Here we describe a novel therapeutic approach for treating cancer using the new cell-permeable fusion protein sTRAIL:FeSOD. The effects of sTRAIL and sTRAIL:FeSOD on apoptosis in LO2 cells in hyperosmotic medium did not show significant differences, suggesting that the FeSOD domain does not change the oligomerization state of sTRAIL (Figures 1B and 1C). When engaging the TRAIL receptors (TRAIL-R1 and TRAIL-R2), the trimeric sTRAIL:FeSOD triggered the TRAIL-induced apoptotic signaling pathway and simultaneously permeated the cell membrane via the receptor-mediated endocytic pathway. Movement throughout the cytoplasm provided FeSOD with the opportunity to scavenge O2
- originating from the mitochondria. Once sTRAIL:FeSOD engaged the TRAIL receptors, TRAIL-induced apoptosis signaling was triggered, seemingly prior to the elimination of intracellular O2
-. However, in the current study, the brief accumulation of H2O2 and the downregulation of p-Akt occurred rapidly enough to enhance the sensitivity of K562 and HL-60 cancer cells to TRAIL-induced apoptosis. The internalization of sTRAIL:FeSOD was rapid (Figure 2) and triggered TRAIL-induced apoptosis signaling and the elimination of intracellular O2
- to occur nearly simultaneously. In addition, it has been suggested that Fas-associated protein with death domain and caspase-8 are not recruited soon enough because of the long delay in death-inducing signaling complex assembly in type II cells . Thus, FeSOD had sufficient time to scavenge intracellular O2
- and then influence the course of TRAIL-induced apoptosis.
The inhibition of H2O2 accumulation or the overexpression of c-FLIPL partially suppressed sTRAIL:FeSOD-induced apoptosis, and thus we infer that H2O2 and p-Akt affect caspase-8 activity via different pathways. Constitutive phosphoinositide 3-kinase (PI3K)/Akt activity has been demonstrated to be one of the most effective antiapoptotic survival pathways in TRAIL-resistant cells, and therefore decreasing p-Akt levels is an important mechanism of averting TRAIL resistance. Tumor cells containing an activating somatic mutation in PI3K are relatively resistant to TRAIL-induced apoptosis [32, 33]. After treatment with sTRAIL:FeSOD, p-Akt was downregulated in three cell lines (Figures 6A to 6C), depressing c-FLIPL expression . However, the LO2 cells responded little to the fusion protein, in contrast to the synergistic apoptosis in TRAIL-resistant K562 and HL-60 cells . The downregulation of Akt activity by LY294002 promoted TRAIL cytotoxicity in LO2 cells (Figure 6D), implying that low oxidative stress induced by FeSOD inhibited TRAIL-induced apoptosis in the LO2 cells. The results of the present study suggest that differences in caspase-8 activity determine the different cellular effects of sTRAIL:FeSOD on the indicated cells based on the interruption of the mitochondrial apoptotic pathway (Figures 4A, B and 5B to 5E). ROS have been reported to regulate caspase activation in TRAIL-resistant human colon carcinoma cells . Perez-Cruz et al.  indicated that intracellular vitamin C can quench some of these ROS, reducing caspase-8 activation, similar to what we observed in the LO2 cells. However, Fas et al.  found that wogonin sensitizes resistant malignant cells to TRAIL by shifting levels of the TRAIL-induced free radical O2
-, consistent with our results demonstrating the sensitization of K562 and HL-60 cells.
ROS levels have been found to be significantly higher in cancer cells than in normal cells, and the activity of antioxidant enzymes such as glutathione peroxidase, catalase and SOD has been shown to be significantly lower in cancer patients than in controls . The higher ROS levels and lower antioxidant enzyme activities (Figures 3C to 3F) make the transient burst of H2O2 possible in K562 and HL-60 cells; however, because of the lower ROS levels and vigorous antioxidant system, the brief accumulation of H2O2 was not detected in LO2 cells (Figures 3G and 3I). The brief accumulation of H2O2 during O2
- scavenging was involved in the sTRAIL:FeSOD-induced apoptosis in K562 and HL-60 cells (Figures 5H to 5K), possibly exerting a direct effect on caspase-8 activation . The low H2O2 levels protected LO2 cells from sTRAIL:FeSOD-induced apoptosis (Figures 7E to 7G). Normal T cells produced very little ROS, and therefore only a small shift in redox state was seen, which may explain why sTRAIL:FeSOD did not sensitize normal T cells to undergo apoptosis. Our studies reveal that sTRAIL:FeSOD reduces the level of intracellular O2
-, with two results: the downregulation of p-Akt arising from a low level of O2
- and the transitory accumulation of H2O2, both of which may increase the amount of available active caspase-8 . However, a low level of O2
-, impairing caspase-8 activation , and stable mitochondria may inhibit apoptosis. Thus, we infer that the relative ratio of these opposing effects described above determines the sensitivity of a cell to sTRAIL:FeSOD and that this relative ratio may be associated with the cell type and the level of O2
-. Without the accumulation of H2O2, the available activated caspase-8 was insufficient (Figures 5C and 7E to 7G). At the same time, caspase-9 activity was inhibited (Figures 4 and 5C), which means that the mitochondrial apoptotic pathway was interrupted. Thus, sTRAIL:FeSOD cannot sensitize LO2 cells to apoptosis when there is insufficient activated caspase-8 to excite the downstream apoptotic pathway in the absence of the intrinsic apoptotic pathway. The induction of apoptosis in K562 and HL-60 cells implies that the net result of intracellular O2
- scavenging by sTRAIL:FeSOD is the presence of a critical amount of activated caspase-8. Thus, in environments of low-level oxidative stress, the cellular context may influence the relative ratio of the opposing effects and subsequently determine whether this low level of oxidative stress favors apoptosis or survival.
Because of the effective protection against the mitochondrial apoptotic pathway mediated by the increased expression of Bcl-XL and the mutation of caspase-8, K562 and HL-60 cells exhibit strong resistance to chemotherapeutic agents [38, 39]. However, after treatment with sTRAIL:FeSOD, K562 and HL-60 cells undergoing apoptosis retained their ΔΨm (Figure 4), and we also failed to detect a difference in the expression level of Bcl-2, Bax, Bid or Bcl-XL (Figures 6A and 6B) or a change in the distribution of cytochrome c (Figure 4D). The unchanged pattern of cytochrome c localization may be associated with a decrease in ROS levels, leading to mitochondrial stability. Depressed caspase-9 expression did not suppress apoptosis, demonstrating that sTRAIL:FeSOD-induced apoptosis was independent of the mitochondrial apoptotic pathway in K562 and HL-60 cells (Figures 5A and 5B). Mohr et al.  indicated that high levels of MnSOD protect colorectal cancer cells from TRAIL-induced apoptosis by inhibition of Smac/DIABLO release. However, this effect may be limited to type II cells, in which death receptor-mediated apoptosis is dependent on mitochondria. In contrast, in the present study, treatment with sTRAIL:FeSOD caused K562 and HL-60 cells to convert to type I cells and to apoptose independently of the intrinsic apoptotic pathway, consistent with the effects of erythroid differentiation . Thus, mitochondrial hyperpolarization does not delay apoptosis when a critical amount of caspase-8 has been activated, which may be the mechanism that regulates type I cell apoptosis independently of the mitochondrial signaling pathway .
Death receptors are expressed in liver tissue as well as in isolated hepatocytes , and for this reason sTRAIL:FeSOD was able to permeate the cell membrane and reduce the oxidative stress in LO2 cells without causing cytotoxicity. Maintaining the balance of opposing effects is important when treating normal cells with sTRAIL:FeSOD, because an appropriate ROS level is extremely important for preserving vital cellular and biochemical functions. Our future work will focus on the task of estimating intracellular ROS production in an individual cell to determine a treatment dose that maintains ROS levels at an appropriate interval, killing cancer cells without inducing normal cell death. In light of the different sensitivities of different cell types to ROS and the hypersensitivity of tumor cells to decreased levels of ROS relative to normal cells, the ability to control TRAIL cytotoxicity through the regulation of intracellular ROS levels will be a breakthrough in the utilization of TRAIL. Undoubtedly, sTRAIL:FeSOD is a good potential therapeutic choice.