Despite the use of a multimodal therapy, the prognosis of GBM did not change in recent years . Cytotoxicity of clinically achievable serum levels of TMZ, which are around 100 μM , was very small in the three glioma cell lines and, furthermore, the primary glioma tumor tested was resistant to TMZ, while Rsv was cytotoxic alone and potentiated the effect of TMZ in these cells. Therefore, drugs that enhance the effect of TMZ have been actively studied [52, 53] and may represent an alternative strategy for combined chemotherapy.
The effect of Rsv in increasing TMZ-induced toxicity and autophagy occurred in all glioma cells tested, demonstrating that it does not involve the p53 pathway, since U251 and U138 cells are p53 mutant. The lack of correlation between autophagy and cytotoxicity in three GBM cell lines and the absence of reversion of the cytotoxicity by 3MA, despite a significant reduction in autophagy, suggest that this is not a central mechanism in the reduction in cell number. Indeed, autophagy induced by co-treatment seems to be protective rather than cytotoxic. In U138 cells, which have high levels of MGMT protein when compared to U251 and U87 cells, TMZ did not trigger G2-cell cycle arrest, while U87 and U251 cells arrested at G2 with TMZ, also suggesting a p53-independent arrest [5, 54]. Furthermore, in agreement with our data in U87 cells, Mhaidat et al. showed that sensitivity of melanoma cells to TMZ was associated with MGMT status, G2 arrest and senescence entry, while no apoptosis was induced .
Rsv induced a transitory S-phase arrest  and a small increase in the phosphorylation of ATM, Chk2 and H2AX in U87 cells, while leading to a more prolonged S-phase arrest in U251 and U138 cells. Others also reported Rsv-induced S-phase arrest via DDR signaling . Interestingly, Rsv induced an increase in nuclear size after 48 h of treatment in 32% of cells, an indication of senescence induction, but after 7 days in DFM no senescent cells remained, suggesting that the undamaged cells repopulated the well. On the other hand, damage induced by TMZ seems to be more long lasting, since 20% of large nuclei were observed after 48 h and 50% of cells were SA-β-gal positive after 7 days in DFM, suggesting that even in the absence of external TMZ, the damage induced in the 48 h of treatment was maintained. Important to notice that the combination of RT induced 25% of large and large irregular nuclei at 48 h and 90% of SA-β-gal positive after 7 days in DFM, suggesting that the presence of Rsv potentiates the long lasting damage that leads to senescence induced by TMZ.
Cell cycle arrest acts as a pro-survival mechanism, since it gives extra time for DNA repair . This is supported by the effect of an inhibitor of Chk2, which enhances the toxicity of TMZ through induction of MC [56–59]. TMZ-induced G2/M cell cycle arrest was totally abrogated by Rsv in U87 cells, without altering DNA damage and ROS levels. However, RT treatment increased the levels of gammaH2AX, pATM and pChk2 in relation to the treatments alone. DNA damage links to the mitotic machinery through phosphorylation of CDC25 and Wee1 by Chks, inhibiting and activating these enzymes, respectively. This leads to phosphorylation of tyrosine 15 and consequent inhibition of Cdk1 . Experiments in Xenopus indicate that xChk1 phosphorylates xWee1 on Ser549, which corresponds to Ser642 in hWee1 . Phosphorylation of this site in hWee1 was shown to be important for 14-3-3 binding and activity, and Chk1 was among the kinases able to phosphorylate this site in HELA cells . TMZ induced high levels of pWee1(S642), which were reduced by the cotreatment with Rsv, leading to a reduction in pCdk1 after 48 h of treatment, suggesting a permissive status for mitotic progression.
On the other hand, several molecular signs in RT treated cells suggest G2 cell cycle arrest or early mitosis: (1) high levels of pRb(S807/811), which is phosphorylated at the G1 to S transition  and dephosphorylated at the M to G1 transition ; (2) low levels of cyclin D1; (3) high levels of cyclin B1, which needs to be degraded for the completion of mitosis ; and (4) low levels of histone H3 S10 phosphorylation, which occurs in the early step of mitosis, being tightly correlated with chromosome condensation during mitosis . This state of conflicted cell cycle signaling, mainly accumulation of cyclin B  and decrease of Cdk1(Y15)  was also observed in GBM cells treated with radiation or doxorubicin and fluorouracil [67, 68] and was the underlying mechanism of MC induction. Indeed, we observed an increase of cells with MC features after RT treatment, which was not observed in the isolated treatments.
In summary, considering the effect of Rsv on TMZ-treated cells we noted that there was no effect on DNA damage. At the early steps of DDR signaling, γH2AX, pATM and pChk2, Rsv increased the levels induced by TMZ. On the other hand, effectors of DDR signaling that regulate the cell cycle were decreased by Rsv in TMZ treated cells, mainly pWee1, pCdk1 and Histone H3 phosphorylation, which are targets of Chk1/2. This indicates that Rsv somehow potentiates the early steps of DDR signaling while blocking the late stages, probably at the level of Chk1/2, which may be responsible for the lack of TMZ-induced arrest. Although the proportion of cells with nuclear irregularities was relatively low (26%) after 48 h, stretched out over the duration of a typical TMZ therapy, i.e. several weeks, this mechanism, together with senescence induction, may be responsible for the observed total reduction in clone formation after 14 days. Indeed, the multiple mechanism of action of TMZ plus Rsv cotreatment in glioma cells is therapeutically promising, since the induction of autophagy , senescence and mitotic catastrophe, rather than high level of apoptosis  seems to be important for the mechanism of action of TMZ on GBM cells.
In conclusion, Rsv increases the toxicity of TMZ in GBM cells through induction of multiple mechanisms, mainly inhibition of TMZ-induced G2/M arrest followed by induction of senescence and MC. The effects of the RT combination are more clearly observed with long term analysis, in which the combination totally abrogates clonogenic growth. Thus, development of combinations of drugs that induce multiple mechanisms of cell death and growth arrest deserves attention and has a high potential to be pre-clinically tested, due to the effectiveness to induce the disruption of tumor cells, thus, reducing the possibility of resistance and recurrence.