LAUSR

Glioblastoma multiforme (GBM) is the most aggressive brain tumor with median patient survival of 12-15 months. To facilitate treatment development, bioengineered GBM models that adequately recapitulate the in vivo tumor microenvironment are needed. Matrix-encapsulated multicellular spheroids represent such model because they recapitulate solid tumor characteristics, such as dimensionality, cell-cell, and cell-matrix interactions. Yet, there is no consensus as to which matrix properties are key to improving the predictive capacity of spheroid-based drug screening platforms. We used a hydrogel-encapsulated GBM spheroid model, where matrix properties were independently altered to investigate their effect on GBM spheroid characteristics and drug responsiveness. We focused on hydrogel degradability, tuned via enzymatically degradable crosslinkers, and hydrogel adhesiveness, tuned via integrin ligands. We observed increased cellular infiltration of GBM spheroids and increased resistance to temozolomide in degradable, adhesive hydrogels compared to spheroids in non-degradable, non-adhesive hydrogels or to free-floating spheroids. Further, a higher infiltration index was noted for spheroids in adhesive compared to non-adhesive degradable hydrogels. For spheroids in degradable hydrogels, we determined that infiltrating cells were more susceptible to temozolomide compared to cells in the spheroid core. The temozolomide susceptibility of the infiltrating cells was independent of integrin adhesion. We could not attribute differential drug responses to differential cellular proliferation or to limited drug penetration into the hydrogel matrix. Our results suggest that cell-matrix interactions guide GBM spheroid drug responsiveness and that further elucidation of these interactions could enable the engineering of more predictive drug screening platforms.