Statistical comparisons shown indicate comparison with phenotype on 0.2kPa gels. total surgical resection unlikely and the probability of recurrence high1. Despite the implementation of new therapies, invasion remains a major impediment to curing GBM. Based chiefly on advances in the breast cancer research field, it is increasingly realized that mechanical cues within the external tissue environment play a major role in invasive capability2,3,4,5,6,7,8. Initial studies suggested that GBM cell invasion is inhibited on soft brain-like matrices and increases with higher matrix rigidities, thus suggesting that GBM are rigidity-sensitive9,10,11. However, these LJI308 findings are contrasted by newer analyses of primary patient-derived GBM lines that were LJI308 shown to be rigidity-independent12,13. It is now realized that GBMs represent molecularly and genetically distinct subclasses14,15,16,17 and the majority of commonly used, repeatedly cultured GBM lines (such as those used in the initial studies of rigidity response) are of the Mesenchymal subclass14. Thus, in order to establish whether all primary patient-derived lines are indeed rigidity-independent or whether there is variation between patients, it is important to analyse a range of primary lines. In the present study we compare the rigidity dependent migration behaviour of 5 primary patient-derived cell lines. In common with other solid tumours, GBM tumours are also subject to different mechanical forces. The matrix proteins secreted by GBM counteracts expansion of the tumour tissue and thereby increases mechanical forces on the tumour18,19. For rapidly growing GBMs and at later stages in disease progression, forces are increased on the whole brain as the rigid skull prevents tissue expansion20. The natural ECM of the brain tissue and in particular the ECM secreted by the LJI308 GBM cells stiffens in response to the increasing pressure and the resulting tissue strain21. Importantly, glial cells do not migrate efficiently in the very soft brain tissue, and strain stiffening of the ECM improves glial cell migration: disabling this effect reduces GBM invasion11. Finally, GBMs stiffen in response to compressive pressure21. Therefore the natural course of the disease can induce mechanical forces and tumour stiffening. Moreover, brain tissue that is encountered by invasive GBM cells is mechanically heterogeneous with micro-regional stiffness values ranging from as low as 0.1?kPa to as high as ~10?kPa22. This mechanical heterogeneity is important in the regulation of normal brain biology22,23,24,25 and thus may also influence tumour cells as they invade different brain regions. In response to external mechanical forces, cells exert increasing Rho-GTPase dependent contractile force through their acto-myosin cytoskeleton6. This leads to greater traction forces on the surrounding extracellular matrix (ECM) and enhanced migration and invasion. The ability of cancer cells to proliferate and migrate on engineered substrates of defined rigidities is further reflected in their LJI308 abilities to grow in specific tissue environments in vivo26. The limitation that most commonly used cultured CLG4B GBM lines are of the mesenchymal subclass14 is overcome by the LJI308 isolation of primary patient-derived GBM lines and maintenance in serum-free, Glioma Neural Stem (GNS) media at low passage27,28. In the present study we reveal diverse rigidity-dependent responses in primary GBM lines. Results JK2, WK1, RN1 and PR1 cell morphology are regulated by substrate stiffness, but SJH1 cell morphology is stiffness insensitive We compared the responses of 5 primary GBM lines (JK2, SJH1, WK1, RN1 and PR1). Cells were grown on polyacrylamide hydrogels (PAM gels) of defined rigidity, corresponding to the reported range.