Endoplasmic Reticulum Stress and Autophagy Markers in Soleus Muscle Disuse-Induced Atrophy of Rats Treated with Fish Oil

Endoplasmic reticulum stress (ERS) and autophagy pathways are implicated in disuse muscle atrophy. The effects of high eicosapentaenoic (EPA) or high docosahexaenoic (DHA) fish oils on soleus muscle ERS and autophagy markers were investigated in a rat hindlimb suspension (HS) atrophy model. Adult Wistar male rats received daily by gavage supplementation (0.3 mL per 100 g b.w.) of mineral oil or high EPA or high DHA fish oils (FOs) for two weeks. Afterward, the rats were subjected to HS and the respective treatments concomitantly for an additional two-week period. After four weeks, we evaluated ERS and autophagy markers in the soleus muscle. Results were analyzed using two-way analysis of variance (ANOVA) and Bonferroni post hoc test. Gastrocnemius muscle ω-6/ω-3 fatty acids (FAs) ratio was decreased by both FOs indicating the tissue incorporation of omega-3 fatty acids. HS altered (p < 0.05) the protein content (decreasing total p38 and BiP and increasing p-JNK2/total JNK2 ratio, and caspase 3) and gene expressions (decreasing BiP and increasing IRE1 and PERK) of ERS and autophagy (decreasing Beclin and increasing LC3 and ATG14) markers in soleus. Both FOs attenuated (p < 0.05) the increase in PERK and ATG14 expressions induced by HS. Thus, both FOs could potentially attenuate ERS and autophagy in skeletal muscles undergoing atrophy.

This study comprises of a large research project investigating ω-3 FAs (high EPA and high DHA) in skeletal muscle atrophy induced by HS; we previously published results on protein synthesis/protein degradation signaling showing that this experimental model promotes a reduction of body weight, fat mass, muscle weight, and soleus cross-sectional areas (CSA), and a decrease in protein synthesis and an increase protein degradation [7]. FOs supplementations did not influence changes in body weight or soleus CSA induced by HS. However, in the percentage of soleus fibers CSA from rats submitted to HS, there was an increase in the number of fibers in the range of 1000 µm 2 by approximately 140%. On the other hand, with respect to fibers in the range of 800 µm 2 , supplementation with both fish oils during HS caused a reduction of about 25% compared with controls. EPAhigh fish oil attenuated the changes induced by HS on 26S proteasome activity, and levels of p-Akt, total p70S6K, p-p70S6K/total p70S6K, p-4EBP1, p-GSK3-beta, p-ERK2, and total ERK 1/2 proteins. In turn, DHA-high fish oil attenuated the changes induced by HS on p-4EBP1 and total ERK1 levels [7]. The results herein reported (UPR and autophagy markers) and ones previously published (muscle size, proteasome activity, and protein synthesis/degradation) contribute to the understanding of the effects of FOs supplementation on skeletal muscle atrophy due to disuse.

Animals
We obtained eight-week-old male Wistar rats (weighing 203 ± 20.1 g) from the Institute of Biomedical Sciences of University of São Paulo (ICB-USP). We maintained the rats under a 12-h light/dark cycle with free access to water and food. We followed the Guide for Care and Use of Laboratory Animals (National Academy of Sciences, Washington, DC, USA). The ICB-USP Ethics Committee approved the study (24/13/CEUA). The same animals and study design was used in a previous study [7].

Experimental Study Design
We randomly divided the rats into six groups of 10 each, totaling 60 in the whole study. The groups are: MO-C, control receiving mineral oil-MO; MO-HS, HS receiving MO; EPA-C, control treated with high EPA fish oil; EPA-HS, HS treated with high EPA fish oil; DHA-C, control treated with high DHA fish oil; and DHA-HS, HS treated with high DHA fish oil. We administered the oils (0.3 mL per 100 g weight) daily for four weeks by gavage (for details of the fatty acid composition of each FO used, please see ref. [7]). The HS was initiated at the end of the second week and continued concomitantly with the treatments with FOs for another two-week period. We removed soleus and gastrocnemius muscles from both limbs in the conditions previously described [7]. The soleus muscle displays more significant mass loss than the extensor digitorum longus in HS [7].

Hindlimb Suspension (HS)
The HS protocol involved attaching the rat's tail to a rolling pulley at the top of the cage and suspending the hind limbs (30 • suspension) with tape. In this situation, the animals can still move using the forelimbs but do not use their hindlimbs. The same HS protocol was used in previous studies from the same laboratory [6,7,52,53].

Statistical Analysis
Results are reported as the mean ± standard error of the mean (SEM). We used twoway analysis of variance (ANOVA) to indicate significant effects (p < 0.05) of FO treatments and HS. A Bonferroni post hoc test indicated differences between groups (GraphPad Prism software version 4.01; El Camino Real, CA, USA) when a significant interaction was found. Grubbs' test-GraphPad software (graphpad.com/quickcalcs/Grubbs1.cfm) indicated outliers.

IRE1 Signaling in the Soleus Muscle
Using western blot analysis and RT-PCR, p-IRE1 protein content (Supplemental Figure S4) and sXBP1 mRNA levels were measured, respectively. As described above, IRE1α cleaves XBP1 mRNA removing 26 nucleotides and generating the spliced XBP1 (sXBP1) that promotes transcription of UPR genes. Treatment with FOs showed higher values in p-IRE1 protein content ( Figure 1A). sXBP1 was detected in the positive control but it was not found in any experimental group ( Figure 1B). interaction was found. Grubbs' test-GraphPad software (graphpad.com/quickcalcs/Grubbs1.cfm) indicated outliers.

IRE1 Signaling in the Soleus Muscle
Using western blot analysis and RT-PCR, p-IRE1 protein content (Supplemental Figure S4) and sXBP1 mRNA levels were measured, respectively. As described above, IRE1α cleaves XBP1 mRNA removing 26 nucleotides and generating the spliced XBP1 (sXBP1) that promotes transcription of UPR genes. Treatment with FOs showed higher values in p-IRE1 protein content ( Figure 1A). sXBP1 was detected in the positive control but it was not found in any experimental group ( Figure 1B).  The contents p-JNK 1, total JNK1, p-JNK2, and total JNK2 contents, and p-JNK 1/total JNK1 ratio were not significantly different among the groups (please see Figure 2 and Supplemental Figure S4). However, p-JNK2/total JNK2 ratio was significantly higher (p < 0.05) in MO-HS (119%), EPA-HS (42%), and DHA-HS (59%) than in non-HS rats (hindlimb suspension effect, Figure 2F).

Discussion
It has been reported that HS mimics spaceflight, bed rest, or the hospitalization state in humans [52]. Previous studies of our group [6,7,53] and others [61,62] described that this experimental model promotes a reduction of body weight, fat mass, muscle weight, and soleus CSA. Concomitantly, there are a drop of protein synthesis and a raise of protein degradation. Previously, we showed that in rats subjected to HS, FOs increase CSA of soleus muscle fibers. Separately, high EPA FO supplementation attenuates the increase in 26S proteasome activity and the decrease on protein synthesis markers. Conversely, high DHA FO supplementation has fewer molecular effects in the protein synthesis pathway [7].

Discussion
It has been reported that HS mimics spaceflight, bed rest, or the hospitalization state in humans [52]. Previous studies of our group [6,7,53] and others [61,62] described that this experimental model promotes a reduction of body weight, fat mass, muscle weight, and soleus CSA. Concomitantly, there are a drop of protein synthesis and a raise of protein degradation. Previously, we showed that in rats subjected to HS, FOs increase CSA of soleus muscle fibers. Separately, high EPA FO supplementation attenuates the increase in 26S proteasome activity and the decrease on protein synthesis markers. Conversely, high DHA FO supplementation has fewer molecular effects in the protein synthesis pathway [7].
In the skeletal muscle, the endoplasmic reticulum regulates calcium concentrations during muscle contractions and plays a critical role in cellular homeostasis [63]. However, few studies have investigated strategies to prevent alterations on ERS and autophagy markers in a muscle disuse condition. The effects of ω-3 FAs (EPA and/or DHA) attenuate ERS and autophagy markers expressions induced by HS were investigated in the present study. HS altered the soleus protein content (decreasing total p38 and BiP and increasing p-JNK2/total JNK2 ratio and caspase 3) and gene expressions (decreasing BiP and increasing IRE1 and PERK) of ERS and autophagy (decreasing Beclin and increasing LC3 and ATG14) markers. The treatment with both FOs decreased the ω-6/ω-3 FAs ratio in the skeletal muscle. In addition, both FOs attenuated the decrease in p-IRE1 content and the increase in PERK and ATG14 mRNA expressions induced by HS, the last two displaying particularly meaningful changes as compared to the other measurements. Although in this study the effect of the groups as a whole must be considered, we also performed an unpaired t-test to examine the effects of HS in each condition (MO, EPA or DHA). We observed that fish oils can attenuate the protein contents changes on p-p38, BiP and caspase 3, and gene expressions of IRE1, PERK, and Beclin induced by two-week HS.
Herein, there was no evidence of XBP1 splicing in any group. However, there was a percentage increase in p-IRE1 protein content in the DHA-HS compared to the DHA-C group. It is known that the IRE1 pathways can also activate apoptosis through JNK activation [24]. In this sense, it is important to point out that an increase in p-JNK 2/total JNK 2 ratio in HS rats was found. Another MAPK family member is the p38, which is activated by inflammatory signals and oxidative stress, and it is involved in muscle atrophy [64,65]. p38 activates transcriptional factors that lead to protein degradation and apoptosis in skeletal muscle [66]. While it was not observed increased p38 activation, a decrease in total p38 content was detected in the HS groups. This reduction in p38 content could account for the observed increase in JNK activity. Notably, an increase in caspase 3 content was detected in all HS groups after 14 days, indicating apoptotic pathway activation.
It is well-known that BiP is an essential protein in the UPR and ERS. The release of BiP from the luminal domain of IRE1, PERK, or ATF6 correlates with an imbalance in the ER microenvironment and activation of the three signaling pathways. High BiP levels could delay the UPR, and low levels could prolong it [67]. Woodworth-Hobbs et al. (2017) reported that muscle atrophy induced by palmitate can be prevented by DHA through an ERS/UPR mechanism. DHA was postulated to activate proteolysis via caspase and to augment expressions of autophagy-associated genes [68]. Herein, decreased BiP expression and protein content and increased IRE1 and PERK expression in the soleus of MO-HS animals were observed, which is indicative of a prolonged ERS. Both FOs suppressed the upregulation of IRE1 and PERK expression. This finding may have a significant effect at the beginning of the UPR signaling, but since there were no differences between groups regarding JNK activation and caspase 3, PDI, and CHOP expression or content, this effect was not translated into significant downstream alterations.
It was previously reported by Chen et al. (2015) that transgenic amyotrophic lateral sclerosis (ALS) mice (G93A*SOD1 heterozygote) displayed increased levels of ERS markers in the white gastrocnemius fibers. These authors found more pronounced effects than those reported in the present study, which is probably due to the ALS-related neurological changes (e.g., degenerative neuromuscular disease) [69]. In addition, the white muscle portion has more glycolytic/type II fibers with less oxidative capacity, whereas the soleus, herein studied, has predominantly oxidative/type I red fibers [70]. Differences between skeletal muscles with different fiber type compositions require further investigation.
Baehr et al. [42] investigated the effects of HS for 14 days followed by the same period of reloading on 9 (adult) and 29 (aged) months old male F344BN rats. ERS markers (BiP, PDI, CHOP) contents did not change by HS [42]. These results are consistent with PDI and CHOP content results herein observed but are in contrast with the decrease in BiP protein content and gene expression described. Similarly, the authors observed no changes in autophagy markers (phospho-Unc-51 like autophagy activating kinase-1 [p-ULK1], p62, ATG7, Beclin, and LC3b-II) in soleus. Despite showing no evidence of autophagy or ERS increase in the soleus of 9-month-old animals during HS, their data suggest that these processes are activated upon reloading [42], contrary to what was herein found. These differences could be due to the strain of rat employed (F344BN vs. Wistar) or animal age (9 months vs. 2 months).
Herein, Beclin, LC3b, and ATG14 expressions, markers of the macroautophagy, were evaluated [33]. All HS groups exhibited increased ATG14 and LC3b expressions in the soleus muscle, markers commonly upregulated during autophagy. However, contrary to the hypothesis of the present study, the expression of Beclin was reduced by the HS. In the dynamics of autophagosome formation, ATG14 recruits PI3K-III complex when subjected to autophagic stress [71][72][73][74]. In addition, it is known that the Beclin homodimer is inactive when bound to the BCL2-complex. However, when BCL2 is phosphorylated, Beclin is released, allowing it to bind to ATG14, VPS34 and VPS15 [71]. Therefore, a concomitant increase in ATG14 and Beclin expression was expected to be observed. However, this was not the case.
Considering solely the gene expression, without protein translation or post-translational modifications, stability changes, protein translocation and assembly of the autophagy complex, there was another study reporting a disconnection between the expression of Beclin and ATG14. In HeLa cells, siRNA-mediated knockdown of either Beclin or ATG14 did not affect the mRNA levels of the other gene [75] [76]. Thus, it is possible that an extreme condition like hindlimb suspension, which modulates several signaling pathways, would interfere with the activity/expression of some of those transcription factors, with independent effects in the expression of Beclin and ATG14. In this work, a part of the entire process was evaluated, and other markers can influence this pathway. The increase in ATG14 and LC3 could have occurred through the PERK-eIF2α-ATF4 pathway [23]. As mentioned above, ERS and autophagy are involved in controlling skeletal muscle mass loss. Despite some markers of these pathways being upregulated, their specific activation during muscle unloading requires further investigation. For example, the involvement of 5 adenosine monophosphate-activated protein kinase [AMPK], a serine/threonine protein kinase/mammalian target of rapamycin complex [Akt/mTORC], FoxO3, NF-κB, and protein kinase C theta [PKCθ] should be considered [77][78][79][80].

Strengths and Limitations
The anabolic effects of ω-3 FAs on protein synthesis were previously described [13]. There is evidence that FO supplementation can improve insulin sensitivity and increase the protein synthesis pathway activity. FO can also attenuate some markers of the UPR in ERS-induced conditions (ex. PERK), which is in contrast to the induction of ER ceramide synthesis by palmitic acid [81]. Herein, high EPA and high DHA FOs attenuated the increased expression of PERK induced by HS. Such an effect could potentially prevent the deleterious activation of UPR and ERS during disuse muscle-induced atrophy.
It is important to point out that our study has some limitations. For example, the experimental protocol was conducted over a fixed period, with supplementation for four weeks and HS in the final two weeks. The authors interpreted that the effects of the FOs were mainly due to an increase in the content of ω-3 FAs, but one cannot rule out a decrease in the proportion of saturated fatty acids plays a role in the results reported. Additionally, we did not perform any of the analyses on skeletal muscles with different fiber type compositions. Some autophagy and ERS markers were not evaluated, and some were evaluated through their gene expression and not their protein content. A possible total JNK reduction in the DHA group may account for the increased phospho/total ratio Further research is necessary to detail how FOs could potentially attenuate ERS and autophagy activities in muscles undergoing atrophy. The supplementation in different doses of FO in patients under a condition of muscle disuse should be evaluated and muscle ERS and autophagy markers should be measured. Additionally, studies measuring protein synthesis/degradation in vivo will be of great interest. However, despite the mentioned limitations, the results are consistent and support the conclusions of the study.

Conclusions
Both FOs attenuated the increase in PERK and ATG14 expressions induced by HS. Thus, both FOs could potentially attenuate ERS and autophagy in skeletal muscles undergoing atrophy.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/10.3 390/nu13072298/s1, Figure S1: Ponceau S quantification of western blot membranes, Figure S2: Images used for quantitative analysis of the western blot assays in this study, Figure S3: Results of the analysis of fatty acids composition in the gastrocnemius muscle used in the Supplemental Table S1 of this study, Figure S4: Results of the quantitative analysis of western blot assays used in the Figures 1-4 of this study, Figure S5: RT-PCR results after 2 −∆∆C T calculation for the relative expression of genes used in the Figure 5; Figure 6 of this study, Table S1: Composition of fatty acids in g/100 g gastrocnemius muscle wet weight.