Go to symposium website → www.slas.org/HighContentAnalysis
Back To Schedule
Monday, October 21 • 2:30pm - 3:00pm
Poster Presentation #8- LaFOS: A Novel 4D Morphometric Phenotypic Profiling Method to Uncover Therapeutic Vulnerabilities in Patient-Derived 3D Tumor Models

Sign up or log in to save this to your schedule, view media, leave feedback and see who's attending!

3D-spheroid cultures are powerful tools for in vitro models for biology, but size and shape diversity within the culture is largely ignored and also their current end point data outputs (like metabolic measurements) are resource intensive and we cannot afford real-time kinetics at the primary drug screening level. We intent to develop robust Label free Oncology Score (LaFOS) using brightfield image of 3D tumor spheroids in different time intervals. We developed a computational image scoring methodology, that integrates morphometric profiling, segmentation, filtering, and analysis for brightfield images of 3D culture and provide us with cytotoxicity level without any metabolic end point readouts like cell viability assay. LaFOS is accurate and flexible like current existing platforms, and we illustrate its potential by studying 600 plus small molecules that are used in the clinic or in advanced phases of development. We tested this computational method in parallel with cell viability assay (cell titer Glo) in 15 primary patient derived models and 3 commercial cancer cell lines in comparison with traditional 2D monolayer models. We found Pearson correlation of 0.81 for LaFOS methodology and 0.95 for cell viability assay in finding the cellular cytotoxicity. Our results suggest that, there is differential drug response between 2D and 3D model for specific class of compounds. The compound connectivity map showed different targets specificity and network structure between 2D and 3D models. We observed pathway specific enrichment of (EGFR, PI3K and mTOR inhibitors) in 3D models and cell cycle/division related enrichment (HDAC, CDK, Topo and Tubulin) in 2D models. RNA-sequence analysis of 13 cell lines showed 151 pan-differentially expressed genes in 2D vs 3D. Immunofluorescence and western blot analysis showed HIF-1α was high in 3D models and Ki67 was found high in 2D models this data correlates to architecture dependent drug response and gene expression data. Also 3D models showed more activity of pAKT and pERK could be a reason for pathway specific inhibitor activity in 3D models. We did a validation of hits obtained from in vitro screen using in vivo PDX model for ZSTK474 a PI3K inhibitor. Overall we developed a computational platform (LaFOS) that can measure 3D spheroid cell viability in real-time fashion using morphometric signature obtained from label free brightfield phenotypes.


Giridharan Periyasamy, PhD

Platform Leader, Genome Institute of Singapore
Dr. Giridharan Periyasamy is a Platform Leader in Centre for High Throughput Phenomics (CHiP-GIS) at the Genome institute of Singapore. He is the domain expert in Assay automation, miniaturization, robotics, and High Content Imaging technology. His research focus is on the Cancer... Read More →

Monday October 21, 2019 2:30pm - 3:00pm BST
Sherry Coutu Seminar Suite Foyer