Enhanced endoplasmic reticulum (ER) stress leads to cell death in various pathophysiological situations. During a search for compounds that regulate ER stress, we identified methoxyflavones, a group of flavonoids, as strong protective agents against ER stress. Analysis in mouse insulinoma MIN6 cells revealed that methoxyflavones mildly activated the eukaryotic initiation factor 2α and nuclear factor erythroid 2-related factor pathways, but not the XBP1 pathway, and induced downstream genes, including glucose-regulated protein (GRP) 78, a molecular chaperone in the ER. The protective effect of methoxyflavones was enhanced by agents that increase intracellular cAMP levels such as forskolin, dibutyryl-cAMP and IBMX, but suppressed by the protein kinase A (PKA) inhibitor H-89, suggesting involvement of the PKA pathway in the regulation of ER stress by methoxyflavones. Consistent with the results in cultured cells, pretreatment of mice with tangeretin, a methoxyflavone, enhanced expression of GRP78 and HO-1 without causing ER stress in renal tubular epithelium and prevented tunicamycin-induced cell death. Furthermore, preadministration of tangeretin in mice enhanced expression of GRP78 in the substantia nigra pars compacta and protected dopaminergic neurons against 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine, a neurotoxin that induces both oxidative and ER stress. These results suggest that methoxyflavones play an important role in the regulation of ER stress and could be a therapeutic target for the ER stress-related diseases.
- neuronal degeneration
the endoplasmic reticulum (ER) is an organelle where secretory proteins undergo posttranslational modification. Exposure of cells to conditions such as glucose starvation, inhibition of protein glycosylation, disturbance of Ca2+ homeostasis, and oxygen deprivation accumulates unfolded proteins in the ER (ER stress) and activates a set of pathways known as the unfolded protein response (UPR; see Ref. 31). The UPR is transmitted through activation of ER resident proteins, such as inositol requiring kinase 1 α/β, protein kinase-like ER kinase (PERK), and activating transcription factor (ATF) 6, and UPR-target genes include molecular chaperones in the ER, ER-associated protein degradation (ERAD) molecules, and anti-oxidant genes (31). If the protein load in the ER exceeds its folding capacity, or dome defects in the ER stress response exist, cells tend to die, typically, with apoptotic features (ER stress-induced cell death; see Ref. 31). Important roles for ER stress and ER stress-induced cell death have been demonstrated in a broad spectrum of pathological situations, including ischemia, diabetes, atherosclerosis, endocrine defects, and neurodegenerative disorders (7, 10, 12, 24, 28, 34, 38). In Parkinson's disease (PD) research, it was recently reported that neurotoxins, such as 6-hydroxydopamine (6-OHDA), 1-methyl-4-phenyl-pyridinium, and rotenone, induced not only oxidative stress but also ER stress, and disturbance of the UPR, which was observed in PERK null mice, rendered the sympathetic neurons more vulnerable to 6-OHDA (28), suggesting the cross talk of oxidative stress and ER stress in the experimental PD models.
Homocycteine-induced ER protein (Herp) is a membrane-bound, ubiquitin-like protein that is located in the ER (9, 14). Although Herp is strongly induced by ER stress or deep hypoxia, it decays rapidly consequent to proteasome-mediated degradation, suggesting the possible contribution of Herp in the ERAD. We recently reported that targeting disruption of the Herp gene rendered F9 embryonic carcinoma cells vulnerable to ER stress (9). The ER stress-induced death in F9 Herp null cells was associated with the aberrant ER stress signaling, structural changes in the ER, and caspase activation. Using this cell line, and other ER stress-sensitive cells such as mouse insulinoma MIN6 cells and rat pheocromocytoma PC-12 cells, we evaluated the protective effect of 103 plant-derived compounds against ER stress-induced cell death. We report here that, unlike hydroxyflavones, methoxyflavones have strong protective effects against ER stress in vitro and in vivo. They mildly activated the UPR branches such as the eukaryotic initiation factor (eIF) 2α and nuclear factor erythroid 2-related factor (Nrf2) pathways without causing ER stress, at least in part, through the activation of the protein kinase A (PKA) pathway. Furthermore, neuroprotective property of tangeretin, a methoxyflavone, was associated with enhanced expression of molecular chaperons in the ER in a mouse model of PD. Our results suggest that the regulation of ER stress by methoxyflavones could be a therapeutic target for preventing the ER stress-related diseases.
Cell cultures and stress conditions.
F9 Herp null cells and rat astrocytes were maintained in DMEM with 20% FBS and minimal essential medium with 10% FBS, respectively. MIN6 cells and PC-12 cells were provided by Dr. Jun-ichi Miyazaki, Osaka University (20) and Dr. Haruhiro Higashida, Kanazawa University, and were maintained in DMEM with 15% FBS and with 5% FBS/5% horse serum, respectively. ER stress was induced by treating cells with tunicamycin (0.75–2 μg/ml; Sigma, St Louis, MO), brefeldin A (0.5 μg/ml; Sigma), or A-23187 (2.5 μM; Sigma) for indicated times (6–48 h). Each compound was added to the cells together with ER stressors. Oxidative stress was induced by exposing cells to 66 μM H2O2 (Nacalai Tesque, Kyoto, Japan). Cells were pretreated with each compound for 24 h, followed by exposure to H2O2 in the presence of each compound for another 24 h.
Compounds and cell viability assays.
One hundred and three plant-derived compounds (IN1-IN103), including stilbenoids, xanthonoids, and flavonoids, were analyzed for their protecting effects against ER stress in F9 Herp null cells. Hydroxyflavones such as IN17, or methoxyflavones such as IN19, IN69, IN72, and IN88 were isolated as described previously (11, 35). Luteolin, forskolin, dibutyryl-cAMP (DbcAMP), IBMX, and H-89 were purchased from Wako (Osaka, Japan), and dantrolene, N-acetylcysteine, and α-tocopherol were obtained from Sigma. Cell viability under ER stress or oxidative stress conditions was measured by 3-(4,5-dimethyl-2-thiazolyl)-2,4-diphenyl-2H tetrazolium bromide assay (Nacalai Tesque) or LIVE/DEAD cell toxicity assay (Molecular Probes, Eugene, OR), as described previously (9).
Northern blotting and RT-PCR.
MIN6 cells were treated with each compound for indicated times (24 or 40 h), and, in some cases, tunicamycin (2 μg/ml) was added to cells 6 h before harvesting. As a positive control for RT-PCR, astrocytes were treated with tunicamycin (2 μg/ml) for 8 h. Total RNA (10 μg) isolated from MIN6 cells or mouse kidney was subjected to Northern blotting using cDNA fragments specific for glucose-regulated protein (GRP) 78, CCAAT/enhancer-binding protein homologous protein (CHOP), heme oxygenase (HO)-1, or β-actin, as described previously (9). First-strand cDNA was also synthesized with the first-strand cDNA synthesis kit (Takara, Tokyo, Japan) and amplified with primers specific for XBP1, as described previously (32).
Cell lysis and Western blotting.
For analyzing the status of phosphorylation of eIF2α, MIN6 cells were solubilized in buffer containing 1% Nonidet P-40 (NP40), 0.1% SDS, 0.2% deoxycholate and subjected to Western blotting with anti-eIF2α antibody and anti-phospho-eIF2α antibody (Cell Signaling Technology, Beverly, MA). For estimating the status of Nrf2, nuclear proteins were extracted from MIN6 cells as described previously (30) and subjected to Western blotting with anti-Nrf2 antibody (Santa Cruz Biotechnology, Santa Cruz, CA). Sites of primary antibody binding were visualized using alkaline phosphatase-conjugated secondary antibodies.
MIN6 cells were treated with each compound for 24 h and metabolically labeled with 35S-Met/Cys (200 μCi; Amersham Pharmacia Biotech, Piscataway, NJ) for the last 3 h before harvesting in Met-free medium. Cells were lysed in buffer containing 1% NP40, 0.1% SDS, 0.2% deoxycholate and subjected to SDS-PAGE, followed by autoradiography.
Animal experimental protocols were approved by the Committee on Animal Experimentation of Kanazawa University (Takara-machi Campus). All animal experiments were performed using C57BL/6 mice (male, 8–12 wk of age). A mouse model of ER stress was created by a single intraperitoneal injection of tunicamycin (1 mg/kg), as described previously (39). For IN19 or vehicle administration, mice were intraperitoneally injected with IN19 (10 mg·kg−1·day−1) in saline, including 10% Cremophore EL (Sigma) and 10% DMSO, or with the dissolving solution (vehicle) daily for 2 or 4 days. A mouse model of PD was created by intraperitoneal injections of 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP, 20 mg/kg) two times at a 2-h interval (15). Histological analysis was performed at 4 and 7 days after MPTP injection as below. When mice were pretreated with IN19 or vehicle for 4 days, MPTP was injected in mice 2 h after IN19 or vehicle administration on the 4th day.
At the indicated time points, kidney and brain were removed from mice after perfusion with 4% of paraformaldehyde. The kidney was embedded in paraffin, and 5-μm sections were cut. Sections were subjected to hematoxylin and eosin (HE) staining, immunostaining with anti-KDEL or anti-GRP78 antibody (StressGen; Victoria, British Columbia, Canada), and TdT-UDP nick end labeling (TUNEL) staining (Chemicon International, Temecula, CA). Coronal brain sections (10 μm) were cut on a cryostat. Sections were processed for immunostaining with anti-KDEL antibody or with anti-GRP78 antibody, together with anti-tyrosine hydroxylase (TH) antibody (Chemicon International). We employed both anti-KDEL antibody and anti-GRP78 antibody to detect GRP78 expression, since anti-KDEL antibody displayed lower background compared with anti-GRP78 antibody in some brain sections. FITC or Cy3-conjugated secondary antibodies were used for visualization of immunolabeling.
Quantitative data analysis.
Laser densitometry analysis was performed to semiquantitate results of Western blotting as described previously (9). Tyrosine hydroxylase positive (TH+) neuronal cells in the substantia nigra pars compacta (SNpc) were counted in two representative sections (bregma −3.16 and −3.64 mm) as described previously (15). Statistical analysis was performed by Student's t-test.
Prevention of ER stress-induced cell death by methoxyflavones.
Preliminary studies with well-known compounds revealed that dantrolene, an antagonist of the ryanodine receptor, and some anti-oxidants such as α-tocopherol partly prevented tunicamycin-induced cell death in F9 Herp null cells (Fig. 1B). Using these agents as positive controls, 103 plant-derived compounds (IN1-IN103) were screened in F9 Herp null cells treated with tunicamycin, and several methoxyflavones, including IN19 (tangeretin: 5,6,7,8,4′-pentamethoxyflavone), IN69 (nobiletin: 5,6,7,8,3′,4′-hexamethoxyflavone), IN72 (5,6,7,4′-tetramethoxyflavone), and IN88 (sinensetin: 5,6,7,3′,4′-pentamethoxyflavone), were identified as candidate protective agents against ER stress (Fig. 1, A–C). Methoxyflavones prevented tunicamycin-induced cell death more effectively than dantrolene (Fig. 1, B and C), while hydroxyflavones, including IN17 (5,7,2′,4′-tetrahydroxyflavone; Fig. 1B) and luteolin (5,7,3′,4′-tetrahydroxyflavone, data not shown), had no effect under the same conditions. Dantrolene-equivalent protective activity of methoxyflavones was obtained at 10–15 μM in F9 Herp null cells. Methoxyflavones, but not hydroxyflavones, also inhibited tunicamycin-induced cell death in MIN6 cells (Fig. 1DI) and, to a lesser extent, in PC-12 cells (data not shown). In contrast, when MIN6 cells were treated with H2O2, an agent that causes oxidative stress, methoxyflavones were less effective compared with hydroxyflavones (Fig. 1DII). The protective effect of methoxyflavones was also observed when MIN6 cells were treated with other types of ER stressors such as brefeldin A and A-23187 (data not shown).
Effect of methoxyflavones on ER stress signaling.
To dissect the mechanism underlying the protective effect of methoxyflavones, expression levels of UPR target genes, such as GRP78, CHOP, and HO-1 (8, 17, 31), were analyzed in MIN6 cells. IN19, IN88 (Fig. 2A), and other methoxyflavones (data not shown) increased expression of all three genes, whereas the effects were somewhat lower compared with hydroxyflavonessuch as IN17 (Fig. 2A) or luteolin (data not shown). It was previously reported that different types of flavonoids activate eIF2α kinases, including PERK, and suppress protein synthesis through phosphorylation of eIF2α (13). Consistent with these observations, all flavone derivatives tested in the current study enhanced phosphorylation of eIF2α in MIN6 cells after 24 h of treatment, determined by Western blotting with anti-phosphorylated eIF2α antibody. However, the effects were lower in IN19 (Fig. 2BI) and other methoxyflavones (data not shown) when compared with hydroxyflavones such as IN17 (Fig. 2BI) or luteolin (data not shown). Accordingly, metabolic labeling with 35S-methionine revealed that general protein synthesis in MIN6 cells was suppressed by IN17, and, to a lesser extent, by IN19 (Fig. 2BII).
Because expression of HO-1 is regulated by both the eIF2α-ATF4 pathway (8) and Nrf2 pathway (31) and because Nrf2 was found to be a direct substrate of PERK (4), nuclear Nrf2 protein was assessed by Western blotting using nuclear extract from MIN6 cells. The levels of Nrf2 protein in the nucleus were enhanced by treatment of cells with IN17 and, to a lesser extent, with IN19 (Fig. 2C). In contrast to the activation of the eIF2α and Nrf2 pathways, IN19 did not activate the XBP1 pathway in MIN6 cells, another key pathway in the UPR (Fig. 2D), indicating that methoxyflavones did not cause ER stress but activated the UPR branches such as the eIF2α and Nrf2 pathways.
Involvement of the protein kinase A pathway in the regulation of ER stress by methoxyflavones.
It was recently reported that methoxyflavones induced neurite outgrowth through the activation of PKA signaling (22, 23). We hypothesized that this pathway may also play a role in the regulation of ER stress. Analysis with MIN6 cells revealed that forskolin, an adenylate cyclase activator, or IBMX, a nonselective inhibitor of phosphodiesterase (PDE), protected cells against tunicamycin-induced cell death, whereas H-89, an inhibitor of PKA, accelerated cell death in the same condition (Fig. 3 AI). Furthermore, the protective effect of IN19 (see Fig. 5AII) or other methoxyflavones (data not shown) against ER stress was enhanced by forskolin, DbcAMP, a cAMP analog, or IBMX but reversed by H-89 (see Fig. 5AII), suggesting involvement of the PKA pathway in the regulation of ER stress by methoxyflavones. Consistently, forskolin or combination of forskolin with IBMX upregulated expression of GRP78, CHOP, and HO-1 in MIN6 cells (Fig. 3B). Similar cell protection by forskolin or IBMX against ER stress was observed in F9 Herp null cells (data not shown).
Effect of methoxyflavones on ER stress-induced cell death in vivo.
To analyze the impact of methoxyflavones on ER stress in vivo, a tunicamycin injection model in mice was employed (39). First, IN19 (10 mg·kg−1·day−1) was intraperitoneally injected in mice for 2 or 4 consecutive days, and expression of UPR target genes in the kidney was compared with vehicle-injected mice. Consistent with the results in cultured cells, expression of GRP78 and HO-1 was enhanced in IN19-injected kidneys (Fig. 4 AI), whereas XBP1 was not activated (Fig. 4AII). Enhanced expression of GRP78 protein, which was determined by immunostaining with anti-KDEL antibody (Fig. 4B) and with anti-GRP78 antibody (supplemental Fig. S1A; the online version of this article contains supplemental data), was observed mainly in tubular epithelium, although the expression levels were lower than those in tunicamycin-injected kidneys (2 days after a single injection; Fig. 4B). Next, tunicamycin was intraperitoneally injected in the mice, and, 4 days after injection, morphological changes and cell death were assessed in the kidneys. HE and TUNEL staining revealed that tunicamycin treatment (1 mg/kg) partly destroyed the tubular structures (Fig. 4Cb) and caused cell death (Fig. 4Ce) in kidneys, as described previously (39). However, when mice were pretreated with IN19 for four consecutive days, these histological changes almost entirely disappeared (Fig. 4Cc, f), suggesting a critical role of methoxyflavones in the regulation of ER stress in vivo.
ER stress regulation in a mouse model of PD.
Because tangeretin (IN19) was shown to cross the blood-brain barrier and was neuroprotective in a rat 6-OHDA lesion model of PD (5), a possible association of neuroprotection and ER stress regulation by IN19 was analyzed in a mouse PD model by intraperitoneal injection of MPTP (39). Consistent with a previous report in cultured cells (28), injection of MPTP decreased the number of TH+ cells in the SNpc (∼50% of control mice) within 4 days but enhanced immunoreactivity with anti-KDEL antibody in TH+ neuronal cells (Fig. 5A), suggesting induction of ER stress in this in vivo model. Intraperitoneal injection of IN19 (10 mg·kg−1·day−1) for four consecutive days also enhanced immunoreactivity of GRP78 in TH+ neuronal cells in the SNpc before MPTP injection (Fig. 5B and supplemental Fig. S1B), and suppressed reduction of TH+ neuronal cells in the SNpc 7 days after MPTP injection (Fig. 5, C and D). These results suggest that the neuroprotective properties of IN19 in this mouse PD model are, at least in part, associated with the regulation of ER stress.
Flavonoids are plant-derived polyphenolic compounds that exhibit a broad spectrum of bioactivities, including anti-cancer, anti-inflammatory, and anti-atherogenic effects (19). To date, beneficial bioactivities of flavonoids have been attributed mainly to their anti-oxidant properties, such as radical scavenging, chelating of metal ions, and induction of anti-oxidant genes (19, 21, 33). Among flavonoids, however, methoxyflavones, which are abundant in citrus peels, have another aspect. They also possess quite beneficial properties, such as a hypolipidemic effect (16), suppression of tumor growth/metastasis (3, 27), anti-inflammation (18), and neuroprotective/neurotrophic function (5, 22, 23), although they have much less anti-oxidant activity compared with hydroxyflavones or other flavonoids (1, 6). In fact, in our experimental system, methoxyflavones were less effective against H2O2-induced cell death in MIN6 cells compared with IN17 (Fig. 1) or luteolin (data not shown). It was reported that activation of the PKA pathway through the inhibition of PDE played an important role in the various bioactivities of methoxyflavones (22, 23, 25), suggesting the mechanisms beyond anti-oxidant property underlying the beneficial effects of methoxyflavones.
In the current study, we demonstrated that methoxyflavones strongly protected cells against ER stress in vitro and (Fig. 1) in vivo (Fig. 4). Consistent with a previous report (13), methoxyflavones activated some branches of the UPR such as the eIF2α and Nrf2 pathway and induced downstream genes such as GRP78, HO-1, and CHOP without causing ER stress, although the levels were somewhat less than hydroxyflavones (Fig. 2). Activation of the PKA pathway was also involved in the cell protection by methoxyflavones, at least in some types of cells (Fig. 3A). Although activation of the PKA pathway gives diverse biological effects on cells (29, 37), to our knowledge, this is the first report for the protective effect against ER stress. It could be an upstream event in the signaling pathway triggered by methoxyflavones, when compared with activation of the UPR branches, since forskolin or combination of forskolin and IBMX enhanced expression of GRP78, CHOP, and HO-1 in MIN6 cells (Fig. 3B). However, further studies are required to dissect the detailed mechanism regarding the effect of the PKA pathway in the regulation of ER stress. At this moment, it is also not clear why hydroxyflavones failed to protect cells against ER stress-induced cell death (Fig. 1DI). It may reflect the fact that downstream genes of the eIF2α pathway include both cell survival- and cell death-related genes (31), and persistent activation of the eIF2α pathway could be linked to cell death (26). Another possibility is that the effect of hydroxyflavones to the PKA pathway was less than methoxyflavones, as previously described (25).
We also presented evidence that the neuroprotective property of methoxyflavones in a MPTP-injected mouse was associated with the enhanced expression of molecular chaperons in the ER before MPTP injection (Fig. 5). Consistent with previous reports in vitro (28), intraperitoneal injection of MPTP in mice enhanced immunoreactivity of GRP78/GRP94 in TH+ neuronal cells in the SNpc (Fig. 5A), suggesting induction of ER stress in this experimental model in vivo. Our preliminary results revealed that similar tendency was observed when 6-OHDA was injected on the medial forebrain bundle in the mouse brain, suggesting the cross talk of oxidative stress and ER stress in mouse PD models in vivo.
Recently, it was reported that a selective inhibitor of eIF2α dephosphorylation protected cells against ER stress and viral infection (2). We speculate that there are natural products with similar activities, and, in this context, methoxyflavones could be good candidates. Methoxyflavones were reported to be relatively safe to laboratory mice (36), and methoxyflavone-containing citrus peels have been used for a long time in the field of Chinese medicine or Japanese remedy. Therefore, one of our further studies may aim to reanalyze the beneficial effects of methoxyflavones and other flavonoids in terms of ER stress regulation.
This work was partly supported by a Grant-in Aid for Scientific Research (15032315) from the Ministry of Education, Science, Technology, Sports, and Culture of Japan.
We are grateful to Dr. Jun-ichi Miyazaki (Osaka University), and Dr. Haruhiro Higashida (Kanazawa University) for providing MIN6 cells and PC-12 cells, respectively. We thank Dr. David Ron (New York University) for valuable suggestions. We also thank Takashi Tamatani, Michiyo Tamatani, and Harumi Nishihama for technical assistance.
↵* K. Takano and Y. Tabata contributed equally to this work.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2007 the American Physiological Society