Application of murine models of osteosarcoma to preclinical target discovery and
validation
Carl Walkley
St.Vincent’s Institute, Australia
Singapore Sarcoma Symposium, 12-13 Sept 2015
Osteosarcoma
• most common cancer of bone
• peak incidence at age 15 years
• affects lower long bones
• current treatment - surgical resection and chemotherapy
• standard of care has not significantly improved in 30 years
• survival - primary only 70% at 5 yrs, metastatic or recurrent disease 20% at 5 yrs
• 5th leading cause of childhood cancer and a high morbidity and mortality rate
Challenges in the Clinical Study of OS
Novel approaches to drug and drug target discovery, with particular attention to drug development for
metastatic osteosarcoma, are needed
Translation of basic research in OS is hampered by:
• relatively rare tumour
• multi-institutional national clinical trials require years of enrolment
• standard of care includes amputation, complicating Phase II/III trials
• in last 20 years only one significant advance in treatment to standard upfront
chemotherapy resulting in a modest improvement in overall survival
Genetics of Osteosarcoma
Chen et al. Cell Reports 7, 104–112, April 10, 2014
Perry et al. PNAS E5564–E5573, doi: 10.1073/pnas.1419260111
Mouse Models of Human Cancer
A valid and biologically meaningful mouse model of human cancer should recapitulate
• genetic basis of disease
• appropriate age of disease development
• tissue restricted nature of disease
• metastasis to same sites
• gene expression signatures
• histology
Provide “robust pre-clinical infrastructure through molecularly defined model systems that are
directly reflective of their human counterpart and sufficient models numbers for one to
approach the disease diversity found in humans” (W Sellers, A blueprint for advancing genetics-based cancer therapy.
Cell 2011,147;26-31)
Bedside to Bench: Osteosarcoma Modelling
Known genetic lesions in familial OS predisposition and sporadic OS: • Li-Fraumeni syndrome (p53+/-)
• Hereditary retinoblastoma (RB+/-)
• Rothmund-Thomson Syndrome (Recql4 mutated; Ng et al PLoS Genetics 2015 Apr 10;11(4):e1005160)
Osx-Cre
Models of two histological subtypes of human OS
Conventional human OS:
60% osteoblastic OS
15% fibroblastic OS
15% chondroblastic OS
10% various rare forms (upto 8 types)
p53 shRNA OS Normal bone p53fl/fl OS
Mutsaers et al., Bone. 2013 Jul;55(1):166-78
Walkley et al., Genes & Dev. 2008 22: 1662-1676
Using models of human OS
Paired primary and metastatic cell cultures at early passage – not available from human biopsy material Use these for:
• define new therapeutic strategies – chemical/drug screen
• identify interactions with chemotherapy that could be used therapeutically – siRNA screen
• test new therapies based on hypothesis driven understanding of OS
Identifying OS susceptibility pathways
Ankita Gupte, Kate Gould, Dr Kaylene Simpson
Whole Genome Screening
siRNA +Doxorubicin
siRNA alone
Gupte et al., Clin Can Res. 2015 Jul 15;21(14):3216-29
Primary Screen Data from mouse OS cells
Ankita Gupte, Kate Gould, Alistair Chalk Gupte et al., Clin Can Res. 2015 Jul 15;21(14):3216-29
Pathways disproportionately enriched as increasing OS cell death
Ankita Gupte, Alistair Chalk Gupte et al., Clin Can Res. 2015 Jul 15;21(14):3216-29
siRNA screen – chemical validation of results
Ankita Gupte, Dr Emma Baker, Scott Taylor
Secondary and tertiary screens completed - validated candidates identified
Established if any chemicals/drugs were available that targeted these pathways
Could identify inhibitors for 3 of the 4 pathways of highest interest (1 a drug, 2 chemical/preclinical)
Drug/Chemical Screening - HTS
Dr Emma Baker, Dr Kurt Lackovic, Prof David Huang, Dr Chris Burns, Soo San Wan
Chemical/Drug screen – with Prof David Huang, WEHI
3 primary and metastatic paired cell lines from mouse fibroblastic OS model
Screened against 131 kinase inhibitors in an 11 point dose response in duplicate
Library composed of clinically approved, preclinical and agents with known activity that didn’t progress
Can use to identify pathways that may be tractable or for rapid translation of the research with agents already in clinical use.
Drug/Chemical Screening - HTS
Dr Emma Baker, Dr Kurt Lackovic, Prof David Huang, Dr Chris Burns, Soo San Wan Gupte et al., Clin Can Res. 2015 Jul 15;21(14):3216-29
Dual targeting of the PI3K and mTOR axis is a species conserved OS vulnerability
Dr Emma Baker, Scott Taylor, Ankita Gupte, Prof Mike Dyer & Drs E Stewart/A Loh/A Shelat/A Karlstrom (St Jude’s)
Gupte et al., Clin Can Res. 2015 Jul 15;21(14):3216-29
Dual PI3K/mTOR inhibitors cause on target inhibition and cell cycle modulation
Dr Emma Baker, Scott Taylor, Ankita Gupte Gupte et al., Clin Can Res. 2015 Jul 15;21(14):3216-29
Dual PI3K/mTOR inhibitors cause apoptosis of primary OS derived cells
Dr Emma Baker, Scott Taylor, Ankita Gupte Gupte et al., Clin Can Res. 2015 Jul 15;21(14):3216-29
Dual targeting of the PI3K and mTOR axis – PI3Ka dependent effects
Dr Emma Baker, Scott Taylor, Ankita Gupte Gupte et al., Clin Can Res. 2015 Jul 15;21(14):3216-29
Combining clinically relevant PI3Ka and mTOR inhibitors
Dr Emma Baker, Scott Taylor, Ankita Gupte, Dr Jayesh Desai Gupte et al., Clin Can Res. 2015 Jul 15;21(14):3216-29
The intersection of chemical and genetic biology – PI3Kca/mTOR inhibitors in OS
Independent screening approaches have converged on the PI3K/mTOR pathway as a tractable candidate in OS
PI3K specific or mTOR specific agents have limited activity in OS
Dual pathway inhibition appears to be required for maximal activity
Activity requires inhibition of PIK3CA specifically
Pathway is conserved across murine and human OS and is subtype independent
The intersection of chemical and genetic biology – PI3Kca/mTOR inhibitors in OS
Proc Natl Acad Sci U S A. 2014 Dec 23;111(51):E5564-73. doi: 10.1073/pnas.1419260111. Epub 2014 Dec 15.
The intersection of chemical and genetic biology – PI3Kca/mTOR inhibitors in OS
Independent screening approaches have converged on the PI3K/mTOR pathway as a tractable candidate in OS
PI3K specific or mTOR specific agents have limited activity in OS in clinical trials to date
Dual pathway inhibition appears to be required for maximal activity
Activity requires inhibition of PIK3CA specifically – BYL719 now in clinical trial in BrCa
Pathway is conserved across murine and human OS and is subtype independent
Demonstrates efficacy of applying GEMM to preclinical target identification and validation
Collectively the data warrant further testing of PI3Kca / mTOR inhibitors in OS
Acknowledgements
St. Vincent’s Institute
Ankita Gupte
Scott Taylor
Dr Alvin Ng
Dr Alistair Chalk
Dr Mannu Walia
Dr Emma Baker
Dr Anthony Mutsaers (U Guelph)
St. Vincent’s Hospital
Dr Meaghan Wall
A/Prof John Slavin
Funding
AACR-Aflac Inc
Colin North
Australian Sarcoma Study Group
Zig Inge Foundation
Walter and Eliza Hall Institute
Prof David Huang
Dr Chris Burns
Dr Kurt Lackovic
Soo San Wan
Victorian Centre for Functional Genomics
A/Prof Kaylene Simpson
Kate Gould
Dr Piyush Madhamshettiwar
St Jude’s Children’s Hospital
Prof Mike Dyer
Dr Elizabeth Stewart
Dr Amos Loh
Dr Anang Shelat
Dr Asa Karlstrom
Peter MacCallum Cancer Centre
Dr Jayesh Desai
SAHMRI, University of Adelaide
Prof Andrew Zannettino
Dana-Farber Cancer Institute, Boston
Prof Stuart Orkin
Dr Katherine Janeway
Dr Jen Perry