ID #695 Uncovering the extracellular matrix of MYC amplified medulloblastoma to develop bespoke disease-relevant 3D hydrogel models
Mike Tseung, Kristy Kehoe, Yana Bevan, Anya Fletcher, Janice Law, Dean Thompson, Jack Goddard, Michael Wilkes, Johnathan Curd, Veronique Marsaud, Olivier Ayrault, Steve Clifford, Rebecca M Hill, Louisa TaylorAbstract
Medulloblastoma is the most common malignant brain tumour in children. Patients with MYC-amplified medulloblastoma (MYC-MBGroup-3) have a dismal prognosis and are refractory to current therapies. Novel treatment strategies have yet to reach the clinic, partly attributable to a lack of disease-relevant modelling strategies. Existing MYC-MBGroup-3 2D in vitro models fail to replicate the tumour microenvironment (TME). In vivo models, although more representative, are laborious and raise ethical considerations. 3D hydrogel models have the potential to more faithfully recapitulate key tumour characteristics including tissue architecture, pathophysiological gradients and the extracellular matrix (ECM), enabling more rapid closer-to-patient modelling.
Unsupervised analysis was performed on a large medulloblastoma RNA-Seq patient cohort (n = 1227) against our custom ECM gene library (n = 912). For the first time, we show that medulloblastoma tumours can be clustered (UMAP/t-SNE) into Principal Molecular Groups utilising ECM genes alone, highlighting a central role in tumour biology. In complementary functional studies, we optimised a permeable hydrogel model mimicking cerebellar stiffness (1.5-2 kPa) that demonstrated stability over a 14day period. When embedded in this baseline hydrogel-model, MYC-MBGroup-3 spheroids exhibited an invasive phenotype and significantly reduced toxicity (CellTox™ Green; p < 0.0001) while maintaining viability (CellTiterGlo®) over 14-days compared with non-scaffolded spheroids, suggestive of extended longevity. To further tune our base hydrogel-model, in vitro ECM binding arrays were integrated with patient transcriptomic differential gene expression analysis (DESeq2) to identify MYC-MBGroup-3 specific incorporable ECM components. Through this dual approach, Hyaluronan, Fibronectin, Collagen I and Collagen IV will next be incorporated into our base cerebellar hydrogels.
In summary, we present a bespoke disease-relevant MYC-MBGroup-3 hydrogel system with early-evidence of enhanced biological relevance and predictive therapeutic capability over standard in vitro culture. Hydrogelmodel longevity will next be assessed by livecell imaging (Incucyte®) and disease-relevance will be evaluated by proteomic and transcriptomic profiling.