TRAIL is a potential anticancer agent, and drug combination therapy to improve its effectiveness has recently garnered much attention. In this respect, its advantaged combination with STI571 has been shown in CML and melanoma. TRAIL and STI571 can mutually overcome respective death resistance in CML [23, 25, 27]. Co-treatment with STI571 also enhances the susceptibility of melanoma cells to TRAIL . Based on previous promising results of this combination effect, we were interested to address whether other types of cancers also confer higher susceptibility towards co-treatment of both antitumor agents. To this end, in this study we chose colon cancer and prostate cancer cells, where STI571 and TRAIL alone have been demonstrated to exert antitumor activity [28, 29, 38, 39].
Here we found that the action of TRAIL in colon cancer cells is sensitive to zVAD, confirming the process of apoptosis. However, a slight reduction in cell viability by STI571 (of around 12% at 0.3 μM) was not affected by zVAD, ruling out the process of apoptosis. Instead, a cell proliferation analysis indicated that STI571 can inhibit HCT116 cell growth (data not shown) as reported in HT29 colon cancer cells . When treating HCT116 cells with STI571 and TRAIL, an antagonistic result was obtained, suggesting that STI571 can regulate the death effect of TRAIL. Such antagonistic effect of STI571 exhibited the concentration-dependency at 0.1 ~ 1 μM. However, a higher concentration of STI571 (10 μM) did not display this effect. Currently we cannot explain the latter observation for the interaction of STI571 and TRAIL, but it is suggested that multiple mechanisms participate in regulating TRAIL's effect by STI571.
Many cytotoxic chemotherapeutic drugs sensitize cancer cells to TRAIL by increasing its receptor expression . In this respect, STI571 did not change caspase 8 activation caused by TRAIL, ruling out STI571's action is related to death receptor expression or activation of upstream death signals. Moreover, we conducted immunoblotting with DR4 and DR5 antibodies and flow cytometry to detect surface DR5 expression. The lack of any effects in these experiments (data not shown) indicates that STI571 does not change expression of the death receptors.
TRAIL-induced apoptosis has been shown to involve p38 and JNK followed by caspase 3 activation in HeLa and HCT116 cells [4, 29]. Thus, sensitizing cancer cells to TRAIL through activating JNK and p38, which subsequently regulate pro-apoptotic and anti-apoptotic Bcl-2 family members and p53, becomes a promising approach to cancer therapy [41–44]. Using pharmacological inhibitors, we showed the involvement of JNK and p38 in TRAIL-induced cytotoxicity and in STI571-induced cell protection in HCT116 cells. Under conditions of p38 or JNK inhibition, TRAIL-elicited cell death was inhibited. Moreover, STI571 also inhibited activation of both stress kinases induced by TRAIL, but no longer exerted its cytoprotection when TRAIL-elicited MAPK activation was already abolished. We thus suggest that inhibition of JNK and p38 are involved in STI571-induced protection.
Activation of c-Abl by certain DNA-damaging agents contributes to cell apoptosis via p53-dependent and -independent mechanisms. First of all, Yuan and colleagues found that c-Abl is activated by infrared and in turn leads to G1 growth arrest via a p53-dependent mechanism. However, they also noted that transfecting p53-/- cells with wild-type c-Abl could still sensitize cell apoptosis in response to DNA damage, whereas expressing the kinase dead c-Abl could not . Later, they identified p73, a homologue of p53, as a downstream mediator of c-Abl for inducing cell apoptosis [37, 45]. c-Abl was shown to stabilize p73 through phosphorylation-dependent posttranslational regulation [17, 18, 45–47]. To determine if c-Abl and p73 are targets of STI571 in initiating cytoprotection, we silenced c-Abl and p73 using the siRNA approach. As the results seen in experiments using the kinase inhibitor, we found that downregulation of c-Abl or p73 rendered cells less sensitive to TRAIL for JNK and p38 activation as well as for cell apoptosis. We therefore conclude that c-Abl-dependent p73 activation is involved in TRAIL-induced apoptosis in HCT116 cells. Moreover, in agreement with previous findings [29, 48], we did not observe effects of TRAIL to increase protein expression of p53 and Bax in p53-proficient HCT116 cells (data not shown).
The major function identified for p73 is induction of apoptosis [19, 20, 49, 50]. Studies demonstrated the cross-talk between p73 and stress kinases (JNK and p38), leading to the upregulation of apoptotic Bcl-2 proteins and cell death. JNK can form a complex with p73 and phosphorylate p73 at multiple residues . Activation of c-Abl by DNA damage was also reported to activate p38, and p38 is then sufficient to induce p73 phosphorylation and enhance its transcriptional activity [49, 52, 53]. Thus, activation of p73 by c-Abl may play an important role in cancer therapy, especially in cancer cells that lose p53 function, but express p73. In this study, our results indicate that a c-Abl-dependent p73 pathway is involved in JNK and p38 activation, and mediates the death mechanism of TRAIL in colon cancer cells. In this respect, activated p73 via caspase pathway has been shown to localize to mitochondria and augment cytochrome c release and cell death . Therefore, in addition to being a transcription factor, p73 is speculated to have novel protein-protein interacting roles which contribute to enhancement of cell apoptosis. Although JNK can directly interact with p73 , it still needs to identify the interactive proteins linking p73 to p38. Apart from the involvement of c-Abl-p73 in stress kinase activation caused by TRAIL, we still cannot rule out other signaling pathways that link death receptors to JNK and p38. In this respect, TRAIL might also activate JNK through the adaptor molecules, TNF receptor-associated death domain (TRADD), FADD, TNF receptor-associated factor 2 (TRAF2) and receptor-interacting protein (RIP) [54–56]. Moreover, mitogen-activated protein kinase kinase 1 (MEKK1) and MEKK4 activated by caspase 8 were demonstrated to be responsible for TRAIL-induced JNK or p38 activation .
In this study, we also demonstrated caspase-dependent c-Abl cleavage and activation in TRAIL-treated colon and prostate cancer cell lines. Many studies demonstrated that the phosphorylation of c-Abl at Tyr412 by receiving signals through Src kinases, receptor tyrosine kinases or autophosphorylation, is an index for full c-Abl activation . Moreover, besides phosphorylation-mediated activation, c-Abl can be cleaved by caspase in the C-terminal region [31–33]. Such cleavage occurs mainly in the cytoplasmic compartment and generates a 120-kDa fragment that can lead to increased kinase activity and/or accumulation in the nucleus [31, 32]. Our present results clearly demonstrate the occurrence of both phosphorylation activation and proteolytic activation of c-Abl following TRAIL stimulation in HCT116 cells. Moreover, both activating mechanisms are mediated by a caspase pathway, and the increase of Tyr412 phosphorylation is occurred on residual non-cleaved c-Abl. Notably STI571 did not alter the c-Abl cleavage caused by TRAIL, but partially reduced the extent of Tyr412 phosphorylation. These results suggest the existence of c-Abl autophosphorylation at Tyr412 in TRAIL-stimulated cells, and also imply a cleavage-independent, but caspase-mediated mechanism for c-Abl activation. In this respect, a previous report showed that TNF-α can activate c-Abl and upregulate apoptotic p73 function via a caspase-dependent elimination of retinoblastoma protein, and thus unleashing the nuclear apoptotic effector, c-Abl . Currently the molecular events linking caspase to non-cleaved c-Abl activation following TRAIL stimulation remains unknown, and further investigation is required.
In contrast to reduced TRAIL sensitivity in colon cancer cells, STI571 did not change the susceptibility of PC3 and LNCaP cells to TRAIL. We ruled out such cell type-specific effects of STI571 being related to c-Abl protein expression. Similar expression levels of c-Abl were observed in HCT116, SW480, PC3, and LNCaP cells (data not shown). Instead, we suggest that the antitumor activity of TRAIL in colon and prostate cancers might involve distinctive regulation and complex apoptotic pathways. In prostate cancers, neither p38 nor JNK activation by TRAIL is involved in cell death, while STI571 can still slightly inhibit TRAIL-induced JNK activation in prostate cancers. Moreover, TRAIL-mediated c-Abl cleavage displayed the same pattern in HCT116 and LNCaP cells. Therefore, these results further support the notion that the cell type-specific effect of STI571 on antitumor activity of TRAIL is dependent on the roles of p38 and JNK in cell death per se.