Ryota Komori and Mai Taniguchi: Both authors equally contributed to this work. To whom correspondence should be addressed: Hiderou Yoshida, Department of Molecular Biochemistry, Graduate School of Life Science, University of Hyogo, Harima Science Garden City, Hyogo 678-1297, Japan. Tel.: +81–791–58–0219, Fax: +81–791–58–0212 E-mail: hide@sci.u-hyogo.ac.jp |
|
Endoplasmic reticulum (ER) stress is a harmful condition, in which the folding of secretory proteins is hampered and unfolded proteins are accumulated in the ER. Upon the occurrence of ER stress, eukaryotic cells activate a cytoprotective mechanism called the ER stress response (Kohno, 2010; Mori, 2010; Ron and Walter, 2007; Scheuner and Kaufman, 2008; Yoshida, 2007), which is conserved from yeast to mammals.
The mammalian ER stress response consists of three signaling pathways, the first of which is the ATF6 pathway. Upon ER stress, pATF6(P), a transmembrane protein located in the ER, is converted from an oligomer to a monomer by the reduction of disulfide bonds between the monomers. These monomers are transported to the Golgi apparatus and cleaved by Golgi-localized proteases. The cytosolic fragment of cleaved ATF6 (pATF6(N)) translocates to the nucleus, binds to an enhancer element called ERSE, and activates the transcription of genes encoding ER chaperones in order to enhance the folding of secretory proteins. The second pathway is the IRE1 pathway. IRE1 located in the ER membrane is activated by ER stress and converts the pre-mRNA of XBP1 (XBP1(U) mRNA) to the mature mRNA (XBP1(S) mRNA) by unconventional mRNA splicing. pXBP1(S), an active transcription factor, is translated from XBP1(S) mRNA and activates the transcription of genes encoding ER-associated degradation (ERAD) components through an enhancer element called UPRE, in order to enhance the degradation of unfolded or malfolded secretory proteins accumulated in the ER. The third pathway is the PERK pathway. When the transmembrane kinase PERK is activated by ER stress, it phosphorylates the α subunit of eukaryotic initiation factor (eIF2α), resulting in translational attenuation in order to halt the further accumulation of unfolded proteins in the ER. PERK also increases the expression of a transcription factor, ATF4, which binds to an enhancer element called AARE and enhances the transcription of CHOP and GADD34, a proapoptotic transcription factor and a phosphatase specific to eIF2α, respectively. Because the mammalian ER stress response is markedly activated in professional secretory cells, including pancreatic beta cells, acinar cells in the salivary gland, and plasma cells (Lee et al., 2005; Reimold et al., 2001), it has been thought that the main function of the mammalian ER stress response is to increase the capacity of protein folding in the ER when the synthesis of secretory proteins is increased. However, cells other than professional secretory cells also exhibit the ER stress response system, suggesting that the ER stress response is activated in other situations in order to protects cells from stressful environments.
Ultraviolet (UV) irradiation is a harmful electromagnetic wave that has become an increasingly serious issue because of the gradual thinning of the earth’s ozone layer. UV is usually classified into three categories, UVA (wavelength 315–400 nm), UVB (280–315 nm) and UVC (200–280 nm). UVA damages both the epidermis and the dermis, whereas UVB affects only the epidermis. UVC hardly reaches human skin because it is absorbed by the ozone layer; therefore, UVA is thought to be one of the major factors responsible for photo-induced damage of the dermis. In this article, we investigated whether UVA irradiation induces ER stress in normal human dermal fibroblasts (NHDFs), to investigate whether the mammalian ER stress response is activated by not only intrinsic stress but also environmental stress.
NHDFs cells were cultured as a culture of HeLa cells (Yoshida et al., 2001). Cells were irradiated with UVA in phosphate-buffered saline (PBS), incubated in DMEM-FCS for the indicated time period after UVA irradiation and then harvested for immunoblotting or RT-PCR.
Immunoblotting (Yoshida et al., 2003) and RT-PCR of XBP1 mRNA (Uemura et al., 2009) were performed essentially as described previously. For immunoblotting, recycled anti-ATF6α antiserum used in other studies was reused. For RT-PCR, total RNA was reverse-transcribed using Superscript VILO reverse transcriptase (Invitrogen) and amplified with MightyAmp DNA polymerase (TaKaRa) using a pair of primers that correspond to nucleotides 493–512 (CGCGGATCCGAATGAAGTGAGGCCAGTGG) and 834–853 (GGGGCTTGGTATATATGTGG) of XBP1 mRNA, respectively. Amplified fragments covering a 26-nt intron (nucleotides 531–556), and flanking exon fragments were separated on TAE-polyacrylamide gels, visualized by staining with Gel Red (Biotium) and detected using an LAS-3000 (Fuji Film).
Transient transfection of HeLa cells was carried out as described previously (Oku et al., 2011). In brief, HeLa cells cultured in a 24-well plate were incubated with precipitates of calcium phosphate containing plasmids for 12 h at 37°C. After washing with PBS to remove CaPO4-DNA precipitates, cells were incubated in fresh medium for 24 h, irradiated with UVA as described above and harvested for luciferase assays. Construction of the ERSE, UPRE and AARE reporter plasmids was described previously (Wang et al., 2000; Yoshida et al., 1998; Yoshida et al., 2000).
We examined the spectral profile of the UVA source used here and confirmed that it was within the range of 315–400 nm (Fig. 1). We first examined whether UVA irradiation (315–400nm) activated the ATF6 pathway. Whole cell lysates prepared from NHDFs irradiated with UVA were subjected to immunoblotting analysis using an anti-ATF6α antibody. It was found that 0.2–10 J/cm2 UVA induced the cleavage of pATF6α(P) and expression of pATF6α(N) (Fig. 2A, lanes 4–9), as well as an authentic ER stress inducer, thapsigargin (lane 10). Because the total dose of UVA irradiation during daytime around the Tokyo area in July is 150 J/cm2, and the penetration rate of UVA (350 nm) at the depth of 120 μm in human skin is 14% (Anderson and Parrish, 1981), the UVA dose used here is physiologically significant. Expression of pATF6α(N) diminished considerably in cells irradiated with 5 or 10 J/cm2 UVA (lanes 8 and 9). It is possible that excessive oxidative stress caused by strong UVA irradiation over-oxidized pATF6α(P), resulting in retention of pATF6α(P) in the ER lumen because reduction of intermolecular disulfide bonds between pATF6α(P) is prerequisite for the conversion of pATF6α(P) to pATF6α(N) (Nadanaka et al., 2007). UVA-induced cleavage of pATF6α(P) was observed at 2–12 hours after 2.0 J/cm2 UVA irradiation (Fig. 2B, lanes 3–6), indicating that UVA activated the ATF6 pathway.
![]() View Details | Fig. 1. Confirmation of the wavelength of the UVA source. The spectral profile of the UVA source is shown. UVA was delivered using a Xenon lamp through a UVA-specific bandpass filter (Asahi Spectra Co.). The profile was measured using a MSR-7000 (Opt Research Inc.). Experimental exposure time needed to deliver 1 J/cm2 of UVA was 10.95 sec. |
![]() View Details | Fig. 2. Immunoblotting analyses of the ATF6 pathway after UVA irradiation. Whole cell lysates prepared from NHDFs (A) irradiated with the indicated dose of UVA and incubated in growth medium for 6 h before harvest or (B) irradiated with 2.0 J/cm2 UVA and incubated in growth medium for the indicated periods were separated by SDS-PAGE, blotted onto a nylon membrane, and probed with an anti-ATF6α antibody. Asterisks indicate non-specific bands. Non-specific bands indicated by double asterisks were used as a loading control. |
Next, the IRE1 pathway was examined. Total RNA prepared from UVA-irradiated NHDFs was subjected to RT-PCR to examine XBP1 splicing (Fig. 3A). Before UVA irradiation, only XBP1(U) mRNA was observed (lanes 1 and 7). After 1–10 J/cm2 UVA irradiation, XBP1(S) mRNA levels were increased (lanes 4–6 and 8–12), indicating that the IRE1 pathway was activated by UVA irradiation and the splicing of XBP1 mRNA was induced. This UVA-induced splicing of XBP1 mRNA was suppressed by treatment with an antioxidant, ascorbate (Fig. 3B), suggesting that oxidative stress is involved in the induction of ER stress by UVA irradiation.
![]() View Details | Fig. 3. RT-PCR analyses of the XBP1 pathway after UVA irradiation. (A) Total RNA prepared from NHDFs irradiated with the indicated dose of UVA and incubated in growth medium for the indicated periods after UVA irradiation was used as a template for RT-PCR to amplify the region of XBP1 mRNA splicing. RT-PCR products were separated on HCHO-polyacrylamide gel and stained with Gel Red. (B) XBP1 splicing in NHDFs treated with the indicated dose of ascorbate and UVA was examined as in (A). |
We also examined whether UVA irradiation activated the PERK pathway. Whole cell lysates prepared from UVA-irradiated NHDFs were subjected to immunoblotting analysis using anti-phosphorylated eIF2α (upper panel in Fig. 4A), anti-eIF2α (middle panel) and anti-GAPDH (lower panel) antibodies. Phosphorylation of eIF2α a target of PERK, was increased by UVA irradiation (upper panel, lanes 1–9) as well as after thapsigargin treatment (lane 10), whereas the expression of eIF2α (middle panel) and GAPDH used as a loading control was not (lower panel). This UVA-induced phosphorylation of eIF2α was again suppressed by treatment with ascorbate (Fig. 4B, lanes 3 and 4). We did not observed suppression of the UVA-induced cleavage of pATF6(P) by ascorbate treatment (data not shown), and it was speculated that this was because ascorbate enhanced the conversion of PATF6(P) oligomers to monomers by hampering oxidative conditions in the ER.
![]() View Details | Fig. 4. Analysis of the PERK pathway after UVA irradiation. The same samples as in Fig. 2A was subjected to immunoblotting and probed with anti-phosphorylated eIF2α (upper panel), anti-eIF2α (middle panel) and anti-GAPDH (lower panel) antibodies. |
Finally, we examined whether transcription from ERSE, UPRE and AARE was induced by UVA irradiation. Reporter constructs harboring a firefly luciferase gene fused with ERSE, UPRE or AARE were transiently transfected into NHDFs, and luciferase activity was measured after UVA irradiation or thapsigargin treatment. UVA irradiation enhanced transcription from these ER stress-responsive cis-acting elements as well as thapsigargin (Fig. 5). From these results, we concluded that UVA irradiation induces the ER stress response in NHDFs. This suggests that the ER stress response is activated throughout mammalian skin in response to UVA irradiation, and that the ER stress response is activated by not only intrinsic stress (increased synthesis of secretory proteins) but also extrinsic environmental stress.
![]() View Details | Fig. 5. Analysis of transcription from ERSE, UPRE and AARE after UVA irradiation. NHDFs were transiently transfected with the indicated reporter plasmids and a Renilla luciferase standard plasmid. After 1μM thapsigargin treatment (grey bars) or 2.0 J/cm2 UVA irradiation (black bars), cells were harvested for luciferase assays. Luciferase activity was corrected with the internal standard Renilla luciferase activity. |
Below we present our working hypothesis on UVA-induced activation of the ER stress response in human dermal cells. UVA irradiation causes oxidative stress and produces reactive oxygen species (ROS), including superoxide anion radicals, hydrogen peroxide, hydroxyl radicals, singlet oxygen, and lipid peroxides and their radicals (Gruber et al., 2010; Sakurai et al., 2005; Tyrrell, 2004). ROS affects protein folding in the ER and finally induces ER stress (Malhotra and Kaufman, 2007; Zhang, 2010). Under these ER stress conditions, the folding of ECM proteins such as collagen and elastin might be impaired, possibly resulting in their enhanced retention in the ER as well as degradation by ERAD. Reduced expression of ECM proteins in the extracellular matrix might lead to acceleration of the aging process of dermal skin. The importance of the ER stress response in protection against photo-induced skin damage in humans will be of interest for future research.
It is interesting that the dose that each of the UPR branches responded to UVA was different. Each branch of the UPR might respond to distinct molecular changes in the ER caused by ER stress. It is possible that ATF6 could not be activated by higher doses of UVA because the activation machinery of the pathway is labile to oxidative stress caused by UVA, as mentioned above. Moreover, the PERK pathway and the oxidative stress response pathway are known to be interconnected (Cullinan et al., 2003) and this may contribute to the high sensitivity of the PERK pathway to UVA irradiation.
Previously, Kaufman and colleagues reported that UVC (254 nm) activates the PERK pathway in COS-1, MCF-7 and HIT cells (Wu et al., 2002), and Maytin and colleagues showed that UV light consisting of 74% UVB (280–320 nm) and 16% UVA induced the transcription of CHOP in murine epidermis (Anand et al., 2005). Recently, Kanekura and colleagues revealed that UVB (around 302 nm) activated the ATF6 and IRE1 pathways, but not the PERK pathway in human keratinocytes (Mera et al., 2010); however, to the authors’ knowledge, this is the first report demonstrating that UVA (315–400 nm) activates the ER stress response in eukaryotic cells.
Anti-ATF6α antiserum and the reporter plasmid of UPRE were kindly provided by Prof. Kazutoshi Mori (Kyoto University) and Prof. Ron Prywes (Columbia University), respectively. This work was supported by Sankyo Foundation of Life Science, the Uehara Memorial Foundation, grants from MEXT of Japan (Nos. 22020019, 22370069 and 23657136), Hyogo Science and Technology Association, the GCOE program for the University of Hyogo and Grant-in-Aid for Scientific Research on Innovative Areas, MEXT, Japan.