We show that RES increase the expression of FoxO1A, FoxO3A, and FoxO4 under chronic oxidative stress conditions (
Fig. 2). This response is consistent with a homeostatic response to protect lens epithelial cells. Baur et al.
28 found that RES extended lifespan in a high-fat diet mouse model and that it appeared to be dependent on activation of Sir2. It was further demonstrated that the Sir2 homolog SIRT1 controls the cellular response to stress by regulating the FoxO family.
29 SIRT1 and the FoxO transcription factor FoxO3 form a complex in cells in response to oxidative stress. In our study, we found that RES increased expression of FoxO3A and also enhanced the expression of FoxO1A and FoxO4 under chronic oxidative stress conditions, suggesting that FoxO1A and FoxO4 may also be involved in the protective effects of RES (
Fig. 2). This is further supported by the finding that knockdown of individual FoxOs by FoxO siRNAs decreases the protective effects of RES against acute oxidative stress. A triple knockdown of all three FoxOs demonstrates the greatest cytotoxic effect that we measured in this study (
Fig. 4B). In other words, the knockdown of FoxO1A, FoxO3A, and FoxO4 expression abolishes resveratrol's protective effect against acute oxidative damage to human LEC cells. Furthermore, it has been shown that FoxO-deficient mice (FoxO1, FoxO3A, and FoxO4) exhibit an increase in ROS generation and changes in the expression of ROS-associated genes in hematopoietic stem cell populations.
17 Although RES increased the expression of the antioxidants (SOD-1, catalase, and HO-1)
21 , which exert protective effects, RES can also influence other genes that have markedly different effects. RES can induce a concentration-dependent inhibition of cell growth by inhibiting DNA synthesis, decreasing the number of viable cells, and increasing the activity of executioner caspases 3 and 7.
30 Therefore, RES appears to have context-dependent effects. Furthermore, since RES increases FoxO expression under oxidative stress conditions and when this effect is functionally decreased with specific siRNAs, we see a small pro-apoptotic effect. This observation is potentially relevant to understanding why RES can exert opposite effects on cell survival in different cell types and conditions. This context-dependent effect may explain the observation of increased RES toxicity in nonstressed cells, while RES prevents cell death in stressed cells by upregulating FoxOs. Although the overexpression of Sir2 in
Drosophila has been shown to promote caspase-dependent (but p53-independent) apoptosis that is mediated by JNK,
31 the pro-apoptotic effects of RES in nonstressed LECs could not be prevented by inactivation of JNK. Other mechanisms such as JNK-independent activation of BIM appear more likely to be involved in the induction of apoptosis by RES in nonstressed cells.
32 Although cell death in H
2O
2-treated cells could be partially prevented by inhibition of JNK, RES effects were additive, suggesting that RES was preventing cell death not only by inhibition of JNK but also through other mechanisms (
Fig. 3). Our results also showed that each of the three FoxOs tested were needed for the protective effects of RES against H
2O
2-induced cell death (
Fig. 4B), which suggests that the observed upregulation of FoxOs mediated by RES might play a critical role in its protective effects.