Therapeutic implications of interferon regulatory factor (IRF)-1 and IRF-2 in diffusely infiltrating astrocytomas (DIA):: response to interferon (IFN)-β in glioblastoma cells and prognostic value for DIA

The precise mechanisms governing the direct effect of IFN-beta, including apoptosis induction, are not yet fully understood. To gain a better insight into these mechanisms, we investigated the signaling pathways focusing particularly on interferon regulatory factor 1 (IRF-1) and IRF-2 in glioblastoma cell lines. Furthermore, we attempted to determine whether or not IRF-1 and IRF-2 act as additional prognostic indicators in diffusely infiltrating astrocytomas (DIA). We first assessed the cytotoxic effects of IFN-beta based on a cell growth study and modified MTT assay, and then quantified the apoptosis using a sandwich enzyme immunoassay following IFN-beta treatment in the cell lines, U-87MG, T98G, and A-172. Subsequently, we carried out an analysis of apoptosis-related molecules as evaluated by densitometric analysis of Western blots, focusing on IRF-1 and IRF-2, and two major initiator caspases, caspase-8 and caspase-9. Furthermore, we assessed the expression of type I IFN receptor, IRF-1, and IRF-2 using immunohistochemical techniques in 63 DIA (15 of WHO grade II, 18 of grade III, and 30 of grade IV), and analyzed their impact on prognosis. An increase in apoptosis was apparent after 48 h of IFN-beta treatment (1 x 10(4) IU/ml) in T98G but not in U-87MG or A-172. IFN-beta treatment for 6 h significantly enhanced the expression of IRF-1 in all three cell lines. However, an enhanced expression of IRF-2 was observed only in the not-most-sensitive, non-apoptosis-induced U-87MG and A-172. While minimal processing of caspase-8 was noted in the three cell lines throughout the experiment, caspase-9 activation was observed in the apoptosis-detected T98G after 48 h of treatment, as indicated by a 1.33-fold increase (P=0.037). On the other hand, the IRF-1 LI and IRF-1/IRF-2 LI ratio were greater in low-grade DAI, and were negatively correlated with the histopathological grade in DIA (P=0.017 and P=0.001, respectively). Furthermore, the IRF-1/IRF-2 LI ratio was negatively correlated with the MIB-1 LI in DIA (P=0.004), and represented an independent and most powerful determinant of overall survival compared to other conventional prognostic factors (P=0.018). However, the relation was not statistically significant when only patients with high-grade DIA were assessed. Our findings suggest that up-regulation of IRF-1 and IRF-2 might be an important determinant of susceptibility to IFN-beta mediated cytotoxicity including apoptosis. Furthermore, the IRF-1/IRF-2 LI ratio may reflect the proliferative state of DIA and constitute an important prognostic marker in DIA. Thus, IRF-1 and IRF-2 could represent one of the therapeutic target sites for the regulation of cell growth in DIA.