Because Src inhibitors can reverse Src-induced suppression of PTEN function [24], the ineffectiveness of bosutinib on these cells actually suggested

a stronger LOF effect of nonsense mutations over missense mutations. Importantly, nonsense mutations of PTEN displayed favorable responses to bryostatin 1 (Sigma), AZ628 CT99021 cost (Sigma), and procaspase activating compound-1 (PAC-1, Sigma; Figure 4, B–D), suggesting that the adverse effects of nonsense mutations might be targetable. Because PTEN loss causes the activation of protein kinase C (PKC), it is not surprising that bryostatin (PKC inhibitor) can suppress the growth of cells carrying nonsense PTEN mutations. Another adverse consequence of PTEN loss is the cooperation with Ras/Raf/mitogen- activated protein kinases (MAPK) for promoting tumorigenesis [25], and this may explain the enhanced effect of AZ628 (a Raf inhibitor) against nonsense mutations. Finally, loss of PTEN inhibits caspase 3 activity, and this may be the underlying mechanism for the effectiveness of PAC-1 (a caspase 3 activator) on PTEN nonsense mutations. Taken together, the drug sensitivity profile of PTEN nonsense mutations is in good consistency with its severe LOF phenotype and may provide important information CX-5461 in vivo for its

targeted therapy. Furthermore, we tested the effect of PTEN mutation and expres- sion on overall survival (OS) of patients with GBM. Cox regression survival analyses revealed a link between increased Pten protein level and shorter OS (HR = 1.23, 95% CI = 1.03-1.47; Figure 5A). Patients with upper quarter Pten protein expression displayed significantly

shorter OS (median, 7.5 months) than the rest of patients (median, 15.7 months; Figure 5B). However, no correlation was found between OS and PTEN mutation, mRNA level or promoter methylation ( Figure 5, A and C ). Interestingly, patients with GBM with unregulated Pten protein showed substantial alterations in signaling pathways involved in insulin stimulus, lipid oxidation, DNA damage and MAPK cascade, and inactivation Baricitinib of cell apoptotic process (Figure 6A). The expression level of Pten showed no correlation with CNA fraction in genome or the total number of mutations present in the tumor ( Figure 6, B and C). These findings suggest distinct mechanisms whereby PTEN mutations and altered protein expression affect DFS and OS of patients with GBM. Although the prognostic value of PTEN in GBM has been con- troversial, here, we have demonstrated strong association between PTEN mutation/expression and survival of patients with GBM. The analysis is based on a large number of patients with comprehensive clinical and genomic data, and the combined analysis on genomic stability, signaling pathways, and drug sensitivity provides mechanistic insight into the distinct effects of PTEN mutations. We experimentally validated the effects of PTEN mutations on genomic instability and p53/Gata3 protein levels, thereby confirming the findings in patients with GBM.