|Home | About | Journals | Submit | Contact Us | Français|
In response to ultraviolet light (UV)-induced damage, cells initiate cellular recovery mechanisms including activation of repair genes and redistribution of cell cycle phases. While most studies have focused on DNA damage-inducible transcriptional regulation of cell cycle checkpoints, translational regulation also plays an important role in control of cell cycle progression upon UV-irradiation. UV-irradiation activates two kinases, PERK and GCN2, which phosphorylate the alpha subunit of eukaryotic initiation factor 2 (eIF2α) and subsequently inhibit protein synthesis. We recently identified an upstream regulator, nitric oxide synthase (NOS), which controls the activation of both PERK and GCN2 upon UVB-irradiation. Our data suggested that UVB induces NOS activation and NO• production, which reacts with superoxide (O2 •−) to form peroxynitrite (ONOO−) and activate PERK. The NO• production also leads to L-Arg depletion and GCN2 activation. The elevation of nitric oxide and activation of PERK/GCN2 have been shown to play roles in regulation of cell cycle upon UVB irradiation. In the present study, we show that the cell cycle phases were redistributed by inhibition of NOS activation or reduction of oxidative stress upon UVB irradiation, indicating the roles of NO• and its oxidative products in regulation of cell cycle. We also demonstrate that both PERK and GCN2 were involved in regulation of cell cycle upon UVB-irradiation, but the regulation is independent of eIF2α phosphorylation. While the mechanism for UVB-induced cell cycle control is yet to be unraveled, we here discuss the differential roles of NOS, PERK and GCN2 in regulation of cell cycle upon UVB-irradiation.
Cells respond to ultraviolet light (UV) by reprogramming signaling circuits that control cellular physiology, including cell cycle regulation and apoptotic cell death.1–5 The roles of transcriptional regulation in UV-induced cell cycle arrest and apoptosis are intensively studied.6–11 In the latest decade, the impacts of translation, especially the phosphorylation on the α subunit of translation initiation factor 2 (eIF2α), on UV-induced signaling circuits, are being elucidated.12–20 The phosphorylation eIF2α is a universal response of eukaryotic cells to various types of stress, and is regulated by different eIF2α kinases.21–23 Two eIF2α kinases (EIF2AK), the dsRNA-dependent protein kinase-like ER kinase (PERK, EIF2AK3) and the general control nonderepressible protein kinase 2 (GCN2, EIF2AK4), were shown to be responsible for UV-induced phosphorylation of eIF2α.16,18 PERK is an ER membrane localized kinase, which is activated by endoplasmic reticulum (ER)-stress.24,25 GCN2 is an amino acid abundance controlled eIF2α kinase, which is activated during amino acid starvation.26,27 While the roles of PERK and GCN2 in UV-induced translation inhibition and downstream signal transduction are elucidated,14,15,20,28 their key upstream activator(s) have not been identified.
In a recent paper published in the September issue of The Journal of Biological Chemistry, we provided evidence that UVB activates constitutive nitric oxide synthase (cNOS), which coordinately regulates the activation of both PERK/GCN2 and sequentially the phosphorylation of eIF2α in human keratinocytes HaCaT.29 We demonstrated that UVB-activated NOS rapidly generates NO• from L-Arg, which had a concentration of less than 2.5 µM in HaCaT cells. Our data suggested that the consumption of L-Arg led to a shortage of the amino acid, which activates GCN2. The UVB-induced elevation of superoxide (O2•−)30 and depletion of L-Arg led to cNOS uncoupling and generation of peroxynitrite (ONOO−), which is a strong oxidant31–33 that can be rapidly produced from NO• and O2•−.32,33 The accumulation of intracellular ONOO− induces ER-stress and PERK activation.34 In addition to keratinocytes, we also studied the regulation of eIF2α phosphorylation in wild-type mouse embryonic fibroblasts (MEF), PERK knockout (MEFPERK−/−), GCN2 knockout (MEFGCN2−/−) cells and the eIF2α Ser51 Ala mutant (MEFA/A) cells. Distinctive from HaCaT cells, all three MEF cell lines have higher background phosphorylation of eIF2α, which was maintained by basal NOS activity and oxidative stress. L-Arg-shortage-mediated GCN2 activation appears to play a more significant role for maintaining basal eIF2α phosphorylation in MEFPERK−/− cells. However, UVB-induced eIF2α phosphorylation in MEFPERK−/− cells mainly resulted from oxidative-stress. In MEFGCN2−/− cells, L-Arg biosynthesis is downregulated due to the lower activity of GCN4, whose activity is positively controlled by GCN2.35,36 The down-regulation of L-Arg biosynthesis leads to L-Arg depletion and sequentially to NOS uncoupling and higher oxidative-stress levels, which prevented UVB from further inducing eIF2α phosphorylation. Our results demonstrated that a chronic elimination of a gene could rearrange the entire signaling circuit in response to a stimuli and that interpretation from data generated from knockout cells could be misleading. In the following chapters, we analyzed cell cycles of HaCaT and MEF cells in response to UVB-irradiation. We demonstrated that NOS, PERK and GCN2 play differential roles in regulation of cell cycle upon UVB-irradiation.
NO• regulates a wide range of cell functions such as cell cycle,37,38 cell proliferation39−41 and apoptosis.41 However, the role of NO• in cell cycle control is controversial. On one hand, it shows that exposure to NO• derived from either endogenous NOS or exogenous NO• donors blocks cell cycle progression on G0/G1 or G2/M phase depending on cell types.42–44 TNFα, IFNγ and IL-1β are commonly used as iNOS stimulators to produce endogenous NO•, which blocks the cell’s entry to the G0/G1 phase.45 NO• donors also induce G1 arrest in different cell types such as smooth muscle cells,46 breast cancer cells,47 human pro-monocytic cells U937,48 and melanoma cells.49 On the other hand, some research also demonstrated that exogenous NO• generated from S-nitrosoglutathione prevents gamma-irradiation-induced G1 arrest by impairing p53 conformation.50
To determine the role of activation NOS in the regulation of cell cycle upon UVB-irradiation, we pretreated the cells with a broad NOS inhibitor, a N-substituted L-arginine analog LNMMA (100 µM) for 1 h before and 4 h after UVB-irradiation (50 mJ/cm2). At 24 h post-irradiation, the cell cycle was analyzed by flow cytometry. Our data showed that the G1 phase fraction increases from 49 ± 3% to 70 ± 2% after UVB irradiation (Fig. 1). With treatment of LNMMA, the amount of cells in G1 phase was increased to 64 ± 4% after UVB-irradiation (Fig. 1). With the consideration of the 4% increase of G1 phase cells by treating with LNMMA alone, the results suggested that NOS plays a role in regulation of G1 phase arrest upon UVB. Besides NO•, UV also induces an elevation of ONOO−, which is a stronger oxidant with a biological half-life near 100 ms.51 To determined the role of ONOO− in regulation of cell cycle, we use a glutathione (GSH) synthesis precursor N-acetyl-L-cysteine (LNAC), which was reported to specifically reduce intracellular ONOO− levels in multiple cell types.52,53 Our data showed that treating the cells with LNAC significantly attenuates UVB-induced G1 arrest cell proportion from 70 ± 2% to 50 ± 1% (Fig. 1), which is similar to the control without UVB-irradiation (Fig. 1). This result implies that ONOO− plays a key role in UVB-induced keratinocytes G1 arrest, and the role of NO• in cell cycle control might mediated by formation of ONOO− in stead of itself.
In addition, the NO•/ONOO−-mediated G1 arrest is accompanied with a reduced S phase, which represents the DNA synthesis phase in the cell cycle, and usually used as a proliferative index for cell growth. Our data showed that after UVB-irradiation, the percentage of cells in S phase dropped from 33 ± 2% to 6.8 ± 2% (Fig. 1), exhibiting a significant halt in DNA synthesis caused by UVB insults. Inhibition of NOS activity by LNMMA recovered DNA synthesis from UVB-irradiation, the cells in S phase increased from 7 ± 2% to 16 ± 4%. Reduction of ONOO− by LNAC increased the cells in S phase from 7 ± 2% to 30 ± 2% in UVB treatment. These results suggested that NOS activation in combination with oxidative stress regulates G1/S phase cell cycle in keratinocytes upon UVB-irradiation.
Both PERK and GCN2 regulate cell cycle and apoptosis in response to various stimuli.14,19,28,54–58 ER-stress-induced PERK activation or overexpression of PERK downregulates cell cycle regulator cyclin D1 and the tumor suppressor p53, leading to G1 arrest.56,59 Overexpression a dominant negative truncated PERK lacking its kinase domain prevented unfolding protein response (UPR)-induced cyclin D1 loss and cell cycle arrest.54 GCN2 activation accompanied with increased eIF2α phosphorylation controls the progression of the G1 phase and delays entry to S phase with fission yeast cells in response to UV-irradiation.28,60
In our article published in the September 4th JBC, we demonstrated that NOS and oxidative stress mediate the activation of PERK/GCN2 and phosphorylation of eIF2α.29 To determine whether the UVB-induced activation of the EIF2AKs and phosphorylation of eIF2α also play a role in the regulation of cell cycle, we analyzed the cell cycle patterns of a series of MEF cell lines after UVB irradiation. These cell lines included: wild-type (MEFWT), PERK knockout (MEFPERK−/−), GCN2 knockout (MEFGCN2−/−) and the eIF2α Ser51 Ala mutant (MEFA/A). Our data showed that MEFWT cells behaved very much different from HaCaT cells. Without UVB-treatment, MEFWT cells have relatively fewer cells in G1/G2 phases and more cells in S phase than HaCaT cells (Fig. 2). Knocking out PERK or GCN2 increased 16.1% or 7.7% of the cells in G1 phase respectively, decreased 16.9% or 8.5% of the cells in S phase respectively, but had almost no impact on G2 phase (Fig. 2). In contrast, total elimination of eIF2α phosphorylation by knocking in a non-phosphorylatable eIF2α (S51A) mutant (MEFA/A) reduced 5.6% and 7.7% of the cells in the G1 and S phases, while an increase of 3.8% was observed in the G2 phase. However, the changes in MEFA/A cell cycle are not statistically significant to wild type MEF cells (Fig. 2). These results suggested that basal PERK/GCN2 activities are required for maintaining an efficient G1 checkpoint. However, it is not clear whether PERK or GCN2 regulates cell cycle via eIF2α phosphorylation since elimination of the phosphorylation had no significant effects on the patterns of cell cycle. After UVB-irradiation, there was a shift of 4.4% of MEFWT cells from the G1 phase to the G2 phase with no change observed at the S phase (Fig. 2). This result was very different from that of the HaCaT cells after UVB-irradiation, which showed a significantly increased portion of cells in the G1/G2 phases and a reduced portion of the cells in the S phase. For MEFPERK−/− or MEFGCN2−/− cells, UVB-irradiation shifted approximately 10% of cells from the G1 phase to the G2 phase (Fig. 2). This data indicated that PERK and GCN2 both played a role in regulation of the cell cycle in response to UV-induced DNA damage and oxidative stress in mammalian cells, as previously reported that GCN2 did in yeast.28,61 Interestingly, while reduction of PERK or GCN2 activity promoted UVB-induced cell cycle shift from the G1 phase to the G2 phase, elimination of eIF2α phosphorylation prevented UVB-induced cell cycle shift (Fig. 2). These results suggested that PERK and GCN2 might regulate the cell cycle in mammalian cells via an eIF2α phosphorylation independent pathway as GCN2 does in yeast.19 The eIF2α phosphorylation-independent regulation of cell cycle by EIF2AKs could be mediated via promotion of p53 degradation,62 which plays a key role in regulation of cell cycle upon UVB-irradiation.63
We thank Mr. Oliver Luke Carpenter and Ms. Molly Monica for reading the manuscript. This work was supported by National Institutes of Health Grant RO1 CA86928 (to S.W.) and R56 CA086928 (to S.W.)