Imoxin inhibits tunicamycin-induced endoplasmic reticulum stress and restores insulin signaling in C2C12 myotubes.

Prolonged endoplasmic reticulum (ER) stress can mediate inflammatory myopathies and insulin signaling pathways. The double stranded RNA (dsRNA) activated protein kinase R (PKR) has been implicated in skeletal muscle dysfunction. However, pathological roles of PKR in ER stress in muscle are not fully understood. The current study aimed to investigate the effect of imoxin (IMX), a selective PKR inhibitor, on tunicamycin (TN)-induced promotion of ER stress and suppression of insulin signaling in C2C12 myotubes. Cells were pre-treated with 5 uM IMX for 1 hr, and exposed to 0.5 µg/ml TN for 23 hr. A subset of cells was stimulated with 100 nM insulin for the last 15 min. mRNA expression and protein levels involved in ER stress were measured by RT-PCR and Western blotting, respectively. TN significantly augmented PKR phosphorylation by 231%, which was prevented by IMX. In addition, IMX reduced mRNA and protein levels of ER stress-related markers including CCAAT-enhancer-binding protein homologous protein (CHOP, mRNA: 95% decrease; protein: 98% decrease), activating transcription factor 4 (ATF4, mRNA: 69% decrease; protein: 99% decrease), cleavage of ATF6, and spliced X-box binding protein 1 (XBP-1s, mRNA: 88% decrease; protein: 79% decrease) which were induced by TN. Furthermore, IMX ameliorated TN-induced suppression of phospho-insulin receptor beta (317% increase) and Akt phosphorylation (by 36% at Ser473 and 30% at Thr308) in myotubes, while augmenting insulin-stimulated AS160 phosphorylation and glucose uptake (by ~30%). These findings suggest that IMX may protect against TN-induced skeletal muscle ER stress and insulin resistance, which are potentially mediated by PKR.


36
The Endoplasmic reticulum (ER) plays a key role in protein folding and calcium homeostasis in many 37 mammalian cell types including skeletal muscle (36). ER stress can be induced when the ER is exposed to 38 physiological or pathological stresses such as accumulation of unfolded/misfolded proteins and 39 imbalanced calcium levels (23,25). As a result, the cells initiate the unfolded protein response (UPR) to 40 restore ER homeostasis and to sustain cell survival (3). The UPR is mediated by three transmembrane 41 sensors: inositol-requiring protein (IRE)-1, RNA-dependent protein kinase (PKR)-like ER eukaryotic 42 translation initiation factor 2 alpha kinase (PERK), and activating transcription factor (ATF) 6 (33). In 43 absence of stress, these proteins are inactive by binding to an ER chaperone called immunoglobulin heavy 44 chain binding protein/glucose-regulating protein 78 (BiP/GRP78) (2). However, when the cells are under 45 stressful physiological conditions, IRE-1, PERK and ATF6 are released from GRP78 and this alleviates 46 ER stress by modulation of gene expression and protein synthesis in the cells (3). However, failure to 47 (Bio-Rad, Hercules, CA) and run on a MyiQ PCR system (Bio-Rad). Each gene was normalized against 120 the housekeeping gene 18S using the ΔΔC T method, and expression of each gene is presented as mean 121 fold-change relative to the non-treated control group. The primer sequences used for each gene are as 122

2-NBDG Glucose Uptake Assay 132
Using a fluorescent D-glucose analog 2-[N-(7-nitrobenz-2-oxa1,3-diazol-4-yl)amino]-2-deoxy-D-glucose 133 (2-NBDG, Cat# 11046, Cayman Chemical Company, Ann Arbor, MI), glucose uptake was monitored in 134 C2C12 myotubes. Cells were seeded at 1 × 10 5 cells/ml in 96-well plates. When the cells reached 95% 135 confluent, cells were differentiated for 5 days in differentiation medium as described above. After 5 days 136 of differentiation, myotubes were pre-treated with or without 5 µM IMX for 1 h and then exposed to 0.5 137 µg/ml TN for 22 h. Cells were washed with sterile PBS twice and then treated with or without 100 nM 138 insulin in glucose-free/serum serum-free DMEM for 30 min. After 30 min of insulin stimulation, 2-139 NBDG was added to the cells at the final concentration of 100 µg/ml for an additional 30 min. The (29). Cells were seeded at 1 × 10 5 cells/ml in 96-well plates. When the cells reached 95% confluent, cells 149 were differentiated for 5 days in differentiation medium as described above. After 5 days of 150 differentiation, myotubes were pre-treated with or without 5 µM IMX for 1 h and then exposed to 0.5 151 µg/ml TN for 22 h. Before determination, cells were treated with resazurin reagent at final concentration 152 of 0.025 mg/ml for an additional 1 h. The fluorescent intensity was measured immediately at wavelengths 153 of 560 nm for excitation and 590 nm for emission using a microplate reader (Fluostar Galaxy, BMG 154 Labtech, Ortenberg, Germany) according to the manufacturer's instruction. Student t-test was performed for comparisons between two groups ( Fig. 3G and 3H). A three-way 164 ANOVA with Tukey's post-hoc test was performed for multiple comparisons in levels of proteins 165 involving in insulin signaling pathways ( Fig. 4 and 5) and glucose uptake (Fig. 6). The differences between groups and the main effects of TN, IMX and insulin were considered statistically significant at 167 p<0.05 and shown in each figure. 168

Effect of TN and IMX on mRNA expression and phosphorylation of PKR in C2C12 myotubes. 170
The effect of TN and IMX on PKR expression was assessed using RT-PCR (Fig. 1A). TN increased PKR 171 mRNA expression (142% increase) compared with the untreated control group. On the other hand, IMX 172 treatment significantly reduced mRNA expression of PKR (by 95%) in cells exposed to TN compared 173 with the TN alone group. In addition to mRNA expression, the effect of TN and IMX on PKR 174 phosphorylation at Thr446 was evaluated (Fig. 1B

Effect of IMX on insulin signaling in C2C12 myotubes. 201
To evaluate the effects of TN and IMX on the insulin signaling pathway, C2C12 myotubes were treated 202 with IMX and exposed to TN, as described above, and insulin was added. As shown in Fig. 4, TN  203 significantly suppressed protein levels of p-IRβ Tyr1150/1151 (by 68%) and p-IRS1 Ser636/639 (by 36%) 204 compared with untreated control in the presence of insulin. However, IMX treatment augmented insulin-205 induced phosphorylation of both p-IRβ Tyr1150/1151 (317% increase) and p-IRS1 Ser636/639 (151% increase) in 206 the presence TN (Fig 4). There were main effects of TN, insulin, and IMX in protein levels of p-IRβ 207 Tyr1150/1151 and p-IRS1 Ser636/639 and interactions between TN × IMX, insulin × IMX, and TN × insulin × 208 IMX were found for p-IRβ and p-IRS1 Ser636/639 . 209 In addition to IRβ and IRS1, TN lowered insulin-induced phosphorylation of Akt at Ser473 (20% 210 decrease), which was fully prevented by IMX (Fig. 5A and 5B). Even though TN suppressed insulin-211 induced Akt phosphorylation at Thr308 (13% decrease), IMX augmented insulin-stimulated 212 phosphorylation of Akt at Thr308 (30% increase, Fig. 5C). Moreover, there was no effect of TN treatment 213 on AS160 phosphorylation, but IMX significantly increased the AS160 phosphorylation in presence of 214 insulin compared to the vehicle (by 46%) and to the TN treatment (by 105%, Figs. 5A and 5D). 215

Effect of IMX on glucose uptake and cell viability in C2C12 myotubes. 217
To monitor the glucose uptake in C2C12 myotubes, 2-NBDG assay was performed as shown in Fig. 6A. 218 There were main effects of IMX, TN, INS and TN × INS on glucose uptake. TN suppressed insulin-219 stimulated glucose uptake compared with VEH (26% decrease). However, IMX significantly increased 220 glucose uptake in presence and absence of TN treatment (26% and 36% increase, respectively). In 221 addition, to evaluate cell viability, resazurin assay was performed in C2C12 myotubes ( Fig. 6B and 6C). 222 TN only treatment group had a significant reduction of cell viability by 35% compared to the untreated 223 control group (Fig. 6B). However, IMX did not show significant difference in cell viability regardless of 224 TN treatment ( Fig. 6B and 6C). 225

227
This study investigated the beneficial effects of IMX on ER stress and ER stress-induced insulin 228 resistance in skeletal muscle myotubes. Our results demonstrate the protective role of IMX on TN-229 induced ER stress and on TN-induced suppression of the insulin signaling pathway. We found elevation 230 of PKR phosphorylation in the cells exposed to TN, but IMX inhibited PKR gene expression and 231 phosphorylation at Thr446 in presence of TN (Fig. 1). IMX suppressed TN-induced gene expression and 232 protein levels of ER membrane-associated sensors and downstream molecules including GRP78 and 233 CHOP ( Fig. 2 and 3). Moreover, we found that IMX promoted insulin signaling which was suppressed by 234

TN (Figs. 4 and 5). 235
In our recent study, we found that PKR phosphorylation was increased by dexamethasone, and that IMX 236 prevented dexamethasone-induced elevation of muscle-specific E3 ubiquitin ligases including muscle 237 RING finger 1 (MuRF1) and muscle atrophy F-box (MAFbx) and subsequent muscle atrophy (9). Skeletal muscle is a highly dynamic tissue that plays crucial roles in locomotion, energy production and 244 endocrine function (1). The ER is a cellular organelle involved in protein folding, lipid synthesis and 245 calcium homeostasis in various cell types, including skeletal muscle (1). Skeletal muscle, in particular, 246 contains the sarcoplasmic reticulum (SR) which is an extensive network of ER and plays a key role not 247 only in muscle contraction but in various nuclear and cytosolic signal transduction through regulating 248 cellular calcium level (32). ER stress and maladaptation of UPS are closely related to skeletal muscle 249 atrophy, insulin resistance, inflammation (1). Hence, it is important to control ER stress in skeletal muscle 250 properly to maintain muscle mass and function as well as to suppress metabolic diseases such as diabetes. 251 During prolonged and uncontrolled UPR GRP78 disassociates from IRE1α, PERK and ATF6 during ER 252 stress because of its higher affinity toward unfolded/misfolded protein than IRE1α, PERK and ATF6 (33). 253 After dissociation from GRP78, IRE1α oligomerization and autophosphorylation promotes splicing of a 254 26-base intron from X-box binding protein 1 (XBP-1µ) mRNA (1). Spliced XBP-1 (XBP-1s) binds to 255 promoter regions of several genes such as chaperones and components of ER-associated degradation 256 (ERAD) (1, 33). Even though the current study did not evaluate IREα phosphorylation, IMX suppressed 257 not only XBP-1µ and XBP-1s, but also TN-induced XBP-1µ splicing (Fig. 2), suggesting that IMX 258 might modulate IRE1α pathway under ER stress conditions. 259 Another ER transmembrane sensor, PERK, is also auto-phosphorylated and activated upon ER stress (3). 260 Along with PKR, activated PERK phosphorylates eukaryotic translation initiation factor 2 alpha (eIF2α) 261 on serine 51. Under prolonged ER stress, phosphorylated eIF2α stimulates translation of UPR dependent 262 genes including activating translational factor 4 (ATF4) which is a transcription factor for 263 CCAAT/enhancer binding protein (C/EBP) homologous protein (CHOP) (3,16,17). CHOP is known as 264 growth arrest and DNA damage-inducible protein 153 (GADD153) and is considered the most important 265 mediator of ER stress-induced apoptosis (36, 42). Elevated CHOP expression may down-regulate the 266 expression of Bcl2 which is the anti-apoptotic protein and up-regulate the expression of pro-apoptotic 267 genes such as BIM (26). Furthermore, increased CHOP expression can perturb the cellular redox state by 268 depletion of cellular glutathione (26). Interestingly, under prolonged ER stress, the ER may also directly 269 stimulate the apoptotic pathway through ER stress-mediated calcium leakage into the cytoplasm and this 270 action can cause vicious cycles of mitochondria-mediated apoptosis (14, 34). In the current study, IMX 271 treatment suppressed protein levels of both p-PERK and ATF4 which were induced by TN treatment (Fig.  272 3D and 3E). Moreover, IMX treatment significantly reduced TN-induced CHOP at both the mRNA and 273 protein level ( Fig. 2A and 3B). These data suggest that IMX prevents excessive activation of the 274 PERK/ATF4 axis induced by TN and consequently reduces CHOP expression.
Upon ER stress, ATF6 is also translocated and cleaved by the action of proteases including serine 276 protease [or site-1 protease (S1P)] and metalloprotease [or site-2 proteases (S2P)] in the Golgi apparatus 277 (3, 36). Cleaved ATF6 can be translocated into the nucleus and interact with ATF/cAMP response 278 element (CRE) and ER stress-response elements (ERSE-1) which increase expression of target genes 279 including ER chaperones, 36). The current study demonstrated that IMX 280 decreased both full length and cleaved ATF6 regardless of TN treatment ( Fig. 3G and 3H). These results 281 suggest that IMX may suppress all the ER stress cascades and consequently reduce expression of CHOP 282 which promotes cellular damage and apoptosis. 283 The current study indirectly measured mitochondrial mediated apoptosis by performing a resazurin assay 284 which is used for sensitive measurement for the viability of mammalian cells (21). Because living cells 285 can reduce the nonfluorescent dye, resazurin, to the fluorescent dye resorufin through mitochondrial 286 reductase, the lower conversion implies higher cell death (21). Our data demonstrate that TN significantly 287 reduced cell viability. We demonstrated that TN increased the protein level and mRNA expression of 288 CHOP which is a main mediator in ER stress-induced cell death. It has been reported that CHOP can 289 directly stimulate the mitochondrial mediated apoptosis (6). Despite the likely impact this may have on 290 insulin signaling and glucose uptake, IMX lowered ER stress and restored glucose uptake, but IMX did 291 not influence TN-induced cell death as shown in Fig. 6, suggesting TN-induced apoptosis is at least 292 partially independent of ER stress. 293 In addition to ER stress signaling, the current study evaluated the effect of IMX on TN-induced insulin 294 resistance in C2C12 myotubes. It has been reported that ER stress and UPR pathways play crucial roles in 295 the regulation of insulin signaling in several tissues (10,19,20,31). Among the ER transmembrane 296 sensors, ATF6 is known to impair glucose transporter 4 (GLUT4) expression in rat and human skeletal 297 muscle cells (31). Ozcan et al. reported that ER stress in obesity promotes serine phosphorylation of IRS1 298 and interferes with insulin signaling (30). Similar to previous studies, the current study found that TN significantly reduced protein levels of p-IRβ and IRS phosphorylation at Ser636/639. However, IMX 300 treatment prevented the TN-induced reduction in insulin-stimulated IRβ tyrosine phosphorylation, a key 301 step in initiating the insulin signaling cascade. Moreover, IMX abolished the suppression of Akt 302 phosphorylation at Ser473 caused by TN. Interestingly, AS160, a mediator of GLUT4 translocation, was 303 highly phosphorylated in IMX treated groups regardless of TN exposure. Furthermore, TN-induced 304 suppression of glucose uptake was significantly ameliorated in the TN-IMX group (Fig. 6A). Collectively, 305 our data showed that IMX attenuated TN-induced impairments in insulin signaling and glucose uptake. 306 The results suggest that PKR may promote insulin sensitivity in skeletal muscle. However, ER stress is 307 known as a contributor to anabolic resistance in skeletal muscle by impairing mTOR complex 1 308 (mTORC1) activity (7). The current study demonstrated that IMX significantly augmented Akt 309 phosphorylation which is also important in anabolic signaling. Likewise, we previously found that IMX 310 augmented anabolic signaling via activation of Akt/mTOR/S6K1 cascades under glucocorticoid-induced 311 skeletal muscle atrophy (9). Thus, IMX might contribute to combat anabolic resistance caused by ER 312 stress and/or insulin resistance in skeletal muscle. 313 In conclusion, the current results provide evidence that TN increased PKR phosphorylation at Thr446 in 314 skeletal muscle cells, which was prevented by IMX treatment. In addition, IMX attenuated TN-induced 315 excessive UPR by suppressing arms of the UPR including PERK/ATF4, IRE1/XBP1 and ATF6 in 316 skeletal muscle myotubes. Furthermore, IMX promoted insulin receptor signaling cascades (IR/IRS/Akt) 317 and insulin-stimulated glucose uptake which were suppressed by TN. This insulin signaling data provide 318 insight into a potential role of PKR in glucose homeostasis in skeletal muscle. These findings support the 319 therapeutic potential of IMX for the improvement of skeletal muscle ER stress and insulin resistance.