br Materials and Methods br Results br Discussion
Materials and Methods
Discussion ROS generated by SFN target many signaling pathways involved in initiating cancer cell death (Sestili and Fimognari, 2015). For example, ROS have been shown to activate both intrinsic and extrinsic caspase cascades in prostate cancer cells (Singh et al., 2005), suppress Akt phosphorylation and telomerase activity in hepatocellular carcinoma Hep3B cells (Moon et al., 2010), induce mitotic arrest and apoptosis in human Actinomycin D cancer 5637 cells (Park et al., 2014), activate endoplasmic reticulum stress and the nuclear factor–E2-related factor-2 (Nrf2) signaling pathway in T24 human urinary bladder cancer cells (Jo et al., 2014), repress PI3K/Akt pathway activity in thyroid cancer cells (Wang et al., 2015), and inhibit TGF-β–induced epithelial-mesenchymal transition and promote apoptosis in hepatocellular carcinoma cells (Wu et al., 2016). In the current study, we present evidence that the cellular effects produced by SFN in NSCLC cells, including cell-cycle arrest (Fig. 2), induction of DNA damage response and/or apoptosis, and downregulation of EGFR (Fig. 3), are largely attributable to SFN-induced production of ROS, since pretreatment with the ROS scavenger NAC prevented these cellular effects. SFN has been shown to trigger apoptosis through downregulation of EGFR protein (Mondal et al., 2016). We have recently shown that SFN inhibits EGFR signaling by promoting proteasome-mediated degradation of EGFR in NSCLC (Chen et al., 2015). High-level induction of ROS has been shown to trigger the overoxidation and degradation of EGFR L858R T790M mutants and cause apoptosis in the TKI-resistant NSCLC cells that harbor these variants (Leung et al., 2016). Therefore, mutations in EGFR are among the factors governing the sensitivity of a cell to SFN. A related factor that may also affect a cell's sensitivity to SFN is the relative level of EGFR expression. In the current study, we employed a pair of cell lines that have the same genetic background but different degrees of EGFR expression to examine the effects of EGFR level on the sensitivity to SFN. We found that, whereas SFN induced a similar increase in ROS production in both cell lines, high EGFR-expressing CL1-5 cells were more resistant to SFN treatment than low EFGR-expressing CL1-0 cells (Figs. 1 and 2). Moreover, whereas SFN induced DNA damage response, autophagy formation, and cell-cycle arrest in both cell lines, it induced apoptosis only in CL1-0 cells and not in CL1-5 cells (Figs. 2 and 3). The lack of apoptosis induction in cells with high-level EGFR expression is likely attributable to the abundance of EGFR. Although SFN treatment greatly decreased EGFR level, reducing it to about 66% of that in untreated controls, the residual level of EGFR was still higher than that in untreated CL1-0 cells (Fig. 3D). We postulated that the residual amount of undegraded EGFR in SFN-treated CL1-5 cells was sufficient to prevent the induction of apoptosis. Employing shRNA to downregulate EGFR in CL1-5 cells, we found that a weak apoptotic response could be detected in these EGFR-knockdown CL1-5 cells, suggesting that cells with a high level of EGFR resist SFN-induced apoptosis. Since the capability to repair DNA damage is known to depend on the EGFR expression and downstream pathways (Kriegs et al., 2010, Kryeziu et al., 2013, Myllynen et al., 2011), it is likely that cells with a high level of EGFR could perform greater repair of DNA damage and thus resist the induction of apoptosis. Taken together with the observation that EGFR is overexpressed in more than 45% of NSCLC cases (Herbst et al., 2008), our current finding that NSCLC cells with high-level EGFR expression are more resistant to SFN treatment in vitro raised the question of whether SFN is capable of inhibiting the growth of high EGFR-expressing NSCLC tumors in vivo. Using a xenograft animal model, we showed that SFN could indeed inhibit the growth of CL1-5 cell-derived tumors in vivo (Fig. 4). Although SFN appeared to inhibit the growth of tumors derived from xenografted TKI-resistant NSCLC cells to a greater extent than that of tumors derived from xenografted EGFR-overexpressing NSCLC cells, it nonetheless showed good antitumor efficacy against EGFR-overexpressing NSCLC. We suggest that the appropriate combination of SFN with another EGFR-downregulating agent could greatly improve the efficacy of SFN in the treatment of EGFR-overexpressing NSCLC.