gold 10 x PCR buffer, MgCl2 solution, TaqMan® Universal and Genotyping master mixes, TaqMan® probes and gene expression assays from Applied Biosystems (Thermo Fisher Scientific, Australia); ExoSAP-IT PCR clean up kit from Affymetrix (Thermo Fisher Scientific, Australia); ribonuclease inhibitor RNasin from Promega Australia (Sydney, NSW); random hexanucleotide primers, deoxynucleoside triphosphates (dNTPs) and primers from Sigma-Aldrich (Australia).
2.1.1.6 Protein isolation and western blotting
Protein isolation and western blotting reagents were purchased as follows: Tris-HCl gradient polyacrylamide gels (4–15%), β-mercaptoethanol and Laemmli sample buffer from BioRad (Sydney, Australia and Milan, Italy); ethylene glycol-bis (β-
Chapter 2- Materials and Methods 50 aminoethylether)-N,N,N’,N’-tetraacetic acid (EGTA), Ponceau S sodium salt, protease inhibitor cocktail, phenylmethylsulfonyl fluoride (PMSF), DTT and Trizma base (Tris) from Sigma-Aldrich (Australia); skim milk powder from Coles Supermarket (Australia); Complete protease inhibitor cocktail (Roche), DTT, EDTA, NaCl and 2-[4- (2-hydroxyethyl)piperazin-1-yl]ethanesulfonic acid (HEPES) from Sigma (Italy). Protease and phosphatase inhibitors (sodium pyrophosphate, sodium orthophosphate, sodium fluoride, PMSF) and nonyl phenoxypolyethoxylethanol (NP-40) from Life Technologies Italia (Thermo Fisher Scientific); BENCHMARK® prestained protein ladder, range 10–200 kDa, from Invitrogen (Thermo Fisher Scientific); SuperSignal® Westdura and ECL chemiluminescence substrates, and bicinchoninic acid assay (BCA) protein assay kit were obtained from Pierce (Thermo Fisher Scientific, Australia); Nonidet P-40 from Fluka (Sigma-Aldrich, Australia); Whatman® 3MM cellulose chromatography paper from Sigma-Aldrich (Australia); Hybond C extra supported nitrocellulose membranes from GE Healthcare (Sydney, Australia); magnesium chloride, Tween®-20, Coomassie Brilliant Blue R-250 and sodium dodecyl sulphate (SDS) from MP Biomedicals (Sydney, Australia); glycerol from Ajax Chemicals (Sydney, Australia). Primary antibodies were sourced from various companies as outlined in Table 2-1.
2.1.1.7 Electrophoresis
Electrophoresis reagents were purchased as follows: DNA grade ultrapure agarose and low melt agarose from Applichem (Darmstadt, Germany); glacial acetic acid from Ajax Chemicals (Sydney, Australia); boric acid and bromophenol blue from Merck (Kilsyth, Australia); glycine from ICN Biochemicals Inc. (Aurora, OH); Ultrapure Acrylagel and Ultrapure Bis-acrylagel from National Diagnostics (Atlanta, GA); ammonium persulfate and N,N,N',N'-tetramethylethylenediamine (TEMED) from Sigma-Aldrich (Australia); and low molecular weight DNA markers (1kb DNA ladder) from Life Technologies (Thermo Fisher Scientific).
Chapter 2- Materials and Methods 51
Table 2-1 Primary antibodies used in Western blotting and immunohistochemistry
Catalogue Antibody Company Species Number mSIN3A Abcam ab3479 Rabbit FLAG Sigma F1804-200ug Mouse MAX (C-17) Santa Cruz sc-197 Rabbit MYCN (B8.4 B) Santa Cruz sc-53993 Mouse MYCN (C-19) Santa Cruz sc-791 Rabbit HDAC1 Abcam ab7028 Rabbit HDAC2 Upstate 05-814 Mouse NCOR1 Abcam ab24552 Rabbit RUNX1T1 Proteintech 15494-1-AP Rabbit ACTIN Sigma Aldrich A2066 Rabbit
Chapter 2- Materials and Methods 52
2.1.1.1 Cloning and transfection
Restriction enzymes and T4 polynucleotide kinase (PNK) were from New England Biolabs (Ipswich, Massachusetts); Herculase II Fusion DNA polymerase from Agilent (Mulgrave, Australia); Vectors p3xFLAG-CMV™-10 and 14 from Sigma (Italy); Mouse Runx1t1 open reading frame, Isoform 1 clones (NM_001111027.2) from Origene (Rockville, Maryland); Lentiviral Vector FH1UTG from Applied Biological Materials, Inc (Richmond, Canada); E. coli strain DH5 alpha Stbl3™ and antibiotics from Thermo Fisher Scientific; Transformation reagents are as follows: Luria Bertani (LB) broth and LB agar from Sigma-Aldrich (Australia); QIAfilter® Plasmid Maxi kit for DNA purification from bacterial cultures was purchased from QIAGEN (Clifton Hill, Australia).
2.1.2 Equipment
2.1.2.1 Tissue culture
A Neubauer Bright Line haemocytometer (Sigma-Aldrich, Australia) was used to count cells. An inverted microscope from Olympus Optical Company (Tokyo, Japan) was used to visualise cells. Cells were pelleted using a Megafuge™ 16R centrifuge (Thermo
Fisher Scientific). Cells were grown in a humidified CO2 incubator from either HERAcell (Heraeus; Thermo Fisher Scientific) or BINDER (BINDER GmbH, Tuttlingen, Germany).
2.1.2.2 Plasmid DNA and nucleic acid isolation
Tissue samples were homogenised using an Omni TH homogeniser (Omni International, Kennesaw, Georgia). For DNA isolations, a microcentrifuge (Model 5415D) purchased from Eppendorf (Hamburg, Germany) was used. The Sorvall refrigerated centrifuge (Model RC-5B), used with either a GSA or SS-34 rotor, was purchased from Dupont Instruments (Wilmington, Delaware).
Chapter 2- Materials and Methods 53
2.1.2.3 PCR and sequencing
Conventional thermal cyclers (GeneAmp® PCR System 9700) used for PCR amplification and sequencing were supplied by Applied Biosystems (Thermo Fisher Scientific, Australia). The ABI PRISM® 7900 Sequence Detection System used for real-time quantitative PCR amplification and allelic discrimination was purchased from Applied Biosystems. Sanger DNA sequencing was carried out by Australian Genome Research Facility Westmead node (Sydney, Australia). Analysis of sequencing data was performed with the Vector NTI® Suite Version 8 and Contig Express Sequence Analysis Software (Thermo Fisher Scientific, Australia). Exome sequencing was performed at the Australian Phenomics Network located at the Australian National University (Canberra, Australia) using either the NimbleGen SeqCap EZ mouse exon array (Roche, California, USA) or Sure Select XT Mouse All Exon array (Agilent Technologies, California, USA).
2.1.2.4 Spectrophotometer
For quantitation of DNA and RNA the NanoDrop 2000 from Thermo Scientific was used. For determination of the concentration of ENU the DU®530 UV/Vis Spectrophotometer from Beckman Instruments Inc. (Fullerton, California) was used.
2.1.2.5 Electrophoresis
For DNA gels, either the Hoefer Mighty Small™II vertical electrophoresis unit (Model SE250) for polyacrylamide gels, or the Mini-Sub® Cell GT System (Bio-Rad, Australia) for agarose gels were used for electrophoresis. For western blotting, a Mini- PROTEAN Tetra Cell, supplied by Bio-Rad was used. A PowerPac™ basic power supply (Bio-Rad) was used for all electrophoresis apparatuses.
Chapter 2- Materials and Methods 54 2.1.2.6 Visualisation of gels and blots
DNA gels were visualised with the Gel Doc™ XR+ System (Bio-Rad) whilst the protein blots were visualised with the ChemiDoc XRS+ Imaging System (Bio-Rad). Both systems utilise the Image Lab software.
2.1.3 Cell lines
The human neuroblastoma cell line BE(2)-C was clonally isolated from the SK-N- BE(2) cell line, which was derived from a bone marrow metastasis of a child with Stage 4 disease (Ciccarone et al., 1989). BE(2)-C cells are classified as I-type (intermediate), are immortal and develop tumours in nude mice (Ciccarone et al., 1989) and were generously supplied by Dr June Biedler (Memorial Sloan-Kettering Cancer Center, New York, NY).
SH-EP cells, with very little to no expression of MYCN, were synthetically engineered to express MYCN. MYCN expression is repressed upon addition of tetracycline to barely detectable levels, and these cells were named SH-EP Tet-21/N (Lutz et al., 1996). They were a kind gift from Dr Manfred Schwab (German Cancer Research Center, Heidelberg, Germany).
CHP-134 (ECACC 06122002), derived from the lymph node of a Stage 4 neuroblastoma patient who was treated with combination chemotherapy prior to biopsy (Schlesinger et al., 1976) and KELLY (ECACC 92110411), from a brain metastasis of a neuroblastoma patient with 100–120 fold amplification of MYCN (Schwab et al., 1983) were purchased from the European Collection of Authenticated Cell Lines (ECACC; Salisbury, United Kingdom) via CellBank Australia (Westmead, Australia).
NBL-W-N, the neuronal subtype of NBL-W, isolated from the adrenal of a Stage 4s patient (Foley et al., 1991) was a kind gift from Dr Susan Cohn (formally Northwestern University, Chicago, Illinois; currently The University of Chicago, Chicago, Illinois).
Chapter 2- Materials and Methods 55 The murine cell lines NHO1A and NHO1S were isolated from a single tumour of a homozygous Th-MYCN transgenic mouse, and represent adherent (A) and suspension (S) populations isolated from this tumour (Cheng et al., 2007).
2.1.4 Patient cohorts
A publicly available dataset, accessed through the Gene Expression Omnibus database (Accession: GSE45480), comprising 649 primary neuroblastoma tumours, was used to examine the expression of various genes, and has been described previously (Kocak et al., 2013). Briefly, the dataset covers all INSS stages (1–4 and 4s) and primary tumours located in the abdomen, neck/chest and adrenal glands. Outcome data for these patients is known. Gene expression profiles were generated on these samples using a 44K oligonucleotide array from Agilent Technologies (Palo Alto, USA) (Oberthuer et al., 2010). To further confirm findings, two additional datasets were interrogated. A set of 101 metastatic tumours not expressing MYCN (Asgharzadeh et al., 2006) and an RNA- seq dataset from the Sequencing Quality Control consortium consisting of 498 primary neuroblastoma samples (Zhang et al., 2015b) were investigated.
Human neuroblastoma tumour microarray (TMA) sections were obtained from the Tumour Bank at The Children’s Hospital at Westmead (The Sydney Children’s Hospital Network). Six TMA slides were used, and their disease pathology and identification numbers are listed in Table 2-2. After filtering, 96 tumour cores could be used, comprising 79 neuroblastoma, 5 ganglioneuroblastoma and 12 ganglioneuroma samples. Clinical outcome data were available for these 96 patients.
2.1.1 Mice
The Th-MYCN mouse model of neuroblastoma created by Weiss and colleagues (Weiss et al., 1997), was imported into Australia and backcrossed for four generations to 129/SvJ (Animal Resources Centre, Perth, Australia). The line was housed at the Biological Resources Centre (pre-2010) and then at Children’s Cancer Institute (2010 onwards), both located at the University of New South Wales, Sydney, NSW, Australia.
Chapter 2- Materials and Methods 56
Table 2-2 Human tissue microarrays used in immunohistochemistry TMA Identification number Pathology TMA-12-001 Neuroblastoma and Ganglioneuroblastoma/Ganglioneuroma TMA-13-002 Neuroblastoma TMA-14-001 Neuroblastoma TMA-14-002 Neuroblastoma TMA-14-003 Neuroblastoma TMA-14-004 Neuroblastoma, Ganglioneuroblastoma and Ganglioneuroma
Chapter 2- Materials and Methods 57 The mice have been maintained by breeding hemizygous mice together to generate mice with two, one or no copies of the MYCN transgene. The line characteristics are described in Chapter 1. Approximately 2 weeks after birth, the mice were tail tipped for genotyping and issued to the project at weaning.
The Runx1t1 heterozygous knock-out mouse model developed by Franco Calabi (Calabi et al., 2001) was thawed from sperm at the Medical Research Council Harwell animal facility (Oxfordshire, United Kingdom). Animals were imported into Australia (Australian BioResources, Moss Vale, Australia), passed quarantine and were shipped to the Children’s Cancer Institute Animal Facility. Mice were maintained by breeding heterozygous mice together, to generate mice with two, one or no functional copies of Runx1t1. Mice were tailed-tipped early after birth due to the lethality of Runx1t1 knock- out, which occurs 1–2 days after birth, to ensure the genotypes were captured before cannibalism could occur.
C57BL/6 and BALB/c mice for backcrossing were purchased from Australian BioResources (Moss Vale, Australia) at 5-6 weeks of age and allowed to acclimatise for 1 week prior to mating.
All mice were housed in a specific pathogen-free, Physical Containment level 2 facility, with the temperature controlled between 22–24 C and a 12-hour day/night. Mice were kept in Ventirack cages (Tecniplast), provided food and water ad libitum, with environmental enrichment.
All experimental procedures were approved by the University of New South Wales Animal Care and Ethics Committee (ACEC) (Approval numbers ACEC 07/58B (ENU mutagenesis), ACEC 10/8B (Backcrossing), ACEC 12/97A and ACEC 14/122B (Characterisation)). The use of genetically modified mice was approved by the Institutional Biosafety Committee (Notifiable Low-Risk Dealings 12/35 and 14/35).
Chapter 2- Materials and Methods 58 2.2 Methods
2.2.1 Cell biology
2.2.1.1 Maintenance of cell lines
Human neuroblastoma cell line BE(2)-C was maintained as a monolayer culture in
DMEM containing 10% FCS at 37C in a humidified atmosphere of 5% CO2. SH-EP Tet-21/N, CHP-134, NBL-W-N and KELLY cell lines were maintained in RPMI-1640 plus 10% FCS. NHO1A (adherent) and NHO1S (suspension) were maintained in RPMI-1640 plus 20% FCS. All cultures were routinely screened for Mycoplasma spp. and found to be free of contamination, and have been screened for cell line identity where possible.
2.2.1.2 Overexpression and siRNA knock-down
Cells were trypsinised, washed and counted using trypan blue exclusion. Cells were plated in 6-well plates at 1 x 105 cells per well in 2mL of media. For siRNA knock- down studies, cells were transfected using SMARTPOOL or Ambion™ siRNA for 4-6 hours at concentrations stated, using either Lipofectamine® 2000 or RNAiMAX. For overexpression, 3µg of purified p3xFLAG-CMV™-10 or -14 plasmid, containing either wild-type or mutant Runx1t1 (described in section 2.2.2.9) were transfected using Lipofectamine 2000 for 6 hours. After this time, the cells were given fresh media containing FCS and left for 72 hours, unless otherwise noted, when they were harvested for RNA extraction.
2.2.1.3 Colony formation assays
BE(2)-C cells, either wild-type or containing the overexpression or shRNA knock-down constructs, were plated at 500 cells per well in a 6-well plate. The media was replaced after 6 hours. For shRNA knock-down, doxycycline (final concentration 1µg/mL) was
Chapter 2- Materials and Methods 59 added to the well and replaced every 72 hours. Colonies were stained after 10 days using 0.5% crystal violet in 50% methanol, and scanned using the Gel Doc XR+ System (Bio-Rad). Colony numbers were counted using ImageJ software.
2.2.1.4 Fractionation and western blotting
Fractionation experiments were carried out in the laboratory of Professor Giovanni Perini at the University of Bologna, Italy, with the assistance of Dr Daniela Erriquez.
Neuroblastoma BE(2)-C cells (4 x 106) were plated into sterile 10cm dishes. Cells were transfected with wild-type or mutant (YH) Runx1t1 (8µg) (described in section 2.2.2.9), using PEI transfection reagent; 1µg/µL, 24µg per dish, for 3 hours in serum-free DMEM. Cells were washed and incubated in DMEM with 10% FCS for 48 hours. Cells were harvested with trypsin and nuclear protein was extracted for protein fractionation. Briefly, cells were pelleted (200g, 5 mins, room temperature, using a swing-out rotor) and washed in 3mL of ice-cold PBS. The pellet was resuspended in 500µL of hypotonic buffer (Table 2-3) and centrifuged (200g, 5 mins, room temperature, using a swing-out rotor). The pellet was incubated on ice for 15 minutes in 500µL of hypotonic buffer and then an additional 500µL of hypotonic buffer containing 0.2% NP-40 was added and incubated at 4C for 15 mins on a rotating wheel. The tubes were centrifuged (3600g, 10 minutes, 4C, using a swing-out rotor). The nuclear pellet was resuspended in 500µL of Co-immunoprecipitation (Co-IP) buffer (Table 2-4) and sonicated for 15 minutes at medium intensity at 4C in a 15mL tube. The samples were transferred to a 1.5mL Eppendorf and centrifuged for 15 minutes at 4C and maximum speed to remove cellular debris and DNA. The supernatant, containing the nuclear protein extract, was collected and quantitated using a BCA assay, according to manufacturer’s instructions.
The nuclear extracts were loaded onto a Superdex 200 10/300 GL column (GE Healthcare Life Sciences), and fractionated into 40x1mL fractions using an AKTAmicro chromatography system (GE Healthcare Life Sciences). Samples (39µL) were mixed with 13µL of loading buffer (Laemmi buffer containing β-mercaptoethanol) and boiled for 10 minutes. Samples for western blotting were loaded onto pre-cast 4– 15% Tris-HCl gradient gels (Bio-Rad).
Chapter 2- Materials and Methods 60
Table 2-3 Hypotonic buffer recipe for nuclear protein extraction Reagent (stock) Final Concentration
1M HEPES pH 7.6–8 10mM (100X)
5M NaCl 50 mM (100X)
0.5M EDTA 1 mM (500X)
1M DTT 1mM (1000X) 0.1M Sodium Pyrophosphate 1mM (100X) 0.1M Sodium Orthophosphate 1mM (100X) 0.5mM Sodium Fluoride 1mM (500X) 0.1M PMSF 1 mM (100X)
Complete Protease inhibitor cocktail (50X) Milli Q Water Up to final volume
Chapter 2- Materials and Methods 61
Table 2-4 Co-IP buffer recipe for nuclear protein extraction Reagent (stock) Final Concentration
1M Tris HCl pH 7.6–8 50mM (20X)
5M NaCl 50 mM (100X)
0.5M EDTA 1 mM (500X)
1M DTT 1mM (1000X) 0.1M Sodium Pyrophosphate 1mM (100X) 0.1M Sodium Orthophosphate 1mM (100X) 0.5mM Sodium Fluoride 1mM (500X) 0.1M PMSF 1 mM (100X)
Complete Protease inhibitor cocktail (50X) Milli Q Water Up to final volume
Chapter 2- Materials and Methods 62 As a loading control could not be used in this experiment, samples were also loaded onto 12.5% polyacrylamide gels for staining with Coomassie Brilliant Blue. Gels were electrophoresed in a Mini-PROTEAN® Tetra electrophoresis tank at 80V until the dye- front was beyond the stacking gel, and then at 130V until the dye-front reached the bottom of the gel.
For Western blotting, the samples were transferred onto nitrocellulose membrane (Hybond®-C extra) using the Mini-PROTEAN® Tetra electrophoresis tank containing transfer buffer (0.025M Tris/Glycine, pH 8.2; 20% methanol) for 1.5 hours at 250mA at 4C. To confirm efficient transfer, the nitrocellulose membrane was stained with Ponceau S (5 min). The blots were washed 3 times in PBS and then blocked for 1 hour in 4% skim milk powder in Tris buffered saline (TBS; 20mM Tris, pH 8; 150mM NaCl). Membranes were incubated with primary antibody overnight at 4C as indicated in Table 2-5. Membranes were washed in TBS containing 0.05% Tween-20 (TBST) three times for 10 minutes, then incubated with secondary anti-mouse or anti-rabbit horseradish peroxidase-conjugated IgG antibody (1/2000 in 4% skim milk in TBST) for 1 hour at room temperature. Pierce™ Enhanced Chemiluminescence (ECL) Western Blotting Substrate (Thermo Fisher Scientific) was prepared by mixing Detection Reagents 1 and 2 in a 1:1 ratio, and visualising on the ChemiDoc XRS+ Imaging System (Bio-Rad) with Image Lab software.
For Coomassie Brilliant Blue staining, gels were fixed in a solution containing 50% methanol and 10% glacial acetic acid for 1 hour. Proteins were stained overnight in staining solution (0.1% Coomassie Brilliant Blue R-250, 50% methanol and 10% glacial acetic acid) and then de-stained with 40% methanol and 10% glacial acetic acid, replacing the solution until the background was fully de-stained. Protein bands were visualised on the ChemiDoc XRS+ Imaging System (Bio-Rad) with Image Lab software.
Chapter 2- Materials and Methods 63
Table 2-5 Antibodies used for western blotting
Antibody Dilution Diluent mSIN3A 1/2000 4% skim milk in TBS FLAG 1/1000 4% skim milk in TBS MAX (C-17) 1/500 4% skim milk in TBS MYCN (B8.4 B) 1/1000 4% skim milk in TBS MYCN (C-19) 1/1000 4% skim milk in TBS HDAC1 1/2000 4% skim milk in TBS HDAC2 1/1000 4% skim milk in TBS N-CoR1 1/1000 4% skim milk in TBS 4% skim milk in RUNX1T1 1/1000 TBST ACTIN 1/5000 TBST
Chapter 2- Materials and Methods 64 2.2.1.5 Histology
Samples for histology were collected in 10% neutral buffered formalin and fixed for 24 hours. Samples were then transferred to 70% ethanol. Specimens containing bone, including day 0 pups, were decalcified in 0.38M EDTA for 48 hours. Teeth and nails are resistant to decalcification and were removed from the samples prior to placing in histology cassettes. For day 0 pups, the milk stomach was removed from the specimens as it is known to contaminate samples during embedding. Samples were embedded in paraffin, sectioned and stained with haematoxylin and eosin at either the Histology and Microscopy Unit at the University of New South Wales (Sydney, Australia) or the Garvan Institute of Medical Research Histopathology facility at The Kinghorn Cancer Centre (Sydney, Australia).
Histological assessment was carried out by pathologist Dr Andrew Gifford. Scoring of neuroblast hyperplasia within autonomic ganglia of mice was undertaken in a blinded study. The scorer did not know the genotype of the mice when scoring was performed, and was unblinded only at completion of the scoring. Hyperplasia was defined as >30 neuroblasts within a ganglion, similar to previous studies (Hansford et al., 2004). The percentage of hyperplasia within ganglia scored was then determined.
2.2.1.6 Immunohistochemistry
Immunohistochemistry of mouse tissue or human TMA was carried out by the Garvan Institute of Medical Research Histopathology facility at The Kinghorn Cancer Centre (Sydney, Australia). Rabbit polyclonal RUNX1T1 antibody, which cross-reacts with both the mouse and human protein, was used at a 1:400 dilution. The positive controls were mouse brain and the human cell line IMR-32, both of which express the protein of interest, and the negative control was SH-EP, which does not express RUNX1T1. Stained TMA slides were scored, in conjunction with a haematoxylin and eosin stained slide, by Dr Andrew Gifford. Cores were omitted from the analysis if one of the following occurred: there was insufficient clinical information available; there was insufficient tumour tissue on the slide; the core was absent; or the core was folded or crushed, with staining that was difficult to interpret.
Chapter 2- Materials and Methods 65 Firstly, the integrity of the core was assessed and scored as intact (0), minor disruption where >50% of the core was present (1), major disruption where <50% of the core was present (2) or absent (3). Each core was assessed for staining using a protein expression score, based on the percentage of the core with positive staining and the intensity of the staining (as outlined in Table 2-6) and applying the following formula:
Score = Percentage of core with positive staining (0–3) x Intensity of staining (0–3)
The range of possible scores was, therefore, 0–9. Determination of the protein expression score was performed by Dr Andrew Gifford, as well as calculations on the positive and negative predictive value that patients with high RUNX1T1 would also be MYCN amplified. Statistics were calculated and Kaplan–Meier curves constructed by Dr Laura Gamble. To determine if there was an association between overall survival and RUNX1T1 score for the patients in the TMA, Kaplan–Meier curves and Log Rank tests were performed using IBM SPSS Statistics 24. The RUNX1T1 score was classified as either low or high at various cut points between the minimum score of 0 and the maximum score of 9 (for example, score <= 3 (low) vs score >3–9 (high)). RUNX1T1 score was also correlated with MYCN status using GraphPad Prism 7 software, and significance determined using a Mann–Whitney test. Fisher’s Exact tests were performed using SPSS Statistics 24.
2.2.1.7 Computational modelling
The solution structure of the MYND domain of human RUNX1T1 complexed with the co-repressor SMRT (RCSB Protein Data Bank ID 2ODD; PMID 17560331; (Liu et al., 2007b)) was visualized using the PyMOL Molecular Graphics System (Schrödinger, LLC). The Tyr682His substitution was made using the PyMol Mutagenesis Wizard, and the rotamer of best fit selected.
Chapter 2- Materials and Methods 66 Table 2-6 Scoring system for TMA Percentage of core with positive staining Intensity of staining 0 0% 0 Negative 1 1–10% 1 Weak 2 11–50% 1.5 Weak-Moderate 3 51–100% 2 Moderate 2.5 Moderate-Strong 3 Strong
Chapter 2- Materials and Methods 67 2.2.2 Molecular biology
2.2.2.1 Isolation of genomic DNA from mouse livers
DNA was extracted from frozen livers of mice using Proteinase K. Samples were incubated overnight at 55C in a lysing solution (20mM Tris-HCl, pH 8, 0.5mM EDTA, 1M NaCl, 0.5% SDS; 300 μL) with Proteinase K (300μg/ml; 50 μL). After complete lysis had occurred, an equal volume of phenol:chloroform:isoamyl alcohol (25:24:1, v:v:v) was added, mixed by inversion, and then centrifuged at 12,000 g for 15 min at room temperature). The aqueous (upper) phase was transferred to a clean tube, without disturbing the interface, and a 0.1 x volume of 3M sodium acetate pH 5.8 and twice the volume of chilled ethanol were added and mixed. Samples were centrifuged at 12,000 g at 4C for 30 mins. The DNA was washed twice in ice-cold 70% ethanol and air-dried for 10–15 minutes. The DNA was resuspended in DNase- and RNase-free sterile water, and the concentration determined using the NanoDrop 2000 at an absorbance of 260 and 280 nm.
2.2.2.2 Genotyping of mice
DNA was extracted from a 2–4 mm mouse tail tip using a simplified Chelex extraction protocol developed by Burkhart and colleagues (Burkhart et al., 2002). Briefly, 500µL of 5% Chelex® 100 resin in sterile water, was pipetted into a sterile 1.5mL Eppendorf tube containing the tail tip. The tubes were vortexed and incubated at 56C in a waterbath overnight. Samples were vortexed and then incubated at 95C for 15 mins in a heat block, followed by centrifugation at 12,000g for 15 mins. The supernatant was transferred to clean Eppendorf tubes and stored at -20C until use.
Real-time PCR sequence-specific assays were designed using the Primer Express™ software, Version 3.0.1 from Applied Biosystems (Thermo Fisher Scientific, Australia), based on the conventional PCR sequences (Calabi et al., 2001; Haraguchi and Nakagawara, 2009; Weiss et al., 1997). Primers were synthesised by Sigma-Aldrich (Australia), and TaqMan® probes (TAMRA™ or minor groove binder-non-fluorescent
Chapter 2- Materials and Methods 68 quencher (MGB-NFQ)) labelled with either 6FAM™ or VIC™ were synthesised by Applied Biosystems (Thermo Fisher Scientific, Australia). Primers and probes for each genotyping assay are listed in Table 2-7.
The genotypes of mice were determined using an allelic discrimination methodology. For MYCN mice, specific assays were designed to the MYCN transgene or to the insertion site of the transgene on chromosome (Chr) 18 which is disrupted during insertion (Haraguchi and Nakagawara, 2009; Weiss et al., 1997). For Runx1t1 knock- out mice, assays were designed around the targeted disruption site of Runx1t1, or to the lacZ bacterial gene used to disrupt and knock–out the gene (Calabi et al., 2001). For genotyping of ENU mutagenesis mice (1590), probes specific to the point mutation (either T or C) were designed for competitive binding.
Genomic DNA (2µL) isolated using the Chelex extraction method was used to genotype the mice in a 25µL reaction volume using the Applied Biosystems ABI Prism 7900 Sequence Detection System (Thermo Fisher Scientific, Australia). TaqMan® master mix (Universal master mix for MYCN and Runx1t1 knock-out, and Genotyping master mix in the case of 1590) comprised half the reaction, along with the specific primers and probes in an optical reaction PCR plate according to the manufacturer’s recommendations. Prior to thermocycling, a pre-read was performed to detect background fluorescence. The thermocycling comprised an initial hold at 95°C for 10 min, followed by 40 cycles of 15 sec denaturation at 92°C and annealing for 1 min at 60°C. Endpoint fluorescence was determined for each sample after PCR amplification to determine the relative levels of either 6FAM™ or VIC™ in the well. Analysis of the genotypes was performed using Sequence Detection System 2.4 (Applied Biosystems) using controls on each plate.
2.2.2.3 Exome and whole genome sequencing
Samples for exome sequencing were sent to the Australian Phenomics Network at the Australian National University (Canberra, Australia). Samples were prepared using either Illumina TruSeq library preparation for use on the NimbleGen SeqCap EZ mouse
Chapter 2- Materials and Methods 69
Table 2-7 Primers and probes used in real-time PCR analysis for mouse genotyping Primer/ Gene Sequence Probe Th-MYCN MYCN F 5’-CGACCACAAGGCCCTCAGTA MYCN MYCN R 5’-CAGCCTTGGTGTTGGAGGAG Probe 6FAM-CGCTTCTCCACAGTGACCACGTCG TAMRA Chr 18 F 5’- CCACAAAAATATGACTTCCTAAAAGATTT Chr 18* Chr 18 R 5’- CATGGGACTTCCTCCTTATATGCT Probe VIC-5’-AACAATTATAACACCATTAGATATG TAMRA Runx1t1 Knock-out genotyping Runx1t1 F 5’-TGAAGACGCAGTCTAGGCTGACT Runx1t1 Runx1t1 R 5’-TACTTACATGTCGTTGGCGTAAATG Probe 6FAM 5’-CAATGCCACCTCCTC MGB-NFQ† lacZ F 5’-CAGAAACAGCACCTCGAACTGA lacZ lacZ R 5’-GCCATACAGCGCGTTGAAA Probe VIC-5’-CCGCGATATTGCC MGB-NFQ Runx1t1 1590 SNV 1590 5’-GACGTGCAGCGGCTGTAA Forward 1590 5’-CCCAGTCTTTATGCTGGCAAA Reverse WT VIC-5’-ACGGCCCGATACT MGB-NFQ Probe Variant 6FAM-5’-ACGGCCCGACACT MGB-NFQ Probe * Chr indicates Chromosome † MGB-NFQ refers to minor groove binder–non fluorescent quencher
Chapter 2- Materials and Methods 70 exon array (Roche), or the preparation kit from Agilent for running on the SureSelect XT Mouse All Exon array (Agilent), according to the manufacturer’s instructions. The pipeline for both involves extraction of DNA, exome enrichment using the kit provided and sequencing on the array. The reads are aligned and filtered by Australian Phenomics Network, removing false positives or negatives where possible and filtering out variants that are not ENU-induced with their software. Lists of variants are provided to the end- user for confirmatory sequencing.
Venn diagrams of exome sequencing results were constructed using InteractiVenn (http://www.interactivenn.net/) developed by Heberle and colleagues (Heberle et al., 2015).
Tail tip and liver samples for whole genome sequencing were sent to the Walter and Eliza Hall Institute of Medical Research. DNA was prepared and sequenced by Dr Janelle Collinge. Analysis of the sequencing results was performed by Dr Collinge and Dr Benjamin Kile.
2.2.2.4 PCR and sequencing of mutant mouse lines
Sequence-specific assays were designed using the Primer Express™ software, Version 3.0.1 from Applied Biosystems (Thermo Fisher Scientific, Australia), based on the location of the single nucleotide variation identified by exome sequencing. Primers were synthesised by Sigma-Aldrich (Australia). To confirm the presence of the variants, PCR assays were designed as outlined in Table 2-8.
PCR reactions were performed using 100ng of DNA and 1 unit of AmpliTaq Gold® DNA polymerase in a reaction mix containing 10x PCR Buffer II (100mM Tris-HCl, pH 8.3, 500mM KCl), dNTPs (250µM), primers (20pmol) and 1.5mM magnesium chloride in a 25µL reaction volume. The reaction was subject to a 3-step PCR cycle for 35 cycles in a thermocycler. An initial pre-incubation step of 94C for 10 minutes was followed by 35 cycles of 94C for 45 seconds, primer annealing at 58C for 45 seconds and an extension of 90 seconds at 72C. A final extension of 72C for 10 minutes was followed by cooling to 4C. All PCR reactions had a primer annealing temperature of
Chapter 2- Materials and Methods 71 Table 2-8 Primers used in PCR for confirmatory sequencing
Gene Forward Primer Reverse Primer
Adam12 TGCACATACAGGCAGACAGG GATGTGGCTCTGGTAACCCC
Rnf121 GGTCGGTGCCAGCTATACTT TTTGCTAGTCCCAGGTCTGT
Grpc5b CAAGTGTTTCCCTTGCCCC AGCAGCACCATGTCGTAGAT
Cct7 AGTTGCTGGTGTTGCGTTCA TGACATGGAAGTGTCCACAGGTA
Tmem199 TACGCCTCAGAAGGGATTAGATCT CACGCGGGCAGTGATAATG
Osbpl3 ATTTCATGCCCCCATTTTACC TTCAGATCCCATGCCGAAAG
Impg2 CCCACTCCTAGTTCTCTGTGCTTT CATCTCTACAAGGCCATGGTTTT
Try10 TCCCTGTGGATGATGATGACAA GGGACCACCACTTACGACTTG
Epha8 TGCCCGCGTGCGG TCTCCTCCATACCTCTTCTGTCTTC
Tubb1 CATGGGCACCCTGCTCAT TGTGTTCTCTATCAGCTGGTGGAT
Nf1 CAGTCTCCTCGCCAACGC GCGACCCGGGGGC
Gsr TCACGGCGACGCTGC ATCACCAGGTAGTCGAAGGAAGAG
Kctd2 ATGGGCCGGCCCG TCCGCGGCCCGAG
Chapter 2- Materials and Methods 72 58C, except for Kctd2, which was run at 55C. The PCR products were electrophoresed on a 12.5% acrylamide gel to confirm amplification, before being sent for sequencing.
ExoSAP-IT PCR clean up kit from Affymetrix was used for purification of PCR products for sequencing. The PCR product (10µL) was mixed with exonuclease I (2µL) and shrimp alkaline phosphatase (2µL) in a PCR tube and incubated at 37C for 15 minutes to degrade primers and nucleotides, followed by 80C for 15 minutes to inactivate the ExoSAP-IT. The cleaned PCR product (7µL) was mixed with 1µL of forward or reverse primer and made up to 16µL with DNase- and RNase-free water. Sequencing was performed by Australian Genome Research Facility Westmead node.
2.2.2.5 RNA Isolation using TRIzol™
For RNA extraction, tissue samples (~50mg) were homogenised in 1mL of TRIzol reagent. Cell pellets (~2 x 106) were resuspended in 400µL of TRIzol. Samples were incubated for 5 minutes at room temperature. Chloroform (0.2mL per 1mL of TRIzol used) was added and incubated for 2-3 minutes. The samples were centrifuged (12000g, 15 mins, 4C), and the aqueous phase transferred to a clean tube, without disturbing the interface layer which contains the DNA. Total RNA was precipitated using isopropanol (0.5mL per 1mL of TRIzol used), incubated on ice for 10 minutes and then centrifuged (12,000g, 10 mins, 4C). The RNA pellet was washed in 75% ethanol (1mL per 1mL of TRIzol used) and centrifuged (7,500g, 5 mins, 4C). The pellet was air-dried for 10–15 minutes and then resuspended in DNase- and RNase-free water in a volume appropriate for the pellet size (between 20-50µL). The concentration of RNA was determined using 1µL on the NanoDrop 2000 at 260nm, and calculated assuming one 260nm unit is equal to 40µg/mL RNA. The purity of the sample was determined by the 260nm/280nm ratio, whereby ≥1.8 indicated a sample with little or no contaminating proteins.
Chapter 2- Materials and Methods 73 2.2.2.6 cDNA synthesis cDNA was prepared from RNA isolated using TRIzol described in Section 2.2.2.5. RNA (2µg) was reverse transcribed in a 10µL reaction volume containing MMLV reverse transcriptase (10U), random hexanucleotide primers (100ng), 5x first strand buffer (50mM Tris-HCl pH 8.3, 75mM KCl, 3mM MgCl2), 10mM DTT, 10mM RNasin ribonucleotide inhibitor and 500μM of each dNTP. The reaction was incubated at 37C for 1 hour, and 40µL of DNase- and RNase-free water was added to each tube to make 50µL of cDNA, which was stored at -20C until use.
2.2.2.7 Real time quantitative PCR for expression studies
Mouse and human gene expression assays were purchased as primer/probe mixes from Life Technologies (Thermo Fisher Scientific), and used following manufacturer’s instructions. The assay numbers can be found in Table 2-9. KAPA PROBE FAST qPCR master mix for ABI Prism (Sigma-Aldrich), containing Taq polymerase, reaction buffer, dNTPs and magnesium chloride, comprised half of the 20µL reaction volume. Control genes, labelled with VIC™, were multiplexed in the same well as the target gene, which was labelled with 6FAM™. Samples were run using the Applied Biosystems ABI Prism 7900 Sequence Detection System (Thermo Fisher Scientific, Australia). The thermocycling comprised an initial hold at 95°C for 10 min, followed by 40 cycles of 15 sec denaturation at 92°C and annealing for 1 min at 60°C.
Triplicate values were obtained for each sample and the amplification threshold value (Ct) defined for each assay. For determination of the expression of each gene, the comparative threshold cycle method, known as the ΔΔCt method (Schmittgen and Livak, 2008), was used to compare the target gene expression normalised to the control gene, to produce relative quantification.
Chapter 2- Materials and Methods 74
Table 2-9 Gene expression assays and control genes used in real-time quantitative PCR
Gene Species Assay Number Mouse Mm00486771_m1 Runx1t1 Human Hs00231702_m1 Mouse Mm00437762_m1 Β2-microglobulin Human 4326319E Glucuronidase beta Mouse Mm00446953_m1 (GUSB) Human 4326320E Human MYCN and Human/Th-MYCN Genotyping assay, Transgene as in Table 2.7
Chapter 2- Materials and Methods 75
2.2.2.8 Fluorescence in situ hybridisation (FISH)
To confirm the MYCN status of progeny with delayed tumours, 50µL of blood was obtained from the lateral tail vein of mice, collected in heparinised tubes. The blood was smeared onto glass slides and stained for MYCN as previously described (Cheng et al., 2007). Briefly, the blood was fixed (cold methanol for 20 mins, then room temperature in Carnoy fixative for 20 mins) then dehydrated in ethanol and air-dried, followed by denaturation in 70% v/v formamide at 72C, dehydration and air-drying. Using a nick- translation kit (Roche), the Th-MYCN plasmid was labelled with digoxigenin-11-dUTP, then denatured in hybridisation solution (10 mins at 72C) and incubated on the slides overnight. The slides were then washed in saline sodium citrate (SSC) (twice in 2x for 10 mins at 64C, once in 0.1x for 10 mins at 64C, once in 0.1x for 10 mins at room temperature) and equilibrated in 4x SSC with 0.1% Tween. FISH signal was detected with rhodamine-conjugated Fab fragments of sheep anti-digoxigenin in PBS and 1% Bovine Serum Albumin, incubated at 37ºC for 45 minutes. Slides were washed in PBS (2x) and PBS containing 0.1% Triton-X-100 for 10 minutes each at room temperature. The nuclear stain DAPI was used as a counterstain, and the signal was analysed by fluorescence microscopy. The staining was carried out by Dr Ngan Ching Cheng, the blinded analysis of the samples was performed by the author.
2.2.2.9 Cloning of overexpression and shRNA constructs
Constructs for overexpression and shRNA knock-down were created by Dr Emanuele Valli, either in the lab of Professor Giovanni Perini (Bologna, Italy) or at Children’s Cancer Institute (Sydney, Australia).
Murine Runx1t1 Isoform 1 (NM_001111027.2) was amplified by PCR using primers listed in Table 2-10 and Herculase II DNA polymerase. The PCR protocol was as follows: 95C for 2 mins followed by 35 cycles of denaturation at 95C for 20 secs,
Chapter 2- Materials and Methods 76
Table 2-10 Primers used to create mutant Runx1t1 constructs
Primer Name Sequence 5’-3’ Forward containing NotI AAGCGGCCGCTCTGTCAAAAGAAACACTTGG sequence (in red) Reverse containing BamHI AAGGATCCCTAGCGAGGCGTCGTCTCTATG sequence (in red) Forward Runx1t1 Mutant (YH) CACTGTGGCTCTTTTTGCCAGC (mutant base in green) Reverse Runx1t1 Mutant (YH) TCGGGCCGTGTTACAGCCGCTG
Chapter 2- Materials and Methods 77 annealing at 60C for 20 secs and extension at 72C for 2 mins and a final extension of 72C for 3 mins. The PCR product was inserted into p3xFLAG-CMV™-10 or -14 using the restriction enzymes NotI and BamHI, to preserve the open reading frame.
The mutant form of Runx1t1 (YH) was created by All-around PCR using p3xFLAG- CMV™-10 containing Runx1t1 and primers listed in Table 2-10. The PCR product was phosphorylated by T4 PNK using the same PCR cycling as above, however the elongation step at 72C was increased to 5 minutes per cycle, and 50ng of the product was self-ligated at 16C overnight.
The plasmid preparations were transformed into E. coli DH5 alpha top 5 and plated onto LB agar plates containing ampicillin (100µg/mL), and incubated overnight at 37C. Colonies were screened by PCR for the insert and mutation, and if present, grown in large scale cultures for plasmid isolation.
For shRNA constructs, FH1UTG, a second-generation lentiviral vector derived from FUGW was used, that contains green fluorescent protein (GFP). Three shRNA constructs were created, namely FH1UTG (empty vector), Ambion™ (based on the Ambion™ siRNA sequence) and an shRNA designed upstream of the Ambion™ sequence (shRNA3). The cloning involved 3 steps. Firstly, the forward and reverse primers (4µL; 100µM of each) containing compatible sticky ends (listed in Table 2-11) were annealed in 92µL of annealing buffer (150mM NaCl; 1mM EDTA; 10mM Tris HCl, pH 8) for 40 minutes in boiling water. The samples were cooled to room temperature and then centrifuged and phosphorylated using T4 PNK. Secondly, the FH1UTG vector was digested using BsmBI (at 55C) and XhoI (at 37C) to produce ends that match the sticky ends in the primers. The vector and primers were ligated at 4C overnight and transformed using DH5 alpha Stbl3 E. coli. Constructs were transduced into BE(2)-C cells using packaging vectors PMD2G and psPAX2 (1:1:0.5) and sorted for the top 5% expressing GFP.
Chapter 2- Materials and Methods 78 Table 2-11 Primers used to create RUNX1T1 shRNA constructs
Name Sequence shRNA RUNX1T1 TCCCGGTTCCTTACTACCCTGCATTCAAGAGATGCAGGGTAGTAAGGAACCTTTTTC Ambion Forward shRNA RUNX1T1 TCGAGAAAAAGGTTCCTTACTACCCTGCATCTCTTGAATGCAGGGTAGTAAGGAACC Ambion Reverse shRNA RUNX1T1 TCCCCCAAATAACGGCTTATCCTACCTTCAAGAGAGGTAGGATAAGCCGTTATTTGGTTTTTC 3 Forward shRNA RUNX1T1 TCGAGAAAAACCAAATAACGGCTTATCCTACCTCTCTTGAAGGTAGGATAAGCCGTTATTTGG 3 Reverse Bases in Blue: sense or passenger (same as transcript) Bases in Red: antisense or guide (real siRNA)
Chapter 2- Materials and Methods 79
2.2.2.10 Cloning of MYND domain
The MYND domain of RUNX1T1 (Liu et al., 2007b) was cloned into a pGEX-6P vector using BamHI and EcoRI restriction sites, and overexpressed as GST fusions. The YH mutant was generated using the Quikchange II mutagenesis kit (Agilent Technologies Australia), as per manufacturer’s instructions. The wild-type and mutant MYND domains were sent to the laboratory of Professor Joel Mackay and subject to Nuclear magnetic resonance (NMR) spectroscopy as previously described (Gamsjaeger et al., 2013).
2.2.2.11 Plasmid DNA isolation
A QIAGEN mini, midi or maxi plasmid purification kit was used for plasmid DNA isolation, as per manufacturers’ instructions. Briefly, a 5mL starter culture was grown in LB broth containing ampicillin (100µg/mL), for 8 hours at 37C, with shaking at 300rpm. This was then inoculated into a larger volume and grown overnight at 37C. Bacteria was pelleted by centrifuging at 6,000g (15 mins, 4C). All volumes used from here on were appropriate for the size of the culture used as per the handbook provided by QIAGEN. The pellet was resuspended in Buffer P1 with RNase A added and vortexed. Buffer P2 was then added, mixed by inversion and incubated at room temperature for 5 mins. Buffer P3 was added, mixed by inversion and incubated on ice for 15–20 mins. The mix was centrifuged at 20,000g (30 mins, 4C), and the supernatant centrifuged again to ensure removal of particulates. A QIAGEN-tip was equilibrated with Buffer QBT and the supernatant loaded onto a column. The column was washed twice with Buffer QC and the DNA eluted with Buffer QF. DNA was precipitated with isopropanol and centrifuged at 15,000g (30mins, 4C). The DNA pellet was washed with 70% ethanol and centrifuged at 15,000g for 10 mins. The DNA pellet was air-dried and resuspended in TE buffer.
Chapter 2- Materials and Methods 80 2.2.3 In vivo mouse studies
2.2.3.1 Fertility testing after ablation of tumours with cyclophosphamide
Male and female homozygous Th-MYCN mice were treated with 30mg/kg cyclophosphamide, intraperitoneally, for 5 days once they developed a 5mm abdominal tumour, which occurred at approximately 5–6 weeks of age. After one weeks’ rest, treated male mice were harem-mated to one cyclophosphamide treated homozygote female and one untreated hemizygote female to test fertility after treatment with chemotherapy. Homozygote offspring were palpated, and time to both first palpable tumour and tumour cull (10mm) were determined. Kaplan–Meier plots of time to first palpable tumour and tumour cull were generated using GraphPad Prism version 6.01 for Windows (GraphPad Software, La Jolla, California, USA). To test significance between the groups a Log-rank (Mantel–Cox) test was applied to the data using GraphPad Prism.
2.2.3.2 ENU mutagenesis
Male Th-MYCN mice were palpated twice weekly to detect abdominal tumours. Once a tumour was 5mm in size, the mice were treated with 30mg/kg cyclophosphamide for 5 consecutive days. After one week’s rest, mice were ready for the ENU injections.
N-ethyl-N-nitrosourea was prepared using the spectrophotometry method (Salinger and Justice, 2008). All handling of ENU was carried out in a cytotoxic cabinet over a tray containing 5% sodium thiosulfate in 0.1M Sodium Hydroxide. Due to its hazardous nature, ENU preparation and injection was carried out by Professor Douglas Hilton, and Drs Eric Sekyere and Neil Davies. A 26 gauge needle was used to pierce the ISOPAC® bottle containing 1g of ENU to prevent pressurisation. Absolute ethanol, 5mL, was injected into the bottle using a 21G needle, and the ENU was warmed gently with agitation under a hot tap. When the solution was clear, 45mL of citrate phosphate (0.1M sodium dihydrogen phosphate + 0.05M sodium citrate in water), pH 5, was injected into the bottle using a 21 gauge needle. The ENU was diluted to 1/50 and 1/25 in citrate phosphate in a disposable cuvette. The absorbance at 395nm was read on a spectrophotometer, with citrate phosphate (1mL) used as a blank. The concentration of Chapter 2- Materials and Methods 81 the ENU solution was determined as follows. To obtain an average 1/25 value of the two dilutions, the following calculation was used: ((1:25 reading) + (2 x 1:50 reading)) /2 = A. Secondly, to obtain a 1:1 value, A was multiplied by 25 (the dilution factor). One absorbance unit of ENU = 1.39mg/mL, therefore the actual concentration is the 1:1 value x 1.39mg/mL. The concentration of ENU was adjusted to 15mg/mL using citrate phosphate, and each mouse was injected with 100uL of the solution, with the needle pointed towards the testis rather than the diaphragm, to administer 66mg/kg, as the average weight of the mice was 24g. The mice were transferred to clean cages containing cotton wool to absorb any urine and faeces for 24 hours and then transferred to new cages containing corn cob bedding and enrichment. All dirty cages were decontaminated with 5% sodium thiosulfate before washing. ENU injections were repeated at weekly intervals for a total of 3 doses at 66mg/kg for each mouse. Approximately 4 weeks after the last injection of ENU, mice were mated with cyclophosphamide-treated homozygote Th-MYCN females, and monitored for signs of ENU-related toxicity, tumour relapse and for pregnancy and births in the box.
Progeny of ENU-treated males were monitored for tumour development by abdominal palpation. To test for heritability of the phenotype, mice with a delayed tumour (>7.5 weeks of age) were treated with 30mg/kg cyclophosphamide to ablate their tumour and mated to cyclophosphamide-treated homozygote Th-MYCN, and their offspring monitored for tumour development. Scatterplots were generated using GraphPad Prism software.
2.2.3.3 Mating for exome and whole genome sequencing
To elucidate the genes responsible for a delayed tumour phenotype, several mating strategies were used, which are discussed in detail in Chapter 4. Briefly, for exome sequencing, mice with delayed tumours were mated either with another tumour-delayed mouse of the same line, or with cyclophosphamide-treated Th-MYCN homozygote mice. Offspring were monitored for tumour development by abdominal palpation. Progeny that developed tumours normally, and those that demonstrated an increase in tumour latency, were collected for exome sequencing. Scatterplots were generated using GraphPad Prism software.
Chapter 2- Materials and Methods 82 For backcrossing, tumour-delayed mice were mated with C57BL/6 or BALB/c mice. Resultant progeny were then either crossed together (intercrossed) or backcrossed to cyclophosphamide-treated Th-MYCN homozygote mice. Homozygote Th-MYCN progeny were monitored for tumour development by abdominal palpation. For comparison, normal Th-MYCN mice were backcrossed in the same fashion as the ENU mutagenesis mice and followed for tumour development. Scatterplots were generated using GraphPad Prism software. Two-tailed Fisher’s exact tests (95% confidence interval) were used to determine differences between backcrossed populations (GraphPad Prism).
2.2.3.4 Th-MYCN cross with Runx1t1 knock-out mice
Hemizygote Th-MYCN mice were crossed with Runx1t1 heterozygote knock-out mice. Progeny that were hemizygote for MYCN and heterozygote for Runx1t1 were then mated. Homozygote and hemizygote Th-MYCN progeny with one or two functional copies of Runx1t1 were monitored for 13 months for tumour development by abdominal palpation. Kaplan–Meier plots of time to medium palpable tumour (10mm) were generated, and a Log-rank (Mantel–Cox) significance test applied using GraphPad Prism.
Chapter 2- Materials and Methods 83 CHAPTER 3 ENU MUTAGENESIS OF THE TH-MYCN MOUSE MODEL OF NEUROBLASTOMA
3.1 Introduction
The Th-MYCN mouse model recapitulates the human disease of neuroblastoma and, as discussed in Section 1.3.2, has provided a tool for researchers to understand the biology of this disease. The 100% tumour incidence in mice homozygous for the MYCN transgene, makes this model ideal for an ENU mutagenesis study, however the early tumour onset of less than 7 weeks precludes breeding of mice with this genotype directly, necessitating maintenance of the colony by breeding MYCN hemizygous mice. Although an alternative would be to mutagenise the hemizygous MYCN mice, since they have longer tumour latencies of 8-16 weeks following birth, unfortunately, their tumour incidence varies between 25-35%, depending on the colony management (Burkhart et al., 2003; Hansford et al., 2004). The hemizygous mice, having only one copy of the transgenic cassette, require a secondary, and as yet unknown, mechanism to develop tumours, although it is known that they increase the MYCN copy number specifically in the tumours as a result (Hansford et al., 2004; Norris et al., 2000). Thus, use of these mice for an ENU mutagenesis study is precluded since the majority of offspring would fail to develop tumours, making it extremely difficult to determine whether the lack of any tumour formation observed might be due to an introduced mutation. Therefore, it was decided to investigate the use of chemotherapy to treat the homozygous MYCN mice, thereby delaying tumour formation in these animals, and thus providing an opportunity for breeding to occur.
The choice of chemotherapy needed to take into account several factors. The first, and most important, was that the ENU would likely cause long periods of sterility before the mice produced offspring, therefore any chemotherapy would need to be effective in eradicating tumours for at least 3-6 months without causing any significant toxicity. In this model, treatment with the chemotherapeutic drugs, VCR and VP-16, only provides a short extension of lifespan, while the use of higher doses results in significant toxicity (Burkhart et al., 2009). Cisplatin treatment gives approximately 30 days extension of life, with some longer-term survivors at 2mg/kg, and could be used if higher doses were achievable; however, higher doses in this model are also toxic (Hogarty et al., 2008).
Chapter 3- ENU Mutagenesis of the Th-MYCN Mouse Model of Neuroblastoma 84 The Th-MYCN mice have been shown to be very sensitive to cyclophosphamide, with long term cure rates (>5 months) achievable at doses between 20-75mg/kg (J Murray, unpublished data). However, sterility in mice (reviewed in (Ghobadi et al., 2016)) and humans (Lentz et al., 1977) following treatment with cyclophosphamide has been reported, raising the issue of dose and timing with this drug treatment, and the effect it would have on spermatogenesis (Lentz et al., 1977; Siimes and Rautonen, 1990). In one study, spermatogenic dysfunction was reported in 40% of patients (6 out of 15) receiving cyclophosphamide either before or during puberty, while all 4 patients that received cyclophosphamide after puberty reported spermatogenic dysfunction (Lentz et al., 1977). The study by Siimes and Rautonen conducted on 66 adult men, including 8 patients with neuroblastoma, concluded that cyclophosphamide treatment resulted in decreased sperm production (Siimes and Rautonen, 1990). Similarly, in mice, a decrease in male mouse fertility after cyclophosphamide treatment has also been reported, with decreased sperm concentration, motility and testis weight noted (reviewed in (Ghobadi et al., 2016)).
In contrast to these findings, no menstrual dysfunction was noted in females following cyclophosphamide treatment in the Lentz study. This was the case in all 13 female patients studied, regardless of whether the cyclophosphamide was given pre- or post- puberty (Lentz et al., 1977). In a second study using combination chemotherapy including cyclophosphamide, the majority of women studied, recommenced regular menstrual cycles after completion of therapy and a subset also delivered healthy babies, with no major birth defects (Gershenson, 1988).
The experiments described in this chapter investigated early tumour onset and fertility following administration of cyclophosphamide to both female and male Th-MYCN mice. Following selection of an appropriate dose and schedule of cyclophosphamide drug treatment, an ENU mutagenesis screen was conducted with the goal of generating offspring with heritable mutations that could delay or abrogate neuroblastoma tumour formation.
Chapter 3- ENU Mutagenesis of the Th-MYCN Mouse Model of Neuroblastoma 85 3.2 Results
3.2.1 Testing fertility of cyclophosphamide treated Th-MYCN mice
Cyclophosphamide has previously been used in our laboratory in Th-MYCN mice, with doses above 30mg/kg resulting in complete cure of mice over a 5 month period. Since fertility was a factor, 30mg/kg cyclophosphamide was chosen as the lowest dose capable of ablating tumour formation for the period of the study. Doses lower than this result in tumour relapse in a number of mice. A dose of 30mg/kg of cyclophosphamide, given once daily for 5 days was administered to both homozygous male and female mice when they presented with a 5mm tumour by palpation. These mice were then mated in harems, whereby, a cyclophosphamide-treated male mouse was mated with two female mice; one female a cyclophosphamide-treated homozygote and the other an untreated hemizygote. Such a study could determine whether any infertility observed arose from the male or female mouse. Five harems were mated and the time to both first palpable tumour and tumour cull (10mm) of the offspring was assessed. For the hemizygote females in this study, only those progeny homozygous for the transgene were assessed. The results demonstrated that each homozygote cyclophosphamide- treated male produced progeny with both the cyclophosphamide-treated homozygote and hemizygote female (Table 3-1). There was no difference overall between the number of litters produced by the homozygote or hemizygote female (mean 4.2 in each case, P>0.9999). The average litter size overall was slightly greater for the progeny born to the hemizygote females compared to the homozygote females, however this was not statistically significant (mean 6.43 for hemizygote versus 5.39 for homozygote female, P=0.253).
Tumour development was also assessed in mouse offspring following cyclophosphamide treatment. There was no significant difference in either time to first palpable tumour or tumour cull for offspring from either the hemizygote or homozygote female in Harem 3 and 5 (Figure 3-1, E and F, G and H and Table 3-2), although there were differences between the time to first palpable tumour for Harems 2 and 4 (Figure 3-1, C and G and Table 3-2) and to tumour cull for Harems 1 and 2 (Figure 3-1, B and D and Table 3-2). For Harem 2, the homozygote offspring from the hemizygote female developed tumours earlier, whilst in Harem 1 and 4, the reverse was true, where the
Chapter 3- ENU Mutagenesis of the Th-MYCN Mouse Model of Neuroblastoma 86
Table 3-1 Number of litters and average litter size of offspring from harem matings
Harem Male Female Number of Average number (+/+)* (genotype) litters litter size
A2708 (+/+) 5 4.6 1 A2700 A2695 (+/-)† 5 9.2
A3011 (+/+) 5 6.8 2 A2993 A2989 (+/-) 3 6
A2992 +/+ 4 6.75
3 A2994 A3009 (+/-) 1 (Tumour)‡ 6
A3447 (+/-) 2 5
A3274 (+/+) 4 4.5 4 A3282 A3448 (+/-) 5 5.4
A3333 (+/+) 3 4.3 5 A3336 A3330 (+/-) 5 7
* +/+ indicates mice homozygous for the MYCN transgene † +/- indicates mice hemizygous for the MYCN transgene ‡ mouse developed a tumour and was replaced
Chapter 3- ENU Mutagenesis of the Th-MYCN Mouse Model of Neuroblastoma 87
Figure 3-1 Survival curves of progeny from individual harems. Time to first palpable tumour (A, C, E, G and I) and tumour cull (B, D, F, H, J) of offspring from Harems 1 (A,B), 2 (C,D), 3 (E, F), 4 (G, H) and 5 (I, J). The progeny from homozygote females is shown in red and hemizygote females in black.
Chapter 3- ENU Mutagenesis of the Th-MYCN Mouse Model of Neuroblastoma 88
Table 3-2 Statistical comparisons of time to first palpable tumour and tumour cull between offspring from homozygous and hemizygous female.
Harem Tumour P value number First Palpable Tumour 0.9818 1 Tumour Cull 0.0106
First Palpable Tumour 0.04 2 Tumour Cull 0.0051
First Palpable Tumour 0.2015 3 Tumour Cull 0.5896
First Palpable Tumour 0.0049 4 Tumour Cull 0.0975
First Palpable Tumour 0.2439 5 Tumour Cull 0.2578
Chapter 3- ENU Mutagenesis of the Th-MYCN Mouse Model of Neuroblastoma 89 offspring from the homozygote female parent developed tumours earlier. However, when all the offspring were combined and compared, there was no significant difference between the progeny from either female parent and the time to first palpable tumour (P=0.2291) and tumour cull (P=0.2214) and all progeny were euthanased with a 10cm tumour before 7.5 weeks of age (Figure 3-2). These results therefore showed that the cyclophosphamide dose used and the timing of administration did not affect fertility in either male or female Th-MYCN mice. Importantly, the chemotherapy did not significantly alter the overall tumour formation in the offspring.
3.2.2 ENU Mutagenesis
Since the cyclophosphamide-treated mice produced viable offspring that developed tumours in the same fashion as a normal homozygote Th-MYCN mouse would, an ENU mutagenesis project was undertaken to identify genetic mutations that could influence tumour development in neuroblastoma. Homozygote MYCN male mice (81) were treated with 30mg/kg cyclophosphamide, as in the pilot experiment and then rested for 1-2 weeks post injection to recover. Common doses used in ENU mutagenesis experiments range from 80-90mg/kg, however 129/SvJ mice, the background of Th- MYCN, appear to be intrinsically sensitive to these doses, with infertility and death reported (Weber et al., 2000). Therefore, a lower dose of 66mg/kg was chosen, in the belief that this would be low enough to avoid infertility and death, but high enough to induce mutations in the sperm. Preparation of ENU and injections were performed as outlined in 2.2.3.2., with male mice injected, intraperitoneally with 66mg/kg of ENU, once per week for 3 weeks and then mated to cyclophosphamide-treated female homozygote mice (mating scheme as shown in Figure 3-3).
The male mice were sterile for approximately 16 weeks post-ENU injection. After the period of sterility, a total of 2525 pups were born to the 81 ENU-treated males (Table 3-3). Of these, only 1716 (67.96%) were viable through to weaning, with the remaining 32.04% either dying before weaning (10.06%), suffering whole litter wipeout (21.90%) or being runty and needing to be euthanased (0.08%). The 1716 viable pups were monitored for tumour development, and this will be discussed in section 3.2.4. Other than changes in tumour development, there were only 2 mice that exhibited obvious
Chapter 3- ENU Mutagenesis of the Th-MYCN Mouse Model of Neuroblastoma 90
Figure 3-2 Combined survival graphs for progeny from both homozygote and hemizygote females. Time to first palpable tumour (A) and tumour cull (B) of offspring from Harems 1-5. The progeny from the homozygote females is shown in red and the hemizygote females in black.
Chapter 3- ENU Mutagenesis of the Th-MYCN Mouse Model of Neuroblastoma 91
Figure 3-3 ENU mutagenesis and homozygote Th-MYCN mouse mating scheme
Chapter 3- ENU Mutagenesis of the Th-MYCN Mouse Model of Neuroblastoma 92
Table 3-3 All mice born to ENU treated males ENU progeny Number As a percentage of mice born
Viable Offspring 1716 67.96%
Pre-wean death 254 10.06%
Litter wipe-out 553 21.90%
Runty 2 0.08%
Total 2525 100%
Chapter 3- ENU Mutagenesis of the Th-MYCN Mouse Model of Neuroblastoma 93 phenotypic changes. The first had an extra fold of skin on its head that created a Mohawk-like appearance (Figure 3-4, A), while the second unusual mouse had fur that was wavy in appearance (Figure 3-4, B). All other mice that were born and survived to weaning appeared normal.
3.2.3 Fate of the ENU Treated Males
Following cyclophosphamide and ENU treatment the mice were normal, other than the sterility mentioned, and showed no ill effects until after they had begun to produce offspring. However, as they aged, side effects from the ENU injections manifested, as previously reported in other ENU mutagenesis studies. Only 5 out of 81 (6.2%) of the mice remained normal and were euthanased when they no longer produced offspring (Table 3-4). The majority of the mice (72.8%) developed thymoma (Table 3-4 and Figure 3-5, A), a carcinoma not usually observed in this model, nor after cyclophosphamide treatment. Although this was the primary reason for euthanasia, upon necropsy, enlargement of the liver (hepatomegaly) and/or spleen (splenomegaly) was also noted in some of these mice (Figure 3-5, B and C, respectively). In 13.6% of mice, the primary reason for euthanasia was the presence of splenomegaly and hepatomegaly, which was accompanied by lesions of the liver or heart and lungs in a small subset (Table 3-4). Three of the mice (3.7%) succumbed to an infection in either the intestines or reproductive organs, while the remainder had either an unusual liver, or a lesion on the skin or spleen, with only 1 mouse in each of these groups requiring euthanasia due to these conditions.
Chapter 3- ENU Mutagenesis of the Th-MYCN Mouse Model of Neuroblastoma 94
Figure 3-4 Unusual phenotypes produced as a product of ENU mutagenesis. These included an extra fold of skin (A) and wavy fur (B).
Chapter 3- ENU Mutagenesis of the Th-MYCN Mouse Model of Neuroblastoma 95 Table 3-4 Primary reason for euthanasia of 81 ENU treated males Primary reason Number Additional condition Number requiring (Percentage upon necropsy (Percentage euthanasia of ENU of ENU treated treated males (81)) males (81)) Splenomegaly 9 (11.1%)
Hepatomegaly and 5 (6.2%) splenomegaly Thymoma 59 (72.8%) Lung lesions 1 (1.2%)
Clear fluid in thoracic 1 (1.2%) cavity Liver lesions 1 (1.2%)
Splenomegaly and Heart and lung lesions 1 (1.2%) 11 (13.6%) Hepatomegaly Spotty liver and enlarged 1 (1.2%) pale kidneys Infection in intestines 3 (3.7%) - or reproductive organs Punctate and rough 1 (1.2%) - liver Splenic lesions 1 (1.2%) -
Lesion on chest that 1 (1.2%) - bled Normal 5 (6.2%) -
Chapter 3- ENU Mutagenesis of the Th-MYCN Mouse Model of Neuroblastoma 96
Figure 3-5 Conditions requiring euthanasia in the ENU treated males. Thymoma (A) was the most common reason for euthanasia, followed by hepatomegaly (B, left) and splenomegaly (C, left). As a comparison, a normal spleen and liver from a similar aged Th-MYCN mouse is shown on the right in each case.
Chapter 3- ENU Mutagenesis of the Th-MYCN Mouse Model of Neuroblastoma 97 3.2.4 Tumour development of ENU Progeny
The 1716 viable offspring were monitored for tumour development by abdominal palpation. Of these, 1665 (65.94% of mice born and 97.03% of mice able to be monitored for tumour development) developed tumours normally (Table 3-5 and Figure 3-6). Normal tumour development was considered to be any mouse that developed a small (5mm) palpable tumour before 7.5 weeks of age. Homozygote mice usually all develop a 10mm tumour by 7 weeks of age, in our hands. The 7.5 week cut-off allowed for drift caused by several possible factors, including mating of homozygote mice instead of the normal hemizygous pairings or effects due to cyclophosphamide treatment. There was a subset of mice that developed tumours earlier than expected, before 4.5 weeks of age (Figure 3-6). Usually, chemotherapy treatment protocols at a small palpable tumour begin between 5.5-6.5 weeks of age, and anything that developed a small tumour before 4.5 weeks of age would be considered more aggressive than normal.
Fifty-one mice (2.02% of mice born and 2.97% of mice able to monitored for tumour development) had a delay in tumour development beyond 7.5 weeks of age (Figure 3-6 and Table 3-6). The time to small palpable tumour ranged from 7.6 to 11.4 weeks, with two mice not developing tumours, of which one was followed through to 61 weeks of age without developing a tumour (1929) and a second that was only followed for 10 weeks (1748) before anaesthetic overdose occurred while re-tailing to confirm its genotype, resulting in death (Table 3-6). A third mouse (1590) developed an unusual tumour at 49 weeks of age (discussed further in chapter 4). Where possible, each of the mice that had a delay was treated with cyclophosphamide to test heritability of the phenotype. Two of these mice (0459 and 1090) were treated when the tumours were only 1-2mm due to the timing of their tumour development. There were 4 mice where it was not possible to treat with cyclophosphamide and test heritability due to the development of breathing problems (0117 and 2010) or hind limb paralysis (0055 and 1470) from a thoracic tumour. A further 4 mice relapsed with a tumour (0358, 0651, 0941 and 1090) before heritability could be confirmed and a final mouse developed a lesion on the dorsal flank (1204) requiring euthanasia.
Chapter 3- ENU Mutagenesis of the Th-MYCN Mouse Model of Neuroblastoma 98
Table 3-5 Viable progeny of ENU treated males split according to normal or delayed tumour development ENU Progeny Number As a percentage As a percentage of mice born of mice followed Normal tumour 1665 65.94% 97.03% development Delay in tumour 51 2.02% 2.97% development
Total 1716 67.96% 100%
Chapter 3- ENU Mutagenesis of the Th-MYCN Mouse Model of Neuroblastoma 99
6 5
6 0 1 9 2 9
5 5
5 0 1 5 9 0 4 5 1 2 . 0
1 1 . 5
1 1 . 0
1 0 . 5
) 1 0 . 0 1 7 4 8 s
k 9 . 5 e
e 9 . 0 C u t
w (
8 . 5 p o i n t f o r e
m 8 . 0 m a t in g i
T 7 . 5
7 . 0
6 . 5
6 . 0
5 . 5
5 . 0
4 . 5
4 . 0
3 . 5
3 . 0
P r o g e n y o f E N U T r e a t e d M a l e s
Figure 3-6 Tumour development of 1716 mice from ENU treated males. Mice that developed tumours beyond 7.5 weeks of age were tumour development delayed and were mated. Line indicates the 7.5 week cut point for mating. All mice developed a tumour () except two as shown ().
Chapter 3- ENU Mutagenesis of the Th-MYCN Mouse Model of Neuroblastoma 100
Table 3-6 Th-MYCN progeny with delayed or absent tumour development Mouse Time to 5mm tumour Heritability tested number* (weeks) 0022 8.6 Yes
0055 8 No - Hind Limb Paralysis
0099 8.4 Yes
0110 8.7 Yes - no viable pups and prolapse
0117 8.3 No - Thoracic Tumour
0146 8 Yes
0148 9 Yes
0306 8.7 Yes
0358 8.1 No - relapse
0456 7.9 Yes
0459 9.3 (1-2mm) Yes
0531 8.4 Yes
0595 7.6 Yes - no viable pups
0607 8.3 Yes
0629 7.9 Yes
0651 7.9 No - relapse
0664 8.3 Yes
0731 9.7 Yes - sterile
0755 9.4 Yes
0915 7.8 Yes
Chapter 3- ENU Mutagenesis of the Th-MYCN Mouse Model of Neuroblastoma 101
Table 3–6 continued Mouse Time to 5mm tumour Heritability tested number* (weeks)
0941 7.7 No - relapse
0987 11.4 Yes
1030 8.8 Yes
1069 9 Yes
1090 7.4 (1-2mm) No - relapse
1134 7.9 Yes
1204 7.9 No – lesion on side
1308 7.9 Yes
1426 9.4 Yes - no viable pups
1448 11 Yes
1470 9.4 No – Hind Limb Paralysis
1504 8.9 Yes - sterile
1508 8 Yes - no viable pups
1550 8.5 Yes
1552 8.5 Yes - no viable pups
1590 49 Yes
1611 7.9 Yes - no viable pups
1666 9.2 Yes
1705 8 Yes
1748 10 (no tumour) No – accident
Chapter 3- ENU Mutagenesis of the Th-MYCN Mouse Model of Neuroblastoma 102
Table 3–6 continued Mouse Time to 5mm tumour Heritability tested number* (weeks)
1835 7.9 Yes
1853 7.9 Yes
1875 8.4 Yes
1929 61 (no tumour) Yes
1978 7.9 Yes
2010 10 No – Thoracic Tumour
2105 8.9 Yes
2204 8 Yes
2209 8 Yes
2210 9 Yes
2246 7.8 Yes
* Mice are presented in mouse order number with their time to 5mm tumour, and the status of heritability testing.
Chapter 3- ENU Mutagenesis of the Th-MYCN Mouse Model of Neuroblastoma 103 3.2.5 Heritability testing of delayed tumour phenotype
The mouse progeny exhibiting a delay in tumour formation were each mated to a cyclophosphamide-treated homozygote from the Th-MYCN colony to test for heritability of the delayed tumour phenotype (as outlined in Figure 3-7). If a phenotype is heritable in a Mendelian fashion, half of the offspring should develop tumours normally, while half should display the same phenotype as their delayed parent. Of the original 51 mice with a delayed phenotype, 41 were able to be tested for heritability. Two of the mice (0731 and 1504) were sterile as no pups were born, whilst 6 mice (0110, 0595, 1426, 1508, 1552 and 1611) produced pups, but these were not viable and died before weaning (Table 3-6). Thirty three mice produced viable offspring that could be monitored for tumour development. The majority of mice tested were not able to pass on the delayed tumour phenotype to their offspring, or there was little distinction between normal tumour development and a delayed tumour phenotype (Figure 3-8). Of note, however, were several mice that showed some inheritance. Mouse 0306 which had a delay to 8.7 weeks had several offspring with as good or better tumour delay (Figure 3-8, A). Mouse 1030, which developed a tumour at 8.8 weeks of age did not produce many offspring, however, two of them demonstrated a delay in tumour development (Figure 3-8, B). Mouse 1705 had 5 offspring that had a tumour delay better than her own of 8 weeks (Figure 3-8, B). Mouse 1666 (9.2 weeks) and 2105 (8.9 weeks) only had one offspring each that had a tumour delay better than the parent, despite producing a substantial number of offspring (Figure 3-8, B and C). Whilst there was some delay observed with these 5 mice, they all proved difficult to maintain as lines as either the delayed phenotype was difficult to distinguish from normal tumour development or the mice failed to produce further offspring in subsequent generations that could confirm the phenotype.
The two mice with the greatest delay in tumour development, mouse 1590 who developed a tumour at 49 weeks of age and 1929, who never developed a tumour and was followed until 61 weeks of age, proved to have the most substantial heritability to the next generation. Mouse 1590 demonstrated perfect Mendelian inheritance to the next generation, with 11/22 mice developing tumours normally (termed ‘unsuppressed’) and 10 not developing tumours despite being followed for 48 weeks or longer (termed ‘suppressed’) (Figure 3-9). One of the suppressed mice did develop a tumour, at 20.6
Chapter 3- ENU Mutagenesis of the Th-MYCN Mouse Model of Neuroblastoma 104
Figure 3-7 Mating diagram for testing inheritance to subsequent generations
Chapter 3- ENU Mutagenesis of the Th-MYCN Mouse Model of Neuroblastoma 105
Figure 3-8 Time to small palpable tumour (weeks) of progeny of 31 mice in mouse number order. The progeny () are plotted alongside their respective parent () as a reference.
Chapter 3- ENU Mutagenesis of the Th-MYCN Mouse Model of Neuroblastoma 106
Figure 3-9 Mendelian inheritance in offspring of Mouse 1590. Half of the mice developed tumours normally () and half failed to develop tumours (). Parent 1590 that developed a tumour at 49 weeks is shown () as a comparison.
Chapter 3- ENU Mutagenesis of the Th-MYCN Mouse Model of Neuroblastoma 107 weeks of age, a substantial delay of almost 14 weeks over its unsuppressed counterparts. Mouse 1929, also demonstrated substantial inheritance in her offspring, however, the heritability was ambiguous. Of the 13 direct progeny, 6 showed no delay in tumour development (5.1 – 7.7 weeks), 5 showed some delay in tumour development ranging from 9.7 to 15.7 weeks of age, while 2 mice did not develop tumours, like 1929 (Figure 3-10). Whilst the inheritance of delayed versus normal tumour development was Mendelian, the distribution of the phenotype was not as clearly defined as that for the offspring from the 1590 mating.
To elucidate the gene(s) responsible for the phenotype a combination of backcrossing, mapping and sequencing was undertaken. For both mouse 1590 and 1929, this will be discussed in Chapter 4.
Chapter 3- ENU Mutagenesis of the Th-MYCN Mouse Model of Neuroblastoma 108
Figure 3-10 Inheritance of tumour phenotype in offspring of Mouse 1929. Approximately half of the mice developed tumours normally, whilst the majority of the remainder had a delayed tumour phenotype (). A subset of these mice did not develop tumours (). Parent 1929 that did not develop a tumour is shown () as a comparison.
Chapter 3- ENU Mutagenesis of the Th-MYCN Mouse Model of Neuroblastoma 109 3.3 Discussion
To undertake an ENU mutagenesis project using Th-MYCN homozygote mice it was necessary to ‘cure’ the mice of their neuroblastoma tumours using cyclophosphamide, to allow sufficient time for the mice to recover from the sterility caused by ENU injection and produce viable offspring. The pilot experiment indicated that this was achievable with the dose of cyclophosphamide chosen, and indicated that there were no overall differences with respect to litter size and tumour development, in treated versus untreated mice. This is in contrast to the literature that reports decreases in sperm concentration, motility and testis weight after cyclophosphamide injection. Although these parameters were not examined specifically in this study, the mice were fertile and produced viable offspring. There are several key differences between these studies and the one reported here. The first difference is the choice of strain. The Th-MYCN mice are on a 129/SvJ background, and whilst it is known that there is reduced implantation of embryos using fresh 129/SvJ sperm over other strains like C57BL/6J and BALB/c (Kusakabe et al., 2001), it is not known whether they are more or less susceptible to the effects of cyclophosphamide over strains that are readily used for these studies like Swiss albino or BALB/c (Tripathi and Jena, 2008; Zhao et al., 2015).
The second key difference is the dose and frequency of cyclophosphamide used. The equivalent human therapeutic dose in the mouse is 200mg/kg, given once per week for multiple doses, which is known to impair fertility in humans and mice (Carmely et al., 2009). However the cumulative dose used in this study over 5 days of 150mg/kg is 8 times lower than that used in the Carmely study, which used a cumulative dose of 1200mg/kg (Carmely et al., 2009). The dose of cyclophosphamide used has been correlated to the severity of the damage observed. At 50mg/kg cyclophosphamide for 5 doses, which is similar to the dose used in this study, testis weight was affected, however it was only at higher doses (100mg/kg and 200mg/kg for 5 doses), that there was also significant differences in body and epididymis weight (Tripathi and Jena, 2008). Perhaps the most compelling evidence that dose is the key difference is that whilst all doses used in the Tripathi study produced abnormal sperm head morphology, and decreased the sperm length and motility, the only dose that did not affect the seminiferous tubules was 50mg/kg, which means that spermatogonia which form
Chapter 3- ENU Mutagenesis of the Th-MYCN Mouse Model of Neuroblastoma 110 spermatozoa, are not permanently damaged by this dose of cyclophosphamide (Tripathi and Jena, 2008), and therefore longer term fertility should remain intact.
The finding that ENU-treated mice were sterile for approximately 16 weeks post ENU injection was similar to other studies using 129X1/SvJ mice. Sterility in this mouse strain, which is closely related to the background strain of the Th-MYCN mice (129/SvJ), was 13.3-20.5 weeks when given 3 doses of 85mg/kg ENU (Weber et al., 2000). Death following ENU injection can be common if the dose or schedule of ENU is too high for the strain used (Hitotsumachi et al., 1985). Although there was no death observed in the current study following administration of ENU, the 129X1/SvJ strain appears to be quite sensitive to the effects of this mutagen, as two doses of ENU commonly used in most strains (4 x 90mg/kg and 4 x 80mg/kg) resulted in death of 5 out of 10 males injected while the remainder were infertile and did not produce offspring (Weber et al., 2000).
ENU treatment is also known to cause health effects to the mice that survive the injection outside of sterility. For instance, ENU is known to suppress the immune system and as a result infections can occur, as was shown in a previous study of ENU- treated males (Justice et al., 2000). Only 3 of the 81 mice treated in the current study displayed signs of an infection. Another side effect of ENU treatment is cancer, unrelated to the disease of interest. In a study of retired ENU mice, Justice and colleagues found that just over half of their ENU-treated cohort (16/30) had some form of tumour, most commonly in the lung (6), thymus (3) and liver (3), as well as one mouse each having kidney, skin, prostate or colon tumours (Justice et al., 2000). The most common problem seen in this ENU screen was the development of thymoma in over 70% of the mice treated, which did impact the health of the mouse, unlike the Justice study. Also noted were lesions of the lung, liver, heart, skin and spleen similar to the study above (Justice et al., 2000).
In the current ENU mutagenesis screen, just under a third of the progeny produced died pre-weaning or suffered litter wipe-out before weaning. Whilst no analysis was undertaken, it is assumed that they were a product of mutations not sustainable to life. There is little information in the literature about the numbers of pre-wean deaths in ENU mutagenesis screens, however, region-specific screens have suggested that this can range between 14-19% (Nguyen et al., 2011). There are also specific ENU screens
Chapter 3- ENU Mutagenesis of the Th-MYCN Mouse Model of Neuroblastoma 111 designed to look for both recessive-lethal and detrimental mutations (Justice, 2000) and viability screens examining a range of developmental defects pre-birth (Buac et al., 2008). The amount of pre-wean deaths in this screen was quite high and may be attributable to the embryonal nature of neuroblastoma, where events occur during the early stages of development.
There were only two obvious phenotypic changes observed in the ENU progeny. The first an extra fold of skin on the forehead and the second, a wavy coat. Spontaneous mutations resulting in wavy fur have been reported in mouse lines from a number of studies, including the following lines: waved-1 (mutation in transforming growth factor); waved-2 (mutation in epidermal growth factor receptor); and wellhaarig (mutation in epidermal-type transglutaminase 3) (Brennan et al., 2015; Luetteke et al., 1994; Luetteke et al., 1993). These studies also describe defects in whiskers and open eyelids at birth, which were not noted here. Several large-scale ENU mutagenesis studies have also yielded wavy coats. Mutations in Velvet, waved-5 and dark skin-5 lines, all resulting from mutations in the epidermal growth factor receptor, also yielded wavy fur, the last without the open eyelid phenotype that the other mouse lines possessed (Du et al., 2004; Fitch et al., 2003; Lee et al., 2004). Given that most of these mutations are in epidermal genes, it is possible that the extra fold of skin was also as a result of a mutation in a component of the epidermal layer, although this has not been reported previously.
A small ENU experiment still requires more than 1000 animals to be screened, even with 100% penetrance of the starting phenotype (Carpinelli et al., 2012). The expected frequency in an ENU screen investigating a dominant phenotype is approximately 40 animals per 1000 screened (4%) (Cordes, 2005) which is similar to the number identified in the screen described here. However, robust inheritance to the next generation was only observed in 5 mouse lines, resulting in an overall rate of 0.29%. It is plausible that treatment with cyclophosphamide could have resulted in mutations that yielded a false-positive effect, however, this would likely be expected to result in increased numbers of mice displaying a delay in tumour development. Cyclophosphamide treatment, is therefore, unlikely to be cause of the delayed phenotypes and these are more likely as a result of ENU mutagenesis.
Chapter 3- ENU Mutagenesis of the Th-MYCN Mouse Model of Neuroblastoma 112 The ENU mutagenesis study also yielded a number of mice with tumours earlier than expected. In hindsight, whilst it would have been difficult to follow these mice due to the very aggressive nature of the tumours, these mice may have yielded additional information about oncogenes or drivers of tumorigenesis in this model. ALK mutations are known to be drivers of neuroblastoma in humans and in a recent study, in the Th- MYCN mouse model (Berry et al., 2012), tumours arose as early as 4 weeks in some lines, similar to the current study. Upon analysis of the tumours in this Th-MYCN/ALK model, not only was there stabilisation of the MYCN protein, but also upregulation of murine Mycn not usually observed in the Th-MYCN mouse (Berry et al., 2012).
However, the most striking result from the current study was the tumour delay of 1590 at 49 weeks and 1929, who never developed a tumour, despite being followed for 61 weeks, and these two lines were selected for further study.
Therefore, in summary, these results suggest that the Th-MYCN mouse is an ideal model for ENU mutagenesis, although care must be taken when selecting a dose and schedule of ENU due to the risk of potential death and sterility in similar strains. An ENU screen requires significant time and resources to carry out, especially if the phenotype of interest cannot be detected pre-weaning, and should not be undertaken lightly as a lack of either will not yield fruitful results.
Chapter 3- ENU Mutagenesis of the Th-MYCN Mouse Model of Neuroblastoma 113 CHAPTER 4 ELUCIDATION OF THE GENES RESPONSIBLE FOR DELAYED TUMOUR FORMATION IN ENU-TREATED TH-MYCN MICE
4.1 Introduction
As the majority of functional mutations from an ENU screen occur within coding sequences, exome sequencing is an ideal starting point to narrow the potential list (Kile and Hilton, 2005; Nguyen et al., 2011). Mapping of the mice, where the phenotype of interest is crossed to a common well-characterised inbred strain, and then backcrossed or intercrossed, should allow narrowing of the region to 20-60 Mb of the chromosome (Wiltshire et al., 2003). There are now at least 25 common inbred mouse strains to choose from, with the 6 best-characterised strains being C57BL/6J, 129X1/SvJ, DBA/2J, A/J, C3HHeJ and BALB/c, and selection depends on the background strain of the ENU mouse (Levitt et al., 2017; Wiltshire et al., 2003). It is already known that the penetrance of tumours in the Th-MYCN colony relies heavily on the background strain. Whilst 129/SVJ is the most penetrant strain, FVB was found to be the least penetrant, with C57BL/6 and BALB/c falling somewhere between these two (Weiss et al., 1997). Therefore to maximise the ability of being able to detect the mutation responsible, two well-described strains were chosen for the backcross, C57BL/6 and BALB/c, as acceptable neuroblastoma tumour penetrance had been shown with these strains (Henderson et al., 2011; Weiss et al., 1997). The two strategies employed, namely backcrossing and intercrossing allowed identification of the gene responsible for delayed tumour development in the 1590 and 1929 mouse lines.
Chapter 4- Elucidating the genes responsible for delayed tumour formation in ENU treated Th-MYCN mice 114 4.2 Results
4.2.1 Mouse 1590: Mapping and sequencing of progeny
4.2.1.1 Appearance and histology of 1590 tumour
Mouse 1590 developed a tumour at 49 weeks of age, which was found to be atypical by comparison with tumours normally developing in Th-MYCN mice. Not only was the tumour avascular (Figure 4-1, A), as opposed to the majority of mice in this colony where tumours are heavily vascularised (Figure 4-1, B), but it also had an irregular surface and looked almost mottled in appearance. Whilst large tumours in the Th-MYCN colony often embed the kidney, which leads to detachment of the adrenal gland from the kidney (as seen in Figure 4-1, B where the left kidney is attached to the tumour and the adrenal gland is seen embedded in the top of the tumour separated from the kidney), the 1590 tumour was not attached to the kidneys and was more ‘free-floating’, anchored loosely to the spine by fat and blood vessels (Figure 4-2). Histologic analysis demonstrated that whilst there were elements of neuroblastoma present in the tumour, as indicated by the presence of neuropil and ganglionic differentiation defining differentiating neuroblastoma (Figure 4-3, A), and the small round blue cells of poorly- differentiated neuroblastoma (Figure 4-3, B), there were also other elements present in the tumour. Squamous and glandular epithelium (Figure 4-3, C) and skeletal muscle and cartilage (Figure 4-3, D) were found alongside neuroblastic cells, indicating an atypical neuroblastoma tumour that possessed elements of a teratoma. Therefore, whilst 1590 did develop a tumour, it cannot be classed as a true neuroblastoma but rather as a teratoma.
4.2.1.2 Confirmation of MYCN genotype
To confirm that mouse 1590 and her progeny were indeed homozygous for the MYCN transgene, and that the phenotype observed was not simply due to the loss of one copy of the transgene during ENU mutagenesis and subsequent mating, fluorescence in situ hybridisation (FISH), using a probe that detects the transgene plasmid (pTh-MYCN) was
Chapter 4- Elucidating the genes responsible for delayed tumour formation in ENU treated Th-MYCN mice 115
Figure 4-1 Comparison of 1590 mouse tumour with a tumour from a normal homozygote Th-MYCN mouse. The avascular tumour of 1590 (A) compared to a vascular tumour in the Th-MYCN colony (B). Tumours are outlined for ease of visualisation. Arrow indicates heart and lungs for reference.
Chapter 4- Elucidating the genes responsible for delayed tumour formation in ENU treated Th-MYCN mice 116
Figure 4-2 Close-up of the 1590 tumour. The tumour is anchored by fat and a few blood vessels, indicated by the arrow.
Chapter 4- Elucidating the genes responsible for delayed tumour formation in ENU treated Th-MYCN mice 117
Figure 4-3 Histology of the 1590 tumour. Regions of differentiating neuroblastoma (A) and poorly-differentiated neuroblastoma (B), were found along with squamous and glandular epithelium (C) and skeletal muscle and cartilage (D), as indicated by arrows in each picture. Histological analysis was performed by pathologist Dr Andrew Gifford.
Chapter 4- Elucidating the genes responsible for delayed tumour formation in ENU treated Th-MYCN mice 118 carried out on blood from these mice. All of the mice born to 1590, as well as 1590 herself, possess two clear human MYCN signals in their blood cells, indicated by the arrows in Figure 4-4. For simplicity, only 1590 is shown here, however, all progeny, both those that developed tumours and those that did not, were scored, and all of them exhibited two MYCN copies. Therefore, the delayed tumour phenotype cannot be attributed to a loss of the MYCN transgene.
4.2.1.3 Exome sequencing of 1590 progeny
To elucidate the gene(s) responsible for the 1590 phenotype, exome sequencing was undertaken. Based on the assumption that the mutation generated by ENU would be heterozygous (ENU is more likely to mutate only one of the two alleles), and in order to increase the potential of identifying the mutant gene, two suppressed 1590 mice were mated together (as shown in Figure 4-5) to generate mice homozygous for the mutation. The mice used were also homozygous for the MYCN transgene (MYCN/MYCN) and each was assumed to possess one copy of the suppressor mutation (S/0, where S is the suppressor mutation and 0 the wild-type state). Following Mendelian inheritance, 25% of the progeny should develop tumours as they normally would (0/0) and 75% should display the suppressed phenotype. Of these, the majority (two thirds) should have one copy of the suppressor mutation (S/0) and the remainder, two copies (S/S). Eighteen mice were generated using this strategy. Seven of these developed tumours normally (39%) and one was selected for exome sequencing (Figure 4-6). Eleven of the progeny had a delayed tumour phenotype (61%), and two were selected for exome sequencing. The higher than expected percentage of mice developing tumours normally, suggested a decreased number of mice with two copies of the suppressor mutation, either due to pre- wean death (not noted due to cannibalisation of the progeny), or embryonic lethality.
DNA was extracted from the liver samples and sequenced using the NimbleGen sequence capture protocol (as outlined in Section 2.2.2.3). Since it was suspected the mutation would be heterozygous given the inheritance described above, the analysis therefore focused on this type of variant. Mouse 0025, with the unsuppressed phenotype, had 251 single nucleotide variants (SNVs) that were heterozygous (Figure 4-7). The two suppressed mice, 0016 and 0025, had 721 and 300 heterozygous variants,
Chapter 4- Elucidating the genes responsible for delayed tumour formation in ENU treated Th-MYCN mice 119
Figure 4-4 MYCN FISH of 1590 mouse blood. Two copies of the MYCN transgene are seen in two separate blood cells (indicated by arrows).
Chapter 4- Elucidating the genes responsible for delayed tumour formation in ENU treated Th-MYCN mice 120
1590 progeny with ‘suppressor’ 1590 progeny with ‘suppressor’ mutation mutation x MYCN/MYCN Transgene MYCN/MYCN Transgene S/0 Suppressor Mutation S/0 Suppressor Mutation
MYCN/MYCN MYCN/MYCN
(S) (0) MYCN/MYCN MYCN/MYCN MYCN/MYCN (S) (S/S) (S/0) MYCN/MYCN MYCN/MYCN MYCN/MYCN (0) (S/0) (0/0)
All mice are homozygous for the MYCN transgene: 25% should carry 2 copies of the suppressor mutation 50% should carry 1 copy of the suppressor mutation 25% should be unsuppressed
Figure 4-5 Mating diagram for Exome sequencing.
Chapter 4- Elucidating the genes responsible for delayed tumour formation in ENU treated Th-MYCN mice 121
Figure 4-6 Offspring from crossing together two 1590 suppressed mice. A portion of the mice developed tumours normally () and the remainder did not develop tumours (). Unsuppressed () and suppressed () mice selected for exome sequencing are shown.
Chapter 4- Elucidating the genes responsible for delayed tumour formation in ENU treated Th-MYCN mice 122
Figure 4-7 Single nucleotide changes in the 3 exomes that were sequenced. Of these, 66 variants were common to the 2 suppressed samples and not in the unsuppressed sample.
Chapter 4- Elucidating the genes responsible for delayed tumour formation in ENU treated Th-MYCN mice 123 respectively. Of these, 52 were common between the 3 mice and were clearly not responsible for the phenotype, while another 24 and 16 variants were shared between the unsuppressed mouse and one of the two suppressed mice. Only 66 variants were unique to the two suppressed mice. The remainder were found in only one of the suppressed samples.
The 66 variants were spread across 14 out of the 20 mouse chromosomes. A list of these genes is shown in Table 4-1, while Appendix 1 contains more detail about the base and amino acid (aa) changes of the variants and the function of the gene and phenotype of the mutation if known. The largest number of mutations was found on chromosome (Chr) 6, with 27 variants in 7 genes, the majority of which represented different members of a subfamily of genes responsible for sugar binding. However, most of the variations on the list have no known function, and in addition, none of the genes themselves has previously been shown to have a critical role in neuroblastoma tumorigenesis.
In order to further narrow down the list of candidates, whole genome sequencing was performed on the backcrossed mice.
Chapter 4- Elucidating the genes responsible for delayed tumour formation in ENU treated Th-MYCN mice 124 Table 4-1 List of single nucleotide variants that were restricted to the two mice with a suppressed tumour phenotype Number of Chr* Gene name variants Ribosome biogenesis regulator homolog (S. cerevisiae) (Rrs1) 3 1 Renin 1 structural (Ren1) 3 Runt-related transcription factor 1; translocated to, 1 1 (cyclin D-related) (Runx1t1) Interferon alpha 12 (Ifna12) 3 Predicted gene 13088 (Gm13088) 1 4 PRAME family member 6 (Pramef6) 1 Predicted gene 13109 (Gm13109) 2 Predicted gene 13128 (Gm13128) 1 5 Ecotropic viral integration site 5 (Evi5) 1 Solute carrier family 25 (mitochondrial carrier, adenine nucleotide 1 translocator), member 13 (Slc25a13) Killer cell lectin-like receptor, subfamily A, member 17 (Klra17) 7 Killer cell lectin-like receptor, subfamily A, member 5 (Klra5) 4 6 Killer cell lectin-like receptor, subfamily A, member 6 (Klra6) 6 Killer cell lectin-like receptor subfamily A, member 9 (Klra9) 3 Killer cell lectin-like receptor, subfamily A, member 7 (Klra7) 4 Killer cell lectin-like receptor, subfamily A, member 1 (Klra1) 2 Vomeronasal 2, receptor 28 (Vmn2r28) 1 7 Vomeronasal 1 receptor 151 (Vmn1r151) 1 HIRA interacting protein 3 (Hirip3) 1
Chapter 4- Elucidating the genes responsible for delayed tumour formation in ENU treated Th-MYCN mice 125 Table 4-1 continued Number of Chr* Gene name variants Purine-rich element binding protein G (Purg) 1 8 Seven in absentia 1A (Siah1a) 1 Glycine cleavage system protein H (aminomethyl carrier) (Gcsh) 1 Solute carrier family 25 (mitochondrial carrier, phosphate carrier), 10 1 member 3 (Slc25a3) 11 Family with sequence similarity 71, member B (Fam71b) 1 12 Serine/arginine-rich splicing factor 5 (Srsf5) 1 RIKEN cDNA 2410089E03 gene (2410089E03Rik) 1 15 TRIO and F-actin binding protein (Triobp) 1 Tubulin, alpha 1A (Tuba1a) 1 Keratin associated protein 16-2 (Krtap16-2) 1 16 Keratin associated protein 16-1 (Krtap16-1) 1 18 Predicted gene 4841 (Gm4841) 1 Integrator complex subunit 5 (Ints5) 4 Transient receptor potential cation channel, subfamily M, member 6 19 1 (Trpm6) Myoferlin (Myof) 1 X L antigen family, member 3 (Lage3) 2 * Chr = Chromosome number
Chapter 4- Elucidating the genes responsible for delayed tumour formation in ENU treated Th-MYCN mice 126 4.2.1.4 Backcrossing for whole genome sequencing
Two strains were chosen for the backcrossing, namely C57BL/6 and BALB/c, both of which have had their genomes fully sequenced. Within each strain two mating strategies were selected resulting in a total of four backcrossed lines. As a control for tumour development on these different backgrounds, normal, cyclophosphamide-treated homozygote Th-MYCN mice were mated in the same fashion as the suppressed 1590 mice, to create 4 additional lines.
Since 8 lines running in parallel were being generated, a simplified designation of 2-4 letters to identify each line, for cage identification and genotyping, became necessary. Therefore, the mutated 1590 line was given the designation ‘X’. Any lines that involved crossing of 1590 mice therefore began with X. For BALB/c and C57BL/6 mouse strains, it was decided to use coat colour to identify them. For any 1590 mice mated to BALB/c mice, the next letter given was a ‘W’, as BALB/c mice are white. Similarly, 1590 mice mated to C57BL/6, were designated ‘Bl’ for black. For those lines involving MYCN mice, they were given the letter ‘M’ to indicate MYCN mice of 129/SvJ background.
The first step in the strategy for all lines involved mating several 1590 suppressed mice to either a BALB/c or C57BL/6 mouse (Figure 4-8, 1st cross). All progeny from this pairing would be hemizygote for MYCN (MYCN/-) and half should carry one copy of the suppressor mutation (S), with the other half not carrying the mutation (0). Since it was not known which mice carried the suppressor mutation, because these mice are hemizygote and tumour penetrance is only ~25%, progeny were randomly selected for mating. The initial strategy was an intercross of the progeny generated from the first cross (Figure 4-8, 2nd cross). This mating strategy generated mice homozygous, hemizygous and wild-type for the MYCN transgene, however, only the homozygote mice were of interest. Of these mice, ~25% would be expected to develop tumours normally, and ~75% should have one or two copies of the suppressor mutation. Mice were followed until they developed a medium (10mm) tumour, or in the event of not developing a tumour, until there was a clear separation of the populations.
Chapter 4- Elucidating the genes responsible for delayed tumour formation in ENU treated Th-MYCN mice 127
1st Cross: 1590 suppressed mice crossed with BALB/c or C57BL/6
1590 progeny with ‘suppressor’ BALB/c or C57BL/6 mutation x MYCN/MYCN Transgene No Transgene (-/-) S/0 Suppressor Mutation No Suppressor Mutation (0/0)
MYCN/- (S/0) MYCN/- (0/0)
50% of hemizygous progeny carry the suppressor mutation
2nd Cross: Intercross
MYCN/- MYCN/-
(S) (0) MYCN/- MYCN/MYCN MYCN/MYCN (S) (S/S) (S/0) MYCN/- MYCN/MYCN MYCN/MYCN (0) (S/0) (0/0)
Only MYCN homozygote progeny were followed
Anticipated Outcome: 25% of the homozygous progeny will be unsuppressed 75% of the homozygous progeny will be suppressed (1 or 2 copies)
Figure 4-8 Mating diagram for 1590 crosses to BALB/c or C57BL/6.
Chapter 4- Elucidating the genes responsible for delayed tumour formation in ENU treated Th-MYCN mice 128 The first populations generated for comparison involved 1590 mice crossed to BALB/c (XW) with MYCN mice crossed to BALB/c (MW) (Figure 4-9). As anticipated, the spread of tumours in the MW population was wider than the normal Th-MYCN colony, with the majority of mice developing 10mm tumours by 90 days of age compared to 50- 55 days in the non-backcrossed colony. However, two mice in the MW line did not develop tumours, indicating that strain background had some effect on the tumour incidence. In contrast, 42% of the mice in the 1590 crossed to BALB/c mice (XW) colony did not develop tumours, signifying that the suppressor mutation was retained even on a different background. For the mice that did develop tumours in this line, the spread was similar to the MW line. Both populations were compared and there was a significant difference between the two lines (P<0.0001) (Table 4-2), indicating that the suppressed phenotype was not confined to the original strain alone and further suggesting that the suppressor mutation was active in the cross.
For the C57BL/6 intercross, the phenotype was less obvious due to the wider spread of tumour development (Figure 4-10). Mice from 1590 crossed to C57BL/6 (XBl) were compared to MYCN mice crossed to C57BL/6 (MBl), and whilst there were mice in each line that developed tumours, there was also a significant proportion that did not develop tumours in both lines. The number of mice developing tumours and not developing tumours in the XBl line was split almost evenly between the two populations. The frequency of no tumour development was much higher in the MBl line than the MW line, with almost a third of mice not developing tumours. Not only was the frequency different, but also the spread of tumour development was wider than the BALB/c lines. Mice in the MBl line developed tumours between 44 – 332 days of age and a similar spread was seen in the XBl line. Despite the increased number of mice not developing tumours in the MBl line, there was nevertheless still a significant difference between this line and the XBl line in terms of tumour development (P=0.0087) (Table 4-2). These results therefore suggested that the suppressor mutation was also active in this intercross although there was a significance influence from the C57BL/6 strain background.
Chapter 4- Elucidating the genes responsible for delayed tumour formation in ENU treated Th-MYCN mice 129
Figure 4-9 Tumour development in 1590 mice crossed to BALB/c mice. Comparison of tumour development of 1590 suppressed mice crossed to BALB/c (XW) with normal Th-MYCN mice crossed to BALB/c mice (MW). Mice with tumours () and those that didn’t develop tumours () are shown.
Chapter 4- Elucidating the genes responsible for delayed tumour formation in ENU treated Th-MYCN mice 130
Table 4-2 Comparisons between each 1590 and MYCN backcrossed lines.
Line Tumour (%) No tumour (%) Fisher’s Exact P value
XW 40 (58%) 29 (42%) P<0.0001 MW 96 (98%) 2 (2%)
XBl 42 (48%) 46 (52%) P=0.0087 MBl 58 (68%) 27 (32%)
XWM 97 (65%) 52 (35%) P<0.0001 MWM 107 (100%) 0 (0%)
XBlM 125 (87%) 19 (13%) P=0.0348 MBlM 109 (95%) 6 (5%)
Chapter 4- Elucidating the genes responsible for delayed tumour formation in ENU treated Th-MYCN mice 131
Figure 4-10 Tumour development of 1590 mice crossed to C57BL/6 mice. Comparison of tumour development of 1590 suppressed mice crossed to C57BL/6 mice (XBl) with normal Th-MYCN mice crossed to C57BL/6 (MBl). Mice with tumours () and those that didn’t develop tumours () are shown.
Chapter 4- Elucidating the genes responsible for delayed tumour formation in ENU treated Th-MYCN mice 132 The second mating strategy involved taking mice generated from the first cross above and instead of crossing them together, to backcross them to Th-MYCN mice (Figure 4-11, 2nd cross). This would allow generation of the suppressor mutation onto a well- described strain, but would also decrease the risk of losing the phenotype by increasing the amount of 129/SvJ in the mix. Therefore, mice from the first cross that were hemizygote for MYCN (MYCN/-), half carrying the suppressor mutation (S), half not (0), were mated with cyclophosphamide-treated homozygote MYCN mice (MYCN/MYCN). This mating strategy generated mice homozygous and hemizygous for the MYCN transgene. Only mice that were homozygous for the transgene were followed, half of which were expected to possess the suppressor mutation and half not. The mice were followed until they developed a medium palpable tumour or until clear separation was observed in the case of no tumour development.
The 1590 mice crossed to BALB/c and backcrossed to MYCN mice (XWM) were compared to MYCN mice crossed to BALB/c and backcrossed to MYCN (MWM). The first striking difference was the 100% tumour incidence in the MWM line, despite having a mixed background (Figure 4-12). Also noted was the tighter window of tumour development in the MWM line, with all mice developing tumours by 77 days of age in contrast to the MW line (Figure 4-9) which had tumour development up until 227 days. In the XWM line, 65% of mice developed tumours, and all but one developed a tumour between 39-116 days of age (Figure 4-12). The slightly less than anticipated number of mice not developing tumours (35% compared to a predicted 50%) was most likely due to the random chance of selecting more mice that lacked the suppressor mutation rather than background given the tight tumour formation in the MWM line. Nevertheless, there was a clear difference in the two populations in the XWM line and this was significantly different to the MWM line (P<0.0001) (Table 4-2).
The final mating strategy compared 1590 mice crossed to C57BL/6 and backcrossed to MYCN (XBlM) with MYCN mice crossed to C57BL/6 and backcrossed to MYCN (MBlM). The MBlM mice had a 95% tumour incidence, with a spread of between 37- 207 days (Figure 4-13). The increase in tumour incidence and the tighter window of tumour development in the MBlM mice confirmed that the C57BL/6 background was strongly influencing tumour development in the MBl mice. Whilst there was a clear difference between mice that developed tumours and those that didn’t in the XBlM line,
Chapter 4- Elucidating the genes responsible for delayed tumour formation in ENU treated Th-MYCN mice 133
1st Cross: 1590 suppressed mice crossed with BALB/c or C57BL/6
1590 progeny with ‘suppressor’ BALB/c or C57BL/6 mutation x MYCN/MYCN Transgene No Transgene (-/-) S/0 Suppressor Mutation No Suppressor Mutation (0/0)
MYCN/- (S/0) MYCN/- (0/0)
50% of hemizygous progeny carry the suppressor mutation
2nd Cross: Backcross to Th-MYCN homozygous mice
MYCN/- MYCN/-
(S) (0) MYCN/MYCN MYCN/MYCN MYCN/MYCN (0/0) (S/0) (0/0)
Only MYCN homozygote progeny were followed
Anticipated Outcome: 50% of the homozygous progeny will be unsuppressed 50% of the homozygous progeny will be suppressed
Figure 4-11 Mating diagram for crosses to BALB/c or C57BL/6 and backcross to Th- MYCN.
Chapter 4- Elucidating the genes responsible for delayed tumour formation in ENU treated Th-MYCN mice 134
Figure 4-12 Tumour development of 1590 mice crossed to BALB/c and backcrossed to Th-MYCN mice. Comparison of tumour development of 1590 suppressed mice crossed to BALB/c and then to Th-MYCN mice (XWM) with normal Th-MYCN mice crossed to BALB/c and then backcrossed to Th-MYCN mice (MWM). Mice with tumours () and those that didn’t develop tumours () are shown.
Chapter 4- Elucidating the genes responsible for delayed tumour formation in ENU treated Th-MYCN mice 135
Figure 4-13 Tumour development of 1590 mice crossed to C57BL/6 and backcrossed to Th-MYCN mice. Comparison of tumour development of 1590 suppressed mice crossed to C57BL/6 and then to Th-MYCN mice (XBlM) with normal Th-MYCN mice crossed to C57BL/6 and then backcrossed to Th-MYCN mice (MBlM). Mice with tumours () and those that didn’t develop tumours () are shown.
Chapter 4- Elucidating the genes responsible for delayed tumour formation in ENU treated Th-MYCN mice 136 there was a dramatic reduction in mice that did not develop tumours, with only 13% becoming long-term tumour free survivors. Considering the normal backcrossed mice had near complete tumour penetrance, the background of the mice cannot be the only cause of the smaller numbers of tumour-free mice. It is more likely that the difference was caused by an increase in the random selection of parents that did not carry the suppressor mutation. Regardless, there was still a significant difference between the MBlM and XBlM populations (p=0.0348) and considerable information has been garnered from them (Table 4-2).
Out of the four MYCN backcrossed lines generated, the tightest tumour phenotype was MWM and the worst penetrance was MBl. Since the clearest separation between the 1590 and MYCN lines occurred when either the BALB/c or C57BL/6 mice were subsequently backcrossed to MYCN mice, it was decided to perform whole genome sequencing on the XWM and XBlM mice. Livers from the XWM and XBlM mice that had already developed tumours, as well as tail biopsies from those that were still tumour-free were whole genome sequenced.
Across the entire genome, the region that showed differences in both the XWM and XBlM mice implicated the proximal end of Chr 4 (Table 4-3 and Table 4-4). The two tables show the variants detected in the samples. Whilst the variants detected in the samples were different between the two sequenced populations (XWM and XBlM), the region implicated remained the same. The mice have been ordered according to their time to tumour development, or if they had not developed a tumour, the age at which their tail biopsy was taken. As a comparison, the germline sequence for BALB/c and C57Bl/6 are shown at the bottom of their respective tables. For the variant to be responsible for the phenotype observed in 1590 line, it must not match the germline sequence for BALB/c (XWM) or C57BL/6 (XBlM). Regions where the germline sequence is different from the backcrossed strain, and therefore must have come from the Th-MYCN mice, are highlighted in blue.
To narrow down the region, further sequencing of more samples was undertaken, concentrating on the XWM line as this had the clearest separation of the phenotypes. All of the samples sequenced in the second round were from livers of either tumour
Chapter 4- Elucidating the genes responsible for delayed tumour formation in ENU treated Th-MYCN mice 137
Table 4-3 Sequencing of XWM samples and proximal end of chromosome 4. rs number* rs4224463 rs6177460 rs13477828 rs13477862 rs3670129
Exosome FERM Gene Ribosomal component domain Unknown Unknown name protein S11 3 containing 3
Time to medium tumour (days)† Chr 4 Chr 4 Chr 4 Chr 4 Chr 4 Chr position 45329692 73674272 87842756 97813178 114605412 XWM 0068 39 T C A A Fail XWM 0267 41 GT CA GA GA CT XWM 0063 43 T C A A T XWM 0065 45 T C A A T XWM 0012 46 T C A A Fail XWM 0058 46 GT C A A T XWM 0235 46 GT CA GA GA CT XWM 0088 48 GT CA GA GA CT XWM 0091 48 T C A A T XWM 0002 49 T CA GA GA Fail
XWM 0061 231 GT CA GA GA CT XWM 0064 231 T C A A T XWM 0066 231 T C A A T XWM 0042 238 T C A A CT XWM 0051 238 T C A A T XWM 0038 251 T C GA GA CT XWM 0026 252 T C A A T XWM 0027 252 T C A A T XWM 0019 253 T C A A T XWM 0009 259 T CA GA GA CT SVJ T C A A T Balb G A G G C The variants in the chromosome run from left to right. * rs number refers to the reference SNP cluster ID accession number † Mice are sorted according to their time to tumour development or age at which the tail biopsy was taken. Numbers below the grey line indicate time to tail biopsy rather than time to tumour development (days).
Chapter 4- Elucidating the genes responsible for delayed tumour formation in ENU treated Th-MYCN mice 138
Table 4-4 Sequencing of XBlM samples and proximal end of chromosome 4. rs number* rs13477541 rs13477543 rs3022975 rs13477553 rs32379175 Thymocyte selection- Syndecan Aspartate- Fucosyl- Gene associated Carbonic binding beta- transferase name high Anhydrase protein hydroxylase 9 mobility group box Time to Medium Tumour (days) Chr 4 Chr 4 Chr 4 Chr 4 Chr 4 Chr position 6320883 6800300 8073046 9390856 25691856 XBIM 0038 45 G T G C A XBIM 0044 45 GA CT GA CT GA XBIM 0008 46 G T G C A XBIM 0283 46 GA CT GA CT A XBIM 0056 49 GA CT GA CT GA XBIM 0031 50 GA CT GA CT GA XBIM 0025 51 G T G C A XBIM 0028 51 GA CT Fail CT GA XBIM 0196 51 GA CT GA CT GA XBIM 0201 51 GA T G C A
XBIM 0217 163 GA CT Fail CT GA XBIM 0175 183 G T G C A XBIM 0176 183 G T G C A XBIM 0180 183 G T G C A XBIM 0146 207 G T G C A XBIM 0147 207 G T G C A XBIM 0116 209 G T G CT GA XBIM 0098 224 G T G C A XBIM 0048 251 G T G C A XBIM 0053 251 GA CT GA CT G SVJ G T G C A C57 A C A T G The variants in the chromosome run from left to right. * rs number refers to the reference SNP cluster ID accession number † Mice are sorted according to their time to tumour development or age at which the tail biopsy was taken. Numbers below the grey line indicate time to tail biopsy rather than time to tumour development (days).
Chapter 4- Elucidating the genes responsible for delayed tumour formation in ENU treated Th-MYCN mice 139 bearing or tumour free mice as they had completed their monitoring period by this stage. The sequencing results again highlighted a region on the proximal end of Chr 4, but this time narrowed the window even further (Table 4-5). Across the top are the regions of Chr 4 and their position on the long arm of the chromosome, with numbering beginning from the centromere. The backcrossed mice are sorted according to their time to tumour development (days) or the age at which they were monitored for, if they did not develop a tumour. Regions of heterozygosity (H) are highlighted in purple. This heterozygosity indicates that the mice possess a mixed genotype arising from both their BALB/c and 129/SvJ (Th-MYCN) parents and is therefore unlikely to harbour the mutation responsible for the suppressed phenotype. Regions of homozygosity, where their inheritance comes from the 129/SvJ parent (129) are highlighted in blue and these regions are more likely to contain the variant responsible for the delayed tumour phenotype. Therefore, the region containing the variant responsible for the delayed tumour phenotype should be ‘H’ in the mice that develop tumours normally and ‘129’ in the mice that don’t develop tumours. From this second round of sequencing, the critical region was specifically narrowed down to the proximal end of Chr 4, in a 9.4- 43.07 Mb region of the chromosome (highlighted in yellow), as most mice that don’t develop tumours are ‘129’ in this region and the majority of those that do are ‘H’.
Comparing the whole genome sequencing to the list generated from the exome sequencing of the 1590 progeny, the only variant found in the exome sequencing that is in this region is located in the gene Runt-related transcription factor 1; translocated to, 1 (cyclin D-related), abbreviated as Runx1t1. To confirm this finding, suppressed and unsuppressed mice were specifically sequenced in this region. In the unsuppressed mice, at co-ordinate 13816819 of Chr 4, homozygosity of thymine (T) was observed. In the suppressed mice at the same co-ordinate, one allele was changed to a cytosine (C) residue. This single base change from TT to CT results in an aa change of a tyrosine (Y) to a histidine (H). To confirm the genotyping in suppressed and unsuppressed mice, a genotyping assay was designed for competitive binding of either the T or C allele. All mice that did develop tumours clustered together towards the Y axis (indicating a TT genotype), whilst all the mice that did not develop tumours clustered together in the middle of the plot (indicating a CT genotype; Figure 4-14).
Chapter 4- Elucidating the genes responsible for delayed tumour formation in ENU treated Th-MYCN mice 140 Table 4-5 Whole genome sequencing of the proximal end of chromosome 4 of backcrossed mice. Chr 4 4 4 4 4 4 4 * Mb 9.4§ 16.76§ 43.07§ 43.56 44.88 45.13 73.67
Sample Tumour (days) XWM 0121 39 129† H‡ H H H H H XWM 0157 45 H H H H H H H XWM 0078 48 H H H H H H H XWM 0080 48 H H 129 129 129 129 129 XWM 0166 50 H H H H H H H XWM 0176 50 H H 129 129 129 129 129 XWM 0177 50 H H H H H H 129 XWM 0047 52 H H H H H H H XWM 0048 52 129 H H H H H H XWM 0129 52 H H H H H H 129 XWM 0045 57 H H H H H H 129 XWM 0050 57 H H H H H H H XWM 0099 57 H H H H H H H XWM 0081 62 H H H H H H 129 XWM 0150 63 H H H H H H H XWM 0035 65 129 129 129 129 129 129 H XWM 0172 70 H H H H H H 129 XWM 0039 77 129 129 129 129 129 129 129 XWM 0097 85 H H H H H H H XWM 0036 91 129 129 129 129 129 129 129 XWM 0120 91 129 129 H H H H H XWM 0075 102 129 129 129 129 129 129 129 XWM 0180 111 129 129 129 129 129 129 129
XWM 0197 333 129 129 129 129 129 129 129 XWM 0199 333 129 129 129 129 129 129 H XWM 0200 333 129 129 129 129 129 129 129 XWM 0203 333 129 129 129 129 129 129 129 XWM 0206 333 129 129 129 129 129 129 129 XWM 0186 335 129 129 129 129 129 129 129 XWM 0158 341 129 129 129 129 129 129 129 XWM 0147 342 129 129 129 129 129 129 H XWM 0127 343 129 129 129 129 129 129 129 XWM 0128 343 129 129 129 129 129 129 129 XWM 0133 343 129 129 129 129 129 129 129 XWM 0116 353 129 H H H H * Megabase (Mb) refers to the distance from the end of Chromosome (Chr) 4 †129 refers to inheritance from the 129/SvJ parent ‡H indicates regions of heterozygosity § Refers to region implicated where inheritance in tumour delayed mice comes from the 129/SvJ parent #Mice are sorted according to their time to tumour development, or age at which they were euthanased without a tumour. Numbers below the grey line indicate tumour-free end point rather than time to tumour development (days).
Chapter 4- Elucidating the genes responsible for delayed tumour formation in ENU treated Th-MYCN mice 141
Figure 4-14 Genotyping of suppressed and unsuppressed 1590 mice. Mice homozygous for the wild-type (T) allele (blue) cluster together, while mice heterozygous (CT) cluster together (green). Mice homozygous for the C allele were not detected in these samples.
Chapter 4- Elucidating the genes responsible for delayed tumour formation in ENU treated Th-MYCN mice 142 RUNX1T1 is a known regulator of protein and DNA binding, acting as a nuclear co- repressor with non-DNA binding transcriptional activity. Disruption of this gene has been shown to result in lethality and growth retardation (Appendix 1). Runx1t1 was therefore identified as the lead candidate for the 1590 phenotype and further characterisation of this gene has been undertaken in Chapter 5.
4.2.1.5 Runx1t1 YH mutation
RUNX1T1 belongs to a family of genes of which there are at least three known members in human and mouse, collectively known as the myeloid translocation gene (MTG) or eight twenty one (ETO) family. The three family members are MTG8 (also known as RUNX1T1 and ETO), MTG16 (ETO2, CBFA2T3) and MTGR1 (CBFA2T2). These proteins are highly conserved between members and species, with 99% homology observed between human and mouse RUNX1T1 (Davis et al., 2003). ETO genes contain four nervy homology regions (NHR), NHR1-4, that are highly conserved and are named after the Drosophila melanogaster homolog of RUNX1T1, called nervy (Wildonger and Mann, 2005). NHR1 and 2 contain protein-interaction motifs while NHR4, also known as the MYND domain (Myeloid, Nervy, and DEAF-1 (deformed epidermal autoregulatory factor 1)) contains a zinc finger motif, and also mediates protein interactions (Davis et al., 2003). Although NHR3 was thought to have no significant role, it appears that NHR4 cannot exert its function without NHR3 (Hildebrand et al., 2001). The 1590 Runx1t1 mutation, of a tyrosine (Y) to a histidine (H), occurs in the highly conserved zinc finger domain, controlling protein interactions (Figure 4-15 and Figure 4-16, A). It is hypothesized that a mutation in this region would ultimately disrupt the function of the protein.
Using computational modelling of the mutation, the effect of the YH 1590 point mutation in the MYND zinc-finger domain was analysed using protein modelling software (see Materials and Methods 2.2.1.7). The wild-type MYND zinc-finger protein is depicted in Figure 4-16, B. The domain, containing two Zinc (2+) atoms (grey circles) is co-ordinated by three cysteine and one histidine residue (top left) and four cysteine residues (bottom right). The light green element at the rear is a peptide from an
Chapter 4- Elucidating the genes responsible for delayed tumour formation in ENU treated Th-MYCN mice 143
Figure 4-15 Nervy homology regions (NHR) of the myeloid translocation gene (MTG) family. The homology between the family members, are shown. The 1590 mutation of an amino acid change from tyrosine (Y) to histidine (H) occurs in the highly conserved zinc finger motif or NHR4. Adapted from Davis et al., 2003.
Chapter 4- Elucidating the genes responsible for delayed tumour formation in ENU treated Th-MYCN mice 144
Figure 4-16 Computational modelling of the RUNX1T1 protein structure. The NHR4 domain is highly conserved across species (A). Wild-type RUNX1T1 (B) was compared to the 1590 mutation of a tyrosine (pale blue) to histidine (mid blue) (C) which changes the confirmation of the protein. The Zinc atoms (grey circles) are surrounded by cysteine (yellow tip) and histidine (blue tip) residues. A peptide from an interacting protein (green) is at the rear.
Chapter 4- Elucidating the genes responsible for delayed tumour formation in ENU treated Th-MYCN mice 145 interacting protein. In the wild-type conformation, the tyrosine at position 682, is buried in the core of the domain and would likely be integral to the structure of this protein. Whilst the domain containing the histidine mutant (Figure 4-16, C) is similar in size, this change is likely to alter nearby zinc co-ordination centres that are integral to the structure. The mutant histidine is no longer buried in the core and would be orientated towards the zinc residue, and depending on the environment, could be charged, which would be highly disruptive. The mutant histidine can act as a ligand to the zinc, which would displace an already bound ligand, further disrupting the structure. These changes are adjacent to the interaction surface of the protein, and this would be expected to disrupt interactions with other proteins. This point mutation is therefore likely to profoundly alter the structure of the MYND zinc-finger domain and impair or abolish its function.
To test the function of the MYND zinc finger domain, this region was cloned into pGEX-6P (as described in Material and Methods 2.2.2.10) and the Y to H mutation generated using a mutagenesis kit. Nuclear magnetic resonance (NMR) spectroscopy was used to determine the structure of the wild-type and mutant domains. The wild-type domain demonstrates good dispersion and sharp signals which are indicative of a well- folded domain (Figure 4-17, top). In contrast, the mutant domain is broad and unattractive and is not in a well-defined conformation, and may be oscillating between many partially ordered states (Figure 4-17, bottom). This result provides further evidence that the mutation present in the 1590 mouse line will lead to a loss of function of the wild-type MYND zinc finger domain.
Overall, it is therefore highly likely that the mutation of tyrosine to histidine would be extremely disruptive to the RUNX1T1 protein and alter its ability to interact with other proteins.
Chapter 4- Elucidating the genes responsible for delayed tumour formation in ENU treated Th-MYCN mice 146
Figure 4-17 One-dimensional 1H NMR spectra of the RUNX1T1 MYND domain wild- type and YH point mutant. The wild-type spectra (top) has sharp peaks and good dispersion. The YH mutant (bottom) is flat and not well-defined.
Chapter 4- Elucidating the genes responsible for delayed tumour formation in ENU treated Th-MYCN mice 147
4.2.2 Mouse 1929: Mapping and sequencing of progeny
4.2.2.1 1929 Exome sequencing
To elucidate the gene(s) responsible for the 1929 phenotype, a similar strategy was used as described above, with exome sequencing being the first step undertaken. However, since the 1929 phenotype already had two populations within the suppressed mice, it was decided to mate a suppressed tumour-free mouse to a cyclophosphamide-treated homozygote MYCN mouse from the colony, followed by sequencing of the mice that exhibited complete tumour abrogation (Figure 4-18). All the mice were homozygous for the MYCN transgene (MYCN/MYCN) and it was assumed that the 1929 progeny carried one copy of the suppressor mutation (S/0). Based on Mendelian inheritance, it would be expected that half of the progeny would develop tumours as normal (0/0) and half would be suppressed, either developing a delayed tumour or being tumour free.
Two suppressed 1929 mice were mated to two normal homozygote MYCN mice and 46 offspring were generated from these pairings, of which, 25 (54%) developed 10mm tumours in 50 days or less (Figure 4-19). The remaining 21 mice were split into two groups. Sixteen mice (35%) developed tumours ranging from 56-169 days of age, whilst the remaining 5 (11%) became long-term tumour free survivors. To increase the chances of finding the mutation responsible, two mice were selected that had reached 344 and 366 days of age respectively, without any signs of developing tumours. For comparison purposes and to increase confidence in the sequencing results, two unsuppressed mice with early onset tumours at 47 and 50 days of age were also selected for sequencing.
The 4 samples were exome sequenced and compared as described in Materials and Methods (Section 2.2.2.3). Across the 4 samples the number of heterozygous mutations was smaller than in the 1590 samples, therefore it was decided to analyse not only the heterozygous mutations, but also the homozygous mutations present. Each sample had around 250 mutations, regardless of whether they developed a tumour or not, 156 of
Chapter 4- Elucidating the genes responsible for delayed tumour formation in ENU treated Th-MYCN mice 148
1929 progeny with ‘suppressor’ 1929 progeny with ‘suppressor’ mutation mutation x MYCN/MYCN Transgene MYCN/MYCN Transgene S/0 Suppressor Mutation S/0 Suppressor Mutation
MYCN/MYCN MYCN/MYCN
(0) (0) MYCN/MYCN MYCN/MYCN MYCN/MYCN (S) (S/0) (S/0) MYCN/MYCN MYCN/MYCN MYCN/MYCN (0) (0/0) (0/0)
All mice were homozygous for the MYCN transgene: 50% should carry 1 copy of the suppressor mutation 50% should be unsuppressed
Figure 4-18 Mating diagram for exome sequencing of 1929 progeny
Chapter 4- Elucidating the genes responsible for delayed tumour formation in ENU treated Th-MYCN mice 149
Figure 4-19 Offspring from crossing a 1929 suppressed mouse with a homozygote Th- MYCN mouse. Approximately half of the mice developed tumours normally and half displayed a delayed tumour phenotype () or did not develop tumours (). Unsuppressed () and suppressed () mice selected for exome sequencing are shown.
Chapter 4- Elucidating the genes responsible for delayed tumour formation in ENU treated Th-MYCN mice 150 which were common to all 4 samples (Figure 4-20). Fifteen were shared between the unsuppressed samples (0154 and 0156) and were clearly not responsible for the phenotype. Between 18-37 mutations were unique to only one of the samples, while a number were common between one of the suppressed and unsuppressed samples. Thirteen variants were common only to the two suppressed samples (0116 and 0126), which was much lower than the list generated for the 1590 line ( Table 4-6 and Appendix 2). Seven mutations were heterozygous, 4 were homozygous and the remaining 2 were designated ‘other’ as the software could detect a difference, but the peaks were uneven and could not be called as either homozygous or heterozygous. The 13 variants were spread across 7 of the mouse chromosomes and 8 of them had no known phenotype. One gene variant that would be unlikely to be responsible for the delayed tumour phenotype was in the Neurofibromatosis 1 (Nf1) gene, since even though this gene is known to be involved in sympathetic nervous system and adrenal gland development, this was a homozygous variant and it has been reported that homozygous mutations in this gene result in death at E14.5 (Brannan et al., 1994). Given that little is known about the majority of the other genes implicated in the delayed tumour phenotype, backcrossing to different mouse strains was undertaken.
4.2.2.1 1929 Backcrossing for confirmatory sequencing
The 1590 backcrossing strategies used above, demonstrated that the suppressed phenotype was tighter and the penetrance more complete when the mice were backcrossed to MYCN mice after the BALB/c or C57BL/6 crosses had occurred. This strategy (outlined previously in Figure 4-11) was therefore chosen for the 1929 line, since the phenotype in 1929 mice was less well-defined compared to that for the 1590 line. The following simplified designation of 3-4 letters was used to identify the 2 new lines: mice that came from the 1929 line were given the designation ‘V’; 1929 mice crossed to BALB/c and then to MYCN mice were called VWM, and those to C57BL/6, VBlM. The mice were compared to the MWM and MBlM (normal MYCN lines crossed to BALB/c or C57BL/6 and backcrossed to MYCN) mice generated previously for the 1590 crosses.
Chapter 4- Elucidating the genes responsible for delayed tumour formation in ENU treated Th-MYCN mice 151
Figure 4-20 Single nucleotide changes in the 4 exome sequencing samples from 1929 mice. Only 13 variants (circled) were common only to the 2 suppressed samples and not in the 2 unsuppressed samples.
Chapter 4- Elucidating the genes responsible for delayed tumour formation in ENU treated Th-MYCN mice 152
Table 4-6 List of variants in genes detected in suppressed samples from 1929 mice that were not detected in the unsuppressed mice.
Chr Gene Name Type of mutation 2 Tubulin, beta 1 class VI (Tubb1) HOMOZYGOUS 4 Ephrin receptor A8 (Epha8) OTHER* Trypsin 10 (Try10) HETEROZYGOUS 6 Oxysterol binding protein-like 3 (Osbpl3) HETEROZYGOUS Chaperonin containing Tcp1, subunit 7 (Cct7) HETEROZYGOUS Ring finger protein 121 (Rnf121) HETEROZYGOUS G protein-coupled receptor, family C, group 5, member B 7 HETEROZYGOUS (Gprc5b) A disintegrin and metallopeptidase domain 12 (Adam12) HETEROZYGOUS 8 Glutathione reductase (Gsr) HOMOZYGOUS Transmembrane protein 199 (Tmem199) HETEROZYGOUS 11 Neurofibromatosis 1 (Nf1) HOMOZYGOUS Potassium channel tetramerisation domain containing 2 HOMOZYGOUS (Kctd2) 16 Interphotoreceptor matrix proteoglycan 2 (Impg2) OTHER* * OTHER refers to the software detecting a difference but it could not be called as either homozygous or heterozygous
Chapter 4- Elucidating the genes responsible for delayed tumour formation in ENU treated Th-MYCN mice 153 The 1929 mice crossed to BALB/c and backcrossed to MYCN (VWM) were compared to the 100% tumour incidence of the MWM line (Figure 4-21). Despite having a more ambiguous phenotype, the 1929 mice showed a clear separation between mice that developed tumours and those that didn’t. Surprisingly, the crossing to BALB/c appeared to enrich for the no tumour population, with 22% of mice not developing tumours in the backcrossed line compared to 11% in the exome sequencing population (Table 4-7). This could simply be due to the selection for mating of more mice carrying the suppressor mutation used in the second cross, however, it was encouraging to know that the phenotype was not lost or diluted in the process. Whilst the spread of the mice that did develop tumours was wider than the MWM mice, all but one had developed a tumour by 177 days, which was similar to the results obtained in the exome sequencing for this line. There was a clear and significant difference between the VWM and MWM lines (P<0.0001) and the 1929 phenotype was retained (Table 4-7).
The final backcrossing was the 1929 mice crossed to C57BL/6 and backcrossed to MYCN (VBlM) and its comparison to MYCN mice crossed to C57BL/6 and backcrossed to MYCN (MBlM). The spread of tumour development in the VBlM line was remarkably similar to the VWM line, where 76% of mice developed tumours up until 233 days of age (Figure 4-22 and Table 4-7). This was also comparable to the spread of tumours in the VBlM line. There was a similar enrichment of the no tumour population in the VBlM mice (24%), however, the phenotype cannot be attributed solely to the 1929 mutation given the incomplete penetrance in the MBlM line, although the two populations were still significantly different from one another (P<0.0001) (Table 4-7).
As the list of genes suspected to be responsible for the 1929 phenotype was very small, the decision was made to forego whole genome sequencing and instead design PCR assays specific to the 13 listed variants, and sequence the products looking for the mutation detected by exome sequencing. Since the tumour penetrance was 100% in the MWM line and the chance that the no tumour phenotype in the VWM line is less likely to be due to strain background, the VWM mice were selected for these experiments. Interestingly, 11 of the 13 variants were not detected in 6 suppressed samples that were sequenced indicating that backcrossing had removed these variants from the gene pool (Table 4-8). Variants were only found in G protein-coupled receptor, family C, group 5,
Chapter 4- Elucidating the genes responsible for delayed tumour formation in ENU treated Th-MYCN mice 154
Figure 4-21 Tumour development of 1929 mice crossed to C57BL/6 and backcrossed to Th-MYCN mice. Comparison of tumour development of 1929 suppressed mice crossed to C57BL/6 and then to Th-MYCN mice (VBlM) with normal Th-MYCN mice crossed to C57BL/6 and then backcrossed to Th-MYCN (MBlM). Mice with tumours () and those that did not develop tumours () are shown.
Chapter 4- Elucidating the genes responsible for delayed tumour formation in ENU treated Th-MYCN mice 155
Table 4-7 Comparisons between each 1929 and MYCN backcrossed lines.
Line Tumour (%) No Tumour (%) Fisher’s Exact P value
VWM 93 (78%) 26 (22%) P<0.0001 MWM 107 (100%) 0 (0%)
VBlM 124 (76%) 39 (24%) P<0.0001 MBlM 109 (95%) 6 (5%)
Chapter 4- Elucidating the genes responsible for delayed tumour formation in ENU treated Th-MYCN mice 156
Figure 4-22 Tumour development of 1929 mice crossed to C57BL/6 and backcrossed to Th-MYCN. Comparison of tumour development of 1929 suppressed mice crossed to C57BL/6 and then to Th-MYCN mice (VBlM) with normal Th-MYCN mice crossed to C57BL/6 and then backcrossed to Th-MYCN (MBlM). Mice with tumours () and those that don’t develop tumours () are shown.
Chapter 4- Elucidating the genes responsible for delayed tumour formation in ENU treated Th-MYCN mice 157
Table 4-8 Number of variants found in backcrossed mice for the 13 genes selected from exome sequencing.
Number of suppressed Gene Name backcrossed samples with variant Tubulin, beta 1 class VI (Tubb1) 0/6 Ephrin receptor A8 (Epha8) 0/6 Trypsin 10 (Try10) 0/6 Oxysterol binding protein-like 3 (Osbpl3) 0/6 Chaperonin containing Tcp1, subunit 7 (Cct7) 0/6 Ring finger protein 121 (Rnf121) 14/14 G protein-coupled receptor, family C, group 5, 9/13 member B (Gprc5b) A disintegrin and metallopeptidase domain 12 0/6 (Adam12) Glutathione reductase (Gsr) 0/6 Transmembrane protein 199 (Tmem199) 0/6 Neurofibromatosis 1 (Nf1) 0/6 Potassium channel tetramerisation domain 0/6 containing 2 (Kctd2) Interphotoreceptor matrix proteoglycan 2 (Impg2) 0/6
Chapter 4- Elucidating the genes responsible for delayed tumour formation in ENU treated Th-MYCN mice 158 member B (Gprc5b) in 5/6 samples sequenced, and in RING finger protein 121 (Rnf121) in 6/6 samples sequenced. Since there was only one sample that lacked the Gprc5b variant out of 6 sequenced, a further 8 suppressed samples were analysed for these two genes. All additional 8 mouse samples possessed the Rnf121 variation, taking the total to 14/14, whilst only 4/7 samples had the Gprc5b mutation, making the total 9/13, with the fourteenth sample failing sequencing for this gene (Table 4-8).
The mutation in Rnf121 occurs at co-ordinate 102031527 on Chr 7, where an adenine (A) changes to a cytosine (C) residue on one allele and suppressed mice have both A and C peaks at this position (Figure 4-23). This base change results in an aa change of methionine (M) to arginine (R). Rnf121 encodes a protein containing a RING finger, which is a relatively common motif that has roles in protein-protein and protein-DNA interactions (Rosa-Rosa et al., 2009). RNF121 was first identified in the Mammalian Gene Collection Program as a cDNA clone in a screen of over 15,000 mouse and human genes (Strausberg et al., 2002) and was subsequently characterised in Caenorhabditis elegans and found to be highly conserved across species (Darom et al., 2010; Maghsoudlou et al., 2016). The RNF121 protein has six putative transmembrane domains, and a RING domain (Maghsoudlou et al., 2016) and the 1929 mutation lies in the fifth transmembrane domain (Figure 4-24). However, little is known of the biological function of RNF121(Zhao et al., 2014b), and there is no reported phenotype known for this particular mutation, although it is predicted to be probably damaging to the protein (Appendix 2). Further investigation into this mutation, as a gene that predominately slows progression of this highly aggressive model, as well as complete abrogation of the tumour in a subset of mice, will yield new information about its role in neuroblastoma biology.
Chapter 4- Elucidating the genes responsible for delayed tumour formation in ENU treated Th-MYCN mice 159
Figure 4-23 Sequence of Rnf121 mutation. Chromatograph showing heterozygosity of A/C at co-ordinate 102031527 of Rnf121.
Chapter 4- Elucidating the genes responsible for delayed tumour formation in ENU treated Th-MYCN mice 160
Figure 4-24 Transmembrane domains of RNF121. The 1929 mutation of an amino acid change from methionine (M) to an arginine (R) occurs in the fifth transmembrane domain. N and C refer to the N- and C-terminus, respectively. Adapted from Maghsoudlou et al., 2016.
Chapter 4- Elucidating the genes responsible for delayed tumour formation in ENU treated Th-MYCN mice 161
4.3 Discussion
Using a combination of strategies involving exome sequencing and backcrossing to different mouse strains, two putative genes, Runx1t1 and Rnf121, were identified as lead candidates responsible for suppressed neuroblastoma phenotypes.
In particular, the backcrossing experiments demonstrated that the selection of the mouse strain to cross the suppressed mice to was a key element to the success of this strategy. In the original paper, reporting the generation of the Th-MYCN mice, several different strains of mice with mixed tumour penetrance were described (Weiss et al., 1997). These authors demonstrated that crossing Th-MYCN mice to FVB mouse strains led to a poor tumour penetrance. However, they also showed that the mixed C57BL/6 and BALB/c strains resulted in 2 out of 3 lines developing thoracic paraspinal tumours, with all lines overexpressing MYCN in the adrenal gland. In contrast, lines with C57BL/6 background only resulted in one tumour line out of three lines established, however, the founder of this line was sterile and developed cerebral neuroblastoma (Weiss et al., 1997).
In the current study, crossing the Th-MYCN mice to the C57BL/6 strain, resulted in a much wider tumour latency and decreased tumour penetrance compared to the pure bred Th-MYCN population. This was not surprising given that a more recent report showed that a mixed 129/SvJ and C57BL/6 background resulted in a decrease in tumour penetrance in hemizygotes and an increase in tumour latency in homozygotes (Henderson et al., 2011). However, what was unexpected was the 100% and 98% tumour penetrance in the mixed 129/SvJ and BALB/c mice. One of the original crosses used to create the Th-MYCN mice line used a mixed C57BL/6 and BALB/c background, as mentioned above, which demonstrated better tumour penetrance than just C57BL/6 alone, however, if BALB/c was used on its own, this was never reported (Weiss et al., 1997). Regardless, of the strain used or of the strategy employed, there were significantly more mice that did not develop tumours in the four 1590 and two 1929 lines generated, compared to the four normal MYCN backcrossed lines.
Chapter 4- Elucidating the genes responsible for delayed tumour formation in ENU treated Th-MYCN mice 162 Although the 1590 line demonstrated almost complete abrogation of tumour formation in mice possessing the Runx1t1 mutation, the original 1590 mouse that gave rise to this line actually developed a tumour. However as described in this chapter, this tumour was not a typical neuroblastoma, but rather a mix of neuroblastic elements with those resembling a teratoma. The Th-MYCN mouse line, which was on a 129/SvJ strain background, was imported into Australia from the Weiss laboratory in the late 1990s (Weiss et al., 1997). Since the penetrance of the phenotype is more complete when using fresh 129/SvJ mice in the mating process, this strategy was continued following their importation. However, the 129 line used for mating in Australia was a sub-strain known as 129/terSv (now designated 129/Sv-+p Tyrc-ch Ter/+). Despite having a common founder with the 129 line dating back to 1928, there are now at least 17 sub- strains that have been created, with a complex history (Simpson et al., 1997). The substrains are divided into 3 main groups and can be distinguished using coat colour: Parental (chinchilla or albino), Steel (pigmented white-bellied agouti) and Ter (lightened agouti dorsal with a cream-coloured belly) (Simpson et al., 1997). Importantly, the 129/terSv sub-strain used for mating in Australia has a history of developing testicular teratomas (Stevens, 1973). Though testicular teratomas have not been noted previously in this colony, the penetrance is quite low and it is something that is monitored as it can affect breeding (Stevens, 1973). Interestingly, the teratomas forming in the 129/terSv sub-strain consisted mostly of nervous tissue, which included ganglia, and whilst 1590 could not have developed a testicular teratoma as it was female, this is perhaps why this mouse developed a teratoma at the site where neuroblastoma usually occurs. Also of note is the location of the genes implicated in the Ter phenotype, which are found on mouse Chr 18 (18qB1-B2), upstream of the random insertion site for the MYCN transgene at 18qE4 (Asada et al., 1994; Haraguchi and Nakagawara, 2009; Youngren et al., 2005), although the insertion of the transgene occurred before the cross to the Ter substrain.
ENU mutagenesis has been shown to preferentially mutate adenine and thymine over guanine and cytosine (87% versus 13%) (Justice et al., 1999). Interestingly however, in the list of mutations in the current study, there was a lower percentage of adenine/thymine mutations with only 40% occurring in the 1590 line and 61% of the base changes in the 1929 line. In a meta-analysis of ENU mutagenesis screens, a similar percentage of guanine/cytosine mutations were noted compared to this study. It was
Chapter 4- Elucidating the genes responsible for delayed tumour formation in ENU treated Th-MYCN mice 163 hypothesized that this difference may be due to several possibilities including the bias for detecting mutations in genes with longer coding sequences, and therefore higher percentage guanine/cytosine content or that mutations in guanine/cytosine are more likely to result in phenotypic changes in the mouse over adenine/thymine mutations, although there is no evidence for the latter statement (Barbaric et al., 2007). The base changes caused by ENU generated a variety of changes at the protein level, consisting of missense mutations with a change in the aa (79% and 77%, respectively), splice variants (18% and 15%) and nonsense mutations that generated a premature stop codon (3% and 8%). These results were similar to reported findings with 64% missense, 26% splice, and 10% nonsense mutations (Justice et al., 1999). Splicing and nonsense mutations result in frame shift, deletion or premature termination of the protein and often lead to a change in the phenotype of the mouse (Barbaric et al., 2007). As a result, ENU screens are more likely to generate missense mutations.
RNF121 has not previously been reported to have a role in neuroblastoma tumorigenesis, and despite belonging to a relatively large family of proteins known as RING (Really Interesting New Gene) finger proteins, very few papers have been specifically published on this particular member. In C. elegans, RNF121 is described as an E3 ligase RING finger protein with 6 transmembrane domains, that is expressed in the endoplasmic reticulum (Darom et al., 2010). In humans the protein is expressed in the Golgi apparatus, and is highly conserved between species including human, mouse, rat and monkey (Zhao et al., 2014b). RNF121 has also been shown to regulate NF-κB activity (Zemirli et al., 2014) and ring finger proteins in general have been implicated as both oncogenes and tumour suppressors (Zhao et al., 2014b). The chromosomal region in which RNF121 lies has been implicated in breast cancer susceptibility (Rosa-Rosa et al., 2009) and high gene and protein expression has been observed in oesophageal cancer (Wang et al., 2014). Despite these findings, further investigation of RNF121 is beyond the scope of this thesis for a number of reasons. Firstly, Runx1t1 had been identified earlier in the mutagenesis screening process and there had also been some delay in elucidating the gene responsible for the 1929 phenotype. Secondly, the tumour phenotype observed was stronger in the 1590 line, where there is complete abrogation of tumours in half of the offspring, whilst the major phenotype in the 1929 line is a delay in tumour development, with a small portion becoming long term tumour-free survivors. Thirdly, and perhaps most importantly, a knock-out mouse model was
Chapter 4- Elucidating the genes responsible for delayed tumour formation in ENU treated Th-MYCN mice 164 available for Runx1t1, unlike Rnf121, where considerable time would be needed to generate such a model. For these reasons it was decided not to pursue RNF121 further but rather to concentrate on elucidating the role of RUNX1T1 in neuroblastoma tumorigenesis.
The 1590 Runx1t1 mutation located in NHR4, is one of four nervy homology regions that are believed to be essential to the function of this nuclear co-repressor. NHR1 was shown to have homology with a number of TATA-binding protein-associated factors, providing some of the first evidence that RUNX1T1 functions as a transcription factor (Feinstein et al., 1995). In deletion mutant studies by Hildebrand and colleagues, removal of NHR1 did not change the repression activity of RUNX1T1, however, deletion of NHR1 and the region between NHR1 and NHR2 significantly impaired the function, which was further disrupted by deletion of NHR2 (Hildebrand et al., 2001). Interestingly, deletion of the two zinc finger domains in NHR4 does not abolish the RNA-binding ability of RUNX1T1, as a second binding site proximal to NHR2 was identified (Rossetti et al., 2008). However, NHR4 interacts with a range of co- repressors including SIN3A (SIN3 transcription regulator family member A), N- CoR2/SMRT (Nuclear receptor co-repressor 2/silencing mediator for retinoid or thyroid-hormone receptors), as well as histone deacetylases and DNA-binding transcription factors (Rossetti et al., 2004). SIN3A appears to have an important role in repressing MYC responsive genes via interactions with MXI1 and also MAD-MAX heterodimers, while at the same time antagonizing MYC oncogenic activities (Farhana et al., 2015; Rao et al., 1996). In addition, NHR4 has been shown to interact with serine/arginine-related protein (SON), a DNA and RNA binding transcription factor required for cell cycle progression (Lam and Zhang, 2012). Given these findings, it is not surprising that RUNX1T1 has been shown to have a role in cancer, although this role has been most well characterized in acute myeloid leukaemia (AML).
The frequently observed t(8;21) translocation in AML produces a chimeric RUNX1- RUNX1T1 (previously known as AML-ETO) fusion protein acting in a nuclear corepressor/histone deacetylase complex that is able to block haematopoietic differentiation and contributes to leukaemogenesis (Lam and Zhang, 2012). The t(8;21) translocation occurs in 12% of AML patients and at 30-50% in a subclass of AML known as M2, first described over 40 years ago (Davis et al., 2003; Lam and Zhang,
Chapter 4- Elucidating the genes responsible for delayed tumour formation in ENU treated Th-MYCN mice 165 2012; Rossetti et al., 2004). RUNX1T1 remains largely intact during this fusion, only breaking at intron 1a or 1b, while the RUNX1 gene breaks at intron 5 and the fusion protein created includes almost all of the RUNX1T1 protein (Lam and Zhang, 2012). This fusion protein cannot cause leukaemia by itself, however, two mutated forms of the fusion protein are capable of causing rapid onset leukaemia in mice. RUNX1-ETO9a, where there is an additional exon at 9a of RUNX1T1 as well as a C-terminal truncated form of the fusion protein, are both capable of causing leukaemia by themselves (Lam and Zhang, 2012). Interestingly, both of these variants lack NHR3 and NHR4, and loss of the NHR4 domain in this region destroys the fusion protein’s ability to bind to co- repressors which accelerates leukaemogenesis (Lam and Zhang, 2012). Importantly, introduction of a point mutation in the zinc finger domain of NHR4 disrupted the ability of NHR4 to interact with SON, resulting in leukaemia induction (Ahn et al., 1998), while another mutation (W692A) near the 1590 mutation, showed that loss of N-CoR2 binding, rather than loss of SON interaction, represents the critical factor in regulating the leukaemogenic potential of t(8:21) AML (DeKelver et al., 2013). Together, these results highlight the importance of the NHR4 domain in mediating the transcriptional repression ability of this protein.
RUNX1T1 is a complex gene with at least 15 human transcripts listed on NCBI that result in 6 isoforms (https://www.ncbi.nlm.nih.gov/gene/862). The majority of the changes occur at the 5’ untranslated region (UTR) with alternate exons and often longer N-termini. In addition to these listed variants, three additional variants have been reported, namely; 031, short and 9a. Transcript 031 is listed on the Ensembl website (http://www.ensembl.org/Homo_sapiens/Transcript/Summary?db=core;g=ENSG00000 079102;r=8:91954967-92103226;t=ENST00000518361) as containing only 4 exons and results in a 10kDa protein. The short isoform was originally discovered in the context of adipogenesis, where the Fat Mass and Obesity-associated protein (FTO) controls alternative splicing of RUNX1T1. During adipogenesis, there are essentially two isoforms of RUNX1T1, the long, which is 496bp and produces a protein of 68kDa and the short, which involves skipping of exon 6 and is 245bp and 27kDa in size (Zhao et al., 2014a). The short isoform appears to promote differentiation of adipocytes. As discussed above, the third variant, 9a, has been specifically associated with AML (Lam and Zhang, 2012).
Chapter 4- Elucidating the genes responsible for delayed tumour formation in ENU treated Th-MYCN mice 166 The role of RUNX1T1 as either an oncogene or tumour suppressor appears to be dependent on tissue context. In AML, as described above, it is a driver of leukaemogenesis, although the fusion protein is not capable of causing leukaemia alone and a second event is required (Lam and Zhang, 2012). Expression of RUNX1T1 is decreased in metastatic primary pancreatic cancer and low levels can predict metastatic lesions to the liver (Nasir et al., 2011). Mutations in the gene have been implicated in colorectal, breast and lung cancers in large-scale studies (Kan et al., 2010; Sjöblom et al., 2006; Wood et al., 2007). In ovarian cancer, RUNX1T1 is a putative tumour suppressor where expression of RUNX1T1 inhibited ovarian cancer cell line colonies (Yeh et al., 2011).
Neuroblastoma has its origins during embryogenesis and it is therefore not unexpected that Runx1t1 is also known to have a key role early in development. In Drosophila studies by Wildonger and Mann, nervy, the homolog of RUNX1T1, is expressed in the developing nervous system and in the mature central and peripheral nervous systems of adult flies, whilst the protein is found in the nuclei during embryonic development (Wildonger and Mann, 2005). In early embryonic stages, nervy is detected in dividing neuroblasts and as the flies age, it is expressed in the mature embryonic central and peripheral nervous systems. Loss of the NHR4 zinc finger domain only mildly affected nervy activity in the flies, and apart from being bald and some limitations in mechanosensory organ function (response to mechanical stimuli such as pressure or vibration), developed normally suggesting other domains of nervy can compensate (Wildonger and Mann, 2005).
In more complex species, such as mammals, RUNX1T1 is also expressed during early development. The factors affecting early development of oocytes in successful somatic cell nuclear transfer (the process by which an ovum is created with a donor nucleus, which is important in cloning) in mice was investigated. A region on chromosome 4, implicating Runx1t1 as the only protein coding gene in the region, appeared to be responsible for the 2-cell to 4-cell oocyte conversion that occurs in the embryo between E1.5-E2.5 and negatively associated with the 4-cell to blastocyst development that occurs from E2.5-E3.5, suggesting an increase in Runx1t1 could inhibit embryo development (Cheng et al., 2013). At E10.5, Runx1t1 is expressed in regions of the ventral telencephalon, which becomes the basal ganglia and by E13.5 it was also
Chapter 4- Elucidating the genes responsible for delayed tumour formation in ENU treated Th-MYCN mice 167 detected in the pre-plate that contains the pioneer neurons (Alishahi et al., 2009). By E15.5, expression in the cortex is decreased and by birth, it is virtually absent. RUNX1T1 protein expression was also detected in the ganglion cell layer of the retina at E13.5 and continued to be expressed at birth (Alishahi et al., 2009), as well as in the mid- and hindbrain, and in the neural cells of the spinal cord at E12.5 (Alishahi et al., 2009; Sacchi et al., 1998). RUNX1T1 is also expressed in the primitive gut of mice and at E14.5, strong expression is found in the outermost gut layers (Calabi et al., 2001). In Xenopus, injection of in vitro-transcribed mRNA encoding human RUNX1T1 into 4-cell stage embryos impaired axis formation resulting in defective dorso-anterior patterning, the region that becomes the head and neural tube (Moore et al., 2008). In human embryonic and foetal tissue, similar patterns were observed with the highest RUNX1T1 expression observed in the brain with strong protein staining specifically in neural stem cells found in the ventricular zone of the central nervous system (Zhang et al., 2009). Taken together, these studies indicate the importance of RUNX1T1 in early development.
These aspects provide further evidence supporting Runx1t1 as the gene responsible for the delayed tumour phenotype in the 1590 line and although RUNX1T1 has not previously been described in neuroblastoma biology, its role in developing nervous tissue makes it a plausible candidate for the phenotype observed in this study. However, one central question relating to the ability of mutant Runx1t1 to abrogate tumour formation in neuroblastoma-prone mice, is whether the NHR4 YH point mutation leads to loss of wild-type function or a gain of mutant function, and this question is addressed in Chapter 5.
Chapter 4- Elucidating the genes responsible for delayed tumour formation in ENU treated Th-MYCN mice 168 CHAPTER 5 CHARACTERISATION OF THE ROLE OF RUNX1T1 IN NEUROBLASTOMA
5.1 Introduction
As discussed in Chapter 4, there is strong evidence indicating that Runx1t1 is responsible for the loss of tumour phenotype of 1590 mice, which is further highlighted by the established role of this nuclear co-repressor in early development and expression in cells that become the neural tube. However, it is still unclear whether the Runx1t1 mutation results in a loss-of-function or gain-of-function of the protein, or if it is a dominant negative mutation, where the mutant allele antagonises the wild-type allele, similar to what occurs in mutant p53 (Zhou et al., 2016).
To create a knock-out mouse model of Runx1t1, Calabi and colleagues used targeted disruption of the Runx1t1 gene at exon 2, which covers all the alternative splice sites, and inserted the E. coli lacZ coding sequencing at the 3’ end of exon 2 (Calabi et al., 2001) (Figure 5-1). This disrupts transcription and allows for genotyping of the resultant progeny using the lacZ reporter. Also inserted was the herpes simplex thymidine kinase (tk) gene and neomycin resistance (neo) genes for selection of positive clones (Hall et al., 2009). The construct was injected into C57BL/6 blastocysts and implanted into mice. One of the clones resulted in high chimerism as well as germ line transmission. Chimeras were mated to C57BL/6 mice and maintained by interbreeding.
Characterisation of the line by Calabi showed that the major species of the gene was absent in the brains of mice that were knocked-out for Runx1t1, however, an alternative splice site upstream of exon 3 produced a product that was not visible in the wild-type mice. At the translation level, using an antibody directed to the C-terminal domain of the protein, the 75 and 90 kDa proteins were absent, however there was an increase in a 55kDa protein, which might result from an alternate splicing site. Mendelian ratios of wild-type, heterozygote and knock-out mice were detected at birth, however, almost all of the knock-out mice died before 2 days post-birth, with those that did survive
Chapter 5- Characterisation of the role of RUNX1T1 in neuroblastoma 169
Figure 5-1 Generation of the Runx1t1 knock-out mouse model. Clones from a 129/SvJ library were manipulated to insert lacZ, thymidine kinase (tk) gene and neomycin resistance (neo) genes between exons 2 and 3 of Runx1t1. The construct was injected into C57BL/6 blastocysts and the resultant chimeras mated to C57BL/6 mice.
Chapter 5- Characterisation of the role of RUNX1T1 in neuroblastoma 170 demonstrating growth retardation and often death before puberty (Calabi et al., 2001). Approximately 25% of the knock-out mice demonstrated an almost complete absence of the mid-gut, from the distal duodenum to the colon, and empty milk-stomachs, with abnormal gut pathology in the remainder (Calabi et al., 2001). Mtgr1/CBFA2T2 knock- out mice also have a defect in the small intestine, with a reduction in secretory cells of the small intestine after 4 weeks of age, and half of the knock-out mice dying between E18.5 and weaning (Amann et al., 2005). Interestingly, the Mtgr1 knock-out mice also showed 2-4 fold upregulation of Myc in the small intestine (Moore et al., 2008). These two models indicate that this family of genes are important in early development, and defects in these genes can fatality impair function.
There are numerous binding partners of RUNX1T1 that can form high molecular weight complexes. Nuclear receptor co-repressor 1 (N-CoR1) and N-CoR2 (also known as silencing mediator for retinoid or thyroid-hormone receptors (SMRT)) requires the presence of NHR3 and 4 to bind to RUNX1T1 but there are also binding sites within NHR2 (DeKelver et al., 2013; Hildebrand et al., 2001; Lam and Zhang, 2012). Transcriptional regulator, mSin3a, binds in the region between NHR2 and NHR3 (Hildebrand et al., 2001) and the interaction with mSin3a is independent of that with N- CoR (Lutterbach et al., 1998). This transcriptional regulator also has enrichment of target sequences for transcription factors like MYC and MAD:MAX as well as up- regulation of MYC target genes such as Odc1 in embryonic mouse studies (Dannenberg et al., 2005). RUNX1T1 can also homo- and hetero-oligomerise with itself and the other family members, MTG16 and MTGR1 (Zhang et al., 2001). HDAC 1-3, all bind to RUNX1T1 within the same region, and this is independent of the interaction of RUNX1T1 with both mSin3a and N-CoR (Amann et al., 2001). Growth factor independence 1 (GFI1), an oncoprotein that also interacts with MYC proteins, can also bind to RUNX1T1, and with other HDACs (McGhee et al., 2003). The DNA- and RNA-binding protein, SON, involved in transcription and splicing, binds in the region of NHR3 and 4, as deletion mutations in this region demonstrated blockage of binding to RUNX1T1 (DeKelver et al., 2013). It has been hypothesized by Amann and colleagues that RUNX1T1 forms complexes that contain HDAC1 and 2 and mSin3a that are independent of the complex it forms with N-CoR and HDAC3 (Amann et al., 2001). All of these interaction partners make RUNX1T1 a complex protein and the regions where the interactions occur is diagrammatically represented in Figure 5-2.
Chapter 5- Characterisation of the role of RUNX1T1 in neuroblastoma 171
Figure 5-2 Schematic of RUNX1T1 and its binding partners. (Adapted from (Falkenberg and Johnstone, 2014; Lam and Zhang, 2012)
Chapter 5- Characterisation of the role of RUNX1T1 in neuroblastoma 172 This chapter focuses on the validation of Runx1t1 as the causative gene responsible for the 1590 mouse line phenotype and investigates the role of RUNX1T1 in neuroblastoma, both in vivo and in vitro. In vivo studies include the crossing of the Th- MYCN mouse model of neuroblastoma with the Runx1t1 knock-out mouse model described above and characterisation of the 1590 mouse line with respect to tumour formation. Forced biology studies aimed at both overexpression and knock-down of RUNX1T1 were also investigated in human neuroblastoma. Finally, human primary neuroblastoma tissue samples were used to address the question of the role of RUNX1T1 in the biology of neuroblastoma patients.
Chapter 5- Characterisation of the role of RUNX1T1 in neuroblastoma 173
5.2 Results
5.2.1 Th-MYCN mice crossed with Runx1t1 knock-out mice
The finding that the 1590 Runx1t1 mutation causes a dramatic change in the folding and function of the MYND zinc finger domain, suggests that the almost complete abrogation of tumour formation in the Th-MYCN mice is due to a loss-of-function of this protein, rather than a gain-of mutant function. To definitively address this, Runx1t1 knockout mice were imported into the laboratories at Children’s Cancer Institute Australia and maintained as hemizygotes for breeding purposes since the Runx1t1 homozygous mice die within a few days of birth (Materials And Methods, Section 2.2.3.4). For this experiment, Th-MYCN hemizygote mice were crossed with heterozygote Runx1t1 knock-out mice to produce mice that were either wild-type or hemizygote for MYCN and wild-type or heterozygote for Runx1t1. Progeny that were hemizygote for MYCN and heterozygote for Runx1t1 were mated together to produce the three MYCN (WT, hemizygote and homozygote for the transgene) and three Runx1t1 (WT, heterozygous KO and homozygous KO) genotypes in the progeny. Those mice that were MYCN homozygous and either Runx1t1 WT or heterozygous KO, were followed for tumour development for up to 13 months by abdominal palpation.
In mice homozygous for MYCN, loss of one copy of Runx1t1 produced a similar dramatic result previously observed for the 1590 mutant mice, where there was almost complete abrogation of tumour development (Figure 5-3). Mice that were homozygote for MYCN and wild-type (two functional copies) for Runx1t1 had almost complete tumour penetrance, with 92.08% of the mice developing tumours (93 tumours out of 101 mice). However, loss of only one copy of Runx1t1 decreased tumour incidence to 6.54% (10 tumours out of 163 mice). Similarly, and despite the overall decreased tumour incidence due to a different introduced mouse strain, this phenomenon was replicated in mice hemizygote for the MYCN transgene (Figure 5-4). Mice hemizygote for MYCN with two functional copies of Runx1t1, had 9.84% tumour incidence (18 tumours out of 181 mice), which dropped to only 0.86% (3 tumours out of 343 mice)
Chapter 5- Characterisation of the role of RUNX1T1 in neuroblastoma 174
Figure 5-3 Kaplan-Meier survival analysis in Th-MYCN homozygote mice crossed with Runx1t1 knock-out mice. All mice are homozygote for MYCN and either wild-type (red) or heterozygous knock-out (black) for Runx1t1.
Chapter 5- Characterisation of the role of RUNX1T1 in neuroblastoma 175
Figure 5-4 Kaplan Meier survival analysis in Th-MYCN hemizygote mice crossed with Runx1t1 knock-out mice. All mice are hemizygous for MYCN and either wild-type (red) or heterozygous knock-out (black) for Runx1t1.
Chapter 5- Characterisation of the role of RUNX1T1 in neuroblastoma 176 with the loss of one copy of Runx1t1. Both of these results were highly significant (P<0.0001) indicating the importance of Runx1t1 in neuroblastoma tumorigenesis in this model, and confirmed the previous results in the 1590 line, where loss of one functional copy of Runx1t1 can also almost completely disrupt tumour formation.
Also of note is the wider window of tumour development seen in the mice. Whilst some homozygote MYCN mice with wild-type Runx1t1 still developed 10mm tumours before 7 weeks of age (the earliest being at 41 days old), similar to the Th-MYCN colony generally, they continued to develop tumours up until 191 days of age, which is much wider than in the normal colony. Presumably this is due to the mixed background of the mice, which is comprised of 129/SvJ from the Th-MYCN and C57BL/6 and 129/Sv from the Runx1t1 creation. Tumours in the homozygotes with only one copy of Runx1t1 developed tumours from 64 days of age, which is much later than the normal colony, and this continued up until 378 days. Mice hemizygote for MYCN, had much lower tumour incidence regardless of the Runx1t1 genotype, and in those wild-type for Runx1t1, this occurred between 99-345 days of age. The three mice that developed tumours with loss of one copy of Runx1t1, occurred at 270, 333 and 392 days of age, all quite delayed and just before their 13 month cull point.
While tumour development was being followed in the hemizygote MYCN mice, it was noted that many mice, particularly the females, had excessive weight gain, where often females would weigh over 50g, which was unusual for a Th-MYCN mouse. When the genotypes of these mice were examined, it was the mice that were heterozygote for Runx1t1 that were heavier than those wild-type for the gene (Figure 5-5). Mice of different Runx1t1 genotypes began life weighing the same, but over time, the mice with loss of one copy of Runx1t1 gained more weight that those that were wild-type (P<0.0001). This observation of heavier mice was also noted in the mice homozygote for the transgene and heterozygote for Runx1t1, however, the comparison was not able to be made to the wild-type mice in this case, as they developed tumours and could not be monitored for the same period of time as their heterozygote counterparts. The fact that the heterozygous Runx1t1 mice were heavier than their wild-type counterparts may be related to the physiological role of this gene in adipogenesis.
Chapter 5- Characterisation of the role of RUNX1T1 in neuroblastoma 177
Figure 5-5 Weights of Th-MYCN x Runx1t1 mice. Mice hemizygous for the MYCN transgene and heterozygote for Runx1t1 (black) were significantly heavier than those that were wild-type for Runx1t1 (red).
Chapter 5- Characterisation of the role of RUNX1T1 in neuroblastoma 178 Although tumour formation could not be followed in the double knock-out Runx1t1 mice, due to the early death that occurs in these animals, nevertheless it was possible to document what happened to these mice. Over the same period of time as the tumour development, 292 double knock-out mice were generated, 70 of which were homozygote for MYCN, 161 hemizygote and 61 wild-type (Table 5-1).
This ratio of genotypes was close to that expected of Mendelian inheritance (1:2.3:0.9) with a slight over-production of mice hemizygous for MYCN, and an under-production of those that were wild-type. The majority of these pups were either noted as pre-wean death (PWD) or cannibalised, regardless of their MYCN status, with approximately 90% of the pups in each case falling into these categories. A small percentage in each genotype was noted as still-born immediately following birth. Of those that survived, they were smaller in size at both one day after birth (Figure 5-6, A) and at weaning (Figure 5-6, B) compared to their heterozygote and wild-type Runx1t1 littermates. After the phenotype assessment was completed and whilst tissue was being collected for study, a litter of 6 mice produced 3 double knock-out mice that survived to 1 week of age (Figure 5-7, A and B). These knock-out mice were homozygote, hemizygote and wild-type for MYCN. The double knock-out mice were smaller than their littermates, even within the same MYCN genotype (Figure 5-7, C and D). This reduced size in double knock-out mice is likely due to the mid-gut deformity described for this model previously, but also could be attributed to the physiological role of RUNX1T1 in adipogenesis.
If any Runx1t1 double knock-out mice survived to weaning, most needed to be euthanized as they were runty or sick (Table 5-1). Of the very small proportion that did survive to 13 months of age and they were always smaller than the rest of the colony, were ungroomed and often exhibited vague behaviour. Occasionally, they were found dead, with no obvious signs on necropsy. Also noted in the double knock-out mice and a small proportion of the mice heterozygote for Runx1t1 were audiogenic seizures, and this could account for the death observed in these mice, although this cannot be confirmed.
Chapter 5- Characterisation of the role of RUNX1T1 in neuroblastoma 179 Table 5-1 Runx1t1 double knock-out colony statistics MYCN/MYCN MYCN/- MYCN -/- KO/KO* KO/KO* KO/KO* Number (%) Number (%) Number (%) Total Born 70 (100%) 161 (100%) 61 (100%)
PWD† 48 (68.6%) 122 (75.8%) 54 (88.5%)
Cannibalised 15 (21.4%) 17 (10.6%) 1 (1.6%)
Still Born 1 (1.4%) 7 (4.3%) 1 (1.6%)
Runt 2 (2.9%) 2 (1.2%) 1 (1.6%)
Sick/Injury 2 (2.9%) 4 (2.5%) 0 (0%)
Found Dead 1 (1.4%) 3 (1.9%) 1 (1.6%)
13 month cull 1 (1.4%) 6 (3.7%) 3 (4.9%)
*KO refers to knock-out of Runx1t1 †PWD refers to pre-wean death
Chapter 5- Characterisation of the role of RUNX1T1 in neuroblastoma 180
Figure 5-6 Comparison of Runx1t1 wild-type and double knock-out mice. Wild-type Runx1t1 mouse (left) compared to knock-out Runx1t1 mouse (right) at 1 day old (A) and 22 days old (B).
Chapter 5- Characterisation of the role of RUNX1T1 in neuroblastoma 181
Figure 5-7 One week old Th-MYCN x Runx1t1 littermates. Six mice from one litter (A and B) at one week of age. The double Runx1t1 knock-out mice are in the middle and right (A) and the right (B). Mice of the same MYCN genotype were placed together for comparative purposes. Hemizygote MYCN mice (C) with either heterozygote Runx1t1 (left) compared to double knock-out Runx1t1 (right). (D) Wild-type MYCN mice with either heterozygote Runx1t1 (left) compared to double knock-out Runx1t1 (right).
Chapter 5- Characterisation of the role of RUNX1T1 in neuroblastoma 182 5.2.2 Neuroblast hyperplasia in 1590 and Runx1t1 knock-out mice
To investigate the tumour phenotype in the 1590 mouse line further, tissue samples were collected from birth up until 6 weeks of age in both suppressed and unsuppressed mice. In mice that were unsuppressed, tumours were detectable at 4 weeks of age both by palpation and visually (Figure 5-8; A, right and C), compared to suppressed mice at the same age (Figure 5-8; A, left). By 6 weeks of age, the tumours in the unsuppressed mice were large, bulky, vascular, and often connected to the kidneys (Figure 5-8; B, right and D), compared to suppressed mice, which were tumour-free at the same age (Figure 5-8; B, left). These tumours from the unsuppressed mice are characteristic of normal neuroblastoma development in the Th-MYCN mouse, with respect to timing, site and appearance.
A previous study examining the mechanism of embryonal tumour initiation in the Th- MYCN mouse model, found that tumour formation was dependent upon the inappropriate persistence of neuroblast hyperplasia within the paravertebral ganglia of these mice by comparison with wild-type mice (Hansford et al., 2004). To examine this possibility in the 1590 mutant Runx1t1 mice, animals were embedded in paraffin, sectioned and stained with haematoxylin and eosin for histological analysis. Hyperplasia in the ganglia was scored and documented for both suppressed and unsuppressed mice at various ages, with between 12-64 ganglia scored per time point, per 1590 mutation. Neuroblast hyperplasia, defined as greater than 30 small, blue round cells in the ganglia was observed at a similar level in both suppressed and unsuppressed 1590 mice at birth (day 0) and 1 week of age (Figure 5-9, A). Whilst ganglia hyperplasia decreased in the suppressed mice at 2 weeks of age onwards, and by 4 weeks of age was undetectable, there was a general persistence of hyperplasia in the unsuppressed mice, peaking at 2 weeks of age and still present at 4 weeks. The persistence of hyperplastic ganglia at 3 and 4 weeks of age, was accompanied by frank tumour formation in these mice, with half of the mice scored having tumours at 3 weeks of age, and all of them by 4 weeks of age (Figure 5-9, B). No suppressed mice had any evidence of tumour formation over this time frame. These results suggest that there is a critical window where Runx1t1 is exerting its effects, presumably somewhere between 1-2 weeks of age, after which the tumour formation is already determined.
Chapter 5- Characterisation of the role of RUNX1T1 in neuroblastoma 183
Figure 5-8 Comparison of suppressed and unsuppressed 1590 mice. A 1590 suppressed mouse (Panel A and B, left) compared to an unsuppressed mouse (Panel A and B, right) at 4 weeks (A) and 6 weeks (C) of age. The tumours in the unsuppressed mice are enlarged for ease of visualisation (Panels C and D).
Chapter 5- Characterisation of the role of RUNX1T1 in neuroblastoma 184
Figure 5-9 Scoring of ganglia and tumours from 1590 suppressed and unsuppressed mice. Percentage hyperplasia (A) and tumour formation (B) in 1590 suppressed () and unsuppressed () mice.
Chapter 5- Characterisation of the role of RUNX1T1 in neuroblastoma 185 Similar to the 1590 cohort, samples were collected from the Runx1t1 knockout colony at birth to investigate hyperplasia in the ganglia, with between 12-66 ganglia scored per genotype. Even at this early time point, marked differences were observed with respect to the percentage of hyperplastic ganglia, depending on the genotype. Thus, a similar amount of hyperplasia was observed (between 53-67%) in mice wild-type (WT/WT) for Runx1t1, regardless of the MYCN status (Figure 5-10), confirming what was also observed in the 1590 cohort of mice at this time point. When loss of one Runx1t1 copy occurred, the percentage hyperplasia in the ganglia, dropped to between 25-45%. Loss of two copies of Runx1t1 further decreased the amount of hyperplasia to less than 20% of the ganglia scored and this did not vary according to the MYCN genotype. Whilst further time points are needed to complete this analysis, these data strongly indicate that the loss of Runx1t1 is having a dramatic effect on the level of hyperplasia in the ganglia, regardless of the MYCN genotype present and therefore also on the potential for tumour formation in these animals.
5.2.3 Inheritance of two copies of the 1590 mutation
To examine if two copies of the 1590 suppressor mutation has the same effect as knock- out of two copies of Runx1t1, two 1590 suppressed mice were mated together to generate mice that had one or two copies of the suppressor mutation. No mice with two copies of the suppressor mutation were detected previously, either due to embryonic lethality or cannibalisation of the pups before they were identified. Knowing that the likelihood of PWD was high, the breeder boxes were checked for still-born and dead pups shortly after birth, and up until weaning age. The genotypes of the pups born followed Mendelian inheritance (1:2.03:0.97), indicating that inheritance of two copies of the 1590 suppressor mutation is not embryonic lethal. There was a high percentage of still-born pups in all three 1590 genotypes, however, those that had two copies of the suppressor mutation had almost 73% of pups either being born dead or dying soon after birth (Table 5-2). Those that did survive birth, died within 3 days of being born and no mice survived to weaning. For mice that were wild-type for the 1590 mutation, those that survived to weaning all went on to develop tumours while the majority of the suppressed mice with one copy of the mutation became long term tumour-free survivors, as expected. This suggests that inheritance of two copies of the suppressor
Chapter 5- Characterisation of the role of RUNX1T1 in neuroblastoma 186
Figure 5-10 Scoring of ganglia in day 0 Th-MYCN x Runx1t1 mice. Percentage hyperplasia in the mice homozygous (MYCN/MYCN), hemizygous (MYCN/-) and wild- type (-/-) for the MYCN transgene split by their Runx1t1 genotype of two functional copies (WT/WT; black), one copy (WT/KO; red) or knock-out (KO/KO; blue).
Chapter 5- Characterisation of the role of RUNX1T1 in neuroblastoma 187
Table 5-2 Outcome table for mice with two copies of the suppressor mutation
1590 1590 1590 Homozygous Wild-type Heterozygote Variant
Total 34 69 33
Still-born 4 (11.8%) 13 (18.8%) 24 (72.7%)
PWD* 8 (23.5%) 10 (14.5%) 9 (27.3%)
Tumour 22 (64.71%) 3 (4.3%) - Cull
Found dead – - 1 (1.4%) - no tumour
13 month - 42 60.9%) - cull * PWD refers to pre-wean death
Chapter 5- Characterisation of the role of RUNX1T1 in neuroblastoma 188 mutation is equivalent to knock-out of Runx1t1, with insufficient RUNX1T1 generated for normal function in the mouse.
5.2.4 RUNX1T1 expression levels in human cell lines and mouse tissues
The expression of RUNX1T1 in human neuroblastoma cell lines was examined to determine if there was a difference between five single copy and five amplified MYCN lines, using real time quantitative PCR (as described in Materials and Methods 2.2.2.7). Although the highest expressing RUNX1T1 lines did come from those that had amplified MYCN (IMR-32 and CHP-134), there was no significant difference between the mean level of RUNX1T1 expression between amplified and non-amplified cell lines (Figure 5-11, A and B). In contrast however, the protein levels of RUNX1T1 were significantly higher in those lines with amplified MYCN by comparison with the single copy lines (Figure 5-11, C and D) (P=0.0016). These results suggest a disconnect between the RNA and protein levels of RUNX1T1.
To determine expression of Runx1t1 in mouse tissues, organs were collected from 3 homozygote Th-MYCN mice with tumours at 6 weeks of age and 3 matching wild-type controls that didn’t develop tumours. Liver, kidney and spleen had relatively low levels of Runx1t1 regardless of the genotype of the mice (Figure 5-12). The highest levels were found in the brain, with equal expression in both homozygote and wild-type mice. Similar levels were found in the tumours of homozygote MYCN mice, comparable to that found in brain tissue. Based on the literature, the heart also has relatively high levels of Runx1t1, and whilst it was not as high as tumour or brain, it had the highest level of expression after these organs. When the 3 mice in each genotype were combined and an unpaired t-test performed, there was significantly higher expression in homozygotes compared to wild-type mice only observed in the heart (P=0.0193) (Figure 5-13). Small and large intestine samples were also examined since Runx1t1 has previously been implicated in the midgut of knock-out mice. Whilst there was a trend towards higher expression of Runx1t1 in these samples from homozygote MYCN mice, this failed to achieve significance (P=0.1341 for small intestine; P=0.0629 for large intestine) (Figure 5-13).
Chapter 5- Characterisation of the role of RUNX1T1 in neuroblastoma 189
Figure 5-11 Expression of RUNX1T1 in neuroblastoma cell lines. Gene expression of RUNX1T1 in MYCN amplified and single copy (Non AMP) cell lines (A and B) compared to protein levels of RUNX1T1 (C and D). Graphs in B and D represent data from A and B, respectively, relative to the cell line SH-EP in each case.
Chapter 5- Characterisation of the role of RUNX1T1 in neuroblastoma 190
Figure 5-12 Gene expression of Runx1t1 across different mouse tissues. Six week old homozygote MYCN mice (coloured bars) were compared to 6 week old wild-type (WT) mice (grey-scale bars) in nine tissue types (wild-type mice do not develop tumours).
Chapter 5- Characterisation of the role of RUNX1T1 in neuroblastoma 191
Figure 5-13 Gene expression of Runx1t1 in intestinal tissue and heart of 6 week old mice. Homozygote (red) and WT (black) mouse groups were compared for the small intestine, large intestine, and heart, with a statistically significance difference only observed in the heart samples (P=0.0193). * P value < 0.05
Chapter 5- Characterisation of the role of RUNX1T1 in neuroblastoma 192 Expression of MYCN and Runx1t1 was next investigated in two unsuppressed 1590 mice, 0010 and 0029, and two that developed delayed tumours, the 1590 founder mouse and one of her progeny (0407) that developed a tumour at ~20 weeks of age. For MYCN, expression in the tumours was comparable, with the exception of the 1590 founder tumour, which had very low MYCN expression (Figure 5-14, A). This is most likely due to the teratoma nature of the tumour, since homozygous copy number in this tumour had been confirmed previously. Runx1t1 expression was detected in all 4 tumours examined, although the expression was higher in the two suppressed samples, potentially indicating a need to increase the Runx1t1 expression beyond that needed for a normal tumour, to form a delayed tumour (Figure 5-14, B). Interestingly, 1590 had increased Runx1t1 expression, despite not having very high MYCN expression.
Finally, RNA was extracted from pre-malignant ganglia dissected from Th-MYCN mice at time points from birth through to 6 weeks of age and examined for Runx1t1 expression. Global expression of Runx1t1 in these cells, covering all three isoforms of the mouse, was assessed and at birth, similar low levels of Runx1t1 were seen in all ganglia samples (Figure 5-15). Although the level of Runx1t1 did not change in the wild-type mice, there was a marked increase in expression of this gene in the MYCN homozygous ganglia at 2 weeks and 4 weeks, which then plateaued at 6 weeks. Interestingly, by 6 weeks of age, a fully formed tumour is evident in MYCN homozygous mice, suggesting a critical time required for Runx1t1 expression that occurs between birth and 2 weeks of age.
Chapter 5- Characterisation of the role of RUNX1T1 in neuroblastoma 193
Figure 5-14 Gene expression levels of MYCN and Runx1t1 in tumours from the 1590 line mice. Tumours from unsuppressed and suppressed 1590 mice were examined for MYCN (A) and Runx1t1 (B) RNA expression levels.
Chapter 5- Characterisation of the role of RUNX1T1 in neuroblastoma 194
Figure 5-15 Runx1t1 expression in ganglia from Th-MYCN mice. Homozygote mice (red) were compared to WT mice (blue) at intervals from birth through to 6 weeks.
Chapter 5- Characterisation of the role of RUNX1T1 in neuroblastoma 195
5.2.5 Overexpression of Runx1t1
Overexpression constructs of Runx1t were created in two plasmids, namely CMV10 and CMV14 which contain a FLAG tag to confirm overexpression (as described in Materials and Methods 2.2.2.9). Briefly, these constructs were generated with the full length murine sequence of Runx1t1 and contained either the wild-type (WT) sequence or the 1590 point mutant (YH). The constructs were transfected into human MYCN- amplified BE(2)-C cells and plated at 500 cells per well in a 6-well plate. Colony formation was examined after 10 days, in 3 separate runs. Transfection with empty vector (EV) decreased colony formation compared to the Lipofectamine alone control, however, there was no significant difference in colony formation with either the WT or YH mutation with either the CMV10 or CMV14 construct, compared to the EV control, despite achieving overexpression of the Runx1t1 constructs, as demonstrated by the FLAG western blot (Figure 5-16). There are a number of possible explanations for this result with the most likely two being that a fully transformed MYCN-amplified cell line no longer requires RUNX1T1 to support its growth, or alternatively that the level of RUNX1T1 in these amplified cells is already above that needed for its growth requirements.
5.2.6 Knock-down of RUNX1T1 in neuroblastoma cells
Since overexpression of full length murine Runx1t1 did not have an effect on colony formation in BE(2)-C cells, the effect of RUNX1T1 siRNA knock-down in this and several other cell lines was examined using a SMARTPOOL siRNA (see Material and Methods 2.2.1.2). Despite using up to 100nM of siRNA, very little knock-down of RUNX1T1 was observed in BE(2)-C (Figure 5-17, A) and CHP-134 (Figure 5-17, B). Whilst there were significant differences observed in BE(2)-C at both 50nM (P=0.0015) and 100nM (P=0.0225) and in CHP-134 at 50nM (P=0.0074) RUNX1T1 siRNA, the gene expression level of RUNX1T1 never dropped below 50% of the control, indicating ineffective knock-down. Knock-down of RUNX1T1 was also examined in the Stage 4s cell line, NBL-W-N, since disease in patients is capable of spontaneous regression and
Chapter 5- Characterisation of the role of RUNX1T1 in neuroblastoma 196
Figure 5-16 Colony formation in BE(2)-C cells transfected with Runx1t1 overexpression constructs. Number of colonies present in each transfection condition (A) and corresponding western blot demonstrating overexpression (B). Lipo only, lipofectamine reagent; EV, empty vector; WT, Runx1t1 wild-type construct; YH, Runx1t1 1590 mutant construct.
Chapter 5- Characterisation of the role of RUNX1T1 in neuroblastoma 197
Figure 5-17 Knock-down of RUNX1T1 using SMARTPOOL siRNA. Gene expression of RUNX1T1 after knock-down using 50-100nM of control or RUNX1T1 siRNA in BE(2)-C (A), CHP-134 (B) and NBL-W-N (C). * P value <0.05; ** <0.01.
Chapter 5- Characterisation of the role of RUNX1T1 in neuroblastoma 198 there was the possibility that loss of RUNX1T1 might have a greater effect in these cells. However, there was no change in the level of RUNX1T1 at either concentration of RUNX1T1 siRNA used (Figure 5-17, C).
Since the knock-down did not appear to be very effective using the SMARTPOOL siRNA, another source of RUNX1T1 siRNA was used (Ambion™), which had previously been demonstrated to markedly knock-down RUNX1T1 at the protein level using 20nM siRNA (Baby et al., 2014). KELLY cells, were transfected with this siRNA and harvested 24, 48 and 72 hours post-transfection and examined for RUNX1T1 expression. Unfortunately, very little RUNX1T1 knock-down was observed at any time point or concentration of RUNX1T1 siRNA (Figure 5-18). These results suggest that it is difficult to alter the RNA levels of RUNX1T1 using siRNA. This could be due to the many transcripts that RUNX1T1 has although the siRNAs have been designed to target regions common to all 15 known transcripts. The fact that good knock-down cannot be achieved even with high concentrations of siRNA, suggests that either the siRNA are not well designed or there is some mechanism able to overcome the knock-down even in a short time frame.
Given the lack of effective knockdown of RUNX1T1 using siRNA, RUNX1T1 shRNA constructs were generated, since knock-down of this construct can be induced using doxycycline. This has the advantage that the machinery is integrated into the cell and transfection efficiencies, or inefficiencies, do not have the same amount of variability. Constructs were designed and created based on the Ambion™ siRNA sequence and a sequence upstream of this (shRNA 3) using the FH1UTG lentiviral vector. These constructs were transduced into BE(2)-C cells. Colony formation was assessed 10 days after plating, with addition of fresh doxycycline every 72 hours. The addition of doxycycline did not change the number of colonies in the FH1UTG, the empty vector control (Figure 5-19, A and B) with no change in RUNX1T1 protein levels (Figure 5-19, C). There was a small but highly significant decrease in the number of colonies in the shRNA3 construct (P=0.0019), as well as fewer large colonies seen (Figure 5-19, B). In addition, a bigger overall decrease in colony number was observed when the Ambion™ shRNA sequence was induced (P=0.0297), with a greater reduction in RUNX1T1 protein levels (Figure 5-19, C) and also fewer large colonies. These results provide evidence that knock-down of RUNX1T1 can influence colony formation in a human cell line that was originally established from a relapsed neuroblastoma tumour.
Chapter 5- Characterisation of the role of RUNX1T1 in neuroblastoma 199
Figure 5-18 RUNX1T1 knock-down in KELLY cells using Ambion siRNA. Cells were harvested at 24, 48 and 72 hours post-transfection with 50– 100nM of control or RUNX1T1 siRNA.
Chapter 5- Characterisation of the role of RUNX1T1 in neuroblastoma 200
Figure 5-19 Colony formation after induction of RUNX1T1 shRNA using doxycycline. Colony formation expressed as a percentage of the no doxycycline control in the empty vector (FH1UTG), and shRNA3 and Ambion shRNA (A) of three independent experiments. Representative images of the colonies (B) and western blot (C), with and without doxycycline (Dox) are shown. * indicates P< 0.05 and ** P<0.005.
Chapter 5- Characterisation of the role of RUNX1T1 in neuroblastoma 201 5.2.7 RUNX1T1 expression after MYCN knock-down in neuroblastoma cells
To determine whether RUNX1T1 expression is influenced by the level of MYCN, two approaches were used; siRNA knockdown of MYCN in BE(2)-C cells and tetracycline- induced repression of MYCN using SH-EP Tet-21/N cells. Firstly, 24 hours after transfection of MYCN siRNA into BE(2)-C cells, the level of MYCN RNA was approximately 25% of the control and this level of suppression was sustained for up to 72 hours (P<0.0005 at all time points) (Figure 5-20, A). In parallel with this result, RUNX1T1 RNA expression was also significantly reduced after 24 hours (P<0.004), although not as dramatically as MYCN (Figure 5-20, B). This was further decreased by 48 hours (P<0.004) and was at its lowest at approximately 50% of the control at 72 hours (P=0.0004). Similarly, in the SH-EP Tet-21/N cells, addition of tetracycline led to MYCN being barely detectable 72 hours after induction (P<0.0001) (Figure 5-21). RUNX1T1 was also markedly decreased at this time point (P=0.0002). Together, these results suggest that MYCN is capable of regulating RUNX1T1.
Chapter 5- Characterisation of the role of RUNX1T1 in neuroblastoma 202
Figure 5-20 Knock-down of MYCN in BE(2)-C cells. Relative expression of MYCN (A) and RUNX1T1 (B) in BE(2)-C cells 24, 48 and 72 hours after transfection with siRNA to MYCN. Control siRNA (black bars) was compared to MYCN siRNA (grey bars) in each case. * indicates P< 0.05 and ** P<0.005 and *** P<0.0005.
Chapter 5- Characterisation of the role of RUNX1T1 in neuroblastoma 203
Figure 5-21 MYCN and RUNX1T1 expression in SH-EP Tet-21/N cells after induction of MYCN knock-down using tetracycline. Control (no tetracycline) was compared to the induction of MYCN suppression (+ Tet) with respect to MYCN (left) and RUNX1T1 (right) expression. *** indicates P<0.0005.
Chapter 5- Characterisation of the role of RUNX1T1 in neuroblastoma 204 5.2.8 Fractionation of Runx1t1 protein complexes
RUNX1T1 is a non-DNA binding nuclear co-repressor and functions by forming protein-protein interactions. Therefore to investigate proteins that complex with RUNX1T1 in neuroblastoma cells and specifically within the NHR4 MYND zinc-finger domain, the wild-type and YH mutant forms of this region were FLAG-tagged and transfected into BE(2)-C cells. Nuclear protein extracts were made and the proteins were fractionated using a size exclusion column, where larger complexes are eluted first. As a control, untransfected BE(2)-C cells were also fractionated. Fractions 1-40 were run on polyacrylamide gels and either stained with Coomassie Brilliant Blue (as it was not possible to run a loading control) or transferred onto a membrane and blotted for known complex partners.
For the Coomassie staining, the three conditions showed similar results. The proteins did not start being eluted until Fraction 8 or 9 in all cases. For the untransfected cells, the proteins were concentrated between Fractions 8-24, with minimal proteins seen between 25-32 and none observed beyond Fraction 33 (Figure 5-22). For the wild-type Runx1t1 transfected cells the proteins were strongest between Fractions 9-28 (Figure 5-23) and for the YH mutant construct between Fractions 8-27 (Figure 5-24).
Western blotting for MYCN was performed on fractionated BE(2)-C cells. MYCN was detected in similar fractions in all protein extracts. In the untransfected cells this oncoprotein could be detected in Fractions 8-19 (Figure 5-25, A), in the wild-type transfected cells in Fractions 10-19, with low detection between 12-15 (Figure 5-25, B), and in the YH transfected sample between 8-11 and 14-19 (Figure 5-25, C). The untransfected cells appeared to have two peaks of MYCN expression, at Fraction 9 and 16. The wild-type transfected cells also had high levels in Fraction 10 and 16, while the YH mutant had the highest expression in Fractions 9 and 10 with very little in Fraction 16 unlike the other two conditions. These results suggest that in the normal state, MYCN is interacting with many partners making different sized complexes. Once the cells were transfected with either wild-type or mutant Runx1t1, MYCN is found in less fractions, but also the intensity of expression is lower, particularly in the YH transfected
Chapter 5- Characterisation of the role of RUNX1T1 in neuroblastoma 205
Figure 5-22 Coomassie Blue staining of protein fractions 1-40 of untransfected BE(2)-C cells
Chapter 5- Characterisation of the role of RUNX1T1 in neuroblastoma 206
Figure 5-23 Coomassie Blue staining of protein fractions 1-40 of BE(2)-C cells transfected with a wild-type Runx1t1 construct.
Chapter 5- Characterisation of the role of RUNX1T1 in neuroblastoma 207
Figure 5-24 Coomassie Blue staining of protein fractions 1-40 of BE(2)-C cells transfected with the YH mutant Runx1t1 construct.
Chapter 5- Characterisation of the role of RUNX1T1 in neuroblastoma 208
Figure 5-25 Western blot of fractionated BE(2)-C cells using a MYCN antibody. Control untransfected BE(2)-C cells (A) and those transfected with either wild-type (B) or YH mutant (C) Runx1t1 constructs were blotted for MYCN.
Chapter 5- Characterisation of the role of RUNX1T1 in neuroblastoma 209 sample, where very little MYCN is observed in fractions 16 onwards. This suggests that MYCN is preferentially making larger, rather than smaller complexes. The wild-type and mutant Runx1t1 transfected BE(2)-C samples were then blotted for FLAG to detect the presence of the transfected construct and therefore RUNX1T1. The FLAG tag could be detected in Fractions 10-13 in the wild-type transfected sample (Figure 5-26, A) and in Fractions 8-12 of the YH transfected sample (Figure 5-26, B). There was also some RUNX1T1 detected in the WT sample in Fractions 17-19 as well as some staining of lower molecular weight bands in Fractions 21-23 that were not detected in the YH transfected sample. It is unknown what these lower molecular weight bands are. These results indicate that RUNX1T1 can be found in the same fractions as MYCN and in similar fractions between the wild-type and YH mutant Runx1t1, suggesting the potential for MYCN and RUNX1T1 to form a complex, although further analysis is required to confirm this.
Since Fractions 9-16 contained both MYCN and RUNX1T1 in the WT and YH transfected cells, the known protein partners that complex with RUNX1T1 were investigated to determine any changes that had occurred when either wild-type or YH mutant Runx1t1 was present. In wild-type transfected cells, RUNX1T1 was found in the same fractions as MYCN, HDAC1 and 2 and mSin3A (Figure 5-27, A). N-CoR1 was also found in these fractions, although the antibody for this protein was not particularly clean despite several attempts to improve the signal. For the YH transfected cells, there was a similar profile with respect to the HDACs and mSin3A, however, N-CoR1 could not be detected in the same fractions as RUNX1T1 or MYCN (Figure 5-27, B). This suggests the introduction of the mutation into NHR4 leads to RUNX1T1 no longer binding to N-CoR1 which in turn will affect the function of RUNX1T1 as a co- repressor. The fact that HDAC1 and 2 and mSin3A are still present in the fractions, also suggests that these proteins are able to bind in other regions of RUNX11T1 as has been reported (Amann et al., 2001; Hildebrand et al., 2001).
Chapter 5- Characterisation of the role of RUNX1T1 in neuroblastoma 210
Figure 5-26 Western blot of transfected and fractionated BE(2)-C cells using a FLAG antibody. BE(2)-C cells were transfected with either wild-type (A) or YH mutant (B) Runx1t1 constructs and fractionated.
Chapter 5- Characterisation of the role of RUNX1T1 in neuroblastoma 211
Figure 5-27 Western blots of RUNX1T1 protein complex partners. BE(2)-C cells were transfected with WT (A) or YH mutant (B) Runx1t1, fractionated and blotted for various potential complex proteins.
Chapter 5- Characterisation of the role of RUNX1T1 in neuroblastoma 212 5.2.9 Correlation of RUNX1T1 expression with neuroblastoma outcome
In the in vivo mouse models, RUNX1T1 was found to be critical for tumorigenesis. To investigate whether RUNX1T1 might also involved in tumour progression in human neuroblastoma, expression of this gene in a cohort of 649 neuroblastoma samples was correlated with outcome using Kaplan-Meier survival analysis was undertaken (see Materials and Methods 2.1.4).
Following this analysis, low expression rather than high expression of RUNX1T1 was found to be predictive of poor outcome (Figure 5-28). Since the mouse studies had shown a loss of function associated with tumorigenesis, it was anticipated that high expression of this gene would predict poor outcome in those children with neuroblastoma. Given this finding was contrary to expectations, another publically available expression array database was interrogated. This database, comprising 101 single copy MYCN samples with aggressive Stage 4 disease, also showed that low expression of RUNX1T1 predicted poor outcome (Figure 5-29). This demonstrated that even in the absence of high levels of MYCN and poor outcome disease, RUNX1T1 was predicting outcome in the opposite direction to what had been predicted.
Human RUNX1T1 has at least 15 known variants and the possibility was investigated that the expression arrays being analysed were using RUNX1T1 probes that only detected specific isoforms. Investigation of this showed that there were multiple probes used on each of the platforms for RUNX1T1, although all appeared to be probing the same variants, possibly due to ease of designing robust assays for these. In short, the 5 probes investigated only detect 4 out of 15 of the transcripts (variants 1-4) covering 3 out of the 6 isoforms (A, B and C). Although this could provide a simple explanation for what was observed, this seems unlikely. In addition, a similar analysis was performed on a database of 498 neuroblastoma samples that had been subjected to RNA-seq, and this indicated that the level of RUNX1T1 overall was low in neuroblastoma and again low levels were predictive of poor outcome (Figure 5-30, A). Furthermore, the level of RUNX1T1 in MYCN amplified neuroblastoma samples was
Chapter 5- Characterisation of the role of RUNX1T1 in neuroblastoma 213
Figure 5-28 Kaplan-Meier survival analysis based on RUNX1T1 expression in 649 primary neuroblastomas. Low levels of RUNX1T1 are predictive of poor outcome.
Chapter 5- Characterisation of the role of RUNX1T1 in neuroblastoma 214
Figure 5-29 Kaplan-Meier survival analysis based on RUNX1T1 expression in 101 Stage 4 single copy MYCN human neuroblastoma samples (Asgharzadeh et al., 2006)
Chapter 5- Characterisation of the role of RUNX1T1 in neuroblastoma 215
Figure 5-30 RUNX1T1 expression levels in 498 human neuroblastoma samples. Kaplan-Meier survival analysis based on RUNX1T1 expression in 498 human neuroblastoma samples using RNA-seq (A). The level of RUNX1T1 in MYCN non- amplified tumour samples is significantly higher compared to amplified samples (B).
Chapter 5- Characterisation of the role of RUNX1T1 in neuroblastoma 216 significantly lower by comparison with non-amplified samples (P<0.0001) and low levels were also observed in MYCN amplified patients (Figure 5-30, B).
5.2.10 Correlation of RUNX1T1 protein in human neuroblastoma samples
To determine whether the relatively low level of RUNX1T1 gene expression observed in MYCN-amplified primary neuroblastomas by comparison with non-amplified tumours was also apparent at the protein level, immunohistochemistry (IHC) was performed on six tumour microarrays (TMA), containing neuroblastoma (n=79), ganglioneuroma (n=12) and ganglioneuroblastoma (n=5) samples. Only cores that were intact, had tumour material present, and had clinical information available were used in the analysis. The microarrays were scored for intensity of RUNX1T1 expression (as outlined in Materials and Methods 2.2.1.6) and correlated both with MYCN amplification status and also clinical outcome. Briefly, the samples were scored for the percentage of the core with positive staining (between 0-100%) which was then given a number from 0-3. This was then multiplied by the intensity of the staining from negative to strong, which was also given a number from 0-3. Therefore the minimum score possible was 0, indicating negative staining in 100% of the core to a maximum of 9 which indicates strong staining in >51% of the core.
The cohort was shown to be representative of neuroblastoma in general, with MYCN amplification (P<0.001), stage (P<0.001), and age of diagnosis (P=0.022), all being prognostic of outcome (Figure 5-31). Representative images of negative and strong RUNX1T1 staining are shown (Figure 5-32) and only nuclear staining was classed as positive staining. The neuroblastoma samples had higher levels of RUNX1T1 compared to the ganglioneuroma and ganglioneuroblastoma samples scored (Figure 5-33, A). Overall, those samples that were MYCN amplified had higher protein levels of RUNX1T1, which was the opposite to that observed at the gene expression level (Figure 5-33, B). When the scoring was correlated to clinical outcome, the overall cohort was significant when the data was cut at an expression score of 0-4 versus >4 (P=0.034) and approached significance with an expression score of 0-3 compared to >3 (P=0.054) (Figure 5-34).
Chapter 5- Characterisation of the role of RUNX1T1 in neuroblastoma 217
Figure 5-31 Clinical characteristics of the tumour microarray cohort. MYCN status, stage and age, with a cut-off at diagnosis of >18 months were prognostic of outcome.
Chapter 5- Characterisation of the role of RUNX1T1 in neuroblastoma 218
Figure 5-32 Immunohistochemical staining of neuroblastoma tumour microarrays for RUNX1T1. Haematoxylin and Eosin staining shows the tumour morphology. Representative cores from the tissue microarray are shown demonstrating either negative nuclear staining (upper right panel) or strong positive nuclear staining (lower right panel) for RUNX1T1. Magnification 600x for larger panels; 100x for insets.
Chapter 5- Characterisation of the role of RUNX1T1 in neuroblastoma 219
Figure 5-33 RUNX1T1 expression in a TMA of primary human neuroblastoma. A score was determined for the level of RUNX1T1 staining by multiplying the percentage stained by the intensity of staining. Neuroblastoma (NB), ganglioneuroma (GN) and ganglioneuroblastoma (GNB) samples graphed according to their RUNX1T1 score (A) where neuroblastoma had higher expression of RUNX1T1. For neuroblastoma samples split by MYCN amplification status, the level of RUNX1T1 was significantly higher in those samples that were MYCN amplified compared to those that were non-amplified (B). Amp, amplified; non-amp, non-amplified MYCN.
Chapter 5- Characterisation of the role of RUNX1T1 in neuroblastoma 220
Figure 5-34 RUNX1T1 expression and Kaplan-Meier survival analysis of neuroblastoma samples from the tumour microarray. A high level of RUNX1T1 was prognostic of poor outcome in samples where the score was >4 (P=0.034) and approached significance when the score was >3 (P=0.053).
Chapter 5- Characterisation of the role of RUNX1T1 in neuroblastoma 221 These results showing that a high level of RUNX1T1 protein confers poor outcome in primary neuroblastoma tissue, is in direct contrast to the results from the gene expression array databases. This difference may relate to either increased translational efficiency or increased protein stability of RUNX1T1 in MYCN amplified tumours, both possibilities that warrant further investigation. Irrespective, the current study suggests that RUNX1T1 is not only involved in tumour initiation, but also in tumour maintenance, offering the potential of therapeutically targeting this nuclear co-repressor.
Chapter 5- Characterisation of the role of RUNX1T1 in neuroblastoma 222
5.3 Discussion
The finding that genetic knockout of Runx1t1 recapitulates the phenotype observed with the 1590 point mutation, provides the definitive proof that inhibition of neuroblastoma tumorigenesis observed in the Th-MYCN mice is due to a loss of function of this non- DNA binding transcriptional regulator. Although three murine Runx1t1 isoforms have been identified, the 1590 point mutation occurs in the NHR4 domain in the carboxy terminal of the protein, suggesting that loss of the long form of the protein is the critical isoform necessary for tumour abrogation. In addition, since allelic loss is sufficient in both mouse models to prevent tumour formation, this implies that there is a threshold RUNX1T1 protein level that must be achieved for tumorigenesis to occur.
Runx1t1 is expressed early in development in dividing neuroblasts and pioneer neurons, in the embryonic central and peripheral nervous systems and in the regions that become the basal ganglia (Alishahi et al., 2009; Wildonger and Mann, 2005). Murine studies have shown that neuroblastoma arises from a defect in the developing sympathetic nervous system, leading to the persistence of neuroblast hyperplasia. The hyperplasia observed in the 1590 ganglia at birth was similar to the original study by Hansford and colleagues using this same Th-MYCN model. Thus at day 0, the suppressed and unsuppressed 1590 mice displayed levels of hyperplasia of approximately 40%, which was consistent with the homozygote, hemizygote and wild-type MYCN ganglia scored previously (Hansford et al., 2004). However, unlike the hyperplastic ganglia observed in the 1590 unsuppressed mice, which followed a similar pattern to the homozygous MYCN mice in the Hansford study (i.e. increasing to 80% before gradually declining but still remaining at approximately 30% by week 4), 1590 unsuppressed more closely mirrored normal MYCN mice with ganglia hyperplasia being undetectable by week 4. Therefore, the suppressed mice exhibit hyperplasia that would be consistent with mice lacking the MYCN transgene. In the Th-MYCN x Runx1t1 mice, the effect of Runx1t1 loss appears to be independent of MYCN transgene status, since a reduction in ganglia hyperplasia in the Runx1t1 knock-out mice occurs both in the presence and absence of MYCN. Therefore, taking all these factors into consideration, a model of neuroblastoma
Chapter 5- Characterisation of the role of RUNX1T1 in neuroblastoma 223 development is proposed (Figure 5-35). High levels of RUNX1T1 in dividing neuroblasts and pioneer neurons allows these cells to remain in an undifferentiated and proliferative state, however, when RUNX1T1 is decreased, through either normal regulation or through gene deletion, such as has been demonstrated here, these cells differentiate or go through cell death. However, when MYCN is amplified, this oncoprotein is able to maintain high levels of RUNX1T1 allowing these cells to remain undifferentiated, survive normal death stimuli and lead to neuroblastoma development.
It was interesting to note the increased weight in the mixed strain background mice in the Th-MYCN x Runx1t1 line when one copy of Runx1t1 was lost. This may be linked to the ability of RUNX1T1 to form complexes with N-CoR, which was absent in protein complex studies when the point mutation was transfected into cells. When adipocyte specific N-CoR deleted mice were fed a high fat diet, they became more obese than those mice with wild-type N-CoR on the same diet, with an increase observed in food intake (Li et al., 2011). N-CoR is also known to change mRNA splicing to influence adipocyte differentiation and this splicing can be switched in vivo to result in increased weight gain (Goodson et al., 2011; Goodson et al., 2014). It could be hypothesized that even though our mice were fed a standard diet, the defect in Runx1t1 through loss of one copy could result in a defect in N-CoR binding and therefore an increase in fat accumulation and hence an increase in overall body weight. This theory is further supported by work in promyelocytic leukaemia, where loss of the promyelocytic leukaemia gene (pml), which is frequently translocated in this leukaemia, inhibits adipogenesis, through a N-CoR-SMRT co-repressor complex (Kim et al., 2011). It was hypothesized that the preadipocytes could no longer differentiate into adipocytes due to the decreased fat accumulation in pml deleted mice. SMRT expression has also been found to be downregulated in obese subjects, specifically in their adipose tissue (Toubal et al., 2013). Also to note is the role of Runx1t1 directly in adipogenesis. RUNX1T1 has been implicated as an inhibitor of adipogenesis (Rochford et al., 2004) and several microRNAs have been implicated to impair the adipogenesis of brown fat adipocytes by increasing Runx1t1 expression (Sun et al., 2011; Zhang et al., 2015a). Any further studies will require investigation of this phenomenon and whether the loss of one copy of Runx1t1 is truly having an effect on fat deposits, either brown or white fat, or whether this is just simply a gain of muscle mass.
Chapter 5- Characterisation of the role of RUNX1T1 in neuroblastoma 224
Figure 5-35 Proposed model of RUNX1T1 in neuroblastoma development. High levels of RUNX1T1 in neural crest stem cells helps maintain these cells in an undifferentiated state. However, decreasing levels of RUNX1T1 in hyperplastic ganglia promotes both nerve cell differentiation and also cell deletion. High levels of MYCN through gene amplification act to keep RUNX1T1 at high levels to maintain an undifferentiated state and allow these cells to progress down the path of tumour development.
Chapter 5- Characterisation of the role of RUNX1T1 in neuroblastoma 225 It has been hypothesized that the Th-MYCN mice, with a background of 129/SvJ, must have a unique modifier that enables the MYCN transgene to be more penetrant than other strains (Teitz et al., 2011). Based on the results presented here one possible candidate for this modifier gene may be Runx1t1. Weiss and colleagues noted the increased tumour penetrance when the background of the Th-MYCN mice was 129/SvJ over FVB, which had very poor tumour incidence (Weiss et al., 1997). While investigating tumour susceptibility loci in the 129/SvJ mice, this laboratory mapped a primary tumour susceptibility loci to chromosome 10, however, they also found a secondary susceptibility loci in chromosome 4 of the mice (Hackett et al., 2014). This loci contains Runx1t1 and interestingly, the expression of Runx1t1 was upregulated in 129/SvJ mice compared to the FVB mice which had reduced tumour incidence. Additionally in the loxP-STOP-loxP-MYCN mice created by the Schulte laboratory (described in Section 1.3.3), on a mixed background of C57BL/6 and 129X1/SvJ, Runx1t1 was also found to be 80-150 fold increased in the tumours that formed over normal adrenal tissue in the same mouse background (Althoff et al., 2015). Thus the loss of one copy of Runx1t1 abrogating tumour development coupled with the high expression found in these two models implicate Runx1t1 as a potential modifier responsible for the increase tumour penetrance in the 129/SvJ Th-MYCN mice.
Modulation of RUNX1T1 in neuroblastoma cell lines produced differing results. Overexpression of Runx1t1, either wild-type or mutant, produced no discernible effect on colony formation, however, knock-down of this transcriptional co-repressor did lead to a reduction in colony number. Although the level of RUNX1T1 in BE(2)-C cells used in these experiments is markedly lower compared to other MYCN amplified cell lines, it may nevertheless be above a threshold level, such that overexpressing this gene in this line fails to alter the phenotype of these cells. Knock-down of RUNX1T1, however, did cause a reduction in colony number, indicating that even in an established tumour cell line, the loss of RUNX1T1 can have an effect. Given the inverse relationship between RUNX1T1 RNA and protein expression and outcome in primary human neuroblastoma, it is perhaps not surprising that RUNX1T1 has not been implicated previously in this disease, since large scale studies often use RNA platforms to identify and characterise new genes, and these studies tend to concentrate on the more highly expressed genes. Even initial studies within this thesis would have implicated RUNX1T1 as being involved in tumour initiation only, however, the high levels of RUNX1T1 protein in the
Chapter 5- Characterisation of the role of RUNX1T1 in neuroblastoma 226 tumour samples and the reduced colony formation in knock-down studies, suggest a role not only in initiation, but also potentially tumour progression.
This disconnect between the RNA and protein levels observed in human neuroblastoma warrants further study to understand the mechanism behind this. Although transcription and translation have traditionally been regarded as completely separate events, a recent study suggests that this may not be true (Slobodin et al., 2017). In this study, Slobodin and colleagues investigated the dynamic between the two processes and found that transcription can influence translation efficiency and that the N6-methylation of adenosines can modify and mediate this process. Even more recently, RUNX1T1 was hypothesized to be a ‘micro-RNA sponge’, where, in the context of AML involving the fusion of RUNX1T1 with the RUNX1 gene, there is an overabundance of the 3’ UTR of RUNX1T1. This overabundance removes micro-RNAs from their mRNA targets, releasing post-translational repression of the mRNA which further drives AML progression (Junge et al., 2017). RUNX1T1 has also been implicated in hypomethylation, chromatin accessibility and epigenetic programming (Boddicker et al., 2016; Ptasinska et al., 2012). Eukaryotic initiation factor 4E (eIF4E), which is involved in trafficking ribosomes to mRNA, can be used by tumour cells during oncogenesis. In a KRAS driven model of lung cancer, a 50% reduction in the level of eIF4E, by using heterozygote eIF4E mice, was able to reduce the tumour burden without affecting normal development (Truitt et al., 2015). This is similar to the effect in this study where a 50% reduction in Runx1t1 was able to completely alter tumour development. Any of these mechanisms, or a combination of them, could be responsible for this phenomena and further investigation is required.
Overall, these studies have, for the first time, provided compelling evidence of a key role for RUNX1T1 not only in neuroblastoma initiation, but also in the progression of this disease, and have opened up new avenues for both the study and also the development of a potential new therapeutic treatment approach.
Chapter 5- Characterisation of the role of RUNX1T1 in neuroblastoma 227 CHAPTER 6 CONCLUSIONS AND FUTURE DIRECTIONS
The outcome for neuroblastoma remains dismal despite intensive research and multi- modal therapy. Relapse in high risk disease is common, and whilst there are new agents and clinical trials being used in an attempt to improve the outcomes of these children, they are often faced with resistant disease and long-term side effects from the treatment if they do survive.
This study undertook an ENU mutagenesis screen with the goal of identifying MYCN- associated drivers of this disease, and therefore new avenues for therapeutic targeting. A well-characterised mouse model of neuroblastoma was used, the Th-MYCN mouse, where all mice homozygous for the MYCN transgene develop neuroblastoma by 7 weeks of age, and looked for mutations that affected tumour development in the offspring of ENU treated mice. As result of this study, through a series of backcrossing, mapping and sequencing, two candidate genes were identified where a single point mutation is capable of delaying or abrogating tumour development. Whilst the confirmation of the Rnf121 gene was outside the scope of this thesis, the role of Runx1t1 in neuroblastoma was investigated.
To confirm the decrease in colony formation of shRNA knock-down cells observed in vitro, and therefore the potential decrease in tumorigencity in vivo, future study will involve taking RUNX1T1 altered cells and analysing tumour growth following xenografting in mice. In addition to the shRNA lines created and described previously, CRISPR/Cas9 heterozygote and homozygote knock-out and 1590 point mutation lines are being developed for in vitro and in vivo investigation. Thirdly, a multi-potent neural crest progenitor cell, JoMa1, transformed by transfection of MYCN, and capable of forming neuroblastoma tumours in vivo (Schulte et al., 2013), will be manipulated to express the 1590 point mutation and allografted into mice. These three studies will provide valuable insight into the tumorigenic potential of loss of RUNX1T1 in cell lines xenografted into mice, and whether early progenitor cells transfected with mutant RUNX1T1 can form tumours.
Chapter 6- Conclusions and future directions 228 In terms of mouse biology and confirmation of the role of Runx1t1, the most definitive experiment was the cross of the Th-MYCN mouse with the knock-out mouse model of Runx1t1. Loss of a single copy of Runx1t1 resulted in almost complete abrogation of tumour development in mice homozygous for the MYCN transgene, and a 10-fold decrease in tumour development for those hemizygous for MYCN. A conditional Runx1t1 knock-out mouse is being developed, that when crossed with a Cre expressing mouse model under the control of tyrosine hydroxylase, will delete Runx1t1 in the developing neural crest cells, the origin of neuroblastoma. This model is estrogen responsive and when fed tamoxifen at various time points after birth, deletes Runx1t1. This will allow us to overcome the issues created following complete knock-out of Runx1t1 pre-birth. This model will then be crossed with the Th-MYCN mouse model and when Runx1t1 is deleted at weekly intervals post-birth, will inform us as to the crucial timing of Runx1t1 expression in the development of neuroblastoma.
Since Runx1t1 expression appears to be critical early in development, studies in both the ganglia and embryos are planned. RNA sequencing profiles in the 1590 and Th-MYCN x Runx1t1 mice will create an RNA signature and determine pathways that involve Runx1t1 in the context of neuroblastoma. For embryo studies, we plan to confirm Runx1t1 expression patterns in whole mount embryos using optical projection tomography (Shi et al., 2016) as well as traditional staining of Runx1t1 in early time points.
In addition, confirmation of in vivo findings in a second animal model would indicate the robustness of the study. A zebra fish model of high-risk neuroblastoma, which is MYCN-driven, could provide more evidence of the role of Runx1t1 (Zhang et al., 2017). Mutant Runx1t1 can be targeted to the adrenal gland analogue of the transgenic fish. As the construct can be GFP-tagged, the fish can be easily monitored for neuroblastoma development and the neuroblasts expressing wild-type or mutant Runx1t1 sorted by flow cytometry and studied in vitro.
In the human disease, low levels of RUNX1T1 transcript were prognostic of poor outcome, however, it was high levels of the protein, particularly in the MYCN amplified patients, that were prognostic of worse outcome. Elucidating the precise mechanism responsible for generating increased RUNX1T1 protein levels will be an important goal
Chapter 6- Conclusions and future directions 229 since this will undoubtedly open up new avenues of research as well as the potential for new therapeutic opportunities.
Identification of genes involved in neuroblastoma biology and therapeutic inhibition of the genes identified in this screen was the original aim of the study. No known inhibitors of RUNX1T1 currently exist. There has been some work performed in the context of the translocation in AML, however, little is known about the direct effects on RUNX1T1 rather than the fusion protein. Targeting of the histone deacetylases within the RUNX1T1 complex may provide some therapeutic benefit. The pan-HDAC inhibitor, panobinostat, has been successfully used to treat the Th-MYCN mouse model, and interestingly, RNA sequencing on the tumour after a single dose of panobinostat demonstrated 2-fold downregulation of Runx1t1 (Waldeck et al., 2016). Olaparib, an inhibitor of the poly (ADP-ribose) polymerase protein family, involved in DNA repair and programmed cell death also is capable of extending life of mice engrafted with RUNX1:RUNX1T1 translocated cells and may provide another option for targeting RUNX1T1 (Esposito et al., 2015). Small molecule library screening to find candidates that target RUNX1T1 specifically would also be useful, however, these types of undertakings can take years to establish, characterise and develop until a drug is ready for the market.
Finally, although mutagenesis studies should not be undertaken lightly, given the cost and time involved to carry them out, this study has shown that ENU mutagenesis can be a powerful biological tool when used with a robust disease model. Further utilisation of this approach will undoubtedly help identify other novel genes contributing to MYCN- associated neuroblastoma and provide valuable insight in how best to target, and ultimately prevent, this aggressive childhood disease.
Chapter 6- Conclusions and future directions 230 REFERENCES
Acevedo-Arozena, A., Wells, S., Potter, P., Kelly, M., Cox, R.D., and Brown, S.D. (2008). ENU mutagenesis, a way forward to understand gene function. Annual review of genomics and human genetics 9, 49-69.
Ahn, M.Y., Huang, G., Bae, S.C., Wee, H.J., Kim, W.Y., and Ito, Y. (1998). Negative regulation of granulocytic differentiation in the myeloid precursor cell line 32Dcl3 by ear-2, a mammalian homolog of Drosophila seven-up, and a chimeric leukemogenic gene, AML1/ETO. Proceedings of the National Academy of Sciences of the United States of America 95, 1812-1817.
AIHW (2009). A Picture of Australia's Children (Cat. no. PHE 112. Canberra: AIHW.).
Alishahi, A., Koyano-Nakagawa, N., and Nakagawa, Y. (2009). Regional expression of MTG genes in the developing mouse central nervous system. Developmental dynamics : an official publication of the American Association of Anatomists 238, 2095-2102.
Althoff, K., Beckers, A., Bell, E., Nortmeyer, M., Thor, T., Sprussel, A., Lindner, S., De Preter, K., Florin, A., Heukamp, L.C., Klein-Hitpass, L., Astrahantseff, K., Kumps, C., Speleman, F., Eggert, A., Westermann, F., Schramm, A., and Schulte, J.H. (2015). A Cre-conditional MYCN-driven neuroblastoma mouse model as an improved tool for preclinical studies. Oncogene 34, 3357-3368.
Amann, J.M., Chyla, B.J., Ellis, T.C., Martinez, A., Moore, A.C., Franklin, J.L., McGhee, L., Meyers, S., Ohm, J.E., Luce, K.S., Ouelette, A.J., Washington, M.K., Thompson, M.A., King, D., Gautam, S., Coffey, R.J., Whitehead, R.H., and Hiebert, S.W. (2005). Mtgr1 is a transcriptional corepressor that is required for maintenance of the secretory cell lineage in the small intestine. Molecular and cellular biology 25, 9576-9585.
Amann, J.M., Nip, J., Strom, D.K., Lutterbach, B., Harada, H., Lenny, N., Downing, J.R., Meyers, S., and Hiebert, S.W. (2001). ETO, a target of t(8;21) in acute leukemia, makes distinct contacts with multiple histone deacetylases and binds mSin3A through its oligomerization domain. Molecular and cellular biology 21, 6470-6483.
References 231 Ambros, I.M., Zellner, A., Roald, B., Amann, G., Ladenstein, R., Printz, D., Gadner, H., and Ambros, P.F. (1996). Role of Ploidy, Chromosome 1p, and Schwann Cells in the Maturation of Neuroblastoma. New England Journal of Medicine 334, 1505-1511.
Ambros, P.F., Ambros, I.M., Brodeur, G.M., Haber, M., Khan, J., Nakagawara, A., Schleiermacher, G., Speleman, F., Spitz, R., London, W.B., Cohn, S.L., Pearson, A.D., and Maris, J.M. (2009). International consensus for neuroblastoma molecular diagnostics: report from the International Neuroblastoma Risk Group (INRG) Biology Committee. British journal of cancer 100, 1471-1482.
Andrews, T.D., Whittle, B., Field, M.A., Balakishnan, B., Zhang, Y., Shao, Y., Cho, V., Kirk, M., Singh, M., Xia, Y., Hager, J., Winslade, S., Sjollema, G., Beutler, B., Enders, A., and Goodnow, C.C. (2012). Massively parallel sequencing of the mouse exome to accurately identify rare, induced mutations: an immediate source for thousands of new mouse models. Open biology 2, 120061.
Asada, Y., Varnum, D.S., Frankel, W.N., and Nadeau, J.H. (1994). A mutation in the Ter gene causing increased susceptibility to testicular teratomas maps to mouse chromosome 18. Nature genetics 6, 363-368.
Asgharzadeh, S., Pique-Regi, R., Sposto, R., Wang, H., Yang, Y., Shimada, H., Matthay, K., Buckley, J., Ortega, A., and Seeger, R.C. (2006). Prognostic significance of gene expression profiles of metastatic neuroblastomas lacking MYCN gene amplification. Journal of the National Cancer Institute 98, 1193-1203.
Attiyeh, E.F., London, W.B., Mosse, Y.P., Wang, Q., Winter, C., Khazi, D., McGrady, P.W., Seeger, R.C., Look, A.T., Shimada, H., Brodeur, G.M., Cohn, S.L., Matthay, K.K., and Maris, J.M. (2005). Chromosome 1p and 11q deletions and outcome in neuroblastoma. The New England journal of medicine 353, 2243-2253.
Ayer, D.E., Kretzner, L., and Eisenman, R.N. (1993). Mad: a heterodimeric partner for Max that antagonizes Myc transcriptional activity. Cell 72, 211-222.
Baade, P.D., Youlden, D.R., Valery, P.C., Hassall, T., Ward, L., Green, A.C., and Aitken, J.F. (2010). Trends in incidence of childhood cancer in Australia, 1983-2006. British journal of cancer 102, 620-626.
References 232 Baby, N., Li, Y., Ling, E.A., Lu, J., and Dheen, S.T. (2014). Runx1t1 (runt-related transcription factor 1; translocated to, 1) epigenetically regulates the proliferation and nitric oxide production of microglia. PloS one 9, e89326.
Bagatell, R., London, W.B., Wagner, L.M., Voss, S.D., Stewart, C.F., Maris, J.M., Kretschmar, C., and Cohn, S.L. (2011). Phase II study of irinotecan and temozolomide in children with relapsed or refractory neuroblastoma: a Children's Oncology Group study. J Clin Oncol 29, 208-213.
Baker, D.L., Schmidt, M.L., Cohn, S.L., Maris, J.M., London, W.B., Buxton, A., Stram, D., Castleberry, R.P., Shimada, H., Sandler, A., Shamberger, R.C., Look, A.T., Reynolds, C.P., Seeger, R.C., and Matthay, K.K. (2010). Outcome after reduced chemotherapy for intermediate-risk neuroblastoma. The New England journal of medicine 363, 1313-1323.
Barbaric, I., Wells, S., Russ, A., and Dear, T.N. (2007). Spectrum of ENU-induced mutations in phenotype-driven and gene-driven screens in the mouse. Environmental and molecular mutagenesis 48, 124-142.
Bassiri, H., Benavides, A., Haber, M., Gilmour, S.K., Norris, M.D., and Hogarty, M.D. (2015). Translational development of difluoromethylornithine (DFMO) for the treatment of neuroblastoma. Translational pediatrics 4, 226-238.
Beltran, H., Rickman, D.S., Park, K., Chae, S.S., Sboner, A., Macdonald, T.Y., Wang, Y., Sheikh, K.L., Terry, S., Tagawa, S.T., Dhir, R., Nelson, J.B., de la Taille, A., Allory, Y., Gerstein, M.B., Perner, S., Pienta, K.J., Chinnaiyan, A.M., Collins, C.C., Gleave, M.E., Demichelis, F., Nanus, D.M., and Rubin, M.A. (2011). Molecular Characterization of Neuroendocrine Prostate Cancer and Identification of New Drug Targets. Cancer discovery 1, 487-495.
Belur, L.R., Podetz-Pedersen, K.M., Sorenson, B.S., Hsu, A.H., Parker, J.B., Carlson, C.S., Saltzman, D.A., Ramakrishnan, S., and McIvor, R.S. (2011). Inhibition of angiogenesis and suppression of colorectal cancer metastatic to the liver using the Sleeping Beauty Transposon System. Molecular cancer 10, 14.
Berry, T., Luther, W., Bhatnagar, N., Jamin, Y., Poon, E., Sanda, T., Pei, D., Sharma, B., Vetharoy, W.R., Hallsworth, A., Ahmad, Z., Barker, K., Moreau, L., Webber, H., Wang, W., Liu, Q., Perez-Atayde, A., Rodig, S., Cheung, N.K., Raynaud, F., Hallberg,
References 233 B., Robinson, S.P., Gray, N.S., Pearson, A.D., Eccles, S.A., Chesler, L., and George, R.E. (2012). The ALK(F1174L) mutation potentiates the oncogenic activity of MYCN in neuroblastoma. Cancer cell 22, 117-130.
Boddicker, R.L., Koltes, J.E., Fritz-Waters, E.R., Koesterke, L., Weeks, N., Yin, T., Mani, V., Nettleton, D., Reecy, J.M., Baumgard, L.H., Spencer, J.D., Gabler, N.K., and Ross, J.W. (2016). Genome-wide methylation profile following prenatal and postnatal dietary omega-3 fatty acid supplementation in pigs. Animal genetics 47, 658-671.
Bown, N., Cotterill, S., Łastowska, M., O'Neill, S., Pearson, A.D.J., Plantaz, D., Meddeb, M., Danglot, G., Brinkschmidt, C., Christiansen, H., Laureys, G., Nicholson, J., Bernheim, A., Betts, D.R., Vandesompele, J., Van Roy, N., and Speleman, F. (1999). Gain of Chromosome Arm 17q and Adverse Outcome in Patients with Neuroblastoma. New England Journal of Medicine 340, 1954-1961.
Brannan, C.I., Perkins, A.S., Vogel, K.S., Ratner, N., Nordlund, M.L., Reid, S.W., Buchberg, A.M., Jenkins, N.A., Parada, L.F., and Copeland, N.G. (1994). Targeted disruption of the neurofibromatosis type-1 gene leads to developmental abnormalities in heart and various neural crest-derived tissues. Genes & development 8, 1019-1029.
Braun, T., and Gardin, C. (2017). Investigational BET bromodomain protein inhibitors in early stage clinical trials for acute myelogenous leukemia (AML). Expert opinion on investigational drugs, 1-9.
Brennan, B.M., Huynh, M.T., Rabah, M.A., Shaw, H.E., Bisaillon, J.J., Radden, L.A., 2nd, Nguyen, T.V., and King, T.R. (2015). The mouse wellhaarig (we) mutations result from defects in epidermal-type transglutaminase 3 (Tgm3). Molecular genetics and metabolism 116, 187-191.
Breslow, N., and McCann, B. (1971). Statistical estimation of prognosis for children with neuroblastoma. Cancer Res 31, 2098-2103.
Brisse, H.J., McCarville, M.B., Granata, C., Krug, K.B., Wootton-Gorges, S.L., Kanegawa, K., Giammarile, F., Schmidt, M., Shulkin, B.L., Matthay, K.K., Lewington, V.J., Sarnacki, S., Hero, B., Kaneko, M., London, W.B., Pearson, A.D., Cohn, S.L., and Monclair, T. (2011). Guidelines for imaging and staging of neuroblastic tumors: consensus report from the international neuroblastoma risk group project. Radiology 261, 243-257.
References 234 Brodeur, G., and Castleberry, R. (1997). Neuroblastoma. In: Pizzo PA, Poplack DG, editors. Principles and Practice of Pediatric Oncology (Philadelphia, PA: J. B. Lippincott Co.).
Brodeur, G.M. (2003). Neuroblastoma: biological insights into a clinical enigma. Nature reviews Cancer 3, 203-216.
Brodeur, G.M., Nakagawara, A., Yamashiro, D.J., Ikegaki, N., Liu, X.G., Azar, C.G., Lee, C.P., and Evans, A.E. (1997). Expression of TrkA, TrkB and TrkC in human neuroblastomas. Journal of neuro-oncology 31, 49-55.
Brodeur, G.M., Pritchard, J., Berthold, F., Carlsen, N.L., Castel, V., Castelberry, R.P., De Bernardi, B., Evans, A.E., Favrot, M., Hedborg, F., and et al. (1993). Revisions of the international criteria for neuroblastoma diagnosis, staging, and response to treatment. J Clin Oncol 11, 1466-1477.
Brodeur, G.M., Seeger, R.C., Barrett, A., Berthold, F., Castleberry, R.P., D'Angio, G., De Bernardi, B., Evans, A.E., Favrot, M., Freeman, A.I., and et al. (1988). International criteria for diagnosis, staging, and response to treatment in patients with neuroblastoma. J Clin Oncol 6, 1874-1881.
Buac, K., Watkins-Chow, D.E., Loftus, S.K., Larson, D.M., Incao, A., Gibney, G., and Pavan, W.J. (2008). A Sox10 expression screen identifies an amino acid essential for Erbb3 function. PLoS genetics 4, e1000177.
Burkhart, C., Fleyshman, D., Kohrn, R., Commane, M., Garrigan, J., Kurbatov, V., Toshkov, I., Ramachandran, R., Martello, L., and Gurova, K.V. (2014). Curaxin CBL0137 eradicates drug resistant cancer stem cells and potentiates efficacy of gemcitabine in preclinical models of pancreatic cancer. Oncotarget 5, 11038-11053.
Burkhart, C.A., Cheng, A.J., Madafiglio, J., Kavallaris, M., Mili, M., Marshall, G.M., Weiss, W.A., Khachigian, L.M., Norris, M.D., and Haber, M. (2003). Effects of MYCN antisense oligonucleotide administration on tumorigenesis in a murine model of neuroblastoma. Journal of the National Cancer Institute 95, 1394-1403.
Burkhart, C.A., Norris, M.D., and Haber, M. (2002). A simple method for the isolation of genomic DNA from mouse tail free of real-time PCR inhibitors. Journal of biochemical and biophysical methods 52, 145-149.
References 235 Burkhart, C.A., Watt, F., Murray, J., Pajic, M., Prokvolit, A., Xue, C., Flemming, C., Smith, J., Purmal, A., Isachenko, N., Komarov, P.G., Gurova, K.V., Sartorelli, A.C., Marshall, G.M., Norris, M.D., Gudkov, A.V., and Haber, M. (2009). Small-molecule multidrug resistance-associated protein 1 inhibitor reversan increases the therapeutic index of chemotherapy in mouse models of neuroblastoma. Cancer Res 69, 6573-6580.
Calabi, F., Pannell, R., and Pavloska, G. (2001). Gene targeting reveals a crucial role for MTG8 in the gut. Molecular and cellular biology 21, 5658-5666.
Carlsen, N.L., Christensen, I.J., Schroeder, H., Bro, P.V., Hesselbjerg, U., Jensen, K.B., and Nielsen, O.H. (1986). Prognostic value of different staging systems in neuroblastomas and completeness of tumour excision. Archives of disease in childhood 61, 832-842.
Carlson, L.M., Rasmuson, A., Idborg, H., Segerstrom, L., Jakobsson, P.J., Sveinbjornsson, B., and Kogner, P. (2013). Low-dose aspirin delays an inflammatory tumor progression in vivo in a transgenic mouse model of neuroblastoma. Carcinogenesis 34, 1081-1088.
Carmely, A., Meirow, D., Peretz, A., Albeck, M., Bartoov, B., and Sredni, B. (2009). Protective effect of the immunomodulator AS101 against cyclophosphamide-induced testicular damage in mice. Hum Reprod 24, 1322-1329.
Caron, H., van Sluis, P., de Kraker, J., Bökkerink, J., Egeler, M., Laureys, G., Slater, R., Westerveld, A., Voûte, P.A., and Versteeg, R. (1996). Allelic Loss of Chromosome 1p as a Predictor of Unfavorable Outcome in Patients with Neuroblastoma. New England Journal of Medicine 334, 225-230.
Carpinelli, M.R., Burt, R.A., and Kile, B.T. (2012). Genetic Modifier Screens in Mice. Current protocols in mouse biology 2, 75-87.
Carpinelli, M.R., Hilton, D.J., Metcalf, D., Antonchuk, J.L., Hyland, C.D., Mifsud, S.L., Di Rago, L., Hilton, A.A., Willson, T.A., Roberts, A.W., Ramsay, R.G., Nicola, N.A., and Alexander, W.S. (2004). Suppressor screen in Mpl-/- mice: c-Myb mutation causes supraphysiological production of platelets in the absence of thrombopoietin signaling. Proceedings of the National Academy of Sciences of the United States of America 101, 6553-6558.
References 236 Carter, D.R., Murray, J., Cheung, B.B., Gamble, L., Koach, J., Tsang, J., Sutton, S., Kalla, H., Syed, S., Gifford, A.J., Issaeva, N., Biktasova, A., Atmadibrata, B., Sun, Y., Sokolowski, N., Ling, D., Kim, P.Y., Webber, H., Clark, A., Ruhle, M., Liu, B., Oberthuer, A., Fischer, M., Byrne, J., Saletta, F., Thwe le, M., Purmal, A., Haderski, G., Burkhart, C., Speleman, F., De Preter, K., Beckers, A., Ziegler, D.S., Liu, T., Gurova, K.V., Gudkov, A.V., Norris, M.D., Haber, M., and Marshall, G.M. (2015). Therapeutic targeting of the MYC signal by inhibition of histone chaperone FACT in neuroblastoma. Science translational medicine 7, 312ra176.
Castel, V., Garcia-Miguel, P., Canete, A., Melero, C., Navajas, A., Ruiz-Jimenez, J.I., Navarro, S., and Badal, M.D. (1999). Prospective evaluation of the International Neuroblastoma Staging System (INSS) and the International Neuroblastoma Response Criteria (INRC) in a multicentre setting. Eur J Cancer 35, 606-611.
Cattanach, B.M. (1974). Spermatogonial stem cell killing in the mouse following single and fractionated x-ray doses, as assessed by length of sterile period. Mutation research 25, 53-62.
Cattanach, B.M., and Moseley, H. (1974). Sterile period, translocation and specific locus mutation in the mouse following fractionated x-ray treatments with different fractionation intervals. Mutation research 25, 63-72.
Charron, J., Malynn, B.A., Fisher, P., Stewart, V., Jeannotte, L., Goff, S.P., Robertson, E.J., and Alt, F.W. (1992). Embryonic lethality in mice homozygous for a targeted disruption of the N-myc gene. Genes & development 6, 2248-2257.
Chayka, O., Corvetta, D., Dews, M., Caccamo, A.E., Piotrowska, I., Santilli, G., Gibson, S., Sebire, N.J., Himoudi, N., Hogarty, M.D., Anderson, J., Bettuzzi, S., Thomas-Tikhonenko, A., and Sala, A. (2009). Clusterin, a haploinsufficient tumor suppressor gene in neuroblastomas. Journal of the National Cancer Institute 101, 663- 677.
Chen, J., Hackett, C.S., Zhang, S., Song, Y.K., Bell, R.J., Molinaro, A.M., Quigley, D.A., Balmain, A., Song, J.S., Costello, J.F., Gustafson, W.C., Van Dyke, T., Kwok, P.Y., Khan, J., and Weiss, W.A. (2015). The genetics of splicing in neuroblastoma. Cancer discovery 5, 380-395.
References 237 Chen, Z., Lin, Y., Barbieri, E., Burlingame, S., Hicks, J., Ludwig, A., and Shohet, J.M. (2009). Mdm2 deficiency suppresses MYCN-Driven neuroblastoma tumorigenesis in vivo. Neoplasia 11, 753-762.
Cheng, A.J., Cheng, N.C., Ford, J., Smith, J., Murray, J.E., Flemming, C., Lastowska, M., Jackson, M.S., Hackett, C.S., Weiss, W.A., Marshall, G.M., Kees, U.R., Norris, M.D., and Haber, M. (2007). Cell lines from MYCN transgenic murine tumours reflect the molecular and biological characteristics of human neuroblastoma. Eur J Cancer 43, 1467-1475.
Cheng, Y., Gaughan, J., Midic, U., Han, Z., Liang, C.G., Patel, B.G., and Latham, K.E. (2013). Systems genetics implicates cytoskeletal genes in oocyte control of cloned embryo quality. Genetics 193, 877-896.
Chesler, L., Goldenberg, D.D., Collins, R., Grimmer, M., Kim, G.E., Tihan, T., Nguyen, K., Yakovenko, S., Matthay, K.K., and Weiss, W.A. (2008). Chemotherapy-induced apoptosis in a transgenic model of neuroblastoma proceeds through p53 induction. Neoplasia 10, 1268-1274.
Chesler, L., Goldenberg, D.D., Seales, I.T., Satchi-Fainaro, R., Grimmer, M., Collins, R., Struett, C., Nguyen, K.N., Kim, G., Tihan, T., Bao, Y., Brekken, R.A., Bergers, G., Folkman, J., and Weiss, W.A. (2007). Malignant progression and blockade of angiogenesis in a murine transgenic model of neuroblastoma. Cancer Res 67, 9435- 9442.
Ciccarone, V., Spengler, B.A., Meyers, M.B., Biedler, J.L., and Ross, R.A. (1989). Phenotypic diversification in human neuroblastoma cells: expression of distinct neural crest lineages. Cancer Res 49, 219-225.
Cohn, S.L., Pearson, A.D., London, W.B., Monclair, T., Ambros, P.F., Brodeur, G.M., Faldum, A., Hero, B., Iehara, T., Machin, D., Mosseri, V., Simon, T., Garaventa, A., Castel, V., and Matthay, K.K. (2009). The International Neuroblastoma Risk Group (INRG) classification system: an INRG Task Force report. J Clin Oncol 27, 289-297.
Cole, K.A., Attiyeh, E.F., Mosse, Y.P., Laquaglia, M.J., Diskin, S.J., Brodeur, G.M., and Maris, J.M. (2008). A functional screen identifies miR-34a as a candidate neuroblastoma tumor suppressor gene. Molecular cancer research : MCR 6, 735-742.
References 238 Cole, M.D. (2014). MYC association with cancer risk and a new model of MYC- mediated repression. Cold Spring Harbor perspectives in medicine 4, a014316.
Collier, L.S., and Largaespada, D.A. (2007). Transposons for cancer gene discovery: Sleeping Beauty and beyond. Genome biology 8 Suppl 1, S15.
Concepcion, D., Seburn, K.L., Wen, G., Frankel, W.N., and Hamilton, B.A. (2004). Mutation rate and predicted phenotypic target sizes in ethylnitrosourea-treated mice. Genetics 168, 953-959.
Cordes, S.P. (2005). N-ethyl-N-nitrosourea mutagenesis: boarding the mouse mutant express. Microbiology and molecular biology reviews : MMBR 69, 426-439.
Cox, B.D., and Lyons, M.F. (1975). The induction by X-rays of chromosome aberrations in germ cells of male guinea-pigs, golden hamsters and rabbits. I. Dose response in post-meiotic stages. Mutation research 29, 93-109.
Dannenberg, J.H., David, G., Zhong, S., van der Torre, J., Wong, W.H., and Depinho, R.A. (2005). mSin3A corepressor regulates diverse transcriptional networks governing normal and neoplastic growth and survival. Genes & development 19, 1581-1595.
Darom, A., Bening-Abu-Shach, U., and Broday, L. (2010). RNF-121 is an endoplasmic reticulum-membrane E3 ubiquitin ligase involved in the regulation of beta-integrin. Molecular biology of the cell 21, 1788-1798.
Daudigeos-Dubus, E., L, L.E.D., Rouffiac, V., Bawa, O., Leguerney, I., Opolon, P., Vassal, G., and Geoerger, B. (2014). Establishment and characterization of new orthotopic and metastatic neuroblastoma models. In Vivo 28, 425-434.
Davis, J.N., McGhee, L., and Meyers, S. (2003). The ETO (MTG8) gene family. Gene 303, 1-10.
De Braekeleer, M., Morel, F., Le Bris, M.J., Herry, A., and Douet-Guilbert, N. (2005). The MLL gene and translocations involving chromosomal band 11q23 in acute leukemia. Anticancer research 25, 1931-1944.
DeKelver, R.C., Yan, M., Ahn, E.Y., Shia, W.J., Speck, N.A., and Zhang, D.E. (2013). Attenuation of AML1-ETO cellular dysregulation correlates with increased leukemogenic potential. Blood 121, 3714-3717.
References 239 Ding, S., Wu, X., Li, G., Han, M., Zhuang, Y., and Xu, T. (2005). Efficient transposition of the piggyBac (PB) transposon in mammalian cells and mice. Cell 122, 473-483.
Downs, K.M., Martin, G.R., and Bishop, J.M. (1989). Contrasting patterns of myc and N-myc expression during gastrulation of the mouse embryo. Genes & development 3, 860-869.
Du, X., Tabeta, K., Hoebe, K., Liu, H., Mann, N., Mudd, S., Crozat, K., Sovath, S., Gong, X., and Beutler, B. (2004). Velvet, a dominant Egfr mutation that causes wavy hair and defective eyelid development in mice. Genetics 166, 331-340.
Eberhart, C.G., Kratz, J., Wang, Y., Summers, K., Stearns, D., Cohen, K., Dang, C.V., and Burger, P.C. (2004). Histopathological and molecular prognostic markers in medulloblastoma: c-myc, N-myc, TrkC, and anaplasia. Journal of neuropathology and experimental neurology 63, 441-449.
Edderkaoui, B., Kesavan, C., Baylink, D.J., Wergedal, J.E., Srivastava, A.K., and Mohan, S. (2013). ENU mutation mapped to a distal region of chromosome 11 is a major determinant of bone size. Physiological genomics 45, 1222-1228.
Erdelyi, D.J., Elliott, M., and Phillips, B. (2011). Urine catecholamines in paediatrics. Archives of disease in childhood Education and practice edition 96, 107-111.
Erichsen, D.A., Armstrong, M.B., and Wechsler, D.S. (2015). Mxi1 and mxi1-0 antagonize N-myc function and independently mediate apoptosis in neuroblastoma. Translational oncology 8, 65-74.
Esiashvili, N., Anderson, C., and Katzenstein, H.M. (2009). Neuroblastoma. Current Problems in Cancer 33, 333-360.
Esposito, M.T., Zhao, L., Fung, T.K., Rane, J.K., Wilson, A., Martin, N., Gil, J., Leung, A.Y., Ashworth, A., and So, C.W. (2015). Synthetic lethal targeting of oncogenic transcription factors in acute leukemia by PARP inhibitors. Nature medicine 21, 1481- 1490.
Evans, A.E., D'Angio, G.J., and Randolph, J. (1971). A proposed staging for children with neuroblastoma. Children's cancer study group A. Cancer 27, 374-378.
References 240 Evans, A.E., D'Angio, G.J., Sather, H.N., de Lorimier, A.A., Dalton, A., Ungerleider, R.S., Finklestein, J.Z., and Hammond, G.D. (1990). A comparison of four staging systems for localized and regional neuroblastoma: a report from the Childrens Cancer Study Group. J Clin Oncol 8, 678-688.
Falkenberg, K.J., and Johnstone, R.W. (2014). Histone deacetylases and their inhibitors in cancer, neurological diseases and immune disorders. Nat Rev Drug Discov 13, 673- 691.
Farhana, L., Dawson, M.I., and Fontana, J.A. (2015). Down regulation of miR-202 modulates Mxd1 and Sin3A repressor complexes to induce apoptosis of pancreatic cancer cells. Cancer biology & therapy 16, 115-124.
Feinstein, P.G., Kornfeld, K., Hogness, D.S., and Mann, R.S. (1995). Identification of homeotic target genes in Drosophila melanogaster including nervy, a proto-oncogene homologue. Genetics 140, 573-586.
Fitch, K.R., McGowan, K.A., van Raamsdonk, C.D., Fuchs, H., Lee, D., Puech, A., Herault, Y., Threadgill, D.W., Hrabe de Angelis, M., and Barsh, G.S. (2003). Genetics of dark skin in mice. Genes & development 17, 214-228.
Flavell, R.A., Sabo, D.L., Bandle, E.F., and Weissmann, C. (1974). Site-directed mutagenesis: generation of an extracistronic mutation in bacteriophage Q beta RNA. Journal of molecular biology 89, 255-272.
Flavell, R.A., Sabo, D.L., Bandle, E.F., and Weissmann, C. (1975). Site-directed mutagenesis: effect of an extracistronic mutation on the in vitro propagation of bacteriophage Qbeta RNA. Proceedings of the National Academy of Sciences of the United States of America 72, 367-371.
Foley, J., Cohn, S.L., Salwen, H.R., Chagnovich, D., Cowan, J., Mason, K.L., and Parysek, L.M. (1991). Differential expression of N-myc in phenotypically distinct subclones of a human neuroblastoma cell line. Cancer Res 51, 6338-6345.
Fujita, T., Igarashi, J., Okawa, E.R., Gotoh, T., Manne, J., Kolla, V., Kim, J., Zhao, H., Pawel, B.R., London, W.B., Maris, J.M., White, P.S., and Brodeur, G.M. (2008). CHD5, a tumor suppressor gene deleted from 1p36.31 in neuroblastomas. Journal of the National Cancer Institute 100, 940-949.
References 241 Fung, W., Hasan, M.Y., Loh, A.H., Chua, J.H., Hwee, Y.M., Knight, L., Sek, H.W., Yoke, C.M., Tew, S.W., Jacobsen, A.S., and Hon, C.C. (2011). Gene expression of TRK neurotrophin receptors in advanced neuroblastomas in Singapore--a pilot study. Pediatric hematology and oncology 28, 571-578.
Gamsjaeger, R., O'Connell, M.R., Cubeddu, L., Shepherd, N.E., Lowry, J.A., Kwan, A.H., Vandevenne, M., Swanton, M.K., Matthews, J.M., and Mackay, J.P. (2013). A structural analysis of DNA binding by myelin transcription factor 1 double zinc fingers. The Journal of biological chemistry 288, 35180-35191.
Gasparian, A.V., Burkhart, C.A., Purmal, A.A., Brodsky, L., Pal, M., Saranadasa, M., Bosykh, D.A., Commane, M., Guryanova, O.A., Pal, S., Safina, A., Sviridov, S., Koman, I.E., Veith, J., Komar, A.A., Gudkov, A.V., and Gurova, K.V. (2011). Curaxins: anticancer compounds that simultaneously suppress NF-kappaB and activate p53 by targeting FACT. Science translational medicine 3, 95ra74.
Gershenson, D.M. (1988). Menstrual and reproductive function after treatment with combination chemotherapy for malignant ovarian germ cell tumors. J Clin Oncol 6, 270-275.
Gherardi, S., Valli, E., Erriquez, D., and Perini, G. (2013). MYCN-mediated transcriptional repression in neuroblastoma: the other side of the coin. Frontiers in oncology 3, 42.
Ghobadi, E., Moloudizargari, M., Asghari, M.H., and Abdollahi, M. (2016). The mechanisms of cyclophosphamide-induced testicular toxicity and the protective agents. Expert opinion on drug metabolism & toxicology, 1-12.
Goldsworthy, M., Hugill, A., Freeman, H., Horner, E., Shimomura, K., Bogani, D., Pieles, G., Mijat, V., Arkell, R., Bhattacharya, S., Ashcroft, F.M., and Cox, R.D. (2008). Role of the transcription factor sox4 in insulin secretion and impaired glucose tolerance. Diabetes 57, 2234-2244.
Gondo, Y. (2010). Now and future of mouse mutagenesis for human disease models. Journal of genetics and genomics = Yi chuan xue bao 37, 559-572.
References 242 Gonzales, N.M., and Palmer, A.A. (2014). Fine-mapping QTLs in advanced intercross lines and other outbred populations. Mammalian genome : official journal of the International Mammalian Genome Society 25, 271-292.
Goodson, M.L., Mengeling, B.J., Jonas, B.A., and Privalsky, M.L. (2011). Alternative mRNA splicing of corepressors generates variants that play opposing roles in adipocyte differentiation. The Journal of biological chemistry 286, 44988-44999.
Goodson, M.L., Young, B.M., Snyder, C.A., Schroeder, A.C., and Privalsky, M.L. (2014). Alteration of NCoR corepressor splicing in mice causes increased body weight and hepatosteatosis without glucose intolerance. Molecular and cellular biology 34, 4104-4114.
Goodspeed, T.H., and Olson, A.R. (1928). The Production of Variation in Nicotiana Species by X-Ray Treatment of Sex Cells. Proceedings of the National Academy of Sciences of the United States of America 14, 66-69.
Guo, C., White, P.S., Hogarty, M.D., Brodeur, G.M., Gerbing, R., Stram, D.O., and Maris, J.M. (2000). Deletion of 11q23 is a frequent event in the evolution of MYCN single-copy high-risk neuroblastomas. Medical and pediatric oncology 35, 544-546.
Guo, C., White, P.S., Weiss, M.J., Hogarty, M.D., Thompson, P.M., Stram, D.O., Gerbing, R., Matthay, K.K., Seeger, R.C., Brodeur, G.M., and Maris, J.M. (1999). Allelic deletion at 11q23 is common in MYCN single copy neuroblastomas. Oncogene 18, 4948-4957.
Gustafson, W.C., and Weiss, W.A. (2010). Myc proteins as therapeutic targets. Oncogene 29, 1249-1259.
Haber, M., Smith, J., Bordow, S.B., Flemming, C., Cohn, S.L., London, W.B., Marshall, G.M., and Norris, M.D. (2006). Association of high-level MRP1 expression with poor clinical outcome in a large prospective study of primary neuroblastoma. J Clin Oncol 24, 1546-1553.
Hackett, C.S., Hodgson, J.G., Law, M.E., Fridlyand, J., Osoegawa, K., de Jong, P.J., Nowak, N.J., Pinkel, D., Albertson, D.G., Jain, A., Jenkins, R., Gray, J.W., and Weiss, W.A. (2003). Genome-wide array CGH analysis of murine neuroblastoma reveals
References 243 distinct genomic aberrations which parallel those in human tumors. Cancer Res 63, 5266-5273.
Hackett, C.S., Quigley, D.A., Wong, R.A., Chen, J., Cheng, C., Song, Y.K., Wei, J.S., Pawlikowska, L., Bao, Y., Goldenberg, D.D., Nguyen, K., Gustafson, W.C., Rallapalli, S.K., Cho, Y.J., Cook, J.M., Kozlov, S., Mao, J.H., Van Dyke, T., Kwok, P.Y., Khan, J., Balmain, A., Fan, Q., and Weiss, W.A. (2014). Expression quantitative trait loci and receptor pharmacology implicate Arg1 and the GABA-A receptor as therapeutic targets in neuroblastoma. Cell reports 9, 1034-1046.
Hall, B., Limaye, A., and Kulkarni, A.B. (2009). Overview: generation of gene knockout mice. Current protocols in cell biology Chapter 19, Unit 19 12 19 12 11-17.
Hanahan, D., and Weinberg, R.A. (2011). Hallmarks of cancer: the next generation. Cell 144, 646-674.
Hann, H.W., Evans, A.E., Siegel, S.E., Wong, K.Y., Sather, H., Dalton, A., Hammond, D., and Seeger, R.C. (1985). Prognostic importance of serum ferritin in patients with Stages III and IV neuroblastoma: the Childrens Cancer Study Group experience. Cancer Res 45, 2843-2848.
Hansford, L.M., Thomas, W.D., Keating, J.M., Burkhart, C.A., Peaston, A.E., Norris, M.D., Haber, M., Armati, P.J., Weiss, W.A., and Marshall, G.M. (2004). Mechanisms of embryonal tumor initiation: distinct roles for MycN expression and MYCN amplification. Proceedings of the National Academy of Sciences of the United States of America 101, 12664-12669.
Hara, S., Tamano, M., Yamashita, S., Kato, T., Saito, T., Sakuma, T., Yamamoto, T., Inui, M., and Takada, S. (2015). Generation of mutant mice via the CRISPR/Cas9 system using FokI-dCas9. Scientific reports 5, 11221.
Haraguchi, S., and Nakagawara, A. (2009). A simple PCR method for rapid genotype analysis of the TH-MYCN transgenic mouse. PloS one 4, e6902.
Hatton, B.A., Knoepfler, P.S., Kenney, A.M., Rowitch, D.H., de Alboran, I.M., Olson, J.M., and Eisenman, R.N. (2006). N-myc is an essential downstream effector of Shh signaling during both normal and neoplastic cerebellar growth. Cancer research 66, 8655-8661.
References 244 Hayward, W.S., Neel, B.G., and Astrin, S.M. (1981). Activation of a cellular onc gene by promoter insertion in ALV-induced lymphoid leukosis. Nature 290, 475-480.
Heberle, H., Meirelles, G.V., da Silva, F.R., Telles, G.P., and Minghim, R. (2015). InteractiVenn: a web-based tool for the analysis of sets through Venn diagrams. BMC Bioinformatics 16, 169.
Helland, A., Anglesio, M.S., George, J., Cowin, P.A., Johnstone, C.N., House, C.M., Sheppard, K.E., Etemadmoghadam, D., Melnyk, N., Rustgi, A.K., Phillips, W.A., Johnsen, H., Holm, R., Kristensen, G.B., Birrer, M.J., Pearson, R.B., Borresen-Dale, A.L., Huntsman, D.G., deFazio, A., Creighton, C.J., Smyth, G.K., and Bowtell, D.D. (2011). Deregulation of MYCN, LIN28B and LET7 in a molecular subtype of aggressive high-grade serous ovarian cancers. PloS one 6, e18064.
Hendel, A., Fine, E.J., Bao, G., and Porteus, M.H. (2015). Quantifying on- and off- target genome editing. Trends in biotechnology 33, 132-140.
Henderson, M.J., Haber, M., Porro, A., Munoz, M.A., Iraci, N., Xue, C., Murray, J., Flemming, C.L., Smith, J., Fletcher, J.I., Gherardi, S., Kwek, C.K., Russell, A.J., Valli, E., London, W.B., Buxton, A.B., Ashton, L.J., Sartorelli, A.C., Cohn, S.L., Schwab, M., Marshall, G.M., Perini, G., and Norris, M.D. (2011). ABCC multidrug transporters in childhood neuroblastoma: clinical and biological effects independent of cytotoxic drug efflux. Journal of the National Cancer Institute 103, 1236-1251.
Henrich, K.O., Bauer, T., Schulte, J., Ehemann, V., Deubzer, H., Gogolin, S., Muth, D., Fischer, M., Benner, A., Konig, R., Schwab, M., and Westermann, F. (2011). CAMTA1, a 1p36 tumor suppressor candidate, inhibits growth and activates differentiation programs in neuroblastoma cells. Cancer Res 71, 3142-3151.
Heukamp, L.C., Thor, T., Schramm, A., De Preter, K., Kumps, C., De Wilde, B., Odersky, A., Peifer, M., Lindner, S., Spruessel, A., Pattyn, F., Mestdagh, P., Menten, B., Kuhfittig-Kulle, S., Kunkele, A., Konig, K., Meder, L., Chatterjee, S., Ullrich, R.T., Schulte, S., Vandesompele, J., Speleman, F., Buttner, R., Eggert, A., and Schulte, J.H. (2012). Targeted expression of mutated ALK induces neuroblastoma in transgenic mice. Science translational medicine 4, 141ra191.
References 245 Hildebrand, D., Tiefenbach, J., Heinzel, T., Grez, M., and Maurer, A.B. (2001). Multiple regions of ETO cooperate in transcriptional repression. The Journal of biological chemistry 276, 9889-9895.
Hirano, S., Kakinuma, S., Amasaki, Y., Nishimura, M., Imaoka, T., Fujimoto, S., Hino, O., and Shimada, Y. (2013). Ikaros is a critical target during simultaneous exposure to X-rays and N-ethyl-N-nitrosourea in mouse T-cell lymphomagenesis. International journal of cancer Journal international du cancer 132, 259-268.
Hitotsumachi, S., Carpenter, D.A., and Russell, W.L. (1985). Dose-repetition increases the mutagenic effectiveness of N-ethyl-N-nitrosourea in mouse spermatogonia. Proceedings of the National Academy of Sciences of the United States of America 82, 6619-6621.
Hogarty, M.D., Norris, M.D., Davis, K., Liu, X., Evageliou, N.F., Hayes, C.S., Pawel, B., Guo, R., Zhao, H., Sekyere, E., Keating, J., Thomas, W., Cheng, N.C., Murray, J., Smith, J., Sutton, R., Venn, N., London, W.B., Buxton, A., Gilmour, S.K., Marshall, G.M., and Haber, M. (2008). ODC1 is a critical determinant of MYCN oncogenesis and a therapeutic target in neuroblastoma. Cancer Res 68, 9735-9745.
Horiuchi, D., Anderton, B., and Goga, A. (2014). Taking on challenging targets: making MYC druggable. American Society of Clinical Oncology educational book / ASCO American Society of Clinical Oncology Meeting, e497-502.
Hrabe de Angelis, M.H., Flaswinkel, H., Fuchs, H., Rathkolb, B., Soewarto, D., Marschall, S., Heffner, S., Pargent, W., Wuensch, K., Jung, M., Reis, A., Richter, T., Alessandrini, F., Jakob, T., Fuchs, E., Kolb, H., Kremmer, E., Schaeble, K., Rollinski, B., Roscher, A., Peters, C., Meitinger, T., Strom, T., Steckler, T., Holsboer, F., Klopstock, T., Gekeler, F., Schindewolf, C., Jung, T., Avraham, K., Behrendt, H., Ring, J., Zimmer, A., Schughart, K., Pfeffer, K., Wolf, E., and Balling, R. (2000). Genome- wide, large-scale production of mutant mice by ENU mutagenesis. Nature genetics 25, 444-447.
Huang, M., and Weiss, W.A. (2013). Neuroblastoma and MYCN. Cold Spring Harbor perspectives in medicine 3, a014415.
References 246 Imamura, J., Bartram, C.R., Berthold, F., Harms, D., Nakamura, H., and Koeffler, H.P. (1993). Mutation of the p53 gene in neuroblastoma and its relationship with N-myc amplification. Cancer Res 53, 4053-4058.
Iraci, N., Diolaiti, D., Papa, A., Porro, A., Valli, E., Gherardi, S., Herold, S., Eilers, M., Bernardoni, R., Della Valle, G., and Perini, G. (2011). A SP1/MIZ1/MYCN repression complex recruits HDAC1 at the TRKA and p75NTR promoters and affects neuroblastoma malignancy by inhibiting the cell response to NGF. Cancer Res 71, 404- 412.
Islam, A., Kageyama, H., Takada, N., Kawamoto, T., Takayasu, H., Isogai, E., Ohira, M., Hashizume, K., Kobayashi, H., Kaneko, Y., and Nakagawara, A. (2000). High expression of Survivin, mapped to 17q25, is significantly associated with poor prognostic factors and promotes cell survival in human neuroblastoma. Oncogene 19, 617-623.
Ivics, Z., Hackett, P.B., Plasterk, R.H., and Izsvak, Z. (1997). Molecular reconstruction of Sleeping Beauty, a Tc1-like transposon from fish, and its transposition in human cells. Cell 91, 501-510.
Ivics, Z., and Izsvak, Z. (2010). The expanding universe of transposon technologies for gene and cell engineering. Mobile DNA 1, 25.
Jeison, M., Ash, S., Halevy-Berko, G., Mardoukh, J., Luria, D., Avigad, S., Feinberg- Gorenshtein, G., Goshen, Y., Hertzel, G., Kapelushnik, J., Ben Barak, A., Attias, D., Steinberg, R., Stein, J., Stark, B., and Yaniv, I. (2010). 2p24 Gain region harboring MYCN gene compared with MYCN amplified and nonamplified neuroblastoma: biological and clinical characteristics. The American journal of pathology 176, 2616- 2625.
Jiang, M., Stanke, J., and Lahti, J.M. (2011). Chapter 4 - The Connections Between Neural Crest Development and Neuroblastoma. In Current topics in developmental biology, A.D. Michael, ed. (Academic Press), pp. 77-127.
Joshi, V.V., Cantor, A.B., Altshuler, G., Larkin, E.W., Neill, J.S.A., Shuster, J.J., Holbrook, C.T., Ann Hayes, F., and Castleberry, R.P. (1992). Recommendations for modification of terminology of neuroblastic tumors and prognostic significance of
References 247 Shimada classification. A clinicopathologic study of 213 cases from the pediatric oncology group. Cancer 69, 2183-2196.
Junge, A., Zandi, R., Havgaard, J.H., Gorodkin, J., and Cowland, J.B. (2017). Assessing the miRNA sponge potential of RUNX1T1 in t(8;21) acute myeloid leukemia. Gene 615, 35-40.
Justice, M.J. (2000). Capitalizing on large-scale mouse mutagenesis screens. Nature reviews Genetics 1, 109-115.
Justice, M.J., Carpenter, D.A., Favor, J., Neuhauser-Klaus, A., Hrabe de Angelis, M., Soewarto, D., Moser, A., Cordes, S., Miller, D., Chapman, V., Weber, J.S., Rinchik, E.M., Hunsicker, P.R., Russell, W.L., and Bode, V.C. (2000). Effects of ENU dosage on mouse strains. Mammalian genome : official journal of the International Mammalian Genome Society 11, 484-488.
Justice, M.J., Noveroske, J.K., Weber, J.S., Zheng, B., and Bradley, A. (1999). Mouse ENU mutagenesis. Human molecular genetics 8, 1955-1963.
Kaatsch, P. (2010). Epidemiology of childhood cancer. Cancer treatment reviews 36, 277-285.
Kan, Z., Jaiswal, B.S., Stinson, J., Janakiraman, V., Bhatt, D., Stern, H.M., Yue, P., Haverty, P.M., Bourgon, R., Zheng, J., Moorhead, M., Chaudhuri, S., Tomsho, L.P., Peters, B.A., Pujara, K., Cordes, S., Davis, D.P., Carlton, V.E.H., Yuan, W., Li, L., Wang, W., Eigenbrot, C., Kaminker, J.S., Eberhard, D.A., Waring, P., Schuster, S.C., Modrusan, Z., Zhang, Z., Stokoe, D., de Sauvage, F.J., Faham, M., and Seshagiri, S. (2010). Diverse somatic mutation patterns and pathway alterations in human cancers. Nature 466, 869-873.
Katzenstein, H.M., Rademaker, A.W., Senger, C., Salwen, H.R., Nguyen, N.N., Thorner, P.S., Litsas, L., and Cohn, S.L. (1999). Effectiveness of the angiogenesis inhibitor TNP-470 in reducing the growth of human neuroblastoma in nude mice inversely correlates with tumor burden. Clinical cancer research : an official journal of the American Association for Cancer Research 5, 4273-4278.
References 248 Katzenstein, H.M., Salwen, H.R., Nguyen, N.N., Meitar, D., and Cohn, S.L. (2001). Antiangiogenic therapy inhibits human neuroblastoma growth. Medical and pediatric oncology 36, 190-193.
Kauppi, M., Murphy, J.M., de Graaf, C.A., Hyland, C.D., Greig, K.T., Metcalf, D., Hilton, A.A., Nicola, N.A., Kile, B.T., Hilton, D.J., and Alexander, W.S. (2008). Point mutation in the gene encoding p300 suppresses thrombocytopenia in Mpl-/- mice. Blood 112, 3148-3153.
Kile, B.T., and Hilton, D.J. (2005). The art and design of genetic screens: mouse. Nature reviews Genetics 6, 557-567.
Kim, M.K., Yang, S., Lee, K.H., Um, J.H., Liu, M., Kang, H., Park, S.J., and Chung, J.H. (2011). Promyelocytic leukemia inhibits adipogenesis, and loss of promyelocytic leukemia results in fat accumulation in mice. American journal of physiology Endocrinology and metabolism 301, E1130-1142.
Kim, Y.H., Girard, L., Giacomini, C.P., Wang, P., Hernandez-Boussard, T., Tibshirani, R., Minna, J.D., and Pollack, J.R. (2006). Combined microarray analysis of small cell lung cancer reveals altered apoptotic balance and distinct expression signatures of MYC family gene amplification. Oncogene 25, 130-138.
Kishida, S., Mu, P., Miyakawa, S., Fujiwara, M., Abe, T., Sakamoto, K., Onishi, A., Nakamura, Y., and Kadomatsu, K. (2013). Midkine promotes neuroblastoma through Notch2 signaling. Cancer Res 73, 1318-1327.
Knoepfler, P.S., Cheng, P.F., and Eisenman, R.N. (2002). N-myc is essential during neurogenesis for the rapid expansion of progenitor cell populations and the inhibition of neuronal differentiation. Genes Dev 16, 2699-2712.
Kocak, H., Ackermann, S., Hero, B., Kahlert, Y., Oberthuer, A., Juraeva, D., Roels, F., Theissen, J., Westermann, F., Deubzer, H., Ehemann, V., Brors, B., Odenthal, M., Berthold, F., and Fischer, M. (2013). Hox-C9 activates the intrinsic pathway of apoptosis and is associated with spontaneous regression in neuroblastoma. Cell death & disease 4, e586.
Kusakabe, H., Szczygiel, M.A., Whittingham, D.G., and Yanagimachi, R. (2001). Maintenance of genetic integrity in frozen and freeze-dried mouse spermatozoa.
References 249 Proceedings of the National Academy of Sciences of the United States of America 98, 13501-13506.
Lam, K., and Zhang, D.E. (2012). RUNX1 and RUNX1-ETO: roles in hematopoiesis and leukemogenesis. Front Biosci (Landmark Ed) 17, 1120-1139.
Lastowska, M., Viprey, V., Santibanez-Koref, M., Wappler, I., Peters, H., Cullinane, C., Roberts, P., Hall, A.G., Tweddle, D.A., Pearson, A.D., Lewis, I., Burchill, S.A., and Jackson, M.S. (2007). Identification of candidate genes involved in neuroblastoma progression by combining genomic and expression microarrays with survival data. Oncogene 26, 7432-7444.
Laug, W.E., Siegel, S.E., Shaw, K.N.F., Landing, B., Baptista, J., and Gutenstein, M. (1978). Initial Urinary Catecholamine Metabolite Concentrations and Prognosis in Neuroblastoma. Pediatrics 62, 77-83.
Laverdiere, C., Liu, Q., Yasui, Y., Nathan, P.C., Gurney, J.G., Stovall, M., Diller, L.R., Cheung, N.K., Wolden, S., Robison, L.L., and Sklar, C.A. (2009). Long-term outcomes in survivors of neuroblastoma: a report from the Childhood Cancer Survivor Study. Journal of the National Cancer Institute 101, 1131-1140.
Lázcoz, P., Muñoz, J., Nistal, M., Pestaña, Á., Encío, I.J., and Castresana, J.S. (2007). Loss of heterozygosity and microsatellite instability on chromosome arm 10q in neuroblastoma. Cancer genetics and cytogenetics 174, 1-8.
Lee, D., Cross, S.H., Strunk, K.E., Morgan, J.E., Bailey, C.L., Jackson, I.J., and Threadgill, D.W. (2004). Wa5 is a novel ENU-induced antimorphic allele of the epidermal growth factor receptor. Mammalian genome : official journal of the International Mammalian Genome Society 15, 525-536.
Lentz, R.D., Bergstein, J., Steffes, M.W., Brown, D.R., Prem, K., Michael, A.F., and Vernier, R.L. (1977). Postpubertal evaluation of gonadal function following cyclophosphamide therapy before and during puberty. The Journal of pediatrics 91, 385-394.
Levitt, R.C., Zhuang, G.Y., Kang, Y., Erasso, D.M., Upadhyay, U., Ozdemir, M., Fu, E.S., Sarantopoulos, K.D., Smith, S.B., Maixner, W., Diatchenko, L., Martin, E.R., and Wiltshire, T. (2017). Car8 dorsal root ganglion expression and genetic regulation of
References 250 analgesic responses are associated with a cis-eQTL in mice. Mammalian genome : official journal of the International Mammalian Genome Society.
Li, P., Fan, W., Xu, J., Lu, M., Yamamoto, H., Auwerx, J., Sears, D.D., Talukdar, S., Oh, D., Chen, A., Bandyopadhyay, G., Scadeng, M., Ofrecio, J.M., Nalbandian, S., and Olefsky, J.M. (2011). Adipocyte NCoR knockout decreases PPARgamma phosphorylation and enhances PPARgamma activity and insulin sensitivity. Cell 147, 815-826.
Liu, T., Tee, A.E., Porro, A., Smith, S.A., Dwarte, T., Liu, P.Y., Iraci, N., Sekyere, E., Haber, M., Norris, M.D., Diolaiti, D., Della Valle, G., Perini, G., and Marshall, G.M. (2007a). Activation of tissue transglutaminase transcription by histone deacetylase inhibition as a therapeutic approach for Myc oncogenesis. Proceedings of the National Academy of Sciences of the United States of America 104, 18682-18687.
Liu, X., Dobbie, M., Tunningley, R., Whittle, B., Zhang, Y., Ittner, L.M., and Gotz, J. (2011a). ENU mutagenesis screen to establish motor phenotypes in wild-type mice and modifiers of a pre-existing motor phenotype in tau mutant mice. Journal of biomedicine & biotechnology 2011, 130947.
Liu, Y., Chen, W., Gaudet, J., Cheney, M.D., Roudaia, L., Cierpicki, T., Klet, R.C., Hartman, K., Laue, T.M., Speck, N.A., and Bushweller, J.H. (2007b). Structural basis for recognition of SMRT/N-CoR by the MYND domain and its contribution to AML1/ETO's activity. Cancer cell 11, 483-497.
Liu, Z., Rader, J., He, S., Phung, T., and Thiele, C.J. (2013). CASZ1 inhibits cell cycle progression in neuroblastoma by restoring pRb activity. Cell Cycle 12, 2210-2218.
Liu, Z., Yang, X., Li, Z., McMahon, C., Sizer, C., Barenboim-Stapleton, L., Bliskovsky, V., Mock, B., Ried, T., London, W.B., Maris, J., Khan, J., and Thiele, C.J. (2011b). CASZ1, a candidate tumor-suppressor gene, suppresses neuroblastoma tumor growth through reprogramming gene expression. Cell death and differentiation 18, 1174-1183.
London, W.B., Castleberry, R.P., Matthay, K.K., Look, A.T., Seeger, R.C., Shimada, H., Thorner, P., Brodeur, G., Maris, J.M., Reynolds, C.P., and Cohn, S.L. (2005). Evidence for an age cutoff greater than 365 days for neuroblastoma risk group stratification in the Children's Oncology Group. J Clin Oncol 23, 6459-6465.
References 251 London, W.B., Frantz, C.N., Campbell, L.A., Seeger, R.C., Brumback, B.A., Cohn, S.L., Matthay, K.K., Castleberry, R.P., and Diller, L. (2010). Phase II randomized comparison of topotecan plus cyclophosphamide versus topotecan alone in children with recurrent or refractory neuroblastoma: a Children's Oncology Group study. J Clin Oncol 28, 3808-3815.
Look, A.T., Hayes, F.A., Nitschke, R., McWilliams, N.B., and Green, A.A. (1984). Cellular DNA Content as a Predictor of Response to Chemotherapy in Infants with Unresectable Neuroblastoma. New England Journal of Medicine 311, 231-235.
Look, A.T., Hayes, F.A., Shuster, J.J., Douglass, E.C., Castleberry, R.P., Bowman, L.C., Smith, E.I., and Brodeur, G.M. (1991). Clinical relevance of tumor cell ploidy and N-myc gene amplification in childhood neuroblastoma: a Pediatric Oncology Group study. J Clin Oncol 9, 581-591.
Lorico, A., Rappa, G., Finch, R.A., Yang, D., Flavell, R.A., and Sartorelli, A.C. (1997). Disruption of the murine MRP (multidrug resistance protein) gene leads to increased sensitivity to etoposide (VP-16) and increased levels of glutathione. Cancer Res 57, 5238-5242.
Luetteke, N.C., Phillips, H.K., Qiu, T.H., Copeland, N.G., Earp, H.S., Jenkins, N.A., and Lee, D.C. (1994). The mouse waved-2 phenotype results from a point mutation in the EGF receptor tyrosine kinase. Genes & development 8, 399-413.
Luetteke, N.C., Qiu, T.H., Peiffer, R.L., Oliver, P., Smithies, O., and Lee, D.C. (1993). TGF alpha deficiency results in hair follicle and eye abnormalities in targeted and waved-1 mice. Cell 73, 263-278.
Lutterbach, B., Westendorf, J.J., Linggi, B., Patten, A., Moniwa, M., Davie, J.R., Huynh, K.D., Bardwell, V.J., Lavinsky, R.M., Rosenfeld, M.G., Glass, C., Seto, E., and Hiebert, S.W. (1998). ETO, a target of t(8;21) in acute leukemia, interacts with the N- CoR and mSin3 corepressors. Molecular and cellular biology 18, 7176-7184.
Lutz, W., Stohr, M., Schurmann, J., Wenzel, A., Lohr, A., and Schwab, M. (1996). Conditional expression of N-myc in human neuroblastoma cells increases expression of alpha-prothymosin and ornithine decarboxylase and accelerates progression into S- phase early after mitogenic stimulation of quiescent cells. Oncogene 13, 803-812.
References 252 Ma, D., Chen, Z., Patel, K.P., Mishra, B.M., Yao, H., Abruzzo, L.V., Medeiros, L.J., Wierda, W., Keating, M., Sargent, R., and Luthra, R. (2011). Array Comparative Genomic Hybridization Analysis Identifies Recurrent Gain of Chromosome 2p25.3 Involving the ACP1 and MYCN Genes in Chronic Lymphocytic Leukemia. Clinical Lymphoma Myeloma and Leukemia 11, Supplement 1, S17-S24.
Maghsoudlou, A., Meyer, R.D., Rezazadeh, K., Arafa, E., Pudney, J., Hartsough, E., and Rahimi, N. (2016). RNF121 Inhibits Angiogenic Growth Factor Signaling by Restricting Cell Surface Expression of VEGFR-2. Traffic 17, 289-300.
Malynn, B.A., de Alboran, I.M., O'Hagan, R.C., Bronson, R., Davidson, L., DePinho, R.A., and Alt, F.W. (2000). N-myc can functionally replace c-myc in murine development, cellular growth, and differentiation. Genes & development 14, 1390- 1399.
Mao, D.Y., Watson, J.D., Yan, P.S., Barsyte-Lovejoy, D., Khosravi, F., Wong, W.W., Farnham, P.J., Huang, T.H., and Penn, L.Z. (2003). Analysis of Myc bound loci identified by CpG island arrays shows that Max is essential for Myc-dependent repression. Current biology : CB 13, 882-886.
Maris, J.M., Hogarty, M.D., Bagatell, R., and Cohn, S.L. (2007). Neuroblastoma. Lancet 369, 2106-2120.
Maris, J.M., White, P.S., Beltinger, C.P., Sulman, E.P., Castleberry, R.P., Shuster, J.J., Look, A.T., and Brodeur, G.M. (1995). Significance of Chromosome 1p Loss of Heterozygosity in Neuroblastoma. Cancer Research 55, 4664-4669.
Marshall, G.M., Liu, P.Y., Gherardi, S., Scarlett, C.J., Bedalov, A., Xu, N., Iraci, N., Valli, E., Ling, D., Thomas, W., van Bekkum, M., Sekyere, E., Jankowski, K., Trahair, T., Mackenzie, K.L., Haber, M., Norris, M.D., Biankin, A.V., Perini, G., and Liu, T. (2011). SIRT1 promotes N-Myc oncogenesis through a positive feedback loop involving the effects of MKP3 and ERK on N-Myc protein stability. PLoS genetics 7, e1002135.
Matsumoto, G., Nagai, S., Muta, M., Tsuruta, K., Okamoto, A., and Toi, M. (2003). Survival benefit of KRN7000 immune therapy in combination with TNP470 in hamster liver metastasis model of pancreatic cancer. Oncology reports 10, 1201-1206.
References 253 McGhee, L., Bryan, J., Elliott, L., Grimes, H.L., Kazanjian, A., Davis, J.N., and Meyers, S. (2003). Gfi-1 attaches to the nuclear matrix, associates with ETO (MTG8) and histone deacetylase proteins, and represses transcription using a TSA-sensitive mechanism. Journal of cellular biochemistry 89, 1005-1018.
McGinty, L., and Kolesar, J. (2017). Dinutuximab for maintenance therapy in pediatric neuroblastoma. American Journal of Health-System Pharmacy 74, 563-567.
McMahon, S.B., Van Buskirk, H.A., Dugan, K.A., Copeland, T.D., and Cole, M.D. (1998). The novel ATM-related protein TRRAP is an essential cofactor for the c-Myc and E2F oncoproteins. Cell 94, 363-374.
Meyer, N., and Penn, L.Z. (2008). Reflecting on 25 years with MYC. Nature reviews Cancer 8, 976-990.
Miskey, C., Izsvak, Z., Plasterk, R.H., and Ivics, Z. (2003). The Frog Prince: a reconstructed transposon from Rana pipiens with high transpositional activity in vertebrate cells. Nucleic acids research 31, 6873-6881.
Miura, H., Gurumurthy, C.B., Sato, T., Sato, M., and Ohtsuka, M. (2015). CRISPR/Cas9-based generation of knockdown mice by intronic insertion of artificial microRNA using longer single-stranded DNA. Scientific reports 5, 12799.
Molenaar, J.J., Domingo-Fernandez, R., Ebus, M.E., Lindner, S., Koster, J., Drabek, K., Mestdagh, P., van Sluis, P., Valentijn, L.J., van Nes, J., Broekmans, M., Haneveld, F., Volckmann, R., Bray, I., Heukamp, L., Sprussel, A., Thor, T., Kieckbusch, K., Klein- Hitpass, L., Fischer, M., Vandesompele, J., Schramm, A., van Noesel, M.M., Varesio, L., Speleman, F., Eggert, A., Stallings, R.L., Caron, H.N., Versteeg, R., and Schulte, J.H. (2012). LIN28B induces neuroblastoma and enhances MYCN levels via let-7 suppression. Nature genetics 44, 1199-1206.
Monclair, T., Brodeur, G.M., Ambros, P.F., Brisse, H.J., Cecchetto, G., Holmes, K., Kaneko, M., London, W.B., Matthay, K.K., Nuchtern, J.G., von Schweinitz, D., Simon, T., Cohn, S.L., and Pearson, A.D. (2009). The International Neuroblastoma Risk Group (INRG) staging system: an INRG Task Force report. J Clin Oncol 27, 298-303.
Moore, A.C., Amann, J.M., Williams, C.S., Tahinci, E., Farmer, T.E., Martinez, J.A., Yang, G., Luce, K.S., Lee, E., and Hiebert, S.W. (2008). Myeloid translocation gene
References 254 family members associate with T-cell factors (TCFs) and influence TCF-dependent transcription. Molecular and cellular biology 28, 977-987.
Moriarity, B.S., and Largaespada, D.A. (2015). Sleeping Beauty transposon insertional mutagenesis based mouse models for cancer gene discovery. Current opinion in genetics & development 30, 66-72.
Moriarity, B.S., Otto, G.M., Rahrmann, E.P., Rathe, S.K., Wolf, N.K., Weg, M.T., Manlove, L.A., LaRue, R.S., Temiz, N.A., Molyneux, S.D., Choi, K., Holly, K.J., Sarver, A.L., Scott, M.C., Forster, C.L., Modiano, J.F., Khanna, C., Hewitt, S.M., Khokha, R., Yang, Y., Gorlick, R., Dyer, M.A., and Largaespada, D.A. (2015). A Sleeping Beauty forward genetic screen identifies new genes and pathways driving osteosarcoma development and metastasis. Nature genetics 47, 615-624.
Moroz, V., Machin, D., Faldum, A., Hero, B., Iehara, T., Mosseri, V., Ladenstein, R., De Bernardi, B., Rubie, H., Berthold, F., Matthay, K.K., Monclair, T., Ambros, P.F., Pearson, A.D., Cohn, S.L., and London, W.B. (2011). Changes over three decades in outcome and the prognostic influence of age-at-diagnosis in young patients with neuroblastoma: a report from the International Neuroblastoma Risk Group Project. Eur J Cancer 47, 561-571.
Muller, H.J. (1927). Artificial Transmutation of the Gene. Science 66, 84-87.
Nair, S.K., and Burley, S.K. (2003). X-ray structures of Myc-Max and Mad-Max recognizing DNA. Molecular bases of regulation by proto-oncogenic transcription factors. Cell 112, 193-205.
Nakagawara, A., Arima-Nakagawara, M., Scavarda, N.J., Azar, C.G., Cantor, A.B., and Brodeur, G.M. (1993). Association between High Levels of Expression of the TRK Gene and Favorable Outcome in Human Neuroblastoma. New England Journal of Medicine 328, 847-854.
Nakagawara, A., Arima, M., Azar, C.G., Scavarda, N.J., and Brodeur, G.M. (1992). Inverse relationship between trk expression and N-myc amplification in human neuroblastomas. Cancer Res 52, 1364-1368.
References 255 Nakagawara, A., Azar, C.G., Scavarda, N.J., and Brodeur, G.M. (1994). Expression and function of TRK-B and BDNF in human neuroblastomas. Molecular and cellular biology 14, 759-767.
Nasir, A., Helm, J., Turner, L., Chen, D.T., Strosberg, J., Hafez, N., Henderson- Jackson, E.B., Hodul, P., Bui, M.M., Nasir, N.A., Hakam, A., Malafa, M.P., Yeatman, T.J., Coppola, D., and Kvols, L.K. (2011). RUNX1T1: a novel predictor of liver metastasis in primary pancreatic endocrine neoplasms. Pancreas 40, 627-633.
Nguyen, N., Judd, L.M., Kalantzis, A., Whittle, B., Giraud, A.S., and van Driel, I.R. (2011). Random mutagenesis of the mouse genome: a strategy for discovering gene function and the molecular basis of disease. American journal of physiology Gastrointestinal and liver physiology 300, G1-11.
Ni, T.K., Landrette, S.F., Bjornson, R.D., Bosenberg, M.W., and Xu, T. (2013). Low- copy piggyBac transposon mutagenesis in mice identifies genes driving melanoma. Proceedings of the National Academy of Sciences of the United States of America 110, E3640-3649.
Nikiforov, M.A., Chandriani, S., Park, J., Kotenko, I., Matheos, D., Johnsson, A., McMahon, S.B., and Cole, M.D. (2002). TRRAP-dependent and TRRAP-independent transcriptional activation by Myc family oncoproteins. Molecular and cellular biology 22, 5054-5063.
Nolan, P.M., Peters, J., Strivens, M., Rogers, D., Hagan, J., Spurr, N., Gray, I.C., Vizor, L., Brooker, D., Whitehill, E., Washbourne, R., Hough, T., Greenaway, S., Hewitt, M., Liu, X., McCormack, S., Pickford, K., Selley, R., Wells, C., Tymowska-Lalanne, Z., Roby, P., Glenister, P., Thornton, C., Thaung, C., Stevenson, J.A., Arkell, R., Mburu, P., Hardisty, R., Kiernan, A., Erven, A., Steel, K.P., Voegeling, S., Guenet, J.L., Nickols, C., Sadri, R., Nasse, M., Isaacs, A., Davies, K., Browne, M., Fisher, E.M., Martin, J., Rastan, S., Brown, S.D., and Hunter, J. (2000). A systematic, genome-wide, phenotype-driven mutagenesis programme for gene function studies in the mouse. Nature genetics 25, 440-443.
Norris, M.D., Bordow, S.B., Marshall, G.M., Haber, P.S., Cohn, S.L., and Haber, M. (1996). Expression of the gene for multidrug-resistance-associated protein and outcome in patients with neuroblastoma. The New England journal of medicine 334, 231-238.
References 256 Norris, M.D., Burkhart, C.A., Marshall, G.M., Weiss, W.A., and Haber, M. (2000). Expression of N-myc and MRP genes and their relationship to N-myc gene dosage and tumor formation in a murine neuroblastoma model. Medical and pediatric oncology 35, 585-589.
Noveroske, J.K., Weber, J.S., and Justice, M.J. (2000). The mutagenic action of N- ethyl-N-nitrosourea in the mouse. Mammalian genome : official journal of the International Mammalian Genome Society 11, 478-483.
Nowacki, S., Skowron, M., Oberthuer, A., Fagin, A., Voth, H., Brors, B., Westermann, F., Eggert, A., Hero, B., Berthold, F., and Fischer, M. (2008). Expression of the tumour suppressor gene CADM1 is associated with favourable outcome and inhibits cell survival in neuroblastoma. Oncogene 27, 3329-3338.
Oberthuer, A., Juraeva, D., Li, L., Kahlert, Y., Westermann, F., Eils, R., Berthold, F., Shi, L., Wolfinger, R.D., Fischer, M., and Brors, B. (2010). Comparison of performance of one-color and two-color gene-expression analyses in predicting clinical endpoints of neuroblastoma patients. The pharmacogenomics journal 10, 258-266.
Ogawa, S., Takita, J., Sanada, M., and Hayashi, Y. (2011). Oncogenic mutations of ALK in neuroblastoma. Cancer Science 102, 302-308.
Ohlfest, J.R., Demorest, Z.L., Motooka, Y., Vengco, I., Oh, S., Chen, E., Scappaticci, F.A., Saplis, R.J., Ekker, S.C., Low, W.C., Freese, A.B., and Largaespada, D.A. (2005). Combinatorial antiangiogenic gene therapy by nonviral gene transfer using the sleeping beauty transposon causes tumor regression and improves survival in mice bearing intracranial human glioblastoma. Molecular therapy : the journal of the American Society of Gene Therapy 12, 778-788.
Okawa, E.R., Gotoh, T., Manne, J., Igarashi, J., Fujita, T., Silverman, K.A., Xhao, H., Mosse, Y.P., White, P.S., and Brodeur, G.M. (2008). Expression and sequence analysis of candidates for the 1p36.31 tumor suppressor gene deleted in neuroblastomas. Oncogene 27, 803-810.
Olson, A., and Lewis, G. (1928). Natural reactivity and the origin of species. Nature 121, 673-674.
References 257 Parbin, S., Kar, S., Shilpi, A., Sengupta, D., Deb, M., Rath, S.K., and Patra, S.K. (2014). Histone deacetylases: a saga of perturbed acetylation homeostasis in cancer. The journal of histochemistry and cytochemistry : official journal of the Histochemistry Society 62, 11-33.
Park, J.R., Eggert, A., and Caron, H. (2010). Neuroblastoma: biology, prognosis, and treatment. Hematology/oncology clinics of North America 24, 65-86.
Park, J.R., Scott, J.R., Stewart, C.F., London, W.B., Naranjo, A., Santana, V.M., Shaw, P.J., Cohn, S.L., and Matthay, K.K. (2011). Pilot Induction Regimen Incorporating Pharmacokinetically Guided Topotecan for Treatment of Newly Diagnosed High-Risk Neuroblastoma: A Children's Oncology Group Study. Journal of Clinical Oncology 29, 4351-4357.
Pasquier, E., Street, J., Pouchy, C., Carre, M., Gifford, A.J., Murray, J., Norris, M.D., Trahair, T., Andre, N., and Kavallaris, M. (2013). beta-blockers increase response to chemotherapy via direct antitumour and anti-angiogenic mechanisms in neuroblastoma. British journal of cancer 108, 2485-2494.
Patankar, S., Lazaroff, M., Yoon, S.O., and Chikaraishi, D.M. (1997). A novel basal promoter element is required for expression of the rat tyrosine hydroxylase gene. The Journal of neuroscience : the official journal of the Society for Neuroscience 17, 4076- 4086.
Porro, A., Haber, M., Diolaiti, D., Iraci, N., Henderson, M., Gherardi, S., Valli, E., Munoz, M.A., Xue, C., Flemming, C., Schwab, M., Wong, J.H., Marshall, G.M., Della Valle, G., Norris, M.D., and Perini, G. (2010). Direct and coordinate regulation of ATP- binding cassette transporter genes by Myc factors generates specific transcription signatures that significantly affect the chemoresistance phenotype of cancer cells. The Journal of biological chemistry 285, 19532-19543.
Posternak, V., and Cole, M.D. (2016). Strategically targeting MYC in cancer. F1000Research 5.
Ptasinska, A., Assi, S.A., Mannari, D., James, S.R., Williamson, D., Dunne, J., Hoogenkamp, M., Wu, M., Care, M., McNeill, H., Cauchy, P., Cullen, M., Tooze, R.M., Tenen, D.G., Young, B.D., Cockerill, P.N., Westhead, D.R., Heidenreich, O., and Bonifer, C. (2012). Depletion of RUNX1/ETO in t(8;21) AML cells leads to genome-
References 258 wide changes in chromatin structure and transcription factor binding. Leukemia : official journal of the Leukemia Society of America, Leukemia Research Fund, UK 26, 1829-1841.
Puissant, A., Frumm, S.M., Alexe, G., Bassil, C.F., Qi, J., Chanthery, Y.H., Nekritz, E.A., Zeid, R., Gustafson, W.C., Greninger, P., Garnett, M.J., McDermott, U., Benes, C.H., Kung, A.L., Weiss, W.A., Bradner, J.E., and Stegmaier, K. (2013). Targeting MYCN in neuroblastoma by BET bromodomain inhibition. Cancer discovery 3, 308- 323.
Quarta, C., Cantelli, E., Nanni, C., Ambrosini, V., D'Ambrosio, D., Di Leo, K., Angelucci, S., Zagni, F., Lodi, F., Marengo, M., Weiss, W.A., Pession, A., Tonelli, R., and Fanti, S. (2013). Molecular imaging of neuroblastoma progression in TH-MYCN transgenic mice. Molecular imaging and biology : MIB : the official publication of the Academy of Molecular Imaging 15, 194-202.
Rad, R., Rad, L., Wang, W., Strong, A., Ponstingl, H., Bronner, I.F., Mayho, M., Steiger, K., Weber, J., Hieber, M., Veltkamp, C., Eser, S., Geumann, U., Ollinger, R., Zukowska, M., Barenboim, M., Maresch, R., Cadinanos, J., Friedrich, M., Varela, I., Constantino-Casas, F., Sarver, A., Ten Hoeve, J., Prosser, H., Seidler, B., Bauer, J., Heikenwalder, M., Metzakopian, E., Krug, A., Ehmer, U., Schneider, G., Knosel, T., Rummele, P., Aust, D., Grutzmann, R., Pilarsky, C., Ning, Z., Wessels, L., Schmid, R.M., Quail, M.A., Vassiliou, G., Esposito, I., Liu, P., Saur, D., and Bradley, A. (2015). A conditional piggyBac transposition system for genetic screening in mice identifies oncogenic networks in pancreatic cancer. Nature genetics 47, 47-56.
Rao, G., Alland, L., Guida, P., Schreiber-Agus, N., Chen, K., Chin, L., Rochelle, J.M., Seldin, M.F., Skoultchi, A.I., and DePinho, R.A. (1996). Mouse Sin3A interacts with and can functionally substitute for the amino-terminal repression of the Myc antagonist Mxi1. Oncogene 12, 1165-1172.
Rasmuson, A., Segerstrom, L., Nethander, M., Finnman, J., Elfman, L.H., Javanmardi, N., Nilsson, S., Johnsen, J.I., Martinsson, T., and Kogner, P. (2012). Tumor development, growth characteristics and spectrum of genetic aberrations in the TH- MYCN mouse model of neuroblastoma. PloS one 7, e51297.
References 259 Rochford, J.J., Semple, R.K., Laudes, M., Boyle, K.B., Christodoulides, C., Mulligan, C., Lelliott, C.J., Schinner, S., Hadaschik, D., Mahadevan, M., Sethi, J.K., Vidal-Puig, A., and O'Rahilly, S. (2004). ETO/MTG8 is an inhibitor of C/EBPbeta activity and a regulator of early adipogenesis. Molecular and cellular biology 24, 9863-9872.
Rosa-Rosa, J.M., Pita, G., Gonzalez-Neira, A., Milne, R.L., Fernandez, V., Ruivenkamp, C., van Asperen, C.J., Devilee, P., and Benitez, J. (2009). A 7 Mb region within 11q13 may contain a high penetrance gene for breast cancer. Breast cancer research and treatment 118, 151-159.
Rossetti, S., Hoogeveen, A.T., and Sacchi, N. (2004). The MTG proteins: chromatin repression players with a passion for networking. Genomics 84, 1-9.
Rossetti, S., van Unen, L., Sacchi, N., and Hoogeveen, A.T. (2008). Novel RNA- binding properties of the MTG chromatin regulatory proteins. BMC molecular biology 9, 93.
Rothberg, P.G., Erisman, M.D., Diehl, R.E., Rovigatti, U.G., and Astrin, S.M. (1984). Structure and expression of the oncogene c-myc in fresh tumor material from patients with hematopoietic malignancies. Molecular and cellular biology 4, 1096-1103.
Rubio-Aliaga, I., Soewarto, D., Wagner, S., Klaften, M., Fuchs, H., Kalaydjiev, S., Busch, D.H., Klempt, M., Rathkolb, B., Wolf, E., Abe, K., Zeiser, S., Przemeck, G.K., Beckers, J., and de Angelis, M.H. (2007). A genetic screen for modifiers of the delta1- dependent notch signaling function in the mouse. Genetics 175, 1451-1463.
Russell, W.L., Kelly, E.M., Hunsicker, P.R., Bangham, J.W., Maddux, S.C., and Phipps, E.L. (1979). Specific-locus test shows ethylnitrosourea to be the most potent mutagen in the mouse. Proceedings of the National Academy of Sciences of the United States of America 76, 5818-5819.
Ryan, S.L., Schwalbe, E.C., Cole, M., Lu, Y., Lusher, M.E., Megahed, H., O'Toole, K., Nicholson, S.L., Bognar, L., Garami, M., Hauser, P., Korshunov, A., Pfister, S.M., Williamson, D., Taylor, R.E., Ellison, D.W., Bailey, S., and Clifford, S.C. (2012). MYC family amplification and clinical risk-factors interact to predict an extremely poor prognosis in childhood medulloblastoma. Acta Neuropathol 123, 501-513.
References 260 Sacchi, N., Tamanini, F., Willemsen, R., Denis-Donini, S., Campiglio, S., and Hoogeveen, A.T. (1998). Subcellular localization of the oncoprotein MTG8 (CDR/ETO) in neural cells. Oncogene 16, 2609-2615.
Salinger, A.P., and Justice, M.J. (2008). Mouse Mutagenesis Using N-Ethyl-N- Nitrosourea (ENU). CSH protocols 2008, pdb prot4985.
Sano, H., Bonadio, J., Gerbing, R.B., London, W.B., Matthay, K.K., Lukens, J.N., and Shimada, H. (2006). International neuroblastoma pathology classification adds independent prognostic information beyond the prognostic contribution of age. Eur J Cancer 42, 1113-1119.
Santo, E.E., Ebus, M.E., Koster, J., Schulte, J.H., Lakeman, A., van Sluis, P., Vermeulen, J., Gisselsson, D., Ora, I., Lindner, S., Buckley, P.G., Stallings, R.L., Vandesompele, J., Eggert, A., Caron, H.N., Versteeg, R., and Molenaar, J.J. (2011). Oncogenic activation of FOXR1 by 11q23 intrachromosomal deletion-fusions in neuroblastoma. Oncogene 31, 1571-1581.
Schlesinger, H.R., Gerson, J.M., Moorhead, P.S., Maguire, H., and Hummeler, K. (1976). Establishment and characterization of human neuroblastoma cell lines. Cancer Res 36, 3094-3100.
Schmittgen, T.D., and Livak, K.J. (2008). Analyzing real-time PCR data by the comparative C(T) method. Nature protocols 3, 1101-1108.
Schulte, J.H., Lindner, S., Bohrer, A., Maurer, J., De Preter, K., Lefever, S., Heukamp, L., Schulte, S., Molenaar, J., Versteeg, R., Thor, T., Kunkele, A., Vandesompele, J., Speleman, F., Schorle, H., Eggert, A., and Schramm, A. (2013). MYCN and ALKF1174L are sufficient to drive neuroblastoma development from neural crest progenitor cells. Oncogene 32, 1059-1065.
Schwab, M., Alitalo, K., Klempnauer, K.H., Varmus, H.E., Bishop, J.M., Gilbert, F., Brodeur, G., Goldstein, M., and Trent, J. (1983). Amplified DNA with limited homology to myc cellular oncogene is shared by human neuroblastoma cell lines and a neuroblastoma tumour. Nature 305, 245-248.
Seeger, R.C., Brodeur, G.M., Sather, H., Dalton, A., Siegel, S.E., Wong, K.Y., and Hammond, D. (1985). Association of Multiple Copies of the N-myc Oncogene with
References 261 Rapid Progression of Neuroblastomas. New England Journal of Medicine 313, 1111- 1116.
Shalem, O., Sanjana, N.E., and Zhang, F. (2015). High-throughput functional genomics using CRISPR-Cas9. Nature reviews Genetics 16, 299-311.
Shi, H., O'Reilly, V.C., Moreau, J.L., Bewes, T.R., Yam, M.X., Chapman, B.E., Grieve, S.M., Stocker, R., Graham, R.M., Chapman, G., Sparrow, D.B., and Dunwoodie, S.L. (2016). Gestational stress induces the unfolded protein response, resulting in heart defects. Development 143, 2561-2572.
Shimada, H., Ambros, I.M., Dehner, L.P., Hata, J., Joshi, V.V., and Roald, B. (1999a). Terminology and morphologic criteria of neuroblastic tumors: recommendations by the International Neuroblastoma Pathology Committee. Cancer 86, 349-363.
Shimada, H., Ambros, I.M., Dehner, L.P., Hata, J., Joshi, V.V., Roald, B., Stram, D.O., Gerbing, R.B., Lukens, J.N., Matthay, K.K., and Castleberry, R.P. (1999b). The International Neuroblastoma Pathology Classification (the Shimada system). Cancer 86, 364-372.
Shimada, H., and Nakagawa, A. (1999). Tumors of the neuroblastoma group. International Journal of Clinical Oncology 4, 123-132.
Shimada, H., Nakagawa, A., Peters, J., Wang, H., Wakamatsu, P.K., Lukens, J.N., Matthay, K.K., Siegel, S.E., and Seeger, R.C. (2004). TrkA expression in peripheral neuroblastic tumors: prognostic significance and biological relevance. Cancer 101, 1873-1881.
Siimes, M.A., and Rautonen, J. (1990). Small testicles with impaired production of sperm in adult male survivors of childhood malignancies. Cancer 65, 1303-1306.
Simpson, E.M., Linder, C.C., Sargent, E.E., Davisson, M.T., Mobraaten, L.E., and Sharp, J.J. (1997). Genetic variation among 129 substrains and its importance for targeted mutagenesis in mice. Nature genetics 16, 19-27.
Sjöblom, T., Jones, S., Wood, L.D., Parsons, D.W., Lin, J., Barber, T.D., Mandelker, D., Leary, R.J., Ptak, J., Silliman, N., Szabo, S., Buckhaults, P., Farrell, C., Meeh, P., Markowitz, S.D., Willis, J., Dawson, D., Willson, J.K.V., Gazdar, A.F., Hartigan, J.,
References 262 Wu, L., Liu, C., Parmigiani, G., Park, B.H., Bachman, K.E., Papadopoulos, N., Vogelstein, B., Kinzler, K.W., and Velculescu, V.E. (2006). The Consensus Coding Sequences of Human Breast and Colorectal Cancers. Science 314, 268-274.
Slobodin, B., Han, R., Calderone, V., Vrielink, J.A., Loayza-Puch, F., Elkon, R., and Agami, R. (2017). Transcription Impacts the Efficiency of mRNA Translation via Co- transcriptional N6-adenosine Methylation. Cell 169, 326-337 e312.
Solberg Woods, L.C. (2014). QTL mapping in outbred populations: successes and challenges. Physiological genomics 46, 81-90.
Souza, D.R.V.d., Sanabani, S.S., Souza, A.C.M.F.d., Odone, V.F., Epelman, S., and Bendit, I. (2011). Prognostic impact of MYCN, DDX1, TrkA, and TrkC gene transcripts expression in neuroblastoma. Pediatric blood & cancer 56, 749-756.
Spitz, R., Hero, B., Ernestus, K., and Berthold, F. (2003). Deletions in Chromosome Arms 3p and 11q Are New Prognostic Markers in Localized and 4s Neuroblastoma. Clinical Cancer Research 9, 52-58.
Stanke, M., Duong, C.V., Pape, M., Geissen, M., Burbach, G., Deller, T., Gascan, H., Otto, C., Parlato, R., Schutz, G., and Rohrer, H. (2006). Target-dependent specification of the neurotransmitter phenotype: cholinergic differentiation of sympathetic neurons is mediated in vivo by gp 130 signaling. Development 133, 141-150.
Stanton, B.R., Perkins, A.S., Tessarollo, L., Sassoon, D.A., and Parada, L.F. (1992). Loss of N-myc function results in embryonic lethality and failure of the epithelial component of the embryo to develop. Genes & development 6, 2235-2247.
Stevens, L.C. (1973). A New Inbred Subline of Mice (129/terSv) with a High Incidence of Spontaneous Congenital Testicular Teratomas2. JNCI: Journal of the National Cancer Institute 50, 235-242.
Stottmann, R.W., Moran, J.L., Turbe-Doan, A., Driver, E., Kelley, M., and Beier, D.R. (2011). Focusing forward genetics: a tripartite ENU screen for neurodevelopmental mutations in the mouse. Genetics 188, 615-624.
Strausberg, R.L., Feingold, E.A., Grouse, L.H., Derge, J.G., Klausner, R.D., Collins, F.S., Wagner, L., Shenmen, C.M., Schuler, G.D., Altschul, S.F., Zeeberg, B., Buetow,
References 263 K.H., Schaefer, C.F., Bhat, N.K., Hopkins, R.F., Jordan, H., Moore, T., Max, S.I., Wang, J., Hsieh, F., Diatchenko, L., Marusina, K., Farmer, A.A., Rubin, G.M., Hong, L., Stapleton, M., Soares, M.B., Bonaldo, M.F., Casavant, T.L., Scheetz, T.E., Brownstein, M.J., Usdin, T.B., Toshiyuki, S., Carninci, P., Prange, C., Raha, S.S., Loquellano, N.A., Peters, G.J., Abramson, R.D., Mullahy, S.J., Bosak, S.A., McEwan, P.J., McKernan, K.J., Malek, J.A., Gunaratne, P.H., Richards, S., Worley, K.C., Hale, S., Garcia, A.M., Gay, L.J., Hulyk, S.W., Villalon, D.K., Muzny, D.M., Sodergren, E.J., Lu, X., Gibbs, R.A., Fahey, J., Helton, E., Ketteman, M., Madan, A., Rodrigues, S., Sanchez, A., Whiting, M., Young, A.C., Shevchenko, Y., Bouffard, G.G., Blakesley, R.W., Touchman, J.W., Green, E.D., Dickson, M.C., Rodriguez, A.C., Grimwood, J., Schmutz, J., Myers, R.M., Butterfield, Y.S., Krzywinski, M.I., Skalska, U., Smailus, D.E., Schnerch, A., Schein, J.E., Jones, S.J., and Marra, M.A. (2002). Generation and initial analysis of more than 15,000 full-length human and mouse cDNA sequences. Proceedings of the National Academy of Sciences of the United States of America 99, 16899-16903.
Stride, B.D., Grant, C.E., Loe, D.W., Hipfner, D.R., Cole, S.P., and Deeley, R.G. (1997). Pharmacological characterization of the murine and human orthologs of multidrug-resistance protein in transfected human embryonic kidney cells. Molecular pharmacology 52, 344-353.
Sun, L., Xie, H., Mori, M.A., Alexander, R., Yuan, B., Hattangadi, S.M., Liu, Q., Kahn, C.R., and Lodish, H.F. (2011). Mir193b-365 is essential for brown fat differentiation. Nature cell biology 13, 958-965.
Svensson, T., Rydén, M., Schilling, F.H., Dominici, C., Sehgal, R., Ibáñez, C.F., and Kogner, P. (1997). Coexpression of mRNA for the full-length neurotrophin receptor trk- C and trk-A in favourable neuroblastoma. European journal of cancer (Oxford, England : 1990) 33, 2058-2063.
Swartling, F.J., Grimmer, M.R., Hackett, C.S., Northcott, P.A., Fan, Q.W., Goldenberg, D.D., Lau, J., Masic, S., Nguyen, K., Yakovenko, S., Zhe, X.N., Gilmer, H.C., Collins, R., Nagaoka, M., Phillips, J.J., Jenkins, R.B., Tihan, T., Vandenberg, S.R., James, C.D., Tanaka, K., Taylor, M.D., Weiss, W.A., and Chesler, L. (2010). Pleiotropic role for MYCN in medulloblastoma. Genes & development 24, 1059-1072.
References 264 Tansey, W.P. (2014). Mammalian MYC Proteins and Cancer. New Journal of Science 2014, 27.
Tchekneva, E.E., Rinchik, E.M., Polosukhina, D., Davis, L.S., Kadkina, V., Mohamed, Y., Dunn, S.R., Sharma, K., Qi, Z., Fogo, A.B., and Breyer, M.D. (2007). A sensitized screen of N-ethyl-N-nitrosourea-mutagenized mice identifies dominant mutants predisposed to diabetic nephropathy. Journal of the American Society of Nephrology : JASN 18, 103-112.
Teitz, T., Inoue, M., Valentine, M.B., Zhu, K., Rehg, J.E., Zhao, W., Finkelstein, D., Wang, Y.D., Johnson, M.D., Calabrese, C., Rubinstein, M., Hakem, R., Weiss, W.A., and Lahti, J.M. (2013). Th-MYCN mice with caspase-8 deficiency develop advanced neuroblastoma with bone marrow metastasis. Cancer Res 73, 4086-4097.
Teitz, T., Stanke, J.J., Federico, S., Bradley, C.L., Brennan, R., Zhang, J., Johnson, M.D., Sedlacik, J., Inoue, M., Zhang, Z.M., Frase, S., Rehg, J.E., Hillenbrand, C.M., Finkelstein, D., Calabrese, C., Dyer, M.A., and Lahti, J.M. (2011). Preclinical models for neuroblastoma: establishing a baseline for treatment. PloS one 6, e19133.
Tivnan, A., Tracey, L., Buckley, P.G., Alcock, L.C., Davidoff, A.M., and Stallings, R.L. (2011). MicroRNA-34a is a potent tumor suppressor molecule in vivo in neuroblastoma. BMC cancer 11, 33.
Tonelli, R., McIntyre, A., Camerin, C., Walters, Z.S., Di Leo, K., Selfe, J., Purgato, S., Missiaglia, E., Tortori, A., Renshaw, J., Astolfi, A., Taylor, K.R., Serravalle, S., Bishop, R., Nanni, C., Valentijn, L.J., Faccini, A., Leuschner, I., Formica, S., Reis-Filho, J.S., Ambrosini, V., Thway, K., Franzoni, M., Summersgill, B., Marchelli, R., Hrelia, P., Cantelli-Forti, G., Fanti, S., Corradini, R., Pession, A., and Shipley, J. (2012). Antitumor activity of sustained N-myc reduction in rhabdomyosarcomas and transcriptional block by antigene therapy. Clinical cancer research : an official journal of the American Association for Cancer Research 18, 796-807.
Toubal, A., Clement, K., Fan, R., Ancel, P., Pelloux, V., Rouault, C., Veyrie, N., Hartemann, A., Treuter, E., and Venteclef, N. (2013). SMRT-GPS2 corepressor pathway dysregulation coincides with obesity-linked adipocyte inflammation. The Journal of clinical investigation 123, 362-379.
References 265 Tripathi, D.N., and Jena, G.B. (2008). Astaxanthin inhibits cytotoxic and genotoxic effects of cyclophosphamide in mice germ cells. Toxicology 248, 96-103.
Truitt, M.L., Conn, C.S., Shi, Z., Pang, X., Tokuyasu, T., Coady, A.M., Seo, Y., Barna, M., and Ruggero, D. (2015). Differential Requirements for eIF4E Dose in Normal Development and Cancer. Cell 162, 59-71. van der Weyden, L., Giotopoulos, G., Wong, K., Rust, A.G., Robles-Espinoza, C.D., Osaki, H., Huntly, B.J., and Adams, D.J. (2015). Somatic drivers of B-ALL in a model of ETV6-RUNX1; Pax5(+/-) leukemia. BMC cancer 15, 585.
Vassal, G., Terrier-Lacombe, M.J., Bissery, M.C., Venuat, A.M., Gyergyay, F., Benard, J., Morizet, J., Boland, I., Ardouin, P., Bressac-de-Paillerets, B., and Gouyette, A. (1996). Therapeutic activity of CPT-11, a DNA-topoisomerase I inhibitor, against peripheral primitive neuroectodermal tumour and neuroblastoma xenografts. British journal of cancer 74, 537-545.
Vita, M., and Henriksson, M. (2006). The Myc oncoprotein as a therapeutic target for human cancer. Semin Cancer Biol 16, 318-330.
Waldeck, K., Cullinane, C., Ardley, K., Shortt, J., Martin, B., Tothill, R.W., Li, J., Johnstone, R.W., McArthur, G.A., Hicks, R.J., and Wood, P.J. (2016). Long term, continuous exposure to panobinostat induces terminal differentiation and long term survival in the TH-MYCN neuroblastoma mouse model. International journal of cancer Journal international du cancer 139, 194-204.
Wang, X.W., Wei, W., Wang, W.Q., Zhao, X.Y., Guo, H., and Fang, D.C. (2014). RING finger proteins are involved in the progression of barrett esophagus to esophageal adenocarcinoma: a preliminary study. Gut and liver 8, 487-494.
Weber, J., Ollinger, R., Friedrich, M., Ehmer, U., Barenboim, M., Steiger, K., Heid, I., Mueller, S., Maresch, R., Engleitner, T., Gross, N., Geumann, U., Fu, B., Segler, A., Yuan, D., Lange, S., Strong, A., de la Rosa, J., Esposito, I., Liu, P., Cadinanos, J., Vassiliou, G.S., Schmid, R.M., Schneider, G., Unger, K., Yang, F., Braren, R., Heikenwalder, M., Varela, I., Saur, D., Bradley, A., and Rad, R. (2015). CRISPR/Cas9 somatic multiplex-mutagenesis for high-throughput functional cancer genomics in mice. Proceedings of the National Academy of Sciences of the United States of America 112, 13982-13987.
References 266 Weber, J.S., Salinger, A., and Justice, M.J. (2000). Optimal N-ethyl-N-nitrosourea (ENU) doses for inbred mouse strains. Genesis 26, 230-233.
Weiss, W.A., Aldape, K., Mohapatra, G., Feuerstein, B.G., and Bishop, J.M. (1997). Targeted expression of MYCN causes neuroblastoma in transgenic mice. The EMBO journal 16, 2985-2995.
Westermark, U.K., Wilhelm, M., Frenzel, A., and Henriksson, M.A. (2011). The MYCN oncogene and differentiation in neuroblastoma. Seminars in Cancer Biology 21, 256-266.
Wiesner, S.M., Decker, S.A., Larson, J.D., Ericson, K., Forster, C., Gallardo, J.L., Long, C., Demorest, Z.L., Zamora, E.A., Low, W.C., SantaCruz, K., Largaespada, D.A., and Ohlfest, J.R. (2009). De novo induction of genetically engineered brain tumors in mice using plasmid DNA. Cancer Res 69, 431-439.
Wildonger, J., and Mann, R.S. (2005). Evidence that nervy, the Drosophila homolog of ETO/MTG8, promotes mechanosensory organ development by enhancing Notch signaling. Developmental biology 286, 507-520.
Wiltshire, T., Pletcher, M.T., Batalov, S., Barnes, S.W., Tarantino, L.M., Cooke, M.P., Wu, H., Smylie, K., Santrosyan, A., Copeland, N.G., Jenkins, N.A., Kalush, F., Mural, R.J., Glynne, R.J., Kay, S.A., Adams, M.D., and Fletcher, C.F. (2003). Genome-wide single-nucleotide polymorphism analysis defines haplotype patterns in mouse. Proceedings of the National Academy of Sciences 100, 3380-3385.
Wood, L.D., Parsons, D.W., Jones, S., Lin, J., Sjöblom, T., Leary, R.J., Shen, D., Boca, S.M., Barber, T., Ptak, J., Silliman, N., Szabo, S., Dezso, Z., Ustyanksky, V., Nikolskaya, T., Nikolsky, Y., Karchin, R., Wilson, P.A., Kaminker, J.S., Zhang, Z., Croshaw, R., Willis, J., Dawson, D., Shipitsin, M., Willson, J.K.V., Sukumar, S., Polyak, K., Park, B.H., Pethiyagoda, C.L., Pant, P.V.K., Ballinger, D.G., Sparks, A.B., Hartigan, J., Smith, D.R., Suh, E., Papadopoulos, N., Buckhaults, P., Markowitz, S.D., Parmigiani, G., Kinzler, K.W., Velculescu, V.E., and Vogelstein, B. (2007). The Genomic Landscapes of Human Breast and Colorectal Cancers. Science 318, 1108- 1113.
Woodard, L.E., and Wilson, M.H. (2015). piggyBac-ing models and new therapeutic strategies. Trends in biotechnology 33, 525-533.
References 267 Wu, X., Northcott, P.A., Dubuc, A., Dupuy, A.J., Shih, D.J., Witt, H., Croul, S., Bouffet, E., Fults, D.W., Eberhart, C.G., Garzia, L., Van Meter, T., Zagzag, D., Jabado, N., Schwartzentruber, J., Majewski, J., Scheetz, T.E., Pfister, S.M., Korshunov, A., Li, X.N., Scherer, S.W., Cho, Y.J., Akagi, K., MacDonald, T.J., Koster, J., McCabe, M.G., Sarver, A.L., Collins, V.P., Weiss, W.A., Largaespada, D.A., Collier, L.S., and Taylor, M.D. (2012). Clonal selection drives genetic divergence of metastatic medulloblastoma. Nature 482, 529-533.
Yeh, K.T., Chen, T.H., Yang, H.W., Chou, J.L., Chen, L.Y., Yeh, C.M., Chen, Y.H., Lin, R.I., Su, H.Y., Chen, G.C., Deatherage, D.E., Huang, Y.W., Yan, P.S., Lin, H.J., Nephew, K.P., Huang, T.H., Lai, H.C., and Chan, M.W. (2011). Aberrant TGFbeta/SMAD4 signaling contributes to epigenetic silencing of a putative tumor suppressor, RunX1T1 in ovarian cancer. Epigenetics 6, 727-739.
Youngren, K.K., Coveney, D., Peng, X., Bhattacharya, C., Schmidt, L.S., Nickerson, M.L., Lamb, B.T., Deng, J.M., Behringer, R.R., Capel, B., Rubin, E.M., Nadeau, J.H., and Matin, A. (2005). The Ter mutation in the dead end gene causes germ cell loss and testicular germ cell tumours. Nature 435, 360-364.
Yusa, K. (2015). piggyBac Transposon. Microbiology spectrum 3, MDNA3-0028-2014.
Zelcer, N., van de Wetering, K., de Waart, R., Scheffer, G.L., Marschall, H.U., Wielinga, P.R., Kuil, A., Kunne, C., Smith, A., van der Valk, M., Wijnholds, J., Elferink, R.O., and Borst, P. (2006). Mice lacking Mrp3 (Abcc3) have normal bile salt transport, but altered hepatic transport of endogenous glucuronides. Journal of hepatology 44, 768-775.
Zeltzer, P.M., Marangos, P.J., Evans, A.E., and Schneider, S.L. (1986). Serum neuron- specific enolase in children with neuroblastoma. Crelationship to stage and disease course. Cancer 57, 1230-1234.
Zemirli, N., Pourcelot, M., Dogan, N., Vazquez, A., and Arnoult, D. (2014). The E3 ubiquitin ligase RNF121 is a positive regulator of NF-kappaB activation. Cell communication and signaling : CCS 12, 72.
Zervos, A.S., Gyuris, J., and Brent, R. (1993). Mxi1, a protein that specifically interacts with Max to bind Myc-Max recognition sites. Cell 72, 223-232.
References 268 Zhang, H., Guan, M., Townsend, K.L., Huang, T.L., An, D., Yan, X., Xue, R., Schulz, T.J., Winnay, J., Mori, M., Hirshman, M.F., Kristiansen, K., Tsang, J.S., White, A.P., Cypess, A.M., Goodyear, L.J., and Tseng, Y.H. (2015a). MicroRNA-455 regulates brown adipogenesis via a novel HIF1an-AMPK-PGC1alpha signaling network. EMBO reports 16, 1378-1393.
Zhang, J., Hug, B.A., Huang, E.Y., Chen, C.W., Gelmetti, V., Maccarana, M., Minucci, S., Pelicci, P.G., and Lazar, M.A. (2001). Oligomerization of ETO is obligatory for corepressor interaction. Molecular and cellular biology 21, 156-163.
Zhang, L., Tumer, Z., Mollgard, K., Barbi, G., Rossier, E., Bendsen, E., Moller, R.S., Ullmann, R., He, J., Papadopoulos, N., Tommerup, N., and Larsen, L.A. (2009). Characterization of a t(5;8)(q31;q21) translocation in a patient with mental retardation and congenital heart disease: implications for involvement of RUNX1T1 in human brain and heart development. European journal of human genetics : EJHG 17, 1010- 1018.
Zhang, W., Yu, Y., Hertwig, F., Thierry-Mieg, J., Thierry-Mieg, D., Wang, J., Furlanello, C., Devanarayan, V., Cheng, J., Deng, Y., Hero, B., Hong, H., Jia, M., Li, L., Lin, S.M., Nikolsky, Y., Oberthuer, A., Qing, T., Su, Z., Volland, R., Wang, C., Wang, M.D., Ai, J., Albanese, D., Asgharzadeh, S., Avigad, S., Bao, W., Bessarabova, M., Brilliant, M.H., Brors, B., Chierici, M., Chu, T.M., Zhang, J., Grundy, R.G., He, M.M., Hebbring, S., Kaufman, H.L., Lababidi, S., Lancashire, L.J., Li, Y., Lu, X.X., Luo, H., Ma, X., Ning, B., Noguera, R., Peifer, M., Phan, J.H., Roels, F., Rosswog, C., Shao, S., Shen, J., Theissen, J., Tonini, G.P., Vandesompele, J., Wu, P.Y., Xiao, W., Xu, J., Xu, W., Xuan, J., Yang, Y., Ye, Z., Dong, Z., Zhang, K.K., Yin, Y., Zhao, C., Zheng, Y., Wolfinger, R.D., Shi, T., Malkas, L.H., Berthold, F., Tong, W., Shi, L., Peng, Z., and Fischer, M. (2015b). Comparison of RNA-seq and microarray-based models for clinical endpoint prediction. Genome biology 16, 133.
Zhang, X., Dong, Z., Zhang, C., Ung, C.Y., He, S., Tao, T., Oliveira, A.M., Meves, A., Ji, B., Look, A.T., Li, H., Neel, B.G., and Zhu, S. (2017). Critical Role for GAB2 in Neuroblastoma Pathogenesis through the Promotion of SHP2/MYCN Cooperation. Cell reports 18, 2932-2942.
Zhao, H., Jin, B., Zhang, X., Cui, Y., Sun, D., Gao, C., Gu, Y., and Cai, B. (2015). Yangjing Capsule Ameliorates Spermatogenesis in Male Mice Exposed to
References 269 Cyclophosphamide. Evidence-based complementary and alternative medicine : eCAM 2015, 980583.
Zhao, X., Yang, Y., Sun, B.F., Shi, Y., Yang, X., Xiao, W., Hao, Y.J., Ping, X.L., Chen, Y.S., Wang, W.J., Jin, K.X., Wang, X., Huang, C.M., Fu, Y., Ge, X.M., Song, S.H., Jeong, H.S., Yanagisawa, H., Niu, Y., Jia, G.F., Wu, W., Tong, W.M., Okamoto, A., He, C., Rendtlew Danielsen, J.M., Wang, X.J., and Yang, Y.G. (2014a). FTO- dependent demethylation of N6-methyladenosine regulates mRNA splicing and is required for adipogenesis. Cell research 24, 1403-1419.
Zhao, Y., Hongdu, B., Ma, D., and Chen, Y. (2014b). Really interesting new gene finger protein 121 is a novel Golgi-localized membrane protein that regulates apoptosis. Acta biochimica et biophysica Sinica 46, 668-674.
Zhou, G., Liu, Z., and Myers, J.N. (2016). TP53 Mutations in Head and Neck Squamous Cell Carcinoma and Their Impact on Disease Progression and Treatment Response. Journal of cellular biochemistry 117, 2682-2692.
Zimmerman, K.A., Yancopoulos, G.D., Collum, R.G., Smith, R.K., Kohl, N.E., Denis, K.A., Nau, M.M., Witte, O.N., Toran-Allerand, D., Gee, C.E., and et al. (1986). Differential expression of myc family genes during murine development. Nature 319, 780-783.
Zirath, H., Frenzel, A., Oliynyk, G., Segerstrom, L., Westermark, U.K., Larsson, K., Munksgaard Persson, M., Hultenby, K., Lehtio, J., Einvik, C., Pahlman, S., Kogner, P., Jakobsson, P.J., and Henriksson, M.A. (2013). MYC inhibition induces metabolic changes leading to accumulation of lipid droplets in tumor cells. Proceedings of the National Academy of Sciences of the United States of America 110, 10258-10263.
References 270 APPENDICES
Appendix 1 Expanded exome sequencing gene list for 1590
SNP Ref Var aa Chr Co-ordinate Exon Gene Name Gene activity Phenotype Base Base change Type
9535596 G C N/A SPLICE condensed nuclear Ribosome chromosome, biogenesis endoplasmic reticulum, NON- regulator mitotic metaphase plate 9535744 C T P->L None SYN homolog (S. congression, cerevisiae) nucleolus, nucleus, (Rss1) protein binding, NON- 9536446 G A R->Q ribosome biogenesis SYN
angiotensin maturation, 135250695 C T N/A SPLICE aspartic-type endopeptidase activity, cytoplasmic part, Mice 1 endopeptidase activity, homozygous extracellular region and knock-out space, insulin-like exhibit growth factor receptor abnormal binding, kidney kidney Renin 1 NON- development, male morphology 135252182 T G S->A structural SYN gonad development, & (Ren1) mesonephros physiology development, peptidase and activity, proteolysis, decreased receptor binding, systemic regulation of MAPKKK arterial blood cascade and blood pressure. NON- pressure, response to 135255559 C A A->E SYN cAMP, cGMP, drug and stress
Appendices 271
SNP Ref Var aa Chr Co-ordinate Exon Gene Name Gene activity Phenotype Base Base change Type DNA-dependent, fat cell differentiation, metal ion binding, Homozygous nucleus, protein disruption of binding, protein Runt-related this gene homodimerisation transcription results in activity, regulation of factor 1; increased NON- DNA binding, 13816819 T C Y->H translocated to, perinatal SYN regulation of 1 (cyclin D- lethality and transcription, related) surviving sequence-specific (Runx1t1) animals show DNA binding severe growth transcription factor retardation. activity, transcription repressor activity, zinc ion binding NON- 88248779 T G K->T SYN cytokine activity, 4 cytokine receptor NON- Interferon alpha binding, defence 88248879 G C L->V None SYN 12 (Ifna12) response, extracellular region, extracellular NON- space, response to virus 88248880 G T D->E SYN Predicted gene NON- 143245377 T G E->D 13088 Unknown None SYN (Gm13088) PRAME family 143487833 G A N/A SPLICE member 6 Unknown None (Pramef6) NON- 143539649 G A R->C SYN Predicted gene 13109 Unknown None C-> NON- 143539935 G T (Gm13109) Stop SYN Predicted gene NON- 143921108 A C E->D 13128 Unknown None SYN (Gm13128) Rab GTPase activator & regulator, cell cycle & division, cytoplasm, Ecotropic viral NON- cytoskeleton, 5 108249512 T C E->G integration site None SYN intracellular, 5 (Evi5) microtubule organizing centre, nucleus, protein binding
Appendices 272
SNP Ref Var aa Chr Co-ordinate Exon Gene Name Gene activity Phenotype Base Base change Type ATP biosynthetic process, L-aspartate & L-glutamate transmembrane Mice transporter activity, L- homozygous Solute carrier glutamate & aspartate for disruptions family 25 transport, calcium ion in this gene (mitochondrial binding, cellular appear normal, NON- carrier, adenine 6 5992258 C G G->A respiration, integral to healthy and SYN nucleotide membrane, malate- fertile, translocator), aspartate shuttle, although they member 13 mitochondrial inner have a number (Slc25a13) membrane, of metabolic mitochondrion, defects. response to calcium ion, transmembrane transport
NON- 129818787 T C N->D SYN
129818852 A G N/A SPLICE
NON- 129822210 G T P->T SYN Mice homozygous Killer cell lectin- for a knock-in like receptor, allele exhibit NON- 6 129822392 T C Q->R subfamily A, binding reduced SYN member 17 plasmacytoid (Klra17) dendritic cell number and NON- 129822393 G T Q->K function. SYN
129823299 C T N/A SPLICE
NON- 129823383 T C I->V SYN
Appendices 273
SNP Ref Var aa Chr Co-ordinate Exon Gene Name Gene activity Phenotype Base Base change Type NON- 129858925 T G K->N SYN Mice homozygous Killer cell lectin- NON- for a knock- 129858935 C T S->N like receptor, binding, plasma SYN out allele subfamily A, membrane, receptor exhibit normal NON- member 5 activity, sugar binding 129858956 T A H->L NK and T cell SYN (Klra5) morphology NON- and function. 129858960 T G N->H SYN NON- 129963408 G A H->Y SYN NON- 129963442 T G K->N SYN NON- Killer cell lectin- binding, cell adhesion, 129963453 A G F->L SYN like receptor, integral to membrane, subfamily A, plasma membrane, None NON- 129963464 C T G->E member 6 receptor activity, sugar SYN (Klra6) binding
6 129963507 G A N/A SPLICE
129963508 C G N/A SPLICE
130132397 C T N/A SPLICE Killer cell lectin- like receptor NON- binding, receptor 130135550 C G K->N subfamily A, None SYN activity, sugar binding member 9 NON- (Klra9) 130141283 G A L->F SYN NON- 130181593 C T E->K SYN binding, cell adhesion, Killer cell lectin- NON- external side of plasma 130181605 C T V->M like receptor, SYN membrane, integral to subfamily A, None membrane, plasma member 7 130181641 G C N/A SPLICE membrane, receptor (Klra7) activity, sugar binding 130181643 A G N/A SPLICE
Appendices 274
SNP Ref Var aa Chr Co-ordinate Exon Gene Name Gene activity Phenotype Base Base change Type
NON- binding, external side of 130327773 A T M->K Killer cell lectin- SYN plasma membrane, like receptor, plasma membrane, 6 subfamily A, None signal transducer member 1 NON- activity, signal 130327857 T C D->G (Klra1) SYN transduction
G-protein coupled receptor activity, G- protein coupled receptor protein signalling Vomeronasal 2, pathway, biological NON- 5432415 A G L->P receptor 28 process, cellular None SYN (Vmn2r28) component, integral to membrane, molecular function, plasma membrane, receptor activity G-protein coupled 7 receptor protein signalling pathway, Vomeronasal 1 biological process, NON- 23284063 T C H->R receptor 151 cellular component, None SYN (Vmn1r151) integral to membrane, molecular function, pheromone receptor activity HIRA Biological process, NON- interacting 134008317 G A R->Q molecular function, None SYN protein 3 nucleus, protein binding (Hirip3)
Appendices 275
SNP Ref Var aa Chr Co-ordinate Exon Gene Name Gene activity Phenotype Base Base change Type
DNA binding, biological Purine-rich process, cellular 34497862 G A N/A SPLICE element binding None component, molecular protein G (Purg) function, nucleus cell cycle, cell differentiation, cytoplasm, cytosol, ligase activity, male meiosis I, metal ion Homozygotes binding, multicellular for a targeted organismal null mutation development, neuron exhibit retarded apoptosis, nucleus, post- postnatal embryonic development, growth, and proteasomal ubiquitin- high pre- dependent protein weaning and catabolic process, post-weaning Seven in protein binding, protein NON- mortality. 89249156 C T D->N absentia 1A homodimerisation SYN Surviving (Siah1a) activity, protein females are ubiquitination, protein 8 sub-fertile, ubiquitination involved having few, if in ubiquitin-dependent any, offspring, protein catabolic while males are process, regulation of sterile due to a multicellular organism block at growth, meiotic spermatogenesis, metaphase I. ubiquitin-dependent protein catabolic process, ubiquitin- protein ligase activity, zinc ion binding aminomethyltransferase activity, enzyme Glycine binding, glycine cleavage system catabolic process, NON- 119517283 C T S->N protein H glycine cleavage None SYN (aminomethyl complex, glycine carrier) (Gcsh) decarboxylation via glycine cleavage system, mitochondrion
Appendices 276
SNP Ref Var aa Chr Co-ordinate Exon Gene Name Gene activity Phenotype Base Base change Type Solute carrier family 25 NON- (mitochondrial 10 90586435 C T V->I mitochondrion None SYN carrier, phosphate carrier), member 3 (Slc25a3) Family with Biological process, sequence NON- cellular component, 11 46220261 G C A->P similarity 71, None SYN molecular function, member B nucleus (Fam71b) RNA splicing, RS domain binding, nucleic acid Serine/arginine- NON- binding, nucleotide 12 82050956 T A S->R rich splicing None SYN binding, protein binding, factor 5 (Srsf5) regulation of cell cycle, response to wounding RIKEN cDNA C- NON- 8197033 T A 2410089E03 gene Unknown None >Stop SYN (2410089E03Rik) Mice homozygous for gene trapped alleles exhibit embryonic lethality. Mice 15 actin binding,biological homozygous TRIO and F-actin process, cytoplasm, NON- for a targeted 78797491 G A D->N binding protein cytoskeleton, nucleus, SYN allele (Triobp) protein binding, ubiquitin eliminating protein ligase binding isoforms 4 and 5 exhibit profound deafness associated with stereocilia fragility and degeneration.
Appendices 277
SNP Ref Var aa Chr Co-ordinate Exon Gene Name Gene activity Phenotype Base Base change Type Heterozygous mutation of this gene results in GTP binding, GTP hyperactivity, catabolic process, reduced GTPase activity, anxiety, cytoplasm, cytoplasmic impaired spatial microtubule, working cytoskeleton, cytosol, memory, and NON- Tubulin, alpha 15 98781165 A C I->S microtubule-based abnormalities in SYN 1A (Tuba1a) movement, nucleotide the laminar binding, protein architecture of complex, protein the polymerization, hippocampus structural molecule and cortex, activity accompanied by impaired neuronal migration. Keratin NON- associated 88869437 C A G->V Unknown None SYN protein 16-2 (Krtap16-2) 16 Keratin Biological process, NON- associated cellular component, 88874217 A T F->Y None SYN protein 16-1 intermediate filament, (Krtap16-1) molecular function acting on acid NON- Predicted gene anhydrides, GTP 18 60430650 T C I->M None SYN 4841 (Gm4841) binding, hydrolase activity, membrane
Appendices 278
SNP Ref Var aa Chr Co-ordinate Exon Gene Name Gene activity Phenotype Base Base change Type integral to membrane, integrator complex, Integrator NON- membrane, molecular 8971720 G A R->Q complex subunit None SYN function, nucleus, protein 5 (Ints5) binding, snRNA processing transferring phosphorus- NON- containing groups, ATP 8971809 C A L->M SYN binding, apical plasma membrane, brush border membrane, calcium channel activity, cation NON- 8971843 G C R->P Transient channel activity, integral to SYN receptor membrane, ion channel potential cation activity, ion transport, channel, kinase activity, membrane, None NON- 8971953 G A V->I subfamily M, metal ion binding, metal SYN member 6 ion transport, nucleotide (Trpm6) binding, protein binding, 19 protein phosphorylation, protein serine/threonine 18846802 T C N/A SPLICE kinase activity, response to toxin, transferase activity, transmembrane transport caveola, cellular response to heat, cytoplasmic vesicle, cytoplasmic vesicle membrane, integral to membrane, intracellular membrane-bounded organelle, nuclear NON- Myoferlin 38007147 G A R->C membrane, nucleus, None SYN (Myof) phospholipid binding, plasma membrane, plasma membrane repair, protein binding, regulation of vascular endothelial growth factor receptor signalling pathway
NON- 71598719 A G S->P SYN L antigen Biological process, cellular X family, member component, molecular None 3 (Lage3) function NON- 71598730 G T A->D SYN
Appendices 279 Appendix 2 Expanded exome sequencing gene list for 1929
SNP Ref Var aa Chr Co-ordinate Exon Gene Name Gene activity Phenotype Base Base change Type Homozygotes have NON- thrombocytopenia microtubule 2 174457062 T C V->A SYN Tubulin, beta 1 resulting from a class VI (Tubb1) cytoskeleton defect in generating pro- platelets. protein kinase activity; cell adhesion; ephrin receptor activity; neuron remodeling; Mice positive regulation of homozygous for MAPK cascade; this targeted regulation of cell 4 136956647 C T N/A SPLICE Ephrin receptor mutation have A8 (Epha8) adhesion mediated by defects in axon integrin; neuron projection in the projection brain development; early endosome membrane; positive regulation of phosphatidylinositol catalytic activity; cellular component; NON- 6 41355493 T A Y->N Trypsin 10 proteolysis; serine- No phenotype SYN (Try10) type endopeptidase activity protein and NON- Oxysterol binding 6 50300981 T A D->V protein-like 3 phospholipid No phenotype SYN (Osbpl3) binding; lipid
Appendices 280 transport
protein binding and folding; chaperonin- containing T- R- NON- Chaperonin 6 85464927 A T containing Tcp1, complex; cellular No phenotype >Stop SYN subunit 7 (Cct7) protein metabolic process; ATP binding
NON- Ring finger zinc ion and protein 7 102031527 A C M->R protein 121 No phenotype SYN binding; (Rnf121) G-protein coupled G protein-coupled receptor activity; G- NON- receptor, family 7 118984352 A G L->P C, group 5, protein coupled No phenotype SYN member B receptor signalling (Gprc5b) pathway Homozygous null SH3 domain mice display binding; cell partial postnatal
A disintegrin and adhesion; zinc ion lethality, NON- 7 133967972 A G V->A metallopeptidase and protein binding; decreased brown SYN domain 12 (Adam12) proteolysis; fat, and impaired metalloendopeptidase formation of neck activity and interscapular muscles. oxidation-reduction A homozygous process; glutathione- mutation between disulfide reductase exons 1-2 results NON- activity; glutathione in a decreased 8 33653615 G A A->T Glutathione SYN reductase (Gsr) binding; retinal artery-to- spermatogenesis; vein ratio. oxidoreductase Another small activity; protein deletion of exons
Appendices 281 homodimerization 2-5 has no activity; cell redox phenotypic effect. homeostasis; NADP binding; flavin adenine dinucleotide binding; cytosol; glutathione metabolic process integral to membrane; cellular NON- Transmembrane 11 78511926 C A R->L protein 199 component; No phenotype SYN (Tmem199) biological process; molecular function negative regulation of cell migration, Homozygous MAP kinase activity, embryos die by Rac protein signal E14.5 (enlarged transduction, head and chest, astrocyte pale liver, differentiation and microphthalmia, endothelial cell cardiac defects proliferation; and delayed regulation of organ Neurofibromatosis 11 79339991 C G N/A SPLICE angiogenesis; spinal 1 (Nf1) development). cord development; Heterozygotes peripheral nervous have elevated system development; astrocyte number, sympathetic nervous predisposition to system development; multiple tumor negative regulation types and of neuroblast learning/memory proliferation; adrenal deficits. gland development; Schwann cell
Appendices 282 development;
Potassium channel NON- protein binding; ion 11 115420219 A G S->G tetramerisation No phenotype SYN domain containing channel activity 2 (Kctd2) integral to membrane;
Interphotoreceptor proteinaceous NON- matrix 16 56252140 G A E->K extracellular matrix; No phenotype SYN proteoglycan 2 (Impg2) hyaluronic acid, protein and heparin binding
Appendices 283